Digitized by the Internet Archive

in 2011 with funding from

LYRASIS Members and Sloan Foundation

http://www.archive.org/details/foundationsofstrOOindunh
FOUNDATIONS OF STRUCTURES
 McGRAW-HILL CIVIL ENGINEERING SERIES
Harmer E. Davis, Consulting Editor

Babbitt •
Engineering in Public Health
Babbitt and Doland Water Supply Engineering
Benjamin Statically Indeterminate Structures

Chow •
Open-channel Hydraulics
Davis, Troxell, and Wiskocil •
The Testing and Inspection of
Engineering Materials
Dunham Foundations of Structures
•

Dunham The Theory and Practice of Reinforced Concrete

•

Dunham and Young Contracts, Specifications, and Law for Engineers

Gaylord and Gaylord Structural Design •

Hallert Photogrammetry
•

Hennes and Ekse Fundamentals of Transportation Engineering

•

Krynine and Judd Principles of Engineering Geology and Geotechnics

•

Linsley and Franzini Elements of Hydraulic Engineering

Linsley, Kohler, and Paulhus -
Applied Hydrology
Linsley, Kohler, and Paulhus •
Hydrology for Engineers
Lueder Aerial Photographic Interpretation
•

Matson, Smith, and Hurd Traffic Engineering •

Mead, Mead, and Akerman Contracts, Specifications, and Engineering

•

Relations
Norris, Hansen, Holley, Biggs, Namyet, and Minami •
Structural Design
for Dynamic Loads

Peurifoy •
Construction Planning, Equipment, and Methods
Peurifoy •
Estimating Construction Costs
Troxell and Davis Composition and Properties of Concrete
Tschebotarioff Soil Mechanics, Foundations, and Earth Structures
Urquhart, O'Rourke, and Winter Design of Concrete Structures
Wang and Eckel Elementary Theory of Structures
Foundations of Structures

CLARENCE W. DUNHAM
Consulting Engineer

Formerly associated with the Civil Engineering Depart-

ment, Yale University; Consulting Structural Engineer
for Anaconda-Jurden Associates, Inc., New York, NY.;
Assistant Chief Draftsman, Bethlehem Steel Co.; Assist-
ant Engineer, The Port of New York Authority; Chief
Structural Designer, Phelps Dodge Corp. Member,
American Society American Institute
of Civil Engineers,
of Consulting Engineers, American Concrete Institute,
Connecticut Society of Civil Engineers, American
Society for Engineering Education

SECOND EDITION

McGRAW-HILL BOOK COMPANY, INC

New York Toronto London

1962
 bMAb

SCIENCE

FOUNDATIONS OF STRUCTURES
Copyright © 1062 by the McGraw-Hill Book Company, Inc.
Copyright, 1950, by the McGraw-Hill Book Company,
Inc. Printed in the United States of America. All rights
reserved. This book, or parts thereof, may not be repro-
duced in any form without permission of the publish-
ers. Library of Congress Catalog Card Number 61-12047

THE MAPLE PRESS COMPANY, YORK, PA.

18214
^

\
According to the grace of God which is given
unto me, as a wise masterbuilder / have laid the
,

foundation, and another buildeth thereon. . . .

I Corinthians 3:10

^
^
 PREFACE

The planning of foundations more of an art than a

for structures is
science. The conditions at any must be taken as they are
particular site
— the result of the processes of nature in the formation and disposition
of soils and rocks, often modified to some extent by the works of man.
These natural materials differ vastly in their physical properties. They
may be mixtures that possess to some extent the various characteristics
of their constituents, and erosion may have assembled them in an almost
infinite number of combinations and stratifications.
In foundation work the engineer must use rocks and soil as structural
materials. He realizes that the planning of the superstructure should be
based upon the characteristics of the materials of which it is to be built,
and he may select and use the materials that will best serve his purpose.
On the other hand, in planning the substructure, he must use whatever
soils are already there, or he must devise ways to improve the situation in

the interests of safety and suitable economy.

Although it is practically impossible to set up rules, regulations, and
detailed recommendations for the solution of all the problems' that arise
in the planning of foundations, there are many basic principles that may
be helpful in such matters. In this book the author has endeavored to
present some of these principles and to show their application in specific
cases. In a field such as this, a person will inevitably have ideas that
do not agree completely with those of everyone else. Nevertheless, the
author believes that lessons learned through his own experiences and those
of his friends will assist the reader to develop engineering ability that he
can use in solving the foundation problems that may lie before him.
The author has not attempted to write a book on soil mechanics. He
has endeavored to show how to take soils as materials and build sub-
structures upon them. The types illustrated are mostly those that are
encountered in ordinary construction. Major and complicated founda-
tions are usually planned and designed by men of long experience who
do not need to read this book. The young engineers and those of limited
experience are the ones whom the author wishes to help particularly. The
successful performance of their structures is of vast importance to them
and to those whom they serve.
viii PREFACE

The curricula of our engineering schools are already so crowded that

many important subjects can be studied in only a brief introductory
manner. The author has planned this book so that after the first four
chapters the reader may study the first few articles in each of the remain-
ing chapters and from them obtain the general and the
principles involved
methods of attack that may be used.
Those who are in engineering practice may wish to study in detail what-
ever else seems to offer assistance. For this reason many of the problems
are worked out numerically so that the reader can be sure, when studying
alone, that he understands the engineering and computations involved.
All calculations were made by the use of the slide rule, and many of the
results rounded off to two significant figures.
Since concrete is used so much in substructures, the design of the
reinforced-concrete members is Because the reinforce-
often illustrated.
ment and strength in shear are usually more critical than compression in
the concrete in such thick, stiff members, the author has used the unit-
stress method of design.
In the detailed designing and analysis the author uses approximations
that he believes are easy to understand and apply. It is believed also
that the methods used yield safe, reasonable, and satisfactorily economical
results. When one considers the broad assumptions that must be made
in the selection of allowable bearing pressures, applied loads, estimations
of the characteristics of soils, and the action of relatively thick members
of reinforced concrete, one will realize that subsequent calculations need
not be made with a refinement that is utterly inconsistent with the
accuracy of the data upon which they are based. However, the approxi-
mations should lean toward the side of safety. The great economies or
expenditures in foundation engineering are "made" at the time that the
basic plan is determined. Refinements of the computations thereafter
affect the cost less than one might suppose.
In practice an engineer should learn all that he can, within reason, about
the conditions at a given site. He should then determine what types of
foundation are practicable in this specific case. Then he should compare
their cost, suitability, ease of construction, and safety. Finally, he
should adopt the foundation which, in his judgment, is the best. After
that, he may proceed with the detailed parts of the design. Repeatedly,
the author tries to show the reader how to size up a situation and how to
tell which of several alternatives is the best.
The author wishes to thank allwho have contributed photographs,
drawings, and data for his use. They have been very helpful. He is
especially grateful to Ayres C. Seaman, who read the manuscript, and to
L. A. Warner for other assistance.
Clarence W. Dunham
 CONTENTS

Preface vii

List of Symbols and Glossary xiii

1. Introduction 1

Importance of foundations — Definitions— Purpose

Economics.

2. Soils as Foundation Materials 7

Introduction — Types of foundation materials— Origin of
soils — Some properties of specific foundation materials
Soil-mechanics laboratories.

3. Exploration of Site 23
Tentative selection of a — Importance exploration —
site of
Shallow explorations— Deep explorations — Soil profiles
Load — Planning subsurface explorations — Importance
tests
ofsubsurface conditions to the planning structures
of
Value of investment engineering— Example — Problems.
in

4. Some Principles of Foundation Action 50

Introduction— Distribution of pressure— Shearing of soil
Settlement— Estimates settlement — Structures founded
of
upon — Danger signals— Allowable bearing pressures
fill

Problems.

5. Spread Footings 95
Introduction — Data for the analysis of reinforced concrete
Pedestals — Pedestal footings — Some basic assumptions
regarding isolated reinforced-concrete footings — Action of a
circular spread footing— Square and rectangular isolated
spread footings— Combined footings—Special combinations —
Problems.

6. Foundation Walls 139

—
Ordinary simple walls Spread footings and grade beams —
Walls to spread a series of equal concentrated loads Walls —
to spread a single concentrated load —
Basement walls
—
Waterproofing basements Miscellaneous details Problems. —
x CONTENTS

7. Mats 176
Introduction — Uniform mat— Ribbed mat— Mats to resist
hydrostatic pressure— Floating foundations — Examples of
planning — Problems.

8. Foundations Subjected to Overturning Forces 213

Introduction— Connections at bases of columns steel
Anchor bolts— Grouting— Action base plates
of steel
Isolated footings subjected to overturning— Unsymmetrical
footings— Footings steadied laterally by structural parts —
Footings subjected to and overturning— Anchors
uplift for
backstays — Pedestals on large mats — Retaining walls
Foundation walls with loads — Anchorages
lateral pipe for
— Problems.
lines

9. Piles 300
—
Function of piles General action of a pile under load
— —
Action of a group of piles under load Pile driving Wooden
piles— Cast-in-place concrete — Precast concrete
piles piles —
Steel — Sheet — Pile-driving formulas—Load
piles piles tests.

10. Pile Foundations 360

Introduction — Choice type of — Allowable bearing
of pile
value a
of — Spacing
pile — Isolated footings on
of piles piles
with vertical loads — Isolated footings with eccentric loads —
Combined footings on — Mats on — Differential
piles piles
settlements— Sequence driving
of — mats and
piles Piles vs.
spread footings — Vertical subjected to horizontal forces
piles
— Batter — Piers— Bulkheads on — Evaluation
piles piles of
pilegroups — Problems.

11. Cofferdams and Other Aids for Open Excavations 447

Introduction —-Shoring —-Single-wall cofferdams in general
Cantilevered single-wall cofferdams — Single-braced single-
wall cofferdams — Multibraced single-wall cofferdams
Double-wall cofferdams — Cellular cofferdams — Miscellaneous
practical features cofferdams — Special methods
for pro- of
tecting excavations — Wellpoints— Piecemeal shoring deep of
excavations— Surcharge pressures on piling — Problems.

12. Caissons .511

Introduction— Open caissons— Pneumatic caissons— Materials
and — Miscellaneous caissons— Special construction
details
procedures and problems — Problems.

13. Bridge Piers 544

Introduction —Architecture — General choice location — of
Scour— Wave action — Ice — Choice of type bridge and pier — of
Forces acting upon piers — Stability computations — Bearing-
details — Grouting bearings — Details
of pier tops — Miscel-
of
laneous items — Problems.
 —

CONTENTS xi

14. Bridge Abutments 604

— —
Introduction Planning locations Abutments for simply sup-
—
ported and continuous bridges Abutments for rigid-frame
—
bridges Abutments for arches — Suspension-bridge anchor-
— —
ages Miscellaneous details Problems.

15. Underpinning 638

— General principles— Underpinning columns
Introduction
Underpinning walls — Problem.
of

16. Machinery Foundations 655

Introduction— General principles—Vibration— Concrete foun-
dations on — Concrete foundations on — Concrete
soil piles
foundations on superstructure — Special problems.

Appendix 695

Bibliography 707

Index 711
 LIST OF SYMBOLS AND GLOSSARY

a = alpha; an angle
a = acceleration, ft. per sec. 2
A = area
A = area of steel in tension, in.
s
2

A = area of steel in vertical stirrups, in.

v
2

b = width of rectangular beam or stem of T-beam, in.; width of

footing, ft. or in.

c = distance from neutral axis to point where stress is to be com-

puted in formula / = Mc/I, in. or ft.

c.c. = center to center

C.G. = center of gravity
compaction = process of increasing the density of a soil by mechanical
means, e.g., by vibration or rolling
consolidation = process of increasing the density of a soil by effect of
weight through natural means of squeezing out the water and
smaller particles working into voids
cos = cosine
5 = delta = deformation
A = delta = differential part or increment
d = depth from compression face of beam or slab to the center of
gravity of the longitudinal tensile reinforcement, ft. or in.

DL = dead load
e = eccentricity; voids ratio
E = modulus of elasticity
/ = coefficient of friction

fe = compressive unit stress in extreme fiber of concrete, p.s.i.

/' = ultimate 28-day compressive unit stress in concrete cylinder
tests, p.s.i.

/s = tensile unit stress in longitudinal reinforcement, p.s.i.

ft = tensile unit stress in concrete, p.s.i.

fv = tensile unit stress in vertical stirrups, p.s.i.

xiv LIST OF SYMBOLS AND GLOSSARY

F = force; frictional force

fetch = clear distance over water through which wind can cause waves
striking structure
g = acceleration due to gravity = 32.2 ft. per sec. 2

h = height
H = horizontal force
/ = plane moment of inertia
Ip = polar moment of inertia
in situ = in place, not having been transported
j = ratio of distance between centroid of compression and center
of gravity of tensile steel to depth d of beam or slab
k = kip = 1,000 lb.

k = ratio of distance between compression face of beam or slab and

neutral axis to depth d
K.E. = kinetic energy
kip = 1,000 lb.
k.s.f. = kips per square foot

k.s.i. = kips per square inch

L = length; span of beam center to center of bearings, ft. or in.

lb. = pound

Lin. = linear

LL = live load
m, M = mass = W/g
M = applied bending moment or internal resisting moment, in. -lb.,

ft.-lb., or ft.-kips
max = maximum
min = minimum
min. = minute
mm. = millimeter
m.p.h. = miles per hour
n = modulus of elasticity of steel to that of concrete
ratio of
N= normal force or pressure; number
co = omega = angular velocity, radians per sec.
o = perimeter of bar, in. (So = total perimeter of reinforcement
needed at a point or supplied)
4>— phi = angle of internal friction of soil; angle of repose of soil

p = ratio of area of tensile reinforcement to that of the effective

area of concrete bd for beams and slabs
P = load, lb. or kips
p.c.f. = pounds per cubic foot
p.s.f. = pounds per square foot

p.s.i. = pounds per square inch

Q = quantity or volume
 LIST OF SYMBOLS AND GLOSSARY xv

R = reaction
2 = sigma = summation
s = spacing of vertical stirrups, in. ; distance
5 — section modulus = I/c
S.F. = safety factor
sec. = second
sin = sine of angle
6 = theta = angle of slope of earth
T = tensile force
tan = tangent of angle
u = average bond unit stress on or allowed on tensile reinforcement,
p.s.i.

v = velocity, ft. per sec.

vl = computed shearing unit stress in concrete based upon depth jd
and assumed to be a measure of the corresponding diagonal
tension, p.s.i.

v'L = allowable shearing unit stress and diagonal tension in concrete

without stirrups, p.s.i.

vt = computed shearing unit stress in concrete based upon depth

kd and assumed to be a measure of punching shear, p.s.i.

V = total transverse shear in the member at cross section being

considered, lb. or kips
w = unit load
W= load or weight; wind force
x = distance; unknown quantity
y = distance; unknown quantity
 1

INTRODUCTION

1-1. Importance of foundations. It is obvious that the foundation

must be adequate if a structure is to be safe and satisfactory. Though
the foundation is inconspicuous and may be unnoticed by the public, the
planning of its and the design of its parts may involve some
basic features
of the greatest engineering skilland the best judgment in connection with
a project. The construction of the foundation may require some of the
most difficult work of all the operations carried on in the field.
It has been said that he who designs and builds important and difficult
foundations does not sleep well at night. This may be an exaggeration,
but it is not without a bit of justification. If a steel girder is too weak and
flexible, it usually can be strengthened found to be inade-
; if a column is

quate, it may be reinforced or assisted by intermediate columns that were

not at first intended; but if the foundation yields unevenly and ruptures a
structure, if settlement causes the building or pier to
tilt badly, if the sub-

sidence is so much not satisfactory and usable, then

that the structure is

little can be done to improve the situation. One can seldom rebuild an
inadequate foundation and repair the edifice without excessive expense.
Although the Leaning Tower of Pisa is known throughout the world and
is even the objective of many a traveler, the modern engineer should not

expect that one of his structures that settles unevenly will compete for
fame with that edifice.
The planner and designer of the foundation must assume the responsi-
bility for its performance. The construction men may have great diffi-
culties in carrying out the intent of the plans, but they should not be held
responsible for unwise judgment that may have been used in the concept
and design of the project. Yet those who plan the structure and who
make the vital decisions are faced with problems beset by uncertainties.
The vagaries of nature have prepared the materials and placed them as
they are. The characteristics and probable actions of these materials
under loads may be difficult to ascertain; floods may wash away the soil
or inundate the structure; moving ice and storms may exert pressures of
l
2 INTRODUCTION [Chap. 1

unknown magnitudes; and frost may cause heaving and subsidence.

Even the works of man may aggravate the difficulties where a sunken
barge or heaps of riprap lie buried beneath the mud and sand, where some
old sewer cuts across the site far below the surface, or where part of the
terrain consists of questionable fill such as old mattresses and bedsprings.
Even the prospective owners often seem to be determined to build the most
important structures in the most inopportune places and for exceedingly —
meager sums ofmoney.
Is the engineer to be cowed by these difficulties? Is he to go home and
wait until the difficulties disappear? It is up to him to obtain all the
information possible about the problems confronting him, to determine
what courses of action are open to him, to study various alternates that
might be used to support the structure, to visualize the probable action
of those alternates, to estimate their approximate costs, to decide upon
the relative feasibility of their construction, to recommend that which he
considers the best, and last, but not least, to explain to his clients the
nature of the problems and the reasons for his recommendations.
Thus, the major decisions may rest upon his judgment and engineering
sense. These qualities are developed and made reliable through years of
study and experience, yet study and experience alone are not guarantees
of expert ability if they have not been of the right caliber. One of the
greatest of an engineer's assets is the ability to visualize and to think
clearly. Another man who does not possess similar ability and under-
standing should not be criticized too severely because he is unwilling to
accede quickly to these decisions that depend so much upon intangible
judgment.
1-2. Definitions. What is meant by planning and design? Hardy
Cross aptly stated that, in substance, planning is the determination of
whether or not the project is worth building, what its general proportions
are to be, how it will fit into its surroundings, what is required for it to
serve purpose properly, and of what materials it is to be built. Design
its

is the next step in the refinement of the plan. It is the determination of

the framing scheme, the choice of the positions and types of members, the
obtaining of a scale upon their sizes to see that the scheme is feasible, the
selection of what is to constitute the many related accessories, and the
perfection of the layout to see that the structure will serve its purpose
well. During the planning and designing, basic decisions are made. If
these decisions are unwise, minor refinements and the detailed calcula-
tions of sizes and dimensions can do little to remedy the situation. In
some cases, the word design is used to mean the entire office part of the
concept and development of the project in contrast to the construction
work in the field.
The word foundation may denote (1) the soil or rock in situ to which
Art. 1-2] DEFINITIONS

are transmitted the forces caused by the dead load of the structure and by
all other loads applied to the latter, (2) the total ensemble of the structural
parts of the substructure that serves as the medium through which the
weight of the superstructure and the forces due to the loads upon it are
transmitted to the supporting soil or rock, or (3) the combination of both
of these. The sense of the text will make the meaning clear in most cases;
Fig. 1-1 will help to clarify the meanings.
The demarcation between the superstructure and the substruc-
line of
ture may For convenience one ma}' define
not be determined clearly. -

these parts as follows, referring to Fig. 1-1 as an illustration:

1. Superstructure denotes that portion of a structure which is built

essentially and directly for the use of man regardless of how the structure

Superstructure,
C3 C
Top of en c ^Superstructure
ground CD C
Substructure
^or foundation Substructure ^Foundation
VT: ^Foundation or foundation
(bh A SMALL GIRDER BRIDGE
(a)- A MULTISTORY BUILDING
Suspenders
Tower Guard'rail'-^ Culvert Embankment
Anchorage Roadway Pavement, \ \ ^Superstructure
^Anchorage ^^jj^M^ |
|
l
|
l
l

'l ^ l'l
|
l
l

f'^

Substructure Foundation
or foundation
Substructure Foundation Gravel soil
or foundation
(c)- A LARGE SUSPENSION BRIDGE (d)~A HIGHWAY EMBANKMENT

FIG. 1-1. Illustrations of superstructures, substructures, and foundations.

is supported. For example, in Sketch (a), the owner desires to have a

building of given size and for specific uses; inand (c) he wants a bridge
(6)
to hold him up while he he wishes to have a road-
crosses a stream ; in (d)
way to carry the public's cars yet to permit the stream to continue in its
course. In general, these superstructures are what they are because of
the uses for which they are intended, although local conditions of many
kinds may temper one's desires regarding what he wants if he is to attain
his objectives with safety and economy.
2. Substructure denotes the portion of a man-made structure that is

needed to hold the superstructure in place and to transmit all forces due to
the superstructure and its use to whatever the supporting material may be.
Thus, in Sketch (a), the footings and foundation walls spread the forces to
the soil; in (6), the abutments transmit the reactions of the bridge to the
rock; in (c), the piers and the anchorages hold the bridge in place even
though the anchorages merely resist sliding and the upward pull of the
4 INTRODUCTION [Chap. 1

cables; in (d) ,
the embankment and the box culvert with its footings trans-
mit the loads to the gravel. In Sketch (d), the pavement may be looked
upon as the superstructure; in the case of an earth dam, the earthwork
itself is the superstructure and needs no separate or extra parts to bear
upon the soil. A telephone pole set in the ground has no separate sub-
structure although its embedded portion serves the purpose. The fine
distinctions between the superstructure and the substructure are not
important; of vast importance, however, is a clear understanding of how
to plan each structure, of how to build it, of how it will function, and of
how the foundation will behave.
1-3. Purpose. It is the purpose of this book to help the reader under-
stand the character of some foundation problems, to show him how to
devise possible solutions for such problems, to enable him to improve his
ability to determine the advantages and disadvantages of these solutions,
and to help him develop his engineering judgment so that he can make his
decisions wisely. Therefore, the endeavor is to illustrate principles and
structural action, to show various possible solutions for specific problems,
and to indicate how an engineer might compare these solutions in order to
make In many cases of actual work, the schemes that were
his decision.
rejectedand the reasons for such action are more instructive than a mere
examination of the adopted design.
Many types of foundation are to be studied. Ordinary footings for
columns and walls are built in vast quantity. Some conditions require
the use of heavy mats; others require piles, deep piers, or even floating
foundations. Cofferdams and deep caissons may be needed, too. All
these are to be discussed.
Much depends upon whether a structure is large or small, heavy or light,
rigid or flexible. Thus, its requirements become the starting point in the
planning of foundations. When the soil conditions are exceedingly bad
and when it will be too difficult to support the structure, a revision of the
plans for the structure itself may be necessary.
Foundation problems vary so greatly that one cannot illustrate all of
them adequately. Nevertheless, the general principles of the action of
materials and structural elements, when properly understood, are tools
that may be used in spite of the variation of the details of the problems
to which they are applied. The substructure and the soils around and
beneath it are all a part of the same general problem the latter should be
;

studied as carefully as the former.

Real economy may be difficult to ascertain. For example, at a power
plant located at tidewater, the boiler feed pumps were founded on the roof
of a reinforced-concrete intake tunnel. Salt water had flowed through
the tunnel for something like 20 years. A short time ago the roof gave
way, and the pumps dropped into the water. An investigation showed
 —

Art. 1-4] ECONOMICS 5

that the bottom reinforcement of the roof had apparently rusted in the
salt water and salty air so as to spall off the concrete cover below them.
The pumps then continued to corrode until they were fatally weakened, at
which time some gave way. The transfer of load to neighboring weak-
ened bars was more than they could stand, so they failed also, and entire
collapse followed. The vibration of the pumps may have aggravated the
difficulties, but this is uncertain. At any rate, what seemed like an eco-
—
nomical arrangement the use of the tunnel as a pump foundation
proved to be expensive in the long run. One should always consider the
distant future when he plans the foundations for equipment upon which
the operation of a plant depends.
1-4. Economics. As D. P. Krynine has said, "When designing foun-
dations, the engineer should consider three signs, i.e., +, — , and $." The
be borne in mind always, and there is much to consider
last of these is to
besides numerical calculations. How much more should one spend in
order to reduce a structure's settlement by 25 per cent? How much more
expense justified in making sure that there will be no appreciable settle-
is

ment? With the money available, what type of foundation will serve
one's purpose the best? Such questions as these are important; good
answers to them are often difficult to find, yet they must be sought; and
the engineer must face the consequences of his decisions because he has no
opportunity to bury his mistakes.
Perhaps a building costing $1,000,000 is to be built upon weak plastic
soil. Will the soil support it safely and satisfactorily? Should $100,000
more be spent to provide piles that will transmit the loads to firm strata
below the questionable soil? Slide rules and computers do not provide
the answers. Solutions must be based upon engineering judgment, judg-
ment that is developed through years of experience and study, judgment
that assists the engineer in weighing the evidence, in predicting the prob-
abilities, and in making a decision. Such problems are not the place to
exercise blind opinion, to resort wholly to precedent, and to do comforting
wishful thinking. It is no wonder that Hardy Cross often said that civil

engineering is an art, not a science.

Almost any reasonable structure can be built and supported safely if
thereis enough money to spend. This, however, is seldom the engineer's
problem. In general, he must do the best he can with limited funds, or he
must accomplish safe results with a minimum of expenditure.
—
The first of these basing one's decisions upon the best that can be done
for a stated number of dollars — does not necessarily produce results that
constitute good foundation engineering. The engineer is entrusted with
the safety of and property. When a structure fails and someone is
life

injured, nobody will consider any shaving of costs to have been justified;
when a structure settles or tilts so as to impair its value and usefulness, the
6 INTRODUCTION [Chap. 1

owner will not believe that he was to blame because he set unreasonable
financial limitations.
If an engineer asked to plan and design a foundation for a specific
is

structure and to keep the cost within a stated limit, it is his duty to use

the utmost skill of which he is master to devise a good safe economical

solution of the problem. On the other hand, if no satisfactory design can
be made for the money available, it is his duty to see that his client real-

izes the situation and the dangers involved. The engineer should, if nec-
essary, refuse to go ahead with a project when he is convinced that it is
unsafe or unwise, in spite of the pressure that will surely be put upon him.
He is sometimes the prospective owner himself; in other cases he is the
owner's adviser and, as such, should give his honest considered opinion
regardless of whether it is pleasing. His position is similar to that of a
physician who may have to tell his client that an operation is necessary
even though such news is exceedingly unwelcome.
The planning and design of a safe foundation for a minimum expendi-
ture form the usual type of problem confronting the engineer. He is
always and inevitably concerned with economics. There is an old saying
that an engineer is one who can build for a dollar what anybody could
build for two. Regardless of economic pressure or of its absence, a good
engineer will strive to make the utmost use of men, materials, and money.
This is engineering!
 2
SOILS AS FOUNDATION MATERIALS

2-1. Introduction. It is assumed that the reader has previously made

at least a slight study of soil mechanics. Therefore, the details of the
analyses and experiments that are customarily made in a soil-mechanics
laboratory are not discussed. Nevertheless it is desirable to give a brief
description of the materials that one is likely to encounter in the field, and
to provide a condensed discussion of their properties and of their probable
actionwhen used for the support of a structure.
The use of soils as foundation materials and for various other works of
man dates from a time far beyond the dawn of his recorded history. The
knowledge of soils that was obviously possessed and used by the engineers
and builders who preceded us was far greater than we sometimes admit.
If one stops to think, he will recognize that these men were exceedingly
keen and capable, and they often accomplished excellent results in
spite of the limited techniques and equipment with which they had to
work.
The modern engineer who engages in foundation work has great need for
a thorough understanding of soils as structural materials. He will need
a keen sensibility to detect danger when it exists. He will need the ability
to obtain and evaluate evidence upon which to predict the probable action
of soils in innumerable varieties of arrangement and properties. The tests
made in a laboratory may be exceedingly helpful, but observation in the
field, a knowledge of the performance of structures upon comparable mate-

rials, and the proper use of imagination are helpful, too.

2-2. Types of foundation materials. For convenience, foundation

materials have been grouped into a series of types. The properties of
those in any one group may vary considerably because of different con-
solidation, moisture content, and physical make-up nevertheless, a knowl-
;

edge of the properties and behavior of one member of a group or type will
assist a person in anticipating what to expect of other similar soils.
The definitions of the groups as used herein and a brief description of
each type of material are as follows:
7
8 SOILS AS FOUNDATION MATERIALS [Chap. 9

1. Bedrock is sound hard undisturbed rock in its native location, of

indefinitely great extent, not broken up by harmful seams and cracks, and
underlain by no material except rock. The geological types of material
composing bedrock may be such materials as igneous, sedimentary, and
metamorphic rocks. Some subdivisions of these are sandstones, con-
glomerates, slates, and shales. The reliability of the last two may be
questionable, depending upon the attitude of the beds and the degree of
weathering. The excavation of bedrock cannot be accomplished by hand
or mechanical shovels but requires blasting.

7M77W7mK r~^--s-~ JiJ^OZyj

(a ^-Ordinary earth (b)-Uud (c)-?ea\

or loam

fa?;- Soft clay feJ-Stiffclay tfV-Silt

(g)-Sift and clay (h)-Fine sand (i) -Coarse sand

Q..0".Q 6 o'.a.'.o
&"*

(//'"Clay and sand fJc)- Grave I /7J-Gravel and

boulders

(m)-Hardpan or (/^-Disintegrated (o)-Bedrock

caliche rock

FIG. 2-1. Examples of probable symbols to represent various foundation materials.

2. Weathered rock is an indeterminate stage between bedrock and soil

(earth material). It is usually in its natural position above or alongside
sound bedrock. It is likely to have seams that are rilled with rock frag-
ments or even with claylike materials. Hard pieces may be separated
completely by layers of soil or disintegrated rock; in other cases, erosion
may have caused much of the fine material to be carried away.
3. Boulders are fragments or pieces of rock that have been broken away

from the bedrock. They may or may not have been transported far from
their original positions relative to the "parent" bedrock, but they are usu-
ally considered as pieces that have been moved somewhat and worn by
water or grinding. For descriptive purposes, the term boulders will be
Art. 2-2] TYPES OF FOUNDATION MATERIALS

limited to pieces that exceed 10 in. in maximum dimension. Pebbles may

be used to denote pieces ranging from 4 mm. to approximately 2 in., while
those from 2 to 10 in. in size may be called cobbles.
4- Gravel is used to denote unconsolidated rock fragments ranging in

size from around 6 in. maximum to 2 mm. minimum. In ordinary cases,

gravels are composed of particles of many sizes of more or less rounded
Ty er standard sieve r um )ers

3 4 6 8 10 14 20 28 35 4 8 6 5100150 200
i
i

V &^_
\'\-

\ \
No. / : Fine sand
and silt
^\ No. 3: Fine sand
No. 4: Medium and
s
fine sand
\1
l\
en
\\
\
-O
'l

\
\ \
§ an- 3\ \ \

\\ \

V
\\

\\
\\
\\ s
u
..
1 0.5 0.1 0.05 0.01
Grain size, millimeters

Medium Fine Coarse Medium Fine Very fine

Silt
gravel gravel sand sand sand sand

U.S. Bureau of Soils Classification

FIG. 2-1 A. Grain-size distribution of some sands.

pieces, and they generally contain some sand, and often some silt. Arti-
ficiallycrushed stone and coarse naturally angular material may be called
stone. Pea gravel is a name that is sometimes used to denote a screened
or rather uniform gravel having particles approximately }4 to ^4 hi. in
size. Bank-run gravel is a term frequently used to denote a sand-gravel
mixture which is taken directly as excavated from an original deposit
without any attempt to screen it or to sort out sizes. The size gradation
may vary widely, depending upon the particular deposit.
5. Sand is composed of small rounded or angular particles of weathered

rock (or artificially crushed rock) that vary in size from 2 to 0.05 mm.,
 —

10 SOILS AS FOUNDATION MATERIALS [Chap. 2

depending upon what system of size classification is used. Sand fre-

quently contains some clay and silt. Even with a clay and silt content
of 15 to 20 per cent, the material is generally called a sand. The follow-
ing arbitrary subdivisions of the broad classification may be useful when
describing sands:

Fine sand . 0.25-0.05 mm.

Medium sand 0.6 0.2 mm.
Coarse sand 2.0 0.25 mm.

6. Silt is composed of fine particles of rock that range in size from

approximately 0.05 to 0.005 mm. Rock flour is another term used to

denote silt-size material, probably resulting from glacial abrasion.
Organic silt is silt mixed with finely divided organic matter, and it has a
bad odor. It is often found in backwaters, lake bottoms, and deltas.
7. Clay, or clay soil, is composed of exceedingly fine particles of inor-
ganic material, less than 0.005 mm. in diameter. These particles are
scalelike, and they are generally poorly consolidated. A varved clay has
noticeable thin strata, or laminae (often alternately light and dark in
formed by variations in sedimentation during each season,
color), that are
year, or some longer time. Clay becomes plastic when wet but loses this
plasticity when dried.
8. Hardpan is a term that is and may denote a wide range
loosely used
of materials. Some persons use it impermeable
to refer to a dense, rather
mixture of gravel and sand with enough clay or silt so that under terrific
pressures such as those caused by glaciers, the mixture has attained a high
density and the particles are partly cemented together. However, the
term hardpan is sometimes used to denote a cemented indurated rocklike
layer that will not soften when wet. Still others use the term to denote

a stratum composed largely of clay cemented by relatively insoluble mate-

rials but which does not become plastic when wet.

9. Till, or glacial till, is often used to describe a mixture of unsorted and

unstratified stones, gravels, sands, silts, and even some clay left by glacial
action. It may or may not be compacted by pressure. It is likely to
contain some boulders.
10. Caliche generally denotes gravels, sands, silts, and clays which are
cemented together by "desert salts," such as calcium carbonate, that are
deposited by ascending moisture and evaporation.
11. Loam is a mixture of sand, silt, or clay, or a combination of any
of these, with some organic matter —
humus and lis the soil that is of
great value to agriculture. It is sometimes called topsail, in contrast to
the subsoils that contain little or no organic matter.
12. Adobe is a heavy-textured alluvial clay, often associated with desert
regions of the Southwest.
Art. 2-3] ORIGIN OF SOILS 11

13. Gumbo is a very fine claylike material that becomes very sticky
when wet. It occurs in some lowlands and central states.
14. Mud is often a slimy, sometimes sticky mixture of earth materials
and water in a fluid or weakly solid state.
15. Peat is partly decayed organic matter, as found in swamps.
16. Muck is generally a mixture of organic matter and inorganic earth
materials, although the terms muck and mud are often used synonymously.
17. Loess is generally a porous, buff -colored, nonindurated, unstrati-
fied deposit of silt-sized material. It is usually a wind-blown (aeolian)
deposit. It is often threaded with grass roots.
18. Bentonite is a decomposition product of volcanic ash. It swells
when it is wetted.
Because of the almost infinite variety of mixtures of soil materials, it
is customary to describe some soils by means of adjectives that reveal the

nature of the most important secondary material that is mixed with the
primary (predominant) material. Such descriptions as sandy clay, clayey
sand, silty clay, and gravelly (or stony) clay are useful. Such a term as
sandy loam denotes a loam that is primarily sand. As a rule, sandy mate-
rials are "light," whereas clay materials are " heavy" soils, these terms
being descriptive of the use of the soils in cultivation rather than from
the standpoint of unit weight.
It is desirable to picture soils by symbols when studying, designing, and
illustrating foundation problems. There seems to be no set standard of
symbols; each office may have a system ofits own, or its men may invent

the picturization for a specific case. In order to secure some reasonable

consistency in the drawings herein, an attempt has been made to use the
symbols shown in Fig. 2-1 additional ones will be used as necessary.
;

2-3. Origin of soils. Soils are chiefly the products of one or more of
the following:

1. Physical (mechanical) disintegration of rocks due to erosion, wear,

and crushing
2. Chemical decomposition of rocks and rock materials, e.g., oxidation
and hydration
3. The mixing of the products of vegetation and other life with inor-
ganic materials

The combination of mechanical and chemical actions that produce soil

is often called weathering. Leachingis the process of removing from soils

the water-soluble materials therein by the action of rainfall or under-

ground flow.
Long-continued and complicated have been the geological processes of
the breaking up of igneous rocks, the transportation and deposition of the
resultant materials, the formation of sedimentary rocks by means of pres-
12 SOILS AS FOUNDATION MATERIALS [Chap. 2

sure, the breaking up of these last rocks again, and perhaps many cycles
of such changes. The materials may have been transported many times
and through long distances by water, wind, gravity, and glaciers; even
man may have contributed to this movement. The soil structure grad- —
ing and arrangement of the particles — may affect the properties of a soil

composed of a given material, and so may the physical pressure that is or

has been applied to the soil. The presence or absence of water, too, may
affect greatly the physical properties of these soils as foundation materials.
Is it any wonder then that soils are found to be so varied in their character
and behavior?
One may derive considerable benefit by visualizing the probable manner
in which the soils that he must utilize were formed. Some soils — even
thick strata —may have been formed in situ by the weathering of the rocks

^K nt rnnrnn
•

''
A - Inorganic and organic materials

{-j B- Inorganic materials disintegrated

:;.'."-
and thoroughly leached

W£o$$. C~ Disintegrated andpartly leached

S*$ii$&^ D -Partially disintegrated
Wmm,E- Bedrock
D . ,
— "

FIG. 2-2. General character and distribution of soils formed by disintegration in place.

through long periods of time. Such a case is pictured in Fig. 2-2. Such
an arrangement may occur in level and slightly rolling territory in the
South. In general, however, the force of gravity causes pieces that are
broken from bedrock to fall and form slopes of talus water carves ravines ;

through weak spots and crevices and carries the particles far away; run-
ning water continually carries particles from higher to lower elevations
and tends to flatten out the slopes and cause the sediment to settle else-
where, as indicated in Fig. 2-3(a); the wind removes the fine particles as
fast as weathering produces them, as exemplified by the rocks in Fig.
2-3(6), forms sand dunes, and even "cuts" many rocks into strange
shapes; and sometimes volcanoes spread lava and ashes over vast areas.
The glaciers that once covered the northern portion of the United States
scraped away a great deal of the soil, ground off much of the disintegrated
rock of that region, and deposited boulders, gravel, sand, silt, and cla}r
over wide areas. Long Island, N.Y., is believed to be terminal moraine
left near where the glacier's face met the sea, where much of the ice

melted and dropped the imprisoned materials, or where it ceased to push

and carry them along. Except for glacial clays, these glacial deposits are
usually neither stratified nor homogeneous. Even in short distances
Art. 2-3] ORIGIN OF SOILS 13

great differences are likely to occur in the "overburden" or soil covering

the bedrock of glaciated regions.
Beach and shore deposits are formed by the action of the waves and
and by the waves of large lakes. These forces of nature
tides of the ocean

FIG. 2-3. Illustrations of nature's products and processes.

cut away the rocks, grind up the pieces, and sometimes carry the sands
far along the coast.
Flowing water has surprising ability to transport materials. One who
has a garden on a hillside can see its effects after a heavy thunderstorm.
A fisherman who wades in a deep fast river that has a slippery rocky bot-
tom will realize this, too. The Grand Canyon of the Colorado, the Pali-
14 SOILS AS FOUNDATION MATERIALS [Chap. 2

sades of the Hudson, the Niagara Gorge, and the Delaware Water Gap
are spectacular examples of the cutting action of running water however, ;

the myriad small ravines, washes, and streams of the country are more
important in their total effects upon the removal, building, and trans-
porting of soils.

When a stream flows into a lake or other relatively still water, sedimen-
tation occurs. This action is pictured in Fig. 2-4. Assume that stream A
flows down a rocky bed and then empties into a pond at B. The velocity
of flow isreduced quickly so that at C it is slight, at D and E it may be
negligible. The solid materials that are carried in suspension or rolled
along the bottom, especially in time of flood, are discharged into the pond.

Stream

PLAN

Silt and clay

if- 4.".'L. .

SECTION A-A

FIG. 2-4. Illustration of sedimentation caused by a stream carrying materials into a pond.

The heavier particles like gravel will come to rest first, the coarse sand
next, then the fine sands and silts, and finally the clay which, being in
suspension and settling slowly, may be transported far into the pond.
Sedimentation in this manner is a classifying process. However, in times
of small flow, the fine materials will settle close to B; then they may be
scoured out again and carried farther on during the next freshet.
The ground at a particular site may be composed of many strata of
varying character, especially in the areas once covered by glaciers. In
other places, the soil may be fairly uniform over large areas. Near Chi-
cago, for example, there is a deep stratum of clay that was deposited when
the entire region was under water. At Los Angeles, the flood plain of the
Los Angeles River and the Rio Hondo contains a deep \ayev of fine sand
that extends over a considerable area.
Sediments may accumulate far more quickly than one would think. In
a rolling, farm and wooded area of Connecticut, a pond about 100 ft. wide,
Art. 2-4] PROPERTIES OF FOUNDATION MATERIALS 15

250 ft. and 3 ft. deep was made to form a skating rink for a boys'
long,
school. One would not have suspected that a babbling brook would carry
much solid material, yet the pond filled up so badly in approximately 15
years that only a few inches of water remained, and the ice would freeze
solidly to the bottom.
These matters of erosion and the formation of soils are only a few illus-
trations of the workings of the laws of nature. They are given to show
the reader that it may be very helpful to him if he will try to ascertain
and visualize the geological history of the site upon which he is to build
a structure involving the safety of life and property, as well as the future
of hisown professional reputation. He may thus learn what to expect,
and determine the probable presence or absence of danger.
When an engineer recognizes or suspects such dangerous conditions that
the assistance of a capable geologist is advisable, he should not fail to call
for help. Doing so is not an evidence of lack of ability on his part. Many
millions of dollars may
be at stake. The incompetent engineer is he who
fails to perceive the dangers when they exist, or " shuts his eyes and takes

a chance" rather than utilize the abilities of the best talent available.
2-4. Some properties of specific foundation materials. The soils
encountered in the field cannot always be classified completely in accord-
ance with the definitions stated in books. This must be remembered
always. Nevertheless it is proper for an engineer to wish to have avail-
able some information as to what he may expect of rather typical mate-
rials, and of the loads that can be placed upon them safely. The data
given here are for guidance only; they are not given as a haven of refuge
behind which the reader can find comfort.
The following are descriptions of the properties of foundation materials
in so far as they affect the planning and design of foundations:
1. Bedrock is usually capable of withstanding any pressure that can be

applied by man's artificial rock, portland-cement concrete. However, it

is important to ascertain the kind of rock, the presence of cracks, the dip

of the strata, and the slope of the surface.

1
There may be cases in which
the quality of the bedrock itself should be investigated for chemical and
physical stability. An extreme case of this arose in connection with the
selection of the site for an industrial plant in northern Mexico. The
white chalklike volcanic rock underlying a thin siliceous rock cap on a
low broad hogback hill was tested crudely by putting small pieces in a
shovel containing a little water. The "rock" audibly sucked up the
water and disintegrated so fast that it became mushy in a minute or two.
This led to the search for another location.

1
Joints in inclined strata may be slippery when wet. Limestone may contain
caves.
16 SOILS AS FOUNDATION MATERIALS [Chap. 2

This extreme case illustrates emphatically another basic principle for

the engineer to remember; i.e., the hills are hills because their material

was more resistant to erosion than were those of the adjacent valleys and
ravines. Therefore, there is the possibility that, if dangerous materials
are found at the edges of the hill, the softer materials may underlie a
harder rock cap on the hill. Is excavation likely to remove the protecting
blanket and expose the weak materials?
Shales disintegrate easily; so may slates and sandstones. Limestone
as a foundation for a dam may fail because of solution along leaking
Temporary wedges,
point up space
u I Slab
aa Slab or shoe
Roughened Shoe
pJ^pfr-iBars * »

- ~-i Recess
.
2"± for grout
J|
"
''^Anchor Hoops °
:
'• •• •'!, Hairpin
lj
iL bolt
l Close top with
waste, wooden
'>~Ma/h reinf.
Main
^ Anchor bolt plug® or
.

similar device
Welded pi. washer
reinf. during concreting

ta)-S0ME PIER ^-ALTERNATE <W- JACK-SCREW CdJ-P\PE SLEEVE ON

DETAILS PIER DETAILS DEVICE FOR ANCHOR BOLT
DRYPACKING

Building

Overburden

Caj-ROCK CUT FOR A HIGHWAY (b)-k BUILDING FOUNDED ON ROCK

FIG. 2-5. Excavation of, and foundations on, bedrock.

cracks. If core borings show large cracks, the rock may be suitable as
the support for an ordinary building but not for a dam or a great bridge.
If the strata dip sharply toward the excavation, as at A in Fig. 2-5 (a),
the side AB of the rock cut A BCD may weather rapidly and cause rock
slides. A structure supported on the berm EA
will be in a relatively
dangerous situation compared to one on DF. However, if the excavation
is a narrow cut running into this hillside, it may be possible for the rock

to "arch" horizontally around it. When the foundation is to be on slop-

ing rock, as in Sketch (6), benching may be necessary at L, M, and to N
avoid sliding of the structure.
The matter of the excavation of bedrock for foundation purposes should
be faced realistically. Rock cannot be carved out with a knife it must be ;
 —

Art. 2-4] PROPERTIES OF FOUNDATION MATERIALS 17

blasted out, and it will break away along its natural planes of weakness
parallel to the stratification or along the cracks. Plans are sometimes
made without proper consideration for these practical matters. For
examples, refer to the following sketches in Fig. 2-6:
(a) Here a retaining wall is to abut into a rocky hillside. Benches are
shown cut out they are relatively short, as pictured, it
of the bedrock. If

is probable that the finished excavation will follow all too closely along the
dotted line AB.
is shown benched at CD in order to save exca-
(6) In this case, the rock
vation and construction. The rock is likely to break somewhat as shown
by the dotted line CE so that the hoped-for bench will not exist.
{Top of wall
-Surface of rock
Bui/ding
, Assumed sfeps
Probable - for footing
lihe of
breakage Probable //he
of breakage
Vj \
ssum^d
6teP

faJ-STEPS FOR SUPPORT OF A WALL E

/W-OFFSET IN FOUNDATION

Assumed A A
trench*' \ t_ J
Probable jj
fineof rfiU

breakage Plan View of Section

Row of Holes A-A
fcJ-TRENCH FOR DRAIN OR SEWER ftf J- ILLUSTRATION OF LINE DRILLING

FIG. 2-6. Excavation of bedrock

This illustrates a drain that is supposed to be near a row of columns

(c)

in an industrial plant. When the rock is blasted for the trench GHJK,
the result will probably look more like FHJL, and the rock under the
columns must be benched farther to avoid sliding of the bases.
(d) This pictures the idea of "line drilling" of rock. Holes are drilled
very close together so that the rock can be pried off, or so that a plane of
weakness is developed with the hope that a blast will break the rock away
along that plane but not beyond it. It is needless to say that such work
is expensive. When benching is desired along CD of Fig. 2-6(6), it may
be possible to line-drill the face DE. However, the blasting performed
in excavating down to CD is likely to weaken the rock near the corner D,
causing the latter to crack off. It is better to plan rock excavations with
generous dimensions and simple outlines than to attempt to secure sharp
reentrant angles as at C in Sketch (b), corners as at D, narrow vertical
chases, sharp vertical corners, and undercut pockets.
18 SOILS AS FOUNDATION MATERIALS [Chap. 2

2. Weathered rock untrustworthy for heavy loads. In general, it

is

should be removed so that the structure may be founded upon the under-
lying bedrock. Its use as the sides of steep exposed cuts is likely to be
dangerous because of weathering and slides. In some cases, pressure
grouting with portland cement will seal the open cracks; when the cracks
are filled with silt, clay, or very fine disintegrated material, the efficacy
of the grouting may be doubtful. Grouting and pointing up of the joints
in a badly cracked exposed rock face may be positively harmful because
water is more likely to be trapped behind the face where its hydrostatic
head and expansion during freezing may aggravate the danger of slides.
It is best to found a structure well back from the top of a badly cracked
rock slope and to keep well away from the bottom of it.
3. and irregular ones, are not usually reliable
Boulders, especially large
for the support ofheavy loads because of the weaker underlying and sur-
rounding materials and because of their tendency to tip. Boulders in or

Caisson \

Crack

Deflected
position

SKIDDING \b) KINKED (c) UNEQUAL SUPPORT id) CAISSON HUNGUP

FIG. 2-7. Examples of troubles caused by boulders.

deep down under plastic or fine granular weak soils in which piles or cais-
sons are to be driven are likely to cause serious difficulties, some of which
are pictured in Tig. 2-7. It may be advisable in some cases to remove
the large boulders entirely; in others, to found the structure upon the
soils above the boulders; or it may be necessary to find a better location.

4. Gravel is an excellent foundation material if it is not underlain by a

weak stratum and if it is situated so that scour will not undermine the
foundation. Shaly gravel is, of course, an exception. The presence of
ground water is not likely to decrease the strength of a well-graded gravel
composed of stable materials. Gravels have little capillarity.
5. Sand is ordinarily a good foundation material. However, when it
is fine and loose, its strength and bearing value decrease. When very
fine, saturated, and subjected to a positive hydraulic gradient, it may
become "quick" and flow out from under even small loads if a way for
its escape is provided. Such sand thoroughly trapped, as within a tied
steel sheet-pile enclosure, may be trustworthy if one can rely upon the
permanence of the restraint. Sites in which quicksand exists should be
viewed with great suspicion. On the other hand, damp fine sand may
 —

Art. 2-4] PROPERTIES OF FOUNDATION MATERIALS 19

be stronger than the same material when completely dried because of the
surface tension of the moisture, as one may realize when walking on an
ocean beach. Medium and coarse sands have low capillarity; when com-
pact and well graded, they have good bearing value; when saturated, they
will still hold substantial loads. Saturation does not seem seriously to
reduce the angle of internal friction of medium sand or coarser granular
soils.

6. Silt (inorganic) is definitely one of the weak unreliable soils with

high capillarity. It is not plastic and has little strength when dry. Its
presence constitutes a warning to the foundation engineer. Its high
capillaritymay increase the possibility of damage from frost action. Silt
has a low frictional resistance and will flow easily into excavations; even
when shoring is used around an excavation, the ground water that leaks
in may carry enough silt with it to cause undermining and settlement of
adjacent areas. Silts — and fine sands, too — are likely to wash or become
quick in the bottom of an excavation if a hydrostatic head exists under
the bottom. Silt once badly disturbed will have a very low supporting

value for a long time. Very fine silt with a little clay — and organic silt

may have considerable impermeability, plasticity, and cohesive strength.

In the construction of the Lincoln Tunnel in New York, these qualities
of the Hudson River silt enabled the contractor to use far less pressure
of compressed air than was needed to resist the theoretical head of water.
7. Clay usually contains and effectively retains considerable water.

When dried, it is and it shrinks and cracks exces-

generally quite hard,
sively during drying, only toexpand again when its moisture is recovered.
In a moist state it is highly plastic and compressible; it may even flow
slowly under long-continued pressure. Surface water may soften it, but
it maynot erode easily.
Clays may be described as soft, medium, and stiff, depending upon
their moisture content and prior consolidation. In some cases, clay will
support considerable load. For example, desiccated (previously dried)
and preconsolidated clays (which supposedly were compressed for long
periods under the glaciers or other overburden) may be quite firm and
may support fairly heavy loads unless exposure and the action of water
cause them to soften. The impermeability of clay may be an asset when
excavations are made because ground water cannot enter the excavation
rapidly. However, deep excavations may blow up at the bottom if sub-
jected to hydrostatic pressure. The cohesive strength of clay enables it

to stand temporarily with steep slopes and to transmit moderate lateral

pressures around a small excavation. Nevertheless, clay used as a foun-
dation material should be studied carefully and utilized with discretion. 1

1
A rough but useful way to tell in the field whether a soil is clay or silt is by means of
the "pencil test." Take a small portion of the material and roll it between the palms
20 SOILS AS FOUNDATION MATERIALS [Chap. 2

Both clays and silts should be tested under the worst moisture conditions.
Sometimes they can be drained by drilling through impervious soils above
or below them.
8. Hard-pan will generally resist relatively large unit pressures. That
formed by glacial or other pressure evidently is not underlain by soft

FIG. 2-8. Sketch map showing centers of radiation and area covered in North America during the
Pleistocene ice age. (After Chamber/in and Sa//sbury. Prom teggef, "Geology and Engineering,"
McGraw-Hill Book Company, Inc., New York, 1939.)

plastic soils, but it may have many boulders in and below

it. Some
materials called hardpan will be weakened by weathering when exposed.
Hardpan may be eroded by surface or subsurface water.

of your hands. If it rolls into a long pencillike mass which sticks together fairly well,
it at least has considerable clay in it if it breaks into pieces and will not hang together,
;

it is largely silt.
 —

Art. 2-4] PROPERTIES OF FOUNDATION MATERIALS 21

9. Till, till, may be quite variable in its supporting power,

or glacial
sometimes being equivalent to a good sandy gravel but at other times
being weakened by its silt or clay content. Loose till may vary in its
character from place to place so as to cause uneven settlements.
10. Caliche, being sand and gravel —
probably with some silt and clay
cemented together, is generally very strong as a foundation material.
However, future leaching of the soluble cementaceous compounds may
restore the soil to its former character. Hence, a steep cut may break
down so that a structure resting close to its top edge may be endangered.
Structures bearing upon "carved" offsets like CDE of Fig. 2-6(6) may
not always have the intended strength of support beneath them.
11. Loam is not a trustworthy material for heavy foundations. Gen-
erally silty and clayey, it partakes of the qualities of these materials;
containing organic matter, too, it is weak and compressible. In most
instances, it should be removed completely and not utilized for foundation
purposes.
12. Adobe and highly cohesive when dense and moist, but
is fairly stiff
its when wet is extremely poor; and wet-
value as a foundation material
ness should be expected at some time or other.
13. Gumbo, as its name suggests, is notoriously poor for a foundation.
If faced with the problem of utilizing it, the engineer should "float" the
structure or plan it so that settlement will not be harmful.
14- Mud is useful for the support of very light loads only, even when
it is moderately compacted.
15. Peat is practically worthless as a foundation material. Although
compressed under a blanket of fill, it should not be trusted underneath
important structures, even for highway embankments.
16. Muck is only a slight improvement over peat.
17. Loess may
be of some value as a foundation material, as a loose
uniform silt or very fine, somewhat dustlike sand might be. It often has
the unusual property of being more stable in a nearly vertical cut than
on a slope, even under moderate moisture conditions. This fact is some-
times of value when construction sheathing, particularly in trenches, is

under consideration.
18. Bentonite, when wet, is plastic and sticky. One should be wary
about trusting it.

Fortunately, the unsatisfactory materials mentioned are localized and

are not very extensive. Along water-front areas and near sluggish
streams, conditions are sometimes very unfavorable, and these locations
are often the necessary or desirable ones for industrial plants. Each
foundation problem is to be studied by itself, and the solution should be
tempered by considerable conservatism.
Tables 4-2 to 4-5 give assumed figures for the permissible bearing values
22 SOILS AS FOUNDATION MATERIALS [Chap. 2

and other properties of various soils, and are not to be accepted as the
absolute values. They are given as qualitative data —as a general scale
on the probable values Local building codes and qualified persons may
state figures that are different the former particularly should be consulted
;

because they should not be violated willfully. However, codes and speci-
fications are prepared for the guidance of the engineer; they do not prevent
him from being more conservative if his investigations convince him that
he should be so in a particular case.
2-5. Soil-mechanics laboratories. The foundation engineer should
not fail any information that can be secured if it will give him
to obtain
important evidence upon which to make his decisions. He should take
advantage of the tests that can be made in a soil-mechanics laboratory.
It is admitted that they may not tell him everything nor answer all his
questions. Nevertheless, he may obtain valuable evidence of the prob-
able behavior of the soils from the test results, such as the determination
of particle sizes and grading, moisture content, plastic limit, liquid limit,
plasticity index, density, shearing resistance, coefficient of internal fric-
tion, cohesion, compatibility, permeability, and physical and chemical
soundness. The compaction tests for silts and clays and the shear
test for all materials are probably the most important to the foundation
engineer.
These be made, and the engineer should be able to inter-
tests should
pret the resultsand not hesitate to confer with those who made the tests.
The engineers and technicians who make the tests personally are likely
to form valuable opinions regarding the probable action of the materials
under the existing conditions and under the forces to be imposed. Their
opinions should be sought and weighed carefully.
Nevertheless, it is important for an engineer to visit the site and to
study the conditions personally. Hardy Cross once said substantially
that it is advisable for an engineer to visit the site and surrounding ter-
rain even though he sits on a stump for several hours just looking them
over. Of course, this means that it is beneficial if he knows what to look
for, if he knows enough to understand what he sees, and if he lets his

imagination freely picture all the probable things to come that relate to
the job. He should make such a visit before explorations are made, and
he should make one or several more after the soils have been tested and
as the plans are being developed. Then, as the excavation is carried out,
he should watch to see that no surprises are encountered. If unfortu-
nate, unexpected, or dangerous conditions are discovered, it is wiser to
face and overcome them immediately than it is to stick obstinately to
preconceived plans in order to avoid the expense and embarrassment of
revising a design.
 3
EXPLORATION OF SITE

3-1. Tentative selection of a site. There are many things to be con-

sidered before the site for an important structure is selected. In the case
of a proposed industrial plant, for example, there are business matters to
be settled before the general locale is chosen. Some of these matters are
the relation of the proposed plant to such as the following:

1. Present properties and business of the corporation

2. Locations, properties, and facilities of competitors
3. Character and quantity of proposed production, and the possi-
bilities of success
4. Source and availability of raw materials, water, and power
5. Location, character, and capacity of the market
6. Labor supply
7. Transportation facilities and costs
8. Climate and living conditions
9. Sewage and waste disposal
10. Space required and cost of real estate
11. Taxes, legal restrictions, zoning, and special privileges

After the general location is determined, comes the problem of selecting

the specific site. There may be several properties among which one can
choose.
A bridge, for example, should be located where it is possible to build
such a structure and where it can fulfill its purpose of supporting traffic in
the best practicable manner. A large office building, a public edifice, or
an apartment house may be rather restricted as to location because of
other structures, zoning regulations, available space, transportation facili-
ties, and the dictates of business policy.
A discussion of these fundamental questions is not intended. Assuming
that a specific has been chosen tentatively for good and sufficient
site
reasons, what explorations should be made?
23
24 EXPLORATION OF SITE [Chap. 3

3-2. Importance of exploration. The person who fails to investigate

the conditions at a specific site before he purchases is inviting trouble. It
may be necessary to secure an option on the property with the agreement
to purchase, subject to proof that conditions are satisfactory. It is prob-
able that the prospective purchaser must stand the expense of any explora-
tions. However, if he fails to have them made before he buys, he may find
that he cannot build as he intended without unduly costly foundations; he
may be forced to build lower lighter structures covering more area so that
there is insufficient space left on the property for other necessary facilities
or for future expansion; or he may be forced to buy another property and
dispose of the first one as best he can.
A businessman or engineer would criticize severely anyone who started a
business project without first determining that the venture could and

probably would be successful. Nevertheless, many a piece of real estate

has been purchased for building purposes without proper investigation of
what was beneath the surface. The engineers were then forced to make
the best they could of the situation. In one case, an industrial concern
was on the point of buying a large tract of land when the president of the
corporation heeded the advice of his engineer and insisted upon subsurface
explorations. These revealed such bad conditions that expensive piling
would have been needed under the entire structure and beneath all of its
important equipment. Another site was selected, and the owners of the
first one admitted that they might as well give the land to the city because

they themselves would not think of building upon it, and they could not
now sell it to anyone.
Wherever it is practically necessary to locate a structure at a specific
place, there should be a practicable exploration of both surface and sub-
surface conditions. The presence of weak soils may force one to choose a
lightweight material for the structure so that the intensity of the applied
loads will be small, to use long-span construction with a minimum number
of individual foundations because each one extending to a firm support
will be so costly, and to adapt the layout of the entire project to the limita-
tions imposed by the foundation conditions. On the other hand, hidden
rock may prevent the intended excavations or make them very costly.
Springs and high ground water may increase the cost of construction, and
they may cause water to leak into basements unless the construction is
waterproofed. Flood waters may cause tremendous damage unless the
worst probable conditions are ascertained and the structure planned to
withstand them. The cost of these explorations is slight compared to the
possible consequences of a failure to make them.
Interested well-intentioned neighbors, friends, and old residents of the
district may give their opinions about surface and subsurface conditions at
a site. These statements may have considerable value in giving one a
Art. 3-2] IMPORTANCE OF EXPLORATION 25

knowledge of past floods, wet areas in the spring, filled areas, adjacent
mud flats that are a nuisance in dry seasons, unusually high tides, material
encountered when digging or drilling wells or when building roads and
and the conditions encountered when adjoining struc-
installing utilities,
tures were built. Although these data may be given with honest inten-
tions, they should be checked. All too often they are based upon mis-
information, hearsay, and general impression; or they may be the opinions
of persons who are not really competent to advise in such matters.
An industrial plant consisted of several small old buildings. The com-
pany decided to build alongside one of them a moderately large addition
of mill type. The company's men told the engineer who was to design the
plant that the soil was good sand. In fact, they dug a hole 4 ft. deep to
prove it and refused to pay for further exploration. The engineer
accepted this information. However, when the excavations for some of
the footings were completed, he visited the site, took a 6-ft. piece of water
pipe, and had some laborers start driving it down into the ground at the
bottom of an excavation. When the pipe was down about 3 ft., the pene-
tration per blow increased. Suddenly, the pipe almost vanished. Fur-
ther driving of more pipe showed that, after penetrating 3 ft. of sand, no
substantial resistance was encountered until a depth of 12 ft. was reached.
Of course, the foundations had to be redesigned.
What would have been the situation if the engineer had not discovered
the presence of this trapped layer of soft soil? When the building settled,
the owner would undoubtedly have blamed the engineer because the
former had not been convinced that thorough exploration was absolutely
necessary. Sometimes an owner insists upon the reduction of construc-
tion costs by using cheaper materials, weaker designs, and skimpy foun-
dations. Nevertheless, he is loath to admit that he is responsible for the
unsatisfactoriness of the structure. The old advertising slogan, "The
remembrance of quality remains long after the price is forgotten," con-
tains an important truth.
Other illustrations might be given to show the many troubles unex-
pectedly encountered when subterranean conditions were not explored.
Two more will be described.
Two small but important industrial buildings were to be located at the
edge of a harbor. "Everyone" knew that the site was covered with muck
and silt and that it was foolish to waste money exploring the area because
the structures would have to be supported by piles anyway, over 300 of
them. After about 50 piles had been driven, hard obstructions were
encountered 6 to 12 ft. below the surface. These proved to be miscella-
neous old slabs and chunks of concrete from some demolition made in the
past. Seemingly, they had been dumped from barges when the area was
under water and had been covered with the muck and silt. Piles could
26 EXPLORATION OF SITE [Chap. 3

not be driven through this debris without damaging them. Having

already driven 50 piles of a type that could not be pulled and reused, the
engineer decided to excavate the area and remove the obstructions.
After considerable delay and with a substantial expenditure, safe groups
of piles for footings were finally driven. If $1,000 had been spent for
exploration, the obstacles might have been discovered, the buildings
could have been located elsewhere, and considerable money would have
been saved.

NEW JERSEY NEW YORK

FIG. 3-1. Soil profile at the site of the George Washington Bridge. The vertical scale is exaggerated.
The drawing is made with a perspective picture of the adjacent river valley in the background.

This is an effective technique when used to present data to the public. (Courtesy of the Port of New
York Authority.)

In another case, an industrial plant was to be built in a foreign country.

Residents of the area said that it was all rocky territory, and it looked so.
However, photographs indicated deep eroded gullies where some of the
structures were to be located. Explorations were finally authorized.
The result was the discovery that part of the area was underlain by vol-
canic materials that softened when exposed to air or water.. It became
necessary to relocate a large part of this $20,000,000 plant.
Failure to spend the necessary .sum for exploration of a site prior to plan-
ning and building structures upon it is extremely shortsighted. Without
this investigation, almost everything has been done but not all to — —
ensure a good job. It is somewhat like the case of the New Jersey man
who commuted to New York via the Lackawanna ferry. One day he was
Art. 3-4] DEEP EXPLORATIONS 27

late. As the ferry was pulling out, he climbed over the gate at the slip and
tried to make a running broad jump onto the deck of the boat. Later,
when dragged out of the water, he complained that all that he missed was
the last six inches. Yes! The but they made all the
last six inches,
difference between success and failure!
3-3. Shallow explorations. There are many structures for which
shallow explorations are probably sufficient, e.g., pavements, small build-
ings, retaining walls, and small tanks. Nevertheless, the adequacy of
shallow explorations is a matter to be considered carefully.

Driving a pipe or rod a few feet into the ground with a sledge hammer
may disclose the depth of muck, peat, or soft silt in a swamp; the thick-
ness of loam or shallow soft clay overlying sand or gravel; and the depth
to rock or other firm material when it is near the surface. However, this
gives only a vague idea of the qualities of the materials, and the depth of
penetration is greatly limited.
By drilling holes with an auger welded to the end of a pipe, one may
attain greater penetration into the ground than can be obtained by driving
down a rod. The information obtained is still likely to be meager because
sandy soils will not stick to the auger so that one can examine them
thoroughly, and surrounding sand may run into the holes. Clay will
stick too much and clog the auger.
Test pits are more costly, but they enable one to examine the materials
in their natural state, to judge their firmness by the resistance to digging,
and to see firsthand what the ground-water conditions are. These pits
should be at least 2 by 4 ft. in plan so that a man can work in. them; if
shoring is required, they should be larger. Of course when they are over
6 or 8 ft. deep, excavation is rather costly.
A test pit should not be dug at the location chosen for an important wall
or footing if the pit is to be dug below the bottom of the future foundation
because the backfill will cause a soft spot. The pits should be located near
important points but not so as to endanger them.
If laboratory tests of the soils encountered are desired when digging test
pits, reasonably good samples of soil may be secured by forcing a bottom-
less tin can down full depth into undisturbed soil, excavating around the
can, cutting off the earth at the bottom of thecan by means of a trowel,
and sealing the specimen in nonshrinking, nonhardening wax. A chunk
of undisturbed clay may be taken by cutting a pair of steps in the clay
with a spade, then gouging out a long vertical slot on each side of the
sample, and finally cutting off the back and bottom with a wire saw.
3-4. Deep explorations. It is obvious that the exploration of the soil
under an important structure should be carried to a sufficient depth to
reveal all the information which is needed or which can be obtained prac-
ticably to enable the engineer to plan, design, and build the entire struc-
28 EXPLORATION OF SITE [Chap.

ture satisfactorily. What this depth may be in feet cannot be answered

by a specific figure for any and all situations. Instead, it should be deter-
mined with proper consideration of such matters as these:
1. At least one boring should penetrate a distance equal to the width of

the structure unless this exceeds 100 ft.

2. When the average pressure per square foot of the basement or first

floor of a building is to be large, the exploration should be carried down far

^y
I > * Ml
1 ft ml
m ^^0^^*
V «* NT

FIG. 3-2. A close-up of drilling procedure of a Gow-type wash-boring rig. (Courfesy of Domes and
Moore.)

enough to penetrate the soils that may be loaded sufficiently to cause

important settlement.
3. The depth to a firm stratum or rock should be ascertained so that one

may know definitely whether such material can be utilized practicably.

4. Prove that dangerous material does not underlie a firm stratum.

5. Carry the exploration to such additional depth as the findings from

the borings may indicate is essential in order to have proper information

for planning the structure.
A wash boring is one means for making deep explorations. One
method will be described:
 ;

Art. 3-4] DEEP EXPLORATIONS 29

A pipe casing is driven into the ground. A pump circulates wash water
from a sludge-receiving tank through a hose, a swivel head, and a l^-in.-
diameter pipe inside the casing. This
wash pipe has a fishtail bit used to
0-0 Surface
loosen the soil, and the water washes Cinders, sand, fill,
crushed slone
the loosened material up and into the ~SkO"., j . ,

Br/ck sand, cinders,

sludge tank, as pictured in Fig. 3-2. some silt, mud, fill

The bit is operated mechanically so as

@ « -lO'-O"

Wood, fi/l,some mud

-15-6" S//t,cinders,mud,
to produce vertical strokes 2 or 3 ft.
-18-0" some fill
long; in firm soils, the bit is rotated ©
slightly at each stroke. These opera- II ©
tions are continued until the penetra- ©
tion has proceeded as far as may be
® k Soft clay,
necessary or possible. The effluent can marine shells
be sampled, as shown in Fig. 3-2, in ©
order to determine the general constit- © a 5'

is -& Si.
18
uents of the soil being penetrated; *
8! S*
155 © 10'

this, however, yields a completely dis- §

-® 36 L|5'
turbed sample. Nevertheless, experts 60

at the business can secure data that © Gray sand,

coarse gravel
are very helpful. @ 96

In Fig. 3-3 is shown the "log" of a ^1$ 69'-0"

clay, some
boring that was made near New York 1| Soft
gravel,she//s
sand,
j4'-0"
City. The operators determined the ll Coarse sand, gravel,
§*• some clay
limits of the various strata by watch- -80-0"
ftS
ing and recording the material dis-
Fine red sand, silt
charged in the effluent, and by noticing @ ISS

the differences in driving resistance. 91-0"

® Fine fo medium
Depths are determined by computing red sand
97L 0"
the length of pipe that is below the ®
^ s
surface. These points may then be ^S o ij
@ 146 Medium brown
sand
located in elevation by taking levels to 128

determine the elevation of the ground,

@ 114

ill
or by interpolating the elevation of the 263
ES Z'/'p^ Boulder{drilled!
X7p//£ J'^Sand 3
:

top of the boring from a topographic Rock (core l /s':

recovery 2 L 0"
map. The data in the fieldman's notes diamond drill)
-I25'-I"
are then presented in this graphical
After rock was drilled, sand ran up
form. in casing 9 L 0". Had to jar back fools.
Losf core. Hit wood at -I6'-O"on
As it is usually very important to original #2 location
ascertain whether or not bedrock has Water table - 3 L "at high tide

been reached, it is desirable to drill

FIG. 3-3. Boring made for Anaconda Wire
approximately 10 ft. into the rock by and Cable Co., Hastings-on-Hudson, N.Y.
means of a diamond or steel-shot drill. (Courfesy of Philip J. Henry, Inc., New York )
30 EXPLORATION OF SITE [Chap. 3

It is advisable to have the cores at least 1^2 m in diameter. In the -

case of Fig. 3-3, there was a small boulder above ledge rock. The re-
covery of only approximately 25 per cent of the rock core indicated that
the top of the bedrock was probably seamy and partially disintegrated.
Failure to make a core boring after the casing has been driven to refusal
can lead to serious results if the proposed foundation is supposed to bear
on rock. Boulders can be present in large numbers, or sometimes slivers
or slabs of rock may be resting in a mass of decomposed rock which is
almost claylike in its plasticity. In one case of the latter situation, 24-in.
caissons sank from 1 to 8 in. before the steelwork of the factory building
—
was even completed the borings had been sunk to "refusal" only.

Acme thd-8/in.

Flat for h
wrench
Flat for wrench
Sample point Split center section Plain coupling

lasted ball.

~i 1
1
"mi 1
D. Steel
pi'n ^"lg.
l

weld Valve seal

Top view
bronze
Sample point Section A~A
Sail valve coupling

FIG. 3-4. One type of soil sampler. [Courtesy of the Raymond Concrete Pile Co.)

"Undisturbed" samples of soil are obtained by removing the inner pipe

from the wash boring, attaching a sampler or spoon somewhat like one of
those in Figs. 3-4 and 3-5, then driving this into the unexcavated soil
near the bottom of the casing. In Fig. 3-3 are shown the number of
blows required to drive the sampler a stated distance and the position
from which each sample was taken. Each sample was then sealed, num-
bered, and labeled with the depth to the place where it was taken. Sam-
ples 23=-^ in. in diameter are preferred because smaller ones are relatively
more disturbed and may not be so suitable for standard testing equipment.
Samples that are not taken skillfully may yield deceptive data. Of
course, it is obvious that the making of wash borings and the taking of soil
samples in deep-lying coarse-gravel strata and boulders are difficult.
Thin metal cylinders about 3 ft. long sometimes called "Shelby tubes" —
—are often used for obtaining samples of the soil, They are driven
Art. 3-4] DEEP EXPLORATIONS 31

directly into the ground, as are the other samplers, then shipped to the
laboratory after sealing. Again, such cylinders may be used as liners in a
split sampler.
The elevation of the ground water is also an important item to be deter-

mined. This may be found by ascertaining the top of the water standing
in the casing when it is left open for Arm and torque meter
some time. However, in drilling
through dense plastic clays, the rela- -Protractor tor
tive impermeability of the soil may measuring angle

make it difficult for water to enter or P ^5

r-Ball bearing
L
leave the casing, and the water can
enter or leave only at the bottom of
the pipe. A perforated pipe, a deep
-
Casing
test pit, or a "telltale" would be
better. If the soil is not readily

drainable, or if there are intermittent

strata of rather impermeable soils,

the last may be made by driving

down a large pipe, cleaning it out,
-Rod
inserting a small, somewhat shorter
empty pipe inside, rilling the annular
space with coarse sand or gravel, and
then pulling out the large pipe. The 4 vanes at 90°
water in the small pipe
level of the
can then be determined, and obser-
vations can be made over a long
enough period to ascertain seasonal
variations in the watertable, which
may be important.
FIG. 3-4A. Illustration of principles of equip-
What is referred to as a "vane
ment for making a vane test of soil.
test" 1
is an attempt to determine the
shearing and cohesive resistance of soils in place without taking and test-
ing undisturbed samples. The equipment, as sketchily indicated on Fig.
3-4A, is so designed that by determining the torque required to rotate a
special vane in the soil at a specified rate of a certain few degrees per
minute, the engineer can compute the shearing and cohesional resistance
of the soil when it is forced to slide along a surface of rotation at the edges
of thevanes as the latter are rotated. This equipment will enable one to
avoid the remolding and weakening of clays and silts which generally
1
Lyman Cadling and Sven Odenstad, The Vane Borer, Proceedings No. 2, Royal
Swedish Geotechnical Institute, Stockholm, Sweden, 1950. Also, G. P. Tschebo-
tarioff, "Soil Mechanics, Foundations, and Earth Structures," McGraw-Hill Book
Company, Inc., New York, 1951.
 32 EXPLORATION OF SITE [Chap. 3

accompany any spoon sampling. The presence of stones, or even of

pebbles, in the ground will probably produce erratic results.
By studying materials encountered as shown in the log of the boring
and by comparing the driving resistances,
one may obtain a fair idea of the soil condi-
tions under the surface at a particular spot.
For example, the figures shown on the left
side of Fig. 3-3, representing the resistance
of the casing when driven, will do much to
help an experienced engineer judge the prob-
able (or at least the relative) load-carrying
Tubular capabilities of the soils. The encircled fig-
head
°o ures showing the driving resistance of the
Valve sampler are even more reliable because the
W ca9e effectof friction on the casing is not in-
volved. However, all or part of the samples
should be tested at a soil-mechanics labora-
tory to obtain a better idea of their qualities.
:; _ If the foundation engineer does not perform
the laboratory tests himself, he should confer
Split
barrel with the one who does in order to obtain an
expert opinion of the qualities of the soils
and their probable suitability for the in-
Core
retainer tended uses.
rings Samples of sandy and gravelly soil taken
from the effluent from wash borings may
indicate that the soil is finer than it really
is. This may be caused by washing up the
small particles whereas the coarser ones tend
to remain in the hole and accumulate in the

Sample bottom. This situation should be watched.

catcher It can usually be detected when undisturbed
ring
samples are taken. Large borings 5 or 6 in.
in diameter may be desirable when such soils
Valve are explored. Then some type of auger or
plates
excavator can be used to obtain samples
that show the proper grading even though
they are greatly disturbed.
3-5. Soil profiles. From the logs of sev-
eral borings in a row, one may draw soil
2.414 profiles, similar to that shown in Fig. 3-6.

FIG. 3-5. A Dames and Moore These pictures are very instructive and
underwater sampler. should be drawn for various longitudinal and
Art. 3-5] SOIL PROFILES 33

transverse vertical sections. how

Nevertheless, one should never forget
they were made and from what data. The borings were taken at specific
locations only, as shown by the partial plan view and by the vertical lines
on the profile. The borings therefore do not tell what lies between them.
A picture is made by interpolating between the information given by the
various borings, assuming the materials are in some reasonable and con-
sistent arrangement. The intermediate conditions are obviously imagi-
nary, but they give a helpful picture of the probable situation. Further-
more, the lines of demarcation between different strata are not likely to be

PLAN

/ Boring No. 7

FIG. 3-6. Imaginary soil profile at site of proposed highway overpass, (Vertical scale may be
exaggerated when necessary to clarify the picture.)

so clear, and the method of their determination when the borings were
made by no means exact to the last inch.
is

In one case, borings for a vehicular structure were made approximately

150 ft. apart. Rock seemed to be at a fairly uniform elevation. How-
ever, when the excavation was made, it was discovered that two of the
borings had been located on opposite edges of a ravine, which was approxi-
mately 30 ft. deep. It was filled with boulders and constituted the bed of
an underground stream. As it was essential to place the foundations of
the structure on the rock, many thousands of dollars were collected by
the contractor for the extra work involved, probably much more than
would have been needed if the presence of the ravine had been known in
advance.
34 EXPLORATION OF SITE [Chap. 3

seems that, once a draftsman has drawn a soil profile from the boring
It
data, an engineer has checked it, and the drawing has been issued, the pro-

file suddenly becomes accurate and authoritative in the opinion of the

users of the drawing. All persons concerned should remember that it is

needlessly costly to bore every few feet in order to make sure that all the
details about the subterranean conditions are known. This may be
desirable in the case of a narrow bridge pier that must be supported upon
rock, provided this rock seems to be exceedingly irregular or sloping. It
is helpful if the soil profiles, as shown in this sketch, are made purposely
with lines at the locations of the borings in order to show everyone which
parts are based directly upon the logs of the borings and which are reason-
able assumptions.
When soil profiles are used for information, the bidders on a contract
should be told by a note on the drawing or by a clause in the contract
papers that the information shown by the profiles is not guaranteed but is
an interpretation of the data revealed by the borings. Furthermore, the
soil samples and the logs of the borings should be available for all to see,

so that each may have the opportunity to judge these matters for himself.
3-6. Load tests. Because so much depends upon the safety of a foun-
dation, it is often advisable for the responsible engineer to make one or
more load tests of the soil in the field before the final design of the sub-
structure made, provided the situation is such that these tests will fur-
is

nish pertinent and important data. This may enable him to make con-
siderable economies as well as to ensure safety. For example, if the
available data without the results of direct load tests indicate that the soil
has questionable supporting capacity, the engineer may believe that he
must limit the allowable unit bearing pressure to 1 ton per ft. 2 whereas,
if tests show that 2 tons per ft.
2
can be applied safely, he may be justified
in using the latter in his design and thus save considerably in the size and
cost of the footings, In other words, the lack of reasonably complete and
reliable information may cause an engineer to be unduly conservative.
On the other hand, he should not be willing to take unwarranted chances
because, if he does so, he invites trouble for both the owner and himself.
Any man may make an excusable mistake in deciding important ques-
tions; making a decision with the knowledge that the results are likely to
be unsafe or unsatisfactory is not excusable in engineering. Furthermore,
a moderate increase in the assumed allowable unit bearing pressure on the
soil for design purposes will seldom reduce the cost of the substructure by

any sum that is comparable to the seeming change because there is so

much other necessary construction that has to be made approximately the
same anyway the footings are the only parts that are affected directly by
;

this basic assumption as to the allowable bearing pressure.

Load tests of soil in the field may tell much about the shearing strength
Art. 3-6] LOAD TESTS 35

and compressibility ground that is tested directly. However, the

of the
test is insignificant compared to the forces that will be applied by the
structure as far as major settlements and the behavior of underlying strata
are concerned. For example, notice the following comments:
1. A load test on a surface stratum of clay or silt that is to support the

weight of a structure directly may show whether local failure of such weak
soilsunder the footings is probable.
2. A load test on a surface layer of sand 10 ft. thick will reveal the direct
bearing capacity of the sand, whereas it will disclose little about the capac-
ity of an underlying stratum of clay or silt because the sand spreads the
pressure over such a large area of the plastic stratum that the unit pressure
on the latter may be almost negligible.

(a)- LOAD APPLIED (^-BEARING PLATE (^-BEARING PLATE

AT SURFACE EMBEDDED SLIGHTLY EMBEDDED TOO MUCH

^Timber plat form Steel or^

J. separate wooden beams*
Jpading posts
A1W//VI

(d)-USE OF TRUCK TO feJ-USE OF TRUCK AND

APPLY LOAD JACK TO APPLY LOAD

FIG. 3-7. Making load-settlement tests of soil.

3. The effect of a local load upon a small area of soil may not be
extrapolated safely to apply to that of the same unit load placed upon
a large area.
4. The effect of a short-time load on plastic soils should not be extrapo-
lated to apply to that of the same unit load acting on a large area for a
long time. This statement does not apply to firm granular soils.
When load tests of the ground at a given site are planned, it is best to
know enough about the proposed structure, its location, and the elevations
of its footings to enable one to test the soil at or near places where impor-
tant loads will be applied, and at the elevation of the bottoms of future
footings. This may
require some excavation by hand. If the bearing
plate of the testing equipment is rested directly upon the top of a rela-
tively large area, as in Fig. 3-7 (a), the test may be unfairly severe on
the soil as far as a shearing and heaving failure is concerned because the
ground has little lateral support and there is no embedment around the
36 EXPLORATION OF SITE [Chap. 3

loaded area. If the bearing plate is embedded somewhat, as in (b), the

effect of the surrounding weight of soil may improve the test results and
bring them closer to a comparison with actual conditions under the com-
pleted structure. On the other hand, if the size of the excavation is too
small, as pictured in (c), and if the embedment of the bearing plate is too
deep, the restraining effect of the surrounding soil may produce test results
that appear to yield larger allowable bearing pressures than the conditions
under the real structure justify. One should always bear in mind the
small scale (or size) of the test compared to that of the proposed structure.
As a general rule, one may say that, the larger the bearing area of a load
test on the soil, the more reliable are the results. In any case, the total
applied load will probably be very small compared to that of the structure
and, as stated previously, the test yields trustworthy data about the soil
at and close to the bearing plate only. It is practicable to make a load
test upon an area of 1 or 2 ft.
2
or somewhat
,
larger. On the other hand,
areas of 4 ft.
2
or more, when same unit pressure on the soil,
tested to the
generally require such a strong loading frame and such costly work in
applying the required loads that they are not worth the extra cost.
Test loads may be applied by the use of metal blocks, by means of the
weight of water pumped into a tank on the loading frame, by filling a box
with a known weight of sand, or even by running the weighed rear axle
of a loaded truck onto beams that rest upon a loading frame with two
supports, as in Fig. 3-7 (d). This last method may be easy but it may not
yield very accurate results. A more reliable method for quick loading
may be that of running a very heavy truck or piece of equipment onto
beams over a large test pit, then applying the test load by a hydraulic
jack that reacts against these beams, as shown diagrammatically in Fig.
3-7 (e). The tests should be made carefully and well, or they will not be
worth their cost.
Load-settlement diagrams should be plotted for each test. One such
diagram is pictured in Fig. 3-8. One should not be too hasty in testing
plastic soilsbut should try to ascertain a definite yield point of the mate-
rial so that the critical loadmay be determined and the proper safety
factor of 1.5 to 2 secured when the maximum allowable unit bearing pres-
sure is selected. If there is no well-defined yield point of the soil, the
ultimate capacity of the ground may be taken as that unit pressure which
causes or accompanies a total settlement of the bearing plate of some pre-
determined amount, for instance, 1 in.
One thing that should be guarded against after issuing the results of a
load test of weak soils is a too optimistic interpretation by the owner.
When the local bearing value of a clay as shown by the test is fairly high,
itmay be difficult to convince the owner of the need for a large safety
factor to guard against harmful settlement under long-term loading.
Art. 3-7] PLANNING SUBSURFACE EXPLORATIONS 37

3-7. Planning subsurface explorations. One may properly ask

how much expenditure for subsurface explorations is justified in a given
case. This cannot be answered specifically because it depends upon
many things. Relatively large amounts may be warranted when failure
or considerable settlement of the structure would be serious, when the
lives of human beings depend upon the structure's safety, when dangerous
soil conditions are discovered or seem to exist, and when the soil conditions

12 hours 12
iL
<r24
^24
^24
,24

*24

,24

5* 1.0

12 Load Ions per

,
3
sq. f+.
4 5

FIG. 3-8. Load-settlement diagram. (Nofe: Figures at vertical breaks give duration of time in hours

between application of given load and that of the next increment of load. Ultimate bearing capacity
2 to obtain
appears to be 4.5 tons per ft. , but soil is so plastic that longer time should be allowed
maximum settlement under smaller loads.)

present one of those borderline cases in which it is necessary to choose

between a costly substructure of undoubted safety and a far cheaper but
somewhat questionable type of construction.
In general, it is best for an engineer to have reasonably good information
regarding the character, location, and weight of the contemplated struc-
tures to be built upon a specific site. He should then have a few widely
scattered boringsmade for the purpose of ascertaining the character of the
soilunder various parts of the area that may be used. Of course, these
borings should be closer together at the places where major structures are
desired. A sample of such a plan is given in Fig. 3-9. This topographic
38 EXPLORATION OF SITE [Chap, 3

map shows the outline and contours of a plot on which a corporation plans
a manufacturing plant. The principal structure is to be a large heavy
mill building; the other structures will be much smaller. The general
layout of the equipment and main building has been made tentatively.
The preferred location is shown by the solid outline one of the alternate
;

locations, by the dotted lines.Very little is known of the details of the

soils beneath the surface. Therefore, a series of borings 200 ft. apart each
way is planned as the first step in the exploration; these are numbered 1

to 13, inclusive.
If the preferred location seems to be the best and if the soils are excellent
and uniform, no further borings need be made. If the conditions at the

Secondary road
^Property fine

® First exploratory boring

o Secondary boring

FIG. 3-9. Layout of borings at site of proposed industrial plant.

preferred location are not uniform or if the findings show that more
borings are desirable, secondary borings, 14 to 24, may be made. On the
other hand, if borings \ , 2, 3, 6, 7, and 1 1 reveal the presence of weak soils,

whereas those farther east show that the ground is much better there,
then the location shown dotted may be investigated further.
Most of the preliminarjr borings may extend to a depth of 50 ft. below
the surface unless hard material is encountered at higher elevations. At
least one or two borings, such as 7 and 9 of Fig. 3-9, should go down 75 to
100 ft., especially at the preferred locations of heavy structures, if no
strong bearing is found at a higher elevation, in order to determine
whether firm materials exist within the reach of piling.
It is generally advisable for an engineer to be on the job during the
making of the borings, or to have some other engineer observe the opera-
tions for him. It is also desirable to have the contractor prepare daily
Art. 3-8] IMPORTANCE OF PLANNING 39

progress reports so that the engineer in charge of the planning may have
the information as quickly as possible and take action if necessary.
The boring contract should be made with provision for a unit price per
foot for extra borings that may be desired, in order to enable one to explore
more thoroughly under any areas that are found to have soils of question-
able value, any areas that appear to offer promise of better materials, and
other locations where further information is desired. A unit price should
be provided also for taking extra undisturbed soil samples that may seem
to be needed as any of the borings are made. This enables the engineer
to obtain more borings and samples if essential, doing so at known costs.
For example, at the site for one industrial plant, two of the preliminary
borings revealed what seemed to be a buried stream bed filled with silt and
clay, whereas the others showed good sand; therefore, a series of shallow
borings was made determine the course of this localized weak
in order to
material across the The main structure was then relocated slightly
site.

in order to minimize the number of foundations that had to be extended

down through this soft material. At the location of a group of deep pits
that were to be a part of the plant, one boring was carried down below
their bases in order to learn whether rock would be encountered, because
the excavation of rock would add greatly to the cost of construction.
At another site, 30 ft. of sand was underlain by at least 20 ft. of very fine
sand. One of the structures was to be a heavy blast furnace. Therefore,
at the preferred location for the furnace, another, deeper boring was made,
and rock was located at a depth of 83 ft. This showed that piles could
be used if it was decided that the soil was untrustworthy.
When preliminary borings are made for a bridge whose location is not
fixed by other considerations, it may be desirable to make a series of bor-
ings along various possible alignments in order to determine which one is

best from the standpoint of foundations. The conditions at the loca-

and abutments along the best alignment can
tions selected for the piers
then be explored more carefully, perhaps by means of a group of borings
spaced 20 or 30 ft. on centers both ways. If the customary 10-ft. core
borings at the location of a heavy pier show that the rock is badly seamed

or poor in quality, some of these borings may

be carried to considerably
greater depths in order to locate sound bedrock.
3-8. Importance of subsurface conditions to the planning of
structures. It is important for an engineer to know the details of sub-
surface conditions at a site before he completes the design of the structure.
There are many reasons for this. These conditions may also have great
effect upon the planning of construction procedures. Unanticipated con-
ditions generally cause additional trouble and expense.
Nonuniform plastic soils may limit the use of continuous and rigid-
frame structures because of the danger of serious unequal settlement,
40 EXPLORATION OF SITE [Chap. 3

thereby causing the designer to plan for the use of simply supported struc-
tures, or those utilizing some other determinate type such as cantilever
bridge construction. These soils may prevent the use of arches on
account of the likelihood of horizontal yielding of the earth. Heavy
multistory structures may be inadvisable because of probable settlement
and cracking. Lightweight flexible structures of steel or wood may be
preferable so that distortions may occur without serious harm.
In some cases, the shape of a building may have to suit the area of good
soil upon which it can be placed. The span of a bridge, or several spans
in a long structure, may have to conform with the conditions under the
ground. For example, rock may be at or near the surface in one or more

FIG. 3-10. Illustration of alternative alignments for a bridge.

places so that these are the most advantageous locations for piers or abut-
ments ; silt may have filled certain old channels that should now be avoided
if possible ; and boulders may cover the rock at certain places so that the
sinking of caissons and the driving of piles would be hazardous there.
Weak soils may require the use of many small light columns and spread
footings on the one hand or, on the other, the use of long spans with a
minimum number foundations but with each one extended down to
of
firm material by means of piles or caissons. Great engineering skill may
be required to secure the best coordination of architectural features,
utility, safety, construction, and economy. This is engineering planning.
happens that someone wishes to build a structure of a certain
It often
character, of a specific size, and of a particular material on a given site.
Unfortunately, it sometimes happens that the plans have progressed con-
siderably before the subsurface conditions at the site are explored. The
engineer then faced with the problems of planning a foundation that will
is

hold the structure safely with a minimum reasonable cost, of persuading

the owner to modify his plans to accommodate the existing conditions, or
Art. 3-8] IMPORTANCE OF PLANNING 41

of convincing the owner that he should locate the structure upon some
other more favorable site. If the engineer accedes to the owner's wishes

against his own best judgment and if the structure is not satisfactory, the
engineer will still be blamed for the results because, in time, the owner is
likely to say, "Why didn't you explain these things to me more fully and
show me that the situation was so bad?"

FIG. 3-11. A primary crushing plant under construction on a mountainside. The structure itself is a
heavy reinforced-concrete buttress designed to prevent any slippage of the rock behind it. (Courtesy
of Cananea Consolidated Copper Co., S.A.)

Many times proper subsurface explorations have yielded information

which affected the design of a structure and which enabled the engineer to
save far more than their cost by avoiding design revisions and charges for
extrawork not included in the original contract. Many times the reverse
has happened when adequate data were not secured.
Consider these cases in addition to those cited previously:
1. The primary crushing plant mine was to be a heavy concrete
at a
structure about 80 ft. high from the base to the main working floor, and a
42 EXPLORATION OF SITE [Chap. 3

steel superstructure was to be above that. The concrete part was to be in

a niche excavated in a rocky mountainside. Diamond core borings, made
to the level of the invert of the substructure, revealed that much of the
upper part of the rockwas seamy with disintegrated claylike material in
the seams. Furthermore, these seams dipped downward at about 70°
from the horizontal, creating the possibility that the uphill side of the cut
would slide into the excavation. A truck roadway to be at this uphill side
of the structure made it inadvisable to remove these upper layers behind
the structure completely. Under these conditions, the structure was
made as narrow as possible so that the niche to be blasted out of the rock
would probably be small enough to enable the remaining rock to arch
around the cut. The structure was then made longer in a direction per-
pendicular to the hillside, and it was designed with two outer walls and
two inner partitions of heavy reinforced-concrete construction which bore
against the rock and acted as huge buttresses that would not let the rock
start to move. These considerations were the starting points in the plan-
ning of the structure which is shown under construction in Fig. 3-11.
2. Another heavy industrial structure of a character somewhat similar

to that just described was to be located in " solid rock." Adjacent hillside
cuts and other excavations for related structures were made, then, belat-
edly, a core boring was driven down to the bottom of the proposed struc-
ture. Here was found an old inactive fault zone of shattered rock cutting
on a slope through the region at and just above the proposed invert.
Neighboring work had proceeded so far that it was not practicable to
choose another location. Although resting a heavy structure across a
fault is not recommended practice, it was done in this case because a major
earthquake seemed to be improbable and, if one should cause a serious
movement of this fault, the rest of the plant would probably be ruined also.
However, since the structure was very heavy, the design was made so as
to provide a thick mat of reinforced concrete at the bottom to spread the
loads. Since the fault would probably be a source of water under high
pressure, means were taken to provide drainage and to prevent the build-
ing up of hydrostatic pressure. These things were provided easily because
known in advance.
3. A
heavy building was designed to be supported upon rock. The
"top" was located by means of wash borings, but core borings
of the rock
were not made. It was discovered during construction that the boring at
one corner had hit a boulder. The foundations in this vicinity had to be
revised and carried down much deeper than intended, with a large and just
claim for extras by the contractor. If the contractor and the engineer had
known the conditions beforehand, both the design and the method of con-
struction could have been adapted to them with much saving in the total
expenditure.
Art. 3-8] IMPORTANCE OF PLANNING 43

4. Sometimes an investigation
of ground-water conditions will avoid
much Figure 3-1 2 (a) pictures one kind of possible situation.
trouble.
Here the substructure was to be supported upon sand. When the clay
stratum was penetrated, it was discovered that the ground water in the
sand layer had been trapped under this clay, and the water was under
sufficient pressure to cause it to enter near A and to fill the excavation to

B. Wellpoints had to be driven to the sand around the excavation, and

heavy pumping was required to remove the water fast enough because of
the large volume of the supply. The entire basement had to be redesigned
to resist the pressuresand attain waterproofness.
Even shown in Fig. 3-12(6), ground water may cause
in a case like that
trouble. Here the basement of a building several hundred feet long was
built on rock. The soil explorations were made during dry weather when
the ground water was negligible, and no attempt was made to ascertain
conditions at other times. The foundation wall above D served as a dam

™*^ |
Basement j
gfc#£-
e

Coarse sand & Hydraulic

head
(a)- EXCAVATION INTO A POROUS, WATER- (b)- FOUNDATION ACTING AS A DAM
BEARING STRATUM

FIG. 3-12. Examples of troubles caused by ground water.

to impound the water flowing down to E. This water would eventually

flow around the building, but it caused dampness and some leakage that
were objectionable and costly to rectify. A knowledge of the conditions
would have enabled the designer to plan for the waterproofing of the uphill
wall, for the installation of collecting drains along the side at or above D
and around the ends of the basement, or for the building of a few drains
from E under the basement floor to C.
In one situation explored during the dry season, a test pit was dug 15
ft. deep so that the soil could be examined carefully. This revealed occa-
sional thin layers of small gravel and coarse sand that were devoid of fine
particles. The engineers concluded that these strata were clean because
they were small watercourses during wet weather. This proved to be the
case.
5. A harbor was to be improved during World War II. A cursory
examination of the general vicinity indicated a muddy bottom and shal-
low water. Because of the great rush, soundings were made to determine
the depth of the water at the site of the pier; however, no time was spent
making borings. A big hydraulic dredge was dispatched to the site, and
material was assembled for the construction of a pier to be founded upon
the assumed materials. When the dredge arrived and excavation work
44 EXPLORATION OF SITE [Chap. 3

was started, it was discovered that

solid rock lay under a very thin stratum
of mud Of course, the entire plan had to be changed
throughout the bay.
radically, and much valuable time was lost.
3-9. Value of investment in engineering. The emphasis upon
careful planning and study of foundation problems serves to bring to the
reader's attention the question of justified expenditures for engineering in
general as well as for the exploration of a site. Sometimes one hears that
the engineering costs of a certain big job were 4.45 per cent of the cost of
the completed work, whereas the engineering cost on some other project
was 7.82 per cent This gives the impression
of the cost of construction.
that the engineers connected with the first project were much better and
more capable than those who worked on the second job. This may be
utterly fallacious for many reasons, some of which are the following:
Engineering costs vary greatly because of the vastly different char-
1.

acter of the work to be done.

2. Engineering costs are reduced in terms of the percentage of the con-

struction cost when there is considerable duplication of parts. For exam-

ple, a steel-frame mill building 100 ft. long with 5 typical 20-ft. bays on
uniformly good soil is likely to cost almost as much for the exploration,
the design work, and the drafting as would a similar building 400 ft. long
with 20 of the same typical bays. On the other hand, the construction
costs of these two buildings would vary somewhat as their lengths.
3. In general, small jobs involve more engineering expense, relatively,

than do large ones of the same basic character.

4. Engineering work that is done hastily and carelessly may be cheap in

itself but expensive to the owner. For instance, a complete project may
have cost $500,000, whereas the engineering was kept to a minimum,
say $25,000, or 5 per cent of the total cost. On the other hand, the
engineering might have been done with careful study and great skill, with
an expenditure for engineering of $40,000 but with a total cost for the same
project of $460,000, giving a relative cost of engineering of approximately
8.7 per cent. Nevertheless, the owner in the second case would have
saved $40,000 in construction costs, and his investment of $15,000 more
in engineering would have paid excellent dividends in cash as well as in
satisfaction.
5. Sometimes various engineering expenses are hidden in the construc-
tion cost. One should be sure that he understands the bookkeeping
involved in the cases that he compares.
If a project is satisfactory because it has been well planned and carefully
worked how is one to compute the monetary value of the engineering
out,
servicescompared to the "cheap" job with which no one is pleased? Sel-
—
dom is an owner grateful for small savings and they are relatively small
— made at the sacrifice of safety, utility, and the general satisfactoriness
Art. 3-10] EXAMPLES 45

of the structure. The importance of careful investigation and study of

foundation conditions and of all other engineering problems connected
with a job is self-evident.
3-10. Examples. An oil company wishes to build an addition to its
tank farm alongside tidewater. Recent dredging operations have pro-
duced an extensive hydraulic fill next to the company's property, and the
area of this new land is adequate for three tanks and their dikes. The
surface of the ground is approximately 6 ft. above mean high water
(M.H.W.). A typical proposed tank is 150 ft. in diameter and 40 ft.
high. A partial plan of the site and the soils as shown by the log of

El. 106.2
•))><)» <IIl_

rNo.7
Hydraulic MHW*£/. 100
fill -
laminated MLW=EI. 9S
'i. Proposed ~r'v
"'.'-

91
tank No. 18
Beach_
\No.6 sand 86.S
Proposed Silly _ 5i-~~ ^

sand 81
(tank No. 20
^y Proposed ^ v Borinq
^tank No. 19 S C6J-L06 OF

— » <
/\ No.l BORING N0.1

I No. 3 \
/ \
y yy
/
~XNo.2 + Ala 4
X

Edge of hydraulic fill

at mean high water

CctJ-PARTIAL PLAN OF SITE

FIG. 3-13. Data for exploration of site for some new oil tanks.

exploratory boring No. 1 are given in Fig. 3-13. Is it safe to support

these steel tanks primarily upon a sand cushion placed directly upon the
fill?

First, one should obtain a general idea of the magnitude of the things
with which he is dealing. Some data for this case are the following:
1. A 40-ft. depth of oil will cause a unit pressure of nearly 2,500 p.s.f.

on the soil.

2. A 150- by 40-ft. tank, when filled, will cause a total load of nearly
22,000 tons.
3. At
15 cents per pound for steel and 12 cents per gallon for oil, such
a tank would be worth over $750,000.
full
When one is faced with the problem of founding a heavy important
structure upon hydraulic fill, he may have something to worry about. He
should remember that such fills are likely to contain lenses of material of
46 EXPLORATION OF SITE [Chap. 3

different characteristics and to be poorly consolidated. The coarse mate-

rials generally settle near the point of discharge, whereas the finer particles
spread out widely. From time to time, the discharge pipe is usually

moved to different locations. This may cause local deposits of coarse

materials to be above or below strata of fine loosely compacted soil. The
magnitudes of the differential settlements are difficult to predict. If silt,
or soft clay underlies the hydraulic fill, add to the uncertainties.
this will
If the watertable is near the top of the heavy pressure over a large
fill,

area may cause plastic materials and very fine sand to flow out from under
the loads.
is some silt in the sand stratum below
Here the boring shows that there
the hydraulic The driving record indicated that this layer permitted
fill.

penetration even more easily than did the fill. However, what are the
conditions and probabilities elsewhere?
The value of three tanks and the cost of deep foundations to guarantee
safety justify the expenditure necessary to explore the soil thoroughly.
The one boring indicates that the sand 15 ft. below the surface is probably
adequate if the upper strata can transmit the load to it. Since the danger
seems to lie in the upper 15 ft. and the silty sand, it appears that at least
six more shallow (25-ft.) borings should be made at locations 2 to 7, shown
in the figure, in order to obtain a better idea of the probable value and
character of the soils. Since none of the original borings was deeper than
25 ft., one new one, like No. 4, might well be carried down 25 to 50 ft.
farther in order to make sure that no dangerous plastic materials underlie
the sand. This is advisable because the extent of the loaded area under a
tank is so large that there is a relatively small rate of decrease in unit
pressure caused by the distribution or spreading of pressure by the soil.
Although load tests of the upper strata cannot simulate the effect of
such extensive loads as those of the proposed structure, a bearing test of
both the fill and the underlying stratum might well be made near one or
two of the borings, as shown in Fig. 3-13 by the symbols T\ and T%.
Further tests should be made if the borings and the results of the first
load tests indicate that they are necessary.
Here is another example. A four-lane highway was built on an
embankment approximately 30 ft. high above the original grade. The
borings taken along the center line of the roadway showed that there was
about 4 ft. of soft material (silt or clay containing vegetable matter) in a
localized swampy area which extended for a short distance off to one side
of the highway. It appeared that this condition was not serious and did
not extend over enough area to require additional borings. Hills of gravel
were on both sides of this little swamp. The fill for the embankment w as r

made without removing the soft material in the swamp. Some 8 or 10

months later, slight evidence of settlement of one side of the embankment
Art. 3-10] EXAMPLES 47

was noticed. Then approximately 8,500 cu. yd. of fill was added as a
blanket over the swampy ground just beyond the toe of the slope, as
shown in Fig. 3-14(a), this being for the purpose of preventing the soft
material from squeezing out. However, most of this fill gradually settled
until only a few feet of height remained above grade. Later on it was
ascertained that the bottom of this fill had settled approximately 15 ft.
below the top level of the swamp.
Shortly thereafter the engineer happened to be examining the site when
one side of the embankment cracked at the top and slumped as indicated
in Sketch (a). At the same time, he actually saw the effect of a mud wave
traveling away from the embankment, because he could see the tops of

{b) PLAN

FIG. 3-14. Example of subterranean surprises.

the trees wave and moved the trees bodily. The length-
as the flow lifted
wise extent of the slump was somewhat as pictured in Fig. 3-14(6). Bor-
ings made later showed that just beyond the embankment, the silt
extended for a depth of 50 ft. Evidently there was a small ravine, or a
large pothole, which had become filled with soft materials after the depar-
ture of the glacier.
How is one to tell in advance what expenditure for soil investigations
is justified? Generally, such work might better be more thorough than
necessary rather than less so.

PROBLEMS
3-1. Draw a soil profile along the center line of the borings shown in Fig. 3-15.
Use a vertical and horizontal scale of 1 in. = 30 ft.
48 EXPLORATION OF SITE [Chap. 3

,1?

244 215 213 192

'^Lpam^\ -242 W.L b"a'rn^ 211 ?i'£'d'ei rrrfr 188
Sandy. Fine sand. 203 F/nesand. 181
230
Clay— 175
E^C/ayEE
189 San id'
211 160
Sand-
203 173 149
% o o "t>
° o
175 \^f^m
° Ora veh Rock
179 ISO No. 5
No. 4
Gravel" Rock No.3
165
Wmm
Rock
No.2

No.l

FIG. 3-15. Logs of borings.

El. 107

^EXISTING BUILDING

FIG. 3-16. Plan at a power plant.

 PROBLEMS 49

3-2. Draw the soil profiles along the lines connecting borings 1, 2, and 3, also 4,
5, and 6, in Fig. 10-51 (Chap. 10). This is alongside a river so that scour from
floods through the ages has washed away portions of the ground, then other
deposits have been made. some of the troubles encountered in
It illustrates
interpreting boring data.
3-3. An electric power company is planning an addition to a generating station
at the water front of an important city. The extension is shown bj^ the hatched

areas in Fig. 3-16. Determine the locations, type, and depth of borings to be
made before the foundations are planned. The present building is supported
upon piles that penetrate through a few feet of fill and 25 ft. of harbor silt to fine
sand.
3-4. A friend is planning to buy a lot on which to build a one-family brick house.
He has asked you to advise him in the matter.The lot is along a suburban road
and has a width of 100 ft. and a depth of 200 Near one front corner is what
ft.

appears to be the top of a flat boulder. Around it for several feet are a few
scraggly bushes and weeds. The ground slopes downward diagonally from this
corner to the opposite rear corner, which is approximately 6 ft. lower. Crossing
this corner there is a small brook amidst alders and a few swamp maple trees.
Miscellaneous trees and bushes are scattered between these two corners. What
will you do, or tell your friend to do, before purchasing the property?

FIG. 3-17. Plan of a proposed industrial plant.

3-5. Refer to Fig. 3-17. Plan and specify the borings to be made so that pre-
liminary studies can be made to determine the general types of foundations for the
structures.
 4
SOME PRINCIPLES OF FOUNDATION
ACTION

4-1. Introduction. Various assumptions and general principles re-

garding the action of under loads are described in this chapter.
soils
These should be understood thoroughly so that they can be used in the
planning and the detailed designing of foundations for a great variety of
structures built in many different situations. Undoubtedly there are dif-
ferences of opinion held by various individuals regarding these assump-
tions and the proposed technique of procedure. Yet, if rightly used as
tools, the principles set forth will yield reasonably satisfactory results,
will enable the reader to visualize his foundation problems more clearly,
and will help him solve these problems safely.
Probably it cannot be stated too often that the exact properties of soils
are unknown, reasonable approximations of them are generally the best
that are available, and these properties vary greatly. A given stratum
is likely to be a mixture of materials that give to it some of the qualities

of each of the major constituents, the properties of each being modified

by those of the other different materials. The degree of compaction and
the moisture content affect these properties, too. Therefore, the factor
J — meaning judgment — is always to be inserted in the equations used by
the foundation engineer.
4-2. Distribution of pressure. Assume that a footing like that
shown in Fig. 4-1 (a) rests upon the ground and supports a total vertical
load P. The footing is in equilibrium; therefore, the resisting pressure
under must equal —P. Is this pressure distributed uniformly over the
it

area of the bottom of the footing, as pictured in Sketch (a)? Is the

intensity higher near the edges, as in (6)? Or is it larger in the center,
as in (c) ? A relatively compressible soil may produce pressures somewhat
like those in (b) ; a rather incompressible soil, like those in (c). Probably
none of the results is exactly as pictured in (a). Nevertheless, for prac-
tical purposes, this uniform distribution may be assumed for the design
50
Art. 4-2] DISTRIBUTION OF PRESSURE 51

of the footing itself and for an analysis of the pressures in the soil under
it,because this assumption is probably as accurate as are those that must
be made concerning the action of the reinforced-concrete footing, the soils,
and the magnitudes of the loads.
If the footing is on plastic soil, the inequalities of pressure may equalize

gradually because of the greater consolidation of the areas that are loaded

384000 #
A B
D
^AP=P AP /'

r~
faj-UNIFORM PRESSURE
cc
\P
\
& / \
A B
//Ali
^ / \ <J

PW1
\
j J \
E1 G 1
ZAP=-P A Q\ \ \N' \0 \ \ \R A
r6;-HIGH AT EDGES
JP
rrfJ-ASSUMED DISTRIBUTION OF PRESSURE

R'
--- -— A
/ N
/ \
1
\
I/ 1 {Footing \
-
XJP=-PJ AP \

r \
1 SP

fcj- HIGH AT CENTER 1

1 | 1*5
P 1 1

t£
S \
\
/
/ <5

\ /
=ss=/ Clay ,y "
nr S n doi
Q'

7v < LM+h=28' ^
Clay
w (e) -VARYING MATERIALS
(f)- SECTION A-A

FIG. 4-1. Assumed distribution of pressures into soil under a footing.

more heavily. If the footing is on granular soils that can withstand the
necessary pressure, an increase of intensity near the center of the footing,
above the average pressure, will not harm the footing; if the soil yields
slightly under the center, the edge pressures tend to rise toward the
average.
When the footing in Fig. 4-1 (d) is loaded, it is obvious that it will not
merely compress a rectangular prism of earth pictured by Shear- LMON .

ing stresses on the soil within LN and will affect the soil outside of MO
these planes and cause a spreading of the load. The angle a will depend
52 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

upon the shearing strength and cohesion of the soil: furthermore, the sur-
face QL may not be clearly delineated at all. In the case of weak soils
with small internal frictional resistance and low cohesive strength, a will
be smaller than for firm granular materials. As a general average, it is
sufficient to assume that distribution of the load will spread out at a slope
of 2 : f , shown in Sketch (d)
as When a series of differing strata is pres-
.

ent, the distribution may vary as pictured by the broken line TUVW in
Sketch (e) ; however, this refinement is unnecessary and, undoubtedly,
still inaccurate.
The area QQ'RR' in Fig. 4-1 (/) indicates that, as the depth h increases,
the assumed affected area of soil approaches a circle with a diameter of
P'= large load
" small load

{a) [b]

FIG. 4-2. Bulbs of pressure assumed for purposes of visualization.

LM + and that the intensity of pressure AP decreases rapidly as h

h,

increases. For purposes of study and comparison of designs, it is suffi-

ciently accurate to assume that the resisting area retains the general shape
of the footing. The corners may be considered as quarter circles if one
prefers to do so. The error caused by this assumption is greater in the
case of small footings than for large loaded areas. However, when the
depth of the plane being investigated is great with respect to the dimen-
sions of the footing, the intensity of pressure is probably decreased to
unimportance anyway. The pressures that exist outside this assumed
pyramid are probably relatively small.
For example, if P in Fig. 4-1 (d) is 384,000 lb., the footing is 8 ft. square,
and h is 20 ft., the intensity of pressure p at is 6,000 p.s.f., whereas LM
at QR it is approximately 384,000/28 2 = 490 p.s.f. At such depths, it is
probable that the unit pressure at the center of the assumed affected area
of the plane QR is greater than the average, perhaps by as much as 50
Art. 4-21 DISTRIBUTION OF PRESSURE 53

per cent. Nevertheless, this is not serious in the case of small foundations
for which the critical soil pressures are directly under and close to the
footing. Large heavy structures and weak substrata are discussed more
fully elsewhere. The principle, however, is illustrated by Fig. 4-2 where
ACEDB pictures a bulb of pressure bounded by a curved surface that rep-
resents an assumed locus of points of equal pressure. This method of
picturization is useful in studying the action of soils.
Now compare the effects of a large loaded area and a small one upon
the unit pressures assumed to exist at a plane in a stratum below the

400 noo' 40,000,000*

Assumed lines »'±t n»\

Gi —tu 7

/\ipkKff\ of distribution 1 1
100'x 100'
// \f. limits i
/
CI \ n A \K
30x30' 120"x 120'
7l
j;
EL Jf
50x50' 140'x 140
(a)-COMPARISON OF RELATIVE DISTRIBUTION OF PRESSURE UNDER SMALL
AND LAR6E FOUNDATIONS
# , „#.
640,000* 640,006 640,000*
. ?/v '.
f
20' •

68'

TV X

'

Overlap of
* ^Bulb ofpressure (assumed pressure lines
for picturization)
Section A-A
(^-DISTRIBUTION OF PRESSURE FROM A ROW OF FOOTINGS

FIG. 4-3. A study of the distribution of pressure in soils under various foundations, assuming 2 :
1

distribution.

structures. Figure 4-3 (a) pictures a footing A B and large structure GH.
Assume that the intensity of pressure at AB and GH equals 4,000 p.s.f.
According to the assumption of 2 1 distribution, the intensities of pres- :

sures at 20- and 40-ft. depths below these structures are the following:

400,000
At CD, p 20 = 445 p.s.f.
30 X 30
40,000,0 00
At JK, p 2 o = 2,780 p.s.f.
120 X 120
400,0 00
At EF, p i()
= 160 p.s.f.
50 X 50
40,000,000
At LM , Pao
140 X 140
= 2,040 p.s.f.
54 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

These estimates indicate that A B does not have any important effect upon
these two lower strata, whereas the pressures from GH are reduced only
about 50 per cent at a depth of 40 ft. Therefore, a weak stratum under-
lying a large heavy structure may constitute a real hazard.
In the case of a row of closely spaced footings like that shown in Fig.
4-3(6), the assumed "cones or pj^ramids of distribution" overlap. If the
pressure at the footing SU is 10,000 p.s.f., that over WT alone would be
640,000/28 2 = 820 p.s.f. Because of the overlapping at VT, the pressure
in this area would seem to be doubled, 1,640 p.s.f. However, it is not
reasonable to suppose that the unit pressure from to V is one value W
whereas that from V to T is twice as much. In such a case, it is best to
assume that, when small overlapping occurs, the average unit pressure at
WX is rQ rz sq — = 1 ,000 p.s.f., but that the intensity through the cen-

tral portion will be larger and that at the ends and edges will be less,
somewhat shown by the bulb-of-pressure line WYX. When there are
as
large overlaps, it is more conservative to add the two computed pressures.
The Boussinesq equation
5/2
3Q 1

Cz — o (4-1)
1-KZ
S2
1 + {r/zY

may be used to estimate the vertical pressure at some point at depth z

and horizontal offset r caused by a concentrated load Q. This formula
assumes that the earth mass is homogeneous, elastic, and isotropic, which
it seldom ever is. On the other hand, it gives a reasonable idea as to the
probable magnitude of the vertical pressure, and this information is very
useful.
On what is usually wanted is the total unit pressure
the other hand,
at one or more points caused by a number of local loads or by distributed
loads acting over one or more significant areas. The Newmark circle 1

shown in Fig. 4-3A is a way of approximating the theoretical pressure at

a point caused by uniformly distributed loads over a large area without
having to use the Boussinesq formula directly, although the diagram is
based upon this formula. The distance z is the vertical depth of the point
to be investigated below the bottom of the bearing area. This distance
is drawn to some scale ab, and this is then used as a unit distance. With
a convenient point as center, circles may be drawn with radii equal to the
corresponding proportions of z that are shown in the table in the illus-
tration. Twenty sectors are then drawn with central angles of 18°. A

1
Nathan M, Newmark, Influence Charts for Computation of Stresses in Elastic
Foundations, University of Illinois, Engineering Experiment Station Bulletin, Series
338, Nov. 10, 1942.
Art. 4-2] DISTRIBUTION OF PRESSURE 55

diagram of the loaded area isalso drawn to the scale of z on a separate

piece of thin tracing paper. This latter sketch is to be laid over the circles
with their centers at the position of the selected point z in the plan. Each
area between two adjacent radial lines and circumferences is called an
influence unit. The number N of these units included in whole or in sub-
stantial part within the outline of the loaded area is determined. If the

Ifz*l
o Pad.
© 0.2.6
C?) 0.40
» 0.52 Influence unit O.OOS z .

€0 0.63
<rz (*/o'J =q(*/o')x (No. ofinfluence unils)xO.OOS
tt 0.76
G) 0.91
0) 1.11
® 1.38
(£) 1.90

FIG. 4-3A. Newmark influence chart for vertical pressure at a point. Draw plan of loaded areas to a
scale such that Z equals ab. Place point of 6 Z on plan at center of Newmark circles. Count influence
areas. [Courtesy of Nathan Newmark.)

uniform load on the bearing area is q p.s.f., then the vertical pressure at
z is computed as

<r s = q XNX 0.005 p.s.f. (4-2)

To assume that an L-shaped apartment house

illustrate the procedure,
is planned. At a depth ft. below the basement slab is the top of a
of 40
10-ft. stratum of soft reddish clay that may cause harmful settlement.
The average bearing pressure directly under the basement slab is 1,500
p.s.f. The intensity of vertical pressure at the center of this layer caused
by the structure
is to be estimated. The pressure under the vertex of the
reentrant angle of the "inner" walls will be computed as one specific case.
56 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. A

The distance ab of Fig. 4-3A therefore represents the 45 ft. to the middle
of the 10-ft. layer to some scale. The outline of the building drawn to
this scale is shown by the dotted lines in Fig. 4-3 A, and it is placed with
the chosen corner at the center of the circles. The approximate total of
the influence units is taken as N = 123. Therefore,

cr, = 1,500 X 123 X 0.005 = 920 p.s.f.

Computation of the pressure at this point by the 2 : 1 method with over-

lap of assumed distribution areas gives

1,500(50 X 125) 1,500(50 X 75)

=
(50 + 45) (125 + 45)
+ '
(50 + 45) (75 + 45)
1,070 p.s.f.

This comparison of the results of the two computations is made for interest
only. It is not to be considered a check of the 2 1 method because prob- :

lems may be compared the results of which vary widely.

A *" (Heaving
f G \
p M
v/,y/t
r~j
'

^T/fr £1 » ffT
Setflemem~^~- Direction of g J
movement ofearth
(a) (b) (c)

FIG. 4-4. Principle of failure of soil in shear when under pressure.

4-3. Shearing of soil. When a soil is subjected to the load of a struc-

ture, as in Fig. 4-4 (a), there it to squeeze out from
is the tendency for
under the load. It may try to go one way as shown in (a), or theoreti-
cally, for a symmetrical structure and perfectly homogeneous soil, it may
move out both ways as in (6). If real movement of the ground does
occur, there will be a sliding movement along some such surfaces as ABC
and DEF, so that the shearing and cohesive resistance of the soil must be
overcome; the structure will settle, and the soil will be heaved up as indi-
cated by the dashed lines. This action is especially to be guarded against
when the unit pressures under the structure are large. It is obvious that
sand and gravel, because of their large internal frictional resistance, are
not so likely to fail in this manner as are plastic soils. In general, the
resistance to both shearing and compressive forces becomes greater as
the depth increases in the case of sands, but this is not so likely to be
true for clays. Vane tests have indicated that, in some instances, the
shearing resistances of clay at various depths did not differ greatly.
A structure placed directly on top of the ground may squeeze out
plastic soils rather easily. If the bottom of the structure is depressed,
or embedded, some appreciable depth GH, Fig. 4-4 (c), this will tend to
increase the resistance to a shearing failure. The weight of the soil above
KL must be lifted if the earth mass KHJL is to move upward and to the
 :

Art. 4-3] SHEARING OF SOIL 57

right. very important, especially when one must utilize

This principle is

weak soils. to embedment tends to increase sur-

The back pressure due
prisingly the unit pressure that can be applied safely to a granular soil.
In fact, moderate pressure tends to make a sand firmer.
Although this may be greatly exaggerated, the benefit of embedment
in granular soils may be visualized qualitatively as follows, referring to
Fig. 4-4 (c)
1. The weight of the load P as a surcharge will cause a unit pressure
of p = PIA at UK. Then p will cause a lateral pressure w h on an imagi-
nary vertical wall at KJ. This pressure, if the angle of internal friction
of the soil is 30°, may be assumed to be perhaps w h = }^P because of the
action of the friction in helping to support the structure.
2. The horizontal pressure Wh on KJ will in turn cause a lateral (in

this case a vertical) pressure acting upward on KL. This pressure w v

may be %Wh because of the internal friction opposing the movement of

the soil.

3. The weight of the soil at KL produces a downward force to counter-

act w v, but it is limited to the weight of the material.
4. Therefore, one can see that if the coefficients were correct,

w = v y3 w —h
i
9P

In other words, the weight of a few feet of granular earth, GH, can be very
beneficial in preventing a shearing failure and upheaval of the soil due to
a heavy structure.
The shape of the loaded area may be important in connection with a
possible shearing failure of the soil. Figures 4-5 (a) and (6) show a long
narrow structure and (d) a square one
;
(c) ,
(a) (c)
of equal weight. For purposes of visual-
j
ization, the shearing action is shown as i -lB
affecting the surfaces ABC and FGH and
the earth masses ADCE and FJHK. The B G
resistances in these two cases may vary
considerably. Since plastic soils are the J
ones that may cause trouble because of
shearing and plastic flow, it seems reason-
C F\ w-
able to assume that the larger perimeter
of the narrow structure will result in more m
area of the surface of possible slippage
rS)-P\ar\A-A CdJ-P\anB-B
from under and inasmuch as the
its edges,
FIG. 4-5. Effect of shape of foundation.
total cohesive resistance depends partly
upon this area rather than upon the weight of the structure alone, there
should be a larger resistance to movement under the narrow building than
under the square one. Since the magnitude of a frictional resistance
58 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

depends upon the normal pressure, this frictional resistance should be

approximately the same in both cases. The inertia of the mass of earth
to be moved in the event of failure should not have much effect in either
case because the motion of the soil would be too slow to bring this feature
into importance.
It may be possible to excavate a stiff clay and a caliche soil so that the
side AB of Fig. 4-6 (a) will stand temporarily in a practically vertical posi-
tion. However, it may not remain so. Some portion, such as BCA, is

Berm 10'orjh min.

Floor //he.

repose <f>

Footing

(d) (e)

FIG. 4-6. Dangers when foundations are placed near cuts or have little embedment.

likely to slide down to FGHB because of its own weight, particularly when
water and weathering soften the soil. Any extra load placed to the right
of C will hasten this sliding. A loose granular soil seldom stands up on
a very steep slope, not even temporarily.
If a structure must be founded near the top of the slope, the earth
should not be steeper than approximately l 1^ as shown in Sketch (b);
: 1 ,

even then the structure should be set back from the top corner K, and the
berm JK should be rather wide because the lack of support to the right
of KM will aggravate the tendency for the load to cause slippage along
LM.
Art. 4-4] SETTLEMENT 59

In the case of the footing NO of Fig. 4-6(c), a trench is to be dug for

the drain Q. Even though the sides of the trench are shored and the
trenchis backfilled, the footing NO is endangered. The drain should be
moved out to some position *S where the slope of OR is not steeper than
1^:1 (preferably flatter), or the footing should be lowered to the level
of the bottom of the trench. Conservatism in such matters as these is

desirable.
When studying this question of embedment and its beneficial action
in preventing a shear failure of the soil, one should look for the weakest
possible link in the "chain." For example, assume a condition like that
pictured in Fig. 4-6(rf), which shows part of the side wall of a feeder com-
partment under a large ore bin. When the bin is full and the pressure
on the footing is greatest, the surcharge above the bearing level of the
footing on the outside in this case was over 50 ft. of ore. This depth
affects the holding-down pressure on the outside only. If the soil is to
fail by shearing, it will of necessity slide inward as indicated in Sketch

(e), so that the embedment or downward holding value of the soil above

the footing is only the weight of the "floating" floor slab and 1^ ft. of
earth. In such a case it would be better to use a continuous mat under
the structure if the equivalent of deep embedment were necessary.
4-4. Settlement. When a structure is built upon earth, some settle-

ment of that structure is practically inevitable. A few general princi-

ples to be borne in mind when studying problems of settlement are the
following
1. Granular materials like gravel, coarse sand, and medium sand gen-
erally attain their maximum settlement under load as soon as the load is

applied, and they retain that settlement. The loads do not cause the
subsidence to continue to increase over long periods of time. Further-
more, the settlement is usually slight and unimportant.
2. Silt and fine sand may attain a large part of their consolidation when
the load is applied, but the subsidence may increase somewhat over long
periods of time because of the squeezing out of the water content. If
^the soil is free to move laterally, saturation may expedite its movement
under the action may even-
of the pressure so that a serious settlement
tuate, especially if such that the water can move upward
the situation is

and cause the soil to "float." When a real quicksand condition develops,
the resultant settlement may be disastrous. In the case of silt, a mud
wave may occur, resulting in the slow movement or flowing of the silt
laterally even when covered with fill.
3. Clays attain a part of their consolidation as the load is applied, but

being plastic, they generally continue to consolidate slowly at a decreasing

rate for a long period of time. Most of this is due to a slow squeezing
out of the water. A time-settlement curve of a local thin stratum of
60 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

<Build/ng

Settlement

^ ^Settlement
Fine sand
fay-MOVEMENT OF MUCK OR MUD FROM UNDER A BUILDING

-Highway
Heaving
„_-/_

x
Muck, mud, peat Partially compacted material
"
or soft clay
Stiff clay or fine sand
f6J-M0VEMENT OF PLASTIC SOIL UNDER IMPROPERLY BUILT EMBANKMENT

FIG. 4-7. Examples of possible flow of plastic material from under a load.

Approx, Load, Tons Per Sq.Ft.

30 40 50 60
Elapsed Time, Hours

FIG. 4-7A. Some load-settlement data.

adobe is shown No. 3. The other test results

in Fig. 4-7.4 for test hole
shown in the figureapply to tests made on sand or sandy gravel. The
tests were made at the bottoms of test pits about 3 ft. deep.
Besides settlements of clays due to pressure, there may be those pro-
duced by drying out. For example, a foundation placed on clay near
the surface may settle as the clay dries in the summer, then rise as the
Art. 4-4] SETTLEMENT 61

clay reabsorbs moisture during a rainy period. A substantial depth of

embedment may be needed to avoid the drying tendencies.
4. Muck and mud are so dangerous for supporting loads that a predic-
tion of the settlement of a structure placed on them is largely guesswork
unless the structure is made so as to displace enough of the soil to equal
the weight of the structure and thus float. Even when such soils are
trapped under a blanket of fill or naturally deposited soil, a substantial
difference of pressure is These
likely to cause disastrous settlement.
materials sometimes move through and cause subsidence
large distances
under load and heaving many feet away, as illustrated in Fig. 4-7. The
movement of soft, plastic soils shown in (c) and (d) was revealed by the
records obtained at a power plant located on the Eastern seaboard.
Sketch (c) pictures the apparent condition when the plant was built and
the coalyard was first filled. Brush mats were placed on the silt before
depositing the fill, the objective being to bind the silt together and to
prevent movement. However, as indicated in (d), subsequent channel
dredging and soil investigation showed that both silt and mats had moved
riverward from 100 to 200 ft.
5. Temporary loads such as wind, live loads, and traction forces gen-

erally act so briefly that their effects upon cumulative settlements are
negligible. In this sense, a week is a brief time. However, a large ware-
house for "permanent" storage is an exception for which it may be desir-
able to assume a long-term live load of 50 per cent of the design live load
when the settlement is estimated.
6. It is generally differential settlement rather than the absolute mag-
nitude of a settlement that causes trouble with a structure. Some illus-

trations of this are given in Fig. 4-8. When

an entire area under and
around a building subsides uniformly, when the magnitude of the subsid-
ence is not excessive, and when the settlement does not damage other
structures, no one may be the wiser, and the settlement may be unnoticed.
However, one is gambling when he trusts this uniformity under conditions
in which the settlement is likely to be 6 in., 1 ft., or more. Not only
should a building be structurallysafe, it should be satisfactory to the
owner, useful for the purpose intended, and salable at an unimpaired
value. One who has carried on heavy construction work in urban areas
owners associate with unsightly cracks,
will realize the financial loss that
and how valuable are photographs showing the existence of these cracks
prior to the start of a construction job when some of these owners file
claims for damages.
The following comments refer to the sketches in Fig. 4-8:
(a) When the ends of a structure settle more than the center, the build-
ing is likely to "break its back," as shown by the crack at A. The damage
is obvious and may be serious.
62 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

(b) When the ends settle less than the center, the top of the building
is compressed. This is not likely to cause serious trouble unless the defor-
mations are so great that horizontal shearing ruptures the structure, or
tension cracks its bottom.
(c) Settlement of one end of a long building with respect to the other

end may crack the superstructure near the top, as at C. 1

(d) Gradually increasing settlement from one side of a tall building to

the other may tilt the entire structure without cracking it. This may
be psychologically if not structurally harmful.
7. Frost action may cause local heaving, subsequent softening of the

soil, and localized settlement. Therefore, the bottoms of foundations

Crack ^rushing zone

'
'ISeftlemenr^ * * * **

fW-SETTLEMENT AT ENDS ^-SETTLEMENT NEAR CENTER

rn
,

D U J"

(c)- SETTLEMENT AT ONE END ftf,J- VARYING SETTLEMENT

FIG. 4-8. Some harmful effects of uneven settlements.

should be placed below the probable frost line. This depth may be 5 to
6 ft. in northern New York, Maine, and Minnesota; 4^ to 5 ft. in slightly

warmer climates; about 4 ft. in New York City; and approximately 3 ft.
in somewhat milder regions. Because of the desirability of proper embed-
ment and the removal of topsoil, no important foundation should be less
than approximately 3 to 4 ft. below the ground surface, even in warm
climates. Interior column footings of structures in the North that may
be unoccupied and unheated for long periods during the winter should
also have reasonable embedment.
Under refrigerated warehouses and cold-storage places, the foundation
should be protected from freezing. The cold from subfreezing tempera-
tures penetrates to great depths when it continues for long periods, and

Heavy reinforcement of a concrete structure may prevent wide open cracks but
1

cannot usually prevent the distortion of the building. It is obvious that when uneven
settlements are probable, the use of simply supported members is better than con-
tinuous ones because of the bending moments impressed in the latter by distortions.
Art. 4-4] SETTLEMENT 63

the formation of ice lenses may cause heaving and the presence of tre-
mendous forces. Then, when refrigeration is stopped for a considerable
time, thawing may cause critical settlements. At the edges of the foun-
dation, deep freezing of the soil may occur in the winter, and then warm
ground water may thaw under the main walls, whereas it cannot
the soil

penetrate rapidly to the center of the foundation. This action may

enable the force of gravity to pull down the outer shell of a building around
the rigid frozen core, thus cracking the structure seriously.
Insulation of a soil-supported floor to prevent the transmission of heat 1

(or cold) may be utterly ineffective. Insulation retards the transmission

of heat, but it cannot stop conduction entirely. When frigid tempera-
tures exist continuously on top of the ground floor and when there is no
way of dissipating the cold, the latter will penetrate slowly in spite of
insulation. A ventilated basement, a network of closely spaced ducts
through which sufficiently warm air is circulated, a system of piers to
hold the insulated floor off the ground and permit natural circulation
under it, and the use of the basement or ground floor for the storage of
fruits and vegetables at temperatures considerably above 32°F. are all
means whereby this problem of freezing under such warehouses can be
conquered.
When a structure is founded upon rock, it does not seem to be necessary
to excavate the rock to the so-called "frost depth" of 4 or 5 ft. and to
replace it with concrete in order to avoid damage from frost action. Even
if a film of water does get between the rock and the bottom of the con-

crete, it will expand less than 10 per cent of its thickness when if, freezes.
The real danger from frost action is the formation of ice lenses under a
structure, with resultant heaving and subsequent settlement. These
lenses form most frequently when the soil is silt or fine sand with high
capillarity, when the watertable is not far below the bottom of the struc-
ture, and when the structure's footings are located above the depth to
which the "frost" can penetrate. When these conditions exist, the freez-
ing of the ground seems to cause a "cold front" at the bottom of the
frozen soil; this freezes the water in the soil and also freezes additional
water brought to it by the capillary action of the soil. Thus, ice lenses
of appreciable thickness may be built up, and successive layers of these
lenses may form. These cause heaving; when they melt, the water from
the lenses tends to cause local softening of the soil, perhaps producing
settlement. Clay does not have the capillary action that is essential for
the formation of these lenses; coarse sands and gravels do not facilitate
their formation either, especially if the ground water can be drained away
to a distance of 1 or 2 ft. below frost depth.
1
A vapor barrier should be used under a soil-supported floor, but this is for the
purpose of stopping the penetration of moisture through the concrete.
64 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

A building under construction or not occupied should be heated suffi-

ciently to prevent heaving of floors placed directly on the ground. In

one case, a contractor closed in a school, then left it for most of the winter
without heat. Some of the floors heaved approximately 3 in., cracking
partitions, etc.
Permafrost is a term used to denote the continuously frozen conditions
in the far north. Building a heated structure directly on such frozen soils
can cause thawing and serious settlement. Frozen strong rock, of course,
will not disintegrate upon thawing; coarse gravel may also be satisfactory,
but fine sands, silts, and clays may cause very serious trouble. Arranging
for cold air to pass under the structure may maintain the frozen state.
8. A compressible stratum below a much firmer one may still consti-

tute a hazard and cause serious long-term settlements. An estimate of

probable settlement should take into account the soil strata that are within
some reasonable distance of the bottom of the substructure. The sug-
gestions in Table 4-1 may be helpful in determining the depth below which
the settlement caused by the weight of the structure may be assumed to
be unimportant.

Table 4-1. Minimum Average Unit Pressures to Consider When Estimating

Probable Settlements

Character of soil Average pressure p, p.s.f.

Deep soft clay 300

Thin stratum of soft clay between firm impervious strata 500
Deep stiff clay 500
Thin stratum of stiff clay between firm impervious strata 700
Deep silt and very fine sand 500

In the table, p denotes the average unit pressure caused by the struc-
ture, assuming a 2 1 distribution. The higher unit pressures for thin
:

strata — —
10 ft. thick or less when trapped between firm impervious strata
are suggested because the water in the plastic layer cannot escape easily
when the pressure is applied; on the other hand, a porous soil will not
greatly retard the movement of the water that is squeezed out of the
neighboring stratum.
9. Consolidation tests of plastic and weak soils should be made in a
soil-mechanics laboratory upon some such basis as the following:
a. One test for each important stratum that has, or seems to have, con-
siderably different constituents, grading, or properties than the stratum
directly above it. These tests are to be made on undisturbed samples
taken from at least one selected boring.
b. One test for each 10 to 20 ft. of depth of thick strata. The number
of tests to be made will depend upon the judgment of the engineer when
Art. 4-4] SETTLEMENT 65

he considers the importance of the case and the properties of the soils as

already partly revealed.

c. When the borings indicate that the soil varies considerably in char-
acter over the site, tests as in (a)
(6) made on the soil
and should be
samples from some one of the borings in each area where important vari-
ations are apparent.
d. If the results of the soil tests show the existence of dangerous or
unsatisfactory materials, additional tests on undisturbed samples from
other borings should be made as the situation seems to warrant.
It is advisable to prepare load-settlement curves from the laboratory
tests, especially when the seriousness of the probable settlement makes it

desirable to have such data for use in studying alternative schemes for
the support of the structure so that the one adopted may be the most
economical plan consistent with safety and permissible settlement. One
should remember again that settlements may be limited to small magni-
tudes but at unjustifiable expense; on the other hand, harmfully large
settlements that ruin or seriously damage a structure may mean, in effect,
that the entire expenditure for the structure has been wasted.
Assume that the site of an important structure is being explored and
that four undisturbed samples have been taken from the first boring.
These have been sent to a laboratory for preliminary tests to ascertain
whether the consolidation and settlement will be serious and to serve as
a guide for further exploration and sampling. Assume also that the sam-
ples are being tested at loads of 1 and 2 tons per ft. 2 only.
Perhaps part of the initial report of the consolidation tests is the
following

Boring No. 1, Sample No. 2: Stiff Blue Clay

Unit pressure p, p.s.f. Voids ratio, e

0.886
2,000 0.864
4,000 0.849

These data may serve to plot a curve of unit pressure vs. settlement by
using the equation

Ah =
e
[
~ €p
h (4-3)
1 + e-i

where Ah = total expected settlementunder load in inches, e, = initial

voids ratio, ep voids ratio when under a pressure p, and h = thickness
=
of layer in inches. Notice that e is to be the voids ratio of the soil under
(

the pressure estimated as existing at the location from which the sample
66 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

was taken, and p is to represent the pressure superimposed upon that

previously existing.
It may be convenient to draw a diagram to show at a glance the esti-
mated settlement in inches for a stratum 10 ft. thick when a particular
unit pressure is applied to it, as judged from the results of the tests. For
this sample,
'0.886 - 0.864
s

A/la.onn — 10 X 12 = 1.4 in.

1 + 0.886
9
and '
!im
-AA, ( °f X~0SM ] '" '" '

The curve for this case is plotted as No. 2 in Fig. 4-8.4. The diagram
should be curved, but straight lines between the reported readings are

V Y /£a

v yV r
b

C|

'/2 2 27 2 3 , 3'/z 4 4Vz 5'/

2
Settlement, inches

FIG. 4-8A. Load-settlement curves for a 10-ft. stratum of soil. Such curves are to be plotted from
results of laboratory tests on undisturbed soil samples.

used here for samples 1 to 4, since the diagram is prepared only to enable
one quickly to obtain a scale on the expected settlements. Further tests
and curves can be made later for use in checking the final design. How-
ever, the methods of computation shown here are for preliminary investi-
gations. Since the 2 1 distribution often used to estimate pressures is
:

only an assumption, the data of Fig. 4-8.4 seem to be suitable for use
therewith. Later on, the use of the Newmark circle of Fig. 4-3.4 and
curves made from the final laboratory tests may be used to make the final
estimates of settlement after the structure has been planned and the loads
Art. 4-4} SETTLEMENT 67

and bearing areas are more fully determined. The general method of
summation is similar to that for the preliminary estimates.
It ishazardous to attempt to show load-settlement curves for a typical
clay or silt because they vary widely. Some organic clays may have voids
ratios equal to 2 or 3, and under a load of 2 or 3 tons per ft. 2 they may
compress to voids ratios that are only 50 to 75 per cent of these voids
ratios, if they do not flow away in the meantime. Clays that have been
compressed for ages under a much heavier overburden than now exists
(as under glacial pressures) may be quite dense and reliable. Test results
made on one sample of clay are shown in a of Fig. 4-8.4, and those on a
silt are given in b. Perhaps these are medium to soft materials.
El. 1742.8
Light brown silly clay loam with
a large amount ofsmall rock
fragments (partially indurated)
SO. "X- Wetter level

20
Maroon clay loam with partially
Rebc un d Sample decomposed rock fragments
-gO

0.07
o ooo
o O OOO
ooo o ooo
o ooo 40
Load, lb per sq.ft.
(fa)-LOG OF BORING
CaJ-LOAD vs. CONSOLIDATION FROM
CONFINED COMPRESSION TEST

FIG. 4-88. Report of consolidation test on undisturbed soil sample taken 24 ft. below surface in

Arizona. (After report by Dames and Moore.)

Figure 4-8A indicates that sample No. 1, which was taken near the

surface, was more compressible than the deeper materials. This may or
may not be the case in nature. It is shown this way purposely in order
to illustrate later on the use of such curves when preliminary estimates
of settlement are made.
Another question in connection with the use of the laboratory tests
showing load-consolidation relationships is what to allow for the effect of
the weight of the overburden. By necessity, a soil sample is removed
from pressure when it is taken out of the ground. How much does this
change it? How much is it remolded? How different will it behave
when in place? The effect of added load on the soil in place should be
known.
Undoubtedly, there will be a slight rebound when pressure is removed
from the sample, as shown in Fig. 4-8B, but its magnitude is seldom
important. The voids ratio, such as shown in Fig. 4-8 C, will not sud-
68 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

denly jump up in magnitude. Incidentally, curve A in this illustration

was taken near the top of a saturated 50-ft. stratum of clay underlying
about 20 ft. of silt and fine sands; curve B, approximately 20 ft. lower;
and curve C, 15 ft. lower than B. The loop in B near e = 1.25 was the
result of a temporary reduction of loading followed by reloading.
Good sands need not be a matter of great concern when loaded because
pressure usually makes them firmer. Silts generally are consolidated to

o o o oo
o o o o ,
O O O Oo
1.60

"- N \/J
"'*».*
>.
>, ^
1.40
\ \
\
\
1.30 \ \s—

1.20

^
\\ \
.10
\
\\
1.00 \

\ V \

.0.90 \\
0.80
v \

\
\
\

\
_l 1 MM
1 5 10 50 100 500 1,000
Pressure p, p.s.f. x 100

FIG. 4-8C. Voids ratio vs. pressure curves for saturated clays in the Connecticut River Valley.

some extent under the pressure any overburden, and they will not
of
"deconsolidate" unless disturbed too much. The same statement applies
to some clays. However, as shown in Fig. 4-8C, some saturated clays
may not be consolidated very well, even when under a considerable over-
burden, because much of the pressure may have been withstood by the
water "trapped" in the pores. However, if drainage or drying causes a
major decrease in moisture content, the settlement may be serious.
In general, it seems that the load-settlement curves as determined by
the tests can be used safely in estimating settlements caused by super-
Art. 4-4] SETTLEMENT 69

imposed loads, except for the time element. This omits the weight of
the overburden as such from the computations. The consolidation of the
samples may generally be somewhat larger than that of the soil in place
under pressure, hence computed results tend toward the side of safety.
Of necessity, these laboratory tests are of relatively short duration.
The element of time should be considered when investigating clays because
it is obvious that, in an actual case, the moisture will continue to be

squeezed out for a period of years. If the pressure is large and the orig-
inal voids ratio is large, the situation is more dangerous than if one of
these factors is relatively small. The results of one's computations of
settlement should be modified according to the factor "J." Again one
should remember that, once the structure is built, it is probably impossible
to change the product of the basic assumptions used in the planning. The
performance of the structure is real, not imaginary.
When problems of settlement are being studied, it is often helpful to
try to visualize what the unit pressures mean in terms of the weight of
some physical object. For this purpose, an imaginary pile or block of
earth that will produce the same unit pressure may be pictured as resting
upon the soil. Therefore, if one plans a unit pressure of 4 tons per ft. 2
under a footing, this is somewhat equivalent to the weight of a column
of earth 80 ft. high, approximately the height of a seven- or eight-story
building. This may be entirely satisfactory on sand and gravel, but not
on a soft clay.
Perhaps it would be helpful to try to visualize this consolidation of
soil. Figure 4-9 has been prepared for this purpose.
A granular soil generally has a considerable volume of voids through-
out the mass even though it is fairly well graded. Such a soil is pictured
to exaggerated scale in Fig. 4-9 (a). When a load is placed on this soil,
the pressure is transmitted by the bearing of grain on grain, and only
slight deformation is likely to occur unless the particles are crushed.
However, the loads are seldom such as to cause a mashing down of a sili-
ceous sand or gravel, although they may do so in a material composed of
flaky weak shale.
On the other hand, strong vibrations may cause particles to slip into
voids or to readjust themselves somewhat as pictured in Fig. 4-9(6). This
mass and, since the total solid matter remains
increases the density of the
the same, the reduction of volume usually causes a vertical settlement.
Practically, this action is not often serious except when the material is

saturated fine sand, or possibly a silt. Any sudden disturbance that

"shakes the solids down" automatically "flushes water out," as illustrated
in Fig. 4-9 (c). This water may then tend to lubricate or "float" very
fine sands and form a flowable material, at least temporarily. The sand
may then become quick, flow from under the load, and cause sudden dis-
70 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

aster. On the other hand, if flow of the solids is prevented, as by enclos-

ing the loaded area within suitable sheet piling, there can be no serious
settlement beyond that caused by the displacement of the excess water
as the sand becomes denser. Of course, the driving of piles may produce
the vibration that will shake down the grains.
In one case, an addition was to be built alongside an old brick factory
in New England. was supported on wall footings resting on loose
It
glacial sand. The new portion was
to have a basement extending sev-
eral feet below the old footings. However, the basement wall was to be

Excess
water
(a)- Loose (b)- Denser condition Loose sand More compact sand
granular of same granular (Voids for uniform (Voids for uniform
soil soil 01 fter vibration spheres =48%) spheres in densest
arrangement =26 % )
((^-Illustration of how vibration may cause
a fine sand to become quick

COMPACTION OF GRANULAR SOILS

warm
Sand-pile
^drains
TTi
ffl^-Original clay,
^Dense clay
feJ-Compacted
showing clay
~ Soft, compressible
flake li ke
stratum
particles Dense^sanoty clay
(f)-Sand piles used to expedite
compaction of compressible soil
COMPACTION OF CLAY SOILS

FIG. 4-9. Illustrations of consolidation of soils.

kept 10 ft. away from the old footings. Steel sheet piling was to be driven
outside the old wall to hold the soil. Contrary to orders, the contractor
drove this piling heavily with a drop hammer. One corner of the old
building settled approximately 1 in. and opened up an ugly crack.
The ordinary soils that may cause settlement troubles are clays, sands
or silts mixed with considerable clay, and sometimes silts alone. Figure
4-9 (d) is an exaggerated picture of the assumed particles in a clay. The
individual particle is supposed to be weak structurally, to be flakelike,
and to be somewhat spongy. The mass is supposed to be in equilibrium
and to be saturated, or nearly so. When a load is applied, the "skeleton"
of the mass may break down in places, the particles rupture or distort,
and the solid material become packed more closely, as shown in (e). The
Art. 4-4] SETTLEMENT 71

water, being incompressible, will be squeezed out and will flow away, if

it can, somewhat as when one steps on a wet sponge.

Since clay is rather impermeable, it may require a long time for this
water to escape completely, and the progress of settlement will be slow
but prolonged. If the clay is trapped within other dense strong imper-
meable materials, escape of the water and appreciable settlement may be
prevented. However, it is often dangerous to trust this because of pos-
sible disturbance or change in the future. An adjacent layer of sand or
gravel, on the other hand, may be like pipes in conducting the water away.
In fact, holes are sometimes made by driving a pipe pile down to a com-
pressible material, as shown in Fig. 4-9(/), and filling it with sand; then
the pipe is withdrawn. These "sand piles" expedite the escape of the
water (and therefore facilitate the consolidation) of a compressible stra-
tum. They are useful when a heavy fill is spread over soft marshy ground.
The total long-term settlement that will occur because of such consoli-
dation is difficult to predict. Consolidation tests made in the laboratory
will give some information. When the conditions appear to be danger-
ous, it is better to plan the structure so that one plays safe rather than
to take chances and be sorry in the future.
Occasionally, one encounters something special. At a plant in northern
Mexico, it was planned to build a bucket elevator structure about 20 ft.
square and 115 ft. high. Adjoining structures were under construction.
When the excavation for the elevator house was started, the soil was found
to be a slightly moist spongy disintegrated volcanic material, a sort of
pothole 50 ft. deep being filled with it. Walking on it felt somewhat like
walking on a thick layer of pine needles in a forest. Relocation was inad-
visable; no piles or pile-driving equipment was at hand; deep excavation
and construction would be costly. Therefore, to avoid danger, the order
was canceled, pumps were substituted to handle the muddy liquids, and
a low pump house on a mat was built to "float" on the spongy mass.
The loads were so light and widely distributed that the settlement was
negligible.
another problem developed at a plant in Chile where large dewater-
Still

ing tanks were to be built on desert

soils, mostly caliche. Samples showed
about 15 per cent of the material to be soluble in water. Would leakage
of water from the tanks remove the solubles, with consequent subsidence
of these costly structures? It was thought that the soils were probably
deposited and consolidated ages ago and that, subsequently, the chemi-
cals were deposited in the voids, largely as a result of evaporation. If so,
solution would remove them but would not cause a settlement of the soils.
Special tests, devised to prove this idea because of the tremendous sums
involved, showed the generally anticipated loss of weight but no loss of
volume when the chemicals were leached out, not even when the leached
 72 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

samples were vibrated by hitting the container with a hammer. There-

foreharmful settlement was not likely.
These cases are given to show again how foundation problems vary.
Each is an individual problem, and the best solution should be found for
the special conditions as they are and where they are.
Estimates of settlement. 1. Settlement of a structure on a thick
4-5.
plastic soil. Assume that the structure shown in Fig. 4-10(a) is to have
a heavy concrete mat at its base, and that the unit pressure at AB is to
be 2 tons per ft. 2 The soil is a 40-ft. layer of clay above a thick stratum
of sand. Assume further that undisturbed soil samples were taken at
10-ft. intervals, as indicated, and that the pressure-settlement curves for
a 10-ft. depth as determined by laboratory tests are as pictured in Fig.
4-8 A. How much settlement may be expected?

siirvi/tv/i-w/ y?> v// -w

<

W^_ ^\ \M
-

50'x50' / 12'xSO' \
'Sample No.!
±P / V-/ s
#2 60x60' ^
Sand/ \
El
3^ \
Medium
cl°iY g!
> #3

#4
70'x70'

— \

\
mf .

C/ayr*
o _&} 37 'x 75' _\
o
„

Sand
T 1 90'x90' t i r |p t \52'x90'} f f
Coarse sand
(a) (b)

FIG. 4-10. Procedure for roughly estimating settlements.

For convenience, the 2 1 distribution will be assumed. The estimate

:

will be made in three ways for purposes of comparison.

a. Use of average of unit pressures at top and bottom of stratum, found

as follows:
P = 50 X 50 X 4 = 10,000 kips
p AB = 4 k.s.f.

Vjk = 10,000/90 2 = 1.2 k.s.f.

Average p = 4 + 1.2
= 2.6 k.s.f.

Average settlement Ah for 10 ft., interpolated for p = 2.6 from the

mean of the curves for samples No. 2 and 3 of Fig. 4-8A, is 1% in. There-
fore, the approximate total A/i = 4 X 1 5 8 = ^2 m -

b. Use of unit pressure at middle of stratum:

p EF = 10,000/70 2 = 2 k.s.f.
 :

Art. 4-5] ESTIMATES OF SETTLEMENT 73

Average Ah for 10 ft. determined as in (a) is 1% in. Therefore, the

approximate total Ah = 4 X l a s = 5*2 in.
c. Use of a series of imaginary 10-ft. layers:

In this case the 40-ft. depth is divided into 10-ft. layers. The compu-
tations are tabulated as follows, with values of the settlement for each
layer scaled from the proper curve in Fig. 4-8A

Stratum Sample no. p, k.s.f. Ah, in. per 10-ft. depth

AB-CD 1 10,000/552 =3.3 2H

CD-EF 2 10,000/65 2 =2.4
EF-GH 3 10,000/75 2 =1.8
GH-JK 4 10,000/852 = ! 4
ZAft = 5%

All these results are estimates only. The computations need not be
carried out to more than two significant figures because even the first one
is not likely to be "correct." The "undisturbed" samples on which the
tests are made are remolded somewhat so that they do not exactly dupli-
cate the properties of the soil in its natural state. The computations do,
however, show whether the settlement is likely to be 6, 2, or 12 in. In
many cases, method (6) is sufficient for one's purposes.
Now look at the results of these computations. If the structure settles
this estimated amount, it is probable that the surrounding soil will sub-
side also and form a saucerlike depression with the structure at its center.
This may cause no serious difficulties, but a settlement of this magnitude
is not satisfactory to the owner or the engineer. If railroad tracks, over-
head bridges, flues, adjacent buildings, and other structural features are
alongside or connected into the building, they may advertise the settle-
ment to all observers.Furthermore, it is probable that the central por-
tion of the assumed loaded mass of earth will settle more than that esti-
mated for the average pressures, perhaps by 30 to 40 per cent. Will this
tendency of the mat to sag in the center cause harm to the structure
because of stresses or deformations?
Because of a tendency to sag near the center, the structure should be
designed as a strong box or block that will retain its desired shape. Then
the rigidity of the structure may cause the pressure to equalize more fully
as the plasticity of the soil produces a readjustment of the latter. How-
ever, because of the long-term loading and the gradual consolidation of
the clay, the eventual settlement of this structure may
be as much as 8 in.

In this case, it is desirable to study some such remedies as these:

a. Make the structure considerably lighter.

b. Spread the load over a much larger area. This may require a large
low structure rather than a narrow heavy one.
74 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

c. Change the type of foundation by using piles to reach into the sand
stratum.
d. Look for a more favorable location.
2. Settlement of a structure on sand with an underlying plastic stratum.
Assume that Fig. 4-10(6) represents the base of a high bridge pier near
the edge of a sluggish stream. Can the pier be supported safely upon the
sand when a 15-ft. layer of clay underlies it?
Assume that the load to be on the base LM will cause an average pres-
sure of 3 tons per ft.
2
One undisturbed sample, No. 5 in the picture, was
taken from the clay stratum. Assume that the clay's properties are simi-
lar toNo. 2 in Fig. 4-8.4, and use the 2:1 ratio for distribution of the
pressure.
The settlement of the 25-ft. stratum of sand will not be important by
itself. Hence the compaction of the clay alone is to be estimated. It is
satisfactory to use the average unit pressure at the center of the clay
stratum and, with this pressure, find the settlement of a 10-ft. layer from
No. 2 in Fig. 4-8^4, and multiply this by 1.5 because the layer is 15 ft.
thick. The resultant computations are the following:

P = 12 X 50 X 6 = 3,600 kips
3,600
= L ° Lsi
P* T = 44.5 X 82.5
-

Ah = 3
4 X 1.5 = \} s in.

This settlement might be acceptable if the superstructure is of a type

that is not harmed by such a subsidence, e.g., simply supported steel or

concrete girders and steel trusses, and (possibly) long shallow steel girders
designed to withstand the deformations caused by the settlement of one
pier. However, some arches and stiff continuous structures are likely to
be endangered by such settlements.
A decision of acceptance or rejection is difficult to make, and one must
abide by it after it is made because he cannot change his mind after the
pier is built. A steel mill building might stand such a differential settle-
ment, and human lives might not be endangered. A large bridge project
that is to support heavy traffic is a more serious matter. Furthermore,
the computed settlement is an estimate only. If all the piers of the
bridge are founded similarly and if the estimated settlements are practi-
cally equal, itmay be safe to accept such foundations. 1

In this case, how large would the base LM, Fig. 4-10(6), have to be in
x
order to reduce the estimated settlement to /i in. ? From curve No. 2 of
Fig. 4-8^4, for a settlement of approximately ?£ m m - the 10-ft. stratum,

1
See Art. 13-4 regarding scour.
 \ :

Art. 4-5] ESTIMATES OF SETTLEMENT 75

p = 0.5 kip. Therefore, the area of the central plane of the clay layer
would have to be 3,600/0.5 = 7,200 ft. 2 The new dimensions of the foot-
ing should then be increased to something like the following, where x =
the new width and x + (50 — 12) = the new length:

+ 32.5) (x + 38 + 32.5) = 7,200

x = 35.5 ft. x + 38 = 73.5 ft,

This size is obviously impractical, and it shows that minor increases in the

area of a footing have little effect upon settlements caused by deep-lying

and thick strata.

20'-O"

'//WAV/Ay//

A)
'

\B / C\ I/?.
/ IO'x/0' \Samp/e No'.l 10x10'
\
E —SF\ G'
I
\H
IS'x/5' \C /S'x/S' \
2
¥^ \l>-
/ 2O'x20' 20'x20'
\
Af/. 7 n
O
22 [6"x25 22'-6"x25'
/ \
RL k.
I 25'xJO' 2S'xJO' \

AV
*4 \
u/-
2T-6"x35' 27L 6"x35'
/
XL- l Z-X-
1 t 30'x40' t Y t 30'x40'\ T
FIG. 4-11. Estimating settlements when "pressure pyramids" of distribution overlap.

Many an engineer has founded structures upon plastic materials and

been gratified when they proved to be successful, but many another has
wished that he had founded them more securely.
3. Settlement of closely spaced footings on plastic soils. Figure 4-11 pic-
tures the proposed two-column pier of a highway viaduct or bridge
approach. The unit pressures under AB and CD are to be 3 tons per ft. 2
How much settlement may be expected?
In this case, the inner assumed 2 1 distribution lines intersect at K. :

Therefore, the assumed areas of resisting pressure at any lower level abut
at the line KY. The arbitrarily chosen zones of pressure are divided into
three layers 10 ft, deep. The average unit pressure at XY = (100 X 6)/
(30 X 40) = 0.5 kip per ft,
2
; therefore, settlement of the soil below this
level will be disregarded. samples were taken only at the
Because soil
locations shown, the settlements are for 10-ft. layers, assuming computed
the same soil characteristics as for samples No. 2, 3, and 4 of Fig. 4-8A.
The computations are as follows, using P = 600 kips
 76 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

Stratum Sample no. p, k.s.f. Ah, in. per 10-ft. depth

AB-JK 2 600/15 2 =2.7

JK-RS 3 600/(22.5 X 25) = 1.1 %
RS-XY 4 600/(27.5 X 35) = 0.6 %
ZA/i = 2%

This settlement seems to be altogether too much for a viaduct pier.

One can see that making the footing continuous from A to D would reduce
the unit pressure directly under the footing, but it would have little effect
upon the total settlement because practically the same volume of soil
must support the structure in both cases. Such a continuous footing,
however, or a strut between B and C, might be useful because inevitably
higher pressures in the vicinity of KY might cause greater subsidence of
the center, with the possibility of tipping of the footings toward B and C.
It is also obvious that, for small footings, the assumption of rectangular
pyramids of distribution of pressure yields appreciably smaller estimated
settlements than would circular cones of distribution. Here again, other
designs should be studied before one accepts so large a settlement.
Structures founded upon fill. As a general rule, one should
4-6.
hesitate to support a structureupon filled ground of unknown quality,
such as former city dumps and swamps that have been filled in with waste
material. Old fills may have been made by dumping soil from a truck
down the face of a slope of previously deposited fill so that the material
will be poorly compacted, uneven in density, partly segregated, and more
or less stratified along inclined planes. On the other hand, it may be
desirable from the standpoint of location and relation to other structures
to place a structure on an old fill which has been proved to be good or to
locate it on a new fill which has been properly placed.
Fills should be made of granular soils, although a little clay binder may
be helpful. Fills should be constructed in thin horizontal layers that are
compacted thoroughly as each one is made, using sheep's-foot rollers or
other acceptable equipment. With proper care and materials, new fills
can be made dense and satisfactory.
Attempting to make a compacted fill in freezing weather may be a
hazardous job. Freezing may prevent proper consolidation of material
already in place; it may also make the newly placed soil chunky, so that
there will be large voids between the frozen lumps. These voids cause 1

large settlements when the soil thaws out.

In one case, a fill 10 or 12 ft. deep had been made near a river's edge
about 40 years ago. The borings indicated that the fill was made of sand,
1
The same sort of problem occurs when using a cemented soil like caliche for making
fills.
Art. 4-6} STRUCTURES ON FILL 77

gravel, old bricks, and miscellaneous earthy materials, and that it was
underlain by medium
clay, silt, and finally by sand. A small but impor-
tant industrial building was to be erected at that spot. It was decided
therefore to use a bearing pressure of less than 1 ton per ft. 2 and to found ,

the structure on the fill about 4 ft. below the ground's surface rather than
to spend many thousands of dollars driving piles to the lower sand stra-
tum. When the excavation was made, it was found that a mass of old
wooden shavings was encountered below the watertable near one corner
of the substructure. These were removed, and the space was filled with
lean concrete. The work was allowed to proceed because it seemed too
late and otherwise inadvisable to alter the contract because of this dis-

°
Lean concrete °
•p.:- ?'
'-
•„. ' :''.?

si
concrete
(c) (d)

FIG. 4-12. Filling of weak spot under a structure.

covery. However, one cannot help worrying about what else may be
hidden below the structure.
Referring to Fig. 4-12, there are two principles illustrated which the
reader should notice. The situation of the shavings is pictured in Sketch
(a). In (6), is shown excavated with a flat area at AB and a
the earth
very moderate slope BC, with the lean concrete fill deposited first so that
the reinforcement of the mat may be placed above a level surface. If

the side BC were made steep, as DE in Sketch (c), the bearing power of
the soil at EF might be weakened seriously. If the lens of shavings had
been farther in from the edge, as indicated in (d) and if the concrete fill
were placed as shown, the slope GH should not be used because it would
aggravate the tendency of the soil to squeeze out from under the footing.
One would be wiser to make the bottom flat, as shown by JH. These
basic principles apply to any earth foundation as well as to the use of fill.
78 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

When a building must be supported upon fill, it is generally advisable

to design it so that any settlement will not endanger it. Of course, the
structure should be as light as practicable. In questionable cases the
frame should be sufficiently flexible to permit uneven settlements if they
are probable; a series of stress-relieving joints might be used through the
substructure and the superstructure; or the building might be made strong
enough to act as an entity without cracking.
Figures 4-13 to 4-15 show the principles used in the design of some
structures when new or old fill was to be utilized. Each case was a mat-
ter for individual study, and the illustrations are simplified in order to
show the principles more clearly. The ideas are also applicable when one
encounters strata of weak plastic soils of similar dimensions and shapes

Steel column Wall

FIG. 4-13. Reinforced-concrete floor supported upon concrete piers.

as the fills pictured. The following comments are given in connection

with the respective illustrations:
Figure J+-13. A large machine shop was desired in a particular location
in an industrial plant, this location being determined largely by the avail-
able space and the problem of getting railroad tracks into the structure.
An area of approximately 60 X 75 ft. at one corner would be over the
head end of a small valley so that the floor of this end would have below
it a wedge-shaped space that would vary from zero to 18 ft. in depth,

somewhat as pictured in the sketch. If this area were filled in and if the
building were supported directly upon it, the settlement of this corner
might damage the structure and, especially, harm the crane runways and
the machinery foundations. Therefore, concrete piers with spread foot-
ings were erected at all columns and under heavy machines, the fill was
placed between and around these piers, and a reinforced-concrete beam-
and-girder floor was laid directly on top of this fill. The floor was seated
on the piers and was designed to carry all the loads so that, if the fill
settled, the floor would not be affected.
It is very difficult to compact fill thoroughly when it is placed between
an assembly of these piers. The fill may be placed first, then piles can
be driven through it if it contains no large gravel and boulders, or drilled
Art. 4-6] STRUCTURES ON FILL 79

piers might be installed. These, however, are costly. Another method

is to sink pipes through the fill, excavate inside them, fill them with con-
crete, and withdraw the pipes. On the whole, it seems preferable, when
the fill is not to be used for supporting loads, to construct simple piers in
10-ft. lifts, place the fill approximately to their tops, and then proceed

with the next lift. One should be careful to avoid unbalanced filling

that will tip the piers.

37 L 6"±
platform

^i. Depressed rib = under piers

(a)- LONGITUDINAL SECTION

RPM No.24

ll pipe
railing

8 'in IZ'pope
Cone, plai form
and sfeps
jy//.v//>vw>V'« |
WA

/Basemen! floor level

(b)-CROSS SECTION

FIG. 4-14. A structure with a mat foundation on a fill 30 ft. deep.

Figure 4-14- This small warehouse in Arizona was built upon new fill
having an approximately uniform depth of 30 ft. The original ground
was good. The structure did not seem to warrant the expenditure neces-
sary for tall piers or piles. Furthermore, it was not desirable to have
the adjacent railroad tracks settle with the fill, whereas the warehouse
remained stationary. It was therefore decided to rest the structure upon
the fill, but to make it so that it would not be damaged if settlement
occurred.
The basement under one end of the buildingmade it probable that this
end —with less fill under it —would not much as the remainder.
settle so
On this account, a stress-relieving joint was made through the entire struc-
80 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

ture at A, Sketch (a). The shallow part was seated upon a shelf on the
basement wall, as illustrated by points B, so that the two portions could
not shear apart. The basement was naturally a stiff boxlike structure.
The section AC was designed somewhat like an inverted boat that floated
upon the fill. The longitudinal and end walls D and E served as stiff-
ening ribs; the fillets at F enabled the floor to act as a continuous part of
these ribs and resist upward or downward forces.
The fill was a mixture and sand. No noticeable subsidence
of gravel
has occurred because the building has remained intact, and the entire
area has behaved uniformly. Even a brick-walled one-story change house
built nearby, on the same material and of the same basic design, shows
no cracking.
West
Concrete deck,
-Deflection join t
{Bituminous pavement

Wooden back wall

\ Direction of
movement
of abutment
Jack
bracket
Compacted fill

F ^Settlement
Soft || \f Precast concrete of clay
clay piles

FIG. 4-1 5. Simplified illustration of viaduct structure and fill at western end of the New Jersey approach
to the Lincoln Tunnel in New York City.

In another case, a one-story building approximately 50 by 180 ft. was

built upon a slightly tapered sandy gravel fill. The structure had brick
walls, reinforced-brick pilasters that acted as columns, steel roof trusses,
a suspended ceiling, and interior brick or hollow-tile and plaster parti-
tions. The substructure was made similar to Sketch (b) except that the
floor was only 6 in. above the surface of the fill so that the side stiffener
walls were not so deep. Because weight of the partitions,
of the localized
cross ribs were used under some
them, approximately 20 to 25 ft. apart,
of
in order to stiffen the structure transversely. No trouble has occurred
with the structure in the 16 years of its life.
Figure J+-15. This pictures a deep fill used as the approach to the steel
viaduct at the western end of the approach to the Lincoln Tunnel, over
the New Jersey Meadows, where connecting roadways converge at the
main structure. The fill was placed on top of a thick stratum of cla}^.
It was expected that the fill would consolidate and that the clay would
do likewise when such a tremendous load was placed upon it. The steel
Art. 4-7] DANGER SIGNALS 81

structure was therefore supported upon a roller Awhich would permit

considerable longitudinal motion; the deck girders were made with a
rocker at B to allow vertical movement of A with respect to B; and the
legs of the adjoining steelbent were made with pivoted supports at C
and D in order to permit lateral motion of D with respect to C if the clay
moved out from under the toe of the fill, as indicated at E.
The abutment itself was a temporary structure, partly of wood, with
a concrete mat on steel tubular concrete-filled piles whose purpose was to
eliminate from the vertical settlement of the abutment any effect of con-
solidation of the fill itself. Considerable consolidation of the clay was
expected, and it was realized that the piles would be carried along with
any movement of the clay in which they were embedded. In 2 years,
the abutment settled downward about 16 in. and toward the west so that
the contemplated adjustments had to be made. The clay seemingly con-
solidated more at F than at E. Furthermore, the pier D moved eastward
slightly, as expected, so that the bent was jacked up and the bearing D
was adjusted.
A large one-story factory was to be built on the site of an old waste
dump underlain by some 20 ft. of silt with sand lenses. The waste was
excavated off the silt, then a mat of well-graded sand and gravel approxi-
mately 8 ft. thick was deposited in thin layers and compacted to 95 per
cent of maximum density. The structure was founded directly on this
mat, using shallow continuous footings under the walls and utilizing the
floor slabs as spreaders for interior column loads since there were no heavy
concentrations due to cranes or equipment. It was the engineer's inten-
tion to equalize the intensity of loading as much as possible, -so that if

settlements occurred, they would be practically uniform. The gravel

itself was upon to spread small concentrated loads so that the pres-
relied
sure on the would be substantially uniform. The structure was light
silt

and flexible so that it could distort somewhat without damage if unequal

settlements did occur. No evidence of harm to the building has become
evident.
The construction of the fill for any such situations as the foregoing is
of prime importance. Success is largely dependent upon field conditions
and the care with which deposition and compaction are executed. Heavy
rains, uncontrollable ground water, and variability of materials can upset
the engineer's plans to attain the desired result. Preloading can be help-
ful if the excess load can be left in place perhaps 6 to 12 months.
4-7. Danger signals. Rock, gravel, and coarse sand are good mate-
rials for most foundation purposes, but irregular surfaces of these firm

materials with varying depths of plastic soils over them may constitute
dangerous conditions from the standpoint of unequal consolidation of the
overlying materials. It is possible to have gravel or sand lenses within
82 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

or between softer strata in glaciated regions and in places where floods

may cause local scouring and deposition of materials. Such conditions
are also conducive to unequal settlements. Some such conditions are

q cBui/ding D
D
J

TTrTV7,1\
' :

I
jjl
B A\
<v)awj [] lit _ __
\B
Y^Trrrr

Old s/reahri^^C/ay-^Sand
(or gra vel and sand) bed fille d in ' -p^-^and grave/
:

(b)
(a)
(c)

Crushing ~U
A >.

\C jD B fus m —— i —!!-
J?
-^^^^Z Clay -^=-p£^=_ Clay

(j) (i)

FIG. 4-16. Examples of causes of unequal settlements of portions of a structure.

pictured in Fig. 4-16, with the possible character of the settlement and
the effect upon the structure shown to exaggerated scale.
The following comments apply to the various sketches in Fig. 4-16:
This pictures a typical case in which settlements can be expected
(a)

to vary somewhat as the depth of the clay beneath the substructure. The
building should be founded upon rock throughout, or a complete contrac-
Art. 4-7] DANGER SIGNALS 83

tion or stress-relieving joint should be provided at CD if cracking at D is

to be avoided.
(6, c) Here the compressible soil will cause sagging near the center C,
with possible cracking there. A joint at CD is not likely to be very effec-
tive. A small structure might be made strong enough to span across
such a weak area, but it is usually impracticable to make a large struc-
ture capable of doing this. The substructure should be supported upon
the firm material.
(d) The settlement of the portion CB of this structure will make it

desirable to have a joint at CD or to support the section CB on the rock.

(e) Here it is difficult to tell just what the action of the tapered layer
of coarse sand E will be. It seems reasonable to expect that it will serve
somewhat as a soil mat to spread the loads from F over a larger area of
the underlying clay G than will be the case under end B. It is therefore
likely that the end CB will settle so as to cause a crack near D. Since
the rock is too deep to be reached easily and since the sand stratum does
not extend under the entire structure, it seems that there are only three
sensible courses to be considered: (1) move the structure farther to the
north where the sand stratum E is thick and uniform and where the clay
layers F and G will not cause large and unequal settlements; (2) make the
structure wide, low, and light so that none of the settlements will be seri-
ous; and (3) cut the structure into two or more sections by means of ver-
tical joints so that deformations can occur at these predetermined points.
The joints should, however, be keyed or seated so as to transfer the nec-
essary shearing forces without vertical displacement at these joints.
(/, g, h)These are conditions that are likely to cause tilting of a struc-
ture. In the substructure might be carried to rock; in (g), piles might
(/),
be used well down into the clay in order to minimize or eliminate the effect
of this varying thickness of the compressible layer below the limits of the
substructure; in (h), piles might be used through E and F into G, or the
building might be moved off the tapered clay layer entirely.
(i) In this case, the chief danger lies in the possibility that the clay

of stratum F may move slowly under pressure toward H so that the struc-
ture would tilt toward end B. The best remedy here may be to set the
building farther back from the face of the hillside or to make the structure
so that the superimposed load on the soil is small.
(j, k, I) These illustrate cases where the unit pressures under the towers

are much larger than those under the remainder. This is inadvisable
when the buildings are founded upon plastic soils. The shape of the
structure should be changed to eliminate this differential, or the substruc-
ture should extend to suitable firm materials.
The results of unequal settlement caused by founding one part of a
structure upon firm ground whereas the remainder is supported upon
84 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

newly placed fill can be very serious. A large steel oil tank was built
on sloping terrain near the water line alongside the approach to a pier.
It was an addition to a tank farm that had been built previously and was
near the pump house. Approximately one-third of the tank was sup-
ported upon a wedge-shaped layer of new till. The tank failed suddenly
when filled with water for testing purposes. The accident caused a dis-
astrous fire and loss of human life. It seems that the fill contained "junk"

that was poorly compacted, and so settled badly. Apparently, the side
plates near the bottom were distorted seriously, and a long strip failed

Tar paper
or felt-

s'^ compacted sand

Compacted granular
fillor original ground,
well drained

A few small ties

Heavy reinforcing to prevent

cracks and to resist local
shearing forces

i Same pressure as
average for tank

FIG. 4-16A. Ring wall under an oil tank.

suddenly. Although be a limit to the

steel is ductile, there is likely to
strains and can withstand safely.
stresses that it

When a steel oil or water tank can be placed on reasonably suitable

soil, a moderate amount of uniform settlement may not cause serious dif-

ficulty nor even be noticed. In such a situation, having a fair granular

material to support it, the tank generally needs no structural foundation
except under the outer rim. Here it is desirable to use a reinforced-con-
crete ringwall somewhat as pictured in Fig. 4-16^4 . This wall serves the
following purposes:
1. Since there is a large vertical pressure inside the tank, there is a
tendency to cause the soil at the edge to shear and move outward. The
ring should be strongly reinforced to withstand this bursting tendency
and to prevent shrinkage and temperature cracking.
2. It spreads the weight of the tank itself over enough area to prevent

harmful local pressure and bending of the edge of the bottom plates.
 :

Art. 4-7} DANGER SIGNALS 85

3. gage for finishing the sand cushion under the tank

It serves as a level
and tank wall.
for the erection of the
4. It helps maintenance against rusting by keeping the steel a few

inches above the surrounding ground.

5. It prevents erosion of soil from under the edge of the tank.

In Fig. 4-17 (a) is pictured a situation that may cause trouble. The
stratum of very fine saturated sand may move out from A and boil up
at B. If the sand under the building were confined by a sort of steel

sheet-pile cofferdam whose top was tied together so that the sides could

'.'•.•.'.{•';:'•'•;
Coarse sand -.'•'.-:•.•.'• Direction of.
movement.'.

(a) -CONDITIONS THAT MAY CAUSE QUICKSAND

Probable
/=
underground flow •*' '
\"
'/?o ck

(b)- CONDITIONS INDICATING PROBABLE SUBTERRANEAN STREAM

FIG. 4-17. Dangers may exist even in granular soils.

not spread and the sand escape, the sand would support considerable load.
Even confined water will do so. Such a remedy, however, may not be
so reliable as piles would be if thej^ were driven into the coarse sand.
Another feature to be guarded against is underground streams. When
deep excavations are to be made, the flow in a case such as that shown
in Fig. 4-17(6) may cause considerable trouble and expense during con-
struction. Clean gravel and boulders in a subterranean ravine generally
indicate the presence of intermittently or continuously flowing water.
A valley containing a sluggish meandering stream may not only con-
tain soft materials but may hide old stream beds below the surface in
different locations, so that the soil will not be uniform. Deltas at the
mouths of streams and ocean beaches are also sites of probable irregulari-
ties and of possible future scouring or deposition of soils.
 :

86 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

Obviously, these discussions of possible harm because of settlement

apply not only to buildings but also to bridges and other structures. The
planning of structures to meet many of these conditions is discussed more
fully in subsequent chapters. However, the fact that the soil is a silt or
clay does not mean that nothing can or should be built upon it, For
example, in a part of the Quinnipiac Valley in Connecticut there is a layer
of flood-deposited sand overlying some 20 ft. of clay which is very plastic
and which has been used for the manufacture of bricks. An industrialist
wished to build a 250 by 300 ft. plant for light manufacturing on the site.
By designing the building as a light, flexible framework with panel walls,
continuous T-beam footings instead of isolated footings, and a rather
heavy reinforced floor, he was able to construct a very economical and
serviceable building.
4-8. Allowable bearing pressures. In order to prepare the contract
drawings for the foundations of a structure, someone must determine what
is to be used as the allowable unit bearing pressure for purposes of design.

This is a very important matter. Local building codes may set values
that are not to be exceeded. They are helpful guides because they are
usually based upon experience with the soils in that vicinity. They do
not, however, protect one against dangerous situations.
The following is quoted from the Foundation Code of the City of New
York, revised 1948, except for the addition of the table numbers, which
has been done for convenient reference: 1

§ C26-377.0 Presumptive bearing capacities of soils, a. Satisfactory bearing

materials shall be ledge rock in its natural bed, natural deposits of gravel, sand,
compact inorganic silt, or clay or any combination of these materials. These
bearing materials shall not contain an appreciable amount of organic matter or
other unsatisfactory material, nor shall they be underlaid by layers of such unsatis-
factory materials of appreciable thickness.
b. Fill material, mud, muck, peat, organic silt, loose inorganic silt, and soft clay
shallbe considered as unsatisfactory bearing materials and shall be treated as
having no presumptive bearing value.
c. The maximum allowable presumptive bearing values for satisfactory bearing

materials shall, except for pile foundations (see section C26-405.0,c), in the absence
of satisfactory load tests or other evidence, be those established in the following
classification
Hard sound rock is rock such as Fordham gneiss, Ravenswood gneiss and trap

sound condition, with some cracks allowed.

rock, in
Medium hard rock is rock such as Inwood limestone, Manhattan schist
and massive serpentine with some cracks allowed and slight weathering along
cracks.

1
Courtesy of Councilman Hugh Quinn, head of the New York City Council's Com-
mittee on Buildings, who has been the leader in the revision of the Code.
Art. 4-8] ALLOWABLE BEARING PRESSURES 87

TABLE 4-2. Classification of Supporting Soils

Maximum allowable
Class presumptive bearing
2
values, tons per ft.

1 Hard sound rock 60

2 Medium hard rock 40
3 Hardpan overlaying rock 12
4 Compact gravel and boulder-gravel formations; very compact sandy
gravel 10
5 Soft rock 8

6 Loose gravel and sandy gravel; compact sand and gravelly sand;
very compact sand-inorganic silt soils 6
7 Hard dry consolidated clay 5
8 Loose coarse to medium sand; medium compact fine sand 4

9 Compact sand-clay soils 3

10 Loose fine sand; medium compact sand-inorganic silt soils 2

11 Firm or stiff clay 1.5

12 Loose saturated sand-clay soils; medium soft clay 1

TABLE 4-3. Explanation of Terms

Descriptive term Blows per foot Remarks

Compaction related to spoon blows; sand

Loose 1 5 or less These figures approximate for medium sand, 2 1 ^-in.

Compact 16 to 50 spoon, 300-lb. hammer, 1 8-in. fall. Coarser soil

Very compact. 50 or more requires more blows, finer material, fewer blows

Consistency related to spoon blows; mud, clay, etc.

Very soft. Push to 2 Molded with relatively siight finger pressure

Soft 3 to 10
Stiff 1 1 to 30 Molded with substantial finger pressure; might be
removed by spading
Hard. 30 or more Not molded by fingers, or with extreme difficulty;
might require picking for removal

Soft rock is rock such as shale, decomposed serpentine, decomposed schist or

decomposed gneiss, with some disintegration and softening and with considerable
cracks allowed.
Hardpan overlaying rock is a natural deposit of a thoroughly cemented mixture
of sand and pebbles, or of sand, pebbles and clay, with or without a mixture of
boulders and difficult to remove by picking.
88 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

TABLE 4-4. Soil Sizes

Descriptive term Pass sieve No. Retained sieve No. Size range

Clay 200 Hydrometer 0.006 mm.

Silt 200 analysis 0.006 to 0.074 mm.
65 200 0.074 to 0.208 mm.
28 65 0.208 to 0.589 mm.
8 28 0.589 to 2.362 mm.
8
Pebble 2.362 mm. to 2*2 m -

Cobble 2 ]
'2 to 6 in.

d. it is shown by borings, or otherwise, that materials of varying bearing

"When
valuesmust be used for the support of structures:
1. The bearing value allowable for footings on the stronger material shall be

unchanged
2. The bearing value allowable for footings on the weaker material shall be

unchanged, provided the weaker material is not more than two classes below that
of the stronger material as established in this section, but
3. If the weaker material is ranked more than two classes below that of the
stronger material as established in this section, the bearing value allowable for
footings on the weaker material shall be reduced by a percentage equal to five
times the number of classes it is below the stronger material in ranking.

Another set of data is given in Table 4-5. It varies from the New
York Code because it represents a different opinion. In general, it is

somewhat more conservative but not necessarily more correct.

TABLE 4-5. Approximate Allowable Bearing Value

of Foundation Materials

2
Bearing capacity, tons per ft.

Material
Approximate Approximate
depth, 3 ft. depth, 6 to 10 ft.

Soft silt and mud 0.1-0.2 0.2-0.5

Silt (wet but confined) 1-2 1.5-2
Soft clay 1-1.5 1-1.5
Dense firm clay 2-2.5 2.5-3
Clay and sand mixed (firm) 2-3 2.5-3.5
Fine sand (wet but confined) 2 2-3
Coarse sand 3 3-4
Gravel and coarse sand 4-5 5-6
Cemented gravel and coarse sand 5-6 6-8
Poor rock 7-10 7-10
Sound bedrock 20-40 20-40
Art. 4-8] ALLOWABLE BEARING PRESSURES 89

The psychological reactions of people are sometimes peculiar. Let

someone write a letter stating that the soil at a certain site is "good for
4 tons per ft.and the recipients of this information generally use it as
2
,"

absolutely reliable. They may have many arguments among themselves

as to the details of the proposed substructure, but they seldom stop to
question the accuracy of this basic starting point for their planning.
However, the safe bearing value to be used in the design is the point that
really needs the careful stud}' and verification.
c
Bearing value, lb. per sq.ft.
2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 20,000
r s
/o °9»° £» ?4 2 2 J
CJ^«. ^ •
9 o fu
c ,'

St 7
.6
Reco mmenateddesigin value >S f J
10
.7 r?"^ *\\
-t-
CD
CD
4- ( >/
•^ 27,000 -

+-
A
D.
CO
H ?oo /U
°20

2*>
25
I
A
30
KEY
• Direct shear fesf
a Unconfined compression fesf
*- Tests at field moisture
— Tests at increased moisture

FIG. 4-18. Example of results of laboratory tests to determine the allowable bearing value of a soil

(Courtesy of Dames and Moore.)

Load tests of soils in the field are one source of helpful data upon which
to make a decision regarding this vexing problem. Laboratory tests may
also be useful.
Figure 4-18 pictures the results of laboratory tests made upon undis-
turbed soil samples taken at the site of an industrial plant. The results
vary, of course; and the plot somewhat resembles a "shotgun" diagram.
However, from them the curve shown in the illustration was drawn in
order to give data regarding allowable bearing values to be used for design
purposes. This curve shows how the bearing value was allowed to
increase with the depth of the footings. The tests were helpful in show-
ing that this could be done, and in setting the values. Load tests in the
field were not considered necessary since the ground was fairly strong.
90 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

The importance of keeping in mind the scale of a load test, especially

on a plastic cohesive soil, cannot be emphasized too strongly. The mag-
nitude of the test load is negligible as far as the earth mass in general is

concerned, and its important effects are localized. Suitable analogies are
difficult to find. However, the following may give some notion of the
basic idea:
Assume that the reader has a 30-lb. weight fastened to a loop of heavy
cord. He hooks one finger into the loop and finds that he can lift and
hold the weight safely off the floor with this one finger and that he can
carry it around the room. Therefore, he concludes that he can hook a
similar weight to each of the four fingers of each hand— not counting the
thumbs — and that he can carry 240 all not take him long to
lb. It will
realize that the strength of his arms, shoulders, back, and legs —his body
as a unit — now enters the picture and determines what he can and cannot
support.
when one makes load tests of the ground, he should not fail
Similarly,
to obtain a good scale upon which to compare his test loads to those to
be applied by the structure. The tests are helpful in showing the strength
or yielding of the soil locally. They have little value in determining the
action of deep layers, of long-term consolidation, and of the effects of
heavy loads over large areas. However, if the ground cannot support
such small loads safely, it will perform even worse under large loads
applied perhaps permanently. This idea is illustrated in principle in Fig.
4-2. The choice of the size of the load is important, but obviously the
cost will limit it.

The allowable bearing value of a cohesive soil for purposes of design

may depend somewhat upon the character of the loading. The dead load
is always present. In a warehouse, a large portion of the live load may
also be present for considerable periods of time. However, as stated in
connection with settlements, ordinary live loads on a bridge, in an office
building, on a pavement, or in a factory are temporary. So are wind
loads, even though they last for a whole day. These things should be
considered when one estimates the permissible bearing value to use in
design. For example, clay may resist safely a short-time load of large
magnitude, whereas it would be unable to do so if the load were applied
continuously. Therefore, when an engineer determines the safe bearing
value of a he should not only consider the values specified in codes
soil,

and indicated by tests but should use good judgment in considering his
special structure and the service expected of it, being conservative when
circumstances show such a course to be wise.
For example, an industrial plant was built on a fairly dense clay.
There were many crane columns having large crane loads but relatively
small dead loads. Alongside part of one row was a large heavy elevated
Art. 4-8] ALLOWABLE BEARING PRESSURES 91

bin supported upon separate columns. For design purposes, a bearing

of 8 k.s.f. was allowed under the footings of the crane columns but only
6 k.s.f.under those supporting the bin. The smaller bearing value for
the under the bin was used because of the fact that heavy loads would
soil

be applied for long periods, whereas the crane loads would be temporary.
The desirability of thinking in terms of the total load when studying a
foundation on plastic soil is well illustrated by the following:
An oil company had a steel tank 40 ft. in diameter and 40 ft. high that
was built directly on the ground near tidewater. The soil was plastic
clay which had supported this tank safely for many years. The com-
pany decided to increase its storage capacity at this port by removing
the small tank and erecting in its stead a tank 140 ft. in diameter and
40 ft. high. Those responsible for the design reasoned that the unit pres-
sure under the bottom of the new construction would be the same as
under the old tank and that the new one would be safe because the old
one had been so.
The construction of the sides of the new tank was performed by erecting
the annular rings one at a time. To avoid the use of costly scaffolding,
the staging was designed so that it would float in the tank when the latter
was partly filled with water. Thus, as each lift of plates was erected and
attached, the water level in the tank was raised accordingly, the platforms
rose automatically, and the construction proceeded step by step. One
day, when the wall of the tank was approximately 75 per cent completed,
a little subsidence was noticed. That night the tank settled nearly 16
ft. Apparently, this total load was approximately 0.75 X 140 2 /40 2 = 9.2
times that of the small tank when filled. Even though this was spread
over a larger surface area, the effect upon the underlying soil mass was
obviously much more severe than was the original load.
This case is cited to remind one that, when building upon plastic soils,
he should try to visualize the structure and the soil as an entity. In gen-
eral, he should remember that it may be advisable to reduce the theoreti-
cally allowable unit bearing pressure as the magnitude of the loaded area
increases. Several times, case histories of oil tanks have been used to
illustrate particular points. This is not done because oil tanks are dan-

gerous structures but because it is easy to visualize the magnitude of

their loads, the uniformity of their distribution, and the actions of the
structures.
There may be uncertainty about the load to be permitted upon a large
area when the deeper strata of soils are plastic or otherwise weak. About
the best that can be done is to make laboratory tests of undisturbed sam-
ples of these soils and estimate their strength and probable consolidation
under various pressures. Then an allowable superimposed pressure may
be selected that is to be permitted upon any particular stratum, and the
92 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

general scale of the expected settlement can be estimated. If the latter

is not harmful, one can estimate backward from the pressure on the lower

soil to determine what he will allow at the surface so that, when the latter
is distributed downward, the critical material will be loaded approxi-
mately as he wishes.
Throughout all the considerations preceding a determination of the
allowable bearing value for a particular site should be the question of
economy as well as of safety. Nevertheless, it must be borne in mind
that the cost of the foundation is only one part of the total cost of the
project,and a percentage variation in this cost does not have the same
relative effect upon the total expenditure, but the safety of the entire
structure will depend upon that foundation.

PROBLEMS
4-1. Figure 4-19 shows pressure-consolidation diagrams obtained from labora-
tory tests of clay soils. Convert these into curves giving the relation between
pressure and settlement per foot of soil. Use Eq. (4-3).

0.92

0.90

^C'4
0.88

0.86
o
"S0.84
/B
i~

</i
Rebo unc/
."5 0.82
o
>

0.78

0.76

0.74
0.2 0.3 0.4 0.6 0.8 1.0 2 3 4 6 8 10 20
Pressure, tons per sq.ft.

FIG. 4-19. Curves showing character of pressure-voids-ratios relationship for two soils.

4-2. Having curve A of Fig. 4-19 and the results of Prob. 4-1, estimate the
settlement of a building 50 X 100 on a mat foundation resting upon
ft. in plan
clay 100 ft. deep. The average pressure under the mat is 2,000 p.s.f. Use the
2 : 1 assumption for distribution of pressures.
 PROBLEMS 93

e?~^p
The settlement Ah - f ) h
Side View End View
P=3600 \P
"/W/AV/AV//* y /7\v//\w/A/vw/vy//'
,

f/^7 r
Sandy'':. \*~ -40- A U 18- >J
clay I
K Sample */„«»,

Medium I VSample #2
El. 220

Dense
Clay I VSample *3

El. 195

Coarse
sand

Pressure, lb. per sq.ft.

Stample No.
2000 4000
] 0.929 0.894 0.870
2 0.872 0.849 0.842

3 0.850 0.837 0.828

FIG. 4-20. A problem for estimating settlement of a large pier.

Pressure, lb. persq. ft. Voids ratio

0.852
2000
4000
6000
0.836
0.824
0.814
Ah •&&
1

#/ #2 #3 *4 #s
°o D a D
,*', P LAh i

h
20 / >+• — 4* —^ *f<
—^ 20'
H

-- J|£-

[
|fc |||^ \\\p ||

JTLayL-

Sand ELEVATION

FIG. 4-21. A problem on settlement under small footings.

94 SOME PRINCIPLES OF FOUNDATION ACTION [Chap. 4

4-3. Assume a building 100 X having an average bearing pres-

150 ft. in plan
sure of 1,500 p. si. By use of the Newmark
estimate the pressure under
circles,

its center and under one corner at depths of 20 and 50 ft. below the foundation.

4-4. The footing of a bridge pier is to be 25 ft. wide and 50 ft. long. The
average bearing pressure is 2.5 tons per ft.
2
Under the footing is 20 ft. of good
sand, then a layer of clay 12 ft. deep, and finally deep sand. If the clay has the
characteristics shown in Fig. 4-19, curve B, estimate the settlement that may be
caused by the stratum of clay. Use the 2: 1 method.
4-5. Estimate the settlement of the structure shown in Fig. 4-20, assuming 2:
distribution and the data shown on the drawing.
4-6. Estimate the settlement of footing No. 3 of a series of column footings
shown Each column has a load of 300 kips, including the footing.
in Fig. 4-21.
Consider the clay but not the sand. The clay is uniform in quality, and the voids
ratios are as shown on the drawing.
 5
SPREAD FOOTINGS

5-1. Introduction. The purpose of this chapter is to show various

constructions that may
be used for footings supported directly upon soil.
Methods are given for proportioning and analyzing these foundations.
The reader may think that the methods suggested here and in other chap-
ters are empirical and that the results are approximations. That is true,
but they are believed to yield safe results, to be easily understood, and
to be simple in application. They are probably just as reliable as are
the broad assumptions upon which foundation designs are generally based.
"Exact" computations may produce excellent psychological results upon
him who makes or observes them, yet they should not be expected to
correct the approximations in the basic data from which they started.
It is well for an engineer to visualize his structures as though they were
made of material that deforms greatly under load action. He should
imagine how a contemplated part may fail, then make sure tnat it cannot
fail that way, or any other way. If he foresees all these possibilities and

if he takes precautions to prevent harmful results, his structure will prob-

ably be safe. In general, if he provides a practicable way in which a

structure can act to withstand the loads safely, it will act in that manner
rather than fail.
There are many important matters affecting the choice of any one type
of foundation for a specific structure. There may be conditions that
require the use of one type under part of a structure, whereas another
type is preferable under the remainder. It is for the engineer to select
which serves him best and most economically. Such questions cannot
be answered specifically for any and all problems, but many suggestions
can be given in the hope that they will be helpful to the reader who must
make a decision.
Spread footings are used for the purpose of distributing concentrated
loads over sufficient soil to enable the latter to support the loads safely.
The principles used in designing them are applicable to many other forms
of substructure.
95
96 SPREAD FOOTINGS [Chap. 5

5-2. Data for the analysis of reinforced concrete. Many of the

computations involved in proportioning substructures require the analysis
of tentatively selected members made of reinforced concrete. To save
space, therefore, and tohave all calculations made upon the same basis,
the symbols given in Table 1 in the Appendix are used throughout this
book. The permissible unit stresses in materials are assumed to be as
shown in Table 2 of the Appendix. These are based upon concrete having
an ultimate 28-day compressive strength of 3,000 p.s.i. This will give the
reader a uniform datum from which to compare the results of various cal-
culations. In practice, he may use whatever materials, methods of analy-
sis, and allowable unit stresses are desirable for each particular case, but

the principles and methods illustrated remain the same. The approxi-
mate formulas that will be used frequently are listed in Table 1 of the
Appendix. 1

Some persons may wish to use larger allowable unit stresses in steel
and in concrete than those given here. However, the author believes that
stiffness is one of the attributes of good foundations. High unit stresses
in reinforcement are accompanied by correspondingly severe cracking of
the concrete if the steel is fully stressed, and their use generally effects
very tiny total savings whereas they may greatly reduce the reserve
strength of the structure. The steel is generally critical in such short
thick heavily loaded members as foundations; diagonal tension, shear,
and bond may be critical also. Parts of concrete that are stressed highly
in compression as beams are generally too thin for practical foundation
purposes. One cannot afford to take chances with members that com-
pose a substructure because the results of failure are so serious and
repairs are so costly.
5-3. Pedestals. In many instances columns are supported upon con-
crete pedestals that transmit the column loads to the footings, or to other
parts of the substructure. Two
such cases are pictured in Figs. 5-1 (a)
and (6). Sometimes the limitations of space restrict the area of the top
of a pedestal; when practicable, however, it is desirable to have this top
equal to at least 1^ or 2 times the area loaded directly by the column if
the sides of the pedestal are vertical and the load is large.
In practice, the following points should be considered when a pedestal
is dimensioned:
1. It is desirable to have the compressive stress in a pedestal consid-
erably less than 0.25/r unless it is beneath a reinforced-concrete column
'

and is itself reinforced as a column.

1
Ultimate-load methods of design may also be used by those who wish to do so.
Since the bending strength of most members of foundations made of reinforced con-
crete is dependent upon the bars, the author has used the elastic theory of analysis

herein for the sake of simplicity.

 V

Art. 5-3] PEDESTALS 97

2. It is almost essential to place the bases of steel columns above the

ground and above the water line. Because of possible rusting caused by
washing the floor, it may be desirable to place such bases 2 to 6 in. above
the ground floor of an industrial plant. The top of the pedestal should
then be proportioned in both these cases so as to provide such an increase

Concrete A -area of column A =area of base pt.

column
A- area of A= area of
pedestal pedestal
Steel column
Base plate
Grout

Pedestal-
Pedestal
Preferred
A=2A'fo3A'

(W-RECTANGULARREINFORCED-CONCRETE ^-SIMPLIFIED VIEW OF STEEL

COLUMN AND PEDESTAL COLUMN AND CONCRETE PEDESTAL

Steel columns- 3"tor

L
Mastic seal — — II

L on col. web :

Z on flange
/ Floor
[;•.?.'
|> *4%6" {Floor "o*? lb o

^:-\i>.
:
t."P:*«. o: .• ''/o.
--—•.."•..'•-' j&.
* ,:\
• «• :-S<: :.'.'!
: :•!•''.

~7tr
I
1
,hl jtn.
6

ir.'f'.ll'.-.-.V-.-.-.V.? "
l>
o. * * \
.•••>:.•:•'•'• •»•••'••.. '•'•.

t^^..^:^..!!
:'P/.-
:
.V-''
: :'-.'
:"./>.-.V.C

!i
:°.ff.-.|j''-'--?"y.'*.

,.i^->
::•:•':•':&£
: .'

s
r "/
:»*, > :

!!/ !
-" ;?:.-. :

;t>: •.
|i

il
i

n ••.••:•.•.•:•«

Pedestal 1 Steel shvb

Grout
rcJ-EMBEDDED COLUMN BASE WITH ftfJ-EMBEDDED COLUMN BASE WITH
RAISED ENCASEMENT ANGLES AS RUST PLATES

FIG. 5-1. Details at tops of pedestals under columns.

of area as horizontal clearances and satisfactory appearance make desir-

able. A
suggested treatment for the bottom of a steel column that must
be embedded below the floor line is given in Fig. 5-1 (c). In Sketch (d),
the long legs of the base angles are used as "rust plates" to protect the
main shaft somewhat against corrosion.
3. The cross section of a pedestal should be ample to spread the load
over a suitable area of the footing or slab that is under it. If the pedestal
98 SPREAD FOOTINGS [Chap. 5

is used on top of a footing as in Fig. 5-14, it may be desirable to make the

width of the pedestal equal (or nearly equal) to one-half the width of the
footing.
4. A pedestal is supposed to be an enlarged extension of the column.
It should therefore be strong and rigid. The height should be so great
relatively that there no likelihood of cracking of the pedestal because
is

of bending or shear. In Fig. 5-2 (a), the offset and the depth of the ped-
estal should be such that angle ABC equals approximately 30° or less;
the absolute maximum value of this angle should be 45°, which may be
satisfactory for small footings.
5. When the column loads are light and the unit stresses in the pedestal
are low, the area of a pedestal may be relatively small; when the column
loads are heavy, it should be more conservatively proportioned.

"|l Steel col. ,

^ Welded pi 3"min
Floor
Floor _ Slab
X J |
Mastic
'/
A
^Pedesfa/j^bor
,,/ LrVTTfr-".-^-.

M
'

/1 WW

Footing^
K r ,.
\j Footing
^
bolt
I
Pedestals Anchor
Footing^ bolt
I

L.

A C
(a) (b) (c)

FIG. 5-2. Some types of concrete pedestal.

6. Although the formwork is more costly, pedestals may be battered

as indicated in Fig. 5-2(6). With slopes on two opposite sides only, it is

simple. If all four sides are to be sloped, one should endeavor to mini-
mize the variety of sizes of pedestals so that he may use the same forms
repeatedly and thus reduce the unit cost.
7. In many cases it may be desirable to seat the floor slabs on the
pedestals, as pictured in Fig. 5-2 (c); in others, where the floor is on filled

ground or on plastic soils, it may be better to isolate the slab and the
pedestal, as shown in Sketch (6).
There is a tendency to crack the upper corners of a pedestal as pictured
when the column loads are large. The dowels
in Fig. 5-3 (a), especially
shown should equal the area of the longitudinal reinforcement in the col-
umn and should be developed by bond. One can realize easily that a
hook at the bottom of such a compression dowel does no special good but
may aggravate buckling of the dowel above it.

Assume that a heavy column is supported upon a pedestal and that

special means are to be employed to prevent harmful cracking at the
abrupt change in section pictured in Fig. 5-3 (a). If the vertical bars a
Art. 5-4] PEDESTAL FOOTINGS 99

are used near the surface of the pedestal, they will not accomplish much
good. Hoops like b may tie in the corners of the pedestal but the straight
parts of the bar along the sides will bend outward too easily to offer much
resistance to sideward forces. A mat of bars like c in Fig. 5-3(6) will
hold these top corners more strongly —at least to prevent appreciable dis-
placement if cracks occur. It is even more desirable to use a combination

Column
Col. re/nf.^v,
I Constjf:
Barsb
Co above
I.

Crack
Bars"aiW'
Jl.

— f-MSft f
'df

Bars "c :

X-i I
-v>v!.!

Dowels- , , .

!
-

Plan oi+ top of

pedestal
Hooks not good
for compression
(a) (b) (c)

FIG. 5-3. Reinforcement at top of a pedestal.

of reinforcement such as that pictured in Sketch (c) where bars b hold the
corners and provide circumferential reinforcement, bars c tie in the cen-
tral portions, and bars a act as supporting columns for holding the others
during the concreting.
Since the centralized load on top of a spread footing and the distributed
pressure below it cause a footing to try to bend to a saucerlike shape,
the top of the footing at DC, Fig. 5-4, shortens and curves slightly, tend-
ing to cause a concentration of pressure at the edges D and C. If these
vertical pressureswere to bend the pedes-
tal, the base DC would lengthen. There Column
is thus a tendency to break the junction
Pedeslal

DC. On this account, a roughened con-

struction joint is desirable at DC. Such
mechanical bond, together with the fric-
tional resistance caused by the vertical
pressure, will generally prevent shearing
of the junction. Furthermore, the unit
FIG. 5-4. Exaggerated picture of the
compression in the top of the footing at
action of a spread footing.
DC, caused by bending, will probably
be larger than the theoretical tension in the bottom of the pedestal, and
the former will actually prevent tension in the bottom of the pedestal
itself. The large edge pressures at C and D, however, make it desirable
to have the pedestal sufficiently high to eliminate weakness in shear.
5-4. Pedestal footings. The most simple individual footing is a solid
rectangular unreinforced concrete block like that shown in Fig. 5-5 (a).
100 SPREAD FOOTINGS [Chap. 5

It may be used when the load is light and when a small area of soil will
support it safely. In this case, the reacting pressure causes some shear
and bending in the footing, but the resultant unit stresses should be small.
An exaggerated picture of such a footing tested to failure is shown in
Sketch (b).
The pressure under the projecting part EGBF of Fig. 5-5 (a) will tend
to cause tension near the bottom of EF, and a combination of compres-
sion and shear from FB up to the column. The exact magnitude of the

Column Deflected
t- Thickness
re/nf. position of
of base plate
baseplate

Shelf as seat
/ for forms

n W

Rock

(d) (f) (9)

FIG. 5-5. Assumptions used in analysis of plain concrete pedestal footings.

shear along EF is indeterminate. Nevertheless, some procedure for anal-

ysis of such a footing should be set up in order to have some scale by
which to compare and study various footings. The following are sug-
gested as general guides, referring to Fig. 5-5:
1. Limit the angle FEB to 45°, with the vertex E taken at the edge of
a concrete column, as in Sketch (a) or, for a steel column, at the base of
a concentric rectangle whose sides are the depth of the column plus 2t
and the width of the flanges plus 2t, as pictured in Sketch (c). This seems
to be a desirable limitation because of the relative flexibility of the steel
billet. Some persons may wish to use unreinforced pedestal footings that
are relatively wider than these limits permit. It seems to be unwise to
do so, however.
2. Compute the shear that might exist if the pressure on the area ADCB

minus KLMN of Sketch (d) were resisted by the total depth EF around
Art. 5-4} PEDESTAL FOOTINGS 101

the perimeter —
KLMN what may be called punching shear. Limit this
computed shearing stress to 0.05/'. Since all these calculations are based
upon assumptions, the full depth of the footing will be used because fur-
ther refinement is not justified. Such shearing action is seldom impor-
tant as long as the footing does not crack in tension.
3. The bending moment M
theoretically might be computed about an
axis OQ, Sketch (d), through the center line of the footing as one-half
the column load P times ^QB minus J^P times one-quarter of the width
of a concrete column or one-half of the distance QJ for a steel column.
However, tests seem 1
to indicate that it is sufficient to assume that the
moment can safely be computed about a horizontal axis in a plane parallel
to EF, Sketch (a), but approximately one-seventh of the width of the
column [or KL, Sketch (d)], inside the edge of the column, neglecting any
local effect of pressure in the column. Since the column itself is very
stiff, some persons compute the bending as a cantilever about E or HJ,

Sketches (a) and (d). Because of uncertainty of the tensile strength of

the concrete, it is better to be more conservative, using the first or the
second of the methods previously described. Then, compute the bend-
ing stress as M divided by the section modulus of the full vertical section

through the footing. However, limit the computed maximum tensile

stress in the plain concrete to 0.03/'.
4. If the footing is trapezoidal — —
a frustum of a pyramid as pictured
in Sketch (e), analyze it for the same punching shear and bending moment
as would exist if it were a rectangular vertical-sided footing enclosing the
actual one, as shown by the dashed lines. When computing the tensile
stress at S, calculate the moment of inertia and the section modulus of the
trapezoidal section through RS, using the lever arm x from the center of
gravity of the area to point S.
5. A pedestal footing may also be used under a heavy column load if the
footing rests directly upon bedrock. If the bond to the rock is trust-
worthy, the bottom of the footing, TV in Fig. 5-5(/), cannot elongate;
therefore, only the punching shear need be computed. However, a little
dust or mud on top of the rock when the concrete is poured may destroy
most of the bond. Therefore, the footing should be analyzed as though it

were on and the limiting 1 1 slope from V to

soil, : U should still be retained
if bearing on the entire bottom of the footing is to be relied upon.
6. If the footing is made as a stepped one, as pictured in Sketch (g), the
slopes of lines WX and XY should not exceed 45° with the vertical. The
footing should be poured monolithically if it is to be analyzed as an integral
stepped rectangular member.

1
Clayton M. Crosier, A Study of Stepped Concrete Footings, University of Kansas
Publications, Bulletin of Engineering and Architecture 33, Lawrence, Kans., 1955.
 102 SPREAD FOOTINGS [Chap. 5

There are many instances where footings are to be designed with the
certainty that they will bear on bedrock. There are others where irregu-
lar rock is present so that some of the footings of a large structure will
rest upon it whereas others will not, unless piers are placed down to it
purposely. This last procedure, or the use of piles, may be advisable if
the soil overlying the rock is highly compressible and the structure can-
not withstand differential settlement without harm.
Does a column of reinforced concrete supported on rock need a footing?
Generally, the reinforcement in such a column is used to resist consider-

able stress. If the bars are merely stopped where the column reaches the
rock, they have no reliable way of transmitting their stress to anything
except the concrete near the bottom, thus probably overloading it.
Furthermore, the bars are generally detailed and fabricated in advance on
the assumption that they will start at a given elevation, whereas the rock
may not have been excavated at the time and it cannot be blasted out that

accurately. Again, such a column is inadequately restrained at its bot-

tom. Therefore it is desirable to use at least a pedestal to receive the
column, to contain dowels to transfer the steel stresses into an adequate
amount of concrete, to provide a reasonable restraint at the base, to serve
as a starting point for the column forms at the desired elevation, and to
spread the load sufficiently in case the rock is not of good quality.
When a steel column is supported upon rock, at least a fair-sized con-
crete pedestal should be used to receive the pressure from the column base,
to contain the anchor bolts, to distribute the load a little, and to enable

the bases of all columns to be at the right elevation regardless of the

irregularities of the rock.
To be considered also is the desired resistance of the base to overturn-
ing. In cases where effective restraint is desired, a footing may be needed
for this purpose regardless of the ability of the rock to resist bearing.
It sometimes happens that a footing comes in a position where it would
be supported on sloping rock. It is costly and troublesome to blast out
enough rock to form a more or less level bench or pocket for the footing.
On the other hand, the footing should be prevented from skidding down-
hill. When the column load is 200 to 300 kips, it is probably best to
prepare a bench for the footing; for lighter loads if may be sufficient to
jack-hammer or blast out a local pocket, which, when the concrete is in
place, will form a lug on the bottom of the footing to prevent sliding of
the base.
5-5. Some basic assumptions regarding isolated reinforced-con-
crete footings. There are three general types of isolated reinforced-
concrete footings used under columns and other structural parts that
transmit concentrated loads to their foundations: a shallow slab pictured
in Figs. 5-6 and 5-7; a sloped footing, in Fig. 5-13; and a stepped footing,
Art. 5-5} SOME BASIC ASSUMPTIONS 103

in Fig. 5-14. The first two are used mostly under reinforced-concrete
J

columns, whereas the last may be used with steel or concrete columns.
The principal reason for this is the fact that steel columns generally
terminate above the floor or ground so that a pedestal is needed to trans-
mit the load from the steel base to the top of the spread footing inasmuch
as the latter is usually placed sufficiently deep to be below the frost line
and to secure proper embedment. In practice, many combinations of
parts are used as substructures.
The word footing is used to denote the slab that spreads the load over
the soil, and also to mean the local substructure as an entity, including
the slab and pedestal. In fact, it is used also in the first sense to refer to

Pattern of reinforcement >

7,\W///,\\V-4nV^WA

/-Footing or
/ slab

(a) SIDE ELEVATION

\b) VIEW OF BOTTOM

AFTER FAILURE

FIG. 5-6. Example of character of cracking of footing tested to failure.

the slab under a wall; in the second sense, to denote a combined founda-
tion for two or more columns as well as for a single one. Isolated footings
are also called spread footings to differentiate them from wall footings
and mats.
Although a spread footing is an important substructure, it is an indeter-
minate structure. The methods used herein for the analysis and design
of such footings are largely empirical. Nevertheless, they seem to have
resulted in footings that have given satisfactory service. The assump-
tions and procedures used here are believed to be suitable for their pur-
poses and sufficiently conservative to yield safe results.
The pressure from a concentrated load is assumed to be distributed
uniformly over the soil in contact with the bottom of the footing if the
load is applied at the center of gravity of the footing's bearing area. This
1
Wide, thin-shelled, highly reinforced cones with a column at the vertex, and other
geometric shapes, formed on earth mounds can be used for footings, but the economy of
such construction for ordinary purposes is questionable. See A. B. Anderson, Precast,
Prestressed Stadium Floats on Hyperbolic-Paraboloids, Engineering News-Record,
Feb. 18, 1960.
104 SPREAD FOOTINGS [Chap. 5

may not be entirely correct for very large footings where compression of
the may not be equal over the entire area, thus causing deformations
soil

of and resulting However, this assumption seems

stresses in the footings.
to be satisfactory when a proper is used in design. On the
safety factor
other hand, the intensity of the soil pressure may vary with different
combinations of vertical and horizontal forces caused by dead load, live
load, wind, and other forces, as will be explained in Chap. 8.
The unit pressure upon the soil under a footing, for purposes of esti-
mating the superimposed bearing pressure and the settlements, is gener-
ally assumed to be the increase of bearing pressure over that initially

/w./ziy-vyi'/w'

2^ J
is
9? S rFooting
TJ"
£-_JSjL _jReinf.'-IB
'Q
T"T"Tjr*f"TTTT~r
?-a 6-0"square dP dP
Section on center line

(a) (b)

FIG. 5-7. A simple reinforced-concrete spread footing.

caused by the soil above the bearing plane, For example, in Fig. 5-7 (a),
the added pressure at AB is

p = 198,000/36 + 1.67 X 50 = 5,600 p.s.f. (approx)

The second term in the preceding equation represents the increase in pres-
sure caused by the weight of the footing at 50 p.c.f. in excess of the weight
of the displaced earth. This additional pressure is generally of little con-
sequence and is omitted in most cases. If the weight of the soil that was
(or is)above AB is included in computing the effective pressure applied
at AB, one will realize that some unwise conclusions may result. For
example, an embedment of 33 ft. instead of 3 ft. in Fig. 5-7 (a) would add
a theoretical pressure of approximately 3,000 p.s.f., thus implying that
the footing AB could safely support less load when deeply embedded than
when near the surface of the ground. Only in special cases, such as exca-
vations in plastic soils, silts, and quicksands, need this feature of over-
burden be a cause for worry. It is generally the increase of pressure above
that originally present that interests the engineer.
Inasmuch as the concrete of the footing AB, Fig. 5-7(a), is placed upon
the ground in a plastic state, the effect of own weight need not be
its

included in the pressures that cause bending moments and shears in the
footing itself. Therefore, in this case, the unit upward pressure for use
Art. 5-6} ACTION OF A CIRCULAR FOOTING 105

in the analysis of the footing may

be assumed to be the concentrated load
applied at the top of the footing divided by the area of the bottom of the
footing. For Sketch (a), the effective bearing pressure p is

p = 198,000/36 = 5,500 p.s.f.

This pressure causes the footing to deflect somewhat as pictured to exag-

gerated scale in Fig. 5-7(6). Thus the footing acts somewhat as a canti-
levered member that bends in radial directions about a centrally located

~L Circular Possible
column cracks in
/ \boftom

r TS "Bar a
faJ-VERTICAL SECTION AT CENTERUNE

(d-PLAN OF RADIAL REINFORCEMENT

Imaginary
contours
*

\
+3
— GH CF
*
T I

\P P
!sL2L£ -V—M ^y
1

-p^ *=
-f-f- t _ ,

^ t +

~3*-AV~Ap
(b)-Pim OF TOP OF FOOTING (W-ASSUMED ACTION OF OPPOSITE
RADIAL SECTORS

C=-T' Wlf_

V=A'p

TeJ-FORCES IF CRACK IS VERTICAL (fJ-F0RCES IF CRACK IS AT 45 c

FIG. 5-8. Picturization of assumed action of a circular footing.

column base. The backfill above the footing is not considered to have
any effect upon this action.
5-6. Action of a circular spread footing. To visualize the action of
an isolated spread footing, assume first a circular concrete column sup-
ported upon a circular flat slab, as pictured in Fig. 5-8. Sketch (a) shows
a vertical section or narrow slice along one diameter of such a footing. It
seems that the pressure under the entire footing will try to curl the slab
upward so as to form a saucerlike shape that is concave on top, as shown
in the qualitative contour plan in Sketch (b). The footing therefore elon-
gates radially near the bottom, whereas it shortens radially near the top.
 106 SPREAD FOOTINGS [Chap. 5

Bar a in Sketch (a) is used to resist the radial tension; if it resists this
tension, it must elongate, and the concrete probably crack somewhat
will
as shown. Furthermore, the radial elongation of the bottom must be
accompanied by a circumferential elongation of the same region. Simi-
larly, the top will have both radial and circumferential compression.
Imagine that the footing is reinforced radially by four bars at 45° angu-
lar spacing, as shown in Sketch (c). The plan of some possible hair cracks
at the bottom is given also. The sectors CDEF and HJKG of the annular
ring outside of the column's perimeter tend to break off and rotate upward
about CF and GH as pictured in Sketch (d). Bar a then serves as a tie
to prevent failure in this manner. If hoop b were used instead of bar a,

it might also prevent an excessive outward movement of the bottom and

thus enable the footing to resist pressure. At the same time, compression
acts upon the uncracked concrete in the vicinity of L and M. There must
be a shearing force acting upon each of these uncracked areas, too, because
the shear cannot cross the cracks below these areas even though the cracks
are infinitesimal.
Assume that an imaginary sector is formed by cutting vertical planes
along linesCD and FE of Sketch (c). The static forces that are assumed
to act upon this part and on a vertical plane through CF are shown in (e).
The bending moment at the face of the column is assumed to be the prod-
uct of the area of the sector DCFE times the unit pressure times the lever
arm to the center of gravity of the area (M = Apr ). If a crack forms
at about 45° with the vertical just outside the column, as is probable, the
forces that may upon the portion beyond the crack are pictured in
act
Sketch (/). There must be shearing forces in the uncracked concrete just
beyond the column.
An examination of Figs. 5-8 (e) and (/) indicates that the following
assumptions may be logical:
1. The critical compression and tension will occur in the plane of the

edge of the column or pedestal if the tensile cracking there is vertical.

This should be assumed as a critical case for bending moments.
2. The greatest intensity of transverse shear -punching shear — will
occur in the uncracked concrete at the edge of the column or pedestal
when the tensile cracking is vertical. This is a critical case for punching
shear. However, failure by diagonal tension generally occurs along a
somewhat conical surface, as indicated by the cracks near L and of M
Fig. 5-8(d).
3. Any underneath the projected area of the column or
tensile cracking
of an adequate pedestal probably be vertical because it will be at
will
right angles to the tensile forces, and these are horizontal. Outside of
the projected area of the column or pedestal, however, the cracks may
be inclined downward and outward from the neutral axis because of the
Art. 5-7} SQUARE AND RECTANGULAR FOOTINGS 107

combination of horizontal tension and vertical shear. As a critical case,

may be assumed to equal 45° with respect
the flattest angle of cracking
to the vertical plane at the face of the column or pedestal.
4. With 45° cracking, the vertical shear causes shearing and tensile
stresses to act upon an imaginary inclined section along with the stresses
produced by bending alone, thereby decreasing the effective resistance of
the section against vertical displacement of the portion beyond this 45°
plane. This condition is therefore assumed to be a critical case for diago-
nal tension.
The sectors in Fig. 5-8 (c) are treated as though they were formed by
sawn cuts along the radial boundary lines. Since this is not the fact,
there may be some question as to how one can have bending moments
normal to a cylindrical surface at the edge of the column when the footing
is a monolith. For ordinary footings it is customary to consider the bend-
ing moments as acting upon a straight vertical cross section through the
footing at an edge of the column or pedestal. However, this radial
arrangement is illustrated for purposes of visualization. It also seems
to be a useful concept when designing very large circular footings such
as for a tall chimney.
5-7. Square and rectangular isolated spread footings. It is not
ordinarily practicable to use circular footings like that in Fig. 5-8 (c) under
columns because the formwork is more expensive, the radial reinforcement
stacks up too much at the center, and bands of parallel bars require the
use of various lengths of steel in each band. Therefore, square and rec-
tangular footings are generally used.
A square footing, such as that shown in Fig. 5-7 (a), tries 'to deform
somewhat as does a circular one. Compare the contours in Fig. 5-8(6)
with those in Fig. 5-9 (a). If the latter footing is not ruptured, the con-
cave upper surface will not have sudden breaks or changes in slope that
would be revealed by sharp angles in the contours. It seems logical to
assume that the corners of the column ABCD will tend to "ride" the top
of the footing, and hence they are likely to be points of high pressure.
It seems likely also that contours near the column will be rounded squares,
whereas those farther away become more nearly circular. The corners
of the footing obviously curl upward the most. In an elongated footing
like that of Fig. 5-9(6), the action is similar to that of a square one except
that the ends must deflect more than the sides.
It also seems assume that the centrally loaded square footing
logical to
have equal pressures acting upon all four of the sym-
in Fig. 5-9 (a) will
metrical parts shown, equal shearing forces on vertical planes through
the column faces AB, BC, etc., and others equidistant therefrom, and
equal bending moments about horizontal axes in these planes. However,
these statements are obviously untrue for the rectangular footing of Fig.
108 SPREAD FOOTINGS [Chap. 5

5-9(6). The more elongated a footing becomes, the more unequal will
be the conditions of stress at sections passed through faces and JK JM
of Sketch (6).
How, then, can one estimate the shearing and bending stresses in square
and rectangular footings? It is not practicable to analyze them exactly
as has been suggested for circular footings; nevertheless, a logical, simple,
economical, and safe procedure is desirable.
The general procedure for the design and analysis of square and rec-
tangular footings may be outlined briefly as follows:
1 Assume a style or shape of footing that seems to serve the particular
purpose required.
Determine an adequate bearing area that the footing must have.
2.

3. Assume
all thicknesses and other dimensions for the trial footing

upon the basis of judgment and precedent.

+6\ F N R

H G Q U Z T P
(a)- A SQUARE FOOTING (b)-h RECTANGULAR FOOTING

FIG. 5-9. Imaginary contours to picture curvature of top of square and rectangular footings.

4. Make the footing so that it cannot fail by bending, shear, or bond

in the direction WX of Fig. 5-9(6) ; make it so that it cannot fail in the
direction YZ either. If these two paths of resistance are strong enough,
the footing probably will not fail.

5. Check the design economy and general suitability.

for
First, consider square footings only. Tests made by Richart indicate 1

that the bending moment at JK, Fig. 5-10 (a), should be the full magni-
tude of the forces acting upon the area KJGH times their lever arms from
the section JK, according to the principles of statics. This bending is
assumed to act across the full width of JK. Some specifications, such
as Building Code Requirements for Reinforced Concrete (ACI 318-56),
hereafter called the code, allow one to multiply the computed bending
moment by the factor 0.85, but it seems to be advisable to stick to the
rules of statics, especially because the assumed axis of bending is taken
at the edge of the column or pedestal instead of through the center of the
1
Frank E. Richart, "Reinforced Concrete Wall and Column Footings," Journal of
the American Concrete Institute, October and November, 1948.
Art. 5-7} SQUARE AND RECTANGULAR FOOTINGS 109

column. For computing the bond on the bars perpendicular to JK,

stress
the full pressure on the area KJGHFor diagonal tension,
is to be used.
all quarters of the footings are assumed to act similarly. Thus, referring
to Fig. 5-10(6), the pressure on the area HPQG outside of a plane, located
at the effective depth d from CD, is assumed to cause diagonal tension
upon a width PQ rather than RS for a square footing.
As a rectangular footing becomes more and more elongated, its action
approaches that of a wide beam cantilevered across the column. It seems
that the computation of bending moment across the short direction should
be computed about the line JK, for pressures acting on KJGH, and that
across the long direction about LM, for pressures acting on LMHE. The
shear for computing bond stresses on the reinforcement should be the
pressure on these same areas.

F J G F S G
i
,>

Q/
i

B c B 1 c/ % v.

' 9. •"'•:. °:
6
5

— A D M A/ D 1 5

1
/ " \ !
ii

1
/
\
H N R H
1 i
1

E K E O
(^-RECTANGULAR SECTION r^-TRAPEZOIDAL SECTION

FIG. 5-10. Square reinforced-concrete spread footing.

The diagonal tension upon a plane through RS of Fig. 5-11(6), a dis-

tance d from the face CD, will be assumed to be caused by the pressure
acting on area HOSRNG outside of RS. Similarly, the diagonal tension
on TS be assumed to be produced by the pressure on the area PTSO.
will
If the depth d is so large that the assumed 45° diagonals intersect the
narrow sides of the footing within the distance d, as in Fig. 5-1 1(c), the
diagonal tension on RS will be assumed to be caused by the pressure act-
ing on area HSRG and affecting a width RS. Theoretically, there will
be no critical diagonal tension acting upon planes parallel to AD if the
cracks occur at 45° slopes. Actually, since diagonal tension is the result
of tensile and shearing stresses combined, some such action will exist
between the edge EH and AD, but it is not likely to be serious.
Seldom does the compression in the concrete due to bending cause
critical unit stresses in properly proportioned footings. The assumption
of the full width of the footing as resisting bending gives "average" results.
The maximum stress near the column may exceed this average.
When footings are placed upon soils of high bearing value, there might
be danger of failing because of punching shear as pictured in Fig. 5-12(a),
110 SPREAD FOOTINGS [Chap. 5

— 1"— tT/?

B r 1

1
\ 5
til

i £ /! i

l (
d , t
d > |

1
^
A D
i
< *

E K H
(a) DESIGN SECTIONS FOR {b) DESIGN SECTIONS FOR
BENDING DIAGONALTENSION

(c) DESIGN SECTION FOR DIAGONAL TENSION

IN RELATIVELY THICK FOOTING

FIG. 5-11. Elongated or rectangular footings.

Assumed
neutral axis Crack

3*1
T
z?mb
f

r«;-POSSIBLE PUNCHING SHEAR FAILURE ^-POSSIBLE DIAGONALTENSION FAILURE

Tie,

f-4- HH-M-r-H-}

Effiiffl
L.g 4 M+M^
i_-Ll_l_LI_l-LI_U
Main band #2^
fcJ-TWO-WAY REINFORCEMENT rrfJ-FOUR-WAY REINFORCEMENT

FIG. 5-12. Shearing action in, and reinforcement of, a spread footing.
Art. 5-7] SQUARE AND RECTANGULAR FOOTINGS 111

causing failure along such planes as AB and CD. However, the action
ismore likely to be a diagonal tension failure as pictured in Sketch (6).
Since tensile cracking of the concrete probably will have occurred below
the neutral axis at E and F because of the elongation of the reinforcement,
the uncracked portion Ac? of the section is all that should be depended
upon to resist transverse breaking. Furthermore, since the steel rather
than the compressive strength of the concrete is critical in most footings,
the footings are underreinforced. Therefore, the factor k will be assumed
to equal 0.3. A magnitude for it seems to be unknown. The
correct
maximum allowable punching resistance may be assumed to equal 0.2f'.
For example, an analysis of the footing shown in Fig. 5-7 (a) will give
the intensity of punching shear v T as follows:

Unit pressure X (gross area — column area)

= column perimeter X kd X vt
5,500(36 - 1.5 2 ) = 4 X 18 X 0.3 X YIv T
vT = 510 p.s.i. (0.2 X 3,000 = 600 p.s.i. allowed)

Inasmuch as it is desirable to make a footing thick enough so that no

web reinforcement is required, in order to obtain stiffness and simplicity
of reinforcement, the permissible magnitude of the diagonal tension (lon-
gitudinal shear) at the edges of the 45° slopes shown in Fig. 5-12(6) should
not exceed v'L = 75
0.03/', or It p.s.i. is, of course, necessary that the
tensile reinforcement be properly anchored by effective corrugations on
the bars, or by other suitable means to provide adequate bond. Hook-
ing of tensile reinforcement (or 90° bends) in heavily loaded footings is
ordinarily helpful as insurance against slippage.
Footings that are relatively small compared with their depth may have
the tensile reinforcement distributed uniformly across the entire width.
However, in large footings where the cantilevered projection is 1.5 to 2
or more times the effective depth, it may be preferable to locate most of
the main reinforcement nearer thecentral portion of the footing. In Fig.
5-12(6) grouped purposely within a band or critical section that is
it is

—
equal to, or not much wider than, GK the limits determined by the
assumed 45° lines from the edges of the column or pedestal. One or more
bars like a should be used for reinforcing the corners of the footing and
for tying the bottom edges together when the main bars are more than
10 to 12 in. from the edges. This arrangement provides two strong bands
as shown in Sketch (c), most of these bands passing under an area near
the column that can be relied upon to provide a trustworthy vertical
reaction.
In the case of a long narrow footing, the bars required in the short
direction should be grouped in the vicinity of the width UR, Fig. 5-11(6),
with a few extras across the outer ends.
112 SPREAD FOOTINGS [Chap. 5

The four-way arrangement of reinforcement shown in Fig. 5-12(d) may

be used for very large footings, but the four layers of bars require con-
siderable depth for packing up where they cross. Furthermore, the parts
of the diagonal bands that are assumed to be effective in assisting the
main bands of steel are the components or projections of their areas par-
allel to the latter. For example, bar b times cos a is relied upon to aid
bar c.

Inasmuch as the bond resistance of bars in footings is so important,

the author will use a permissible bond unit stress for numerical problems
in this text equal to 0.07/', or 210 p.s.i.

The magnitude computing unit stresses

of the factor j for in the con-
crete and reinforcement may be assumed equal to 0.88, or 0.9. The latter
will Because of other uncertainties, no theoretical
be used generally. cal-
culation of j seems to be justified.

Assumed width
{ for compression
Shelf for J Lui j J"
-

T
"forms' \£_3£ C \J K\ D i
'
J

Width
bJP-
r effective

A G Assumed width of MFj

main band ofreinf. f±) ***

(a) (b) (C)

FIG. 5-13. A sloped-fop spread footing.

When a footing is to be used on bedrock it might seem that reinforce-

ment in the bottom is valueless because the footing cannot deform and
stretch the bars as pictured in Fig. 5-7(6). However, it is generally advis-
able to use a moderate unit pressure (perhaps 20 k.s.f.) and to design the
footing as though it were on earth. This is because of the likelihood of
overblasting, backfilling,and grading to the design elevation; shattering
of the rock without complete removal of the weakened portion; disinte-
grated rock to start with; and coating of the rock with enough dust or
dirt to break the bond and good contact. Because of the high bearing
pressures in such a case, the bond on the bars has to be developed very
rapidly; hence, hooks are desirable to ensure safety.
Round columns not having square or rectangular pedestals are to be
considered equivalent to square columns having the same cross-sectional
area as far as the design of footings
is concerned.

Sloped footings, such as that shown in Fig. 5- 13 (a), are somewhat more
economical in their requirements for volume of concrete than are flat-
topped ones. They also have greatest depth where the bending moments
and shearing forces are the largest, and they are often desirable when no
Art. 5-7] SQUARE AND RECTANGULAR FOOTINGS 113

pedestal is to be used. The tangent of the angle 5 generally should not

exceed 0.33(1 :3 slope) if the use of top forms is to be avoided even then —
the concrete should be relatively dry. The assumed width of the main
bands of reinforcement may be determined in the same manner as for
Fig. 5-12. However, as seen in Fig. 5-13(6), the cross section of the
assumed cantilevered beam is not rectangular. The compressive stresses
may be determined by the use of the transformed-section method, but
the compression in the concrete is seldom critical. For practical purposes
it should be sufficient to assume that the section is an imaginary rectan-

gular beam of width (CD + GM) /i x

as pictured in Sketch (6) and having
an effective depth of d. The neutral axis may be assumed to be at %d

Concrete Preferred
/
column minimum=2 *h
.t;
Step or ^s :.»:'•'•. :",':
short __^ i-
<i)

pedestal -0
k
\
\ (JO
Footing-
cz.
I

:<C3 ^
<\
\
Critical section tbr~J\
fs>fc,VT and ju ^
V
V
a.
i

(b)-10P VIEW
Critical section for v^
fa;- SIDE ELEVATION

FIG. 5-14. A stepped footing.

from the top of the footing, this being conservative. If fc seems to be

critical, a more careful analysis may be made.
For estimating the diagonal tensile stresses in a sloped footing, the pres-
sure under MF of Fig. 5-13(6) may
be assumed to act upon a vertical
section that is a rectangle whose width equalsT U of Sketch (c) as for ,

slab footings, and whose effective depth is that of the footing at section LM.
A stepped footing like that of Fig. 5-14 may be analyzed in a manner
similar to that used for the footing in Figs. 5-7, 5-10, and 5-12 except
that the edge of the pedestal is used as the axis for computing bending
moments, and for reference when estimating shearing forces. The pedes-
tal should be analyzed to see that it is adequate to serve its purpose of
spreading the column load over a large area of the footing. It permits
the use of a much shallower footing and less steel than would be required
for a simple slab footing supporting the same column.
One of the strangest misconceptions of the action of a spread footing
or else an illustration of dishonest construction — is the case pictured in
114 SPREAD FOOTINGS [Chap. 5

Fig. 5-15. A large mill building was built about 1900. The original
column foundations were plain-concrete pedestal footings, as shown by
CDFE. Sometime around 1925 the structure was rebuilt for heavier
duty, and the floor and yard were raised to avoid a repetition of past
flooding. The new spread footings AB were built on top of the old
footings.
It is obvious that the soils under AC and DB are far more compressible

than the concrete block under CD. Therefore, the column load will be
transmitted directly down to the old
Brick bearing area at EF, and the spread foot-
ing will do little good in supporting loads
unless the settlement of EF becomes so
Grade beam bad that the latter departs for unknown
between lower regions. Loads are transmitted to
footings
the strongest practicable reaction points
by the most direct and stiffest path that
is available.

Old The reader can imagine the problems

footing involved when this condition was dis-
covered during the planning of a second
remodeling and strengthening of the
structure during the Second World War.
Some of the foundations had to be sur-
FIG. 5-15. An example of unwise con
struction.
rounded by a wide reinforced-concrete
collar under and beyond AB and extending
down to the level of EF in order to have proper bearing area. This made
the remodeled footings a sort of enlarged block of concrete of strange
composition but of trustworthy safety. The unit cost was very high.
Example 1. Part of the superstructure of a highway viaduct is to be
supported upon a concrete tied column that is 2 ft. 2 in. square. The ver-
tical load at the base of the column 400 kips. The soil profile is shown
is

in Fig. 5-1 6(a). Frost depth ft.

is Design a sloped-top footing
at least 4
for this column. Allow the unit stresses in Table 2 of the Appendix.
The allowable bearing pressure on the soil will be assumed to be 4 kips
per ft. 2 for the sandy clay and 8 kips per ft. 2 for the medium sand. The
approximate required bearing areas are 40 %
= 100 ft. 2 for the clay and
40 % = 50 ft. 2 for the sand. The latter material is so much stronger that
it seems preferable to use a footing 7 ft. 3 in. square extending to El. (ele-

vation) 395.5 rather one 10 ft, square at El. 397.5, approximately 4 ft,
below ground at its southern edge.
The reinforced-concrete column may be extended into the ground with-
out the use of a separate pedestal if the cover o^er the reinforcement is
sufficient. The trial footing is sketched in Fig. 5-16(6).
Art. 5-7} SQUARE AND RECTANGULAR FOOTINGS 115

In this case, the punching shear around the edge of the column is

estimated as follows:

p av = 400,000/7. 25 2 = 7,600 p.s.f. (net pressure)

400,000 - 2.17 2 X 7,600 _ 364,000
Vt = 730 p.s.i.
kd X 26 X 4 0.3 X 16 X 26 X 4
(greater than 600)

Therefore, the footing is probably too thin for other purposes also. A
new trial d is

364,000 364,000
d = = 19.4 in. (assume d = 20 in.)
vTk X 104 600 X 0.3 X 104

This new footing is shown in Fig. 5-16(c).

North

I 1 ,
Assumed b for. fc -4-1
2'2sg

2ty" 2 L8sq

„ ^
Mi ___ '
~~-—~~^ JJ.;,

^Nl
^
Medium sand
fa;- SOIL PROFILE
7^3"scl

f6;-FIRST TRIAL FOOTING

*4
"

_
5-6" EI.39S.S^
7'-Z"sa

fci-SECOND AND FINAL FOOTING

p
^1
3

\ ? ,
\
['
[7
y/9-*5@4'c.c.6oM\
ways
"|

/ 1

D
i

\
2*4' c

MOMENT (e)-PLM FOR DIAGONAL TENSION

/8@4"=6'-4"

fiV-PLAN OF REINFORCEMENT
-4-

m
frtV-PLAN FOR BENDING

FIG. 5-16. Footing for a viaduct column.

From Fig. 5-16(<f), using the net pressure under area ABCD without
any 0.85 reduction factor,

V = 7,600 X 7.25 X 2.54 = 140,000 lb.

Mad = 140,000 X 2.54/2 = 178,000 ft.-lb.

A<
M 178,000 X 12
= 5.93 in.
2

fsjd 20,000 X 0.9 X 20

V_ = 140,000
)o required = 37 in.
ujd 210 X 0.9 X 20
116 SPREAD FOOTINGS [Chap. 5

From Table 4 Appendix, 19 No. 5 bars will give A s = 5.89 in. 2

in the
and So = 37.2 in. This is near enough. If these bars are placed in a
central band at approximately 4 in. c.c, this should be satisfactory. It
illustrates how bond stresses may control the size of bars used.
Using an assumed rectangular strip (32 -f 66) /2 = 49 in. wide as indi-
cated in Fig. 5-1 6(c), a rough approximation of the compressive stress in
the concrete is

2M_ 2 X 178,000 X 12
h kjbd 2 0.3 X 0.9 X 49 X 20 2
p,s " 1-

This is so much lower than the 1,350 p.s.i. allowable that no further check
of fc is necessary.
A check of the diagonal tension v L at E, Fig. 5-16(c), using the force on
the area FGHJ of Sketch (e), gives

Vl
V_
=
[(5.5 + 7.25)/2]0.88 X 7,600
== S L
P * "

bjd 66 X 0.9 X 16.3

This is lower than the 75 p.s.i. allowed by Table 2 of the Appendix and is

therefore safe. The measure of diagonal tension, to

restriction of vl, a
75 p.s.i., whereas 90 p.s.i. is permissible in beams, is conservative and
advisable, because the shear V in two-way footings is computed on the
basis of the pressure acting beyond the 45° line GF in Fig. 5- 16(e) whereas
it is computed up to the edge of the support when designing beams. The
use of 75 p.s.i. instead of 90 allows a little for the fact that the cracks
p.s.i.

may not be quite as flat as 45° and therefore V might be larger than
assumed.
if it is dimensioned as shown
Therefore, the footing will be satisfactory
in Fig. 5-16(c)and reinforced as in Sketch (/).
Example 2. Figure 5-17 shows an interior steel column that an
engineer wishes to use in an industrial plant. Alongside the column is
to be part of the foundation of a large machine. It is essential to keep
the column's foundation separate from that of the machine. The engi-
neer in charge of the design has ordered the permissible bearing pressure
2
to be limited to 2 tons per ft. Design a concrete spread footing to serve
this purpose.
What type of construction is desirable and practicable? To answer
this, consider the following conditions that are necessary or seemingly
desirable, referring to Fig. 5-17:
1. This column is in the interior of the building, hence freezing will not
control the depth of embedment of the footing.
2. The bottom of the footing should be placed at the same level as that
of the adjacent machinery foundation so that deeper excavation alongside
Art. 5-7} SQUARE AND RECTANGULAR FOOTINGS 117

will not weaken the soil under either one. The column foundation will
probably be built before the machinery foundation.
3. The steel base plate is to be raised above the floor. Therefore, a
concrete pedestal is necessary, and its top should be as small as practi-
cable in order to minimize the obstruction around the column and above
the floor.

4. The be seated upon the pedestal.

floor slab is to Therefore, a 3-in.
shelf is desirable around the pedestal.

1-8'

fa>>- PLAN OF BASE PLATE

Top of I2WF65*
concrete -$
SE Jffor
grout
Elev.124^

I?'
AO

Concrete floor slat?

(seated on fdn.)

S andy clay
3^6"

(^-VERTICAL SECTION AT COLUMN

FIG. 5-17. Base of a steel column in an industrial plant.

5. The footing should be symmetrical, but one side dimension should

not exceed 7 ft. because of the adjacent machinery foundation.
The minimum area of the footing should be slightly larger than 275/4 =
69 ft. 2 7 by 10 ft. will be used as the trial dimensions of the footing.
;

The top of the pedestal will be made 2 ft. 6 in. square and provide an
area of concrete in bearing that is 2.5 times that of the base plate. The
main shaft of the pedestal is therefore 3 ft. square, after adding two 3-in.
shelves.
How thick should the footing be? Since the large pedestal may elimi-
nate any damage from punching shear, the bending and the longitudinal
shear in the long direction will probably control. A reasonable guess for
118 SPREAD FOOTINGS [Chap. 5

footings with heavy pressures is a cantilevered projection from the ped-

estal ofone to two times the effective depth of the footings. For smaller
bearing pressures, like this one, the projection may be 1.5 to 2.5 (or per-
haps 3) times the effective depth. The maximum projection is 3 ft. 6 in.
therefore try d = 18 in., with a cover of 3 in. over the bars. A sketch of
2-6 "sq
< t

3'-0''

El-
111
fc.st
ism 1

i
r
h

1 \sf*
l i

» -/4S
r
I

WA~\ I! I
^sk "V.

i-Vf'W—» • 4' • i—I -w—»^-r rt r

^
5JTr i
A 7
1"
Main band 5-/0" \7 gl /-£". I
frggf band 6-0" . |
/-£" Igj y?
7-0' IO L
(a)-CROSS SECTION 76;- LONGITUDINAL SECTION

A A^ 6> 1,

I
6-0'
-Jz^vo
"3 hooked-
«•"
Q _^_\£€.° £7 *6,
hooked-

44-KhfH4H4ffl-r-

.V5< 7^
rr+T-Hfl-TlTJPT
Ti-HTlTrtrrh-l+
I I I
|! II I
l| l| I

T-l+rTt4-H-H4-H-+-
J W M /4@5± = 5-/0"
re) -PLAN OF FOOTING (d)-?LM OF FOOTING REINFORCEMENT

FIG. 5-18. A spread footing to be used in an industrial plant.

the footing is shown in Fig. 5-18. Assume that the stock reinforcement
to be used is No. 4 and No. 6 bars.
The assumed increase of pressure caused by the weight of the footing
itself is

AP for the footing = (7 X 10 X 1.75 + 32 X 2.58)50

= 7,300 lb. (approx)

Total p = - =tt — - = 4.03 k.s.f. (near enough)

Art. 5-7] SQUARE AND RECTANGULAR FOOTINGS 119

Effective p = 27 5
; 7<) = 4 k.s.f. (approx) (This is used for analysis.)

- 9)4,000 = 01
VT =
Mm
(70 - 9)p =
4
(70
X 36 X 0.3 X 18
3U .

^ . ,

(appr0X but

than the 600 allowed)

,
mUch
, ,

leSS

Long direction. The bending moment in the long direction, based upon
the pressure under area PKLU about line PU, is

MPU = 4 X 7 X 3.5 X 1.75 = 171.5 ft.-kips

-
A _
^
M - miX = b 35 in m .
s
fjd
~ 20 X 0.9 1.5
6 "^ -

= —V— = 4,000 X» 7 X 3.5

= OQ _ . , . „ f
,, .. , . .

So ,
n v
'..
, » v , ir) 28.8 in.- per in. (henceforth called in.)
ujd 210 X 0.9 X 18

Fifteen No. 6 bars will be satisfactory, spaced 5 in. c.c. (A s = 6.6,

2o = 35.4), as shown in Fig. 5-18(a).

.
= 2M = 2 X 171,500 X 12 = 56
«.«„•/ n
/c
Wd* 0.3 X 0.9 X 84 X 18" ° P S L (aPPr ° X) " -

This is very conservative.

For computing the diagonal tension, use the pressure under area
YKLXba, Fig. 5- 18(c). Then vL = shear /bjd, or

6 7
4,000 7 Xl.5( t )o.5J
Vl - [
= 47 p.s.i. (75 allowed)
72 X 0.9 X 18

Short direction. Using the area OLMW, Fig. 5- 18(c), the bending
moment about OW is

Mow = 4 X 10 X 2 X 1.0 = 80 ft.-kips

A
As = M = 80
= O
3 m
' . /
(aPPr ° X)
X

fjd 20 X 0.9 X 1.5 '

Vn
20 = V = 4,000 X 10 X 2 QQr5
= 23 '
.

in "

ufd 210 X 0.9 X 1 8

Therefore, use 15 No. 4 bars (A, = 3, So = 23.6) as shown in Fig. 5-18(6).

It is obvious that fc and diagonal tension will be even less than for the
long direction and therefore will be safe.
This footing will be accepted, and it is drawn in more detail in Figs.
5- 18 (a), (6), and (d). The reader should realize that this is not the only
footing that will serve the required purpose. In a large project there may
be many footings, and may
be desirable to prepare designs for several
it

different sizes and use them whenever they are safe and not too wasteful.
 120 SPREAD FOOTINGS [Chap. 5

5-8. Combined footings. There are conditions under which it is ad-

visable to use a large slab or combined footing under two or more columns
instead of an isolated spread footing under each one. Some such condi-
tions are illustrated in Fig. 5-19:
1. When a row of columns is to be so close to a building line that spread
footings would be loaded eccentrically. A common footing may be placed
under each outer column and the corresponding one in the adjoining
interior row, as pictured in Sketch (a).
2. When two columns are so close together that isolated footings would

almost overlap, e.g., footing A in Sketch (6). The footing under a

double column at an expansion joint in the superstructure, as shown by B,
is practically an isolated footing rather than a combined one.

3. When the designer wishes to tie two footings together to resist uplift,

overturning, and opposing horizontal forces, as illustrated in Sketches (c),

((f), and (e).

4. When such a structure as a bridge or viaduct pier composed of two

or more columns rests upon soil of low bearing value or upon piles, so that
it is desirable to have one large common base, as shown in Sketch (/).

5. When the principle of combined action seems to offer the most desir-

able construction. The special conditions encountered in engineering

practice vary amazingly, and each design should be the best for that par-
ticular case.
For example, Sketch (g) illustrates a situation that was encountered in
an industrial plant located on a hillside. The floors were placed at differ-
ent levels. The position of wall C was fixed by the machinery, whereas
the column locations were determined upon the basis of the preferable
structural framing. Because of the probable weakening of the soil in area
F on account of the excavation for the lower floor, it would be desirable to
place the footing for the heavy column D down at some such position as
shown by the dotted lines at G. This excavation would, in its turn, make
it advisable to build the footing of the lighter column E at approximately
the dotted location H. This construction would involve considerable
cost for excavation and substructures and would require that the founda-
tions of the heavy machines be extended down to undisturbed soil. Even
-the fact that portions of the concrete floor would be placed upon soil that
was disturbed in some cases, firm in others, and backfilled around the
, columns would be disadvantageous because of the uneven settlements that
might occur unless the floor, too, were structurally supported.
It was therefore decided to use a strong combined footing or beam under
both columns. This was a deepened portion of the floor built integrally
with the adjacent floor slab. The footing and the floor slab were seated
upon the wall at J The load from column D was thus spread across to J
.

and to the undisturbed earth between F and K. The load from column E
Art. 5-8] COMBINED FOOTINGS 121

was included in the forces affecting the soil FK. This scheme proved to
be satisfactory.
Incidentally, there is sometimes a question as to whether an elongated
member like the dotted one at G, Fig. 5-19(g), is a column or a pier. In
this case, if used at all to support column D, the cross section of the mem-
ber should be large enough to receive the column or the base plate of a
steel column, and a few inches larger to provide a seat for the floor slab.
Although it is subjected to compressive forces, the unit stresses in the
Expansion
Braced bay joint *-*
:

0- "0 HH
& p |
|

Outer col. Interior col. -JT

Ca)-K RESTRICTED CLEARANCE

(b)-VML CONDITIONS ENCOUNTERED IN MILL-BUILDING CONSTRUCTION

~
T7F77I& MWW.w/,w/ v>/
\j .

\
[y/f'r i r^n^\>m/m>).'m .
i>/ .iiiM>^
.
\/Ai. ^Fri\~\/>wi-ww/>-WJ/m \\ »/'»'. i

* Footing

IS LU e
Cc;-BASE OF A TOWER OR BENT (W-BASE OF BRACED BAY feJ-CRANE COLUMNS

D weak
Floor, £1 area A *-|
™f
|

Wa/IC, II /A'
/. A ^ I
I
I
SECTION A-A
I I

/ i

3 Floor
I

I
X
I

H
^ie±k i/
6
(f)-k VIADUCT PIER OR BENT f0/-A STEPPED FLOOR IN AN INDUSTRIAL BUILDING

FIG. 5-19. Some conditions where combined footings may be advantageous.

pier are likely to be small, and it need not be reinforced in accordance with
the Code requirements for columns. In substance, the problem may be
handled as follows:
1. If the slenderness ratio of the member is 10 or more, it approaches
the character of a column regardless of its low unit stress.
2. If the unit stress is less than the 0.18/' ordinarily allowed in tied

columns, there is no need for heavy reinforcement, and the member is

primarily a pier.
3. If overturning moments act on the member, and if the restraint of
the surrounding soil is inadequate to support the pier horizontally, the
member should be designed to resist the bending or tilting regardless of
the magnitude of the direct compressive stress.
122 SPREAD FOOTINGS [Chap. 5

The planning of a combined footing requires that the designer deter-

mine how he is going to "make" the structure act. Undoubtedly, the
footing will behave as an entity that supports the loads as efficiently as it
can, and the actual stresses may be unknown. However, it is the safety
of the structure that is important rather than a knowledge of the mag-
nitudes of the stresses in it. If the designer provides a definite, adequate,
and efficient system that is capable of resisting all the anticipated forces
and he has anticipated them wisely and well, the structure will behave
if

substantially as he intended before it will fail. It is to be understood that

the engineer will provide a system that will be as simple, direct, and posi-
tive in its action as he can devise.
Several designs and analyses of combined footings are made in this
chapter and following ones. By studying them, the reader will learn how
to attack such problems.

400* 500 k
20-0' 4 L 6"
Column 2'sq.
*V£t]
n
I
I I
Pedestal 4s<t
W
,

Pedestal 4 sq.
I

-///.WAW/^'i oarrTw ww
Sdrncty ctay ^i ELEVATION A-A
<r Bottom of footing aJ
FIG. 5-20. Load diagram and column dimensions for a viaduct bent.

Example. In Fig. 5-20 is shown the lower end of an assumed two-

column bent for a reinforced-concrete viaduct. The columns are close to
a building line that limits the width of the footings along one side, although
an overhead easement has been secured for the superstructure. The
engineer in charge has had a study made of the quality of the soil. As
a result he has limited the allowable bearing pressure for vertical loads to
2 tons per ft.
2
The allowable stress/., in the reinforcement is to be 18,000
p.s.i. Although the column loads differ because of the weight of a can-
tilevered sidewalk on one side of the superstructure, the columns and ped-
estals are to appear alike. Design and detail a substructure to support
this viaduct, keeping the edge of the concrete at least 6 in. back from the
building line.

If isolated spread footings are used, the approximate area of the larger
one will be

A =
500 + 16
129 ft.
2

where the 16 kips is a guess at the added load caused by the footing itself.

A square footing would thus be nearly 1 1 ft. 6 in. on a side a rectangular ;

 :

Art. 5-8] COMBINED FOOTINGS 123

one 8 wide would be 16 ft. long. It therefore seems that a combined

ft.

footing may be preferable to separate ones.

The total area of a combined footing may be, approximately,

_ 900 + 30 =
4

Therefore, try a symmetrical footing as shown in Fig. 5-21(a).

The center of gravity of the loads of the columns is located 1.1 ft. away
from the center of the footing. Using this figure, since the effect of the

Center of gravity
of 900*
LhB

5'-0" l/.l't 8.9 '±

5-0'
JO'-O"
r«;-PLAN

(b)- PRESSURE DIAGRAM

FIG. 5-21. A symmetrical combined footing.

footing's weight is relatively negligible, the soil pressures at the ends are
found as follows, assuming uniform variation in intensity : ,

930 6 X 1.1
i+ + =
/'max
A (
930
9X
A " 6 1.1\
8 X 30
1
30
4.73 k.s.f.

p* = = 3ono
02 k si
,
,

—go—; - - -

240 (1

This, foreseeable at a glance, is unsatisfactory because of the excessive

pressure near one end.
How may the footing be shaped to secure uniform pressure ? Although
not applicable in this case, a shape like that pictured in Fig. 5-22 might
be used. The trial dimensions TD, DL, and are chosen, then MN NP
iscomputed as follows upon the basis that the centers of gravity of the
bearing area and the loads should approximately coincide

X 15 X 7.5 = 8 X 5 X 2.5 + 8 X 9 X NP
NP =11.1 ft, (call it 11 ft.)

930
P = = 3.75 k.s.f.
~ 8 X 20 + 8 X 11
124 SPREAD FOOTINGS [Chap. 5

This pressure is slightly below the allowable, and the footing might be
reduced a little. However, the proximity of the building line in this case
prevents the use of this sideward spreading of loads.
Another method is to use a rectangular footing like that of Fig. 5-21 but
relocated so that its center coincides, at least approximately, with the
center of gravity of the loads, as shown in Fig. 5-23 (a). A trial footing is

then drawn as in Fig. 5-23(6). Because of the probable large upward

bending moment in the central portion of the footing, the two pedestals
will be joined by a 4-ft. rib, forming an inverted T-beam, thus obtaining
considerable depth without the use of an excessive amount of concrete.

c1 P
T S
A A Approx.c.g. B 3a
r> of 900^ 1
\-
1
1

1 ! 1

^n
L
M 0"
'

n JV -,

//-O" . 9-

4tf /6 L 0" 4k)" 4-0

2S-0"

FIG. 5-22. An unsym metrical combined footing.

Neglecting the weight of the footing, the pressure on its bottom is

8 X 3.75 = 30 kips per lin. ft,

The extra applied weight of the footing in excess of the weight of the earth
is, approximately,

(8 X 30 X 1.5 + 4 X 24 X 3)0.05 = 32 kips

Therefore, the gross pressure under the footing is

Total p = 932/240 = 3.9 k.s.f. (satisfactory, less than 4)

For the design of the footing the net pressure of 3.75 k.s.f., or 30 kips per
be used.
lin. ft., will

With a uniform load of 30 kips per lin. ft, and point loading at the
columns, the shear diagram is as shown in Fig. 5-23 (c); the bending-
moment diagram, in (d). The ordinates are approximately but not
technically correct because the center of the footing is to be placed 0.1
ft. from the point C of Fig. 5-23 (a) in order to avoid unimportant refine-

ment of dimensions. Although the projection of the pedestal above the

be satisfactory to assume that the column loads are
rib is small, it will
spread uniformly over the areas of the pedestals. This reduces the
theoretical bending moments under the columns as shown by the dashed
 \

Art. 5- COMBINED FOOTINGS 125

400* Resultant 500 k

900* 89 >
A c B
4-0\ l/'O" 9-0" 6-0"
t c
$ of footing
E
/s-o" /S-O" '

f«;-DIAGRAM SHOWING POSITIONS OF LOADS AND FOOTING

i<400 k

•S^T
y^W
4L
M 20'0"

,4- ft. rib n cm

S0° k 6-0'

eV
^.
JL^L^L. m 3*
^
i-3.75 k/°'

(b)- DIAGRAM OF TRIAL FOOTING AND PRESSURE

> »
2
/2 ° k
trfY
d ^v\ a
For uniform
pressure L (ch SHEAR DIAGRAM
atpedestal
/Zv uniform pressure at pedesta '/

foU-BENDING-MOMENT DIAGRAM

P-0" 4-0" <?4?"

Z!0"2-'0^
R s
V <NT

7"W, T
-4- E
(e)- SECTION A-A (f J-AREA ASSUMED TO AFFECT
SIDE OF PEDESTAL

FIG. 5-23. Analysis of a trial ribbed combined footing.

126 SPREAD FOOTINGS [Chap. 5

lines in Fig. 5-23 (d), and these values will be used for design purposes:
85 near A and 315 ft. -kips near B.
ft. -kips

The determination of the necessary reinforcement at critical points and

the analysis of the trial footing are as follows, referring to Fig. 5-23:

7
A at K = = 155 h1 '
2
'
18 X 0.9 X 4.25

For convenience, the units used are foot-kips and feet. If 10 Xo. 11
bars are used, ^4 S = 15.6 and So = 44.3 in. These are to be in the top of
the rib as indicated in Sketch (e).

If the shear at the edge of the pedestal 2 ft. left of B, Sketch (c), is used,

V (10.7 - 2)30,000
= 128 p.s.i.
(Xo)jd 44.3 X 0.9 X 4.25 X 12

Since this is below the 210 p.s.i. allowed, it shows that

so far it is not neces-
sary to extend bars throughout the top of the rib.
all

Assuming the footing to be a T-beam,

.
at+ J. A _
- 2 X- 1,070,000
- X- 12
-_ 6W
U Q 3 Q g Q6 5p
p.s.i.

which is naturally very safe.

9 40
A atG =
*
= 1L9in -
2

18X0.9Xl.25
Below B, neglecting the shallow top of the pedestal, the required steel may
be assumed to equal

As = = 4-6 m '
2

18 X 0.9 X 4.25

The bond requirement at G, Fig. 5-23(6), is

J_ 4 X 30,000
w°
ujd 210 X 0.9 X 15 "
m
.

"

Obviously, the bending at the section through G will control the reinforce-
ment, and at least 20 No. 7 bars at 6 in. c.c. will be needed.

,
n =
,

&tG
2 X 240,000 X 12 = ___
l > 25
fe
0.3 X 0.9 X 76 X 15' ° P SA ' '

A careful examination of Fig. 5-23(6) and a consideration of the results

of the computations of A sand fc at G show that this point G is a relatively
weak spot in the design. It will be remedied easily by sloping the end of
the pedestal or rib at B as shown by the dashed line to G' in this sketch.
 :

Art. 5-8] COMBINED FOOTINGS 127

This will make the projecting portions of the slab at both ends alike.
Therefore,

As at G =
'
= 3 in "
2

18 X 0.9 X 1.25
_ r = 60,000
20atG
,
,

210 X 0.9 X 15
= 2L1 m
.

It will be very satisfactory to use 11 No. 6 bars at 8 in. c.c. The rein-
forcement near A will be made the same as at G'. Furthermore, fc will
now be small at F and G' and need not be recomputed.
For the side projections of the footing,

3.75 X 2 X 1
A
.

PD = at, = n 4m . .
2

F8X 0.9 X 1.17

- -

_ 3,750 X 2 __
ZoatPp = = 2 83in
, .

2l 0X0.9Xl4
- -

Therefore, No. 4 bars be used at 6 in. c.c. may

Assume that the reinforcement in area VRST, Fig. 5-23(/), is the same
as a corresponding portion of a spread footing. Extend the reinforce-
ment determined for P, Sketch (e), to the edge ST. Similarly, extend
that computed for point G' clear across to edge RS.
The critical shearing unit stresses probably will not exceed the following

vt at F and G', Sketch (b) = " ' ' = 280 p.s.i. (safe)

rr atf Jt ,
-qi t
Sketch w^
[a) = - 2 X 30,000 = Q _ n
320,00
^
n „ X ol 350 p.s.i.
'
. , .
(sale)
.

X O.o 4o
000 X 75
H. ^M = 30 '— - = OO
"

yL
4.If Gl
at M, Sketch
'
f
22 p.s.i. (safe) / N
(6)
76

where the assumed width shown in Sketch (e). Here it does not seem
is as
main section ABasa spread footing with the
to be desirable to treat the
shear determined by that computed at the left of the 45° line to point
L, but the shear should be calculated as for a T-beam. From Sketch
(c), the shear at the left edge of the pedestal at B, Sketch (6), is 320 — 2
X 30 = 260 kips. Therefore,

260,000
VL = 118 P S L
48 X 0.9 X 51 = - '

Using No. 6 vertical, U-shaped stirrups,/,, = 18,000 p.s.i., and the allowed
v'i = 90 p.s.i.,
A v fv = (v L- v'L )bs
2 X 0.44 X 18,000 = (118 - 90)48 X s

s = 11.8 in.
128 SPREAD FOOTINGS [Chap. 5

Therefore, use stirrups between the pedestals as shown in Fig. 5-24,

with a few under the columns to serve as chairs to hold up the top bars.
All these calculations indicate that the thickness of the footing might
be reduced slightly. The central rib might be narrower also, but it is
easier to construct if it is the same width as the pedestals. Such changes
will not cause much saving of money. Therefore, the trial design will be
accepted as far as the ribbed type is concerned, and the steel will be
arranged as shown in Fig. 5-24.
Another possible design for a footing to be used for the conditions in
Fig. 5-20 is pictured in Fig. 5-25 (a). The footing is made narrower than

24'-0"
»
6" lL6"l'-6" 4-0" ,
4'-6" = Cols, and dowels 2L
1 j?
not shown
S?
2£
*6n !i !i
\ch^ r
\Choiirs
;: y:
i ij
j j ( i| i|
is
#6L i i
i
i
i
•

1 1 1 1

^T-r—4j 'I 'l 'I 'l

'I 'l .I i i i
1

i
1
i
i
1

ZiJ J_U-LJ__k A&£&i\

JO'-O"
LONGITUDINAL SECTION

6-*llx23 L0"str.

I- ?"cover

HALF PLAN OF TOP MAT PARTIAL PLAN OF BOTTOM MAT

FIG. 5-24. Details of ribbed footing.

the one in Fig. 5-23 so that the cantilevered ends beyond F and G will
cause more restraint and reduce the bending moment in the central por-
tion K. This is done in order to eliminate the need for a central rib,

thus permitting the use of a heavy simple footing. The footing is

primarily a one-way beam.

The shear and bending-moment diagrams for this footing are shown in
and (d). The pedestals are assumed to spread the column
Figs. 5-25 (c)
loads uniformly.
The magnitudes of the bending moments under the pedestals are really
questionable. In the case of a symmetrical isolated footing, the bending
moment was considered to be constant across the bottom of the pedestal.
This was permissible because the opposite sides were balanced so that rota-
Art. 5-8] COMBINED FOOTINGS 129

Approx. center of
gravity of 900 k

400 k soo>

\A\ If J
^Pedestal

S'-6" ll'-O" 9-0" 7-6'

33-0"
fa;-PLAN OF FOOTING

3'-6" 4-0" !
16 L 0" ., 4 L0" ,
6-6"

*|
{

|j t
S^IOOk/lm.ft. /25k//in.ft~
\
J
^ jjjj
f"

*-27.3 k///n. ft or J. 9 W
C6J-ASSUMED LOADING AND PRESSURE DIAGRAMS

iWik .„-i

?#*" -

rcj- SHEAR DIAGRAM, AS LONGITUDINAL ONE-WAY BEAM

+I68'\ F A

(rfj-BENDING-MOMENT DIAGRAM, AS LONGITUDINAL ONE-WAY BEAM

10.83' 7-6" Pedestal 4-0 "wide

2*0*
mx >-'<?
'
y s 4'-0"

\ / 1
1

$t 1

u
1

1
1
r i

jikf
\ "1?
"^ CM
\ II

V x
I
'v/yt'lifyp

\
\
!

V
:'
c
a »" o

7-0"
. »

M
oT 4^0" d
rej-PLAN OF TRAPEZOIDAL PRESSURE AREAS tfV-SIDE VIEW OF SLOPED PEDESTAL

FIG. 5-25. A combined slab-and-pedestal footing.

tion of thecolumn did not occur. In this combined footing, however, this
assumption does not seem to be valid.
For example, Fig. 5-26 shows the shear and bending-moment diagrams
when the footing under the pedestal is assumed to be rigidly fixed.
Sketch (b) indicates that the bending moments at F and H and at G and ,
130 SPREAD FOOTINGS [Chap. 5

J, are unequal. This fixed condition is improbable. Unless the columns

are very stiff, will rotate until they conform with whatever slope the
they
footing under them takes when it is in equilibrium. It is also obvious
that the assumption of point loading at the columns gives results that are
unreasonably large for the bending moments at the columns. It there-
fore seems desirable to adhere to the principles of the calculations used in
computing the ordinates in Fig. 5-25 (d). Continuity and frame action
of the columns and the footing combined may be considered when the
columns are so stiff that their resistance to rotation becomes important.
Generally, however, the columns are relatively flexible.

Uniform load -27.3 kf,in ft

-

^-2/8*
^m 96 k
J
F 2l8 h
ISO

3-6' 4^0" 16-0' 4 L0" S'6"

Pedesfal Pedestal
rfzJ-SHEAR DIAGRAM

Pedestal Pedestal
r<V-BENDIN6-M0MENT DIAGRAM

FIG. 5-26. Improbable shear and bending-moment diagrams, assuming ends fixed at pedestals.

When a combined footing is large and relatively flexible, the pressure

under it may not be so uniformly distributed as pictured in Fig. 5-25(6),
even when there is no eccentricity of loading. Upward bending between
A and B may tend to reduce the intensity near K, but much depends upon
the soil modulus of the ground. Uniformity of pressure seems to be a
reasonable assumption.
The maximum shear in this footing is at J, but the area of effective
resistance to transverse shear is questionable. Some broad assumptions
must be made. Therefore, referring to Fig. 5-25(e), the following pro-
cedure might be adopted for such cases:
1. Let RW represent the point of zero shear in Sketch (c).
2. From the corners of the pedestal draw lines at 45° to the sides of the
footing, as shown by the dashed lines Ux and Ty.
Art. 5-8] COMBINED FOOTINGS 131

3. At first, assume that the pressure under the area RxUVx'W causes
shear on UV, and that the pressure under the area SyTXy'Y causes
shear on TX.
4. Similarly, assume that the pressure under the areas yTUx and
x'VXy' causes shear on TU and VX, respectively.
5. If the estimated force on UV greatly exceeds that on face UT or VX,
as it obviously does here, there is a general tendency for the whole width
of the footing in the plane UV to resist shearing. The projecting parts
beyond U and V tend to transfer the shears around the corners to the adja-
cent portions of sections UT and VX. In such a case, compute the shear
on UV as though it alone resisted the force applied by the pressure under
area RxUVx'W, then temper the result by using good judgment as to
how excessive the computed stress may be.
Use these previously designated areas for estimating the diagonal
6.

tension beyond any corresponding face of the pedestal. For example,

at L of Fig. 5-25 (e), assume the effective width to be that of the section
across the assumed area at a distance (d) from the face of the pedestal.
7. Use the computed shear at such an edge as UV to estimate the bond

unit stress on the bars perpendicular to that edge.

In order to obtain some ideas for assuming a trial depth, compute the
approximate minimum permissible depths for shearing resistance at UV
and mn of Fig. 5-25 (e). Therefore, compute first the assumed pressure
under area RxUVx'W.

- 2 X L 5 X L5\
Vuv = ( 7 X 8.83 3 Q0() = 332,000 lb.

, _ V 232,000 =
.

ln "

v T bk 600 X 48 X 0.3
Then test the diagonal tension at section mn on the basis of this thickness.
Line mn is beyond x and x' , hence, the pressure under area RmnW is
V = 7 X 6.58 X 3,900 = 180,000 lb.

V_ 180,000
VI P - S - L
bjd 84 X 0.9 X 27

It thereforeappears that an effective depth of 27 in. is the minimum desir-

able. This will be tried. The areas of reinforcement are to be deter-
mined, and the compressive strength should be checked.

720
A at K= = 2 in 2
s
18X09X 2^5
-

Since the largest bending moment is at the center of the footing, this is the
critical point for compression in the concrete. Therefore, instead of
132 SPREAD FOOTINGS [Chap. 5

assuming A = 0.3, use Fig. 1 in the Appendix and find k = 0.34, using
-

n = 10 and p = 20/(84 X 27) = 0.0088. Then j = 1 - %k = 0.89.

Therefore,

K= 2 X 720,000 X 12
= 930
fe at =2 p.s.i. (satisfactory)
0.34 X 0.89 X 84 X 27

The maximum bond stress on the basis of the change in bending mo-
ment will occur between J and N, Fig. 5-25 (d) Thus the required surface .

area of the reinforcement is, using j = 0.9,

ZiO = change in M per lin. in.

(jd)u

ZO = 420,000 X 12
41.1 in.
2

24 X (0.9 X 27) X 210

Therefore, use 14 No. 11 bars for top reinforcement (.4 S = 21.8 in.
2
,

So = 62 in.). This is a bit generous, and all bars do not have to extend
for the full length. They will be arranged as shown in Pig. 5-27.
{

J3'0'
5-6" lO'-O" 2-0". *9—= 2"cove, —->i

—
s
,*llstr. Col. and pedestal relnf.
4 — h -e=!E=!==!=3J- -fe^^fek ==fefi£ i_i not shown,

IZ I" zk^E^=«r:ife€.^Efe5^4lh££
a fK s *
ST
1

ft{

^sfe *5 str. ties

r
5=
S<S>2-6" 18'spcs.i /lis 8
'

CROSS SECTION

PARTIAL PLAN OF TOP MAT PARTIAL PLAN OF BOTTOM MAT

FIG. 5-27. Reinforcement for combined slab-and-pedestal footing.

Assuming that the pedestal will be poured after the footing has set and
that the longitudinal shearing strength at the junction is not sufficient

to cause the pedestal to participate effectively in the beam action of the

footing, the effective depth under the pedestal is assumed to equal that
of the footing alone. Therefore,

520
A s at B = = 14.3 in. 2
18 X 0.9 X 2.25

Using V 240 kips from Fig. 5-25 (c),

V 240,000
So at 11 V = 47 in.
ujd 210 X 0.9 X 27

Therefore, use 16 No. 9 bars placed as indicated in Fig. 5-27. A few of

these could be omitted under point A since the shear and bending moment
at H are somewhat smaller than at J, Fig. 5-25(c).
Art. 5-9] SPECIAL COMBINATIONS 133

The transverse reinforcement in the footing under and near the ped-
estals is important. Elsewhere, it serves mostly as ties. Assuming that
a width xy, Fig. 5-25(e), is to serve as a header to resist the reactions
brought to it by the longitudinal steel outside UT and VX, the total
force on each of these sides is

Force = RS X 1.5 X 3.9 = 18.33 X 1.5 X 3.9 = 107.2 kips

Then
M xy = 107 X 0.75 = 80.4 ft.-kips
80.4 _q
a
As —
~ ~_ m
.

"

18 X< 0.9 X 2l7

107,200
= 21.8 in.
210 X 0.9 X 26

Use 12 No. 5 bars. This reinforcement will be placed within the width
xy of Sketch (e), and the same amount will be used under the other
pedestal. If this transverse bending were objectionably large, the sides of
the pedestal might be sloped as shown in Fig. 5-25 (/), thus enabling the
main footing to act entirely as a beam that is reinforced in one direction.
5-9. Special combinations. Some illustrations of special construc-
tion are pictured in Figs. 5-28 to 5-30. The purpose of each is to show a
design or plan for a foundation that was used to support a structure that
could not be supported very well on ordinary isolated footings. The
drawings are greatly simplified in order to emphasize the principles
involved rather than the details. However, the main reinforcement is
shown by dotted lines so that the reader may see what paths of resistance
were provided.
When planning such special construction, the designer should approach
each problem with an open mind. He should find out what the existing
conditions are, what is wanted, and how he may accomplish it. Amazing
are the instances of an individual's reliance upon precedent what was —
done on some other job. Too often he does not know the reasons for the
other construction, and these may be very different from those that should
affect his problem. Creative engineering thinking, a thorough knowledge
of structural action, and a keen sense of practicality these are attributes —
that he should possess and exercise instead of thumbing through books
and looking over blueprints in search of a design already prepared by his
predecessors.
In Fig. 5-28 (a) is shown a condition that occurred in a machine shop
where a conveyor tunnel was to run under a building and close to one line
of crane columns. If the footing for a column were made separate from
the tunnel, it would have an objectionable eccentricity. If it were sup-
ported along one edge of the tunnel by a seat at A, it would rest partly
134 SPREAD FOOTINGS [Chap. 5

upon backfilled or disturbed soil so that it would probably settle and rotate
about the seat as an axis. Therefore, the column foundation was incor-
porated in the conveyor tunnel construction, as shown in Sketch (b).
The footing B was extended sufficiently to prevent any serious tipping
because of the large resistance of the soil under the long narrow bearing
at C under the tunnel wall. The floor D was poured later so as to avoid
its participation in the original distribution of the heavy dead loads.

Figure 5-29 (a) shows the typical foundation used under one pair of
columns for a long elevated coalbin that was incorporated in an extension
of a boiler house. Column A had to be close to the existing wall because
of the coal chutes and aisle clearances. The additional load could not be

Direction of
rotation

Column ,F/oor s/ab

' 'Y4

1
Footing
Con veyor
tunnel
-Pilaster
'/ Construction joint under
footing
W/V/' V// W '- //A/*yAW(*/WVW.-
Footing
at column
fa/-C0lUMN FOOTING ON "HARD SPOT" ^-COLUMN FOUNDATION EXTENDED DOWN TO
ATONE SIDE-UNWISE CONSTRUCTION FIRM GROUND- BETTER CONSTRUCTION

FIG. 5-28. Column foundation when near edge of excavation.

supported by the existing wall and its footing; neither should this footing
be cut off on one side to permit the new construction to encroach upon it.
A combined footing was therefore used. The pedestal for column A was
set close toone end of the slab, the latter being isolated from footing C.
Column B was set similarly with respect to end D in order to equalize the
pressures under the footing. The footing was primarily a one-way beam
with tension in its top, and the slab was thick enough to withstand the
shears caused by the concentrated loads at the pedestals. If the loads

had been sufficiently heavy, the pedestals might have been sloped as
shown in Sketch (6), or a ribbed footing could have been used.
Sometimes, in a situation like that of Fig. 5-29(a), the footing may be
shaped as in Sketch (c), where end E is cantilevered in order to cause
restraint under column B and thereby reduce the tension in the top of the
slab. The dimensions are chosen so as to keep the center of gravity of the
footing under the resultant of the applied loads. The reinforcing band F
acts somewhat as in an isolated footing. This shape of footing is also
Art. 5-9] SPECIAL COMBINATIONS 135

when the load on column A

useful is much larger than that on B, and even
when B exceeds A.
Still another case is illustrated in Fig. 5-29 (d). An extension was built
alongside a building, but the existing framing and foundations could not
support the added loads. A new row of columns G was required as close

Mew column
-y//vew Stepped pedestal ^
g \X W/H/f&t/i&f/w&WWM'W'-
Square pedes fa/ Square pedestal

{Footing

D Side Elevation
Scheme of rein f.

Plan
^-COMBINED FOOTING WITH SQUARE PEDESTALS
Top reir,if Pedestal CI

Bottom rein f. shown

f6i-F00TING WITH TRAPEZOIDAL PEDESTALS rcMRAPEZOIDAL COMBINED FOOTING

Column
LVX Top rein f RibB^ f~\G
Existing wa/i
*and footing

Footing J^ *M
Side Elevation KNNN/V1/M
mhcfK-* Gymnasium
III
i 1

I'M
i 1

1 II
-t
~T SEE! MMiu
I IIIII
! !

Mill
*- Heavy flight

Bottom reinf.shown
P oi n
1
A„ Mill
CrfJ-SPECIAL COMBINED, RIBBED FOOTING

(e)

FIG. 5-29. Examples of combined footings on weak soils.

to the old wall as possible in order to support the new framing and
provide a vertical deflection joint at the junction of the new and the old
structures. A large ribbed footing was therefore used under G and L,
the latter column serving as an anchor for the cantilever. In order to
obtain reasonably uniform pressures, footing J was made with the wings
136 SPREAD FOOTINGS [Chap. 5

K. A clearance of 2 in. was provided at M in order to have space for the

small but inevitable settlement that would occur as the soil was com-
pressed by the new structure. Another case is illustrated in Fig. 5-29 (e).
If combined footings, like A of Fig. 5-30 (a), are used under a structure
that is close to building lines or other obstructions, special construction is

required at the corner B. One possible construction to hold the corner

column is that used in this case, where a pair of combined footings joins B

f Bui/ding tine
D < ? BU
1

, I&
.£
Pier

1
03
"R
Q Substitute piers
>s

Bi i

faJ-COMBINED FOOTING TO SUPPORT CORNER

L>P ^-QUESTION OF CHOICE OF
VS. COMBINED FOOTING
PIERS

(Foundation walls are even stronger)

N
At_ Pit
Solid wal Thin wall
with footing and pilasters
with footings
M
O
Plan Section A-A
CcJ'DEEP PIT COMBINED WITH (d)- SHALLOW PIT COMBINED WITH
COLUMN FOUNDATIONS COLUMN FOUNDATIONS

FIG. 5-30. Some examples of special foundations and footings.

with C and with D, each footing being sufficient to transfer its share of the
load back onto adjacent adequate areas. Each footing resists the tend-
ency of the other to twist under the eccentric load of B. Sometimes,
heavy foundation walls serve this same purpose even better.
When combined footings are on plastic soil above sloping rock or gravel,
as pictured in Fig. 5-30(6), harmfully unequal settlement should be pre-
vented. It is possible to increase the area at E so that the bearing pres-
sure is so much
than at F that the estimated settlements of the
less there
footing are equalized. Another remedy is to increase the projection at E
and decrease that at F to accomplish the same purpose. If the rock is
sufficiently close and if the importance of the structure justifies the
Art. 5-9] SPECIAL COMBINATIONS 137

expense, it may be preferable to extend the pedestals to rock, as indicated

at J and K.
At a large war plant, a deep pit was required close to crane columns, as
pictured in Fig. 5-30(c). The pit was sunk as a caisson, then a heavy
cantilevered beam LAI was incorporated in the top of one side wall.
Crane loads on the columns were eccentric with respect to the bottom
OQ. However, the pit was so deep (over 50 ft.) that the abutting resist-
ance of the earth on the side walls, such as NQ, was so great that no
appreciable tilting could occur. If the pit had been shallow (perhaps

10 ft.), it would have been desirable to extend beam N to the bottom as a

wall or as two pedestals joined to a thin side wall, and to have added a
footing at Q, as shown by Sketch (rf), in order to avoid too much eccen-
tricity of load on the bearing area.

PROBLEMS
5-1. A stepped footing like that of Fig. 5-18(a) has a pedestal 2
ft. 6 in. square

and 3 ft. 6 in. high. above the ground but has no seat for the floor
It projects 6 in.
slab. The spread footing is 8 ft. square, 1 ft. 9 in. deep, with an effective depth of
1 ft. 6 in. The column reaction is 450 kips. Design the tensile reinforcement for
this footing. Check the bond, shear, and diagonal tension in the concrete.
5-2. Assume a stepped footing with a pedestal exactly like that of Fig. 5-18(a).
The main footing is 9 ft. square, is 27 in. deep, has 3 in. cover over the bars, and is

to support a column load of 600 kips. Design the reinforcement and check the
bond, shear, and diagonal tension.
5-3. Assume a footing exactly like that of Fig. 5-18 except that the depth of the
spread footing increased 6 in. and the main bars are No. 7.
is The column load
is 550 kips.Check the footing completely.
5-4. Assume a footing like that of Fig. 5-7 (a). The column is 24 in. square and
has a load of 350 kips. The footing is 7 ft. 6 in. square, is 24 in. deep, and its top
is 3 ft. below the ground. The reinforcement is two perpendicular bands of 13
No. 6 bars at 6 in. c.c. with a cover of 3 in. Is the footing safe?
5-5. Design a footing like that of Fig. 5-7 (a) to support a load of 300 kips if the
soil is a rather soft silty clay and the frost depth is 5 ft. The column has a diam-
eter of 30 in. Use square column having the same area.
5-6. Assume that a spread footing is similar to that of Fig. 5-7 (a) but is 6 ft.
wide and 8 ft. long. The column is 24 in. square and has a load of 320 kips. The
footing is 1 ft. 9 in. deep. The lengthwise reinforcement is 13 No. 6 bars at
5 in. c.c; that across the narrow way is 13 No. 6 bars at 6 in. c.c. centrally
located. The cover over the bars is 3 in. Is this footing safe to hold a load of
320 kips?
5-7. Assume a sloped-top footing like that of Fig. 5-16(c) except that it is 8 ft.

square, 1 ft. 6 in. thick at the edge, and the load is 380 kips. Is the footing safe in
all respects?
5-8. Design a sloped-top footing to support a rectangular concrete column
28 X 32 in. having a load of 700 kips. The allowable bearing value of the soil is

6 k.s.f.
138 SPREAD FOOTINGS [Chap. 5

5-9. A slopecl-top footing for a 36 in. square concrete viaduct pier has the fol-
lowing dimensions: width, 12 ft.; length, 14 ft.; thickness at center, 3 ft.; flat
portion around pier, 6 in.; edge thickness, 2 ft.; cover over steel, 3 in.; reinforce-
ment in long direction, 23 No. 8 bars at 6 in. c.c; reinforcement in short direction,
17 No. 8 bars at 8 through the center with one extra rod 8 in. from each
in. c.c.

end of the footing. Is this footing satisfactory if the load on the pier is 850 kips?
5-10. A footing for a steel column with a base plate 18 X 30 in. is limited to a
width of 5 ft. parallel to the narrow side of the billet. The load on this crane
column is 450 kips. The allowable bearing value of the soil is 3 tons per ft. 2
Next to one of the narrow sides of the footing is to be a duct trench with its bot-
tom 5 ft. below the floor. Design a footing to support this column.
5-11. Assume two concrete columns as in Fig. 5-20 except that both are 2 ft.
6 in. square and have loads of 480 kips each. They are spaced 18 ft. c.c. The
bearing value of the soil is 4 k.s.f. Frost depth is 5 ft. Design a combined foot-
ing to support these two columns.
5-12. Two steel columns at an expansion joint in a mill building are 2 ft. 6 in.
c.c. The base plates are 18 in. wide and 36 in. long. There is 12-in. clear space
between the billets, and their bottoms are 6 in. above the floor. The column
loads are assumed to be 300 kips each. The soil is a fine sand with thin lenses of
inorganic silt. Design a combined or common footing for these columns.
5-13. Assume the columns, spacing, pedestals, and loads shown in Fig. 5-20.
Then assume the footing of Fig. 5-22 to be used under these columns. Determine
the thickness of concrete and the reinforcement needed for this footing.
5-14. A combined footing is to be used under three concrete columns 36 in.
square. The columns are 15 ft. c.c. The load on the central column is 600 kips;
that on each of the outer columns is 500 kips. The allowable bearing value of the
ground is 3 tons per ft. 2 Frost depth is 5 ft. Design a footing for these three
columns.
5-15. A heavy machine weighing 750 kips is supported upon a base that is a
hollow square 12 in. wide and 10 ft. square in outside dimensions. It rests upon a
footing 18 ft. square and 24 in. deep. The reinforcement at the bottom consists of
two perpendicular bands of 17 hooked No. 11 bars at 12 in. c.c. with 3 in. of cover.
No top reinforcement is used. Assume the pressure on top to be spread uniformly
over the hollow square bearing area directly under the machine. The soil pressure
is also assumed to be distributed uniformly. Is the footing satisfactory?
5-16. A brick chimney for a three-story house is 2 X 5 ft. at the basement floor.
It weighs 30 kips. The soil is soft clay. Excessive settlement will crack the walls
of the house. Design a footing for this chimney.
5-17. Design a square sloped-top footing to support a 2 ft. 6 in. round concrete
column supporting a load P = 450 kips. The soil is a 10-ft. top stratum of fine
sand. Under this is a 20-ft. layer of clay like that represented by curve A, Fig.
4-19. Assume frost depth is 5 ft. Allowable pressure on the sand is 5 k.s.f.
Estimate the settlement of the footing.
5-18. Design a rectangular combined footing 10 ft. wide to support two col-
umns 18 in. square holding loads of 250 kips each. The columns are 18 ft. c.c.
with 3-ft.-square pedestals. The allowable pressure is 2 k.s.f. Frost depth is
4 ft. Estimate the settlement if the soil is a silty clay layer 22 ft. deep under the
footing and has the characteristics shown by curve B, Fig. 4-19.
 6
FOUNDATION WALLS

6-1. Ordinary simple walls. Houses, some industrial buildings, and

many other light structures without basementsmay be founded upon
shallow simple concrete walls. Their depth may depend upon the esti-
mated penetration of frost below the ground's surface, the desired embed-
ment beneath this surface, and the distance down to suitable soil, as illus-
trated in Fig. 6-1 (o) ; the elevation of the top, upon the position of the top
of the soil or of the ground floor of the structure. The thickness may be
controlled by such matters as the thickness of the masonry or other wall
resting upon the substructure, the minimum desired for pouring concrete,
the total of the thickness of the supported wall plus the allowances for such
details as a watertable and a seat for the floor construction. Some
detailed sketches at the tops of foundation walls are given in Fig. 6-1.
The loads may be so small that firm sand and gravel will support the
foundation wall and the superstructure without any footing, as' indicated
in Fig. 6-1(6). In other cases, a plain concrete footing may be used as in
Sketch (c), and its width may be determined by the requirements of the
building code, by the bearing value of the soil, or by good judgment.
Heavy brick walls on weak soils may necessitate the use of wider footings
reinforced as indicated in (d). However, in these simple cases of wall-
bearing construction, the loads may
be considered as uniformly dis-
tributed along the wall and across the footing unless truck doorways and
other large irregularities occur that are sufficient to prevent the foundation
wall from spreading the loads satisfactorily.
In cold climates it is generally desirable to extend foundation walls
under truck doorways when the soil and moisture conditions are likely to
cause frost heaving and subsidence of a concrete floor and driveway. An
apron slab, as illustrated in Fig. 6-l(/), is advisable to span across any
weak area near the This combination will generally avoid inter-
wall.
ference with doors and cracking
of pavement.
Shrinkage of concrete, caused by drying and chemical action during
curing, and shortening due to a decrease in temperature make it desirable
139
 r

140 FOUNDATION WALLS [Chap. 6

to reinforce concrete foundation walls longitudinally. The amount of

steel required is problematical. The shrinkage occurs gradually but is

usually completed in 4 to 8 weeks. A wall will try to shorten somewhat in

accordance with its length, a figure of ^
in. in 100 ft. being useful for

obtaining a scale on the probable deformation. This contraction is

assumed to be uniform over the wall's cross section. Not only are con-
traction joints necessary, as explained in Art. 6-8, but the reinforcement

Bricks or
S concrete
blocks

si/iv/AW
WW77&

• Sloped bottom may be used if

grade is less than I '-3, otherwise
make bottom in steps of I to 2'
CaYlW ELEVATION OF A SMALL BUILDING
ti*=? i*?± 1

(b)-k SIMPLE WALL fc>-WALL AND FOOTING

Inside of_
Concrete blocks I"rounded corner building
''or hollow tiles Apron slab 3 to 10" V Dowels

®MB====mW^
g
mmrn*
J
"^Cinders or ravel :•£&. "seats
2 shelf

Weepers
ff)- APRON AND WALL AT TRUCK DOORWAY

(J I r-\ — —s— — i

Bars or mesh
(d)-\NALL AND HEAVY FOOTING teJ-WALL WITH BRICK TOP

FIG. 6-1. Some examples of small foundation walls.

should be sufficient to prevent localized visible cracks. A steel area

0.002 times that of the cross section of the concrete is usually sufficient
for this purpose, and the reinforcement may be distributed evenly over the
cross section, but it is desirable to place a large portion of it near the top
and bottom, as indicated in Fig. 6-1 (d), or at any point of maximum
exposure to temperature changes. If a crack cannot start at the edges,
it will not be likely to occur at all.

When there is considerable surface area of a foundation wall exposed to

wide and rapid variations of air temperature, thermal change may cause
Art. 6-1] ORDINARY SIMPLE WALLS 141

curvature or cracking of the wall. The longitudinal deformation may be

a total variation in length of 3 4 in. per 100 ft. per 100°F. change of temper-
ature. For example, refer to Fig. 6-1 (d) again. When the top of the wall
is cooled quickly by the air, the top shrinks and bar b tends to distribute
the deformation as hair cracks. In the meantime, the wall tends to curl
upward near the ends because the bottom is not shortened so rapidly as
the top. The weight and stiffness of the structure would probably pre-
vent this curvature so that, were it not for bar b, one or two serious cracks
might open up. Similarly, when the air temperature rises suddenly, the
expansion of the top of the wall causes the wall to try to bow upward
(convex on top) near the center. Bar c then resists the cracking that
might occur when curvature is prevented. It is obvious then that rein-
forcement is needed near both the top and the bottom just as it is for
resisting shrinkage. The steel area specified for shrinkage will generally
resist the thermal effects as well. However, in extremely exposed situa-
tions, the reinforcement should be increased to 0.0025 or 0.003 times the
cross section of the concrete.
Vertical reinforcement of small foundation walls may not be necessary
when they are primarily compression members between the superstructure
and the soil. A few bars may be useful as spacers and supports for the
longitudinal steel, and they may be needed if anchor bolts cause sufficient
uplift.
Foundation walls are sometimes built of concrete blocks, and even of
bricks. Although it may be slightly more costly, poured concrete (rein-
forced) below ground is preferable, especially in cold wet climates.
Brickwork at the surface of the ground at A in Fig. 6-1 (e) is also likely
to disintegrate. Any saving caused by a substitute for a good foundation
is certain to be unimportant compared to dissatisfaction with the service

performed by a structure that is not built properly.

Seldom is there need to calculate unit stresses in a foundation wall
under a wall-bearing superstructure, except to determine the reinforce-
ment needed in the footing, and for the bearing pressure upon the soil.
There is, however, a very real opportunity for sensible design and
planning.
When foundation walls are built it is generally very desirable to install
open-joint drains (V.C. bell-and-spigot type is preferred by the author)

around the outside of the footing to intercept and remove ground water.
These should be somewhat as shown in Figs. 6-1.3(c), 6-13(rf), and 6-15(a),
or even deeper, with strategically located Y's and capped vertical risers
for cleanouts. One or more inspectable manholes in the line will be use-
ful to enable one to see that the drains are functioning. It may also be
desirable to place the floors on a top layer of compacted sand or gravel
(sometimes adding agricultural tile drains) with porous backfill inside the
142 FOUNDATION WALLS [Chap. 6

walls and weepers to allow rainfall during construction or water from

spring effects to reach the drainage system beneath the footings
if the soil

israther impermeable. Otherwise, rains during construction or ground

water under pressure may soften fine-grained weak soils before the floor
is placed or afterward.
Example. A businessman wishes to build a fire-resistant commercial
garage. The walls are to be 12-in. brick unlined, the roof is to be precast
concrete slabs with insulation and built-up roofing on top. The roof is to
be supported by wall-bearing, steel bar joists or trusses 45 ft. long. The
average estimated load per linear foot on the side walls is 4,100 lb. at the

Brick bearing wa/J

Assumed span =8'-0 "
Water
1'
table
V;
3W 5-0'

A
> — —ft: a .."...t-*-4 -
, Jfca
JB

Sloped
Surface /
""
ofground Assumed limit of/' 6" to 12"
weakened area caused-*. / compacted
by excavation / sand or gravel

/-Weepers
a /
E .iitG Iff /
P a t J> ..

/
\>&J ^trr-p*- ttv-:
m$/
tu I
^j \K

FIG. 6-2. Foundation wall for a commercial garage.

floor line.The floor level is to be 2 ft. above the ground for most of the
wall,and the 6-in. reinforced-concrete floor slab is to be supported at its
edges by the foundation wall. The probable frost depth is 3 ft.; the
allowable bearing value of the soil, 2,500 p.s.f. Design a typical section
of the foundation wall for this structure.
The first step in the planning is the determination of various structural
features that are desired. For example, a 1-in. watertable is to be used
because of appearance, a 2-in. shelf is to support the floor slab, the founda-
tion is to be poured concrete, and the brickwork is to begin at the floor
line. These features are pictured in Fig. 6-2.
The next step is to select the depth of the foundation. This is made
3 ft. 6 in. in order to be sure that it is below the danger line from frost.
Art. 6-1] ORDINARY SIMPLE WALLS 143

The third step is to determine the necessary footing. Some allowance

should be made for the fact that a truck wheel or other heavy load may be
placed on the floor slab near B. Since the shelf CD holds the edge, the
floor will have to support any load P and to span from CD to some area of
soil R. The floor slab will be thickened along the wall, as shown in Fig. 6-2.
The slab and the foundation wall will distribute any concentrated load P.
Therefore, assume that P represents a rear wheel of a loaded 10-ton truck,
that the load is 3.5 tons, that the span of the floor slab is 8 ft., and that the

load is distributed over a distance of 5 ft. along the shelf CD. The
average load per linear foot of wall caused by P is then (7,000/5) X — %
880 lb. The weight of the floor slab carried by the shelf may be 90 X 4 =
360 lb. per Lin. ft. Since part of the wall is above ground and part below,
its weight will be considered as that of ordinary concrete with no reduc-

tion for the displaced earth. The projections of the footing will be
excluded from these calculations. The 15-in. wall at 150 p.c.f. will weigh
5.5 X 1.25 X 150 = 1,030 lb. per Lin. ft. Therefore, the approximate
load per linear foot of wall at JK for design purposes is

Superstructure 4,100 lb.

Live load 880

Floor slab 360
Foundation 1,030
6,370 lb.

The width JK should be 6,370/2,500 = 2 ft. 8 in. The dimension a, in

Fig. 6-2, then equals S}4, in. If t is 10 in., the tensile stress in a canti-
levered plain-concrete footing may be

U = M = 2,500 X 0.71 X 0.35 X 12

= 7,450
= P ' S ' K
•

S 12 X (10 /6)
2
200

This is small, but, if tension in the concrete is not to be trusted, use

reinforcement.

,
As ~
M __ 7,450 _
" nrr
U>Ubb m
.

*
2

fjcl 18,000 X 0.9 X 7

^o = —Vr, = 2,500(8.5/12)
_., _ v , „ „ , , _ = 1.34 in.
ujd 210 X 0.9 X 7

Therefore, use No. 3 bars at 10 in. c.c.

The floor slab is to be tied to the foundation wall. The amount of

steel to use ischosen arbitrarily as No. 4 bars at 24 in. c.c. If placed at
the top, they offer a small bending resistance; if near the bottom of the
slab, they permit the end of the latter to be simply supported, which is
generally preferable. These ties eliminate any danger from lateral pres-
144 FOUNDATION WALLS [Chap. 6

sures of earthand surcharge that might otherwise affect the design of the
foundation. the floor is cut loose from the wall, the latter will have
If

to act as a small retaining wall and foundation combined, but the floor
load need not be added to the vertical forces on the wall.
The longitudinal reinforcement in the wall will be computed as
A =
s 0.002(1.25 X 4.67 + 0.83 X 2.67)144 = 2.3 in. 2

Five No. 6 bars will be used, two in the footing and three in the top as
pictured in Fig. 6-2.
Special details and modifications of such a design as this will have to be
made any practical job.
in
6-2. Spread footings and grade beams. Most large structures have
some kind of structural framework with columns which transmit heavy
localized loads to the substructure. It may be that the loads from the
"curtain walls" of the first story and from the first floor are relatively
light so that it is desirable to use large spread footings under the columns,
whereas a reinforced-concrete beam between them is sufficient to support
these other loads. The footings are then designed independently except
for provisions for the grade beams.
This type of construction is especially adapted to industrial mill build-
ings in mild climates, when supported upon a stratum that lies several
feet below the surface. It minimizes the excavation and the materials
required to build the substructure. There may be many modifications
and refinements to suit the peculiar requirements of a particular structure.
One case will be illustrated by an example.
Example. Assume that Fig. 6-3 (a) pictures some of the details of the
construction to be used at the ground floor of a small industrial building in
the Los Angeles district of California. The structure has ten 20-ft. bays,
steel columns holding bracketed crane runways, 75-ft. roof trusses, and
strong longitudinal roof and wall bracing. Assume that the foundation of
an ordinary side-wall column, not part of a bracing bent, is subjected to
the vertical loads shown in Sketch (a). The heavy concrete floor is to be
cut loose from the foundations. The soil is fine sand capable of holding a
unit load of 3 tons per ft. 2 but this material is from 4 ft. 6 in. to 5 ft.
,

below the surface of the ground. The soil above it is unreliable for
supporting heavy loads. Design a typical spread footing and grade beam
for this structure.
The pedestal under the steel billet should be at least 2 ft. wide in the
direction of the wall. The span of the grade beam to support the brick-
work will thus be approximately 18 ft. This member should be designed
first. Figure 6-3 (a) shows that the bottom of the base plate of the col-
umn is to be 4 in. above the floor line. An allowance of 1% in. will be
made for grouting under it. This elevation 2% in. above the floor will be
Art. 6-2] SPREAD FOOTINGS AND GRADE BEAMS 145

selected as the top of the grade beam as well as of the pedestal. The
grade beam under the brickwork, but a width of 8 in., equal to the
is to be
thickness of the wall, seems too narrow for an 18-ft. member. There-
fore, it will be made 12 in. wide as shown in (6). Its depth is made 2 ft. in
order to extend below the ground somewhat and to be sure that it is stiff
enough. The bottom of the beam might be made triangular as shown by
the dashed lines. The reasons for doing so might be to reduce the bearing
resistance under the beam if appreciable settlement of the main footing
is expected and to reduce frost heaving. However, this "cutting edge"
Surface
ofground 4. ofcol.
__£_ /"key (Reinf. lie
400*/'-~\ _^
8" 3"
8 brick wall Floor
6" slab
5L0"fo sill of
conf. sash—+.
Dowel if ' Grade beam
psijmx needed fl2"x 12"plain
Floor line^ i premolded concrete seal
=-? filler
floor
fi-S"
slab s±i Pedestal

Footing
anchor bolls
(a)- DETAILS AT COLUMN BASE (W-GRADE BEAM fcZ-GRADE BEAMS SEATED ON FOOTING

4. of col. ^. 12 "x24 grade beam <£ of col.

rt-rf— +-4—
'

'

3rd pour

-H- \ Pedestal
.

Const, jf.
2-J"sa.

JS-*6 @S"c.c.
2ndpour
both ways 3sets*4
Isi pour 2-*4 n
(~i « •_»- »_«. _. *_ mi±*Sl at top of *
pedestal
L 7-0 "sq. fooling

(W-GRADE BEAM AS PART OF PEDESTAL fe^PLAN OF PEDESTAL AND FOOTING

FIG. 6-3. Spread footing and grade beams.

has little value, especially in cold climates. The grade beam really
ought to be deep enough to get below frost depth; otherwise heavmg may
occur. In this case, a narrow flat bottom can be used since there is no
danger of frost.
A grade beam should be stiff and reinforced as a simply supported mem-
ber if it is to be unaffected by any unequal settlement of the footings.
On the other hand, the series of beams and footings may be tied together
except at contraction joints. If so, it may be best to use a continuous
beam with adequate stiffness so that it will support the masonry wall
without having its deflection cause cracks in the masonry. The beam
should be designed as a self-supporting member even though the earth
might offer some bearing resistance under it.
146 FOUNDATION WALLS [Chap. 6

The load on the grade beam in this case is dead load only. The mem-
ber will be designed as a continuous beam having an end moment equal
to that of a beam with fixed ends, but the center of the beam will be
reinforced equally.

iv = 400 + 1 X 2 X 150 = 700 lb. per lin. ft. of wall

M = v
^f = 700 X ^= 18,900 ft.-lb.

*L 18,900
4
s
- m '
2

f jd
s 20,000 X 0.9 X 1.75

T at+ ,u
So A
the ends = -^ =
y
210
700
x
X
, 9
9
x 21
= ^ .

m.

The stress in the concrete is

,
= 2M = 2X 18,900 X 12 01
317 p
_
S
. ,

L (appr ° x),
/c
6W 2
12 X 0.3 X 0.9 X 21* =
- '

The diagonal tension stress is

= V = 700 X 9
= oq •
/
yi
P 12 X 0.9 X 21
28 P - S - L (appr ° x)n

Therefore, No. 5 bars can be used in the top and the bottom, as indi-
cated in Fig. 6-3(6). These will be extended the full length of the mem-
ber in order to serve as shrinkage and temperature reinforcement also.
The top bars may be lapped near the center of the span; the bottom ones,
at the column. To avoid troublesome long bars during the pouring of
the pedestals, horizontal dowels may be used at the columns even though
the extra laps on each side cause the use of more steel.
One might well ask how the two top bars of Fig. 6-3(6) can be fully
stressed by dead-load bending and still resist the tension caused by a drop
in temperature. The bending moment will cause a small compressive
stress in the lower bars. As the grade beam tries to shorten, the com-
pression will be relieved and the upper bars will be stressed more severely
in tension; then, when the strains have increased far enough, both sets of
bars will be in tension. may
be that the top bars will be stressed
It
beyond their yield point. they will stretch but will not cease their
If so,

resistance to bending action although the elongation may upset completely

one's assumptions as to the magnitude of the negative bending moment
at the column. This is one reason why the center of the beam has been
reinforced for the same bending as the ends because, as the restraining
moments at the ends are reduced, the positive bending moment at the
center will increase. The safety factor will enable the steel at the center
Art. 6-2] SPREAD FOOTINGS AND GRADE BEAMS 147

to hang on safely even though the beam is compelled to act as a simply

supported one. The use of simply supported ends avoids this trouble.
When corrugated or other flexible sidings are used, it is desirable to
design the grade beams as simply supported members. If the floor is also

cut loose from the grade beam, as in (b), the keys shown in (c) may be
sufficient to hold the beams in line. The footings may be looked upon
as strong anchors that prevent the shortening of the grade beams from
moving the footings, hence continuous construction is not recommended
except in warm climates. Even with brick walls, the simply supported
beams are preferable. This idea of intermittent anchoring is discussed
more fully in the next article.
When planning the spread footing at the column, one should consider
an adequate bearing for the grade beam. This may be a seat upon which
the concrete of simply supported beams poured after the footings are is

constructed, as shown beams may be precast. For

in Fig. 6-3 (c), or the
continuous construction, the beams and the tops of the pedestals may be
poured monolithically, as in Sketch (d). The latter method is chosen
here because of the assumed continuity.
The approximate total load upon this footing is the column load plus the
reactions of the grade beams plus the excess weight of the footing com-
pared to the original earth.

P = 280,000 + 2 X 6,300 + 4,000 = 297,000 lb. (approx)

297
Area of footing = -tt = 49.5 ft.
2

A footing 7 ft. square will be close enough.

The outside edge of the pedestal is automatically 15 in. from the center

of the column. The inside edge will be placed 12 in. from the center line
in order to have the projection obstruct the floor area as little as prac-
ticable. This will cause an unimportant eccentricity of loads of about
1 in. The top of the pedestal will also be made 27 in. wide parallel to the
wall.
The maximum projection of the footing beyond the pedestal is 2 ft.

4^ in. A trial d of 15 in. and a cover of 3 in. will be selected. The

analysis of the trial footing yields the following results:
The bending moment on the basis of an axis through AF is

CA
M = p(BE) (CA)^f = 6 X 7 X 2.38 X 1.19 = 112 ft.-kips

The area of reinforcement needed is

A =
s
—=
fs jd 20 X
112
0.9 X 1.25
_
"
4 o 8 in 2
148 FOUNDATION WALLS [Chap. 6

The stress in the concrete is

2M = 2 X 112,000 X 12
fc
kjbd 2 0.3 X 0.9 X 84 X 15 2

/, = 526 p.s.i.

The required area for bond is

= AF
—
So T for the rectangle leit of -r- 7
ujd

= 6,000 X 7 X 2.38 =
20 353 m<
210 X 0.9 X 15
The diagonal tension stress based on the pressure on area BEHG of
Sketch 0) is

V 6,000 X 6.64
ojrf 5/ X 0.9 X 15

The punching shear based on the pressure on area BEFA is

= V = 6,000 X 11
54 ° P L
VT = S
bM
- -

27 X 0.3 X 15

Therefore, the footing will be accepted with the reinforcement shown for
the pedestal and footing in Figs. 6-3 (d) and (e).

Walls to spread a series of equal concentrated loads. Some

6-3.
conditions that may make the use of grade beams undesirable and inadvis-
able are the following:
1. In a cold climate, the grade beam should extend below the frost line,
otherwise heaving may crack it. Thus it is a large deep member auto-
matically.
2. When reliable soils are near the surface, a grade beam will try to
spread the concentrated loads to the soil even though it is not supposed to
do so.

3. When the structure has concrete or masonry walls, when the soil is

plastic or ofunequal bearing value, and when cranes or other equipment

cause large live loads, it may be best to have a strong continuous founda-
tion under the wall to distribute the loads more evenly and minimize
unequal settlements. Live loads may cause a "pumping" action on a
foundation composed of separate spread footings that will tend to disin-
tegrate the walls and the substructure.
4. The column loads may be small anyway so that shallow walls are
sufficient to spread these loads, and they may be able to do so without
difficulty.
5. When the floor line is 2 to 3 ft. above the ground outside, a grade
beam becomes almost a wall by necessity, and it might as well be used as

one.
 6 .

Art. 6-3] SERIES OF EQUAL CONCENTRATED LOADS 149

6. When the position of the ground varies considerably along a building,

it may be necessary to use walls where the height is large, and advisable to
maintain the same type of construction throughout.
7. When there is a basement, some type of wall is naturally essential.

In general, unless special conditions control the design, relatively heavy

column loads make it advantageous to use spread footings and grade
beams, whereas small columns and heavy walls make it desirable to use
continuous foundation walls. Also, large bay widths (25 ft., for example)
make the first scheme preferable; short ones (like 15 ft.) favor the
second.
Because of the size and location of the columns, it is often necessary to
use a pilaster on the side of a foundation wall to receive the column in
order to avoid the use of a continuous wall of unreasonable thickness. It
may be desirable also to have a localized spread footing under this ped-
estal, as pictured in Fig. 6-4(a), in order to decrease the eccentricity of the
column load with respect to the center of gravity of the bearing area.
Assume the construction shown in Figs. 6-4(a) and (b). What is the
maximum pressure caused by the column load alone on the soil? Sketch
(c) shows the bearing area that may be assumed to support the load of a

typical column if all are loaded equally. Even though the columns are
not loaded simultaneously to the maximum extent, it is usually sufficient
to assume that they are so loaded unless the concentrated load is very
large.
Using CD as an axis, the center of gravity of the assumed bearing area
may be located as follows:

Part Area, ft. 2 Lever arm Moment, ft, 3

CDEK 20 X 2 33
.
= 46 . 1.16 54.2
FGHJ 2X4= 8. 3.33 26.6
54.6 80.8

x = 80.8/54.6 = 1.48 ft.

The moment of inertia of this bearing area about its centroidal axis is,

by the parallel axis theorem,

= 20 V 2 33 3 4 V
* 93
I in + 46-6 X 0.32 2 + + 8 X 1.85 2 = 56 ft,
4

The maximum pressure at GH caused by the column appears to be

P Mc 200,000 200,000 X (1.83 - 1.58) X 2.85
V +
.

+
.

= 7,230 p.s.f
A I :»!.<; 56

Similarly, p at CD = 100 p.s.f.

150 FOUNDATION WALLS [Chap. 6

These values are for the column load alone. The weight of the side
wall and the foundation itself will affect the total bearing pressures, but
these weights, being practically uniform per linear foot, do not cause
longitudinal bending in the founda-
4-0
Footing^ \'$'2Jo"/Z tion wall itself.
Pilaster
<tofbay * T
t of bay- At first glance, it would seem that
PS this eccentricity of load and the
large local pressure near GH will be
lO'-O"
I
201-0"
10-0" 4J
troublesome. This may be so if the
faJ-PLAN OF TYPICAL PORTION OF WALL AND FOOTING soil is weak and if the floor is sepa-
Jz2!UiP=200 k rated from the wall at A and B, Fig.
a3T cB_
6-4(6). However, in practical con-
/WV/AW/WW W
struction, the floor may be attached
to the foundation as pictured in Fig.
Pilaster
WalU 6-2, or it may be made so that it is

in contact at A and B, Fig. 6-4(6).

In either case, the top of the founda-
tion cannot tilt inward, hence the
overturning moment cannot affect
-R
m
W-SECTIONA-A
the base seriously, and the resisting
4'-0" soilpressure may be assumed to be
Id
Center of G spread equally over the bearing area.
E gravity J
F K If the lateral forces caused by the
x\
tendency of the walls at opposite
20-0"
sides of the building to tilt do not
fcJ-PLAN OF FOOTING ASSUMED FOR EACH COLUMN
counteract each other directly, the
FIG. 6-4. A foundation wall with pilasters.
frictional resistance of the floor slab
will generally be sufficient to restrain the walls. For example, from Fig.
6-4(6), the lateral force Fl needed to resist tipping is

Fi
Pe 200,000 X 0.35
= 14,700 lb.
h 4.75

where e is the eccentricity of the column load P. A 6-in. floor slab 20 ft.
wide and 20 ft. long, and a coefficient of friction of 0.5, will produce a
frictional resistance of 15,000 lb. The abutting power of the earth will
also help resist this overturning. Nevertheless, this tendency to tilt
should not be overlooked.
In general, a foundation wall like that of Fig. 6-4 acts like a deep con-
tinuous beam that spreads most of the concentrated loads longitudinally.
The weight of the superstructure's wall, or that of the first story of a cur-
tain wall, is usually distributed sufficiently uniformly on the substructure
Art. 6-3] SERIES OF EQUAL CONCENTRATED LOADS 151

so that its effect upon bending foundation wall is negligible. When

in the
the soil is firm granular material, it may
be that most of the load will be
resisted by the soil close to the column. However, if the wall is designed
to resist the bending caused by uniform distribution, any excess load
taken near the column will be on the side of safety.
Example. Assume that the wall in Fig. 6-4 is held laterally by the
floor slab, that a 12-in. brick wall on the foundation averages 2,400 lb.
per lin. ft., and that the trial dimensions are as shown in the illustrations.
What are the added and the total pressures on the soil, and what rein-
forcement is needed in the substructure? Use / = 18,000 p.s.i. s

The unit pressures for bearing may be found as follows, considering that
tilting is prevented:

Part Weight, lb.

Added Total
Brick wall 20 X 2,400 = 48,000 48,000
Concrete X 1 X 20
4 = 80
2.33 X X 201 = 47
1.33 X 2 X 4 = 11
2 X 4 X 1 = _8
146 X 50 = 7,300
146 X 150 = 22,000
Earth above footing (approx) 10,000
Column 200,000 200,000
Total 255,300 280,000
Added pressure = 255,300/54.6 = 4,700 p.s.f.

Total pressure (neglecting the floor slab and live load)

= 280,000/54.6 = 5,100 p.s.f.

The wall will be assumed to have the uniform cross section shown in
Fig. 6-5 (a), neglecting the local effect of the pilaster. The upward pres-
sure per linear foot of wall for design purposes is

w =
200,000 X
„
2.33
= Q
___ „
8,600 lb.
r .

54.6

This assumes that the wall is so stiff that it will spread the load uniformly
over the bearing area. It is also assumed that the pressure under the
area FGHJ of Fig. 6-4 (c) is transmitted directly through the pilaster and
does not affect the bending moment in the wall appreciably.
If the wall is considered to be a continuous beam with all spans loaded
equally, and with the column load spread uniformly over the width of
2-ft.
152 FOUNDATION WALLS [Chap. 6

the pilaster, it is sufficient to assume that

M J = _ l
column load X X
column width =
8,600 X 20 2
at <2 4 of
12 12
100,000 X 0.5
= 237,000 ft.-lb.

M at K = V^L = 14.3,500 ft.-lb.

Assuming that the friction caused by pressure, the chemical bond, and
the mechanical bond at FG, Fig. 6-5 (a), are sufficient to make the footing

200k , „ -
20Qk , „ 20O*
2Q 2o o

20 -\ T 1. „.
/'O' *\ ^2-0 '
Wa//X ^_±-?-0
A vS K J
1^ *-8.6 k/'

f/W-PARTIAL ELEVATION OF WALL

"I

7T
dA J-fi

(c)SHLAR DIAGRAM

Jj EF G H Parabola

&!
5>
VJ
c D
2-4" ~%
faJ-TYPICAL CROSS SECTION

Crfj-BENDING-MOMENT DIAGRAM

FIG. 6-5. Foundation wall supporting equal column loads.

and the main wall act as a unit, and that the bars are to be placed as
shown in the sketch,
237,000
A s at ./ = = 3.1 in. 2
18,000 X 0.9 X 4.75
143,500
A s at K = = 1.9 in. 2
18,000 X 0.9 X 4.62

2o near ,/ = 8,600 X 9
= 7i.z
9. in.
210 X 0.9 X 57

Therefore, four No. 8 bars in the bottom and three in the top appear
to be sufficient. The maximum/,, will be at the top near J . A sufficient
Art. 6-3] SERIES OF EQUAL CONCENTRATED LOADS 153

estimate of its magnitude is

2 X 237,000 X 12
U 12 X 0.3 X 0.9 X 57 2
p,S L '

which is entirely safe.

But what should be done about reinforcement for shrinkage and temper-
ature? If some of the bars are extended continuously through the top
and bottom of the wall, they will prevent automatically any serious local-
ized cracks due to shrinkage, even though they cannot stop that shrinkage.
A large drop in temperature, however, may be a different matter. If the
building is 200 ft. long, it is obvious that, with friction on the footing,

bearing of earth against the sides of the pilasters and their footings, and
any resistance caused by the floor slab, the ends cannot be drawn in
toward the middle of the wall as a neutral point when the wall shortens.
Therefore, the deformation will probably be taken up by a multitude of
hair cracks. These effects will tend to add tension to that in the tensile
reinforcement and relieve or eliminate compressive stresses in the concrete
at and near the cracks.
The wall might be cut into units 40 to 100 ft. long by means of con-
traction joints as explained in Art. 6-8. The length of a unit will depend
upon the importance of the case and the expected severity of temperature
changes. In addition, a few more bars than required for beam action
alone should be used so that there will be some reserve strength to resist
the tendency to crack. The longer the unit, the more need for these
extra bars. The reinforcement should be arranged so that part of the
steel of the top and the bottom extends the full length of the unit.
In this case, the 200-ft. length of the building is not enough to compel
the use of a central contraction joint in the superstructure. If such a
joint is placed in the foundation under the 12-in. brick wall, the movement
of the former will tend to crack the wall in the same vicinity. Further-
more, this wall is not severely exposed. It seems desirable, therefore, to
knit the entire substructure together as a unit.
The reinforcement may be proportioned to resist bending alone, then
0.002 times the cross section of the wall may be added as extra steel to
make allowance for thermal effects. In this case, the added amount will
be 6.33 X 144 X 0.002 = 1.82 in. 2 Three extra No. 8 bars will be added.
Therefore, bars a and b will be extended, with laps, for the full length of
the wall; bar d will be added and used similarly. be added to Bar c will
increase the top reinforcement locally, and bar e will be used similarly
near the column. Will this arrangement be satisfactory? No one can
be sure, but it should prevent any harmful cracking. A suggested mini-
mum area of steel for such conditions is 0.004 to 0.005 times the normal
cross section of the foundation wall. If there are no pilasters or other
154 FOUNDATION WALLS [Chap. 6

projections, the usual 0.002 to 0.003 times the area of concrete should be
sufficient. In Fig. 6-6(a) is shown the reinforcement for a typical "inte-
rior" bay of this wall. The footing of the wall is so deep relatively that
little transverse reinforcement is needed in it. However, the footing
under the pilaster is designed like part of an isolated footing, and the
steel to be used in it is shown in Fig. 6-6(6).
The question of the amount of reinforcement to use for the prevention
of cracks caused by a drop in temperature is one of those intangibles that

£ of span <£ of col. <£ ofspan

#4 4'c.c
r
\ I a. andf
.cane

IUa-_t ii n , d \

~±=&=-.
Longitudinal bars*6 j
eandg
7- -*4 6'c.c
^4-*4 6"c.c,

f«J-BARS IN MAIN WALL (b)- PLAN OF BARS IN

FOOTING AT COLUMN

FIG. 6-6. Arrangement of reinforcemen' in a foundation wall.

requires the exercise of good engineering judgment. To guide the reader,

the following suggestions are summarized:
1. Determine the steel required to enable the substructure to support
the assumed loads.
2. Estimate the probable range of temperature change in the particu-
lar locality. A wide daily change is more serious than a merely seasonal
one.
3. Investigate the degree of exposure. Parts that are exposed to direct
sunlight during the day and to cold air during the night are more endan-
gered than are those that remain in the shade.
4. Cold climates with freezing temperatures are more dangerous than
are mild climates.
5. If a small portion of the top of a wall is subjected to the effects of
the atmosphere, whereas the bulk of the interior of the wall is exposed

to rather uniform temperatures — as

heated basement the conduc- in a —
tivity of the concrete will probably prevent serious chilling of the exposed
portion. Light reinforcement will then be satisfactory.
6. If the wall is in long sections that are anchored considerably by

projections and offsets, the structure should be reinforced strongly.

7. The reinforcement should be spread over the cross section instead of

being concentrated at one spot. The top and bottom of the wall are the
most important locations for continuous reinforcement.
Art. 6-4} SINGLE CONCENTRATED LOAD 155

8. An important structure deserves more conservative design than does

an unimportant one. Walls that are exposed to view advertise their

cracks to the owner and the public.
6-4. Walls to spread a single concentrated load. When a continu-
ous foundation wall is used in an industrial building equipped with bridge
cranes, the crane load may cause a heavy concentration at any one of the
columns. The wall will spread the load, somewhat as is done by a stiffen-
ing truss in a suspension bridge. If the soil is firm, the wall may not have

to spread the load very far before the bearing power of the soil supports it.
If the soil is weak and plastic, the wall may spread the load over con-
siderable length, and this is probably the reason why an engineer would
use such a type of construction. The stiffness of the wall naturally affects
the probable longitudinal distribution of a load.
Exact analysis of the distribution of a local load by a foundation wall is
probably impossible because of the many uncertainties involved, espe-
cially the unknown resistance and deformation of the soil under tempo-
rary loads. However, the following assumptions and approximations are
given for the guidance of the reader, and the sketches referred to are in
Fig. 6-7:
1. Assume that the temporary concentrated load P is spread uniformly

over a length of wall equal to the corresponding dimension of the pilaster

if used or to the width of the base plate of the column.
2. Unless the height of the wall h is less than one-fifth of the length of

the bays L, assume that 75 per cent of the load P is distributed as a uni-
formly varying pressure over the two adjoining bays, as illustrated in
Sketch (6), regardless of the fact that the wall may be continuous under
a series of columns. This allows a 25 per cent reduction for the effect
of any projecting local footings at the column points, for the fact that
such a stiff wall will probably spread some of the load beyond the two
adjacent bays, and for the fact that the soil near a temporary load will
not have time to have the moisture squeezed out of it so as to cause
appreciable consolidation.
3. Assume that the wall is a continuous member fixed at the adjacent
columns B and D and with a constant cross section.
4. Assume that the magnitudes of the shears are as given in Sketch
(c). Because the maximum shear at F and G is critical and because it

may occur at any column, shears elsewhere need not be computed. One
might ask what resists the upward reactions of the soil pressures at B
and D. At least at one of the adjoining columns there will be a reaction
caused by the two or more wheels of the crane trucks. The strength of
the next portion of the continuous wall will also resist upward forces.
5. Assume that the magnitudes of the bending moments are as shown
in Sketch (d). The ordinates near K and J are the critical ones for ten-
156 FOUNDATION WALLS [Chap. 6

sion in the bottom of the wall ; those at M, for tension in the top near the
center of a bay. For purposes of plotting the shear and bending-moment
diagrams, the ordinates in Sketches (c) and (d) may be used.
6. Add the shear and bending moment atany given point algebraically
to those computed for the same point for any other conditions of loading
that do or may exist simultaneously, and be sure that the structure can
resist the combination safely.
Be conservative in the choice of the allowable unit stresses for bend-
7.

ing and shear because of the probable repetition and reversal of stresses
caused by the movement of the load back and forth in the structure. If

4 ofco/. atC.
4 of cot. atB - ^.Approx. slope
Co/ JP Continuous wa/i r^i width of base of tangent
\c / LP \E
f or pi/aster =45°
-0.025P'
YJfi a/JI ^
-0.I2P'
~ {

Magnitudes syrn.
faJ-RARTIAL ELEVATION OF WALL

Width of base
5 'about <£ of this co/.

or pi/aster x fd-ASSUMED SHEAR DIAGRAM

,A ,B
^'width of base orpi/aster- 4 of co/. a/C
'Sym. about 4

(W-ASSUMED PRESSURE DIAGRAM

\K, Approx. siope

JST of tangent =30
(d)- ASSUMED BENDING-MOMENT
DIAGRAM AND COEFFICIENTS OF P'L

FIG. 6-7. Assumed shears and bending moments caused by a concentrated load.

vl exceeds 0.02/', some web reinforcement is desirable. One should

remember that these suggested values for shears and bending moments
in acontinuous foundation wall are for the purpose of giving the designer
an easy way to obtain a reasonable estimate of the scale of their magni-
tudes rather than to have him guess their magnitudes or neglect them
entirely.
At a corner of a building with continuous foundation walls, the crane
load on the corner column will generally be less than those on the inter-
mediate columns. The right-angular corner prevents the continuous
beam action that has been assumed elsewhere, but the footing of the end
up the column. In this case, it is suffi-
wall will help the side wall hold
assume that the maximum shear in both walls near the corner
cient to
column because of the temporary crane load P is 0.25P. The bending
moment causing tension in the bottom of the walls under the corner
Art. 6-4] SINGLE CONCENTRATED LOAD 157

column may be and the bending moment near the

called equal to zero,
center of the bays caused by P
and producing tension in the top of the
walls may be estimated as 0.04PL. The second column from the corner
may be assumed to cause the same shear and bending as any other inter-
mediate column.
If a continuous side wall at a corner has no intersecting end foundation

wall, it is advisable to use a large footing or extension of the wall at

the corner so that the substructure at this point may be designed as
though the column were supported practically upon an isolated or elon-
gated footing. The column next to the corner one may be assumed to
be a typical intermediate column.
A column near a contraction joint that has no adequate keys for trans-
mitting shear may be supported as suggested in the preceding paragraph.
If the joint has proper keys for resisting shearing forces that act up at one
time and down
at another, the two bays of the substructure adjoining the
joint may
be designed for the same shears and bending moments as those
suggested for the end bays at a corner having continuous side and end
walls.
Special or additional reinforcement in a wall for the prevention of local
large cracks caused by a drop in temperature was discussed in the pre-
ceding article. Although moving crane loads are temporary, they should
be considered when one plans the steel to be added to resist this cracking.
The suggested increase of 0.002 times the cross-sectional area of the wall
should be made over the requirements of the combined bending moments
caused by all loads.
Example. Assume that the wall pictured in Fig. 6-4 supports columns
having a dead load and uniformly distributed live load of 120 kips each,
and a maximum crane live load of 80 kips at any one column at a time.
Design a typical portion of this wall for these conditions.
The shears and bending moments shown in Figs. 6-5 (c) and (d) may be
multiplied by 12 /200 = 0-6 and later combined with those assumed to
be caused by the 80-kip load. The resultant diagrams for the 120-kip
load are shown in Figs. 6-8(c) and (/).
The assumed uniformly varying pressure diagram for two bays is pic-
tured in Fig. 6-8(a), with 75 per cent of the 80-kip load assumed to be
spread by the wall in the adjacent two bays. The ordinate b in the shear
diagram of Sketch (6) is obtained from Fig. 6-7 (c), and the rest of the
diagram is plotted for purposes of illustration. The maximum shear is
combined with c of Fig. 6-8 (c), and the combined shear diagram is pic-
tured in (d). This is really the addition of ordinates at two different
points, but great accuracy or refinement is not necessary because of the
empirical nature of the assumptions.
Similarly, using Fig. 6-7 (d), the diagram in Fig. 6-8 (e) may be plotted,
158 FOUNDATION WALLS [Chap. 6

with the assumed net load of 60 kips. Its ordinates combined with those
of the bending-moment diagram in (/) give the values shown in (g). The
wall should be made strong enough to resist safely these approximated
shears and bending moments, with additional steel provided to enable the
structure safely to withstand temperature and shrinkage effects. The
final design is shown in Fig. 6-9. The amount of reinforcement used is
conservative because of the alternation and repetition of stress conditions

L=20 lO' L=20 1O"

p =30 k/ /in. ft.
+2-0"

a -
40 -
J
faj-ASSUMED LOADING DIAGRAM

b 2l k

C6J-SHEAR DIAGRAM FOR 80-KIPLOAD (eJ-BENDING- MOMENT DIAGRAM FOR30 k

Causes tension
k
h 86[ / in top
4SH
|P^p
fcJ-SHEAR DIAGRAM FOR 120-KIP LOADS (fj- BENDING-MOMENT DIAGRAM FOR 120 k

67kA d

(W-TOTAL SHEAR DIAGRAM (g) -TOTAL BENDING-MOMENT DIAGRAM

FIG. 6-8. Shear and bending-moment diagrams for a foundation wall.

caused by the moving load. It can be argued that the infrequent occur-
rence of the maximum live loads with the most severe temperature condi-
tion justifies a high allowable unit stress in the reinforcement. The
author does not try to refute this argument, but the possible saving in
steel is small and it is not generally worth hazarding the safety and satis-
factoriness of the structure to secure it.

6-5. Basement walls. When there is a basement in a building, the

reinforced basement wall, if made of poured concrete, is generally so strong

and stiff that, without difficulty, it will spread the column loads that rest
upon it, provided the footing and the bearing power of the soil are ade-
quate. Seldom need these walls be analyzed to see that they are safe
as longitudinal beams if the longitudinal reinforcement recommended for
the prevention of cracking (0.002.4) has been used. The magnitudes of
the shearing and bending stresses are very uncertain when members are
Art. 6-5} BASEMENT WALLS 159

so deep compared to their spans — spans between columns equal to two

or three times the depth of the wall. Probably such walls cannot curve
(deflect) so easily as assumed in the common theory of flexure, and shear-
ing resistance may be unusually effective. However, there are other
matters to consider.
Light columns may be supported directly upon the top of a foundation
wall. This is most likely when the superstructure is made of reinforced
concrete with monolithic or curtain walls between the columns. It may

also occur when the foundation walls are made very thick (perhaps 24 in.)
to obtain mass, rigidity, or great lateral and longitudinal strength. In
most cases, the walls of the basement need not be over 10 to 15 in. thick,
and then the column bases are likely to project inside the plane of the wall,
as illustrated in Fig. 6-10(a). It is usually a needless waste of materials

,a \JkofcoL ,,

M\**4Ug. p--r-——xr— 1—«f~-i ±=— 1=.= = =±

bT
*
f |
^-b | |
1 |

!
4'c.c. yg
—r-^-1—+—
\ff y& Y9
i
Vg ^A
1-^--4

tt t+ \-£ hnv
nf bay
V-tt of
\
/ v.. J.'
Longitudinal / L
bars—#
#7
1
edde
ta;-CR0SS SECTION C/W-SIDE ELEVATION

FIG. 6-9. Arrangement of reinforcememt in a continuous wall.

and space if the wall is thickened enough to bring its inner face to some

such position as shown by line AB. It is not common practice to step

back the outer face C to some such position as shown by the dotted lines
near D, although this might be done.
One arrangement for the support of such a column base is shown in Fig.
6-10(6). This utilizes a pilaster at the column. The illustration also
pictures a heavy floor slab that is seated on the wall, an arrangement that
might occur in the case of heavy flat-slab construction for the first floor.
If there are large reinforced-concrete beams under the columns, the top of
the basement wall, the top of the pilaster, and the first floor may be poured
monolithically, as shown in (c) for a case where the beam is so strong that
the pilaster may be omitted. If the floor is poured after the wall and
pilaster are completed, the pilaster should provide a seat for the beam.
Figure 6-10(cT) pictures a case in which encased steel beams are used to
support the first floor. It is customary and advisable to construct the
basement wall and pilasters first, then to erect all of the steelwork, and
finally to build the floors. It is therefore desirable to have seats, local
pockets, or shelves in the concrete construction upon which to support
the steel members.
160 FOUNDATION WALLS [Chap. 6

In Fig. 6-10(e), the column base is below the structural steel framing
of the first floor in order to eliminate projections at or just above the floor.
This makes a neat appearance but, if the floor is likely to be wet because
of operations orfrom washing, the steel is likely to rust at its junction
with the concrete floor. Rust plates like (E) may be used to protect the
main steelwork, the concrete may be sloped upward locally about Y2 in.
against the steel in the hope that it will shed water, or a V-shaped groove

l L 0'
H f/2"WFco/.
Bricks-
£
l fF/o or

yAnchor bolts
\

J
I
'-qua Inside ofwall
S
Beam

\B
faJ-GENERAL ARRANGEMENT (b)-mil AND PILASTER (W-WALL AND BEAM

Direction of
s rust pi E rotation

Thrust

* Small beam
Encased G ^ Steel beam
Corbel under col.
beam Pilaster
Pilaster Wall
Seat

<a!J-$TEEL ON PILASTER (e) -COLUMN BASE DEPRESSED W-WALL AND CORBEL

FIG. 6-10. Some details of columns on basement walls.

may be made at the junction of steel and concrete so that the joint may
be sealed with mastic. When the columns are depressed as indicated
here, it is best to have a construction joint in the wall at FG, then to pour
the top of the wall monolithically with the floor rather than to leave
pockets in the wall for the columns alone. The latter arrangement causes
the joints to show prominently in the finished work.
If the appearance and the need for usable space make it desirable to
eliminate pilasters in the basement and if there is not a strong beam at
the column as indicated in Fig. 6- 10(c), a stepped or sloping corbel may
be used as pictured in (/). This assumes that the floor will steady the
Art. 6-5} BASEMENT WALLS 161

wall laterally so that the wall need not act as an eccentrically loaded
vertical cantilever.
There may be cases in light construction where it is desirable to sup-
port steel or wooden platforms and floors on the side of a concrete wall.
Figure 6-11 shows some details that may be advantageous in such a situ-
ation in order to avoid costly projecting shelves built of concrete, deep
recesses that weaken the wall, and local pockets in which it may be diffi-
cult to erect the steelwork. One should remember that anchor bolts
embedded in concrete, as in (6) and (/), are no stronger in resisting shear
than the bearing value of the concrete against the outer 1 or 2 in. of the
embedded bolt. The embedded structural pieces shown enable the bolts
to bear against steel and the concrete against a part having considerable

IL

(b) Cf)
(Weak) (Weak)

FIG. 6-11. Connections for steel framing to concrete walls.

bearing area. These structural parts should be so anchored that they

will not rip out because of eccentric loading and tensile forces.
The lateral pressure of the soil outside of a high basement wall may
cause considerable bending in the latter as a vertical slab. The wall will
be supported at its top and bottom ordinarily, as pictured in Fig. 6-12(a).
The wall should ordinarily be designed as a simply supported member
with a loading diagram similar to that shown in this sketch (or with sur-
charge added). The bending-moment diagram may be assumed to be as
pictured in Sketch (b). The vertical load will relieve a little of the com-
puted tension in the vertical reinforcement, but this may be neglected
because its effect is small compared to the uncertainties in the assumption
of loads. End restraint of basement walls may be provided in some cases,
but its assumed effect upon the economy of reinforcement is likely to be
deceptive. When a large part of the height of a basement wall is above
ground, the lateral bending moment will be small and possibly negligible,
or it may be assumed to be as shown in (c). When the basement is two
stories deep, it is probably desirable to design the wall as a two-span
continuous vertical slab with a uniformly varying lateral load.
162 FOUNDATION WALLS [Chap. 6

If water pressure is be applied against the basement wall in

likely to
addition to the soil pressure, the intensity of the total pressure is that of
the full hydrostatic pressure of the water plus the pressure due to the
buoyant weight of the earth. The combination of the two materials act-

ing together may be assumed to be equal to that of a fluid with a unit

weight of 80 to 85 p.c.f. It is safe enough to use the pressure diagram
shown in Fig. 6-12(rf). If the watertable is very low, the hydrostatic

celt
6L

Max Mas
simply supported
beam
*0.3Lto0.4L

Bending moment

faROADING toJ-BENDING-MOMENT ("c; -PARTIAL LOADING

DIAGRAM DIAGRAM

\&-Horiz.
beam

*
Assumed footing
(d J-LOAD FROM (e)-\NALL AT FLOOR (f)- RETAINING WALL
SATURATED EARTH OPENING

FIG. 6-12. Pressure diagrams and bending moments assumed to be applicable to a 1 -ft. vertical slice

of a basement wall.

pressure may
be neglected. The effectiveness of drainage is a matter to
be estimated by the engineer for each special case.
When part of a basement wall is opposite an opening in the first floor,
the top of the wall may be reinforced as a beam to span horizontally across
the opening. If the open space and the loads are too great, the top may

be enlarged as shown in Fig. 6-12(e), or the wall may be designed as an

L-shaped retaining wall somewhat as indicated in (/). In the latter case,
the vertical load of the superstructure should be included when the sta-
bility of the assumed effective unit is tested and when the intensity of
pressure upon the ground is estimated.
Art. 6-6} WATERPROOFING BASEMENTS 163

Pipes and ducts for utilities generally must pass through basement walls
and probably through other foundation walls, too. It is frequently desir-
able to provide large rectangular holes through the walls when they are
poured so that the piping can be installed later, thus avoiding delay and
minimizing errors of location of pipes. The spaces around pipes can then
be packed with concrete or mortar after all connections are made.
The details to be used at the bottoms of basement walls may be affected
by special local conditions. Figure 6-13 shows some construction that
may be suitable for many cases. Heavy basement floors to resist uplift
are discussed in Chap. 7.
6-6. Waterproofing basements. Admixtures in the concrete of
basement walls may be beneficial in preventing leakage if they increase
the density or reduce the porosity of the concrete. However, they do
not stop leakage at construction joints and cracks. It is obvious that
the joints in a basement wall built of precast blocks are likely to be planes
where leakage will occur easily. Coatings applied to the inner side of a
basement wall that is found to leak may be helpful, but this remedy is a
last resort and of dubious efficacy. It is best to prevent leakage by proper
planning in the original design.
Drainage lines may
be placed outside of basement walls as indicated in
Figs. 6- 13(c) and To be effective, these pipe lines should have
6- 15 (a).
open joints or perforations that will let in the water but will not become
clogged, and will not permit the pipe line to be closed by silt. The pipes
should be of vitrified clay, cast iron, or other noncorrodible material.
Uncalked bell-and-spigot joints, wrapping of joints with tar paper, embed-
ment of the pipes in coarse screened gravel or crushed stone, and the
insertion of burlap, tar-paper sheets, salt hay, and similar materials
between the stone and the earth backfill are all effective means for secur-
ing a minimum Furthermore, the drainage lines should dis-
of silting.
charge at an outlet which will let the water flow away, which will not
become submerged, and which will not be closed by freezing. It is helpful
if the lines can be cleaned or flushed out by having suitable risers and

Y-connections.
Incidentally, drains placed in and under a basement floor with inlets
to prevent ponding should have closed joints, and the operation of their
discharge facilities should be reliable. Too frequently, such installations
have been made without regard to these matters and to the elevation of
the ground water. The result has been the flooding of the basement and
the realization that water may flow either way in a pipe.
A bituminous emulsion or mastic coating without a membrane may be
used upon the outer surface of a wall as shown in Figs. 6- 13(c) and (d).
This means that space must be available for its application. Such a coat-
ing is very beneficial, but it cannot be depended upon to resist leakage
 —

164 FOUNDATION WALLS [Chap. 6

permanently when subjected to hydrostatic pressure that acts for at least

a day or two at a time. It is desirable to have such coatings built up by
two successive moppings.
When it is important to prevent the penetration of dampness and
actual leakage of water through a basement wall, the basement should

Waterproofing
without
membrane \P

<Wall

Concrete
Dirt floor -*
{ floor _ j_f
-HTV
Friction' Earth is weak Soit may be
in bearinq if displaced here
baa is if I3 is large
1

lateral

" ^-FOOTING OMITTED

'"sfmall^^
^-FOOTING AND CONCRETE FLOOR
fa)-F00TINGAND DIRT FLOOR

W.P. without~~ F
membrane -Pilaster
(Floor and combined
Ifooling poured first

— '
> o.° «,i,:7*—
y/ixwt>//*i»'
{^Footing at
VWVll-W/WIM/fii 1
pilaster
frfJ-COMBINED FOOTING AND FLOOR feJ-PILASTER AND FOOTING

Caulking compound
.Weill Mortar
Wall
Concrete plug in / ,Reinf.
Pi/aster
rectangular hole
W.P-

J.
Floor an Pipe 2nd pour^g^^r1st pour
m Footing ZK-
'Rough joint fhhOUE METHOD OF
ftv//ty//V?>*s»>vwr/,
SEALING A
(f)- STRAIGHT FOOTING (g)- DETAILS OF PIPE CONSTRUCTION
BETWEEN PILASTERS JOINT AGAINST
LEAKAGE

FIG. 6-13. Some details at bottom of basement wal

be designed and built so that the water can be stopped at the outside
surface of the structure and so that the basement walls and floor can
resist any resultant hydrostatic pressure. One of the most effective
means of waterproofing a basement is the use of a continuous bituminous
membrane forming a sort of waterproof pan that fits under and around
the basement. In this membrane, tar or asphalt is generally used to
secure the waterproofing properties, whereas fabric is included to provide
toughness and reinforcement.
Art. 6-61 WATERPROOFING BASEMENTS 165

The general nature of a membrane waterproofing is illustrated in Fig.

6-14. Much of the quality of the product depends upon the skill with
which the surface of the structure smoothed and dried, the hot mastic
is

is applied, the plies of fabric are smoothed down and overlapped, the junc-
tions are made at horizontal and vertical corners, and the finished prod-
uct is protected. On the other hand, much should be done in the plan-
ning of a structure to make it possible and easy for the contractor to
secure good results.

:»;.'(;.»:;« : o; :.-0:.{D> .:".

';.':jf

FIG. 6-14. Detail for substructure waterproofing — five-ply construction. (Courtesy of the Barrett
Division, Allied Chemical Corporation.)

Some suggestions are given below, and the sketches referred to are in
Fig. 6-15:
1. When conditions are not sufficiently serious to cause waterproofing
of the basement floor, the membrane used on the side wall should seal
the construction joint at the top of the footing, as indicated at A in
Sketch (a).

2. Adequate clearance should be provided between the concrete wall

and the side of the excavation, as shown in (a). Otherwise, the work-
manship may be poor, as one can learn for himself if he tries to apply hot
asphalt and fabric in a restricted space.
3. If the membrane is not protected, backfilling and other operations

may damage it. Old bricks laid on edge in lean mortar as pictured in
(a) are suitable for this, and so are 2-in. precast concrete blocks. A
poured-concrete covering 2 in. to 4 in. thick is even better, but it is more
costly. Sheets of plywood and pieces of old lumber may offer temporary
protection for the membrane, but their use in important construction
should be questioned.
166 FOUNDATION WALLS [Chap. 6

Backfill
Brick and mortar
protection
Membrane
waterproofing
Basement wall

~W.R \HV.P.
\-2"to4"
Drain
Ca^-DRAINAGE AND WATERPROOFING fi&i-INADVISABLECeAWATERPROOFING CaW-PROTECTION
OF SIDE WALL CONSTRUCTION RECESSED TIED TO WALL

Wail
Ifefc Protection
jtM^A WR -

Bottom W.R
^Concrete and mesh
(e>-CHAMFER AT W-FILLET AT VERTICAL r^J- WATERPROOF ON (h)- JUNCTION WITH
VERTICAL CORNER REENTRANT CORNER BOTTOM AND SIDES CORNER FLASHING

Protection

fi fes Waterproofing
'
fV-Net line
Protection or
"sand wall"
f/J-PROJECTING FOOTING CJh WATERPROOFING (jy-WATERPROOFING BOTTOM CORNER
WATERPROOFED SIDE WALL IN ROCK IN ROCK EXCAVATION
EXCAVATION

(7J-DRAINAGE OF Contemporary C«;-JUNCTION BETWEEN

WATER-BEARING sump under INSIDE AND OUTSIDE
SEAMS IN ROCK basement floor APPLICATIONS

FIG. 6-15. Some details of construction for membrane waterproofing of basements and other under-
ground structures.

4. The top simply brought

of such a protective coating should not be
to the top of the waterproofing and then
because thisleft there, as in (b),

may fail to help prevent the eventual penetration of water behind the
top of the membrane. Some details for better construction are shown in
Sketches (c) and (d).
Art. 6-6] WATERPROOFING BASEMENTS 167

5. Projecting corners of the substructure should be chamferred as

shown in (e) so that the membrane can be built without the danger of
tearing and serious wrinkling of fabric.
6. Reentrant vertical corners should be widely filleted as pictured in (/)
so as to avoid air pockets, poor overlaps of fabric, and harmful
wrinkles.
7. When the basement floor is to be waterproofed, the membrane under
the structural slab should be laid upon a thin concrete base, as pictured in
(g). The membrane on the may
be lapped over the horizontal layer
wall
later on, as indicated at B. this junction may not be effective
However,
in resisting water under pressure. It is likely to be safer if soldered copper
flashing is lapped into the bottom membrane as shown diagrammatically
in (h) then the flashing may be bent up, folded (or possibly soldered) at
;

the corners of the building, and sealed into extra plies used at the bottom
of the vertical membrane. It is apparent that heavy pressures applied
by the structure may cause the mastic to squeeze out from under the
walls. Therefore, wide copper sheets soldered at the joints may be used
2
as flashing, but the copper should be thin (8 oz. per ft. ), otherwise it

may be too stiff to lie flat and in complete contact with the bottom. On
the other hand, a footing may be used as in (i) to reduce the unit pressure
on the membrane. This trouble may also apply in the case of interior
columns that carry very heavy loads. Another feature to notice in Sketch
(g) is the 2-in. protection of concrete placed over the waterproofing so
that the erection of reinforcement, the pouring of concrete, and other
operations will not damage the membrane.
8. When the basement
is in rock excavation, a concrete "sand wall"

should be poured against the rock, as indicated in Sketch (j), so that the
membrane waterproofing can be applied to a smooth dry even surface
before the main structure is built. This minimizes the amount of rock
excavation by the elimination of outside working space. Of course, the
rock will be irregular, and the concrete should not be expected to stick if
it is merely plastered upon the rock; at least 3 or 4 in. should be provided

between the waterproofing and the net line the line within which the—
—
rock must not project in order to make sure that the sand wall will be
self-supporting. It is often necessary to build the sand wall in successive
layers 5 or 6 ft. high because of the cost of forms and the difficulty of
placing concrete in such thin sections.
When the bottom as well as the sides of a basement in rock excava-
9.

tionmust be waterproofed, the construction may be made as shown in

Sketch (k). The bottom membrane is extended up the sides at F after
the rock is lined with concrete. The protecting concrete cover G is placed,
and then the waterproofing is applied on the side walls. If the floor or
mat H is poured before the membrane is applied on the sides, the bars
168 FOUNDATION WALLS [Chap. 6

J projecting up to bond with the wall reinforcement K to reinforce the

corner will seriously interfere with the construction of the waterproofing.
10. When water-bearing seams are in the rock, hydrostatic pressure is
likely to push off the sand wall. One way to prevent this is to install
split-tile or sheet-iron drains over the seams, as indicated in Sketch (I).

These drains may then be connected to a temporary sump and pumps, as

indicated in (m) ; later the sump may be
In such a case, it is
sealed.
probable that the sand wall should not be constructed more than a few
feet high before the membrane and the pressure-resisting main structure
are built inside it, and within 3 or 4 ft. of its temporary top.
11. Generally there will be a position near the top of the rock excava-
tion at which it is possible and desirable to change from inside water-
proofing to normal outside operations, as shown in Sketch (n). Copper
flashing in the lower membrane may be bent outward at L, then it may
be bent up again and joined into the upper membrane. This work is
difficult, and its results may not be reliable. It may be advisable, in an
important case, to extend the sand walls upward as a 6-in. reinforced-
concrete box, then to apply the waterproofing on the inside only.
12. Pasting waterproofing overhead on ceilings and surfaces that slope
forward (as when undercut) is almost impossible and should be avoided.
It is often advisable to increase an excavation so that shelves and slopes
can be waterproofed from the top and outside. Of course, the top of a
conveyor tunnel or compartment in the bottom of an open excavation
can have its waterproofing joined to the side-wall membranes by simply
lapping it over the latter, or making the junction by means of the prin-
ciples illustrated in (n). The membrane on the top of the structure
should be protected by a layer of concrete. It is preferable to have mesh
reinforcement in this layer, and to extend the top protection over that
of the sides.
A membrane should be composed of at least two plies of fabric if it is

to be reliably waterproof. Three or four plies may be desirable when the

hydrostatic pressures are large and when some ductility of the membrane
is needed because of possible strains and small motions of the structure.

A four-ply membrane will be approximately in. thick. %

When piping must pass through a membrane waterproofing, special
details are needed at the junction. Pasting of the membrane around the
pipe may be sufficient when there is little hydrostatic pressure. In serious
cases, welded plate flanges may be used to lap under or into the membrane
as indicated in Fig. 6-15(w). The rod welded to the pipe in this illustra-
tion is to prevent rotation of the pipe when connections are made.
Bricks laid in and coated with mastic have been used as waterproofing.
Such construction is difficult to build, the numerous joints add to the
uncertainty of its tightness, and settlement of the bricks because of the
Art. 6-7] MISCELLANEOUS DETAILS 169

plastic material in the joints may rupture the waterproofing near

its top.

When dampness occurs in a basement because of condensation (sweat-

ing), membrane waterproofing will not stop this action.
Proper ventila-
tion may remove the trouble except when warm, moist
is brought in air
contact with cool walls, whereupon the trouble may be aggravated.
Slight heating of the basement will cause the desired evaporation in most
cases.
6-7. Miscellaneous details. There are various details of long foun-
dation walls that should be planned in accordance with the requirements
of practicable engineering and construction. One of these is the location
and type be used in a wall.
of joints to
The quantity of concrete that can be poured in a limited time or as one
continuous operation may be inadequate for the construction of the wall
as a complete unit. Economy and reuse of forms make it desirable to
build the wall in relatively short lengths, or in short lifts of 6 to 10 ft.

The permissible construction joints should be located and shown on the

plans by the engineer instead of leaving such matters to be determined
by anyone who happens to think of them.
Horizontal construction joints may not be troublesome unless they are
located at points where large shearing forces exist. Even then the friction
caused by the vertical pressure, and the bond of the concrete to the pre-
vious pour, may be sufficient without the use of special keyways. In
Figs. 16-13 and 16-15 are shown some arrangements for construction
joints between footings, walls, and floors; and they indicate that a base-
ment may be used as a strut between opposite walls. However, con-
floor
struction joints may be influential in their effects upon the planning of the
reinforcement.
It is desirable to avoid the necessity of .supporting long vertical bars
while a little of their lower ends
is covered with concrete; they are also

likely to interfere with further formwork. However, a bar may ordinarily

be used through two pours if the projection above the top of the lower
pour is not over 150 to 200 times the diameter of the bar; greater lengths
may need lateral support. When vertical bars are spliced, it is convenient
to arrange the reinforcement as suggested in Fig. 6- 16 (a). This joint is
assumed to be at a point where the wall is not laterally supported. At
points A and B the edges of the first pour may be troweled for a depth of
1 in. when a neat straight joint is desired, or a V-cut may be used at these

points to separate the pours so that any difference in color and texture
of concrete will not be too apparent. Figure 6-16(6) shows an arrange-
ment for construction joints that may be used when a light intermediate
floor or platform is to be inside the wall. Sketch (c) pictures construction
that may apply when the intermediate floor is heavy.
 4 3 '

170 FOUNDATION WALLS [Chap. 6

Vertical construction joints are likely to endanger the strength of a

foundation wall. It is obvious that such a joint located as shown in Fig.
6-17(a) constitutes a plane of weakness that endangers the spreading of
the load from the adjacent column because the reinforcement should not
be relied upon as dowels to serve this purpose. A joint located under the

Wall

$% Dowels
A X,
^Intermediate
^Const. joint floor

Const, jt.

(a)

FIG. 6-16. Some details of intermediate horizontal construction joints in heavy concrete foundation
walls.

Column
Reinf, (L Column
Jt i

Pour
Pour 2 Const, jh
Construction joint

i
* Pour J ^-Reinf. Pour
(a)- JOINT NEAR COLUMN (b)- JOINT UNDER COLUMN

2) \ ® \

View A-A View B-B

17 If0\/
tioriz.key ^Fnn4it
l
Footing f/oriz. key
fc) -ORDINARY KEYS AT JOINT fdj-OFFSET JOINT re) -INTERMITTENT KEYS

FIG. 6-17. Vertical construction joints in a continuous foundation wall.

center of a column, as in (b), may be safe, but this is often inadvisable.

It is preferable to have a construction joint at the center of the space

between columns, or at other points of small shearing stress.
Keys may be used at vertical construction joints, as shown in Fig.
6- 17(c). The vertical key will not be effective in resisting vertical shear-
ing forces, and the key in the footing alone may be too weak for this pur-
 .

Art. 6-7] MISCELLANEOUS DETAILS 171

pose. The scheme shown in Sketch (d) may be useful if the shearing
force always downward at the right, but it is weak when the reverse
is

occurs. The intermittent keys shown by B in Sketch (e) are useful in

resisting both horizontal and vertical shearing forces. They may be made
easily by attaching short pieces of boards or planks to the inside of the
end form of the first pour.
Contraction joints (and so-called expansion joints) in heavy foundation
walls are generally troublesome. It is often difficult to determine whether
to cut the foundation into units of reasonable length or to endeavor to
knit it together as an entity. If the superstructure is sufficiently flexible
and if the walls are composed of units that permit slight adjustment, the
foundation walls should be cut by contraction joints into units from 60
to 100 ft. long, with joints in the floor slab to match those in the walls,
although the floor should have many more such joints. If there is a base-
ment, and concrete floors supported by the basement walls, it is preferable
to try to make the substructure act as a unit unless its length exceeds 200
or 300 ft. In such a case, it is possible to minimize the harmful effects
of shrinkage by pouring the walls in alternate units, then filling in the
intervening sections 1 to 3 weeks later. It is probable that basement
walls that are not largely exposed to the weather will not be subjected to
a wide range of temperature change because the air inside remains rela-
tively uniform in temperature. It is apparent that building foundations
are not affected so violently by changes in weather as are retaining walls
and similar long exposed structures.
If contraction joints are to be used, they should be located at proper
points. A
few suggestions regarding their locations are the following:
1 They should be
at points having small shearing action. A location
similar to that of the joint in Fig. 6-18(a) is inadvisable because of the
interference with proper distribution of the column load.
2. When a double column is used at a contraction joint in the super-

structure, the joint in the substructure should be in the same plane, as

indicated in (6).

3. has offsets and corners, as in (c), the joints should be at

If the wall
such positions as A, B, and C so that the substructure is divided into parts
that permit small motions to occur without cracking the wall. Corners
and offsets may act like anchors or as points of weakness. The longi-
tudinal motions should open the joints, not shear them as at D.
4. Keys should be used when vertical or horizontal shearing forces are
They should be strong but not too deep, and
to be resisted at the joint.
they should have beveled edges. The end of the first section of wall and
the keyways should be painted with asphalt, paraffin, oil, or some other
material to prevent bonding the two sections together. A sample detail
is shown in Sketch (d).
 )

172 FOUNDATION WALLS [Chap. 6

Contraction (and construction) joints may leak. One type of water-

5.

stop used to prevent thisis pictured in Sketch (e). The continuous sol-
dered copper strip should be stiff enough to hold its shape during the
placing of the concrete. Such a strip placed in a horizontal construction
joint is likely to be bent down seriously by construction operations. Some
Footinq

i Direction of'motion «_4

/ 1
Plan A-A Plan B-B

VfWall \TyCol. i r n n "1

D (bad)
~\f\ Br ~1& ^
Contraction
Joint Wall* B
Contraction joint
fa^-POOR LOCATION (6M1 DOUBLE-COLUMN Cc;- JOINTS IN LONG WALL
EXPANSION JOINT
V-cut if
desired
Reinf.^ 1 h 2 covers Ins,ole

1
(
("» ' ——
l

»•?

Membrane IV. P.
/ *
V. f)/imnr\mr

I" to
2"
_ // -
Copper. Dampproofihg Protection Copper
-A k-
on outside
deep
foW-KEYED JOINT feJ-WITH WATERSTOP tfV-WATERSTOP ON OUTSIDE

flnside Nails (copper)

o t F7^
Joint -* Asphalt
Paint paint

'
Caulk

* Membrane Cork or premoulded

waterproofing ' Protection mastic filler

Copper
(g -CAULKED OUTSIDE WATERSTOP (h) -EXPANSION JOINT

FIG. 6-18. Some details of contraction joints in foundation walls.

other types of waterstop used with membrane waterproofing are shown

and (g).
in (/) "Dumbbell" rubber junction pieces, special calking com-
pounds placed in grooves at the joint, "Igas," and certain plastic sealing
compounds are other materials which are used to waterproof contraction
and expansion joints.
If a compressible filler is to be used in an expansion joint, as shown
6.

in Sketch (h), the filler should be fastened mechanically so that it will

remain in place initially and cannot fall out later.
Art. 6-7} MISCELLANEOUS DETAILS 173

7. Of course, reinforcement should not pass through a contraction joint.

When planning the reinforcement of walls that are subjected to longi-
tudinal and lateral forces, the designer should be sure that the structure
is tied together properly. The following comments refer to the corre-
spondingly lettered portions of Fig. 6-19:
(a) Hooks on reinforcing bars are useful under conditions that do not

afford adequate development of bond otherwise. However, when used as

at A, they are harmful because of the concentration of stress at one place.
It is better to splice the bars by unfolding the hooks and so secure a simple
lap of adequate length, as at B.

C
K'Reinf. „ few >„j° |h
LA. Reinf,
j
Pressure
mE ry
y
^-Pressure r ••' ^
D I?: -I"

(a)- SPLICES OF CW-WRONG METHOD OF INNER TcJ-BETTER ARRANGEMENT OF

REINFORCEMENT REINFORCEMENT INNER REINFORCEMENT

Cross wall-

Main wall-

E7TTT1
Pressure
rTTTTTTT
^-Pressure
^-REINFORCEMENT FOR ^-JUNCTION OF MAIN AND f/y-REINFORCEMENT FOR
TENSION IN OUTSIDE CORNER CROSS WALLS FILLET AT INSIDE CORNER

FIG. 6-19. Some details of reinforcement for heavy basement walls.

(b) Reentrant corners that are subjected to forces that try to separate
wall C from D,
or to increase the angle between them, tend to crack.
Tension in bar E causes it to try to straighten out as shown by the dashed
line, thereby producing a tendency to spall off the concrete and permit
large cracks before appreciable resistance is offered.
(c) Reentrant corners may be reinforced better if made as shown here.
Bars F reinforce the inner corner. The lap at G is desirable but is not
required to resist bending.
(d) When tension exists around a projecting corner, bars H should be
lapped.
(e) At such an intersection as this, bars J should be bent as shown or

form a U; hooks are less desirable. Bars K and L should go straight

through or be lapped.
 :

174 FOUNDATION WALLS [Chap. 6

(/) Large fillets in a reentrant corner are beneficial when subjected to

compression. If the corner is in tension, they avail little unless bars M
are used.
Foundations for wooden floors or superstructures should be planned so
as to prevent the access of termites to the woodwork. Suggested con-
struction is shown in Fig. 6-20, and the following suggestions are given

Copper termite guards should be placed beneath the woodwork and

1.

on all pipes, piers, and parts that may serve as a means of access for the
insects.
2. All guards and woodwork should be as high above the ground as it

is practicable to have them, 1 to 2 ft.

l
r
W!m Barrier

r«;-STUCco and frame (b) -FRAME rcJ-BRICK VENEER (W-SOLID MASONRY

CONSTRUCTION CONSTRUCTION CONSTRUCTION CONSTRUCTION

FIG. 6-20. Some construction used to prevent access of termites to timbers of a building. (Courtesy
of American Brass Co.)

Enclosed spaces without basements should be accessible for inspec-

3.

tion, and a clearance of 2 to 2}4 ft. is none too large for one to crawl
around to examine the structure. If inspection is not easy, it will seldom
be done.
4. Cracks in foundation walls may become tunnels for the termites.

It is obvious, therefore, that a hollow-block foundation wall is especially

hazardous unless the voids are filled with concrete or the upper part of the
wall made of solid brickwork or concrete.
is

Poison of the right kind placed in the ground alongside a foundation

5.

wall will be helpful in protecting the structure for 1 to 5 years, but this is
a temporary expedient.

PROBLEMS
For all the following problems, assume that the cover over the reinforcement is
3 in. and that the minimum frost depth is 4 ft.
6-1. A basement wall is 10 in. thick and supports a load of 4 kips per lin. ft. of
wall. The soil is silty clay. Design a footing for this wall.
6-2. The basement wall under a small apartment house is somewhat similar to
Fig. 6-1 (d). The wall is 14 in. thick and 10 ft. high. The load of the brickwork,
 PROBLEMS 175

etc., is 6 kips per lin. ft. of wall. The allowable bearing on the soil is l -^ tons per 1

2
ft. Design a footing for this wall.
6-3. A factory is to have a foundation composed of spread footings and grade
beams like Fig. 6-3(d). The columns are 25 ft. c.c. and support loads of 300 kips
each. The pedestals are 2 ft. 6 in. square. The brick walls weigh 1 .5 kips per lin.
ft. of wall. The grade beams are 14 in. wide and 30 in. deep with three No. 8 bars
in the top near the column and in the bottom near the center but only two in
the top and bottom elsewhere. The footing is 8 ft. square, 21 in. thick, and rein-
forced with two bands of 17 No. 6 bars at 5 in. c.c. The top of the concrete is 1 ft.

above ground; the bottom, 4 ft. 6 in. below grade. Is this satisfactory?
6-4. Design grade beams and footings for a factory wall having concrete
columns 18 in. wide, 24 in. deep, 22 ft. 6 in. c.c. The column loads are 350 kips
each; the wall load, 2 kips per lin. ft. The floor is 6 in. above ground. The
2
allowable bearing value of the earth is 3 tons per ft.

6-5. A foundation wall is to be of the type shown in Fig. 6-4. The columns are
22 ft. c.c. and support loads of 240 kips each. The wall load from the super-
structure is 1.2 kips per lin. ft. The floor is 1 ft. above ground. Column bases
are 16 X 24 in. The allowable bearing value of the soil is 4 k.s.f. Design the
foundation.
6-6. Assume a foundation exactly like Fig. 6-4 except that the wall footing pro-
on each side of the wall and the column loads are 275 kips. The foot-
jects 10 in.
ing reinforcement is No. 3 bars 12 in. c.c. for the wall and No. 4 at 6 in. c.c. under
the pilaster. The wall reinforcement is as follows:
Two No. 7, full length top and bottom
One extra No. 7, 14 ft. long in top for central portion of bay
Two extra No. 7, 12 ft. long in bottom at the column
Is this construction satisfactory?
6-7. A row of crane columns in the center of a factory is to be founded upon a
continuous wall. The columns are 20 ft. c.c, have dead loads of 100 kips each,
and are to be on base plates 16 X
36 in. Assume the crane loads to be equivalent
on anjr one column, or 150 kips on two adjacent
to a concentrated force of 300 kips
columns. There is no wall in the superstructure between the columns. The
ground has a safe bearing capacity of 4 k.s.f. Design the foundation.
6-8. A reinforced-concrete warehouse has exterior columns 21 ft. c.c. The con-
struction at the first floor is similar to that of Fig. 6-10(6) except that the columns
are 20 X 24 in. Their outer faces are covered with one layer of bricks so that
they are 6 in. inside of the outer surface of the concrete wall. The column loads
are 400 kips each; the wall and floor loads average 4 kips per lin. ft. The distance
from first floor to basement floor is 11 ft. The first floor is approximately 4 ft.
above the ground, and a small window is in each bay of the wall. The soil is
medium sand and is well drained. Design the basement wall.
6-9. It is planned to have the exterior foundation wall of a factory 5 ft. (over-
all) deep with a 1-ft. projection above ground and a footing 10 in. thick projecting

12 in. each side. The wall is to be 13 in. thick with 6- by 20-in. pilasters at the
columns 20 ft. c.c. The masonry on the foundations weighs 500 lb. per lin. ft.
The dead load on all columns is 40 kips; the live load, 50 kips on all columns and
a moving load of 60 kips on any adjacent pair of columns. Determine the neces-
sary reinforcement, and check the concrete.
 7
MATS

7-1. Introduction. The word mat (also raft) is used to denote a sub-
structure that transmits its loads to the soil by means of a continuous
slab that covers the entire area of the bottom of a structure, like a floor.
The use of a mat may be looked upon as the result of a situation in which
the low bearing value of the soil would result in such large isolated foot-
ings that it is logical to join them all together and transmit the loads to
the entire area directly under the superstructure and where watertightness
is needed.
There are several variations of this basic construction, a few of which
will be discussed as though they were separate structures. In engineering
practice, there may be the need for combining and modifying the general
features presented in order to meet the exigencies of particular problems.
The engineer's first step in any case is to plan the substructure so that it
can be built practicably and economically, and so that he thinks it will
support the loads safely. The tentatively chosen dimensions and mem-
bers are to be tested by analysis, and then revised if necessary. A pre-
liminary analysis is used to obtain guiding data regarding the unit work-
ing stresses in the materials, and the thicknesses and reinforcement needed.
However, the methods of analysis and the assumptions should be such
that the designer is reasonably certain that they will enable him to obtain
a structure that will serve his purpose safely.
Here again great refinement of calculation is not always justified or
practicable because of the uncertainties of the action of soils and of such
short thick members that are often arranged in complicated and multiple
systems. The assumed supporting systems and the suggested methods of
preliminary analysis shown herein are believed to be practicable, to be
easily understood, and to yield safe results. The illustrations are pur-
posely simplified, and the general scheme of reinforcement is merely indi-

cated. Other procedures utilizing greater refinement of calculation may

be used as a final check of the preliminary design, recognizing the mat as
a highly indeterminate structure.
176
Art. 7-2] UNIFORM MAT 177

7-2. Uniform mat. A type of mat that may be constructed easily is

one of uniform thickness supporting individual columns, or columns that
are on individual pedestals, as shown in Fig. 7-1. Assume that this illus-
tration pictures a proposed substructure to be used as the basement of an
apartment house founded upon plastic soil, such as soft clay. The fact
that a basement floor is needed anyway is a secondary reason why one
might wish to combine this with the structural supporting system in the
form of a continuous mat. The illustrations are simplified by the omis-
sion of waterproofing because the latter should be outside of the structural
mat, and the waterproofing and its appurtenances would contribute little
or nothing to the support of the superstructure.
Figure 7-1 (a) shows an assumed vertical section through this basement.
The outside walls have pilasters at the columns, but the mat itself is to
serve as the footing, and the outer walls can distribute the column loads Pi
along the edge of the mat. The interior columns are assumed to go down
to the 5-ft. -square pedestals shown projecting above a portion of the base-
ment floor. These pedestals serve the same purpose as the capitals and
drop panels of a conventional flat-slab floor; i.e., they spread the critical
shearing and bending sections over enough perimeter to prevent harmful
concentrations in local spots. If the columns are used without any
enlargement of the base, the column is likely to punch through the slab
unless the latter is very thick or has some type of embedded structural
steel or reinforcing shear head which might serve the same general pur-
pose as a pedestal. The first of these is illustrated in Fig. 7-6(a); the
reinforcing-bar type is somewhat like the arrangement of bars a and c
of Fig. 7-4 (c), even though the mat may have a uniform thickness. With-
out some such spreader as a pedestal or shear head, the tension in the
bottom of the mat under the column is likely to be excessive. A flared
base on the column, like an inverted capital, would also serve to spread
the load on the mat.
Such a mat as this in Fig. 7-1 (a) should be thick, stiff, and strong. It
will generally be loaded with an upward pressure of 500 to 1,500 p.s.f., or
perhaps more. It seems, therefore, that it is not practicable to analyze
it strictly in accordance with the empirical codes often used for the analy-

sis of flat-slab floors, e.g., the Building Code Requirements for Reinforced

Concrete (ACI 318-56). Those floors are thin, relatively flexible, and
designed for much lighter loads.
It is reasonable to assume that the mat is so stiff and the load so con-
stant that plastic compress and adjust itself so that each column
soil will
load will spread almost uniformly under the mat in the general vicinity
of that particular column. For example, the total unit pressure under
the rectangular area DEFG, Fig. 7-1(6), may be assumed equal to one-
fourth of the total loads on the columns at D, E, F, and G divided by the
178 MATS [Chap. 7

area of DEFG, plus the weight of the mat per square foot. As the last
item does not affect the shears and bending moments in the mat, the net
load caused by the columns alone will be considered when analyzing the
mat except for the portion close to the walls, where the weight of the
basement walls is to be added to the assumed loads Pi, P 5 etc. For ,

the purpose of computing assumed average pressures under the slab, near
the walls, the outer column loads will be treated as though they were
concentrated at the pilasters. Therefore, the unit pressure for use in the
design of the mat is

'^ 2P +
p for DEFG = *
2Pa)
- '^°° + 550)
= 1.19 k,i.

Assuming the side wall to be 12 in. thick and 10 ft. high, and adding this
to the loads in the side-wall columns,

p for DETS = K(P* + P*) + ^(jj +P + 6) wall

A
34(500 + 550) + 3^(240 + 260) + 10 X 22 X 0.15
(1 + 20) X 22
= 1.18 k.s.f.

When the loads on adjacent columns (or in various areas of the base-
ment) differ considerably, such a mat as this
be undesirable owing may
to the likelihood of local settlement because the mat cannot spread
unequal heavy loads very far over compressible soil. The bays in either
direction should be reasonably equal in length, the larger spacing not
exceeding 1.2 times the smaller one, and the columns should be arranged
in reasonably straight rows. Furthermore, a spacing of columns exceed-
ing 20 to 24 ft. may require an unduly thick mat.
One method of making a preliminary analysis of such a mat is on the
basis of an assumed supporting system of column strips that constitutes
a grid of beams along the column rows in each direction. If this grid is

adequate, can support the loads, assuming the portions of the slabs in
it

the central areas to be supported by this grid.

Referring to Fig. 7-1(6), assume that 45° diagonal lines are drawn from
the column centers. These do not intersect at a point near H, J, and
K, but because of the 22 ft. of the longer span, there is a 2-ft. space
between the intersection points along the bisectors of the 20-ft. bays.
Using a single letter for each of these 2-ft. lines, one can say that the
pressure under area DHEJ is to be resisted by the portion of the mat
between D and E. Assume also that the similar area EKFJ is to be
supported by the mat between E and F. Next, assume that the rein-
forcement is so arranged as to form the chosen set of beams (column
strips) extending each way along the column lines, as pictured by w s in
Art. 7-2] UNIFORM MAT 179

Fig. 7-1(6). The effective width w s of these strips or shallow beams will

have to be assumed. If the pedestals are thick enough to support the mat
properly, w may be assumed to be wider than the pedestals, w p A
s
.

reasonable width may well be w p + 3d or slightly larger, but not over one-
"
22-0' 22 '-0' 22-0

Col.
£ Col. £ ir
{1st. F/oor
i » * ' scsaa « * '
**
ip
P faster p3 --sso k
' '
Pedestal
{Mat K t 2f80_ m
pressure p*/o' (a)- SECTION A-A
S* fCol.line

1 T

#
a
A kfc

£ H A
X T*S

/K
J

fe D
*

(6J-PLAN OF ASSUMED COLUMN STRIPS AND DISTRIBUTION OF LOADS

\B D E Q
TTTTT

flr.

rcJ-ASSUMED UPWARD LOADING ON COLUMN STRIP BDEQ

""
A\ \ Z> 1 /• \ /; / \ q

(d)- ASSUMED BENDING-MOMENT DIAGRAM FOR COLUMN STRIP BDEQ

FIG. 7-1. Assumptions for approximate analysis of a large mat.

half the average of adjacent bay widths, where d is the effective depth of
the mat. This gives an assumed strip a little wider than the width deter-
mined by 45° lines from the bottom of the pedestal to the lower reinforce-
ment in the mat. Perhaps a wider strip might be assumed. At any
180 MATS [Chap. 7

rate, this requires an assumption of the thickness of the mat. If the

effectivedepth d is assumed as equal to approximately 1 in. per ft. of clear
span between pedestals for heavy loads (less for light loads), d = 22 ft.
minus 5 ft. for the pedestals divided by 12, or d = 22 — 5 = 17 in.
Use this value, and add 3 in. of cover. These column strips are to be
designed strong enough by themselves to support the total net loads.
The width of the pedestals preferably should be one-fifth to one-quarter
of the column spacing.
A central "suspended" panel like MNOR of Fig. 7-1(6) may be designed
as a two-way rectangular slab with fixed edges. However, longitudinal
deflection of the column strips and angular rotation or curvature near
their edges may modify this assumed condition. Being poured integrally
with the column strips and reinforced continuously, these central portions
cannot well be simply supported along the edges of the strips. An ultra-
conservative way to design the suspended portions is to proportion the
bottom reinforcement as though these parts were fixed at or inside the
edges of the column strips, then to determine the top bars as though the
suspended part were simply supported along the same lines. However,
this conservatism seems to be a bit extreme. A safe rule seems to be to
assume that the suspended portions are fixed at imaginary supports
located at a distance of 20 per cent of the width of the column strips (but
not exceeding the effective depth d) inside the assumed edges of these
strips, then to design the outer portions of the suspended or central areas
accordingly. Having the bottom reinforcement thus determined, use the
same reinforcement in the top of the central areas like MNOR.
When the sides MR
and MN
are unequal, as in this case, the load will
not be distributed equally in both directions, but it may be a sufficient
approximation to assume that it is so distributed when the dimensions are
reasonably equal. Another assumption is that the load is divided be-
tween the two directions as shown by Fig. 2 in the Appendix. Assuming
the latter method, and adding 2d to the span for fixity, with the ratio of
imaginary spans equal to

MN + 2 X 1.42
= 14.84
=
MR + 2 X 1.42 12.84

Fig. 2 gives ps = 0.6p carried in the short direction and p L = OAp in the
long direction. Therefore, for area MNOR,

ps = 0.6 X 1.19 = 0.71 k.s.f.

Vl = 0.4 X 1.19 = 0.48 k.s.f.

Hence, the bending moments per foot of width for computation of the
Art. 7-2] UNIFORM MAT 181

tensile reinforcement at the edges are

Mmn = 0.71 X K2(12.84)

2
= 9.76 ft.-kips per ft.

MMr = 0.48 X K2(14-84)

2
= 8.8 ft.-kips per ft.

The steel (per foot of slab) parallel to the short direction is

9 76
As = = ° 382 in 2

X 1X2
-

20 X 0-9

The critical area to satisfy bond-stress requirements is

710 X 6.42 u
=
210 X 0.9 X 17
= 1 ,
m.

'
Pei
"
,, , . ,

Therefore, No. 5 bars could be used' at 9 in. c.c.

The steel area parallel to the long direction is

As = = °- 344 h1
2

X O^X
-

20 1.4 2

Therefore, for this reinforcement, No. 5 bars could be used at 10 in. c.c.

However, for simplicity and to avoid errors, these bars will be used in the
bottom at 9 in. c.c. in both directions. All this reinforcement should
cross under the full width of the column strips and be anchored or spliced
in the suspended panels. At least 25 to 50 per cent of this bottom rein-
forcement should be extended clear across the central panels in order to
have it wherever it may be needed and to knit the entire mat together.
The remainder of these bars may be stopped where theoretically not
required beyond the edges of the column strips, but they must be extended
enough to anchor them.
Inasmuch as any yielding of end restraint will cause less tension in the
bottom bars but more in the top of the mat near J, the use of the same
reinforcement in the top as in the bottom will provide a reasonable
At least one-half of the top steel
allowance for such possible increase.
should extend to or somewhat beyond the assumed supporting edges of
the central panel — 1.42 ft. outside of MNOR in this case; then at least
25 per cent should extend across the column strips and be lapped for
splicing in the latter.
Both top and bottom reinforcing bars in the central panel MNOR, etc.,
should be continued at the computed spacing s until the bars along the
edges are approximately a distance s from the column-strip reinforcing to
which they are parallel.
The purpose of the preceding requirements is to make sure that the mat
is well tied together outside of and across the column strips which are

actually indefinite in extent. Remember that assumptions are assump-

182 MATS [Chap. 7

tions. The cost of the steel will be small; its service value will be great
if need arises.
Assume that the column strip BDEQ of Fig. 7-1(6) is to be analyzed.
The total load on each span is found from the applicable areas such as
DHEJ for column strip DE. Even though DHE and DJE are slightly
trapezoidal instead of triangular, the load acting upon strip DE will be
assumed to be triangular as pictured in Fig. 7-1 (c). Using the average
unit pressure as 1.19 k.s.f.,

W = (2 X 20 X 10
2 :
+ 20 X 2)1.19 = 286 kips

Assuming that the ends of the column strips are fixed and using the data
given in Fig. 7-2,

M' = MD M
= E = 0.104JFL = 0.104 X 286 X 22 = 654 ft.-kips
M c at mid-span = 0.063 WL = 0.063 X 286 X 22 = 396 ft.-kips

The bending-moment diagram is pictured in Fig. 7 -1(d).

jTotal /oad=W

1
fa J- SYMMETRICAL TRIANGULAR LOADING DIAGRAM, SPAN Li

f6J-0RDINATES OF BENDING-MOMENT DIAGRAM AT

EACH 0.) OF SPAN TO BE MULTIPLIED BY WL
FIG. 7-2. Data for bending moments in rectangular beam with fixed ends. (Data from Hardy Cross
and Portland Cement Association.)

The column strip BDEQ may be analyzed by moment distribution if the

variations of loading or spans make this desirable, using the data in Fig.
7-2 for fixed-end moments. However, in this case, it is designed as a
series of fixed-end beams.
Should this bending moment M' at the supports be reduced because of
the column and pedestal? The column will certainly spread its load over
its area. If so, this will modify the bending-moment diagrams as in-
dicated in Fig. 7-1
(d). If the pedestals are deep enough, they will also
spread the column loads, and they are assumed to be deep in this case.
Art. 7-2] UNIFORM MAT 183

Therefore, the modified bending moment at point D, for example, is

approximately

M = 654 - 28% x 0.07 X 2.5 = 414 ft, -kips

where the downward force providing the end reaction is assumed to act
at the center of gravity of the neighboring triangular quarter of the 5-ft.
pedestal.
The reinforcement in the bottom of the column strip DE, Fig. 7-1(6),
under and near the columns will be

414
^ _
A
s
= ifi 9 in 2

20 X 0.9 X 1.42

Assuming the shear on both sides of the pedestal at D to be balanced

and equal to

VD = 2 %% - (5 X 2.5/2)1.19 = 136 kips

the requirement for bond is

= _ 136,000
=
210 X 0.9 X 17

Therefore, in the 10-ft, strip, use 13 No. 10 bars at approximately 9 in. c.c.

Since 71/,. at the center of DE is 396 ft. -kips, this nearly equals 71/ at D, so
that the same reinforcement will be used in the top of the central portion
of DE.
The maximum compressive stress in the concrete is near D and E.

,
fe
=
0.3
2 X 414,000 X 12 =
X 0.9 X 120 X 17'
1 '° 6 °
^ Me)
The punching shear at the face of pedestal D is assumed to be

VT = VD = 136,000
= ...
444 P SJ
.
,
(safe)
, ,

Wd 60 X 0.3 X 17
-
-

As indicated in Fig. 7-3 (a), the diagonal tension at a point 45° (1.42 ft.)

beyond the bottom of pedestal D may be estimated as

7 84 4- 20
1,190
.
1
(10 - 3.92) + 20 X 1

vL = = bi p.s.i.
(60 + 2 X 17)0.9 X 17

This is safe according to a beam specification but exceeds the 75 p.s.i.

allowed in two-way footings. Since the stress condition at D is similar to

that at the capital of a flat slab floor, one might assume that the Code
specification of 0.03/c with a '
maximum of 90 p.s.i. would apply. How-
184 MATS [Chap. 7

ever, this point in the mat is very important, and a little strengthening to
avoid a diagonal-tension failure should be provided in this case, inasmuch
as Vl exceeds 75 p.s.i. and since it seems that the vicinity of the pedestal
should be looked upon as similar in action to that of a two-way footing.

^-(j of bay

-Wall

ax <[ of column
"Column
strip

£ of bay

(a) SUGGESTED ASSUMPTIONS FOR ANALYSIS OF COLUMN STRIP

OF CONTINUOUS MAT

J L stirrups Edge of
pedestal
\__
ZZ ZZ /^Pedestal
Top bars

1 1 1
c\ /.Shear head — /
^
1 1 1

1 1 1 l l l
n I -*y /-/Bottom

1
1

1
1

1
l

l
1

I l
I -H— i—\±^- ^z_— -Zll .

(f of col.

id) PLAN OF STIRRUPS (c) SECTION SHOWING SHEAR HEAD

FIG. 7-3. Some details of a uniform continuous mat.

One method of improving the condition is to make the mat thicker

about 3 in. — and to revise the reinforcement accordingly. However, this
requires a large amount of concrete in order to aid a few localized areas.
A second method of reinforcing this region of high diagonal tension is
by the use few double U-shaped chairs, or U-bars with a 90° bend at
of a
the ends, placed slightly outside the pedestal. These are needed to hold
up the top reinforcement of the mat anyway. In this case two No. 5
U-bars placed 6 in. beyond the edge of the pedestal and two more sets 8
—
and 16 in. outside of the first one and the same for all four sides of the
pedestal, as shown in Fig. 7-3(6) should be sufficient. —
Art. 7-2] UNIFORM MAT 185

A third method is the bending up at 45° of two or three of the main

bars under the pedestal (in both directions), starting the first bends
approximately one-half the effective depth d outside of the edge of the
pedestal and any others about %d beyond the first bends. An equivalent
but economical arrangement can be made by adding a reinforcing-bar
less
shear head as pictured in Fig. 7-3(c). This adds considerable steel across
the bottom under the pedestal, unless some of the shear-head bars are
counted as main reinforcement, an idea which should not be carried too
far.

In this case, the third method (bent-up main bars) seems to be desirable
from the standpoint of tying the concrete together, but the second one
would probably be easier to install.

Pilaster Col.
Wall ttt cMcit n n Z7~
B -<t:2 or """*'
"'•'"' flatter*
C
fa)- PARTIAL VERTICAL SECTION
Probable
J± ) crack

c E W cF ^itttttttt
r^-SECTION THROUGH DEPRESSED FOOTING (^-DEPRESSED FOOTING WITH VERTICAL EDGES

J±L i Shear CtL

swm \ rr

t it t

^FTTTTTT^T
(d)- POSSIBLE CRACKS CAUSED BY SHEAR feJ-EXAGGERATED DEFLECTION OF MAT

FIG. 7-4. Continuous mat with depressed footings instead of pedestals.

Other spans, including the exterior bays, may be designed similarly.

For EF the loading is triangular.
Figure 7-5 shows the construction of a thick uniform mat that now sup-
ports a very heavy industrial building made of reinforced concrete. As
indicated in the picture, the slab is embedded some 8 or 10 ft. so that the
bottom story serves the purpose of a basement with conveyors and other
equipment in it. The mat has a system of drains under and alongside it.
The soil is a somewhat spongy volcanic material. There has been no
harmful settlement of the structure.
When a wall occurs on a mat and along an intermediate column row, it
seems to be satisfactory to design the column strips normal to the wall as
usual, considering the edge of the wall to be equivalent to the edge of a
column pedestal. The portion of the assumed triangular or trapezoidal
areas delivering load to the wall can be handled somewhat as for a con-
186 MATS [Chap. 7

tinuous two-way slab, except that the shears and moments at the wall
should be increased (perhaps 20 per cent) to allow for the fact that the
wall is so much stiffer than the imaginary column strips in the mat.
Along an edge wall, such as AST in Fig. 7-1(6), when suitable continuity
is not obtained, the perpendicular column strips may be designed with a

simply supported end at the wall. The intervening portions of the mat
such as SHT may then be designed as parts of a two-way slab, with a
hinged support at the wall plus an increase (perhaps 20 per cent) in the
shear and bending moments to allow for the "infinite" stiffness of the wall
compared with that of the nearest parallel column strip.
It takes little imagination to realize that this mat, shown in Fig. 7-1,
will require a large quantity of concrete and that the large pedestals in
the basement may be a hindrance to use of the space by the occupants of
the structure. The first is inherent in the plan and perhaps unavoidable,
but it should be judged in comparison to the cost of alternate types of
foundation. The second might seem to be remedied by thickening the
slab locally at the bottom, as pictured in Fig. 7-4 (a). If the thickness
near the column is enough to resist shear, if the compressive
to be great
stresses in the footing or mat around the column are to be reasonable, and
if the tensile reinforcement under the column is to be moderate, the

thickening under the column must be very considerable. This requires

the careful excavation of a pocket at each column as indicated by DEFG
of Sketch (6).
might seem that the reinforcement of this mat could be arranged as
It
indicated in Fig. 7-4(6), having bars a and 6 lapped as shown. This might
be done for the bars in the flat portion, as shown by EF for the bars at 90°
to the cut section. However, the bars c in the sloped portions such as DE
and FG will have varying bends, and all of them make a complicated situa-
tion where a and c in both directions intersect bars 6 which are also in both
directions. the mat beyond D and G is rather thin, bars a and c
When
may have hooked or bent in order to develop bond properly. This
to be
complicates the reinforcement still more. If bars a and c are bent down
parallel to and alongside 6, their tendency to straighten under tension may
spall the concrete and cause failure of the mat.
The sloped sides DE and FG of Fig. 7-4(6) are made this way in order to
ensure good bearing on the soil. Assume that the footing is made as in (c)
with a steep offset. The soil near H and J is likely to be weakened. Fur-
thermore, the upward shearing forces at H and ./ caused by the mat are
likely to crack the structure as shown. Even if bars d were bent vertically
upward and hooked would be
into the top of the mat, their effectiveness
questionable. Concrete construction can stand pressures easily; when it
must be stitched together with reinforcement to prevent being pulled to
pieces, the advisability of the basic plan is questionable. Even with the
Art. 7-2] UNIFORM MAT 187

sloped bottom, there is still the tendency to load the edges of the thickened
section as illustrated in Sketch (d). An exaggerated picture of possible
failure is shown in (e).
Another variation of the construction is the use of enlarged strips or
beams shaped like the section shown in Fig. 7-4(6). These are to extend
continuously in both directions under the columns. Thus they form a
grid of column strips or beams that will act even more closely in accord-
ance with the assumption that the column strips constitute a system of
beams that supports the entire pressure under the mat. In effect, it is a
checkered pattern of shallow wide beams with thin central panels attached

FIG. 7-5. Construction of a heavy mat under a secondary crushing plant. The soil is of volcanic origin
and is somewhat spongy. (Courtesy of trie Cananea Consolidated Copper Co., S.A.)

to them. The excavation is somewhat more troublesome than is that for

a uniform mat but this is usually of little consequence. This system may
be very useful.
In Fig. 7-6 is shown another arrangement for a continuous mat. This x

general scheme was used, for example, in the Kansas City, Kan., plant of
Sunshine Biscuits, Inc. Beneath the structure is about 12 ft. of gumbo.
The grillages and welded column stubs shown in Sketch (a) serve as a sort
of footing to spread each column load over a sufficiently large area of the
mat to avoid excessive bending, shear, and diagonal tension, with the help
of the U-stirrups when necessary. In this way, heavy billets under the
steel columns are avoided, projections above the floor are prevented

1
Patented by Walter H. Wheeler, Minneapolis, Minn. See Multistory Factory
Floats on Gumbo, Engineering News-Record, Mar. 3, 1949.
1! MATS [Chap. 7

without increasing the thickness of the mat, and a slab of uniform and
moderate depth can be used, thus facilitating preparation of the subgrade.
Many other details are shown in Fig. 7-6. Special attention should be
given to the following features:
1. The assembly of the grillage on seat angles or shims and a concrete
pad, with space for the reinforcement to be inserted.
2. Most of the load from the column reaches the concrete through the
bearing of the top flanges of the channels, although bond may help con-
siderably also. Notice that downward pressure on the bottom flanges of
the channels will tend to spall off the concrete whereas the stirrups tie
back the bottom to the main body. Proper design of these grillages is
important.

Stub column, splice 2-0'to3-0" . ,-

A
,

above mat Edgeofmat ,

^Construction joint
,

<£
x

Welded /s forming/ /W for setting

d^Ztzndset grillage
on shims 1st pour 2nd pour 1st pour 2nd pour

faj-DETAILS OF GRILLAGE AND MAT (AJ-Pim OF POURING MAT TO REDUCE SHRINKAGE CRACKS

FIG. 7-6. Patented system using steel grillages to spread load into uniform mat. (Courtesy of Walter
H. Wheeler, Minneapolis, Minn.)

3. The general arrangement of the reinforcement. The principles of

design are similar to those used for flat-slab floors, or as stated herein.
4. The construction
joints, shown in Sketch (b). To minimize shrink-
age, the slab waspoured as bands in alternate strips across the width
first

of the building, then the remainder was filled in after initial shrinkage
occurred. The concrete was made with low-alkali content, type II cement
to reduce shrinkage in general.
A system of underdrainage was installed to remove ground water to
sumps and automatic pumps. This system was to prevent hydrostatic
pressure under the slab. It works well in impermeable soils, but porous
materials tapping large sources of ground water might require too much
pumping.
A mat is a required structural member used to spread loads. There are
cases where the construction appears to be a mat but is not one. For
example, the bottom of a concrete tank placed directly on the ground
may have a continuous concrete floor. This, however, is not a structural
Art. 7-3] RIBBED MAT 189

mat to distribute concentrated loads; it is a pavement to prevent leakage.

Such a slab may well be only 4 to 8 in. of concrete, or gunite, but it

should be strongly reinforced, perhaps with a steel area of at least 0.5 to

1 per cent of the cross-sectional area of concrete, and with the bars extend-

ing in both directions. In such a case, the slab is purposely made thin
and can accommodate itself to unequal settlements of
flexible so that it
the ground without too harmful cracking. It has no influence upon the
pressure under the tank or the settlement of the soils below it.
It has been stated that a mat may not be stiff enough to spread the
loads properly if the forces on the columns differ greatly. The loads will
go as directly as possible to the nearest reaction points which can sup-
port them. Much depends upon the strength and stiffness of the soil
under the mat as well as of the latter itself. This may also apply to
problems of overturning (as for resisting earthquakes). For example, a
power plant founded on a mat 80 ft. wide and 120 ft. long cannot resist
tipping about one edge like a tremendous footing under a huge retaining
wall; it will be more likely to form a series of bends between columns.
In other words, the pressure under the edges or elsewhere cannot be com-
puted by use of the formula

V = P ± Mc
,

A ~T
where A is the entire area of the mat, because the latter is not stiff enough
to act as a unit. The best that seems acceptable is for one to devise a
probable (or possible) distribution of loads on the soil, then provide a
structural system in the mat itself capable of distributing the applied loads
as planned.
The reaction of the depend on the soil modulus as well as on
soil will

the stiffness of the mat. When

the column loads vary greatly, or the
spacing of columns differ considerably, or edge or interior walls modify
the uniformity of the loadings and action of the mat, one should try to
determine the probable relative consolidation of the soil and the distor-
tion of the mat. From the latter, the otherwise computed shears and
bending moments may be modified as seems necessary.
7-3. Ribbed mat. The preceding article shows, when one examines
the bending moments in the mat, that it may be desirable to strengthen
the grid of assumed supporting slabs by using a system of heavy beams
or walls along each column row and in both directions, then to design
the mat as a system of two-way slabs supported by these beams. Of
course, these beams will generally have to be very heavily reinforced for
bending and shear. This arrangement has many obvious difficulties. If
the beams are deep ribs placed below the basement floor or mat, as in
Fig. 7-7 (a), the bottom of the excavation becomes badly cut up with
190 MATS [Chap. 7

trenches, the work be costly, and the bearing value of the soil may
will

be impaired because ground next to the ribs. The

of disturbance of the
stirrups in the rib should be strong and stiff because tension in them tends
to crack the rib or open it up at the construction joint.

2-way mat. {Floor slab ^ Const jt.

rvT \Pilaster at column,

|
1]

Compacted
"
Rib at 90°-
Pressure Rib
Y \fcconst.
fill t, $
/M>^b*d jt r
« :
'
1{ y Roughened const, jt.
'

fftibat90° I
/lj 2-way fnati)

±zaaas
fa,l- RIBS BELOW MAT
C6)-R\BS ABOVE MAT

/First floor slab

»::.:, M?.--..:.^
P»-5

*-Beam
Roughenec i i

-Pi/aster at col.
const jt
i

i-
Wall- Wall at 90
l

Const.
>nsr.~^.i |
—
jt
it. I) „ 2-way mat One-way mat^
,__u_3ii_.
(cHWO-WAY MAT WITH INTERIOR WALLS rrfJ-ONE-WAY MAT WITH INTERIOR WALLS

^-CONSTRUCTION JOINT IN 2-WAY MAT ft7 -CONSTRUCTION JOINT IN ONE-WAY MAT

FIG. 7-7. Some construction for ribbed mats.

If the construction is inverted so that the mat is at the bottom, the

ribs will destroy the usefulness of the basement unless the main founda-
tionis depressed and a basement floor is placed over it, somewhat as

shown in Sketch (6). On the other hand, walls with doorways through
them might be used as indicated in (c) if a series of small rooms is accept-
able — which isseldom the case. Any of these grid systems will generally
distribute a moderate variation in column loads satisfactorily if the
detailed design of the ribs provides for this. Any such construction
should be strong and stiff. It is a poor place to skimp in the use of steel
Art. 7-3] RIBBED MAT 191

and concrete. Furthermore, the owner seldom consults the designer when
he intends to increase the loads applied to the structure at some future
time.
An alternate method that may be used sometimes with a continuous
mat under a basement is a system having continuous walls (with occa-
sional doorways) extending along the column lines in one direction only.
A possible arrangement for this construction is shown in Fig. 7 -7(d).
Since the mat is primarily a heavy one-way slab, it may be desirable to

FIG. 7-8A. Ribbed-mat substructures of the Central Mill Boiler House, Anaconda Wire and Cable
Company, Hastings-on-Hudson, N.Y. The interior columns will rest upon the ribs at their intersections.

The cantilevered portion of the mat will be covered with earth.

thicken it have greater resistance

as indicated in the sketch in order to
to bending and shear at the walls and economize on the use of con-
to
crete in the central region of the spans. The walls are usually so deep
and strong that, if the doorways are at the centers of the bays and if they
are not more than 4 or 5 ft. wide, one may assume that the column loads
are distributed uniformly along the walls, including the space below a
doorway. As this system does not have great strength to distribute loads
in a direction perpendicular to the walls, the column rows should be loaded
reasonably equally.
In Fig. 7-7, the general arrangement of the principal reinforcement only
is shown. The slabs and ribs or walls should be designed as continuous
members. Of course, construction joints are necessary. Suggested loca-
192 MATS [Chap. 7

tions for these between ribs and mats are pictured in (a) and (6). If

vertical construction joints are to be used also, it is generally desirable to

locate them at points of small shearing forces, as indicated in Sketch (e).
When vertical construction joints are to be in a one-way mat construction,
it is preferable to locate them at the centers of bays in which there are no
doorways in the walls. Horizontal keys in the mat and intermittent keys
in the ribs and walls are desirable at such joints.

IliiplS^lililliipB**8*

FIG. 7-86. The structural framework of the Central Mill Boiler House and the construction of the brick
walls. The floor slab was not poured until after the installation of equipment and piping. (Courtesy
of Anaconda Wire and Cable Company.)

The and walls of a rectangular grid system such as that of Fig.

ribs
7-7 (e) may
be analyzed as fixed-end or continuous beams with triangular
loading based upon the principles of distribution shown in Fig. 7-1(6) for
column strips. The reaction points are at the columns and outside walls.
The slabs may be proportioned to resist uniformly distributed loads as
two-way slabs with fixed or continuous edges.
The walls of a one-way mat system like that of Fig. 7-7 (/) may be
analyzed as continuous or fixed-end beams loaded uniformly along their
lengths, with reaction points at the columns. If the walls are cut by rela-
tively small shallow doorways at the centers of the bays, these openings
may be neglected; if the doorways are large or high, it is best to consider
Art. 7-3] RIBBED MAT 193

the wall as hinged at these points as far as bending is concerned but to

reinforce the top opening enough to prevent cracking.

and bottom of each
The mat, of course, is to be designed as a one-way, fixed-end, or continuous
slab uniformly loaded by the net upward pressure.
A two-way mat with ribs for a small boiler house is shown in Figs. 7-8A
and 7-9. The general dimensions of some of the construction are given
in Fig. 7-9 in order to show some scale regarding expected sizes for net
pressures of approximately 1,000 to 1,500 p.s.f. with considerable spread-
ing of the heavy loads required from the central boiler columns toward
the outside.

4 of cols.
S1!"/ 18-0" 16- 3" ,IO'-0" 5-1 Brick wall
sMaf
1^~ .% ft
2"
0-0"
Fill of
sand, gravel
Fijorm*-*5 12-0"

"ter^nthrt Slab on Y77V77V

backfill Y
| |

{.Holes for
Sand.sheils.
trace of mud
:' Clay and
25-0"
i 1-M utilities shells
42'
0"

E 9 CCS
"I
Mi-Li
I

'O
Main bars *8
;

;
Sand, gravel
and shells
SJ-O"
°.°* Gray sand
CW -SECTION A-A and gravel
70-0"
^ 70 Fine sana
75-0"
k 56
85 \-\ fine sand
/Floor slab
-fc
« 98 and silt
37-0"
V*-~— 'TrwTTWtpfrjpzW**
ZTio
T
-S 121 ?-;\n sand and r,n f
94-0"
•o IIS

A -i iv.'' I
•Compacted $100 Brown sand
03
Main bars \ \ backfill 161
II2 !0"..
IS8 -r.- 6 ravel
'8 JB- Sf
w US'II

4k)\" 46'S" jz~i

E1T5I 1 -9"c.c.
Rock

J-B0RIN6 NEAR SITE

OF BOILER HOUSE
'

125-/7"

^/-SECTION B-B
54-5
(ahVLNH AND HORIZONTAL SECTION THROUGH RIBS

FIG. 7-9. Ribbed mat for Central Mill Boiler House, Anaconda Wire and Cable Company, Hastings-
on-Hudson, N.Y.

A brief outline of the evolution of this design may be instructive:

1. At the first site selected near the river, the borings showed approxi-
mately 20 ft. of fill, a tapered stratum of some 40 ft. of soft clay, a few
feet of silt, some fine sand, and then coarse sand. It was decided that
90-ft. concrete-filled pipe piles should be used to support the structure.
2. An alternative shoreward location was The old fill
investigated.
here was only about 12 ft. deep, the clay stratum was firmer and only
about 15 ft. thick, fine sand underlay it, and then there was coarse sand
at a depth of 40 ft. It seemed to be safe to use a mat foundation here
and thus avoid the expense of using piles. The ribbed type was adopted
in order to spread the heavy interior column loads.
3. The original design for the superstructure is shown in general in Fig.
7-10. The two small wings at the right and left illustrate a principle to
be borne in mind when planning mat foundations the desirability of fairly :
194 MATS [Chap. 7

uniform loading or of great stiffness of the structure in order to spread

the loads. The central loading here is heavy, whereas the weight of the
wings is light. —
As the structure settles and some consolidation of these
soils is inevitable — there would be a tendency for the center to settle more
and to shift loads to the wings, probably breaking the substructure near
the outer walls of themain building because of the upward bending
moments. The entire mat will try to act as a unit. One should not
assume that the ground under one part can be loaded heavily whereas

fjBf- ff/ar/ftw/swH

FIG. 710. Preliminary study of small boiler house.

the remainder is loaded lightly unless unequal settlements occur. There-

fore, the layout was revised to eliminate the extensions and place the
pumps and compressors inside the main walls.
Level readings to show the settlement of this structure were taken once
a month for over 2 years. The results are given in Table 7-1. As ex-
pected, settlements occurred, but they seem to have been reasonable and
to have, been substantially completed in 2 years.
When deciding upon such important matters as the use of a mat or of a
deep more costly substructure, one should not forget that there is little to
be done if the future proves that this decision was unsound.
Another case is illustrated by a small laboratory building that was to
have brick walls and a steel trussed roof. It was to be founded upon a
20-ft. blanket of new fill placed over soil of questionable bearing value.
It did not seem to be worthwhile to drive long piles to hold up the struc-
Art. 7-3] RIBBED MAT 195

TABLE 7-1. Settlement Record of a Small Boiler House Supported by a

Ribbed Mat on Clay and Sandy Strata

Date Sta. A Date Sta. A

3 9
Jan. 14,1947 1
32 Aug. 21,1947 1
16
5
2 8 2 1

16
23 11
3 32 3 1
1

16
13 27
4 32 4 32

1 17
Feb. 12,1947 1
8 Oct. 17, 1947 1
32
25 11
2 32 2 I-
1

32
31
2 I 3 32 3 I1
11
32
PLAN- 17 27
4 32 4 32

3 19
Apr. 17, 1947 1
8
June 1, 1948 1
32

2 l& 2 111
3 \k 3 ift
23 27
4 32 4 32
SECTION
1 5
June 26,1947 1
2
Mar. 24, 1949 1
8

2 iA 2 11
^
8
A= Settlement in inches 3 i£ 3 11
11
32
be/o w El. 9.26 27 13
4 32 4 16

ture. Therefore, the design shown in Fig. 7-11 was developed. In

principle, it is somewhat like an inverted concrete barge with a continuous
mat as a floor and with heavy ribs around the edges and across the width.
The substructure planned so as to have sufficient rigidity to prevent
is

local sagging. was expected that the entire structure would settle a
It
little. This, however, has not been noticeable, and no objectionable
cracks have appeared in the concrete or the brickwork.
This construction in Fig. 7-11 proved to be so successful that the
designers used the same scheme for a 60- by 240-ft. clear-span structure
that was to be supported upon new but good fill that varied from 9 to 14 ft.

in depth. The intermediate cross ribs were omitted. Columns were

along the outside only. The men in the field proposed that concrete piers
with footings be used at 15 or 20 ft. c.c. under the peripheral ribs. These
would have been harmful for the structure because the edges could not
settle with the fill, whereas the central part of the floor would have to do
so because the slab could not span a distance of 60 ft. safely. The basic
idea of the design was that it should accompany the settlement of the fill
without injury. If piers were used around the edges, it would be better
to cut the floor slab loose from the outer wall so that it could settle with
the fill. This, however, might cause trouble at doorways and partitions.
When planning any mat foundation, the engineer should explore the site
196 MATS [Chap. 7

thoroughly to make sure that conditions are suitable. It is very impor-

tant to avoid relatively hard spots and soft spots under the mat. For
example, plans were made for a mat foundation under an industrial
building that was to be placed close to an existing structure. The main
foundation consisted of an 18-in. slab with a heavy grid system of rib walls
around the outside of and across the mat along the column lines. When
the excavation was made, a forgotten old box drain was encountered,
somewhat as shown in Fig. 7-12(a). The drain cut across one corner of

T
r 1 [ Oufer rib\ \
I

-
A « B B II CM

L |J i| LiJ !
Sr

1
HI
1 11,$ II 1
1
1

1 II II II 1

1-0" 21-0" !9'-0" l9'-0" 2/-0" /-'0

-«=
\ 32-0" . =».

(ct)-PLM OF SUBSTRUCTURE

{Floor

Ww^SISB^^-

!'-6"

(b)- SECTION A-A (W-SECTION B-B

FIG. 7-11. Construction of ribbed mat for small industrial building.

the area for the mat, as pictured in Sketch(6) The contract had been let, .

and it was too late to make extensive changes in the plan. The old drain
was therefore removed, and a by-pass was built as indicated, then the
trench was cleaned out, and the sides were sloped to avoid leaving
weakened material alongside. If the space thus made were rilled with
weak concrete, it seemed that there might be a wedging action along the
sloped surface AB, causing the soil to move from under AB and BC.
If so, the column loads at D and E might cause the structure to "break
its back" over the ridge A. It was therefore decided that the earth out-
side of AB should be excavated down to the level shown by the dashed
line AF, and that the entire space below the mat should be filled with
Art. 7-3] RIBBED MAT 197

weak concrete. Even then, the situation is one to worry an engineer.

Were the extra excavation and concrete in the space FABC necessary?
Remember it is not wise to wait and see what happens before applying a
remedy. If the drain had been discovered before the design was made,
the mat should have been lowered to avoid this bulge, or it might have
been sloped gradually upward away from F and A.
When the use of a mat foundation is contemplated, whether it is a flat
slab like that of Fig. 7-1 (a) or a ribbed one as in Fig. 7-9(a), the engineer
should remember, especially when the underlying soils are plastic, to
endeavor to equalize the loads on the mat so as to obtain very nearly
uniform distribution of loading applied to the top of the mat as well as to
the bottom through the bearing on the soil. If he does not do this, one

Weak--
concrete
Old drain remo ved
// //
faJ-PARTIAL VERTICAL SECTION Cd)-PLAU OF CORNER OF SUBSTRUCTURE

FIG. 7-12. Unexpected difficulties caused by inadequate preliminary exploration.

part of the mat which is heavily loaded may sag and settle more than
other areas which are more lightly loaded. This may crack the super-
structure, and it may even rupture the mat itself because the mat may
not be sufficiently strong and stiff to spread the unbalanced loads. For
example, if one portion of a building which has a mat foundation is three
stories high whereas the rest is five stories, the former portion may tend
to "hang up" with respect to the latter. Again, a tall central tower may
depress the center of the mat, or a small lean-to may break away from the
main building because it does not settle as much as the heavy part. The
central portion of a large heavy building having uniform weight may also
settle more than the edges because of the greater pressure actually
existing under the former, where the intensity of vertical pressure in the
soil below the mat may be 1.5 times the average pressure and where the
consolidation may vary accordingly.
A surprise developed in the use of a mat with ribs like those of Fig. 7-7
for a boiler house. This may well serve as a warning. One of the rec-
 "

198 MATS [Chap. 7

tangular spaces between the upstanding ribs or walls near the center of the
structure was used as a sump. The engineers started testing the boilers
one winter day when the temperature outside was below freezing. After
the test, they emptied boiling water into the sump. A short time later,
the cold exterior walls opposite the sump were found to be cracked notice-
ably. Apparently the expansion of the central portion of the concrete
forced the cold outer concrete and the steel of the narrow side to crack
because the former acted as a tremendous jack.
When working on the foundation of a power plant or any structure that
has equipment for generating electricity, one should consider the possible
action of the reinforcement in conducting stray currents of electricity.

40'-0"

k
240 max.., k
270 max.,
k 90 k min.
X80 min.
G\ I
i Floor line iEI.108 |
\J
Inside ofpit wali~ Water fable Ei. /OS
PA 'fine sand ,£/_ /o/
Mach'yxi PB
'^SHiyic/ay <EI. 98.5

SECTION A-A Fin e

(b)- '

W,E /.92.5

Loads:
Deep medium
and small
PA =8kmax. 3.5 kmin.
1
yjsdnd
iigravel
PB 10 kmax, 4 k min.
=

*-2'-0"
j

Col. line 5-
UrCol. line 4
faJ-PARTIAL PLAN

FIG. 7-13. Layout showing large pit to be built as part of an industrial plant.

Not only may but the electrolysis may damage the

this cause loss of power,
steel. mat reinforcement and the
Separation of various layers of the
splices of bars by an intervening 2 or 3 in. of concrete may be helpful.
7-4. Mats to resist hydrostatic pressure. In the planning of base-
ments, pits, depressed tanks that may be emptied, and other structures
that are to be watertight, one frequently encounters the problem of
resisting hydrostatic uplift. The soil may be satisfactory sand or gravel
so that the downward loads on columns and walls may be supported by
footings; nevertheless, the water pressure causes upward forces that affect
the floor somewhat as do the soil pressures discussed previously. The
case may be complicated, however, by buoyancy.
Example. To illustrate some of the problems involved in the planning
and detailed designing of such foundations, assume the simplified case
shown in Fig. 7-13. A manufacturer is to build a new plant near tide-
Art. 7-4] MATS TO RESIST HYDROSTATIC PRESSURE 199

water. Under one of the crane aisles he wishes to have a deep pit with
certain machinery in it. The soil, the column loads, the weights of the
machines and their pedestals, and the probable maximum elevation of the
watertable are shown in Sketch (b). How may the pit and the adjacent
column foundations be designed? The safe bearing value of the soil is 3
tons per ft. 2 Assume that pumps and wellpoints are to be available to
lower the watertable temporarily so that construction work may be
carried on "in the dry." Use fs = 18,000 p.s.i.
In attacking this problem, the following assumptions may be made,
basing them upon the situation shown in Fig. 7-13:
1. The bottom and sides of the pit are to be covered with three-ply
membrane waterproofing.
2. Columns G and H will be so close to the pit that their foundations
may be incorporated in the construction of wall CF, or they may be
immediately outside of this wall, whichever seems advisable.
3. When placed below the stratum of silty clay, the footings for columns

J and Kwill be too low to cause harmful effects on the pit construction.
4. The floor slabs surrounding CDEF will be seated upon and tied to

the pit walls in order to act as horizontal beams that will support the tops
of the pit walls against the horizontal pressures produced by the sur-
rounding soil and water.
5. Since the elevation of the watertable is not a completely known and

trustworthy figure, the pit walls and floor will be designed to resist lateral
and upward pressures based upon the tentatively determined maximum
height shown, letting any excess caused by a serious flood be taken care of
by the safety factor provided in the customarily used unit stresses in the
concrete and steel.

6. For resistance to buoyancy, the pit will be designed to resist hydro-

static pressures caused by water up to the floor line, the point of flooding.
The downward weights to be assumed for equipment A and B will be the
weights without the machines themselves. Similarly, the minimum dead-
load reactions only of columns G and H will be relied upon, if they are
supported upon the pit construction. The safety factor to be attained
under this condition need be only slightly greater than unity. Since the
access stairway is to be bolted steelwork, to permit future rearrangement,
its weight is small and will be neglected.

7. The combined horizontal unit pressure caused by the soil alone will

be assumed equal to 30 p.s.f. per ft. of depth; that of saturated soil, 80

p.s.f. per ft. of depth below the watertable.

8. The upward unit hydrostatic pressure will be assumed equal to 63

p.s.f. per ft. of depth below the surface of the water.

To start the solution of this problem, first obtain a general idea of the
seriousness of the buoyancy problem. Assume the side walls to be 12 in.
200 MATS [Chap. 7

thick; the bottom, 12 in. ; the upper floor slab for a 10-ft. width around the
pit walls, 8 in. Columns G and H do not rest upon the pit.
The maximum weight of displaced water when flush with the floor is
found as follows:

Pit = 26 X 30 X 11 X63 540,000

Upper floor = (46 X 50 - 26 X 30)0.67 X 63. . . . 64,000
604,000 lb.

The minimum downward force exclusive of columns G and H is

Bottom = 26 X 30 X 150 1 17,000

Walls = (48 + 60)10 X 150 162,000

Upper floor = (46 X 50 - 26 X 30) -5- 76,000

Machine bases 15,000

370,000 lb.

It would seem that the dead loads of columns G and H may be used to
help counteract the excess of upward over downward forces. However,
if they and the essential masonry under them are included, the center of
gravity of the downward forces will be nearer wall CF than will that of the
upward forces, thus causing a rotation about that edge and a tendency to
lift wall DE. Furthermore, it would be wise to make sure that, after the
pit is built and the pumps removed, a flood cannot cause trouble before
the superstructure is completed.
How can the deficiency of weight be overcome, and what remedy will be
both reliable and economical? Here are a few possibilities:
1. Might it be necessary to increase the thickness of the pit floor to

withstand the bending moments, thus securing more weight automat-

ically? The net upward pressure under the floor for design is

p = (11 _ 3) x 63 - 150 = 350 p.s.f. (approx)

Using Fig. 2 in the Appendix, p' for the 24-ft. direction = 200 p.s.f.;
p" for the 28-ft. direction = 150 p.s.f. (approx),

M 24 = ^—
8
= - ^-^— =
8
14,400 ft.-lb. (counting no end restraint) for

a 12-in. strip
,,
ill 28 =
p"U-
*—r — =
150 X 28 2
- q
— = tl .
14,/ m
, u
00 ft.-lb. (approx),
,

8 8
, 2M = 2 X 14,700 X 12
= . .

f< = l' 350 PSK (rather large)

kjbd* 0-3 X 0.9 X 12 X 92
14,700
As = M. m
.

'
2

fa jd 18,000X0.9X0.75
Art. 7-4] MATS TO RESIST HYDROSTATIC PRESSURE 201

It seems that, from the standpoint of strength, a 12-in. slab is sufficient.

However, it might be made thicker to obtain more weight, but this in-
creases the excavation and the pressure on the invert, as shown in Fig.
7-14(a).
2. The might be thickened as in (6) but this, too, gains only
side walls
about 90 each cubic foot of concrete used under water.
p.s.f. for

3. The pit floor might be cantilevered outward like a footing, as in (c),

so that the weight of earth on top of it may be included. If the soil

weighs 100 p.c.f. and solid stone weighs 165 p.c.f., the volume of solids in
the soil per cubic foot is 100/165 = 0.61 ft.
3
Therefore, the unit weight
considering buoyancy is 100 — 0.61 X 63 = 62 p.c.f. Such a projection
will hit the foundations of columns G and H. The presence of the mem-
brane waterproofing makes it difficult to transmit the net weight of the
earth to the pit construction unless the membrane encloses the projections.
The water must be assumed to penetrate to the outside of the membrane

r Floor line Thickened

^L=T "!_T TL_X

^-Thickened Thickened Extended
(
-*'U"*
Extended
=

1UX
(a) (b) (c) (d) (e)

FIG. 7-14. Methods of increasing weight of pit construction.

and keys, dowels, shelves, and such shear locks or ties may cause
itself,

leaks in themembrane.
4. The main floor around the pit might be thickened, as in (d), but

buoyancy and the fact that at least half the weight of this floor will rest
upon the surrounding soil render this method less economical than some
of the others.
5. Ribs or walls extended from DE to the foundations of columns J and

K in order to utilize their weight would have to be very strong and would
be outside of the waterproofing. This might be a good way to obtain
extra weight if the columns were in line with walls CD and EF. However,
for the situation as it is, this remedy may not be advisable.
6. It is probable that frictional forces and a downward component

caused by the lateral pressures exerted by the soil will offer resistance to
upward displacement of the pit. These, however, are not absloutely
trustworthy in their action, especially because they are applied outside of
the rather frictionless surface of the membrane waterproofing.
The most trustworthy and practicable solution seems to be the fol-
lowing, which is shown in Fig. 7-14(e) and 7-15:
1. Increase the pit floor to a thickness of 15 in.

2. Increase wall CF to 3 ft. 6 in., and use it as a foundation wall for

columns G and H.
202 MATS [Chap. 7

Cantilever the pit floor as a footing beyond sides CD, DE, and EF.
3.

Arrange the details of the footing and wall construction as shown in

4.

(c). Although the excavation is increased somewhat and the water-

proofing of the projecting footings is more troublesome, these matters are
not too serious.

J V . <9 Rein forcement {Wall J-p/y membrane

a-forDL&LL waterproofing
If -.•'.* b- for uplift a-
"ct
:.•„::

D \ c- for uplift U i:
m
C o pa c ted
«:•:'. ^Walls c d- for vert, beam '"'backfill
:

-H":'
-C; e- 1 at corners
-2-0" o f- for upward £ Concrete or
k pressure /brick protection

7"

A 3-6 24 '-0' l^0\ *S "a ?^r

L J J concrete Copper flashing
(iH«
F >H- jEM
rc^- DETAIL OF FOOTING
L :0.:-: .:i>:U

Q s: "

31'- 6" P ^Railing

^ Floor
faJ-PLAN OF PIT J£ .£
rr
I? '»>W>V&>
'Compacfeol -'
=57 w <W
'

M backfill
-3-ply membrane W. P.
-Concrete or bricks
-Wail
f {? ^^concrete
Membrane W. P. (copper
fSheet
needed
if
piling rrfJ-DETAIL AT TOP OF WALL
flashing in bottom corners) for construction

(bh SECTION A-A

FIG. 7-15. Proposed construction for large pit.

An approximate computation of the added weight to resist uplift is the

following, assuming 87 and 62 p.c.f as the
. net unit weights of concrete and
earth, respectively:
Pit floor and footings:

(31.5 X 36 X 1.25 - 26 X 30 X 1)87 = 55,000 lb. (approx)

Wall CFLM:
(3.5 - 1)30 X 10 X 87 = 65,000 lb.

Earth on footing (minimum)

(2 X 31.5 + 30)3 X 9.33 X 62 = 161,000 lb.

Art. 7-6] EXAMPLES OF PLANNING 203

Total added weight = 281,000 lb., which exceeds 604,000 - 370,000 =

234,000 lb., the former deficiency. This is safe without relying upon the
dead load of columns G and H. If the weight on these columns is utilized,
the cantilevering of the footing might be decreased, but the resultant
economy will not be large because one must be careful that the side DE
cannot lift up.
After this general design has been approved, the analysis of it may be
completed by the reader.
7-5. Floating foundations. There are occasionally situations in
which it is inadvisable to apply any major increase of pressure to the soil
at a specific site. One principle that may be utilized, however, in such
cases is that of the "floating" foundation. This means that the estimated
weight of soil, removed permanently in order to build the substructure,
should equal the load to be applied later by the entire structure.* It is
reasonable to expect that serious settlement will not occur unless addi-
tional pressureis applied to the soil, basing this belief upon the supposition

that the has been subjected to the present pressures for so many years
soil

that, for practical purposes, equilibrium has been reached and further
consolidation would be too slow to be important.
For example, assume that the total dead load of a building equals
1,500 p.s.f. over the horizontal projection of the structure. If a base-
ment with some type of mat construction is then built under the entire
structure and if the bottom is set approximately 15 ft. below the surface,
the lower strata should be unaffected. As the excavation is made, there
will be some elastic springback of plastic soils, but this is not likely to
result in serious recompression. This procedure naturally implies that
the center of gravity of the applied loads should coincide closely with that
of the mass of earth excavated. In computing the weight to be supported,
the live load may usually be neglected, except possibly for the inclusion of
25 or 50 per cent of that to be expected in large warehouses.
The planning and detailed designing of the substructure will be similar
to those of other basements supported upon a mat. Waterproofness will
generally be essential.Without this quality, seepage and staining will
probably cause the owner to be dissatisfied with the results even though
the building is structurally safe. The effect of the uplift caused by a
varying elevation of the watertable will seldom be harmful if the depth of
the excavation is determined upon the assumption of a low water level
whereas the detailed design of the structural side walls is based upon the
pressures caused by flood conditions. Since the dead load of the building
is constant, it makes no mat whether the upward pressure
difference to the
is that of the soil and water acting in unison.
alone or of soil

7-6. Examples of planning. The reader has been and will be cau-
tioned repeatedly to look upon a structure as a whole when he plans the
204 MATS [Chap. 7

foundation for it. A minor case that illustrates the psychological side of
such matters is shown in Fig. 7-16. Sketches (a) and (b) show the original
plan for a compartmented concrete bin for the storage of steel and cast-
iron balls for use in the ball mill (grinding equipment) at a metallurgical
plant. It was located where three very heavy building columns came
along one wall and some light columns for an elevated platform rested on
the opposite wall. All six columns were planned to have concrete piers

62-0"
C-i
Col.
line'

Joint
^T>
Joint
-&
Joint
w\
Joint

m S m
raJ-PLAN AT TOP OF BIN

Col.

Dumping,
Col
d
slot
Yard level
x /

^ t
W/W
Bottom
Xj ot bin

Footing i Footing Mat full length

(bh ORIGINAL SECTION C-C fcJ-REVISED SECTION OC
FIG. 7-16. Construction of a concrete bin and column foundations.

that extended from the top of the bin to footings under its bottom.

These piers had keyways to tie them to the bin walls laterally, but they
were to be separated from them by joints as shown so that the walls
would not participate in resisting the column loads, and vice versa.
Referring to Fig. 7-16, notice the following points:
1. Sketch (b) shows the footings for the A -line columns to be under a
part of the bin floor so that the same area of ground is forced to resist both
column and heavy bin loads. The excavation for the footing will almost
inevitably weaken the bearing value of the ground under the bin just be-
yond the edges of the footing.
Art. 7-6] EXAMPLES OF PLANNING 205

2. The joints between the piers and the walls are so deep that freedom
of action is almost impossible to attain without the use of some proper
type of joint filler.

3.The end wall of the bin at column line 38 is not tied to the side walls
to keep them from spreading.
4. The bin will be heavily loaded at some times but empty at others,

and any individual compartment may be filled. Because of the heavy

crane live loads, the reactions on the building columns will also vary
greatly and rapidly. There will be no coordination between these load-
ings. This will tend to injure the structure.
5. The long sides of the bin are cut into isolated sections. At column
lines 36 and 37, the walls act as cantilevers to resist bin loads, and the
walls must steady the piers under the columns.
6. There are very heavy transverse shears on the columns in line A.

That is why they are made so deep. These shears must be resisted by
pieces of the bin walls, an action which is undesirable.
7. The construction work itself will be unnecessarily complicated.
Figure 7-16(c) shows the changes that were made to remedy these
troubles. The bin was designed as a boatlike structure with a heavy mat
under and cantilevered outside of it enough to enable it to spread the
it

loads safely to thesoil no matter what the combination of live loads might

be. The wall along column line A is in effect an elongated foundation for
the heavy columns. The bin is now tied together thoroughly, and con-
struction joints are omitted except for the horizontal ones between the
monolithic pours. The mat was protected by a sand cushion above it.
Nevertheless, it was made excessively thick to prevent serious conse-
quences from wear. The structure now acted as a unit, as it would try
to do anyway.
Figure 7-17 shows the general dimensions of, and the loads caused by,
a coal-storage installation near the boiler house of an industrial plant in
Indiana. It also contains the log of a boring taken at the center of the
proposed construction. A foundation is to be designed for these steel
silos.

Notice that a conveyor trench and a pit for the bucket elevator are to
go between the silos. The bottom of each silo is to be a concrete slab
supported upon a sloping earth fill in order to avoid the fire hazard from
dead storage of combustible coal below the drawoff position. The silos
must be able to withstand a heavy gale when empty or full.
The main clay stratum is moderately good. Since the silos may be
almost fully loaded for long periods, the allowable bearing on this clay
was with 3 tons maximum unit edge pressure from
set at 2 tons per ft.
2
,

overturning and vertical loads. Tilting and considerable settlement

must be prevented. The watertable was not definitely determined.
 :

206 MATS [Chap. 7

Conveyor fo bo/'/er room

rY Concrete french
Data:
Sfbs /S-0 "i. D, 18-6 "o. D.
:

Weights in tb
Silos + snow = 298000
Coal -1252,000
Stoping floor - 54,000
Fill'under floor = 416,000
Elevator Equipment, etc. = 16, 000
Wina load:
1

& track 40 lb. per sq.ft. of

Concrete vertical projection
trench
fttJ-PLAN OF SILOS
El. 26.4
/ o
If///® J 12

/ k Claywith 7
somez-
^SfChute •sand* 7 6

i/F y\iS/ope4: 12
— 9

/K
i

r £' :
]
1
s
/ ^N \>
/\

Clay
7 S
fe

I
1 k with Q
£ some 8 £
K 5 traces £
45 -M •52 of Qi

sand
7
I8L6"0.D.
1 i ".j '

12-0 1
12-0'^ to
5
^ \
& *
Ci
Qq
ta m
\ iJ

^i \
\
! \
LJ

/
ri
=>iifrdi
41
\
9

///w/v /A*&v l///^)«A^

> <£.. .*Ele i^/cr p/t
Sandy
clay
IS

*°1

\6 ¥
(c)- LOG OF BORING AT
(b)- ELEVA1"ION OF SILOS fe OFS TRU ZT JRE

FIG. 7-17. General design for two coal silos.

However, little trouble from water is expected since most of the water is

surface runoff or held by the relatively impermeable clay.

One suggested solution is the use of a 3-ft. ring of concrete 5 ft. deep,
centered under the silo walls, as pictured in Fig. 7-18. A second is the
scheme shown in Fig. 7-19. A third design is that pictured in Fig. 7-20.
Art. 7-6] EXAMPLES OF PLANNING 207

Which should be used? After examining these three proposals, the fol-
lowing comments should be studied:
Figure 7-18. 1. The weight of the silo would rest upon the wall.

What would hold the weight of the coal in it? Generally, this would be
transmitted through the floor and fill to the ground and to the top of the
inner portion of the ring. However, if the coal tends to arch and hang up
during drawoff, much weight may be transmitted by friction into the

Silo wall,

(a)-PLkH OF RING FOUNDATION

L o ads Loads .,

q
Fill
Clay iisandj-^

Clay I

"^^UiU •rttftNj ttftffi LlU^'

4i
'^Overlap \
2\/dfl>2lirjes K

CcJ-ASSUMED GENERAL CHARACTER OF PRESSURE DIAGRAM

AT SECTION B-B FOR VARIOUS DEPTHS

FIG. 7-18. One study of foundations for two coal silos.

structure and thence to the ring. The latter should be able to support a
large part of the total weight if necessary. This would cause a larger soil
pressure than allowable.
2. The ring walls have to be notched out at the conveyor trench and
pit, causing local weakness.
3. The pit is deeper than the ring wall. This is not desirable.
4. If the trench-and-pit construction is not isolated from the rings, the
former will tend to transmit unintended loads to the soil, and it may
break off or unbalance the resistance.
 —

208 MATS [Chap. 7

5. There is a question as to whether the permanent dead load is suffi-

cient to prevent the empty from tipping over in the wind.

silos
6. On the whole, this foundation seems to be inadequate and unde-
sirable.
Figure 7-19. 1. The is to make the rinjs thin but
idea in this design
deep, thus clearing the trench and engaging considerable earth to
pit,
resist uplift, and transmitting the load to deeper stronger soil. The last
does not appear to be accomplished because of the plastic nature of the
clay. The deep overburden is not so beneficial with such material as it

would be with sand.

te;-PLAN OF TOP OF DEEP RING FOUNDATION

yil-WHWWH V/AVV/,m
Fill 1--
C'iay'joiha's'and\

ESSl
C/ay i~'
:
7] C ..1
m
£fg] Ring
footing
(b)- SECTION A-A
FIG. 7-19. Study of deep ring foundation for two coal silos

2. The excavation and construction are considerably more costly.

There is no necessity for placing the footings at this depth.
In general, the distribution of the pressure on the underlying clay
3.

may be somewhat like that shown in Sketch 7- 18(c). Compaction of the

deeper clay is therefore likely to cause the silos to tilt toward each other
something that should not be permitted.
4. The pressure from the contents of the silos may tend to load the top

of the inward projection of the footings and cause a twisting tendency

because this weight is offset from the reaction of the soil under the footing
at any given section.
Figure 7-20. 1. In this design, the bottom of the elevator pit is made
the position of the top of a continuous mat extending under both silos as
Art. 7-6] EXAMPLES OF PLANNING 209

well as under the trench and pit. Whether the load is applied mostly
through the fill under the drawoff floor, through the silo walls, or through
both in varying amounts, the mat will distribute it over a large area at
the top of the thick stratum of clay. As this area is considerably larger
than that of the silos, the intensity of pressure is low.
2. The corners of the mat are nipped off, and the central portion is de-

liberately enlarged, in order to cause a tendency to produce tension in the

top of the mat. This may counteract somewhat the tendency to pro-
duce tension in the bottom of the mat that will be caused by larger
settlement of deep portions of the clay under the central region. Since

•Ik?" S>
foJ-PLAN

Ring wall
rMat.
|^>/\o: •;;». /« o'-;
;\o::o:
»:-::'.-'--a::-?.-0.:.
m
(b)-SZCHObl A-A

FIG. 7-20. Mat foundation for two coal silos.

the top of the clay has a rather low unit pressure, the lower layers will
have still less, and the probable settlement will be smaller than for the
other designs.
3. The trench and pit walls and the cross rib are used as stiffeners for
the mat. This foundation will tend to resist relative tilting of the silos.
4. The base and fill on it provide adequate weight to resist overturning

of the silos and prevent large edge pressures.

5. This plan may require more concrete than the others, but its total

cost is probably less than that of Fig. 7-19, and it is inherently much more
suitable for the job.
6. This plan should now be analyzed to see that it meets all require-
ments and to determine the reinforcement.
210 MATS [Chap. 7

PROBLEMS
7-1. Design a waterproofed continuous mat as a basement floor for a building
similar to that pictured in Fig. 7-1 for the following conditions: The soil is soft

clay; the basement floor is 6below the ground and 9 ft. below the first floor;
ft.

the bays are 20 ft. square; Pi = P h = P 6 - Pi = 300 kips, P 2 = P, = 400 kips,

Pz = 450 kips, and the corner column = 250 kips; no ribs or walls are to project
inside the outer foundation walls.
7-2. Design a waterproofed mat for the conditions of Prob. 7-1 except for the
addition of reinforced-concrete walls along the column lines in one direction only.
Assume a 4-ft. doorway in the center of alternate bays along each wall, the center
of the first door being 10 ft. from the outer foundation wall. Is this arrangement

as economical and as safe as that designed for Prob. 7-1?

B C

Inside
A ofp/'t —*- A
J
A !6'-0" D
foj-PLAN AT TOP

Floor line
j K ^V*|l-i ,li7>T /g>':
.

:Clciy:idarn
^
^Sanely clay
*X

F
Em
/^-POSSIBLE SECTION A-A •
-£*nj!Y ^
'SSi Q R W3M

M is:mm AT

rc^-ALTERNATE SECTION A-A

FIG. 7-21. Design for a deep pit.

7-3. Design a ribbed-mat foundation similar in type to that of Fig. 7-9 for the
following conditions: There are five rows of columns 18 ft. on centers in one direc-

tion and six rows of columns 20 ft. on centers at 90° thereto; the soil is medium
 PROBLEMS 211

silt for 10 ft. below the surface, 18 ft. of soft clay below that, then 12 ft. more of
silt underlain by deep fine sand; there is to be no basement; the floor is to be 2 ft.

above the adjacent ground; the assumed loads on each of the four corner columns
= 200 kips; the other outer columns = 250 kips each; all interior columns =
350 kips apiece; the "cells" are to be backfilled with compacted gravel; an 8-in.
reinforced-concrete floor is to cover all the space inside the outer 12-in. brick walls.
What is the increase of unit pressure applied to the silt? Assume that the loca-
tion is near Norfolk, Va.
sump 18 ft. deep that is to be built
7-4. Figure 7-21 (a) shows the plan of a pit or
below the floor of a newThe soil profile is given in (b). The
industrial building.
column foundations are not shown because they are at least 25 ft. from the pit, and
therefore will not complicate the problem. Develop a plan for this pit, and show
how it may be constructed.

El. 103,

El. I00
\ Fin
Wafer table El. WIS max.
Sofi red clay

5 10 15 20 o Sand and ''

°

Scale
.i

in feet
small gravel o

wmmm Rock
FIG. 7-22. Soil conditions at site of a settling tank and sump

Suggestions: One possible type of structure is pictured in (6); another, in (c).

(1) Make a design that will resist buoyancy when the watertable is as shown;
(2) analyze the sides and bottom, and plan the reinforcement; (3) plan the details
of the membrane waterproofing; (4) plan the construction methods. If steel sheet
piling is driven around the excavation and well into the sandy clay, will the latter
prevent "blowing of the bottom when the excavation is unwatered?
up"
7-5. The tank pictured in Fig. 7-22 is one of three proposed for the
settling
conditions and in the position shown; the one sump will serve all three tanks.
The tanks are to be on an extensive, low area near a river, and the flood waters
have been known to reach El. 101.5. Furthermore, any tank may be empty at
the time of high water. Design the structures for the climatic conditions at
Buffalo, N.Y.
Suggestions: Some details of the construction of such a tank are shown in Fig.
7-23. Can the buoyancy problem be overcome economically? Is it advisable to
increase the over the general area in order to raise the tanks to a higher eleva-
fill

tion? tank so that the top of the footing of the ring wall is
Is it safe to raise a
level with the ground? Will frost ruin a tank if the latter is emptied during
winter weather? Can an automatic check valve or other flooding device be
212 MATS [Chap. 7

.X-Rinq wall

Sg^6
(a )- SECTION AT RING WALL

W-BASE OF CENTER PIER

r Ji.A

frfJ-CONSTRUCTION JOINT feJ-PLAN OF SLAB REINFORCEMENT

fcJ-PLAN OF
CONSTRUCTION JOINTS

FIG. 7-23. Some details of construction of settling tank.

depended upon in case of floods at any and all seasons? If the bottom of the
sump is set upon the rock, will this eliminate buoyancy acting upon the sump?
Do the tanks have to be waterproofed? Does the sump have to be waterproofed
when the pumps and motors are supported upon its bottom? A good answer to
each of these questions may suggest a solution for the problem.
 8
FOUNDATIONS SUBJECTED TO
OVERTURNING FORCES

8-1. Introduction. Most foundations are subjected to lateral forces

in addition to vertical ones. In some cases, these lateral forces are im-
portant and should be considered in the design; in others, theymay prop-
erly be "absorbed" by the safety factor. In no case, however, should
they be disregarded as unimportant until the designer has satisfied him-
self that this may be done safely.

This chapter deals with some typical problems that are frequently en-
countered in the design of ordinary foundations. The principles illus-
trated are applicable to a wide range of engineering work; they will assist
the reader to make sure that the foundations he designs will not slide,
tip over, or be pulled out of the ground.
Critical conditions may be caused by wind on tall narrow light struc-
tures; by lateral loads from cranes and rapidly moving vehicles; by the
unbalanced pull of cables and such equipment as belt conveyors; by
pressures produced by earth, other granular materials, and water; by large
eccentric loads;and by earthquakes.
The foundations discussed in this chapter are supposed to have footings
or bases that rest directly upon the ground. Pile foundations are con-
sidered in Chap. 10.
The overturning effect caused by earthquakes is difficult to predict with
any certainty. Local building codes (as for Los Angeles, Calif.) may
specify the magnitude of the horizontal force to assume. This is usually
expressed as some portion of the weight of the structure, such as 0.1 or 0.2,
and it is often stated as an acceleration of 0.1<7 or 0.2g. This force is
assumed to be applied at the center of gravity of the mass of the structure.
In general, structures with foundations on rock are less seriously endan-
gered by earthquakes than are those that rest on deep granular or alluvial
deposits which seem to magnify the vibrations somewhat as would a
gelatinous mass.
213
214 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

For the purpose of analysis, overturning is considered in this chapter

without giving special consideration as to how
the magnitudes of the
forces and moments are obtained.
8-2. Connections at bases of steel columns. Consider the action
of the base of a steel column that is subjected to an overturning moment.
Some details for such steelwork are pictured in Fig. 8-1. The following
comments refer to the correspondingly lettered sketches:
This shows the assumed condition greatly simplified. This situa-
(a)
tion might occur in the case of a side-wall column of a tall mill building in
which the frame action of each bent (columns and trusses in one row
across the building) is to resist the wind pressure and other lateral forces
and in which the bottoms of the columns are to be strongly restrained.
The vertical load P and the bending moment M
are to be resisted by the
foundation, but the first requirement is the provision of sufficient strength
and stiffness in the steelwork to transmit these to the concrete. The
column shaft must be restrained directly, or indirectly through the base
plate. It is customary to depend upon anchor bolts as the direct con-
nection for resisting any tension caused by overturning. Columns may
be embedded in the concrete, but this is seldom effective in resisting
bending unless the embedment is 3 or 4 ft. deep, and the concrete is
designed to provide the necessary reactions.
(6) A upon a base
milled column shaft resting directly and squarely
some resistance to rotation because of the "riding" of the
plate will offer
leeward flange B as shown here. The greatest resistance, however, is
limited to the moment Pd/2, in which P should be assumed equal to its
—
minimum value the dead load only. This resisting moment is usually
too small and too questionable to be trusted completely unless P is very
large compared to M.
(c) Here base angles are connected to the column and the anchor bolts,

the base plate being for the distribution of pressures only. Unless angle
C is very stiff, it will bend as shown to exaggerated scale, it will yield
excessively, and its distortion may cause the top of the anchor bolts to
bend or break. It is obvious that these bolts must provide any necessary
downward reaction T, and that the pressure P -f- T tends to be concen-
trated near the leeward flange. make the connections
It is desirable to
strong enough so that M can be resisted by the moment TL unless the
moment Pd/2 is relatively large and trustworthy, in which case one may
assume M = TL -f- Pd/2. However, the magnitude of thewind pressure
is uncertain. It may be assumed to equal 20 p.s.f. upon the
acting
vertical projection of buildings under 50 ft. high; 30 p.s.f. on taller
structures; and sometimes 40 p.s.f. in regions subject to hurricanes
and cyclones. The local effect of gusts is anybody's guess. In any
case, one can easily imagine the concentration of pressure in the steel-
 yd j&/n6uvu±

Z3E ~P

; h !''. t b
ITS

LAJ
"®1
© .5ft

IV1

215
216 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

work and masonry in the vicinity of D, and this pressure should be

provided for.
(d) This sketch shows the ineffectiveness of web angles on the column,
because of both the short lever arm L and the warping of the outstanding
legs of the angles. When the wind is at 90° to the web of the column,
these angles are also relatively weak.
This pictures a base plate welded to the column shaft. If the plate
(e)

is thick enough,if the welds are adequate, and if the concrete can resist

safely, this arrangement may provide a greater lever arm L and a more
effective resistance. However, this involves field welding or difficulties
in setting the column, as will be explained in connection with grouting.
(/) This illustration shows one arrangement for "boots" at a column
base to prevent weakness caused by the yielding of details. Angles G are
connected by rivets or welds to the billet and do little except transmit
shearing forces when necessary. The base angles J should be milled with
the column shaft after fabrication, or be set very carefully and made flush
with the milled end of the column. Angles H
serve as a shelf for the
washers and nuts of the anchor bolts, but their resistance depends mostly
upon the stiffeners K. Not only may L be increased in this manner, but
reliability and stiffness are secured. However, if P and M are large, there
may be an unsafe concentration of pressure at the outer tips of angles K
on the leeward side.
(g) This is a minor modification of Sketch (/). It is arranged for
welded construction and to avoid thin outstanding material at the point
of maximum compression at the leeward edge. The welded connection
of angles N or the compressive stress at the corner of these parts may then
be the critical feature. All the horizontal shear must be resisted by the
web angles and friction because the anchor bolts stand free of angles N.
Of course, the bolts and steelwork inside of these angles cannot be painted
after erection.
(h) Here the column is set so close to the face of the exterior wall that
one pair of anchor bolts placed inside the outer flange.
is It is obvious
that the lever arm anchor bolts for resistance against tipping is
of the
probably less the overturning is as shown than when the wind acts
when
in the opposite direction. The flange plates Q may be added when more
metal is needed in bearing than is available in the column flange alone,
especially when the column tends to rotate about the tips of the flanges,
90° to the direction shown here. When d is large and when obstructions
outside of the column are objectionable, both sets of anchor bolts may be
placed inside the flanges.
This pictures a case where the wind is acting perpendicular to the
(i)

column web. In some cases, it may be desirable to flare the reinforcing

plates Q to provide a greater lever arm L. The angles J should be
Art. 8-3] ANCHOR BOLTS 217

extended, too, and both J and Q should be thick and milled to bear.
Even so, the edge R is likely to yield under the concentrated compression
if the anchor bolts are stretched appreciably. Sometimes, as for canti-
levered poles supporting electric lines, the entire base may be widened so
that Q, H, J, and K may accommodate three or four anchor bolts in a
row.
(j) Here a heavy column isshown with channel cover plates and
stiffening plates S so that large overturning moments may be resisted in
any direction. Plates S are to be thick enough to enable the compression
to act safely at the edge,and the channel flanges should enable the section
to resist local buckling. This sketch shows triangular welded plates to
stiffen the base angles. These plates are helpful but are not a complete
substitute for the much stronger "boots."
8-3. Anchor If an anchor bolt is to resist tension, its connec-
bolts.
tion to the concrete must be adequate. A plain straight rod embedded
in concrete may not develop adequate bond to prevent the bolt from
pulling out, or even from twisting when one tries to loosen a badly rusted
nut. When an anchor bolt is not completely embedded in concrete or
masonry, the exposed end may corrode badly, hence the critical section at
the root of the thread should be designed so that the computed maximum
tension is only 12,000 to 16,000 p.s.i. A strong, plate washer is generally
desirable under the nut; and double nuts are sometimes used when it is

advisable to "lock" them in place.

Shouldered bolts are used occasionally
w hen tension is to be resisted, but the connection must be loose enough
T

to permit a slight lateral movement of the connected part. One should

not rely upon a workman's "tightening a nut lightly" so -that some
shearing movement can take place; he will generally draw it up tightly.
Hexagonal nuts are preferable to square ones because they can be engaged
by a wrench more easily in narrow spaces.
Referring to Fig. 8-2, the following comments are given as a general
guide:
(a) This bolt is "hooked" to increase its anchorage and to prevent
turning. The projection should be 4 to 6 diameters. The embedment L
should preferably be at least 40 diameters, or more for machinery subject
to sudden or repeated tensile loads.
Plain rods are almost necessarily used when making anchor bolts.
(b)

However, such a rod may be deformed cold or when heated to make a

swedged bolt. Such a type is useful when anchor bolts are to be grouted
into holes drilled in rock or concrete. Because of the danger of trapped
air pockets that prevent thorough grouting, it is often desirable to fill the
hole with grout first, then to force the anchor bolt into position. Lead,
or iron filings mixed with sulphur, may be used in lieu of portland-cement
grout. Another danger to be guarded against in such drilled anchorages
 Pirn
fe-.... % )

^
i

Evs£ 8*
xa
3Ad3/S 3dIJ

KHH]
©Jdl/SVM
9p/SUJ

^ ID
t— <
LL_ O
' ' ti-i

pm~
EE? -*~l
^^
o <->

UdlfSVM 34&/d>

I'-'.'
"'.' '"" sgggBS

nnnni

aj"
dBpaM uspooM o^

. yoo/ a/ 'sjnu % 9S/y 9A39/S- Qd/d-t

I ^7
P/S^
218
Art. 8-3] ANCHOR BOLTS 219

is the presence of seams or cracks that limit the resistance of the anchor
bolt to the weight of a relatively small volume of rock. The diameter of
the drilled hole should be at least 2^2 m - or 3 to 4 diameters, and its depth
should be 40 to 60 diameters.
(c) This hacked bolt is for service similar to that shown in (6). How-
ever, the former is generally easier to manufacture.
(d) Seldom are anchor and short that it is feasible to use
bolts so small
regular formed heads on them. This drawing shows the substitution of a
rod threaded at both ends. The washer is to resist pulling out, the nut is
to bear against the washer, and the welding is to prevent turning of the
rod. This bolt is generally more costly than (a).
(e) A
threaded rod bent into a U may seem to be desirable when two
bolts are to be near together. It is difficult, however, to fabricate it
exactly right, and deep embedment is still needed if the concrete or rein-
forcement is to develop sufficient tensile resistance.
(/) Two rods may be attached to a large common anchor as shown.
Welding is desirable here as in (d). In all cases, the threaded length and
the projection above the concrete should be adequate. Here t\ is the
allowance for grouting; t2, for the connected steelwork; and x, for washers
and nuts plus some leeway.
(g) In the case of machines, the anchor bolts may have to be re-
movable. Here a threaded rod is extended through a pipe sleeve to a
niche where a washer and nut can be attached. When machines are
anchored to floors or beams, the pipes may go clear through the concrete
so that the rods may be inserted from above or below. If this is not

done, be sure that the bolts can be inserted from above after erection of
the machine. Another type of removable anchor bolt is one with a short
head similar to a hammer. This may be placed through a slot in an
embedded insert, then rotated 90°, and tightened to hold it in position.
(h) There is an old saying that "if the anchor bolts are not set correctly,

the structure will not fit." This sketch shows a rod with a 180° hook
—
seldom necessary and a pipe sleeve. The latter is to permit a moderate
amount of springing of the bolt to fit the holes in the attached base.
The washer is to close the bottom of the pipe against the entrance of wet
concrete during pouring; the welding is to hold the washer and sleeve in
proper position; the H
_m projection is to raise the pipe sleeve above the
-

level of the poured concrete but it should be considerably less than the
allowance for grouting. The washer and nut may be used temporarily to
assist inkeeping out the concrete. The effective anchorage of the bolt is
only the part below the pipe sleeve. Cardboard tubing and wrappings
of tar paper are generally poor substitutes for pipe sleeves, except in such
a case as that in (g). However, conical sleeves of sheet metal fitting
tightly at the bottom around the anchor bolts are very useful and econom-
220 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

ical. The top of the inverted cones can be closed with temporary annular
wooden plugs to keep out the wet concrete during pouring.
(i) This shows other ways of closing the ends of pipe sleeves. Waste
is obviously weak and unreliable, and wooden washers may drop off when

the concrete is compacted. The third illustration shows an arrangement

for anchor bolts that must necessarily be relatively short. Here the large
plate washer must support the pull in the bolt except for whatever bond
resistance may be developed on the outside of the pipe. If the nut is

not welded to the rod, this bolt may be unscrewed, removed, and then
replaced. However, if the threads become damaged, these operations
cannot be continued, and remedial measures are difficult.
(j) This sketch pictures the absurd results that may be obtained if one
pursues too far the philosophy, "if a little of something is good, more of it
is still better." To be specific, long pipe sleeves permit one to spring an
anchor bolt sidewise more easily than do short sleeves. However, when
the depth of embedment is small and limited, the anchorage of the bolt
itself may be ineffective. As shown here, the hook cannot develop the
proper anchorage because its strength as a cantilever is relatively small,
and the length is too little to develop appreciable bond resistance.
(k) These drawings show a few details for use when the anchor bolts for
machines are attached directly to steelwork. In the first illustration, field
welding is avoided by inserting the threaded rod upward through the hole
in the shelf angle, then dropping the pipe sleeve and the centralizing
washer over the bolt, and finally tightening the nut to hold the assembly
during the concreting. The cross rod at the bottom is to prevent twisting
of the bolt when the machinery is erected or removed. The nut and wide
washer at the bottom of the second sketch should be shop-welded to the
bolt, then the washer may be field-welded on top because bottom welding
is more difficult. When the steel is bare, some physical connection is
needed to hold up the bolt except when it is supposed to be removable; it
may drop out or be knocked out when the machine or steelwork is placed
over its top. In the third sketch, the two nuts are supposed to lock the
bolt in position. This principle may be used with various modifications
when the bolt must be erected from above.
(I) It is generally essential to use some sort of template of wood or steel

to hold anchor bolts in the correct positions vertically and horizontally

during the placing of concrete. This sketch illustrates a simple method
for supporting the anchor bolts in a pedestal thac is to be under a steel
column. When planning the structural details, one should endeavor to
arrange a group of four anchor bolts so that they are spaced equally in
both directions or so that the inequality is obvious. For example, if the
spacing in one direction is ^ in. less than in the other, the erectors may
easily make the mistake of rotating the group 90° from the position
Art. 8-3] ANCHOR BOLTS 221

intended. Furthermore, varieties of pattern and spacing should be

minimized. There have been cases in which the wrong group of bolts was
put in a particular place, necessitating chipping of concrete, bending or
burning of bolts, welding of offset extensions, drilling of holes and inser-
tion of expansion bolts, or even the omission of some or all of the intended
connection to anchor bolts. In any case, templates for erection should be
strong, easily placed, well supported, undisturbed by concreting, and such
that the bolts can be located properly by the surveyors or others. If one
can be sure that the anchor bolts are set accurately, the need for trouble-
some pipe sleeves will be eliminated because the shopwork done on the

^S %
{Anchor bo/f 03\ \ i
K'
'

§ <*•

cvt ."^s
cv
rij

k SS 1 L_
&
:;5;!o
r
V'\
o ...» ••..

(a)- PLAN OF BOLTS ^-DETAIL OF ANCHOR BOLT

FIG. 8-3. Details of anchor bolts that pulled out.

(aJ-PLAN AT TOP OF PIER W- SECTION A-A

FIG. 8-4. A pivot pier that failed because the anchor bolts were not embedded properly.

metallic parts when drilling the holes in them is generally trustworthy.

The pipe sleeves are to allow for inaccurate fieldwork.
Figure 8-3 shows the general conditions prevailing in one case where the
anchor bolts at the head end of a heavy metallic conveyor pulled out.
The tractive force caused an upward pull on bolts A and the first time ;

that the machine was operated under full load, these bolts gave way.
Holes were then drilled through the reinforced-concrete beams under the
bearings, and through bolts were installed —
all at considerable expense

and loss of production.

Another unfortunate case is pictured in Fig. 8-4. This illustrates the
trouble that may result from a "little" knowledge. A 5-ft. cylindrical
pier was to support the machinery to operate the rotating equipment in a
100-ft. -diameter settling tank. The equipment was to be attached to the
222 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

top of the ring of concrete. The designer realized that a construction

joint at A was essential. He thought that it would be economical to pour
the main shaft and then place the anchor bolts on top of this concrete
before pouring the ring. (This was contrary to specific orders.) When
the machinery was started, the large horizontal torque sheared off the
ring at A. The ends of the seven vertical bars b were hopelessly inade-
quate to "tie" the ring in place, and the circumferential bar a was, of
course, merely a spacer. The final "economy" was the cost of complete
dismantling of the equipment, cutting off 3 ft. 6 in. of the top of the pier,
the erection of long anchor bolts, an increased amount of reinforcement,
the pouring of more concrete, and a serious delay in starting production.

/; -
* Weld "I
SS^gL Base pi. g%%
•Jfc*3
fcs
•d b r
Grout

''Anchor bolt P/'pe

sleeve
(a)

Base pi
Grout- Fx"~\ TOP Of :-

concrete

-\H

(d)

FIG. 8-5. Shearing resistance of anchor bolts.

Furthermore, the same work had to be done to a duplicate tank alongside

the one that failed.
If the space around an anchor bolt in a pipe sleeve empty, water
is left

may enter and cause some corrosion and, if a structure exposed to freez-
is

ing, cracking of the concrete may result. It is desirable under such con-
ditions to fill the sleeve with grout, as pictured in Fig. 8-2 (j).
Anchor bolts that must resist large transverse shearing forces require
special attention by the designer. In most ordinary structures, the ver-
tical loads are likely to be so large that frictional resistance under pressure
is withstand all probable lateral forces. In the case of
sufficient to
machines and some special structures, large transverse shears may be
applied to the anchor bolts.
As shown in Fig. 8-5 (a), the holes in the steelwork are usually made
/4 to 3^ in. larger than the anchor bolts. If the structure or frame is
fastened to a base plate by some adequate connection A, the shear S will
be transferred to this plate without depending upon the anchor bolt.
Art. 8-3] ANCHOR BOLTS 223

Otherwise, slippage may occur either along surface B or along C until the
space D is closed and the bolt offers resistance. However, this resistance
is be dependent upon the local crushing strength of the grout or
likely to
concrete at F. When the bolt is in an empty pipe sleeve, the anchor bolt
will not develop effective resistance until the slippage occurs and the
space G in Sketch (b) is closed. Bearing against the threads of an anchor
bolt will damage them, but this weakness is generally inconsequential.
The pipe sleeve and spaces D and E, Sketch (a), may be filled with grout,
sulphur and iron filings, lead, or any material that may be packed in and
will ultimately have sufficient strength in bearing.
./ in Sketch (c)
It is possible, in special cases, to use small angles like
with holes H
having a clearance of only 3^6 m
The bolts may bear -

against the steel at H, and this angle will bear against a sufficient area of
concrete. It is also possible to set the bolts in a permanent embedded
frame. However, one should be sure that the frame is properly encased.
For example, if a channel K with its flanges turned down is substituted for
angle ./, as shown in Sketch (d), the lower portion L may not be filled
with concrete. Details utilizing the principles shown in Fig. 8-7 (d) may
also be employed to lock the steelwork and concrete together mechani-
cally. In some cases a diagonal bracing member attached close to the
base of a column may cause a large lateral shearing force. This may be
resisted by welding the shaft to the base plate or by connecting these
parts by a pair of strong web angles and by having a cross made up of
bars welded on the bottom of the base plate as indicated in Fig. 8-7 (d).
These bars then serve as lugs, and the recessing of the grout provides a
mechanical lock so that the grout cannot move laterally.
When pipe sleeves are used in a foundation, they tend to cause planes of
weakness, especially if two or more are close together. This is especially
serious in the case of vibrating machinery and when there are thrusts or
shocks that cause the anchor bolts to bear upon the inside of the sleeves.
The concrete around the anchor bolts should be well reinforced. It is
desirable to tie the pipes in one row together, and to tie those on two oppo-
site sides of a pedestal, by using "hairpin" bars looped around the sleeves.
Three foundations for large machines cracked along the sleeves because
this reinforcement was too light. Two rows of hoops around the top of a
pedestal were insufficient; three or four rows with at least two rows of
intermediate hairpins at the pipes probably would have been satisfactory.
Expansion bolts such as those pictured in Fig. 8-6 are useful when the
attachments of machines or steelwork must be made to old concrete, and
when the locations of anchor bolts cannot be ascertained before new con-
crete work is done. The following comments should be considered:
1. Expansion bolts may be able to resist a moderate tension if the con-

crete will withstand the pull —

sometimes a questionable matter.
224 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

2. They are generally weak in resistance to transverse shear because

their tops are not supported laterally unless well grouted.
3. Usually, the steelwork must be set, the holes located, the steel
removed, the holes drilled, the bolts installed, and the steelwork reerected.
There must be space for performing these operations.
4. The cost of labor for fieldwork may exceed that of setting standard

anchor bolts.
5. Some types of expansion bolts may be removed an asset in the case —
of machinery.
6. Drilling of holes and the wedging action of expansion bolts are likely

to crack or spall narrow concrete members, such as curbing and the ring
shown in Fig. 8-4. The same results may occur if the bolts are near the
edge of wide concrete sections.

Expansive sleeve
of soft metal
rm .

1 1

'

~» : Uj^J
<y
^Casf iron
(a)-EXPANSIVE SCREW ^-EXPANSION UNIT WITH (c)-lYJO EXPANSION ^-EXPANSION SCREW
anchor nuf
ANCHOR WITH A A PLAIN SLEEVE AND UNITS WITH TWO WITH A FIBRE ANCHOR
THREADED ANCHOR NUT PERMANENT BOLT PLAIN SLEEVES AND
A PERMANENT BOLT
Ackerman -Johnson Co. The Rawlplug Co.

FIG. 8-6. A few types of expansion bolts.

7. If the drill strikes a reinforcing bar near the surface of the concrete,
relocation of the bolt may
be necessary. Therefore, the reinforcement
should be planned to expedite the installation of the expansion bolts.
Near edges or corners, the bars should be arranged to tie the concrete
together. They may prevent serious spalling but are not likely to stop
the cracking of the concrete.
8. Expansion bolts are generally small in size and should seldom be used
when the tensile forces are large. Occasionally it may be possible to leave
wedge-shaped pockets in large concrete bases of machines so that the
equipment may be set, the bolts placed, and the pockets filled with con-
crete to lock the bolts in position.
It is obvious that reinforced-concrete columns or piers subject to over-
turning moments, as in Fig. 8-10, may and should be tied into their foun-
dations by means of dowels or extensions of the main reinforcement.
Great refinement of fieldwork, however, is seldom necessary.
8-4. Grouting. The use of portland-cement grout and mortar under
bearings has already been referred to. One can easily realize that work-
 —

Art. 8-4] GROUTING 225

men should not be expected to finish the top of a concrete foundation to

the exact desired elevation and to a true level plane surface. Wet shrink-
age of concrete, troweling around anchor bolts, the formation of laitance,
vibration of neighboring concrete, hasty surveys to determine elevation
and true level — these make accurate concrete work difficult and almost
unobtainable. For minor parts, it may be satisfactory to set steelwork
directly upon a finished concrete surface with a thin intervening layer of
dry cement or a paint coat of neat cement grout. It is possible but —
—
inadvisable to pour the concrete slightly high and then grind or bush-
hammer it down to the desired elevation. Accurate workmanship, how-
ever, is next to impossible amidst anchor bolts, and the cost is beyond all
reasonableness.

Billet^ I
(A Sfeel wedge
!

ill-
fTT
jzt
;?:;. |
Pedestal \

j
°.

v. '
i
i
anchor bolts ->i I
.'

fa>BiLLETAND PEDESTAL FOR COLUMN ^-ORDINARY GROUTING

n~~
E
Sfl» Mortar'-:-,
5
±
Mortar
%
I

te II
\

fcJ-ADJUSTING SCREW foW-RECESSED PEDESTAL

FOR DRY PACKING

FIG. 8-7. Arrangements for grouting a heavy base.

A simple case of grouting is illustrated in Fig. 8-7. A

large column
generally requires a heavy base plate, as shown in (a). Such a billet
cannot be placed by hand. Angles A are useful for the attachment of a
chain or hook so that a crane can lift the plate into position over the
anchor bolts. The some predetermined allow-
bolts should be set with
ance, t\, word being synonymous with mortaring). A few
for grouting (this
of the procedures and possibilities of grouting are shown by the following
explanations of the sketches in Fig. 8-7:
(a) A portland-cement mortar (1 1^ or 1:2) might be spread over the :

top of the pedestal, then the base plate might be set and forced down to
proper grade and level. Of course, this method is inadvisable because it
is difficult to secure sufficient accuracy of erection. It is also apparent
that the erection of a tall column with the base plate attached would be far
226 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

more difficult because the column would have to be steadied laterally for a
day or two, and the great weight would probably squeeze out the mortar
to an unpredictable extent. In any case, some reliable control is needed.
It is practicable to set J^-in. leveling plates and to grout them first, then
to erect the columns with the billets welded to them after the grout
has set.

(b) Here the base plate is supported upon

steel wedges that permit
proper placement of the billet. A
form is built around the base, and a
"wet" mortar is placed around and under the latter. There is likely to
be an air pocket near the center so that the plate is not properly sup-
ported. up the plate and fill any low places with mor-
It is possible to lift
tar, then to replace the billet. This extra work costs money, and it may
be difficult to secure proper contact with the wedges the second time. If
the plate is adjusted and removed, the mortar placed, and the base
reerected, similar difficulties ensue. After the mortar has set, the wedges
should be removed and the spaces pointed up. Sometimes, one or two
large holes are made near the center of a large base so that grout may be
placed through them, or so that the elevation of the mortar may be
observed and (it is hoped) the latter may be made to bear against the
steel.
Special compounds are manufactured for use in grouting, either as an
admixture or as a substitute for cement mortar and grout. Wet shrink-
age and slumping of mortar may be troublesome chemical shrinkage will
;

probably be very small; and the quality of the workmanship may be more
important than slight variations of materials.
(c) By threading the anchor bolts sufficiently, two nuts C and D may

be used as adjusting screws. The base plate may be erected and posi-
tioned accurately, then a stiff, "dry" mortar may be packed under the
base and rammed into place with a stick and mallet. Space for this work
is necessary around the base. It is almost essential to have four anchor
bolts in any base plate because two bolts may not prevent rotation of the
base about the line between them when the mortar is packed on one side
more than on the other. It is obvious that some pressure will be exerted
on nut C by the structure, but this is inconsequential when the mortaring
is done well. In this way, any sensible workman can obtain a trust-
worthy bearing. It is also possible, when this screw device is used, to
remove the base, deposit and smooth off the mortar, replace the billet,
and tighten nuts D so as to force the base back into position, except for
whatever mortar may remain between the plate and nuts C.
(d) Under very heavy loads, and especially under machinery subjected
to shocks, it may be desirable to have means for clinching the mortar in
place in case it cracks. This pedestal is recessed at E; the base may be
dry-packed; but the mortar pad cannot become loose. A scored or
Art. 8-5] ACTION OF STEEL BASE PLATES 227

roughened top finish of the pedestal might provide similar results. Wire
mesh in the mortar may prevent the latter from falling apart, but it prob-
ably will not prevent the cracks in the first place.
The customary allowance for grouting is 1 to 2 in. ; if less than that, the
grout is difficult to place. Under very large machines, 3- or 4-in. thick-
nesses may be desirable. In the construction of the bearings of the
Bayonne Bridge, the main structural shoes were riveted to a steel frame,
and then a 2-ft. layer of reinforced concrete was placed below the shoes,
as indicated in Fig. 8-8.

Bottom
strvfatLl
Concrete
filling

ENLARGED DETAIL
AT PIN

FIG. 8-8. Details of the bearings and erection frames of the Bayonne Bridge. (Courtesy of the Port

of New York Aufhoriiy.)

Action of steel base plates. The minimum area of a steel base

8-5.
plate under a column or under a bridge bearing is generally determined by
dividing the computed load by 600, 800, or 1,000 p.s.i., or by some other
specified or selected allowable unit stress on the concrete. The base plate
is usually relatively thin, made of a ductile material, and designed for a

large unit stress in bending, and the load is generally applied to it at a few
points or lines of large concentration. It is therefore probable that a base
plate under a column, for example, cannot spread the concentrated load
uniformly over the concrete and mortar or grout directly under the steel.
It seems reasonable to believe that the plate will deform into a saucerlike
shape, the edges curling upward. This would tend to cause a high unit
pressure on the concrete under the shaft of the column, with decreasing
intensities toward the edges of the plate.
228 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

When there is a large bending moment that tries to rotate the column,
column tends to bear heavily on one
as pictured in Fig. 8-9 (a), the side
and cause a very highly concentrated load at some line such as F. It is
probable that the base plate will tend to deform, as pictured in exagger-
ated manner in (6). At least, it is evident that the resultant of the com-
pressive stresses in the concrete must counteract and line up with the
applied load P, and this indicates the probability of large stresses in the
vicinity of G. The edges A and B may not lift up as pictured, but the
former particularly will not apply much if any pressure. It is obvious

Possible crack

faJ-ASSUMED LOADING (^-EXAGGERATED DEFORMATION

P P+T
-^TrtlfMl

-(P+T)
^-IMPROBABLE PRESSURE DIAGRAMS CeJ- PRESSURES WITH ANCHOR BOLTS

G D AT
•
o ' * Pedes fa I
' Pedestal

*(P+TJ
^-POSSIBLE REACTION DIAGRAM tf>PRESSURES IN BODY OF PEDESTAL

FIG. 8-9. Illustration of possible action of a steel base plate and a concrete pedestal.

that the large compressive stresses and the deformation near G will tend
to cause cracking of the corner D, and possibly of C.
This exaggerated picture in Fig. 8-9(6) indicates that the probable
pressure diagram under the base plate is more like that pictured in (d)

than the one in (c). In general, a large area of thick concrete well tied
together will withstand a very high localized and centrally applied pres-
sure, but most practicable concrete pedestals and bases are not large
enough to trust this action. Large loads causing high localized pressures
near edges and upon narrow concrete sections should have the foundation
concrete proportioned conservatively.
If the overturning moment is sufficient to cause tension in the anchor
bolts, as pictured in Fig. 8-9 (e), the total resultant pressure under the base
Art. 8-6] ISOLATED FOOTINGS 229

plate and on the concrete by the magnitude of T. Down in

is increased
the body of the pedestal, the pressure diagram may be somewhat like that
in (/), where AT represents local tensile stresses caused by the bond of the
concrete to the anchor bolts. The compressive stresses in the pedestal
are generally unimportant if the top is strong enough.
8-6. Isolated footings subjected to overturning. In the design of
practicable structures there are many instances in which isolated footings
are subjected to vertical and overturning forces. When the latter are
small compared to the former, the overturning may be neglected, its
effect being assumed as taken care of by the safety factor. However,
overturning should be investigated to see whether or not it is important.
When the maximum bearing value of the soil is known with reasonable
certainty and when a safety factor of 1.5 to 2.0 has been used in estab-
lishing the allowable unit bearing pressure p for design purposes, one may
properly ask whether p may be increased by some percentage when
designing foundations for dead, live, and lateral loads all acting simul-
taneously. It would seem that the improbability of the absolute maxi-
mum loads and the temporary nature of these combined loads justify
some increase in the allowable bearing pressure. The author suggests
the following specifications, where p is the allowable bearing value for
direct loads:

For dead load plus live load p

For dead load plus wind and other lateral loads p
For dead load plus live load plus wind and other lateral
loads 1.2-1.3p

As a simple assume the footing pictured in Fig. 8-10(a), with the

case,
loads as shown. vertical load P will be assumed to equal the weight
The
of the superstructure, its loads, and the weight of the complete sub-
structure, including the earth above the footing GH. Let represent M
the overturning moment without consideration of any accompanying
horizontal forces. Probably the intensity of the pressure under GH will
not be distributed uniformly, nor will it vary as a straight line. Never-
theless, it is safe to make these assumptions. Therefore, there may be
two varieties of pressure diagram:
In the first, using L for GH and 6 for the width of the footing at 90°
thereto, the uniform pressure caused by P is

=
V £ (8-1)

The pressures p' supposedly caused by the overturning moment will be,
for a rectangular footing,

P
Mc _
j
M
bL ,
k° w
230 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

When p' is less than p, the combined pressure diagram is represented by

Fig. 8-10 (e). If the eccentricity of the resultant pressure at the base is

M (8-3)
P
then Pi ^ L)
bL\
(8-4)
P ( 6A

The second case occurs when p' , of Sketch (d), exceeds p, the eccen-
tricity exceeds %L, and the resultant load hits outside of the middle

Loads caused
by superstructure

Roughened
const, jt.

Reactions shown
in diagrams

-?» L NJ M
(a)
H J (b)
K
H J K
J "^tii! i ,
»||

4!
(c)
LJ3 J —
(d)
^
L ,,

X
-p\ H (f)

|I
$
(e)

FIG. 8-10. Assumed pressures under an isolated concrete footing with direct and overturning forces.

third of the footing GH. Since tensile resistance of soil sticking to the
footing near J of (6) cannot be relied upon, it is satisfactory to assume a
triangular distribution of pressure, as illustrated in (/).. The maximum
pressure at K is computed as follows:

e = -p (as before)
Art. 8-6] ISOLATED FOOTINGS 231

This locates the position of the resultant reaction — P, which will be at

the center of gravity of the pressure diagram. Therefore,

2
- e = -g
1
and Li = %(- - e\ (8-5)

2P
y2PlL b=P 1 or Pi "
mL/2) _ e]
(8-6)

To test the footing for safety against overturning

(assuming a safety
factor of 1.5), multiply the overturning moments by
compute a new 1.5,

e = 1.5M/P, and determine the new pressure diagram, which may be

like that of Fig. 8-10(,f). Probably the critical case will be that with a
minimum load P. If the resultant, — P, is inside the edge K, Sketch (/),
it would seem that the footing is safe for the assumed conditions. How-
ever, if pi exceeds the ultimate strength of the soil (which it should not
do), the latter may cause the footing to fail or to tilt badly. Even when
Pi is lessthan the ultimate strength, one should investigate the possi-
bility of unequal settlement, which may cause harmful tilting of the
structure.
In connection with Figs. 8-10(a) and (b), notice the following details in
the respective sketches:
(a) In any case, dowels A should be large enough and long enough to
splice bars B and to distribute their stresses into the massive concrete.
If bars B are in tension, it is advisable to hook (or 90° bend) the dowels
at the bottom to engage the main footing; these hooks are of little value
in resisting downward compressions. The construction joints at C and
D are roughened, but no actual keys are necessary because of the large
pressure and friction caused by P.
(6) Dowels E and F are shown bent toward and ./, respectively. K
This may seem to be efficient, but it is usually undesirable because of
questionable bond stresses in the vicinity of N, and regular bars are needed
alongside E and F anyway in both directions. A long U-bar from E,
across N,and up F is better.
The maximum pressures pi in Figs. 8- 10(e) and (/) are gross intensities.
The weight of the displaced earth and backfill above H and may be K
deducted from pi if one wishes to consider the increase of pressure only.
In such problems as these, this refinement is not generally justified as far
as pressure on the soil is concerned, but it may properly be considered
when proportioning the concrete and reinforcement of the substructure.
Figure 8-11 shows some types of footing that may be used. In these
cases, the overturning is assumed to be caused by the shear *S that may be
the result of frame action, wind, or bracing connected to a steel column.
The eccentricity e may be found as follows, using h for the total height of
232 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

the footing:
S = Sh
or e (8-7)
V
The load P equals the weight on the column plus that of the foundation
concrete and the earth above the footing. In the case of (a), the effective-
ness of the earth around the concrete and above C and D may be question-
able because of the wedging action of the trapezoidal shape.
In the sketches in Fig. 8-11, notice the following:
(a) The anchor bolts are extended close to the bottom of the footing so
as to engage the entire weight of the footing without the use of vertical
reinforcement. The only bars that may be of appreciable value are the
hoops a and b to tie inthe top and bottom of the footing; even these are a

p Anchor bolts
E F S

J If K

raw
DL
^M
raJ-TRAPEZOIDAL flW-UNREINFDRGED
FOOTING FOOTING (cJ-REINFORCED FOOTING

FIG. 8-11. Isolated footings for resisting small overturning forces.

sort of insurance rather than a necessity. The resistance of this footing

to overturning is generally small because the base CD is usually narrow.
However, the passive pressure of the soil at the right may help prevent
actual failure when the footing is not too shallow; it will be very effective
if the footing is rather deep.
(b) The anchor bolts in this unreinforced footing are also extended
close to the bottom. They reinforce the pedestal automatically. Both
P and S are supposedly small, otherwise there may be danger of cracking
the footing.
(c) When P and $ are large, it is probable that a bigger stronger footing
will be needed. Here the footing is depressed in order to get the benefit of
cheap weight in the form of earth on top of TU. The anchor bolts are
shorter, bars c tie the top together, and bars d are made strong enough to
transfer the pull from the anchor bolts to the bottom of the footing and to
withstand the bending at the base of the pedestal QR. Bars e are to resist
any tension caused by uplift of one side of the footing. This tension
may be caused by the weight of the footing itself and the earth on it.
Art. 8-6] ISOLATED FOOTINGS 233

Although it may
be small, this tension should be provided for when it is
likely to exist. Bars / are for the normal action of the footing, to resist
the overturning combined with the vertical load.
Example. Assume a footing like that of Fig. 8-1 1(c) used as part of
the central bent of a small two-span steel overpass. Assume the following
data: The net column load caused by the superstructure on the foundation
is 160 kips for dead load combined with the theoretical uplift caused by

wind on the structure tending to lift the footing when it is on the wind-
ward side of the structure with no live load. When dead load plus
maximum live load is combined with the greatest downward pressure
produced by wind acting on the structure, the total load is 320 kips.

FIG. 8-1 1A. Final assembly hangar, Douglas Aircraft, Inc., Santa Monica, Calif. These are three-
hinged, trussed, rigid frames that cause an outward as well as a vertical force at their bearings.
Architects and designing engineers, Edward Cray Taylor and Ellis Wing Taylor, Los Angeles, Calif.
[Courtesy of the Beihlehem Steel Co., builders of the steelwork.)

Assume that the steel base plate is 16 by 24 in.; S = 20 kips; the soil is

1 ft. 6 sandy loam underlain by deep fine sand. Design the founda-
in. of

tion for this column if the structure is near Rochester, N.Y.

From Table 4-5, assume that the allowable maximum pressure on the
sand is 6 k.s.f. for DL -f- LL and 7 k.s.f. for DL + LL + W. Because of
frost, the footing will be placed so that YW is at least 5 ft. From the
size of the base plate, the pedestal will be assumed to be 2 ft. 6 in. square.
The anchor bolts are to hold the steelwork in position but not resist large
overturning moments. Because of snow, the pedestal will be placed with
its top 2 ft. above the ground.
How can one obtain a reasonable estimate of the size required for the
footing? The following is one way to attack the problem:
1. The area at VW
will exceed 320/7 = 46 ft. 2 because of the weight ,

of the foundation and the effect of overturning.

234 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

2. Use a slightly elongated footing to get greater resistance in the direc-

tion of S, and then add 8 or 10 ft. 2 to this minimum area. Therefore, try
VW = 8 ft., the width b = 7 ft., WU
= 2 ft,, and UY = 3 ft. Then the
effective weight of the footing is computed as follows:

Pedestal = 2.5 X 2.5 X 5 X 0.150 4.7

Footing = 8 X 7 X 2 X 0.150 16.8
Earth = (8 X 7 - 2.5 X 2.5)3 X 0.1 FL9
Total W 36\4 kips
3. M = 20 X 7 = 140 ft.-kips
M 140
6
P+ W 320 + 36

From Eq. (8-4),

P + W {, ,6e\ 356 A ,
6 X 0.39\ qo1 ,

This issomewhat too large, since only 7 k.s.f. is allowed.

4. Try L = 9 ft. and b = 7 ft. Then W= 41 kips and e = 0.39 ft.

?>i =
361 /, ,
1 H
6X 0-39 \
) = 701
7.2 k.s.f. (near enough)
q 7
I
q

361 A 6 X 0.39\ , . ,

It is also essential to investigate this trial footing for the condition of

minimum vertical load combined with the lateral force to see that it will
not overturn. The magnitude of P + = 160 + 41 = 201 W kips.
Therefore,
M 140 _
e
P + W 201
A1+,6
P1
= 201
<IX"71 "X0.7N
g-j- , 71
4.7 k.s.f.
.

201 A
- 6 Xg—
0.7\
= _ ,

9X ^1 -- j
.
t
P2 1.7 k.8.f.

Since this case is found to cause less serious results than the maximum
loading, proportion the footing to withstand the latter.
From the magnitudes and character of pi and p 2 it is apparent that ,

no tensile reinforcementis needed in the top of the footing. The bars at

the bottom may
be determined and the footing analyzed upon the basis of
the pressure diagram shown in Fig. 8-12(a). The upward shear at R for
the investigation of punching shear along this face of the pedestal caused
by the pressure diagram of Fig. 8- 13 (a) may be approximated by assuming
that the pressure diagram is a rectangle having the average ordinate of
Art. 8-6] ISOLATED FOOTINGS 235

6 k.s.f. This unit pressure is assumed to act on the area RR'X'U'UX of

Sketch (6). Then,

VR = 6
2 5
'
+ ?
X 2.25 + 7 X l) = 106 kips
( 2
106,000
Vt = 560 p.s.i.
30 X 0.3 X 21

This seems to be a little high, so the footing will be strengthened, because

bending and diagonal tension will probably be serious. One way to do
this is to slope the top of the footing so as to deepen it 6 or 8 in. at Q and
R, Fig. 8-12; another method is to slope the sides NQ and OR of the

2-6"
N
R^

Q R
9-0"
V
%T
W W
(b)- PRESSURE DIAGRAM ^-REINFORCEMENT TO BE USED
ASSUME FOR DESIGN IN SUBSTRUCTURE
OF REINFORCEMENT

faJ-ACCEPTED PRESSURE DIAGRAM

FIG. 8-12. Pressure diagrams and reinforcement.

pedestal to reduce the cantilever; a third is to slope the other two sides of
the pedestal so as to increase the resisting width; a fourth is to make the
pedestal 3 ft. square or 3 ft. 6 in. square. The first method will be used
because this will also increase the bending strength of the footing. The
total depth will be increased to 2 ft. 6 in.

106,000
Vt = 440 p.s.i.
30 X 0.3 X 27

The bending moment at R, using the net pressure diagram of

Fig. 8-13 (a), is

M =
1.1 X 3.25 w< 2
=
( 5.5 X 3.25 X 1.62 -i
2 i
x 3
-) 7 230 ft, -kips

M 230
= 5.68 in.
2

fsjd 20 X 0.9 X 2.25

V=
1.1 X 3.25
=
5.5 X 3.25 + 2
7 138 kips

So = V 138,000
= 27 in.
ujd 210 X 0.9 X 27
236 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

Use 13 No. 6 bars, arranged as shown in Fig. 8-12(c).

2M 2 X 230,000 X 12
= 330 p.s.i. (very safe)
/-
kjbd 2 0.3 X 0.9 X 84 X 27 2
Using d = 23 in. at XX' and an average pressure of 6.43 k.s.f. acting on
area XX' VV of Fig. 8-13(6), the diagonal tension

V_
=
6,430 X 7 X 1 = 26 (very safe)
Vl p.s.i.
bjd 84 X 0.9 X 23
If the intensity of pressure under the footing varies rapidly in such a
diagram as Fig. 8-13(a), one can "cut" it up into parts and use the com-
puted values instead of averaged ones for estimating V as was done in
the preceding calculations.

3.25'
R-*.
.
i

\\\\
I!

-*
J5°
t:
^
^S / _L *

<JSJ
S3
~~ ~ :
-_ 1 U l0 '

Ic j

\ M i

[a) NET PRESSURE DIAGRAM

W-V
R B 1
\
c

45°\

, x U
325
(b) ASSUMED TRAPEZOIDAL AREA

FIG. 8-13. Pressure diagram and trapezoidal area for analysis of footing.

The preceding methods of analysis are equally applicable to footings

that are loaded eccentrically with respect to the rectangular bearing area.
It makes no difference to the footing whether the direct load times an
eccentricity causes the bending or whether the same M
is due to other

causes except that no increase in the allowable bearing pressure is advis-

able when the moment is caused by live and dead loads that are off center.
When overturning forces act simultaneously about two rectangular
axes, the footing will tend to bear heavily on one corner. If there is
Art. 8-6] ISOLATED FOOTINGS 237

compression under the entire footing, the maximum pressure may be

approximated as follows:
P M xcx M2c 2
Pi =
z+ -j- +
,

-^ f
(8-8)

The minimum is

?2 - J~ T x 17 (8 " 9)

If 2>2 is theoretically tension, then p x may be approximated by trial about

a diagonal axis, assuming triangular distribution and no tension.
In the discussion of Fig. 8-11, the resistance to sliding is assumed to be
provided by friction on the bottom of the footing. One may properly ask
"Will the resistance of the earth against a footing prevent its over-
turning?" In general, the abutting power of the soil resistance to being —
—
pushed sidewise may prevent actual overturning and sliding. How-
ever, the deformation of the soil and the rotation required to bring the
earth into maximum may
be such as to cause uncertainty
resistance
regarding how much dependent upon friction combined
of the stability is

with unequal pressures under the footing and how much upon lateral
resistance of the soil.

As an example of the theoretical lateral resistance that may be pro-

vided by the soil, assume the foundation shown in Fig. 8- 14 (a), and
assume the gross load P = 144 kips and S = 40 kips. It is sufficient to
use Rankine's formula for the lateral resistance of the soil, i.e.,

p = wh , 1 + sin .
fQ in>1
(8-10)
-z

1 — :

sin

where p = resistance in pounds per square foot, w = unit weight of earth

(assumed here as 100 p.c.f.), h = depth in feet to the point considered,
and 4> = angle of internal friction of the soil, assumed to be 30° in this
case. Therefore,

p = lOO/i
+ = 300/1 (8-11)
I
— jjj
1 0.5

Thus, the assumed triangular pressure diagram for a vertical strip of soil
1 ft. wide will be as shown in Sketch (b). The computed diagram of total
pressure on the 2-ft. -square pedestal is then DEF of Sketch (c), whereas
that on the 6-ft.-wide footing is EGKH. Summing up the lateral resist-
ances, the soil will withstand a lateral force of approximately 17.6 kips,
leaving 40 — 17.6 = 22.4 kips to be resisted by friction under MN. The
required coefficient of friction is, therefore,

22.4 = 144/ or / = 0.16 (very safe)

238 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

Applied loads from

y^. Appliec
ipersiruclure

Reactions
not shown

$.025*

M '6 L
0"sq. N
^-DIMENSIONS AND LOADS ^-PRESSURE fcJ-RESISTING FORCES AND
DIAGRAM LEVER ARMS

3-0" 3-0'
\e--/S/'P

^-PRESSURE DIAGRAM UNDER FOOTING

FIG. 8-14. Approximation of maximum lateral supporting power of earth on an isolated footing

subjected to overturning.

r
The moment of these lateral resistances about point A is

DEF, 4.8 X 2.83 13.6

EGJH, 10.8 X 0.75 8.1

GJK, 2.0 X 0.5 1.0

Total 22.7ft.-kips

The resultant overturning moment that must be resisted by the footing

MN is then
40 X 6 - 22.7 = 217.3 ft.-kips

The corresponding eccentricity of the resultant at MN is

M 217.3
= 1M .

P = 144
ft '

This is shown in Fig. 8-14(rf), and L x = 4.47 ft. From Eq. (8-6),

= 2P 2 X 144
= 10.7 k.s.f.
V\
3b[(L/2) - e] 3 X 6(3 - 1.51)
Art. 8-6] ISOLATED FOOTINGS 239

From the preceding computations, it is apparent that the benefit of the

surrounding earth in resistance to sliding is considerable, whereas it is
relatively small as far as overturning is concerned unless the substructure
iswide and deeply embedded. It is generally advisable to design founda-
tions subjected to overturning so that they will be safe without depend-
ence upon the lateral resistance of the soil. On the other hand, excep-
tions occur. For example, piles in mud, deep bridge piers in clay,
narrow cylindrical caissons in silt, and tall pedestals extending up through
earth fill are cases in which the lateral supporting power of the soils may
have to be relied upon.
One may compute the end restraint at the base of a concrete or steel
member, and he may depend upon the footing to develop the required
resisting moment. Theoretical calculationsmay show that the selected
footing can offer this resistance safely. However, when the footing is on
weak or plastic soil and when the overturning moment is applied steadily
or for a relatively long time, compression and flow of the soil caused by
the large edge pressure under the footing may produce sufficient angular
rotation to reduce or even eliminate the assumed restraint. The angular
movement required to transform the fixed end of a stiff member into
what approximates a hinged end is often small. This should be borne in
mind when rigid-frame structures are designed.
What does the embedment of a foundation have to be in order to enable
one to rely upon the lateral support of the surrounding earth to prevent
overturning? This is almost impossible to answer. There is no sharp
dividing line between the depth that is sufficient and the one that is not.
The following are given as a general aid in drawing one's own conclusions:
1. An appreciable deformation of ordinary earth, especially backfill
around a foundation, will occur before a large passive resistance is
developed. Coarse sand and gravel are fairly dependable, but clay, silt,
and fine sand may yield too much.
2. If a slight angular rotation of the substructure is permissible, the
resistance of the surrounding soil will probably come into action strongly
before failure by tipping can occur. However, appreciable movement is

not generally permissible.

When failure would be serious, one should be careful about depend-
3.

ing upon such uncertain resistances.

4. Consider the possibility of someone in the future unwittingly

removing the soil upon which the structure depends for lateral support.
5. The depth for reliability is relative. For example, an embedment of
10 ft. is a lot for an isolated pedestal and footing 5 ft. square; it is very
small for a bridge pier 60 ft. high with a base 15 by 40 ft. in area.
In some cases where large horizontal thrusts must be resisted, "wings"
can be built out on the sides of the pedestal of a footing in order to
240 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

obtain a large area to push against the earth. For example, the founda-
tions used for supporting the trussed rigid frames of a large hangar, some-
what like Fig. 8-11 A, were sunk down 6 ft. into the ground and the grade
beams were deepened and reinforced as beams to spread the thrust side-
ways. In addition, cross-struts between the bottoms
of the footings
were used for the same purpose.
There may be cases in which a substructure that must resist overturn-
ing is supported directly upon bare rock. The rock will probably with-
stand safely more pressure than will the concrete, hence the magnitude of
the pressure is unimportant. Sliding may be resisted by roughening the

WFstee/post
3 Anchor bo/fs

.
Rock
f

(ct)-A STEPPED FOOTING (b)- A TRAPEZOIDAL FOOTING

FIG. 8-15. Two types of foundation on rock for railroad electrification.

rock surface if it is not already so. Overturning, however, must be with-

stood by the substructure itself.
As an illustration, consider the stepped footing in Fig. 8-15(a), which
is to support a steel post used in the power lines for the electrification of a

railroad. Assume that there is a reversible overturning moment M =

100 ft. -kips, a horizontal shear S = 6 kips, a gross load P = 25 kips, and
that L/2 = 4 ft. and h = 5 ft. Taking moments about A, the over-
turning moment is 1

M = 100 + 6 X 5 = 130 ft.-kips

whereas the righting moment is

MR = 25 X 4 = 100 ft.-kips

Therefore, the footing is not safe. If the footing is increased to 11 ft. in

width and all steps widened to make P = 45 kips,

MR = 45 X 5.5 = 250 ft.-kips

1
Since the rock is very strong, the corner of the footing can be used as the point of
rotation.
Art. 8-6] ISOLATED FOOTINGS 241

The safety factor against overturning is then 250/130 = 1.9, which is

satisfactory.
apparent from the above that these footings must be very broad
It is
and The pull on the
heavy.
anchor bolts will also be large. If
the post is embedded in the footing,
as pictured in Fig. 8-15(6), the lat-
eral pressure at B may be im-
portant. Here the concrete is

sloped up to help reduce corrosion

of the steel. Rust plates C and D
may help to protect the main ma-
terial. The reinforcement shown
in Fig. 8-15 is designed to knit the
concrete together and permit the
footings to "ride either edge"
safely. If M and S are not revers-
ible, an unsymmetrical footing
may be useful, but it may not re-
sult in important economies be-
cause most of P may be caused by
the weight of the footing itself.

On the other hand, this is a case

where the top pedestal of Fig.
8-15 (a) might be placed directly on
the rock and the anchor might
bolts
be made and
like that in Fig. 8-2 (d)
buried in the concrete, with heavy
reinforcing bars overlapping them
and extending down into the rock
in drilled holes as pictured in Fig.
8-2(6), using a length of perhaps 10
ft. so as to engage enough rock to
supply the necessary weight. In
this case, if the rock is not too FIG. 8-16. A steel transmission tower that was

seamy, one may count on the damaged when the boom of a railroad crane
collided with in the darkness. Notice the
it
weight of an inverted cone with
crack in the foundation.
its apex at the bottom of the group

of bars and with theoretical outward side slopes of 30 to 45° with the
vertical.
Figure 8-16 shows construction that was made for the electrification of
a railroad at a mine. The steelwork was erected during a certain day.
That night the operator of a railroad crane was ordered to take his
242 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

machine out to a particular place, but he was not told about the new
overhead trusses. In the dark, his boom collided with the steelwork and
caused the wreckage shown. Notice how the heavy base details held
together, and how the tensile shock delivered to the footing by the anchor
bolts caused the concrete to crack in two. The footing is shown in
Fig. 8-17, and the approximate location of the crack is indicated. Of
course, footings are not designed —
and cannot be economically for such —
conditions. However, as a general principle, it is desirable to make them
sufficiently strong to cause the superstructure to fail without wrecking
the foundation. If only the former fails, it may be patched up; if the
latter gives way, then both parts have to be rebuilt.
8-7. Unsymmetrical footings. Occasionally a foundation
is to be

unsymmetrical, and the center of gravity of the bearing area will not coin-
cide with the resultant of the vertical loads. This may be caused by the

4'-0"
P4-fif'each side
r 214" 3'-t
4Jl £ \P
tf 1ft

6-0

FIG. 8-17. Footing that cracked.

limitations of space and by interference of pipes, pits, and machinery

foundations with footings. Eccentricities of varying magnitudes and
directions may be caused by large movable loads that may be in one loca-
tion at one time and example, an
in a different place at another; for
interior footing in a factory may
support separate columns under two
adjoining crane runways, as pictured in Fig. 8-52. It sometimes happens,
too, that alterations make it desirable to cut a piece off from an existing
footing. It is desirable, therefore, to be sure that the increased unit
pressures P/A' are not excessive and that these foundations are safe.
Heavy loads times moderate eccentricities may produce large moments
and surprisingly big edge pressures.
The eccentricity of a load causes the equivalent of an overturning
moment. Therefore, the method of analysis of an unsymmetrical founda-
tion is similar to the procedures described in the preceding article.
As an illustration of the assume the footing shown
method of analysis,
in Fig. 8-18. It is for anan industrial plant. One
interior column in
corner is to be cut as shown in Sketch (c). Assume that P = 275 kips.
Compute the gross pressures under the footing, and draw the pressure
diagram.
Art. 8-7] UNSYMMETRICAL FOOTINGS 243

The gross weight W of the footing and earth is computed as follows:

Footing = (48 - 3)225 10,100

Pedestal = 6.25 X 2.67 X 150 2,500
Earth = (48 - 3 - 6.25)200 7,800
Total W 20,400 lb. (say, 21 kips)

The eccentricities are

~2
ei = * L5 X 3
= -0.2 ft., or 0.2 ft. from O for axis 1-1
— 3
48
_q v 9 25
e2 = = -0.15 ft., or 0.15 ft, from O for axis 2-2
45

Therefore, about the center of gravity axes,

V V 93
/a = ^^
6
_ 83 1
X J5'

* t _ 3 x 32 - 45 X 0.2 2 = 226 ft,

4

8V 6 3 9
w v i S3
/2 = -~^ - * ° - 3 X 2.25 2 - 45 X 0.15 2 = 127 ft.
4

Assuming P and W to be practically colinear, the overturning moments

are

M i
= (P + W)ei = 296 X 0.2 = 59 ft.-kips
M 2 = (P + ]F
/
)e 2 = 296 X 0.15 = 44 ft.-kips
From the use of Eq. (8-4) in its general form, the computed corner
pressures are

Ve = 4.6 k.s.f.
45 226 127
296 59 X 3.8 44 X 3.15
= 6.7 k.s.f.
Vf
45 226 127
296 59 X 2.2 44 X 3.15
= 8.2 k.s.f.
Vg 226 127
45 '

296 59 X 2.2 44 X 1.65

=
Ph
45
+ 226 '
127
7.7 k.s.f.

296 59 X 4.2 44 X 1.65

= 8.2 k.s.f.
Vj
45 226 '
127
296 59 X 4.2 44 X 2.85
= 6.7 k.s.f.
Vk
45 226 127

The pressure diagram is pictured in Fig. 8- 18(d).

The reinforcement of the footing in Fig. 8-18(6) may be made in the
form of one strong band perpendicular to AD and BC, and another run-
ning across AB and CD. For computing the bending moments and
shears, the methods explained in Art, 8-6 may be used.
244 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

By similar procedures, the pressure may be approximated for various

combinations of forces and conditions that produce unequal bearing as
long as none of the pressures are theoretically negative. If the assumed

or actual loads and conditions cause serious variations of bearing with a

col.
H
Pedestal-* e2 o
Footing*
C.G.axis
f*i I

r«;-SIDE ELEVATION
E K
,
2-0"
(W-PLAN OF BEARING AREA

H
J
\D
h
2'-6

E K
S'O"
(b)-Pim OF FOOTING faW-DIAGRAM OF COMPUTED PRESSURES

Y Y'
RU=IX
US=Iy
s J*£$^ TO '
>

iy-rx
0'v=0'Ucos2e+UVs/n20
Ix >=TV,Iy> = VTV,e=ZO'S

^-LOCATION OF PRINCIPAL AXES (/./-GRAPHICALMETHOD FOR LOCATING PRINCIPAL AXES

AND DETERMINING PRINCIPAL MOMENTS OF INERTIA

FIG. 8-18. Analysis of pressure under an unsymmetrical footing

tendency for large uplift under a corner and an appreciable area of the
footing, an engineer might well spend time and thought first upon the
best way to change the basic construction so as to eliminate such question-
able construction.
Art. 8-7] UNSYMMETRICAL FOOTINGS 245

The preceding computations may not yield the critical magnitudes of

the pressures under the footing. To obtain these in cases where it seems
to be desirable to do so, the principal axes and moments of inertia may
be used. The procedure then is the following, referring to Fig. 8-18:
1. Draw the footing in outline as in Sketch (c) or (e). The latter will
be used purposely to illustrate this.
2. Compute the bearing area of the footing, locate its center of gravity,

and draw the rectangular axes X-X and Y-Y through the center of
gravity 0'.
3. Assume that positive angles are to be measured counterclockwise
from O'Q to locate the principal axes of inertia.
4. Compute the moments of inertia I x and Iy about the chosen rec-
tangular axes.
5. Compute the product of inertia I xll about the center of gravity,
bearing in mind the signs of the coordinates.
6. From the equation

tan 29 = r ^V5 (8-1 la)

obtain the angle 9 to locate the principal axes X'-X' and Y'-Y'. If tan 20

is measure 9 above axis X-X; if negative, measure it clockwise

positive,
from O'Q. Of course, if the tangent of 29 is negative, the angle can be
measured counterclockwise from O'Q into the second quadrant. Then 9
will locate one of the principal axes as before, neglecting any uncertainty
as to which is X'-X' and which is Y'Y'. Measuring clockwise from O'Q
for negative values of tan 29 may avoid some confusion.
7. The principal moments of inertia are

/ max = ~^ + ~ /x r
* (cos 29) - Ixy sin 29 (8-116)

/ min = h±±« _ k^Ii (cos 29) + I X1I sin 29 (8-1 lc)

8. Compute the components of the overturning moments normal to

the directions of the principal axes. Call these M x> and M y >, repectively.
9. Compute or scale the lever arms from the principal axes to points
at which the pressure is desired, as c and d in Sketch (e).
10. Compute the intensity of pressure under seemingly critical points,
such as G of (e), as follows:

=
P M x M y 'd
-j ± ~j— ± —~-
'c .
n
(8-1 Id)
Vg A 1 X ly' '

With the values previously computed for the footing of Figs. 8- 18(c) and
(e), the computations for critical pressures at G and J are the following:
246 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

I xy = 48(-0.2)(0.15) - 2 X 1.5 (-3.2) (2.4) = +21.6 ft.

4

tan 26

26
=
^^g
= -
=
(23°37')
-0.437

sin 26 = -0.401 cos 26 = 0.916

= — (11°48') sin = -0.204 cos = 0.979
226 + 127 226 — 127
/ max = + (0.916) - 21.6(-0.401) = 230 ft,
4

3 2
226 4- 127 226 — 127
/ min = - (0.916) + 21.6(-0.401) = 123 ft.
4

^
M x = 59 ft.-kips and M y = 44 ft.-kips

Remember that the load acts at O, Sketch (e). Then, neglecting signs
of angles but considering directions of rotation, the new coordinates of
point are
x' = 0.2 sin + 0.15 cos 9 = 0.188
y' = 0.2 cos 0-0.15 sin 6 = 0.165
M x > = 296 X 0.165 = 49 ft.-kips
M y > = 296 X 0.188 = 56 ft.-kips
Scaling the lever arms from Fig. 8- 18(e), or computing them if desired,
the estimated pressures at G and J are

_ 296 49 X 1.5 56 X 3.6 _

_ , .
PG — ,r- + r,or\ ' I
190 "" K.S.I.

= 296 X 49 3.75 56 X 2.5

= D _ . .
+ ~"23CT~ +
, .

Vj 8 "° k si - -

^5 123

These values are only slightly greater than the 8.2 k.s.f. computed pre-
viously. However, this may not always be the case.
The Mohr-Land circle 1 is a graphical method for obtaining the positions
of the principal axes and for finding the principal moments of inertia.
Referring to Fig. 8-18(/), the procedure is the following:
1. Draw a circle with radius (I x I y )/2, using the values of I x and Iy+
about any rectangular axes such as X-X and Y-Y of Fig. 8- 18(e).
2. Scale off RU to equal I x ,
and US to equal I y .

3. Compute I xy as for (e).

4. From U, draw UV = Ixy and normal , to O'S. If tan 26 is positive,
draw UV above O'S; if negative, draw it below O'S as for this problem.

5. O'U =
2

Therefore, ^ = I xy + ^-^ = j^j = tan 26

1
Courtesy of C. J. Posey and Angel Lazaro, Jr., State University of Iowa, Iowa City,
Iowa.
Art. 8-7] UNSYMMETRICAL FOOTINGS 247

Then 20 = SO'W and 6 = SO'Z.

6. TV = In
VW = L
The preceding methods of analysis assume that there is compression
under the entire area of the footing. If this is not the case, they are
inapplicable.
When necessary, one may compute the pressures under a footing sub-
jected to eccentricities and a serious diagonal overturning, as indicated in

IE (a)

s/^r
(b)

FIG. 8-19. Footing with unsymmetrical loads and overturning.

Figs. 8-19(a) and (b) by trial. A rectangular footing is shown for sim-
plicity. The procedure is outlined as follows:
1. By means of the general application of Eq. (8-4), compute the pres-
sure diagram as though there were tension under part of the footing.
2. By graphics or algebra locate the position of the line of zero pressure
under the footing.
3. Select a trial zero axis, parallel to the one in item 2 but somewhat
nearer the high-pressure corner, as shown by EF in Sketch (c).

4. Draw a plan of the footing to scale, and show area ABEFD.

Using the plan view as in Sketch (c), scale off the perpendicular dis-
5.

tances from A, B, and D to the axis EF. Then compute the moment of
inertia of area ABEFD about axis EF, using / = bh 3 / 12 for a large tri-
angle with its vertex at A and base along EF minus a similar / for each
of the triangles with vertices at B and D.
6. Assume the total overturning moment M
= Pa + Sh; equate this
t

to pJef/c; and solve for px .

 >A
c ^
^N \ N
^
^f
N N
"

IT) «;r to CJ - o 8
B/J

'£ CS

A
//,

/
,4 / -o
c
3
O!

n
n/ ft <y/ uj 01

k /// ki|^ -o i5

{
O
I =
4)
c
a>

1i
1
il
1
1, >
3
77 %
U4 1 »h-
o
' h7/// I 5
1 /// / ^h
mi
'fl

'I W
1
D

"a
7/ n
/; lim
// L
>
<u

<l
<D
-o
0)
l~

// •- jf

1 ///
1
'll li

1
7
1

O "S

248
Art. 8-7] UNSYMMETRICAL FOOTINGS 249

7. Assume that HJ of section A-A in Sketch (c) is a straight line.

From it the pressure diagram may then be determined sufficiently for
practical purposes. Check to see if the volume of this pressure diagram
approximately equals the total load P. If the discrepancy is large,

assume a new location for the axis EF, and try again.
8. If the bearing area is very irregular, I E f may be approximated by
dividing it into arbitrary strips as indicated in (c). From / + Ax 2 for

each strip modified into a more or less equivalent rectangle or triangle, a

value for 1 E f may be obtained.
The maximum pressure under rectangular footings can often be com-
puted quickly and with sufficient accuracy by the use of Mattson's

abed, is lower
surface of foundation,
ef = line ofp=0 &

Case I Case II

FIG. 8-1 98. Assumed typical pressure diagrams for use with Fig. 8-19A.

graphs in Fig. 8-19A. Figure 8-195 shows the assumptions for two cases.
The procedure is outlined as follows:
1. Case I is when the load acts within the kern and there is pressure

over the entire base; hence, P/A + Mc/I can be used.

2. Case II is when the line of zero pressure cuts one corner. Referring
volume of the pressure diagram equals the total
to Fig. 8-195(a), the net
load P equals the volume of the largest pyramid minus the volumes of
two "negative" pyramids, and the centroid must be at the line of action
of P. Then
(1 ~ B/y)A (1 - A/x)B
J =
A/x
and K B/y
QP
xy[l - (1 - B/y)* - (1 - A/x)*]

3. Case III is when the line of zero pressure cuts two opposite sides.

ra x — A
rd

Use the same equation for p max as in step 2, but one binomial equals zero.
The effective B/y = 1 if F/B < 0.25.
250 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

4. Case IV is when the line of zero pressure cuts two adjacent sides
close to the corner nearest the load P.

x = 4:E and y = AF
Pmax = 5?
xy

To illustrate the preceding steps and the use of Fig. 8-19A, assume the
following
No. 1. A= 12 ft., B = 10 ft., E = 3.6 ft,, F = 3.5 ft,, and P = 100
kips. Using £7^4 = 0.30 and F/B = 0.35 in (d), their intersection lies
in the zone for Case II. Using the trial-and-error procedure, assume
A/x = 1; then, from the upper graph for F/B = 0.35, find B/y = 0.64.
From the lower graph with B/y = 0.64 and E/A = 0.30, find A/x =
0.86. Now, from the upper graph, with A/x = 0.86 and F/B = 0.35,
find B/y = 0.66 + Returning to the lower graph with B/y — 0.66 and
.

E/A = 0.30, find A/x = 0.86. Then, from A /x = 0.86 and B/y = 0.66,
x = 14 ft. and y = 15.2 ft. Use all these values in the equation for p max ,

Case II.
6 X 100,000
7W -= u x 15 _2[l - (1 - 0.66) - 3
(1 - 0.86) 3 ]

No. 2. 12 ft,, B = 10 ft,, E = 3.6 ft,, F = 2.0 ft,, P = 100 kips.

A =
E/A = F/B = 0.20. Then, from (d), the intersection lies in Case
0.30,
III. Using the minimum value of F/B = 0.25 instead of 0.20 and A/x —
1.0, the upper graph gives B/y = 1.0. Using this in the lower graph with
E/A = 0.30, A/x = 0.82. Therefore, x = 12/0.82 = 14.6 ft. Then,
using a pyramid similar to that in Fig. 8-195(6), with the resultant pres-
sure at 2 ft. in from ad and 3.6 ft. in from cd, the approximate length of
the effective B (labeled df here) is

(J/ 3X2 6 X 14.62

8 ft. +
x x - 3.6
J
11.02

Since 1 — B/y = for this case,

6 X 100,000
^ max - -
14.6 X 8[1 (1 0.82) 3 ]

To check the preceding computations the reader can compute the

volumes and centers of gravity of the large pyramids minus the "nega-
tive" ones and compare the results with P, E, and F.
8-8. Footings steadied laterally by structural parts. Except for
the lateral resistance provided by surrounding soil, the preceding discus-
sions have assumed that a footing must resist rotation by unequal pres-
sures under its base. What happens when the substructure, or the pedes-
Art. 8-8] FOOTINGS STEADIED LATERALLY 251

tal, is incorporated in the structure so that its top is steadied? It may

be that a resistance is provided that will prevent overturning so that the

footing should be designed for vertical loads only.
Assume the exterior column footing shown in Fig. 8-20. It is subjected
to load P together with a reversible overturning moment and a shear S. M
The concrete floor slab is seated upon the pedestal C and the grade beam E
at B and D, respectively. Bars a tie the floor to the foundation. Of
course, the floor will not contribute appreciably to the support of load P

-M +M

F/oor slab A f H4 Floor

t^LJ +s
.J..., +&
±S A {slab
w& J
Compacted Paint with
fill asphalt t
w Pedestal C
Pedestal
1 Original
(b)
ground
^Footing Reactions
not shown

F m.
(a.)

K Ai
{Upper floor
ly/AV/W/W-uy u? gf "
MBars c ,
•*v/^»v/ AV7/^v//i>y

Wetf/M
Original ground line
{Lower floor
(c) £<//>iV,At

'*&

FIG. 8-20. Walls and footings steadied by floors.

and the substructure the reverse

; is probable. The shear S, however, will
be transmitted by direct pressure or by tension in bars a to the floor slab.
If the frictional resistance of this floor on the ground is sufficient or if the
floor is connected to anything that will serve as an adequate anchorage,
the shear S will not cause the footing to overturn. Furthermore, if the
bottom of the foundation is embedded and if the floor can hold the top of
the pedestal steady, the moment M can cause no angular movement of the
substructure, hence it will be unlikely to do more than add to the force S
another lateral force S' = M/h. Without bars a, tensile resistance is

unreliable without the shelf, vertical loads on the floor slab

; may bend this
252 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

reinforcement or spall the concrete because of possible settlement of the

floor.
How much of a floor can be relied upon to resist lateral forces? Assume
that the columns are 22 ft. c.c, the nearest real contraction joint in the
floor of Fig. 8-20 (a) parallel to beam E is 20 ft. away, the floor is 6 in.
thick, and k' equals 3 ft.

The soil for approximately 1.5h' back from E depends largely upon the
grade beam and footing for its lateral support. Therefore, reliance upon
this strip of soil to support E and C is unsafe. The weight of the effective
concrete of the slab is, approximately,

22 X 15 X 75 = 25,000 lb. +
With a coefficient of friction of 0.7 and a safety factor of 1.5, the tension
that can be relied upon in bars a is

T .
26£MX07 ._ U6001b
1.5

Of course, the grade beam should be keyed to the pedestal, and no trans-
verse contraction joint should be located so that it will prevent the
intended action.
In an industrial plant, the floor was not supported upon the pedestals
of interior column foundations but was made as shown in Fig. 8-20(6).
No thick compressible fillers were used at H and J. Therefore, the top
of the foundation could not rotate without pressing against the floor at
H or J. Since the floor was sufficiently heavy and large in area, it was
relied upon to resist the effects of M and S.
In another plant, there was a long depressed pit or lower floor 10 ft.

deep, asshown in simplified manner in Fig. 8-20 (c). Lateral loads from
columns N and from the earth backfill behind wall R were resisted by
bars c that tied this wall through the floor to the foundations of columns
K. The vertical loads from columns N were distributed to the soil by
wall R and its footing U. The earth pressures on the bottom of R were
resisted by both the earth and the lower floor bearing against U.
In many
instances it is possible and desirable to plan a structure so that
large overturning forces upon footings can be avoided by the action of
other parts of the structure. In some cases, this is not safe because of the
likelihood that the supporting floors or parts may be removed or partly
incapacitated in the future. In still others, it is desirable to plan the
foundations so that they are sufficient by themselves. Which course to
pursue is a question that requires an answer based, at least partly, upon
good engineering judgment.
8-9. Footings subjected to uplift and overturning. The foot-
ings of tall light structures may be subjected to such relatively large
Art. 8-9] UPLIFT AND OVERTURNING 253

overturning the foundations on one side must resist con-

moments that
siderable uplift. In some instances, the uplift may not overcome the
downward dead loads, but the resultant vertical force may be so small
that the footings are not able effectively to resist horizontal shearing
forces and tipping. These foundations may need special care in their
design to assure safety. The prin-
ciples involved will be illustrated by
means of a few specific structures.
Suppose that a high-tension
power line is being built to supply
electric power to a new factory.
The cables are to be supported by
a steel dead-end tower at the sub-
station. Figure 8-21 shows the
Wind
loads)
/
dimensions and assumed forces act- 0.5 k
ing upon one bent of this tower.
HH= I8.2 1*

There are two such bents 12 ft.

c.c, and the front and back of the
tower are both braced strongly.
The front bracing in the plane of
GK a portal frame so that a large
is

transformer may be placed inside 0.3 * Ji

the tower. Two other transform-

ers are to be installed, one on each
side of the tower. How may the
FIG. 8-21. Dimensions and loads for one bent of
foundations be designed to resist the
a dead-end tower.
required forces? Their resistance
to transversewind loads should be tested after the substructure has been
designed to withstand longitudinal wind and the cable pulls T\ and TV
The and the cables is W. Assume
resultant of the dead load of the bent
that the sandy and able to support a bearing pressure of 7 k.s.f.
soil is

It is obvious that the tower will tend to tip about point D even though
the trussing terminates at G in order to simplify the steel details. Taking
moments of all forces about D, the reaction at A is
Ra X 16.5 = 6 X 37.75 + 10 X 32.25 + 0.4 X 28.75 + 0.4 X 23.75
+ 0.5 X 17.5 + 0.6 X 10 + 0.3 X 1 - 5 X 2.5
R A = 35 kips uplift
Rd — 40 kips downward
If the cable pulls 7\ and T 2 are the ultimate strengths of the wires, the
safety factor against overturning may be some such value as 1.25; if these
pulls are the maximum
probable forces caused by dead load, ice, and wind
— as they are in this problem —
the safety factor of the tower and its foun-
254 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

dation should be such that the cables will almost certainly fail before the
tower will collapse or tip over.
Assume that the is to be 1.6. Then the downward force
safety factor
required at A is 35 X 1.6 =
56 kips. Furthermore, if the tower members
shown in Fig. 8-21 are properly designed, the shear will be divided be- H
tween A and D if the foundation of each can resist these forces. For de-
sign purposes, assume 9.1 kips at each pedestal. If member JG is not
supplied in the tower itself, a concrete strut between A and D may serve a
similar equalizing purpose. If members JG and LG are both omitted, the
bearing at A must transmit the entire shear into the substructure.

Top of ground

F F'
fa)-REAR FOOTING fW-SIDE ELEVATION AND COMPUTATION OF RESISTANCE OF SOIL
Port Force in lb. Lever arm, ft. Moment, ft- /b.
1 3.5x2x300 = 2,100 x S 10,500
2 5.5x2x900= 9,900 x J = 29,700
3 15x2x1,500= 22,500 x I = 22,500
S= 34.500 S= 62,700

FIG. 8-22. A trapezoidal footing for support of uplift at rear of a dead-end tower.

Many different foundations may be used for this tower. Figure 8-22
shows one type of isolated footing that will be considered. The rear
footing in Sketch (a) is a trapezoidal or pyramidal concrete block. The
soil around the sides and above the edges EF will be included as useful

weight although the wedging action of the sloping sides may cause one to
question seriously any reliance upon this ground above E and F. By
trial, choose dimensions for the footing until the weight of the concrete
and earth equals at least the desired 56 kips. The concrete footing shown
in the figure weighs approximately 37 kips; the earth within GEFJ 20 ,

kips; the total, 57 kips. This footing will be accepted tentatively for
further analysis.
The net downward load at A is 57 — 35 = 22 kips. Then, applying
the safety factor of 1.6, M = 1.6 X 9.1 X 6.5 = 94 ft.-kips, and e =
94/22 = 4.3 ft. Therefore, the footing cannot offer the necessary resist-
Art. 8-9] UPLIFT AND OVERTURNING 255

ance without being about to turn over unless the passive resistance of
the soil near JF is relied upon or unless more weight is theoretically added
by assuming that the frictional resistance of the soil along planes like GE
and JF is sufficient to enable one to assume that the useful holding-down
weight of earth is that of the soil within planes sloping outward from E
and F at 30° with the vertical. This last might be a safe assumption for
a suddenly applied load. However, the seriousness of a failure and the
possibility of someone digging around the footing for some purpose make
it seem desirable to rely only upon the weight of the earth within vertical
planes around the bottom of the footing.
If the abutting power of the earth is included, with unit resisting pres-

sures of 300/i acting upon the side CF of the footing, the moment of these
resistances about F
approximately 63 ft. -kips, and the resisting force is
is

34 kips, as indicated in Sketch (b). If the soil can be relied upon, it can
resist much of the overturning moment and all of the shear. The moment
to be resisted by the bearing at EF is 94 — 63 = 31 ft. -kips, and e = 31/
22 = 1.4 ft. Then,

22 6 X 1.4 \
Pl
8.5 X 8.5
1 + 8.5 /
0.6 kip (approx)

p2 = (approx)

It does not seem wise to depend upon the lateral resistance of the earth
to such a large extent when failure of the structure might shut down the
plant, and cause loss of life, too. Therefore, although the footing at D
may be made satisfactory in itself, it would have to be large to resist
overturning and, since it is desirable to investigate a different type of
foundation at A, a safer design will K
T=35 If
be sought for both. B C JZZ 1
Would the foundation shown in G
Fig. 8-23 be satisfactory? The
Pedestal
footing is made somewhat deeper
2-6"sq.
and wider than the bottom of Fig. :

c
30
8-22 (a), and the total weight is \30'\
65 k I 7
approximately 65 kips. However, \ i
Footing •

K
I

\i t 1/
a comparison with the analysis of
the preceding trapezoidal footing
E
shows that this one is not suffi- 9-0"sq. AJjT
ciently broad and heavy to remedy
FIG. 8-23. A proposed isolated spread footing.
the overturning danger, and the
narrow pedestal will offer even less sideward bearing resistance against the
soil. The ground that is actually in contact with the concrete pedestal is,
of course, backfill, and it may be rather compressible. The soil on top of
the footing but beside the pedestal cannot be relied upon for lateral resist-
256 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

ance even though its weight is useful for stability. These isolated footings
are more suitable when the chords of the tower trussing are spread much
farther apart. Therefore, look for a design that is inherently more
satisfactory.
In Fig. 8-24 (a) the pedestals at A and D under each side of the dead-end
tower of Fig. 8-21 are placed upon a common footing. The reactions at

J^K. D^\

{Footing
sRi'b
53
1 r\A 9\ z>-h 1

A
L 3 L 9" l6'-6"c.c. J-'?" Mat-*
t
24-0" ,
fey- PLAN OF TWO RIBS AND MAT
(ah PLAN OF TWO LONG FOOTINGS

J-4. P/'bj^ ground^

r^-SIDE ELEVATION A-A AND PRESSURE DIAGRAM (d)-PLAN OF FOUR RIBS AND MAT

FIG. 8-24. Combined footings for a dead-end tower.

A and D cause a tendency to break the footing between them, and also
(indirectly) the part in front of point D. When the forces are large, it is
desirable to use a rib under the bearings as shown in Sketch (b), the foot-
ing merely being a cantilever to resist upward pressure in front and down-
ward weight in the rear. A floor to support a transformer within the
tower may be placed across the two ribs, but this will not be counted upon
for stability. For the dimensions shown in (6), W = 98 kips. The
safety factor (S.F.) against overturning about F is
1

S.F. = 98 X 12 + 40 X 3.75 1.57

35 X 20.25 + 18.2 X 7.5
1
The center of moments might be a foot or so to the left of F when on earth.
Art. 8-9] UPLIFT AND OVERTURNING 257

The resultant vertical load is 98 + 40 - 35 = 103 kips, and is 2.24 ft.

from D. The pressure diagram is shown in Sketch (b). Friction alone
under the footing is able to resist sliding without dependence upon the
abutting power of the earth in front of the substructure. This design
seems to be a more satisfactory solution.
A possible modification of this design is that in Sketch (c) where the two
footings are joined as a single mat. This may be desirable if the tower is
sufficiently narrow. When large forces may be in two directions, as in-
dicated in (d), and when H' may be reversible, the four-rib and mat con-
struction shown may provide a simple and reliable foundation.
The construction shown in Fig. 8-25 is still another variety of founda-
tion for this tower. It consists of a heavy cross wall and footing at the

Rear footing

-m B
a_|"§L.

L
Front^
footing

fcJ-SECTION B-B rrfJ-SECTIONC-C

Z'-O" 3'-0" .
3^0"
J .| ,|

l6'-6"
(b)- SECTION A-A
faJ-PLAN OF FOUNDATION

FIG. 8-25. Design for foundation of a dead-end tower at an industrial plant.

rear, a smaller cross wall and footing in front, and two ribs to resist the
tipping tendency of the footings. Therefore, for practical purposes,
the structure may be analyzed as though it tends to tip about rib CD, the
leverage of the rear footing is AD, the pressure under the front footing is

reasonably uniform, and friction and the resistance of a large volume of

earth prevent sliding. The weight of the rear part minus the ribs is 144
kips; the ribs, 13 kips; and the front part (CD) 24 kips. Upon these
assumptions, the safety factor against overturning about CD is, taking
moments for the entire structure about the bottom and middle of the
front footing (a conservative procedure),

S.F. == 144 X 16.5 + 13 X 7.75

= 1 g6
70 X 16.5 + 36.4 + 5

Then, taking moments about AB at the level of the bottom of the front
footing, the unit bearing pressure under CD is

= 13 X 8.75 + (80 + 24)16.5 + 36.4 X 5

= 2.5 k.s.f.
Pi
16.5(3 X 16)
258 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

This foundation may be made slightly smaller if the safety factor is only
1.6. It should be noticed that the necessity of holding down the back of
this tower automatically results in such large footings that the bearing
pressure upon the ground is small. For such a structure, the wind should
be classed as a live load, and the allowable unit stresses should not be
increased because of this simultaneous action; the structure is too
important.
Comparing the volumes of excavation and concrete, and the practica-
two suggested designs, the footing in Fig. 8-24(6) seems
bility of these last
to be more economical, but one of the general size and shape of that in

FIG. 8-26. Conveyor trestle under construction at the Morenci Reduction Works, Morenci, Ariz.
Notice the single bent and the braced anchor tower. The rectangular chute above the railroad tracks
will be used for filling cars. (Courtesy of the Phelps Dodge Corp.)

Fig. 8-25 will be safer and will not depend so largely upon the weight of
the earth above There have been instances in which a workman has
it.

removed the soil on top of a footing while installing pipes, ducts, or an

adjacent structure, and he did not suspect that the designer had depended
upon that soil remaining undisturbed.
There are many other structures which must resist considerable over-
turning because of wind, but for which the action of the latter is not so
critical. For example, consider the foundation for a single steel bent to
support a large elevated conveyor gallery like that shown in Fig. 8-26,
assuming good sandy soil.
The bent is intended to support vertical loads and transverse wind loads
only. In some cases, where the overturning moment is small or where the
legs of the trussing are flared apart to provide large leverage, isolated
Art. 8-10] UPLIFT AND OVERTURNING 259

trapezoidal or spread footings may be adequate and economical. In fact,

lateral spreading of the columns to eliminate any net uplift at the wind-
ward bearing is basically good planning. When the bent is narrow, how-
ever, it is advisable to use a single narrow footing under the pedestals, as
shown in Figs. 8-27 (a) and (6).
1.

/-o"/^
., PedesfaI Footing,
**-
u
Jx,
f
LT^s
fa.)- PLAN M fe,l-PLAN

S~ A
Rotation

Roughened Roughened
l

fa, X
n j|'ij'\\
const jt. consf.jt.~~^0.\ |° 'Adc

e a
teJ-SIDE ELEVATION faJ-SIDE ELEVATION

FIG. 8-27. Some footings for a single bent.

A feature that may

be important in the design of some bents is the
participation stress in the diagonals.For example, the structure may be
braced as shown in Fig. 8-28, with member AB missing. When members
FC and GD are compressed, the diag-
onalsAG and FB are also compressed
and may have a unit stress approxi-
mately equal to one-half that in the
columns. There is then a horizontal
component H
that may try to spread
the pedestals slightly. However, the
greater danger lies in the severe bend-
ing and shear caused in CAF and DBG.
On the other hand, if AG and FB are
very slender so that they have little
value except as tension members, the
wind shear may be concentrated at
by
the windward pedestal. Member AB FIG. 8-28. Overturning forces caused
participation stresses.
should be used as a part of the bracing
system. A strut between the pedestals as a part of the foundation
will support the pedestals, but it will not relieve the local stresses in the
steelwork.
Long
trestles, conveyor galleries, and viaducts that cannot be anchored
to abutments or other structures generally must have an anchor tower to
resist longitudinal wind as well as transverse forces. The foundation
shown in Fig. 8-29 is only one type that may be used when side clearances
260 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

require the use of a narrow superstructure.

This superstructure's trusses
are supposed to be parallel to AD
The substructure should be
and BC.
analyzed separately for transverse, longitudinal, and diagonal wind.
For the last, the general method indicated by Eq. (8-8) may be used,
although the transfer of "longitudinal" shearing forces between parts
of the substructure to bring into action all the bearing areas may be
questionable.

12-0
r Direction of rotation
Aj2Anchorbo/fs flu.

B
Consfr.
joint
I
T\ 5i mV
A Id
B m
AL_ _fA
(a)-PLAhi W-SIDE ELEVATION A-A

FIG. 8-29. One type of foundation for a tall trestle tower.

rcy-SECTION B-B

fe^-PLAN OF FOUNDATION

FIG. 8-30. Foundation for a tall conveyor tower at edge of slope or on a hillside.

Sometimes it is necessary to found bents and anchor towers on the

slopes of cuts or on steep hillsides. The bearings may be placed at dif-
ferent levels down the slope, or the substructure may be built to provide
level bearings for the steelwork, as shown for one instance in Fig. 8-30.
The overhead structure may be parallel or perpendicular to the slope.
The center of the bottom of the footing F in (b) should be considered
as the point of rotation for downhill tipping. As the supporting power of
the earth against downhill shearing at footing E is questionable, the ribs
are made so that all the force may be transmitted to F. For tipping
parallel to the hillside, the side AC will rotate about F, whereas will BD
 :

Art. 8-10] UPLIFT AND OVERTURNING 261

try to rotate about E. If the front and rear portions are safe when con-
sidered as independent halves "cut" along the center between A and B,
the combined structure will probably be safe also. However, it is some-
times desirable to analyze the structure as an entity and to see that it
cannot tip about line FE or about either downhill corner.
In some cases, when the soil is suitably stiff, footings for towers may
be made by drilling holes in the ground and "belling out" the bottom
somewhat as pictured in Fig. 9-7. The hole is then filled with concrete
containing the necessary reinforcement.
Example. Figure 8-3 1(a) shows a steel A-frame that is to be used to
support one side of an outdoor crane runway. Assume that Pi = 10 kips
DL and 70 kips LL, and that *S = 10 kips. The soil is deep medium clay
that has been covered with 3 ft. of cinder fill. The location is Baltimore,
Md. The longitudinal bracing is not connected to this particular frame.
Design a suitable foundation for this superstructure, using a safety factor
of 2.
The cinder fill is not suitable for the support of a structure like this one
where appreciable subsidence, uneven settlement of neighboring footings,
and tilting caused by unequal settlement of the ends of any individual
footing will affect the alignment of the rails seriously. The footing should
therefore be placed well down into the clay. A maximum allowable unit
pressure of only 4 k.s.f. will be used in order to avoid serious pumping
action and compaction.
Since the top of the frame is Pi should be
free to tilt sidewise, the load
assumed to act fully upon point A =
because participation-stress
(S 0),
action will not affect B unless the top cannot move horizontally. The
web members and back leg are to brace the front column and obtain
anchorage to keep the column from tipping over.
The critical case for tipping may be when P is small but S is large, as x

when the trolley is at the opposite end of the crane bridge. For pressure
at A, the critical loading condition is likely to be when Pi is a maximum
and S acts toward the left in Fig. 8-31 (a). Analyze the latter situation.
Consider the following designs for the substructure
1. Try separate footings under A and B. The uplift at B for design
purposes is UB = 2 X 10 X (28/8) = 70 kips. This requires the equiva-
lent of approximately 70/0.15 = 467 ft.
3
of concrete, close to an 8-ft. cube
and obviously unreasonable. The downward reaction at A is

10 * 28
pA = 80 + o
= 115 kips

or 2 X 70 + 2X10X28
o
+ io = 220 kips

when including the safety factor. This footing must then have an area in
excess of 115/4 = 29 ft.
2
262 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

//\*
Cor faced £
*7$r
o"o'«
I

2&
I!

is A $&

X^Anchor bo/ts Anchor bolts

s '°", Reactions
<
\

Ground not shown

faJ-ELEVATION OF A-FRAME ^-DETAILS OF BASE OF A-FRAME

4*
>

S b\
u° \
#< ' \

2'd 2-0" 410" 2-0 ' 6-0" 2-0 4-0"

5>

M^
Front Elevation
-s
9-0" *&
/8-0"

Side Elevation
rW=S3.4k 9Iq»

fcJ-JML FOUNDATION FOR A-FRAME

4.9' ,
9.8'

•3.7**'-
P,
R=l63.4 k
frfJ-PRESSURE DIAGRAM FORDL+LL
3.4' ,
6.8'

p^i63 k/eL,
~R=233.4 k
feJ-PRESSURE DIAGRAM T0RDL+2LL
P. ~*%
s^-M *1!

2-0 t-6'
!'
:
b" 8 L0" W2'0"%^

S'O" W=72.3 k 90 ± >

18-0"

Front Elevation Side Elevation

tfV-ALTERNATE DESIGN FOR FOUNDATION OF A-FRAME

S.6' .,. 11.2'

n LLiiU — -^T/S2J k
(g)- PRESSURE DIAGRAM FOR DL+LL
4.5' 9.0'

^ afi5
^lMI - '^<R=222.3 k
fAJ-PRESSURE DIAGRAM FOR DL+2LL

FIG. 8-31. Support for outdoor crane runway.

Art. 8-10] ANCHORS FOR BACKSTAYS 263

These footings will practically bump into each other, and no allowance
has yet been made for the additional effects of overturning caused by the
shear S/2 applied at the tops of the pedestals at A and B. It seems,
therefore, that a combined footing should be used.
2. Try a foundation like that shown in Fig. 8-3 1(c). The pressure dia-
gram dead load -\-Pi + S is shown in (d) that for dead load
for the ;

+ 2Pi + 2$, in (e). These are computed as though the shear at the top
of the substructure is S or 2S, that it acts upon a solid structure, and that
it has a lever arm of 7 ft. above the bearing area. This foundation may
be improved.
3. Try the foundation shown in Sketch (/). This provides relatively
more bearing area in the vicinity of A and has a continuous rib, and the
footing is narrower and thinner. The pressure diagrams for the two load
cases are shown in (g) and (h). The rib wall now serves as the backbone
of the structure, and the footing is not subjected to serious bending and
shear. The longer projection in front of A also reduces the maximum toe
pressure.
The third design seems to be the best of the three. It is probably im-
practical to use a foundation that will have absolutely uniform pressure
under it for the load Pi alone. These loads are temporary and will not
produce flowing of the soil unless the pressures are excessive. In a certain
industrial plant, a high edge pressure was disregarded. The clay yielded
progressively. According to the latest reports, the settlement of the bear-
ing that corresponds to A of Fig. 8-31 (a) amounted to 6 or 8 in. for some of
the foundations so that successive adjustments have been made by shim-
ming, welding on extensions to the anchor bolts, and building and rebuild-
ing concrete pads under the base plates in order to line up the crane
girders. The results of this faulty design have been both unfortunate and
expensive.
8-10. Anchors for backstays. A column and backstay, such as pic-
tured in Fig. 8-32 (a), constitute a simple and economical structure to
resist horizontal and inclined forces applied to its top. In some cases, to
secure lateral stability, it may be desirable to use two columns that are
sloped to form an A-frame; in others, to use a single column and two
backstays that are splayed at an angle of 60 to 90°, or more.
For simplicity of illustration, the structure shown is assumed to consist
of a singlecolumn BC that is steadied laterally in a plane perpendicular to
the drawing, and a single wire-rope backstay AC. When a force is applied
at C with components H and V, the former must be resisted by the back-
stay because the column is practically free to rotate about its base. The
tension in the backstay is therefore T, as shown by the force diagram, and
the forces applied to the foundations by the frame are as indicated, the
dead load of the structure being neglected.
264 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

It is obvious that the footing under B may be designed for vertical loads
only. When BC is an A-frame, this foundation should be designed to
and whatever transverse forces may exist. The
resist the vertical loads
A-frame may be considered to constitute a single bent, and it generally is
braced to act as one.

/
b&
Force D(
diaqram E~^ ,

WfKR Friction F
must equalIf
fW-RECTANGULAR BLOCK

V^J
V'-Y.V
fc)-0IAGRAMOF N
fai-ELEVATION OF POST AND BACKSTAY VERTICAL PRESSURE

(d) -TRAPEZOIDAL BLOCK

G
r/7-COMBINED ANCHOR
R
<r
Headwa//

Footinq E \ °

feJ-PEDESTAL AND FOOTING ^-COMBINED ANCHOR

FIG. 8-32. Anchorage for a backstay.

The forces applied to the foundation at A try to pull it out of the

ground, to drag toward B, and to tip it over. The action of a simple
it

—
gravity anchor one dependent upon its weight alone is pictured in Fig. —
8-32(6). The horizontal component of T is resisted by the friction F, and
these forces constitute a couple H X DE that tends to tip the block about
G. The force W is the weight of the block treated as a single force at the
center of gravity, then W-V is the vertical component of the resultant R.
The righting moment that resists tipping is (W-V) EG, unless the pressure
under G is so large that possible yielding of the earth makes it advisable to
 .

Art. 8-10] ANCHORS FOR BACKSTAYS 265

assume a point of rotation closer to E. If the forces are as shown to scale

in Sketch (b), the pressure diagram may be assumed to be as pictured in
the upper sketch in (c). The safety factor against uplift is against W/V ;

overturning, W X EG/( V X EG + H X DE) ; against sliding, Wf/

(V'f +
H), where / is the allowable coefficient of friction, or friction plus
cohesion. However, the last two may be deceptive. It is advisable to
increase T by the required safety factor, then to compute the maximum
pressure at G. When the latter is excessive and the pressure diagram is

supposedly like that pictured by the solid lines in the lower sketch in (c),

the pressure diagram may be more nearly like that shown by the dashed
and the actual point of rotation may be at some uncertain point J
lines,
Rather than try to locate J theoretically, one might better modify the
design to ensure safety.
In such anchorages as these that are not on rock, it is usually desirable
for reasons of economy ground so that the
to sink the anchorage into the
abutting power of the soil in front of the foundation can be utilized to
assist in the resistance to sliding and overturning. For analysis, one may
use the methods explained in the previous article. However, when test-
ing for the desired safety factor, T' = T(S.F.) should be used, and the
analysis made accordingly. In fact, if one assumes T' in the first place
and then designs a foundation that will not quite fail under those condi-
tions, he will know that his structure is safe, and he need not care what
the pressures and stresses may be under operating conditions.
Many are the varieties of anchor that may be used to hold a backstay.
Some are shown in Fig. 8-32. For the correspondingly labeled sketches
the following comments are given:
(d) This is designed to shift W
toward the rear, to increase the leverage
EG, to equalize the pressure on the soil somewhat, and to utilize the resist-
ance of the earth H a The soil above KL produces useful weight, but the
.

possibility of its sliding along KL may cause doubts as to the reliability of

that part of the soil above L in producing lateral resistance. The friction
of soil on smooth concrete may be less than that of earth on earth.
Notice that the sliding is to be resisted by a friction force equal to Wf and
the abutting power H a of the soil in front of the anchorage. It may be
that the engineer will want to put a "cutoff wall" or concrete projection
under the left end of the bottom beyond E so as to increase the shearing
resistance. However, he should be careful not to depend upon the
frictional resistance of the soil to develop abutting power and, at the
same time, rely upon the same frictional forces to produce the frictional
resistance Wf.
(e) This design shows an attempt to minimize the quantity of concrete,
and to utilize the weight of earth above MG
as a part of W. This plan is
not likely to be desirable for large pulls.
266 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

(/) This is an L-shaped concrete slab with a central stiffening rib. It

attempts to utilize a large volume of earth for weight and to have con-
siderable breadth to bear horizontally against the soil. The height of
point N may cause large eccentricity of the resultant at the base.
(g) Here the structure in (/) is practically reversed. The advantage
gained is the low position of point and the decrease in the eccentricity
of R.
In connection with all these anchorages shown in Figs. 8-32 (d), (e), (/),
and (gf), notice the following:
1. The wire rope backstay is assumed to have a clevis at its lower end.

Between the clevis and the steelwork of the anchorage there should be a
turnbuckle or some other device to permit adjustability of the backstay.
2. The steelwork of the anchorage should be thick so as to reduce the

danger from rusting at its junction with the concrete.

3. The steel should terminate in such a manner that it engages the con-

crete mechanically rather than through bond alone.

4. The concrete should be reinforced so that it cannot fail as a whole

or in detail.
Sometimes a deadman is used as an anchor for a backstay. This may
consist of any one of various arrangements. For example, a concrete
block somewhat like Fig. 8-32 (d) may be buried several feet into the
ground so as to get a lot of earth on top of it and to develop large abutting
power. Such a block should have a flat face normal to the pull, because
beveled faces might enable the block to act somewhat like a wedge
between masses of earth. Sometimes these footings can be precast with
long pedestals and set in holes in the ground. Sometimes a group of piles
may be used somewhat as pictured in Fig. 10-40 (a).
When bedrock is at or near the surface, a backstay may be anchored
directly to it. At first glance, it might seem that the anchorage shown in
Fig. 8-33 (a) is ideal. To be the question of how
considered, however, is

the inclined hole can be drilled. If the eyebolt AB

with an upset end at
B is to be grouted or concreted in place thoroughly, the diameter of the
hole should be large. The vertical hole shown in (c) is therefore more
practicable because it can be made with a churn drill. The anchor should
be bent to a long-radius curve at F where it is embedded in the concrete.
The rock near G should be broken off or roughened to provide adequate
shearing resistance so that the block J will not slip on the rock.
The following comments also apply to the construction shown in Fig.
8-33:
1. Instead of the upset end B, the Fox bolt arrangement in (6) may be
used. Unless the wedge E is welded to D, failure of the bond along the
rod may cause the latter to pull away from the wedge. A large welded
bead around the end of the rod may serve as an even better anchor.
Art. 8-11 PEDESTALS ON LARGE MATS 267

2. The threaded end H, if long enough, will be an excellent anchorage

because of the high strength in bond.

3. It may be cheaper to thread the rod and put on a standard clevis or

other fitting, E, than to forge and drill the rod as at A.

4. The rock must be able to resist the pull without cracking and

spalling.
5. The drilled hole should be deep enough to provide a reliable anchor-
age. The top foot of the hole may have little value because of local
weakness of the rock.
6. The concrete pad around the top of the anchor rod in (a) is useful
principally to make a workmanlike job, to protect the steel near the rock
line, and to prevent serious bending in the rod due to any cause.

t
i 6 lT
Rock\ a*
*Wchurn-dri/lecl
\ hole
%
I
1

faJ-BOLT IN INCLINED fW-FOX BOLT TcJ- BENT BOLT IN

-
HOLE VERTICAL HOLE

FIG. 8-33. Anchorage in rock for a backstay.

When planning the structural connection between a foundation and a

member thatis subjected to violent and sudden uplift, one should remem-

ber that long highly stressed anchor bolts or steel ties will deform appre-
ciably. If anchor rods 10 ft. long connect the end of such a member to a
buried beam or girder near the bottom of the concrete foundation and if
these rods are stressed an average of 15,000 p.s.i., the elongation may be

approximately 3a 6 in- Under shocks and repeated or reversible loads, as

in a large headframe like that in Fig. 8-55, this may cause objectionable
vibration and "chattering" of the superstructure even if it does not crack
the concrete. Prestressing of the rods may be desirable.
8-11. Pedestals on large mats. It is possible to have a condition in
which a group or series of structural parts or machines may be founded
upon a large common mat. One, several, or all parts may be loaded at a
given time, and they may cause overturning moments and shears as well
as vertical forces. A common mat may have been chosen because of the
268 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

low bearing value of the soil, the desire to tie all parts together as a unit,
the advisability of spreading the effects of any local and variable loads,
and the need of a concrete floor over the area so that it is economical to
incorporate this floor in the substructure itself.

When critical cases are encountered, careful analysis of a tentative plan

may be necessary. However, empirical design is generally sufficient for
most practicable structures as far as overturning effects are concerned.
If the use of a mat entails serious uncertainties, one might better change
the basic plan to some more trustworthy and determinable arrangement.
Assume for illustration that Fig. 8-34 (a) is the plan of the substructure
for a battery of tall vertical cylindrical steel tanks 5
ft. in diameter, which

may be full or any given time. Assume further that they are
empty at
part of the equipment of an industrial plant in a warm climate where there
is no danger from frost and that they may be subjected to hurricane

winds.
Each tank
is on a base or pedestal of the dimensions shown in (a) and

(6). For supporting vertical loads, the mat is to have two longitudinal
and four transverse reinforced strips 10 ft. wide. These strips may be
proportioned to spread the localized loads as already explained in Chap. 7.
However, if the loading on each longitudinal strip is as shown in Sketch
(6), what be the effect of the overturning forces upon the mat?
will
is shown an exaggerated picture of the deformations that
In Fig. 8-34(c)
might occur because of the moments M
and shears S of Sketch (6). As
each tank tries to tip toward the right, the end A of the mat will tend to
lift up, causing little restraint except for its dead-load moment. End D
will tend to act one side of a footing subjected to overturning.
like
Between E and H the adjacent
tanks cause bending and transverse shears
in the mat, and the uplift at R is restrained through the mat by the com-
pression at Q. The shear caused by F between QR is pictured by —V n ;

that caused by E, by V n Somewhere between adjacent pedestals there

.

will be a point of inflection in the mat.

The bottom reinforcement under Q should be proportioned to resist all
of M
e the top steel at R and the bottom steel under T, for jHjM/; the top
;

at U and bottom under W, for }>i,M g and the top bars at X and bottom
;

below Y for }4Mh- Any bending caused by the shears $ should be

resisted similarly. Of course, the mat should be reinforced for wind from
the other directions also.
The preceding recommendation is based upon the assumption that the
mat is stiff enough to prevent distortions that would affect greatly the
bearing on the soil. The rotations might cause some redistribution of
pressure as indicated by Fig. 8-34 (d), but if the mat is made strong
enough to withstand the moments by itself, the structure will not fail, and
moderate variations of the pressure upon the soil need not cause worry.
Art. 8-11] PEDESTALS ON LARGE MATS 269

If the pedestals are strong enough, the critical bending will occur at
their bottom corners; otherwise, the greatest bending will probably be
near the leeward side of the tanks and near the windward anchor bolts.
Transverse reinforcement is also needed to spread the effects of localized
forces.
Strip I0'-0

foJ-PLAN OF COMPLETE FOUNDATION

Reactions r~^Me r-<Mg

not shown ,pe
E\ se f\ _ Sf.
I
'iwV'V/v-w-w W.W'AVVA 1

C6J-SIDE ELEVATION OF FOUNDATION

/Shearing forces
E
iq >fcrif -i >!/W

Va Vt2 Vi3

fcJ-EXAGGERATED PICTURE OF DEFORMATION OF MAT

a r \Q R\ \T U\ \W X\ \TD
mvj\
mTrmrJ^S SS
\Bl \&2 \&3 \E4
rdMMPROBABLE DISTRIBUTION OF PRESSURE

FIG. 8-34. Overturning action of a series of pedestals on a large mat.

Some further pictures of the suggested construction for the preceding

foundation are shown in Fig. 8-35. Notice the following.
1. The vertical load will cause tension in the bottom steel of the mat,

and the bars are extended clear across under the pedestals.
2. The anchor bolts should be embedded thoroughly. If they are short
like g in Sketch (c), the bond stresses on the bolts and reinforcement may
be too large.
270 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

3. The top bars, like b and d in the mat, are extended clear across the

pedestal to obtain proper anchorage for tension both at R and at T.

4. The pedestal is poured after the mat has set. The bars / are
merely to knit the top together, whereas h does this and also ties down
the pedestal.
5. The shearing stresses in the mat may be assumed to be equal to

XM/L, where L is the spacing of the pedestals except in the case of E, the
one on the windward side. Here the shear near Q may be considerably
higher, perhaps by 50 per cent.

rE -ZM
19 JP
U4J&__JU^Li
—K
f!

— Th u
(ah DEAD L0AD+ LIVE LOAD
PRESSURE DIAGRAM. E
(c)- SHEARING REACTION ON PORTION AT
ASSUMED TO BE TYPICAL

j:m
F *1—-—
I A f£" T* F
I 1 I I

_u_.
V, ,r?c Roughened <e V-V+
t2
Tt 5^
const, jf.

r^J-SHEARING REACTIONS ON PORTION AT.P rrfj- SHEARING REACTION ON PORTION AT if

FIG. 8-35. Assumed forces on mat because of overturning.

6. The transverse reinforcement is not shown in detail. It may be

determined in the same general manner as is the longitudinal steel.
7. In general, it is advisable to determine the required top and bottom
reinforcement of the arbitrarily chosen strips — 10 ft. wide in Fig. 8-34 (a)
— for dead load and live load; test for the combination with overturning,
allowing 1.3 times the permissible unit stresses; increase the reinforcement
of the mat to resist the combined loads if necessary; and then extend the
reinforcement clear across the mat, except for minor local additions
that may be needed in the bottom under the end pedestals, such as E
and H
One cannot expect to illustrate all possible problems that a designer may

encounter. However, problem shows how one may assume a definite

this
system in an indeterminate mat, then make this system able to resist the
applied loads, relying upon the probability that the structure will not fail

if there is a way in which it can stand up. On the other hand, the choice
Art. 8-12] RETAINING WALLS 271

of such a system should be made with good judgment and with reasonable
probability that the structure can act as intended.
When a single local overturning moment of large magnitude is applied
in an intermediate position upon a large mat, or when numerous such
moments are applied too far apart to cause trustworthy counteracting
resistance in the mat as illustrated in Fig. 8-34, then it is probably advis-
able to design the mat near the load so that it will be strong enough to act
somewhat like an isolated spread footing. A strong pedestal on top or a
suitable thickening below, as in Fig. 7-4(6), may be satisfactory. Even
the construction of a depressed rib, as in Fig. 7-7 (a), to form a stiff beam
in the desired position may be advantageous. In any case, one should be
sure that the substructure cannot tear apart locally near the edge of the
structure that applies the loads and that the concrete will not fail there by
shearing.
8-12. Retaining walls. A brief discussion of the design of retaining
walls is necessarily a part of this chapter, but more complete data are
1
published elsewhere.
In Fig. 8-36 (o) is shown a T-shaped reinforced-concrete retaining wall
that is to hold back a sloping embankment. The lateral pressure of the
earth and of any water impounded behind the wall tend to tip it over
about A, to slide it along the soil under the footing AB, and to break its

individual parts as indicated in (c). Referring to the various sketches in

Fig. 8-36, the design procedure may be described briefly as follows:
Choose a height and shape of wall to meet the general requirements
1.

of foundation, property lines, and any other local conditions that influence
the problem. The width of the footing L should be approximately one-
half the height EJ of the wall, the thicknesses of parts should be substan-
tial, and AB should be below frost depth.

2. For stability and resistance to sliding, 2W equals the weight of the

concrete and of the earth above the footing AB, applied at the center of
gravity x .

According to Coulomb's formula, 2 the lateral thrust of the

3. soil may
be assumed to be
COS0
Pe = V whl
2 . .
=- (8-12)
(1 + V2 sin 2
— 2 sin cos tan 5)
2

where the various terms are as shown in (a), and w is the unit weight of
the soil. When = 30°, 5 = 0, w = 100 p.c.f., and hi = the depth from
the surface of the ground to the point considered, Eq. (8-12) becomes

Pe = \bh\ (8-13)
1
Clarence W. Dunham, "The Theory and Practice of Reinforced Concrete," chap.
8, McGraw-Hill Book Company, Inc., New York, 1953.
2
A somewhat different expression is used in connection with Fig. 5 in the Appendix.
 -

272 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

topjZT""

Front \>

of wall

^-ADDITIONAL PRESSURE ^-EXAGGERATED PICTURE

DIAGRAM- CAUSED BY OF DEFORMATION
IMPOUNDED WATER

raj-GENERAL DIAGRAM

NK

ACTING ON "Pressure
(tf J-FORCES STEM
^Ufi diagram

feJ-FORCES ACTING Pressure

ON HEEL diagram
69- FORCES ACTING ON TOE

FIG. 8-36. Data for the analysis of a reinforced-concrete retaining wall.

The direction of the thrust is assumed to be inclined downward at the

angle of repose or of internal friction of the soil, and the distribution of the
pressure assumed to be triangular or hydrostatic. If the embankment
is

supports a roadway close to the wall, the vibration makes the magnitude
and direction of P e uncertain, and it is more conservative to assume the
full value of P e acting horizontally.
Art. 8-12] RETAINING WALLS 273

4. When upon rock or fairly impervious soil, water may

the wall rests
collect behind it removed by longitudinal drains or weepers. The
unless
additional thrust caused by water pressure may be assumed as equivalent
to the hydrostatic pressure caused by a liquid with a unit weight of 45
p.c.f., as shown in (6). This gives approximately the same computed
total pressures when combined with those calculated from Eq. (8-12)
using the dry weight of earth as would be found by using the water as
62.5 p.c.f. and the buoyant weight of earth in Eq. (8-12).
5. The total horizontal force, neglecting surcharge and frost action, is

XH = Hi -f- P w and it is applied as their resultant at y

, .

6. The safety of the structure may be tested by multiplying 2// by the

safety factor, usually 1.5 to 2.0, and analyzing the wall for this condition.
For illustrative purposes, a safety factor of 1.75 is used, which is satis-
factory for most purposes. For these ultimate conditions, the yield point
stress of steel, 1.75 X /< for concrete, the ultimate compressive resistance
of the soil, and the maximum coefficient of sliding friction / may be used
as limiting values.
7. For testing the stability against overturning, take moments about A,
or a point somewhat behind it if the pressure at the toe too large.
is Then
the righting moment should at least equal the overturning moment, or

S!F(L - Xo) > 1.75(2H)y (8-14)

8. For testing the stability against sliding, the allowable frictional

resistance should at least equal the active horizontal force, or

(ZW)f > 1.75(2//) (8-15)

When a large wall rests upon plastic soil, the designer may have difficulty
in providing sufficient resistance to sliding. In some cases, he may
include with the friction the abutting power H s of the earth in front of the
wall, as shown in Sketch (a), assuming

ff
«^/l+J»nA (8 _ 16)
2 \1 — sin cj)/

as computed from the Rankine formula, Eq. (8-10). However, this may
be dangerous if someone excavates in front of the wall.

9. The eccentricity of the resultant R may be found as indicated in

Sketch (a) and as explained in Art. 8-6. The soil pressures and the pres-
sure diagram may then be determined by means of Eq. (8-4) or (8-6).
10. The maximum bending and shear in the stem of the wall are at OD,
Sketch (</). Neglecting the weight of the stem and multiplying the
274 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

forces by the safety factor,

Mod = 1.75 (h 2
^ + P'w ^\ (8-17)

and Hod = 175(ff 2 + P'J (8-18)

11. The critical section of the heel is at S, Sketch (e), where bars a and b
cross each other. The maximum bending at S for 1.75 times the hori-
zontal forces is

M s
— W e X lever arm -j- W c X lever arm — U h X lever arm (8-19)

The shear at S is

V =
s W +We c - Uh (8-20)

These two formulas are left in a generalized form because the shapes and
magnitudes of the earth mass e the heel c W
and the pressure diagram
, W ,

Uh vary with specific cases. It is also desirable to analyze the heel for
the condition using the ordinary computed values of the horizontal forces,
the accompanying pressure diagram, and customary working stresses
because this case may control the design.
12. The bending in the toe
critical will be at OU of Sketch (/). Using
symbols similar to those in item 11,

M = U X t lever arm — W' X lever arm — W' X

e
lever arm (8-21)
and V = U - W' - W't e (8-22)

This part of the structure should also be analyzed to see that it is safe for
the ordinary forces and unit stresses.
When a retaining wall is to be 25 ft. or more high, the designer might
well consider making it a counterforted structure because of the large
bending moments which would occur at the junction of the stem and the
footing of such a large wall if made T-shaped. This statement is based
on practical reasons rather than economic reasons alone
Let Fig. 8-36A(a) picture the cross section of a counterforted wall,
and Sketch (b), a partial plan. The following is an outline of the proce-
dure for the proportioning and analysis of such a retaining wall
1. The footing GH should be somewhat over one-half of the height
CI as for a T-shaped wall, the dimension GH being larger when the soils
have low bearing or shearing strength.
2. The toe EB should be proportioned as for a T-shaped wall— prob-

ably 25 to 33 per cent of GH. It should be designed as a cantilever with

a fulcrum at B, but the heel (or the stem also) must be able to "back up"
the toe in order to avoid actual twisting of the footing.
3. The spacing s of the counterforts should not be too large perhaps —
about 10 ft.
Art. 8-12] RETAINING WALLS 275

4. The stem is to be designed as a continuous vertical slab with hori-

zontal earth pressures acting on its back. It can therefore be rather
thin, should be thickened near the bottom and reinforced to tie
but it

it into the footing and to enable the stem to offer enough resistance to

twisting so that the toe will develop the necessary bending moment along
line B. Another way is stem loose from the footing between
to "cut" the
counterforts as far as bending moments
are concerned and to assume that
it offers a vertical reaction line for the footing but no resistance to twisting

at its bottom. In general, tensile reinforcement will be needed at J and

K near the back of the stem and at N between counterforts.

\:
<;
II

1
1
J
1 1

N 1

i| K P ,

ii
<o

1 1

I"

id) PLAN

FIG. 8-36A. Counterforted retaining wal

5. The heel is to be designed to act as a continuous slab with supports

at the counterforts and along the stem, with a net load equal to the
differencebetween the loads on top of the heel and the upward-bearing
pressures below it. Design the rear edge (12 in. width) across the bot-
toms of the counterforts, with tensile steel ordinarily in the top near
and P and in the bottom near Q. Then use this same reinforcement
most of the way between OJ and PK, with some reduction close to the
rear of the stem. The footing steel a should be carried well into the heel;
if the toe is to be tied into the stem, some bars a should be bent up into
the latter.
6. The counterforts are really somewhat like tapering stems of T-beams

with heavy tension brought into bars b by the reactions from the heel.
There are a shear along the junction CD and a compression in the stem.
It is very important to have the bottoms of bars b hooked and well
embedded; or U-shaped dowels may be placed in the footings so as to
276 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

avoid holding long bars before the footings are concreted; then b can
be bonded to these dowels.
7. Stability and sliding can be investigated as for other types of walls,

but it is convenient to test a unit extending from one counterfort to the

—
middles of the spaces on each side 10 ft. for counterforts at 10 ft. c.c.
8. Test all parts for bending and shear as for other reinforced-concrete

construction.
9. At contraction or expansion joints, locate the joint between two
counterforts, which are perhaps 5 ft. c.c, so that the stems and footings
can cantilever from the counterforts to the joints.
Walls of many types and shapes may be used. Construction joints,
contraction joints, paneling for architectural appearance, batter of the
front for aesthetic reasons, drainage, cutoff walls under the footings,
sloped or stepped copings and footings, waterproofing, and surface
finish — all and founda-
these are important, as in the case of basement
tion walls.
8-13. Foundation walls with lateral loads. A wallmay be in part
a foundation, in part a retaining wall. Lateral loads may be caused by
wind, cranes, earth, frame action, and many other things; they may be
applied by the superstructure along with its vertical loads, or independ-
ently. The longitudinal bending produced by concentrated loads may be
approximated as explained for foundation walls; the lateral bending, as
described in the preceding article provided the additional vertical and
horizontal forces are included in the calculations.
As a general illustration of combined action, assume the condition
shown in Fig. 8-37 (a). This represents one side of a large steel mill
building to be located on sloping ground. The area A is to be filled in
and used as a plant roadway between this building and another one
farther up the hill. The soil is a clayey sand that has a safe bearing
value of 6 k.s.f., and the area has been used as second-growth woodland.
Each column is to be a 24-in. wide-flange section with the crane girder
seated on the inner flange, and with a 10-in. column spliced on top of the
other portion to support the roof. These columns are to be 20 ft. c.c.

What type of construction is desirable?

The may
illustrate one solution for the problem
following outline
1. A needed to hold back the ground.
continuous wall is

2. The columns are so deep that pilasters are needed under them, since

the top of the wall need not be thick enough throughout to support the
columns directly.
3. The bottom of the corrugated siding is at E. It is desirable, there-
fore, to place the top of the wall so as to form a continuous line at E,
Fig. 8-37(6), with the pilaster projecting inward a distance CD. This
provides more space within the building than would be the case if the
 . :

Art. 8-13] WALLS WITH LATERAL LOADS 277

wall lined up with C, and it avoids the widened top that would be neces-
sary to close across to E.
4. The wall must serve as a combined foundation and retaining wall.
Therefore, a T-type will be used, the width of footing b will be made a
little wider than Sketch (c), the inside of the wall and pilaster will
i-^/i x of
be battered y± to avoid the appearance of leaning inward, and
in. per ft.

the toe will project a little beyond the base of the pilaster. This causes
the column load to be rather close to the toe so that large toe pressure
may cause an increase in the width of the footing.

M ,62
k>lr,i/'
,10"

P
ty)^ DSPzi=l37
k
Corrugated
Column. siding Inner ^\7i^4^pnr=j7/r 2'-3"l-'0"2'-9"
flange^
H-
1 i E ^U^QAiQ "roadway'A
i
'
L
l 6'\
fA<
///All
\>
,

Pilaster-^

Fill 12"

FIgorB
J Siem

i'6 "o'c"
2-6
T / |
}. HppI

'3:1
V.^C 7 "•d;.\ "

Toel. b=7-6"
W- DETAILS AT TOP OF
(W-GENERAl REQUIREMENTS WALL AND PILASTER fcJ-TRIAL DIMENSIONS OF WALL

FIG. 8-37. Combined foundation and retaining wall for a mill building.

5. The dimensions of the wall are shown in Sketch (c).

trial

6. The This permits the

footing will be placed below the floor slab.
installation of the floor after the heavy construction work is completed,
and the floor can then be finished and sloped for drainage as desired.
This floor, being wide and heavy, may be relied upon to assist in the
resistance to sliding of the wall if necessary.
The pavement A will be a flexible bituminous type because of econ-
7.

omy and the fact that much of it is on fill.

The loading conditions for which the wall should be analyzed are the
following
1 Temporary safety as a retaining wall with the fill in place, the floor B
omitted, and the weight of the superstructure not applied.
2. Safety of the completed structure with the dead load of the super-
structure combined with maximum w and w caused by wind. H
In M
Sketch (6), the position of this dead-load reaction Pdl is shown.
3. Safety of the completed structure with the dead load and the maxi-
mum live load combined with the possibly coincident live-load shear H LL
278 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

and overturning moment LL M

The new position of the vertical reaction
.

isindicated in (b). There may be cases also in which the omission of live
load upon the roof causes more critical conditions.
4. Safety of the completed structure with dead load, maximum live

load, and wind all acting simultaneously. A 30 per cent increase in the
allowable unit stresses is then permissible.
Safety of the completed structure with dead load, minimum vertical
5.

crane live load, maximum H

LL caused by the crane at the wall when the
trolley is at the other end of the bridge, and the corresponding LL M .

6. The safety factor may be applied in the live and lateral loads as
illustrated in the preceding article when testing for safety.
By an inspection of the forces, it is generally possible to pick from the
load cases previously listed the one or two that will cause the greatest
effects upon the substructure. If the wall can withstand these conditions,
it can ones more easily, and the latter need not be
resist the less serious
investigated individually except when the results cannot be predicted
with reasonable certainty.
The from the superstructure may be assumed to cause
vertical loads
vertical bending in the stem as a beam, but this may not be serious when
the wall is so deep. The wall may generally be analyzed as though it were
a huge T-beam. Extra longitudinal reinforcement in the top of the stem
is all that is required in most cases. When the wall is shallow, compres-
sion in the top near the pilasters, caused by longitudinal bending, may be
important.
The horizontal loads from the superstructure may be assumed to be
resisted by a considerable length of wall even though they are applied at
the pilasters. It is true that the stem is much stiffer laterally at the
pilaster than at some vertical section beyond it, nevertheless the mono-
lithic construction will tend to act as a unit unless it cracks. The magni-
tude of the torsional stiffness of such a wall is uncertain, but the following
assumptions are suggested regarding how far along a wall a local concen-
trated horizontal shear or overturning moment applied at a pilaster may
be assumed to spread, as far as the stability of the wall and the strength
of its parts are concerned:
1. One-half bay each side of the center of the concentration; 20 ft. in
this case.
2. The height of the stem each side of the center of the concentration;
24 ft. in this case.
3. Eight times the thickness of the top of the wall, DE in Fig. 8-37(6),

each side of the edge of the pilaster, or the center of the concentration if no
pilaster is used. This gives 18 ft. in this case.
4. By using an enlargement of the top of the wall, somewhat as indi-

cated in Fig. 8-40(6), the wall can be stiffened laterally so as to distribute

horizontal loads even farther than specified above.
Art. 8-13] WALLS WITH LATERAL LOADS 279

The minimum distance computed from any one of these three gives the
limiting spread; 18 ft. here. Then the concentrations applied by the
superstructure may be divided by 18, and a typical 1-ft. slice of wall may
be analyzed for these loads combined with those of the earth and other
forces that are assumed to be acting uniformly along the length of the wall.
It is obvious that horizontal wind and live loads that act toward the earth
behind the wall will not be important in ordinary cases.
For example, assume that Fig. 8-38(a) shows one load case for which a
typical section of the wall in Fig. 8-37 (c) is to be tested for stability and
soil pressure. The moment M
may be replaced by shifting the load P to
the left a distance e' = M/P. The earth pressure P e per foot of wall will

ffLtr0.4S\

Reactions
not shown k
-EH--2.89

F G
foJ-LOADS ON WALL

fW-RESULTANT LOADS AND

PRESSURE DIAGRAM

FIG. 8-38. Analysis of wall for one load case.

be computed from Eq. (8-13) and it will be assumed to act horizontally

because of the vibrations caused by traffic on the roadway. Surcharge is
not to be included in this problem. The resultant of and P e is I,H; H
that of the dead load of the wall and earth W
and P is 27. With these,
the eccentricity e and the ordinates of the pressure diagram may be calcu-
lated. The results are shown in Sketch (6). Taking moments about a
point 6 in. to the right of point F, the safety factor against overturning
seems to be

S.F.
14.7 X 3 = 2.7
2.89 X 5.73
These results and the pressure diagram in Fig. 8-38(6) indicate that the
proposed wall is perhaps unnecessarily conservative for the load case
tested. It may not be so if impounded water, wind, and the surcharge
caused by a truck are included in the loading.
280 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

By no means is the concrete of the pilaster in Fig. 8-37 (c) negligible as

far as stresses in the concrete and reinforcement are concerned. Never-
theless, the effect is somewhat localized, and it will probably stiffen the
stem somewhat. However, the pilaster cannot act as the stem of a
T-beam that resists all the overturning because the footing near the
and the soil under it cannot offer the necessary resistance, whereas
pilaster
rock might do so. In planning a foundation, one should remember that
he cannot push on something any harder than the latter can react.
Another possible arrangement, instead of that used in Fig. 8-37 (c), is
the moving of the stem of the wall toward the fill sufficiently to clear the
column. The latter may then be extended down to the footing. This
may be suitable when the footing is on rock or very strong gravel. Other-

rra r Floor or pavement

Floo

Ro'ck
foj-Z.-SHAPED WALL W-STANDARD T-SHAPED (c)-WALL AS LINING (W-CURTAIN WALL,AND TOP
WALL OVER ROCK SLAB AS LATERAL TIE

FIG. 8-39. Possible construction in rock excavation.

wise, themoments and shears at the top of the wall may be less trouble-
some than the heavy concentration that tends to break the toe of the foot-
ing locally.
Assume that a wall must be built along a rock cut, as shown in Fig.
8-39(a). To minimize excavation, a heelless wall may be used that can-
not be waterproofed properly, and this is a situation in which water may
be troublesome. The design of the wall should be made with the inclu-
sion of at least some earth pressure because of probable overbreakage.
This wall must be thick in order to secure the necessary weight unless Pdl
is large. Gripping of the rock at the rear of the footing of the wall to
resist overturning
is unreliable unless holes are barred or chipped in the

rock. A better design is the extension of the excavation and heel about
2 ft. as shown in (6). This provides space for waterproofing the wall, for
the installation of a drainage system, and for the addition of some weight
of earth on the heel.
If a rock excavation has been made and looks like that in Fig. 8-39 (a),
or if line drilling and careful blasting can be trusted to produce such
 .

Art. 8-13] WALLS WITH LATERAL LOADS 281

pour the concrete against the rock as shown in (c)?

results, is it safe to
Should one assume that there will be little or no lateral pressure against
the concrete? Here are some comments about such construction:
1 Ground water will almost inevitably pass downhill through seams in
the rock. When the wall is long, it tends to act as a dam so that large
hydrostatic pressure may exist.

2. Leakage through such a wall is almost certain because the keying of

the concrete to the irregularities of the rock will result automatically in
shrinkage cracks. Admixtures in the concrete may improve its density
but not eliminate these cracks, and waterproofing applied to the inner
surface of the wall is likely to be costly and unreliable.
3. Surface water will almost surely penetrate behind the wall, auto-

matically becoming ground water.

4. Freezing of the ground water, if freezing can occur behind the wall,

will force the wall to yield or even crack badly.

5. Longitudinal drains, or local vertical drains passing under the floor

and discharging at the downhill side, may not be completely effective,

although helpful.
6. When waterproofing is essential, it may be best to put up a "sand

wall," apply a membrane on it, and then make the structure inside the —
left side in Fig. 8-39 (a) —
capable of resisting the water pressure as well as
the horizontal and other loads applied by the superstructure.
7. Drilled holes in the rock with grouted dowels therein to anchor the
concrete to the rock may not be helpful permanently. The steel may
rust off, and any pressures from ice are too great to be resisted.
8. A construction like that in Fig. 8-39 (d) may be used if the excava-
tion is not too irregular and if a reliable, inspectable, and cleanable drain-
age system can be installed.
9. Unless a high-level yard is needed alongside, it may be advisable to
excavate the rock to a line farther back so as to eliminate the need for the
wall and to obtain better light and ventilation in the structure. The
areaway should be wide enough to prevent falling pieces of rock from
entering the structure a continuous concrete wall 2 or 3
; ft. high as a base
for the side wall may be useful also for this purpose.
There are in which a structure may be designed as a box.
many cases
For example, assume the basement shown in Fig. 8-40 (a). The wall at
the uphill side will be like a slab spanning vertically between A and B.
The floor CA, particularly, may be a strong horizontal diaphragm that
will transmit horizontal forces to the end walls. When there is a large
opening in the floor AC alongside the wall A B, the latter may be built with
a top that will serve as a horizontal beam, as pictured by EF in (6). This
beam will then support the top of the wall and transmit its end reactions
into the adjoining floor. This principle of planning a structure as an
282 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

entity that is inherently strong and stable should be utilized when

practical.
To illustrate a planning problem, assume the general construction
shown in the isometric view in Fig. 8-40 A. This is modeled after a real
case that occurred at an industrial plant. Assume that the column loads

Opening
c
Floor, infloorj_E ' F
Column
Basement
/floor
^m EST "BBS
B
faj-GROSS SECTION THROUGH A BASEMENT (b)- SECTION ATA
LARGE HATCHWAY

FIG. 8-40. Illustration of boxlike action of a substructure.

FIG. 8-40A. Problem in planning concrete construction.

shown where the steel H's are pictured are 300 kips each; the floor loads,
200 p.s.f. The walls and floors are of reinforced-concrete construction.
Plan the construction, assuming the soil to be firm sandy gravel that can
support safely a bearing of 4 tons per ft. 2 The original ground under DE
is and slopes upward toward J where it is at El. 2096.
at El. 2088
Using the 24-ft. wall from E to F as a specific case, how might this be
planned? Here are a few alternatives for consideration:
Art. 8-13] WALLS WITH LATERAL LOADS 283

1. Make it a heavy retaining wall with the columns supported upon

pilasters projecting from its back face. Fill in behind it to hold the raised
floor.
2. Make it a curtain wall with pilasters for the columns, supported
laterallyby the upper and lower floors. The earth backfill is to be sloped
from F back toward J so that the raised floor must be reinforced-concrete
beam-and-slab construction supported upon local piers and the edge walls.
Consider the form work for the floor; it is costly.
3. Make it a strong vertical slab supported laterally by the lower and

upper floors, tying the latter together by reinforcement that extends clear

FIG. 8-40B. Retaining wall nearly 50 ft. high along approach ramp to new Rankin Bridge over the
Monongahela River, Pittsburgh. Notice the grooving, the contraction joints, and the sequence of
pours. (Courtesy of Parsons, Brinckerhoff, Hall, and Macdonald, Consulting Engineers, New York,

and Department of Public Works, Allegheny County, Pa.)

back to the column bases beyond L. The upper floor is then placed on
backfill.
4. Make it a slab spanning horizontally between ribs or walls at the

columns and at the middle of the bays. These ribs are to be triangular,
and their bottoms are to slope upward from the footings (or to be made in
steps) toward J. They serve as a series of counterforts to support the
front wall. Fill is to be placed between them, and the floor is to span
from the outer wall DJ across these ribs.
5. Make it, the side wall from GJ to CD, the opposite one beyond EF,

and the upper floor like an inverted box that is inherently stable. The
bottoms of the walls will be supported laterally by the various lower floors.
Support the upper floor on fill.
6. Since the upper floor is only two bays wide, a central counterfort

might be built under the middle column with a keyed contraction joint on
one side of it to cut DE in two. Then one corner portion could be made
284 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

as an L-shaped structure that is inherently stable because the wing JD

serves as a counterfort for the front whereas the front does likewise for
the side. The rear part will be U-shaped. The small wall for the inter-
mediate floor will have a joint between itself and the raised portion.
It is obvious that there are many ways to make this structure. Choose
the system that seems to be the most practical, then develop it. The
plans for the other parts should be selected along with that for the high
portion because each may be affected by the other. Joints should be pro-
vided to prevent cracking, but their locations are influenced by the nature
of the system itself.

A*|

F u n ace

_J ....' y.y :///.'//,

*l B-l
^Mastic at junction

^Movable wall
A -J B*-1

faJ-GENERAL PLAN OF LAYOUT

rFloor Furnace

Floor 77&Za0O890!&

I
^-FOUNDATION OF BRACED TOWER, (W-SECTION THROUGH MOVABLE WALL,
SECTION A-A SECTION B-B

FIG. 8-41. A situation in which the main structure is isolated from the part that is subjected to serious
lateral forces.

Sometimes a designer is faced with a situation where the lateral forces

are so great that inevitable movement should be provided for without
endangering the main structure. Figure 8-41 shows, in simplified form,
part of the layout of a large smelter where this was done. The drawings
illustrate the basic features of the design. The reverberatory furnaces
are over 100 ft. long. When such a furnace is in operation, the heat
causes expansion, and the force will be tremendous if motion is, not per-
mitted. Furthermore, when a furnace is cooled down, it will seldom
return to its original length but will crack. Rebuilding and reheating a
furnace are likely to cause longitudinal creep of the foundation. There-
fore, the main steel superstructure is supported upon towers that have a
Art. 8-14] ANCHORAGES FOR PIPE LINES 285

heavy combined foundation, a sort of island, as indicated in Fig. 8-4 1(a).

In front of each furnace is a light retaining wall, as shown in principle
in (c). This wall may slide or tip as a unit without endangering any-
thing except itself, and it may be rebuilt if necessary.
8-14. Anchorages for pipe Pipe lines that contain rapidly
lines.
flowing liquids, or even static fluids subjected to large pressures, need
special foundations. The vertical dead loads may be moderate; the
thrusts caused by pressure and flow may be relatively important.
As a starting point for a study of these problems assume the portion of a
pipe line shown in Figs. 8-42 (a) and (6). The pipe is 24 in. O.D., 3
^ in.

thick, welded steel, and the maximum flow expected is Q = 80 c.f.s. The
pressure head h at B when the water is shut off at the discharge end is
190 ft.; when the water is flowing, it is 175 ft.
First, consider the vertical load of the pipe and its contents. Of course
the span A B should not exceed that which the pipe can stand safely as a
beam to hold up the weight of the water and the steel. A small pier or
pad would be suitable for supporting the vertical load of a straight pipe
and its contents at such a point as A, since F = (95 + 2.95 X 62.5)50 =
14,000 lb. (approx). Nevertheless, the pipe should not rest upon a flat
concrete surface as shown in Fig. 8-42 (d) because of the local concentra-
tion of pressure. When a cradle is built of concrete, as in (e), there
should be some protective or cushioning material between the pipe and
the masonry, and it is generally desirable to anchor the pipe more or less
securely by some such means as the anchor-bolted strap shown in (e).
Next, consider the effect of changes in temperature. The pipe is made
pg in. thick in order to have strong beam action and a reserve allowance
against corrosion, not just because of pressure. Assume that the range of
temperature for span AB is only ±25°F. from normal. Then the total
change in length between A and B due to expansion and contraction is
approximately AL = 0.0000065 X 50 X 12 X 50 = 0.195 in. If the
pipe is fastened firmly at B, it will try to slide on the top of the founda-
tion at A, or to tip this foundation slightly. One can see that, if a con-
tinuous pipe extends over a series of spans, this movement near the end
may become considerable and troublesome. On the other hand, if the
pipe is supposedly anchored immovably at each support, the theoretical
unit stress in the pipe for a change in temperature of 25°F. either side of
the average is

fs = Eb = 30,000,000 X 0.0000065 X 25 = ±4,900 p.s.i.

The maximum total force that the entire restrained pipe might exert upon
its foundations is, therefore,

F = ±4.9 X x X 23.62 X 0.375 = 136 kips

286 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

An examination of the preceding paragraph shows that provisions

should be made to permit longitudinal motions and thereby eliminate
these large temperature stresses and forces. One way to do this is to
install some kind of slip joint in the pipe line, as indicated by J in Fig.
8-42 (c). One type of joint for such use is shown in (/). However, as
indicated to exaggerated scale by the dashed lines in (c), these joints may
interfere seriously with the beam action of the pipe unless the joints and
the foundations are located properly with this in mind.

Ffow
r~ —2T
•CL-.30-'
g 7
E
(cr)-PLAN OF A PORTION OF PIPELINE

Head h'

''
r
^™$?77?V777V77W7Kti*fj?F'

(^-DEVELOPED SIDE ELEVATION

Follower,

Rod in
pipe
fcj-USE OF EXPANSION JOINTS Holes or sleeve
pockets in (Wooden
Band concrete block
^ J* Pipe pier
Al \ A <L of pipe * Plain end ofpipe
(fj- PARTIAL SECTION OF DRESSER
(d) (e) COUPLING (AFTER "HANDBOOK
OF WELDED STEEL PIPE',' BY fgTJ-PIPE ON WOODEN SADDLE
CALIFORNIA CORRUGATED CULVERT CO.)

FIG. 8-42. Pipe lines and foundations for them.

One possible arrangement for a straight pipe line is the use of an anchor-
age near center of a long length of pipe, with rocker bents on each side to
resist vertical and transverse forces but to permit longitudinal motion to
be taken up at the expansion joints. Another is the use of the anchor pier
with adjacent foundations on which the pipe may slide. Unless the pipe
is on rollers or can slide upon a wooden or other saddle as shown in Sketch

(g), the pipe is likely to be worn harmfully. If the pipe is embedded in

soil, as shown at D in Fig. 8-42(6), this may be a suitable anchorage in

some respects. The frictional resistance of the soil on the pipe will hold
it is overcome.
unless the friction
Next, consider what happens at an elbow or curve, such as at B in the
pipe line of Fig. 8-42 (a) because of pressure. Assume that the pipe is
welded as shown in Fig. 8-43 (a). It is clear that the hydrostatic pres-
Art. 8-14] ANCHORAGES FOR PIPE LINES 287

sures on the inside of the pipe will counteract and be in equilibrium on

the left side of the section GJ and on the right of HK. The pressure on
the junction piece, however, is not balanced. This piece is shown
unrolled in (6). A resultant force Ri is therefore present at the elbow.
Then,
Hi = 2pA(sin y a)
2 (8-23)

where p = hydrostatic pressure in pounds per square foot or 62. 5h, A =

the cross-sectional area of the inside of the pipe in square feet, and a =

J K

faJ-RESULTANT OF PRESSURE AT ELBOW

J K
^-DEVELOPED JUNCTION
PIECE

(c)- DYNAMIC FORCE AT ELBOW

uFlow
Branch
'l/he
Flange^ Bolted
I 1- Plate \J Q Flow

^
(

6 l:
-\ )
y Reaction
WJ-THRUSTAT DEAD END ^-THRUST AT TEE

FIG. 8-43. Forces at elbows and bends in pipe line.

the deflection angle of the elbow. Therefore, for the maximum head of
190 ft. and a = 30°,

Ri = 2 X 62.5 X 190 X 2.95(sin 15°) = 18,150 lb.

For a head of 175 ft., R[ = 16,700 lb. This resultant force acts in the
plane of the center lines of the two pipes, and along the bisector of the
angle /3, Sketch (a). The components of the pressure normal to the plane
of theelbow counteract each other. When the bend has a long radius, one
should not forget that the lateral pressure is spread over the region repre-
sented by the space between GJ and of Fig. 8-43 (a). HK
Therefore, the
anchorage should embrace this area.
288 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

When an encountered by flowing water in a pipe line,

elbow or bend is

the inertia force of the moving

fluid causes a force that tends to push the

bent portion of the line outward, as represented by R> in Fig. 8-43 (c). If

Q = the quantity of water flowing past the elbow in cubic feet per second,
w = unit weight of the fluid in pounds per cubic foot, v = the velocity of
flow in feet per second, and g — the acceleration caused by gravity,

D
Rz = mv sin
5-7—
cos y%a
a
= —
Qw
g
v
sin
cos y^a.
a
Y7~ or 2
Qw
—
-5

g
v
.

sm
-.,
ya2
fo nA .
,

(8-24)

This force is coincident with R\ of Sketch (a), neglecting eddies and fric-
tional losses. Then, for point B of Fig. 8-42 (a) and a = 30°,

_ 80 X 62.5 80 0.5 _
Ri 32^~~ X 2~95 X 0965 ~ 2 18 ° lb
'
"

Therefore, the greatest thrust at B occurs when the water is flowing, since
R[ + Ro exceeds R x .

Of course, a dead end like that in Fig. 8-43 (d) will have a longitudinal
thrust Ri = pA because the flow is necessarily zero. A* tee like that in
Sketch (e) must be anchored to resist both the pressure and the dynamic
force as determined by Eqs. (8-23) and (8-24), using the cross-sectional
area of the branch line. In the case of a diagonal branch and other special
situations, the magnitude and direction of the thrust should be determined
for the particular angles, sizes, pressure, and flows that apply to each
specific problem.
If the pipe line has expansion joints in it to relieve the anchors of the
thrusts produced by changes in temperature, the principal forces to be
resisted by the anchorages are those represented by Eqs. (8-23) and (8-24).
The frictional resistance of the pipe to the flowing liquid, water hammer,
the hydrostatic pressure on the exposed end of the pipe at the expansion
joint, and the frictional resistance of the pipe if it is forced to slide over

adjacent piers — these are forces that may be of importance in extreme

cases. However, a conservative safety factor of 2 in the design of the
anchorages generally covers all these minor forces with a satisfactory

margin.
When a pipe line is buried in a relatively deep trench, the weight of the
soil backfill is generally sufficient to hold a straight pipe line in place with-
out frequent expansion joints. At an elbow or sharp bend, a pad of
concrete should be placed as shown in Fig. 8-44 (a). When a horizontal
bend, as shown in (b), is made by a series of pipes connected by expansion
joints,and each pipe is skewed slightly at the junction within the limits
permitted by the type of fitting used, compacted backfill along the out-
side of the pipes is generally sufficient to resist the thrusts.
Art. 8-14] ANCHORAGES FOR PIPE LINES 289

An elbow in a vertical plane, if the bend is concave up as shown in

Fig. 8-44 (c), is generally safe automatically because the thrust is down-
ward. However, it is important that the soil bear firmly against the
underside of the pipe at the bend. If tamped soil is not considered

trustworthy or if points of rock may cause localized punching, a concrete

pad or cradle under the pipe in the vicinity AB is desirable. If the soil
is muck or other unreliable material, a wide concrete anchor, a group of

Dresser,
coupling \

Concrete
(^-HORIZONTAL ELBOW

#>>- STRAIGHT PIPES

AS CHORDS

y^zz OF LONG-RADIUS CURVE

(d-VERTICAL ELBOW IN TRENCH

Anchorage

(d)- VERTICAL ELBOW CONVEX ON TOP

FIG. 8-44. Pipes set in trenches.

piles with a concrete pad, a stone-filled timber crib, or some other device
should be used to spread the thrust and support the structure safely.
An elbow in a vertical plane with the bend convex upward may be diffi-
cult to restrain. Figure 8-44 (d) illustrates such a case. The earth over
the region CD may not be sufficient to resist the upward thrust unless the
trench is very deep or a large mound of earth is built over the bend. Even
then, the weight of the soil should not be trusted because, when some
maintenance man thoughtlessly removes the soil, the pipe line may be
damaged. It is advisable to anchor the line safely as though it were
exposed.
The schematic drawings in Fig. 8-45 are intended to provide a few sug-
gestions that may help the reader in designing pipe supports and anchor-
ages. When a pipe line is built over rough terrain, one will probably
290 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

encounter a great variety of conditions. The bends are likely to be in

inclined planes, and one must design the supports and anchorages to fit
the local requirements. Pipe trestles and bridges may be needed; tunnels
in earth or rock may be worth their cost; even a catenary action of the

Plan Section B-B

Section A-A
fb)- SOLID BLOCK AT HORIZONTAL BEND
faJ-SOLID BLOCK AT HORIZONTAL ELBOW

5 {
Rock
Plan Side Elevation Plan End Elevation
(c)- BLOCK AND FOOTING AT HORIZONTAL BEND (W-BLOCK AND FOOTING AT L0NG.H0R1ZCNTAL BEND
Bead Welded ring
-
-| weld

1
dL kJ
Plan
Plan Elevation 1V>f faJ-SOLID BLOCK ATA
(e)- ANCHORAGE AT SMALL Section C-C VERTICAL ELBOW
HORIZONTAL ELBOW
(fJ-H-TYPE ANCHORAGE AT VERTICAL ELBOW

c: 1

Weepers Footing
Plan Side Elevation Plan
Plan Section D-D
(i)-A SIMPLE ANCHORAGE AT HORIZONTAL T &V-ANCH0RAGE AT
HORIZONTAL Y
(hh BOX-TYPE ANCHORAGE AT VERTICAL BEND

FIG. 8-45. Some types of anchorage for welded steel pipe lines on earth or rock. (Courfesy of A. L

Offo, Anaconda Co.)

pipe with special anchors might be used across a deep ravine. Con-
itself

trol valves,branch lines, siphons, points for access and egress of go-devils
or other cleaning devices, removal of sections for repairs or maintenance,
by-passes, manholes, loads from overpassing traffic, floods, erosion, frost,
 PROBLEMS 291

and corrosion from without and within — these are among the factors that
the designer of a pipe line should consider.

PROBLEMS

8-1.Assume the column shown in Fig. 8-46. Are the anchor bolts satisfactory
if P = 290 kips, e = 6 in., and .1/ = 220 ft.-kips? Assume that the allowable
tension in the anchor bolts is 12 k.s.i. at the bottom and 18 k.s.i. at the threads.
Assume the point of rotation to be under the back of the 12-in. channel.

i-t
'
' ,tteavy,cont. 5-0"
r-H washer
1
p/.

\,'2-l'/2"*anchor
j^
bo/ts each side
1-0*
I
I

Pedestal
2 mortar 1
J2% (Floor

ML .1

Reactions not shown faJ-SIDE ELEVATION (W-END ELEVATION

FIG. 8-46. Base of an interior crane column in a FIG. 8-47. Footing for a viaduct column.

steel mill building.

8-2. Check the footing shown in Fig. 8-47 for safety under the following
conditions:
P — 90 kips min; 300 kips max
S = kips min; 42 kips max from wind
Soil = sandy clay having allowable p = 5 k.s.f.
Safety factor = 2 (applied to LL and wind only)
Max coefficient of friction / = 0.75
Suggestions: Test for overturning, sliding, and edge pressure for the following
cases
1. P min and 2 X Sw, with limiting p = 2 X 5 k.s.f.

2. P max and «S max , with limiting p = 1.3 X 5 k.s.f.

3. Finally, test for P min , 2(P max - P min ), and 2 X 5 raax ,

with limiting p = 2 X
5 k.s.f.

Case 1 is the only one in which sliding may be critical.

An easy way problem with sufficient accuracy is to compute the

to do this
weight W
of the footing and the soil on it, then divide W, P min Pmai Smax 2(P max , , ,

— -Pmin), and 2 X S max by 7, the width of the footing normal to S. Use these
quotients as loads on a strip of footing 11 ft. long and 1 ft. wide. On a drawing of
this strip of footing draw the loads to scale for various combinations, and scale off
eccentricities, etc., as shown in Fig. 8-48. Compute the ordinates of the respec-
tive pressure diagrams.
Ans. Edge pressure for case 3 is 10.9 k.s.f., a bit too high. Otherwise the
footing is safe.
292 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

Force diagrams
are superimposed
and placed as
shown in order
to aid comparison

6-0"
M
6.9
k/°'
Case1:DL+2S

k
2.8 f°'-^m.
6.4 k/°'

Case2:DL+LL+S

3.8*1°'-
10.9
k/o '

Case 3-DL+2LL+2S

FIG. 8-48. Analysis of footing, using graphics to determine eccentricity.

8-3. Design an isolated spread footing for the steel viaduct column shown in
Fig. 8-49, assuming the following data:

P DL = 140 kips; Pll = 180 kips

Tensions in bracing Fi max = 50 kips for longitudinal braking forces and wind
combined as a live load.
Soil = medium sandy clay having allowable p = 6 k.s.f.

Safety factor = 1.5

Location — Cleveland, Ohio
Suggestions: The and size of the top of the pedestal are shown. The
elevation
bottom of the footing must be below frost depth. The force F\ is the load in mem-
ber B. Its horizontal component is the force to be used as S, and S is assumed to
be transmitted through the gusset C to a working point at the intersection of the
center lines of A Assume that this working point is 8} 2 n above the
and B. i -

bottom of the billet F. Then S acts 10 in. above the pedestal. The vertical
component of Fi counteracts some of P. Make a guess at the weight of the W
footing, add it to Pdl + Pll, divide the sum by p to compute an area, and use
a trial area some percentage larger than this (perhaps 20 per cent). Assume
dimensions for the footing and recompute W. Test for Pdl + Pll and Fi max,
with a limiting p = 6 k.s.f,, since Fi is not entirely a wind load but a force that
exists with wind and live load P m ^, or with wind combined with live loads in
 PROBLEMS 293

adjoining spans and affecting this braced bay. Finally, test for P DL , 1.5Pll,
and 1.5Fi max; then for P min and 1.5F X max.

Anchor
bolts ,

Floor,
"i 1-4" i
1-4"%

faJ-FRONT ELEVATION (b)- SIDE VIEW 3s6 "x 2-6' Pedesfal

FIG. 8-49. Base of a braced column. FIG. 8-50. Base of an interior crane column
in an industrial plant.

8-4. Assume the construction shown in Fig. 8-50. Neglecting the lateral
supporting value of the floor and earth, assuming the pedestal to be 3 ft. high, and
assuming a footing 9 ft. square and 2 ft. thick centrally located, compute the
maximum pressure on the soil for the following case:

P = 375 kips e = 6 in. M = 50 ft.-kips

Ans. 7.2 k.s.f.

8-5. Assume the construction shown in Fig. 8-46. Neglecting the lateral
supporting value of the floor and earth, design an isolated spread footing for this
column for the following conditions:

P = 320 kips e = 7 in. M = 90 ft.-kips S = 5 kips

Allowable p = 7 k.s.f. Safety factor = 1.5

8-6. Design an isolated spread footing for the combined interior steel columns
shown in Fig. 8-51. Column A, a building column in an industrial plant, supports

M
WF WF
14
n 41
12

.-.*.. |, >-,:, .
;,;..
f t
11 X If 10 I Floor line
Reactions nol shown

FIG. 8-51. A double column for two adjacent crane runways.

294 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

a bracketed crane girder on its left flange ; B supports a crane girder that is seated
upon its top over and parallel to the web. The diaphragms C will not transmit
vertical loads from one column to the other. Assume the following design data:

Pi = 60 kips DL; 110 kips LL

P =
2 DL; 250 kips LL
10 kips
M = ±100 ft.-kips LL; ft.-kips W (W wind)
S = ±10 kips LL;0 kips W
The bulk by roof bracing and vertical bracing located
of all lateral forces is resisted
elsewhere. Assume that the diagonals of the longitudinal bracing towers for the
columns are not connected to these particular columns. The soil has a safe
bearing value of 8 k.s.f. The safety factor is 1.5.
Suggestions: In this case, the live load can be on either column alone, or on
both. Design the footing for the maximum loading conditions, then analyze
it for the maximum live load upon B only. Assume that 30 per cent of M and S is

caused by live loads on A

70 per cent on B.
;

8-7. Figure 8-52 pictures a double steel column to be used as the support for two
adjacent crane runways in the interior of a large industrial plant. The roof is

9" SE
JTfyFloor
3-3 c.c.ofcols. line
and anchor bolts
Reactions
not shown

FIG. 8-52. A heavy double column for two adjacent crane runways.

supported by a stub column connected to a pair of heavy channels across the tops
of the crane columns. Either crane may be loaded at a time, and both may be
loaded simultaneously. Assume the following design data:

P = l 35 kips DL; 170 kips LL

P =
2 35 kips DL; 170 kips LL
M= ±900 ft.-kips LL; ±150 ft.-kips W (W = wind)
S = ±30 kips LL; ±5 kips W
Safety factor =1.5
 PROBLEMS 295

The soil is a sandy clay that will support safely a unit pressure of 6 k.s.f. The
columns are laced together so strongly that the participation stresses in the lacing
are capable of practically equalizing the live load in the columns for varying load
conditions on the cranes. Furthermore, the cantilever action of the braced col-
umns is relied upon to withstand lateral live load and wind forces. The columns
are braced strongly in the direction of the crane girders by towers in special bays,
not at the point shown. Design a combined isolated footing to support these
columns.
Suggestions: Determine the maximum load combination, and design the footing
to support the vertical and overturning forces without reliance upon lateral sup-
port from the soil or the floor slab. Notice that the wind forces may occur with
dead load only and with live loads as well.
8-8. Figure 8-53 shows a tall steel bent that is to be used to support two long
trussed spans of a heavy twin-conveyor gallery at an industrial plant. Two
railroad tracks pass diagonally under its center. Assuming the steelwork to be
designed properly, check the pressure and uplift on the proposed footings if the
loads and conditions are the following:

Pi = truss reactions = 100 kips DL and 100 kips LL

P =2 weight of bent =16 kips DL
S = wind load = 30 kips
W = weight of footing and earth = 25 kips assumed at A and B
p = 8 k.s.f. allowable
Safety factor = 2

Suggestions: The legs or columns of the bent are inclined, and the direct loads in
them will be assumed as parallel to their longitudinal axes. Assume that the
^24'-0"

12"WF

Hole clear through

member for drainage

CahELEVATION VIEW fW-DETAILS OF COLUMN BASE fol-SIDE VIEW OF COLUMN BASE

FIG. 8-53. A tall bent supporting long twin conveyors over railroad tracks.

shear S is resisted equally by A and B as a horizontal force applied at their tops,

and that each footing is set normal to the axis of its column so that the horizontal
component of the column reaction will not tend to tip the foundation. The sloped
bottom of a column avoids the use of bent details at the base of the steel. The
anchor bolts are to be parallel to the column shaft whether the top of the pedestal
is sloped or the inclination is secured by using a wedge-shaped pad of mortar or

grout under the steel.

296 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

8-9. A
conveyor trestle is to be built across part of an old clried-up tailings
pond. A
proposed design for the foundation of a typical single bent is shown in
Fig. 8-54. The original ground is fine sand, which can withstand a pressure of at
least 6 k.s.f. The tailings material can be excavated easily and will stand on
practically vertical slopes temporarily, whereas its supporting value is very small
and untrustworthy. No piles and pile-driving equipment are available. Is the
design for the substructure satisfactory for a safety factor of 1.5? The lateral
supporting power of the tailings around the concrete may be utilized.

Steel work —
-W-
I3'-Q"

—
S=±30
* *
k
El. 2024'
\b Anchor
bolts ^^jT
M
*j~f

Struts'

40 kDL
2 10"
pA/7 k LL
\±90 k W 21o"

i Original ground line

Reactions \—
IcviSSrIISl not shown "<\ I

E F I
,
S10"sc/.\
r«;-FR0NT ELEVATION r/b;-SIDE ELEVATION

FIG. 8-54. Tall pier embedded in deep fill.

Suggestions: Compute the maximum pressure under a footing, as though located

directly at A or B.The lateral supporting power of the backfilled tailings will
hold the substructure as though it were an integral part of the earth, and the pier

cannot tip about its bottom.For resistance to uplift, one-half of the pier and the
weight of the soils above a footing like EF may be relied upon. Since the sand is
confined and its bearing value is probably increased by the weight of the over-
burden, the footing appears to be adequate for pressure requirements. It is
needed for securing the necessary resistance to uplift, and the footing should be
increased because the safety factor of 1.5 is not obtained. In such soft material,
the top strut is probably worth its cost.

8-10. A steel tower or headframe with two bents and transverse trusses sup-
ports two large sheaves around which run the cables for a two-car inclined skip-
way from the underground workings of a large mine. Figure 8-55(a) shows the
general framing, dimensions, and forces to be used for design purposes. The
rock is sloping as shown, but the area around the structure will be filled in to
form a level working area. An engineer has designed the foundations shown in
(f), (d), and (e) for the structure. The forces shown are for one bent of the tower.
If the safety factor is at least 1.5, check these foundations to see if they are
satisfactory.
 PROBLEMS 297

Suggestions: The bearings at A and B are planned to resist vertical forces and
some transverse wind shears, whereas that at C is to transmit both vertical and
horizontal forces into the rock. The framework is assumed to be adequate in
strength. The forces applied to the foundations are indicated in Sketch (6). The

Concrete block
4'-0"w/de

Rock Anchor bolts

,-, .-,, .-, - Reactions ^-FOUNDATION UNDER SLOPING
notshown FRONT COLUMN
32W
(a)-S\DE ELEVATION OF TOWER FRAMING

uA Ur D,
I
A B -*C
f6J-F0RCES ACTING ON FOUNDATIONS

2-'5" *

Side Elevation Transverse Elevation

^-FOUNDATION UNDER INTERMEDIATE COLUMN
Side Elevation Transverse Elevation
(tfJ-FOUNDATION UNDER REAR COLUMN

FIG. 8-55. Foundations for a headframe at an inclined mine shaft.

tower will tend to overturn about C, a rigid support. Multiply the resultant R of
the cable pulls and the wind load by 1 .5. Take moments about C, and include the
righting moment of the dead load W. Compute the moment of inertia of the
bearings about C as follows: / = 1 X 32 2 + 1 X 50 2 The uplift at A is . MX
298 FOUNDATIONS SUBJECTED TO OVERTURNING [Chap. 8

50/7 at B,; M X 32/7. A and

This assumes that the bearings and anchor bolts at
B and that the tensile unit stresses in them are propor-
are substantially equal,
tional to their distances from C. Notice that this assumption is different from
what it would be for a section cut through an elastic framework or medium.
Ans. UA = 50 kips; UB = 32 kips; Dc = 136 kips; H= 122 kips. The
foundations are satisfactory.
8-11. A notched footing similar to that in Fig. 8-18 has the following dimen-
sions and loads: footing, 9 ft. square; notch in one corner, 1 ft. wide and 2 ft.
6 in. long; weight of footing and column combined, 440 kips, assumed applied at
the center of the pedestal. Compute the ordinates of the pressure diagram.
8-12. A stepped footing is 10 ft. square and 2 ft. thick. The pedestal is 3 ft.

square, 4 and centrally located, with its top 1 ft. above the ground. It
ft. high,
supports a concrete column having a maximum load of 500 kips, not including the
footing. The soil is good for a bearing pressure of 6 k.s.f. The owner wishes to
cut a strip 12 in. wide and 18 in. deep off one entire side of the spread footing so
that he may install a conduit trench. Should you approve such action if the
safety factor for the soil is 1.5?

8'-0" 7-6" \%;P" Columns 22 16"c.c.

\
Machinery jL i Px = 350 k from cranes
1
I
foundation at\ _ \\\\\ » flr " ltO\from structure
°
3
\*f
ofeach bay- P3 40 k"fota/on 12 'base
=
P k
4 = 20 total on 12 'base
Soi/= F/'rm sandy gra vel
Floor loads fordesign=
t50 #/o 'beyond machines

, „ Bases of
d'-O'
9\ ^machines
Lower floor

2-0
Electric
conduit
funnel
^5
6-0 min.

FIG. 8-56. Problem in planning a retaining wall to support an offset floor in an industrial plant.

Suggestions: The 6-in. concrete layer theoretically remaining under the cut will
be so badly damaged that it should be considered worthless after the other con-
crete is knocked off. Find the maximum pressure as though the footing were
10 X 9 ft., with the load 5 ft. from one side and 4 ft. from the other. Consider
the revised weight of the footing and its eccentricity.
Ans. pi = 8.1 k.s.f. The footing may not fail but the proposed change
should not be made. It will reduce the safety factor to (6 X 1.5)/8.1 = 1.11.
 PROBLEMS 299

8-13. Design a gravity anchorage at a 30° vertical bend in a pipe line similar to
that shown in Fig. 8-45 (/). The static head is 180 ft. when the discharge is 200
c.f.s. The pipe O.D. and %-in. -thick welded steel, and the elbow is made
is 30 in.

by welding sections so that the inner radius approximates 5 ft. The safety factor

Gravel ,

3 $ steel pipe
5\ 5 10 20
j i_

Head at A= 310' of
wetter

Pipeline is in a vertical plane

>2°-30'

FIG. 8-57. Problem in planning anchorage: for a pipe line.

is to be 2. The soil is a layer of gravel 3 ft. thick overlying rock. The location
is at the edge of a ravine in New running down the 30°
Mexico. The water is

slope of the ravine. The expansion joints on each side of the anchorage are 100
ft. away, and the pipe is supported upon several intermediate piers having a

small coefficient of friction. The anchorage is to be designed for pressure and

dynamic forces only.
8-14. Plan the construction to support the columns and machinery bases shown
in Fig. 8-56 and to form the 15-ft. wall. The tunnel is to be of reinforced concrete.
After developing a plan, design a typical portion of the structure. The electric
tunnel runs the full length in front of and below the wall.
8-15. Assume that the pipe line shown in Fig. 8-57 is in a vertical plane.
Design the anchorages at A and B for the head shown, using the static pressure
only.
Compute the edge pressure at AB of the footing of Fig. 8-17A if P is 250
8-16.
kips and S is 6 kips. Include the weight of the earth and the substructure.
8-17. Assume a 9-ft.-square footing like Fig. 8-1 1(c) with a 2-ft.-square pedestal
located 6 in. off center so that RU = 3 ft. and TQ = 4 ft. Assume TV = 2 ft.,
X T = 3 ft., and NQ = 3 ft. 6 in. If P = 400 kips and S = 20 kips, compute the
pressure diagram and design the reinforcement.
8-18. Assume the same footing as in Prob. 8-17 except that another horizontal
force S' = 30 kips is applied at the center of NO acting 90° to S (into the paper).
Compute the pressure diagram, and design the reinforcement.
8-19. A footing is like that of Fig. 8-18 except that the pedestal is 3 ft. 8 in.
deep above ground), EK = 9 ft., EF = 7 ft., GH = HJ = 2 ft. The load
(8 in.
P = 300 kips, and a horizontal force of 30 kips acts at the top and center of the
pedestal toward the left parallel to AD. Determine the pressure diagram and
the reinforcement.
 9
PILES

9-1. Function of This chapter deals with piles primarily as a

piles.
material or element in a foundation. Basically, they are part of the sub-
structure. They may be used to transmit forces from the superstructure
and the remainder of the substructure through weak unsuitable soils,
water, or air to a lower stratum that has sufficient bearing value to sup-
port the completed structure and all the loads that are expected to be
applied to it or they may be used to spread these loads through and over
;

sufficient relatively weak soil to enable it to support the structure safely.

Obviously, piles are used when the subsurface conditions are not suit-
able for the use of spread footings and mats, or when it is not economical
and preferable to build these last in the positions where they should be
located if the piles were not used. Piles are therefore generally associated
with foundation problems and with dangerous subsurface con-
difficult
ditions. Nevertheless, this does not mean that pile foundations are
hazardous. This is a warning to inexperienced and unwary persons,
particularly the owner and the person responsible for the design. Plan-
—
ning a pile foundation and frequently the construction of it requires —
that one obtain all the data about conditions at the site that can be
secured reasonably, that he investigate and compare all feasible plans for
the foundation, that he eliminate all uncertainties that can be avoided,
and that he respect and use the sound engineering judgment that is
developed through study and experience.
"Piles are good as foundations. Therefore, these piles are good."
This is logical. However, piles are no better than the materials to which
they transmit their loads. One cannot accomplish much by pushing on
a rope, neither should he expect a pile to offer adequate resistance when
it has nothing to produce the necessary reaction. In pile-foundation
problems, therefore, the soils under and around the piles are the things
that need to be studied carefully. Obviously, the piles themselves must
be long enough, strong enough, durable, and suitable for the particular
300
Art. 9-2] GENERAL ACTION OF A PILE 301

conditions in which and for the structure for which they are to be used.
Furthermore, the intended construction should be practicable.
What is meant by failure of a pile? Not necessarily does a pile have to
vanish into subterranean depths, nor does it have to break and fold up
like a jackknife to be a failure. When a pile cannot support the intended
or needed loads without a subsidence that renders the structure unsafe,
unusable, or so damaged as to impair its value greatly, then it may be said
to have failed. Furthermore, a settlement that constitutes practical
failure in one case may not be vitally harmful in another. Some settle-
ment of a pile foundation not carried to rock is almost inevitable; if exces-
sive expense is entailed, the settlement may be kept very small. How-
ever, unnecessary cost and uneven subsidence are to be avoided. What
the particular circumstances are makes a tremendous difference. An
accurate predetermination of the settlement is difficult to make. Unfor-
tunately, one cannot try out a pile foundation, then change it at will if it
proves unsatisfactory. If it is safe and serviceable, the owner will be
pleased; if not, he will forget that he may have applied pressure to have
the costs reduced.
In general, service proves whether or not a pile foundation is safe.
Whether a given foundation could have been made safely at less cost is a
subject for argument without much chance for proof. Many and
frequently difficult decisions have to be made in the planning of pile
foundations.
9-2. General action of a pile under load. A pile may be driven into
a deep stratum of granular or cohesive soil, as pictured in Fig. 9-1 (a).
W hen the load P
T
is applied to the top or butt end, the pile tends'to pene-
trate farther into the soil. Compression and displacement of the material
under the tip B offer some resistance, but most of the reaction is caused by
tangential skin friction along the surface of the pile, as represented by AF.
Hence it may be called a friction pile.
The magnitude of the coefficient of friction for the contact surface of the
soil and pile depends mostly upon the properties of the soil. The char-
acter and smoothness of the surface of the pile may have some effect, but
they seem to be relatively unimportant except where the bark is left on
wooden piles, in which case the bark may loosen and isolate the pile from
the soil. Naturally, the frictional resistance per unit of area of the pile
will depend upon the intensity of the normal pressure, which is likely to
vary in some relation to the depth to the point considered. It is upon
this basis that Table 9-1 was prepared. It is intended as a qualitative
guide, not as correct information to be used in any and all cases. Actual
values depend upon the properties of a particular soil, and its resistance
to shear and compression should be studied by means of laboratory tests
302 PILES [Chap. 9

on undisturbed samples. The combined effects of friction and cohesion

are classed herein as frictional resistance.
The total load P on the pile in Fig. 9-1 (a) must be resisted by XAF if

this can be done. The total resistance will then be

R = -P = 2(AA)(p/+ c) (9-r

where p and / are the unit normal pressure and the coefficient of friction,
respectively, acting upon a specific unit of area and c is the unit cohesion.
The area of the surface of the pile per foot of length may or may not

-= zMud

h
clay
- \
apSO
f1 W-END-BEARIN6 fcJ-FRICTION NEAR
PILE BOTTOM, ONLY
faJ-FRICTION PILE

Baiter
pile

1& Ji\
H-pile Corrugated pile i

frfJ-USEFUL SURFACE (eh FRICTION PILE (f)-

AREA5 UNDER ACTION
OF LATERAL FORCE

FIG. 9-1. Approximate action of piles when resisting loads.

vary, depending upon its dimensions; when tapered as in Sketch (a), the
taper itself may have
some, but a rather minor, beneficial effect upon the
friction, but the smaller tip and gradually changing cross section may have
great influence upon the ease of driving the pile. This taper, however,
may cause the resisting normal pressures on the surface of the pile to have
an appreciable vertical component that is helpful, although the taper is
usually too small for this action to be dominant.
Not only is the frictional resistance at the surface of the pile important
but so is the shearing resistance of the surrounding soil against sliding
upon itself. Imagine that the pile in Fig. 9-1 (a) settles slightly into the
ground. Even though the soil may not slide along the actual surface of
the pile (the corrugated shell of a Raymond concrete pile, for instance), it
 .

Art. 9-2] GENERAL ACTION OF A PILE 303

TABLE 9-1. Approximate Allowable Value of Skin Friction on Piles*

Skin friction, p.s.f.

Material
Approximate Approximate Approximate
depth 20 ft. depth 60 ft depth 100 ft.

Soft silt and dense muck 50-100 50-120 60-150

Silt (wet but confined) 100-200 1 25-250 150-300
Soft clay 200-300 250-350 300-400
Stiff clay 300-500 350-550 400-600
Clay and sand mixed 300-500 400-600 500-700
Fine sand (wet but confined). . 300-400 350-500 400-600
Medium sand and small gravel 500-700 600-800 600-800

* Some allowance is made for the effect of using piles in small groups. Values are average at
depth stated.

may slide upon itself along a curved area just outside the pile's surface.

Successive concentric areas located outward from the pile must also
transmit the forces from one assumed thin film to the next, but the areas
of these surfaces of possible failure increase so that the critical surface is
likely to be that one located just outside the pile. As a result of the
transfer of loads from one curved layer to the next curved layer by shear
and cohesion, and from horizontal stratum to horizontal stratum by com-
pression, one may assume that, for convenient visualization on his own
part, the pile in Sketch (a) distributes its load through some volume of
material, such as that shown by section AEDCA. This will be called the
cone of distribution of the load. In other words, the concentrated load P
is spread by the use of the pile to an area of soil below the circle whose

diameter is EC. The angle BAC is not known exactly, and AE and AC
may not be straight lines, but this concept of the spreading of the load is
useful in planning and analyzing pile foundations.
Probably the unit pressure is greatest near B of Sketch (a), hence
assume that the circular arc EDC represents the pressure diagram for the
soil at the general level of B. bottom
Call this imaginary hemispherical
the bulb of pressure produced by the pile and its load. Such a picture as
that of Fig. 9-1 (a) is not intended to imply that the pile has no effect upon
the soil below EDC. However, as the pressures are distributed outward
and downward, the intensity of pressure upon the soil decreases to a value
too small to cause appreciable settlement. It may be that area XYZ is a
truer representation of the character of the pressure diagram on the soil

at the level of the point of the pile. However, the illustrations are to
give the reader qualitative rather than quantitative data.
304 PILES [Chap. 9

Sometimes a pile is primarily a column that supports a load on its top

and in turn bears upon a firm stratum at and near its tip, as illustrated in
Fig. 9-1 (6). This is called an end-bearing pile. To understand its action,
assume that it is driven easily through the mud and silt to the gravel,
where it is stopped abruptly by strong resistance, then a heavy load P is
applied. This load will try to push the pile downward. The soft mud
will probably compress so easily that it offers little resistance to downward
motion of the pile. The silt will develop skin friction, but this soil, too,
is so relatively compressible that it will not provide a large resistance
unless the settlement of the pile isThe pile cannot settle
appreciable.
unless its end shoved farther into the gravel or unless the gravel com-
is

presses. Neither of these two can be done without offering large resist-
ance. It is obvious, therefore, that the greatest part of the reaction to
the load P will be under the tip of the pile, whereas some small amount
may be applied as frictional resistance on the surface of the portion of
the pile in the silt. Basically, the pile is to be classified as an end-bearing
one.
Obviously, the pile in (b) should be strong as a column, and its tip
should not crush. Such a pile may well have a uniform cross section
throughout its length to give it strength as a column and to provide a
large bearing area at its bottom.
Piles need lateral as well as vertical support. If the pile of Fig. 9-1 (b) is

long, it may This buckling cannot happen unless the pile shoves
buckle.
aside the surrounding soil. Even a moderately dense mud or a weak silt
will usually offer sufficient resistance to lateral displacement to prevent
serious sideward buckling and bending. This statement is based upon
experience and upon the theory that the force required to prevent initial
buckling is so small that even weak soils are able to offer sufficient resist-
ance to lateral displacement to prevent buckling of the pile. If the piles
pass through water and their tips are embedded little or none in a dense
stratum, as illustrated by the gravel alone in Sketch (6), the piles will tip
over easily, and they must act like long columns.
In practice, an engineer encounters a great variety of conditions. One
is pictured in Fig. 9-1 (c). Here the pile is primarily a friction pile, but
the resistance is confined to the lower portion only so that the part above
F is still a column. If the stratum above F were soft clay underlain by
silt and then by medium sand, the would cause relatively large resist-
last
ance so that the load will be withstood by a combination of friction and
end bearing. It really makes little difference to the superstructure how
the pile obtains its resistance; however, it makes considerable difference
to the pile as a member of the substructure.
The area of the surface of a pile that may be relied upon to resist skin
friction is usually its actual embedded surface. However, this should not
Art. 9-2] GENERAL ACTION OF A PILE 305

be greater than the minimum area of slippage, whether along the pile itself
or along a weaker shear plane in the soil. Figure 9-1 (d) illustrates two
cases. The dependable area is the minimum area that will circumscribe
the pile.

Assume that the pile in Fig. 9-1 (c) is driven, then earth fill is added
around it to consolidate the mud. The weight of this fill will undoubtedly
compress the mud and the silt, and it may compress the clay somewhat.
Furthermore, the fill and mud particularly may continue to consolidate
for several years. The
effect upon the pile is settlement along with the
consolidation of the and clay because this friction pile must move with
silt

the material in which it is driven. The greater settlement of the mud and
fill will probably cause a downward frictional force on the portion of the

pile above F, this being equivalent to adding to the load P. Therefore,

the fill may cause greater penetration or settlement of the pile.

If fill is also added around the pile in Fig. 9-1(6), the silt may be com-
pressed, but the gravel is not likely to deform appreciably. The result
will be that considerable downward frictional forces may act on the por-
tion of the pile that is in the mud, fill, and silt and cause the pile to deliver
a larger concentrated pressure at its tip unless it can penetrate the gravel
still farther. This is called negative friction.
When large lateral loads are to be resisted, a friction pile like that in
Fig. 9-1 (a) will bear against the adjacent soil and act somewhat like a
cantilevered beam, as pictured in Sketch (e). The probable character of
the distribution of the pressures against the soil is shown also. When the
lateral forces are large and when the unsupported top portion of the pile
is long, batter piles are desirable, asshown in Sketch (/). As the hori-
zontal force H tries to tip the structure toward the left in the picture, all
the piles bend somewhat, but the batter pile at the left tends to press
downward, whereas the one at the right tends to pull out. The direct
resistances along the axes of these two piles have components that oppose
H, making them more effective for this purpose than vertical piles would
be. Of course, if the tension to be developed in any pile exceeds the dead
load on it, means should be provided to anchor the pile effectively into the
cap. The holding-down power of UT may be questionable unless the
movement of the cap is large. It is also obvious that vertical loads on
1

RS alone would tend to force the superstructure to the right, whereas the
actions of RS and UT in a symmetrical arrangement counteract each
other for vertical loading.
Assume that a pile has been driven down to a stratum of sand and gravel,
as pictured in Fig. 9-2 (a). When a load P of short duration is applied,
the end bearing of the pile transmits it to the sand the latter
; may spread
1
Screw piles can be used to resist uplift and to support temporary loads, but they
are not desirable for heavy, long-term forces.
306 PILES [Chap. 9

it over sufficient area of the underlying silt and clay to prevent failure.
However, if the load is continued, the compressible materials in the vicin-
ity ofS may consolidate as shown by the dashed line. The sand stratum
may not be strong enough to act like a structural slab to bridge over such
an area but must conform to the shape of the underlying layer, and settle-
ment of the sand and pile occur at R to equal and accompany that at S.
Consider further the question of the action at S of Fig. 9-2 (a). The
settlement will probably be caused by the squeezing out of some water
from the compressible soil. The overlying sand is porous so that this
water can escape. If this soil were a stiff dense clay, it would be relatively

P
Pile No. I 1 Pile Ala 2^ £

MWd- -Mud
R-->
Sandand^'^
small gravely. wm
m
^Sand andsmall
\'&':4ra vel
li.:Vi.y-.-i

Si/iand^v^^ •fii-fl
yVvl
..-
j.y.i-u.-J Sift
{
and
clay _»ii±JVi fi clay
Ljypjj
\'-'.'-
-X-;-
'.V
':
Coarse '.•''•[.Coarse
sand :' ."

T f sand

r«;-SH0RT PILE (b)-UM, PILE

FIG. 9-2. An illustration of action of an intervening hard stratum.

impervious, then the water could not escape so easily, and serious settle-
ment might not One should remember that water trapped in a
occur.
filled tremendous compression. However, he should also
bottle will resist
be sure that the bottle cannot leak and that no one can smash it.
Dependence upon this principle of trapping the plastic layer may or
may not be wise. Can one be sure that the piles will not. break up the
overlying stratum or serve as potential leakage channels? Is there an
outlet for the water from S somewhere else through a weak portion of the
upper layer? Will neighbors build a structure or drive a well that changes
the entire situation?
A second pile is shown in Fig. 9-2(6). Of course it is longer and it costs
more than the other, but it is driven clear through the upper layer of sand
and the compressible soil to a stratum of deep coarse sand. The question
of harmful settlement is thus eliminated. Is this pile almost entirely an
end-bearing one with a reaction at T, or is it supported in part by skin
friction along the embedded portion below the mud? Initially, the latter
is probably the case, but one can see how a slow consolidation of the silt
Art. 9-3] ACTION OF A GROUP OF PILES 307

and clay might occur, with the result that more and more of the reaction
will be at T unless settlement of the pile occurs, too, so that it continues
to apply some load to the upper materials. The spreading of the load is

pictured qualitatively by the dotted lines in the sketch.

9-3. Action of a group of piles under load. How will several piles
act if they are driven near together and all are subjected to simultaneous
vertical loading? If the piles are end-bearing ones, as pictured in Fig.
9-1(6), each will act as a column and deliver its load directly to the

PP P P PPPP
iff i i
A. E J O
\ \
/

/ \
)
/ \

/ \

/ \
/ l\ \
I l
-B, \N
i i

\ \
< y B \
I

C ^Bu/bs ofpressure
assumed forpurpose
M
of visualization

/
i
{

\ /
s ,y o o o
Section A-A o o o
\
SS
Section B _ B
faJ-SMALL ROW OF PILES (b)-Y-hWl GROUP OF PILES

FIG. 9-3. Action of group of friction piles

gravel. Ifsize, material, length, straightness, and penetration of the

the
piles are the same, each may be assumed to carry the same load. If one
pile can support safely a load of 30 tons, 10 piles can support 300 tons.
This situation does not apply, however, in many other cases.
A group of three friction piles is shown in Fig. 9-3 (a), and the spreading
of load by the cone of distribution of each one as an individual is pictured
by the dashed lines. Obviously, the affected volumes of soil overlap so
that material near F is affected by the load from two piles; and G by three
loads. The soil is what it is, and it probably will compress more under
the heavier loads than under the lighter ones. As pile HG tends to settle
308 PILES [Chap. 9

more than AF and EI, load will be shifted from HG to the others because
all will ordinarily be forced to settle equally if they are under a single cap.
The general result may then be pictured as though the load affects a
volume represented by ABODE, and the soil under the area shown in
section A- A must resist this force. Since the angle BAF depends upon
soil, the action of the pile group cannot increase it
the properties of the
appreciably above that applying for the case of a heavily loaded single
pile. Therefore, the safe resistance of this group of three piles should not
be considered equal to three times the safe load upon one pile alone
because there is less soil supporting the group than there would be under
three individual and widely separated piles.
In the action of a large group of piles like that shown in Fig. 9-3(6), the
overlapping of the cones of distribution and the areas of soil affected by
one pile are still more apparent. The total safe load of the no
group is

more than the bearing capacity below some such area as PQRS
of the soil
of section B-B. Therefore, the careless use of data regarding the safe
load carried by a single test pile, or even by a small group of piles, in the
design of large groups of friction piles is likely to be dangerous; so is similar
use of data obtained from the resistance of individual piles when they are
being driven. In such cases, it is best to study the structure as a whole,
including the soil below the piles as a part of the structure. The need for
adequate evidence regarding the properties of this soil is obvious. Many
piles are used in such a case as that of Sketch (6) because the load is large
and the soil is weak, yet the situation is one in which the resistance of each
pile may be relatively small and unpredictable. Thus, with the greater
importance of the structure and the greater difficulty of supporting it may
go greater uncertainty regarding the safe bearing resistance per pile, and
greater need for sound engineering judgment to attain safety without
unwarranted cost. It may
be that fewer piles at wider spacing would be
just as safe as many piles jammed
closely together since the load-resisting
capacity of the underlying soil may be the weakest part of the system.
In Fig. 9-3(6), the piles are pictured as being loaded equally. This may
not be true; those at the periphery of the group may have more resistance
per pile than those near the center, but, in cohesive and compressible soils
and with a properly designed cap structure, the stiffness and strength of
the latter will compel all piles to settle practically the same amount so
that, as the load tends to shift from the central to the outer piles, the
outer ones bear more heavily on the soil and tend to compress it more or
penetrate farther into it. However, considering the other uncertainties
involved, it is generally satisfactory to assume that uniformly applied
loads on a group of piles, such as in Sketch (6), are resisted equally by each
pile. It is the magnitude of the average safe load per pile that is uncer-
tain. This, of course, does not apply when the loads are eccentric, the
Art. 9-4} PILE DRIVING 309

character and lengths of the piles considerably different, and the soils
nonuniform in any important respect.
Another situation to consider when planning foundations with large
groups of piles may be illustrated by Fig. 9-2. Assume that the first pile
is driven into the sand and gravel stratum or crust but not through it.

The shearing resistance of the crust may spread the load over a consider-
able area S of the weaker stratum below it, as explained before. This
may be satisfactory for one or two piles, whereas a large group designed
for thesame load per pile may overload the crust, cause the substratum
to compress excessively, or compel the weak material to flow laterally.
On the other hand, if the piles are driven through the crust and the weak
stratum, then well into the coarse sand as in (b), the piles will support
most of the load by end bearing, and the structure will probably be
satisfactory.
A rather broadly interpreted analogy of the action of the crust in the
vicinity of R of Fig. 9-2 (a) may be the following: Assume that this sand
stratum is like a thin layer of ice floating on a pond. A boy skates on the
ice, and it bends down until it displaces enough water to counteract his
weight, provided it has sufficient strength as a structural slab to do so
without rupture. Now along come two or three more boys who join him,
temporarily. No longer can the ice bend sufficiently and transfer enough
forces to support all these concentrated loads; hence there is an accident.
9-4. Pile driving. By pile driving is meant the operation of getting a
pile into the position that it is to occupy as a part of the completed sub-
structure. Pounding it down may or may not be a part of the process,
although it generally is. The details of the method to be employed, or
the combination of methods, will generally depend upon the type of pile,
the characteristics of the soil into which it is to be placed, the conditions
under which the operations must be performed, and the equipment
field
that is available for and best suited to the job. Some of the most prev-
alent methods are described here.
1. Drop hammer. This is one of the methods and is relatively
oldest
elementary. Somewhat like Fig. 9-4, the equipment consists of a tall
framework with a heavy weight or hammer that can slide up and down be-
tween guides, means for lifting the pile into position between the leads or
guides ready for driving, a cable system running over a top sheave and
thence to the drum portion of a hoisting engine or other power unit, and a
clamp and trigger arrangement that permits attachment of the cable to
the hammer for lifting and then automatic tripping or disconnection of the
weight at some desired height from which it will drop onto the top of the
pile.

The operation of such a hammer requires considerable time between

blows, and each blow may deliver a violent shock to the pile. Perhaps
310 PILES [Chap. 9

unusually long piles cannot be driven well at first if the top (butt) end is
so high that there is insufficient distance left for the fall of the hammer.
Widely spaced heavy blows may tend to damage the butt or the tip of the
pile, or to buckle it sidewise. Be-
tween blows, the pile is at rest.
When a pile has not been driven so
far as the operator believes it should
go, there is the possibility that he
will raise the hammer far too high
and thus damage the pile by too
severe a blow.
2. Steam hammer. Figure 9-5
shows a steam hammer in operation.
Generally, the hammer itself consists
of a short steet frame with guides on
which a movable weight can travel.
This weight is attached to a piston
and cylinder device that operates
under steam (or air) pressure and
has proper valves so as to lift the
weight, then not only drop it but
accelerate it still further by steam
pressure on the opposite side of
the piston. The remainder of the
equipment must have means for lift-
ing a pile into position, guides for
the steam-hammer unit, and means
for raising and lowering the latter. 1

Raymond step-taper piles for

FIG. 9-4. Driving
A steam hammer delivers short
the extension of a power plant. The soil is
rapid blows that tend to keep the
approximately 30 ft. of silt over fine sand.
[Courtesy of Westcoit and Mapes, more nearly in continuous mo-
Inc., Con- pile
sulting Engineers, New Haven, Conn.) tion and vibration. It delivers a
practically uniform blow for any
given steam pressure. It is placed upon the butt of the pile and follows
the latter down. Such a hammer is less likely to damage a pile than
is a drop hammer, and it may cause less violent vibrations of the soil.

Outside the corner of an old wall-bearing factory in New England, the

blows from a drop hammer caused vibrations of the sand that consoli-
dated the latter enough to cause such subsidence of the soil that the

1
A vibrator type of pile driver has been developed which utilizes the centrifugal
force of a rapidly rotating eccentric weight resting on top of the pile. This causes
very rapid vibrations which "shake" the pile into the ground.
Art. 9-4] PILE DRIVING 311

corner of the old structure dropped more than 2 -, with serious crack- ym
ing of the walls. The designers had intended to use a steam hammer.
Whether the same settlement would have occurred with a steam
hammer is questionable but not probable.

FIG. 9-5. Steam hammer driving prestressed-concrete piles. (Courtesy of C. W. Makes leeand Sons,
New Haven, Conn.)

rf/ou/ of wafer
Precast-
concrete p/'/e

f
So/7 washed
upward—*:
Sand
^Flowing

Water pipe ^^>.' P

FIG. 9-6. One arrangement illustrating sinking of a pile by means of a water jet.

3. Water jet. This method of sinking piles is illustrated in principle

in Fig. 9-6. It consists primarily of a pipe inside the pile or alongside
itthrough which water can be forced under high pressure. This stream
ofwater at or near the tip of the pile tends to loosen and wash out the soil
so that the pile can settle easily into or be driven without difficulty
through the ground to the desired position. It is usually best to drive
down the pile a little to "seat" it into final position after jetting.
312 PILES [Chap. 9

Naturally, a water jet will turn clay into mud, and silt or rock flour into
a semiliquid. This method is therefore most useful when a crust must be
penetrated, as in Fig. 9-2, and other methods are, or are thought to be,
inadequate or undesirable. It is useful, too, when piles are to be driven
into a gravel stratum, as shown in Fig. 9-1(6). Of course, this method is
used primarily with end-bearing piles because the water disturbs the sur-
rounding soil so much. Granular soils, when dis-
placed this way, will settle back around the pile
> Footing'
7/N> fairly well; clays and silts may remain disturbed
and lubricated for many years.
I
4. Drilling. Figure 9-7 shows in principle a
method that has been used by the Texas Highway
£ X
k Department. In the summer the clay will dry
I-
(j
out, shrink, and crack considerably. When rains
^
^ <;
.Q.C come, the water penetrates quickly and deeply and
Cs t,
t\l
Q.^
-5 causes swelling and lubrication of the soil. By
C^ t
bis using an auger or by "carving" out a hole through
~~ ~<<o
<b
5. 3 the clay down to firm material, using a casing if the
presence of sand or water-bearing
intermediate
strata require it, and filling the hole with concrete,
one may construct an end-bearing pile. By means
of special equipment the Highway Department has
had these holes undercut so as to form a pedestal or
sort of footing that is very helpful. It is advisable
to extend the piles deeply enough to make sure that
skin friction along the part of the pile that is in the
FIG. 9-7. Principle of area of swelling clay will not lift the pile up. This
drilled concrete piles used
pile is a sort of caisson rather than a true pile.
by Texas Highway De-
partment to prevent
A modification of this general idea was used to
movement caused by clay penetrate a new fill containing scraps of rock mixed
that shrinks when dried with a physically unstable volcanic excavated
and expands when wetted. material. A large churn drill was used inside a pipe
casing, as for the drilling of deep wells. The hole
was drilled down to firm material, cleaned out, and concreted to form an
end-bearing pile.
5. When a pile is to be driven inside an existing structure
Jacking.
where headroom for ordinary pile-driving equipment is not available or
where vibration cannot be permitted, it may be possible to force the pile
down by using a hydraulic jack that reacts against a heavy weight above
it. Of course, this is a highly special procedure and generally requires
that the pile be jacked down in short pieces. Short lengths of steel pipes
that can be welded together as needed are useful for this purpose.
When a solid pile is driven into a granular soil, it will displace some of
Art. 9-4} PILE DRIVING 313

the sand and may cause consolidation of the soil because of the vibration,
as stated previously. Ifand cohesive,
the soil is plastic e.g., clay, it will
probably be forced to move sideward and upward, the latter being the
line of least resistance.
Assume that a single pile is being driven into a thick layer of clay, as
shown in Fig. 9-8 (a). As the pile displaces the clay so quickly, little of
the water can be squeezed out of it, whereas the clay must be displaced
by the pile. The arrows picture the general directions that the clay will
probably take, and the lines ACD and EFB indicate a new surface to
which the original dotted surface AB may be forced. This may not be
harmful for a single pile, for a small group of piles, or for a narrow row of

J <f Lf

Direction of
movement of
/ displaced soil

Sand
K M
(a)- DRIVING OF ^-EFFECT OF DRIVING MAY
SINGLE PILE LIFT PILES PREVIOUSLY DRIVEN

FIG. 9-8. Exaggerated picture of displacement of soil caused by driving of solid piles.

them. It may, however, cause the lifting of previously driven piles of a

large group.
For example, assume that several piles represented by JK and LM of
Fig. 9-8(6) have been driven through clay into sand as shown. Then
another series of piles represented by GH is driven close to them. It is
probable that these last ones will cause much of the clay around JK to
move upward slightly so that JK is lifted some distance d. Pile may LM
be lifted a smaller amount d' This is likely to weaken their end bearing.
.

Under such conditions, redriving of the piles may be advisable in order to

ensure proper bearing. The movement of the soil above the sand may
tilt the piles or injure them if the sideward movement of the clay is
sufficient and if the bottom ends of the piles are held laterally by the sand.
This is a feature to be guarded against in the Of course, any con-
field.

struction that merely removes the soil in the space taken up by the pile
avoids this uplift and lateral movement.
As a pile is driven, the adjacent soil is greatly disturbed. Sand will
314 PILES [Chap. 9

adjust itself quickly; clay may be

remolded almost to the state of lubrica-
tion so that some time is required for
it to reconsolidate. However, even
in the case of clays, a pile that may drive easily, when put down, may
appear to have the clay "set" around it so that it is difficult to start the
pile down again after one or several days of rest whereas, after it is started,
it may be driven easily once more.

Sometimes it is specified that piles are to be driven to refusal. This

is generally assumed to mean that they are to be driven until the resist-

ance to driving and the driving force are in equilibrium or as near so as

they can be without damage to the pile and there is little or no penetra-
tion. Driving to such extremes is hazardous; the piles may be injured
permanently. It is difficult to determine what should be the stopping
point to ensure the maximum resistance to load without damage from
overdriving. As stated so well by Fred D. Hartford " an engineer
1
. . .

sometimes consents to the public flogging of an innocent pile that cer-

tainly does his profession no credit."
In a contract for pile driving it may be desirable to state that the
contractor is to drive one or more piles to test the driving conditions and
depths before ordering the piles for the job. This is to prevent entire
dependence upon estimates of the required lengths, because the latter
may result in providing piles which are too short and cannot be used or
too long so that they are wasteful.
If the borings show the existence of heavy boulders where piles are to

be driven, one should investigate the driving problem very carefully.

Here is the story of one case:
A bridge was being built across one end of a lake in glacial territory.
The bottom was covered with 10 to 20 ft. of silt and peat; under that was
a varying stratum of sandy material, which, in turn, was underlain by
rock. However, in one portion there was a ravine from 20 to 30 ft.
deeper than the adjacent areas. It apparently was a channel for a
stream ages ago. In this ravine there was a layer of glacial till or com-
pacted sand several feet thick. This was directly on top of the rock, but
the remainder of the former channel had apparently been filled with
boulders by the glacier. The engineer designed the piers to be supported
upon heavy 14-in. steel H piles, each to hold a load of 90 tons. It was
thought that these piles would be able to penetrate through the boulders
to rock.
When the piles were driven in this ravine, it was found that they would
appear to "fetch up," then penetrate a few feet farther under hard
driving, until finally they refused to go deeper. Naturally, it was ques-
tionable as to whether they bore on bedrock or on boulders. The yield-
ing under persistent driving seemed to indicate that the piles either
1
Notes on Driving Timber Files, Engineering News-Record, Nov. 29, 1945.
Art. 9-4] PILE DRIVING 315

displaced obstructing boulders or penetrated between them. However,

the engineer was obviously uncertain as to the reliability of these piles
for supporting such heavy loads.
In a similar situation nearby, one of the H piles which failed to pene-
trate as far as thought necessary was pulled out. Its end was found to
be split and peeled back. Apparently, a boulder had wedged between
the flanges and web on one side, and under the hard driving, it split the
end of the pile. This gives an excellent idea of what may happen under
extreme conditions of driving piles and hitting obstructions. One cannot
help wondering if continued slight penetration of a pile under hard driving
means actual penetration into rock, brooming, or deflection and bending
as the pile finds some path of least resistance among the boulders.
The driving of these heavy H piles revealed another problem. The
single-acting hammer originally used proved to be inadequate to over-
come the inertia of these heavy 72-ft. piles and still have enough energy
left to secure proper penetration. The situation was somewhat like
trying to drive a spike with a tack hammer. Fortunately, the contract
clauses were so worded as to enable the engineer to demand that a heavier
hammer be used because the penetration of the piles seemed inadequate
when compared with that of the borings. Piles that had been driven to
l

refusal by the former hammer were redriven by the heavier one, with the
result that some piles penetrated several feet farther.
Piles driven in bouldery soil are likely to skew sidewise. It is probably
better to drive piles in such cases (especially batter piles) by means of
proper leads on the pile driver. The use of a template or frame for
guiding piles under such conditions may actually aggravate -damage to
the piles by bending if they are deflected by boulders. The possibilities
of this are illustrated in Fig. 9-8A. This shows how the frame may
actually force a curve into the pile, thereby not only contributing to poor
bearing but facilitating buckling of the pile under load.
Anyone who stands on the ground close to a displacement pile which
is being driven will realize that the driving operation causes considerable

vibration of the ground in the immediate vicinity of the pile. This

experience poses the question whether or not it is safe to drive piles
alongside an existing structure. If unsafe, how close to the structure can
they be driven without harm?
This question arose when Westcott and Mapes, Inc., 2 were designing
extensive additions to the Taunton Municipal Lighting Plant, Taunton,
Mass. The existing power plant is founded upon spread footings which
rest upon the soil shown in the log of the boring in Fig. 9-85. Inasmuch
as some differential settlement of the present structure had occurred in
1
These had been made by drilling, not merely by washing.
2
Consulting Engineers and Architects, New Haven, Conn.
316 PILES [Chap. 9

the past, the engineers decided that pile foundations seemed to be neces-
sary for supporting the new construction. However, they thought that
pile driving close to the existing structure might cause further settlement.
They therefore arranged to have a driving test made.
The engineers arranged with the Raymond Concrete Pile Company to
drive a step-taper pile as a test before driving any piles near the structure.
This pile was driven 40 ft. into the ground and was 70 ft. from the existing

Leads

FIG. 9-8A. Illustration of bending of a pile being driven among boulders and restrained by a guide
frame.

building. The hammer used was a Raymond No. 1 single-acting, deliver-

ing 15,000 ft. -lb. of energy. F. J. Crandell, Assistant Vice President and
Chief Engineer, Liberty Mutual Insurance Company of Boston, had
charge of obtaining readings of both vibrations and heaving during the
driving operation. He set up instruments 7, 37, and 67.5 ft. horizon-
tally from the test pile. The results of the test are the following:
1. In Fig. 9-8B(a), the curve marked A shows the general relative
magnitude energy generated by the driving of the pile as the point
of the
of the pile penetrated farther into the ground. These data were obtained
from the instrument located 7 ft. from the pile. The energy transmitted
Art. 9-4} PILE DRIVING 317

Instrument
50 "AW'
7A\V'/A\V/A\\''//»

Water
Z.005I? /00/77, 5tfrt</
1 table
oflO' gravel {fill) _j_
A-^
40 10
/7/7£> 5O/70',
Approx. T
e/ev. of
//>V/<? c/c/ footings
30 / 20-

/
/ /.ewe
some
Coarse sand
///?<? 50/?tf',
clay

and gravel
o 20 30-
/ <=>

Sand, gravel,
4
boulders, little
shale
40-

50-
0.01 0.05 0.1 0.5 1.0
Energy ratio

(a) (b) LOG OF BORING

FIG. 9-8B. Relation between energy transmitted to instrument and depth of penetration of pile.

0.07

0.06 1

0.05

••£ 0.04
\

,
c 0.03
,

0.02
-B

0.01

10 20 30 40 50 60 70
Horizontal distance from pile, feet

FIG. 9-8C. Decrease in energy vs. distance from pile.

318 PILES [Chap. 9

0.003

L = longitudinal [push] wave

0.002 V = vertical [surface) wave
T - transverse [shear) wave

-L
0.001

">
T
T
10 20 30 40 50 60 70 80
Horizontal distance from pile, feet

FIG 9-8D. Displacement vs. distance from pile. L = longitudinal (push) wave; V = vertical (surface'

wave; T = transverse (shear) wave. '

to the instrument increased as the depth of the penetration of the pile

increased. However, Crandell reports that a magnitude of 3 is considered
the safety limit for masonry structures. The recorded values are far
below this.
2. In Fig. 9-8C, the curve marked B shows the decrease of energy
1

transmitted as the distance from the test pile increased. The transmitted
energy declines somewhat as the square of the distance from the source;
it is very small at 37 ft., and it is negligible at 67.5 ft.

3. In Fig. 9-8Z>, curve L shows the attenuation of the longitudinal

(push) wave as the distance from the pile increased. This wave was the
strongest, but it declined rapidly to 37 ft. and then approached zero at
67.5 ft. Curve V shows the vertical (surface) wave, which declined
slowly but was negligible at 37 ft. from the source. Finally, curve T
shows the transverse (shear) wave, which declined slowly but was negli-
gible at a distance of 67.5 ft. All three curves show the results when the
pile had penetrated 40 ft.

4. Figure 9-8£' pictures the heaving of the surface of the ground in the
vicinity of the test pile. Of course, when a large group of piles is driven,
this heaving will be cumulative, and therefore considerable.

o 5 10 15
Distance from pile, feet

FIG. 9-8E. Curve showing heave of surface of ground from driving single 40-ft. pile.

1
This is the resulting energy in all three planes: radial, tangential horizontally, and
vertical with respect to the pile.
 PILE DRIVING 319
Art. 9-4]

These curves show that it was desirable to avoid driving piles near the

existing structure because of both vibrations and heaving. The engineers

finally decided that no piles should be driven closer
than approximately
20 ft. from the present building. All closer ones were to be jacked
down
even though this might cost eight to ten times as much as driven piles.

power plant. (Courfesy of Raymond

FIG. 9-8F. Equipment used for jacking pipe piles alongside a
Concrete Pile Co., New York.)

Therefore, it was decided to use pipe piles within a distance of 18 ft.

from the old foundations. These pipes were to be jacked down with open
ends. The pipes were 10% in. O.D., with }i-'m. thickness of metal,
except for a piece of 10 34-in. pipe \i in. thick about 2 in. long which
was
welded to the bottom and ground so as to produce a cutting edge. A
large reactor weight was constructed, consisting of a large frame sup-

porting steel billets and mounted on wheels, as pictured in Fig. 9-8F.

This device was developed by the Raymond Concrete Pile Co. The jack
320 PILES [Chap. 9

is at the top, and the links of the hold-down chains are such that one
complete link equals the stroke of the jack, so that, as the pile is forced
down, one pair of links is removed and the next pair is connected to the
frame. Jacking forces used were 85 tons; the design loads to be applied
to the piles were 50 tons.
The pipes generally admitted about 65 per cent of the soil penetrated.
They were cleaned out by water jets, then rejacked again to make sure
that no undermining of the bottoms of the piles had occurred. The pipes
were not cleaned a second time, but a 2-ft. plug of tre,mie concrete was
placed in the bottom. The pipes were later unwatered and concreted.
Another test conducted at the Taunton plant was a pull-out test of a
jacked pile. The one tested was 10% in. O.D., 34 thick, and approxi- m -

mately 26 ft. 5 in. embedded length. The results of this test are shown
in Table 9-1A. They reveal a maximum skin frictional resistance of
about 700 p.s.f. However, this was a single pile.

TABLE 9-1A. Pull-out Test of a Pipe Pile

Upward load, tons Up ward deflection,

9
18
22 0.001
24 0.010
25 0.025
26 0.040
27 9 1 J£ 6 (failure)
Note: Pile is 10% in. O.D.; length of embedment is 26 ft. 5 in.

If a group of piles were subjected to uplift, one should estimate the

weight of the soil which might resist the upward pulling. One way to
do this is the following
1. Estimate the total skin friction of the group of piles at perhaps

600 p.s.f.
2. Estimate the volume and weight of the imaginary frustrum of a cone

or pyramid of soil which may tend to be lifted out by the piles. For this,
use a slope of 1 horizontal to 4 vertical for silt and clay and 1 3 for fine :

to medium sand and small gravel, starting with a geometrical figure which
encloses the pile group at the tips of the piles and is tangent to their
outer edges, then sloping upward and outward to form an imaginary
solid of earth resting on its small end. Consider the effect of buoyancy
when computing the weight of submerged soil.
3. Use the smaller of these resistances as the ultimate strength of the

pile group, then divide this by the desired safety factor (perhaps 1.5 to
2.0) to compute the safe working resistance.
Art. 9-5} WOODEN PILES 321

9-5. Wooden piles. 1 Centuries ago men used the trunks of trees as
piles to hold up water-front and other These products of
structures.
nature are still used, although now in competition with other types of pile.
For the purpose of convenient comparison, the characteristics, advan-
tages, and disadvantages of wooden piles and other types will be discussed
under headings arranged in the same sequence but in separate articles.
1. Availability. Trees are a renewable resource and can be grown in
many parts of the world. However, a tree that is suitable for a large pile
may require 75 to 100 years (or more) for growth. The people of the
United States have begun to awake to the long-term advisability of curb-
ing the waste of our timber resources and of conducting planned measures
for their replenishment.
It is becoming difficult to obtain wooden piles that are sufficiently long
and straight. Lengths of 30 to 50 or 60 ft. are reasonably available in
pine; those over 60 ft. are readily available in the West Coast area.
Wooden piles from 60 to 1 15 ft. long are obtainable in Douglas fir in the
Pacific Northwest.
2. Shipment. Wooden and easy to handle.
piles are relatively light
Since they float, transportation by rafting may
be very helpful in the case
of water-front structures. Transportation by ship may be made economi-
cally over long distances; by railroad, over fairly long distances.
3. Handling. Their lightness and strength as beams facilitate the han-
dling of wooden piles. Their tensile strength also permits them to be
pulled readily and reused as parts of construction accessories such as
trestles, falsework, fenders, piers, and working platforms. The ease with
which they can be cut and worked in the field is also a great asset.
4- Driving. Hard driving is likely to "broom" the butts of wooden

piles, and the tips may be crushed when driven through or into dense
gravelly materials. The lack of straightness may cause them to deflect
from the intended position when driven, and hard driving may crumple
them so that yielding may be mistaken for further penetration. Some
such cases are pictured in Figs. 9-9 and 9-9 A.
5. Adjustability of length. It is very easy to cut off a wooden pile that
does not penetrate to the expected depth. When it proves to be too
short, can be driven farther by means of a follower, a short piece of
it

timber placed on top of the pile. The permanent extension may then be
made of concrete or a short length of wooden pile spliced on the lower one.
Wooden piles may also be cut off at a specified elevation under water by
means of a saw supported upon a framework above water level.
6. Strength. Wooden piles are good as friction piles; they are not so

1
See also American Society of Civil Engineers, Timber Piles and Construction Tim-
bers, Bulletin 17, Manuals of Engineering Practice, 1939.
322 PILES [Chap. 9

desirable as end-bearing ones because the compressive strength of the

wood is relatively small, the tips in bearing are small also (unless the pile
isdriven with butt down, a difficult and unusual procedure), and they are
not inherently strong columns. Ordinarily, a wooden pile should not
have to support more than 30 to 40 tons under the best conditions.
It can be braced by bolting on diagonal members and struts above or
below water.
7. Flexibility. Wooden piles are relatively flexible and springy lat-
erally. They will deflect considerably, offer strong lateral resistance,
and spring back into position again, thus cushioning the shock against a
colliding ship. This is a great asset in such structures as small piers,
ferry slips, fenders, and dolphins. For such service, it is difficult to find a
satisfactory substitute for them.

Rock
(a)-EW BROKEN BY CW-PILE BUCKLED CcJ-END BROOMED (d)- FIBERS
SKIDDING OFF AND THEN AND SPLIT CRUSHED
SLOPING ROCK •
CRUMPLED BY GRAVEL
OR BOULDER

FIG. 9-9. Some types of failure of wooden piles from overdriving after hitting an obstruction.

8. Durability. one of the grave questions connected with the

1
This is

use of wooden piles. may be destroyed by fungi,

Their usefulness
marine borers, and mechanical action.
—
Fungi. These a low form of plant life cause what is ordinarily —
termed rotting. For growth, fungi need food, air, moisture, and a mod-
erate temperature. The food is found in the wood itself. If the piles are
continuously immersed in water, the necessary air is excluded. If the
wood is perpetually and thoroughly dry, the requisite moisture is missing.
If the temperature is frigid or very high, the fungi cannot survive. The
conditions for their growth are generally most advantageous at or near
the surface of the ground.
Inasmuch as piles are a part of some foundation, and the necessity for
them usually arises because of the presence of water and of weak wet
plastic soils, about the only means of preventing fungus growths is to have

1
See also R. D. Chellis, "Pile Foundations: Theory, Design, and Practice," chap. 13,
McGraw-Hill Book Company, Inc., New York, 1951.
Art. 9-5] WOODEN PILES 323

the piles completely immersed, or to add poison to the wood so as to kill

the organisms or greatly retard their growth.
This matter of submerging the piles below the "permanent" watertable
may seem easy and the results trustworthy. However, it is fraught with
dangers in many cases, and extremely good engineering judgment is
needed to determine the wise course of action. Two illustrations are the
following: 1

1. The Boston Public Library on Copley Square, Boston, was built on untreated

wooden piles. Some years later settlement was found to be caused by decayed
piles. It is reported that the ground-water level had been lowered, probably as
the result of building and draining a subway.

FIG. 9-9A. A damaged wooden pile. (Courtesy of L. D. Luck, Pullman, Wash.)

2. In the Brooklyn (N.Y.) Navy Yard, buildings which had been standing
began to settle. The untreated wooden foundation piles were
safely for 40 years
found decayed below the original ground-water table. The original water level
had been materially lowered by pumping in the western end of Long Island, which
had been going on at an increasing rate to meet industrial and other demands.
In some areas the ground-water level was below tide level, but inflow of salt
water was prevented by the. mud cover on the shore.

One of the best methods for preserving wooden piles through poisoning
of the fungi is pressure creosoting of the piles. In the full-cell process, a
1
American Society of Civil Engineers, Pile Foundations and Pile Structures,
Bulletin 27, Manuals of Engineering Practice, 1946.
324 PILES [Chap. 9

vacuum created around the piles, then the creosote is applied under
is

pressure. This is best for marine piles and timbers. The empty-cell
process for foundation piles and other timbers utilizes initial vacuum,
then the application of the preservative under pressure, and finally the
reapplication of the vacuum. Paint coatings and surface creosoting are
inadequate. The wood should be as completely impregnated as possible.
Bolt holes, notches, cutoff of butts, and even shrinkage cracks may provide
access for the fungi to any untreated portion
wood. If possible, all
of the
fabrication should be done before creosoting but this is not always —
practicable. However, field applications should be used to protect such
cuts as much as possible. Proper creosoting will generally prolong the
lifeof wooden piles many years.
Marine borers. These creatures occur mostly in sea water or brack-
ish water. Among the destructive varieties are the Teredo and Limnoria.
The following also is quoted from Pile Foundations and Pile Structures:

The severity of attack varies greatly in different localities. Marine borers are
most active in warm waters and in the tropics, but they have done serious damage
to timbers in waterfront structures almost to the northern and southern limits of
the respective temperate zones. Borers are found in Norway, practically in the
Arctic, and a dock has been destroyed in 18 months in Alaskan waters, latitude
60°N.
Where marine borers are not present, consideration must be given to the possi-
bility of their introduction. For many j'ears the waters of the upper part of San
Francisco Bay, in California, were considered free from marine borers, but a
sudden infestation caused the failure of waterfront structures. Boston and other
New England harbors (borers first noted in 1933) are other such examples. The
water of New York harbor is now practically free of borers possibly because of —
the low oxygen content due to sewage pollution. The introduction of sewage
treatment on a wide scale in such a harbor may result in the admission of marine
borers in destructive quantities.
Since the water conditions which govern attack are not completely understood,
a prediction as to the future is uncertain, and conservative practice normally
justifiestreatment of piles and timbers in marine water. In fresh- water lakes and
rivers, marine borers are not present, although in southern streams Sphaeroma
may be destructive during periods when the water is brackish. Teredos are
destructive in some fresh-water harbors in India and Australia.

The following data are quoted from "Protection against Marine Borer
Attacks," by William G. Atwood: 1

Many years of research and study have not resulted in finding any method of
preservation as economical and effective as impregnation with coal-tar creosote.
To use this material properly it is necessary to obtain the best creosote, use enough

of it, select the timber having the required strength and other qualities which can

1
Wood Preserving News, December, 1935.
Art. 9-5] WOODEN PILES 325

be properly treated, and see that the treatment specified is obtained and that the
timber is not damaged after treatment.
The most satisfactory timber for treatment is southern yellow pine, especially
that with a large amount of sap wood. Heartwood can not be penetrated to any
appreciable depth and, therefore, any heart requirement in the grade of timber
used should be waived. The U.S. Forest Products Laboratory has proven that
there is no difference in strength between heartwood and sapwood, or between the
different species of pine. Specifying longleaf timber instead of southern yellow
pine is a pure waste of money, and frequently obtains timber with a large amount
of heartwood, which is resistant to treatment. The use of standard specifications
of the American Society for Testing Materials or of the American Railway
Engineering Association will generally result in lower prices and more uniform
material.
Considering the cost of the pile alone, a 16-lb. per cubic foot treatment with
which is as light as should ever be used for piles in salt water,
coal-tar creosote,
about doubles the cost of a pile delivered in New York. One of the recently built
New York City one-story piers cost about $622,000, of which about $62,000, or
10 per cent, represented the cost of the untreated piles themselves. If these piles

had been treated their cost would have been 18 per cent of the total, or an increase
of 8 per cent. The two-story piers recently built cost about $933,000, with an
untreated pile cost of about 6.6 per cent. If these piles had been treated the pile
cost would have been about 12.5 per cent, or an increase of only 5.9 per cent in
total cost.

Driving of wooden piles into gravel or soil containing boulders may

damage the piles seriously so that the treated portion is crushed or
cracked to such an extent that decay of the interior can occur. Even
above the soil, hard driving may cause or increase cracks that perrnit fungi

or borers to attack the pile. Therefore, one should be careful to avoid

anything that will nullify the benefits of the protective treatment that he
has spent so much to obtain.
A brief outline of some of the data given by Chellis relating to the 1

physical protection of wooden piles against attack by borers is the

following

1. Complete protection is necessary:

a. Good material and excellent workmanship are essential.
b. Bark is only temporary protection. Limnoria will enter through
knots.
c. Packed fill is good protection, even riprap may be helpful.
d. It is essential to seal the bottom in soil. Scour may remove soil

relied upon.

1
R. D. Chellis, Finding and Fighting Marine Borers, Engineering News-Record,
Apr. 1, 1948.
326 PILES [Chap. 9

e. Abrasion may remove or damage coverings.

/. The wooden piles should be creosoted properly even though they
are to be covered by some protective coating.
2. Metallic armor:
a. The must be made as smooth as possible.
surface of a pile
b. Metal may
be nailed on in advance over bituminous membrane.
c. The material of the cover and nails must resist corrosion well.
Copper and zinc are costly.
d. Scupper-nailing with copper nails may resist Limnoria if done
properly.
e. Iron and steel sheathing and mesh are not practicable.
/. Such coverings are not used frequently now.
3. Concrete encasement:
a. Metal shields with spacers may be used as forms.
b. Precast covering sections may be suitable. There may be a
question regarding proper sealing.
c. Encasement poured under water may not be uniform and strong.

d. Hay process (Cement Gun Co., Inc.). The concrete is placed in a

watertight form above water, then both are lowered to position.
The form is removed later.
e. Presscrete method. A
metal form is used, the water is ejected,
and concrete placed by compressed air, and the form is removed
later.

/. Johnson-Western method. A circular reinforced gunited jacket,

5 ft. long, is made above water with felt over spacer battens as a
back form. This form is lowered into mud, others are placed on
top of it, and the annular space is grouted.
g. Shotcrete or gunite may be used to make a reinforced casing put
on before driving. Arsenious oxide (AS2O3) seems to reduce solu-
bility of portland cement. Pressure-jacketed concrete may also
be used, this method being controlled by The Presscrete Co., Inc.

Marine borers cannot live below the mud line. They need the supply
of oxygen furnished by open water. Their growth might therefore be
prevented by encasement of the piles in reinforced concrete or noncorrodi-
ble metallic shells. The former
is likely to crack to pieces because it is

relatively rigid andwhereas the wooden pile is flexible; the metallic

brittle
covering is difficult to apply properly and to protect from damage.
Mechanical action. Wooden piles may be damaged by abrasion
caused by ships, ice, and driftwood; also by granular soils that are pro-
pelled along the surface of the ground by wind, current, or waves. Piles
that are lashed together to serve as fenders may be worn badly by chafing.
Fire, too, may destroy wooden piles when they are above water.
 1

III!

!z o
oz
s
ooo

L/jBua/ Auy

fj-bs$£ U'te££ U'telS W

hs6Z W^IZ U**Si
uoyoas /tonpiAipui joj svajio aovjjns ajvw/xojddy
k
-
* S-itk
S <o ho? "o Q

^
<0 K
li ,«)
^
o s s
10 ^ S

~U 'teS£ U 'te££ U 'fc/P V 'b?6Z

'

' Wb?d U bsSi

'

U'bs£Z
SUO/J33S /enp/Aipui joj lodjv aoojjns afvu/ixojddy

vdJK> gjvjjns 1VJO4 9j.0un)<ojddj/ I

5^
u'fcm U'b?0£Z VT'b$L6l Wbsgy TubmMJ'bs Oll YW
bsS9\U'bsZ9 WbsQi,

jfbss£ lfbs££ u>bs>/£yb?6Z Wb?LZ U'bsSZ 'U'bs£Z 'U'b?ZZ Wb */Z 'U'^61

U0IJ035 /Vnp>/AipUI JOJ P3J0 33k)JjnS 3j.lOUJjX0jdd\/

an
jl£ ,0£ oz .SI
t' Id
^
lojiJ
Q UJ
WMMSMMM1 •55t
*>0
o <=J

u'U b?£9
4J-bs(,pi ybsoit U 'bss9 ~

'•U-'teSU
L/idu3/ paieuBisapjo 9/ id joj &9jv aonjjns 3jvujixojddy

327
328 PILES [Chap. 9

Cast-in-place concrete piles. There are several variations of

9-6.
this general category,but they may be divided into two classes: (1) those
having a metallic shell that is driven and left in the ground, then filled
with concrete, and (2) those having the shell removed after or during the
deposition of the concrete.
In the first class are the following:
1. The Raymond tapered or step-tapered piles, such as those shown in
Fig. 9-10. The shell is driven down by means of a mandrel that is with-
drawn after the shell is in its final position. The concrete
is then poured. These are primarily friction piles.
2. The Monotube pile shown in Fig. 9-11. These shells
are strong enough to be driven as a metallic pile, then they
are filled with concrete. These, too, are generally best as
friction piles.
3. The MacArthur corrugated-shell pile pictured in Fig.
9-12. A strong open-ended pipe and an inside core are
driven first, the core is removed, a corrugated shell is

dropped inside the pipe, the shell is filled with concrete,

and the pipe is withdrawn. These are primarily end-
FIG. 9-11. Union
bearing piles.

Metal Monotube 4. The closed-end steel-pipe pile shown in Fig. 9-13(a)

thin 5 steel-shell, or (6). This is driven into the ground with a steel shoe of
concrete-filled
some sort or a concrete button to close the bottom end.
piles. (Shell is

capable of being
The pipe is then filled with concrete. These shown are
driven as a unit.)
especially good as end-bearing piles but may be of con-
siderable value as friction piles also. Some engineers
prefer to drive a pipe pile with a flat plate welded (watertight) on the
bottom instead of a conical point. They believe that stones or other
irregularities in the soil will not so easily cause a flat-end pile to deflect

li ||| s
^
•SU w
5- I o 5 *

<a ~'t
P;
°
^
ia *

;?.'.'>•
1 i 1
4
:?•
to < .>..-.:.: !

Steel tip-
1 '""
Sfeelii
caD ^m<m <M
•0 [>

raJ-CAP AND PIPE ^-CONCRETE FILL (cl INITIAL STAGE (W/-FINAL PEDESTAL feAPILEWITH r/V-PEDESTAL AND
DRIVEN BY RAMMED DOWN IN MAKING PILE OF PLAIN THIN STEEL THIN STEEL SHELL
MANDREL AS PIPE IS SIMPLE PEDESTAL CONCRETE SHELL COMBINED
WITHDRAWN PILE

FIG. 9-12. Some cast-in-place concrete piles made by MacArthur Concrete Pile Corp. (Similar con-

struction is used by others.)

Art. 9-6] CAST-IN-PLACE CONCRETE PILES 329

sideways during driving and that the value of the end bearing may be
superior to that of the pointed ones.
5. The open-end steel-pipe pile pictured in Fig. 9-13(c). It is driven
with the bottom end open. This enables the pipe to cut through gravelly
crusts and to penetrate to rock or hardpan without displacing the soil as
much as would a closed-end pile. The soil inside the pile is then removed
by water and compressed-air jets, or by direct excavation, and the pipe
is filled with concrete. It is often desirable to clean out such a pipe pile
intermittently as the driving progresses in order to avoid having the soil

jam so hard inside the pipe that no more can enter and the pile will act
as a displacement pile in shoving the earth aside.

P °:

» o
* ••'»•'."•'

^ft

^ .'

U
p
jjs
qi

$
«i
p

S a »
Is

o
s
<o
D
0-
^
?^ £>
8
*o
-^
"^
1
^
1CZJ <0
9
i/j

%^w
\A ra)-DRW\NG PIPE
LEFT IN PLACE
/WHIN SHELL TO
HOLD CONCRETE.
faJ-NATIONALTUBE (^-CLOSED- (c)-OPEH- IF NEEDED f Driving pipe
CO.'S PIPE PILE END PILE END PILE Mot customary) removed)

FIG. 9-13. Some steel-pipe, concrete-filled piles. FIG. 9-14. Western Foundation Corp.'s but-

ton-bottom piles.

6.Another variety of poured-in-place concrete pile is shown in Fig.

9-14. This has been developed by the Western Foundation Corporation
and the Western Concrete Pile Company. It consists of a precast-
concrete "button" as shown in Sketch (a) that is driven down by means of
a heavy detachable pipe. A corrugated shell dropped inside this pipe
may then be filled with concrete whereas the driving pipe is withdrawn.
'These are especially useful for end-bearing piles under conditions where
one must penetrate unreliable material overlying a good stratum of
dense sand, gravel, or hardpan. There is no doubt as to the reliability
and symmetry of the button, and its large area usually provides adequate
bearing for heavy loads. Of course, the settlement of the surrounding
material into the space disturbed by passage of the button and occupied
by the driving pipe renders this type inadvisable for general use as a
friction pile.
330 PILES [Chap. 9

In the second class — piles without metallic shells — are such as the
following
1.The simple cylindrical concrete pile shown in Fig. 9-12(6). This
may made by driving a pipe with a closed end or core, filling it with
be
concrete, and then withdrawing the pipe. The wet concrete will settle
against the soil as the restraining effect of t»he pipe is removed, or the con-
crete may be rammed down. This is especially useful as an end-bearing
pile, but may have considerable value as a friction pile also if there is
it

good contact between the soil and the concrete along the sides.
Another variation of these cylindrical, poured-in-place concrete piles
is sometimes called an auger 'pile. This is made by "drilling" a hole in
the ground by means of some type of rotating excavator. The hole is 1

then filled with concrete that is poured directly against the earth. After
the concrete has set, the pile can be used as an end-bearing or as a friction
pile, depending upon the circumstances. It is obvious that the quality
of such a pile depends upon one's ability to bore the hole and to keep it
clear of soil and water before and during the placing of the concrete.
Clay soils, and those with sufficient clay or other binder, may be suitable
for auger piles and other types where the concrete is poured into an open
hole in the ground. —
Sandy soils both dry and saturated tend to run —
into the holes so as to make the drilling impossible, the placing of the
concrete difficult or impossible, and the quality of the piles uncertain.
Gravelly and bouldery soils — including such as glacial — render the use
till

of these piles difficult or impossible.

Still another variation of this sort of poured-in-place concrete pile is
made by means of an auger attached to the end of a pipe or hollow shaft.
The auger is rotated and worked down through the ground, leaving the
disturbed soil above the cutting tool. Then cement mortar (or even
concrete) is forced down through the shaft under pressure as the shaft and
auger are withdrawn. The mortar is to displace the disturbed soil as
the auger forcing the soil out of the top of the hole and leaving a
rises,

column mortar or concrete in its stead. The depth of such a pile is

of
usually limited to 40 or 50 ft. If the soil is fine saturated sand, there is

some question as to whether or not it will mix with the concrete or even
displace some of the mortar, causing a weakened portion in the pile,
especially when below the watertable.
Any cast-in-place concrete pile not having a permanent metallic or
other shell, when it is used in "runny" soils such as fine sand, loose silt,
or even coarse sand and gravel, may become defective where the soil

displaces the concrete as the shell is withdrawn, forming the equivalent

of a rupture of the pile.
1
For convenience, we may look upon the boring device as a sort of overgrown bit,
such as those used by carpenters.
Art. 9-6} CAST-IN-PLACE CONCRETE PILES 331

Plug.

Casing

Bulb

[a) [6)

id) [e)

FIG. 9-14A. Franki pile, or displacement caisson

2. The MacArthur pedestal pile pictured in Fig. 9-12(d). After the

pipe is driven and the core removed, some concrete is deposited in the
pipe. The core is reinserted and forced down so as to press the concrete
into a bulb or pedestal below the pipe to provide a greatly increased end-
bearing area. The pile is then completed as in item No. 1.
3. The Franki pile, or displacement caisson, is a modern application of
the use of a bulb or pedestal formed in place at the bottom of a pile.
332 PILES [Chap. 9

The principles of the construction of this type are the following, referring
to Fig. 9-14A
(a) A steel pipe casing is spotted at the desired location and held in a
vertical position. A plug of 3 to 5 ft.
3
of dry concrete is placed in the
bottom of the pipe and packed tight by means of a 7,000-lb. ram inside
the casing.
(b) The ram is then raised 10 to 20 ft. and dropped freely on the plug.
This forces the plug into the ground, but the resistance between the plug
and the casing in effect drags the pipe along with the concrete. The
earth is displaced by the plug and pipe, and it is compacted into the sur-
rounding soil or shoved aside.
(c) When the desired depth is reached by the casing, the latter is

anchored to the driving frame, the casing is raised slightly, and the con-
crete plug is driven down. Additional concrete is inserted in small
quantities around the ram, and it drops to the bottom as the ram is raised,
then it is pounded down, thus forming a bulb of concrete which is not
only compacted itself but compacts the adjacent soil under the effect of
the blows of perhaps 150,000 ft. -lb. energy.
(d) When the proper resistance is attained, the casing is gradually

withdrawn and successive small quantities of concrete are rammed down

to compress the neighboring soil and to form the shaft of the pile (or
caisson)
(e) When the loads are not too great, the shaft may be capped by a
concrete extension which serves as a pier or pedestal for the direct sup-
port of the superstructure.
The following information gives the author's ideas regarding the advan-
tages and disadvantages of cast-in-place piles:
1. Availability. The metallic shells and pipes are generally available,
and the concrete can be secured or made almost anywhere. The materials
are trustworthy. The piles without permanent shells require little or no
metal.
2. Shipment. The light metallic shells and pipes can be shipped easily;
short lengths may be welded together. The Raymond step-tapered pile
has parts that are fabricated to fit together so that the shells need not be
shipped as long units.
3. Handling. The shells and pipes are so light and strong that han-
dling in the field is relatively fast. The concreting operations are simple
and efficient also.
4- Driving. The driving of these piles is generally relatively simple and
reliable. Boulders, large gravel, buried objects, and very hard driving
may tear thin shells. Jetting is seldom resorted to unless done in advance
to loosen the soil.must be broken through, the mandrel or a
If the crust
pipe may be driven part way and withdrawn before the shell is driven
Art. 9-6} CAST-IN-PLACE CONCRETE PILES 333

down to final position. These piles with shells can often be inspected
reasonably well before concreting by lowering electric lights in the shells.
The open-end pipes are especially useful where movement of the soil dur-
ing the driving of solid piles is feared, e.g., next to existing structures or
piles. Pipe piles are also very strong so that they may be the best for
driving in bouldery soil without failure. In one job, such piles occasion-
ally deflected off the edge of a boulder so that they actually tilted. How-
ever, the pipes were not damaged badly, the questionable piles were
detected, and they were pulled out or not relied upon. The shells and
pipes may
be used as a permanent part of batter piles.
5. In this respect these piles have a great
Adjustability of length.
advantage. Excess lengths of shells may be burned off, extensions may

(a)- STEEL PIPE

"PROJECTILE" W-WOODEN PILE
EXTENSION EXTENSION

FIG. 9-15. Two types of composite pile used by the Western Foundation Corp.

be welded on unless the total exceeds that permitted by the mandrel or

core, the concrete may be poured to the desired elevation, and reinforce-
ment may be used in the concrete sometimes, except in the shell-less
type where settlement of the concrete might make this difficult.
6. Strength. In most cases the strength of these piles is excellent, if
the soils permit its development. The concrete is the main supporting
material. However, the following points are raised as questions to be
borne in mind:
a. Will the driving of piles alongside newly concreted ones damage the

latter? Several empty pile shells should be left between the concreted
ones and those being driven.
b. Can the piles be redriven (reseated) if heaving of the ground occurs?

c. Will water inside the shell or pipe cause the concrete to be poor in

quality?
d. Will granular soils settle amidst the concrete of shell-less piles as the
334 PILES [Chap. 9

pipe withdrawn, causing constrictions or eccentric sections? Will water

is

in the soil weaken the concrete under similar circumstances?

e. Will the bottom of a pedestal pile be shaped as hoped for? Will it
be forced to one side to cause an eccentric bearing, and will the displaced
soil injure adjacent piles?

These piles are heavy, and their own weight should be included in the
loads that are applied to the surrounding and underlying soils.
7. Resiliency. These piles are a rigid type and, except for the pipes,
may crack if subjected to lateral bending and shocks.
8. Durability. These piles are inherently durable because of the con-
crete, unless the soil contains acids. At one industrial plant, the soil near
a slag dump was found to contain considerable sulphuric acid. Even
though the shells may rust away, the concrete should be durable in other
situations.
9-7. Precast In Fig. 9-16 are shown some details of
concrete piles.
precast concrete piles as presented by A. E. Cummings, Research Engi-
neer of the Raymond Concrete Pile Co. These are strongly reinforced
1

as columns and designed to resist the bending and shocks that may be
incident to field operations. The tapered piles are generally limited to
the shorter lengths, 35 to 40 ft. The parallel-sided ones are those used
for the larger sizes, which may be as much as 24 in. or more in width and
over 75 ft. long.
Precast concrete piles are now made by using pretensioned cables or
wires as reinforcement in 5,000- to 6,000-lb. concrete. These usually can
be driven with flat ends. They can be made with square or octogonal
cross sections and reinforced with spirals. This type is very tough, and
shrinkage cracks are avoided.
1. Availability. Precast concrete piles may be made of high-strength
concrete placed in forms at or near the site of the job. Their reinforce-
ment may be very strong, properly encased, and well tied together. The
materials may generally be secured conveniently. adequate Means for
curing, handling, and storage must be provided. If there are to be
insufficient piles to warrant the provision of necessary forms and equip-
ment at the site, or if there is inadequate space there, then these piles
may not always be procured easily. Prestressed ones are usually
factory-made.
2. These piles are so heavy that transportation is costly
Shipment.
except when
can be done by barges. The piles should be cast at or near
it

the location of any large job. The transportation cost may be a great
disadvantage if only a few piles are to be used, and it is necessary to set up
special equipment to load and unload such long heavy members.
1
Pile Foundations, reprinted in 1940 from the Proceedings of the Purdue Conference
on Soil Mechanics and Its Applications.
Art. 9-7] PRECAST CONCRETE PILES 335

3. Handling. A precast concrete pile 18 in. square and 60 ft. long may
weigh approximately 10 tons. This requires heavy equipment for han-
dling, and it generally necessitates a long reach also. The piles must be
picked up by slings or special accessories that prevent excessive bending
when the piles are raised from a horizontal position. Special strong
equipment is needed also for the transportation of the piles about a large
site. The pile driver itself must be able to handle these heavy members.

^/"chamfer

mm
\l6"h24' 12'% 16

-8

\U 1/
\
IO"±

faJ-SQUARE,UNIFORM ft>;-TAPERED,OCTAGONAL (W-UNIFORM.OCTAGONAL

PILE PILE PILE

FIG. 9-16. Precast concrete piles. (Prestressed concrete piles are generally made in square or
octagonal shapes with uniform cross section.)

4. Driving. The inertia of a heavy pile tends to resist the shock from
a hammer blow. On the other hand, even though the pile is given a rela-
tively small initial velocity by the blow, its momentum is still large, and a
correspondingly strong resistance to penetration is necessary to stop it.

However, the use of a cushion head is usually essential to pre-

on the pile
vent local spalling of the concrete under the impact of the hammer.
Otherwise, since these piles are strongly reinforced, they can be driven
safely in most cases. If one glances off a boulder, the lower portion may
be cracked, or the tip spalled. These piles can be reseated if swelling of
the ground makes this necessary. They are also so strong that they are
unlikely to be damaged badly by the driving of adjacent piles. The use
336 PILES [Chap. 9

of a water jet is especially advantageous with precast piles when they are
to be driven through or into gravelly materials.
5. Adjustability of length. This is one of the disadvantages of precast
concrete piles. they prove to be too long and cannot be driven to the
If

anticipated depth, it is difficult to cut f,hem off. Even if cut off, the top
of the remaining concrete is likely to be cracked. If the piles are too

short, it is difficult to extend them so as to secure any effective splice with

the original pile. This feature may be especially bad when the piles pass
through air or water and are to serve as columns that may be subjected to

FIG. 9-16A. Driving of prestressed concrete batter piles by use of hanging leads, steam hammer, and
guide frame to control direction of pile. (Courfesy of C. W. Blakeslee and Sons, New Haven, Conn.)

lateral bending and buckling. An irregular rocky bottom, or one contain-

ing boulders, may almost impossible to anticipate the proper
render it

lengths when the piles are ordered or cast. In such a case it may be
possible and desirable to order the piles a few feet short, to leave about
3 ft. of the longitudinal reinforcement projecting for splicing, to use a
is designed to fit over the bars during driving, and
special follower that
then to cast on them whatever reinforced extensions are necessary.
6. Precast piles, especially those having a constant cross sec-
Strength.
tion,can be excellent columns, and they can be good friction piles, too.
They are excellent when needed to resist uplift, provided the reinforce-
ment or the pile as a whole is anchored properly into the main substruc-
ture. The tips may have a large bearing area, especially when jetted
Art. 9-7] PRECAST CONCRETE PILES 337

down. One can be reasonably certain of the strength of such a pile as far
as column action is concerned. Another possible advantage of great
value is the fact that a precast pile may be subjected to a heavy load
immediately after driving. This is very helpful when a trestle is being
built and the erection equipment is to build each unit of the structure in
series, being supported each time upon the part previously erected. It is
a great help also when a large group of precast (or wooden) piles is being
driven if the piles can be used promptly to support the pile driver because
such equipment cannot "flit" from place to place. Of course, the weight
of these piles should be included in the loads applied to the underlying
soil.

7. Resiliency. Since they are reinforced so well, these piles are strong
in beam action if However, severe bending as vertical
held properly.
cantilevers projectingfrom a sandy bottom is likely to cause tensile crack-
ing of the concrete and lead to deterioration. Sharp lateral blows may
crack and kink them. They should not be subjected to service that
requires true resiliency, although prestressed ones are better in this
respect.
8. Durability. In general, embedded precast concrete piles are very
durable. One source of possible deterioration of exposed piles is the
spalling of the concrete at and just above the water line because of the
freezing of absorbed moisture, the crystallization of salts in the pores, the
rusting of reinforcement, and the attack of chemicals in sea water. Acids
and other chemicals in the soil may injure them. These chemicals may
be caused by leakage from industrial operations and sewers; by leaching
from piles of coal, slag, and cinders; and by organic decomposition.
Occasionally electrolysis may be harmful. It is not desirable to ground
a structure to one or two piles that have steel as a part of their structural
system. Ice, driftwood, colliding ships, and wind-blown and wave-
washed sand may cause physical damage. Special cements, dense and
"rich" concrete, adequate cover over reinforcement, and bituminous
coatings are means of reducing disintegration.
It has been reported that square precast piles used in the bents of a large
trestle in sea water had visible shrinkage hair cracks in the portions above
the water. This is not improbable in the case of such heavily reinforced
members that are not subjected to a sufficiently heavy dead load because
the column strength of the longitudinal steel may exceed the tensile
strength of the concrete, whereas shrinkage during setting and curing
compels the concrete to shorten. Loads that compress the reinforcement
would probably close the cracks. Whether these cracks will cause ulti-
mate rusting of the steel and spalling of the concrete remains to be seen.
A paint coat of bituminous material or some other suitable one might
seal the cracks sufficiently. Pretensioning prevents such cracks.
338 PILES [Chap. 9

9-8. Steel piles. Under the classification of steel piles are included
those that depend primarily upon steel for resistance to longitudinal and
transverse loads. The pipes and shells used as a part of poured-in-place
concrete piles will resist some loads, especially the thicker pipes. If the

latter are designed as the chief load-supporting constituent, then these

piles might well be classified as steel piles. However, this name is used
here generally to denote steel H piles and those in which a steel member
is the principal part, whether it is encased in concrete or not.
Among the advantages of steel piles are their great strength as columns,
the high unit strength of the metal itself, the long lengths that may be
used, the ability of the piles to cut through obstacles and penetrate into
the underlying materials, and their strength as beams. When certain of
these qualities are necessary, steel piles may
be desirable regardless of
cost. In Fig. 9-17 are shown some details that may be applicable for
work involving steel H piles. These are discussed in the following
paragraphs
1. Availability. Steel piles are procurable in many places; in others,
they are hard to get; in times of steel shortage, it may be almost impossi-
ble to secure them. This is a matter to consider when a structure is in
the planning stage.
2. Shipment. Long heavy steel piles may be shipped by water or rail,
although this is likely to be expensive. Hauling them by trucks is likely
to be both difficult and costly. Short pieces may be welded together in
the field to make long members if necessary, but this is seldom done by
preference.
3. Handling. Steel piles are so strong as beams that they can be
handled safely if equipment with sufficient load-lifting capacity and
reach is available.
4. Driving. These piles are so strong as columns that they will gen-
erally withstand any driving forces and impacts that can be applied to
them. The cuts A of the flanges shown in Fig. 9-17 (a) accentuate the
cutting properties of the end of an H pile, but these may not be necessary.
They tend to prevent bearing of the tip of a flange on a piece of rock that
would cause eccentricity of bearing, and they enable the piles to cut into
disintegrated and weak rock more easily. However, the pile may not
bear on all of the area at its tip. These plain H piles do not displace
much of the soft soil into which they are driven sometimes a great —
advantage. In denser materials, the soil may pack so hard between
the flanges that the piles in effect become equivalent to square ones in
their resistance to driving and their displacement of the soil. Steel piles
can be used as batter piles as well as vertical ones. Of course, the driving
and handling equipment must be suitable.
It is sometimes assumed that steel H piles can be driven through prac-
Art. 9-8] STEEL PILES 339

tically anything except rock. This is not necessarily a correct assump-

tion, even though such piles have been reported to have penetrated 2 or
3 ft. into disintegrated rock, shale, and soft limestone. In one case, to
the author's knowledge, 12-in. H piles were driven approximately 75 ft.
into sand containing some small gravel. When a mixture of sand and
gravel not exceeding 2 or 3 in. in diameter is well compacted, one can
expect to encounter such hard driving that it may be difficult to penetrate
such a soil for more than 40 to 50 ft. When the soil is compacted boul-
dery gravel, one may not be able to drive the piles through it at all, or
perhaps to penetrate it more than 25 to 30 ft.
In one case a deep fill at a bridge abutment was made of a top layer of
35 ft. of compacted river-valley sand and gravel containing a considerable
number of boulders from 6 to 18 in. in diameter, together with enough
silt to make the fill compact firmly. Under this layer was approximately
a 20-ft. stratum of sandy fill with some silt and a few stones. This latter
layer was placed over about 10 ft. of the same kind of soil originally in
place. Finally, under all these was a natural stratum of 10 to 15 ft. of
bouldery sand and gravel overlying bedrock. To prevent possible danger
from scour, it was decided to drive light 12-in. H piles through all this fill
to bedrock. The contractor tried to drive the piles as ordered, with
unreinforced square ends. Practically all the first 15 attempted hung up
after penetrating 30 to 35 ft. It was then decided to remove the top
35 ft. of bouldery fill and to drive the piles from that level, splicing on
extensions and later filling around the piles. When the previously driven
piles were dug out with the top stratum of fill, most of the bottom ends
were unbelievably battered and torn. Even though some of the piles
reached the layer of finer material, they had been stopped by friction
along the shaft and by bearing on the damaged ends. In the later driv-
ing from the lower level it was found to be impossible to drive all the piles
to bedrock through the bouldery layer overlying it. 1

Once the boulders cause curving of the flanges of an H pile, the trouble
rapidly gets worse, until the pile assumes various queer shapes, wdiich
may vary from knuckled-over to peeled flanges resembling wide coiled
springs. Reinforcing the tips with 3 ^-in. plates welded on the outside
of the flanges and on both sides of the web may be very effective in pre-
venting the start of serious deformation of the tips of the piles.
5. Adjustability of length. Steel piles that prove to be too long can be
burned off readily at the proper elevation, even under water, although this
1
The author does not recommend the use of bare steel H piles in a case such as the
embankment previously described because of the possibility of long-term corrosion.
Rain water will repeatedly drain down through the fill around the piles, thus causing
a renewal of oxygen and moisture, with air present in the voids of the fill. After
40 or 50 years of service, the piles may become badly corroded.
340 PILES [Chap. 9

may be wasteful. Those that are too short may have scarfed extensions
butt-welded to them, or field-welded splices may be used to join the parts.
These junctions may be very strong.
6. Strength. As indicated previously, the strength of a heavy steel H is
tremendous, provided it cannot buckle. The load-carrying ability may
be reduced, however, if the load cannot be delivered to the pile properly or

*1
hdt=
II

|l

Heavy |l

welded
~i — I plates
II

II

MMii

ii

A'
Elevortion A-A B^- 1

Section B-B

faMVELDED PLATE LUGS ^-LAGGING

Boiled, riveted
,D (orwelded

^-^ ^I P°°°I1F
0\> .

n l°°°l K;
ls
^dn^

(c)- PLATE CAP (W-ANGLE LUGS feJ-WELDED ROD LUGS

AND LUGS

FIG. 9-17. Some details that may be used on steel H piles.

if the necessary bearing resistance cannot be developed. In this connec-

tion, notice the following points illustrated in the sketches in Fig. 9-17:
(a) The welded plates B, or a series of them, may be used to increase
the effectiveness of the bearing if the pile is driven through soft soil into
sand or gravel. This is intended to provide more resistance than that of
friction along the surface of steel only. In the case of gravelly soils,
plates B may be very effective if they are thick and the welds are strong
enough to withstand the driving. The upper plates are of questionable
value unless the lower ones fail. Of course, such plates make the driving-
more difficult.
(b) The hardwood blocks C constitute lagging that is also used to
Art. 9-8] STEEL PILES 341

increase the bearing resistance of the pile. This is no stronger than its
bolted connection to the steel.

(c) This shows a cap plate D

that is welded to a pile to increase the
bearing area of the concrete of the footing on top of the steel. The
welded plates E are for the same purpose. If the embedment di is 5 or
6 ft., the bond of the concrete may be of real value, but it is not likely to
be adequate or absolutely trustworthy for loads of such large magnitude.
Grease, dirt, or paint may weaken the bond. The bearing of the concrete
directly upon the end of the H is limited by the crushing strength of the
concrete.
It may
not always be practicable to use details like those shown in
Fig. 9-17 at the ends of steel H
piles, especially if they are driven to
irregular rock so that their tops must be burned off in the field, or if they
are driven to varying and unpredictable elevations under water. Tests
made by the Ohio Department of Highways indicate that if the thick-
1

ness of the concrete footing above the ends of the piles is 2 or 3 ft., there
is great resistance of the concrete against penetration of the end of the
plain pile without any cap plates. Of course, the footing should be mas-
sive and reinforced to prevent splitting and cracking. The bearing on
the end of the pile is most useful; the bond along its side may be helpful
but, as stated previously, it should not be trusted unless the embedment
is several feet. On must be sure that the intended
the other hand, one
load can be delivered to the piles without any uncertainty.
(d) Here angles F and G may be riveted or welded to both the web and

the flanges of the pile. The use of very wide outstanding legs on these
angles is ineffective because they have little shearing resistance except
near the connected leg. It is better to use a series of narrow but thick
angles closely spaced.
This sketch shows a series of welded pieces of reinforcing bars that
(e)

serve as lugs. The amount of welding is probably large compared to the

bearing area obtained. The concrete footing in any case must be strong
enough in bearing and in shear to transmit the required load to the pile.
7. A steel pile should not be called resilient; it is strong and
Resiliency.
elastic. Large lateral deflections may cause overstressing and permanent
deformation of the steel. Nevertheless, such a pile will not crack and
shatter. Even though bent or somewhat kinked, it may still support a
large load.
8. Durability. The question
of the durability of an unencased steel pile
frequently causes argument. Here are some comments upon the subject
a. Steel piles completely embedded in dense clay will probably be per-

manent. Rusting requires the continued presence of a supply of oxygen.

1
Research Report No. 1, Ohio Department of Highways, Dec. 1, 1947.
342 PILES [Chap. 9

When the soil is nearly impermeable, the water will not circulate through
it to As soon as the oxygen in the available water
any appreciable extent.
isused up, corrosion will be stopped.
b. In porous soils, the ground water jnay be moving slowly so that the

supply of oxygen is replenished. Some rusting may then occur.

c. Soils and moving water containing acids or other injurious chemicals

may cause serious corrosion. Bituminous and plastic coatings may be

helpful if they can be applied and are not damaged.

d. Sea water is likely to be very injurious.

almost impossible to
It is
predict accurately the service life of steel piles and sheet
piling exposed
to sea water. In the harbor at New Haven, Conn., a group of unencased
steel piles supporting a T-dock corroded badly in about 10 years. In 1

one location sheet piling about 25 years old was in fairly good condition,
whereas similar piling 30 years old in another location was corroded
entirely away in spots.
e. Fresh water and air may cause moderate rusting.
/. Electrolysis may cause serious trouble if it can occur.
When bare steel H piles are to be used, the possibility of electrolysis
should be considered. Stray ground currents moving horizontally (or
even vertically through the piles from one soil stratum to a different one)
now or in the future may cause serious loss of material. It is good policy
in all cases to isolate the steel piles from each other (above and below ground
and in concrete footings) and from all reinforcement in the pile caps by
providing 2 or 3 in. of concrete between the piles and any other metal.
It may even be desirable to insulate the anchor bolts from the steelwork
of the superstructure by fiber or other insulators so that electric currents
cannot pass from the bridge or building into reinforcement and thence
into the piles.
g. The sand-blast effect of water-borne or wind-blown sand may dam-
age exposed steel close to the ground.
h. Encasement with concrete is beneficial if the concrete is dense, of
good quality, and properly reinforced so that cannot break off. When
it

piles are to be encased in concrete, they should be carefully cleaned to

aid in securing bond on the steel. However, when the exposed portions
of corroded piles or sheet piles are to be coated with bitumastic as a
maintenance aid, it is the author's opinion that the coating of rust should
not be cleaned off to bare metal because this aggravates further corrosion.
Only the loose scale should be removed by wire brushing or similar means;
then the bituminous coating should be brushed or sprayed on. 2
In general, steel piles have a very important place in engineering con-
struction. When one needs very long strong end-bearing piles, this type
1
Cathodic protection was later installed at a great expense.
2
This procedure does not apply for ordinary painting.
Art. 9-8] STEEL PILES 343

may be the best that can be found. A few such piles supporting 50 tons
or more apiece may be far more desirable than a large number of weaker
piles,provided the steel will be durable.
Consider further the problem of corrosion of steel H piles. It is likely
that such materials as organic silt, peat, and "meadow mat" (partially
decayed vegetable matter in swampy areas) will be slightly acid because
of the decomposition of carbonaceous materials. This may cause serious
corrosion of bare steel piles in contact with such soils. If excavation
would not be excessively expensive, such harmful materials should be
removed and replaced with compacted clean sand before driving the
piles. Cinder fill, coal, slag fill, organic matter, and "junk" may also
cause serious corrosion because of acids leached from them. These
materials should also, if possible, be replaced by clean sand fill, or the
piles should be well protected. Wood should not be left lying in contact
with the steel.
Bare steel piles exposed to sea water, especially in the tidal range,
should be considered to be temporary. In New England, a concrete dock
supported on steel H piles was examined 11 years after construction. It
was found that the average loss of area of piles in the tidal range was
approximately 10 per cent. Seven years later the piles were examined
1

again. The engineers were shocked to find that the average loss of pile
area was then approximately 45 per cent. In other words, the rate of
corrosion apparently accelerated as time passed. In few of the
fact, a
flanges were "eaten away" so badly as to be Further
frightening.
investigation showed that the losses caused by corrosion 3 ft. below low-
tide level were considerably less than those in the tidal range and that
the loss of section was not yet serious at the mud line 10 to 15 ft. farther
down. However, extensive repairs had to be made to save the structure.
Encasement with dense concrete of at least 5,000-lb. strength should
extend from above the splash line to 2 or 3 ft. below the top of the sand,
gravel, or clay supporting the piles. In the banks close to shore, the
concrete should be extended down below low-tide level.
As a general guide it can be assumed that a steel surface exposed to
sea water at least 2 or 3 ft. below low-tide level will corrode approximately
0.005 in. in thickness per year (on an average). In and just below the
tidal range this rate of loss of metal may be 0.010 to 0.015 in. per year, or
even much more at local spots. One can readily see that an exposed
H pile will be attacked from both sides of the web and flanges, and it can
therefore be weakened badly in 20 to 25 years. Foul harbor waters may
be even more corrosive than clear sea water. It is therefore obvious
1
Incidentally, the pile caps made of 3,000-lb. concrete were considerably disinte-
grated on the surface, especially at corners. At least 5,000-lb. concrete with a
bitumastic surface coating should have been used for this construction.
344 PILES [Chap. 9

that a steel pipe pile filled with concrete will last much longer than an
H pile having the same thickness of metal because only one surface of the
member is exposed to corrosion. Furthermore, the concrete inside will
carry load even if the pipe becomes ineffective.
The region just above mud line is also likely to be a point where cor-
rosion may be accelerated. In the body of the soil of the bottom under
sea water, the rate of corrosion of an exposed steel surface may be con-
siderably less than in the clear water. However, it seems that a great
deal depends upon the contents of the soil itself — the oxygen content, the
alkalies and salts (especially sulphates) present, the organic or other acids
available, and the porosity of the ground.
Another feature which accelerates corrosion is a rise in temperature
of the water. For example, a rise of 50°F. will probably double the rate
of attack.
Mechanical abrasion due to wave action, floating ice, driftwood, and
water-borne sand may remove rust from a bare steel surface, thereby
accelerating loss because of the exposure of the metal.
It is probable that Monel metal and certain bronze or other nonferrous
metals are the best for use in the tidal range not stainless steel or —
wrought iron. Copper-bearing steel seems to be helpful when piles are
exposed to air, but not to be of much value when they are submerged in
sea water or when they are in the tidal range. However, the engineer
should consult a specialist when confronted with problems of this nature.
If one plans to install cathodic protection of exposed steel, he should
consult someone who is experienced in such matters. A few comments
regarding this protection are the following:
1. Piles in the ground will probably be protected satisfactorily.
2. Piles in water below low-tide level will probably be protected well.
3. Piles in the tidal range will not be properly protected.
4. Piles to be protected must be positively and adequately connected
to the electrical system.
5. Other steel members in the ground or water which are not connected
in the circuits may suffer serious loss, even when some distance away.
6. The current needed is usually proportional to the exposed surface
area.
7. Ice, driftwood, etc., may rupture the connections to the piling.
If so, the disconnected sections may suffer badly because they may be
acting as anodes.
9-9. Sheet piles. Some types of steel sheet piling are pictured in Fig.
9-18. As indicated by their name, these are designed for the special pur-
pose of forming a wall or sheet which, when supported properly, will resist
lateral pressures. They might be of some value in supporting vertical
loads, but this is not an efficient use of the material in them. The joints
 CO -<t K to so K o N lO K
co N. -<* o 00 o *oo- o
CN CO CO
•*-

o u-> in CO lO o 00 CO O o <n
>o o -<* O0 N lO ND o CO n »o
o K -t -* o oo T -<f oo n -*
' CN CN i R
a --

o
co
->*
00
o CO o
^t o c:
K o
IV o CN
CN o o
>o
K
CN CO

IT) O o •o o o o _ O >o o
© 00 o r- _ (V CO O o >o ->*
— >
co CO -<* lO u-oo lO lO CO in o Z 0)
' CN '

CO •o ^* •o-t o o K
>o •o -* o K o o o o- o CO

v -c -z
o n
o o o o O o CO m o o
->* L0 0- c > =
o o -* CO _ _ oo lO o
o o CO c n
CO o -t
E
' " co
-<* -<J- K " CO
-"*
CN
o a 0)

<u k
a
F a o
CO oo CO
< > %
00 u> 4)
_c -C fc
CN o E

o o CN
> -
CN
o K o
c-
CO
• K
c i: —
* .*? rf£
<
t o
U"> CN ->* 00
co CO N3 CN -<*•
-t
CN CO CN E * rS £
^ U £
^3 ^, » J2 *
o o >o o o CO o CN o oo
o CM u-> o o~ lO CN K o •o
K CN
CO CN
-t ~o CN CO K CN o <o

5 Z i
o N
K K o N o o
u- co ro o- "*
-* CO oc o o oo o "* S3
-D £

<oo >o >o u- o o u-> l/-> >n lO

o _
"*
o CO N CO
K -t
-<r -T 00
K
•o O 2 = " O
" co -<* -«r o> ->*
" M- E CN
-
D •*
01 ~
-D 4> ^
K
o o
«o CN in o o
K >o -<* o a CO >o o o CO
co
> — f — £
Q o -§ s
-"=

o T> o >o c o ID o o o «1 o- »> t; £ 4)

h- -° M- "
-to _ o CN _ •o _ .TO ,_ O E ~°
O
" K t *
rn
lO o co -t ^t K u-> -* « o «, c
' CN
u Q. — o) t;

1t)
<8 f 2 ?
B .

^ T3 "8 T3 "e *-. CB CB CD riS ^ o fc

*
o o wo o
4)
T3 **-, •^ 5
co
o >o o o
o o >o c
ir i
o
co K <* !/•> CO
o
•o 6
4)
c Z) o.
.E
10 Z
CN O)
c

351
352 PILES [Chap. 9

reduction of safe load that should be applied to the results of the tests
because of the group effect of the large number of piles to be used near
together in the actual substructure.
3. The test loads are applied for such a short time that no data are

given regarding the slow consolidation and flow of the soil that may occur
because of the future long-term loading of the area.
4. The settlement of the test piles to be used as that constituting prac-

tical failure is difficult to establish without being chosen arbitrarily. Per-

haps the applicable building code specifies it, but that does not mean that
it is correct for all cases.
5. It may be necessary to set some average load as that to be considered
to cause this critical settlement if the data obtained from the tests differ.
6. A safety factor should be applied to the ultimate bearing capacity of
the piles as determined by the tests. Perhaps this should be 2, perhaps
something larger, or something smaller. Excellent engineering judgment
is needed in making this decision.

7. It is probable that the more nearly the piles act in end bearing, the

more reliable the results of the tests will be, unless the tips of the piles are
underlain by plastic or other compressible materials, or soils that may
flow.
8. Pull-out tests to determine the average magnitude of the skin fric-
tion may be very helpful and reliable provided the same size of pile and
depth of penetration are used and provided the soils do not vary across
the site.

To be made in advance of the planning of the

preferred are pile tests
structure. Then, besides information indicating load-bearing capacity,
these tests may yield valuable data regarding the best type of pile for a
particular site, the length of pile needed, the pile-driving equipment
required, and whether jetting is necessary. All these will probably
enable the designers to develop a safe and economical design for the
structure. However, it is sometimes difficult to convince an inexperi-
enced owner that the money spent for such tests is a good investment.
Too often he believes that the cost of getting equipment to the site, driv-
ing piles that may not be a part of the permanent structure, and making
the tests is wasted and unnecessary. He seldom pays attention to the
additional expenditures that he does not see —
those covered in the con-
struction costs of a design that is and should be made very conservatively
because of the lack of proper information. Furthermore, he will complain
vehemently if the engineer guesses wrong and designs a structure that is
unsatisfactory.
When jobs are small and when pile-driving equipment cannot be
obtained in advance (which occurs all too often), it may be necessary to use
the data obtained from borings and soil tests for the determination of the
Art. 9-11] LOAD TESTS 353

and bearing value of the piles used for design purposes.

type, length,
Then arrangements should be made in the contract on a unit-price or —
force-account basis — for the testing of a few of the first piles driven.
This will enable the engineer to obtain data on which he can base the
instructions for driving and penetration, and he can check
his assumptions
in time to revise the design he finds that he has been too optimistic or
if

too pessimistic in his ideas about the loads that the piles may support
safely.
Figure 9-20 shows one method of making a load-bearing test of a pile.

A few points to be considered are the following

1. The loading platform should be strongly built.

H Metal billets, Rod

Rod or pipe I

for check ^caisson weights,

levels
sandbags, or
other materials
to apply load

M {Platform -*-

~Y\^Ho/d- down
1R
rod
% w°m
y/AiWV/Wcw: iof—Ji
*riesi
Test- hold-down
Hold-down
pile DO pile
V Hold-down
pile piles

piles
(ah END VIEW (61- SIDE VIEW

FIG. 9-20. One arrangement for a loading platform for making static load test of a pile. (Adjust-

able hold-down rods to anchor piles prevent overturning. Rods must be slack and balance main-
tained when settlement is measured.)

2. The pile should be steadied laterally by bracing that will not partici-
pate in the resistance to vertical loading.
3. Settlement readings should be made by taking levels upon some

point or points of the platform which are continuously accessible and

which yield data giving the actual settlement deformation
of the pile, not
of the platform, tilting of the pile, or movement of the soil. A proper
bench mark should be used.
4.The load should be applied so that its magnitude may be determined
accurately. Weighed metallic or concrete blocks, bags of sand, and mem-
bers of structural steel may be suitable. Water in a tank may serve the
purpose.
5. The load should be applied in predetermined, and preferably uni-
form, increments. For piles that may be expected to support safely
only 15 or 20 tons, the increment may be 5 tons; for much stronger piles,
 00'0=MHW Mopq$A3|3 ||y

-1 s tP?| ill s| ^e^l

— 1 1 U 1 1 l 1 1 M M M 1 1 1 |-iy? '.-I-..
.:..:. : :v| . \ I -< !

louiJoq ^ o q j d 3 j
o- a:

00'0=MHlA|MO|aq -s,A3| 3 [|V lZ 2

354
Art. 9-11] LOAD TESTS 355

it may be 10 tons. increment may shorten the duration

The use of a large
of a test, but when approached and when the last large
the critical load is

increment produces a settlement that is called failure, one may not be

sure that the pile was not previously almost at the point of yielding, so
that only 2 or 3 tons were necessary to constitute the "last straw." The

n 4 hr duration of load
L--,
V/5 hr
>

^4hr
T
f-4 hr
N
:

hr
i<
N
X
N
50

^4 hr
o
o
o r

s
:

100

-^ "-£ Retwund 50hr

•^m

'v.^
*»

""•«,

^~> :

150
5P 100
Load on pile in tons

FIG. 9-22. Load-settlement record of a pile tested by VVestcott and Mapes, New Haven, Conn.
(Length 36 ft. 7 in. Ten blov/s per foot for last foot. Pile driven through silt into thin gravel stratum

above fine sand. Raymond step-taper pile with larger section welded to end to increase end diam-
eter to 1 1 in. Net settlement 0.1 15 ft.)

application of a safety factor of 1.5 or 2 to this final test load may yield
a result that on the side of danger.
is

6. Each increment of loading should remain on the test pile until settle-

ment for a period of 1 or 2 days is negligible, except when the load proves
to be the critical one, and serious, perhaps continued, settlement occurs.
7. Settlement readings should be taken before and after the application

of any load increment, and probably at 6-hr. intervals between loadings.

 356 PILES [Chap. 9

Steel beam

Pipe to
transmit load
-2 rods
CT Hydraulic jack
Casing
Reinf..
Welded Anchor pile

- Anchor pile ^Test pile

WW' VKWW/>W*V//WW-Vm Concrete
Nil button
Nil
llll

UliJ
^-WESTERN FOUNDATION CORPORATIONS METHOD
(a)-OUE ARRANGEMENT FOR LOAD TEST OF PILE BY JACKING OF TESTING BEARING VALUE OF BUTTON-BOTTOM PILES
(By putting rods at center and jacks at ends, this can be
(Casing used for driving eliminates friction so that
used as a pulling test to measure friction and cohesion)
real bearing value of button is measured

* f Double beam
.Loading platform Loaded truck .
v , .Timber deck
II I II I I I I I
pdednver.etc. , > , f< , , ]> rT f , , , ri , ,

Sieel beams Wedges

JIW/WW.'W'- '.AW.V/'n'/WW.'W'W/*
rPiles 'Jack
rTestpile
'/l*W/v///<y»vm-yM b/lVIPW/WinVIHW'i
rrrvn

fct- CANTILEVER LOADINGEQUIPMENT TO REDUCE

WEIGHTS TO BE HANDLED WHEN MAKING TEST
(dJ-JACKING TEST AGAINST HEAVY MOVABLE LOAD

FIG. 9-23. Miscellaneous loading devices for testing the bearing value of piles.

MODIFIED ENGINEERING NEWS FORMULA

2Wh For Raymond single -action No. 1 steam hammer

R=
s+a,£ W=S,0O0lb., h=3ft
For step taper pile,
P - 8,500 lb. for mandrel
and 10001b. for shell

Note:
Enter final
penetration
in inches, also
length ofeach
pile, on
record sheet

0.33" 0.28" 0.25" 0.22" 0.20" 0.18" O.I67"O.I52"O.I42"

Penetration in inches per blow

I i l i

3 3Vz 4 4'/z 5 l\ 6 6'fe 7

Number of blows per inch

FIG. 9-24. Example of a reference chart for field use during pile-driving operations when a particular

formula is used as a guide.

Art. 9-11] LOAD TESTS 357

The results should be watched closely and checked carefully, especially

when inconsistencies seem to occur.
8. important also to obtain settlement or rebound readings and
It is

the corresponding loading, as the test load is removed. The final reading
after the load is off will enable one to determine the permanent settlement
of the pile.
9. Those who are making the test should observe and record all other
data that may have any probable usefulness when the results of the tests
are studied. Among these are such things as the number of blows per
foot of penetration for the complete driving.
Naturally, such a load test as that previously described takes time.
When necessary, some useful data may be obtained from a more rapid
test. However, a rapid load test is likely to yield reliable information in
cases where the piles reach granular soils and where they are largely end-
bearing ones. These tests may not provide good information on the
safe loads for friction piles in plastic cohesive soils. It is obvious that
group effect and probable consolidation of soil are revealed even less than
they are by the longer tests. Nevertheless, the latter are so poor in this
respect that the rapid tests may be as accurate as other uncertainties
justify.
One method of making a rapid load test of a pile is indicated in Fig.
9-23. This assumes that a row of piles has been driven for the permanent
foundation. A heavy beam is attached to one or two of the piles on each
side of the one to be tested. A powerful hydraulic jack is then placed
between the pile top and the beam. Pressure is applied to the pile by
increased amounts at short intervals. By means of the gage pressure
and level readings taken on the top of the pile, the load-settlement curve
may be plotted. Certain questions should be raised concerning this
procedure:
1. and shearing resistance along the
It tests primarily the frictional
pile's surface and in the soil between the tested pile and the piles that
anchor the beam. No reliable measure of soil consolidation is found
because no extra load is added to the area.
2. When the piles are primarily end bearing, there may not be enough

resistance to uplift to prevent pulling of the anchor piles.

The anchor piles should not be too close to the test pile, perhaps 5 or
3.

6 from it.
ft.

4. There may be difficulty in attaching the beam to the anchor piles

unless the former can be bolted or welded to steel H's, shells, or anchored
reinforcement.
When piles are to resist uplift, a pulling test may be made by attaching
abeam to the top of the test pile, then using jacks under the ends of the
beam to apply upward pressure. Here again one should be careful how
358 PILES [Chap. 9

he extrapolates the result of a test on one pile to a large group because the
latter can be no better than the weight of the soil that anchors them.
Another variation of making a load test is the building of a platform on
which a heavily loaded truck or the pile driver itself may be placed. A

'Wf!\w.vj>--vf>.\v//.\\//f.*vrrv>w>f.\v>ij,\j}>\w>3yr'i-

Sandy fill

> ;

^
Soft clay

<3
i k\ Drilled Silfy clay
Rock~% v. I" socket
~ I' -

I Dense c/o>y
T
Plate Concrete
r
t. Open
excavated
end,

FIG. 9-25. Drilled-in pile or caisson for FIG. 9-26. Three ways of making sand-pile
very heavy loads. [Affer data of drains by driving steel pipes, filling with sand

Western Foundation Corp.) or gravel, and withdrawing pipes.

fof-PIPE CASING DRIVEN (b)- WOODEN PILE (c)- COMPLETED SPLICE,'

WITH CORE INSIDE; DRIVEN INSIDE CASING REMOVED
CORE REMOVED CASING

FIG. 9-27. Construction and details of a splice in a composite pile.

jack may then be placed on top of the pile so that it can react against the
load on the platform. One must be sure, however, that he can obtain
accurate readings of the settlement of the test pile.
In general, load tests of piles are so small compared to the forces applied
to a structure that they aid little in estimating the settlement to be
Art. 9-11 LOAD TESTS 359

expected from the consolidation of plastic soils around and below a large
group of piles. Nevertheless, if the test pile does not support the desired
load safely, it is a reasonably sure bet that the average bearing value of
the group of piles will be even less, and perhaps very much so.
The effect of vibrations upon the safe bearing value and settlement of
piles isanother source of uncertainty when one attempts to interpret the
results of static load tests. Probably this feature is negligible in the case
of end-bearing piles but not for friction piles. Tests made for the General
Electric Company at Schenectady, N.Y., indicate that vibrations 1

increased the settlement of heavily loaded friction piles that were driven
into the following strata: approximately 30 ft. of old fill; a soft organic
sandy clay; and rather uniformly graded sandy silt and silty clay,
together with some compressible fresh-water clay. The increased settle-
ment caused by vibrations under working loads is probably considerably

Reinf. cage of mesh and

rods or spiral and rods welded I

\4"min.

(a)- ATTACHMENT OF f6J-ENCASEMENT OF (c)- PLATE TO PROTECT raj- RECTANGULAR CASING USED
CONCRETE PILE EXPOSED PORTION WOODEN PILE BY WESTERN CONCRETE
TO FOOTING OF STEEL H-PILE AGAINST BROOMING PILE CO. TO BE DRIVEN PART
DURING DRIVING WAY TO ENABLE EXTRA LONG
PILE TO BE LOWERED INSIDE,
DRIVEN.AND CASING REMOVED

FIG. 9-28. Miscellaneous details.

less than that of test piles loaded to capacity. However, the group effect
in the compaction of granular materials supporting many piles that are
all affected by vibrations simultaneously may be serious. This problem
should be considered when one is planning pile foundations for generators,
heavy machinery that is subject to shocks, railroad trestles and bridges
when the live loads are relatively large, and very heavy cranes when
settlements caused by them may be harmful to the structure in general
or to costly equipment nearby.
With rapid tests as well as with slow ones, somebody must determine
what constitutes failure of the piles, what is to be taken as the critical
load, and what the safety factor should be for design purposes. When
one considers the empirical nature of these things as well as their great
importance, he will begin to realize the character of some of the decisions
that must be made in the practice of foundation engineering, and the
responsibility that someone must assume.
1
C. F. Dodge and W. F. Swiger, Vibration Testing of Friction Piles, Engineering
News-Record, May 13, 1948.
 10
PILE FOUNDATIONS

10-1. Introduction. This chapter contains discussions and illustra-

tions of general principles and procedures that are likely to apply to pile
foundations. There are many other considerations and details that are
dependent largely upon the special conditions at a given site and the
requirements of any particular structure. Many of these will be included
in subsequent chapters.
It is impossible in a single chapter to tell all that should be known about
the planning and designing of pile foundations. Different engineers use
different methods, as is to be expected. The illustrations, however,
should give the reader some help in inventing a plan of attack of his own
upon his special problems.
10-2. Choice of type of pile. In the planning of a pile foundation,
one of the first decisions concerns the type of pile to be used. Several
types may be practicable, but the problem is the selection of the best one.
Then this type is generally used in the design and specified in the contract
unless alternative proposals are asked for in the hope of securing lower
bid prices.
Each problem is a specific one — a given structure, with given loads and
dimensions, at a particular site, to be supported upon the soils that
exist at that place. Knowing these things, how can one determine the
best type of pile to use?
The advantages and disadvantages of various types may be judged to a
certain extent from the data given in the preceding chapter. At least, it
may be possible to eliminate a few that are obviously less suited to the
conditions than are others. A list of things that may influence the deci-
sion is the following:

1. Length required
2. Bearing value desired per pile
3. Accessibility of site
4. Means for handling piles
5. Materials on hand or easily available
360
Art. 10-2] CHOICE OF TYPE OF PILE 361

6. Elevation of watertable with respect to pile caps

7. Presence of marine borers
8. Comparative costs
9. Degree of permanence required
10. Experience with construction used in other similar structures
11. Ease of installation
12. Speed with which piles can be secured
13. Speed of installation
14. Piles best suited to methods and conditions of driving
15. Possibility of damage to piles after driving
16. Adaptability to varying lengths if found necessary
17. Ease of cutting off or otherwise securing proper elevations of tops
18. Ease of increasing number of piles if necessary
19. Personal preference of the engineer or owner because of past experi-
ence with a particular type
20. Availability of a contractor and his equipment able to install a
certain type
21. Desire to utilize services and products of a particular contractor
for business reasons and because of favorable experiences with his work
in the past
22. Driving above or below water level
23. Flexibility or resistance to bending required
24. Value of a pile as a column if end bearing
25. Area of tip of pile if end bearing
26. Resistance of pile to hard driving when this is necessary
27. Piles previously used if new work is an extension of an old structure
28. Minimum disturbance of soil if this is a factor
29. Ease of driving at an angle when batter piles are needed
30. Presence of acids or other materials in the soil that would injure
certain types

If the type of pile is left to be settled after the receipt of competitive

proposals, there may be many and heated arguments in arriving at a deci-
sion. Each contractor or manufacturer believes in his own product, and
each may
guarantee that his piles will support the specified loads safely.
Judgment based upon price alone may not be advisable. Each type of
pile has special advantages for certain uses. If this were not so, competi-
tion would eliminate many of them. On the other hand, a contractor
engaged in foundation work usually installs various types of pile in the
course of his work, and he can do so readily. It is desirable for the engi-
neer to decide for himself in advance which type is best, obtaining counsel
in this matter when he needs it. After that, the competition is secured
through the desire of various contractors to get the job.
362 PILE FOUNDATIONS [Chap. 10

10-3. Allowable bearing value of a pile. The determination of the

allowable (or presumptive) bearing value of a particular type of pile to be
used under the specific conditions existing at a site is a very serious but
uncertain matter. Stating a figure fs one thing; making this figure the
correct one is something else. This may as well be admitted.
An engineer may tell his men to use 65-ft. thin-shell poured-in-place
concrete piles for a particular structure, and to allow 30 tons per pile.

The men then go ahead wholeheartedly with their design work

will
because of their confidence in him. Later, a chosen pile-driving formula
in which are applied the data obtained during driving of piles at the site
may yield a computed safe load of 35 tons each for these piles. Then
everyone is stillhappy because of confidence in the formula.
The selection of the allowable load on a pile in cohesive soils is probably
one of the most difficult decisions that a foundation engineer has to make.
This is because there are so many uncertainties involved. What should
the penetration be? What is the magnitude of the skin friction? What
will be the group effect? What will be the settlement under various con-
ditions? What load can the underlying soils withstand safely? How
can the greatest economy be secured consistent with safety? In a large
job, too much conservatism can cost a lot of money; too optimistic a
decision can result in a tremendous amount of dangerous or unsatisfactory
construction.
The presumptive bearing value may be based upon the results of load
tests of piles driven in advance at the This is a wise procedure
site.

and, in many cases, it may result in considerable economies when piles

can be driven to firm materials and can be proved to have large load-
carrying capacity. Such tests are far better than mere estimates. At
the expense of reiteration, the following warning is given:
Load tests of piles that are wholly or largely end-bearing piles supported
upon firm soils will generally yield reliable results; those made on friction
piles in and underlain by cohesive or weak soils will generally yield results
that are considerably (or even radically) larger in value than the actual
average safe bearing value per pile when used in a large group.
Table 10-1 gives a general scale of values that may be useful in deter-
mining the safe bearing value of piles and in comparing types. The
data should be used for preliminary designs only. Tests should be made
where such action is practicable. In small jobs, where small groups of
piles are used in widely scattered locations or where soil conditions vary
considerably, it may be unduly costly to make proper tests. Conserva-
tive bearing values should then be used in design, and these should be
compared with such additional data as can be secured during the driving.
 o
CO
u->
CO * o
o o o
lO T> S3
in
CN
o
CO
u~i
CO
o
* o o
T -<*

O
CO
"O
CO
ir>
CO
o
^ O O"^" "^"
I

u->
CN
oI

o
CO CO
I I

ko
CO
I

io
CO
I

>o
CO
o
o O o o o""> o o
E - n io io o
-<t

tO -D IO o "> o o o
IN O O * « t

o >o o o o o
co co ->*

c o
UOO
m o n"O o o o
* t * CN CO CO
io jjj

-O -2
O "O o o
o io uo CN CN CO CO
I

IO
.—
OI

CN
I

<T>
CM
I

UO
CN

w- o o >o o o
o o
u->
I/)

0) a>
3 D
o vi

> >o o O o
D)
C < IO
CN
O O "O
CO -t "*

D
a> m
CD 'n o O o
CO CO ^t . . .
co IO
CN
o O
CO CO
a> CN
«-t-
o o *" Y o
D a CN CO
u->
CN CN CN
to c
Da)

D 4) CN CO CO ._
O "O IO
CN CN CN
IO
CO
E ^ o o o
I I I I

>o
CN CN CN •— CN CN CN

CO
<

*- w
>*
o o E
C TJ TJ » a u
E TJ It <D

ll. O co X co oo u CO 3 CO lC

<U .K

363
364 PILE FOUNDATIONS [Chap. 10

An interesting and instructive case showing an apparent change in the

bearing value of piles was reported to the author by A. G. Beaulieu of
Hamden, Conn. The story is approximately as follows:
An extension of a power plant along the Eastern seaboard was started
before the war. The piles were driven, tested, and capped with concrete,
then the work was stopped. They were Monotube piles driven down 40
or 45 ft. into sand. They were driven to practical refusal. After the
war, whenthe extension was reauthorized, the equipment was quite
different from that previously intended. As so often happens in such
instances, the existing piles were now in the wrong places and they were
not sufficiently numerous. The caps were to be knocked off some of them
so that they could be incorporated in the new construction. Others piles
were to be abandoned, and many new ones were to be driven.
The first few of the new piles were driven next to the existing structure.
To everyone's surprise the new piles went down 8 to 12 ft. deeper than the
old ones. Mr. Beaulieu ordered that one of the old piles be tested under
load by jacking against girders fastened temporarily to four adjacent
piles. This test pile settled considerably and easily, indicating that it

should not be loaded as planned. The next step was to try to drive this
pile after welding on an extension. It was driven down nearly 10 ft.
farther. Redriving caused most of the other old piles to go down from a
few to 13 ft. Load tests thereafter indicated that the redriven piles were
safe for use as intended.
In trying to find the reasons for Mr. Beaulieu discovered that,
this,
when company was experi-
the original piles were being driven, the power
menting with pumping its water supply from some deep wells that had
been sunk on the property. It seems that the previous piles were driven
when the watertable was lowered temporarily beneath them. The sand
was then merely damp, and it was stiff because of this moisture. Of
course, the piles were driven as far as they could be at the time. When
the later piles were put in, the abandonment of the wells had allowed the
water to return to within a few feet of the surface. The saturated sand
then did not offer the same resistance that the moist sand had, and the
piles could be driven much deeper. This seems to be a reasonable
explanation of the phenomenon. The implications and importance of
this may be very important in connection with future work where well-
points are used to lower the water level temporarily, and where the water-
table is subject to considerable variation.
It is unsatisfactory to leave such an important matter as the bearing
value of an individual pile with only vague generalities as a guide. Yet
it is difficult to do otherwise. Past experience is one of the best guides
available in the making of a decision regarding the allowable load on an
individual pile. Much should be done to assemble data in the future and
to get them in such form that the depths, strata, and properties of soils
Art. 10-4] SPACING OF PILES 365

can be compared with those of the site at which the proposed structure is
to be built. Nevertheless, in Art. 10-17, the author hazards the presen-
tation of some procedures and practical cases that may help to illustrate
how to select the type of pile and the safe load for each.
10-4. Spacing of piles. When one plans the location of, or the pat-
tern for, a number of piles in a group, there are several things that should
be considered. Among them are the following:
1. End-bearing piles may be placed as close together as it is feasible to
drive —
them at least as far as the bearing capacity is concerned — if the
stratum upon which they rest can support the loads.
2. Since friction piles are no better than the bearing capacity of the
which they transfer the
soil to loads, closely spaced piles may be inefficient
and uneconomical.

v//w/i

y
1
_J

1 1 1

\ApileA ]A-P/7eB
SIDE ELEVATION SIDE ELEVATION (c)-PLMi OF 3-PILE FOOTING fdJ-PLMi OF 3-PILE FOOTING
AS DESIGNED AS BUILT

(£> 3 1
vg
® -\ffi

fa >- PLAN OF 2-PILE FOOTING W-PLAN OF 2-PILE FOOTING re> A2-PILE FOOTING SHOWING HOW MOST
AS DESIGNED AS BUILT OF LOAD MAY REST ON ONE PILE

FIG. 10-1. Illustration of small footings on piles, showing possible effects of inaccurate driving of piles.

3. It is usually inadvisable to attempt to drive piles closer together than

twice the diameter of the butts, or a minimum of 2 ft. 6 in. A minimum
of 3 ft. to 3 ft. 6 in. is preferable for cast-in-place concrete piles. Under
certain conditions it may be possible to drive steel H-type piles and steel
pipes more closely than stated herein, but this should be investigated
carefully before it is planned.
4. Since it is often difficult to drive piles exactly in the position shown
on the plans, the design should be so made that small variations in loca-
tion are not harmful. For example, assume that a small column load is
to be supported by two piles, as pictured in Fig. 10-1 (a). The column
must be in the position shown. Now assume that the piles when driven
are in the position given in (6). The eccentricity of the column with
respect to the piles is very undesirable. It would have been better to
plan upon a three-pile group in the first place, somewhat as illustrated
in Sketch (c). Then, if the piles were not spotted accurately and if they
were as indicated in (d), the result would not be too serious. Again, if
366 PILE FOUNDATIONS [Chap. 10

the two-pile group were driven as shown in (e), the pile G would have to
support almost the entire load because F could not offer much resistance
without causing a tilting of the cap and probably a harmful bending of
the column.
5. A reasonable spreading of the piles may be desirable to attain more

adequate stability to resist overturning forces.

6. Close spacing of piles may reduce the size of the footings and the
reinforcement needed in them. This may be important when a group of
piles is large.

_ T -- © o o o
-*--© o o o
(9>

Q^OvQ

FIG. 10-2. Some patterns for isolated footings.

7. rows rather than

It is generally desirable to drive piles in "straight"
in staggered pattern when there are a great many
one group or area.
in
However, this may not apply in the case of isolated footings such as shown
in Figs. 10-2 and 10-3 where some sketches show the latter arrangement.
Furthermore, in Fig. 10-4 (a) it would be sufficient to use a single row of
piles under the center of a light foundation wall except for possible tilting
when the piles are not in the right location; a staggered pattern such as
(6), or the scheme shown in (c), would be more stable. When there is a
lateral overturning tendency, the wall might have the piles arranged as in
Sketch (d), but some plan like (c) or (e) might be better.
8. When cast-in-place concrete piles are to be driven, it may be desir-

able to use a sufficiently large spacing to ensure that no harm will be

Art. 10-4] SPACING OF PILES 367

o o o o o o o o o o o
o o
o o o o o o o o o o
o o
o o o o o o o o o o o
faJ-ELEVEN PILES r&l-TWELVE PILES (^-THIRTEEN PILES

o o o o o o o o o o o o
o o o o o o o o o o o o o
o o o o o o o o o o o o o
o o o o o o o
frfJ-FOURTEEN PILES reA FIFTEEN PILES (£}- SIXTEEN PILES

o o o o o o o o
o o o o o o o o o
o o o o o
o o o o o o o o o o
o o o o o
o o o o o o o o o
o o o o o
o o o
^-SEVENTEEN PILES 6<zJ-EI6HTEEN PILES ^-NINETEEN PILES

FIG. 10-3. Grouping arrangements for piles under large isolated footings.

o o o f- o o ,'
*i_£ "
5
(a)
(b)
4' 2-6"

O O
o o
(c)
o &
O^ O
raj
Q—i
4' ,2-6"

o
o o
o o
(f)
(e)

FIG. 10-4. Pile patterns for foundation walls.

368 PILE FOUNDATIONS [Chap. 10

done to adjacent empty shells or to concrete filling that has not cured
adequately.
10-5. Isolated footings piles with vertical loads. Except whenon
a single pile customary to use some sort of reinforced-
is sufficient, it is

concrete footing or cap to spread the concentrated load from a column

over a group of piles. A stepped footing, pictured in Fig. 10-5, is used
as a specific case to illustrate details and procedures. The pedestal
is 4 by 4 ft,

Assume that this design is to be checked for the following conditions:

P = 475 kips; bars a are 9 No. 6, and b are 8 No. 9; the maximum allow-
able superimposed load per pile is 25 tons; the piles are wooden friction
piles approximately 35 ft. long; the dowels are 12 No. 8 bars to match the

Pile

7
j/Ar
""
\4 n7 s
Jo Column reinf.-^.
\\j —
5 '8
•
Ties
Dowels
"B
U !.r io"

r^2 in"
24xZ4"column ^^^^M^lJcolstjt.
Pedestal
d
X
\ q (Footing E — — —X-^FI I
l
H
V
>
V }6 \ J9\\{ )
m*-L,- L^Itz, £Jh&.°~)
/'
6"A. 3 0"
L
3 0"L
3 0" \l-6 "
L N3
.

^ I2n'n'i
|

L
.1,

toJ-PLAN (b)~ SECTION A-

FIG. 10-5. A spread footing on piles with vertical loads.

column reinforcement; and all dimensions are as shown. The steps in

the checking are as follows:
1. Load on piles. The footing weighs 56 kips (approx), including 1 ft.
of earth on top. The load per pile is (475 + 56)/ 12 = 44 kips (approx),
which is conservative but satisfactory.
2. Arrangement of piles. The 3-ft. spacing is reasonable. The diam-
eter of the butt of a pile may be about 14 to 18 in. The edge distance of
18 in. should be adequate, and so is the embedment of 6 in. for the ends
of the piles. The minimum cover of 3 in. over the reinforcement across
the piles is desirable since they are wooden.
3. Pedestal of column. The pedestal appears to be satisfac-
and base
tory. The angle a
approximately 30°, as it should be. The arrange-
is

ment of the dowels appears to be satisfactory because it matches the steel

in the column.

4. Shear in footing. The gross load on each pile is 44 kips. Techni-

cally, the weight of the spread footing GH and the earth on it may be
deducted from the total pile loads as far as the stresses in the footing are
Art. 10-5] ON VERTICAL PILES 369

concerned. As an approximation, assume that the net load per pile is

(475 + 5)/12 = 40 kips. The critical condition will be at E and F.
The estimated average punching shear around the 4-ft. pedestal will be

V 10 X 40,000 ,,
= Wd = = (< no
.
., .
33 ° P L f
VT S °' 2/ Safe)
(4 X 48) X 0.3 X 21
" -

As a maximum, assume that the footing at the edge F of the pedestal must
resist the pile loads that are outside 45° lines from the center of the
column, as shown in Fig. 10-5 (a). Then,

VT = V = 3 X 40,000 = 397 P S
.

L
Wd
- -

48 X 0.3 X 21
Assume that the loads are acting at the centers of piles No. 10, 11, and
12. They are farther than 45° from F (1 ft. 9 in.) . Therefore, the full net
reactions on these piles will be assumed to cause longitudinal shear on the
width mn of Sketch (a). Assuming 45° angles and d = 21 in.,

mn = 2(24 + 21) = 90 in.

The diagonal tension (longitudinal shear) at this section outside of the 45°
line FN in (6) may then be computed as

VL = V = 3 X 40,000 = 7n
70 P - S -
.

L (safe)
, . .

bfd 90 X 0.9 X 21
In this case could the full width qr of the footing be used for b in the above
formula? Preferably not.
On the other hand, some specifications assume that, when piles are used,
the diagonal tension shall be computed as though the critical section is

that beyond FQ, where angle OFQ is 30°. In such a case, mn would equal
approximately 71 in. and v L would be at least 100 p.s.i., which is a bit too

large. None of these assumptions should be considered as perfect; they

are approximations at the best. The differences in computed values pro-
vide the engineer with some figures to use as a guide when he determines
what dimensions to assign for the structure. In this case, the 21-in. depth
seems to be none too large. Incidentally, if a footing is square, one may
assume that one-half of the loads on the corner piles will be resisted by
each of the adjacent trapezoidal sections when computing shears.
It is now evident that a designer might wish to place the point F, Sketch
(b), so that the 45° line would hit the top of the outer piles by using a

larger and deeper pedestal. However, since these piles may not be
driven exactly as shown, since a lower vl is desirable, since web reinforce-
ment is impracticable, and since any possible weakness should be avoided,
it seems to be advisable to increase d to 24 in. and, since the tops of
370 PILE FOUNDATIONS [Chap. 10

wooden can be cut fairly accurately, to decrease the embedment

piles
of the pile ends to 3 in., keeping a 4-ft. -square pedestal. The sides of
the footing over piles 4 and 7 and J3 and 9 are obviously safer than the
other two sides.
5. Bending in footing. For bending about F, of Fig. 10-5(6), piles
10, and 12 will have a lever arm of 2.5 ft. from the edge F of the
11,
pedestal. Then,

MF = 3 X 40,000 X 2.5 = 300,000 ft.-lb.

.M 300,000
.

s
= m
.

"
,

fsjd 20,000 X 0.9 X 2

20 = V = 3 X 40,000
= op
26 5 - m
.

bfd 210 X 0.9 X 24

The reinforcement may then be nine No. 9 bars instead of eight. In the
other direction,

Mr = 2 X 40,000 X 1 = 80,000 ft.-lb.

80,000
s
_ = in "
2

20,000 X 0.9 X 2
_
20 = 2 X 40,000
= 176 m '

210 X 0.9 X 24

Eight No. 6 bars could be used.

In general, the preceding analysis shows that the position of the piles is
very important when planning the footing because of the possible serious
effect of the large concentrated pile loads upon the shear and bending
in the cap. Even though poured monolithically, the concrete below the
tensile reinforcement should not be trusted to help on account of tensile
cracking.
In further consideration of the question regarding whether or not a pile
may cause serious diagonal tension in a footing, refer to Fig. 10-6 (a).
Assume that the design calls for pile 11 of Fig. 10-5 (a) to be located so
that, when equals 45°, the line FJ intersects the steel above the top of
the edge of the pile. It might be argued that some of the reaction p n
will reach the pedestal without causing the part FHKJ to crack off.

However, the pile may actually be driven as shown by the dotted lines.
Then the argument no longer seems to have value. Unfortunately, the
designer cannot be sure in advance just what the situation will be.
Incorrect driving will affect bending as well as shear in the footing. If
one provides strength enough to allow for moderate inaccuracies, he is
conservative;if the pile is actually much closer to F, he is a bit wasteful.

His opinion may not be asked before the piles are capped, and he should
not risk important matters affecting life and property for insignificant
economies.
It is desirable to make the contract drawings of important footings so
Art. 10-5] ON VERTICAL PILES 371

that typical details illustrate what should be done if irregularities arise.

For example, an adequate edge distance is desirable for the part of the
footing beyond the outer piles. Perhaps the edge should be moved HK
to LN in Fig. 10-6(a). Then the note calling for a minimum edge dis-
tance will control the construction. Furthermore, an increase of the
leverage of the piles beyond F may affect the bending moment and reduce

^D
Pedestal
Desiredposition offloor^
FQ _f#____
T

Mgljy i!>'°
'»:'•'.
K
-Pile #17

of pile #7/ p:
X
edge distance IMPOSSIBLE WEAKNESS OF A FOOTING
fa;- SHOWING HOW INACCURATE DRIVING WHEN TOPS OF PILES ARE TOO HIGH
MAY HAVE LARGE EFFECT UPON THE
THEORETICAL STRESSES IN A FOOTING

2 3 4^
©'
5 ,6
+

J
+
Pedestal
/ ,8
+
\ / ^
+6
^ ^
7
+
\
© 9)
f
/ + +
s +9
]

40 n, 42 +/J 14 /2 44
|
+/^| +// + \J3
1

45 I6
+ +17
-he 19 4s L
+17
-1 „
+/8
49
+
1 46 1

20 21 22 J3 ?o 22
4' +'

(c)- PLANNED LOCATIONS OF PILES rrfj-POSITION OF PILES AS DRIVEN,

UNDER A LARGE FOOTING SHOWING ECCENTRICITY OF LOAD

FIG. 10-6. Illustrations of how inaccurate driving of piles may endanger large footings.

the safety factor more than is desirable. A note covering this might
state that, any of the piles are driven more than 1 ft. (or some other
if

figure) farther from the center of the footing than shown on the plans,
the chief engineer should be notified so that the footing may be redesigned.
Another possible weakness is shown in Fig. 10-6(6). Assume that pile
11 is precast concrete, and that the top RS is 1 ft. too high. If bars b are
placed so as to straddle thepile, will this be satisfactory? The principal
danger comes from the tendency of the pile to punch out a section repre-
sented by QRST, and this may be very harmful. The pile should be cut
off below bars b. If RS is 4 in. higher than intended in the design, should
372 PILE FOUNDATIONS [Chap. 10

the pile be cut down, should the bars be raised 4 in. with the position of FH
unchanged, should the whole footing be raised 4 in., should the bars be laid
directly on RS with FH raised 1 in-higher? The answer is a problem for
the engineer to settle in accordance with good engineering judgment. Of
course, when the piles are below the intended elevations, they may be ex-
tended or the footing may be lowered or padded down to cover their tops.
Figure 10-6(c) pictures the design of a footing, whereas (d) shows
approximately the pattern of the piles as actually driven because of a
hidden obstruction or gravel pocket. Was this poor workmanship? It
was done by experts who had considerable difficulty doing even this well.
The engineer is shown the result and is asked whether or not he approves
proceeding with the footing which, incidentally, is to support a load of
1,400 kips. Is the effect of eccentricity dangerous? Yes. Should a
combined footing be made and extended across to the next column at the
left of the sketch so that a few intermediate piles can be driven? The
latter seems to be reasonable. On the other hand, should the obstruction
be dug out or blasted to pieces so that some piles can be driven down as
intended? Probably this is too difficult and dangerous when the other
piles have already been driven. At any rate, the eccentricity is too
serious to be neglected. It seems best to try to cut through the obstruc-
tion by driving down a steel H, pipe, or mandrel in three places at the left
side of the footing of Sketch (d). Then try to drive typical piles in these
places after the steel has been withdrawn. If this is unsuccessful, the

problem has to be solved in whatever practicable manner the engineer

can invent.
10-6. Isolated footings with vertical and horizontal loads. Con-
sider again the 12-pile footing shown in Fig. 10-5 (a). A longitudinal sec-
tion is given in Fig. 10-7 (a). Assume that the footing supports one leg
of a braced steel bent which causes vertical and horizontal forces to be
applied to the pedestal. The footing and piles are surrounded by clay
that has sufficient abutting power to prevent lateral tipping of the piles.
The maximum bearing value of a pile for dead and live loads is assumed
to be 25 tons. Inasmuch as wind forces are temporary loads, this allow-
able value will be increased 30 per cent when maximum vertical and wind
loads are combined.
In order to illustrate a method for the computation of what will be
called the load on each pile, and to analyze the footing, it will be assumed
that Pll+dl = 475 kips, P w = 60 kips, S = 40 kips, b = 10 No. 9 bars,
and d — 2 ft.
1. Total forces. The weight of the footing and the soil on top of it
will be included in the vertical loads acting on the piles. Therefore,
W = 475 + 60 + 55 = 590 kips (DL + LL + wind + footing).
2. Overturning. The position of the axis of moments for computing
the overturning is questionable. Somewhere a resistance of 40 kips to
Art. 10-6] FOOTINGS WITH VERTICAL AND HORIZONTAL LOADS 373

counteract the force S must be developed. Assume that the clay is

reasonably stiff and passive resistance are equivalent

so that its cohesion
to a soil having an angle of internal friction equal to 25° and a unit weight
of 100 p.c.f. Then Fig. 6 in the Appendix indicates that the passive
resistance can be assumed to equal
250 p.s.f. per ft. of depth. Fur-
thermore, assume that the com-
bined resistance of the 12 piles, as
explained later in Art. 10-13, is at
least equivalent to that of an ex- C—^JL -oLf&til-. K
tension of the width of the
9-ft.

footing. Include the 1-ft. depth of

earth on top of the footing. Then
faJ-DIAGRAM OF CONCENTRATED PILE LOADS
the depth x of soil to develop an
abutting power of 40 kips should be

0.250.r 2 40
9
x = 5.74 ft.

Assuming a triangular-force dia-

gram, the resultant will be applied
(2 X 5.74)/3 = 3.82 ft. from the (b)- SUBSTITUTE DISTRIBUTED PILE LOADS
ground surface. This would seem
FIG. 10-7. A spread footing on piles with vertical
to justify a lever arm
from of 4.82 ft.
and horizontal loads.
S to the center of resistance. Ac-
tually, the resisting pressure on the piles is likely to decrease from the top
down, somewhat as shown in Fig. 10-25 (c). Therefore, the actual center of
gravity of the resisting pressures is very questionable. In silts and soft
clays, it seems to be desirable to take moments about a point a few feet
below the bottom of the footing but not more than 10 ft. below the ground
when the footing is somewhat like that of Fig. 10-5, with no batter piles
or special provisions to resist horizontal forces. With firmer soils at the
top, the center of moments may be higher; it may be estimated as being
at the center of resistance of passive pressures computed as described
above, but it should seldom be taken as higher than the tops of the piles.
This will be discussed further in Art. 10-13. In this case, assume that
the lever arm of S is large, e.g., 8.5 ft. Then

M = 40 X 8.5 = 340 ft.-kips

3. Vertical loads on piles. It is assumed that the load on any pile is

W ,
Mc
374 PILE FOUNDATIONS [Chap. 10

where N = the numberunder the footing, c = the distance from

of piles
the center of gravity of the pile group to the pile in question, and / = the
plane moment of inertia of the pile'group about an axis through the center
of gravity of the group and perpendicular to the plane of rotation. In
this case, the axis is at the center line of the footing. Therefore, treating
the piles as units that are concentrated at their longitudinal center lines,
and neglecting the moment of inertia of each pile about its own center,
/ for the three rows is

I = 3 X 2(1. 2
+ 4.5 2 ) = 135 pile-ft. 2

Therefore, for pile 2,

2?2 = 590
-yj H
340 ,

™X 4.5 Ano
= 49.2 +,-,-, Q =
11.3 rn ,
60.5 kips
,
•

Similarly, pb = 49.2 + 3.8 = 53 kips, p 8 = 45.4 kips, and p u = 37.9

kips. The maximum is less than 1.3 X 50 = 65 kips allowed.
These computations seem to show that the footing is conservatively
designed.
It is possible, for the purpose of analyzing a pile footing, to treat the pile
loads as distributed pressures similar to those that would be caused if the
footing were supported upon soil. This substitute pressure diagram for
the preceding problem is pictured in Fig. 10-7(6). However, the piles
necessarily cause localized loads of large magnitude. It seems advisable
to plan and analyze moderate size upon the basis of
isolated footings of
these concentrated loads.Large groups of piles under combined footings,
mats, and large foundations may be replaced by assumed distributed
loads when the designer believes that this is a more suitable and efficient
procedure.
The stresses in the footing may be computed upon the same assumptions
as those that were discussed in the preceding article. When the pile loads
are computed, they may be used directly, or the average load per pile
caused by the footing may be deducted, and the remainder used as the net
pile load.Here again the angle OEQ in Fig. 10-7 (a) is 30°; OEN, 45°.
These are for consideration when computing the longitudinal shear.
When the loads include dead, live, and wind forces, it may be permissible
to increase the allowable unit stress in steel and concrete 30 per cent.
The reader may test the footing for himself, to see if it is safe for this
loading.
10-7. Isolated footings with eccentric loads. There may be cases
in which a foundation is not or cannot be made so that the principal
pile
load coincides with the center of gravity of the group of piles. Of course,
an engineer tries to avoid arrangements like this, but such a plan may be
the most economical one to be devised under certain circumstances. If
Art. 10-7] ISOLATED FOOTINGS WITH ECCENTRIC LOADS 375

made properly, a footing with more or less permanent inequality of pile

loads need not be feared. However, it is desirable to bear the following
in mind:
End-bearing piles that rest upon a firm stratum
1. may be loaded
unequally without serious danger.
2. Friction piles that are in soft cohesive soils may eventually settle

unequally if the loads on the piles in a group vary considerably and

permanently.
3. The inequality of pile loads should be kept to a minimum.

Footing of
{existing building
P=400><
f~~fcll £1.1/3

New col.

P=400>< 2L6"
D
faJ-GENERAL LOCATION

EUJ2\ fc;-ELEVATI0N OF FOOTING

%J8§£ Fin El. 98

tss/t.
-£1.9/
Fine sand
-El. 85

-i-|-f:
-EI. 7

Fine
sand
-El. SO
Coarse-
;sdnd-:'.

taj -SOIL PROFILE (d)-Pim OF FOOTING

FIG. 10-8. An eccentrically loaded footing.

4. Permanent downward pressure on the friction piles at one edge of a

footing and uplift on those at the opposite edge should be avoided.
As an illustration, assume that Fig. 10-8 (a) shows the location of a pro-
posed column near the reentrant corner of an existing structure that is
supported upon poured-in-place concrete piles 45 ft. long. The soil pro-
file is shown in (6). Poured-in-place concrete piles with metallic shells
and 143^-in. tips are to be used in the new construction. If driven so
that the tips are approximately at El. 58, each of these piles will be
assumed to have an allowable bearing equal to 30 tons. The top of the
pedestal is to be at El. 113. From Sketch (a), it is obvious that the
column will be near one corner of a rectangular footing or pile cap.
To obtain an idea of the number of piles required, assume that the
376 PILE FOUNDATIONS [Chap. 10

footing weighs 40 kips, and that the average load per pile is 50 kips.
Then,

N 400 + 40
=
_ ..
9 piles
50

To minimize the eccentricity, assume that the piles are to be driven rather
close together, as shown in Figs. 10-8(c) and (d). The trial footing in the
former weighs 36 kips. Its weight is slightly eccentric, but this will be
neglected.

® (&
-

e--e--@
F
(a)

FIG. 10-9. Two methods of determining pile loads caused by eccentricity.

Figure 10-9 (a) shows the piles as assumed, and the point A where the
column load of 400 kips will be assumed to be concentrated. The footing
load of 36 kips is at J. The moments of inertia of the piles about axes EF
and GH are the same, and

I = 3 X 2 X 2.5 2 = 37.5 pile-ft. 2

M EF = 400 X 1.25 = 500 ft.-kips
MGH = 400 X 0.75 = 300 ft.-kips

The load on pile 3 is the largest, and it may be estimated as follows:

436 500 X 2.5 300 X 2.5

= =
Pi
9
+
,

37.5
+
,

37.5
48.5 + 33.3 + 20 101.8 kips

This is obviously too large.

Another method of computation is illustrated in Fig. 10-9(6). The
group of piles is drawn to scale in position. The line J A from the center
of gravity of the group to the center of gravity of the column load is drawn,
and its magnitude is scaled or computed. The axis KL is drawn through
J perpendicular to J A. The perpendicular distances from the axis to the
Art. 10-7] ISOLATED FOOTINGS WITH ECCENTRIC LOADS 377

piles are scaled and recorded as shown. The moment of inertia of the
group is the sum of the squares of these distances, i.e.,

I = 2(0.85 2 + 1.3 2 + 2.15 2 + 3.45 2 ) = 37.9 pile-ft, 2

= 436 400 X 1.46 X 3.45 _

KO _ 1A1 - ,
= 10
+ = .
.

^g
,

Ps -g- H - 48.5 53.2 101.7 kips

This result checks the preceding calculations, as it should. The method is

shown only for the purpose of indicating its usefulness when a group of
piles is very irregular, or when it turns out to be so, as actually driven.
It is now necessary to increase thenumber of piles, to change the design
to a combined footing, or to devisesome other means of supporting the
structure safely. The problem should now be examined much more care-
fully. This will be done somewhat as an engineer might do when looking
for and judging the merits of various solutions.
1. Increase the number of piles. To an experienced engineer, the
arrangement of piles in Fig. 10-8(d) should automatically show a danger
signal. When the resultant load is applied at the outer edge of the middle
third of a rectangular footing on the assumption of a uniform varia-
soil,

tion of pressure would cause the intensity at the edge of the footing on the
"high" side to be twice the average unit pressure, whereas that on the
other side would be zero. It seems reasonable to think of the action of a
group of piles similarly. It is evident that point A of Sketch (d) is near
the edge of the middle third of the group. Therefore, one should expect
that the maximum load per pile would be somewhere near double the
average load, and an assumption of 50 kips for the latter is obviously too
large.
Using one-half the allowable load of 60 kips per pile, then, adding some-
thing for the weight of the larger footing,

N = 450/30 - 15 piles

If, in Fig. 10-8(d), six piles are added as pictured in Fig. 10-10(a), will this
be satisfactory? By inspection alone, one can approximate closely the
position of the axes EF and GH through the center of gravity of the group
of piles. Although the figures are given, it is apparent that the time spent
in computing them is wasted. At a glance, one can see that the load is far
over to one corner of the group. The piles in the vicinity of A will take
most of the load; those far away will resist uplift if they do any good at all.

Just to show the reader the results of computations to determine the

critical pile loads in this case, assuming that he does not believe the
preceding statement, the following are recorded: p 3 = 91 kips downward;
P12 — 18 kips uplift; pi3 = 18 kips uplift. The extra piles are a waste of
money, they do little to relieve the pressure on pile 3, and this plan is not
satisfactory.
 '

378 PILE FOUNDATIONS [Chap. 10

2. Driving the piles nearer to the existing footing. From Fig. 9-10, piles
50 ft. long may have a diameter of over 15 in. at the butt. It is there-

fore unreasonable to expect to drive than approximately 8 in. them closer

from BC and CD Furthermore, examine Fig.
to the center of the pile.
10-10(6). Assume that KLM
is the outside of the foundation wall and

NOQ is the projection of the eaves overhead. To be investigated now are

Edge ofo/d

jg ' .
.- E I footing .
r Existing foundation wall

&~0 r-3
cA

efko & e
e 5 -^ ^
/'0
<3

e G e13 14 15

F2'-3" 2'-

2'-6' 2'-6" 2'-6"

'k
2 k
W±6Jkl~=I=/l0pi/e-ft. M,, =1,050 ,M
EF GH ""r"" ' >-"£F''" .
r jS20
'-™aff" — (b)-<MW OF 9 PILES CLOSE TO
OLD FOUNDATION
(a)-TR\AL GROUP OF 15 PILES

E &
Gj C? <D
'$$•

& ^
'*•»

* e^e^B
•A 5,

2-6"
i-H
& &H
G, W,

&-&-Q «.*P
Edge of
"old fooling
2> » 2-0" iW'
D
e 7X1
ikf
V)
fc;-6R0UP OF 9 PILES CLOSELY SPACED /W-GR0UP OF 6 LARGE PILES TO
MINIMIZE ECCENTRICITY

FIG. 10-10. A study of a pile foundation under eccentric loads.

the clearances required for the pile-driving equipment. Can a pile driver
snuggle alongside the building wall to drive such piles as 1, 2, 3, 6, and 9?
Figure 10-1 1 At best, one should not assume that
will help to judge this.

the clearance RS can be less than 1 V£ to 2 ft, To reach

of Fig. 10-10(6)
pile 3, the equipment may approach the corner at an angle so that TU
represents a line parallel to the leads. In general, TU should be at least
3 to 4 ft. For this situation and to encase the tops of the piles, it is prob-
able that the piles should not be driven closer than 12 in. from the old
footings. If they are placed 2 ft. from NOQ because of clearance needed
Art. 10-7] ISOLATED FOOTINGS WITH ECCENTRIC LOADS 379

for the equipment and if the piles are 2 ft. 6 in. c.c, the effect of eccen-
tricity will be remedied only slightly.
In such troublesome situations, one should remember that certain
things called for on drawingsmay not be within the range of possibility.
However, when he has been told subsequently that the piles have been
driven as close as possible, yet they are 6 to 8 in. outside of the positions
called for, he must accept them. He must now permit overloading of
some of the piles or change the design of the structure. It is better to
be sensible and conservative in the first place and not attempt to do the
impracticable.
3. Suppose that the spacing x in Fig.
Drive the piles closer together.
10-10(6) is reduced to 2 ft. The point A would then be 6 in.
; and y is 1 ft.

from pile 5, as shown in Sketch (c). Then I E f = Igh = 24 pile-ft. 2 ;

M EF = 400 X 0.5 = 200 ft.-kips; and p 3 = 65 kips. Theoretically, this

could be accepted, but it is practically impossible to guarantee that the
piles will actually be this way when driven. Experience shows that, even
though the pile driver could do this, the lateral pressure caused by driving
pile 5, for example, so close to 2, 3, and 6 is likely to cause the shells of
these previously driven piles to collapse. If they are filled with "green"
concrete, the concrete may be damaged. A spacing of 2 ft. 6 in. is close
enough 3 ft. is better.
;

4- Use a smaller number of stronger piles. The coarse sand below El. 50
of Fig. 10-8(6) very firm. Assume that piles are to be driven down to
is

El. 48. they support safely a load of 40 tons? The resistance

If so, will
to penetration into the coarse sand will be large. Jetting would assist in
getting the piles down, but the action of the water might endanger the
bearing value of the existing piles that now end at El. 62, 14 ft. above the
desired position of the tips of the new piles. However, open-ended pipe
piles might cut into the sand sufficiently without great difficulty. Pre-
cast concrete piles probably could not be driven. Poured-in-place piles
with shells and small tips might not have sufficient end area or strength
as columns. Pedestal piles would be difficult to make in a closely spaced
group when the deep fine sand or coarse sand must be displaced. It seems
to be advisable to use steel pipe piles with open ends and with the pipes
left in place for protection of the newly placed concrete. An analysis of
the group shown in Fig. 10-10(rf) indicates that the probable maximum
load on piles 1 and 2 will be at least 90 kips each, which is probably some-
what too large. Before this is accepted, one such pile in the vicinity
should be subjected to a load test.
5. U se a combined footing another column is along wall BC or CD of
. If
Fig. 10-8(a) and
near enough, a combined footing might be used.
if it is

This type of construction is illustrated more fully in Art. 10-8. Probably

this construction is the most practicable.
380 PILE FOUNDATIONS [Chap. 10

30- ft. extension section

can be added

Working weight SO tons

Leads
1
Min.64 ,

Max. 80'

Bedframe
Runnincfbeams
Top View
(Quarter Scale)
Area required for operation = 40x4€'

Bedframe
Main skid
be„ m
Turntable <
}

Side Elevation
(a)-OUTLINE DIAGRAM OF A TURNTABLE SKID PILE DRIVER W-OUTLINE DIAGRAM OF RAYMOND CRANE PILE DRIVER
(Courtesy of Western Foundation Corp., New York, N.Y) (Courtesy of Raymond Concrete Pile Ca.NewYork.N.Y.)

FIG. 10-11. Some information regarding pile-driving equipment.

General Data for Turntable Skid Pile Driver

Wt., tons
1 two-drum hoist and 50 H.P. boiler 11
1 leads — 60-ft. main section and 30-ft. extension 9
1 bedframe 36 long ft. 5
1 turntable 16 ft. X 16 ft. (with circular rail) 5
2 main skid beams 40 ft. long j4
34

1 No. Vulcan steam hammer 8

5 40 -ft. running beams and 15 mats 10
Total shipping weight 52

Notes: Pile-driving equipment is very much specialized. drivers may have reach of 100 ft. and
Some
be able to drive 36-in. -diameter piles. Others are made to drive short sections beneath 8- or 10-ft. clear-
ances. Drivers are usually designed to handle only certain types of pile most efficiently. They may be
made for railroad, floating, skid, or caterpillar supports.
Lengths of pile that may be driven if proper equipment is available (approx. limits):

Length, ft.
Wooden 75
Precast concrete (not over 24"0) 70
Composite wood and light-shelled cast-in-place concrete 125
Composite pipe and light-shelled cast-in-place concrete 200
Cast-in-place light-shelled concrete 110
Cast-in-place compressed concrete 60
Concrete-filled pipe (not over 24"<£) 200

6. Revise design of superstructure. Sometimes the most practicable

solution for troubles of this sort, especially when loads are large, is one of
the following:
a. Set the new structure far enough from the old one so that reasonable
footings can be built easily and economically.
Art. 10-8] COMBINED FOOTINGS ON PILES 381

b. Offset the columns in the lower portion of the structure so that the
foundation condition is satisfactory. Then support the upper portion of
the columns on girders cantilevered across these stub columns.
c. Set all columns as desired for the foundations, and cantilever the

floor and wall framing past these columns.

7. Conclusion. The owner should be shown the advantages of action in
accordance with item 6a. If this is impossible, the most practicable solu-
tion is likely to be the use of a combined footing. This problem demon-
strates the desirability of planning foundations and of making trial cal-
culations in the early stages of a project instead of neglecting such matters
until the design of the superstructure and equipment has progressed so far
that changes will not be accepted, at least not without an argument.

20-0
1
'

——rX
Existing footing
20-0" /., 20 L "
////,/ ////////////////////]//////////////,///,-,//-»
2-6"

y £

m
A-400^- B,C,ctndD=500!<; EandF-580 k ^
FIG. 10-12. Columns, locations, and loads for a new building adjacent to an existing one.

When the pile cap is so thin or so wide that itcannot act as a rigid body
in spreading the loads over the piles as required, the formula P/A ±
Mc/I should not be used to compute the pile loads. The forces will then
have to be assumed to be distributed locally to various piles as the char-
acter of the situation seems to justify.
10-8. Combined footings on piles. There may be situations in
which it is desirable to use combined footings with piles under them. As
a specific illustration of the planning of such construction, assume con-
ditions similar to those shown previously in Figs. 10-8 (a) and (b). Figure
10-12 pictures the corner column A and some of the neighboring columns
of a large addition to be built alongside an existing building. Of course,
the piles under the present structure are inadequate to support the new
one if new loads are added to the present loads. Supposedly there are
good reasons why it is desirable to locate the new columns in the positions
shown. The column loads given in Fig. 10-12 do not include the weight
382 PILE FOUNDATIONS [Chap. 10

of the substructure. Again assume that the new piles are to be designed
for a bearing value of 60 kips each.
By adding an assumed weight of 50 kips to each of the column loads to
allow for the weight of the substructure, the trial number of piles needed
under each column is the following: at A, 450/60 = 8; at B, C, and D,
550/60 = 9; at E and F, 630/60 =11.

20 -0" 20-0 2-6"

3i0'[A 3'-0"}3L0' 3L0" 9L9"

'
J'-O" 3-0" 8-0" 3i0"/'3\73''
A
If
k 'K -g iiu — T5S 35
h dikb
4 i

. t
Pedestal Wall rooting Atf _JA

s-'o

K o D,

CaJ-PlM OF WALL AND FOOTINGS

1-6"
400"
500 k

,B
n
n 1

2p 2p 2p 2p 2p
1 1 TTT
2p 3p/2 p/2

(c)-U)hV DIAGRAM FOR PORTION A~E

(b)- SECTION A-A

FIG. 10-13. Illustration of a continuous, combined, exterior footing.

What is the best arrangement for the piles? As usual, different

schemes are possible. shown in Fig. 10-13(a). This utilizes walls
One is

along the outer column rows, and the piles are arranged in two rows under
narrow footings so that the typical construction is shown in (b). The
piles under A are grouped so as to be as close to the column as possible,
and to have those that are off center so placed that they cause bending
but no twisting in walls AB and AD. Thus at B, C, and D, the piles are
Art. 10-8] COMBINED FOOTINGS ON PILES 383

arranged so that the substructure constitutes a sort of elongated footing

that is symmetrical about the center of the column. The footing between
the pile groups might be notched back as shown by the dashed lines in
Fig. 10-13(a) if the bending strength of the wall will permit it, but this
economy is not great. Of course, the piles could be spaced equally along
the wall, but this would increase the bending moment in the latter and
would accomplish no useful purpose here. If the wall were the outside of
a deep basement, it would be so strong anyway that the effect of bending

would be unimportant even when the piles were equally spaced.

In the computations for bending moments, shears, and bond in the
walls, it is generally satisfactory to assume that the dead load of such a
section of substructure as AB of Fig. 10-13(a) is distributed uniformly.
Then the net pile loads may be used in the calculations. In this case, it

is sufficient to average the loads on the piles under A and B, then use this

figure as the load p shown in the load diagram in Sketch (c). Here the
corner loads under A are assumed to be divided between members AB
and AD as though the members were cut along the diagonal GH of
Sketch (a).
Another variation is pictured in Fig. 10-14(a). This shows a long
narrow thick footing extending from A past B. The piles between these
points are distributed equally and so that the resultant of the loads on A
and B almost coincides with the center of gravity of the pile group. At
first glance this might seem a reasonable design. However, one must
guard against surprises when planning heavily loaded members like these.
An approximate check of this footing should be made before the scheme
is adopted. This may be done with sufficient accuracy as follows:

Average
....
load =
pile
400 + 500
= rn kips
50
.

^o

Substitute uniform load = 2

- —X-—
50
= 33 kips per ft.

Approximate M at center = 33 X= 20 2
= ft. -kips
1,650
o

Approximate A =
s = 37 in.'
18 x *ff| 2J5
Approximate vl at 45° line from bottom edge of pedestal, using the
uniform load, equals

—
33,000(10 - 1.25 - 2.5)
= 1 13 D
p S 1

72X0.9X28
Both A and
s v L show that this footing is too shallow for good design, and
the computations need be carried no further. It is therefore desirable to
384 PILE FOUNDATIONS [Chap. 10

20-0"
J O'
l
\
6@3'-0"=/8'-0"
JJQ^jf/%"
JJ°jPJ
v, ///////////////////////////////////Zh/////////. y//)vA/.

4) A. 9 9—Q—Q^"'^'X'W^
—o <k4& Existw footing

K \By2 6sq.
L
2L6"sq.-^\ T
l
L 9" 2'-6" i-9'
A^T

C-:
^i%'
/# " i
j4?" I
/-<?

B-l
^-SECTION A-A

© z>

B^
(a)-lW COMBINED FOOTINGS

-Existing
V/, footing
New-*
column
Yy.

C
/

1
'

Q €> X
,

Q o

teJ-ANOTHER ARRANGEMENT OF COMBINED FOOTING (dhk SPLIT FOOTING

FIG. 10-14. Some ether arrangements of combined footings.

increase the depth, to use a T section as in Fig. 10-13(6), or to shift the 10

piles forB farther to the left so as to develop more negative bending under
B and thus relieve the moment and some of the shear in AB. A combina-
tion of these ideas is preferable.
Elongated footings parallel to the outer walls might be used for columns
C, D, etc., of Fig. 10-12. On the other hand, an arrangement like DE of
Art. 10-9] MATS ON PILES 385

Fig. 10-14(a) may be adopted with section B-B similar to A-A if conditions
make this desirable. Here the portion under E is planned primarily as an
isolated footing to carry the load at E, then the few piles between D and E
cause bending in the portion DE. In other words, the elongated footing
is joined to E in order to anchor that end and to eliminate the twisting
that would occur otherwise.
Figure 10- 14(c) is prepared to show one scheme that might be used if

there were some obstruction between between A and D so

A and B and
that no continuous wall or footing could be used between them. The
footing at A is planned first as though it were an isolated one with eccen-
tricity. That at E is planned as a symmetrical isolated footing with its
long axis parallel to AE. The beam between them is then designed to
resist the bending caused by the unsymmetrical loading at A.
It often happens that the piles under an existing foundation are grouped
under a footing that projects, as pictured in Fig. 10-14(tf). Sometimes
new ones can be grouped at each side in this manner with concrete or
encased steel beams to carry the column load across to the caps.
In such problems, an engineer has to study the particular case and then
invent an arrangement that will serve his purpose most advantageously.
The illustrations here are merely for the purpose of showing possible ways
of attacking such problems.
10-9. Mats on piles. In many respects, the design of a heavy rein-
forced-concrete mat on were supported
piles is similar to its design if it

directly on soil. Except for local shearing stresses and punching action,
piles that are spaced uniformly under the mat may generally be replaced
for purposes of computation by an equivalent uniformly distributed load.
For example, assume that the wooden piles under a mat are to be driven
in rows 4 ft. on centers both ways, and that the net load on them exclusive
of the mat is 18 tons each. This is equivalent then to an average upward
pressure of
18 X 2
= 2.25 k.s.f.
16

However, the bending moments in a mat so loaded and having columns

20 ft. c.c. are surprisingly large. One must be sure, too, that the concen-
trations can be spread safely and that the mat is stiff enough to avoid
overloading of the piles near the columns because of local sagging of the
mat at and near each column.
It is generally desirable to group the piles more closely under the por-
tion of the mat near and beneath the columns, just as though
that is

isolated footings were used. The mat may be needed as a floor for a base-
ment, or even as a first floor. It may then span from pile group to pile
group somewhat as though it were a flat-slab floor on column capitals,
386 PILE FOUNDATIONS [Chap. 10

or a few widely spaced piles may be placed under it to relieve the bending
when the live loads are heavy. However, one should be careful how he
does this.

Figure 10-15(a) shows a typical portion of a thick mat that was used
under a power plant. The left-hand half of the picture shows the details
originally proposed. The poured-in-place 30-ton piles under the thick-
ened portion were designed to support the column loads of 1,000 to 1,500
kips, just as though there were no mat. Then additional piles were
driven about 4 to 5 ft. c.c. in the intervening spaces to support pump
foundations, a above the mat on which the first floor was to
5-ft. backfill

be placed, and the heavy loads on the latter floor. These intermediate

Steel sheet piling

Slab
PvIRE U
t£_UU-l>ifclr
1 1
i i
X T-beam
u Gfu u a j~ Const, jt.
Footing
Longitudinal reinf.
foJ-DEPRESSED FOOTING
not shown

I I
P I o i ,-- Construction joint
=^i=l Structural
-Pedestal
Mat ?.
v
;V r '
:
r
slab
J£Jh*r—-*'—-+

TJ
m ^
U^Pi/es
tTremie
concrete

(b) -CONTINUOUS MAT WITH RAISED PEDESTAL (c)- INTAKE TUNNEL POURED ON
TREMIE-CONCRETE MAT

Fig. 10-15. Some arrangements of reinforced-concrete mats on piles.

piles were assumed to support no more than 10 to 12 tons each, but they
were the same as the others and were driven to the same depths.
Now try to visualize what might happen when the final loads are
applied to the substructure and the 50-ft. friction piles are fully loaded.
When the mat is poured, all piles will have small and almost equal loads.
As the column loads are applied, the piles under the "footing" will settle
slightly. The 3-ft. mat will try to spread these loads over the intermedi-
ate piles also. If it is not strong enough to do this, it will crack around
the footing. If the mat is strong, it will distribute the loads about equally
to all the piles. Thus the average load per pile may be 20 to 25 tons, and
the intermediate ones will cause severe bending and shear in the mat.
This is another one of those cases in which the structure should be con-
sidered as an entity instead of an assemblage of independent parts.
Refer again to the details shown in Fig. 10- 15 (a). It is obvious that if
Art. 10-9] MATS ON PILES 387

the reinforcement is and as shown on the left-hand

as detailed originally
side of the drawing, the mat can crack between the ends of bars a
easily
and b. None of these can develop the required tension at the bottom of
the mat where it joins the footing. The bars 6 will probably be over-
loaded because the upward shear from the mat will be applied at the edge
of the footing where it will have a maximum lever arm. The dowels c are
not detailed properly either.
The revised design is shown in the right-hand half of Fig. 10- 15 (a).
The piles being already driven, it was necessary to use them. Bars b
were therefore increased. were made straight, and some of
Bars d and e

them extended as shown. Then bars

g were added around the edges to
serve as stirrups and as a sort of hanger to attach the mat to the footing.
The dowels / were made so that they could be supported on b during
construction and so that the inadvisable use of a hook to resist compres-
sion could be avoided.
When it is possible to do so, the construction shown in Fig. 10-15(6)
should be used instead of that in (a) for the following reasons:
1. The load from the column can be distributed more directly into the

mat by shear and compression. It is not advisable to have to use sus-

penders like g in Sketch (a) if they can be avoided.
2. The tension in the bottom of the mat can be taken by full-length

band bars m, with extra bars k under the column region. This avoids
carrying heavy tensions around corners, as with bars e, b, and g of
Sketch (a).
3. Bars j and h can be used as bands to resist tension in the top of the

mat —the intermediate areas.

4. The pedestal can be made whatever area and height are necessary
of
to prevent excessive punching shear and diagonal tension in the mat.
5. The tops of the piles can be cut off at the same elevation so that no

depressions and special forming are required. Of course, as a general

principle and as stated previously, it is desirable to group the piles under
the mat so that they are near the column points, thus reducing the bend-
ing in the mat itself. The point emphasized here is that the piles at
intermediate locations will resist loads if they can. This fact should
not be forgotten.
Another example of the use of a mat is shown in Fig. 10- 15(c). This
illustrates the construction used at the intake to a power plant at tide-
water. Heavy sheet piling was driven around the area required for the
structure, the fill and muck were excavated considerably below the invert,
and the wooden piles were driven under water by means of a follower. A
tremie-concrete mat some 3 or 4 ft. thick was then placed below the future
invert, and its top was roughly screeded to grade. When this had hard-
ened sufficiently, the cofferdam was unwatered safely. The weight of
388 PILE FOUNDATIONS [Chap. 10

this pad did not theoretically counteract the 10-ft. head of water, but the
resistance of the piles and sheet piling to upward pull probably took care
of the difference.
The structural slab of the invert consisted of a heavy mat that extended
under the outer walls and under a heavy middle longitudinal wall. When
the concrete substructure was completed and the superstructure was
built, the structural mat acted as a one-way two-span slab to spread the
loads from the walls to the piles through the now inactive tremie concrete.
When the gates were closed and one-half of the twin structure was
unwatered, the empty side resisted the upward and lateral hydrostatic
pressures as a box. The structure was not waterproofed because minor
seepage would not be troublesome.
Another case of what seemed to be a mat was really not a mat in this
sense. A one-story warehouse with a large floor area was to be built for a
grocery concern on a former low river flat having a thin blanket of new fill
over 20 ft. or more of silt underlain by very fine sand and then by medium
sand. The structure was planned with light columns approximately 28
by 54 ft. c.c. to hold the roof. The reinforced-concrete floor was designed
as a flat slab. Single 50 ft. poured-in-place concrete-and-shell piles were
driven in rows 14 ft. apart and 18 ft. each row except under
on centers in
the columns, where two piles were used. Thewere capped
single piles
with 3-ft. -square capitals and 6-ft. -square drop panels, whereas similar
rectangular ones were used under the intermediate building columns.
Continuous grade beams were built under the outer columns. All were
poured monolithically with the floor slab. Each pile or pair of piles
served practically as a column, whereas the "mat" was in effect nothing
but a self-supporting floor.

The situation may be very different for pier platforms, relieving-plat-

form types of bulkheads, large one-story warehouses, industrial plants,
and other structures that have widely spaced columns with light super-
structures but must support large trucks, trains, localized heaps of heavy
goods, and machines. In these cases, the piles are necessarily located so
as to support directly heavy fixed loads such as tracks and machines.
Under extensive floor areas, the piles should be driven in rather widely
spaced rows or in whatever arrangement suits the conditions so that there
are sufficient piles to hold up a large load in any possible location without
overstressing the mat or floor. It is advisable to make such a mat thick
and stiff, with reinforcement in both directions at the top and bottom.
The principal weakness is likely to be the diagonal tension in the concrete
at the surface of an imaginary 45° frustum of a cone sloping upward and
outward from the edge of the top of the pile. It is usually desirable to
make the mat above the piles thick enough to avoid the need for web
reinforcement.
Art. 10-10] DIFFERENTIAL SETTLEMENTS 389

The computation of pile loads on the basis of the eccentricity of the

resultant of the applied loads with respect to the center of gravity of the
pile group and on the use of the moment of inertia of the pile group as
illustrated in Art. 10-7 has its limitations. This procedure is satisfactory
when the group is not excessively large and when the structure is rigid
enough to act as a unit; it is not generally applicable when the piles are

scattered over a large area and when the structure will not act as a rigid
unit, even though a mat foundation is used. For example, consider
this case:
A boiler house is approximately 60 by 120 ft. in plan. It has a mat
foundation with piles grouped under the mat but in the general vicinity
of the columns and equipment foundations, each group being propor-
tioned in accordance with the local loads. The coal bunker is near one
end of the structure. This bunker may be full or empty, causing a
considerable shifting of the resultant of the loads, whereas the piles, of
course, remain in constant position. The effect of eccentricity of load in
this case can be neglected because the entire structure cannot tip as a
unit. Although the theoretical eccentricity under maximum load was
5 ft., the designers were justified in neglecting its effect upon the maxi-
mum load on the extreme piles.
10-10. Differential settlements. Besides those differential settle-
ments that may occur because of differences in applied loads and in the
qualities of portions of the substrata, sometimes there are differences in
the piles themselves and in their action. For example, Fig. 10-16 shows
the case of some wooden piles driven in 1931 for an extension of a power
plant. Because of the business situation the work was discontinued.
When the extension finally went ahead in 1946, recent developments in
water-jacketed boilers had been so great that the new structure was to be
vastly different from that originally contemplated —
seemingly a charac-
teristic of one's plans for future extensions. The new loads were so much
heavier than the original ones that Raymond concrete piles were decided
upon.
When planning the new work, the engineers were faced with the prob-
lem of deciding what to do with the existing 190 wooden piles. Should
they be used, discarded, or pulled out? Since it was possible to make the
layout so that the new screen house and intake could be placed upon the
wooden piles, the engineers decided to use them. The adjoining switch
house was to be founded upon new Raymond piles. From the standpoint
of the superstructure, it was desirable to make the screen house an integral
part of the switch house. However, would the old concrete-capped
wooden and the new concrete ones compress and settle equally? If
piles
they did not, the concrete foundations and the brick walls would inevi-
tably crack. Since the wooden piles were driven for loads of 15 tons each,
390 PILE FOUNDATIONS [Chap. 10

whereas the new piles were to support 25 tons apiece, equal movement at
the tops seemed to be improbable? and reinforcing the concrete to prevent
cracking was practically impossible; at least, the results would be uncer-
tain and cracking would be
difficult to remedy. Therefore, the two
shown in Fig. 10-16(6), in order that vertical
structures were isolated, as
movement could occur without harm to the structure. Two expansion
joints were used at the ends of the large cast-iron intake pipe that passed
over the joint so that this piece could rotate slightly in a vertical plane;

Superstructure not shown

Membrane WP-
MHI/ir= El. 100-

New machine shop area

Monel metal wcter-stcp Jl

Intake w eil
New mat
Old, unrein forced mat,
urEL 80.5

-55 ^

—yuuyr
-^ — —
^-Wooden
-.

-Wooden piles
piles'
*rj-
01 It

E/.56 +
A A I ~>t
tl +J $ Fine sand
+ J

T Existing cofferdam

"
(a) PARTIAL PLAN '
[6) VERTICAL SECTION A-A

FIG. 10-16. Pile foundation for part of a power plant.

otherwise, settlement might crack the piping and cause a shutdown of the
plant.
Another problem that faces one when making extensions to existing
piled structures is illustrated in Fig. 10-17(a). This pictures an existing
plant that is shown by the solid lines; the proposed extension is indicated
by the dotted ones. The original footings and piling projected beyond
the end wall in order to avoid or reduce the eccentricity of loading pro-
duced by the end columns and the heavy brick walls. The removal of
the latter will reduce loads on the piles somewhat, but the principal
rebound of the foundations will be only that of the elastic recovery of the
Art. 10-10] DIFFERENTIAL SETTLEMENTS 391

piles and ground. The new piles to be driven alongside will be loaded
gradually as the addition is built. Compression of the new piles may
not be large, but consolidation of the soil supporting them may be
appreciable, and it will undoubtedly affect the neighboring soil under the
adjoining end of the old structure, especially if the subsoil is plastic.
In such a case, at what elevation should the floors in the new structure
be built compared to those in the existing plant? If they are placed at
the same elevation and if the new construction is connected to the exist-
ing framing, the settlement of the new struc-
ture will cause the beams to tilt slightly. New 2nd floor
1

Original
buildina
!h* I

r,
This may
not have serious effects upon the jitewJsff/oor_
!

\Xa+&l
but it is likely to crack masonry walls
floors, tx^,
,
d_
and cause severe stresses in vertical bracing
! !

that joins the new and the old steelwork. If 1

<->>//«

fa/- POSITION OF EXTENSION

the two structures are isolated and if new col-
umns are placed alongside the old ones, as
shown in Fig. 10-17(6), this arrangement will
have many obvious advantages. However,
some means should be provided for preventing
small offsets in the floors over which persons
may stumble. Some type of adjustable or
"floating" panel should be used at these junc- ^-EXTENSION ISOLATED FROM
ORIGINAL BUILDING
tions to attain a ramp effect.
Cross walls at col. lines
obvious that driving piles alongside an
It is Extension, j_{ L
existing structure like that of Fig. 10-17 (a) is

likely to disturb the piles and the soil under

the building. When an extension is antici-
pated some time in the future, one might
consider the following alternative procedures: (W-OMISSION OF PROJECTING MAT

1. Drive the piles under the exterior wall of

FIG. 10-17. Extension of an ex-
the original building as in Fig. 10-17 (a), using isting building when on piles.

enough of them for all probable future loads

and pouring the cap or footing to deliver the present and the future loads
to them. In the future, connect the framing to the existing structure,
using relatively flexible siding.
2. Use a mat foundation or continuous narrow footings under the
outer bay (at least), and cantilever the end wall, as in Fig. 10-17 (c). In
the future, drive the new piles and construct the addition as an isolated
structure. The advantage of this scheme is the fact that the future
construction may be what and where it needs to be, and it is not affected
by present wrong guesses.
3. Use the scheme in item 2 but drive at least the adjoining three or
four rows of piles to serve as a buffer when the additional ones are driven.
392 PILE FOUNDATIONS [Chap. 10

Preferably, these piles should not be capped until the future work
proceeds, thus permitting additional piles to be driven and the footings
to be designed when the complete requirements are known.
Of course, the trouble with all these suggestions is the fact that they
require extra cost at the time of the initial construction. Many owners
object to this.
The action of an organic silt surrounding end-bearing piles when the
silt isunder a heavy surcharge may be illustrated by the following case.
At the site of a power plant alongside a tidal river there was approximately
35 ft. of organic silt overlying fine to medium sand. This silt was soft
enough so that when the soil samples were taken, the spoon could be
pushed through the silt by the weight of the hammer. The top of the
silt was approximately 3 to 4 ft. below mean low water, and rock was

some 60 to 70 ft. below this level. A sort of sand dike was built in a
trench along the bay side of the site; then a sand fill about 25 ft. deep was
spread over the site and left there for a few months. The excess fill was
later removed, and the first unit of the plant and its accessories was
built on end-bearing concrete piles driven to (or close to) the rock. The
main plant seems to have remained intact except for a settlement of
about 1 in. near one corner. However, some of the light structures which
were on only a few piles have not settled, but they have "drifted" hori-
zontally riverward. This was evidently caused by flow of the silt under
the pressure of 15 to 18 ft. of permanent fill left on the site, probably
causing the piles to bend near where they enter the sand. This condi-
tion obviously can be dangerous. Also, within 3 years, more fill had to
be added around the structures to maintain the elevation of the yard
surface as the silt consolidated. One can readily realize that this settle-
ment must cause some "negative friction" on the piles. About 3 years
after the initial construction was completed, prestressed concrete piles
were driven to rock for an addition to this power plant. Once in a while,
as a pile was being driven, water would squirt up alongside the pile.
This appears to have been caused by the pile puncturing a water pocket
which was under pressure and which was apparently produced by squeez-
ing of the water into local areas as the silt consolidated. This action
occurred in only a few cases, thus seeming to show the localization of
concentrations of water.
10-11. Sequence of driving piles. As stated previously, it often
happens that a large number of piles are to be driven for a new structure
alongside an existing one, or that the new work is to be an extension of an
existing building. Of course, disturbance of the old structure is to be
minimized. How should the piles be driven and in what order? The
answer to these questions depends upon the conditions at the site, possibly
upon the type of pile to be used, and how they are to be driven.
Art. 10-11] SEQUENCE OF DRIVING PILES 393

Assume, first, that the project is the extension of a pier, that the plans
ends
of the of the new and the old structure are as shown in Fig. 10- 18 (a),
that the soil profile is as pictured in Sketch (b), and that creosoted wooden

piles are to be driven. In this case, it would seem that, if the pile driver

Areas for piIfnq

r\-T—
I
J 2
Tt-
! \ /
Extension
/ Existing pier.

5 4 ^Present wall
J! and projecting
II J
mat with
pi/es below

7 6
Medium c/ay
I

.1.
^-EXTENSION OF A PIER

iaJ-PLAU OF FUTURE PILING

Extension,

Water I

fab/e

—mSZZ'. ~r

-[Siit
S and
SPlSPiT Fine sand
Rock
fcJ-END-BEARING PILES frfJ-PILES INTO FINE SAND

I Pilebe/no/dris/en
tieavinq \Buiiding
Newpites \

/ /
Siitor
\K
soft clay.
it-< H—
1-
New^'

piles

Movement
lC/ayH 7
O/dpi/es
ofsoil
feJ-POSSIBLE DAMAGE TO PILES f/V-HEAVING AND UPLIFT OF
BECAUSE OF LATERAL PILES BECAUSE OF DRIVING
PRESSURE DUE TO DRIVING INTO PLASTIC SOIL

FIG. 10-18. Some problems encountered when driving piles alongside existing structures.

were mounted upon a barge, the sequence of driving would not be very-
important as long as the work proceeds outward from the end of the exist-
ing structure, the order of driving being areas 1, 2, 3, 4, 5, 6, and 7, or
1, 2, 4, 6, 7, 5, and 3. If the pile driver is supported upon completed
piling, the latter order may be preferable in order to minimize movement
394 PILE FOUNDATIONS [Chap. 10

of the equipment. The movement of the mud and clay of the bottom
will not be likely to disturb the existing or completed work seriously.
Next, assume the soil conditions pictured in Fig. 10-18(c), the existing
structure being heavy and on end-bearing steel H piles or concrete-filled
pipe piles. The pile driver must be supported upon piling or upon a
heavy timber grillage or mat that is borne by the fill; for economy, it

should utilize the new H piles as far as possible. The order of driving
them depends very much upon what arrangement will best provide for a
minimum of movement and shifting of the equipment. The general
order might well be areas and 3 or 6, 4, 2, 1, 3,
1, 2, 4, 6, 7, 5, 5, and 7 of
Fig. 10-1 8(a). The material displaced by the new H piles will not be
great, and all piles will reach a firm stratum of rock without excessively
hard driving. Serious disturbance of the existing structure is improbable.
Assume a third case like that shown in Fig. 10-18(c?). The original
plant on wooden friction piles; the extension, being much heavier, is to
is

be supported upon thin shell cast-in-place concrete piles that are to sup-
port twice as much load per pile as do the wooden ones. The first plan
considered is the driving of piles in the order of areas 3, 5, 7, 6, 4, 2, and 1
in order to relieve pressure upon the bulkhead due to displaced silt, and
to compact the silt next to the old structure so that the piles in areas 6, 4,
2, and 1 might have an increased frictional resistance. If this is done, the
piles driven in these last areas will displace considerable silt, and pressures
may be exerted northward against the old structure and southward
against the newly driven piles, as illustrated in Sketch (e). The former
may bend the wooden piles severely and cause some upward pressure
under the existing mat; it is unlikely to cause movement of the entire
structure. On the south, the side of least resistance, the movement of
the very likely to displace the new piles sideways or to tilt them as
silt is

shown in Sketch (e), breaking them at EF, the top of the sand. Because
of this danger, the piles were driven in the order 1, 2, 4, 6, 7, 5, and 3.
Hard driving alongside an existing piled building is likely to disturb the
present structure because of vibrations; it may also cause subsidence of
the adjoining sandy or silty soil, and of the structure. Jetting of the piles
may or may In the case of plastic soils, vibrations and
not be advisable.
movements of the supporting material may cause the adjoining piles to
settle because of the impairment of their frictional resistance and of the
shearing strength, cohesion, and compressive resistance of the soil upon
which the Later on, the application of more loads upon the
piles react.
underlying strata below the piles, Fig. 10-18(.f), may cause further sub-
sidence because of compaction of this material.
When planning a pile foundation and when specifying the sequence of
pile driving, a designer should bear in mind the nature and magnitude of
pile-driving equipment. On a drawing it is easy to make circles to repre-
Art. 10-11] SEQUENCE OF DRIVING PILES 395

sent the piles without bothering to think how they will get there. Hence
3 piles in one group, 10 in another 20 ft. away, 6 in a third 25 ft. off in
another direction, a row of 10 at 3-ft. spacing somewhere else, etc. these —
may seem simple enough. With a floating driver, such scattered work
may not be difficult; neither is it difficult on reasonably good soil. But
in muck, soft clay, or loose and saturated silt, it is another matter. A pile
driver weighs many tons; dragging it out of the mire is both difficult and
costly. It may have to be supported on many temporary or permanent
intermediate piles. No wonder then that the contractor's bid price for
such work is high.
When piles are to be driven in entirely new work, the principal features
to consider in planning the operations are the following:

1. Point of delivery of piles

2. Transportation of piles from stock to point of installation
3. Minimum temporary construction to hold equipment
4. Minimum movement of equipment
5. Parts needed first for construction of portions of superstructure
6. Ability to use driven piles as temporary supports for equipment
7. Portions needed as temporary supports for other purposes than
holding pile-driving equipment
8. Adjustment of driving to supply of materials
9. Allowance of time for setting of concrete in piles
10. Use of available equipment if a special driver must be obtained for
part of work but is not at hand, as for handling extra-long piles

If subterranean obstructions are known to exist in an area where piles

are to be driven in large numbers, the planning as well as the sequence of
driving may be affected. Wlien piles are to be driven along old water
fronts, thorough investigation first is good economy. Here is one
example
Several large steel tanks were to be built on hydraulic fill placed over
oil

a former shallow area at the margin of a harbor. The contents and the
steel of each tank were to be supported by a concrete mat built on poured-
in-place friction piles under the tank. A boring was made at the location
of each tank, but no future trouble was disclosed. About half of the piles
for the first tank had been driven when a serious obstruction was encoun-
tered about 12 to 15 ft. below the surface of the fill. After unsuccessful
attempts to break through the obstruction by driving down the mandrel,
the engineer ordered that the rest of the piles be driven around the edge
of the tank and thence the driving was to be conducted so as to work in
toward the obstruction. The mat was then to be bridged over the obstruc-
tion. As it turned out, the area of the latter was too large to make this
remedy practicable, but this was not realized until the surrounding piles
396 PILE FOUNDATIONS [Chap. 10

had been driven. It was finally necessary to dig out the obstructing
materials. This proved to be "a slow and costly job because of the
restricted space. The obstruction proved to be a lot of pieces of slabs of
reinforced concrete that had been dumped there from a barge when
someone demolished an old structure, and then they had been forgotten.
After this, the engineer made careful explorations under the other tank
sites. More obstructions were encountered under one of them, but these
were dug out immediately and easily while they were accessible.

FIG. 10-19. Raymond step-taper piles for the new boiler house at English Station of the United
Illuminating Company, New Haven, Conn. Notice the filled piles, the splice in the shell in the fore-
ground, the steam hammer driving the one at the right, and the timber trestle on wooden piles for
the support of the pile driver. The wooden piles will be cut off and left in place. The soil is approxi-
mately 30 ft. of silt over fine and medium sand. (Designed by Westcott and Mopes, Inc., Consulting
Engineers, New Haven, Conn.)

10-12. Piles vs. mats and spread footings. In engineering practice

there are often situations in which there is a question as to whether piles
are needed or whether the use of a mat or spread footings will be sufficient.
Much may depend upon the correct answer. Each problem should be
studied by itself. What is to be built? On what is it to be supported?
How is it to be constructed?
One previously mentioned but useful concept for studying such prob-
lems is that of the equivalent uniform loading and its probable effects.
Art. 10-12] PILES VS. MATS AND SPREAD FOOTINGS 397

For example, assume that nine wooden piles holding 20 tons each are to be
used in rows 3 ft. 6 in. c.c. both ways under a footing. This is equivalent
to a load of 40,000/3.5 = 3,300 p.s.f.
2
If the soil can hold this locally

without squeezing out, the strata below may be able to do likewise

because of the distribution of the load by spreading from the footing. If
the top stratum is not safe, then piles may be needed to transfer the forces
down to a plane of suitable bearing value.
In order to illustrate what unit loads mean, assume that the footing
referred to above is 10 ft. square with a load of 9 X 40 = 360 kips. Just
below the footing this is equivalent to placing an imaginary column of
earth 10 ft. square and about 36
high on the soil. At 2 1 assumed ft. :

distribution, the loaded area 10 below the footing is 20 ft. square, and
ft.

the imaginary superimposed load is a column of earth 20 ft. square and

~l Floor
k
ISO DL ZOO hDL
40k U 80 k LL
? Ordinary max.
<ISO
kDL ,200 kDL it
wafer fab/e=
h60 kLL it/SkLL a
£/. 206±; min
£l50 kDL 6200 kDL D Ei. 198+
60 k LL U5kLL

<ISO kDL ,200 k DL

® J40 k LL J80 kLL a

e S&22'0"=//0'0"
@ w /40 ; '•
:

Deep fine sand

A ^J ; -

(a)-PlMi OF COLUMN LOCATIONS <W- SECTION A-A AND SOIL PROFILE

FIG. 10-20. Soil and superstructure data for small office building.

9 ft. high. Similarly, at a depth of 20 ft., the fictitious applied load is a

mass of earth 30 ft. square and 4 ft. high. This analogy may help to
visualize what may happen to a structure.
How one to start the solution of these problems? For example,
is

assume that the owners of an industrial plant alongside tidewater wish to

construct the small fireproof office building, shown in Fig. 10-20 (a). It
is to be founded upon the soils shown in Sketch (b). The average esti-
mated dead weight 670 p.s.f. on
of the structure, including the first floor, is
the area of the building; the maximum
probable live load, 300 p.s.f.
There is no basement because of the high watertable, and steam will be
piped in from the power plant. The building is to have a reinforced-
concrete frame and brick or masonry-block walls. The typical loads
for column footings are shown Sketch (a). in
Is a mat foundation advisable? In the first place, this seems to be a
light building. Nevertheless, its weight is equivalent to a mass of earth
approximately 10 ft. deep piled on the site. The top layer of silt and fine
sand might not be harmfully compressible, but the thick tapered stratum
398 PILE FOUNDATIONS [Chap. 10

of soft clay will consolidate considerably —

and unequally tilting the build-
ing toward the left in Fig. 10-20(6) and perhaps cracking it. This is
serious. Furthermore, other matters make a mat undesirable:
1. floor should be an integral part of the mat for economy.
The This
would require deep walls at the edges to get below the frost line, thus mak-
ing doubtful the bearing of the mat on the fill adjacent to these walls.
2. The bays are rather large for effective spreading of the column loads

over the intervening areas unless the mat is very thick or heavily ribbed.
3. There is so much load on the central columns that local sagging here

may crack the structure.

4. This cannot be a floating foundation without a basement, and there
is no necessity of building a basement for that purpose.
5. The structure is too valuable and too inflexible to put it on a mat of

doubtful stiffness.
Is it practicable to use spread footings under the interior columns with
footingsand grade beams or heavy foundation walls under the exterior
columns? No, for the following reasons:
1. The localization of pressure on the clay is likely to be greater, with

larger differences in settlement.

2. The floor should be connected to the column footings rather than

independently resting upon fill, if uneven settling and cracking are to be

avoided.
3. With a bearing value of 2 or 3 k.s.f. on the soil, the footings must be
large anyway.
4. Again, the building should not be risked upon questionable supports.
If piles are then to be used, how should the construction be made?
Here are some suggestions:
1. The variable watertable makes wooden piles undesirable unless the
footings are very deep.
2. Poured-in-place concrete piles will be durable. Their lengths can be
varied so that they penetrate a few feet into the deep layer of sand, but no
deeper than necessary.
3. The soils are not really fluid, but the use of permanent thin shells on
the piles seems to be desirable to make sure that the fine sand and silt will
not run in and weaken uncased concrete.
4. The should be reinforced beam-and-slab construction that
floor
delivers its loads to the piles under the columns.
5. Place the bottoms of the footings at approximately El. 204 for
exterior columns; 207 for interior columns.
6. Support outer walls on a deep grade beam between footings, below
frost line.
7. Use a combined piled footing for the pairs of interior columns.
Now assume that the small office building shown in Fig. 10-21 is under
Art. 10-12] PILES VS. MATS AND SPREAD FOOTINGS 399

consideration. This is somewhat smaller than the preceding one in plan,

but has a basement. The soils are also different.
it Assume that, includ-
ing the basement, the average dead load is now 750 p.s.f. and the live load

is 400 p.s.f. What should be done in this case?

Piles might be used here — wooden ones if desired. They should be
driven into the firm sand. A safe structure can be made in this manner
with piles under the outer walls and under the pairs of interior columns,
and with a self-supported basement floor.
Could exterior foundation walls and interior spread footings be used
safely? Perhaps so. At an allowable bearing value of 3 k.s.f., there
would have to be a footing about 12 by 18 ft. under each pair of central
columns. They are big. As for the pressure on the soft silt stratum
under one of the central footings, its intensity will be approximately
1.2 k.s.f.This ought not to be harmful to the silt because it is trapped
under a rather impervious layer, it is already consolidated under a pres-

Max wafer-
Afax. W"ier
WSkJ*—
^^
n , Fl in? n -

iBI-99 EI.99 ,t*bleEI.98

J— C
<

^r^Silty clay
W-SECTION A-A

£167- sand F/ne

3r£Tr Medium clay

EI.54-
Deep fine to
medium sand
(W-SOIL PROFILE
foJ-PLAN OF COLUMNS

FIG. 10-21. Data pertaining to planning of foundation of a small office building.

sure of some 1.9 k.s.f., and it is thin so that the magnitude of the future
consolidation should not be large. Nevertheless, there will also be a
tendency for the clay to consolidate near the center of the building. The
basement walls and their footings can easily distribute the loads from
the exterior columns.
On the other hand, this site is low and flat. At times, the watertable
will be above the top of the clay, and surface water will almost inevitably
seep down around and under the basement walls. If the basement is to
be kept dry and since French drains outside of the walls are assumed to
have no reliable and accessible outlet into which they can empty, in this
case, the construction should be made watertight. If so, the water pres-
sure under the basement floor may be 300 p.s.f. or more. The floor should
therefore be a structural waterproofed slab capable of resisting this pres-
sure. This applies equally to the use of pile foundations.
To illustrate, assume that ordinary walls and footings are to be used and
that they are to be constructed first. The waterproofing and floor slab
 '

400 PILE FOUNDATIONS [Chap. 10

are then to be built as shown in Fig. 10-22 (a). It is obvious that water
under pressure may leak through the construction joint from A, or under
B and D. The protection layer C will not stop the water. The pressure
will tend to Lift the slab that is why the keys are used at B and D.
; These,
however, are not very reliable as tight joints, especially around the
columns or piers, and the slab will be weak in bending and shear. Dowels
to tie the slab to the footings are undesirable because they will interfere

WP = 2- ply membrane waterproofing

Wall Column^
WP-A. or pier

A H Floor WP D
EM
?o.;».;V:.'.r-;
\&P *C »:.•:!>';

::'•»•:'•.?'::»

Fooling
Footing
(a)-WALL AND COLUMN FOOTING WITH FLOOR ABOVE THEM

m Exterior
Swll
Column r-^_
WP_.\ '
orpier-

m
>.-t

Footing WP to
'

Footing
(b)-WLl AND COLUMN FOOTINGS POURED WITH FLOOR

Protection of WP
Column-A
Exterior wail
fa
WPAikW* Pedes
Pedestal-'.
-fa,
) h
rMmH
I

S\\ l l Floors

W> v//.v//"WW
\
'

W VSA
-WP [
WP
(W-FOOTINGS AND FLOOR COMBINED AS A MAT
FIG. 10-22. Alternative construction schemes for the foundation of a building.

with the membrane. This construction is not desirable to resist real

uplift.
Next, assume that the base course E in Fig. 10-22(6) is placed first and
that recesses are made for the footings. The waterproofing is laid over
the bottom to the edges F. Then the footings and floor slab are poured
monolithically. The shear istransmitted from the slab to the footings
at the sections G and H. The trouble with this is that the entire floor
must be poured before any loads are applied. It will therefore try to
act as a continuous mat, an action for which it is not designed.
Of course, this arrangement also brings heavy pressure on the water-
Art. 10-12] PILES VS. MATS AND SPREAD FOOTINGS 401

proofing under the footings. One scheme for avoiding this is shown in

Fig. 10-23. Here the loads are large, piles are necessary, and the hydro-
static uplift is Notice how the steel base plate is
small but not negligible.
set so that the waterproofing can be bonded to it. The upward reaction
from the slab is resisted by the column encasement; the downward floor
loads, by the top of the footing.
A glance at Fig. 10-21 (c) shows that the excavation for the basement
will be 5 to 6 ft. deep. The live load in the building may be looked upon
as a temporary load that will not cause serious consolidation of the silt

and clay if the latter does not flow locally. Since the dead load of the
structure practically equals the weight of the excavated ground, the

3"min. encasement

Steel Mesh
column l°i (2ply membrane waterproofing

9"floorslab

Billet ~-i^ Grout •

."-":
'. >
:

'~M ^"protection
ravel blanket

Poured-in -place
concrete piles

I ^01 I &- J

FIG. 10-23. Details of a nine-pile footing and waterproofed floor used at a multistory industrial plant.

structure will nearly be floating in the clay if the basement floor can make
it act like a boat. It makes no difference to this floor whether the upward
pressures are hydrostatic or are upward reactions from the earth. At
least, the floor must resist the former. If it can do this, it will, as stated
previously, try to act as a continuous mat in spite of the designer's
wishes. Therefore, it seems desirable to design the floor to resist these
forces, and to use construction like that of Fig. 10-22(c). In this scheme,
the projecting footing beyond / is eliminated, thus simplifying the water-
proofing problem. With
this arrangement and the utilization of the
"floating" foundation, seems that piles are unnecessary. Apparently,
it

this principle of a floating foundation was used successfully for the Loteria
Nacional Building in Mexico City. 1

Of course, if the water could be removed by trustworthy drainage

outside the walls, the ordinary walls and footings might be satisfactory
for this building. However, if the loads were considerably heavier, the
1
Engineering News-Record, Dec. 12, 1946, p. 95.
402 PILE FOUNDATIONS [Chap. 10

footings would become so large that piles would probably be advisable.

Consider an entirely different situation. An industrial company had
built some large oil tanks along the Northeastern seaboard. A large new
tank was later added to the tank farm. It was near the shore so that part
of the concrete mat rested upon original ground whereas the outer portion
was upon a few feet of fill that seemingly covered softer material formerly
the edge of a muddy flat. The steel walls were completed, and the tank
was with water to test for leakage. Suddenly one side of the tank
filled

failed. The water gushed down and over the surrounding dike and
flooded an adjoining pump house. Apparently, this short-circuited some
of the motors and caused explosions of gasoline that resulted in a very
disastrous fire, with loss of life as well as of property. There is some
disagreement regarding the cause of the failure, but it seems probable
that it was due to the distortion of the tall stiff wall produced by unequal
settlement of the portion on the filled area. This case is one in which the

Possible shear Possible failure

failure in tension
^3

Saq^ ^Possible rupture of Settlement

junction with bottom '

(a ) -SETTLEMENT IN FORM OF LOCAL SAG (^-SETTLEMENT OF ONE WHOLE SIDE

FIG. 10-24. Character of possible damage to an oil tank caused by uneven settlement.

mat cannot do any real spreading of loads. It is merely a sort of pave-

ment that necessarily accompanies any settlement of the ground under
it. The soft material should be removed and replaced with good fill.
Perhaps there are two lessons to be learned from the preceding illus-
tration :

1. One should obtain a

scale on the magnitudes of the total loads
applied to the soil. An
tank 150 ft. in diameter and about 30 ft. high
oil
may cause an average unit pressure of about 1,900 p.s.f. This seems
small. However, the total load is something like 33,500,000 lb., or
17,000 tons. What could this do to cause movement of weak cohesive
soil?
2. Small unequal settlements, or a small change in the angle of settle-

ment, may cause very severe strains and accompanying stresses in stiff
structures. The wall of the tank referred to is a sort of stiff vertical plate
girder in a curved position. Ordinarily it has tomerely the ring
resist
tension produced by the fluid pressure. However, an unequal settlement,
as shown in Fig. 10-24(a), may rupture the junction of the bottom and the
Art. 10-13] HORIZONTAL FORCES 403

wall, or it may deform the wall excessively. like thatA change in angle
in Fig. 10-24(6) is likely to tear the top apart.
Supposedly, this is what
caused failure in this instance. Therefore, when settlements of the vari-
ous parts of a structure are likely to be unequal and when they may cause
serious trouble, it is best to use piles or some other type of substructure
known to be trustworthy, but the structure should be considered as a
unit. Local hard spots may hurt a structure somewhat as a marble hurts
one's bare foot when he steps on itin the dark.
Cost estimates of alternate plans should be made before one decides
upon the solution for a specific problem. One should not jump to con-
clusions too quickly.
10-13. Vertical piles subjected to horizontal forces. There are
many special features that should be borne in mind when one plans pile
foundations. By necessity, the information herein is largely qualitative
because quantitive data depend upon the soils, piles, forces, and dimen-
sions involved.
If the pile shown in Fig. 10-25 (a) has a continuously applied horizontal
force H acting as in (6), and if the pile is surrounded by weak cohesive
soils or muck, tend to bend as shown to exaggerated scale in the
it will
latter sketch. The pressure diagram may be of the character shown in
(c). The resisting pressure may vary somewhat as the horizontal deflec-
tion of the curved pile from its original vertical axis and as affected by
the depth to any point in question. The passive pressure acting on the
portion AD may vary with depth so that the resisting pressure diagram
may be somewhat like the bulging line ANDQB, where CA represents
themaximum displacement. The pile will bear against the soil on the
left side A to some point D. Below D a small pressure on the right
from
will resist the tendency of the pile to rotate counterclockwise and to curve.
If the soil is surrounding the full length of the pile, it can resist small
lateral loads, but not if the ground is practically fluid. But when the pile
projects far above the soil, as in Fig. 10-25(d), the resistance of the soil
and the bending strength of the pile may become important.
In actual practice, the lateral force will be resisted as efficiently as
possible, i.e., as near the top as the strength and stiffness of the soil and
pile permit. Quantitative values of pressures and bending moments
are difficult to ascertain. Cylindrical reinforced concrete piles and steel
H piles have uniform stiffness along their lengths whereas this property
varies in tapered piles. If the pile is embedded in fine sand or firmer
soils, the point of maximum bending may be somewhere near 5 ft. below
the top of the soil; if in soft clay, 10 ft. or more below; if in mud or muck,
a few feet below the top of the highest stratum that is soft clay or better.
For a single pile, the abutting resistance of the soil may be much greater
than that caused by a strip of earth equal to the width of the pile. The
 —

404 PILE FOUNDATIONS [Chap. 10

pressure may spread through the ground as indicated by GAJ of Fig.

10-25 (e) if the soil is not too weak, causing a very considerable resistance
to finite displacement.
On the other hand, the pressure from a row of piles AA h Sketch (/),
will spread sidewise through no more than some volume represented by
KAAiL, whereas a large group will not affect much more soil than does
the single row, as pictured by K' A-iAJJ '
'.
This shows that the results
of a lateral pulling test on one pile may not yield reliable data for the
resistance of a group of piles unless the spacing is large or the soil is fairly
firm or almost fluid near the top. It is advisable to assume that the
resisting mass of earth for a closely spaced group of piles in reasonably
resistant soil is a vertical portion not more than 6 to 10 ft. wider than the

K'
AH K o i

o o o o o
G
"o o o o o
'JT
o o o o o -H
Spread assumed
o o o o o
reJ-SPREADING
OF LATERAL cxooo *A*
PRESSURE r /
IN SOIL
L'
ta;-PILE BEFORE (bhWL BENT fc/- PROBABLE ^-PRESSURE ffJ-SPREADING OF
LOADING BY LOAD TYPE OF DIAGRAM PRESSURE ON PILE
PRESSURE FOR PILE GROUP INTO SOIL
DIAGRAM IN PROJECTING
COHESIVE SOILS ABOVE SOIL

FIG. 10-25. Vertical piles subjected to horizontal loads at their tops.

distance center to center of piles across the group perpendicular to the

active horizontal force AA of Fig. 10-25(/). The abutting power of
X

this earth may

then be estimated. The upper 10 to 15 ft. should be able
to resist the applied force H.
No great accuracy should be expected from such computations of lateral
resistance. One should be very conservative in design. Seldom will
the situation be dangerous when the pile is fully embedded in soil of
moderate stiffness. On the other hand, vertical piles should not have per-
manent horizontal forces applied to them when there is any doubt of the
resistance of the surrounding soil. Temporary wind and live-load shears
of smallmagnitudes are usually harmless if the soil is at least as good as
moderately compacted silt or rather soft clay.
Groups of piles may be lashed together, as in the case of dolphins, to
resist large horizontal forces. If the piles can slip past each other, the
Art. 10-13] HORIZONTAL FORCES 405

longitudinal shearing resistances will be inadequate to make the group

act as a large single member;
be only as good in bending as the
it will
sum of the strength of the individual piles. Furthermore, the whole
group will be no better than the lateral resistance of the soil in which the
piles are driven.
Vertical piles that are braced together to form towers or bents may
have considerable lateral strength initially. However, such bracing
members and their connections are often difficult to install and are not
likely to be permanent unless they are well protected and maintained.
Now refer again to the isolated footing shown in Figs. 10-5 and 10-7.
There is a real question as to how the horizontal force S = 40 kips will
be resisted and where the center of moments for overturning should be
assumed to be located. Probably no exact solution can be found, but
there should be some way to estimate these things. The following is
suggested as one way to handle the problem, referring to Fig. 10-25A
1. Assume that
the 1 ft. of soil over the 2-ft. footing will be held by
friction so that can be assumed to be an integral part of the 9-ft.-wide
it

footing itself. As the structure tends to move toward the left, the passive
pressure can be assumed to act against a surface 9 ft. wide and 3 ft. high.
2. If the clay is such that the passive pressure can be assumed to

equal 250 p.s.f. per ft. of depth, the soil alongside the footing may'
1

develop a passive resistance of

E x = 3 X 250 X 1.5 X 9 = 10,100 lb.

located as pictured in Sketch (a).

3. Because of the spreading action of the soil around and in front of

the 12 piles, as illustrated in Fig. 10-25(,f), the group can be assumed to
be at least equivalent to a "wall" as wide 2 as the footing, and this will be
assumed to bear against the soil. If so, the theoretical pressure diagram
for resistance to horizontal displacement may be assumed to be bounded
by lines EB and EX, Sketch (6).
4. Assume that the piles tend to bend as shown in Fig. 10-25(6) but
that only approximately the top 25 to 33 per cent of the embedded pile
length Lp is effective in pushing sideways against the soil.
3
Then assume
that the theoretical passive pressure varies from zero at D to the com-
1
If the soil is below the watertable, the effect of buoyancy must be considered in
reducing the unit weight.
2
It may be safe to assume that the tangent of the angle AA 2 K' of Fig. 10-25(/)
should not exceed the following: 1 q for soft clays and silts, 4 for medium clays,
J
for ^
stiff clays and fine sands, l^ for coarse granular materials.
3
For most cases, this length should not be assumed to be much greater than 10 to
15 ft.
406 PILE FOUNDATIONS [Chap. 10

(c) id)

200*
mufsesA jgrnss^

^5;
OS 1)

Section /-/

Iff) If)

FIG. 10-25A. A study of lateral resistance of piles.

Art. 10-13] HORIZONTAL FORCES 407

puted value at CA, Sketch (c), as follows:

Top of pile: CA = full computed value at point A

Top = 0.75LF
quarter point: IF
Middle quarter point: JG = 0.50MG
Bottom quarter point: KH = 0.25NH
Bottom point: D: pressure =
Compute the total magnitude of the pressure diagram for a vertical slice
1 ft. wide and find the location of the resultant, as pictured in (d), assum-
ing the top 10 ft. of the piles to be helpful in this case. Then the maxi-
mum possible passive resistance is about 9,800 lb. per ft. of width applied
3.28 ft. below the bottom of the footing. If the top 15 ft. of the pile is
assumed to be bearing on the soil (length AD), the similarly computed
force H
2 = 16,900 lb. and is located approximately 5.28 ft. below the
bottom of the footing.
5. Next, referring to Sketch (d), assume that the line of action of H 2

is the fulcrum about which the footing and piles tend to rotate. On this
assumption, the moment to be used for computing the overturning pres-
sures on the piles might be assumed to be M = *S(4.5 + 3.28) = 40 X
7.78 ft. -kips instead of the 40 X 8.5 assumed in Art. 10-6 for the footing
shown in Fig. 10-7 (a). For AD = 15 ft., the lever arm would be 10.22 ft.

A lever arm of about 8 ft.seems to be acceptable.

6. Since S/9 = 4.44 kips per ft. of width, the abutting resistance H 2

of Fig. 10-25 A (d) is larger than necessary. However, assume that the
resistance Ri in Sketch (e) is as far below D as H 2 is above it. Then the
required magnitude of the passive pressure for equilibrium is *

„ SL = X 4.44 X 21.22
=
#2 = ~jj~ 7 kips (approx)
J344
The computed safety factor is

S.F. = 9.8/7 = 1.4

7. When the passive pressure against the footing as computed for Hh

Sketch (a), is sufficient to seems to be desirable
counteract the force S, it

to take moments about the tops of the piles when computing the vertical
pressure on the piles. When Hi is not sufficient, assume the lever arm
of S to be located as computed in item 5 above.
8. For single piles or poles, one might estimate the resisting moment

due to embedment in a somewhat similar manner, using 4D as the top

one-half to two-thirds or three-quarters of the embedded length when the
member is and strong, where A is the surface of the
relatively stiff

ground and where the assumed resisting width is twice the diameter of
the pole. For example, assume the pole shown in Sketch (/). If the
408 PILE FOUNDATIONS [Chap. 10

passive pressure is 300 p.s.f. per ft. of depth and the bottom pressure
diagram is triangular, H 2 = 1,500 lb. (approx) and the safety factor is

X4 =
1,500
200X22
As far as the footing of Fig. 10-7 is concerned, and as a result of the
preceding computations and when the soils are too weak, it seems to be
advisable to provide some better way of resisting the large force S.
Batter piles might be used as explained in the next article.
Figure 10-25B is a nomograph which is very convenient for use in
1

determining the required embedment for poles to hold a given load at a

stated height. based upon an assumed yielding of
This is in. at the ^
top of the ground. It is suitable for designing
telephone poles, billboard
supports, and similar structures where appreciable movement is not
harmful. It yields much larger safe loads than does the method described
in connection with Fig. 10-25^4 (/), which is intended for use in the design
of structures where practically no yielding is permissible.
To use Fig. 10-255, try the following example, as stated by Kinney:
P = 3,000 lb., H
= 20 ft., soil = average (2,500 p.s.f. allowable), and
width b = 14 in.
1. Draw a line from *Si = 2,500 through P = 3,000 and extend to find

C = 1.2.

2. From C = 1.2, draw a line through b = 14 in. to L, the line of depth

coefficients (L= 1.0 in this case).
3. From L = 1.0, draw a horizontal line to the curve for = 20 ft., H
then draw vertically downward to the bottom of the diagram and find
the required depth of embedment, 8.5 ft. in this instance.
If the horizontal forces acting on a group of piles are caused by the tend-
ency of the surrounding soil itself to move laterally, the piles can offer
little or no resistance since they depend upon the supporting soil for their

own strength. If the piles go through a deep layer of unstable soil into a
stable sand stratum, they will usually tilt or be broken by the deformation
caused by any lateral movement of the top layer.
A good example of the fact that piles embedded in a deep plastic soil
will merely move with that soil is the case illustrated in Fig. 10-26.
When this viaduct was planned, it was considered desirable to terminate
the superstructure near the flare in the roadways shown. This required
an extensive fill a little over 30 ft. deep above the original ground, but it
would avoid the use of complicated and costly structures. The soil was a
layer of peat 8 to 10 ft. thick over a very deep stratum of fairly soft clay.
1
Edwin E. Kinney, Correct Embedment for Pole Structures, Wood Preserving
News, October, 1959. (Chart prepared for the Outdoor Advertising Association of
America, Inc. by P. C. Rutledge.)
 CO
,. Q Dm w =
£2 < o
co O
-
g m
O 1 w 5
o:
° n
^ 1

§ 5 a
- t " °
£ Z 3 of,
-2
5 I £ ? £ d
089V OZ£V m ° > S z "

SlzSo
h ° o S
llJ
H WbpH a wdao

m T3 V
>/, // y
O E // z^ /?/
7

+
=e •2
\
\
///
'//
//
//
a&£
7-Xtw
0, S ii \ '//A V-lLT-l
\ Vj/a
n \^ Va& $$v
7
-/
£?/5 t
wj "^.C
\ -*%i, /////
///// '//// //
ZZ > ®-9 •*- ^
9 ? \J
R7/; 4?22 //
/

pwW7K'/ //// /
6 y)
y/,"/ y>
w4/A* ///,
(ft/,
y<y.
//// // /
'/, Vf 2V/V M/
/>
A//a <>/, 'A
^
V* Sa/* 4?k
^<>// /*
//A <y. s, s,
^ '/< '/A
v/. <y. fs
/>
/
y/, // y
// st

*/,
4U3IDIJJ30D
'
M+dap = 7
o odd —

(S3U,0U()

psppsqws
jsod jo U0|;39s
jo tupiM = q
ol'TTT '

m io

A+n iqo+s jsod ^, r I

'
I
'
I
'
I

JO JU3p!JJ300 = J O O 0"> 00 r-
— d ' 6 ci

spunod u| jsod
uo pooi WnwiXDW = j
ro cvj eg —

Ul
'44.~bs J3d-q|
SS9J(S I0S |
^r
36DJ9AD 3|qDM0||D - ^
I/OS //OS l/os //"f
1
PJDI/ Aj3fy P00Q 3bOJ3A\f VOOc/ //0J Jjg/l
I'l 'I
SAop PUD S||IS JOJ o oo o o o
oo o o o
O
m
CO S- U3
<T>

|T -
S||0S A||3ADJ6 ' '

O
I

oo o o
I I

PUD ApuDS JOJ

o o
o en o o
oo N

J96nD JO|D0|pu! j3i9LUDip ^/,-l uo spunod Ul ||n,J

409
410 PILE FOUNDATIONS [Chap. 10

The main portion viaduct consisted of a series of rigid-frame units

of the

like EH in Fig. 10-26, with hinged suspended spans between them.

Toward the abutment, however, the structural character but not the
appearance of the viaduct was changed, as shown in (b). This permitted
movement of the abutment A and footing D to occur without injury to

the girders because of the roller at A and the hinges at B, C, D, and E.

The peat was first removed from the area to be covered with the fill.
The latter was then placed and compacted. Steel pipe piles filled with
concrete were installed and capped by the concrete footing and bridge
seat of the abutment. These piles were used merely to ensure that

XSidewalk

(Sidewalk

faj-PLAN OF ROADWAY

(Abutment
Embankment
Direction of
V settlement

rtW-SIDE ELEVATION AND SOIL PROFILE

FIG. 10-26. A viaduct planned for anticipated settlement of abutment.

future consolidation of the would not cause settlement of the

fill itself

abutment, However, the clay was expected to settle and possibly to

flow laterally. The backwall of the abutment was made of bolted
timbers so that it could be rebuilt later if necessary. The bearing at A
was a geared roller on an excessively wide seat with provisions for jacking
up the girders and raising the bearing surface. Jacking and adjustment
were also provided for at D.
After about three years, the abutment at A had settled approximately
16 in. downward and a few inches backward. Of course, the piles were
carried with the clay in which they were embedded. As expected, the
greatest settlement occurred under the greatest mass of the fill, hence the
backward rotation of the top. Some movement of the clay toward D

also occurred. The first revision of the abutment was made as intended,
and the roadway on the fill was raised. Undoubtedly, other alterations
Art. 10-14] BATTER PILES 411

will be required in the future, but the magnitude of the settlement per
year should decrease progressively.
10-14. Batter piles. When large horizontal forces are or may be
applied to the tops of piles, the use of batter piles is generally desirable.
If driven to the same horizontal level as the tips of vertical piles, batter
piles may be relied upon to resist the same longitudinal load as the verti-
cal ones, provided, of course, the soil conditions are the same.
Assume that shows a cross section through a reinforced-
Fig. 10-27 (a)
concrete retaining wall or bulkhead supporting the edge of a yard area at
an industrial plant near the water's edge. The horizontal component H
of the earth pressure tends to tip the wall about pile a and cause it to
slide toward the left. If piles are used as shown in Sketches (a) and (b),

horizontal movement of the wall will cause the vertical piles to bend or tilt
leftward whereas piles d, e, and / will tend to be forced farther into the
sand or to tip about their bottom ends. It is obvious that only the
horizontal component of the resistance of the batter piles to penetration
should be relied upon to counteract the thrust H.
When using a combination of vertical and batter piles, one should
remember that the structure will try to act as a unit. One should not
assume that the vertical piles resist all the vertical loads whereas the
batter piles resist all the horizontal forces. Referring to Fig. 10-27 (a)
again, it may be that the applied loads cause piles c and / to produce a

small upward vertical or even a downward reaction; piles b and e, an

upward resistance; piles a and d, a large upward reaction. In each case,
the forces will be assumed to be parallel to the longitudinal axis of the
pile. It is obvious that the batter piles are the effective ones in resisting
H, but they are not unaffected by the vertical load W.
The method used to compute the forces upon batter piles should be
simple, logical, and easy to apply. Its results should be reasonably near
reality. Tedious computation amidst so many uncertainties is not neces-
sary, nor are its results more accurate than the assumptions of loads,
soil pressures, soil resistances, etc., upon which it is started. The follow-
ing procedure is therefore suggested:
1. By approximations or guesses obtain a trial design for the structure.
Give all dimensions and assumed loads.
2. Compute the center of gravity and moment of inertia of the pile
group at the tops of the piles, or a typical portion of it, as though all piles

were vertical.
3. Compute the overturning moment and eccentricity of load at the
plane of the tops of the piles.
4. Compute the vertical load on each pile.

5. Draw a force diagram showing the horizontal force and the com-
puted vertical loads on the piles.
412 PILE FOUNDATIONS [Chap. 10

6. Draw the hypotenuse of each right triangle of which the computed

vertical force on a batter pile is the vertical component, sloping it at the
assumed batter of the piles. If these, when drawn in the force polygon,
close the polygon, the computations may be assumed to check. If the

batter of the piles not assumed, the slope needed to close the force
is

polygon may be assumed to be the desired batter.

•>>\v// s \y/AW/\>.<.

F/n

°
U'ii6
4- & e.g.
(b)- PLAN OF TYPICAL
,,_ ,_?, \
?o" f %'o"f 1 SECTION OF FOOTING

faJ-TYPICAL SECTION
THROUGH WALL

58 * unbalanced
p I

Re/feving
p/afform

Toe wa//
rC ;-FORCE ftf J-SECOND FORCE and
DIAGRAM DIAGRAM embankment

feJ-ALTERNATE
CONSTRUCTION

FIG. 10-27. Analysis of loads on batter piles.

Art. 10-14] BATTER PILES 413

Apply this to Fig. 10-27 as an illustration. Since the pile rows are
assumed to be 6 ft. c.c, assume a slice of wall 12 ft. wide as a typical case.
If the loads and their positions are as shown in (a), H = 12 X 7.4 =
89 kips and W = X =
252 kips. Neglecting the moments of iner-
12 21
tia of the individual piles about their own centers,

I = 4 x 52 = 100 pile-ft, 2

The eccentricity of the load is

e = 3.25 + 7 X 8
H 52 - 5 = 0.72 ft, left of pile b
M = 252 X 0.72 = 181 ft. -kips

252 181 X—o = ni-

pa = pa = —n~ H Tqq 42 + 9 = 51 kips

= =
252 , 181 X = lo
42
.
kips
.

pb Pe -g- H
Jqq—
= pf =
252 181 X 5
= ,
- _
=
._ . .

Pc -g j^— 42 9 33 kips

The force diagram is drawn in Fig. 10-27 (c). Line JP represents the
inclined reactions of the batter piles since it is drawn parallel to them.
The magnitudes are scaled. Line PC represents the portion of H that is
not resisted by the horizontal components of the batter piles. Some of
this might be resisted by bending in all six piles if it is withstood at all.
Since the silt around the piles will tend to be forced out by the pressure of
the fill behind the wall, it is unwise to place any dependence upon the silt
to support the piles against lateral bending. Therefore, something else
should be done.
Line JC of Fig. 10-27 (c) shows the batter needed by the three piles if
they are to resist the entire thrust H. On the other hand, EC shows the
slope required if all six piles are battered, and EQ gives the resultant
thrust in pile a. However, if all piles are battered, the weight of the wall
alone before the fill is placed will cause the structure to tend to tip back-
ward and bend the piles. This silt may be strong enough to resist this
before loaded otherwise, but such a condition is not generally desirable.
it is

Now suppose that all six piles of Fig. 10-27(6) were driven at a batter
of 2.5:1. Incidentally, a slope of 2 : 1 is a practicable limiting angle for
driving. Line EU of the force diagram in Sketch (d) represents the
reactions of these piles. The distance CU
a theoretical excess of pile
is

thrusts above the value of H. This, however, is not harmful. All it does
isto tend to push the wall against the fill until enough passive pressure is
developed to cause equilibrium. Therefore, if the piles can resist safely
the maximum active pressures for any combinations of loading, a smaller
active load will merely cause the development of enough of the tremendous
potential passive resistance to produce equilibrium.
414 PILE FOUNDATIONS [Chap. 10

Increasing the weight of the wall will merely require more piles. Some-
thing has to resist H. It therefore seems best to modify the construction
by using a relieving platform, a heavy concrete slab, or a smaller wall and
sloping embankment, as indicated in Fig. 10-27 (e).
Another way to look at this problem is to assume that the structure in
Fig. 10-27 (a) will move toward the left and tend to rotate about the tips
of the batter piles until equilibrium is attained by a shifting of the vertical
loads from the vertical piles to the batter piles, or even producing an
actual upward pull on the vertical piles. The relief of vertical load can
usually be considered effective, but in order to have the piles actually
resist uplift,they must be physically attached to the footing a difficult —
thing to do with wooden piles. If this shifting of load is to be resorted to,
Fig. 10-27 (d) shows that, in order to have enough horizontal components
to resist H, the batter piles must support almost the entire structure.
This is not efficient use of piles because the vertical ones are supporting
little or nothing, whereas the batter piles will be overloaded. This seems
to indicate that one should use more batter piles and only enough vertical
ones to hold up the structure before the fill is placed.
As another illustration, assume the bridge pier shown partly in Fig.
10-28. Sketch (a) shows a typical row of piles; (6), a plan at one end.
Assume that all outer side piles are battered as shown, that the four at
each end are battered transversely, and that all others are vertical. For
a typical row, W
= 290 kips, M
= 230 ft. -kips, and a shear S at the
bottom of the footing = 5 kips. Then,

/ = 2(1. 2
+ 4.5 2 + 7.5 2 ) = 158 pile-ft, 2

p a = -»b
290
— ,

—
2 30 X 7.5 =
r^
loo
48.3 + 10.9 = .....
59 kips

t™ — =
290 X
230 4.5 ._ _ _ _ ._
+ = , .

ph = +
, ,

-7T- 48.3 6.5 55 kips

290
p c = -g- H
.230 X
—= Ysg
1.5 .__
48 3 -
+ 2OQ2 =
.

-
K1 kips
51
..

Pd = 46 kips pe = 42 kips pf = 37 kips

The force diagram is shown in Fig. 10-28(c), where CO is the sum of the
computed vertical components. Lines CJ, JK, KL, LM, MN, and NP
represent the reactions of the piles. Line PA is negligible. If it had
appreciable magnitude, it would indicate a desired change in the batter.
However, small amounts of unbalanced shear could be resisted by the
piles in bending or by lateral bearing on the soil.
The end row of piles can be treated the same in the analysis as any
interior row. The reactions on the four central batter piles will be
slightly larger than the computed vertical components. For this loading
Art. 10-14] BATTER PILES 415

case the piles at the two ends counteract each other's horizontal com-
ponents so that no harm results. In some cases, it may be advisable to
use more batter piles sloping transversely with respect to the bridge.
As a third illustration, assume that Fig. 10-29(a) pictures a typical row
of piles under the footing of a rigid-frame rib in a hangar. The concrete
ends at ABC. Extend AB to D. The intersections of this plane with the

—— U
Footing
— ?
U-
1
N VH
rft :
'\'n i.'h ''fi

VA- /:*

S@3-0=/S-0' b
Ta;;-VERTICAL SECTION M<
fXT
S 2> 9> £> $ -

Ep q p p p
u u u u
— a:
d
K Kt-E
W=290k

(d)-PLAU NEAR ONE END D

f
1
CiB

Ccj-FORCE DIAGRAM

FIG. 10-28. Piles supporting a bridge pier.

centers of piles a and b are shown. Assume them to be in these positions.

Then, about the center of the group,

M = 60 X 6 - 200 X 2.5 = - 140 ft.-kips

/ = 2(2 2 + 5 2) = 58 pile-ft, 2

Pa
200
-j-
140
^X—5
58
= _
50n — 10
12 = „
fiV
38 kips

1 ±f\ y o
Vb 50 - Kr - = 50 - 5 = 45 kips
58
P< = 50 + 5 = 55 kips and Va 50 + 12 = 62 kips

The force diagram is then constructed in Sketch (b). Line QG shows

that approximately 44 kips of the thrust H
remain to be counteracted.
Piles a andmight have their slope increased to match line MG if this is
b
not too steep for practicable driving. The footing might be redesigned to
set W back farther, thus increasing the forces and horizontal components
in piles a and b. However, this would avail little unless pile c were bat-
416 PILE FOUNDATIONS [Chap. 10

tered also. Because of variations in live load and wind reactions, it seems
desirable to keep pile d vertical. Of course, passive pressure behind the
footing would help also, but a footing for a rigid frame should be designed
so that it is entirely trustworthy. Perhaps the best solution here is to
change the type of superstructure so as to reduce H.

JV=200*

faAFOOTINGAND ACTIVE FORCES <W-F0RCE DIAGRAM

FIG. 10-29. Bearing of a rigid frame.

10-15. Piers. Many piers are founded upon piles; in fact, most of
them are. Piers (and many bulkheads) are not so much a problem of
pile foundations as they are of pile structures. Here the piles are pri-
marily stilts down through air and
or columns to reach from the pier deck
water to the underlying soils. That these soils are often silt, clay, and
fine sand is to be expected because of the geology and topography of the
sites. Where bare rock or a thin gravel stratum over rock exists at a
proposed pier, some construction other than piles alone is likely to be
more economical, and also desirable to attain lateral stiffness.
The live loads on pier decks are likely to be heavy in terms of pounds
per square foot —
500 to 1000 p.s.f Columns of double-deck transit sheds,
.

warehouses on the piers, truck passageways, and railroad tracks usually

require strong pile foundations. The planning and design of such struc-
tural foundations are similar to others except for the fact that the piles are
more like long columns, may be subjected to considerable lateral bend-
ing, and are more exposed to disintegration.
The problems of pier construction are largely those of practical struc-
tural framing rather than of the attainment of adequate support. Steel
piles may be useful when the piles must be very long; creosoted wooden
piles, when flexibility against lateral impacts from ships is necessary; pre-
cast and prestressed concrete piles, when strong column and beam action
 — —

Art. 10-15] PIERS 417

are essential but the lengths are not too great. When rock is easily acces-
sible and a heavy structure is a part of the pier, may be advisable to use
it

drilled-in caissons with steel H's, large pipe piles rilled with concrete, or
1

even a number of 4- or 5-ft. caissons placed under the columns and resting
on the rock.

MHW Sump and

MLW Open
pumps if cells
needed

'•m'l^'/i-wmw I pHYtCivI — — '^wf

I

| to^/A'-I-'^ I — I

—
r7TTr7T"7i I '^imwm
-Batter piles
in between
vertical piles

FIG. 10-30. Illustration of the principle of using buoyancy to reduce the number of piles required to
support a heavy pier.

Recesses for Curb except^ Column of light

railroad tracks 'transit shea'
r at doorways'
-i

Vo'to/2"
-ev. Bolts in pipe
\2 L0"fo2 !6"
2-0 i
| sleeves for
j>
Stringer
timber guard "-\v--y

\2'6x2-'6%
\\

TX T? \3'-0"x3 L0"

'Fit
Encasement TAJ
needed YJ Batter
Steel piles shown.
'Precast- concrete
Cap 20-0 c.c.

pile close piles useful also if

'75~'±)
to cap not too long I

FIG. 10-31. A poured-concrete deck for pier on long steel piles.

There are so many special details of construction used for piers that
minute descriptions of them here seem unwarranted. Some are illus-
trated by the accompanying drawings. Local conditions, the materials
available, the desired permanance of the structure, and the loads to be
supported are likely to influence the choice of the general type of pier
construction. Piers of prestressed precast concrete piles and con-
made
crete decks are especially worthy of consideration because of durability,
fire resistance, low maintenance, and long-term economy. Precast mem-
bers in the decks may also save considerable in the cost of forms.

See Fig. 12-13(c).

418 PILE FOUNDATIONS [Chap. 10

A pier should be stable against horizontal forces. These forces may be

caused by wind, waves, ships, ice, or floating debris. Bracing of the
structure near the tops of the piles may be sufficient when the latter are
firmly embedded in deep stiff soil. In other cases, batter piles trans-

Bronze anchor bolts fo ho/el guard Bituminous pavement

I

f %^*
f Poured fascia girder ~\

<l;-...'.'.

T
J Mortar key

'

mi
Key and dowel

Precasi units Pour- <W/der cap

line at rows of
baiter piles

t
'Poured cap A-*—

Precast-concrete pi/e
w u

il Batter piles in
symme trical
arranqemen t
faJ-PARTIAL TRANSVERSE SECTION (b)- SECTION A-A

FIG. 10-32. Suggestion for reinforced-concrete pier with precast deck units.

ra

FIG. 10-33. Some construction for a structural-steel pier with concrete deck. (Double pile bents to
be used at transverse expansion joints. Cathodic protection is desirable.)

versely and longitudinally may accomplish this. They may be placed

between or next to the rows of vertical piles. In any case, they should
be attached firmly and arranged so that they do not cause unbalanced
horizontal thrusts against the deck when vertical loads are applied.
Art. 10-16] BULKHEADS ON PILES 419

10-16. Bulkheads on piles. The problem of earth pressures against

high bulkhead walls has already been somewhat in connection illustrated
with Fig. Heavy
10-27. masonry may be suitable on rock, hardpan,
gravel, or even on firm and coarse sand but not on weak cohesive soils
where piles are needed. In these last cases, the relieving-platform type
of construction is useful.

Bolts in pipe sleeves

Composite floor-2x4 "and 2"x6''
'Laminated timber floo, '
Shear lock

faJ-TIMBER STRUCTURE AND DECK ^-TIMBER STRUCTURE WITH COMPOSITE DECK

FIG. 10-34. Some construction for wooden piers.

Pipe sleeves r!8 to 24 ffat-slab deck Double rows ofpiles at expansion joints 100 i c.c.

\y° '
'

' KJcap/w'^F^'
I
]
'

X7 1

/
,
/
xf / D/f,
'
'

~4
\y] I

Pilesare pipes
/ monotubes
or other thin
shells filled
with concrete,
or precast Riprap placea
concrete after pi/es
are driven

FIG. 10-35. A flat-slab relieving platform.

Several varieties of relieving platform are shown in the accompanying

illustrations. These are intended to show many different details, and
they may be combined in various ways. The use of the earth fill in Fig.
10-36 requires more piles per square foot to support the vertical load than
do Figs. 10-35 and 10-37, but this type is convenient when utilities and
railroad tracks are to be placed in or on the fill. On the other hand, the
platform in Figs. 10-35 and 10-37 must be wider for a given freeboard and
420 PILE FOUNDATIONS [Chap. 10

depth of water because the slope of the earth below starts almost at the
surface of the ground whereas, in Fig. 10-36, it starts from 10 to 15 ft.
below grade.
The batter piles are usually in rows placed between the vertical ones.

Bulkhead
wall, Bituminous Railroad
/'pavement { tracks .Grave/

"
>tj| Earth fit/ or stones and grave/
ii _
0j
°
yflkvMyM'/f-Avs, /'rlat-s/ao deck ^mv/z-m, .......

r«;-CONCRETE-AND-EARTH-FILLED CONSTRUCTION

Stone-masonry wall
3 y and facing
^Bituminous pavement or turf
4.'>'J!).*W>.'>'*V.'J.!
-

Wop soil if a park area

of gravel and sand
Fill
with stones at bottom Concrete slab (may be
'omitted ifpermanently
TaG planking below water)
.w .'iwj.vlm^wilv/iwv't w v \\w,h\\ vn*.\&//t\v<7& szzsssmzmas

7
Wooden
cap
Wooden
piles

(b)- CRE0S0TED WOODEN PILES AND DECK WITH CONCRETE

AND EARTH COVER
FIG. 10-36. Relieving platforms with earth fill.

They should not project into the water in front of the bulkhead. There-
fore, their tops are near the back edge of the deck slab. In this construc-
tion, the batter piles are to resist any forces tending to move the structure
waterward; e.g., wind blowing against a ship tied to the wharf, and active
Art. 10-16] BULKHEADS ON PILES 421

earth pressure. The abutting resistance of the soil will prevent landward
movement. The vertical (plumb) piles, especially those embedded deeply
in the sloping earth, generally have considerable lateral stiffness, espe-
cially when the slope is considerably flatter than the angle of repose, when
the composed of gravel, and when it is armored deeply with riprap.
fill is

Fenders or bumpers are generally desirable to prevent damage to ships

when they strike or scrape against the bulkhead. It may also be advis-
able to extend the fenders or a wall down below low-tide level in order to
keep small boats and driftwood from catching under the structure.
One danger to be guarded against is the placing of fill on silt or clay
that may move slowly waterward because of the weight and the small
cohesion and low angle of repose of these soils when saturated. Tides

jConcrete Bituminous concrete R.R. track , L aminated timber

Fenders -T~

FIG. 10-37. A creosoted-timber bulkhead of relieving-platform type.

and waves may also tend to move the soil away and destroy the resistance
to lateral movement of the structure. Gravel and riprap coverings help
to reduce these effects, but it may be advisable to excavate dangerous
soils before the structure is built.

The sheet piling at the rear of the bulkhead in Fig. 10-36(6) is used to
shorten the necessary width of the platform, especially when the waterway
is deep. However, if the piles pene-
This adds to the outward thrust.
and if the fill is made of good material well compacted,
trate well into sand
the batter and plumb piles generally permit the use of a moderate pressure
against the sheet piling.
Figure 10-38 pictures a type of relieving platform with a sheet-pile wall
at the front. Some advantages of this construction are the following:
1. Wooden piles are embedded in fill or soil, kept continuously wet, and
protected from marine borers.
2. Sheet piling forms the front wall at the top.
 '

422 PILE FOUNDATIONS [Chap. 10

3. Sheet piling can be driven first, then used as a cofferdam for the

remainder of the construction.

4. If the soil or fill is firm enough, the concrete deck can be poured

directly upon the ground without expensive forms or a timber platform.

A structure like that of Fig. 10-38 should be investigated carefully
8x8"YRCont
to™ 3
bolt
I J6x4 x /8x0'-6"/g
"
s ri 33Scont. 8"x8"Y.R
M^JyS. ///
:
y//s ://fW \\0/r///\//y-='l^ =
'

-6"x6V/8LxO'67#
I"////.-,

8"x8"YR

6x6"x££cont,
fM.LW.EI.-6.3l'
I'-C"

y U.S. Harbor line

';* <*
cy.cs,

10
IS

-
Proposed channel
^Mud line -29.0'
Mud line varies
}ffrom -29.0 to -33.0

Batter piles z
38 'to 40'
lg

FIG. 10-38. A type of relieving platform with sheet-pile face and batter piles. (Courtesy of Arthur G.
Beaulieu, Bulkheads: Their Purpose, Design and Construction, 56th Annual Report, The Connecticut
Society of Civil Engineers, The Quinnipiac Press, Inc., New Haven, 1940.)

to make sure that it Although this bulkhead is a

is stable laterally.
relieving-platform type, the under the platform exerts considerable
fill

horizontal pressure against the sheet piles, which has to be resisted by

the batter piles. It usually requires many of these to provide sufficient
strength to prevent overturning.
A sheet-pile and anchorage bulkhead is shown in Fig. 10-40. This is a
Art. 10-16] BULKHEADS ON PILES 423

FIG. 10-39. Tied sheet-pile bulkhead construction at English Station of the United Illuminating Com-
pany, New Haven, Conn. The old bulkhead is shown at the right with the tie rods anchored to it.

The new sheet piling is at the left with timber wales and pile bents to line it up. The tie rods are
adjusted by turnbuckles and are permanently supported upon timbers and piles as shown. Before
the fill is placed the tie rods will be boxed in concrete for protection. (Courtesy of Wesfcoff and
Mapes, Inc., Consulting Engineers, New Haven, Conn.)

Steel sheet piling

MHW
/ Double Owale attached to sheet pi/ing
E Tie rod] T/'erod±
MLW 7rf ^Concrete
'dead man"
Steel-,
on piles sheet
piles

(b)- WOODEN PILES

ALONE AS frfJ-SHEET-PILE
ANCHORAGE WALL AS
Buttresses
ANCHORAGE
Dredged ifneeded "X

Tie rod ^/

fcJ-CONCRETE WALL
(aJ-VERTICAL SECTION THROUGH BULKHEAD ANCHORAGE
B
U

FIG. 10-40. Sheet-pile bulkhead construction.

practicable structure when the subsoils are firm enough and when the fill

is not too deep. It is not generally suitable for large ships and deep
waterways. It consists primarily of a continuous wall of deep-arch or
Z-shaped steel sheet piling attached to steel tie rods uniformly spaced.
These rods are connected to a bolted or welded wale placed so that the
cantilever effect of the top AB reduces the maximum bending moment
in the portion BD. It is very important that the embedded toe CD be
424 PILE FOUNDATIONS [Chap. 10

sufficient for the soil to develop the necessary passive resistance to the
outward thrust of the sheet piling.
The anchorage of the tie rods is very important. Sketch (a) shows the
use of a continuous wall on vertical and batter piles. If the line CE is at
the angle of repose </> and soil, the wall itself cannot do
of the saturated fill

much to anchor the bulkhead because it will push against the soil which is,
in general, pushing against the sheet piling. The plumb piles will tend to
resist being pulled out and tilted, and the batter piles will be very effec-
tive. If a wall is used alone, as in (c), it should be placed considerably

Future

TYPICAL CROSS SECTION

FIG. 10-41. Bulkhead construction along the edge of the coal-storage area at a power plant. The
tracks are for one end of the loading and unloading bridge. Courtesy of Arthur G. Beaulieu,
Bulkheads: Their Purpose, Design and Construction, 56th Annual Report, The Connecticut Society of Civil

Engineers, The Quinnipiac Press, Inc., New Haven, 1940.)

back of line CE. Other details of anchorages are shown in (b) and (d).
The latter, too, should be placed 10 to 15 ft. beyond the plane of the
angle of repose if this sheet-pile wall is to develop the necessary passive
resistance.
The tie rods should be adjustable by means of turnbuckles in order to
line up the bulkhead properly. These rods may need intermediate piles
above them. In one case this
for support against the weight of the soil
was not done, in spite of the engineer's advice. Furthermore, the fill
was dumped from the shore outward instead of in reasonably horizontal
layers. The tie rods were bent down so badly in places that they pulled
the bulkhead wall far out of line. One tie rod failed.
Art. 10-17] EVALUATION OF PILE GROUPS 425

Tie rods might be encased in concrete for preservation, but this adds to
the weight and the transverse vertical earth and live loads carried by the
rods. use large-sized rods to allow for corrosion, and it is
It is better to
advisable to paint them with hot asphalt or some other suitable coating
or to wrap them in membrane waterproofing.
The bulkhead shown in Fig. 10-41 is subjected to heavy loads. The
coal-storage pile might be 40 ft. high. The tracks shown are to support
one end of a gantry crane having a 300-ft. main span and a 120-ft. canti-
lever over the water to unload colliers having a draft of 30 ft. Hurricane
winds might cause a load 600 kips on the leeward wheels of the gantry,
of
also a horizontal thrust of 167 kips. These forces might be applied along
the track. Notice the combination of steel and wooden piles, the trench
for the tracks and the braced sheet-pile anchorage.
The tie rods shown in Figs. 10-40(a) and 10-41 may be seriously endan-
gered if the soil under them is very compressible; e.g., an organic silt or soft
clay. The effect of the fill or other soil above the tie rods, as well as that
of any live-load surcharge, will cause the compressible soil to consolidate.
As a result, the material above the tie rods tends to "ride" them as it
settles, causing extremely large bending and increase of tension in them;
a few scattered piles cannot support these tie rods adequately. In addi-
tion, the settlement may rupture the connections of the tie rods. Fur-
thermore, such settlement will tend to add very heavy vertical loads on
the anchorages. It is advisable to remove the compressible soil and
replace with a sand or gravel fill.
it In Fig. 10-41 the exposed steelwork
should have cathodic protection.
10-17. Evaluation of pile groups. As stated previously,' tests of
piles that bear on firm soils generally yield reliable results for use in deter-
mining the safe load per when used in large groups, unless they are
pile
underlain by weak cohesive The tough problem is the selection of
soils.

the presumptive bearing power when the piles are in large groups in and
above plastic cohesive materials. Another problem related to this is

estimating the probable settlement of the foundation, the determination

of whether or not this settlement will be harmful, and the making of a
decision regarding what should be done about it.

Some This term

engineers refer to the efficiency of the piles in a group.
isused to denote the ratio of the safe average bearing value of a pile in the
group to that of an identical individual test pile driven to the same depth
in the same soil. This is another way of allowing for the fact that the
average bearing capacity of one pile in a group of friction piles, as limited
by the strength soil, is usually less than that of an
or settlement of the
isolated test pile. obvious that attempts to determine this efficiency
It is
are, and have been, unsatisfactory because local conditions differ so much
at various sites, and they are not dependent upon the number and type
 ,

426 PILE FOUNDATIONS [Chap. 10

of the piles. It is better to analyze a problem upon the basis of the

estimated bearing value and stiffness of the soil, looking upon the piles
primarily as a device for getting the load "into" the earth.
Theoretically, for a particular required bearing area just below the
piles, the greatest economy will result when the number of friction piles
and their spacing are such that the frictional resistance along the surface
of each pile is almost completely utilized in the transfer of the pressure to

IO L 0"

^"tbolis
** I2"xl2"cap

l"<px26"drift

^ *| Section A-A
PLAN OF WHARF AND CONNECTING PIER AT JUNCTION

24-0*

3-6" _
2"x6"with % "open
open joints 3'-6"

m_ S / {
2"x6"bridging ^'decking, £ j ,EI. 12.0

T~S\MLW
• El. 0.0
^i V.-.-J

CROSS SECTION OF WHARF

FIG. 10-42. Creosoted timber pier and wharf construction at the Charleston, S.C., Navy Yard. (Cour-
tesy of Wood Preserving News, January, 1 945.)
Art. 10-17] EVALUATION OF PILE GROUPS 427

this soil. This tends to influence the use of wide spacing and long piles
under many
conditions. This, however, may cause the use of wider and
more expensive footings, and of more costly piles and equipment. The
engineer who is planning a specific foundation dares not let his desire for
economy carry him so far that his structure will be endangered. On the
other hand, it is impracticable to make load tests of large groups of piles
at various spacings because of the great cost and delay. Furthermore,
the results of small-scale model tests are likely to be delusive. Service
records of actual foundations are almost the only way to increase our
knowledge sufficiently to avoid danger on the one hand and too costly
conservatism in design on the other.

FIG. 10-43. An example of precast concrete bridge construction. Intermediate bents consist of four
concrete piles with a capacity of 25 tons each. The roadway and sidewalk supports are precast
sectional slabs. The curb, sidewalk, and handrail are cast in place. (Courfesy of E. S. Fraser,

"Precast Concrete Slab Bridges in Florida," Reinforced Concrete, Portland Cement Association, Chicago.)

The following procedures are given as suggestions that may be used in

addition to loading tests in obtaining quantitative data as a basis for
judgment in the selection of the allowable load per pile for purposes of
design. These are based upon the idea that the supporting power of the
earth below the tips of the piles is the critical part of the design. This
includes the avoidance of excessive consolidation under load as well as the
ability to hold up that load. Naturally, the results may not be accurate,
but they may be helpful in comparing various possible solutions for a
problem involving groups of piles in cohesive soils.
1. Select a type of pile and a depth of penetration that seem to be

suitable for the conditions.

2. Estimate the static frictional resistance that may be developed along
the embedded surface of the pile, using the area of the surface of the pile
 .

428 PILE FOUNDATIONS [Chap. 10

and the unit frictional resistances taken from Table 9-1. If the result of
a pull-out test is available, that will be even more useful.
3. Deduct the weight of the pile, neglecting buoyancy, because the

weight of the pile itself is not "useful" load but requires some of the
friction to hold it up. Wooden piles have the advantage of lightness.
4. Having the result of item 2 minus 3, discount somewhat the total

load per pile because of the group effect, if this seems to be necessary,
although the values given in Table 9-1 already have allowed a little for the

5-0" S !0" 9 L 0" S'0"s'0"

17

r«;-PLAN OF PILES
G H U
A B rcAASSUMED PRESSURE DIAGRAM

o'
Q
I
\
I

V h-

%
\ \
\
M-
c'
7J&
>
^ \F D
I!
\

\
\

\
-s^ I
I
^,1
-\

% P/'/es^. \ \ Tips 1

I T
T \ \

Cl V E \o 'n a
10-0" n'n"
9-0' IO-0'
,

J/S)J--?i)" lO'-O'
(W-ASSUMED AREA UNDER PRESSURE
(b)- VERTICAL SECTION

Assumed
diagram
^-COMPARISON OF ASSUMED AND MORE
PROBABLE PRESSURE DIAGRAM

FIG. 10-44. Scheme for estimating pressure on soil at tips of piles of a group in cohesive soils.

effect of small groups. Compare with Table 10-1. From these select a
trial safe load per Determine accordingly a tentative number and
pile.

arrangement for the piles, and make a rough sketch of the scheme in plan
and vertical section.
5. From the top of the earth that has reasonable supporting value,
draw assumed distribution lines like AC and BD, Fig. 10-44(6). These
intersect the plane CD at the tips of the piles so as to delineate the area
CC'O'OD'DNN' in Sketch (d).
6. Assume that the pressure diagram on the area outlined in Fig.
10-44(d) is shown in cross section in (c) as GCDH. This is flat under the
Art. 10-17] EVALUATION OF PILE GROUPS 429

piles and decreasing in intensity from area EE'F'F to the perimeter. It

is probable that the actual distribution of pressure will be more like the
dashed line PQ in (e), but the difference will be neglected. However, for
the purposes of computation, it is satisfactory to assume that the diagram
of intensity is represented by RSTU in Sketch (c) and that SS'T'T in (d)
represents the area over which it is theoretically applied. Notice that
y
RG is 2 CE, etc.
To avoid confusion, it is desirable to use one method uniformly. The
rectangular diagram is so much more simple that it will be assumed.
The batter lines therefore will be considered to be those that determine
the supposed edges of this diagram, SR and TU in Fig. 10-44 (c). These
are shown as AS and BT in Fig. 10-44(6) in contrast to lines AC and BD
orginally assumed. Ifvery soft clay or weak silt, assume a
the soil is

batter of 10:1 to 12:1; if moderately soft clay or compacted silt, 6:1 to

8:1; if medium clay or stronger, 4:1 to 6:1. The choice of one of these
values of batter will be affected largely by the cohesion and angle of fric-
tion of the soils around the piles — their ability to resist vertical shearing
action and to spread the load. These values are based upon the assump-
tion that the soil is fairly uniform. If its strength and stiffness increase
with the depth, the center of gravity of the frictional resistance will be
lower and there will be more tendency to concentrate the load nearer to
the tips.

7. the volume V of the pressure diagram.

Compute The volume of
this pressure diagram equals the superimposed load W. This need not
include the weight of the piles themselves except when this item is signifi-
cant, because they can be considered as replacing equivalent earth even
though they may compact the soil somewhat during the driving. The
weight of the overburden is not considered to increase the pressure applied
to the soil because it is already there.
8. The computed maximum soil pressure p v is then

total load on foundation W

area A
9. Compare p with
v the load p that is considered to be safe for the soil

at the tips of the piles as determined by tests or otherwise. If p exceeds

v

p, reduce the assumed safe load per pile accordingly; if pv is less than p,
the safe load per pile might be increased if settlement, the frictional resist-
ance, and load
tests of individual piles seem to warrant it.
Revise the layout to utilize the piles to maximum practicable
10.
advantage if the test load per pile is considerably greater than the com-
puted value as a member of the group. This generally means that the
use of wider spacingis economical except as limited by practicable sizes

of footings or caps.
430 PILE FOUNDATIONS [Chap. 10

To paragraph in reverse, assume that the soil below

illustrate this last
a group of 16 piles at spacing seemingly overloads the soil below
4-ft*.

them. The situation will not be improved by driving more piles between
the others to obtain a group of 49 piles at 2-ft. spacing. Spreading the
spacing to 5 ft. would help slightly but might make an unduly large
footing. A better remedy would be the deepening of the penetration to
obtain more spreading or to reach firmer ground.
Another illustration of these principles might be made by again assum-
ing a group of 16 piles as pictured in Fig. 10-44(a) but at 4-ft. spacing,
for which the estimated spreading of loads may be somewhat as pictured
in Sketches (b) and (c). Assume also that the estimated safe bearing
value of the "effective area" of soil under these piles is 640 kips. This
will give an average of 40 kips per pile for design purposes. However,
assume further that a test pile in this area has shown that the estimated
safe load per individual pile (with friction and some end-bearing value)
is 54 kips.Then, since the spreading of load into the soil occurs inside
a pile group as well as outside,it would seem that the four center piles

could be omitted because the load of 640 kips divided by 12 gives 53.3 kips
per pile. This may be possible, but the spreading of all the load around
the perimeter of the footing would be a disadvantage to the latter, and
the omission of the four central piles might be carrying the idea a bit
too far.

The next problem is the invention of a procedure for estimating the

settlement of a pile foundation. The best that can be expected is a
determination of the general scale of the magnitude of the future settle-
ment. Whether it will be 2 or 2^ in. means little. Whether it will be
2 in., 6 in., or 2 ft. means much.
The following procedures are suggested for making an estimate of the
settlement of a group of friction piles supported by plastic cohesive soils:

1. By laboratory tests for consolidation under load, obtain data giving

the relation of pressure vs. consolidation, and plot curves giving these
data. Do this for each 10 ft. of depth of soil around the piles, and for
considerable distance below them, when the properties of the soil vary
considerably.
Compute the pressure diagram as outlined previously.
2.

Assume the intensity p„ to act upon the soil around the piles for the
3.

lower half of the embedded depth, h, of Fig. 10-45 (a). Find the com-
puted settlement for the depth d\, or the individual settlements for each
of the 10-ft. layers in the distance d\ if the soils vary.
4. intensity p v and the pressure diagram computed in
Assume the
item 2 to act over the area directly under the group of piles. This is
pictured as the width ST at the top in Fig. 10-45(6). Assume a theoreti-
cal spreading of this load at some batter x:l, where x is consistent with
Art. 10-17] EVALUATION OF PILE GROUPS 431

the properties of the soil below the piles. An assumed magnitude for x
may be 3:1 for medium clay and well-compacted silt; 2:1 for firmer
plastic soils and fine sand. Then compute the resultant pressure p'v for
similar pressure diagrams at the centers of the 10-ft. layers below the
f
piles, including as many layers as the p v and the soil con-
magnitude of
ditions seem to warrant. Notice that pressures below ST in (b) are to
be computed at the centers of the assumed layers.
5. Compute the compaction of each of these layers, then add all the

computed results of items 3 and 4 to obtain the estimated settlement.

Of course, these settlements estimated on the basis of consolidation of
plastic soils are based upon long-term loading. Temporary loads do not
have time to squeeze the water out of clays.

B R U
Pressure diagram Elevation of
assumed

//
// X ili.
ji tips ofpiles
\*rrAssume<d \~.
I
\
// \X distribution -'
'-'

// I
\ lines
//_ V —TT-
\\
\ \
\ —V \

<!
// \ \
s
I \ ?
>i
\ \*
I
I I

I
',

\ \ tn r T ? ,

1
—Q
\

E F T D
:

C S
^"Assumed lines of distribution
r«)-VERTICAL SECTION rW-ASSUMED DISTRIBUTION OF PRESSURE
THROUGH PILES THROUGH SOIL BELOW PILES

FIG. 10-45. Scheme for estimating settlement of a group of piles in and above cohesive soils.

6. Mentally back off and look at the whole problem in proper perspec-
tive. Are these results seemingly reasonable? Do they show any danger
signals? Test the results again with a different (greater) assumed
spreading batter. Estimate the settlement by any other applicable
method. If the results persistently show computed settlements that will
be harmful to the structure, devise some way to remedy the situation,
even to modifying the superstructure, changing the substructure scheme,
or putting the structure somewhere else.
Paragraph 3 needs explanation. A tapered pile has more surface area
near its butt than near its tip. Therefore, it would seem that the pressure
transmitted to an annular area of soil near the butt, pictured by BC of
Fig. 10-46, would have a greater increment of compression per square foot
than would a corresponding ring LM near the bottom, assuming the prop-
erties of the soil to be the same at both places. The pressure at BC will
tend to decrease in intensity as it is transmitted downward over the
432 PILE FOUNDATIONS [Chap. 10

increasing assumed area to HM. Nevertheless, the soil below BC cannot

resist compression without* deformation. If A and M were to remain
stationary, since the pile any shortening of is relatively incompressible,
the ground from BC
would have to be taken up by cracks or
to HM
tension, both of which are inconsistent with what must occur. As the
lower soils compress, the pile must penetrate farther into them in order
to have compressive stress in the ground around it. Since AH is assumed
to be a limit to the lateral spreading of significant compressive stresses
in the ground, it would seem that the forces on ring BC
would spread
over the whole of ring HM; EO over JM; PN over and LM over
KM;
itself. Therefore, the intensity on HM will probably decrease from M
Max h/gh water
fFloor line El. 306 rEL303
A Pile

\ 290 -Pear:
l\n\\
280 -i ofcolumn
b!c\

"
W
h \
270

260 Dark blue

\ \ clay, moderately
250
Pi' .t-v-
\ \ \ 240
stiff
! 1 4
i /<*h \ \
230
FIGjPjX
UP
Ilk r
\ \ \ 22Ci

2jq_
II 1 \ \ \ \

\ \ \ ,\ 200 Dense clay w/lh some tine sand

\K\L\M\ f f (Rock at El. 143)

FIG. 10-46. Assumed action of soil around a fric- FIG. 10-47. Soil profile at a proposed industrial
tion pile. plant.

outward to, and perhaps beyond, H. The lower strata are also often
more compact and resistant than the higher ones, and the driving may
tend to remold the clay somewhat. Therefore, with so many uncertain
elements, it seems to be sufficient to assume a broadly interpreted com-
pacting effect equivalent to the computed maximum pressure p acting v

through a depth equal to the lower half of the embedment.

There remains much to be done in correlating the results of experience
with any proposed mathematical procedure for predicting the safe loads
and settlements of groups of friction piles. Numerical values given
herein for the assumed batter, and the suggested procedures, in their
entirety, are proposals. Their accuracy cannot be expected to be per-
fect. Even past performances seem to have been exceedingly variable.
Now, test all these procedures by application to real situations.
Art. 10-17] EVALUATION OF PILE GROUPS 433

Example 1. The soil pictured in Fig. 10-47 is at the location of foot-

ings that are to support heavy columns 30 ft. c.c. in an industrial plant.
One total estimated load, including the footing, is 240 kips DL and
320 kips LL. Select the type and length of pile to be used, plan the
pile layout, The clay is fairly uniform.
and estimate the settlement.
Assume the data in Fig. 10-48(a) for its compaction, using curve A.
Assume further that the engineer in charge believes that the deeper clay
should not support a load of more than 2 k.s.f. because of settlement.
Friction piles are necessary. The peat should be removed before the
fill for the floor is placed in order to avoid large future settlement of the
floor and machinery. This, since drainage of the fill is to be installed,
will permit a low watertable during a dry season, making plain wooden
piles undesirable. Steel H's, prestressed concrete, and pipe piles are not

± 0.1

^ 0.08
fe 0.06

12x8
Tips ofpiles
| 0.04
c Tips ,
q/J&J'*22J A R

I 0.02 III J3'x29'

E
£
12 3 2Jx/9'
& 39.7'x35.T
^\7T

Load in kips persq.ft

(a)S0ME ASSUMED CURVES
SHOWING COMPRESSIVE
DEFORMATION OF SOILS VS. /W- SECTION THROUGH (W-LAYERS FOR COMPUTING
COMPRESSIVE LOAD PILES SETTLEMENT

FIG. 10-48. Data for an estimate of settlement.

suitable because their great strength cannot be developed by the friction

unless they are excessively long. Creosoted piles or thin-shelled cast-in-
place piles of Raymond Monotube type seem to be advantageous.
or
The Raymond step-taper pile shown in Fig. 9-10 will be selected, with
the cutoff at El. 301. Assume 56-ft. lengths as suitable; great length
does not seem to be necessary. The tips will be at El. 245.
With the data in Fig. 9-10 giving surface areas, and a unit friction of
300 p.s.f. from Table 9-1, the estimated frictional resistance of a single
pile is

F = 166 X 0.3 = 50 kips

The fill is pile. The buoyant weight

not relied upon to help support the
of the pile approximately 4 kips. Therefore, the net supporting friction
is

F' = 46 kips. Because of the group effect, call the bearing value 40 kips
per pile. This requires
434 PILE FOUNDATIONS [Chap. 10

NA7 =
240 +
j?r
320 ,,
=14 ..
piles
,
(assume 16)

Try a square group of 16 piles at 4 ft. c.c. (12 ft. X 12 ft. over-all).
Now test for the pressure on the soil. To be very conservative, assume
distribution at an 8 : 1 batter from the top of the clay, El. 290. Then
W = 560 kips = p v (12 + 11)(12 + 11)

pv = 1.1 k.s.f.

This is less than the 2 k.s.f. allowed, so that the supporting power of the
soil is not critical.

It is now desirable to see whether a higher load per pile can be used,
since the soil seems to be strong enough. If the full 46 kips per pile were
allowed, the revised number required would be
N' = 56()
46 = 12.2 (call it 12)

For the present, assume these as three rows of four each, all at 4-ft. spac-
ing. The dimensions are shown in Fig. 10-48(6). Compute the revised
soil pressure
560
Pv ~ 23 X 19
" L6 k ' S - L

In estimating the long-term settlement of the piles, the live load is

neglected because crane loads constitute most of it, and these are tempo-
rary. The effect of the added fill over the whole area is not included in
this particular problem, but it should not be overlooked.
The computations are as follows, referring to Figs. 10-48 (a) and (c) and
using a batter of 3 : 1

p v for dead load = 1.3 X 24

%eo = 0.6 k.s.f. (approx)

Above tips, using curve A

Ai = 10 X 0.013 = 0.13 ft,

A 2 = A x = 0.13 ft.

Below tips: Compute the pressure at QR, the middle of layer No. 3.

240
Vi= = ° 4 ksf -

26.3 X 22.3
A = 3 10 X 0.01 = 0.10 ft,

?40
Similarly, p4 = rz = 0.25 k.s.f.
g3 29
A = 4 10 X 0.007 = 0.07 ft,

2
7)5= = °- 17 k si - -

39.7 X°35.7
A5 = 10 X 0.004 = 0.04 ft.
Art. 10-17] EVALUATION OF PILE GROUPS 435

The total computed settlement thus far — the sum of the five 10-ft.
layers used — is 0.47 ft.obvious that the unit pressure on the still
It is

lower strata will decrease to insignificance. Therefore, it is reasonable to

conclude that, according to these computations, the long-term settlement
of this footing will not exceed 6 or 7 in. Is this objectionable? It cer-
tainly would be if one column alone were to settle this much more than
its neighbors. If all settlements could be kept the same, no serious harm
might result. This, however, will be difficult since the column loads
vary considerably. Besides this, the live loads are large and include
some vibrations. The footings should not seem to sink into the surround-
ing floor. Furthermore, overlapping of the effects of loads on adjacent
columns has been disregarded. More conservatism seems to be desirable.
Assuming the clay to be as compressible as Fig. 10-48 (a) and these com-
putations indicate, settlement of footings and general settlement of the
area under the weight of the new fill will be troublesome. Of course, the
assumption of a flatter distribution in Fig. 10-48(6) would yield a smaller
computed settlement, but this would not change the general character
of the result.
Spreading the piles farther will gain little advantage. The addition of
more piles also gains littlecompared to the extra cost. Driving the 12
piles 8 to 16 ft. deeper is more advantageous. If this is done, what effect
will it have upon the computed settlement ? It looks as though the result
would still be unsatisfactory unless considerably longer piles could be used
to reach the dense clay and sand farther down. This might be a good
place to use composite piles with wooden piles at the bottom and poured-
in-place concrete for the upper portion.
Example 2. Figure 10-49 shows the soil profile at a small bridge pier
consisting of two concrete shafts 6 by 8 ft. in section and 30 ft. c.c. at
El. 200. The total load at El. 200 for this study is 1,650 kips DL and
450 kips LL. It is thought to be advisable to use piles under two separate
footings or under a combined footing. Determine the type, length, and
arrangement of piles. Estimate the settlement and the soil pressure
below the piles.
Here, 100-ft. H piles or pipe piles could be used directly down to rock.
Precast concrete piles to firm end bearing would be rather long and heavy,
but they could be used if the base of the pier were set at approximately
El. 180, which might very well be desirable. Wooden piles into the soft
clay could be used also, but they would have to be too long if they were to
reach the sand and gravel. Poured-in-place thin-shell concrete piles
could be used in the clay, but they too would be rather long if they were
to reach the sand. Button-bottom piles could be driven to the sand,
and a thin shell could be used on the main shaft to keep out the silt. The
choice is between long strong end-bearing piles and shorter weaker fric-
tion piles.
436 PILE FOUNDATIONS [Chap. 10

Before making a decision, test 50-ft. wooden piles with their butts at
El. 185. These are to be driven inside a cofferdam that permits the
excavation of the mud and the upper portion of the silt. Assume that
these piles are 8 in. at the tip and 20 in. at the butt. Assume 20 ft. in
silt and 30 ft. in clay. From Table 9-1, assume 100 p.s.f. for skin friction
in silt; 400 p.s.f. in clay. Use the area of the surface of a cylinder having

MHW E/.200 \S of P/er

El. 190 -^-^-Mud^z-^-Z-
180

170

160

150
~ Medium fo
140
stiff clay
130

120

110

100 « •: -Coarse CTffa^g'gwi?/

1
Ledge rock
FIG. 10-49. Soil profile at a proposed bridge.

the mean diameter of each of these portions of pile instead of the tapered
sections.

F s for silt = 0.1 X 20 X x X 1.47 9.3

F c for clay = 0.4 X 30 X r X 0.97 3Jx5
Total F 45.8 kips, or 46 kips each

Because of their light weight, no deduction will be made for this from F.
However, because of the group effect, and assuming the clay to be safe for
only 2 k.s.f. because of settlement, call the safe load per pile 40 kips.
To the loads given, add a guessed net weight of 300 kips for the pier
shafts and footing below El. 200. The number of piles required would
then be
N = 2,400/40 = 60

This seems to require a combined footing under both shafts. Therefore,

try four rows of 15 piles each. With four rows at 3 ft. 6 in. spacing, and
with the piles at 3 ft. c.c. in each row, this would make a group 10.5 ft.
wide and 42 ft. long.
Using an average distribution batter of 10:1 in the 20 ft. of silt and
6:1 in the 30 ft. of clay, the estimated pressure at El. 135 is
Art. 10-17] EVALUATION OF PILE GROUPS 437

2,400 _'
p* " SJ
+ +
-

(10.5 14) (42 14) '° -

This appears to be fairly large, but the settlement should be investigated

before a decision is made.
To make this estimate of settlement, assume p v to affect only 15 ft. of
clay above the tips, because the silt does not resist much anyway.
Assume a distribution of 2 : 1 for the clay below the tips, El. 135. Let
curve B of Fig. 10-48 (a) represent the load-deformation relationship for
this clay. It is so nearly a straight line thatassumed to be one, and it is

the 30 ft. of clay from be used as one single 30-ft.

El. 135 to El. 105 will
layer. The area of the assumed pressure diagram at El. 120 the center —
layer— is 71 by 39.5 ft. Use the dead load
of the 30-ft. of 1,950 kips
when computing the settlement. Then,
Above tips: p„ = 1.75 X 1,950/2,400 = 1.4 k.s.f.
Ai = 15 X 0.017 = 0.26 ft.

Below tips: p'

v
= '
= 0.7 k.s.f.

A 2 = 30 X 0.009 = 0.27 ft.

The total computed settlement is 0.53 ft. or 6 to 7 in.

Is this satisfactory? It raises serious questions because greater settle-

ment is not guaranteed by any of these figures. It is not a matter of
weakness of the piles but of consolidation of the soil under and around
them.
Suppose that end-bearing piles driven to rock are considered now. At
50 tons per pile, the number would be

AT = 2,400/100 = 24

These could be arranged as two groups of three rows each at 4-ft. spacing
under each pier shaft, provided the lateral stability will be satisfactory.
In this case, there is no question of settlement. Is the cost greater or less
than for the wooden piles? If greater, is the security worth it?
If the end-bearing piles cost $10 per foot in place whereas the wooden
ones cost $4 per foot, the latter would be cheaper, but some additional
expense would be entailed in the cost of the larger footing.
The final choice will depend considerably upon the method of construc-
tion desired, but it seems advisable to eliminate all question of harmful
settlement by using long piles that reach the sand and gravel. Assuming
that floating equipment cannot be used without too much dredging, the
proper procedure seems to be the construction of the cofferdam from a
temporary wooden-pile trestle, the excavation to El. 185 for the base of
the pier, then the driving of 16- or 18-in. closed-end pipe piles with the
ends capped and welded watertight. These pipe sections could be made
438 PILE FOUNDATIONS [Chap. 10
6

of two pieces welded in the field, if one piece is too long to handle. The
pier might then have two groups of 12 piles each, with a bottom strut of
concrete joining the two sides.
Example 3. Figure 10-50(a) shows part of the soil profile at the site of
a fireproof warehouse. The design loads and column spacing are given in
(6). Assume that 70 per cent of the loads given represent live loads.
Select the type of foundation to be used, and design a typical support for
column C3.
The top stratum is too weak to support spread footings. The column
loads and spacing are too great for the practical use of a mat foundation.
Strata 2 and 3 are better soils but are too far down to reach with footings.
Stratum 4 is about 10 to 12 ft. thick and reasonably firm. However, it

is underlain by nearly 30 ft. of clay. What should be done? Can piles be

R.R tracks Fl.206.5

soo k 920 k 1000 k 920 k )r
D a SOOa

500 k 920 k I000 k k

9Z° r„„ k
D u ° 500°

500 k 920 k W00 k 920 k „

I
D a £00°

270 k 530 k 560 k S36*mk OH

~~iZoarse sand, gravel and s0"
•'•
o ^.u^-g- some boulders-Strafum 7
'/^\Mii/,=w=im
VJ0-0"\ J5-0"
+ J5-'0"Tj0
\

(6J-PLAN OF ONE END

"Bed rock

raJ- SOIL PROFILE AT COLUMN LINE©

FIG. 10-50. Soil profile at site of proposed fireproof warehouse.

supported by stratum 4, or must they be driven clear down to stratum

6 or 7?
Before attempting to answer this, estimate the settlement of the brown
clay if the piles penetrate to El. 165, just above stratum 4. The average
load of the column when spread over the 25 by 35 ft. area is 1.2 k.s.f. For
settlement computations assume 50 per cent of the live load is applied for
a long time. The average superimposed load of the superstructure is then

p for DL + Y LL 2 = 0.3 X 1.2 + 0.7/2 X 1.2 = 0.78 k.s.f.

To this add 75 p.s.f. for the weight of the first-floor slab that is supported
upon the fill, and add 100 p.s.f. for the live load on the first floor. Assume
also that the average fill placed over stratum 1 is 8 ft. at 120 p.c.f., or
960 p.s.f. The total superimposed load then is

p = 0.780 + 0.075 + 0.100 + 0.960 = 1.92 k.s.f.

Assume at first that no piles are used and that the load is distributed
Art. 10-17] EVALUATION OF PILE GROUPS 439

laterally through the ground at a rate of 2 : 1. Then the added pressure at

El. 135 is, approximately,

1.92 X 135 X 255

^^
Pl35 " (135 + 70)(255 + 70)
~ iU

Assuming curve B of Fig. 10-48 (a) to represent the load-consolidation

curve of the brown clay, the estimated settlement is

A = 30 X 0.013 = 0.39 ft.

This might be satisfactory if it is fairly uniform over the entire area.

Furthermore, nearly one-half of the settlement is caused by the fill. The
fill is tapered slightly, but this should not be serious. On the other hand,
if the piles penetrate far into stratum 4, they will tend to concentrate the

pressure on the top of stratum 5. That is why their tips are to be near
El. 170 — penetrating slightly into the sand and gravel in order to secure —
at least the benefit of end bearing in stratum 4.
Assume that the length of piles below the top of stratum 1 is approx-
imately 30 ft. The minimum watertable is a little too low for plain
wooden piles. The great strength of prestressed concrete piles cannot be
developed by the friction alone, but they would have a large surface
area and a good end for bearing. Furthermore, they are a good type for
driving through the thin but dense crust of stratum 2. Poured-in-place
concrete piles without shells do not seem to be desirable, partly because
of thedanger of penetration of the saturated sand and gravel of stratum 2
when concreting the piles. Metallic shells seem to be desirable if they
can be driven through stratum 2 without rupturing. Steel H piles do
not seem to be advantageous here. The choice appears to lie between
creosoted wooden, prestressed concrete, and cast-in-place concrete piles.
Since the surface areas are given in Fig. 9-10, use Raymond standard
piles in this study.
The estimated average frictional and end-bearing resistance, using
figures from Table 9-1 and Fig. 9-10, is found as follows:

Stratum 1 : 250 p.s.f. friction, 14 ft. thick

Stratum 2: 600 p.s.f., 5 ft. thick
Stratum 3: 400 p.s.f., 11 ft. thick
Stratum 4: end bearing
Stratum 1: 0.25 X 63 = 16 kips
Stratum 2: 0.6 X 14 = 8 kips
Stratum 3: 0.4 X 33 = 13 kips
Stratum 4: estimated = 6 kips

43 kips (or call it 44)

Deduct 4 kips for the weight of the piles.

440 PILE FOUNDATIONS [Chap. 10

The number of piles for footing C3 is then

N = 1,000/40 = 25

Use these as a square group at 3-ft. spacing. Have the tops at about
El. 204 for interior footings and El. 201 for exterior ones.
There will be a tendency to have larger pressures on stratum 5 near the
ends of the piles than in between the footings. However, stratum 4 will
spread the loads somewhat so that the resulting settlement should be
negligible. There seems to be no necessity for deeper penetration of the
piles. It is also probable that the group effect will not weaken the
assumed supporting value of the piles because strata 4 and 5 are good
materials.

PROBLEMS
Some of the data in the Appendix is useful in working out problems. Where
additional data are desired, they are to be assumed and the problems are to be
worked out accordingly.
10-1. A footing 9 ft. square and 2
deep has a centrally located pedestal
ft.

2 ft. square and 3 ft. The pile pattern is

high flush with the top of the ground.
shown in Fig. 10-3(e). A vertical load P = 400 kips is centered on the pedestal,
and a horizontal load of 20 kips acts at the top of the pedestal and toward the
right. Compute the pile loads, and design the reinforcement, taking moments
about the bottom of the footing.
10-2. Assume a footing like that of Fig. 10-5. The pedestal is 3 ft. square and
3 ft. high, projecting 6 in. above the ground. The total depth of the footing is
3 ft., and d — 2 ft. 3 in. The vertical load P = 600 kips. A horizontal load
H— 25 kips acts at the top and center of the pedestal and parallel to the 9-ft.
side. Determine a proper center of moments for overturning forces (as in Fig.
10-25-4.), compute the total loads on the piles, and design the reinforcement.
10-3. Assume the same footing as in Prob. 10-2 except that the force H = 30
kips. Piles 1, 4, 7, 10, 3, 6, 9, and 12 are battered 1 :3 away from the center of
the footing parallel to the 9-ft. side. Compute the total loads on all piles. Is
the footing safe?
10-4. footing of Prob. 10-2 with P = 500 kips,
Assume the = 30 kips, and a H
similar force H' =
10 kips acting at 90° to H, but place the pedestal 6 in. off
center toward pile No. 8 and batter piles 1, 10, 3, and 12 diagonally at 1 :3 and 45°
with respect to the sides. Compute the total loads on all piles. Is the footing
safe?
10-5. A steel column supporting a vertical load of 1,400 kips has a pedestal
5 ft. square and 4 ft. high. If Raymond piles supporting 30 tons each can be
used at 3 ft. 6 in. c.c, design the necessary footing.
10-6. A mast 12 is to support a vertical load of 400 lb. and a
in. in diameter
horizontal load of 150 above ground. Determine the necessary depth
lb. 25 ft.

if the soil is sand weighing 110 p.c.f. and = 30°. <$>

10-7. Assume a bulkhead wall similar to Figs. 10-27 (a) and (b). If = 24 W
 OM o- — oj 9° cr>

co
— E —c while
and
TS a"£
C— little
clay
o c Fine
sand

X:

C3 c2 cvi ft, r "*

— <=t- <N ro oo i-O
"C3
i_
s «J c
S Q3 telf
O
co <a "to W
J-ts >
2U QJ
Sip Sea
O t L £ c ? ISfi
o

fO CO c—J — r~ n- l£ CD
l«
111 Soft gray
silt
CV3 §sS""l
c £"1
o «s
1Z o
to

13 = "a- "te "s: S r%*

£ o
K5 1— «* OJ OJ «*
1

"T5:§
Ec?
d ""15
CO
Soft
gray Silt
||1
TS >--^
"SI 5 OLE
o U- 3| £.5 :fc

3 -cr- CS j Pi p= «- 4-
4 S pt s s:

O "d" CO W o ro ^° O OO

r-~>
CO or g C JZ S ^ 5; lb I^v
* a*
TS- \g>

O en
1° ll ^
a?
t>

*>
o Sep -<?
_:
=^
- , ab — O
-X in
CO co CO
^
O
E2|

ll| |-B
>
CO
"C —
CO |li 15 1
UJ o 8=1 2 w
Jl 1-1'

S To =Z LH OJ no ^"
d-
>
r~s a~ J
co t2 ¥
-cs

§—
some wet and
and clay
W3O Co«rse gravel
Coarse sand little
gravel
Coarse
gravel

it "S or CS gravel

o c^- sand little

Course
sand

441
 1 il
"oJ "s is
>* — CO o
Cinder fill in 4j ~§ ~S Q

rO C*J — o er to
>-T3 « 2-*
T5 =
3: tr >j
o >a

3 § "~ "S "S r5 "S "§'S <3

- [ai - ^J- •* iS CD

1:1

l-o
9\ » vo C^J <*- oo o
C*3

OJ-K = -c •"= ^ 5 ^ 15
Soft gray silt

— 1
g o f
2 s~g
> a
CS-i w <5 n: o

o j» — O
*J- CO
H s^f a. o

5 1 SlT-5
if o — O
o — ECS III * IS

r~ kO W <o r-> <N ^

Loose
fill
granulated x> ±" > Oc
M- —
i- — •

W
slag 2~ w o CO CT 1

. ?:

ro CH 1 »* oj ro ro tl- m K
>. C j2 1^1
trig
^- cs _ c c
o c
—JE
cy
_S iff
"1.1
His
J "S-
"i3

Sic

CO a- — n i^ r~ H2

"3 it
ID C
o CO
||1 U3 c

442
 PROBLEMS 443

kips, H — is 1 :3, compute the loads on all piles.

9 kips, and the batter of the piles
Is the structure safe? what should be done to remedy the situation?
If not,

W and H are for one foot of width. Multiply by 12 to get the totals.
10-8. Figure 10-51 shows the logs of some borings that were taken at the site
of a proposed steel mill building to be located in the yard of an industrial plant
alongside a river near the coast of New The
building is to house a few
Jersey.
heavy but widely spaced machines to be served by 50-ton cranes. Make a
preliminary investigation to determine the type of foundation that is best. If
piles are used, determine the type of pile, their length, and the arrangements for
typical foundations.
The borings are not spaced to give as comprehensive information as one might
desire, but they are taken from a real job just as thej^ were presented to the
engineer.
Suggestions: Using Table 5 in the Appendix, the dimensions shown in Fig.
10-51, and an over-all dead and live load of 100 p.s.f., approximate the load on an
exterior column and that on an interior one.
Now draw soil profiles along the column lines and across the building site
through borings 3, 2, and 6; also through 4, 5, and 7. These may show the
difficulties of fitting the data from the borings into what seem to be reasonable
profiles. The past effects of scour, the localization of deposits, the presence of
lenses of and the general unpredictability of what lies beyond the borings
soil,

may become apparent. These are often characteristics to be expected in the

lower valleys of rivers, especially in glacial territory.
Some questions to be answered are the following:
1. Is it safe to found these crane columns on the cinder fill?
2. Is it safe and practicable to use spread footings on the uppermost stratum
of sand?
3.Should piles be used to reach the strata below the silt?

4. Can piles be used safely for some of the bases and spread footings on soil for

the others when cranes are to be supported?

5. What type of pile should be used, and what might the length be?
10-9. In Fig. 10-52 are shown the logs ofsome borings that were taken at the
site of a proposed extension of an existing shop at an industrial plant at the mouth
of a river along theEastern seaboard. The existing structure at the north is
founded upon 40-year-old spread footings. The structure is to be a steel-frame
mill building with a 30-ft. clearance under the flat and monitored roofs. The
walls are to be 8-in. brick.There will be a 25-ton crane between column rows A
and B and a 10-ton crane between B and C. Approximate the column loads and
plan the general character and details of the foundations.
10-10. Figure 10-53 shows the soil profile and other data pertaining to a pro-
posed industrial plant. Typical details at the bases of the columns are given,
and so are the loads. Select the type of pile to be used to support these columns.
Design and detail the complete supports for the two bases shown.
10-11. The shown in Fig. 10-54 pictures the conditions at the site
soil profile
of a warehouse. A
basement is at one end only, the remainder of the first floor
being directly on the ground. Plan the foundations for the structure, and esti-
mate the probable settlement.
444 PILE FOUNDATIONS [Chap. 10

Suggestions: Assume that the columns both ways. Assume that

are 20 ft. c.c.

the dead load on the interior columns in the basement is 220 kips; elsewhere,

175 kips. The exterior columns have 140 kips dead load at the first floor.
Assume that the unit live load for design is 30 p.s.f. on the roof, 300 p.s.f. on all
main floors, and 200 p.s.f. in the basement.
10-12. Plan the foundations for the power plant outlined in Fig. 10-55, if it is
to be on the soils shown.

I0@20'-0"=200 i 0"
fW-SIDE ELEVATION OF BUILDING

Pf22S^ 6 ,.
P^225 k „P2-/75k
, 2
24"WFt 3L0"

vel i

=
10
ll6"grade beam
\
I 5:4"pedestal
faJ-SOIL PROFILE AND TYPICAL CROSS SECTION OF BUILDING W'^LfccA./
HWpeclestoi^ I
I
C2-6"w/a/e)
'
1 '
W6"mde) '

W-BASE OF EXTERIOR (W-BASE OF INTERIOR

COLUMN DOUBLE COLUMN

FIG. 10-53. Soil and superstructure data for small industrial plant.

4-story concrete fiat -slab warehouse

240' long x 80' wide

Ground water/
20'
Yellow- brown soft clay
40'

60'

Siiff brown clay

100

FIG. 10-54. Problem in planning foundation for a warehouse at a tidewater terminal.

is much higher than the

Suggestions: Notice that one portion of the structure
rest, and much heavier. Assume that the 200-ft. low portion has a turbine aisle
60 ft. wide and 200 ft. long with an overhead crane and heavy turbogenerators.
The center of the high portion (boiler house) has two rows of columns spaced as
follows: 20, 40, 20, 40, 20 ft. All other columns, interior and exterior, are
20 ft. c.c.

Assume that the maximum column loads in the boiler house are 1,400 kips,
that many columns support 400 to 600 kips, and that the smallest loads are
 PROBLEMS 445

150 kips. Consider the character of the structure. It is steel frame with con-
crete floors and brick walls; and it holds very costly heavy equipment.
10-13. Figure 10-56 pictures in outline a large U-shaped apartment house, and
itshows a typical portion of the soil profile. It has a basement throughout.
Plan the foundations of the structure, and estimate the settlement.

200

Key Plan

Heavy steel
frame

M HW -

20'
Fill
Silt M±M^
Seine Tana
—
small gra vel
40'
Soft brown clay TZ3==H_
60'

WmmK^^ Sands ton t

FIG. 10-55. Problem in planning foundation for a heavy steam power plant.

lOO'x 100 'lower

10 stories high

2-200'xSO 'wings forming U (3)

5 stories high,concrete frame

Ground water Basement +

t

20'

FIG. 10-56. Problem in planning foundation for a large apartment building.

Suggestions: Notice the central tower, twice as high as the two wings. Assume
that the columns average 20 ft. c.c. The walls are brick. Assume that the dead
load averages 100 p.s.f. for each floor and the roof, including partitions; that the
exterior walls average 75 p.s.f.; and that the average live load is 60 p.s.f.
446 PILE FOUNDATIONS [Chap. 10

Is this structure suitable for a mat when the loads vary so greatly? Is a float-
ing foundation suitable? Are spread footings safe? If piles are used, what
type and length are suitable?
10-14. In Fig. 10-57 is pictured the soil profile at the site of a large store. The
structure has a basement and subbasement. Plan the foundations, and estimate
the settlement.

8 -story department store 100 wide x 240 'long,

steel frame

Water table* j
Ba_sement ^^^^ Sand^clai
20' £L .
~^5up-Jx»sement ^.'._i '-_^-il
: J
F/'ne sand
40'
~—- — Medium yellow clay
60'
'-^- Sand with some ciay
so'
-^jn/mjfwmnm
too'

J20'

Rock

FIG. 10-57. Problem in planning foundation of a large department store.

Suggestions: Assume that the columns are, in general, 25 ft. c.c. across the build-
ing and 20 ft. c.c. along its length. Assume the following loads: dead load of all

floors, roofs, and basements, 120 p.s.f. ; walls above ground, 75 p.s.f. ; live loads,
150 p.s.f.

Is this a good situation for a floating foundation? Are spread footings and
walls safe? Will the medium yellow clay be dangerous?
 11
COFFERDAMS AND OTHER AIDS
FOR OPEN EXCAVATIONS

11-1. Introduction. It is the purpose of this chapter to describe vari-

ous temporary means used to support the sides of open excavations that
are made for the purpose of building substructures. The main discussion
is about the features that are essential for the proper planning of these
devices prior to construction. As usual, the problems are special ones,
each dealing with a particular foundation, built on or in specific soils, and
under special pertinent conditions.
The equipment and facilities used for making these excavations are
adaptable to a wide range of detail variations. It is desirable to know
what means of support are available, what are their possibilities and
limitations, and how they can be used to accomplish specific purposes.
Some of the reasons for the need of such temporary supports of the
ground are the following:
1. Loose granular soils may have such a flat angle of repose that, in

order to make the desired excavation, an unduly large amount of earth

would have to be removed unless supports are provided, thus causing an
uneconomical procedure.
2. Cohesive soils may stand on steep slopes temporarily, but protection

is needed to ensure the avoidance of a sudden cave-in and accident.

3. Slumping of the adjacent earth may undermine adjacent structures,

such as buildings, walls, and pavements.

4. The removal of soil that provides lateral support for adjacent ground

loaded by structures may cause settlement and cracking of these struc-

tures even though a real cave-in does not occur.
5. When work is
done below ground-water level, or below river, lake, or
ocean necessary to surround the excavation with a tempo-
level, it is often
rary structure that will hold out the adjacent soils but will also enable
the excavation to be un watered.
6. When there is danger of flooding, a protecting structure may be

needed to keep water out of an excavation.

447
448 COFFERDAMS [Chap. 11

11-2. Shoring. Shoring denotes temporary structures of wood or

excavation proceeds. It is used to
steel that, in general, are installed as
keep the sides of the excavations from caving in. Ordinarily it is used in
connection with both large and small but shallow excavations, or for small
but moderately deep ones. For example, the trench for a sewer in a
street, if not shored, might cause damage to the pavement and so stop
traffic; or the sides of a deep pit, if not protected, might cave in and injure
workmen. This has happened all too frequently.
A few arrangements of shoring are pictured in Fig. 1 1-1. The following
comments refer to the respectively numbered sketches:
(a) This shows wooden shoring installed in a trench. A portion of the
trench is dug, and its sides are trimmed to make them reasonably even.
The shoring consists of boards or planks a that are set horizontally as
panels or as single pieces. Vertical beams b are added with struts c to
brace them. The excavation is then deepened, and another piece or tier
of shoring is added. This is repeated until the proper depth is reached.
This applies to excavation done by hand. If a trencher is used and if the
material stands on a vertical slope temporarily, the shoring may be erected
after the excavation if the trench is wide enough. Sometimes the shoring
need not be extended to the bottom of a trench but can be used only to
keep the top from caving in. Again, if the upper material is clay and the
lower stratum is fine sand, only the latter need be shored to keep it from
running in.
(6) This sketch shows wooden members d that are erected outside of
wales e and struts g after the upper portion of the trench is excavated.
As the digging progresses downward, the vertical pieces are driven far-
ther. Additional wales may have to be added with separators /. Posts
of some sort are generally needed to hold up the bracing system.
(c) The sides of a trench are seldom excavated evenly. The struts c
and g of the preceding sketches bracing the two sides are therefore of
variable lengths unless some adjustment is provided. Furthermore, the
sheathing should be pushed firmly against the earth if it is to prevent
slight slippage of the soil, or some packing should be used behind the
sheathing. Nailing the struts to wales or studs would provide some
adjustment but not much lateral resistance. A better arrangement is the
provision of struts h that are somewhat too long. These can be placed
diagonally with toe nailing or a cleat at one end. The other end may then
be driven along the wale to wedge the sides apart, and then held by nailing
cleat i to the wale. Again, wedges j and shims may be used for the same
purpose, and partly nailed to keep them in place. These braces may be
used in vertical as well as horizontal planes. Since braces interfere with
excavation, the side members should be strong so that the number of
braces can be minimized.
Art. 11-2] SHORING 449

(d) It is usually difficult to erect shoring directly under an upper por-

tion, as forSketch (a). If the excavation is deep, the sheathing d in (6)
will be too long or difficult to drive. Therefore, a telescoped arrange-
ment as shown diagrammatically here may be advantageous. The first
set is made as deep as practicable, then the excavation is deepened, and
a second set is installed inside the first, etc.

(e) The shoring of pits requires proper details at the corners. Rab-
beted ends for pieces m may seem to be ideal, but an excavation can
seldom be made so that they fit properly, can be erected easily, and are

£pW~ 7m,
\ \
7m
'
i

'S

1
o -ft
.N k
s £ ^
t-S>
^K I
i

(a)- TRENCH WITH (b)- TRENCH WITH (c) -MEANS FOR W-TELESCOPED
HORIZONTAL SHEATHING VERTICAL SHEATHING PRESTRESSING ARRANGEMENT
CROSS STRUTS OF SHEATHING

r<?y-C0RNER DETAILS tfJ-SH0RIN6 SIDE OF for;- SHORING SIDE BY DRIVING

FOR PIT SHORING WIDE EXCAVATION SHEATHING PROGRESSIVELY

FIG. 11-1. Some arrangements of shoring for sides of excavations

wedged sufficiently. Members n can be long; parts o, short. Wedges

can be inserted between o and n
n sideways; corner pieces p can
to drive
be driven against o to force the latter into bearing; then p can be nailed
in place. Another method is to use sides q of proper length and to wedge
both sides at a corner against the soil by wedging and nailing blocks r.
(/) Many shallow excavations are too wide to justify cross bracing.
As shown here, sheathing s may be set vertically outside of wales t and
diagonal braces u. As the edge of the excavation is completed, the
sheathing can be driven down, and more wales and braces can be added as
necessary to make sure that the sheathing cannot rotate or fail in any way.
450 COFFERDAMS [Chap. 11

(g) This shows a somewhat similar scheme used for supporting the sides
of a deeper large excavation. Steel sheet piling v is driven down first to
sufficient depth for toe hold below the finished grade. The central part of
the excavation is completed and that portion of the concrete mat is

poured. Then the sides are excavated, and the wales and bracing members
are erected, the latter bearing against the concrete. Sometimes, posts y
may be needed to support the long diagonals x and z.
There are many variations of detail, but the danger to be guarded
against is the erection of shoring so loose that it permits sufficient move-
ment of the earth to destroy its supporting power under whatever adja-
cent structure is to be protected. This movement may occur during the
dismantling of the shoring, before its erection, or during its period of

Surcharge

/
M- 2
a
9 £L 3
/t= 4
/ s S
/ 6 '
6
/ 7 7
/ 8 8

\l+sin<t>'

faj-PRESSURE DIAGRAM 6W-PRESSURE ON Or '-PRESSURE ON WALES ftfJ-PRESSURE ON BRACES AND EMBEDDED

PER LINEAR E00T HORIZONTAL OR STRUTS TOE PER LINEAR FOOT
SHEATHING

FIG. 1 1-2. Assumed pressures on shoring.

service. Therefore, provision for easy and safe conduct of all phases of
the operation is essential.
Shoring should be strong and stiff. The minimum probable force for
use in its design is the active lateral pressure of the restrained earth.
When shoring is pressed against the face of the cut, a portion of the pas-
sive earth pressure will be developed, the amount depending upon the
severity of the wedging. If the active pressure can be resisted with a
safety factor of at least 2, this reserve will usually take care of the excess
when good judgment used in the erection of the shoring.
is

An attempt to compute with exactness the forces affecting shoring is

impractical. Broad safe approximations, easily made, are sufficient
except in very special cases. The data shown in Fig. 11-2 are believed to
be conservative and sufficient for ordinary use, and the following com-
ments regarding the illustrations are for general guidance:
(a) This gives the pressure diagram, assuming an angle of repose of 30°
and a unit weight of earth of 120 p.c.f. This does not include any water
 1

Art. 11-3] SINGLE-WALL COFFERDAMS IN GENERAL 451

pressure because the sheathing is not assumed to be watertight. Sur-

charge is to be treated as indicated by A'C in the diagram.
(6) Horizontal sheathing is assumed to resist the force represented by
that portion of the pressure diagram which acts directly against it, or that
part which is bounded by lines midway between adjacent members if the
planks are not in contact.
(c) When a series of horizontal wales or braces is used, assume that the

pressure upon each is represented by the portion of the pressure diagram

between mid-span points (or ends) on each side of a given member, neg-
lecting the effect of continuity. Long heavy wales permit the shifting of
struts to accommodate excavation operations.
(d) When sheathing is braced in part by driving it down a little way to
get a toe hold as shown here, assume the pressure diagram to be OPQ.
The maximum reaction R 6 is limited by the passive resistance of the earth

A
C ,d

"
^ I
1

c
D _U LL

ySfeel
KW/w/f.wwH'V/r
sheet
piling Bottom of
excavation
B
(a)- NBRACED rtW-BRACED ONE WAY rW-BRACED IN TWO DIRECTIONS

FIG. 11-3. Single-wall cofferdams.

for the depth d in front of the toe. The theoretical resistance Re is

reduced purposely, as indicated, to allow for light loose saturated soils,
assuming a unit weight of 80 p.c.f. for the soil. It is satisfactory to
assume that R e acts at }^d from the bottom Q when the embedded depth
is too small to offer more than a shearing resistance.
11-3. Single-wall cofferdams in general. The word cofferdam de-
notes a temporary structure used to protect an excavation against lateral
earth pressure and water during the process of excavation and of building
the foundation or structure that goes in it. In almost all cases, the coffer-
dam is removed when no longer needed. Sometimes the excavation inside
the cofferdam is not unwatered, but generally one of the most important
functions of this temporary structure is to facilitate conduct of the work
"in the dry." To a certain extent, a cofferdam is similar to shoring except
that it is much larger, is made of heavier members, and is generally
installed (at least in part) prior to the making of the excavation.
One simple type of cofferdam is shown in Fig. 11-3 (a). It consists of a
452 COFFERDAMS [Chap. 11

single wall of steel sheet piling which encloses an area and which is driven
into the ground below the bottom of the proposed excavation far enough
to enable the earth at BD to hold the piling in place. The lateral pressure
of the ground above D bends the piling as a beam that is cantilevered
above this vicinity. The supporting soil is not hence the upper
rigid,

resultant reaction is below D. The lateral pressure is assumed to be

TABLE 11-1. Estimated Horizontal Active Pressures Acting upon Cofferdams,

Level Top
(Pounds per square foot per foot of depth)

Material Moist condition Saturated condition*

62.5
64
75
35 85
30 85
35 85
30 85
25 80
0-25 65

* Includes water pressure and earth pressure. Deduct 60 to estimate earth pressure alone.

TABLE 11-2. Estimated Horizontal Passive Resistances of Earth Masses,

Level Top
(Pounds per square foot per foot of depth)

Material Moist condition Saturated condition*

250 210
300 250
375 280
425 325
250 170
325 250

* Does not include water pressure. Add 60 to estimate the combined pressure of earth plus water.

hydrostatic in character. Some suggestions for the assumed unit hori-

zontal pressures caused by various materials acting on cofferdams are
given in Table 11-1. However, the most uncertain part of such a coffer-
dam is the resistance of the soil below D.
To avoid serious uncertainty when deeper excavations are necessary, it

is desirable to use one or more planes of bracing to support the sheet piles,
as indicated in Fig. 11-3(6). If the cofferdam is small, wales a may be
used alone with the corners mutually supported. Long narrow structures
Art. 11-3] SINGLE-WALL COFFERDAMS IN GENERAL 453

may have a series of struts b to brace the long wales. Larger cofferdams
have to be braced both ways, one scheme being the double-member one
shown in (c). It is necessary to support the bracing vertically by posts c,
by using diagonals between the upper and lower members to form trusses
that span the opening, by having seat angles tack-welded to the inside of
the piling, or by some other device. Lower sets of bracing may be erected
as the excavation proceeds but, if so, it is desirable to plan the arrange-
ment so that the members form a honeycomb pattern with cells 8 to 10 ft.
or more wide through which the excavation can be conducted. If the

depth of water is sufficient, all or part of the wales and their bracing may
be fabricated and floated to the site as a unit, then the sheet piling may
be driven around it.
The pressures given in Table 1 1-1 are likely to cause dispute. They are
believed to be conservative and are based upon the following reasoning:

FIG. 11-4. Explanation of assumed pressures on cofferdams.

1. Assume that a sheet pile wall AB of Fig. 11-4 (a) is driven through
water into dense clay. Then large rectangular blocks of stone a, b, and c

are set outside the wall as shown. The water will still fill the joint
between the piling and the stones and cause full hydrostatic pressure
against the wall. The stones do not tend
to fall down, hence they cause
no added pressure to the wall (neglecting any effect of compression of the
clay beneath them).
2. Assume that, as in Sketch (b), the big stones are now removed and

replaced by a mass of heavy solid spheres. The water will still cause
pressure against the piling because the spheres cannot seal it off. Fur-
thermore, the portion of the mass of spheres above some surface of failure
CD will fall down as soon as the piling is removed. They must therefore
cause some additional sideward pressure against the piling if the latter
prevents this failure.
3. If the spheres are removed and replaced by a granular soil, as in (c),

the water will fill the voids and will, for practical purposes, cause the same
454 COFFERDAMS [Chap. 11

hydrostatic pressure against the piling. Some wedge of earth like ECD
soon as the piling is removed. It, too, must therefore have a
will fall as
supporting reaction provided by the piling if this failure is prevented.
The pressure diagram for the earth may be anywhere from something like
EFC to EGC; that of the water, JKC. Both act simultaneously.
In Sketch (d), JKC again represents the water-pressure diagram.
4.

That for the earth pressure is pictured as EGC where GC is an assumed

hydrostatic pressure caused by a granular mass having a unit weight equal
to that of the submerged
and having a coefficient of friction that is
soil,

applicable for this saturated ground. For example, assume that a fairly
uniform fine sand has 40 per cent voids, and = 30°. The solid matter
<f>

when submerged would weigh about 0.6(165 — 62.5) = 62 p.c.f. Then

the intensity of the soil pressure, using Fig. 5 of the Appendix, is ph =
62 X I3 = 21 p.s.f. per ft. of depth. This added to 62.5 equals approxi-
mately 84 p.s.f. per ft. of depth. The total pressure is pictured by JMLC.
5. When the watertable is below the ground line, the pressure diagram
NOPC in (e) is computed similarly, except that OJ for moist soil is super-
imposed upon the pressures from water and saturated ground.
6. Cohesive soils may be able to prevent full contact of the water

against the sheet piles. However, this is unreliable because of the violent
disturbance of the ground in the immediate vicinity of the piles as they are
driven. Stiff clays tend to stand up fairly well temporarily, so that it
seems safe to assume that the pressure caused by them on the piling is
approximately that caused by the water alone.
One of the serious troubles that may occur in connection with coffer-
dams is unwatering them. The soil inside may be excavated to the
desired depth without removing the water. The interlocks of the piling
are then likely to be somewhat loose. As soon as pumping has lowered
the water level inside the cofferdam sufficiently, the external pressure
tends to pinch the interlocks together so that leakage through the pile
walls becomes small. However, as the water level inside is lowered, the
unbalanced pressure increases and tends to cause leakage up through the
bottom, or to blow the bottom up.
As illustrations of such troubles, refer to Fig. 11-5. In Sketch (a), the
piling penetrates into deep sand and gravel. The inside is to be excavated
and unwatered to DD. The gravel is very porous, and an almost infinite
amount of water can percolate through the silt, sand, and gravel so that
unwatering may be practically a hopeless job. As the water moves
through the gravel, the fine particles of sand are likely to be carried with
it. Prolonged pumping may then undermine portions of the gravel within
the cofferdam and even outside it. The finer the materials are, the
greater is this danger of pumping out the soil with the water. In the case
of Sketch (a), grouting of the gravel inside the piling might be one way of
Art. 11-3] SINGLE-WALL COFFERDAMS IN GENERAL 455

decreasing leakage sufficiently. Lengthening of the piling would avail

little.

In Fig. 11-5(6) the piling penetrates into deep dense clay. For tem-
porary purposes, the clay is practically impermeable. When the coffer-
dam is unwatered, the water outside cannot get through the clay, neither
can it generally work along the junction between the sheet piles and
sticky clay. When the clay is deep, a moderate embedment DB of about
20 to 40 per cent of CD will generally be safe for sealing, although experi-
ence with particular soils is the best guide. On the other hand, if a
stratum of porous material lay slightly below B and if water could pene-
trate into this stratum from some source, the unbalanced hydrostatic
pressure under the plug of clay BDDB might heave up the latter blow —
up the bottom. There should be enough weight of clay to counterbalance
the uplift, or deeper piling should be driven down to cut off the dangerous
layer.

A A A A A A A
C C C C C C C
Wafer Wafer Water Wafer
Si/t
S/'/f Silt Soflclay ===: Softclay
Whl:
D D Fine Softclay Sand
Sift

bottom of. sand D Sand

excavation
-^^Bothmof
~ excavation
Deep,
dense
Dense clay MB
|1|$
mm _ Gravel
sandand clay Rock
gravel -**~B B
(a) (b) (O (d)

FIG. 1 1-5. Illustrations of some problems involved in unwatering cofferdams.

Figure ll-5(c) pictures another case. Here the structure is to be

founded upon the rock. The piling is driven through dense clay to rock,
automatically cutting off the sand stratum and sealing the end of the
piling. The Leaks through
clay also serves as a seal against the rock.
seams be serious, but this possibility should
in the rock are not likely to
be investigated in advance by making several core borings. As long
as the clay is sufficiently dense and thick, this situation is favorable to
pumping.
A general situation that is more frequently encountered is pictured in
Fig. ll-5(cT). The founded on the rock, which is over-
structure is to be
lain by porous water-bearing sand and gravel. If the rock is irregular or
if the gravel has a considerable amount of boulders in it, the junction

between the piles and the rock will not be watertight. Pumps may or
may not be able to handle the leakage. It might be desirable to grout
the gravel outside the pile tips or put in a sealing plug of tremie concrete
several feet thick, before pumping. The latter should be heavy enough
to prevent uplift.
456 COFFERDAMS [Chap. 11

Figure 11-6 shows another source of difficulty with a cofferdam. If the

sheet piles are driven clear down, one at a time, there may be trouble in
completing the enclosure because of jamming of the last members and
interlocks. This can best be prevented by starting all the sections and
driving them lightly until the enclosure is complete, then driving each
piece in series a few feet, going around the wall as many times as may be
necessary.
The driving of sheet piling into coarse sands and gravels may be very
difficult. For example, at an industrial plant in Indiana the soil w as
r
fine
to coarse sand, with considerable gravel interspersed with the sand. It
proved to be impossible to drive the sheet piling more than 18 to 22 ft.
As a test of driving resistance, a single piece of sheet piling was driven
first; then a 12-in. steel H pile was driven alongside to a depth of 75 ft.

The driving records are shown in Table 11-3.

Closure Closure Closure

sheetpi/e sheeipi/e sheetpi/e

Last
Las I preceding I, -1st:

preceding Jst.

Last
mm
/$ Boulder
Last piece / Is/piece preceding -Stone or ledge
preceding
Flared at bottom Flared at top

f«J-SHEET PILES NOT VERTICAL BUT IN W-SHEET-PILE (cJ-SHEET-PILE

PLANE OF THE WALL WALLS NOT IN BENT BY
SAME PLANE SKIDDING
OFF ROCK

FIG. 1 1-6. Some sources of trouble in closing sheet-pile walls.

The great cost, as well as the possible loss of life, that would result from
the failure of a cofferdam is too serious for one to take chances with such
structures. Not only should they be strong but they should be stiff.

The ground outside should not be allowed to slip or start to fail. If it

does, the pressures will probably be considerably larger than those
assumed here. Wedges and jacks should be used to make sure that all
wales and bracing are pressed into proper bearing. Members should be
strong enough to resist the extra loads caused by their use to support
equipment, by excess pressures created by wedging and jacking, and by
bumps from buckets and other equipment.
11-4. Cantilevered single-wall cofferdams. Assume a section 1 ft.
wide of the cantilever piling AB shown in Fig. ll-3(a), with saturated soil
outside at C and inside at D. This is reproduced in Fig. 11-7. Let p a
equal the unit active pressure caused by the saturated soil; p p the unit ,
Art. 11-41 CANTILEVERED SINGLE-WALL COFFERDAMS 457

passive pressure or resistance; p n ,

the difference p p — pa . Line DF to
some scale equals p a h. Below Z), the active pressure on the inside annuls
any further increase of that outside. Therefore, CDBGF is assumed to
represent the net pressure diagram acting on the outside of the piling.
When the soil is saturated, the pressure outside includes that from the
water itself and the ground. The values given in Table 11-1 combine
them. If the cofferdam is not unwatered, the external pressures should
be reduced to those of the soil alone. This is assumed to be 60 p.s.f. per

TABLE 11-3. Driving Records of Single Steel Sheet Pile and H Pile*

Single sheet pile H pile

Type: M.P.-116, 36 lb. per ft., Type: 12 H 53 lb., flanges cut off at
A = 10.59 in.
2
, section modulus 45° at point
per pile = 14.3 in.
3

Total penetration, Blows per ft. Total penetration, Blows per ft.

ft. (average) ft. (average)

9-11 14 7-10 4
11-13 20 10-15 6
13-16 25 15-20 12
16-18 42 20-25 14
18-19 57 25-30 25
19-20 59 30-35 26
20-21 72 35-40 32
40-41 59
41-45 38
45-50 33
50-55 44
55-60 36
60-65 37
65-70 38
70-75 35

* Soils were fine to medium sands with considerable gravel interspersed. Hammer was No. 1

single action; plunger, 5,000 lb.; fall, 3 ft.

f Stopped driving to weld on second section of 12-in. pile.

ft.less than the pressures given for the saturated conditions. Further-
more, the passive pressures given in Table 11-2 for saturated soils are
those produced by the soil alone. When the water pressure assists them,
add 60 p.s.f. per ft. to the values given in this table.
The earth inside the cofferdam from
tendency of D to B will resist the
the piling to rotate clockwise in the figure, and do so as efficiently as it will
possible. It therefore seems reasonable to assume that, at and somewhat
below D, the soil will resist to its maximum — its passive resistance.
Since a pressure equal to the active pressure has already been utilized in
458 COFFERDAMS [Chap. 11

the determination of line FG, the net resistance p n is all that can be
depended upon. At D, this resistance will start to oppose the active
pressures. Line DN, drawn at the slope determined by p n times the
distances, intersects FG at K, a point of zero pressure on the piling.
Below K, KN, and KG picture the limit of the resisting intensities of p n
on the right.
Since there is a moment trying to overturn the piling, equilibrium
requires that a counteracting moment must exist. It is reasonable to
assume that the passive pressures on the right below K and the opposite
passive pressures on the left near the tip B of the piling constitute a force
couple that causes equilibrium. The forces near K can be the pressures

Assumed pressure diagram

Hi

a B G
FIG. 11-7. Development of method of analysis of cantilevered cofferdam.

p n times the corresponding depths, but they will be no larger than neces-
sary. Those on the left near B can be equal to pressure p n times the
height (h + d). This is such a large intensity that the resistance may be
assumed to be almost a concentrated force P 2 applied a short distance
above B, its magnitude being what is necessary but not exceeding the
passive resistance. Below K, the pressures on the right probably increase
as fast as they can at first, then increase at smaller rates near J, and
finally decrease to zero in the vicinity of M. The pressures on the right
probably cannot increase steadily toward N, then suddenly reverse to the
left side.

The upper resultant passive resistance below K will be called Pi. It is

assumed to be represented by the isosceles triangle KJM of Fig. 11-7,
where MJ = KJ. The base x is determined by K and the distance a
assumed in the location of P 2 If the values of Pi and P 2 can cause equi-
.
Art. 11-4] CANTILEVERED SINGLE- WALL COFFERDAMS 459

Librium without exceeding their respective limiting passive pressures, the

piling will not tip over; if their full strengths are not needed, the design is
safer, and it should be so. In all cases, XH = and 2M = 0, if there is

equilibrium.
The procedure for analysis of the piling under the preceding assump-
tions is the following:
1. Assume distance a = 1 ft.
2. Compute the pressures and centers of gravity of portions of the
pressure diagrams:

H, = CDF = Pa ^2 H, = DFK = ^-
2
—pn
p = KJM = - FK - d ~ F& ~ a
=
^^
1
(d a) ^ x

Zoo
1

3. Assume the moment at E = 0. Compute lever arms of the forces

about this point:

x
= ,
— FK
— = — h
7
o = - c a a e a
,
a +
,

4. By moments, compute Pi required:

Hie +Hc - 2 PJb =

_ H^ + H c 2
Fl
b~
If this computed value of Pi than x 2 p n /4:, not all the resistance of
is less

the earth is needed. In fact, the latter should provide a safety factor of
perhaps 1.5.
5. Test the required magnitude of P 2 :

2H = Hi +H +P -P =
2 2 x

where Pi is the computed required magnitude. If this computed mag-

nitude of P than approximately p n (h + d — a)2a and is less by
2 is less

the desired safety factor, the structure will be safe in this respect.
6. Compute the shears and bending moments in the piling by using the

diagrams for the forces Hi, H%, P h and P 2 as computed. The ordinate
OJ of KJMshould be reduced to make the area of the triangle equal to
Pi. It may be that the real pressure diagram will be curved with a maxi-
mum above JO, but this would cause less bending in the piling than will
the assumed diagram. The structure will therefore be safe.
Cantilevered cofferdams can be used to advantage in ordinary earth,
but other methods may be applicable and more economical for such con-
ditions. In saturated and submerged ground they are particularly
needed. The water adds greatly to the outside pressure and reduces the
460 COFFERDAMS [Chap. 11

abutting resistance of the soil inside because of buoyancy, as indicated in

Table 11-2. The safe height h of Fig. 11-7 is then greatly restricted.
Usually, d should equal somewhere near 2h. These facts will be illus-

trated in three examples.

Referring again to Fig. 11-7, it is possible to set up an equation to deter-
mine the minimum depth of penetration for any particular soil conditions,
having the cofferdam safe under the assumed pressure conditions. The
procedure is the following;
1. Determine DF, FK, Hh Hz, and a for the assumed properties of the
soils. Compute distances/ and g to the centers of gravity of Hi and H%.
2. Assume the safety factor (S.F.) desired against overturning, and
divide the expression for the required Pi by it.

3. Set up the equation for moments about E in terms of x:

H + FK + x) +H {FK - + x) - ~r (pn = (11-1)

|j ||
1 (f 2 g

Solve this by trial.

This last procedure is preferable for design; the preceding one, for
analysis.
Example 1, Assume that the sheet piling in Fig. ll-8(a) is driven
through moist silt and well-graded sand as shown. Assume silt and sand
weigh 100 p.c.f.: p a = 35 for both materials; p p = 375 for sand. Is the
piling stable, and what is the maximum bending moment in it?
The computations of pressures and distances are given in Fig. 11-8(6).
Using these, and taking SM = about E,

P x b = H +Hie 2c

p needed =
Pi
, , 3,940
— X 29 + 405 X 23.5
= 1 ,nnn n
11,000 lb.
1 1 , — •

This is much less than the available Pi of 43,000 lb.

P 2 needed = Pi - H - H, =
x 11,000 - 3,940 - 400 = 6,060 lb.
P 2 available per sq. ft. of area = 375 X 39 = 14,600 p.s.f. (safe)
Assuming a revised value of a = 340 X (11,000/43,000) = 87 instead of
340 p.s.f. per ft. the point of zero shear and greatest bending might be

#i + H, - 87 X | =
y = 10 ft. below K
Notice that the full value of p n is used in computing Ho.
87 V
* 1A2 in
M v
= 3,940 X 16.5 + 405 X 11 - X y = 55,000 ft.-lb.
Art. 11-4] CANTILEVERED SINGLE-WALL COFFERDAMS 461

As stated previously, the diagram for Pi is likely to have its maximum

ordinate above OJ, and this would yield a smaller computed bending
moment.
Example 2. Assume
the same conditions as pictured in Fig. 11 -8 (a)
except that the watertable is 5 ft. below the top of the silt. Assume that
pa for the moist silt is 35; p a for saturated silt, 85; p p for the saturated
sand, 280; p n for the sand, 280 - (85 - 60) = 255. Will the cofferdam
be safe?
DF=3SxlS*52S*/o i
H, = 525xf=3940*
FK'%-1.54' ff2 'S2Sx^'40S*
^w^ -
x = 24-1.5=22.5'
OJ'340xll25^3,820 ,,/o'
Allowed P, =3,820 x 11.25=43,000 *
'

Silt 55
.0 e = 24+5=29'
ft 0=24-^=235'
b* 11.25'

Sand .55
fc

<Q
1

faJ-SOIL DATA AND « W-DIAGRAM FOR MOIST SOIL

DIMENSIONS

Pa =35 or 85
Pp--280 QR=35xS=l6S*/o' SF=85x/0=S50*/o'
P„--2S5 DF =850*165 =1,015 "/a'
I65xf = 4l0, x 11.67 - 4,800
165x10=1,650, x 5. = 8,200
= 14.200
850xf -4.250, x3.33
H, =6.310 6.310 /27.200
4.3'toC.G.

FK-^-4' f*U>
> J Hy- 1,0/5xj-^ 2,030*
x=24-4=20'
6 = 10'
c=24-/.3=22.7'
e = 24 + 4.3=28.3'
<j OJ=/0x255=2,550*/a'
rd-DIAGRAM FOR WATER TABLE Alltmed PfZSSOx IO*2$$00*
5-0" BELOW TOP OF SILT

FIG. 11-8. Analysis of a cantilevered cofferdam.

The new pressure diagrams, and the computations for pressures and
dimensions, are given in Fig. ll-8(c). Assuming these data,

p needed
P\
,
= 6,310 , X 28.3 +
— -
2,030'-
X 22.7
= 00 _
An
22,o00
..
lb.

This is slightly less than the 25,500 lb. available. Hence the piling is safe
against rotation, but the safety factor is only 25/22 = 1.13.

P =
2 22,500 - 6,310 - 2,030 = 14,160 lb. required

The value of P 2, counting a 2-ft. width, is approximately

39p„ X 2 = 39 X 255 X 2 = 19,900 lb.

462 COFFERDAMS [Chap. 11

Example Assume the moist soil and conditions given in example 1

3.

and Compute the minimum embedment of the piling to

Fig. 11-8 (a).
give a safety factor of 1.25- Assume the active pressure diagram CFKD
in Fig. 11-8(6), and refer to Fig. 11-7.
Let KJM represent the pressure diagram for P x. Use Eq. (11-1).

1 a
Fx(/ + FK + x) +H 2 (FK - g + x) - Vn
S.F.
1
3,940(5 + 1.5 + x) + 405(1.5 - 0.5 + a;) X 3404r =
1.25
x = 13.6 ft. (approx)

The embedment then equals, approximately,

FK + x + 1.0 = 1.5 + 13.6 + 1.0 = 16.1 ft. (call it 16)

In Fig. 11-8A are shown some data for approximating the depth of
cantilevered sheet piling necessary to resist the lateral pressure which may
be caused by earth alone or by earth with a uniform surcharge thereon,
these being shown in cases 1 and 2, respectively. Case 3 is for a concen-

4 6 8 10 12 14 16 18 20 22 24 26 28 30
/
TO FIND VALUE X F ordinal e
seclies foi
1. Compute ordinate F CU) ~ves, mi 'asure
F =\M<L\ he
^^
\Pp-Pa\
*
A
I i
2. Set point on ordinate " B C
line for the /"volue,
using scale A.
1,200 2,400 12,000 1 1 /
/
3. Set point on the d ijoo 2,200 i 11,000
abscissa line for the /

^
1
s
V e f /
2,000 i 10,000 /
4T hro ua h fhesetwoooints
draw a line to intersection
I

with A curve. 900 1,800 1 9,000 J

\
5. The reading on the X [
J \
/
abscissa of the point of 800 1,600 = 8,000 1
intersection is the / '
\

reguired value of X. '

7rin 1,400 1 7,000 /

/
/ \
If the computed /"value / /
1,200 = 6,000 /
/
f }
the A scale, or if the /
line drawn does not 500 1,000 = 5,000/
r I
/
/
intersect the A curve, E /
try B scale and # curve. 40 o 800 I 4,o6o
/
use C scale and Ccurve.
300 600^? 3,000 /
'
Example /
J

For 4= 5.0 ft., Z.= 2,970 lb. 200 400 2^000

andPp -Pa =293p.s.f.,
:
J
f
/ /
^=305 and X= 10.6 ft. ,
00
'
200 E
1,000 /
/ /
/
1

1
1 i
1
1
1 i
1 1
1 1 1
1
1 1 1
1
1 i 1
1
1
1
1 i
<f i |
i
| I'M
8 16 14 12 10 8 6 4 2 \ I) 3 1 D 1 2 1 4 1 5 1 8 2 22 24 26 28 3

L din feet >« X in fe et -4

(a)

FIG. 11-8A. Diagram for approximating required depth of cantilevered

Art. 11-4] CANTILEVERED SINGLE-WALL COFFERDAMS 463

trated load at the top of a

pile. All are in terms of a section of pile wall
1 ft. These data are based upon copyright material prepared by
wide.
A. H. Malkin. The reader will notice that the assumed pressure dia-
grams are not exactly similar to those in Fig. 11-7, but the differences
are minor.
Uni om
surcharge-^

j
^-^ s?
7?
k

.Backfill**'''*'
M <3

i
-.i* i / A
SrvW^ y -a x Soil ^«£±E
Soil -
Soil y^ A 5i
N «= ^
N
J, j^
"
'
Soil
NJ ;
'
— "

(a) CASE 1 (b) CASE 2 [e) CASE 3

Symbols:

We = weight of soil, p.c.f. ; Wf = weight of backfill, p.c.f.

S = uniform surcharge, p.s.f. ; S.F. = safety factor

$ = ongle of internal friction of soil in degrees
= angle of internal friction of backfill in degrees
<f> f
Pa = active horizontal pressure of soil, p.s.f. per ft. of height
Pp = passive horizontal pressure of soil, p.s.f. per ft. of height
Pf = active horizontal pressure of backfill, p.s.f. per ft. of height of backfil
P = uniform horizontal pressure on piles for surcharge, p.s.f.
s
L - resultant ofall horizontal forces on piles above A, where the soil

pressure intensities on both sides of the piles are equal

h = height of soil or backfill above a point of computed pressure
U = depth of pile penetration, ft.

Formulas
cos cos <j>
2 We_
W Wrf
'H
<f>
Pa = " ce or y
\ + /sin (sin 56-cos /tan/?)J " '

|_1 -/sin ^(sin ^-cos tan ,5:1 S.F

(f> <f>

For 3 = <f>,
Pa = cos z <p(We orWf ); for 3 = 0, Pa = tan 2 (45°- \){We or Wf )

2 2
For S = j>, ^=cos <f>^\ forS =0, Pp = tan (45° + |-)
^f-
PS =WS tan
2
(45°--^)
HP, 5
Y = 7—^ for Case 1 ; Y* / Q for Case 2

£/ = ( Y + *) for Cases 1 and 2 ; d/ = 1.1 X for Case 3 ; with X determined

from curves in Sketch (a)

2
=
(Pp-Pg) X -ZL 1L
Z 2(Pp-Pa )X+ ^+Y)Pp -Ypa for Cases 1 and 2 or Z- 2U+H (2U+H)[Pp -Pa )

7 U
= *-- L
for Case 3
2 UKPp-Pa)
\b)

sheet piling. (Courtesy of A. H. Malkin, New York.)

464 COFFERDAMS [Chap. II

The reader is cautioned to make sure that he correctly uses the curves
in Fig. 11-8.4. The three sets of scales at the left of the diagram are to
be used with the correspondingly labeled curves. The dashed straight
lineson the diagram illustrate the procedure for the case shown. All
ordinates for values of F are to be measured along the same line, where
the small subdivisions are shown. The safety factor to be used in the
computation of Pp should be determined in accordance with the reserve
strength which the particular situation makes necessary.

WWc

teJ-COFFERD AM rtW-BOTTOM SUPPORTED (c)- BOTTOM

HORIZONTALLY FIXED
1
HORIZONTALLY
C
k <
S
i

h h
2 ^\ Hi

n F \f
\ a rr.
m
{ V-^ 2
•S<x^~^ t e
d K c
PC
T"^^^ ^^ t>\
B M N
\

^-ASSUMED PRESSURE DIAGRAMS FOR BOTTOM

SUPPORTED HORIZONTALLY BUT NOT FIXED

FIG. 1 1-9. Single-braced, single-wall cofferdam.

11-5. Single-braced single-wall cofferdams. When the unbal-

anced head on a single-wall cofferdam is to be large, the top should be
braced. Consider for the present that this bracing is limited to a single
layer. This produces a structure similar in action to the tied bulkhead
of Fig. 10-39 except that the bracing takes the place of tie rods and anchor-
ages. The same method of analysis of pressures on and bending moments
in the piling may be used for both types of structure.
Figure 11-9 shows such a cofferdam wall with a wale and brace at T.
The end DB is embedded in the ground. If this embedment is small, the
piling may bend somewhat as shown in (b) with a point of contraflexure
Art. 11-5] SINGLE-BRACED SINGLE- WALL COFFERDAMS 465

at U. As the length of DB is increased, the ground restrains this portion

more fully as in (c), and it will cause another point of contraflexure at
some point V. The pressures and the bending are somewhat different in
these two cases.
Assume that the lack of leakage problems permits the use of a short
embedment that is just strong enough to prevent the bottom of the piling
from kicking inward. CFMB represent the diagram
In Fig. ll-9(d), let

of active pressures as partly counteracted by the earth inside and below D.

Let DN be drawn at the angle a determined by p n since the passive pres-
sure below D will be developed as fast as possible. Assume that KNM
represents the diagram of effective passive pressure, with the resultant Pi.
Both S and Pi are passive forces and will be no larger than necessary to
produce equilibrium.
To determine the distance x and the force Pi in (d), assume a section of
wall 1 ft. long, and take moments of external forces about T. Then,

H x {h, - f) +H 2 (h 2 + g) - ^ p n x | (h 2 + FK + ~\ = (11-2)

As usual, the term S.F. is the safety factor desired. Solve this equation
by trial. Then,

Pi = P«f (H-3)

This gives a minimum value for d = FK + x.

In order to estimate S, assume that Pi represents the position of a reac-

tion point with negligible bending, and treat as the concentrated KNM
force Pi. Take moments about this point.

Ht(e - b) +H 2 (c - b) = Sm = S(h 2 + d - b) (11-4)

This is usually a close enough estimate of the pressure per foot on the
wale. The load on a strut (or tie) is then $ times the spacing of the
supports.
Let Fig. 1 1-10 represent the case in w hich avoidance of leakage requires T

deep embedment of the piling. As in the case of cantilevered piling,

assume that CFME is the diagram of active pressures; KJM, of passive
resistances. The point of inflection V in Fig. ll-9(c) may be somewhere
near the point of zero pressure K in Fig. 11-10. As a first trial, assume
them to coincide. Taking moments about K, for diagram CDKF,
H^FK + /) + H,(FK - g) = Sn = S(FK + h 2) (11-5)

Assuming this value of S, take moments about E, the assumed line of

action of the passive resistance P 2 .

H +Hc-Pb -
ie 2 x S(h + d - hx - a) = (11-6)
466 COFFERDAMS [Chap. 11

Compare the required Pi with the maximum allowable p„(.i" 2 /4). If Pi is

safe, draw a revised diagram KJM to represent its required magnitude,
letting DK remain unchanged. Now compute the moment diagram for
the bending under the action of these forces. If the required
in the piling
Pi is larger than that allowed by the assumed properties of the soil, the
ground will probably yield slightly near D, and the point of inflection may
move downward. However, it is probably desirable to place the struts in
a lower position or to add more bracing. In other words, if it is desired
to have the earth restrain the piling fully, at least one should design the
structure so that, according to the computations, it can do so.

FIG. 11-10. Assumed pressure diagrams for cofferdam with bottom fixed horizontally.

When the proportions of the structure make one suspicious that the
point of inflection will not be close to the assumed point K, repeat the
calculations with the assumption that it is 2 or 3 ft. higher, then 2 or 3 ft.

lower. Proportion the structure and its bracing so that it is safe under
any of these conditions. Check to see that the conditions 2H = and
2M = for equilibrium are reasonably satisfied. When part of the dia-
gram KJM enters the equation for S, use as the one found to be neces-
it

sary for equilibrium after the solution of Eq. (11-6).

The results of these computations may not be exactly correct, but they
seem to give results as accurate as the unavoidable uncertainties in the
assumed data warrant.
A cofferdam may have to be used in sloping ground. The pressure on
the high side will then exceed that on the low one. This unbalance
should be provided for. Ordinarily this can be done easily because the
Art. 1 1-5] SINGLE-BRACED SINGLE-WALL COFFERDAMS 467

bracing will transmit the larger force to the downhill side where the sheet
piles willdevelop the passive resistance of the soil as far as necessary.
This, however,may change the maximum bending moments in the piling.
Example 1. A cofferdam is to be built as shown in Fig. 11-1 1(a).
Assume p a = 80 p.s.f. per ft.; sea water = 64 p.c.f.; p p = 250; and p n =
250 — (80 — 60) = 230. Find the absolute minimum penetration d to
hold the bottom of the piling (S.F. = 1) compute the maximum bending ;

moment in the wall and find the pressure in the struts if they are 10 ft. c.c.
;

Sea water=64pcf p ^80,p-=2SO,

pn =250-(80-6O)--230
RU=5k64--320, VF=ISi80=l,200
DF- 1200+320=1520 <-5
320x§ = 800, x/6.67=/3J50
320x15 '4,800, x 7.5 =36,000

IJOOx'j --
9,000, x5. --
45,000
H^ 14,600* 14.600194,350

f=6.5'

FK
rrc '.S20_ 6
. rr „.£i.=2.2
~l30' 6 9-T
-
'

6.6x1,520
H, -5,020*

(a)- SOIL ^-PRESSURE DIAGRAMS FOR

PROFILE BOTTOM SUPPORTED a--l'd" t
B
HORIZONTALLY BUT
fW-PRESSURE DIAGRAMS FOR BOTTOM
NOT FIXED
FIXED HORIZONTALLY

FIG. 11-11. Analysis of single-wall, single-braced cofferdams.

The assumed pressure diagrams are shown in Fig. 11-11(6),, with the
preliminary computations and computed dimensions. Taking moments
about T,

#!(14 - 6.5) + H,(U 4- 2.2)

- 230x | (14 + 6.6 4- %x) = 64 X 6 2 /2 X 2 (approx)

x3 + 30.8x 2 = 2,440
x = 7.9 ft.

Assume x = 8 ft. (approx)

and d = 6.6 + 8 = 14.6

Use x = 8.4 ft, and d = 15 ft,

Taking moments about Pi,

#!(5.6 + 6.6 + 6.5) +# 2 (5.6 + 4.4) = £(5.6 + 6.6 + 14)

S = 12,300 lb.
Shear at S = 12,300 - 1,160 = 11,1401b. and TZ = 400 p.s.f.

3^[400 + (400 $0y)]y +

11,140 =
y = 12 A ft. for point of max below M T
468 COFFERDAMS [Chap. 11

M = 12,300 X 12.4 - 800 X 15.07 - 320

X 13.4 X 6.7 - 80 X 13.4 X 6.7 X 4.47
M = 80,000 ft. -lb. per ft. of cofferdam (very high)

Force in a strut = 10S = 123,000 lb.

Example 2. Assume the same cofferdam as in Fig. 11-1 1(a) except

that fine sand replaces the dense clay, and the piling is to be driven so that
d = 25 ft. in order to have the bottoms of the sheet piles reach a dense
sandy clay. What is the estimated load per foot on the wale at T, and
what is the maximum bending moment in the piling? Assume that
p p = 250 p.s.f per ft. p n = 250 - (80 - 60) = 230 p.s.f per ft.
.
;
.

Having the same values as for example 1, the diagrams above K in

Fig. 11-11(6) have been copied in (c).

x = 25 - 1 - 6.6 = 17.4 ft.

OJ = ^ X 230 = 2,000 p.s.f.

P lmax allowed = 2,000 X =^ = 17,400 lb.

Taking the first trial moments about K as a point of contraflexure,

14,600(6.5 + 6.6) + 5,020 X 4.4 = £(14 + 6.6)

S = 10,300 lb.

Taking moments about E, in Fig. 11-11 (c),

14,600 X 30.5 + 5,020 X 21.8 - Pi X 8.7 - 10,300 X 38 =

P = 1 18,700 lb.

This is a little over the maximum Pi allowed under the assumptions.

The point of zero shear is found to be 11 ft. below T. Then,

M = 10,300 X 11 - 800 X 13.67 - 320 X ~ -

M = 56,000 ft. -lb. per ft. of cofferdam

This would require heavy Z-shaped piling. More bracing should be 1

added, or the position of the bracing should be lowered considerably.

Test these results in comparison with those found by assuming the point
of inflection to be at points otherthan K, using the same general method
of analysis. Provide for what seems to be the worst reasonable case.
11-6. Multibraced single-wall cofferdams. It is generally desir-
able, or even necessary, to support the sheet-pile wall of a deep cofferdam
1
Generally, the maximum computed unit stress should not exceed 16,000 p.s.i.
Art. 11-6] MULTIBRACED SINGLE- WALL COFFERDAMS 469

by several sets of bracing to prevent excessive bending in the piling, and

to avoid dependence upon the somewhat uncertain restraint of earth
around the embedded ends. The braces may be spaced in whatever way
the designer prefers, or as construction procedures in the field require.
In some ways, it is desirable to arrange the bracing so that the maxi-
mum bending moments in the piling at various points utilize the strength
of the piling and are nearly equal. Bracing is very expensive, and econ-
omy is essential. Therefore, piling that has great strength in beam action
is desirable for this feature, whereas sections with straight or shallow

arched webs may be better for driving, especially for penetrating gravel
or soft rock. On the other hand, there are advantages in arranging the
wales and bracing so that each layer is composed of corresponding mem-

bers of the same size instead of having them vary. Deep cofferdams
may require heavy steel beams as wales, with steel or wooden struts.
For example, one given size of beam may be available and its use through-
outmay be desired.
No matter what is planned regarding the equality of bending moments
in piling or of reactions on bracing, there are likely to be considerable
variations and uncertainties besides those caused by the uncertainties in
the combined soil and water pressures themselves. Wedging or jacking
of struts and wales may cause some reactions to be larger than the theo-
retical; some may be smaller. The piling may be pressed against the
soil in places so that it develops more than the water and active earth
pressure, thus causing higher stresses. In other cases, it may be slack,
even to the extent of permitting small movements with unknown conse-
quent pressures because the resisting friction and cohesion have been over-
come by such movements. It is also very difficult to drive piling and to
erect bracing with absolute perfection. Therefore, conservatism in design
is desirable. How, then, should such a cofferdam be designed?
As in the preceding cases, assume that the earth pressure is hydrostatic
in character. Let Fig. ll-12(a) represent a cofferdam with four sets of
bracing at unequal spacing. The sheet piling is driven down to rock.
Since the piling is continuous, it will act like a beam that is continuous
over a series of supports, with the top cantilevered above G and the bot-
tom assumed to be cantilevered below Q. For simplicity, represent the
piling and its loads as in Sketch (b). The bending moments may now be
found by moment distribution, and this method is recommended. For
this, the fixed-end moments can be found by means of the data in (c) for
triangular loading, and by (e) for trapezoidal loading.
If the use of moment
distribution is not desired, a rough estimate of the
bending moments may
be made as follows, provided the lengths of canti-
levers and intermediate spans are proportioned reasonably and successive
spans do not change too rapidly:
470 COFFERDAMS [Chap. 11

1. Compute the bending moments in the ends GC and QD as canti-

levered beams.
2. Assume the piling to be fixed at all bracing supports.
3. Compute the moments at the right and the left sides of each inter-
mediate support K and N.
Loading diagram
*hP mm nm>
Loading Diagram

D K
O 'W
L 4 L3 ,
^rrrnTftrT>.

iW-LOADING ON CONTINUOUS BEAM p>wL u>L'

Moment Diagram
(-^-FIXED-END BEAM,
UNIFORM LOADING

Loading Diagram

0A23L Loading Diagram

CaJ-PRESSURE DIAGRAM 0.043 wL 2
PER FOOT OF
COFFERDAM •i:
2
^
-.ouL ±ivL wW2 )L
(iu>i + >)L<

Moment Diagram Moment Diagram

(ch FIXED -END BEAM, (e)- FIXED- END BEAM,
TRIANGULAR LOADING TRAPEZOIDAL LOADING

FIG. 11-12. Data for analysis of multibraced cofferdam, cantilevered at top and bottom.

^nHHHJZIa;
-<]

L
U &— Ja=lengfhi
terms ofL
L
Loading Diagram

^(S-Ja)L 2 ^(I0-10a+3a 2)L

Moment Diagram
B ^-MOMENT DIAGRAM FOR BEAM WITH
PARTIAL TRIANGULAR LOADING
(a) (b)

FIG. 11-13. Assumed pressure diagrams at bottom of embedded piling of multibraced cofferda

4.Average the computed fixed-end moments at K and N. Use the

cantilever moments at G and Q. If the cantilever moment at a support is

small as at G, add 10 to 20 per cent to the average bending moment com-

puted for the next support K in this case. If the cantilever moment is
anywhere near equal to the fixed-end moment at the other side of the
support, the moment at the adjacent support need not be increased.
Call these the maximum bending moments for design. Those in the por-
Art. 11-6] MULTIBRACED SINGLE-WALL COFFERDAMS 471

tions of the piling between supports will be smaller and need not be com-
puted, unless the lengths of spans are very irregular.
If the endD is supported by the rock and if this support can be trusted,
the span QD may be treated as a beam fixed at Q and simply supported
atZ>.
It is sufficient to estimate the reactions or forces upon the bracing as
though intermediate spans of piling were simply supported, with
all

trapezoidal loading. For the end braces, add the full weight of the load
on the respective cantilevers, increased for leverage.

FIG. 11-14. Cofferdam arranged in cells for excavation to different levels. Use of wooden bracing
is shown. (Courtesy of Wescoff and Mapes, Inc., Consulting Engineers, New Haven, Conn.)

There are many cases in which a multibraced cofferdam will have the
bottom of the sheet piling embedded in soil below D, Fig. ll-13(a). If
this embedment is slight and if it cannot completely support the end B,
the portion QB may be considered to be a cantilever with the loading
diagram QRFEB for active pressure and DVB for passive pressure. This
is seldom the case, however.

When the embedment of DB is considerable, it can support the shear at

the bottom of the piling, and perhaps fix that end against rotation. In
this case, the bending moment at Q and in the portion QB may be esti-
mated in the following manner, referring to Fig. 11-13(6):
1. Assume the active-pressure diagram to be the rectangle QUEB.
 .

472 COFFERDAMS [Chap. 11

2. Assume the passive-pressure diagram to be the triangle DVB, where

angle a represents p n p.s.f. per ft.
3. Assume that the passive pressure will be developed as rapidly as

possible to support the piling. Excessive length of DB over that required

will be assumed to fix the piling in position but not to add materially to
the reaction or the bending moments.
4. Find the distance x such that the minimum passive resistance

DVB' will support one-half of the pressure caused by the diagram

QUE'B'. Thus
ry 5leeI sheet piles
QUE'B'
= DVB'
2

E'B' = pji and B'V = PnX

1
p a h(L 5 + x) Pn

PnX' — p hx
(l
— p a hL b = (11-7)
Bracing sunk as cage
io bottom of channel
Solve for x.

5. The resultant passive pressure

Pi will act at x/3 above B' . Assume
this line of action to be the support
point for the piling, thus determin-
ing the span L = QW.
W<Soff red rock
Trench excavated
6. Assume the span QW to be
by divers fixed at both ends, and compute the

FIG. 11-15. Cofferdam with concrete seal used

fixed-end moment at Q as ^2
for Ferry St. Bridge piers over Quinnipiac River p a hL
2 - (p nDW/W)a*(5 - 3a)L\
by C. W. Blakeslee and Sons, New Haven, Conn. where the latter term is explained
by reference to Fig. ll-13(c).
Assume the reaction at Q from span Q to be }^p a hL — p„(DW) /6L.
7. W 3

The cofferdam in Fig. 11-15 shows a different solution for supporting

the bottom of the piling when excavation is to be carried to rock.
The cofferdam pictured in Fig. 11-16 also deserves examination. It
was built to a depth of 60 ft. in saturated volcanic sand. A few points to
notice particularly are the following:
1 The bottom
set of wales and bracing is separate from the main brac-
ing cage. This was weighted by old rails to overcome buoyancy. It was
made with 2-in. clearance and was suspended on wire ropes attached to
hand-operated winches so that it could be lowered as fast as excavation
proceeded, thus guarding against any sudden failure of the bottoms of
the sheeting.
2. Batter piles on the outside held the sheeting in line until considerable

excavation in the wet was completed. The main bracing cage was then
 o i> o>

^j

£ O

1 *
H_'

J "J!

£ O U

* F
X <i>
'^ X
5^
XI Lu
T3 2

a Q.
.c >•

'I
V
> -D

_ >Z -a

<j E

U ID D 2

O'-En 4)
<-
0>
i-

473
 s

474 COFFERDAMS [Chap. 11

erected inside at a higher position than the final. Although weighted, it

was driven to proper position by jacks that reacted against beams placed
across the cofferdam and hook-bolted to the sheeting. The lowest set of
bracing also had to be forced down to position because deflection of the
piles took up the clearance. This was then strutted against the main
cage.
3. Divers placed wedges between the sheeting and wales to ensure
bearing.

dSx5--42S*/°')

-mumw-- 255

(aJ-SOIL PROFILE (b>- PRESSURE DIAGRAM

AND DIMENSIONS PER FOOT OF WALL

QK'1.67, NK'I.2S, KK-I.O

F.E.M. -11,900 +6.800 qy6.500 +8950 Q>. 8.250 1-7,790 <&,6,370 +1770

1st Did. +5.100 -1,400 -1,050 +260 +200 4.600

1st CO. -700 +2.S00 + 130 -500 ^2,300 +100

Ind.Dist. +700 -1,510 -1120 -1,000 -100

2nd. CO. -750 +350 -500 -560 -50 -400

3rd Dist. +750 +90 +60 +340 +270 +400

3rd. CO. +40 +370 +170 +200 +130

4fh.Dist. 40 -310 -230 -100 -130

2 +11,900 -6,4/0 +6,4/0 -9,810 +9,810 -,',770

M 11,900 6,410 9,810 1,770

^-MOMENT DISTRIBUTION

FIG. 11-17. Analysis of a multibraced cofferdam.

4. The foundation piles were then driven by a hammer equipped for

underwater work, assisted by jetting.

5. The tremie-concrete seal of over 1,800 cu. yd. was then placed in one

operation.
6. Unwatering and pier construction then followed.
Example 1. Assume that a cofferdam is to be built as shown in Fig.
ll-17(a). Estimate the bending moments in the piling at the wales, and
the pressure per foot on the wales. Assume p a = 85 p.s.f. per ft. Neg-
lect any lateral support at D.
The pressure diagram will be as shown in Fig. 11-17(6). The fixed-end
moments, as computed from the data in Figs. ll-12(c) and 11-17(6), are
as follows:
Art. 11-6] MULTIBRACED SINGLE- WALL COFFERDAMS 475

Point M, ft. -lb

Gtop 425X2.5X1.67 = 1,770

G bottom (K 2 X 425 + 30 X 1
850) 10 2 = 6,370
K top (K2 X 425 + 1 20 X 850)10 2 = 7,790
K bottom 012 X + } 30 X 680)8-
1,275 = 8,250
N top (K2 X ,275 + 2 o X 080)8
1
1 2
= 8,950
JV bottom (K2 X 1,955 + 30 X 510)6 3
2
= 6,500
Q top (K2 X 1,955 + 20 X 510)6 i 2
= 6,800
Q bottom 2,465 X 3 X 1.5 + 255 X 1.5 X 2 = 11,900

The stiffness factors — the relative values ofI/L of the beams are —
given as K in Fig. ll-17(c). The carry-over factors are shown in the
circles. With the fixed-end moments previously estimated, the moment

'•flSMWf!
Top wale

1 st brace

^Finished grade 1st pour

la) FIRST EXCAVATION \b) SECOND EXCAVATION (c) FIRST TWO POURS
AND BRACING AND BRACING

Sheet pi 'ing cut

1
••?^j^
below grade 4 th pour

Finished floor >.

\

id) THIRD POUR AFTER ie) FOURTH POUR AFTER REMOVAL

REMOVAL OF SECOND OF FIRST WALE AND BRACING.
WALE AND BRACING FLOOR FINISHED

FIG. 11-18. Illustration of bracing of cofferdams and pouring of concrete structures when against the
piling.
476 COFFERDAMS [Chap. 11

distribution is carried out as shown in (c). The summations give the

moments for design purposes.
The reactions on the wales are estimated as follows, assuming simple-
beam action:
Total reac-
Point One end, lb. Hon, lb.

Gtop 425 X 2.5 = 1,060

G bottom 425 X 5 + 850 X 5 X ~

3.33
= 3,540 4,600

K top 425 X 5 + 850 X 5 X ^~ = 4,960

K bottom 1,275 X 4 + 680 X 4 X ~-

2 67
= 6,010 10,970

iV top. 1,275 X 4 + 680 X 4 X —^ = 6,920

N bottom 1,955 X 3 + 510 X 3 X ^ = 6,380 13,300

Q top 1,955 X 3 + 510 X 3 X | = 6,880

Q bottom 2,465 X 3 + 255 X 1.5 = 7,780 14,660

As a matter of curiosity, see how the bending moments computed in

compare with the approximation of averaging the fixed-end
Fig. ll-17(c)
moments instead of using the moment-distribution method. The results
are the following;:

Fixed-end M, Averaged M, Adjusted M,

Point
ft.-lb. ft.-lb. ft.-lb.

G as cantilever 1,770 1,770 1,770

K top 7,790
K bottom 8,250 8,020(+20%)* 9 620,

N top 8,950
A bottom
r
6,500 7,720 7,720
Q as cantilever .... 11,900 1 1900, 1 1900,

* Increased because point G is not fully restrained.

Example Assume that the cofferdam of Fig. 11-1 7(a) now has the
2.
sheet piling drivendown into stiff saturated clay instead of onto rock.
Assume the pressure diagrams to be as pictured in Fig. 11-13(6). Then
Vn = 250 - (85 - 60) = 225 p.s.f. per ft. Find the depth to which the
piling should be driven to provide adequate lateral support. Also esti-
mate the bending moment in the portion QB of the piling, and find the
approximate revised reaction on the wale at Q.
Art. 11-7] DOUBLE-WALL COFFERDAMS 477

From Eq. (11-7), since pa h = 2,720 p.s.f. and L 5 = 3 ft., find x.

225x 2
- 2,720a; - 2,720 X 3 =
x = 14.5 ft. (call it 15)

Then QB' = 15 + 3 = 18 ft.; QW = 3 +% X 15 = 13 ft.

MQ = i X 2,720 X 13 s - ^°
X 0.77 3 (5 - 3 X 0.77) 13 2 = 30,600 ft -lb.

End shear:
225 X 1
3

y 2 x 2,720 x 13 - ~
\; Yo— =
6 X 13
14 800 lb
>
-

Reaction at Q:

14,800 + 6,880 (from example 1) = 21,700 lb.

These computations show that the forces and moments in the bottom
wale and the lower portion of the piling are very severe compared to those
higher up. Therefore, the estimate of moments and reactions, as com-
puted in example 1, should not be used but should be revised because of
the large moment at Q. The trouble comes from the inability of the clay
to develop sufficient passive resistance fast enough to make a short effec-
tive span below Q.
11-7. Double-wall cofferdams. It is often necessary or desirable to
use cofferdams with double sheet-pile walls. One such is shown in Fig.
ll-19(a). The need for this generally is not the depth required inside the
cofferdam but the greater probability of securing watertightness. A
single wall of piling driven to irregular rock, into gravel, or onto boulders
will almost inevitably have irregular openings below the ends of the piles
through which water can pass in great volume provided a granular water-
bearing stratum exists at that position. Grouting outside of the pile tips
may be helpful but is not completely dependable.
A double-wall cofferdam is usually built by driving down two walls of
sheet piles around the area to be excavated. be 4 to 6 ft. These walls may
or more apart in order to have room enough to excavate between them.
The earth inside is generally removed by excavation "in the wet," with
struts to separate the two rows at top and lower down if necessary.
When the bottom of the space is cleaned sufficiently and when the piling
is seated firmly in case the removal of boulders makes this necessary, the
space between the rows like AB and A'B' of Fig. ll-19(a) is filled Avith
clay. Better yet, the bottom with a layer of tremie concrete
is filled

several feet thick, the space above it is later unwatered, and then stiff clay
is packed between the walls. This provides a dense fairly impervious
w all and a reasonably good
7
seal at the bottom.
478 COFFERDAMS [Chap. 11

The next procedure is that of excavating, bracing, and unwatering the

cofferdam as usual. If local leaks are discovered under the bottom,
grouting outside of the bottom of the structure may be beneficial in
sealing them.
Double-wall cofferdams may be used to depths of 70 to 75 ft., whereas
single walls should generally be limited to 40 or 50 ft. (or perhaps 60 ft.)
when the excavation goes to rock. With a single wall there is greater
danger of a blow-in if the piling hits steeply sloping rock or boulders so
that it cannot get a toe hold on the bottom. For deep double-wall coffer-
dams, the concrete seal on the bottom is very desirable because the mortar
is likely to penetrate between any remaining gravel or boulders and fill
A' Bracing Tie rod
i
•W/ll'

Steel sheer
Spiling
^Fi/I
clay ^kAvm^kmmwm^m w=W==W^
fill ACROSS SECTION OF SHALLOW COFFERDAM
Concrete

faJ-CROSS SECTION OF DEEP COFFERDAM

r—-—-— ///W//H
Ties

(C)~ PART OF EARTH LEFT INSIDE (d)

FIG. 11-19. Double-wall cofferdams.

crevices. The clay, especially when placed through water, is not likely
to do this.
The two rows of sheet piles shown in Fig. ll-19(a) should not be
assumed to share equally in resisting the outside pressure unless concrete
fill or firm bracing is used between them. When merely packed with clay,
the row A'B' tends to press against the fill, and the latter in turn trans-
mits the load to AB. In any case, the bracing must resist the total pres-
sure. Both piling and bracing may be designed or analyzed by the
methods previously explained for single walls.
Here is a case that emphasizes the safety element in design. A large
cofferdam was built to surround the excavation for a large bridge pier to
be founded upon sloping irregular rock. The bottom of the piling on the
deep side was approximately 70 ft. below M.H.W.; on the landward side,
40 ft. The riverward side and most of the two sides normal thereto were
made double- wall construction with a concrete seal 10 or 15 ft. deep and '
Art. 11-7] DOUBLE-WALL COFFERDAMS 479

clay fill above it, but the rear corners and landward side were made with
a single wall in the interest of economy. After excavation and unwater-
ing, the bottom portions of some of the single-wall piling near one rear
corner failed. This caused a blow-in that caused the loss of lives as well
as of money. The intended economy proved to be extremely unwise and
expensive.
Another type of double-wall cofferdam for use with shallow excavations
to rock is shown in Fig. 11-19(6). This was used for the abutments of the
Bayonne Bridge. The double sheet-pile enclosure is driven first. The
space between the walls may often be left filled with the original earth,
the tops of the piles being connected by tie rods between the two walls to
prevent spreading. It may then be desirable to leave an embankment
or berm of earth along the inside to support the bottom of the inner wall,
asshown in Sketch (c). This, however, increases the required size of the
cofferdam in plan. Another method is to drive the piling, excavate

Top decked with wood or concrete

I , Wooden struts and steel tie rods
Sand and gravel f/'/A -Wooden sheeting
Normal water levelj^ < -Wooden wales outside

Riprap to
/fc_
prevent ix^ffSand andgraveL
scour \t£*&'°' berm

Bottom of river

FIG. 11-20. Box-type cofferdam of wood formerly used by the U.S. Army Engineers for low heads.

between the walls, brace them temporarily or permanently, tie them

together by tie rods and welded connections, and then backfill the space.
A mixture of clay and gravel at the bottom will help to prevent leakage
along the rock. Generally, when this type is used, the hydrostatic head
is small so that the main portion of the backfill may be gravel and sand,
or even riprap, that has dependable internal friction.
The stability of the cofferdam of Fig. 11-19 is obtained through its

action as a sort of earth dam, with steel sides to provide vertical faces.
The forces are pictured in (d). Piles AB and A'B' should be relatively
far apart in order to provide adequate weight of fill, shear resistance in
the fill, and frictional resistance along the bottom. The cofferdam should
be stable as an earthen structure. The method of analysis can be similar
to that for the cellular cofferdams to be described in Art. 11-8. In gen-
eral, the maximum height of such a cofferdam should not exceed its width.
One thing to be guarded against is the kicking out of the toe of the inside
piling unless tie rods are provided or a strong berm is used.
480 COFFERDAMS [Chap. 11

The cofferdam shown in Fig. 11-21 was designed for repeated use. It
was made with double steel-plate walls, structural-steel diaphragms, bulk-
heads, and deck, so that it consisted of several tanks that could be
unwatered for floating or flooded for sinking. It was floated to position
and sunk to the bottom, where it was landed on wooden piles that had
been driven and cut off previously. Notice the use of sheet piling driven
outside the cofferdam and the excavation being carried on in the dry.

POURING CONCRETE REMOVAL

FIG. 11-21. Floating cofferdam for piers of Storstrom Bridge, Denmark. This shows the sequence of
operations when impermeable ground permitted unwatering the area inside the cofferdam. (See

S. M. Koefoed, Four Cofferdams for Forty-one Foundations, Engineering News-Record, July 6, 1939.)

The seal between the piling and cofferdam was made by fitting wooden
blocks into the troughs of the piling. When the cofferdam was unwa-
tered, the pressure caused these to bear against a bulge around the outside
bottom edge of the cofferdam. Divers then placed rope impregnated with
tallow between the blocks and this projection. This proved to be suffi-
cient. Jacks were built into the bottom of the cofferdam so that, when
ready to float, they could lift the cofferdam vertically far enough to break
away the bottom and beveled inside from the concrete.
Art. 11-8] CELLULAR COFFERDAMS 481

11-8. Cellular cofferdams. Another type of cofferdam for large

areas and high heads is the cellular construction pictured in Fig. 11-22.
This is used mostly in connection with the construction of dams, locks,
wharves, and other large water-front structures. There are various
shapes and arrangements of cells, some being illustrated in the drawing.
When filled with sand or gravel, they constitute a sort of earth dam bound
together by the piling.
The advantage of the circular cells is that, even though they are 40 to

60 ft. in diameter, the earth pressure on the inside causes only ring tension
in the sheet-pile walls. These tensile forces are transmitted through the

iW-DIAPHRAGM (W-CLOVERLEAF 6W- CLOSELY

faJ-CIRCULAR CELLS
TYPE TYPE SPACED
(INCOMPLETE CIRCLES
>s
COFFERDAM)

b/2

f/i J-FORCES AND PRESSURE DIAGRAM (7J-DIAGRAMS Cjj-SMALL BERM

OF PRESSURE
ON BASE

FIG. 1 1-22. Types of and miscellaneous material for cellular cofferdams.

interlocks, and they should ordinarily be limited to 8,000 lb. per Lin. in. of
interlock. The ultimate tensile strength of the piling is usually specified
as a minimum of 16,000 lb. per lin. in. The sheet piles should have
straight webs because of the heavy tension across them. No wales or
bracing are needed.
When anunbalanced outside pressure acts upon a filled cell, the former
over as indicated by the dashed lines in Fig. ll-25(a).
tries to tip the cell
The earth inside resists in both horizontal and vertical shearing action;
even the friction on the interlocks of the piling tries to resist the distortion
shown in Fig. 1 l-22(gr).
The problem of providing stability and shearing resistance in a large
cellular cofferdam is an important one. Notice that, in Fig. 11-22(6), the
482 COFFERDAMS [Chap. 11

cross-walls or diaphragms are under tension even though the earth pres-
sures 1 against them counterbalance each
other. This is because they tie
the scalloped sides together. Of course, the main shearing resistance is
provided by the fill in the cells; hence, it should be coarse sand, gravel, or
stones, if possible. Sand and gravel taken from a dry borrow pit are
usually reliable. Dredged fill must have the water "bled" off. Such fill
may be exceedingly variable in quality, and layers of fines which settle

FIG. 1 1-23. View inside the completed steel-pile cellular cofferdam erected by the Dravo Corporation,
Pittsburgh, for construction of the new navigational lock at Morgantown, W.Va., on the Monongahela
River. Water has been pumped out, rock is being removed, and the first concrete is being placed
for the bases of the land and river walls. The project is under the supervision of the U.S. Army Corps
of Engineers.

out during off shifts may cause planes of weakness having poor shearing
resistance.
Watertightness is also important. The sheet piling should provide
fairly good tight joints for the walls. The leakage at the bottom can be
resisted mainly by the sealing action of the fill on the rock or other bottom.
In Fig. 11-22 (a) it is apparent that leakage is probable between adjacent
cells unless they are connected, and there is little chance to seal the bottom

at the junction. The portions a and b in the other illustrations are pri-
1
Be careful to avoid unequal filling of adjacent cells and consequent distortion of the
diaphragms.
Art. 11-8] CELLULAR COFFERDAMS 483

marily to prevent leakage rather than to provide structural resistance to

overturning. From the standpoint of tightness, clay fill would be desira-
ble; from that of strength, a coarse granular fill is better. Furthermore,
the latter should be drained by providing weep holes near the bottoms of
the inside piling. This adds greatly to the resistance of a sand or gravel
fill. Any berm used inside should be drained also. Of course, the
greatest tensile stress in the interlocks would come when the fill in the
cell is saturated but the enclosure is dry. This eventuality should not
be able to wreck the structure.

w
c=^?

4x4*4lf
^
TEE
k
CROSS 120 DEG. WYE

if
l"R

^
7^
L
U- £ pile

30 DEG. WYE 10 DEG. BEND

90 DEG. WYE

FIG. 1 1-24. Details of fabricated connections for steel sheet-pile cofferdams used for Kentucky Dam
(See A. F. Hedman, Cofferdam Design for Kentucky Dam, Engineering News-Record, Jan. 1, 1942.)

The stability of such a cofferdam should be entirely reliable and able

to resist floods. The top should be covered with riprap or a 4- to 6-in.
concrete slab to prevent scouring out of the fill in such an emergency.
Some details of junction pieces for piling 1 are shown in Fig. 11-24.
Similar pieces may
be fabricated by welding. The dimensions of these
and depend upon what type of
of the cells will sheet pile is adopted and
how many pieces are to be used to construct a cell.
According to Hedman, the following data regarding coefficients of fric-
tion were found as the result of tests made for Kentucky Dam:

Gravel on rock 0.5

Gravel on steel 0.4
Steel on steel at interlocks 0.3

1
A. F. Hedman, Cofferdam Design for Kentucky Dam, Engineering News-Record,
Jan. 1, 1942.
484 COFFERDAMS [Chap. 11

The type of cofferdam shown in Fig. ll-22(d) has the advantage of

permitting the filling of cells more or less independently, whereas this can-
not be done in the case of the diaphragm type in (6). Furthermore, if a
diaphragm in (b) should give way, the entire cofferdam might fail, whereas
rupture of a closure piece in (d) is very unlikely and probably would not
harm the main cells. The closure pieces a in (d) should not be too large
or have too big a radius of curvature because the pull on the wall of the
main cell may tend to "kink" the latter.

A cellular cofferdam may be founded upon soil if the sheet piles can be

driven down to reach material that is sufficiently impermeable yet able to

o' E E' D D EE

w \ \

H
1
y \

\ c U_.
TT~ < — c *
B \
y\^f///f^^f^
Sliding slices

U)

FIG. 1 1-25. Analysis of a cellular cofferdam.

resist the applied loads. The abutting power of the earth below the plane
of excavation inside the cofferdam will resist horizontal shear, thus aiding
friction along the bottom, and it will serve to counteract the bursting
tendency of the embedded portion of the structure. However, the effec-
tive height for the computation of overturning moments will be consider-
ably greater than the distance from the water outside down to the level
of excavation inside the enclosure, and the pressures causing interlock
resistances should be only those active ones producing tension in the rings.
Cummings recommends 1
a method of analysis of cellular cofferdams
1
E. M. Cummings, Cellular Cofferdams and Docks, Paper 1366, American Society
of Civil Engineers, Journal of the Waterways and Harbors Division, September, 1957.
Art. 11-8] CELLULAR COFFERDAMS 485

which, for the action of the fill alone, may be outlined as follows, referring
to Fig. 11-25:
1. If the force H is applied near the bottom of the cofferdam in Sketch
(a), the latter will slide F, where F = if H>
tan 6 and the angle of W
internal friction of the assumed to equal that of the latter sliding on
fill is

the rock. As H is raised above some height y, the cell will start to distort
as shown, instead of sliding. This must cause a slippage on horizontal
planes in the fill.

2. Assume A B, Sketch (b), is at the angle of repose so that the fill above

it isabout to slide down. When the cell is distorted, A moves to A' and
the fill above A'B tends to slide down over this sloping surface. There-
fore, the soil below this in CA'B is acting in passive resistance under the
surcharge effect of A'D'E'B. As movement occurs, the mass CA'B serves
as a buttress, and sliding may occur on horizontal planes in it, as indi-
cated in
(c). The confined material stiffens the cofferdam, and the latter,
wrapped around CA'B, causes the lateral forces to be applied to this
buttress.
3. The total frictional resistance R, referring to Sketch (d), is

R = W tan 6 = wbh tan 9

for a 1 ft. vertical slice, where w is the actual unit weight of the fill in air
when dry or the buoyant weight if the cell is full of water. This assumes
that any additional shearing resistance of the bottoms of the piling catch-
ing into rock is too unreliable to be trusted. Let b = an averaged width
of the cell to provide the same cross-sectional area as the cell.

If c = b tan 8 and h = a -f- c as in (d), R = w{ac + c

2
).

Referring to Sketch (d), make BF = JG = c, FG = wa, and JK = 2wc.

Then let the volume of BFGJ = Ri and the volume of JGK = R 2 .

Hence,

Ri = wac and 2? 2 = 2wc X „ = wc 2

Thus the volume represented by BFGK = Ri +R = 2 R.

:.R = w(ac + c
2
)

The location of R is

_
= R lC /2 + 7? 2 c/3
y
R
In other words, the moment of resistance of the imaginary buttress about
the base CB is

MB = Ry = w(
\2
u
^- + '-)
37
(11-8)
486 COFFERDAMS [Chap. 11

The safety factor against sliding is S.F. = R/H; that against over-
turning for the resistance of the fill only is S.F.= M B /Hy.
The resistance of the interlocks to the shearing action shown in Fig.
1 1-22 (g) is also important. It may be assumed to equal the tension caused
by the total active pressure of the contained soil acting on a vertical 1-ft.

slice times the coefficient of friction of the interlock — perhaps 0.3.

Therefore, we may assume that
2

Fi = (C.4
ICA w j \ r(0.3) (11-9)
)
If the cell contains water, the hydrostatic pressure can be added, and the
coefficient Ca and the weight w should be those for the saturated condi-
tion below water level. This friction Fi is assumed to act equally on all
interlocks, tending to hold the piling together like a diaphragm. Hence,
the resisting moment Mi for one side may be assumed to be Mi — Fjb/L,
where L is as shown in Fig. ll-22(/). Therefore, the total safety factor
against overturning for a circular cell having two sides acting in interlock
friction is

S.F. -
M ' + 2M <
(11-10)
Hy
When cellular cofferdams are used for docks or bulkheads, it is usually
desirable to build a concrete deck over at least a large part of the outboard
half of the tops of the cells, cantilevering the edge and equipping it with
bumpers to protect shipping. A few points to be considered in such con-
struction are the following:
Provide expansion joints in the deck, preferably at the center of each
1.

that the concrete can span across the recesses in the outer face but
cell so

not be cracked by relative movement of the cells with respect to the

concrete.
2. Support the concrete on piles within the cells, the piles being installed
in a manner to prevent soil bulging from bursting the cells.
3. Keep the concrete above the cells so that live loading behind the cells
will not cause the deformation of piling to crack the concrete.
Fill behind the cells before pouring the concrete deck.
4.

Allow for bulging of the cells during filling and for considerable
5.

movement of the cells when the fill is placed behind them. These motions
are sometimes considerable because, when driven, the interlocks may be
jammed together, but when subjected to tension, each one may "open"
M to }& m - when highly stressed.
6. Consider cathodic protection of the steel if in sea water.
7. The cells will probably be full of water up to a point close to high-
tide elevation.
Art. 11-8] CELLULAR COFFERDAMS 487

8. With earth fill on the inboard side, one should allow for an unbal-

anced head of water at least equal to the tide range (probably more)
unless adequate drainage is provided, which means that the water will
flow into the drains during a rising tide and out during the ebb.
9. Live-load surcharge behind the cells can be exceedingly serious, so

that an adequate safety factor (perhaps 1.5) should be used for design.
Figure 11-25^4 shows the safety factors found in the analysis of a proposed
design for a diaphragm type of wharf.

3.0

2.5

rchoSSZ^
nojh
A_
2.0

,1.5 -
j^"^
,-horgL-
j/m^>^' D_

1.0 With su

0.5

28 29 30 31 32 33 34 35
Angle of internal friction, deg.
Curves A and B based on stability
Curves C and D based on vertical shear

FIG. 1 ]-25A. Study of effect of angle of internal friction of soil and 1.5 k.s.f. surcharge on the safety
factor of a large diaphragm-type wharf.

10. Hydraulic filling of the cells and behind them may be convenient,
but the material is likely to be variable and to contain many fines, so that
one should be conservative in the selected angle 4> used for analysis.
11. One might ask if the active soil pressure and the effect of surcharge
will act at an angle $ with the horizontal instead of horizontally as assumed
in Fig. 11-22(7?,). Using a coefficient of friction of 0.4 for the sliding of
soil on piling, one might assume an angle of about 22° for this slope. The
horizontal component of the earth pressure would then be less than the
total, but not much so. The vertical component should not be used as a
stabilizing force because this will be delivered directly to the rock or other
bottom by the piling instead of being applied to the contents of the cell.
488 COFFERDAMS [Chap. 11

Example. Analyze the stability and shearing resistance of a cellular

cofferdam like that of Figs. ll-22(d) and (/), and determine the corre-
sponding safety factors. Assume d is 50 ft.; L, 52 ft.; the width of the
closure pieces e, 24 ft.; the height h, 40 ft.; the freeboard h', 5 ft. The
cofferdam is to rest on rock. Assume h x is water 10 ft. deep at 62.5 p.c.f.
h-2 is sand 30 ft. deep at 100 p.c.f. dry and 65 p.c.f. saturated. The piling
is to be driven first, the cells filled with dredged sand, then the excavation

inside the enclosure is to be made. Assume that the angle of internal

friction (/> of the sand is 30°; the coefficient for active pressure Ca, 0.33.
Let the coefficient of sliding friction of gravel on rock be 0.5; of steel on
steel at the interlocks, 0.3. Neglect friction between the soil and the
sheet piling. A minimum safety factor of 1.33 is desired. The interior
of the cells is to be drained into the enclosure.
It is satisfactory to substitute an imaginary rectangular section having
the same area as the circle and junction pieces for one unit. In this case,
assume
Lb = 7r X 25 2 + one whole junction piece

or
.

o = — 1,960
^+ 120
= 40 ft.

The averaged dry weight per foot of length of a typical unit, neglecting
the weight of the piling, is

W = 45 X 0.1 X 40 = 180 kips if entirely dry

The total lateral pressure H is

Water = 62.5 X 40 2 /2 = 50,000 lb. = 50 kips

Sand = (0.33) = 9,650 lb. = 9.6 kips Say, H= 60 kips

The center of gravity of H from the bottom of the cell is

=
50 X 40 X 0.333 + 10
-^ X 30 X 0.333
- = 10 _
12.8
.,
y ft.

1. Stability. Referring to Fig. ll-25(d), with </> = 30° and b = 40 ft.,

c = 40 X 0.576 = 23 ft,; a = 45 - 23 = 22 ft. From Eq. (11-8),

(22 V
2
9^2
+ ~
9Q3\
J
= 988 ft.-kips per ft. of length

The active soil pressure for a full cell is

H' = C A w~ = 0.33 X 0.1 X 45V2 = 33.4 kips

The ring tension T = H'r = 33.4 X 25 = 835 kips

Art. 11-8] CELLULAR COFFERDAMS 489

The interlock friction F = T X

t 0.3 = 835 X 0.3 = 250 kips
The resisting moment of interlock friction per side is

Mi = ^=
Li
250 X 40
52 = 192 ft. -kips per ft., averaged

Therefore, the safety factor from Eq. (11-10) is

&
M B + 2Mj 988 + 2 X 192 _
Hy 60 X 12.8 "

Notice that, if the cell is flooded to the level of the outside water, the
buoyant weight of the soil reduces Mb but the water pressure increases Mi.
In this case, the average unit weight is

40X0.065 + 5X0.1- = n Ar>n

0.069 p. c.i.
,
r=
45

— * 4- r|_j = 681 ft.-kips

#' = 0.33 X 0.069 X 45 /2 + 0.0625 X 40 /2 = 23 + 50 =

2 2
73 kips
Mi = (73 X 25 X 0.3) 4 % 2 = 421 ft.-kips per ft, for one side
rp,
Therefore,
,
Q v = 681 + 2 X 421
b.K
OU X l2.o
— = 1.98—OQ

2. Sliding. The frictional resistance to sliding is Wf. Therefore,

a „
S.F.
TT/
= -^ = 180 —X 0.5 = t
1.5
_ ..
if dry
,

or a „
S.F. = (0.069 X 40
T77T
60
X 45)0.5
— = ,
AOE
1.035
.,
if
„
flooded
, ,

Obviously, flooding is to be avoided but leakage is likely to let in some

water so that effective drainage is needed. If necessary, a gravel berm
like that of Fig. ll-22(j) might be used inside the enclosure. The ends of
the sheet piles, if driven to clinch the rock, would also help resist sliding.
3. Shearing along vertical center on a
line. The greatest shearing force
on the one at the center. Figure 1 \-22(i)
vertical plane in the cell will be
pictures the pressure diagram which might apply if the cell were a rigid
solid. Assuming the dry condition, the pressure would be

W ,
Mc

- ™-
v= lx40 1
+ (60 X
12.8)25
= kS L
or P ± 1(40 3 /12)
4 5 ' ± d b " "

The total force represented by one of the triangles in Fig. ll-22(r) is the
shear to be withstood along the vertical plane through E. Hence,
490 COFFERDAMS [Chap. 11

V = 3.6 X 2
?2 —
For the soil alone in a dry condition, the
36 kips.
frictional resistance to shearing on this plane is F s = H tan d = 33.4 X 1

0.576 = 19.2 kips. Since the friction on the interlocks helps to oppose
this shearing action, the resistance from this per foot of length is 2F /L = 8

2 X 250 52 = 9.6 kips. Therefore, the total resistance is 19.2 + 9.6 =

28.8 kips per ft. Hence, the safety factor for this case is

S.F. = 28.8/36 = 0.8, which is unsafe

Examine this action further. It seems that the pressure H acting as in

Fig. ll-22(/i) will bear against the cell on the bottom and that a friction
must resist it. Since a frictional resistance equals fN, the high compres-
sion half of the bottom would seem to offer more resistance than the left
side, FE, in the sketch. However, a force will be resisted by the first thing
which can do so; hence the side FE may be fairly effective. It seems to
be conservative if one assumes that this friction is evenly distributed
along the bottom. Then one-half of H can be considered to be a pressure
acting on the vertical plane through E and increasing the shearing resist-
ance an amount equal to

^ tan <$> = 60
2 X 0.576 = 17.3 kips

Adding this to the other resistances already computed would give

QT?
b.l H = 28.8 +—17.3
^ - = 46.1
= ,„„,.,.
which is
1.28,
.
f
not quite
-

enough
,
. -FT7T-
36 36

In the case of the flooded cell, H' = 23 kips for the soil alone since the
buoyant weight of the soil must be used. The interlock resistance for
both water and earth pressures will be Fi = /(pressure)r = 0.3(73 X
25) = 54.8 kips per side. Therefore, the total shearing resistance on the
central plane may be (H' + H/2) tan <t> + 2F /L =
( (23 + 6°
/
/
2 )0.576 +
2 X 548 52 = 30.5 21.1 + = 51.6 kips. Therefore,

n „
S.F. =
resistance
- -y- - =
w
51.6
= . n
1.43
.

The preceding stability and sliding computations seem to be reasonably

reliable. The vertical shearing resistance might well be investigated, but
its reliability is questionable. The cell is not a solid and cannot tip about
one edge. If the radius of a circular cell is increased while the height
and H
remain unchanged, the stability and resistance to sliding will be
increased also. V will be reduced somewhat, but the vertical shearing-
resistance of the soil will be unchanged whereas the increase in interlock
friction will vary as the radius, but the average effectiveness of this will
be offset somewhat by the larger spacing. In the case of a diaphragm
Art. 11-9] PRACTICAL FEATURES 491

type of cofferdam with b increased but L unchanged, the vertical shearing

resistance will not be greatly increased.
If the diaphragm type of construction shown in Fig. 11-22(6) is used,
only one diaphragm is The tension through it will be
available per cell.

the parallel components of the two ring tensions coming at one end of the
diaphragm. For a 120° junction, these add up to the same as the ring
tension.
11-9. Miscellaneous practical features for cofferdams. A few of
the many practical matters concerning the planning and building of coffer-
dams not previously described are given for the guidance of the reader.
They will not be discussed in great detail. Each job has to be made to
suit particular conditions.
1. Floods. These may raise havoc with cofferdams. Nevertheless,
economy requires that there be some practicable limit to the height and
freeboard of the structure. Floods that are likely to occur frequently
during the months when the cofferdam will be in the building or use stage
should be guarded against, and with some reserve to spare. Those floods
that occur regularly but at an inapplicable season may be discounted if
one can be sure that the useful life of the cofferdam will not extend close
to that period. Major floods that come once in several years or decades
generally may be discounted because one cannot afford to provide against
such possible but improbable events. However, the establishment of the
elevation of the top is a very serious matter.
2. Currents. Because of its large area, the dynamic effect of flow-
ing water may be a real force to consider in the design of a cofferdam.
Records of stream and tidal currents should be searched for, and estimates
should be made of the acceleration of velocity to be caused by the obstruc-
tion of the waterway when the cofferdam is in place. Scour is another
effect that may be harmful. The pressure may be estimated by the
formula
Ft = M(v - v )
(11-11)

where F = force in pounds, = unit of time (taken as 1 sec), M =

t

W/g — mass of water impinging upon the structure per second, v = origi-
nal velocity in feet per second, and Vo = final velocity in feet per second
(usually taken as zero). The force is to be reduced to allow for any
streamlining of the cofferdam that may be applicable.
3. Waves. Storms may cause waves that are hazardous for some coffer-
dams because of their large area and light weight. The trouble may
come from flooding as well as from shock. This subject is discussed
more fully in Chap. 13.
4- Unwatering. When a cofferdam is flooded, the interlocks of the
piling may be loose because of the lack of pressure to close them. It may
492 COFFERDAMS [Chap. 11

then be difficult to get centrifugal pumps of sufficient capacity to over-

come the effects of so many leaky joints. After a head of several feet is
acting upon the piling, the pressure is likely to close the joints and reduce
the leakage in spite of the fact that the head increases as the unwatering
proceeds. The movements of sheet piles produced by waves during the
first part of the pumping may accentuate leakage and greatly interfere
with securing of the initial lowering of the water. Dirt, sand, cinders, or
similar materials piled around the cofferdam may help to "silt up" the
interlocks. Another serious danger is the blowing up of the bottom dur-
ing or after unwatering or during the driving of piles inside the cofferdam.
Personal experience with the particular soils, or with similar ones, and the
experiences of others are exceedingly helpful in estimating the head of
water that can be withstood safely and without harmful seepage.
5. Length of sheet
piles. The ordinary maximum length of sheet piles is
80 ft. Some work has been done using welded extensions on standard
piling, but this should be avoided in general.
6. Sheet piles are very reliable material.
Stresses in sheet piles. Ordi-
narily the yield point from 30,000 to 35,000 p.s.i. the ultimate strength
is ;

60,000 to 70,000 p.s.i. In such temporary structures, it might seem to be

justifiable to use a high working unit stress. This, however, is not
advisable generally because of the uncertainty about unit pressures,
bracing reactions, and the stress conditions in the piling. The cost per
pound of recoverable and reusable piling is so small compared to the cost
of a failure that it is unwise to skimp on this particular item.
7. Safety factor. Cofferdams are temporary structures. They do not
need the reserve strength that is customarily required for permanent ones,
yet they must not fail. When a cofferdam is studied carefully and when
all reasonably probable loads are wisely estimated and provided for, the

safety factor may be 1.25 to 1.5. When there are unpredictable elements
to be withstood, the reserve may well be larger.
8. Reuse of materials. A
cofferdam should be planned with proper
consideration for its So much material is generally involved
dismantling.
that it is worth while to arrange the details of the structure so that the
bracing can be taken apart and the piling pulled with a minimum of labor
and waste.
11-10. Special methods of protecting excavations. The freezing
process has been used successfully in some cases to form a sort of cofferdam
around an area to be excavated. This is usually done by setting up a
refrigeration plant, driving large pipes around the site, as shown in Fig.
ll-26(a), inserting smaller pipes inside the large ones, connecting all into a
closed system, and down through the small pipes
circulating cold brine
and up through the This causes the ground to freeze around
large ones.
the pipes and, if planned properly, the frozen regions around the adjacent
Art. 11-10] SPECIAL METHODS 493

pipes will eventually constitute a solid wall of frozen earth several feet
thick. In certain cases, this may be strong enough to permit open
excavation within the enclosure.
A few points to bear in mind regarding this freezing process are:
1. The area of the excavation should not be too large.

2. The layout of pipes should be such as to constitute or approximate a

circle in order to utilize the arch action of the frozen earth when the area
is to be large. Small areas can be surrounded by square or rectangular

Inner pipe

Firm, impervious bottom

(a)-FROIEH WALL AROUND EXCAVATION

Line of
excavation
F/ow

"
Water-bearing
stratum
(W-FROZEN WALL ON UPHILL SIDE OF EXCAVATION

FIG. 1 1-26. Illustration of use of frozen earth as protection for excavation.

layouts, but the frozen earth is not dependable when it has to act as a

beam.
3. Water-bearing sand, gravel, and silt are suitable materials for

freezing. Clay may be frozen, but other methods may be more desirable
for such soils.
Conditions must be such that serious leakage in or blow-up of the
4.

bottom will not occur. The frozen region should therefore extend down
to rock, hardpan, or dense clay unless the incoming water can be handled
by pumps.
5. The freezing process is best suited to jobs of short duration, partly
because of the cost of operating the plant and partly because the depend-
494 COFFERDAMS [Chap. 11

ability of a frozen earth structure is not very good. Equipment or power

may fail temporarily; floods or hot weather may cause conditions that
exceed the capacity of the plant.
6. The process is particularly useful in protecting excavations in or at
the foot of slopes where slides may be a menace. In this case, a portion
of the slope is frozen to constitute a temporary dam to support the soil
behind it. Such a situation is shown in Fig. 11-26(6).
Cement grout may be used to protect an area to be excavated if the soil
is porous and the excavation is small. The difficulty is the inability to
control where the grout goes. It may not form a complete and tight
enclosure or cutoff wall. It can be used sometimes to assist in such special
jobs as underpinning. It is primarily one method of consolidating the
ground.
The uncertainty of control is illustrated by a case where grout was
pumped down to a point some 20 ft. below ground. The operator realized
that an unusual amount of high-pressure grout was being used. A helper
soon discovered that it was coming to the surface as a boil many feet
from where it was introduced. It apparently followed a porous seam or
stratum as the line of least resistance.
In planning grouting, one should remember that saturated sand cannot
be grouted well without expelling the water already in the voids. Forcing
grout into the sand rapidly may serve only to heave it and to form
local seams or pockets of grout. Fine sands can seldom be grouted
satisfactorily.
Chemical consolidation is another method that has limited usefulness.
In general, with the Joosten method, one chemical is forced into the soil,

then another one is added to cause the formation of solids or jells that
stiffen the soil. This may be of value.
The construction of the foundations of the floodlight towers at the
Kezar Stadium in San Francisco's Golden Gate Park is an example of such
consolidation methods applied to a local area. ' A sodium silicate solution
was pumped into the ground under a pressure of 200 p.s.i. to solidify the
soil — a running sand — fill in the vicinity of the points where the bases of
the pier columns were to be. This was 16 below the surface. Work-
ft.

ing through 30-in. casings, the soil was excavated by hand to form belled
bottoms somewhat like that of Fig. 9-7. The solidified sand developed a
compressive strength of approximately 75 p.s.i.
11-11. Wellpoints. Much trouble in the making of excavations is
caused by the presence of water, especially when the soil is fine sand or
silt. When the water is allowed to flow into the excavation or to sumps,
it is likely to cause movement of the fine particles, slumping of the sides,

undermining of adjacent areas, or even softening and blowing up of the

1
Engineering News-Record, Nov. 10, 1949, p. 48.
Art. 11-11] WELLPOINTS 495

bottom. It may also cause a large increase in the amount of excavation

because of the flat angle of repose of the water-bearing slopes. When the
bottom is clay, water brought in by higher granular strata may cause
serious muddying of the bottom.

FIG. 11-27. Detail picture showing the sand filter around wellpoint riser pipes. This is part of the
system for unwatering the site of the Neversink Dam Project in New York. The sand is placed around
the wellpoint screen and riser pipe to aid in the drainage of fine soils and to prevent clogging the

screens. This is an important feature in such soils as silt. Built by S. A. Healy Co., White Plains, N.Y.
(Courtesy of the Moretrench Corp., New York.)

In many cases, the use of wellpoints may be a means of lowering the

watertable so that excavation can be made in moist soil that will stand on
fairly steep slopes temporarily. A wellpoint system consists of a series of
driven or jetted pipes, several feet apart, as shown in Fig. 11-27, with
screened intakes. These pipes are attached to headers that are connected
to centrifugal or other pumps which remove the water from the pipe line.
The water runs through screens that are fine enough to keep out the par-
ticles of soil. The effect is somewhat as pictured in Fig. 11-28 (a). The
496 COFFERDAMS [Chap. 11

water flows in the direction of least resistance, to the screened intakes in

this case. The result is the lowering of the watertable to some such sur-
face as ABODE, the slopes depending upon the permeability of the
ground and the amount of water to be removed.

,-^Line of excavation
,Oriainal
watertable . p /^ Ground'line j

> Dense, impervious stratum

(a) ACTION OF WELLPOINTS AROUND AN EXCAVATION

^Ground line Original water table

Bottom of
excavation
\
Water table

\b) WELLPOINTS INSTALLED IN STAGES

FOR DEEP EXCAVATION

(c) TRENCH id) DRAINING OF A WATER-BEARING

EXCAVATION STRATUM

• Ground line
w" ^//,,wyAvv/\i" y^',7^7
wronr^ Fine sand
\r- L ine of excavation -y or silt
/
'Sand
Clav z-~-_
p~T tttt Gravef^^i
Hydrostatic
pressure

{<?) DANGER OF BLOW-UP OF BOTTOM

FIG. 11-28. Use of wellpoints in aiding open excavations.

As an example, refer again to Fig. ll-28(a). The excavation to be

made is indicated by the dashed lines FGHJ, and the soil is medium sand.
The watertable is ordinarily at the level KL. If the pipes a and the well-
points are driven to B and D and if pumping is carried on continuously,
the water from the surrounding region will flow to and into the wellpoints
until a balance between supply and pumping is secured. Then, if water
comes from the sides only, the portion BCD of the watertable will become
Art. 11-11] WELLPOINTS 497

and remain rather flat because there is no renewal of its supply. If water
also flows upward from below, BCD will be curved more sharply. The
system should be planned so that the watertable is below the bottom GH
of the excavation. After the structure is completed, the pumping is
stopped and the wellpoint system is removed. The water will then return
to its natural level.

FIG. 1 1-29. A Moretrench wellpoint system for unwatering the excavation for a pumping station for
the city of Muskegon Heights, Muskegon, Mich. This shows the header pipes at different levels.

Notice how close the excavation is to the lake. (Courfesy of the Moretrench Corp.)

Whenplanning a wellpoint system, one should study the permeability

and the quantity of water to be handled. A pumping test may
of the soils
be very helpful in judging whether the ground will drain properly and in
estimating the capacity of the pumps required. The advice of those who
make and use wellpoint equipment should often be secured before one
starts a major installation of this character.
Some points to bear in mind regarding the installation and action of
wellpoints are the following, referring to Fig. 11-28:
1. Suction pumps cannot lift water much above 20 ft. Therefore, a
practical limit should be set as 15 or 16 ft.
498 COFFERDAMS [Chap. 11

2. When greater lifts are necessary, deep-well pumps may be used.

These are lowered into the holes and can push the water to substantial
heights. However, this installation is expensive.
3. As shown in Sketch (b), it is sometimes practicable to make a large

excavation in stages. A row of wellpoints is placed around the site at

position M the watertable is lowered, and the first portion of the excava-
,

tion is made to N. Another row of wellpoints is then added at N, and the

work is continued. A third row may then be placed, etc. At the bottom,
a final row like that at may be necessary to keep the bottom sufficiently
dry. By the use of high-pressure water jets, as illustrated in Fig. 1 1-29^4,
1

it has been possible to avoid the installation of wellpoints in 20-ft. stages,

attaining "lifts" of over 40 ft. Here the vacuum created by the jet "sucks"
the water in and the velocity head in the
Return low-pressure
line
return pipe carries the ground water along
with the jet water to the discharge pipe.
4. The ground around shallow trenches

and other narrow excavations may be

High -pressure
upply line drained sufficiently by the use of one row
of wellpoints P, as shown in (c).

5. Excavations on hillsides may be diffi-

Eductor " housing jet, cult to make because of one or more porous
as made by Wellpoint
Dewatering Corp.
water-bearing strata, as shown in (d), Fig.
11-28. One row of wellpoints at Q may be
Flow from wellpoint able to intercept the water. If the excava-
tion is near the base of the slope, wellpoints
FIG. 11-29A. Wellpoint installation
using water jet and vacuum. R may also be necessary because of back
pressure on that side.
6. Sketch (e) shows a condition where wellpoints S and T may drain

the sand layer satisfactorily, However, hydrostatic pressure in the gravel

stratum below the clay may heave up the bottom. Deep borings should
be made to reveal such conditions, and especially to see if there is likely to
be artesianal action due to water under high pressure trapped in a porous
layer under an impervious one. Deep jetted holes, sometimes called deep
wells, may be used to tap such a layer and to relieve the pressure, or they
may prove to be like leaks in a water main, that cannot be stopped with-
out great difficulty.
7. Very fine sands and silts tend to hold water by capillary action, and

they offer considerable resistance to percolation. Wellpoints in them

may be ineffective unless they are very close together. The vacuum
method 2 applied through a wellpoint system may then be of value.
1
See Water Removes Water from Deep Cut-and-cover Tunnel, Engineering News-
Record, Oct. 1, 1959.
2
K. Terzaghi and R. B. Peck, "Soil Mechanics in Engineering Practice," John
Wiley & Sons, Inc., New York, 1948.
Art. 11-11] WELLPOINTS 499

8. Clays cannot be drained by wellpoints because of their impermea-

bility.

9. made as in (6), there may be danger of insta-

In deep excavations,
bility of the slopes because the unwatered soil is only a relatively thin
layer along the surface with water under pressure behind it.
10. If too great a quantity of water has to be pumped, the flow through
the may remove or disturb too much of the soil itself.
soil

11. Once having started the pumping of wellpoints, it should be kept

up continuously. Hence, provisions should be made for breakdown of
equipment or power.
An interesting example of the use of wellpoints occurred recently. An
institutional building was to be erected on a site that the new owners
inspected "while driving by in their automobile." They saw that a
swale extended diagonally across it and that skunk cabbage and similar
plants grew there, but they made no attempt to investigate conditions.
After the contract was let, the contractor started digging for footings.
He found that, just under the mucky top, the soil was a very fine satu-
rated sand that flowed almost like thick gravy. He managed to get
down a few feet to hardpan and to build two of the footings. As the
runny sand was much deeper at the next position, he thought that he
would dig a sump behind the structure, install pumps in it, and drain the
ground that way. He started to do this and connected the pumps to a
suction pipe. As he lowered the water level a little, the sand ran in.
He was soon pumping sand and water and ruining the pumps. Further-
more, the sump was soon as big as a house. He finally gave up and hired
someone else to install wellpoints. This was soon done, the area was kept
free of excess water even though the 6-in. header was running full of clear
water. It was then revealed that the water-bearing sand was a few feet
deep at one end of the building and 22 ft. deep at the other. The footings
were placed on hardpan inside deep pits, piers were extended to the first-
floor level, and the floor was made a self-supporting reinforced-concrete
structure. Of course, the owners "paid through the nose" for all this
extra work.
The may lead to unwise conclusions if the bearing
use of wellpoints
value of a judged by its action during excavation after the soil has
soil is
been ''dried out." Allowable pressures and depths of foundations should
be determined by tests and studies of the soil in its normal condition.
The firming action of capillary moisture retained in the earth after unwa-
tering may cause a weak, fine, runny sand to appear to be strong whereas
it is weak when saturated. One example of disastrous results occurred
recently. No borings had been made at the site of a group of school
—
buildings almost an unpardonable neglect. When excavation was
started, underground water and springs were encountered. Wellpoints
were installed at extra cost to the owner. When the excess water was
500 COFFERDAMS [Chap. 11

removed and kept down, the excavation was continued. The

fine sand
seemed to be firm, and the footings were built as planned. A
few months
after completion of the buildings, unsightly cracks appeared, and they are
obviously the result of uneven and considerable settlement.
There can be various conditions in which the use of wellpoints will not
be sufficiently effective. These occur when the soil is so fine and saturated
that it will flow with the water, causing the pumping to remove a large
amount of fine material so that undermining of adjacent areas may take
place. Another such situation occurs when the ground is too impervious

\
Seepage

*fcj^ Water

ia)

FIG. 1 1-30. Drainage wel

to drain properly. This is likely to be the case when the soil is a glacial
hardpan, or other material, with enough clay or silt to prevent easy
till,

seepage. In such situations as these, other means for removing the water
must be resorted to.
One possible solution for such aproblem is the use of wells such as
pictured in Fig. 11-30. This method was employed in the construction
of a deep pit inside an existing factory. The top
10 ft. of soil was silty
sand. Then came a and boulders, under which was
layer of sand, gravel,
a dense hardpan containing some stones. The watertable was only about
6 ft. below the floor line. A row of columns and some machinerv were
Art. 11-11] WELLPOINTS 501

alongside the site of the pit. After excavation was carried down a way
below the watertable, it was found that water seeped in at the rate of 80
to 100 g.p.m.
To intercept the water, a series of wells was constructed a few feet
outside the sheathing. One of these is shown in Fig. 11-30. The general
method of procedure is the following:
1. As indicated in Sketch (a), a large steel pipe (perhaps 24 in. or more)

is driven and excavated, penetrating as far down as the bottom of the

contemplated excavation or as far as conditions make it desirable or pos-

sible. In the case cited previously, the wells penetrated only into the top
of the hardpan stratum.
2. A long cylindrical cage of heavy }4r or 3^-in. wire mesh is dropped in
the center of the pipe and the annular space between the pipe and the
cage is rilled with uniform crushed stone or screened gravel of proper size
— say, \'2 or %
depending on the size of the mesh.
in.,

3. The pipe
withdrawn so that water from slowly draining
is soils will
have a large surface area through which to enter the well.
4. A suction pipe, wellpoint, or deep-well pump is lowered into the cage
as indicated in Sketch (6). By pumping, the water is removed from the
wells as fast as it enters, thus intercepting it before it can reach the
excavation.
In some excavation problems it is practically impossible to keep all the
water from entering the excavation. There may then be a question as to
how to keep the water from interfering with the concrete work, especially
when cannot be used. One method which has been
steel sheet piling
resorted to is In this case, the pit is shown in
indicated in Fig. 11-31.
plan in Sketch (a). "Soldier beams" and wooden sheathing were used
1

to hold back the earth because the bottom 15 ft. had to be excavated
through hardpan. Wells were used around the pit, but boulders in the
hardpan prevented the contractor from sinking them much below the top
of the hardpan itself. Some 30 g.p.m. of water seeped through the hard-
pan, which, when disturbed, tended to soften. The contractor con-
structed a steel box somewhat as shown in Sketch (6), to serve as a sump.
This was anchored to a concrete pad as indicated. Then pipes were
welded to the box, and valves were installed on their inner ends as pic-
tured. These pipes were connected to a sort of French drain at the bot-
tom edges of the excavation, as pictured in the sketch, so that water
w ould drain down behind the sheathing, into the collectors, and thence
r

into the sump. Matched boards driven around the inside, plus a little
mortar where necessary, served as a dam to force most of the water into
the pipes. The sheathing over the inside of the collectors was sealed so
that wet concrete could not enter and clog the system. A local pit was
iSee Art. 11-12.
502 COFFERDAMS [Chap. 11

dug for the bottom of the sump in order to have the latter low enough for
the water to drain into it.
When ready to pour the concrete of the bottom of the pit, the con-
tractor pumped out the water which had by-passed the drains; then he
deposited the concrete. Thereafter, the water had to enter the pipes and
the sump, from which it was pumped. Leakage through the wooden

-Drain of bottom

1 bell and spigot

.-V.C.
pipe where needed

ryJL- Gravel or crushed

MM stone

Wire mesh basket

over ends of pipes

SECTION A-A THROUGH

OUTER DRAIN
FIG. 11-31. Details for draining deep pit.

sheathing was checked by cleats and calking so that it was not trouble-

some and so that mortar from the concrete of the walls could not seep
out behind the boards and cause honeycombing.
Since this pit was 45 ft. deep with a future head of water equal to 36 ft.,
it was necessary to let the drainage system function until all the walls

were poured and properly cured. Then the valves were closed, the 1

bottom of the sump was filled with concrete to encase the valves, and the
structure was put into service.
1
Plugs could have been used to close the pipes without the valves and the enlarged
bottom of the sump, but the engineer wanted to be certain that future corrosion of the
steel would not cause leakage.
Art. 11-12] PIECEMEAL SHORING OF DEEP EXCAVATIONS 503

In this case, steel waterstops were used at all construction joints.

After the concrete work was completed, the voids (hay, etc.) outside the
wooden sheathing were low-pressure-grouted from the top so as to mini-
mize the possible slumping of the soil around the pit, since the ground
might settle into the voids in the course of time.

17-2"
r UL L^L LA LUI
t
I4"xl6

2-/8"
lr <r
47/bH

/Ox 12"

X
s^ c* 1
1
ggjgggg
Deta/'/A 36"-230/b.H
'^PLAN-TEMPORARY SUPPORT h PLAN-COFFERDAM
FRAME WALES, BRACING AND SHEETING

Detail B

r//2-/"d/c*m.
bolts fastening
sheet pile
and 36"
JC&.-29
Top of seal

297 woodpiles
approx. tip El.- 1IS

DETAIL- SECTION B-B

FIG. 1 1-32. Construction of a pier on piles inside a cofferdam using steel sheet piling with heavy steel

beams as master piles to reduce interior timber bracing. (See Puree//, Andrew, and Woodruff, Deep
Open Caissons for Bay Bridge, Engineering News-Record, Aug. 23, 1934.)

11-12. Piecemeal shoring of deep excavations. One method of

shoring deep excavations when sheet piling is not available or is not
desirable is the use of steel beams or steel H piles as pictured in Fig. 11-33.
These are sometimes called soldier beams. This general scheme is useful
in underpinning jobs and in other instances where it is important to avoid
504 COFFERDAMS [Chap. 11

settlement of the surrounding soil and of the structures presently sup-

ported on it. However, it is not advantageous when ground water can-
not be removed readily by pumping or when the soil is a fine sand that
may flow too easily or become quick.
This method has many variations in detail, but one arrangement is
shown in plan in Fig. ll-33(a). Steel piles are driven at a selected
spacing and in any desired pattern. These are driven down far enough
below the intended bottom of the excavation to secure the desired toe
hold for future use. Excavation proceedings are then started at the top,
as indicated between A and B. Planks are cut to such a length that they
can be inserted between the flanges of the piles as pictured at BC. These
planks are set against the steel as shown between D and E\ and soil, hay,
or some other suitable material should be packed behind the planks.
Excavation and shoring are carried on by progressive steps from the
top downward, as indicated by NO in Sketch (6), until the bottom is at
the required elevation TU in the sketch. Much improvisation is needed
in practice when doing such work. To begin with, it is next to impos-
sible to drive the piles exactly as desired. Contact at the corners AM,
CD, etc., may not be attained; hence, wedging with strips of wood or
stuffing crevices with hay may be necessary if the soil runs out. The 1

planks have to be cut to suit the spacing of the piles, and these planks
may have to be held up by nailing strips attached to bracing at the top if
necessary, as shown near Land M
of Sketch (a).
The piles have to be braced as circumstances require. In Sketch (a),
steel wales are shown near F and G. These have to be made to fit the
conditions, generally by burning to proper length in the field. They also
have to be held in place by welding to the soldier beams, by bolting, or by
means of a series of struts and tie rods as pictured for Q in Sketch (6).
These members are suspended from supports like X at the top; they must
be strong and well connected because excavating equipment may bump
them and cause them to be displaced. Furthermore, it is frequently
necessary to drive wedges between the wales and the soldier beams because
the latter are seldom in perfect alignment. The use of timber wales as
shown near J and K is often more advantageous than steel beams if the
lengths and pressures are not too great. Whatever the material may be,
sets of wales are to be added as the excavation proceeds downward. The
lowest set should be placed so that the reactions against the ends TV and
UW will not exceed the passive resistance of the soil or the safe cantilever
action of the soldier beams. All the details naturally have to be made to
suit the materials used for the soldier beams and for all other parts.
If the excavation is made by the soldier-beam method, it is probable
1
Generally a clayey, medium sandy, or gravelly soil will arch across a considerable
opening without flowing in; hence, the joints may not need to be tight.
506 COFFERDAMS [Chap. 11

that the earth pressures from the outside will be more or less concentrated
on the beams rather than spread over the concrete walls of the completed
structure. This is because of the straw and various soft materials often
used as "stuffing" behind the lagging. If the pit is to contain some fluid
under pressure, it may be desirable to design the walls to span between
the soldier beams and so be strong enough to resist the full internal pres-
sure. In the case of cylindrical subsurface tanks, it may be advisable to
design the walls for a large part of the ring tension caused by the internal
pressure because of the tendency of the soil to arch around the excavation
and therefore not produce the full theoretical uniformly distributed active
lateral pressure. If the tank is to be empty, one should nevertheless

design it for the full active earth pressure from the outside, because
slumping of the soil and filling of the voids behind the lagging may even-
tually cause the earth to bear fully on the sides of the structure.
The soldier beams may be designed on principles similar to those
explained for multibraced cofferdams.
11-13. Surcharge pressures on piling. In practical engineering
one often encounters special problems for which he has to invent solutions.
Not always are they subject to exact analysis. As stated many times,
the designer has to devise a structure or system to serve his purpose, he
has to estimate the magnitudes of the loads acting on it, he has to deter-
mine how the structure is likely to act under the influence of these loads,
he has to make certain that the structure will be safe, and he has to make
sure that it can be built with safety and reasonable economy. Each
engineer may have his own ideas regarding what is best under the cir-
—
cumstances and he has to assume the responsibility.
The principles involved in a problem which may arise in work requiring
deep excavations are illustrated by the partial layout pictured in Fig.
11-34, which shows a large pit to be built as part of a metallurgical plant.
The pit was to be 40 ft. deep, and it was close to two large column foun-
dations. Other columns in the same row were at 40-ft. spacing, so that
only the two shown were affected. Should the column footings be placed
at the level of the bottom of the pit (40 ft. down), should they be sup-
ported by long piles, or should they be placed on the adjacent soil next to
the pit? The last solution is, of course, the easiest.
There are many how the soil will behave in
uncertainties regarding
such a situation as These uncertainties are not removed by "wait-
this.
ing until tomorrow," hoping that they will vanish. Assumptions and
decisions have to be made. They may be right, or they may not be. In
any case, they should be conservative and reasonable.
In this case, to estimate the effect of the surcharge pressures from these
columns, the following assumptions were made, and they are illustrated
in Fig. 11-35:
Art. 11-13] SURCHARGE PRESSURES ON PILING 507

8-0" 20'-0" .
8'-0"

FIG. 1 1-34. Plan of pit alongside columns.

18'-0"
-,x
|

hj\ \n
4 4 6\
/lO'-O"
'-
<r
4 '- 6 "^-
>l \
'$
§^
^
,

5.
It
1 c
k/a
J
til
/ 1 P=1.85 'vert.\

\ pD /8'xl7'=136°' % _1_ \Q
/;
i:
^ \
/ <=si \
R
k/o ' \
Area = 8'y.2l'=168°' _j/.5 -V
S -

\
\
k/a ' \
8'x23.5'=/88°' \J.34 I
A
\
\
k/a \
8 'X 26 '=208°'
'

±_7.2/ \
\
\
k/a '
\
8'x 28.5 '=228°' ±J.1 \
\
\
\
°'
U
8' x 31' = 248°' ±!.02 k/
w \
WAWW WAW/WAV

_ Y

[a) TRANSVERSE SECTION AT (b) LONGITUDINAL SECTION ALONG

CENTERLINE OF COLUMN CENTERLINE OF COLUMN ROW

FIG. 1 1-35. Study of surcharge pressure on side of deep pit.

508 COFFERDAMS [Chap. 11

1. Since the total load on each column was 504 kips, one-half of this
load was to be supported by the soil between the center line of the column

row and the adjacent sheet piling of the pit.

line of the
2. The bearing areas of the footings were to be 5 ft. below the floor,
and the footings were 8 by 9 ft.

3. The 1 rates shown until the

pressures were to be spread at the 2 :

piling was reached at D in Sketch (a) and the center line of the pit and
column group was reached as shown by R in (6). Distribution was
allowed to continue to Q, T, and V.
4. The effect of vertical friction on the piling was neglected, causing

computed vertical pressures to be larger than they would be otherwise.

5. The horizontal pressure at any level was computed to be one-third

of the computed vertical pressure at that point.

The space below point D in (a) was divided into imaginary 5-ft. layers.
The area of soil assumed to be effective in resisting pressure at each level
is shown in (6), as for PQ, ST, etc. The vertical pressure in each case is
also shown there. The pressure diagram for a vertical strip 1 ft. wide is
shown by LMKD in Sketch (a). With these surcharge loads and the
computed active pressures exerted by the soil itself, the concrete lining
of the pit was designed as a continuous frame made up of imaginary
horizontal layers. The piling and its bracing also had to be designed for
these complete surcharge loads since the pit could not be concreted until
the cranes were in service.
This scheme was adopted. The sheet piling was left in place, and the
concrete was poured directly against it. The bracing of the cofferdam
was removed one set at a time as the concrete work was completed in
successive layers.

PROBLEMS
11-1. A single-braced single-wall sheet-pile cofferdam is to be constructed
around the base of a bridge pier. The soil profile is shown in Fig. ll-36(o).
Design the cofferdam.
Suggestions: Do not overlook the fact that the pier is to rest on sand. Assum-
ing the buoyant weight of the sand to be 70 p.c.f., the piling should extend down
at least far enough to have this weight balance the hydrostatic head of 15 ft.
Perhaps it should extend much farther.
11-2. A can tile vered single-wall cofferdam is to be built around a large excava-
tion. The piling goes through a saturated top layer of silt 12 ft. deep. Under
the silt is a deep layer of fairly dense clay. The bottom of the excavation is to be
2 ft. below the top of the clay. It is proposed to drive the sheet piles 24 ft. below
the bottom of the excavation. Is this adequate? If so, estimate the maximum
bending moment in the piling.
11-3. Analyze the single-wall cantilevered cofferdam pictured in Fig. 11-36(5).
Will it be safe? Estimate the bending moment in the piling.
 PROBLEMS 509

11-4. Estimate the maximum bending moment in the sheet piles of the bulk-
head shown in Fig. ll-36(c), the bending moment in the wales as simply supported
beams, and the tension in the tie rods. Is the embedment safe? Ties are 8' c.c.
11-5. Assume a double-wall multibraced cofferdam, as shown in Fig. ll-36(d).
Estimate the maximum bending moment in the inner sheet piling as though it
acted alone. Estimate the pressure per foot on each of the wales.
11-6. Assume the same cofferdam as in Fig. ll-36((i) except that the wales are
located as follows: one at El. 300, one at El. 290, one at El. 280, and one at El. 274.
Estimate the pressure on the wales and the bending moments in the piling, assum-
ing the inner row to take the load.

El. 2 10

El. I9S
*- Tie rod
W/ "Dead man"^

El. 180 ''Sheet piles

-Sand Sat rid

:•;.: 5 a'rid:..
(El. 16
-
HaTrid~pan
faf BRIDGE PIER C6;-CANTILEVERED COFFERDAM rcA TIED BULKHEAD

^Sheet piles
IS'-O" . I20'x200'
-El. 106

^I1IEjIZZmEI.I02 &«ier table El 103 .

if! 96-, ?0la'-'

foundation , ^ New structure
cEI.88
Silt
r rrr
rEl.81

R o c k '

Sdnd :

fdJ-DOUBLE-WALL COFFERDAM TeJ- BUILDING ALONGSIDE OLD STRUCTURE

FIG. 1 1-36. Miscellaneous data for problems.

11-7. Assume a diaphragm type of cellular cofferdam like that in Fig. 11-22(6).
Assume a radius of curvature of 30 ft., a maximum width of cell of 70 ft. (10 ft.
between centers of the two circles), a spacing of diaphragms of 30 ft., a depth of
excavation of 32 ft., to rock, and saturated fine sand outside the cofferdam with
4> = 30°. Test its safety. The fill is gravel, with 4> = 33°. No berm is used
inside.
11-8. Assume a cellular cofferdam like that of Fig. ll-22(/). The diameter is

60 ft. ; the spacing, 64 ft. ; the width of closure sections, 50 ft. ; the depth of excava-
tion, 35 ft.; the soil outside, 35 ft. of fine, saturated sand (<f>
= 30°); the fill, 40 ft.

of gravel (0 = 33°). Is the cofferdam safe?

11-9. A large new building shown by the dotted lines in Fig. ll-36(e) is to be
founded upon cast-in-place concrete piles under a mat. It is to be alongside the
existing building shown at the left, and a street is 20 ft. from the side. How
510 COFFERDAMS [Chap. 11

would you conduct the work so that the piles can be driven in the dry, and the
building and street be protected?
11-10. A viaduct pier is to be built, as shown in Fig. 11-37. How can the work
be done to best advantage?
11-11. A contractor proposes the use of a sheet-pile wall temporarily to protect
an excavation 9 ft. deep. The top 9 ft. of soil is silt (w = 100 p.c.f., and <f> =
25°). Below that is deep medium sand (w = 100 p.c.f., and 4> = 32°). The

40-Qc.c.
-i^-Pier shaft S'-0"sq. ± -
0\y/W'v/wwvsA High water-' Low wafer, Peat and muck

10'-
Si It

Fine sand,
rFoof/nq
4 12'sa +
some pebbles

:
j '':?&
'
~Hardpan and boulders

FIG. 11-37. Problem in planning construction of a viaduct pier in a swamp.

piling is to be driven 18ft. below the bottom of the excavation. The water level
will be 3 below the bottom of the excavation. Will the wall be safe? If not,
ft.

what depth should be used ? What size of piling is necessary? Allow a surcharge
of 100 p.s.f. behind the piling.
11-12. Assume the same conditions as in Prob. 11-11 except that a surcharge
of 300 p.s.f. is to be provided for. Struts and a wall are to be used to brace the
top of the sheeting, and the bottom of the piling will be assumed to be fixed.
Will this be safe? Compute the size of piling necessary.
 12
CAISSONS

12-1. Introduction. The term caisson denotes a structure that is

used as a protection and aid in the making of excavations for foundations
but remains in place as a part of the permanent structure. In general, a
caisson is sunk into position as the excavation proceeds inside of or
under it.

Caissons may be large or small, deep or shallow. Large ones are espe-
cially useful for heavy bridge work; small ones, for buildings and miscel-
laneous structures. They are exceedingly useful in the making of deep
foundations where other methods are inadequate or undesirable. Their
use in shallow excavations is generally limited to cases where some special
conditions make them necessary.
The and sinking of caissons are parts of engi-
design, construction,
neering in which much is by experience. Many things are
to be gained
done in certain ways to meet certain conditions because they have been
found to be successful in the past. Some of the work involves theory but
much is practical engineering, and it requires "know-how."
In this chapter an attempt is made to discuss the kinds of caissons that
have been used, to show the general principles underlying the planning of
them, and to illustrate some of the conditions for which each type is best
suited. Much will be covered by drawings of various arrangements that
have been used. By necessity these drawings have been simplified con-
siderably in order to show the basic ideas without including too much
detail. A collection of the working drawings giving all the details of just
the commonly used types would book in itself. This cannot
constitute a
be attempted here. If it were, the expert would not need to look at it,
and the beginner would be lost in a mass of material because he would not
understand the reasons behind it all.
12-2. Open caissons. As its name implies, an open caisson is one
that has no top or bottom during its use as a protection for the excava-
tion process. It is like a box with vertical sides only, a honeycomb
structure with open ends. In plan, it may be square, rectangular, cir-
511
512 CAISSONS [Chap. 12

cular, oval, or whatever shape best suits the situation. If small enough,
the center may
consist of one opening or well; if large, it may be sub-
divided by vertical partitions into a series or group of wells.
Figure 12-1 illustrates a very simple open caisson and the operations of
sinking it. This is for the foundation of a tall approach pier that is to be
part of a high-level bridge. It is near the shore line where the muck has

& ofpier
Cutting edge Water
El. 203
'EI.200
C5

j-'o" J'O"

/6-0"
(c) (d)
r«;-PLAN of
CAISSON

Sand /20

Bedrock \

Pier-*,
f6J-S0IL PROFILE shaft

^m %<*

(e) ffj (A)

FIG. 12-1. Sinking of an open caisson.

been covered with a blanket of fill. The structure of the caisson is of

reinforced concrete.
The following paragraphs refer to the sketches in Fig. 12-1
(a) This is a plan of the caisson. Assume tentatively that the dimen-
sions are as shown.
(6) This gives the soil profile at the center line of the pier. The pier is

to be founded upon the rock at approximately El. 120. The datum is

chosen so as to avoid any minus elevations because their use is likely to

cause errors in the office and field work.
Art. 12-2J OPEN CAISSONS 513

(c) not so difficult to start a caisson when there is ground at the

It is
surface to work
on, as there is here. A small excavation is dug at the
surface, then the cutting-edge portion of the caisson and a small section
above it are built in the position where the pier is to be. In this picture,
some excavation under water has been carried on, and the caisson has
started its downward course.
(d) It sounds simple to say that the further sinking of the caisson is
accomplished by excavating under it and letting it settle into the hole
because of its great weight. This is substantially what is done. How-
ever, one can realize that it is difficult to control the excavation because
it has to be made by equipment working under water and within the limits

of the area that can be reached by clamshell buckets (or other devices)
lowered through the open area way. If one side slumps in from under the
cutting edge but the other does not, there is a tendency for the shaft to
hang up on the latter side. If this persists, it may cause tipping of the
caisson. Once tipping has occurred, it is difficult to correct. Adding
weight to the high side and trying to undermine the higher cutting edge
may cause the caisson to straighten up as it sinks farther. This is obvi-
ously an uncertain control. It is therefore very important to have the
caisson started in the right location and kept in a vertical position during
this early stage of the sinking process.
(e) The caisson is now shown in a position well advanced toward the
rock. By time the supporting effect of so much earth around it will
this
tend to compel it to continue its course. This is fine if the sinking has
been done accurately thus far; it is correspondingly difficult to correct any
errors in its position. Friction of the soil bearing against the outside may
now constitute very real resistance to further sinking. Pipes for water
jets should be built into the caisson so that the jets can be used to loosen
the material next to the caisson and reduce its frictional resistance.

(/) Here the caisson is shown with its cutting edge on the rock. The
last several feet of progress may
be extremely difficult when boulders,
hardpan, or irregular rock is encountered. When a cutting edge is hung
up on a large embedded boulder that is under one side, there really is
trouble on hand. What can be done with it when it lies from 100 to 200
ft. below the surface and cannot be reached by the buckets or dislodged

by water jets? If not too deep, divers may

be able to undermine it or
blast out.
it If off near it may
too deep for that method, small blasts set
do some good, or they may do harm to the caisson or the surrounding
areas. At best, they are a dangerous stab in the dark. Hardpan may
also be difficult to remove from under the portion of the caisson that is
shielded by the latter's thickness. A cutting edge can cut into soft soil,
but it is not a pickax to chop away resisting materials. Compacted sand
or gravel under one side and running sand under the opposite one may
514 CAISSONS [Chap. 12

also create a difficult situation. When the rock is irregular or steeply

sloping, it is usually very hard to excavate the rock so as to seat the caisson
properly. In addition, disintegrated and seamy rock may not be removed
and cleaned properly to secure adequate and even bearing. This phase
of work —
seating the caisson and cleaning the bottom is a very impor- —
tant one.

S J* —33 '
-I ,
^ fTf^W^H* .
- , ,

6
%F/n£UI4.00
\

Splice

QUEENS TOWER
SECTION A-A

FIG. 12-2. Details of construction of caissons under towers of the Whitestone Bridge. (See George L
Freeman, Deep Caisson Work at the Whitestone Bridge, Engineering News-Record, Aug. 3, 1939.)

(g) In this view, the inside of the caisson has been filled with a deep
plug of tremie concrete, and the water above it removed. This latter
may not be necessary since, if tremie concrete is good enough for the
lower part, it should be sufficient for the upper portion of the fill or body
also. Notice that the top of the caisson is below the ground level so that
a sort of footing of the pier can be extended over the walls and the con-
crete fill without projecting above the surface. In some cases, it may not
Art. 12-3] PNEUMATIC CAISSONS/ 515

be necessary to use more than the tremie-concrete plug in the bottom,

but solid concrete seems to be desirable here.
(h) This indicates the completed structure with the pier in place. The
caisson is now merely an integral part of the heavy solid shaft that sup-

ports the bridge.

Open caissons may differ greatly in shape, material, and details but the
preceding paragraphs show the general principles involved. The caissons
for the towers of the Bronx- Whitestone Bridge at New York are pictured
in Fig. 12-2. The Queens anchorage for this bridge consists of four cais-
sons. The riverward pair were 33 by 100 ft. in plan, were built of rein-
forced concrete with 5-ft. walls and partitions, and had steel cutting edges.
They were sunk approximately 150 ft. below water. The rear pair of
caissons were 24-ft. -diameter cylinders of reinforced concrete with a'6-ft.
wall. These were also sunk approximately 150 ft. In all four cases,
excavation was carried several feet below the cutting edge because of
sloping rock. The bottoms of the cells of all caissons were sealed with
concrete plugs extending 40 ft. above the cutting edges in some cases and
full height in others. These cases show how varied the shapes and uses
of caissons may be.
12-3. Pneumatic A
pneumatic caisson is a caisson with a
caissons.
permanent or temporary roof or near the bottom so arranged that
floor
men can work in the compressed air trapped under it. In this way, the
difficulties described in connection with Fig. 12-1 (/) can be overcome.
On the other hand, the work may be much slower, and it almost certainly
will be more expensive. Nevertheless, its successful completion and the
safety of the structure may require the use of compressed air. •

When the borings show the likelihood of serious and perhaps hazardous
difficulties for the sinking of an open caisson, a pneumatic one should be
used, or an open caisson may be made with provisions for the installation
of compressed-air work if it becomes necessary.
Figure 12-3 illustrates the principles of a pneumatic caisson and the
operations in the sinking of one. This caisson is now supporting the New
York River Ventilation Building of the Lincoln Tunnel. The sketches
are greatly simplified. The following comments refer to the correspond-
ing drawings:
(a) The caisson is composed of an inner and outer skin plate of steel,
with steel trusses or girders as horizontal supports to form a boxlike
structure. The cutting edge is fabricated steelwork with strong triangu-
lar brackets to support it. A
working chamber, approximately 10 ft.
deep, is made by installing an heavy beams and buckle
airtight roof of
plates as shown. Access to the working chamber is secured through the
tubular shafts and air locks, which will be described later. In this case,
the lower portion of the approximately 40- by 50-ft. caisson was fabri-
516 CAISSONS [Chap. 12

Shafts to locks
r=c V*i Double sieel shells
and trusses
Bulkhead
Roof /Water
""' Tit Bulkhead,
Cutting
edge girders not
X -LWorkinq^- shown
chamber
mm
'- ~
.l>
5
<o-3 7ff? —
i-,-"5 fit .Concrete
fill

y Compressed x~
C *' r
< - -J

Rock
^^m^^^^
(a)-LOWER PORTION OF STEELWORK (W-CAISSON BEING SUNK
FLOATED INTO POSITION THROUGH SILT

Stone
facing

s-S

m?$ ;
ii s ...rrr^y^

Compressed IKHtti
air in
Tunnel- working
chamber
and funnels
LI"™'

Concrete seal
fcJ-CAISSON SEATED ON ROCK frfMUNNELS CONNECTED,WITH
AND SEALED WITH CONCRETE ROOF AT HIGHER LEVEL

FIG. 1 2-3. New York river shaft of the Lincoln Tunnel built by the Port of New York Authority under the
Hudson River at New York City.

cated as a unit, floated to the site, and anchored. The sinking was
started by filling the spaces between the skin plates with concrete after
the steelwork was built high enough to maintain freeboard.
(6) By maintaining compressed air in the working chamber at a pres-
sure sufficient to counteract that of the silt and water at the cutting edge,

men could excavate the silt from underneath the caisson. This was dirty
Art. 12-3] PNEUMATIC CAISSONS 517

laborious work. When a few feet of silt had been excavated, the caisson
was usually sunk by adding more steel and concrete, and when the men
had left the working chamber, by temporarily decreasing the air pressure.
Rapid release of air pressure under the caisson is relatively equivalent to
giving it a downward shove. This process was repeated many times as
the caisson progressed downward.
(c) When the caisson was close to the rock, the latter had to be cleaned

off and excavated to provide a firm level support for the concrete seal or
invert that was placed under compressed air. This work was carried on
about 90 ft. below the surface and in an air pressure of approximately
45 p.s.i. An ordinary limiting depth below water for compressed-air
work is 100 to 110 ft,

(d) Circular bulkheads of steel plates and girders had been provided in

two opposite walls of the caisson. After the latter was seated and sealed,
it was necessary to extend the working chamber so as to have the roof
above these bulkheads. The work in the tunnel shields was also done
under compressed air. They approached from opposite sides until they
were close to the bulkheads, then the latter were burned out, the tunnels
advanced, and the tunnel linings joined to the caisson walls as indicated
here. After that the compressed air was released, the roof removed, and
other construction carried on. The caisson now made "an open hole in
the river."
Compressed-air work is rigidly controlled by law as far as working
time, maximum pressures, and, generally, wages are concerned. It is
hazardous work, but it can be done without serious consequences if
planned and executed properly. A large item of cost is for manual labor
when pressures are high. For example, the regulations affecting the
work on the caisson of Fig. 12-3 permitted a man to work a normal 8-hr.
shift with a lunch period when the air pressure was at or below 18 p.s.i.
For higher pressures, the working time was decreased. For example,
under a pressure of about 45 p.s.i., a man could work 45 min., then he had
to lock out and rest for approximately 5 hrs. After that he could go back
and work for another 45 min. Not only was a man paid his regular daily
wage when he worked this shorter time under high air pressure but he was
paid a bonus also.
The general principles of the operation of the air locks are illustrated by
Fig. 12-4. In Sketch (a), a man who wants to enter the working chamber
has gone into the man lock through door a. Door b is kept closed so that
the compressed air in the working chamber cannot escape. The man
sits down while a lock tender closes door a, turns on the air pressure, and
lets the pressure in the lock build up to equal that in the chamber. The
door b is then opened, and the man descends to the working chamber, as
shown in (6).
518 CAISSONS [Chap. 12

When leaving, the procedure is reversed. Door a remains closed while

b is open. The man climbs through the shaft to the lock and sits down,
door b is closed, the exhaust valve is opened, the air pressure drops slowly
until the lock hasnormal pressure like that outside, door a is now opened,
and the man goes out. Going under air pressure can be accomplished
fairly rapidly. Coming out is a slow and chilling process, but slowness is
necessary in order to allow the body time enough to eliminate the excess
air absorbed or taken in when under pressure. Too rapid a decompres-
sion may cause bubbles of air to be trapped in muscles, joints, or the
blood, producing caisson disease or " the bends." For example, the law
may require that the rate of decompression be 1 min. for each pound of
pressure up to 30 p.s.i. and 1% min. for each pound in excess of that,

Exhausf (From compressors

Airlock, |
Door °Pen Door closed^
\

Ladder
/ Bench,

Tubular
shaft
^ ^Normal air
pressure inside
Same air pressure as
"Jo working
in working chamber
chamber
in caisson

^ Roof of working Roof

w///< chamber '/////<

b -^y~ Door closed ''Door open

Compressed our in Compressed air in

working chamber working chamber

taJ-FIRST STAGE OF "LOCKING IN" f&l-SECOND STAGE OF "LOCKING IN"

FIG. 12-4. Simplified illustration of operation of air lock for access to working chamber of a pneumatic
caisson.

because this rate has been found to be successful in most cases. Rigid
control and observance of regulations will almost always prevent serious
trouble.
The handling of materials through the locks is similar to that for the
men except that it can be much when no men accompany the
quicker
materials. It is, of course, a slower and more costly process than when
the operations are performed in open air.

The material locks (and occasionally the man locks) used in caisson
work are generally vertical cylinders. They are provided with a trap
door at the bottom, means for passing buckets and cables, and a side door
for the discharge of material dumped from the buckets. Such a lock is
shown in Figs. 12-5 and 12-18.
When it is so expensive, why could not the caisson in Fig. 12-3 be sunk
as an open caisson? The chief reason is the necessity of making sure that
Art. 12-3] PNEUMATIC CAISSONS 519

FIG. 12-5. Views of the caisson for the New York River Ventilation Building of the Lincoln Tunnel-
(Courtesy of the Port of New York Authority.)
520 CAISSONS [Chap. 12

12 gage,
/! "° mesh
1

Recesses for
well points
Ladder and plat-
forms installed
after sinking
and sealing
caissons for
access to well
point system —
a
Continuous
steel ladders

*A- 4 discharge pipes

Typical well point
assembly and
recess
Typical wash pipe
assembly 2" wash pipes
8" stand pipe
8" pipe unused

Pump room--^^
6 "plumb pipe
Centrifugal pump,
Platforms spaced 7 stage -300 g. p.m.
20 + ft. vertically
Vacuum tanks 10" screen
5 "suction pipes 2" wash pipes -

6"suction pipes Working chamber

Underpinning Cutting edge
3"grout pipes 3" grout pipes
Underpinning \Sump[ r~
drilling and for future i

for future grouting Gravel filter

grouting of
foundation
Elevation 500 Concrete seal Culvert, perforated-

FIG. 12-6. Two pneumatic caissons like this were sunk through 180 ft. of alternating water-bearing
and impervious materials at the Merriman Dam site. The air pressure was kept to a maximum of
32 p.s.i. by lowering the ground water. (See Caissons for a Cutoff Wall, Engineering News-Record,
Dec. 5, 1940.)

it can be sunk without insurmountable difficulties, and that it can be

seated and sealed properly because it is to support a 150-ft. building that
will house ventilation equipment. Old piles, riprap, sunken timbers, for-
gotten bulkheads, and miscellaneous debris are among the things that
may cause trouble when working at old water fronts. Furthermore, the
rock was known was essential to get a level bearing, to
to be sloping, it

remove disintegrated rock and any unexpected boulders or knobs, to

clean out any crevices, and to have the bottom thoroughly concreted.
Art. 12-4] MATERIALS AND DETAILS 521

Pneumatic caissons may be used for purposes other than foundations

and sometimes are sunk more than 100 or 1 10 ft. below ground. One such
case, illustrated in Fig. 12-6, shows one of two exploratory caissons that
were sunk as part of the core wall of the Merriman Dam of the Delaware
water supply for New York City. Since such caissons had been used for
the cutoff wall of the Quabbin reservoir, it was desired to learn whether or
not similar construction would be suitable in this case where the ground
varied from runny sand or silt to glacial till and boulders, and most of it
was water-bearing. There was artesian action in the upper seamy part
of the rock and in some of the soil.
The caissons were reinforced concrete 15 by 38 ft. in plan, and ulti-
mately 180 ft. high. They were located with a clear space of 48 ft.
between them. They were started in a 17-ft. trench, sunk about 35 ft.
in free air, then continued under compressed air except when in imper-
vious ground or that with little water. As shown in Fig. 12-6, the caissons
were equipped with pumps to lower the ground-water level, operating
like wellpoints. This equipment was able to lower the watertable about
120 ft. so that the caissons were sealed under an air pressure of only
approximately 26 p.s.i. Special gravel filters and perforated pipes were
installed for future pumping.
Jet pipes were installed in these caissons but were not used. No top
load had to be applied although, when under air, the caisson had to be
"blown" to start it downward. Estimates of sliding skin friction with
the caisson moving gave values from 570 to 924 p.s.f.
12-4. Materials and details. Caissons are generally made of steel,
reinforced concrete, or a combination of the two. This is natural because
they are needed where excavations are deep and pressures are heavy.
Wood may be used to some extent. The structures must be tight, gen-
erally must be built in successive lifts as the sinking progresses, should be
durable, and should be heavy for aid in sinking. These are things for
which steel and concrete are particularly adapted. In many cases, as in
Fig. 12-1, the interior is filled with concrete so that the caisson is not
heavily loaded in the finished substructure. In others like Fig. 12-3, the
caisson itself is the final substructure and must support large loads.
The choice of steel or concrete depends upon many things. The former
can be fabricated in advance and erected quickly, and boxlike construc-
tion eliminates the need for forms. It is also strong and ductile. It will
withstand great inequalities of loads without collapse, as when a cutting
edge strikes an obstruction. It can be planned so as to provide for
attachments of many kinds, such as the provision of connections for the
roof of a working chamber in an open caisson if the need for compressed-
air work arises. As in Fig. 12-3 (a), a large portion of the caisson can be
floated into place when there is sufficient depth of water. The steelwork
522 CAISSONS [Chap. 12

may be riveted and calked for watertight ness, or it may be welded. When
the caisson must be started water and muck, it is an advantage
in shallow
to be able to assemble the steel shell by working from temporary plat-
forms, to lower the steelwork alone until it is self-supporting or is borne by
the muck, and to add the heavy concrete fill when it is convenient to do so.
On the other hand, reinforced concrete alone is sufficient in many cases,
especially when the caisson can be started on ground, as in Fig. 12-1 (c).
Forms are necessary, but sliding forms, especially when made of steel, can
be used efficiently and without unreasonable cost. Quick-setting cement
will also reduce the time required before the forms can be stripped and the
concrete put under stress. The use of concrete alone for caissons is shown
in several of the illustrations. Naturally, as far as materials are con-
cerned, caissons of moderate size built of reinforced concrete alone are
often more economical than are those made of structural steel plus con-
crete fill.

The size, shape, and depth, the construction procedures used, the char-
acter and weight of the structure to be supported, the difficulties that
may be encountered, available materials, transportation, past experience
under reasonably similar conditions, and relative economy — all these are
influences that bear upon The
the choice of the material to be used.
entire program should be thought through carefully before a decision is
made.
One of the important details of construction is the cutting edge. This
should be simple but strong. Nothing much can be done about it if a
cutting edge fails under the local pressure caused by some obstruction
100 ft. or more under water. A triangular or trapezoidal shape is gen-
erally desired in order to reduce the bearing resistance of soil trapped
under the caisson walls, and to give more chance to remove the ground
that is not directly under the shaftways. Great strength is needed to
resist compression and twisting from a tendency for the beveled faces to
cause the walls to spread out.
Figure 12-7 illustrates some details for the construction of cutting edges
for caissons. Some of these are satisfactory; some are of questionable
value but are shown purposely. The followingcomments refer to the
various parts of the drawing:
(a) This shoe is The outside plate is to provide
for a concrete caisson.
stiffness and to spread any The 8 by 8 angle is stiffened by
localized load.
welded plates to hold the downstanding leg straight. The horizontal
stiffener angle serves as an aid in spreading loads upward into the con-
crete, and as an anchor if the cutting edge is forced inward. The con-
tinuous tie angle and the top cross ties help to anchor the bracket against
forces that tend to push the cutting edge outward. The reinforcement
below the cross tie consists of short lapped bars that are threaded through
Art. 12-4] MATERIALS AND DETAILS 523

Reinf. threaded ^ o Intermittent

through and welded seats for seal
during assembly Bentpl.
Bracket 2 L 6 "to 4 L c. c.
Brackets 2-0"± c.c.
Tcut from WFbeam
-Welded V pis. I2"±c.c.
^8x81
faJ-PLATE-AND-ANGLE CUTTING W-TRIPLE-PLATE, RIVETED ^-INTERMEDIATE WALL WITH
EDGE WITH WELDED BRACKETS CUTTING EDGE ANGLE AS SHOE

Bars welded
j§; fto brackets
Straps
fcH
K
Continuous
seat for seal
#11 welded to T
f
J
Welded brackets
"Welded stif.
^4
(d)-l AS SHOE WITH fal-ALL-STEEL (f)- ALL-STEEL (g)- FULLY SHOD
BRACKETS ONLY CAISSON HOLLOW INTERMEDIATE
FOR LIGHT CONSTRUCTION WALL
Horizontal- CONSTRUCTION
truss Roof beam.

Buckle pis. over

Horiz. girder
working chamber
Holes for
Keyway [concreting ^Inner skin pi
for seaI
/

Intermediate
pockets in
concrete as
seat for seal Bent or
welded pi.
Welded
steelshoe
Conf. pi.

(h)-USE OF PLATES f ^-CUTTING EDGE FOR A LARGE (J) -STEEL-SHOD WOODEN

WITHOUT BRACKET PNEUMATIC CAISSON CONSTRUCTION

FIG. 1 2-7. Some details of cutting edges for caissons.

holes in the T's. seems that the weakest part of the construction is the
It
resistance of the concrete to tension at A because the general tendency
produced by the beveled face is to force the cutting edge outward.
(b) This shoe is strong in itself but very short. Boulders might crum-
ple the plates at C, and the anchorage of the steel to the concrete is not
very effective. The section is relatively weak just above the brackets.
524 CAISSONS [Chap. 12

The same scheme made less skimpily would be much better. The offset
at B supposed to reduce the skin friction on the face above it. This
is

might be advantageous in highly cohesive soil that is not disturbed; when

water or jets practically liquefy it, the offset is of negligible value; and it
is not likely to be useful in granular soils. Contact with the soil along
the outside faces is usually needed in guiding the caisson. The small
seats are supposed to help deliver vertical loads to the concrete seal, but
they are relatively insignificant.
(c) This illustrates a very simple construction for a cross wall. Since
there is access of a sort from both sides, there is not so much need for a
heavy shoe. A rounded or sloping concrete section may be sufficient
unless boulders are encountered. The cutting edge of such a cross wall
may be placed at a higher elevation than those of the outer walls. This
may be helpful when boulders or irregular rock are expected to be
encountered.
This cutting edge is a heavy T with welded stiffening plates and
(d)
brackets. The brackets are connected by heavy welded bars that serve
as spacers, as anchors against rotation of the cutting edge, and as regular
reinforcement. The continuous seat near the upper corner is a better
support than that in (b), but the corner is likely to cause trouble by
catching the excavating equipment. The horizontal web of the T would
be a good seat, but the space under it is difficult to fill with concrete when
placed by tremie or underwater buckets. The brackets seem to be fairfy
effective,but they would be better if the T's were extended to reach to
and above the inside corner. In the sketch, the welded bent bars are
supposed to help remedy this.
(e) This shows an all-steel construction for a small caisson. The tri-
angular cutting edge is made by welding an angle to the skin plate. This
plate is stiffened by means of vertical i/'s that cross the wales E and F.
Plate G is to hold the cutting edge to wale F. Notice the following weak-
nesses and defects:
1. The skin plate is the only real stiff ener for the support of the cutting
edge.
Sideward pressures against the angle are supposed to be resisted
2.

largely by plate G, but this merely twists F and, through the straps, the
other wales. Without vertical diaphragms, the wales cannot offer proper
resistance.
3. Plate G will almost inevitably bend badly unless it is stiffened by
concrete backing.
4. When is placed, the spaces between the lower
the concrete seal
flanges of E
and F cannot be filled. The webs should be open, as with
lacing, but this would weaken them too much.
5. The lightness of the caisson is likely to hamper its sinking. Con-
Art. 12-4] MATERIALS AND DETAILS 525

crete fill would be useful for this purpose, and for stiffening the construc-
tion in general.
This sketch shows another type of all-steel construction for small
(/)
caissons. As shown here, the outer skin plate is not well stiffened, the
splices in the skin plates are weak in resistance to large vertical forces.
The wales J and K can be connected to the outer skin plate but, assuming
that the inner plate L is erected last, the connections of L to the wales are
weak. Also, the cutting edge H
is not stiffened satisfactorily. A concrete
fill placed by pumping or chuting through holes in the webs of J and K
would improve the construction. On the other hand, a reinforced-
concrete caisson might be more economical.
(g) This cross wall has a wide shoe of plates M
with angle anchors.
One extra use for the plates may be as a tie between the outer side-wall
cutting edges.
(h) The cutting edge in this case is not very narrow, but this may be an
asset because of its stiffness and resistance to compression. A sharp edge
is not really needed unless very firm soils like gravel and hardpan overlie
the rock that has to be reached. A certain amount of flatness may be
useful in providing a temporary support for the caisson on the ground
below it. The plates with their welded anchors and channels are tied into
the concrete. The channels and P are especially good anchors because
they are continuous and are tied together by the flats. Rods are N
threaded and attached to the cutting edge. This is primarily in order to
have them serve as reinforcement that is well anchored at the bottom.
On the whole, this construction is simple and sturdy.
(i) This sketch illustrates one way of making the cutting -edge and

working chamber for a large pneumatic caisson built of steel with con-
crete fill. The steelwork is generally erected in horizontal lifts, riveted
and calked for watertightness, or welded. The concrete is then packed
in thoroughly in the dry. Since men can get at the ground directly, there
is not much likelihood of excessive localized loads under the cutting edge.
The entire construction, however, should be very strong. The roof has
to act like thebottom of a boat with upward loads that may be 2 or 3 tons
per ft. 2 and it must be practically airtight.
,
Flooding the inside above
the working chamber, or adding "muck," will counteract some of the
uplift if necessary, but such steps should generally be reserved as aids in
sinking the caisson if unexpected resistances develop.
0") Open caissons may also be made of wood when this is the most
available or practicable material. Steel shoes may
be used as the cutting
edge with heavy timber backing as shown here. The details of the con-
struction should be adapted to the use of commercial timber of standard
sizes and lengths in order to avoid needless carpentry work. The details
are so special and wood is so often replaced by steel or concrete for the
*

526 CAISSONS [Chap. 12

construction of important caissons, that no attempt is made here to show

more than this one example.
All the details of caissons should be strong. Design merely to resist
external pressures is seldom adequate.
Corners should be tied together
thoroughly. Cross walls should be integral with exterior ones, or very
strongly connected thereto. The walls should be capable of resisting all
probable distortions caused by unequal bearing or by anything else that
seems to be necessary or probable.
It is important to have the right piping installation for water jets.
Experiences with jets in different cases seem to conflict somewhat. It is
apparently desirable to have the piping arranged so that the jetting can be
localized. For example, referring to Fig. 12-8, the piping for the lubricat-
ing jets on the outside is separated from that for the inside. Further-
more, separate installations controlled by valves should be used for each
side of a rectangular caisson, or for portions of each side if the caisson is

'!!.•»;
i!^'

II!
ft

Nozzle

(b)

FIG. 12-8. Some arrangements for jetting.

large. In (a), the jets on the outside are equipped with nozzles to produce
high velocity and to assist in the elimination of clogging, but the inside
pipes are large so that portable jetting equipment can be dropped through
them. If desired, they can be used to handle a large volume of water and
to produce merely a simple washing effect. A series of outer jets is
shown in (6). The one at the bottom points downward, supposedly to
help undermine the region near the cutting edge, whereas the higher ones
point upward in the direction that the water will have to take to escape.
Portable jets, air lifts, choppers, and other devices may also be handled
by derricks at the top. These enable the men to attack a local resistance
wherever it may occur. Of course, divers may be employed if the depth
permits it.
The jets should be spread along the cutting edge at spacings of a few
feet. Those to reduce the skin friction outside should be spaced similarly
and in rows perhaps 10 to 20 ft. apart vertically.
 *

Art. 12-5] MISCELLANEOUS CAISSONS 527

The big pipes in Fig. 12-8 (a) can be rodded if they become clogged. If

jets get stopped up, there is little that can be done about it. Another
difficulty with jetting is that of control. Except right at the nozzles, the
water will go in the path of least resistance. It may therefore wash out
local channels instead of lubricating the entire surface as intended. It
may also disturb the surrounding soil sufficiently to interfere with the
steering of the caisson. At the cutting edge, the jets may be entirely
ineffective against heavy gravel or boulders that are jammed between
them. Compressed air and steam under pressure have been used in lieu

13'-
0"
2'-0"
Back of Is

FIG. 12-9. Detail of false bottom in outside cell of caissons for Tacoma Narrows Bridge. There are
two of each buckling thrust sets in each cell, 2 ft. each side of the center line of the cell. (Courtesy of
Victor J. Brown, World's Deepest Bottom-door Type Open-well Caissons, Roads and Streets, May, 1939.)

of water. In spite of these troubles, jetting is a great help in most cases

if it is and used wisely, but it is not a cure-all.
installed properly
12-5. Miscellaneous caissons. A caisson may be made as a box
with cylindrical or rectangular walls and a bottom. As such, it may be
floated to position and then sunk to the natural bottom or to bearing on a
previously dredged and prepared area. This sounds easy. However, a
level, is very difficult to
even, firm bearing surface at the right elevation
obtain. Such caissons may be suitable for sea walls or bulkheads, for
which they can be sunk and filled with dirt, gravel, rocks, or even concrete.
Great accuracy is not usually essential for them, and unequal and con-
528 CAISSONS [Chap. 12

siderable settlement may not be too harmful. On the other hand, the
dangers accompanying their use as supports of major structures are too
obvious to need further discussion.
To enable a portion of a caisson to float into place, it may be built with
a false bottom that is removed later. The construction of the false bot-
toms used in the caissons of the Tacoma Narrows Bridge is shown in Fig.
12-9. This is called an open- well bottom-door type of caisson. The
construction of the false bottom deserves careful study. After the cais-
son was floated to position, anchored, and sunk to the prepared level
patch in the river bottom, the cells were flooded to eliminate buoyancy.

SECTIONAL VIEW OF E-5

PIER AND PEDESTAL

SECTION B-B

FIG. 12-10. Details of caisson and false bottoms in dredging wells and general design of pier. (See
Puree//, Andrew, and Woodruff, Deep Open Cassions for Bay Bridge, Engineering News-Record, Aug. 23,
1934.)

Then a hook was lowered into them individually to catch hold of the wire
ropes. When the hoist yanked the ropes upward, the sloping kick blocks
were removed and the whole bottom of a cell was "disintegrated" and
pulled up. The excavation then proceeded by the use of clamshell
buckets working through the water.
Figure 12-10 illustrates somewhat similar false bottoms used success-
fully for part of the foundations for theSan Francisco-Oakland Bridge.
It also shows the general features and pier construction.
of the caisson
A very interesting type of caisson for work in deep water is shown in
Fig. 12-11. It was invented by Moran and Proctor and was used for the
San Francisco-Oakland Bridge. It shows a means of controlling a cais-
Art. 12-5] MISCELLANEOUS CAISSONS 529

son through the use of "false tops" instead of false bottoms. It consists
of a structural-steel open caisson with cylindrical wells to which steel
domes can be welded. Compressed air forced into the capped wells tends
to buoy up the caisson. In this way a large portion of it can be fabricated
and floated to position. Partial release of the air will permit the caisson
to be sunk until it reaches the bottom. Having many cells, it is possible
to build up the steelwork in lifts with compressed air in part of the wells to
provide buoyancy. The open ones can then be extended at the top,
capped, and filled with air. The caps can then be removed from some of
the remaining wells, the tops extended, and the domes welded on again.
This is repeated until a complete lift has been added. Thus the caisson

FIG. 12-11. General picture of caisson used for constructing pier of San Francisco-Oakla.nd Bay Bridge,
showing caisson, working docks, and temporary floating fender system. (See Car/ton S. Proctor, Con-
structing Foundations of Trans-Bay Bridge, Civil Engineering, February, 1935.)

can be water-borne, even when being sunk, until the bottom is reached,
and even later if necessary.
Thereafter, excavation can be made through some of the wells while the
others remain under air pressure. Capping
open ones, removal of
of the
the domes from others, and excavation through the newly opened wells
permit more of the ground to be removed from under the caisson. By
continuing such work in series until the desired excavation has been made
under the caisson, permanent tipping and loss of support can be avoided
or largely corrected. When it is ready to sink to another level, the air
pressure can be reduced and the caisson's sinking controlled. Such work
can be repeated until buoyancy is no longer needed. Then the excava-
tion and sinking can be carried on as for an open caisson.
A substructure may be built by means of a caisson and a cofferdam
combined. One such case for the construction of a bridge pier is illus-
trated in Fig. 12-12. The cofferdam fitted on top of the caisson as shown.
530 CAISSONS [Chap. 12

Ribs 2-6 c.toc.

4''-\—fZ'-6\

SECTION THROUGH *>\

OUTER WALL \=L
'Aek- SECTION THROUGH
HALF PLAN OFCAISSON HALF PLAN OF PIER

El. 117.5

/ ^anchors
{"^anchors 4' c.toc. 2 'c.toc.
El 8025-

Y\ TremJe
I

xl SECTION /-/ SECTION Z-Z

•
/ V £7 /4^5

r ~1
«{!n*i $

AS **-

El. 20
SECTION A- HALF ELEVATION
OF PIER

Al3-9y* 44-8'/i

HALF PLAN OF COFFERDAM

El 99 75^

SECTION B-B THROUGH CAISSON

HALF ELEVATION OF COFFERDAM

FIG. 12-12. Details of north pier caisson and cofferdam for the Lion's Gate Bridge at Vancouver.
(See Convertible Caissons for Bridge Piers, Engineering News-Record, Feb. 16, 1939.)

It was removed later. This combination may be economical for caissons

of large area when is sometimes a real
conditions permit it. There
advantage in keeping the projecting caisson out of sight, provided it is
below the draft of vessels which might otherwise be wrecked upon it.
Caissons need not be large. A type that is very useful for the founda-
tions of large buildings consists of a steel pipe from 2 to 6 ft. in diameter
that is sunk into the ground. The earth inside is removed by ordinary
Art. 12-5] MISCELLANEOUS CAISSONS 531

methods of excavation under water, by washing with water jets, or by-

blowing out the contents by the explosive action of compressed air. In
some cases, the rock at the bottom can be cleaned by divers but, when
the depth is too great, washing and clamshell buckets may have to be
depended upon. It is obvious that boulders may cause serious difficulties
in sinking such small caissons. After cleaning, the caissons are filled with

1
Sfeef
pipe
form

°3
r
kwiwiJ/ —^—
S§

Working chamber
taJ-ROTATING STEEL f6J-CAISS0N WITH fcJ-DRILLED-|N CAISSON
CAISSON BELLED BOTTOM OF WESTERN
FOUNDATION CO.

'S

frfJ-ROTATING HEAD FOR ^-TELESCOPED

CUTTING FIRM SOIL CASINGS

FIG. 12-13. Some details and construction methods for small cylindrical caissons.

concrete. When the soil is sufficiently cohesive, the steel shells may be
withdrawn as the concrete is deposited.
A variation of these small cylindrical caissons is shown in Fig. 12-13 (a).
The steel pipe isequipped with hardened teeth at the bottom. Special
equipment is used to rotate it as it sinks, thus cutting its way through
obstructions and firm soil to rock. On the other hand, boulders and
coarse gravel may interfere seriously with the sinking process because of
their tendency to act like ball bearings.
The small caisson shown in Fig. 12-13(6) pictures a type that was used
for a pier at Seattle, Wash. The overburden on top of good gravel was
532 CAISSONS [Chap. 12

too thin to support piles laterally, and the load was so heavy that fewer
but stronger supports were desirable. The overburden was first removed
by dredging the area. Then the working chamber of concrete was pre-
cast, the removable steel form for the main column was attached thereto
and calked, and the reinforcing cage was inserted. The assembly was
lowered to the bottom by a floating derrick, plumbed, and braced tem-
porarily. Weight for sinking was provided by an 80-ton water-ballast
tank surrounding the air lock that was bolted to the top of the form.
Compressed air forced the water out of the shell. The workmen then
excavated down to firm material and prepared for the concreting. All
concrete was placed in the dry and compacted by the workmen. The
concrete in the bell was allowed to set before the main shaft was poured.
The forms were stripped later, and the cylinders were inspected by divers.
Sometimes this belled bottom can be utilized in stiff clays without the
need for forms. Many of them were installed near New York City by
using a pipe caisson several feet long near the top to cut off the muck and
some fine sand below it. Underneath them was about 20 ft. of clay, which
could be taken out by the use of an auger. The bottom was successfully
reamed or belled out and concreted.
Figure 12-13(c) shows a way of preparing the bottom of a pipe caisson
driven to rock. Some well-drilling type of head is used under water to
grind or chop to pieces any gravel, boulders, or weak rock. After clean-
ing, the concrete fill is placed directly upon the sound rock. The pipe
may or may not be removed. This general method has also been used to
install drilled-in caissons by using 24- to 36-in. pipes, drilling a few feet
into bedrock at the bottom, putting a heavy steel column down to the
rock as shown in Fig. 9-25, and filling the pipe with concrete. Such a
caisson can generally support a load of several hundred tons.
Another type of drilling head is shown in Fig. 12-13(tf). It was used to
sink the shafts for the Montgomery Ward & Co. building in Chicago.
They penetrated 70 to 80 ft. of clay and boulders. The cutters ground
out the clay so that the jets of water emitted from the hollow shaft could
wash it away.
A telescoped arrangement of the pipe casings, as shown in Fig. 12-13(e),
may be helpful in reducing the skin friction to be overcome when sinking
pipe caissons. The upper section may be driven as far as it is practicable
to do so, then a slightly smaller shell may be driven inside of the first, by
means of a follower. If leakage of ground between the two pipes is feared,
the clearance may be reduced, or some type of closure ring and calking
may be used at the bottom of the higher section. In the Gow system, the
bottoms of these caissons are belled out to form a pedestal and increase
the bearing area.
When planning the spaces that may become future working chambers,
Art. 12-5] MISCELLANEOUS CAISSONS 533

one should be careful to give a man space enough in which to work. A

height of 8 ft. is minimum. A width of 4 ft. is about the nar-
a desirable
rowest for a man handling a shovel. Even this provides very inadequate
space when he is supposed to shovel dirt and lift rocks into a bucket under
the shaft.
At the South Meadow Station of the Hartford (Conn.) Electric Light
Co., cylinders 5 to 63^ diameter were to be sunk approximately 110
ft. in
ft. to rock. Above the rock there was about 10 ft. of hardpan that had
water under considerable pressure, 65 ft. of very soft clay, 25 ft. of sand

> ' .
E

3
1
:>' »•;.«

>:'.' St

3 •". E
'»'•',

:
Working :

: chamber \

"Timber '
|

-
"II

fai-STEEL AND /^-CONCRETE toMA/OOD AND

CONCRETE CONCRETE

FIG. 12-14. Some small open caissons for heavy building work planned for pneumatic work if found to
be necessary.

and and miscellaneous fill and sand. The Powell method was
gravel,
used. procedure was the following:
Briefly, the
1. A hole slightly larger than the caisson was drilled down to the hard-

pan by means of a three-pronged rotating bit equipped with five bolted

teeth on each arm.
2. The hole was kept filled with heavy mud to prevent caving-in of the

surrounding soil.

3. When hardpan was reached, the bit was withdrawn, and a steel pipe
or shell was dropped down through the mud. The bottom of this shell
was a double-wall core barrel.
4. The shell was rotated, and chilled steel shot was fed into the cutting
edge. This continued until the cylinder was supported on sound rock.
Most of the waste material was washed out.
5. The hole was pumped dry, cleaned to rock, and prepared for con-

creting. This was done while entering water was being pumped out
continuously.
534 CAISSONS [Chap. 12

6. The hole was then filled with water to stop inflow and washing of
cement out of the concrete to be placed later, a seal of tremie concrete was
placed, the rest of the water was
,Brackcf welded to
4 th f I.
[*""|
building column n fi'47-?
" pumped out, and the shaft was filled

with concrete. Provisions had been

made for compressed air, but its use
proved to be unnecessary.
3rd fl. EI.*3S'-I" 12-6. Special construction
procedures and problems.
When a concrete caisson is to be
started on top of soft ground, it is

^htt'-o" generally necessary to hold it up in

some manner until it has been built
up to a height of perhaps 10 or 15
ft. One way to do this is to set the
steel shoe on a series of closely
spaced transverse timbers that will
serve somewhat as a footing when
Normal
ground Mater. the few lifts of concrete are
first

poured. These timbers should have

blocking added to support the heel
of the beveled edge of the caisson as
soon as it is built, thus spreading
the load and relieving some of the
pressure on the cutting edge. A
seat like that in Fig. 12-7 (d) is use-
ful for bearing on this blocking.
When the caisson is ready to be
sunk, the blocking timbers should
be removed individually in such a
manner as to maintain reasonable
v'.r>Vv"/fT-"
equality of pressure per linear foot
of the cutting edge and to prevent
FIG. 12-15. Sinking pneumatic caissons by means
of four 150-ton jacks reacting against the weight
tipping of the caisson. Heavy
of a building through a special jacking frame for tackle may be needed to pull out
a viaduct built through the Bell Telephone the grillage timbers, especially the
Laboratories in New York. (See George T.
last few that will be scattered un-
Gilman, Railroad Viaduct Carried through Oc-
der the cutting edge and probably
cupied Building, Engineering News-Record, Nov.
squeezed into the ground.
30, 1933.)
Jacks and shims, or wedge jacks,
are very useful in assembling the cutting edge. They enable the opera-
tors to make sure that the cutting edge is level.
The sand island is a very helpful device during the sinking of a caisson,
 a.

—
-a
Z

— "D

.£ %

u c

O iu
Z1 £
*. o

o .

3 .= .O
o u

£ a Z
<" "• (U

O • 2

535
#
536 CAISSONS [Chap. 12

especially where shallow water, tides, floods, waves, current, or deep

muddy bottom makes the work hazardous or difficult otherwise. An
may be built above high water so that the
island caisson can be started in
the dry, and a good working space can be available for future operations.
The caisson will be thus embedded and under control at all times.
A more spectacular sand island is pictured in Fig. 12-16. The shell for
such an island may consist of a circular ring of steel sheet piling or steel
plates with a sand fill placed inside it and resting on the river bottom.
This makes a temporary island with its top above water. The caisson
can be erected on this ground which constitutes an excellent working
platform. Then it can be sunk down through the fill and the soil below.
The great mass of the island prevents disturbance. Once the caisson is

mms^^^
(«;-W0RKING PLATFORM ON PILES
~ 2oS.B°°Hardpan and boulders

C6J-YJ0RK CONDUCTED FROM BARGES

FIG. 12-17. Two schemes for conduct of work of starting light caissons

well started and "aimed" correctly, it can be controlled more easily for
the remainder of its course. After the caisson is in its final position and
the bridge pier is finished, the piling and sand can be removed.
Guiding an open caisson during the early stages of its sinking may be
difficult if it is erected in fairly deep water or in shallow water over deep

mud or silt. Figure 12- 17 (a) illustrates the use of a pile platform around
a light steel or wooden caisson. Temporary beams a are anchored to the
platform and cantilevered to support the cutting edge and the first few
lifts of the shell. When enough of the caisson is erected, it is lifted, the
beams are removed, the shell is lowered to rest on the bottom, more lifts
are added, and some of the concrete fill may be placed. Guides are then
connected to the platform. They, through the strength of the batter
piles, are supposed to hold the upper part of the caisson in line until it is
Art. 12-6] PROCEDURES AND PROBLEMS 537

3"xJ"x A"L 1

6-10"*16\24' bent towers

S Trestle
stringers

3 protection bents /5'x 20 'protection bent

12' square

Steel caisson
Stop log guides
\5"3 @
6.7 lb.
Still water basin used on all columns
on some bents
PLAN OF TRESTLE AT A PIER
10"* 16" stringers ^ p
2-5" B-

4"air hose -

El. 72-

^vK5^^^Xv V^Avj*j55 l

END ELEVATION TRESTLE

SECTION A-A Details of trestle bent tower

FIG. 12-18. Details of preframed steel tower bent trestle construction used on Caughnawaga Bridge.

(See George L Freeman, River Piers Built with Land Plant from Trestle of Unique Design, Engineering
News-Record, May 17, 1934.)

deeply embedded so that the soil will steady it. It is obvious that excava-
tion, and especially jetting, may disturb the support for the inside row
of piles. Furthermore, the piling cannot resist the force caused by any
serious tipping.
Another case is shown The water here is deep enough
in Fig. 12-17(6).
to permit the use of floating equipment. The lower portion of the caisson,
if not too large and heavy, may be erected directly upon an anchored
538 CAISSONS [Chap. 12

barge, or this may be done at some dock and the barge may be towed to
the site.Derrick barges may then be anchored alongside. The derricks
can lift the caisson shell so that the supporting barge can be pulled out,
and they can then lower the shell until it rests upon the mud and silt. It
is obvious that the site for such work should be well protected from large

waves and that good anchors are needed.

PLAN OF CRIB, FRAME AND ANCHORAGES

IgCba/n, 7-Part steel coble fackle

1-Part steel cable tackle.

SECTION A-A

FIG. 12-19. Anchorages and guides for sinking caissons of Carquinez Strait Bridge. (See George J.

Calder, Deep Water Caisson Sinking, Carquinez Strait Bridge, Engineering News-Record, Mar. 24, 1 927.)

Another way to handle the field work is shown in Fig. 12-18. A series
of prefabricated steel tower bents with hollow posts was made so that
timber spuds could be driven through them, and the steelwork could be
supported by these piles. A trestle was thus built out to the site of the
pier, and a working platform was constructed as indicated. Land equip-
ment could be used. This was a great advantage because of the current,
storms, and floating ice. Fenders and cribs protected each caisson against
the current. The working chambers of these pneumatic caissons were
assembled on timbers spanning transversely across the sites for the piers.
 Art. 12-6] PROCEDURES AND PROBLEMS 539

When built to a height of 14 ft., they were lifted, the timbers removed,
and the caissons lowered into the water.
In eases like that of Fig. 12-19, it is necessary to depend upon ships'
anchors or some other kind of anchors to hold a caisson in position. Six
30-ton structural-steel guide columns 120 ft. long were driven as shown.
After five of these were in place, the caisson was floated into position.
The column was then driven. Steel girders, tie rods, and timber
last
trusses were attached to the columns to form an enclosing frame. Four
1-in. steel cables were attached to yokes on the respective columns and to
heavy anchored chains on barges to keep their ends out of water. The
sheaves permitted the lines to be tightened as necessary. The caisson
was then sunk within this framework.
Once a caisson gets out of line, correcting the tilt is likely to be a very
serious problem. To illustrate some of the troubles, refer to Fig. 12-20.
Jet pipe or hose

\ Frame for
Barge ropes

foJ-CONDUCTlNG EXCAVATION fW-CAISSON IN TILTED rcJ-CAISSON BEING RIGHTED

THROUGH CELLS POSITION

FIG. 12-20. Illustration of difficulties of righting an open caisson that has tilted.

Sketch (a) is greatly simplified to show one case schematically. Sup-

posedly, the excavation in cells B and C was considerably ahead of .4.

This lack of equal support then caused the caisson to tilt to the right as
pictured in (b). The top edge D moved away from its guide, whereas E
pushed its guide along, slackened the guy, and became submerged.
The first remedial step probably would be the addition of the next lift
in order to prevent submergence, although this may not be strictly neces-
sary. Sketch (6) shows the difficulty in trying to excavate cell A by clam-
shell buckets in order to undermine that edge. The bucket will reach
only the space next to the partition. Some possible procedures are shown
in (c). The soil atA may be loosened by jetting under that edge. The
ground outside at F is removed in order to reduce the passive pressure
against counterclockwise rotation back to the vertical. Cables attached
to anchors or to shore may be able to pull the top over as the undermining
proceeds.
540 CAISSONS [Chap. 12

An
unusually tough problem of correcting a tilted caisson is illustrated
in Fig. 12-21. Uplift by a barge, weighted booms, pull on cables, hung
weights, and undermining by an ingenious bucket system eventually
succeeded in righting the caisson so that it was finally only 12 ft. away
from its intended position. This illustrates foundation problems that
fully tax the ingenuity of engineers.

Dredged by c/amskel!-,
filled by dragftne— i<

Dra0jfne bucket
'Dredge anchor block

FIG. 12-21. Methods used in righting a large caisson. Weighted booms and pulling tackles anchored to
sunken cribs were rigged to apply a pull of 1,500 tons at top of caisson. (See Glenn B. Woodruff,
An Overturned 19,000-ton Caisson Successfully Salvaged, Engineering News-Record, Feb. 12, 1931.)

PROBLEMS
The following problems are for practice in the planning of construction proce-
dures for the building of deep foundations. It may be that some type of caisson
is desirable in all these problems, but not necessarily so. If a caisson is to be used,
what what general dimensions, and what procedures are the most desirable?
type,
All bridge piers have considerable transverse and longitudinal overturning
moments at their bases because of wind forces. This should be borne in mind
when planning the foundations.
12-1. Figure 12-22 shows the soil profile at the site of a tall bridge pier. The
section is taken parallel to the bridge axis. The pier load at the base of the shaft
 PROBLEMS 541

^t ofpier
fMHW=E/.200

Lu & ofpier
Pier shaft 10'x 40'
{MLW=Ef.m I

i i

Caisson'30x70

Wafer

Sand and
'
. ° o +129 gravel
Coarse
113 s cin ^

w
Rock
FIG. 12-22. Site for a bridge pier. FIG. 12-23. Soil profile at site of a large bridge
pier.

4 Ofpier
;MfiW=E/.300
sMLW=E/.2?4 35i0"c.c.
IO'xIO'-
Piershaff
v4\
4<>W>V>

•/:<:-l-:;':"l/r-.''.' r 7(?
3"/'//
~
$and
I .'-'', V
ILfilL

Sand
Fine 24
173
S®nd
Grave/ and °a »°o°oyo 12 Boulders
° ° o O 0_X-g-°k- .u
°Vo°o° °jA heavy bou/ders .

<£ ofpier
'f
FIG. 12-24. Soil profile at site of a bridge pier. FIG. 12-25. Soil profile at bridge pier.
542 CAISSONS IChap. 12

^ ofpier
{HW=E/. 220

tLW=E/.ZOO

Woifer
160

Sift and
c/ay
125

•dS6__Si/fy clay
'
~V.- ':.
22 . 1 139 Sand, some c/ay

Very f/'ne
sand
°
\
oo tiardpan arntgravei
Deep
coarse sand
FIG. 12-26. Soil profile at site of a large bridge FIG. 12-27. Soil profile at part of site for large,
pier. multistory building.

20-0"^ 2Q'-0"spaces

fFloorline.EUOO
'""Loairi'^\
Sandy-^r\
n (Water ta ble, EL 92 ^4: ofpier
J
n
day..;^ ii \ / HW=E/.436
ii

:<--,,
/ LW=E/A20
ii
' II

."I-. -.'.•|I<K
•
\% .407
:.•::- :::; i * : . : 0.

1
*> a'.-.o
-.ex- -.a-vd; Gravel and boulders,
1-5 !!
Deep II greatly disturbed
';.
fine !l

sand

FIG. 1 2-28. A deep pit at an industrial plant. FIG, 1 2-29. Soil profile at site of pier in swift river.
 PROBLEMS 543

is approximately 3,200 kips. Tidal current is slight. Wave action is not impor-
tant. Plan the foundation.
12-2. It has already been determined that a large caisson is to be investigated
for the construction of the pier to be built on the rock shown in Fig. 12-23. Tides
and tidal currents are large, but the waves are moderate in size. The bridge is
very heavy. Plan the construction, and explain the reasons for it.
12-3. Figure 12-24 shows a longitudinal profile at the site of a bridge pier across
a protected portion of a meandering river near tidewater. The pier is to have two
shafts, each 8 ft. longitudinally and 6 ft. transversely of the bridge at mean high

water. The shafts are 30 ft. c.c. with a connecting portal strut near the top.
The vertical load at the base of each shaft is approximately 1,500 kips. What
is the best construction?
12-4. Plan the foundation for the pier indicated in Fig. 12-25 if the vertical load
on each shaft at El. 100 is 2,000 kips. Notice that this section is taken trans-
versely with respect to the bridge axis. Assume that ground water is ordinarily
at El. 95, but occasional floods may reach El. 106. The site is at the bank of a
wide sluggish inland river.
12-5. A main pier for a cantilever bridge across a navigable tidal river is to be
built on the soils shown in Fig. 12-26, which is a view transversely with respect to
the bridge axis. The pier is to have two large shafts on top of a solid base 15 ft.
wide and 45 ft. long with its top at El. 225. Assume the total vertical load at El.
225 to be 5,200 kips. Plan the construction.
12-6. A large, multistory office building is to be constructed near the water
front of a large port city. The soil profile at a portion of one row of columns is
shown in Fig. 12-27. Assume that the middle column shown has a total load of
1,350 kips. it, and plan how to build it.
Design a foundation for
12-7. Figure 12-28 shows the soilsand neighboring construction near the site of
a deep pit needed for manufacturing processes at a new industrial plant. There is
no superstructure resting upon the pit. The adjacent crane columns are each to
support a vertical dead load of 120 kips and a live load of 230 kips. The pit will
have pumps in the bottom to remove leakage water, but buoyancy must be
resisted. How can the pit be constructed and the adjacent columns supported?
12-8. Figure 12-29 shows a transverse soil profile at the site of a bridge pier in a
rapid northern river. The pier is to be a solid shaft 14 ft. wide and 45 ft. long at
El. 420. The total vertical load at this elevation is 6,200 kips. Plan the best
construction procedure for founding this pier upon the rock.
 13
BRIDGE PIERS

13-1. Introduction. It is the purpose of this chapter to discuss some

of the engineering problems that occur in the planning and detailed
designing of piers for bridges and viaducts. In such planning, the method
of construction — or at least one feasible method — may be the chief feature
that influences the choice of the type of structure to be used below the
surface of the ground or water. The character of the superstructure, its
dimensions, and its weight and loads are also basic influences. Of course,
the soil conditions at the site must be taken as they are and may greatly
restrict the number of alternate practicable designs. Again, there may be
special conditions that are of importance in a particular situation. Some
of these will be described also.
This planning of the foundationsis an integral part of the development

of any major bridge project. The general features of the superstructure

and of the substructure have to be worked out together in any actual case.
Careful planning in advance may save many thousands of dollars, as well
as avoid severe headaches for the engineers and owners.
In such major projects, safety of the structure is vital because life and
tremendous property values are involved. Thorough investigations of
the local conditions are absolutely essential, and so is careful study of all
probable future developments that may affect the structure and its site.
Except for certain temporary construction, the engineers who plan a
bridge should look upon it as a structure whose useful service is to be
almost eternal. Certainly, structural insecurity should not be the cause
forabandoning or replacing an important structure. Obsolescence may
develop because one can seldom anticipate all future developments.
However, wise planning may minimize the probability of such an
eventuality.
The principles involved in planning pier construction are to be illus-
trated rather than the vast number some of the last
of details, although
that may be useful will be shown Most of the problems discussed
also.
relate to bridge piers because viaducts and bridge approaches include
similar but perhaps lesser difficulties.
544
Art. 13-3] GENERAL CHOICE OF LOCATION 545

13-2. Architecture. Cooperation of architects and engineers is

extremely important in the early planning of a bridge and all its major parts
so that the completed structure may present the most satisfactory appear-
ance consistent with the service required, the local setting, the subsurface
conditions, and practicable economy. Basic proportions, harmony of
parts, consistency of style, suitability to its purpose, and obvious effi-

ciency — these are generally the things that cause a bridge to present a
pleasing aesthetic effect. If the structure as a whole is not well suited to
its site and service, minor cannot camouflage this fact.
details
The preliminary studies of the topography, subsurface conditions, and
superstructure are likely to suggest a few alternative possibilities for the
bridge. These should be presented in perspective drawings as well as in
projection so that all concerned with major decisions may judge, the
relative merits of the appearance of each suggested structure. The time
and effort given to these studies are likely to be of great benefit.
An engineer should remember that the general public sees the structure
as an entity in its surroundings, and they like it or they do not. Few are
able to perceive the engineering problems involved, especially those con-
nected with the subsurface work. Sufficient study will usually produce a
design that fulfills the requirements of architectural and engineering
suitability,and that attains reasonable economy.
The general shape of a pier above water and its architectural features
are largely dependent upon the type, size, and dimensions of the super-
structure. Thus, it often happens that a specific type and size of pier are
desired at a certain spot. It is the engineer's duty to try wholeheartedly
to devise safe and economical means of supporting this structure at the
place where it is wanted. Sometimes a careful study of the problems of
safety and construction at this spot reveals so many difficulties and the
need for such costly work, that the preferred general plan is modified.
Therefore, study and restudy may be necessary. The urgency to get
work going should not induce one to fail to respect the importance and
great value of the planning that leads to the best design.
The shapes and details of piers that may be used are almost infinite in
number. Various illustrations are given to suggest ideas so that the
reader may benefit by examining them when planning piers of his own.
13-3. General choice of location. The alignment of a bridge may,
for practical purposes, be determined by conditions beyond the structure.
For example, in a densely built-up city, it may be almost necessary to
place a bridge in line with a main traffic artery, because of the cost of real
estate and the diversion of traffic if a new alignment were used. The
foundation engineer then must meet and overcome the conditions that
exist there if he can do so within reason.
In other cases, there may be a few neighboring sites that are possi-
546 BRIDGE PIERS [Chap. 13

bilities.Such a condition may exist where a city on one side of a river

wishes to build a bridge extending one of a series of streets across the
stream to facilitate traffic to a suburb on the other side. Then studies of
each site may reveal data that render one of them preferable to any other.
When a highway or railroad crossing over a waterway or other obstruction
is to be made in rural territory, the range of possible sites may cover sev-

eral miles. Then a very general survey may enable engineers to select
one or more sites that are advantageous and should be studied in detail.
If the necessary data are not available, it is desirable first to obtain a
general idea of conditions at a preferred alignment somewhat as follows:
1. Erect a few markers on both sides along the tentatively selected
alignment.
2. By triangulation surveys determine the distances between two
fixed points on opposite sides of the waterway if accurate maps are not
available.
3. Obtain sufficient topographic data to enable the drawing of a map
of a narrow strip of terrain along the entire alignment.
4. Take photographs of such features of the site as may seem to be use-

ful to those who will study the problem away from the site.
5. Take a few soundings across the waterway and along the proposed

center line, so that a profile can be drawn.

6. Study the general geological conditions at the site as far as can be

done visually.
7. Obtain information regarding flood heights, current, ice, wave
action, traffic on the waterway, and any governmental requirements for
channel depths and clearances.
8. Obtain such detailed information regarding approach connections,

real estate, traffic estimates, and subsurface conditions as may be avail-

able and essential.
9. Take a few borings along the center line to reveal the general nature

of subaqueous soil conditions. These may be spaced at wide intervals

and at such points as seem to be obviously desirable; e.g., at edges of
channels, at pierhead and bulkhead lines, and at shoals that are likely
locations for piers and abutments.
From the preliminary maps and profiles, the first studies of the struc-
ture can be made. After the planning has progressed far enough to reveal
what seems to be the best design, and after this plan has been tentatively
approved by the responsible persons, then careful and detailed explora-
tions of the soil should be made at the proposed locations of all piers and
abutments. Sometimes, one boring and several undisturbed soil samples
from it will be sufficient if taken on the center line of each proposed sub-
structure. If the bottom is irregular, two borings at each such location
may be enough, these being about 50 ft. apart and placed symmetrically
about the center line of the bridge and along the transverse axis of the
Art. 13-3] GENERAL CHOICE OF LOCATION 547

proposed pier. In other cases, it may be advisable to make at least one

boring at each of the four corners of the pier. If the construction of a pier
may many more borings may be needed
involve sinking a caisson to rock,
in a closely spaced checkered pattern over the probable bearing area.
This may be desirable also when a pier or abutment is located in a present
or former harbor frontage in order to explore for hidden obstructions.
Such specific and adequate information not only helps to plan wisely but
is likely to produce great economies by removing some of the uncertainties
that would otherwise have to be covered by contingency allowances in a
contractor's bid on the construction work.

FIG. 13-1. Bridge over the Susquehanna River at Havre de Grace, Md. A Wichert continuous span
of 456 ft. with 332-ft. anchor arms and deck-truss approaches of Wichert continuous type. The foun-
dations of such important structures, though unseen or unnoticed by the public, may involve some of the
most difficult planning and construction of the entire project. Designing engineer, J. E. Greiner Co.,
Baltimore, Md. (Courtesy of Bethlehem Sieel Co., builders of the superstructure.)

Careful, fairly accurate studies of the most feasible sites for a large
bridge project cannot be made hastily. Relative serviceability and prac-
ticability may have to be settled by expert judgment. The determination
of relative economy generally requires the comparison of estimates based
upon reasonably complete preliminary plans. A few thousand dollars
spent in making these studies will usually save relatively large sums in the
cost of the project.
It is sometimes said that the cost of substructure work of a large bridge
project should approximately equal that of the superstructure. What is

meant is that, if the cost of a single pier will be large, it is economical to

use relatively long spans so as to minimize the number of piers. When
the foundations are deep and the superstructure is at a high level, a pier
is likely to be large, heavy, and costly in itself whether it supports a small
548 BRIDGE PIERS [Chap. 13

load or a heavy one. Local conditions generally are such as to make

certain spots the natural or almost necessary locations for at least part of
the foundation structures; others must automatically be avoided. The
rest will usually work out to produce a plan that coordinates with these
control points. Such points are the following:

1. Streets that must be cleared

2. Real estate that must be avoided
3. Real estate that should be utilized
4. Channels that must be spanned
5. Pierhead and bulkhead lines beyond which substructures must not
project

Channel Highway *2
\ clearance Highway #J
Railroad,
<t of'bridge \

r«;-PLAN ALONG BRIDGE CENTER LINE

Profile Deck Deck

grade line girders truss Main cantilever bridge, ®eck
Deck girders

Abutment

Pier numbers 9 Rock 10

(W-PROFILE ALONG BRIDGE CENTER LINE

FIG. 1 3-2. Example of planning locations of piers and abutments of a highway bridge, Roadway is to

connect to Highway No. 1 and to grade east of Highway No. 3.

6. Rights-of-way of railroads, highways, power lines, and other utili-

ties that must be avoided

7. Islands and peninsulas that are situated advantageously
8. Subsurface as well as visible rocky ridges or gravel bars that pro-
vide economical sites for piers
9. Deep subsurface valleys, troughs of muck or silt, and lenses of

weak materials that are undesirable for supporting structures

10. Points that are automatically needed to produce symmetry of the

superstructure. For example, a three-span continuous deck bridge may

have the two inner piers located just outside of the channel lines. The
pier or abutment at one end may be restricted to a certain point because
of some local condition. Therefore, the one at the other end may neces-
sarily be situated so as to provide a side span that is equal to the first one
even though the foundation conditions here are not so desirable as they
are 50 ft. beyond this point.
Art. 13-4] SCOUR 549

Of great value is the ability to see the important features that affect a
project, to devise the best plan or alternative plans quickly, to grasp a
picture of the construction problems of the entire structure, to estimate
the relative costs reliably, to weigh wisely the intangible values of one
plan vs. another, and to make wise and prompt decisions. Such ability
springs from a background of experience and study combined with com-
mon sense and excellent engineering judgment. It is not attained
overnight.
13-4. Scour. The undermining of bridge piers by the scouring action
of flowing water is one of the serious dangers to be guarded against when
one plans such foundations. It is generally difficult to predict just what
the future may bring forth in this connection. It is a subject that war-
rants careful study in each situation, because past failures of bridge piers
and abutments from scour have been all too numerous and serious.
There is much more to be learned about this phenomenon.
The required area of the waterway is not usually critical in the case of
long bridges that have just a few tall intermediate piers, at least as far as
the flow of water alone under the bridge is concerned. This is because
such a bridge will generally have a superstructure with moderately long
spans. However, the following are cases in which the provision of ade-
quate and unobstructed cross-sectional area for stream flow during floods
may be very important.
1. A bridge at such a low level that the superstructure may become

immersed if the current is appreciable and flow is hindered. A minimum

number and width of piers may then be advisable.
2. A narrow gorge containing a stream that is sometimes a raging tor-

rent. Here it may be best to use no piers at all in the stream bed.
3. A small low bridge across an ordinarily dry stream bed located in an

area where cloudbursts occur, such as in the Southwest. In such cases

the required area is indeterminate. The common use of paved sags in
lieu of bridges for many highways in that region may be the best proce-
dure, letting the stream flood over the road temporarily when a cloud-
burst occurs rather than building a bridge that tends to dam the valley.
4. A narrow tidal estuary used by boats where the accentuation of
tidal currents and the presence of obstructions may be undesirable.
5. Streams that are filled with heavy ice cakes during the spring break-
up.
6. Streams that are used for transporting logs in rafts or singly because
the piers may cause log jams.
The determination of the best type and spacing of piers is a matter to be
studied carefully for any particular site where the area of waterway is

important. Naturally, closely spaced piers tend to facilitate the jamming

of driftwood and debris during a flood, and such jams may cause a severe
550 BRfDGE PIERS [Chap. 13

FIG. 13-3A. The Hildago-Reynosa Bridge across the Rio Grande, after one pier tilted because of
scour. It was righted and supported upon new concrete piles. (Courtesy of the tone Star Cement
Corp.)

FIG. 1 3-38. Santa Ana River Bridge in California after subsidence caused by scour during the floods of
March, 1938. (Courtesy of Ralph W. Stewart.)

nozzle effect locally that speeds up the velocity and increases scour. It
is necessary to make a special study of the flood records of the stream, 1 to
secure all available evidence of past flood heights, to determine the prob-
1
Past records may not always hold. In the Naugatuck River at Waterbury,
Conn., the hurricane in August, 1955, caused an estimated flow about four times
greater than any recorded in the past
—
"a 300-year flood." In October of the same
year another hurricane caused an estimated flow double the record before 1955.
Art. 13-4] SCOUR 551

able worst annual flood conditions of the future, to get a fair idea of the
velocity of the current before and after the bridge is built, and to compare
the estimated costs of various reasonable arrangements for the structure.
With such data, the engineer must then use his best judgment in making
a decision between conservatism at high cost on the one hand and econ-
omy with possible danger on the other.
It may seem that,if a river bottom is composed of muck or silt or even

of fine sand, this is evidence that the currents at that site are too small to
cause scour. Nevertheless, these fine materials may be disturbed to con-
siderable depth, moved downstream during a flood, and then replaced by
new and similar deposits as the current decreases. It is somewhat as

2^2 diameter
cylindrical
glass graduate

Irregular
/ ripples

Line of deepest scour

FIG. 13-4. One-liter glass graduate used as model of tubular pier to reveal character and extent of
scour caused by flowing water, (a) Longitudinal section; (b) half plan. (Courtesy of C. J. Posey, State
University of Iowa, Iowa City, Iowa.)

though part of the bottom flowed like a muddy stream, especially close to
the piers. Figure 13-4 shows some experimental data on scour. One 1

should remember, too, that fine particles are transported easily and that
it may much increase of current to produce very noticeable
not require
effects in the deepening of the waterway.
One might think that a stream with a bottom of boulders and coarse
gravel would not be endangered by floods. However, the transporting
power of rapidly moving water in large volumes is tremendous. During
a flood, boulders and gravel may be disturbed to considerable depth. As
an extreme example, the gravelly bed of the Colorado River is apparently
churned up or moved along at a depth of many feet during a major flood,
1
C. J. Posey, Why Bridges Fail in Floods, Civil Engineering, February, 1949.
552 BRIDGE PIERS [Chap. 13

this apparently being the cause of the grinding action that wears away the
bedrock below and deepens the canyon.
Some qualitative data regarding the ability of flowing water to move
solid particles may be obtained as follows, referring to Fig. 13-5:
1. Assumptions:

The weight of submerged solids = 100 p.c.f. (buoyant weight = W)

The weight of water = 62.5 p.c.f.

Impulse exerted by flowing water = F =

W
— (v — 0) (13-1)

where F = force in pounds, W — weight of water impinging upon body

per second, g = 32.2 ft. per sec. 2 , and v = velocity of water in feet per
a

a/2 "c 4
v. :

u w 3
+-\-
O (U
^-OVERTURNING
OF CUBE
o
\W' £ 1

d 2 4 6 8 10 4 8 12 16 20
\d/2 (c)- SAFETY FACTOR S.F FOR (d.)- SAFETY FACTOR S.F S FOR
OVERTURNING IF t> = 3 ft/sec SLIDING OF ONE-INCH CUBE
H=W'f
^-SLIDING OF CUBE

FIG. 13-5. Transportation of rock by flowing water.

second. This assumes that the water is deflected at 90° to its former
path — more severe than the probable case.
Particle is a cube with side = d.

2. Rolling. As shown body will tip about point A.

in Fig. 13-5 (a), the
Taking moments about A, the safety factor against overturning is

S.F.o = W X d/2 = W lOOd 3 l.Qdg

(13-2)
F X d/2 ''

F '
Q2.5d 2 v 2 /g v
2

Assuming =
3 ft. per sec. as a sample case, the curve in Sketch (c) can
v

be plotted. This shows that, for a current of this magnitude, a cube

having a side of approximately 2 in. would be at the point of rolling,
safety factor =1. A 1-in. cubical piece would start to roll in a current
ofapproximately 2 ft. per sec.
3. Sliding. Figure 13-5(6) pictures the sliding conditions for a cube.
For equilibrium,
Art. 13-4] SCOUR 553

H W'f = F
Then, for this case,

10(W 3/ = 62.5dV
100t/ 3/ l.Qdgf
S.F., = (13-3)
Q2.5d 2 v 2 /g

This, compared with Eq. (13-2), shows that the body will slide before it
will roll if / is less than unity, and a value of 0.6 to 0.75 may be assumed
as a maximum.

Tests
Competent velocity at
" bottom"
By
0.025 ft. 0.050 ft. Sand Gravel Havdite
Ho -tt 4 A A A 6.0
Tu -Q- a a 5.0
Liu O 1 y
X X 4.0 '>
<_> Gilber t
CO i^ *

««K^ ^ <>
Schaffernak + 3.0 4
a5 ™Rffi33i
;-• .

£2.0 2.0
i
=3 1.5 Jv> % .

EnST X
-£
£
1.0
1
ft
x.
Jfl
tgL>jr
^1 !
° i4 A +
0.8 B> ^xtjta
1 IT
§ 0.6 3 5aI9 r* D
P 14

?0.5 1
d
L V j*

0.4
Jj

0.3
Spec/ f/'c c. &w 'it) '= 2.6i
0.2
2 0. \ £ 16 8 1.0 2 : 4 i 8 10 2C ) 3() 4 6 ) 8(J l( )0

rf = 6rain diameter, millimeters

FIG. 1 3-6. Experimental data showing relation of size of grains and velocity of current such as to cause
bed of stream to start moving. Also proposed formula to assist in predicting results. (Courtesy of

F. J. Mavis and L. M. Laushey, Carnegie Institute of Technology, Pittsburgh.)

4. Probable conditions. In most cases, the material in a stream bed is

worn considerably so that the particles are somewhat rounded. The
safety factor against rolling is then far less than the value in Eq. (13-2)
because point A of Fig. 13-5(a) is much closer to the line of action of W.
If the lever arm is reduced, the safety factor will be decreased accordingly.
The force F is probably decreased also. At least, one can see that small
rounded particles may move rather easily in even moderate currents.
In Fig. 13-6, Mavis and Laushey show the results of experiments and 1

propose a formula for "competent" velocity material in a stream about to —

move with the current. This, as they say, may "only suggest an answer."
1
F. J. Mavis and L. M. Laushey, Formula for Velocity at Beginning of Bed-load
Movement Is Reappraised, Civil Engineering, January, 1949.
554 BRIDGE PIERS [Chap. 13

Alluvial beds like that of the lower Mississippi River may scour seriously
with only a slight increase of current. Erickson reported that an increase
'

of 40 ft. in depth occurred alongside the 111- and 121-ft. sand islands
during the construction of the 660-ft. main span of the bridge over the
Mississippi River at Baton Rouge, La. This was in spite of the use of
mattress work on the bottom.
When the stream bed is composed of fine materials and when the cur-
rent occasionally has considerable velocity, this matter of the obstructing
effect of piers may be of great importance in disturbing the existing bot-
tom through scour. As a qualitative example only, assume that it is pro-
posed to build a bridge having a series of 100-ft. spans with piers 8 ft.
wide at and below the water line. The original flood velocity V is 6 ft.
per sec. Neglecting variations in depth, the clear area of waterway A' is
then 0.92 A, where A is the original area of a portion of the waterway 100
ft. wide. The effective area may be even less than this. The result will
be a rise of water level upstream until there is sufficient head to increase
the velocity under the bridge enough to maintain the required discharge
Q. Then, approximately,

Q = VA = V'A'
The new velocity

V = -jr = Q-^J = 6 5
' ft "
per SeC "

This is an increase of more than 8 per cent in velocity and may cause con-
siderable local scouring.
The necessary depth of embedment of a pier supported directly upon
ground depends upon the maximum current, not just upon the height of
flood water. One cannot specify safely a set of dimensions for these
embedments for any and all conditions. Such figures should be deter-
mined by the engineer after a thorough study of each case.
The bottom of a pier that rests upon soil rather than upon bedrock or
deep piles should be far enough below the probable scour line to ensure
safety. Yet who can be sure what distance is "far enough" ? It has been
stated 2 that, for the American Southwest, "scour extends two times flood
water depth in the channel." Figure 13-7 and Table 13-1 show data
reported by R. W. Stewart 3 after the storm of March, 1938. He states
1
E. L. Erickson, Some Measurements of Velocities and Scour at a Mississippi River
Bridge Pier, Public Roads Administration, Proceedings of the Twenty-sixth Annual
Meeting of the Highway Research Board, December, 1946.
2
F. D. Hartford, Notes on Driving Timber Piles, Engineering News-Record, Nov. 29,
1945.
3
R. W. Stewart, Safe Foundation Depths for Bridges to Protect from Scour, Civil
Engineering, June, 1939.
Art. 13-4] SCOUR 555

that the Atchison, Topeka, and Santa Fe Railway Co.'s policy is to keep

its footings 25 below stream bed, and that its record of safety is gen-
ft.

erally good. However, depths of 15 to 18 ft. in fast streams with gravel

and boulders were not sufficient.
Excavation during construction may disturb the adjacent material
sufficiently to enable the current to move it more easily and to produce a
narrow channel alongside the pier as shown in Fig. 13-8. The water may
even attack the soil under the base.
The action of water flowing past a rectangular pier is shown qualita-
tivelyand rather empirically in Fig. 13-9. The increased velocity of the
water as it swirls and flows past the pier tends to wash away particles of
the soil locally. The eddies caused in the stream tend to suck up parti-
cles from the bottom and carry them away. These particles may be
deposited again in the slack water beyond the downstream end of the
pier, but that is irrelevant. The removal of the supporting soil is the
danger. Furthermore, the scouring away of material alongside the pier
may cause shearing failure of the ground below the pier because of the
lack of embedment. A layer of riprap dumped around the pier is benefi-
cial sometimes. In other cases, scour may undermine the stones so that,
eventually, they are not where they were originally placed and they may
not offer fully the intended protection.
A very careful study should be made of the direction of currents. If
the piers are skewed with respect to the direction of flow, as pictured in
Fig. 13-10, their effective obstructing value is obviously increased from w
to some amount more nearly equal to w v The reduced clear width w 2 ,

therefore, tends to cause a larger velocity of flow and greater scouring

tendencies, especially along sides A of the piers. Considerable suction
from eddies may also exist along sides B. In severe cases, it may be wise
to use skewed spans for the superstructure and so minimize the inter-
ference of the piers with flow.
Another matter for study is the shifting of the channel in a stream that
is inclined to meander so that the channel now is not where it was 10 years
ago, and it may be in still another position 10 years hence. Winding
streams in wide flat valleys are likely to do this. Such action may seem
to be too slow to worry about. Nevertheless, the life of a bridge is
planned to be 50 years, 100 years, or more. Much may happen in such
periods. Bars that deflect the stream now may erode away, and others
may appear in new positions so that the deep channel will migrate lat-
erally. It is difficult to predict these things, but they should be guarded
against as far as practicable. The Mississippi River is an excellent
example of this action.
An important detail in the planning of piers is streamlining. If the
upstream end of a pier is made with a cutwater, as pictured in Fig.
 > 2^ _>.
JO CO CO CO CO O)
3 s 3 3 3 "5 "5
•a
O .£ ° .2 X.
° .2 .£ CL a
4) J3 -O JO
c 41 cc 4) 41 41
O «)
a
JQ
41
JC T)
~0
-D ~l\ TJ a a.
4> 3 E -o >* 41 >^ 4) >^ 41 >^ 41 >^ 41

l*
E ° JJ
D -°
O
J3 JJ 3
TJ
4)
3
c -°
u § "8 E 4)
E
1 E TJ
41
E T)
41
41
JJ
41
3
J! 8
3
o
3 ^ 3
o o
3 3 3
O
3 3
C
3
O
3 41
3 41
3 41
3 TJ
—3D £
•>

ci
i;
u TJ ci CI ci u u 1 1 E ci 1t CI E u E
lO CO CO CO CO CO CO CT CO CO CO CO CO CO U. li.

"3 "41
> >
%4_
TJ TJ TJ
41
o -a "5 D> CO c 4) 41 41 4) 41 c e
« > E > > > > > D a
._ 4) 41
c
5 -a E E a O
4> JC 4) "5
« E CO c > o > JC CO CO CO CO co li
O
2
o
5 TJ
C
JC JC
co
TJ
c c i Oc T3
c
-0
c
-0
c
O £
jc jc
«J
o
CO
>. -0
4;
a _ -0 O O a CO 41
"5
>
a 41
?
TJ TJ
tj TJ TJ tj > 3 4) -a 3 "0 -O TJ 73 TJ TJ 3 Tj" 3
c c c c ? c a c E c c c C C
o 4) _Q 3 a JJ D D JO O JO
00 to CO CO O X LJ X O CO CO CO CO CO CO CO

N. t
1 -•
-
"•
4) H- c
S .£ § CN
adj
ch O
CN 8 O
O
spans >o o CO K O N 0" CN O
-*
O CO
of to *~ co CN CN CN <n •O -<»
O CN
CO
o"
D ,— co
*CO *
p
o
o
o"
-t

u D
JS

</5
tsi

J £ if
on
s
Q.
—
a. iD

'a 4) 'a £
a X
"3i 4) 41
4 4) _4J
TJ
4) 41
« JC 1 TJ
41 u '5)
c T3 c
3 c c v ">- c
C — CI
3 _£ J S s _4J S c a CI
4)
o a a. 'a. 'a a a 41 41
s c "a
"5
jG J 4 —a
'o. 4
TJ
p ^ 3 "5
c o
3 3
O
3 '5.
a x a ci
_4)
c <j>
1/1

41
m C
D 41 JC j: JC JC JZ JC JC JC -C 4) C
3 41 co C
Q a.
>- 1 g -0 CI
c
c
3
4)
JO 4)
i— 4)
ii
a 4) 4) 4, 01 41 1 J = i
4)
CI c 4)
V, TJ 3
"5
4
"5 4) J 4) 4)
"3
a •* cj
E 41

c Cl c CI u "4 £c "4
2 c
41
JJ 4)
41
J3
E "5
o
c c
C
c c c c c c |? c
c 4)
3 4
41
JJ E a E
D
p
u O U U U U U O IS u Of if < p i- CO

> D
* s.
T3
C TJ
o' TJ
u * #
15
4) *
>» -Q CO > * cc
c c a c # CO
TJ > O
o >
"o
X E cc
> u
6
c
o
c
>
c
O
Q_ j:
c
a
6 >
a.
<

4
„
<
"c z
g 3 CO C 4 "5
o m -C
cI) u CO
c
oT > J
<
c c
c
>i u 3 41
c
u c X 4 c CO C
_o o i JC _c
c < c 0"
41
>'

to CO co c c C u >
c
C D S < c
CO
c
<s
D CO
c co
c
CL
O
TO u c a
-* z <
c co
4)
E
jc
r
| 4> 4)
„
4)
cc co cc J"
CO CO
CO
Z
o a 4) cc
O CC at t >
>' > ca
u c >'

4 ci
i

cc
ac
5 be
CC >•
i ct c at
> C 4)
ac
CO
4
'c e D < a E c
D > CI CL
c O c u c c O
E c c < * e
4
j u < < D CO
E
o
2 > 3 < c LU _c
j
i
pc D 4)
O
ci
3
; f c c c! c JC > c c c c Cc
s
U 1—
: a> a c) r c
u 3 : > CO I a. CO 5 Cf) _ 1 CO cr CO lO CO <

f, o e< CO t u- >o K cc C »— CN co T to •0
jg z
556
 SCOUR 557

+. 30
J 20
l0
c"
-2
§-10

-30 _ wras Lowest point ofstream bedprior to Mar 2, 1938

-40
Lowest point ofstream bed after Mar. 2, 1938

FIG. 1 3-7. Diagrammatic cross sections of piers and abutments that failed because of scour in the flood
during March, 1938, in California, (Courfesy of R. W. Stewart.)

sgiy/nw/-w//AWA

Ground '

^iy/iw/w/Wf 7

weakened by excavation Weakened material easily

or actually backfilled removed by current
(a) (b)

FIG. 13-8. Scour at a pier.

\*~~~ Direction
of- — -(
current

Probable regions
v.- :-
ofdeepest scour
'it>:
»:-:

9
::}

9
':'•'
3
»•'•.'.'•

?
Possible scour
j by eddies
"if,
} Possible sedimentation
beyond wide piers

faJ-POSSIBLE CHARACTER (^-POSSIBLE SCOUR AND

OF FLOW PAST A FILL NEAR A
RECTANGULAR PIER RECTANGULAR PIER
FIG. 13-9. Possible action of water flowing past a pier.

13-1 1(a), this reduces the suddenness of the deflection of flow that occurs
with a square upstream end. The increased velocity from point A along
the sides will probably cause some scouring. If the downstream end is

square, the eddies beyond B are still harmful. A cylindrical shape at

both ends, as in (6), is fairly effective when the current is not very rapid.
558 BRIDGE PIERS [Chap. 13

The most effective shape to minimize resistance to flowing water is more

like the streamlining shown in (c), but this refinement is seldom necessary.

It may seem desirable to use pile fenders or submerged sheet piling

placed upstream from the pier, as shown in Fig. 13-12. This is done with
the idea that these obstructions will prevent loss of the soil around
and under the pier itself. However, the following comments should
be noticed in connection with each of the following sketches in this
illustration

i
Direction of,
current *~H

FIG. 13-10. Flow past skewed piers.

J*"
Direction of
current
On 1e\
A
1
lis

V/
1//
(a) (b) (O
FIG. 13-11. Streamlining of piers.

(a) If there is a small group of piles as shown here, the eddies behind
the piles and the current along the sides of the pier may cause as much
scour as though the piles were not used. Furthermore, if the piles are
not close together, the speeded-up flow between them may cause severe
local scouring.
(b) The wide sheet-pile fender shown here increases the effective
obstruction to flow. The violent eddies at the upstream corners and the
rapid inflow against the sides of the pier are likely to be dangerous.
(c) In this case, the sheet piling has been carried downstream as a con-
tinuous barrier slightly beyond the pier. It may even enclose the down-
stream end. The result is that the effective obstruction to stream flow is
Art. 13-5] WAVE ACTION 559

increased from w to W\. If the piling is not close to the pier, overtopping

and eddies may remove some of the intervening soil. Hence this space
should be filled with riprap or a concrete pad when the velocity of flow is
considerable.
(d) This vertical section through the pier in (c) shows how failure may
occur even though the footing of the pier is poured against the sheet piling.
Scour that occurs along B may remove the lateral support for the piling so
that the latter will yield and cause the pier to settle. Rods across the
base to tie the piling together may resist this action only until they rust
off. probably better and cheaper to deepen the pier itself.
It is
Beach erosion may also be a problem in some cases. This applies
mostly to the design of foundations for water-front structures other than
bridges.

h/Direction of
.

/ ,
I Piles V current
/q/ooV
fOJ pD|

Sheet
'piles

(a) (d)

FIG. 13-12. Flow past piers with fenders.

Long piles under a pier founded upon fine sediments usually prevent
dangerous scouring because they necessarily penetrate to considerable
depth. However, piles are expensive. When are they necessary, and
when not? Answering this question is one of the responsibilities that the
engineer must assume.
—
In the whole problem of bridge foundations both piers and abutments
— this matter of scour is one to be treated with great respect and care.
13-5. Wave action. Wave action against large bridge piers is seldom
serious although it may be very troublesome in its effect upon cofferdams
and caissons during the construction of those piers. Pile trestles and
causeways having short spans and with superstructures near the water
level are much more vulnerable to damage from wave action during maxi-
mum floods, high tides, and hurricanes. These structures are light, and
waves that strike them may impinge against relatively large areas so that
the resultant forces may be serious. Providing a little extra height of
structure and strength of foundations may be good engineering.
560 BRIDGE PIERS [Chap. 13

The greatest exposure will probably occur when the wind is blowing at
an angle with the axis of the pier, as indicated in Fig. 13-13. The compo-
nent of the wave impact parallel to the bridge may then exert an impor-
tant longitudinal force along the bridge axis. However, the fetch —
distance or length of clear water over which the wind blows is usually —
small when the angle a in Fig. 13-13 is large because bridges are built
across the narrowest portions of waterways. Furthermore, the wave
moves along the pier instead of striking all of the side simultaneously.
The real magnitude of the impact produced by a wave striking a rigid
obstacle is not known. A breaker hitting a sea wall evidently exerts a
much larger force than does a roller hitting a pier in rather deep water.
The magnitudes of the wave forces shown in Fig. 13-14 are empirical, but
they give the engineer something to use in the comparison of designs. In

FIG. 13-13. Open water for estimating force of waves.

using these data, the tabulated forces may be reduced to about 50 per cent
for triangular or rounded ends of piers when the wind is perpendicular to

the bridge. The same reduction may be used for winds at 45° to the
superstructure, if the assumed force per linear foot is applied to the pro-

jected width of the pier normal to the wind direction.

13-6. Ice. In some places ice may have a real influence upon the
planning of bridge foundations. This may be important in the location
of piers, the magnitude of their spacing, and the detail design for their
stability.
The lateral pressure of static ice freezing upon a pond or lake is seldom
—
troublesome because it is usually balanced if it exerts any real sideward
—
pressure at all because it does not move horizontally. Thick ice frozen
to the piles of a trestle or the bents of a causeway may produce uplift when
the water rises or, more particularly, extra downward load when the water
Art. 13-6] ICE 561

level drops. However, the weakness of ice in resisting the bending caused
by bulging upward or by sagging will generally cause it to crack so that
there is little harmful effect, even though its adhesion to the piles is not
broken.
Ice in motion is another problem. Anyone who has seen the spring
Lawrence River or the breakup of ice in some other large
floods in the St.
northern stream will have considerable respect for its power. The author
has seen ice form great jams in which large cakes were heaved as though
shoved around by subaqueous giants. Finally, the jam would break up

Ar Undisturbed
\ [water /eve/

d
Bottom

Wave Pressure Diagram

Maximum pressure Total pressure,

Wind Pm, p.s.f. lb./lin. ft. of wall
ho, h, a, hi,
ft. ft. ft. ft.
m.p.h.
d = 20' d = 30' d = 50' d = 20' d = 30' d = 50'

30 10.0 9.3 5.0 1.1 1.700 2,100 2 500 , 13,800 17,000 20 200,

70 16.6 14.2 8.3 1.9 2.100 2,200 2 , 500 26,200 27 500

, 31,200

100 21.0 17.0 10.5 2.0 2,200 2,300 2 500

, 33,500 35,000 37,100

FIG. 13-14. Estimated values of maximum normal pressure caused by wave action. Assumed fetch =
100 miles. (Based on formulas given by Paul Andersen, "Substructure Analysis and Design," The Ronald
Press Company, New York, 1956.)

with a terrific rush and grinding. How can a bridge pier be designed to
resist such forces, and what are their magnitudes?
A force equal to the crushing strength of ice, even at something like
1,000 when applied to the upstream end of a deep-founded pier is
p.s.f.,

bad enough. The impact of a large floe colliding with the pier is still
worse. And the wedging effect of ice jammed between two piers, or the
river bank and a pier, may be extremely serious. Either the piers must be
kept out of the dangerous portions of the river by the use of a long span or
else some means must be devised to ensure the safety of the structure.
Fortunately, rapidly flowing rivers are generally shallow, and rock may
often be within practicable reach. A pier then is to be designed not
562 BRIDGE PIERS [Chap. 13

FIG. 13-15A. Air view of bridge over Indian River Inlet near Rehoboth Del., subjected to pressure of
drifting ice. Portion collapsed when a truck and an auto were passing over it, killing three persons
and injuring two. (Courtesy of Acme Photos.)

FIG. 13-158. The ice jam in the Des Moines River destroyed this three-span bridge near Chillicothe,
Iowa, in the spring of 1949. The water carried the center span down the river. The south (near)
span later collapsed. The north span buckled but remained fast to its piers. (World Wide Photos.)
 Art. 13-6] ICE 563

merely as a support for the vertical load of the bridge but as a rigid
streamlined mass that can withstand these other forces. Resistance to
sliding may be obtained by keying the structure to the rock; resistance to
overturning downstream, by sufficient mass and length and resistance to ;

lateral tilting by mass and width of base below w ater. r

Of course, a narrow cutwater at the upstream end of a pier is advanta-

geous, but not so much so as one might think. A starling is one of the
best devices to use for severe conditions. As indicated in Fig. 13-16(a),
the cutwater is sloped as an inclined wedge or be on a curve as in
it may
(b). The ice then tends to slide up the starling until one edge is clear out
of water and its weight is sufficient to break the cake over the "knife-
edge." The current then carries the pieces past on both sides of the pier.
Sometimes the cutting edge of a starling may be shod with a heavy steel

Water,

(ah STRAIGHT PROFILE (W-CURVED PROFILE (ch SECTION A-A .TRIANGULAR -(d)-\ZL SLIDING UP ON STARLING
OR ROUNDED

Strap Old rails Granite Anchors

Cast
Steel ^j0;--**-Welded 'anchors 77^Boiled
angle \°5"£_!- anchors
iron
bolted to
cast tugs
M' ^strap
w rods
.

Tie
,/"*§&£ anchor,
staggered

feJ-ANGLE NOSING (fh CAST NOSING f^-USE OF RAILS tf?;-ST0NE NOSING

FIG. 13-16. Details and action of starlings as icebreakers.

angle or with old rails, or it may be made of rounded stones or concrete.

What is desired is sliding and breaking, not cutting of the ice. Of course,
the position and slope of the starling are to be determined in accordance
with probable flood levels and ice conditions at the particular site con-
cerned. Plenty of length and a sufficiently flat slope for the starlings may
cost some extra money, but the owners of the structure will be grateful
when they see them function safely during some spring when the river is
on a rampage.
There may be places where ice (or log) jams are likely to occur in spite
of the starlings and piers that are as narrow at and above the water line as
it is practicable to make them. This is a difficult problem to handle.
Stone-filled cribs or concrete blocks with starlings may be located upstream
from the piers in the hope that they will break up the ice before it reaches
the bridge, or that they may cause the ice jam to occur between them
rather than at the piers.
564 BRIDGE PIERS [Chap. 13

may also be
Drifting ice in large northern lakes and in tidal estuaries
troublesome. This may not justify the use of starlings but should be
considered in the planning and design of the piers. The following illus-
trates how such ice may act.
There is a lake about 5 miles long in a north and south direction and 2
miles wide. In the spring floods, the water level rises and the ice tends to
break into large cakes before melting. These are driven by the wind like
large floes. If a northerly gale blows for a day or two at the right time,

FIG. 13-17. The Pie IX Boulevard Bridge over the Des Prairies River in Montreal, Canada. The arch
ribs are box sections of steel painted gray and sprayed with white sand. The spandrel columns are
structural steel encased in concrete, and the deck is reinforced concrete. Notice the design of the
piers to resist ice and the current of close to 5 m.p.h. The greatest water depth varies from a low of
17 ft. to about 25 ft. at flood. Two of the piers were built with compressed-air caissons; the other
three with open caissons. The bridge was built for the Province of Quebec by the Dufresne Con-
struction Co., Ltd., Montreal. S. A. Beaulne was consulting engineer, and the Dominion Bridge Co.,
Ltd., subcontractor for the steelwork. (Courtesy of the Dufresne Construction Co., Ltd.)

the ice moves toward the the wind then shifts to the south, the
south. If
ice drifts Although the motion is slow and the
back to the north shore.
ice may be somewhat "rotted," the author has seen huge cakes 12 to 18 in.
thick shoved over a shoreside highway embankment and some 50 to 100
ft. beyond it, forming heaps of ice cakes.

When a highway bridge was planned across an arm of this lake, a design
for the piers was made with this ice condition in mind. The piers were to
be founded upon piles in a mucky bottom under 30 ft. of water. A
cheaper design was finally built, consisting of slender piers on piled
footings.
One spring, some years later, it was discovered that the ice had tilted
Art. 13-6] ICE 565

one pier by wedging action or diagonal drifting so that the bridge was
endangered seriously. To remedy the situation (at a cost of many thous-
ands of dollars) extra piles were driven beyond both ends of the pier, steel
girders were erected between caps on these piles, and all were concreted

(o) The ice beginning to break up at 8:50 A.M., Mar. 27, 1947. The action of the pier nosing is shown
clearly.

iiBiiiiii

Si*"- :
*p.v:<: :
""i||fci ..»""' " KPWtj ;
. • -,

(b) General breakup of ice at 10:10 A.M. Pier No. 1 is in the foreground under the pile-up of ice.

FIG. 13-18. Piers for the Garrison Dam construction bridge across the Missouri River near Bismarcki
N.D. These are founded upon 24-in. -diameter, ^4-in. welded-steel tubular piles. They have safely
withstood the tremendous pressures caused by ice as thick as 3 ft. (Courtesy of Frank Harrison, vice
president, Missouri Valley Constructors, Inc., builders of the piers, and E. E. Howard of Howard, Needles,
Tammen, and Bergendoff, Kansas City, Mo., the designers.)

against the old pier, thus in effect holding it up "by the neck." The
bearings were then reset and the bridge was saved.
In most such cases where drifting ice is a. hazard, a pier can be made
long enough (perpendicular to the bridge) and heavy enough so that it
can withstand the pressure applied at one end. It may be difficult, how-
 566 BRIDGE PIERS [Chap. 13

ever, to resist the thrusts from wedging or from ice moving diagonally
with respect to the bridge because the piers and their footings are usually
relatively narrow in this direction. The plans for the structure may be
made purposely, by the proper use of fixed bearings, so as to have the
bridge itself act as a strut to prevent tipping of the piers in the direction
of the bridge. Some possible arrangements are shown in Figs. 13-19(c)
and Of course, the piers should be sufficiently tall and slender so
(d).

that they will deflect easily with thermal and live-load deformations of
the superstructure, as indicated in the sketches. The bearings should be
designed to resist the necessary shearing and overturning tendencies. If

the structure is arranged as in Fig. 13-19(c0, the piers are tied together by

3AL dL V 2AL ~~34L

(a)-F\XED AT ONE ABUTMENT; ROCKERS AT AT ALL POINTS EXCEPT ONE

(c)- FIXED
BOTH PIERS AND OTHER ABUTMENT ABUTMENT TO STEADY PIERS

-Hf-JLZi
T 2AL
\Exp. Fixed Exp. Exp-

^^^
(bh FIXED AT ONE PIER ROCKERS AT BOTH
, BOTH piers TO tie THEM TOGETHER;
r«r;-FixED AT
ABUTMENTS AND OTHER PIER SOME MOTION PERMITTED AT ABUTMENTS

FIG. 13-19. Illustration of some arrangements for bearings of three-span continuous bridge. Move-
ments shown for rise in temperature.

the central span, but wedging at one side might tip them until the super-
structure bumps into an abutment, causing a longitudinal pressure in the
superstructure itself.

The mechanical action of ice rising and falling with the tide may cause
spalling of the corners of concrete construction which is in the tidal range.
For example, refer to Fig. 10-41. The corners of the pile cap near the
outboard edge of the dock platform will be subject to severe abrasion.
The concrete should be made as a rectangular shaft extending from below
M.L.W. to above high tide.
When an engineer who is not a resident of the locality is studying the
ice conditions at a particular site he should consult the people who have
lived there for many years. He may also obtain data from the Weather
Bureau and local town and county officials. The position of driftwood,
scraped marks on shore and on trees, and damage to any boat docks and
bulkheads nearby will sometimes yield useful information.
The following data are given only to assist the designer by making
Art. 13-6] ICE 567

suggestions regarding magnitudes to be used for the horizontal forces

from ice pressures in rivers. Each engineer is to judge whether they
are inadequate or too severe for his own case because they are entirely
empirical
1. Crushing of ice against end of pier. Pressures per horizontal foot of
width of upstream end of pier at water line, as given in Fig. 13-20.
2. Drifting lake or sea ice moving at angle {3 with long axis of pier. Total
force P, 1,000 lb. per horizontal foot of projected width of pier at water

1234 5 67 8910
Pressure in kips per horizontal foot of width of pier
fltwaterline perpendicular to current

FIG. 13-20. Empirical values for pressure of ice against end of a pier.

Piles

Direction of\
movement \

f«J-S0LID SHAFT (6J-TN0 COLUMNS Cc)-P\LE BENT

FIG. 13-21. Pressure areas on piers caused by ice drifting diagonally in a lake or other open water.

line, as shown in Fig. 13-21. Notice that this width in sketch (a) includes
the side BC as well as end AB even though may
tend to skid off the
the ice
former. Furthermore, in (b) the width used generally should be w t

instead of 2w because of the jamming of ice cakes unless the columns are
very far apart. The same principle applies for the row of piles in (c).
3. Ice jams. Equivalent to static hydraulic head of 5 ft. uniformly
across stream applied to a length equal to the average of the span lengths
supported by the pier being considered. If this is less than twice the pres-
sures given in Fig. 13-20, use double the values given by this diagram.
568 BRIDGE PIERS [Chap. 13

4. Forces on starlings. Forces as in Fig. 13-20 for the corresponding

rounded or triangular ends times the sine of the slope angle a, Fig.
13-16(d), of the starling with the horizontal. No reductions are to be
assumed for item 3.
13-7. Choice of type of bridge and pier. From among all the
things that could be done, how
does one choose the best plan for a par-
ticular bridge? In practice, the planning of the foundations piers and —
abutments —and of the superstructure constitutes one general problem,
with each part influenced by and somewhat dependent upon the others.
First of all, the bridge must have a certain capacity and strength to
accommodate and support the desired traffic. Next, the entire structure
should be the most suitable, economical, and attractive one that the
planners can conceive. The problem to be dealt with herein is primarily
that of planning the foundations. A few of the major things that may
influence the choice of the basic features of the structure as a whole are
the following:
1. Large vertical and horizontal clearances for shipping may so affect
the planning that only certain types of high long-span structures are
feasible, such as suspension bridges, arches, or cantilever and continuous
trusses.
2. A high-level structure with uninterrupted traffic may be desired so
that the use of a few foundations and long spans is obviously desirable,
tall
such as cantilever and continuous trusses or girders, arches, or perhaps a
series of simply supported trusses.
3. Long high approaches may be so costly and undesirable that a low-
level structure with some type of movable span is preferable. Then a
series of short girder or truss spans may be the most practicable with a
vertical lift, bascule, or swing bridge as the dominant unit.
4. Deck bridges afford a better view of the scenery than do through

structures, especially for highway traffic. Furthermore, if vertical clear-

ances are not important, deck trusses, arches, or girders are generally
more economical than are through ones having the same span lengths.
This is due to economies in the superstructure as well as to the smaller
height of piers.
5. The use of steel, concrete, or wood, if the use of one of these mate-
rials is necessary or highly desirable, will automatically affect the planning
because each material is particularly adaptable to certain types of
structure.
6. Rocky terrain with sound bedrock at shallow depths may permit
the use of many types of short-span structure or of heavy long-span ones.
Soft soils and shallow depths may render light short-span girders or
trestles most desirable. Deep foundations may be individually so costly
as to compel the use of a long-span superstructure. Deep rocky ravines
Art. 13-7] CHOICE OF TYPE OF BRIDGE AND PIER 569

JT
(aJ-VERTICAL SHAFTS WITH HEAVY iW-SHAFTS BATTERED ON THREE fcJ-SHAFTS STEPPED WITH VERTICAL
PORTAL AND BASE OUTER FACES.THINNER PORTAL SECTIONS. FLUTING TO AVOID
AND BASE RIB PLANE SURFACES.TH IN PORTAL

ft

fc
o
a
&>
s ^
.t;

CQ
\f

(dJSOilD SHAFT WITH teM-TYPE WITH SOLID f/7-CANTILEVEREDTOP (^-SHAFTS PARTIALLY STEPPED
VERTICAL.STEPPED WEB, RIBS BATTERED SHAFT SLI6HTLY ON THREE SIDES
SECTIONS ON TWO SIDES BATTERED ON
FOUR SIDES

1- i ¥
(h)- CIRCULAR SHAFTS CfV-TRIPLE COLUMNS (/7-TRIPLE COLUMNS tiW-END VIEW
AND BOTTOM RIB FOR FOUR TRUSSES FOR THREE TRUSSES OF FLARED
OR GIRDERS OR GIRDERS SHAFT OR
COLUMN

FIG. 13-22. General shapes of some piers for large bridges.

may be ideal for arches. Thus, the topographic and soil conditions at a
site may practically limit one's choice of structure to two or three general
possibilities.
7. Foundations on weak soils may render the use of statically deter-

minate structures preferable to arches and continuous bridges because of

the effect upon the stresses in the superstructure if unexpectedly large
differential settlements should occur. However, one should not build
bridge foundations that are expected to be unreliable.
570 BRIDGE PIERS [Chap. 13

8. The amount of available funds may control much of what an engi-

neer can do, but this should never be permitted to result in construction
that is not safe. However, if a certain type of structure is needed,
proper planning will usually enable one to produce an attractive structure
at practically the same cost as a similar but homely one.
9. The type of traffic may freedom of choice somewhat. For
limit the
example, railroad service requires a stiff structure so that heavy steel
trusses are more advisable than a flexible suspension bridge.
10. The personal preferences of the owner, of the architect, and even of
the engineer himself may be of vast importance when a choice among
types is made.
There are many ways in which the type of superstructure to be built will
either control or suggest the type of pier construction to be used, at least
for the portionabove water or ground. In order to illustrate some of the
many and to suggest types of construction for both
varieties of bridges
superstructures and substructures, an extensive list of photographs is
included herein. The captions accompanying them are to be treated as
part of this text.
There may be situations in which the type of pier to be used is prac-
tically dictatedby soil conditions and construction procedures. The prob-
lem that is usually presented to the engineer, however, is more likely to be
stated thus: "This is what I want. How can it be supported safely?
What will it cost? How can it be built?" However, when he finds that
the proposed structure will be impracticable, uneconomical, or unsafe, it

is duty to
his make concerned and to do his
this clear to all utmost to
assist in thedevelopment of the most feasible plan.
A pier should have sufficient area at its top to receive the bearings. It
should appear strong rather than weak or flimsy. It should be capable of
supporting the lateral and longitudinal loads as well as the vertical ones.
It should be practicable and durable, requiring a minimum of mainte-
nance. And it should be attractive.
In general, a bridge with two trusses or girders is well adapted to sup-
port upon a two-column type of concrete pier; a wide bridge with three or
four deck trusses or girders may be held by a corresponding number of
pier shafts; a superstructure of closely spaced concrete or steel girders may
be placed upon a solid pier, a two-column and portal type, or a series of
columns; where mass is required, a solid shaft may be used regardless of
the number and position of the bearings.
The preceding statements undoubtedly still leave the reader uncertain
as to how to proceed in the creation of a plan. Observation of existing
structures, study and experience in bridge design and construction, ability
in the making of freehand sketches or more elaborate drawings to picture
various possibilities, and the development of some aesthetic and engineer-
Art. 13-7] CHOICE OF TYPE OF BRIDGE AND PIER 571

ing sense to judge the merits of these possibilities — all these are helpful,
and some are developed largely from practice.
Perhaps the following story is relevant. A young graduate engineer
asked a consulting engineer of vast experience in bridge construction how
the latter created a plan for a major bridge. The older engineer replied,

r«;-souD SHAFT, (b)- SOLIDSHAFT, (W-TWO COLUMNS rtfJ-TWO SEPARATE

CURVED END TRIANGULAR END AND PORTAL COLUMNS

(eh SIMPLE COLUMNS ^-COLUMNS AND PORTALS r^-CANTILEVERED PIERS

AND PORTALS FOR DOUBLE BRIDGE

End Side
fAJ- SOLID SHAFT r^-SLENDER.SOLID SHAFT O'J-CANTI LEVER AND
WITH STARLIN6 WITH ROUNDED ENDS PORTAL COMBINED

n n
End End End
CAM-TYPE WITH I7J-RIGID-FRAME tol-STEEL COL, (iiJ-STEELCOL, fd-BRACED
TRIANGULAR ENDS RIGID ROCKER TOWER
FRAME BENT

FIG. 13-23. General shapes of some piers for small bridges.

"I do not know how can tell you but, if you will come to work with me
I
for ten years, then not need to tell you."
I shall
In Figs. 13-22 and 13-23 are shown several types of pier. The drawings
are made to picture the basic characteristics of the visible portions of the
structures and to show various details rather than the methods of support
below water or ground. In general, a pier is most satisfactory if it is
572 BRIDGE PIERS [Chap. 13

simple, neat, and obviously appropriate. Perhaps this will be evident

if the reader will study the structures shown in Figs. 13-24 to 13-31.
Many times there are special problems that call for special solutions.
For example, Fig. 13-32 pictures a structure on 200 to 300 ft. of Jersey
Meadows by 8 or 10 ft. of peat. Lightness of piers was
clay overlain
desirable, was an attractive appearance in keeping with the
and so
$90,000,000 project of which the viaduct is a part. The footings are
generally continuous under four bearing pedestals constituting one pier,
and they are supported upon precast concrete piles. Longitudinal and
transverse stiffness is provided by the rigid-frame action of the girders

**4
V ,
1
'

M 1

\
/ \

y^ /"
j

FIG. 1 3-24. The Ambassador Bridge over the Detroit River at Detroit, Mich. The main span is 1 ,850 ft-

Notice the style of the anchorage and the use of light steel bents and towers to support the side spans.
Consulting architect, Smith, Hinchman, and Grylls, Detroit, Mich.; designing engineer, McClintic-
Marshall Co. (now Bethlehem Steel Co.). (Courtesy of Bethlehem Sfeel Co.)

and tapered legs. Thus the bending is resisted by the superstructure

whereas the shears are applied to the footings at the tops of the short
pedestals only so that they do not cause large overturning moments on
the pile foundations. The general arrangement of the structure is shown
in Fig. 10-26.
13-8. Forces acting upon piers. It is difficult to set up general
specifications for the forces to be applied to all bridge piers when testing
their strength and stability. This is because of the differences in local
conditions, the special characteristics of a particular structure, and the
loads to which the bridge is to be subjected.
If other regulations are lacking, the following specifications may serve
the purpose when one estimates the loads to be applied for the preliminary
Art. 13-8] FORCES ACTING UPON PIERS 573

FIG. 13-25A. Thomas A. Edison Bridge over Raritan River at Perth Amboy, N.J. The central plate-
girder span is 250 ft. Designing Engineer, N.J. State Highway Department; bridge engineer, Morris
Goodkind. (Courtesy of Befh/ehem Steel Co., fabricators and erectors of the steelwork.)

FIG. 13-258. Piers of the north approach of the St. Georges Bridge at St. Georges, Del. The maxi-
mum height is 123 ft. above top of base. When piers are so tall, the bending moments in the shafts
and portal caused by wind may be severe. The bending in a continuous footing or bottom strut
would also be very large. For such conditions the designers used the idea of having a hidden member
across the bottom that was designed to permit a small angular deformation, yet to limit the rotation
and to offer appreciable restraint. {Designed by Parsons, Brinckerhoff, Hall, and Macdonald, Con-
sulting Engineers, New York.)
574 BRIDGE PIERS [Chap. 13

FIG. 13-26. Bridge on the Pulaski Skyway between Newark and Jersey City, N.J. This is a cantilever

bridge with a 550-ft. main span and 350-ft. anchor arms. Designing engineer, S. Johanneson, N.J.
State Highway Department. (Courtesy of Bethlehem Steel Co., builders of the superstructure.)

FIG. 13-27. The Peace Bridge over the Niagara River. The piers are founded upon rock. They are
designed for a swift current having a normal velocity of 9 m.p.h. and a maximum of approximately
12 m.p.h. The normal depth of water near the center of the river is 22 ft., the maximum about 27 ft.

Notice the provision of ramps as icebreakers. The truss at the far end is over the canal. (Designed
by the Edward P. Lupfer Corp., Buffalo, N.Y., Wm. Russell Davis, Consulting Engineer.)
Art. 13-8] FORCES ACTING UPON PIERS 575

FIG. 13-28. A bascule bridge across the Shark River Inlet, Belmar, N.J. The pier bases are approxi-
mately 22 by 69 ft. and are approximately 10 ft. below the river bed, which consists of sand with
small pockets of clay. The piers were constructed in steel sheet-pile cofferdams, and they are sup-
ported upon creosoted timber piles driven into hard-packed sand. (Courtesy of Howard, Needles,
Tammen, and Bergendoff, Consulting Engineers, New York.)

FIG. 13-29. Mill Basin Bridge, Shore Parkway, Brooklyn, N.Y. The bascule span is 165 ft. Notice the
types of pier under the approach spans. Designed by Waddell and Hardesty, New York. (Cour-
tesy of Bethlehem Sieel Co., bu/7ders of the superstructure.)
FIG. 13-30. The Siickel Bridge on Route 25A over the Passaic River at Newark, N.J., built by the N.J.

State Highway Department. The piers of the lift bridge were built with pneumatic caissons to rock.
One pier base is 60 ft. below mean water; the other, 40 ft. The foundations were designed by
Morris Goodkind, bridge engineer, under Spencer Miller, Jr., Commissioner of the N.J. State Highway
Department. The superstructure was designed by Hardesty & Hanover, Consulting Engineers,
New York. (Courtesy of Hardesty & Hanover.)

FIG. 13-31. Alameda Creek Bridge and overhead in California showing central circular piers. (Cour-
tesy of the California Division of Highways.)

576
Art. 13-8] FORCES ACTING UPON PIERS 577

analysis of bridge piers and abutments. The diagrams referred to are

intended to give a general scale of values; they cannot be exact because of
the great variations in structures. In the descriptions, the word trans-
verse denotes a direction perpendicular to the longitudinal center line of
the bridge whereas longitudinal, of course, means parallel to that center
line.
1. Dead load. a. W
the estimated weight of the superstructure
s ,

applied at the bearings. This should include the horizontal component

of any unbalanced thrust caused by arch or rigid-frame action, including

FIG. 1 3-32. Close-up of the supporting structure of the steel viaduct of the New Jersey approach to the
Lincoln Tunnel at New York City. The tapered steel legs are terminated in interlocked, hinged bear-

ings. The concrete pedestals rest upon continuous concrete footings that are supported by precast
concrete piles. Horizontal forces are resisted by the frame action of the superstructure. (Courtesy of
the Port of New York Authority.)

the resultant produced by the reactions of unequal spans or structures on

adjacent sides of the pier. Figure 13-34 is for use in making preliminary
estimates of the weights of some superstructures.
b. W
p the dead load of the pier itself.
,
This should allow for buoyancy
of the submerged portion at 62.5 p.c.f. for fresh water and 64 p.c.f. for sea
water. If the water level varies greatly, the resultant weight for both low
water and flood conditions should be computed. Buoyancy should be
considered in stability computations even for submerged foundations
resting upon rock since water under pressure will probably penetrate
seams or the junction of the concrete and the rock.
578 BRIDGE PIERS [Chap. 13

45'2" /j-'5"

EI.+/J4.7
^
i
i

i
i

JS"
"k

fe
^
5

1
£
%!

MLW —
.I, i, I, i.
. .
'i' i'i'i'

^'VW
j
WW
I8 L 0"
El.- 22.

ir "1
|

K i! |H v ^ IS T ""
S HI
"' '1 IP ^ IS I If 1 R!tf\ !
flfl ^ tf [n |[! rn rfi rn iji^. i

9 iSBiifii i ii i iiiiSSiS^ pii)y)pli)li)iii

75-tf''
r

Scale
10 20 30 40 feet

iiiiiiiiiiiiiiiiiiiii
IIIIIIIITIIXIIIIIIIII
IIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIII
IIIIIIIIIIIIIIIIIIIII (Symm. abi <t

FIG. 13-33. The tall piers at the ends of the main span of the Chesapeake City Bridge, Chesapeake
City, Md. (Courtesy of Parsons, Br/nclcerhoff, Hall, and Macdonald, Consulting Engineers, New York.)

c. W
E the weight of the soil above projecting footings.
,
This should
consider buoyancy for both low- and high-water conditions. It is useful
when computing maximum bearing pressure under the pier, and when
testing for overturning.
2. Live loads, a. Rll, theoretical live-load reaction including no
allowance for impact for highway bridges but with some allowance in the
 Art. 13-8] FORCES ACTING UPON PIERS 579

case of short railroad bridges. This should include the reactions, at the
bearings, produced by the theoretical traffic loading condition that gives
the critical load for the particular step being considered in the analysis.
Just as for dead load, it should include any accompanying unbalanced

3000

2500

2000

1500

Change 'n scale

E 1000

500
/o

t\ J/
B- J /
^ -A

100 200 300 400 ) 100 200 300 400 500

Span c.c. of bearings, feet Span c.c. of bearings, feet

A- Single track, deck girder

B= Single track, half through girder
C= Double track, half through girder
D= Single track, truss
E= Double track, truss
Concrete slab and 2 sidewalks assumed Wooden ties, no bal/ast assumed
(a)- HIGHWAY BRIDGES fW-RAILROAD BRIDGES

FIG. 13-34. Dead-load end reactions for highway and railroad bridges. (For preliminary estimates

of loads on substructures only. Double the reactions to estimate total weight of bridge.)

1000 4000
A = Pier reactior.
with equalspan
1* 3000
each side,
single track
,600
A* %^ *< B= Pier reaction
A UJ> <tf^ ?s 2000
track
with equal span,
Ui* '-<%'loin ~Jr1.f each side,
5400 (double track '

cz/ ttr& 1000

q200

100 200 300 400 500 100 200 300 400

Span c.c. of bearings, feet Span c.c. of bearings, feet

raj-HI6HWAY BRIDGES WITH TWO ORDINARY SIDEWALKS fW-RAILROAD BRIDGES WITHOUT SIDEWALKS
(Modified somewhat from A.A.SiH.O. Specs.,1949 (Based in general upon E~60 loading
Some allowance for impact

FIG. 13-35. Approximate total live-load reactions of bridges for preliminary analysis of substructures.

horizontal components. Figure 13-35 is for use in making preliminary

estimates of live-load forces. The impact allowance for railroad bridges
is included in the short spans, then reduced arbitrarily to provide none for
spans of 200 ft. or more.
580 BRIDGE PIERS [Chap. 13

b. Rlo, longitudinal force caused by live loads. Its position may be

assumed to be at the bottom of the bearings. If applied at or above the
roadway or rails, the force will cause a change of the reactions at the ends
of the span plus a shear at the bearings. only
It is sufficient to include
the latter because the magnitude This force is caused by
is uncertain.
traction or braking forces and may be estimated from the data in Fig.
13-36. Of course, it must be consistent with the theoretical loading con-
dition assumed for the simultaneous value of Rll-
c. Rlt, transverse force caused by live loads. This will be neglected
for highway bridges entirely. For straight railroad bridges, the lurching
of a locomotive or cars will be assumed as 20 kips at the top of the pier,
with allowance for one track only. Centrifugal force is to be included

when a railroad track is on a curve and supported by superstructure.

^
50 250
Longitudinal thrusf
40 200 is assumed to act
(/3
1/3
Q.
~4\ at fixed bearings
Q.
.2 30
&/ '-*
150 atone end of bridge
-t-"
1/3 bf 1/3

I 100
p^
170
t—
10
/ ^-J^jTgS— (-

50
5i^

%f[\c(y

100 200 300 400 500 50 100 150 200 250

Span, feet Span, feet
r«;-F0R HIGHWAY BRIDGES (b)- FOR RAILROAD BRIDGES
FIG. 13-36. Total longitudinal live-load thrusts for analysis of substructures, (Note.- Transverse live

load for highways = 0; for railroads, 20 kips per track on one track only.)

3. Wind loads, a. HT s, by wind blowing on

transverse force caused
the superstructure. wind pressure, w w on a vertical
The intensity of ,

surface may be assumed ordinarily as 30 p.s.f. but, in the case of some

large structures and in areas where hurricanes occur, it may be advisable
to increase this 50 per cent. The total transverse force applied to the
pier may be computed as the reaction produced by ww acting upon an
area equal to approximately lj^ times the estimated area of the side
elevation of the spans on each side of the pier.When deep trusses are
30 ft. or more apart, it is safer to assume both trusses as completely
exposed except at the floor system. Although this force Hts should be
applied at the computed center of gravity of the assumed side areas of
the superstructure, magnitude is so uncertain that it is usually suffi-
its

cient to apply it at the bottom of the bearings except in the case of through
trusses. Then it is satisfactory ordinarily to estimate the center of
gravity by eye from a sketch of the side elevation of the structure. For
these computations, all spans may be assumed to be simply supported
even though this is not the fact. Some data for use in preliminary esti-
Art. 13-8] FORCES ACTING UPON PIERS 581

mates Hrs are given in Fig. 13-37. However, more careful estimates
of
should be made for large structures.
b. Hls, longitudinal force caused by wind blowing diagonally against

the superstructure. This and a corresponding transverse force, H'TS will ,

200
f /
Railroad through trusses -*/
^ 160
/ /
cl20
o /// Highway
/thiough trusses
§80
i_

is
uj
6/rders Deck trusses
40

100 200 300 400 500

Span cc. of bearings, feet

FIG. 13-37. Transverse normal wind reactions on highway and railroad bridge structures. (Note.- For
live loads, add 0.2 kip per lin. ft. for highways and 0.3 kip per lin. ft. for railroads.)

200 400 w
c.&
o _*
y
160 320 « v
Railroa Jthi-ough s'/
trusses \y
120
\y 240
Highwaythro ugh) russe l

m 160 &
I^Di?ck t r-usse s
S,
40 *^*
^S 80
6irders O w
)

100 200 300 400 500

Span cc of bearings , feet

FIG. 13-38. Transverse and longitudinal wind reactions for 45° wind on highway and railroad bridge
structures. (Nofe: For live loads, add 0.1 kip per lin. ft. for highways and 0.15 kip per lin. ft. for
railroads.)

act simultaneously. The data in Fig. 13-38 are for use in making pre-
liminary estimates of these forces. The minimum magnitude of Hls is
not zero, even with longitudinal wind. The direction of the wind that
generally causes the worst condition at the pier may be assumed to be 45°
from the normal to the bridge center line. This gives transverse and
longitudinal forces that are each equal to ^Hts except as affected by
wind forces on exposed transverse members such as trusses. For the
582 BRIDGE PIERS [Chap. 13

web members, it is assumed that these components are each approximately

equal to the full wind on their side elevations.
In the case of tall piers and bents where overturning may be a hazard,
it is desirable to increase the wind forces by 50 per cent, then to test the
pier without live load on the structure. This is to allow for hurricanes.
The safety factor against overturning in such a case may be reduced to
1.25 or thereabouts.
c. H T p, transverse wind on the pier. This may be assumed to equal
ww times the estimated exposed area of the pier in side elevation at low
water plus any secondary exposed area, as for example, the leeward
column of a two-column pier. The resultant is to be applied at the esti-
mated center of gravity of the forces concerned.
d. Hlp, longitudinal wind on the pier when the wind is blowing diag-

onally. This force and the corresponding transverse force, H'TP may be ,

estimated on the basis of the total projected area of the pier normal to the
wind direction times w w This force should then be resolved into the
.

proper components parallel and perpendicular to the bridge.

4- Special forces. Earthquake forces. These are difficult to esti-
mate and should be included where
in one's analysis only in regions
seismic shocks are probable. If on rock, the vertical
the structure is

forces from earthquake shocks may be neglected as far as piers and abut-
ments are concerned, and the horizontal forces may be assumed to equal
10 per cent of the weight of the superstructure (omitting live loads) acting
at its estimated center of gravity plus 10 per cent of the weight of the
pier (with buoyancy considered) applied at the latter's center of gravity.
The surrounding earth and water may usually be neglected since their
action is so problematical and since they are present on all sides of the
piers. If the substructure is supported upon deep soils, whether granular
or cohesive, and whether piles are used or not, the shocks from earth-
quakes are likely to be more serious since the soil may magnify the
vibrations somewhat as a dish of jelly which is shaken slightly. The
earthquake problem is one deserving careful study. In this case, it is
probably best to build a structure that is basically strong in its resistance
to earthquakes.
Other forces. Wave action, the thrust from currents, and ice pres-
sures should be included when they are important. These have been
discussed previously. They should be assumed at their critical condition,
acting at flood stage.
Collision. In navigable waters, boats might collide with piers during
storms and foggy weather. This should be prevented, not so much for
the sake of the pier as for the protection of the boats which are relatively
like eggshells compared to massive foundations. The best protection
is a fender of wooden piles a few feet on the channel side of the pier and
Art. 13-9] STABILITY COMPUTATIONS 583

curving around to protect both ends so as to deflect a ship channelward.

The resilience of the wood renders it better for this use than steel or con-
crete. The current is generally moderate in navigable channels so that
the fender will probably cause no serious scour or obstruction to flow.
Friction. When the expansion bearings under a superstructure do not
function easily, large longitudinal forces may be set up at the top of a
pier. Sliding steel bearings that are rusted badly may practically
"freeze"; round rollers and segmental rockers may become so clogged and
rusted that they cease to operate properly. The longitudinal forces
caused by thermal expansion and by live-load deformations may then be
very large unless the pier can tilt sufficiently or the superstructure can
deform to accommodate the impressed deformation. Ordinary sliding
bearings may cause frictional forces equal to 0.3 to 0.4 of the dead-load
reaction as a minimum or 0.3 to 0.4 of the live-load and dead-load reaction
as a maximum. The resistance of deep rocker bearings to movement is
problematical, but it is usually assumed to be negligible. Obviously,
therefore, the type of bearing to be used is important. When a pier is
purposely anchored to the superstructure in order to steady its top, the
pier should be sufficiently flexible to prevent harm to it and the super-
structure as the latter deforms.
All the forces listed in the preceding paragraphs may make the design
of piers seem very complicated. This is not really so. Some of the forces
may not apply to a particular case, or they may be so small compared to
others that they may be For example, the forces caused
neglected safely.
by wind and lateral live loads on a low massive pier in the rapids of a
river may be tiny compared to those of ice, current, and the dead load.
The next article will illustrate how these forces may be used.
In any case, good judgment is needed in deciding which forces are
applicable and how they should be combined. For example, a continuous
structure may be anchored to one pier or an abutment as in Figs. 13-19(a)
and (b). Then the longitudinal wind and live-load forces on that entire
unit of the superstructure must be resisted by this anchor pier or abut-
ment as intended. If fixed at one abutment as well as at the intermediate
piers, the stiffest and strongest unit will offer the most resistance, this
probably being the abutment, as shown in Fig 13- 19(c). If, on the other
hand, the structure is fixed at two piers as in (d), these should share the
resistance and be able to hold the structure without dependence upon
bumping at the abutment.
13-9. Stability computations. A pier must be strong enough in
itself to hold up the superstructure without collapsing. This is seldom a
serious problem because most piers that are reasonably proportioned are
relatively massive so that they act as blocks rather than as slender col-
umns. Naturally, the soils under a pier must also be able to resist the
584 BRIDGE PIERS [Chap. 13

vertical and all other loads without tilting, sliding, or harmful settlement.
This is where the difficulties usually arise in the planning of piers.
Of course, the bearing area of the soil at the base of a pier must support
the dead load of the structure always. For design purposes, the vertical
live loads should be included.
The wind and other lateral loads with vertical live loads and
inclusion of
dead loads introduces the question of the probability of these maximum
combinations. The figures in Table 13-2 are believed to be suitable for
the allowable maximum computed bearing pressure on the soil, in terms of
the safe allowable bearing value p, when the theoretical loading conditions
are combined.

TABLE 13-2. Allowable Pressures on Soils for Various Combinations

of Loading

Dead load p'us vertical live load (including impact when necessary) p
Dead load plus wind p
Dead load plus wind plus ice and waves 1 .
5p
Dead load plus live load plus wind 1 . 33p
Dead load plus live load plus horizontal live load 1 .33p
Dead load plus live load plus wind plus horizontal live load 1 .5p

The pressure from current is to be included with any of the preceding

loads when it is important.
In the estimating of the long-term settlement of bridges on cohesive
soils, the full live load and wind load are so problematical and temporary
that their effects may be very small or negligible. In general, the dead
load only will be considered in the case of bridge foundations. In such
computations, wind, horizontal live loads, and all others of temporary
duration will be neglected also
The determination of the depth to which the substructure is to extend
is a very important starting point that should be based upon careful
explorations of the site, relative economy, safety, and the problems of
construction. The allowable bearing pressure p should be selected as a
starting point. This important decision is often difficult to make. This
matter has been discussed in Art. 4-8.
Practical problems are so varied that each must be studied by itself.
The method of attack is illustrated by the following examples. One of
the points to be remembered particularly is the fact that there are usually
several combinations of loading to be investigated for each problem. It
is assumed here that the general size and shape of each pier have been

determined and that its safety is to be tested. The analyses are abbre-
viated somewhat in order to save space. The general specifications for
loading set forth in the preceding article are to be used if no other data are
given.
Art. 13-9] STABILITY COMPUTATIONS 585

Example 1. Plan the general features of the base of proposed pier B

of Fig. 13-39. Then test it for maximum soil pressure and stability.
The maximum edge pressure p' on the soil is not to exceed 1.5 times the
allowable average pressure p for the critical combination of loads.
The ultimate strength of the sandy clay will be assumed to equal 14 k.s.f.
In general, the safety factor against overturning is to be 1.5 for the
critical combination of loads. The soil profile and the assumed bearing
values of the different strata are given in Sketch (c). The values for p
4% up
TTExp. Exp.
Fixed t El. 277
Exp
100' 150' 100'

(a)-SlDl ELEVATION OF APPROACH

(cj-PARTIAL SECTION A-A

(rfJ-LONGITUDINAL
SOIL PROFILE

FIG. 13-39. Site and general arrangement for a pier in an approach to a large bridge.

are those of the superimposed unit load that may be added to the soils
in excess of the pressures to which they are already subjected by the
overburden.
This is a preliminary design to establish the general features of the
structure. A check analysis is assumed to be necessary after the super-
structure has been completely designed.
This pier is part of the approach to a high-level bridge across a river
near the Atlantic Coast. A main highway passing at a skew between
piers B and C necessitates a longer span than the typical 100-ft. ones if
586 BRIDGE PIERS [Chap. 13

the piers are to be normal to the axis of the bridge, which is desirable for
simplicity of the superstructure. The piers above ground are all to look
alike. The steelwork from A to D is to be the same depth, and continuity
is to be used in this three-span unit.
The superstructure is to be "anchored" to pier B. Therefore any
longitudinal forces acting upon this 350-ft. section of superstructure must
be resisted by this pier. The expansion bearings at A C, and D are to be
,

W= RWLL==440
k
780 k
s
1220*

26-0" I40 k
0"

4"-

U'-O"thick
portal

WT,=945 k

I
Top of
\ground.
El. 215

El. 202
20'-0" 20-0" 2Q 0"
L
V \
faASIDE ELEVATION (W-END ELEVATION

FIG. 13-40. An anchor pier for bridge approach.

deep rockers that will cause negligible frictional resistance to longitudinal

movement. Assume a wind load at 45° to the axis of the bridge.
The general type of pier and its top dimensions are shown in Figs.
13-39(6) and (c).
The weight of the pier above El. 215 is approximately as shown by P W
in Fig. 13-40(6). The estimated forces to be applied to the pier (as taken
from the data in Art. 13-8) are given in the figure. These loads include
longitudinal live loads and a 45° wind. In connection with them, notice
the following:
Art. 13-9] STABILITY COMPUTATIONS 587

1. The live load is estimated from Fig. 13-35 (a) for three lanes as
though the loaded length affecting the pier were 150 ft. on the right and
120 ft. on the left because of the continuity. This gives R LL = 220 +
200 = 440 kips. This method of adding the adjacent span lengths (plus
a little here) and considering them as simple spans to get the reaction is
satisfactory because a concentrated load has been applied at the reaction
point in making the diagram, and such a concentrated load is used in each
case separately.
2. The dead load of the superstructure is estimated from Fig. 13-34 (a)
for one span of 150 ft., plus one of 100 ft., plus a little for continuity,
giving Ws = 780 kips.
3. The longitudinal live load Rlo is estimated from Fig. 13-36(o) for

a 350-ft. span since all braking forces on the three spans must be resisted
by this pier. This gives R L o = 18 kips. The probability of this load is
questionable, but it is included for the sake of safety.
4. The wind load at 45° on the superstructure is estimated as follows

from Fig. 13-38:

a. Transverse wind on one 150-ft. and one 100-ft. span is about 13 kips
for live load plus 37 kips for the structure. Then HT s for 45° wind =
50 kips.
b. The longitudinal component HL s on the 350-ft. structure is then
approximately H LS = 50 X 350/125 = 140 kips. It is not correct to
compute this as for a single 350-ft. bridge, the exposed area of which
would be much larger.
5. The transverse wind forces on the pier are estimated as follows:

a. Assume two full end areas and compute their center of gravity.

b. Assume one full side elevation, as in Fig. 13-40(a), compute its area

and center of gravity.

c. Add these areas to find 2A, compute their combined center of
gravity as shown on the drawings, and compute the total diagonal wind
pressure on the pier as

H = ?A X 0.7 X 30 = 40,000 lb., or 40 kips

Then each component will be

HT p = 40 X 0.7 = 28 kips
HL p = 40 X 0.7 = 28 kips

Now, how may the pier be supported when the soils are as shown in
Fig. 13-39(rf)? Here are some questions to think about:
1. Is the soft clay layer suitable to support spread footings under the
pier? No. The structure is continuous, it is too heavy, the overturning
forces are large, and better materials are near at hand below it.
588 BRIDGE PIERS [Chap. 13

2. Will the top layer of fine sand spread the loads over the soft clay?
No. There is not enough sand to mean much after the footing is embed-
ded sufficiently, and the clay is dangerous.
3. Should piles be driven down into the coarse sand? This seems to be
a needless expense, and it is to be done only if nothing more practicable
can be planned.
4. Is the 5-ft. stratum of silty sand dangerous? No. It is buried
under a strong deep blanket of firm soils so that it cannot escape. It is
so far down that the unit load on it will be small, and it is so thin that the
total settlement will not be serious.
5. is it cheaper to use large foot-
Disregarding the settlement problem,
ings at about 4 or 5 below the surface and above the soft clay, or to
ft.

extend the concrete columns down to footings on the sandy clay 12 or

14 ft. below the ground? The former will require nearly twice as much
bearing area as the latter needs. The deeper smaller footings seem pref-
erable. Plan therefore to place the footing at El. 202, which seems to be
below the soft clay for the necessary width.
6. Is it better to use two separate footings, one under each shaft, or a

combined footing under both? Since this pier is an anchor pier for three
spans of the approach, it should have good resistance against tilting longi-
tudinally. It therefore seems best to have one large combined footing
because this will be more stable than two separate footings would be.
The portal will hold the tops of the two shafts together.
The weight of the soil between El. 202 and 215 will be approximately
1.3 k.s.f. If 6 k.s.f. is the permissible superimposed load which may be
placed on the sandy clay as indicated in Fig. 13-39 (d), then the gross
average pressure at El. 202 may be

p = 6 4- 1.3 = 7.3 k.s.f.

The gross maximum edge pressure may therefore be

V' = 1.5 X 6 + 1.3 = 10.3 k.s.f.

The trial dimensions of the combined footing are shown in Fig. 13-40.
The weight of concrete and earth below El. 215 is approximately W b
=
1,260 kips. Therefore,

2W = W + WP + W b = 1,220 + 945 + 1,260 = 3,425 kips

The longitudinal overturning moment above El. 202 is

ML = (H L0 + H LS )7o + H LP X 45 = 158 X 75 + 28 X 45 =
13,110 ft.-kips

Then the eccentricity at the bottom of the footing is

Art. 13-9] STABILITY COMPUTATIONS 589

_ ML 13,110 _ QQ

This is outside of the middle third. Therefore, the edge pressure may be

^ X 3(10 - 3.83) = 3,425 and p' = 9.26 k.s.f.

The overturning moment transversely above El. 202 is

M t
= H'TS X 75 + HTP X 45 = 50 X 75 + 28 X 45 = 5,010 ft.-kips

Assuming, as an approximation, that there is pressure over 3(10 — 3.83) =

18.5 ft. + of the width of the footing, the edge pressure caused by M t is

„ = Mtc =
5,010 X 20 u^ k
V I X 40 3 " J - - s - t -

K2(18.5 )

Then the maximum total edge pressure at one corner is

p max = 9.26 + 1.02 = 10.28 k.s.f. (satisfactory)

To test the safety factor against longitudinal overturning, assume first

that the edge pressure P

u cannot exceed 14 k.s.f., the ultimate strength of

the soil. Then, for a triangular pressure diagram and full load,

2W = |- X 40 X width

3,425 = 1
% X 406, and b = 12.2 ft.

Then assume that the axis of rotation is the center of gravity of this
pressure diagram, 10 — 12.2/3 = 5.9 ft. off side the center of the footing.
Therefore,

Q
S.F. = —
2W X
lir
5.9
- = 3,425
13>no
X 5.9
=
.
1-54 (safe)
, ,.

For no live load and full wind on the structure,

(780 + 945 + 1,260)5.9 _ b {S e)

140X75 + 28 + 45 ~
The proposed plan for the footing will be accepted tentatively. After
the superstructure has been designed and the actual loads are known, the
pier should be checked. Because of the twisting required to develop
pressure under the edge of the central part of this combined footing, it

may be desirable to add the wide pedestals shown by the dotted lines in
Fig. 13-40. If the bending moment in the footing between the shafts is

too large, a thick vertical rib may be added between the shafts.
590 BRIDGE PIERS [Chap. 13

This problem shows that tall slender piers used as anchor piers may
requiremore of a footing than one suspects at first.
Example 2. Assume that the pier shown in Fig. 13-41 is to be ana-
lyzed for bearing pressures and stability, for dead loads, live loads, and
horizontal traction or braking and lurching forces. This is a single-track
railroad deck-girder bridge across a swampy unnavigable arm
of a reser-
voir in northern New
York. Current, waves, and ice are negligible.
The ordinary allowable bearing value p of the sand and gravel is 4 tons

u. 4.ofpier
Span = 92-6 " ^ Span = 92-6 "c. c. ofp/ers
/Top of ret//

Cc)-?\M OF FOOTING

',c,°;:C&D
Coarse sand
with some
small gravel
taj-END ELEVATION OF PIER fW-SIDE ELEVATION OF PIER

FIG. 13-41. A pier for a deck-girder railroad bridge.

per ft.
2
For this load combination, according to Table 13-2, p =
1.33 X 8 = 10.6 k.s.f. Is the proposed structure satisfactory?
For maximum pressures on the soil, the pier has the greatest effective
weight at low water. The minimum effective weight at high water is
critical for stability computations. The dead loads are relatively light,
whereas the live loads are heavy. The smaller value of Rlo in Fig. 13-41
is for loads on one span only; the larger, for loads on both spans. For
one span loaded, Ell = 550 kips; for two, 900 kips.
The bridges have expansion bearings at their far ends and fixed bearings
at this pier. Hence, all longitudinal loads on both spans are resisted by
this pier.
Art. 13-10] BEARING DETAILS 591

The computations are as follows:

One span loaded; high water:

W = 275 + 2 X 95 + 550 = 1,015 kips

MY = 65 X 22.5 + 550 X 1.25 = 2,150 ft.-kips
Mx = 20 X 22.5 = 450 ft.-kips
20 V
^ * x14*
3

IY = 4
= 4,600 ft.

14. V
* 90
" u3
Ix = tl = 9> ,350 ft.
4

A = 14 X 20 = 280 ft.
2

1,015 2,150 X 7 450 X 10

= n . . , , ,

"2W + + ^300"
.

,
1A Lsi
,

Vc = 4,600 -
(safe)

p F = 3.6 - 3.3 - 0.5 = -0.2 k.s.f.

Neglect the slight inaccuracy of relying upon tension at this corner.

Stability safety factor for the longitudinal direction, using an assumed
fulcrum 1 ft. inside the edge of the footing:

a „
S.F. = (550 + 95)4.75 + 95 X 7.25 + 275 X 6
=
_ - ,
3.7 (safe)
, -

65*X 22.5

Two spans loaded; low water:

W= 340 + 2 X 95 + 900 = 1,430 kips

MY = 105 X 22.5 = 2,360 ft.-kips (assume equal loads on bearings)

Vc

pF
=
=
W1,430

5.1 -
,

+
3.6
2,360

-
4,600
X

0.5 =
7
+
. 450
-9^300-
1.0 k.s.f.
X 10
= _
9 2 k si
"
_ .

-
,

-
,
(safe)
, ,

Q
S.F.
_,
=
(900 + 190 + 340) X 6
= _ _
3 6
. .
safe)
.

2360 -
(

13-10. Bearing details. Knowledge of the dimensions and details of

the bearings supporting the superstructure is important in the planning of
a bridge pier. The accompanying drawings of some bridge bearings are
given for the purpose of suggesting types and possible details. The
proportions can be established when the loads and details of the super-
structure are known.
The following are a few general comments regarding bridge bearings
and pier tops:
1 Welded or cast metallic bearings for steel bridges
.

Except for small spans with light loads, special care should be used
a.

to see that the centralization of the load is maintained even when the
superstructure deflects under the action of live loads. In other words,
the pressure should not "ride one edge" of the bearing.
592 BRIDGE PIERS [Chap. 13

b. Pins in steel bearings centralize the load in general. However,

angular movements or tendencies of the end of the superstructure to
rotate slightly may cause wear on the pin and pin supports if rotational
motion occurs, a large restraining moment if friction keeps the parts from
rotating, or some combination of the two. The bearing in Fig. 13-43(/)
is wear the pin and pin holes. Deflections from live
especially likely to
loads as well as thermal changes will cause small movements repeatedly.

Rolled beam Flanged

StifMIE washer

Masonry Spring Nuf^Billet

plate ^^ weisher
Bevel Wdd '
Bottom
Beveled » . I „ •.
\ „ >
Anchor billet
Projecting
pl. bolts
maSOnr>''P'^Arfchorbolt
Side End Stpin
ffaJ-PLAIN PLATE FOR <W- BEVELED MASONRY PLATE (Cj- INTERLOCKED, (W-PIN AND GROOVED
VERY SHORT BRIDGE TO ACT AS HINGE. SHEAR MACHINED BILLETS, BILLETS JOINED BY
TAKEN BY ANCHOR BOLTS HINGED EFFECT FLANGED WASHER

L ink and tap bolts at 4

Welded girder Top billet
tor lip and hold-down
^Bolts

Sole Casting Bolts

pl- TSk ^Anchor bolts
Jffc
% Bottom
at rib
One f ype of connection
End Side >Ahchor^ between top and bottom
bolts
feM SIMPLE, FLAT-TOPPED SHOE tfJ-LOCKEO SHOE WITH HINGE EFFECT

Bolts

'VjCasf or
Cotter ^welded
€>\ pin

*% rAnchor' 4
_ ,
bolls
End Side End doits Side Cross Section
(gh PINNED BEARING WITH PIN IN SHEAR (A J-PINNED BEARING WITH BEARING FULL LENGTH OF PIN

FIG. 13-42. Some details of fixed bearings for steel bridges.

c. Pins fully supported in half-round grooves in the bearings for prac-

tically their full length are generally better than those having large shear-
ing stresses and high bearing pressures on local pin plates. Compare
Figs. 13-42(0) and (h).
d. Single rockers permit practical centralization of bearing pressures
and expansion movement at the same time. However, a single-line (or
long narrow area) bearing cannot be used for extremely large loads with-
out excessive length. Figure 13-43(t) shows a twin-rocker arrangement.
e. Roller nests under heavy expansion bearings are useful in order to

spread the load over a large area of concrete through several line bearings.
Art. 13-10] BEARING DETAILS 593

Small cylindrical rollers have a low bearing value per inch of length and
are likely to be difficult to maintain properly. Large circular rollers make
the bearings unduly large. Therefore, segmental rocker nests, as in Fig.
13-43(/e), are preferable.

Slotted
hole \ Masonry pi.
Sole pi. Mach/ned Weld
phosphor- i^iMS groove
l^^
f|

ud„*. bronze
'beveled
Anchor bo/f "^-Anchor bolt-
Anchor bolt
faJ-FLAT PLATES FOR VERY SMALL BRIDGE (b) -HINGED EFFECT WITH SLIDING PLATES fcJ- PLATES WITH
GUIDES

Welded or
\Machined, D° wel bolted tooth

Dowel
^Anchor bolt Anchor bolt
(dKAST ROCKER WITH DOWELS ('eJ-SOLID ROCKER JOOTH GUIDE (7V-PIN IN SHEAR

Plate and r
C
tap bolts, \
?sf,n9
Casting
Billet

Rocker^

Drain
^hole
'"I Anchor bolt
(#)-R0CKER CLINCHED f/ij-LINKS FOR PRESSURE AND UPLIFT Cjj-DOUBLE ROCKER AND
AT TOP CENTRALIZING PIN
Gusset

Tooth guide in
Billet ,
machined groo ve
^Pin with
full bearing

Anchor bolt,
lock washer
at top
DgnEteg
^-Anchor both
Casting
al
rockers
f

W
Bar to hold
rockers in line
0>ROCKER AND PIN IN FULL BEARING 6t; -BEARING ON SEGMENTAL ROCKERS

FIG. 13-43. Some details of expansion bearings for steel bridges

/. With large single rockers, dowels or some other means should be used
to prevent creep or twist of the rockers, and to resist transverse shearing
forces. Friction alone may resist the latter, but some mechanical device
is preferable. Bar guides are desirable to hold rocker nests in the right
594 BRIDGE PIERS [Chap. 13

alignment, and side links are needed to force all rockers to maintain
correct relative position.
g. Anchor bolts are generally used in masonry plates of expansion
bearings. They may down mechanically,
also be used to tie the structure
as in Figs. 13-43 (a) and Shouldered bolts prevent pinching when the
(6).

nuts are tightened. Flanged caps like that in Fig. 13-42 (d) will hold the
parts together. The anchor arm of a cantilever bridge may be tied down
strongly, as in Fig. 13-43(/i). This holding down is important where
floods or gales may possibly cause uplift. A short heavy viaduct super-
structure does not need such an attachment.
h. Bearings should be accessible for proper maintenance.

i. Fixed bearings should be fastened to the concrete by anchor bolts

sufficiently to ensure the prevention of sliding and tipping under the
action of any forces that are expected to be applied to the superstructure.
Anchor bolts should be of large size to allow for some corrosion.
j. Phosphor bronze or some other nonferrous material is desirable for

sliding-plate bearings because of lower friction and prevention of "freez-

ing" by corrosion. Sliding bearings should be used only for light struc-
tures because they do not usually move easily. They should also be
limited to use with short spans of 50 ft. or less in order to avoid the
tendency to cause large edge pressures.
k. Bearings should be sturdy and strong, with metal thick enough to

avoid weakness in case of considerable eventual corrosion. Cast steel,

riveted structural steel, or welded structural steel are best; cast iron is for
use in very light structures only.
I. Bearing plates over 2 in. thick should be planed to secure even bear-
ing of steel on steel, unless the shop can guarantee proper bearing without
this. Bearing surfaces of castings should be planed also. Similarly, the
proper finishing of rockers, rollers, and pins is essential. Steel plates or
castings bearing on masonry need not be planed if the bearing is grouted.
m. For light bridges, blocks made of "rubbery" plastic compounds have
been used for bearings at both ends. These are supposed to yield enough
to allow for deflections and temperature changes but to resist enough to
withstand the forces caused by wind, braking, and traction. More long-
term service tests are needed to make sure that these compounds will
keep their elasticity indefinitely.
2. Bearings of concrete structures

a. Sliding bearings under heavy concrete beams or flat slabs are not

very satisfactory, not even with short spans. If used, it is desirable to

have nonferrous bearing plates, one anchored to the pier and the other to
the superstructure. Concrete will not slide well on other concrete, not-
even when an asphalt layer separates them, because the asphalt may
harden or be squeezed out.
 '

Art. 13-10] BEARING DETAILS 595

b. Special care should be used in planning the reinforcement at fixed as

well as sliding bearings in order to prevent cracking locally because of
the tension required to move the sliding bearing when the superstructure
shortens.
c. Metallic rollers or rockers may be installed at expansion bearings if
sufficiently strong bearing plates are used to spread the localized load over
a suitable area of concrete. Precast concrete rockers, as in Fig. 13-44(cT),
may be used if the loads are not too heavy.

Premolded mastic Welded anchors

j Fixed slab
Movable slab i Beam =/.
Dowels
Steel ^*mm>-
MasticTWf^WWP't Pier '»- '«
»^i"l- »:
^«
Steel,
WKhoFjiy*phosphor bronze. ° /t:e'\
i*v , VppI r«»Jtp
asbestos.or N I RougheneJ{ Y
holts or Muntz metal
tar paper Trowelled
^'Anchor bo/ts
(a)- MASTIC JOINT FOR ^-SLIDING JOINT FOR (c)- STEEL PLATES
HEAVY SLAB BRIDGE T-BEAM OR GIRDER AND ROLLER
Solid deck
Solid deck Sotid deck
Solid deck I Dpwels Dowels,
\>
>f^i.|
y
Dowels ~+.\
•''>::i':i;y.:.o:-i.D-\

Temporary
mastic
mm
-Precast
Mastic '
W^\Sira/ght K
!
°
^ — Roughened..
"

joint
r
!
•
"

Welded', rocker,
cage ''.'long sections

Dowels in recesses
for alignment
(^-CONCRETE ROCKER (e)- HINGED (fh FIXED (g) -HINGED
END JUNCTION JUNCTION
Premolded mastic-, (2nd. pour /Stab s Open slot
I 1st \
our\,
togur R
m
.i\i embedded
Nails

Xgirder
m
i
' i

i i

i i
Girder
~4"±
Coi-

tfi)-D0UBLE WALL OR SERIES (t)-DOUBLE COLUMNS (7V-ROCKER WALL

OF DOUBLE COLUMNS OR COLUMN

FIG. 13-44. Some details of bearings for reinforced-concrete superstructures.

d. an expansion
Flexible double columns or walls are often preferable at
joint in a series of heavy concrete spans.
e. When an intermediate or end bearing is not reinforced to form a

rigid joint, high edge pressure caused by the deflection of the super-
structure may be avoided by the use of deep V cuts, strips of premolded
asphalt filler, or open slots at the edge of the bearing area.
/. Hinges in concrete construction may be made by the use of metallic

bearings. Hinges may be approximated by central or crossed reinforce-

ment and slots that prevent bearing near the edges. However, the stress
 2

596 BRIDGE PIERS [Chap. 13

condition at such a point is uncertain. Some details as applied to rigid-

frame bridges are shown in Fig. 14-20.
g. In planning any concrete bearings, special care should be taken to

prevent cracking of corners near pressure points, and tensile cracks caused
by undesired restraint.
13-11. Grouting of bearings. Tops of concrete piers can be poured
to exact elevation and made level and smooth. However, shrinkage, cost,
and practical difficulties in obtaining good workmanship with the usual
labor employed make this unsatisfactory except for minor cases. The
concrete may be poured high, bushhammered down to the proper eleva-
tion and condition, and the bearings set and leveled on a "paint coat" of
neat cement, but this is expensive work and is usually impracticable.

Temporary wedges,
point up space
u / Slab Slab or shoe
Hoops ,=,
Grout Shoe
''„
:>\^r-nBars 2 i for grout
'^Anchor Hoops
Close top with
'

L bolt
waste, wooden
\~Main reinf. plug© or ,

similar device
during concreting

(a)-yML PIER ^ALTERNATE rr;-JACK-SCREW ftf;-PIPE SLEEVE ON

DETAILS PIER DETAILS DEVICE FOR ANCHOR BOLT
DRYPACKING

FIG. 13-45. Some details of pier reinforcement, grouting, and anchor bolts.

More often the concrete of a pier is poured low, then a grout or mortar
fillerused under the bearing to bring it to the proper position. A 1 :

portland-cement mortar may be used; admixtures of various kinds may

be added to reduce shrinkage; or a "rust joint" of iron filings and sulphur
that tends to expand slightly may be used to form a hard pad. An allow-
ance of 1 3^ or 2 in. for grout is usually a minimum.
Metallic bearings and anchor bolts may be installed as explained in
connection with columns in Arts. 8-3 and 8-4. Some details are also
given in Fig. 13-45.
13-12. Details of pier tops. The tops of bridge foundations are
usually subjected to very large localized pressures under the bearings.
The concrete can generally resist such action if the pier top has sufficient
area and strength to confine the compressed region so that it cannot "go
any place." That means that the shearing and bursting tendencies pro-
duced by local compression and lateral expansion (remembering Poisson's
ratio) cannot crack the concrete. The general action of the localized
pressure is indicated in Fig. 13-46(a).
 ;

Art. 13-12] DETAILS OF PIER TOPS 597

A metallic bearing may be designed to produce a specified unit pressure

on the concrete. However, a large steel billet under a heavy rocker will
probably cause a much higher pressure under the vicinity of the line
bearing than under the outer edges because of the deflection of the steel as
a beam, unless plastic flow permits local yielding of the concrete. Fur-
thermore, the localized compression in the concrete will spread out as fast
as the concrete can distribute it. It seems that an average allowable unit
pressure of 800 to 1,000 p.s.i. under a bearing is a desirable maximum for
design purposes for any concrete of reasonably good quality. The lower
figure is recommended because the prevention of cracking of the concrete

is more a matter of avoiding local deformations than of utilizing com-

pressive strength.
The area of a pier top around a large metallic bearing should be appreci-
ably larger than the required area of the bearing itself, or of the pair of

45" A
H 6 a A iS

"5^ N O
N /' G>
.vt
>, b\
a ci a «s>
^ a
<\ 1 ^t- .^ .S: "
-

o t C
&
'~-j
<1> <s is
/•; //
Shear cracks f
^ A D L
-CJ <o
>-^£°_. M £ Kl
03

b\
Bulging J
Make a<b for line bearing Areas beyond JKLM Area beyond NO not
a -b(i) for distributed not counted as counted ordinarily
bearing immediately effective

faJ-CRACKING TENDENCY (b)-ZQL\MW (cJ-SOUD PIER fdhZGLUMN AND PORTAL

FIG. 13-46. Pier tops and positions of bearings.

bearings when two are adjacent. No inviolable ratio of the areas can be
stated. The top of the pier should be planned to provide a reasonable
area, to look well, and maintenance men a chance for a foothold
to give
when working around The shape of this area with respect
the bearings.
to that of the bearing is also to be determined with good judgment. How-
ever, if other controlling factors are absent, the arrangements illustrated
in Fig. 13-46 are suggested. The following descriptions refer to the cor-
responding sketches:
(b) This pictures a single rocker bearing on a rectangular shaft Most
such bearings are longer in a direction transverse to the bridge than in that
parallel to it. If the top of the pier is very large for architectural or other
reasons, try to have the excess area symmetrically placed with respect to
the bearing.
(c) This shows a massive rectangular pier with a single cast or welded
shoe. The minimum edge distances are again determined by judgment,
598 BRIDGE PIERS [Chap. 13

and portions outside of JKLM are classed as temporary excess. Of

course, this does not mean that the latter areas resist no load as the com-
pression spreads downward in the pier; they are assumed to have little
beneficial effect at the top.
(d) This assumes a pair of bearings on top of a two-column portal-
braced shaft. The edge distances a and b may be determined as for a
single bearing. The area of the top of the portal is not assumed to be
effective, although it might be so if designed for the support of an inter-
mediate bearing.
The preceding suggestions are given in order to prevent the placing of
bearings on unduly slender shafts or close to the edges of long narrow pier
tops where two edges will be compressed heavily whereas the areas on the
other two sides are not very effective. This may seem to require piers of
large cross-sectional area. It does, and they are usually so for other
reasons: to provide an appearance of strength and proper proportion. If
more slender piers are desired, they should be designed somewhat as
columns, with special provisions to prevent harmful cracking of edges.
Reinforcement of the top of a pier to prevent cracking from localized
loads is helpful in the avoidance of visible cracks and spalling. It is not
very effective in spreading loads. It cannot offer resistance to tensile
forces except as there is deformation —
and this may be when hair cracks
have formed. The steel is also rather ineffective as dowels to transmit
shearing stresses. The schemes of reinforcement shown in Figs. 13-45 (a)
and (b) However, they are designed primarily to hold the
are suggested.
top corners and edges in place even if tiny cracks do occur.
If the depth of a coping is four to six times its projection from the

body of the pier and if it is reinforced properly, the area of its top may be
assumed to be the top of the pier when considering the area of pier vs.
bearings. However, no overloading of the shaft should occur under the
coping.
It is generally desirable to slope thetop of a pier away from the bearings
sufficiently to ensure good drainage.
13-13. Miscellaneous items. A facing of granite masonry on that
portion of a pier in and a few feet each side of the tidal range constitutes a
good protection against deterioration caused by sea water and freezing.
However, such masonry must be held adequately. A thin veneer applied
after the concrete has set is usually a waste of money. Heavy stones
bonded and anchored to the concrete are essential. Some details of such
construction are shown in Fig. 13-16(/i).
If wooden guards or other objects are to be fastened to bridge substruc-
tures, noncorrodible bolts should be embedded in the concrete. It is
even better if through pipe sleeves are used so that long bolts can be
inserted and later removed when desirable or necessary. Plain steel or
Art. 13-13] MISCELLANEOUS ITEMS 599

galvanized bolts rust off surprisingly fast, especially in and near salt
water.
The drains from roadways should be located so that their discharge will
not cause staining of the substructure. Drainage pipes extending down
through the concrete to or near the water line may become clogged with
debris or closed by icicle action during cold weather. If such pipes are
used, they should be of large-sized cast iron (or formed holes), and they
should terminate in a projecting 45° bend well above the water line.
There should be special provisions for protection and drainage at joints
in the roadway, somewhat as shown in Fig. 14-32.
Adequate provisions for drainage and for the avoidance of unsightly
stains are also essential in such cases as the intermediate piers in a series
of spandrel-filled barrel-type concrete arches. Rain water will almost
inevitably get through the pavement and tend to collect at the piers as
though in a sump. After a time it will seep through construction joints
or the concrete itself, causing efflorescence, stains, and even spalling.
What reinforcement should be used in a concrete pier? As an example,
assume the simple construction shown in Fig. 13-41. This is just a thick
concrete wall that transmits loads in compression from the superstructure
to the ground. It is not an anchor pier that is subjected to large longi-
tudinal bending as a vertical cantilever; neither is it a vertical beam that
participates in bending caused by rigid-frame action of the structure.
Since the unit compression is also small, vertical bars are structurally
unnecessary. Shrinkage and temperature changes will not cause the pier
to "jump up" into the air; nor will loads on the structure do so.
If the pier is long, horizontal steel may be useful in preventing visible
vertical cracks resulting from chemical or thermal shrinkage because the
friction on the bottom may not let the wall shorten easily. This may lead
to the use of a few bars, perhaps an area of steel equal to 0.0025 times the
vertical cross section of the pier. Then a few vertical bars will naturally
be used to hold up the horizontal ones prior to pouring. These should be
at least No. 6 in order to have sufficient stiffness. They might be used
3 ft. on centers. A layer of bars near each face of the pier is better than
a single set at the middle. The bars may then be of some help in pre-
venting major spalling of the concrete, although spalling that goes deeper
than the usual 2 or 3 in. of cover over the steel means that the pier looks
like a wreck anyway. Good quality of workmanship and the use of dense,
rich concrete are the best means of preventing bad spalling.
In a pier like that of Fig. 13-22(0), the frame action of the portal and
strut, or rib, will ordinarily require some reinforcement in them and in the
neighboring portions of the shafts, at least near the junctions. Cantilever
action at the junction of pier shafts and footings may also make the use
of reinforcement desirable here. It is then sensible to continue some
600 BRIDGE PIERS [Chap. 13

reinforcement for the full height of the pier. In many cases, however,
this reinforcement seems to be overdone. In brief, it should be used
where it serves a useful purpose, but it should not be wasted.
When reinforcement of the junctions of heavy concrete construction,
like the portal of Fig. 13-40, with the pier shafts is planned, the steel
should be located at edges, corners, or other places where cracks may
start. Other bars scattered in between will not be very useful.

100 'girder
100 'girder
(fixed bearing)

Rocker -iSL g , El. 2H

Sfa. 113 + 40
Approx. flood level El. 207

Bedrock
Soil Profile at Center Line of Bridge

Load data:
2 girders 24 ft. c.c. ; 100 ft. long c.c. of bearings
Roadway 26 ft. ; 2 bracketed sidewalks 5ft. wide
Assume depth of structure (minus railing)- lift.
Assume top ofpier to be 6 ft. wide

FIG. 13-47. Data at a pier for a deck-girder bridge in Connecticut,

When a long pier is built on rock, it isnot free to shrink because of bond
or mechanical gripping of the rock. Horizontal steel then merely helps to
compel cracks to be tiny unnoticed hair cracks instead of localized large

ones.
So-called surface reinforcement below tide water
and splash level of sea
may actually accentuate spalling if the concrete is porous to sufficiently
permit the bars to rust badly and thus "pry" off the concrete over them.
In the approach piers of the Bayonne Bridge, which are somewhat like
those shown in Fig. 13-40, structural-steel towers were used as reinforce-
Art. 13-13] MISCELLANEOUS ITEMS 601

ment. These greatly facilitated construction because they could be

erected as units, equipment could be supported upon them, and the
placing of bars was eliminated.
It may seem economical to make pier shafts hollow in order to save con-
crete and to minimize weight. The former may not result in economy
because of the cost of the inner forms. If the hollow portion is extended
below the water line, seepage may cause water to collect in the void, then
freezing of this water may crack the pier. The hollowed portion should
stop above water and should be drained, if used at all.

If a pier or steel bent for a viaduct is alongside a railroad track, it is

desirable to plan the construction so as to resist derailment. No magni-

tude is specified for the force to be resisted, but the shape and strength of
the construction should be such that the safety of the structure is assured.
A colliding car should glance off the side of the structure, not catch upon
projections, hence a smooth streamlined contour is preferable. If the pier
consists of steel columns, these should be based upon a concrete wall 8 to
12 ft. high, or should be encased in a concrete wall up to this height. A
wreck is bad enough without having collapse of the structure add to the
disaster.
Small projections on concrete construction are undesirable because of
their tendency to crack off. Projecting corners should be beveled.
Grooves or V cuts may well be used at construction joints to conceal
them. When such markings are a part of the architectural treatment,
they should be sufficiently broad and deep to cast shadows that will
produce the desired effect. The same comment applies to recessed panels.
Considerable discussion has occurred regarding the question of whether
the sides of a pier should be battered. Of course, batter increases the
difficulties with forms, but it is intended to give the impression of strength
— increased area near the base for a structure cantilevered from the
ground. It seems appropriate for massive solid piers. It is less so for
those that are primarily independent or braced columns. The use of
steps or projections as in Figs. 13-22(c) and (g) is another method of
attaining the general effect of a reduction of area with height.
Copings are questionable features. With stone masonry they were
appropriate as a top finish and a means of protecting the top joints from
weathering. They are still advisable for such masonry and for stone-
faced concrete, whether the copings are of stone or concrete. Stone is
preferable. For concrete structures there seems to be no real or imagi-
nary need for copings. It is better to design concrete structures as
honestly made of concrete, not as an imitation of something else.
Special care is needed to obtain properly formed surfaces free of waves,
offsets, fins, honeycombing, and streaks. Unusual surface treatments
may be desired in certain cases, but a clean plain surface is usually pre-
602 BRIDGE PIERS [Chap. 13

ferred. Straight troweled edges at construction joints, or V cuts as

pictured in Fig. 13-33, often add much
appearance of a pier.
to the
A discussion of piers for movable bridges has been omitted purposely.
These are generally special piers planned to suit the particular super-
structure that each is to support. They are generally low and broad.
Their design involves the same general principles as those described for
other piers.
In general, the shape and details of a pier should be such that the
completed structure gives the observer the idea that it is simple and
functional, neat and attractive, strong, and suitable for its purpose.

PROBLEMS
13-1. Figure 13-47 pictures the soil conditions found at the site where a pier is

desired. It is in glacial territory. Ordinarily the stream has a moderate current,

/Top ofraii

Loading and General Data:

L
Trusses 28 0"c.c.
Max. DL +L L +Ion two bearings
cs at each side ofpier = 2200 k

V U>
Braking force at top ofra/i=240k
Wind load on trains:
Max. transverse =60kon pier
45" wind =30 k trans, and 30k/ongit.
Wind load on trusses:
ss (El. 240 Max. transverse = 80k
45°wind = 56 k trans, and 56 k iongif
-I U J-9' DL of bridge on each bearing=250k
Current at flood-=3mph ±
Heavy ice in spring floods
Bottoms of bearings 3-6"/ongit.
£ ofpier and 4-0" trans, of bridge

Approx. flood stage t fEI. 2/5

fApprox. low water -El 204 ±

P*fj'°f\ Heavy gravel

°o Vo? ?: and boulders
Rock

FIG. 13-48. Site for a railroad-bridge pier in a shallow, rapid river.

asshown by the gravelly bottom. Design a pier that is suitable for this structure.
Assume the loads from the data given in the illustration and by use of Figs. 13-34
to 13-38.
Suggestions: There is some drifting ice in the spring breakup but this is not
severe in Connecticut. Consider scour. Is the gravel sufficient to confine the
sand? If the pier is founded upon the sand, will it be safe?
 PROBLEMS 603

13-2. Design a pier for the railroad bridge and conditions pictured in Fig. 13-48.
13-3. Figure 13-49 shows the soil profile at the site of a pier to be built along the
Atlantic Coast. Assume the same superstructure and loads as shown in Fig.
13-48, except that there is little or no ice and current.

A£> El. 240

,£ ofpier

A4HW=EL200
lMU/\/=BU9S

El. 183

Muck and si'/f-

Ei i7i ~y~\ i
ij

Soft brown clay a 162 ^3^-

Si/ty sand m
El. 157

Medium gray -brown

j

clay with some

fine sand El I43 ^f^S
Finesand with
some clay
El. 130
is
Deep coarse sand '''.: '.:*•

(Rock = El. 70+)

FIG. 13-49. Site for a railroad-bridge pier in a tidal estuary.

Suggestions: The longitudinal loads may now become very serious. Assume
allowable bearing values of soil or loads on piles that seem to be consistent with
the soil data. Do this before designing the pier, then adhere to the original
assumptions and see what are the results.
 14
BRIDGE ABUTMENTS

14-1. Introduction. A bridge abutment may be defined in general as

the foundation that supports one terminus of the superstructure of a
bridge. Usage varies somewhat but this is not important. The dis-
tinction in character between a pierand an abutment will be obvious as
the reader proceeds with this chapter. Naturally, however, a bridge
must have a beginning and an end when considered by traffic using it,
and these will be called the abutments.
Modern bridge abutments are generally made of reinforced concrete.
Stone masonry may be used occasionally, but it is ordinarily incorporated
as a facing backed with concrete, this being the case w hen special archi-
r

tectural effects and durability of surface are desired.

Much of the information given in the preceding chapter is also appli-
cable to the planningand design of abutments, e.g., scour, bearings, forces
applied by the superstructure, and many details. These things will not
be repeated but are taken for granted.
If the type of abutment to be used at a given site is not automatically

determined by architectural considerations, the owner or the engineer

may have a personal preference based upon past experience, or upon what
has been used in a somewhat similar case. However, in any situation,
safety, economy, and the practicability of construction are always impor-
tant. Alternate designs and estimates are generally justified in the case
of large structures.
The dimensions and character of the superstructure have considerable
influence upon the type of abutment that is most suitable in a given case
so do local topography and subsoil conditions. In this chapter, therefore,
an attempt has been made to discuss the planning and design of abut-
ments somewhat in accordance with the type of superstructure in terms
of the materials used or the structural action of the bridge. The illustra-
tions may be helpful in aiding one to develop a plan that will be satis-
factory in all respects
respects.
As with other foundations, the engineer should try to visualize how his
604
 Art. 14-2] PLANNING LOCATIONS 605

structure .might fail, then try to make sure that it will not do so. An
abutment may settle unevenly because of the larger pressure caused at the
toe by the earth behind the abutment and by the heavy bridge load in
front. If the earth under it is not uniform, the abutment may tilt side-

wise. It may even slide forward into the water as a whole, or the bottom
_may skid forward while the top tilts backward. Such are the important
things to guard against, and they may be troublesome. An abutment as
a'structure is usually so large that there is little difficulty in making it
sufficiently strong and stiff of itself. Keeping it where it belongs may be
more "difficult.

(Profile grade line

(b)
FIG. 14-1. One of the features to consider in the location of an abutment is the relative cost of fill

and superstructure.

14-2. Planning locations. The determination of the locations of the

abutments of a bridge is only one of the problems incident to the general
planning of the structure. Injsome cases, an abutment may be a minor
part that merely holds the end of one span of the approach portion of the
bridge; in others, a large and costly foundation that supports one end of a
major unit of the superstructure.
It may be that an abutment is located at a point where local conditions
make it almost automatic. Again there may be considerable latitude for
the selection of its position. The following are a few of the considera-
tions that may
be important in establishing a position:
1. An abutment may be placed at the point where an approach embank-

5 ment becomes so high that the use of elevated structure is more economical
than a continuation of the embankment. This is not very definite. In
general, if the height h of the fill, as shown in Fig. 14-1 (a), is 20 ft. and if a
continuation of the embankment would cause h to increase considerably
606 BRIDGE ABUTMENTS [Chap. 14

and progressively, then it is probably desirable to consider seriously the

use of superstructure rather than a continuation of the embankment.
This, however, may depend upon the width of the roadway and the
magnitude of the live loads.
As indicated roadway of width Wi may require the fill
in Fig. 14-1(6), a
A BCD. If doubled to w 2 the fill ABEF is by no means
the width is ,

double that of A BCD, whereas the cost of the superstructure will be much
more nearly increased as w%/w\. This would indicate that higher fills may
be relatively more economical for wide roadways than for narrow ones.
On the other hand, a fill is seldom more costly when it supports heavy
loads than when it has light ones (it should be well compacted anyhow).

FIG. 14-2. A multigirder steel bridge carrying the Wilbur Cross Parkway over Whitney Avenue,
Hamden, Conn. The bridge is on a skew, and the wings of the abutment are to accommodate this.

(Courtesy of the State Highway Department of Connecticut.)

But this is not so for the superstructure. Thus, it may be more eco-
nomical to use higher embankments for railroad construction than for
highways.
2. Soft clay or silt may
be so compressible, and subsequent settlement
or lateral flow under the weight of a largefill may be so serious, that it is

advisable to extend the superstructure much farther than would be done

otherwise in order to avoid the use of a massive embankment.
3. If the layout of roadways is such that a wide flaring area is required

for side connections, even high fills may be preferable to costly com-
plicated elevated construction. Thus the abutment may almost auto-
matically be placed at the point where the roadways merge into the width
of the main bridge roadway, or at least where the flare is small.
4. A transverse highway or railroad that is to be overpassed by the new
structure may be a natural obstacle that limits the abutment to a position
Art. 14-2] PLANNING LOCATIONS 607

just beyond the landward (outer) right-of-way line if there is vertical

clearance enough to permit the superstructure to pass overhead. If the
clearance is inadequate, or if the horizontal position proves to be an
obstacle, the main structure may have to terminate beyond the water-
ward side, making it necessary to use a special flanking structure over

the road or tracks. Such a situation establishes one control point in the
planning.
5. Symmetry, or at least reasonable similarity of both sides or
approaches of the superstructure, is generally desired for the sake of
appearance. Therefore, when the abutment and piers of one approach to
a main span are located to meet special requirements, it may be desirable
to use similar positions for those of the other approach unless special local
conditions there prevent.
6. Local good soil conditions may make it important to locate a heavy
abutment so as to take advantage of them, whereas adjacent weak soils
are to be avoided as much as possible.
7. Sometimes, the topography and the presence of sound bedrock leave
almost no question as to where the abutments ought to be.
8. Horizontal and vertical clearances that must be maintained are

likely to cause the abutments to be as near the limits of the former as the
shape and type of structure permit. This is often demonstrated by rigid-
frame and other highway overpasses, such as those of Fig. 14-23. The
government's clearance diagram for the channel may limit the position
of the bridge. Then the general shape of the bridge must be fitted over
this, and the abutments placed where needed to hold the structure.
9. The possibility of damage by scour, shore erosion, ice, and floods
may cause the abutments to be located back out of danger, even though
this causes the superstructure to be longer.
10. When an abutment is to be of some particular type and to present
a certain architectural appearance, then its location may be determined
largely by the position where it will have the proper size, height, and
setting in the topography to produce the desired aesthetic effect.
There are many cases in which the best locations for the abutments of a
bridge are very uncertain. This is particularly true in the case of stream
crossings when such conditions as the following exist:
1. Wide streams with large areas of shallow water near shore.
2. Streams with extensive marshy areas along the shores.
3. Streams with low, flat, or gently sloping ground for long distances
back of the shore line.
4. Streams which frequently flood wide areas of adjacent low shores and

for which the area of waterway should not be constricted.

5. Streams which meander in soft alluvial soils and which may change

radically the locations of their channels.

608 BRIDGE ABUTMENTS [Chap. 14

Railroad bridges and low-level bridges whose approaches have little

6.

or no grade and which remain nearly parallel to the ground for consider-
able distances.
For such cases, an engineer will probably locate the abutments where
the resultant total cost of the project is a minimum, or wherever, in his

opinion, they ought to be.

14-3. Abutments for simply supported and continuous bridges.
Seats for the bearings of steel and concrete bridges are to be provided as in
the case of piers. This is one of the starting points in the planning of an
abutment. The area of the seat should be adequate, and the bearings of

FIG. 1 4-3. Three-level traffic interchange at Hudson County Boulevard West, the New Jersey approach
to the Lincoln Tunnel. The lower structure is a two-span reinforced-concrete rigid frame. The
upper one is structural steel encased in concrete. (Courtesy of the Port of New York Authority.)

steel bridges should be so placed that they can be inspected and main-
tained easily. There may be special cases like that of Fig. 14-3 where the
architect preferred that the bearings be hidden. Here the steel girders
supporting the boulevard are encased to make them appear as though
they were concrete construction, and the bearings are in pockets which
are covered by removable steel plates attached to the girders and which
slide over the tops of the concrete walls around the pockets. Another
instance of concealed bearings is shown in Fig. 14-4.
Consider first some of the ordinary types of abutments used with steel
and concrete bridges other than arches and rigid frames. The details of
construction vary widely in practical cases to suit the local conditions.
However, for convenience, various abutments will be grouped and named
in accordance with some basic principle underlying the design of each.
The drawings are simplified to show essential features only.
Art. 14-3] FOR CONTINUOUS BRIDGES 609

1. Straight-wing abutments.Figure 14-5 shows two abutments of this

type. In general, they are retaining walls modified so as to support the
superstructure. They are used with embankments of moderate height.
They are often built at stream crossings where the wings prevent the fill
from blocking the stream, and where they tend to restrict the scouring
action of water eddying around the main support. In the case of under-
passes, the wing walls may actually be replaced by long retaining walls
when necessary. This type is adaptable to use with skewed as well as
with normal bridges. Such abutments are usually massive and must

FIG. 14-4. Abutment between two steel girders of the New Jersey approach to the Lincoln Tunnel.
This shows an attractive treatment with stone facing and shielded or recessed bearings. (Courtesy
of the Port of New York Authority.)

resist large overturning moments. Therefore, they are suited best to use
-onr-sand^arKTnrm soils.

Sketch _(aX shows a. continuous projecting bridge seat. This is simple

construction. The bearings are exposed, and the backwall is a continua-
tion of the wings. With through trusses, pony trusses, through girders,
and deck trusses, pilasters may be used under the bearings only instead of
having a continuous seat. This, hojvyeyer, because of overturning, tends
to cause large local pressures under the toe of the footings if they project
locally in front of the pilasters. Even with a continuous bridge seat,
beveled or offset footings of the wing walls tend to concentrate the pres-
sure at HJ.
In Sketch
(6), the bridge seat is recessed by making the wing walls flush
with the front face of the central structure. This may be a preferred
610 BRIDGE ABUTMENTS [Chap. 14

arrangement for underpasses where the cut for the lower roadway is to be
minimized, for cases where the wings are needed to distribute heavy bridge
loads over large footing areas, and where the toe is to be in one line in
order to minimize the critical bearing pressure at the front edge.

Contraction joint Stem Footing

'

i /

Stem H es
Front face
'Bridge seat J Wing
Plan Rackwall

/Seat
A B

* Footing

Front Elevation End View

(ah PROJECTING BRIDGE SEAT

Bridge seat
Plan
-y~
' Backwall
-Seat

Sfe.

Front Elevation End View

f6J-RECESSED BRIDGE SEAT

Seat

Counterfort

Section at<fc End View Section at <t End View

(ch GRAVITY SECTION (W-COUNTERFORTED CONSTRUCTION

FIG. 14-5. Straight-wing abutments.

^
Such abutments are inclined to be long. The bridge seat AB, in Fig.
14-5 (a), is dependent upon the width of the superstructure. This may be
40 to 60 ft. or more. The corners C and D should be somewhat beyond
the limits of the shoulders of the fill— perhaps 6, 8, or 10 ft. beyond the
edge of the roadway pavement. The slope CE of a wing wall depends
Art. 14-3] FOR CONTINUOUS BRIDGES 611

upon that of the embankment. The distance EF will be determined by

frost depth, and by embedment to reach firm soil or to guard against
scour.
Figure 14-6 shows a very different design but one that, above ground,
appears to be similar to the heavy one of Fig. 14-5. The backwall and
bridge seat constitute a heavy beam spanning between buttressed piers on
piles, and the wings are its cantilevered extensions. This is a far more
economical design when supported upon weak soils. It may be used
advantageously on good soils if considerable depth is needed to reach firm
ground or to avoid danger of scour. The basic idea may well be borne in
mind in connection with other types of abutments.

Scale:j = l-0"
SECTION A-A
FIG. 14-6. An abutment on two pile-supported piers. The shallow backwall and cantilevered wings
minimize the horizontal earth pressure that the structure must resist. (Courtesy of Parsons, Brincker-
hoff, Hall, and Macdonald, New York.)

If contraction joints are not provided in such a long concrete structure,

nature will usually create them because of shrinkage and temperature
action. Therefore, joints may be advisable, as shown in Figs. 14-5(a)
and where the section changes abruptly.
(b), If the seat is over 40 ft.
1

long, a vertical joint may also be advisable at its center. In general, all
such joints should be keyed to hold the parts in line; those in the stem
should be flashed to prevent leakage and staining. When the length is
small enough and when the wings are to be used to spread bridge loads, it
may be desirable to have a joint at the center of the abutment only. In
other cases, when the abutment is high, keys may be used in the joints in
the footings but not in the stem, thus permitting the wings to deform
differently as thin retaining walls than would the massive central portion.

1
The volume of any single pour should be limited to perhaps 200 cu. yd.
612 BRIDGE ABUTMENTS [Chap. 14

Steel bridge construction will generally permit the tiny movement at

contraction joints without harm to the superstructure. In the case of
concrete slabs and T-beam construction, one should take precautions to
avoid having a contraction joint in an abutment cause a corresponding
crack in the bridge deck.
In general, abutments should be strong, rather massive, and stiff.
However, adequate horizontal reinforcement and vertical contraction -

joints are essential to prevent cracking.

2. Beveled-wing abutments. In some ways, the construction shown in
Fig. 14-7 is a minor modification of that of Fig. 14-5. However, it has
considerable difference in structural action when the superstructure is

narrow and the angle a of the wings is large, perhaps 30 to 45°. This is
because the entire abutment is of a shape that is basically stable; i.e., the

Top of
embankment

fb)- FRONT ELEVATION e^-SIDE ELEVATION

'
)

FIG. 44*7» Beveled-wing abutment.

wings tend to serve as counterforts so that the structure is stronger than

would be a straight T-shaped retaining wall. When the length requires a
contraction joint at the center, a portion like ABCDEFA still has a con-
siderable resistance to being tipped over as a unit. Of course, if contrac-
tion joints are used in the vicinity of CF, each part must then stand up by
itself as a retaining wall.
A beveled-wing abutment may be used at a crossing over a stream when
the wings are useful to deflect the stream toward the opening, provided
serious scour is not probable. Sometimes such a structure may- seem _to
fit the terrain better and to be more attractive than one with straight
wings. This type may be easily adapted for use with skewed structures.
It may reduce the fill somewhat, but this is seldom important. When the
fill may be allowed to run around the end of the wing as shown by GH in

Sketch (c), the amount of concrete required may be reduced a little. In

general, such an abutment is heavy and is most suitable on firm soils.
Art. 14-3] FOR CONTINUOUS BRIDGES 613

3. U-shaped abutments.The construction shown in Fig. 14-9 has the

wings perpendicular to the front face' They serve as effective counter-
Torts if the Bridge is not too wide. Therefore, this is an inherently stable
structure when it is designed and founded properly. The illustration
indicates what might be used for supporting a deck highway structure
composed of heavy beams or girders placed longitudinally under the
roadway.
Such an abutment is advantageous for grade-crossing eliminations in
rural areas where the sides of the cut for a lower roadway are to be sloped
to avoid the use of the costly retaining walls that are often necessary along
such cuts in urban areas where land is valuable. In other cases, they

FIG. 14-8. A beveled-wing wall at the abutment of a parkway overpass. Notice the pylon and
the curved top of the wall.

constitute a desirable terminus for the embankment of the upper road-

way because they appear strong, as well as being so. This visual effect
is often advantageous as a suitable end and support for the superstruc-
ture. This type is not so advantageous for crossings that are skewed
sharply.
When the upper roadway is on an embankment, the wings may be built
as in Fig. 14-9 (c) so that the footings are on good undisturbed soil, and the
fill allowed to slope along the outside.
is It is generally desirable to have
the overlap distance AB at the top of the embankment at least 2 ft. in
order to prevent settlement and erosion from making the end of the wall
seem to be cut too short, and in order to support the fill just beyond the
end. The distance CD beyond the toe of the slope should also be suffi-
cient to avoid the appearance of weakness and skimpiness.
614 BRIDGE ABUTMENTS [Chap. 14

When the lower roadway is in cut, the wings may have stepped footings
somewhat as pictured in Sketch (d) because it is not necessary to excavate
deep trenches back into the undisturbed ground and in them to build walls
that have to resist little difference in lateral pressure. Of course, the
bearing resistance of the soil at such corners as E may be weak, but there
will generally be enough resistance under the rearward portion of each
step to hold the structure. The embedment EF should exceed the

^Parapet

Wing

Backwall
Contr. jt.

~F> (W-SIDE ELEVATION

Seat'
(a)-P\an

±
(b)-FROm ELEVATION toW-SIDE ELEVATION WITH PYLON
AND STEPPED FOOTINGS
'Backwall s Back wall
11

mmm&*
Wing Wing

feJ-DETAILS OF REINFORCEMENT (//-OFFSET BACKWALLTHAT MIGHT

AT CORNERS BE USED WITH (c) AND (d)

FIG. 14-9. U-shaped abutment.

expected frost depth even though natural drainage may prevent the for-
mation of harmful ice lenses. The rearward portion of the wall footing
may be parallel to the slope instead of stepped, if the slope is flat enough
for the soil utilized. This, however, may require undesirable top forms
when the footings are poured.
A central contraction joint is an abutment because
desirable in such
the embankment fills the inside of the U. Such fill is assumed to be
used with this type. After the fill is placed, traffic tends to compact it
against the three sides. When the temperature drops, the concrete
Art. 14-3] FOR CONTINUOUS BRIDGES 615

(o) These abutments are supported upon a very deep fill. They are designed with a wide base to
reduce unit loads, but they are expected to settle slightly with the embankment. No attempt is

made to hold back the main fill itself.

(b)A concrete roadway slab is to be placed on this steelwork. There is to be a joint in the road-
way above the far side of the tower in the background, so that the end span can rotate about a
pin on top of the post if the abutment and fill settle.

FIG. 14-10. A steel highway bridge with steel towers as piers and with small U-shaped abutments
at the Morenci Reduction Works, Morenci, Ariz. (Courtesy of the Phelps Dodge Corp.)
616 BRIDGE ABUTMENTS [Chap. 14

cannot shrink without tending to squeeze the warmer fill inside, and the
latter's abutting power will usually exceed the tensile or shearing strength
of the abutment, as illustrated in Fig. 14-11.
If the abutment is so wide that one central joint is not sufficient, one

must be careful not to locate the joints where they destroy the mutual
counterfort action that is so beneficial. For example, clear joints through
the wings at their junctions with the front and parallel to it compel each
of the portions to act by itself as a retaining wall; shear keys in these
joints will almost surely be broken by the shrinkage action, and they do
not permit the wings to hold the front against tipping outward. If two
joints are needed, as in an abutment 60 ft. wide, they might be keyed
joints through the front or bridge-seat
portion and located about 10 ft. from the
junction with the wings. The keys can
then be used to enable the angular end
portions to help steady the middle piece.
The wings and the front should be strongly
tied together at their junction.
In some cases, concrete may be saved
by the use of an offset backwall, as shown
in Fig. 14-9 (/). In others, pilasters may
be used at the bearings only, constituting
buttresses for the front wall. Again, extra
concrete may be used deliberately in order
to attain a desired architectural effect, as
illustratedby the pylon indicated in (d).
In any event, U-shaped abutments
FIG. 14-11. A crack a large bridge
in
should be and can be made attractive in
abutment. This seems to be the result
appearance. V-cut markings, paneling,
of shrinkage and drop in temperature.
special form finishes, pilasters and pylons,
attractive balustrades or railings to reduce apparent height, horizontal
—
and fluting these may yield excellent results without much added
offsets,
expense when they are used expertly. Stone-masonry facing on monu-
mental structures may produce such attractive results that it is worth
its cost.

4. Box abutments. Outwardly, the abutment shown in Fig. 14-12 may

appear to be U-shaped because of the wings. It is, however, a partial box
resting upon the ground. The one shown here might be used under a
through plate girder or truss bridge.
The two pilasters are practically piers on spread footings. The back-
wall behind the bearing is extended down as a curtain and may be utilized
to help spread the bridge loads. Part of this wall might be omitted, but
its use is desirable in order to avoid recesses that might become filled with
Art. 14-3] FOR CONTINUOUS BRIDGES 617

debris. Sometimes, a temporary opening may be left in it for the removal

of forms. The wings are also curtain walls that may or may not have
footings. At the rear, a secondary wall and footing are supported upon
undisturbed ground near the top of the slope, or they may be placed on
piers as shown in Sketch (c) by dashed lines when the upper roadway is on
embankment. The roadway deck is an integral part of the abutment.
It and the wings span from the front to the secondary wall. When the
bridge is narrow, the deck and secondary wall might span crosswise
between well-supported wing walls.

\X Joint painted with asphalt

n :
[
f" Floating "slab

!l

Jv.l°j Line of offset

-k
fcli& yjoint in wall
Curtain , Joint painted
wall and r with asphalt
wing
i
Q
mm
A Footings ^-*- lZj
-Line ofjoint
in wall
(dJ-SOME POSSIBLE DETAILS
OF CONTRACTION JOINTS
ftt)-PLAN IN FLOOR IF NEEDED

-"II II II IT"
Seat
'ft'
f
Combined backwa/l ft
Secondary
and curtain wall
Cut wait

(bh FRONT ELEVATION (c)- SIDE ELEVATION

FIG. 14-12. Box abutments.

A
box abutment eliminates the fill that required heavy retaining walls
in the preceding cases. It may, therefore, reduce the weight, and thus be
suitable for use on weaker soils while appearing to be massive. Of course,
the structural deck costs money. With wide structures and deep cuts,
this may offset other savings.
Referring to Fig. 14- 12(c), the rear footing should be placed far enough
back from the edge of the slope so that the soil can spread the loads prop-
erly. In general, AB should be at least 5 or 6 ft. If the wing walls span
from front to rear, the embedment CD may be 1 or 2 ft. — enough to make
sure that erosion does not expose the bottom.
It is usually best to have a box abutment built as an integral structure.
If the front is over 40 ft. wide, it may then be desirable to have a keyed
618 BRIDGE ABUTMENTS [Chap. 14

contraction joint along the center line. Furthermore, a transverse row of

intermediate piers may be
used to support the center of the deck when the
latter is too long to be economical as a single span.
It is obvious that a box abutment may be skewed readily when the angle
is not too sharp for proper appearance. It is a type worthy of study when
the cut is deep or the fill is high.

Flanking-span abutments. It is logical in many cases to carry the

5.

idea of reducing the lateral pressures on abutments and of securing econo-

mies still further. Figure 14-13 shows this in what will be called a flank-
ing-span abutment. Here, the wing walls are frankly omitted; when
feasible, the front wall may also be reduced by the use of rectangular or
arched openings.
This abutment reduces the weight to that of the structural essentials,
yet it is automatically sturdy. The deck and secondary wall steady the
structure longitudinally; the front should be sufficient in frame action or

Reinforced -concrete deck (Balustrade or railing

Sidewalk -s

(Bridge seat -^Secondary

tu wall

fPier ^AContjt. if Depths of footings to

needed Cut suit conditions
-Grade
7W f
'W//SW/*Vtt\

(a)- FRONT ELEVATION r^-SIDE ELEVATION

FIG. 14-13. Flanking-span abutment.

in resistance to overturning to provide transverse strength without

depending upon the deck to resist twisting. Naturally, contraction joints
must not be used where they will interfere with the structural action. If
the bridge is very wide, the abutment might then be built somewhat as
two abutments side by side. Although this type may be better than many
others for use on poor soils and on piles, the abutment will move as the
earth supporting it moves.
6. Wingless abutments. A flanking-span abutment may become too
long or too costly to be a practical structure. The end span of the super-
structure, or a special short terminal span, may then be used. It may be
supported upon a simple wall with a bridge seat, backwall, and end wall as
shown in Fig. 14-14(a). This is merely supported upon a shelf excavated
in the side of the cut and then surrounded by backfill. If soil conditions
or the danger of scour require it, the wall may be deepened as necessary.
When the roadway is on an embankment, two or more piers may be used
to support the bearings directly or to hold a long bridge seat, and the back-
Art. 14-3] FOR CONTINUOUS BRIDGES 619

wall may span across them. The wings may be very small, and they may
be perpendicular to the bridge seat or extended at the ends parallel to it to
keep the fill from the bearings, an important matter.
Here there no attempt to terminate the bridge by means of any
is

massive appearing or special structure for aesthetic effect. The support

is classed as an abutment only because it holds up the end of the

superstructure.

-Wing
-Seat
Plan at Top
Abutment on Original Ground
of Good Bearing Value
Winn- —
v r.

*
Original
ground

'
Footing
Front Elevation Side Elevation
CaJ-ABUTMENTWITH SMALL REAR WINGS

'Wing
Sear
Plan at Top

XOrig/na/
ground
-jr0^

^Footing
Front Elevation Side Elevation
f6J-ABUTMENT WITH SMALL SIDE WINGS

FIG. 14-14. "Wingless" abutments.

If the end span of the bridge is a simply supported one, a shallow wall

like that of Fig. 14-14(a) may be founded directly upon well-compacted

embankment. If the abutment merely moves
latter settles a little, the
with it, and the span tilts slightly on This is sometimes
its bearings.
called a floating abutment. It may be advantageous where the soil is
weak and some settlement is inevitable, but where this will not be notice-
able or harmful. Of course, the wall should be strong enough to avoid
breaking its back if the fill settles unevenly.
620 BRIDGE ABUTMENTS [Chap. 14

Figure 14-15 shows the use of a wingless abutment. Of course, such

construction needs to be sketched in advance to make sure that the
appearance will not be one of weakness and lack of it is more finish. Also,
adaptable to overpasses and viaducts than to use at streams where floods
may wash away the toe of the embankment.
There are situations in which abutments are so small that they may be
of any or no special type. Such a case is pictured in Fig. 14-16. The

FIG. 14-15. Inkster Road Bridge, Detroit Industrial Expressway. This is an excellent example of the
use of side spans, sloping cuts, and small abutments at the tops of the slopes instead of massive
abutments with heavy wing walls. Notice how simple and neat the structure appears. (Courtesy
of Michigan State Highway Department.)

Wing
Gravity section.

FIG. 14-16. Small abutments seated on bare rock alongside railroad cut.

rock alongside the cut is left bare, and the abutment is nothing much more
than a bridge seat and a back wall. Small gravity-type construction may
be all that is needed. On the other hand, much depends in this case upon
the quality of the rock.
14-4. Abutments for rigid-frame bridges. Reinforced-concrete
rigid-frame bridges are both attractive and economical for many condi-
tions where short single or double spans are needed. The legs of the
superstructure, whether barrel or ribbed construction, serve as the front
wall of this abutment. This is shown in Fig. 14-19(a). Here the lateral
pressure of the earth in the direction of the bridge is resisted by the super-
structure itself; the abutment needs to hold the sides of the embankment
Art. 14-4] ABUTMENTS FOR RIGID-FRAME BRIDGES 621

only. Of course, there must be a footing to support the vertical load V

and the horizontal thrust H of the frame. The latter naturally is opposed
to the horizontal component of the earth pressure.
Since rigid frames are so dependent upon practically unyielding bear-
ings, their use on weak soils should be questioned and studied carefully

(o) A single-span, concrete, rigid-frame bridge (b) A two-span, concrete, rigid-frame bridge
with architectural treatment of surface and with with curved, flaring wing walls at abutment.
pylons and U-shaped abutment.

(c) A two-span, skewed, steel, rigid-frame bridge (d) A two-span, concrete, arch bridge with

with simple, boxlike abutment. U-shaped abutment and vertical markings.

FIG. 14-17. Some examples of highway bridges used to pass over the Merritt and Wilbur Cross
Parkways in Connecticut. (Courfesy of the Connect/cut State Highway Department.)

before being accepted. Vertical settlement of one end will cause a single-
span frame to tilt but not to fail; horizontal displacement of the
rigid
abutment may have serious consequences. Variations in loads and tem-
perature cause some changes in the magnitudes and directions of the
bridge reactions. These should not cause harmful movements under any
conditions. It is possible to tie the bearings together to resist the hori-
zontal thrust, but this is not likely to be economical or advisable.
Figures 14-19(c) and (d) show to exaggerated scale the structural action
622 BRIDGE ABUTMENTS [Chap. 14

of the rectangular pinned frame in (b) when acted upon by a

live load and
rise of temperature, respectively. and (g) picture
Figures 14-19(e), (/),
a simplified U-shaped abutment for such a bridge, assuming the upper
roadway to be on an embankment. The leg is supported upon a hinged
bearing along BB', Sketch (g), and it is free to deform between the fixed
retaining-wall sides of the abutment. These walls are generally tied
together by the footing under BB', but not elsewhere in this type. Joints
AB should be flashed to prevent leakage.
This provision for movement of the legs of the frame is very important.
Although the movements are small in angle and displacement, they are
essential to the structural action.
If the legs are keyed to or made a

monolithic part of the abutment

walls, something will break, or the
structure will not act as intended.
The be about as satisfac-
result will
tory as though a man were to try to
walk with his knees in two pieces of
cast-iron pipe or encased in concrete.
It might seem more simple to fix
the bottom of the leg along line BB'
in Fig. 14-19(gf) than to hinge it.
However, this section is generally
thin because a certain amount of
springing of the leg is essential, and
since it is so narrow at BB', any
restraint there will not have much
relieving effect upon the bending
movements in the main portion of
FIG. 14-18. An attractive treatment for the abut- the frame. A point of inflection
ment of a rigid-frame bridge over a roadway
will occur above BB' anyway. Any
in a rock cut. (Courtesy of the Port of New
yielding of the footing or of the
York Authority.)
materials in the leg will reduce the
supposed restraining moment; it may even damage the structure locally
and tend to produce a hinge where it was not intended to be. Unless
the legs are unduly long so that more stiffness is desired, it is generally
advisable to design the structure with hinged supports. Some suggested
details for hinged bearings are shown in Fig. 14-20, whereas some fixed
ones are pictured in Fig. 14-21.
The details of abutments for rigid frames vary widely, depending upon
the architectural style and treatment desired. When an underpass in a
city is lined with retaining walls, the bridge may merely rest in a slot in
the walls without any special abutments, as indicated in Fig. 14-22 (a).
When the upper roadway is on an embankment, it is possible to use sloping
 .Deck (c)- DEFORMED
FRAME,
,_ Horiz. component VERTICAL
k ofearth pressure V~ LOAD
Reactions t 1
4L JL
2 h- r (d)-DEFORMED
not shown ^;- L1NE DIAGRAM
4,
~
M (

OF RIGID FRAME
FRAME,
TEMPERATURE
Footing
RISE
(a)-SECTION THROUGH HINGED RIGID FRAME

fParapef
r E
{Pavement B' Sym. about <£>

Deck
Sym.
Deck
Wing
J$f Assumed mi abt. t- f. "'wall
Leg- ^f original
r Footing L eg
Joint

B
I

Wing watt fi' Wnqe*

s
Joinf^
Footing

CeJ-SIDE ELEVATION OF ABUTMENT f/7-HALF SECTION A-A ^J-HALF SECTION B-B

FIG. 14-19. Rigid-frame bridge and one type of abutment.

Sides of key way

Exp. joint
painted with
cement in asphalt
groove
Curb.
-I-
/TTt-
Tk
e "fo '//sheet lead
'/

Cover of asphalt ]i u-*ir— -/ P dowels Heavy

L Ji °" c C footing I \*~% "dowels 2-0 ±c. c.
2"x'/
4 'welded^ - -

More suifab/e with

anchors l'0"±c.c,
mastic or asbestos
faJ-HEAVY STEEL PLATES AS lining ofgroove
BEARING SURFACE r^-CURVED.LEADED KEYWAY

Backival/

J <• V •

'^,fh\^-ji\%hot
If Fabric
asphalt below
and
to!"
JPadded * ,

Foot, premolded
filler

'.(" a , c ,
I „. » " •
Paint or oil
„
Asbestos
•

filler
on key wall
or asphaltic paint
CcJ-FLAT, ASPHALTED SHELF Re inf. not shown
ftfJ-CURVED, CENTRALIZED KEYWAY
FOR RIBBED LEG
Premolded
filler
Const, jt,
Large V-cul oiled or Curb
h —tvJr— -?\ Oil bear/ng
j.
»
til" Vw surface

Premolded

W Heavy dowelst-- 1
Cj shrrupf^-^ Heavy f00 ffn9
filler

feJ-HINGEAND KEY DESIGNED AS A tfV-HINGE EFFECT WITH CENTRALIZED CORE

SIMPLE LINE OF WEAKNESS IN BENDING FOR SUPPORTING VERTICAL LOADS

FIG. 14-20. Some arrangements for hinged bearings of concrete rigid-frame bridges.

623
624 BRIDGE ABUTMENTS [Chap. 14

Membrane waterproofing
1 Plywood, mason lie or
2 coats ofemulsified
Pavement J"
I
1
•'

"
•

'

'
"d M
}l%
other protection
'asphalt or other
ff: ; tiM Mortar or bricks dampproof/hg
'

¥>-<-$ F/tt or backfill

C~y-Drain ifneeded

(a)- FOOTING ON ROCK (^-FOOTING ON EARTH

FIG. 14-21. Details of fixed-end footings for barrel-type concrete rigid -frame bridges.

Sym, ab'f <£ Exp.joint

7}
f
Pavement
Leg- Deck^ / {Deck ofabutment /
0-
^?P
footing
\\-Intermediate
under y
^ Fit/ ^columns if
bridge '$. Joint \ I

necessary
X t>

*s

k ')
Footing
5
^s Footing of w/ng watt
<S \
Q
^ " r^-BOX-TYPE ABUTMENT
k
^
i Retaining
o \ "_
"watt footing Exp. joint
"vj

Wing waitI
Deck. I ,Deck Pavement
1 -' > * r. . .
jL

foJ-STRAIGHT WING OriginaT^riVrA Wall^&Earth

rock iinej_:„

Pavement Exp.joint -Rock

Ji Col. if needed
Barn-door Drain
Slab
Membrane
water- rc^-BOX-TYPE ABUTMENT
proofing ON ROCK
fdJ-ONETYPE OF JOINT
BETWEEN BRIDGE
AND PAVEMENT

FIG. 14-22. Straight and box-type abutments for rigid-frame bridges.

retaining walls parallel to the bridge legssomewhat as with straight-wing

abutments. This, however, may not yield a pleasing appearance; the
wings might better flare back at least 45°. Hollow box abutments can be
used advantageously in some cases, thus eliminating large earth pressures
against the leg and abutment. A few schemes for these are indicated in
Figs. 14-22(6), (c), and (d).
 CD CC

yd

19 E

625
626 BRIDGE ABUTMENTS [Chap. 14

Rigid-frame bridges are deservedly popular for use as single or double

spans at grade-crossing eliminations. A few suggestions of various
architectural treatment for the abutments of concrete bridges are shown
in Fig. 14-23.
Longer rigid frames may be made of riveted or welded steel ribs with
equalizing diaphragms between them, and with a concrete deck on top.
It is possible to extend this deck down the rear of the legs as a curtain to
hold back the earth from the steelwork. The motion of the steel, how-
ever, is likely to crack this cover noticeably. Another and better method
is to seat the bearings on a footing near the base of a retaining wall or cur-
tain wall at the front of an abutment of wing, U, or box type. Naturally,
this arrangement requires a more costly abutment than the former. A
few suggestions for such construction are pictured in Fig. 14-24.

Welded leg Riveted leg-

Transverse
Transverse
bracing bracing
or struts or struts
not shown not shown
Bil/et Vr=a
welded to leg^
Billet >S
'
^Anchor Grout 2
/ bolts
Anchor boils
(a)- PINNED BEARING (b)- HINGED BEARING

FIG. 14-24. Two types of bearing for steel rigid-frame bridge.

14-5. Abutments for arches. The horizontal component of an arch

reaction is generally large with ; many ordinary arches it exceeds the verti-
cal component considerably. Furthermore, arches are used for spans
that are far greater than those of rigid frames. Hence, heavy loads and
large thrusts are to be supported by the abutments. Therefore, good
foundations are needed for this type of structure.
Rocky gorges are often ideal situations for arch bridges. The splendid
structure shown Niagara River below the falls.
in Fig. 14-25 crosses the
The steeply sloping rock is strong and sound. The abutments, therefore,
are practically nothing more than inclined footings that receive the thrusts
of the bearings and transmit them to the bedrock. In this case, however,
ice jams might be serious, hence the abutments are kept high enough above
the water to avoid damage by the terrific ice conditions in spring floods.
Three-hinged arches are statically determinate structures. Therefore,
slight yielding of the foundations may cause no serious harm. If the soil
is weak and compressible, however, it may be difficult to determine

whether or not the movement will continue until it becomes harmful. On

Art. 14-5] ABUTMENTS FOR ARCHES 627

general principles, such soils are not suitable for arches unless special
provisions are made to prevent harmful movements.
Of course, an arch may be desired on the basis of its attractive appear-
ance. One such case is illustrated in Fig. 14-26. Rock was available at
one end but was some 20 ft. below ground at the other. Therefore, the
latter was supported by a heavy extension or thrust block capable of
resisting any combination of vertical loads and horizontal thrusts that
might be applied by the arch. The wings of the abutment holding the fill
at such a point as this should be supported upon piers or piles that reach
the rock. Otherwise, differential settlement and cracking of the unequally
supported parts may occur.
If an arch is designed with fixed ends, it may be difficult to make the
abutments such that they will certainly offer the necessary restraint.

TTIHTTS'

FIG. 14-25. The Rainbow Bridge over the Niagara River below the falls. This is a fixed-end steel
arch founded upon rock. It was built for the Niagara Falls Bridge Commission, Niagara Falls,

United States— Canada. Notice the ice in the foreground. (Courtesy of Hardesty and Hanover,
Consulting Engineers, New York.)

Foundations on rock be suitable; those on earth or piles may twist

may
assumed restraint without any-
sufficiently to relieve a large part of the
one's realizing it unless the structure gets into difficulty. One might
think that reinforcing bars or heavy tie rods could be grouted in drilled
holes, thus fixing the abutment as strongly as needed. Some problems
however: the stretching of long rods under high tension, the possi-
arise,
weakening produced by corrosion, inadequate grouting,
bility of eventual
weakness in bond or anchorage, and weakness in the rock caused by cracks
and seams.
Most large reinforced-concrete arches are supportedupon fixed or
abutment blocks which the arch
restrained bearings that consist mostly of
rib joins somewhat as would an inclined column. The reinforcing is
carried into the abutment to provide the necessary resistance to bending.
One type of fixed (restrained) and one hinged bearing for steel-rib arches
are shown in Fig. 14-28.
628 BRIDGE ABUTMENTS [Chap. 14

Suspension-bridge anchorages. The anchorages of suspen-

14-6.
sion bridges differ considerablyfrom ordinary bridge abutments. Their
planning and detailed design for a major bridge usually require the work
of several competent engineers with specialized knowledge and experience
with such structures. However, the basic principles are relatively simple.
{Roadwcty-4 lanes

''/SrefiS<&i,. Subterranean rocky gorge

Tim. fi/led with weak so/7
FIG. 14-26. A reinforced-concrete arch bridge supported upon rock at two different elevations.

FIG. 14-27. Details at the ends of the Bayonne Bridge, a trussed arch. The maximum end reaction
is approximately 30,000 kips. (Courfesy of the Port of New York Authority.)

Figure 14-30(a) illustrates an anchorage built into a rock mass of prac-

tically infinite magnitude. This is almost an ideal situation. Figure
14-31 (a) shows a "gravity" anchorage, this being a massive block of con-
crete that holds down the end of the cables. These two drawings repre-
sent the ordinary types of anchorage. Of course, variations and combina-
tions may be made to suit special conditions.
 Art. 14-6] SUSPENSION-BRIDGE ANCHORAGES 629

Welded Welded seat

arch rib (and stiffeners Bracing not
shown
Dry-packed
'grout Tap bo/ts
Pin with Lomas
nuts or capsC
One or
fwo plaies Grout
of
-2 sets
anchor bolts
and double -
channel
anchorages

Abutment
foJ-A"FIXED BEARING (b)-k HINGED OR PINNED BEARING

FIG. 14-28. Two arrangements for the bearings of steel-rib arch bridges

FIG. 14-29. View of a bearing, abutment, and approach piers of the Rainbow Bridge at Niagara
Falls, N.Y., as seen when looking upward. Notice the details of the bearing, the fluting, and the
rough, ridged surface of the massive piers. (Courfesy of the Niagara Palls Bridge Commission,
Niagara Falls, United States-Canada.)
630 BRIDGE ABUTMENTS [Chap. 14

Wire rope cables of small bridges may be fastened directly to anchorage

masonry through sockets and structural ties. Parallel-wire cables of big
bridges terminate in strand shoes outside the masonry. In Figs. 14-30
and 14-31, each 36-in. cable is built as sixty-one 4^-in. cables or strands.
Riverward of the anchorage saddles A, these are squeezed together and
wrapped to form the seemingly solid circular cables. Landward of these
saddles, the strands are splayed vertically and horizontally so that each
can be looped over a strand shoe. The saddles may be on rollers sup-
ported by masonry, or they may be on steel rocker bents. In any event,
the saddles can move slightly with stress and thermal deformations of the

TTrrrTTTTTTTTI I M I I ' M I
I
I I
ITTmr
Rock cut im'jI m
/Roadway— II 1
1
1

Elevation of future I Anchorage saddle}

lo wer deck-j Strand

Possible
line of rye bars
failure
\J ,

of rock I

Anchorage
girders

(bh ACTION OF ANCHOR PLUG

FIG. 14-30. Simplified section through New Jersey anchorage of the George Washington Bridge.

cables so that the latter will not slide in the castings. The strands are
bent downward more or less at the saddles so that they will not lift out
under tension, Everything beyond the strand shoes is finally embedded
in concrete.
The strand shoes are fastened by pinned connections to eyebars or
fabricated heavy plates that transmit the cable pull back to anchorage
girders near the rear of the masonry. These girders bear on the concrete
and are designed to transmit to it a compression equal to the cable pull.
In Fig. 14-30(a), the portion BDEC is somewhat like the reverse of a
cork in a bottle. As shown in (b), the cable tries to pull the concrete plug
out of the rock.
In Fig. 14-31, over 100,000 cu. yd. of concrete constitute the U-shaped
Art. 14-7] MISCELLANEOUS DETAILS 631

block into which the anchorage steelwork extends. The cables try to tip
this block about some point near its front edge. The moment Wd must
exceed the overturning moment Hh by at least the desired safety factor.
The resistance to sliding should be equally safe.
14-7. Miscellaneous details. The details of the joint between the
end of a superstructureand an abutment are important; they may be
rather troublesome also. Many arrangements are possible, some being
illustrated in Fig. 14-32. Not only should the joint be made so that its
parts are structurally strong enough to support concentrated loads and

Arch over
Riverside
Drive

v~Anchorage
girders
faJ-SECTION THROUGH ONE SIDE OF ANCHORAGE

T^

T^
f&J-PLAN OF BOTTOM 0FANCH0RA6E fcJ-ACTION OF ANCHOR BLOCK
FIG. 14-31. Simplified section through New York anchorage of the George Washington Bridge.

impact, it it will not let through the tire of a

should be so designed that
bicycle, a pedestrian's heel, or anything that may cause an accident. It
should also be made so that drainage water will not flow over and stain
the abutment, thus spoiling its appearance. This last may be difficult to
avoid, but it should be accomplished when attractiveness is desired.
Fills placed behind abutments for the support of traffic will almost
inevitably settle somewhat. It may be advisable to use a bituminous
pavement on the fill from the backwall of the abutment to a point 50 or
100 ft. away. When the settlement causes a depression (and bump to
traffic) next to the backwall, this temporary pavement may be built up or
U&Myovg IIVMyovgT
632
 _

Art. 14-7] MISCELLANEOUS DETAILS 633

otherwise repaired easily. When compaction has ceased some years later,
a permanent pavement may be built.
A better remedy is the use of a "barn-door" slab that can bridge from
the backwall to the fill far enough away so that the angular rotation of the
slab will cause an acceptable ramping effect only. This construction is
illustrated in Fig. 14-33.

Asphalt or paint Typical pavement '

d " " \ »'

V
'•'
1 -
r ', ' ° 'I "
|p t , /Dowels /:p . , » ) {jJ>—R -*-
\\\>\\S\<\\

3i. p.. '

f" 'Slab
v^JL, Backfill
V
'

^Backwall /0 io"/o/5-0"

r«>SEAT,SLAB DOWELLEDT0 BACKWALL

Bituminous macadam-.

Slab

Back ft//
U-2-~±^ Back wall
(b)-SLA2> EXTENDED ACROSS BACKWALL
"Barndoor "s/ab.

.D : o

I ft//
Slab as part of abutment :•;»•'„ '•;

M
^

i. Backwall Secondary
wall _
_
| :V»^;i,;;: ^W< .

Hollow space °i:'

:
:
:
/
:
:°

Rock
fcJ-HOLLOW ABUTMENT AND AUXILIARY SLAB
FIG. 14-33. "Barndoor" slabs spanning from abutments to ground or fill.

Adequate waterproofing of abutments may seem costly but it is impor-

tant when durability and appearance are to be at a maximum. It is
usually sufficient to use a two-ply membrane with adequate protection,
and to have waterstops at the contraction joints. Drainage of the space
or fill back of an abutment should always be as adequate as circumstances
permit. Weep holes alone may become clogged with dirt or closed tempo-
rarily by icicle action.
Should an abutment be battered? A slope of ^4 to Y^ in. per ft. is
sometimes desirable for the front and wings. In other instances, as in
Fig. 14-23, a greater batter may be used to create a particular architec-
634 BRIDGE ABUTMENTS [Chap. 14

tural effect. Some an extensive high wall

architects have stated that
that is absolutely vertical may
appear to lean forward.
In building an abutment, as with other concrete structures, care should
be used to obtain a dense durable concrete and to have a surface finish that
produces a pleasing effect. The use of V cuts or straight-edge finishing at
construction and contraction joints will generally avoid ragged unsightly
junctions of pours. The entire design should be planned to facilitate con-
struction, provide adequate strength, and yet produce a structure that is
attractive and serves its purpose well.
-£ of bearings (fixed)
Top ofroadway-*
4%down

Lake,

,300 k 340*.

Longitudinal force - 10 k per bearing

W-HALF CROSS SECTION OF BRIDGE

FIG. 14-34. Problem in design of an abutment for a highway bridge.

PROBLEMS
14-1. Design an abutment for a bridge to meet the conditions shown in Fig. 14-
34. This highway bridge consists of two symmetrical portions with a longitudi-
nal joint separating them. The bridge is to cross the narrow portion of a long
shallow crooked fresh-water lake in a mild climate.
Suggestions: Floods and ice are not important here. This is a six-lane boule-
 « 5±

PROBLEMS 635

*EU48 /% up- 1% grade up ,Ef. 148

Design bads per bearing Sfeel girder

Sfeel girder
Vertical'= 175 El 138^
x, , Transverse =8" Sta. 182+81.5
sta.imis El. 138 Longitudinals*
Fixed bearing
Fixed bearing High water - El. 135

Low water El. 123 (Low wafer El. 123^ Fine loose sand
:
.
• •
\ 120^

:
!v^» no
Loose sand and gravel
'•' e
°o'°' -^ o°o'
inn vf'i-'-iL boulders •^%?'°Z
Bedrock T^^ffi^F
Hard sand and gra vel
AND BRIDGE SEAT AT LEFT ABUTMENT
feJ-SOIL PROFILE
^ "V.V 90
5'0" 32'-0'-
Sand ~~~^
U_ l^ qV,°»° Hardpan witn'of'Oi

1
32-0 ex. bearings J
^;-DIMENSIONS OF ROADWAY AND GIRDER SPACING fe/-S0IL PROFILE AND BRIDGE SEAT AT RIGHTABUTMENT

FIG. 14-35. Conditions at the abutments of a proposed bridge.

Base ofrait
r~ To o fl° o

( ° <

fo r

M--
!. 6-0" \. 8 L 0' 6'-0" ki_

I4 L 0" - Bridge seat

El. 243.
r«;-cR0SS section of superstructure to*
%f ofrail E/.2S/.67 Side Elevation End View

^-DETAILS OF FIXED BEARING

El. 222.
):°O.';'0.':.o o
-<£ of boring A/o. 8
Scale in feet
Coarse sand with some gravei-deep stratum
fcj-ASSUMED PROFILE ALONG CENTER LINE OF BRIDGE

FIG. 14-36. Data showing superstructure and soil conditions at site of a proposed abutment of a
railroad bridge.
636 BRIDGE ABUTMENTS [Chap. 14

vard supported by four girders spaced as shown. Assume the allowable bearing
pressures from Table 4-5. Use concrete construction.
14-2. Figure 14-35 pictures the conditions at both abutments of a three-lane
deck-girder bridge, over a river approximately 250 ft. wide with moderate current
at flood stage.Design the abutments for this structure.
14-3. Figure 14-36 shows the general construction and the bearings for a
double-track deck-girder railroad bridge. The end span is 60 ft. long. Design
an abutment for this bridge. For preliminary purposes, use the load data in

Figs. 13-34 to 13-38.

Top of pavement, EI.4/6

9v
Span =40 L "clear
* Original ground line.
fw* El. 400
SaMand] —
"
i
f Roadway, & 398
qravelf^J19
El. 390
~Rock
faj-VERTICAL SECTION ALONG CENTER LINE

/Railing
4' sidewalk
tCurb

<Curb

(b)-?im OF STRUCTURE AND ROADWAYS

FIG. 14-37. Problem in planning abutments and foundations for a rigid-frame, concrete-barrel

highway overpass.

14-4. Design the abutments for the highway bridge shown in Fig. 14-37. They
are to be built of concrete and are tohave pylons at the corners for architectural
purposes and to hold light standards. Assume hinged bearings, with a vertical
load of 16 kips and a horizontal load of 4 kips per lin. ft. applied at these bearings.
The railing is and two pipe rails with welded
to be pipe posts on a 6-in. curbing
vertical spindles between them. Invent two abutments that are architecturally
attractive and harmonious even though perhaps different because of the marginal
roadway at the right.
14-5. Figure 14-38 shows a three-ribbed reinforced-concrete arch bridge to be
built across a small river near the Atlantic Coast of southern New England. The
architect wants this type of structure. The loads are given in the plan view.
Design the foundations for the structure.
637
 15
UNDERPINNING

15-1. Introduction. The word underpinning is used in foundation

engineering to denote the process of transferring the weight of an existing
structure from its original foundation to temporary supports, and from
them new permanent substructure. It may also denote the construc-
to a
tion of new foundations under the structure without the intermediate use
of temporary supports. When the work is completed, the structure is
said to be underpinned. This may be necessitated occasionally by the
inadequacy of the original foundations, or by their harmful settlement.
In most cases, however, it is caused by the fact that new work in the
immediate vicinity of the structure, or even in the building itself, requires
new foundations at a lower level.
Some of the modifications and construction likely to make underpinning
of existing structures necessary are the following:
1. Provision of a basement in a building that had none before, or of a

subbasement where only an ordinary basement existed previously.

2. Installation of tunnels, sewers, water pipes, electrical ducts, and air

ducts under or near columns and walls. This is likely to occur when
alterations are being made in an industrial plant.
3. Construction of pits, tanks, sumps, wells, and elevator pits.

4. Lowering of basement floors.

5. Construction of subways, and lowering of adjacent streets for grade-

crossing eliminations and other street improvements.

6. Lowering of the watertable by adjacent excavations or pumping

that endangers the safety of footings on wooden piles.

7. Increase of loads on columns or walls that requires replacement of

the present foundations. A revision that merely adds to the size and
strength of a footing to permit greater loads without taking the original
load off the footing is not underpinning in a strict sense.
8. The construction of deep foundations in positions that endanger
present ones.
638
Art. 15-2] GENERAL PRINCIPLES 639

9. Lowering ground surface by regrading so that footings have to

of the
be set down damage from frost action.
farther to prevent
10. Scour, or the threat of dangerous scouring, caused by flowing water.
15-2. General principles. The conduct of underpinning work is
likely to require the exercise of great ingenuity and engineering judgment.
Conditions, usually seem to be special for each job. The soils, the water
conditions, the character of the new work, the details of the old structure,
the maintenance of use of the structure during the conduct of the work,
the magnitude of the loads — all these vary greatly.
It iseasy to say that the work should be done so as to support the exist-
ing structure safely, and to avoid damage to it. This is the essence of the
problem, and the problem is a tough one. The work will almost always
be slow and expensive. As illustrated later, the underpinning has to be
done a little at a time and usually involves a great deal of hand labor.
Many times the work has to be done with whatever materials are at hand
or readily procurable.
Generally, the load has to be taken off the existing footing and held by
temporary means while a new support is provided. The load then has to
be transferred to the new bearing. It is important to do all this in a man-
ner that does notrequire or produce raising or lowering of the structure
sufficiently to cause cracking or other harm to it. Adjustment to proper
elevation is necessary as well as strength of supports.
One means for obtaining adjustment is the use of wedges. The amount
of their allowable movement is generally small unless the wedges are
driven, shims or blocks inserted to hold the load, the wedges taken out,
and the wedges redriven on top of additional shims. Large forces can be
raised or supported by wedges, but the force exerted is not easy to control.
For example, an enthusiastic workman told to drive in the wedges with a
sledge may hit them so hard that he lifts the structure too much, and
therefore cracks it. A lazy man may do the opposite, with the result that
the structure settles and cracks when its is removed.
original support
Double wedges are preferable to single ones. Any wedges should be
arranged so that they cannot out or be knocked out accidentally.
slip
Another method of adjustment is the use of jacks. Screw jacks are
good and will hold their load indefinitely if used properly, but a single one
has very limited capacity. Several may be used in a battery, but great
care is necessary in operating them so that the load is distributed satisfac-
torily to each one in the group. Hydraulic jacks may easily have an
individual capacity of 100 to 200 tons. However, the loads should not be
left on them for long periods because of possible leakage or breaks in

piping. It is best to use the jacks to hold or raise the load, then block up
the structure with shims and wedges. Later on, the wedges can be driven
out or the jacks can again take the load so that the shims can be removed
640 UNDERPINNING [Chap. 15

and the load transferred to the new supports. Both types of jack give a
fairly good control over the force exerted by them.
Turnbuckles, bolts, and threaded rods are useful in obtaining adjust-
ment when the loads are suspended from some overhead structure or
framing.
Ground water is often troublesome in underpinning operations. Sumps
and pumps may have to be used, or it may be necessary to install well-
points. Underpinning operations in wet runny soil are especially
dangerous.
/ When work requires underpinning of structures that are the property of
others (or even work near these structures), the owner of the new work
should examine these structures carefully, take extensive photographs
showing their present condition, and make a report upon the subject, hav-
ing the report reviewed and approved by someone who is competent to do
so. He is then in a position to defend himself against unjust claims for
damages that are allegedly caused by his construction operations. It is
a reflection upon human nature that, after such operations, the cracks in a
building suddenly become prominent and important. Photographs show-
ing their previous existence are very helpful. In fact, a contractor is wise
ifhe takes pictures and makes a report on his own account if the owner
does not do so. The contractor can then protect himself if necessary.
It is obvious that underpinning work should be done with extreme care.
Nevertheless, here is the story of one actual case:
A tunnel was being built for a new conveyor under the first floor of an
industrial plant. The building was unencased steel framing with several
floors or heavily constructed platforms. It housed and supported equip-
ment for roasting (drying) ore. The tunnel was to pass close to one of the
interior columns. The excavation had proceeded past this particular
column which, with its old footing attached, was blocked up by heavy
vertical timbers under the concrete. Blasting was required for the exca-
vation, and it had been conducted carefully. One day, a workman appar-
ently thought that, "if a little dynamite is good, more will be better."
He therefore put in much heavier charges than usual, without the knowl-
edge of anyone else.
When the blast was set off, it knocked all the shores out from under the
column footing. The column settled about 3 in. but did not fall down.
Seemingly, the beams and roof framing acted as a set of cables, cantilevers,
and whatnot that transmitted the load to surrounding portions of the
structure. The timbers were hastily wedged back under the footing
temporarily.
The next day, the men tried unsuccessfully to jack the column back to
proper elevation. They then collected all the jacks that they could find
around the plant. These were placed under the footing. The only result
Art. 15-3] UNDERPINNING COLUMNS 641

was that several of them broke without budging

the column. The dis-
tortion of the building could not be eliminated.Some tracks that were
supported by the floor near this column were shimmed up later and the
column footing was underpinned with concrete. The building is still in
service, but such operations are not recommended.
15-3. Underpinning columns. Although the procedures used are
similar in many respects for underpinning columns and for underpinning
walls, the former will be considered first because it involves supporting
a single isolated load.
Assume a steel-framed factory building with columns supported upon
isolated spread footings. A small basement is to be built under the first
floor. This requires underpinning for one of the interior columns.
In this case, the dead load of the structure is relatively light. As shown
in Fig. 15-1, it is feasible to drill a few holes in the column flanges and to

Sieel column
Bolted block
X2£s
P- ©||0|
_o_uj-

Wedges*

Ti'mber
Wooden
"mud posf"^

MdaH"
i/mmvrw/wv/rh
KOldpedesfal
I
I
WV/'V//C.'W/'WV/'W
i

removed

ir FIG. 15-1. One method of shoring steel column.

bolt two heavy channels These can be supported upon timber

thereto.
posts and grillages called mud sills,
with wedges (or jacks) under the chan-
nels at the top of the posts as shown, or under the posts. The load can be
transferred from the footing to the channels by driving the wedges suffi-
ciently. The supporting system should be braced carefully to make sure
that no disturbance of it occurs during the work of demolishing the old
footing, excavating a shaft for the new footing that is to be below the base-
ment floor, and building the new substructure. When the new concrete is
set, the base of the column can be grouted or dry-packed, then the wedges
can be loosened so that the load of the column is transferred to the new
pier and footing. It is generally inadvisable to try to pour the concrete
clear up to the billet because adequate and even bearing may not be
obtained.
Many variations of the general scheme illustrated here may be used.
Wedges or jacks at one end of the temporary beams may be sufficient. If
a small initial settlement of the structure is not harmful, the wedging
642 UNDERPINNING [Chap. 15

might be done under the column base plate after the new pier is ready for
grouting. However, this procedure is not advisable. It is better to pick
up the load first to make sure that there will be no weakness in the tempo-
rary supports or harmful settlement of the mud sills before the old founda-
tion is removed.
Another method that may be used for steel-framed structures, and some-
times for reinforced-concrete ones, is illustrated in Fig. 15-2. Inclined
posts supported upon mud sills are wedged under beam or truss connec-
tions to form an A-frame that will support the column until the new

Hardwood
b/ock

^- Bo* torn of trench

FIG. 15-2. Wedged, inclined posts under beams. FIG. 15-3. Inclined shoring of a column.

foundation is ready. This may be dangerous if the beam connections are

inadequate. In the case of concrete beams, the reversal of the end shear
and its large magnitude may be especially serious.
One method of underpinning a concrete column is shown in Fig. 15-3.
A trench is to pass next to the footing, and the latter is to have a concrete
pad placed below it to reach to the level of the invert of the trench. It is
easy to make steel, but it is difficult in the case of
connections to structural
reinforced-concrete columns. In this case, two chases are gouged out of
the concrete as far as the reinforcement. Two steel channels a are erected
as a yoke and tied together by bolted angles b. They are then grouted in
place to bear on the concrete. The operation may then be conducted as
explained for Fig. 15-1, or an A-frame may be used as shown here with
Art. 15-3] UNDERPINNING COLUMNS 643

posts c on wedges or jacks. The best procedure is to pick up the load on

the A-frame, excavate under half of the footing, pour half of the concrete
pad d, and then dry-pack between this new concrete and the footing.
When this is done, repeat the process for the other half, then release the
load onto the footing, dismantle the shoring, and patch up the column.
However, the column will be permanently weakened by the portion that
is chipped out because plaster patches will remedy the appearance only.
In general, such a procedure is dangerous.
One abutment (or two-column pier at the end of a main truss) of the
Summit Bridge near Chesapeake City, Md., settled and tilted when heavy
rains washed out part of the bank. A serious slide threatened. The
Army Engineers saved the structure by doing the following:
1. The approach structure was dismantled and some 20,000 cu. yd. of

earth removed from behind the abutment in order to relieve the lateral
pressure.
2. A concrete deadman was buried 50 ft. back of the abutment. This
was tied to the bottom of the substructure to help prevent sliding
riverward.
3. A wellpoint system was installed to a depth of 20 ft. to remove the

ground water, thus preventing lubrication of the soil.

4. Eight steel H piles were driven alongside each of the two shafts.

5. Slots were cut in each pier shaft, as indicated in Fig. 15-4. Beams
36 deep were then erected on top of the piles and concreted to bear.
in.

Long rods between the beams were tightened to hold them together.
This steelwork formed a pile-supported grillage under each column so
that they could support the structure if the soil under the footing yielded
further.
6. The bridge was then raised by jacks placed on the tops of the col-

umns, concrete pedestals were poured, and the bearings were reset.
What may be called underpinning of a column is illustrated in Fig. 15-5.
The soil is dry or moist gravel. The floor alongside the row of columns is
to be lowered several feet to make a large shallow pit and a retaining wall
is to support the earth. Under and in the vicinity of the footings the
ground is first grouted. Pipes are driven down in various positions as
indicated, and the cement grout is forced in slowly under high pressure
until a sufficient mass of gravel is thought to be converted into a sort of
very weak concrete. When this is set thoroughly, the excavation of a
trench for the wall footing is made, using shoring where it seems neces-
sary. This excavation is done by careful picking of any grouted soil that
must be removed. The wall is then built. Below the footing, the con-
crete of the wall might well be poured directly against the ground to ensure
support for the latter. If this is not done, sand or gravel should be packed
into the intervening space. The rest of the excavation can be made at
will.
644 UNDERPINNING [Chap. 15

By no means can most ordinary soils be grouted successfully. It is

obvious that silts and clays will not allow the grout to penetrate, even
under high pressure. Sands which have a considerable proportion of fine
material will probably resist grouting so strongly that the results will be
merely the creation of a few pockets
of grout rather than a really densified
material, even this causing displace-
Shaft
ment of soil rather than filling of the
Concrete voids. As a general rule, sands and
gravels can be grouted if they have
Projecting a large percentage of voids and are
beams coarse, having perhaps no more than
10 per cent (by weight) of the parti-
Steel Hpiles cles smaller than 3^16 m -> ° r possibly
1 mm. In finer material, the cement
particles tend to clog the adjacent
interstices instead of penetrating
through channels. 1

In the remodeling of an industrial

FIG. 15-4. Use of building there was an interesting ex-
permanent yoke.
ample of underpinning with cement
grout and of the inadequacy of hearsay and assumed information. This
structure had been built on fairly level ground and approximately 75 ft.
from the foot of a steep outcropping of rock. New heavy equipment

, Column

XS Footing ,/
'#
¥~~~^ t'
»;.M
New wa/L //
..'/>

Line of — H \\Crout p/pes //

excavation i
V- \
i
V
\
. --Grouts.

FIG. 15-5. Use of grout as underpinning.

was to be installed in the structure, and the engineer inquired about

the adequacy of the soils. The former operators said, "The building is
all founded on rock. Can't you see it right there?"
1
If certain chemicals are used, it may be possible to consolidate somewhat finer
materials than can be grouted. However, the results are likely to be uncertain.
Art. 15-4] UNDERPINNING OF WALLS 645

There was already in the basement a heavy slab that formerly sup-
ported some small compressors. Taking the information given by the
operators as correct, the engineer built on the basement slab a heavy
pedestal for a large compressor. It consisted of two solid concrete walls

and a thick cap. The was to constitute the footing.

old slab
When the compressor was started, it acted somewhat like a rocking
horse. It had to be shut down immediately. Investigations then showed
that rock was about 100 ft. below the building and that the ground below
the thin sandy top stratum was a fine silty sand of poor bearing power.
What was to be done? After considerable study, the engineer decided to
grout the soil to a depth of about 30 ft. under and around the machine in
order to stiffen a large volume of ground and to form a sort of enlarged

r Original ground //'rre_

A*J
f«J-SIDE ELEVATION ^-SECTION A-A

FIG. 15-6. Underpinning a continuous wall alongside a new subway.

footing. This was done by using a pressure of 100 p.s.i., by starting at

the deeper level, and by grouting in the hope of forming something like
horizontal layers. After the job was completed, the machine operated
but no one knows just how thorough the grouting was.
satisfactorily,
15-4. Underpinning of walls. Figure 15-6 illustrates one method of
underpinning brick, masonry, and concrete walls, or concrete foundation
walls gthat support intermittent exterior columns. This shows a case
where a vehicular subway or underpass is to be built close to and along-
side an old building.
The general method the excavation of pits some 3 or 4 ft. wide at
is

intervals of 15 to 30 This allows the load to be resisted temporarily

ft.

by the intervening undisturbed soil. It illustrates the utilization of the

safety factor in the bearing capacity used in the original design. Thus,
pits represented by A x and Bi are dug, using sheathing generally to make
646 UNDERPINNING [Chap. 15

sure that the intervening soil does not slip. Then a solid concrete pier
(or a footing and a thin pier) is made up to about 2 or 3 in. below the old
wall or footing. After the concrete has set properly, steel or hardwood
wedges are driven between the JieiiL-pier and the old structure, as shown,
in order to force load into the pier.. The rest of the space is then packed
with stiff mortar^
The next step is the excavation of pits next to the first ones, as shown
by Ai and B 2 The adjacent concrete piers are built as before. Then
.

follow other pits and piers. The last ones, pictured by A 4 in Fig. 15-6,
are finally excavated and concreted, thus forming what amounts to a
continuous foundation wall. It is seen that no attempt is made to lift
the old structure, unless the wedging does so unintentionally.
In'the^case of a concrete foundation wall supporting light columns
20 ft. c.c, it may be most advantageous to construct the piers under the

Wall. Concrete
{Needle beam ^/or mortar

Post Packing
.Wedges
Mud sill
Underpinning
pier
Pit
Fooling.
-
y//\\ \>/A V>»ty?/\WAt>

FIG. 15-7. Illustration of use of needle beam.

columns first, making them perhaps 4

ft. wide so that the bearing area

will be considerable. A
middle piers may be constructed next to
series of
hold the central portions of the wall. Then the remaining spaces can be
excavated and filled in.
The use of needle beams is shown in Fig. 15-7. The inside of the
building assumed to be accessible for work. A series of holes some 6 to
is

10 ft. or more apart is made in the wall. Steel beams are placed through
them and supported upon mud sills and blocking of some sort. The
spaces between the wall and the beams are filled with concrete or mortar
to obtain strong bearing. The beams may then be wedged up to transfer
the load of the wall to them, or the wall may be allowed to settle until the
beams support it (a dangerous procedure). The excavation, concreting,
wedging, and packing can then be done more or less in the open and in
quantity. The needle beams and their supports can be removed later,
and the wall can be patched up.
Art. ]5-4] UNDERPINNING OF WALLS 647

A variation of this procedure is shown in Fig. 15-8 (a). Here the needle
beams are placed under the footing. Part of the basement floor is
removed. The general excavation outside is carried to some such line as
EFGH. That inside is made along line A BCD. Slots along JK are dug
out next at intervals, the needle beams / are slid through and placed on
their supports b and c. The outer ends of the beams are then jacked to
pick up the load, and blocks e are put in to support the beams. The
excavation GLMN is then made, the concrete footing and wall are poured,
any wedging and packing are completed, the shoring is dismantled, and
the job is finished.
A footing is shown purposely in Fig. 15-8 (a). On account of the
appearance, it should not be left exposed as shown. If it has a good

Wall

Hydraulic
jack d Basement
Fool/nq (floor a
Steel
w ^ll»
w
jr
C
t G ^ \)W//|
Nem ,

I /
,-

\ [ j /
New foof/hgr^"^ _
L M j

(a)

FIG. 15-8. Needle beams under footing of a wall.

formed outer edge, the new wall may be made thick enough to line up with
this edge, and to look like a regular foundation wall, especially if it is all
covered with stucco or a cement wash. If the footing has a ragged edge
because it was poured against earth, the projection may have to be
knocked off and the exterior refmished.
Sometimes it is feasible to place beams or channels in grooves cut into
an old wall, as shown in Fig. 15-8(6). When grouted in and tied together
to prevent spreading, they form a strong beam that will prevent local
crushing of the wall at a supported point. It is also possible to place
continuous beams longitudinally under the needle beams of Sketch (a).
This use of a strong continuous beam is helpful in underpinning, but
especially in supporting decking of a street when excavation must be done
while traffic continues. The scheme is illustrated in Fig. 15-9. This is
648 UNDERPINNING [Chap. 15

to show the case of Fig. 15-8 (a) where the general area, inside and outside
of the building, is to be lowered.
After the stringers of Fig. 15-9 are placed under the needle beams, local
pits are dug for supports at A and B. Mud sills and posts are erected,
and the beams are wedged to bearing. The post at C is erected and
wedged next. The intervening earth is now excavated down to some
lower elevation. When the supports at D and E are in place and wedged,
those at A, B, and C are removed, the excavation at these points is carried
down still farther. Once more the supports can be installed at these
three points so that D and E can be deepened still farther. These opera-
tions can be repeated until the excavation is completed without dropping
the structure. new foundation is finished, the temporary
After the
system can be removed. This method is especially useful when deep
excavations must be made and when rock has to be blasted out.

Needle beams
TSfrftTgen J

\ i I
/

c
_LJ_ n

E
FIG. 15-9. Use of continuous beam or beams to permit complete excavation for underpinning a wall.

Figure 15-10 shows the use of inclined shoring. The excavation is first
made along ABCD The braces a are shoved against niches that are cut
V

in the wall. By driving the bottom of a brace inward along timber b, the
shore itself acts like a wedge. Blocks c should be nailed on to prevent
slippage. A floor or some lateral support should be opposite or near the
top of a to prevent tipping the wall over. Small auxiliary needle beams d
are^shmvn also. Their use is to prevent dropping of the bottom portion of
\

the wall. The excavation CEFG and the underpinning can then be
completed.
What is called a "figure -1" is
illustrated in Fig. 15-11. There are many
variations in details. Here is shown the use of a fabricated frame. _The„
installation of sheathing a and the local excavation ABCDEF constitute
thejirst step. The frame is then erected upon sill b and two or three
screw jacks c. The channels/ are slipped under the wall or through holes
in it. Jacking the frame upward lifts the wall on the cantilevered mem-
ber/. The excavation EHGF and the rest of the job are then completed.
Of course, the wall should be supported against the inward force at the
top of the frame.
Art. 15-4] UNDERPINNING OF WALLS 649

Instead of a fabricated frame, brace d may be separate and set directly

on the mud sill. Member / may be hinged to d in some manner. Long
rods with threaded ends, or chains with turnbuckles, may be used instead
of a rigid member for e. Tightening the nuts or the turnbuckles then
serves as a jack to raise / and the wall.
Sometimes it is not possible to have access to the area outside the build-
ing that is to be underpinned. The work is then likely to be even more
costly,but some method can usually be found to perform it. The special
case in Fig. 15-12 illustrates the use of cantilevers. The excavation

Blocking^
9

Wa/f

Beam d
2/sjLe

Dry
*&packing

FIG. 15-10. Inclined shoring of a wal FIG. 15-11. Figure 4 shoring used to hold a wal

ABCDE is made with sheathing a in place. Trenches are then

first,

gouged out along JKBFGH.

The supports b and beams c are placed.
Posts d are erected at the ends of the beams and braced against the struc-
ture so as to hold down the back ends. When the load is transferred to
the jacks, the excavation KLMC
is made and the job is completed simi-

larly to other cases.

An addition was to be made alongside an old wall-bearing building with
heavy brick walls and no basement. The addition was to have a base-
ment, as indicated in Fig. 15-13. Steel sheet piling was first driven out-
side the old wall. A concrete plug was placed between the footing and
650 UNDERPINNING [Chap. 15

Underpinning pier

FIG. 15-12. Use of cantilever supports.

Bricks

Back f/'/led
Yard later
leveh

+-/»- Basement
Masonry
r

1st
Pit
^Footing
Wedges Jacking
Old building wall space
Sheet piling
r Future concrete Wooden 2nd l — ^Concrete
Old shoring^ pit I ! fill
footing
Form
Braces ir
/ Inside sleeve
Pipe in short welded to
sections filled
\A top piece
with concrete
i

New
i

^Retaining
U wall floor
'
|7
_^
WW?

"t> ' P
D
" o • °
-Cap
»

Gravel hardpan
or rock
FIG. 15-13. Use of sheet piling left in place. FIG. 15-14. Underpinning with steel-pipe piles
jacked in short sections.
Art. 15-4] UNDERPINNING OF WALLS 651

this piling, as shown. The wale was bolted on and braced horizontally to
heavy stakes some distance to the right of the picture. Narrow pits 4 ft.
wide were then excavated at intervals so that pieces of the retaining wall
could be built up to the bottom of the old footing. The concrete was
poured directly against the piling. When these blocks were in, the pits
were extended so that the wales could be supported by diagonal braces at
each piece of wall. These supported the piling while the rest of the
excavation and wall were completed. The braces and wales were
removed, and the piling was burned off at the top of the old footing.
The wall top and a closure piece of concrete then completed the embed-
ment of the piling. Such piling must be driven carefully.
Figure 15-14 shows one way of placing new deep permanent foundations
under a building that is settling or is to have greater loads added to it.

Wall

Old fooling
Concrete
fencasemeni
•

J
-
' •

•
|

I > -
-~» g; --^

m Beams b
j.f.'ti;

Wedges c
:'"" '
:

. i

-Piles a
b-i
FIG. 15-15. Permanent bridging to adjacent piles.

The local pits arenow so deep that they are excavated and sheathed in
series as shown. In this case, capped steel pipes were jacked down in
short pieces to form piles that were filled with concrete. The jacks were
placed on top of a pile section, and they shoved against the wall overhead.
A concrete pier was built around and above the pile as an extension
thereof. Wedges and packing forced permanent load into each pile.
The pipe sections in this case were fitted with sleeves but not welded.
Welding may be better where leakage of water is important.
Another method of replacing an old foundation with new piling is indi-
cated in Fig. 15-15. The piles a outside may be driven if a pile driver can
operate there. Those inside will probably have to be jacked down.
Beams b can be inserted one at a time, and then wedged up to bear.
When they are all in place, the whole set can be concreted for protection,
thus forming a sort of pile cap.
In any underpinning problem like those illustrated here, it is important
652 UNDERPINNING [Chap. 15

to bear in mind that the load must be forced "into" the supporting piles
or other members before they can be considered to replace the former
supports. The reason for this is that otherwise the new members will
settle a little before they "pick up" the load; this may cause cracking of
the structure.
In Fig. 15-12 is shown the use of cantilevered needle beams when the

upward reactions at the rear of the needles can be resisted by the weight
of the structure. Figure 15-16 pictures a case where a power plant had to

Jacking frame attached / I A Bricks

—^
\

to needles / U!-J

New floor-
~\ ^-Needle beam

-Mortar Wedges
Mat 1 Threaded Water table -**
dowels — Spreader
-
Old foundation

H
Anchor piles
'Underpinning pile

[a) VERTICAL SECTION

Spreader- ffl
BS
^±7^ -*=?-
-Needle

Kb) PLAN AT NEEDLE BEAM

FIG. 15-16. Example of permanent cantilevered underpinning.

be underpinned (before making an extensive addition) by means of canti-

levered needle beams outside of the structure while maintaining plant
operations. The intake and discharge pipes shown, and other equipment,
prevented any work inside the existing structure. Fine sand, silt, and a
high watertable made it inadvisable to dig pits below the old foundations
or to use wellpoints to lower the ground water. Therefore, the following
procedure was devised, and it is pictured in Fig. 15-16.
1. Jack down two anchor piles, using closed-end, 12-in. steel pipes,
1

then fill with concrete and insert threaded dowels for hold-downs.
1
Driving the piles seemed to be too dangerous because of the effect of vibrations
upon the soil under the existing structure. Jacking was done against a heavily loaded
reactor frame mounted on rollers.
 PROBLEM 653

2. Cut a notch in the concrete of the foundation wall, insert and dry-
pack the needle beam, and attach the end to the hold-down dowels.
3. Repeat this operation for other needle beams approximately 8 ft. on

centers for the lengths of the walls to be underpinned.

4. Construct an overhead jacking frame to be bolted to at least two
adjacent needle beams.
5. Jack down the 12-in. steel-pipe underpinning piles by reacting

against the jacking frame, using two such piles for each needle beam.
Fill piles with concrete.
6. beam spreader across each pair of under-
Place and grout a steel
pinning then apply jacks between spreader and jacking frame.
piles,

With the load thus delivered to each needle beam, drive w edges between T

the needle and spreader so as to maintain the load in the needle beam.
7. Repeat these operations for all needle beams, then encase all steel-

work except the piles in concrete for protection.

Many are the ramifications of underpinning work. Such jobs provide
an excellent opportunity to mechanics, to invent
utilize the principles of
special ways of doing special things, to do them efficiently and economi-
cally, and to meet successfully some of the many challenges of the
profession.

PROBLEM
Figure 15-17 shows the conditions at an industrial plant located close to a large
river. A large concrete pit is to be built to support a new casting wheel that is to
receive molten metal directly from the adjacent furnace. The inside outlines of
the new construction are shown between column lines A and B in 'Sketch (a) .

The elevations of the various levels are shown there as well as in (b). The soil

conditions are also pictured in (6). To be maintained during construction are

service of the overhead cranes, the railroad tracks along line A, and the tracks
outside of line A The last are close to the columns and constitute a
(not shown).
storage yard for cars. How
can the construction be conducted safely?
Two 15-ton cranes operate in this aisle of the building. The service of one can
be utilized occasionally during the day shifts and continuously for the rest of the
time. The furnace and waste-heat boiler will be shut down, but these are very
costly and important pieces of equipment and must not be damaged. The fur-
nace supported by a continuous concrete mat as shown by the hatching. The
is

building is steel frame with corrugated siding and roof and has a concrete floor.
There are small grade beams of concrete between the footings of column line A.
The concrete for the pit is to be protected on the bottom and sides by two-ply
membrane waterproofing.
Suggestions: This problem can be subdivided to make several separate problems.
Consider line A first. in this line, assuming their
Underpin the necessary footings
bottoms to be at El. 100. Next, provide the necessary supports for columns B5
and B9. Then plan the underpinning for B7 and B8, remembering that this work
must permit the construction ofthe shallow portion of the pit next to them. Plan
654 UNDERPINNING [Chap. 15

16' 0'' I6 L 0" J2'-0 l6'-0' I8L0" ,

32-0"

v
Waste heat boiler
K overhead /\

Col.t/neB
jj£0«

^Bottom of a//
footings = El. iOO

15-ton crane
overhead

, Footings „® , .

i i i i I i i I I i

f«J-PLAN OF NEW AND PRESENT STRUCTURE

Present floor
\ El. 106

El 102 \ *%FjlLam
-El. 100
El. 100 Medium:'-
err:
[sandi.y.
'El. 91
n-^mSttk -El. 88 Inside of
El. 84 concrete "1
Coarse sand
El. SO

Dense sandy clay

(deep) ^-SECTION A-A

FIG. 15-17. An underpinning problem at an industrial plant.

the underpinning of the furnace along with that of the adjacent columns, assum-
ing the furnace to be 20 ft. wide. The bottom of the mat under it is at El. 101.5,
the tops of the spread footings.
Now plan the construction of the pit itself. Consider both temporal and per-
manent buoyancy. Assume that the machinery in the circular pit weighs 100
tons, including all accessories and supports.
 16
MACHINERY FOUNDATIONS

16-1. Introduction. The design

of machinery foundations, impor-
tant as it is, is more of an
than a science. There are theoretical con-
art
siderations to be borne in mind, but much helpful information can be
acquired by experience. Of course, the school of experience is excellent,
but the tuition is very high. Therefore, one should try to benefit from
the results obtained by other engineers in the past rather than choose to
learn the hard way by trial and error on one's own. The manufacturers
of various equipment are among those who should be consulted. The
designer should not hesitate to seek their advice.
This chapter attempts to discuss some of the principles to be remem-
bered when designing foundations for machines and to describe some
illustrations of both satisfactory and unsatisfactory results which the
author has encountered.
16-2. General principles. The foundations of machines are gener-
ally prepared in advance so that the equipment can be placed upon them
at a later date. Such foundations should be designed so that the machines
can be erected (often by cranes) and leveled up, the couplings connected,
and the entire equipment with motors and auxiliaries lined up accurately
and put in good working order. The arrangement should also be such
that the machinery in part or in its entirety can be removed, repaired, and
reerected.
Inasmuch as machinery foundations are generally made of concrete, it
should not be expected that these concrete foundations will be constructed
so accurately that the equipment can be erected upon them directly.
Provision for adjustment should be made through grouting and some
"play" in the anchor bolts, the latter being placed in pipe sleeves or even
made removable by having full-length sleeves and pockets for the bottom
nuts and washers, as indicated in Fig. 16-11. In Fig. 16-7, notice that
the grout spaces are partially recessed into the top of the pedestal so that
the grout cannot crack and work loose under the effect of shocks and
vibrations.
655
656 MACHINERY FOUNDATIONS [Chap. 16

Concrete foundations for machines are generally heavy. They should

be erected upon firm, undisturbed soil if possible; otherwise an installation
which is made up of interconnected machines or parts should be founded
upon a common base slab that is sufficiently strong and stiff to cause the
group to settle as a unit (if at all), instead of permitting differential
settlement or unequal tilting, with resultant damage to the equipment.

Of course, the foundations should seldom rest directly on top of the

ground but should have sufficient embedment to develop good vertical
and lateral support. Their tops are often considerably above the floor

FIG. 16-1. Photo of pumping station. (Courtesy of American Lumber and Treating Co.)

surface, serving as pedestals, as pictured for the large ball-mill foundation

of Fig. 16-12.
Perhaps one of the chief sources of trouble with concrete foundations
for machinery is the attempt to use individual high narrow pedestals for
motors and gear boxes, then to use other separate ones for the main
machines. This arrangement may result in misalignment and in vibra-
tions which are out of phase, causing cracking of shafts, failure of cou-
plings, and uneven wear on bearings.
In general, it is desirable to isolate concrete foundations for machinery
from adjacent concrete floor slabs by means of premolded fillers. This
will prevent vibrations of the foundations from being "telegraphed"
through the floor to other parts of the building. It will also prevent
disturbance of the foundation and cracking of the slab because of expan-
Art. 16-2] GENERAL PRINCIPLES 657

sion or shrinkage of the floor. Joints in the floor concrete should be

placed so as to prevent such cracking.
Of course, there are many cases where light machines, conveyor sup-
ports, and minor equipment may be set directly upon the floors (with or
without local thickening of the floor slabs), but these are seldom impor-
tant to the designer. Naturally, no expansion joints in the floor should
come under the machine's frame. Since it is so difficult to set and hold
anchor bolts properly when a large area of floor slab is poured, the use of
expansion bolts is desirable in such instances to hold the machine in place.
The equipment can be aligned after the floor is in service, then the expan-
sion bolts can be spotted, the equipment removed, the expansion sleeves
installed (if care has been used to avoid interference from reinforcing
bars), and the machinery finally erected.
Some features to be investigated and things which it may be necessary
to provide in machinery foundations are the following:
1. Resistance to vertical settlement. It is generally good practice to limit
the bearing pressure on the soil to 40 to 60 per cent of the safe value which
could be used for structural foundations. This will minimize any defor-
mations which may be caused by live loads, shocks, or impulse forces.
Weak clays and silts should not be trusted unless the machinery founda-
tions can spread the load enough to cause a very small unit pressure;
otherwise piles may be necessary.
2. Differential settlement. This can be very destructive of shafting
unless flexible couplings are provided. It can also rupture machine
frames if the distortion is sufficient. In general, the center of .gravity of
the bearing area should coincide as nearly as practicable with the resultant
of the applied loads, including the foundation itself and any live load and
impact forces. The foundation should be stiff enough to have the desired
rigidity, and it should be thick enough so that warping of the concrete due
to moisture or temperature changes which cause unequal deformation of
the top vs. the bottom will not produce appreciable vertical movements
equivalent to differential settlements.
3. Overturning. Many machinery foundations are subjected to over-
turning forces. Belt conveyors are a common example of this. The
head and tail pulleys, as well as the drives shown in Fig. 16-14, can be
subjected to large belt tensions.The foundations of the cable drums of
hoistsand skipways are other examples. Even the ordinary motor can
cause heavy starting torques as well as those existing during ordinary
operation. It is obvious that spreading the foundation to get a large
width in the plane of the torque will be very beneficial, and the edge
pressure applied to the soil by the base should be low enough to avoid
appreciable movement of the center of the shaft unless flexible couplings
are provided.
658 MACHINERY FOUNDATIONS [Chap. 16

4. Torsion. Equipment such as that applying to Fig. 16-24 may cause

large torques in a horizontal plane, tending to twist the foundation.
This may require special means to resist such action, and it will be dis-
cussed more fully in Art. 16-7.
5. Provisions for accessories. It often happens that the foundation of
a machine has to be designed to accommodate various accessories. One
of these maybe the ventilation of motors by natural means as indicated
in Fig. 16-11 or by ducts connected to fans. Another may be trenches
for conduits, so that the conduits can be accessible for future alterations.
In any case, it is good policy to have markers in the floor to locate conduits
for possible future alterations when these conduits are under the floor or
buried in the concrete. may also be quite com-
Piping of various kinds
plicated, and neighboring pits or trenches as pictured in Fig. 16-1 can
affect depths and other dimensions which have to be met. There may
be need for supporting racks, benches, conveyors feeding materials to the
machine or taking them away, and many other such auxiliaries.
6. Expansion. Sources of heat can cause expansions which produce
deformations of the foundation, the equipment on it, or both. If the
frame machine becomes much hotter than its foundation, this
of a long
may cause rupturing of one or the other.Although furnaces and refrig-
erated spaces may not be machines, their foundations can also be impor-
tant, and ventilation of the foundation or provision for expansion of the
equipment may be necessary.
7. Vibration. This problem can be very troublesome. It will be dis-
cussed in Art. 16-3.
8. Protection. One of the things to be guarded against is corrosion of
machine frames and anchor bolts. One way to help avoid this is to raise
the concrete above the floor, as in Fig. 16-1, so as to prevent wash water
or drip from directly wetting the metal. Machines that are out of doors
should have the concrete high enough to keep the equipment above the
level at which water may collect and stand. In some cases the founda-
tions should be extended so as to provide for the attachment of weather-
proof housings. There should also be guard frames to protect personnel.
16-3. Vibration. A drop hammer for forging is an obvious illus-
tration of a machine which causes a heavy shock to be applied to its
foundation. This may be partially transmitted through the ground or
structure to other things in the immediate vicinity. However, such
equipment is seldom operated rapidly and continuously enough to cause
neighboring objects to vibrate harmfully, and the shock may be cushioned.
It is apparent that such a machine founded upon rock probably will not
cause enough impact to set the rock in motion. The sound, however,
may be annoying.
Harmful vibrations in machinery foundations generally occur with
Art. 16-3] VIBRATION 659

high-speed machines having reciprocating or rotating parts. Although

equipment like reciprocating pumps may be designed so that the forces
are theoretically self-contained, this ideal may not be perfectly attained,
or pulsating forces may be applied through the connected piping.
In
such a case, the designer should ascertain the magnitudes of the forces
and their period or frequency. He should then endeavor to make sure
that the supporting structure will not have a natural frequency close to
that of the machine and that resonance will be avoided. Rotating equip-
ment, such as a high-speed fan, may have the center of gravity of the
rotating part slightly eccentric, or the bearings may wear enough to
cause a slight "play," both of which actions produce a centrifugal force
as the machine rotates.
It seems that if the frequency of the impulses is very large (perhaps
over 1,000 cycles per min.), the supporting structure is likely to have a
natural frequency so much lower than this figure that the two will not
vibrate in harmony, i.e., be in resonance. Probably the danger zone
usually lies in the range of 200 to 500 cycles per min. However, although
the frequency of the machine can be stated by the manufacturer, it is

likely to be very difficult to make an accurate estimate of the natural

period of vibration of the foundation or other supporting structure. It is
desirable to have the latter slow enough never to exceed 30 to 50 per cent
of the frequency of the machine, or so high as to exceed this frequency by
200 to 300 per cent. Of course, when the motor driving such equipment
starts to increase the speed of the machine from rest, there will be a time
when the frequency of the impulses may equal that of the supports, but
this is likely to be so transitory that its effects will be negligible. How-
ever, variable-speed motors may happen to run sufficiently slowly to
cause troublesome vibrations; hence, their foundations should be designed
for the worst (slowest) case.
A few fundamentals to bear in mind regarding vibration of foundations
are the following:
1. Increasing the weight of the foundation decreases the rate of change
of motion caused by a given impulse. This is obvious from the formula
for plane motion,

F = Ma (16-1)

whereF is the applied force in pounds; M, the mass (W/g) of the weight
W; and a, the acceleration in feet per second per second, g is the acceler-
ation due to gravity — 32.2 ft. per sec.
2
at sea level. Also,

where v is the resultant velocity in feet per second; t, time in seconds; and
660 MACHINERY FOUNDATIONS [Chap. 16

s, distance in feet. Therefore, when the applied force is a reciprocating

one causing plane motion, an increase of the weight will reduce the deflec-
tion and vibration movement. The determination of the necessary
magnitude of the weight to reduce the amplitude of the vibration suffi-
ciently is a difficult matter. One of the things which the author prefers
to do as a starting point in solving this problem is to obtain the recom-
mendations of the manufacturer of any particular equipment, because
he is probably best informed, through past experience with his own
machines, and he is therefore best able to determine what weight to use
under a given condition. This course is better than following some
empirical rule stating the weight to use per horsepower, per cycle, or per
weight of moving parts.
2. Considering the kinetic energy (K.E.) for plane motion,

K.E. = ^j- (16-3)

which means simply that, for a given applied energy, greater mass reduces
the velocity produced. Similarly, an impulse Ft produces a change of
momentum. Thus,
Ft = M(v - v ) (16-4)

where v is the initial velocity prior to the application of the impulse, and
v is the final velocity of the body afterward.
In the case of rotary motion, a mass whose center of gravity is a dis-
3.

tance r away from the axis of rotation will cause an outward, radial
(centrifugal) force F n Then, .

Fn = —
Mv
r
2

(16-5)

It is obvious that if the foundation is strongly supported vertically, as

pictured in Fig. 16-2 (a), the force F„ will tend to lift up the machine when

the center of gravity b of the rotating part above the axis of rotation 0,
is

and it will cause only a downward pressure when b is directly below 0.

However, as b rotates clockwise, a component of F n becomes horizontal,
and this increases to the full value of F n when b is horizontally to the right
of 0, as shown in Sketch (6). At this point there is an overturning
moment about point C of the base. Then e = FJi applied to the bot- M
tom AB. Neglecting any lateral support from the earth alongside, this
causes an edge pressure p which may be computed for a rectangular base
as

v = ^jg aM)
Art. 16-3] VIBRATION 661

Of course, this assumes that — p cannot annul the downward pressure

caused by the dead load. As the machine rotates, the force F n alter-
nating from right to left and back again will cause a rocking tendency
somewhat similar to that which would occur if the force were applied
horizontally by a reciprocating action. This is in addition to the tend-
ency to produce the up-and-down motions previously referred to. Of
course, friction on A B is assumed to keep the foundation from sliding
horizontally.
Letting Fig. 16-2(6) represent a case in which the horizontal force can
be applied by either a rotary or reciprocating motion, one can see that the
bending in the footing AB may cause yielding and curving of this member,
with a resultant horizontal movement of point 0. The alternation of
this bending will cause horizontal vibration of 0; hence, the footing AB
and pedestal D should be very stiff as well as heavy. Furthermore, the

-<£_ of axis

[a) [b)

FIG. 16-2. Illustration of some principles of vibration.

pressure +p will cause the soil or other supporting structure to compress

or deflect near B while — p may produce a rebound at A. Such deforma-
tions alone can result in a rocking of the base about C and a vibration of
0. These motions combine with any caused by the applied vertical
will
impulses. Therefore, the soil modulus and the deformation of the ground
can be important, and so can E and the deflection of any other supporting
structure. A long thin footing on springy soil or flexible beams, under
the action of this alternation of bending of the concrete and yielding of
the support, can accentuate the vibrations.
Very weak, compressible soils under machines can lead to harmful
results in another they actually deform permanently near A and B
way ; if

of Fig. 16-2(6), the rocking about C may become excessive. Therefore,

good stiff soils or some equivalent is essential if vibration is to be mini-
mized. Grouting the soil with portland cement or chemicals can be used
to stiffen the soil if the latter is sufficiently permeable. However, this
662 MACHINERY FOUNDATIONS [Chap. 16

quality would mean that the material is granular and probably fairly good
compared with nongroutable silts and clays.
Considering the same sort of uplift forces at A and B as indicated in
Fig. 16-2(6), one can see that if there is a strong uplift, the anchor bolts
should be so large that their elongation under stress will be negligible; in
fact, considerable prestressing to take up the slack may be desirable, or
the use of double nuts may be essential to prevent loosening under the
action of pulsating forces.
It is extremely difficult to determine the natural frequency of vibration
of a rather simple body; it is much more so in the case of a complicated

F 1
L

U) SIMPLY SUPPORTED BEAM [6] FIXED-END BEAM

Lowest natural frequency = w\ — Lowest natural frequency = 22.3733

L Spring -

[c) CANTILEVERED BEAM ^77777777777

Lowest natural frequency = 3.51602 /-rrt id) BODY ON A SPRING

V ML
1 K_
Natural frequency
Note: E = Young 's modulus ZlT \f L
I = moment of inert ion where K- spring constant.
of the cross-section Unit is Ib./in.
M= ib.x sec. z/in.

FIG. 16-3. Some formulas for computing natural frequency of a body. (Courfesy of Dr. Shan S.
Kuo, Department of Civil Engineering, Tufts College, Medford, Mass.)

structure. Figure 16-3 shows some formulas for computing the lowest
natural frequency of certain bodies, with the character of support indi-
cated in each case. These may be useful in a study of instances in which
machines are supported upon some kind of superstructure.
Consider again the fact that friction on the base A B of Fig. 16-2(6) will
not let the bottom slide. Therefore, F n causes a tendency to rotate the
machine and base about C. For rotary motion,

Angular impulse = Ftr = Jp(co — co ) (16-7)

where I p is the polar moment of inertia about the axis of rotation ; co, the
finalangular velocity in radians per second; and co,„ the initial angular
velocity in radians per second. Since I p depends upon the mass and
dimensions of the body and the distance of the center of gravity of this
Art. 16-3] VIBRATION 663

mass from the axis of rotation, it is again obvious that, for a given impulse,
greater mass reduces the angular velocity and the motion created; hence,
it decreases the amplitude of the vibrations impressed upon the foundation.
One way problem of preventing the transmission of vibra-
to attack the
tions to theground or to a structure is by the use of insulators or some
other device that will permit the machine to vibrate but so cushion the
reactions that the forces applied to the structure will not be severe. From
Eq. (16-7) and the formula for plane motion,

Impulse = Ft (16-8)

it is apparent that anything which increases the time t of application of the

force F will cause a reduction of the impulse, and therefore of its effects.

However, the insulators do not change t, because this depends upon the

F/oor.

Building

FIG. 16-4. Types of vibration insulators.

machine, but the inertia of the body will resist both translation and rota-
tion so that, move, the body will start to go one way, then its
if free to
motion will be reversed as the force changes direction. Thus, even when
supported by springs as in Fig. 16-3(rf), the body's inertia resists a change
of motion first one way and then the other as the alternating forces act
upon it. By permitting some of this movement through yielding of the
supports, the insulators can greatly reduce the magnitudes of the pulsating
reactions, and hence of the forces affecting the foundation.
Some principles used to accomplish this cushioning are shown in Fig.
16-4. Referring to the sketches, notice the following:
This shows the use of a compressible material which deforms under
(a)

a variation of the pressure. Of course, it must always support the dead

load. The device must be strong enough to prevent squeezing out or
crushing of the material, yet springy enough to yield sufficiently to
664 MACHINERY FOUNDATIONS [Chap. 16

reduce greatly the impulse applied to the foundation. Since the area in
this case is so large, the unit pressure on and the deformation of the cush-
ioning material will be small, and therefore the "damping" effect will be
small also. Cork is useful for this purpose, but it may crush easily or
disintegrate if it becomes wet. Asbestos fiber, natural rubber, or one of
the appropriate synthetic compounds may be satisfactory.
(6) This sketch shows the use of localized cushions under special
bearings, causing high pressures and large deformations under the action
of the alternating forces. The damping effect is therefore greater than
for (a). Natural and synthetic rubber are among the materials used in
such insulators.
uu
Excellent

80 Good

Fair

60

40

20 Poor

None
4 6 10
Insulation ratio

FIG. 16-4A. Absorption of vibration. (Based on data of the United States Rubber Co. for U.S. Royal
Rubber Mountings and Structural Rubber.)

(c) This sketch shows a bearing in which the impulse causes a shearing
force and a shearing deformation. A strong yet resilient rubber or plastic
is needed for this use.
id) This arrangement is the obvious use of steel springs to support the
machine. Of course, some lateral steadying of the equipment is necessary.
(e) This sketch pictures the use of a primary concrete mat, or footing,

for the machine. This mat is supported upon brackets resting on insula-
tors. The insulators bear on a curb that is part of a secondary concrete
mat which is resting on a cushion as well as having a yieldable material
around the edges inside of a structural curb. Under the cushion and curb
is the main structural foundation. This scheme utilizes both mass and
insulators to reduce the forces applied to the main structure, and it is
especially useful when compressors or large fans have to be supported
high up in a building.
Art. 16-3] VIBRATION 665

The manufacturer should be consulted in connection with the use of

any one type of insulator in order to make sure that the equipment will
be satisfactory. The Mechanical Goods Division of the United States
Rubber Company, in advertising its U.S. Royal Rubber Mountings and
Structural Rubber, has stated that vibration insulation results may be
somewhat as indicated by the curve in Fig. 16-4.A.
All piping, ducts, and other stiff members connected to the insulator-
mounted machine and the stationary structure should have flexible joints
or expansion devices to prevent the vibratory motion of the machine from
being transmitted through these connected parts to the building.
A simplified illustration of what impressed vibrations can do is the case
of the brackets pictured in principle in Fig. 16-5. These were on the end

[a) SIDE ELEVATION [b] PLAN

FIG. 16-5. A bracket that failed.

of the foundation of a turbogenerator. Bars a were hairpins bonded into

the concrete of the main foundation. The vibrations of the latter pro-
duced a sort of "whipping" of the bracket, which eventually caused a
failure of the bond, so that the bars pulled out and the bracket fell. The
trouble evidently came from by repeated application of
fatigue caused
varying tensions. Bars a should have had some type of positive anchor,
such as embedded plates, to which they were strongly attached as done —
foranchor bolts. Bars b did little good. A few inclined stirrups c would
have been helpful to prevent shearing of the keys.
Figure 16-6 shows some simplified illustrations of cases which have pro-
duced unsatisfactory results because of vibration of the building. Refer-
ring to the sketches, the stories are the following:
(a) This shows the general type of the construction used for a crushing
plant foundedupon a mat on rather soft soil. It was built of concrete
and wood because of the shortage of steel during the Second World War.
666 MACHINERY FOUNDATIONS [Chap. 16

The top crusher a was supported upon two heavy piers b extending to the
mat and tied into the concrete floors. The crane columns c were made of
reinforced concrete and were very strong. The roof was made of wood
with posts / on top of columns c. When the crusher was operated (espe-
cially when empty), the roof shook so badly that it frightened the oper-
ators. The tiny vibrations at e were magnified by the cantilever
(whipping) action of the concrete columns. The crusher had to be
changed so as to support on a cushion. it

(6) This pictures a case in which each of two crushers g was mounted
on a heavy concrete pedestal about 18 ft. wide and over 20 ft. long, bearing
on rock. The electrical control room and change room were alongside at

l
—

b-^ ^-b

[a)

FIG. 16-6. Some illustrations of vibrations of structures.

h. In order to economize on the steelwork, steel columns j were sup-

ported on top of the pedestal. Operation of the crusher made these
columns vibrate and caused annoying vibrations and "chattering" of the
roof. Therefore, a 12-in. concrete buttress k was built and doweled and
keyed into the pedestal. It greatly stiffened the pedestal and remedied
the difficulty.
It is apparent, from the action of the cases just described, that it is not
desirable to place the columns of a building (or the beams if fixed or
rigidly attached) on a machinery foundation which may vibrate. It is
better to isolate the building from the equipment foundation or to place
any beams on rocker bearings if they must be supported by the machinery
foundation.
Concrete foundations on soil. Most foundations for heavy
16-4.
machines in modern mill-type structures are made of concrete and sup-
Art. 16-4] CONCRETE FOUNDATIONS ON SOIL 667

ported directly on the ground. A few types and ideas regarding their
design will be discussed. These should establish some of the principles to
be borne in mind when planning such foundations. The designer should
plan with due consideration for vibrations and deformations rather than
for vertical pressures and the safety factor against overturning alone.
The steel reinforcement in some machinery foundations may be light,
or even negligible; in others, heavy reinforcement may be necessary. A
few typical examples are the following:
1. A pedestal-type foundation The anchor
is pictured in Fig. 16-7.
bolts are rested on the footing during the pouring The of the pedestal.
bars a need be sufficient only for transmitting the anchor-bolt pull and
any overturning forces into the footing. The hoops b near the top are to
prevent the formation of any crack across the top where the pipe sleeves

Grout
v^
Recess-
WW+ -b
Pipe sleeve
rr-

HM ±jS,H*
W-
ii
I!iii
hi
"Hf,

-Premoulded
~*i mm filler
I

T Y>::\^F/oor
I !l
I

n
V— Anchor bolts
m^ slob

4=
-Footing
I I
,-t

FIG. 16-7. A pedestal-type foundation.

weaken the section. Bars b also serve to hold a in place during construc-
tion when wired to the latter. Bars c are nominal in size and are used to
tie both ends of the foundation together. No intermediate bars are
needed in this case. It is advisable to place the bars near the top and
bottom so that a crack cannot get started. If the pedestal were short
enough, the anchor bolts could extend down into the lower portion of the
footing, thereby making bars a unnecessary. However, it may be diffi-
cult to hold the anchor bolts in the right position at this stage of the work.
2. A heavier, two-pedestal foundation, such as might be used for a

centrifugal fan, is pictured in Fig. 16-8. In this case, the heavy footing
is placed at the floor line. 1
The pedestals are reinforced as cantilevers
1
A disadvantage of this isthe finishing of the top of the footing to match the floor
in position and texture. It is generally preferable to complete the heavy concrete
work first, then, after all piping and electrical work is done, to pour and finish the
floor up to the pedestals as in Fig. 16-7.
668 MACHINERY FOUNDATIONS [Chap. 16

projecting from the base. Here the reinforcement is important. It

would be possible to use two separate footings terminating at AB and CD,
but it is clear that two slender, isolated pedestals would not be desirable
in this case. The equipment here is assumed to be set on a neat-cement

Pedestal

Painted
joint-.
-
Floor line

.'
• "-

D
FIG. 16-8. Foundation for a centrifugal fan.

rt i — U
-\-r\ — i— r—
ill
-\-\-\

—
—1—|-H —
L__l_

-| 1-

_U_I__L_L.
A
_4_L-L — M--4 — J_T1
12-4

(a) PLAN OF TOP (c) DETAIL AT RECESS

'Prennoulded filler

naftttTT Anchor bolts and conduits

Mill!
mj_L_4-i-
not shown

Kb) SIDE ELEVATION

FIG. 16-9. Details of a pump foundation.

paste on top of the pedestals instead of on 1}4 to 2 in. of grout or dry-

packed mortar as is customary. Incidentally, notice the sloping faces of
the outsides of the pedestals. This is easily done for one or for two
opposite faces, but sloping all four sides causes more costly formwork,
although the appearance may be better than with vertical sides.
Art. 16-4] CONCRETE FOUNDATIONS ON SOIL 669

3. Figure 16-9 pictures a heavy mat which might be used under a pump
or compressor. It is assumed that notches A and B are necessary, but
they cause points of relative weakness. If bars a and b are cut and bent
as in Sketch (c), one can see that any tensile forces T tend to split the
concrete. Therefore, extra bars c should be used to strengthen such a
weak spot.
4. When machinery foundations are very long, as for a 125-ft. draw-
bench, either the whole foundation should be tied together thoroughly
with more than the usual shrinkage reinforcement (A s = 0.0025bd), or else
it should be separated into two or more self-sufficient parts. This idea is
illustrated in Fig. 16-10. The heavy ends A and B tend to act like
anchors which will not let the foundation shrink, and the local enlarge-
ments C, D, and E do so to a lesser extent. Unless made very strong, a
crack is likely to occur in the concrete at some weak spot. It would be
better to put contraction joints at least at points F and G. If there were

a full-length thrust or tension in the foundation due to the machine, it

might be advisable to use more concrete and to provide a continuous flat

^X~Tn~w-w M ^u u "ru^r
"y £
v-'^
F c D EG
FIG. 16-10. Foundation for a drawbench.

bottom as shown by the dashed line between A and B, using a fair amount
of reinforcement.
5. Figure 16-11 is a simplified picture of a proposed foundation for a
large rod mill for grinding ore. The engineers wanted to have the mill
high enough above the floor so that lift trucks could operate under the
steel working platform for cleaning up spillage. That explains the great
height of the pedestals. Nevertheless, it was necessary to make sure that
the foundation was stiff enough to prevent harmful relative displacement
of the mill bearings, gear reducer, and motor.
In a somewhat similar but smaller installation it had been found that
tall, slender cantilevers were inadvisable for supporting such equipment.
In that particular case, it seems that, for one thing, there was a thrust
between the mill and the gear-reducer pinion sufficient to cause the gear-
reducer pedestal to be tilted slightly away from the mill, causing failure
of the gears to mesh properly. It was important to avoid such troubles,
and since the plant was to be located in Chile, resistance to the effects of
earthquakes also had to be provided for. The mill itself, with its con-
tents, weighed 400 kips; the motor, 30 kips; and the gear reducer, 10 kips.
To meet the requirements discussed in the preceding paragraph, the
design pictured in Fig. 16-11 was developed. Notice the following
features:
670
Art. 16-4] CONCRETE FOUNDATIONS ON SOIL 671

a. The main pedestals A and B are made quite wide and are braced by
the strong portal C.
b. The motor pedestal D is made an integral part of B so that both
parts buttress each other.
c. The motor pedestal E is a heavy block with a hollcw top to permit
ventilation (cooling) of the motor. It is joined to B by portal F.
d. All pedestals are keyed and strongly reinforced into footing G so as
to act as cantilevered piers.
e. The anchor bolts have pipe sleeves and pockets so that the bolts can

be removed.

FIG. 16-12. Foundations for ball mills, Morenci Reduction Works, Morenci, Ariz. (Courtesy of Phelps
Dodge Corp., New York.)

f. The footing G is made with beveled sides instead of reentrant corners

so as to avoid local points of weakness. The reinforcement in top and
bottom is designed as for a combined footing under all' pedestals.
A smaller but somewhat similar series of ball-mill foundations is shown
in Fig. 16-12.
6. Figure 16-13 illustrates some of the possible means of avoiding heavy
pressures produced by machinery foundations on walls nearby. In this
case the electrical tunnel shown in Sketch (a) had already been built. It
had shelves for supporting an "apron" floor slab like that indicated on
the left side near A.
This construction relieved the tunnel walls of large
surcharge pressures, and it avoided the possibility of settlement of the

floor slab next to the tunnel if the backfill consolidated. The pressure
diagram in Sketch (b) shows that the only active earth pressure against
the wall might be represented by the triangle BTP, but if the floor slab
were not seated on the wall, a surcharge pressure somewhat as shown by
BSUT would be added.
672 MACHINERY FOUNDATIONS [Chap. 16

When the design was made, a heavy machine was to be located approxi-
mately 10 ft. to the right of the tunnel so that it would be too far away to
affect the wall BP. However, after construction of the tunnel, the
mechanical layout was revised so that the edge of the machine foundation
was only 8 in. from the wall. The following questions had to be
investigated
a. Would the pressure of 1,500 p.s.f. under the machine's foundation,
if located as shown in Sketch (a), cause excessive pressure against wall
BP?
b. Would the relative weakness of the backfill near the wall, compared
with the firm, undisturbed soil beyond the excavation lines, cause the
concrete of the machine foundation to "break its back" somewhere
between points C and D, or at least tilt the machine?

jLFloor line % -Machine

foundation

A<
\
'C,
r E

^Electrical
/V
o W ,

tunnel

._ M _
L
h-
/
/
f
~G
/? Y
J
U P X P
•«'W"'<V wjj.v/'

(a) SECTION THROUGH TUNNEL (6) ic)

FIG. 16-13. Machine foundation alongside an electrical tunnel.

analysis of the wall BP revealed that both the concrete and the
An
steel would be seriously overloaded. Therefore, the surcharge must be
kept from causing dangerous lateral pressure against the wall. This con-
clusion was arrived at by assuming that the pressure under CD would not
be fully effective in producing lateral loading beyond (to the left of) the
line CQ drawn at a slope of 30° with the vertical. The assumed pressure
diagram on the wall is pictured in Sketch (c), where point V is level with
C, point W
is opposite Q, and QW = PX = 500 p.s.f. (1,500/3).

One possible remedy was to seat the edge of the machine foundation on
the shelf at B, then make this concrete span across to the undisturbed
soil. The width of the shelf and the thickness of concrete above point B
were inadequate for supporting such a heavy load, and the floor of the
tunnel was not strong enough to serve as a footing for the anticipated
pressure.
Art. 16-4] CONCRETE FOUNDATIONS ON SOIL 673

Another remedy to consider was the construction of a wall EFHG and a

footingJK to support one side of the machine foundation as pictured in
Sketch (a). Notice that (1) this footing would have to be almost as deep
as the tunnel in order to reach undisturbed soil; (2) the bearing pressure
under the footing would be very great unless the latter were made very
large; (3) the new excavation lines would extend still further in the direc-
tion of point D; and (4) the concrete CD would have to bridge a consider-
able distance from EF to a firm support on the soil.
The solution finally adopted was based upon the simple reasoning that
"if something causes trouble, remove it." Therefore, the concrete of the
machine foundation was merely thickened to bring its bottom to point L,
Sketch (a), producing a lateral pressure on BP somewhat as pictured by
YZP in Sketch (d). The bottom was sloped slightly upward as indicated
by LM. This required the use of considerable concrete, but its extra cost
was mainly for the material alone and very little formwork was required.
Because of the thickness, a moderate amount of reinforcement along the
bottom was sufficient. 1

7. Figure 16-14 is a greatly condensed and simplified drawing of a

tandem drive for two big conveyor belts at a metallurgical plant. The
head pulley a must resist a large belt pull. The belt then passes over
drive pulleys b and c, pulley d, the take-up e, the pulley /, and then over
the return idlers at the right. The other belt at the left is similarly
passed over the tail pulley g and the take-up pulleys h, i, and j and then
drives k and m. From an examination of the arrows picturing the
character of the forces applied to the pulleys and their supports, one can
see that the various foundations are subjected to large overturning forces
— in the order of 100 kips for some of them.
If the equipment is founded on rock, the pits for the take-ups can be
mined out and all foundations placed on rock, and no trouble will be
likely. However, if all must be on sand or gravel, the excavations for the
take-ups will cause adjacent soil to slump into the pits so that backfill will
later have to hold the upper foundations. Even with sheet piling driven
around the pits in advance, parts of the foundations will have to rest on
ground while others are on the pit walls.
The parts which need firm supports and should be kept in proper rela-
tion to each other are the motors and drive pulleys. Each set should be
mounted on a common, rigid foundation. The head and tail pulleys also
must have a strong base to resist overturning and shear.
One way to handle this problem if on sand is to excavate for the pits
and for walls extending from them to the vicinity of E and H or to build
Unless the machine foundation is very wide and long, the usual shrinkage rein-
1

forcement (A s = 0.0025bd) is not necessary since the base is primarily an integral

block.
 HEI

674
Art. 16-4] CONCRETE FOUNDATIONS ON SOIL 675

A A
D
B B

(a) PLAN [b\ PLAN

A A B B

PLAN PLAN

6'

H H H H
ELEVATION A-A ELEVATION BS
[c) id)

C D I i D
t_
Li l
PLAN J I PLAN J
1

1
n n^ —
-i

i
a a.
K M

P
ELEVATION C-C ELEVATION D-D
le)

FIG. 16-15. Illustrations of possible weak spots

676 MACHINERY FOUNDATIONS [Chap. 16

separate piers at these points. Then a strong slab can be supported on a

shelf at D
with a simply supported edge along LE and spanning onto the
ground near end JK. Even though the slab CD tilts, the motors, speed
reducers, and the drive pulleys cannot get out of line. A strong slab can
also span between seats at A and B. A wall or pier under HO will pro-
vide a support for this end. Additional piers should be under the motor
at G. The main slab can span from ON to PM
with a wing to GH. Use
a four-legged portal frame QRST of reinforced concrete as shown in
Sketch (a) to hold pulley a. A shallow pedestal will hold g. Finally, the
drive pulleys can be mounted on the sloping faces of two strong pedestals
as shown, each bearing on its respective slab.
8. Figure 16-15 pictures some possible sources of weakness in concrete
foundations on soil, especially when subjected to shocks and vibrations.

Keeping mind that reinforced concrete is not a good material to resist

in
fatigue, and also that the cost of the concrete in the foundation is not
great compared with that of remedying unsatisfactory results, the follow-
ing comments may be worth considering in connection with the respective
sketches
(a) This foundation is relatively weak across AB and CD. It might
be better to make it more like that shown in (6).
(c) Here the central pit causes weakness at EF and GH unless these are

made thick and strong. One alternative might be to thicken G'H' and to
use only two heavy pedestals on the footing as shown in (d).

(e) In this case there are weak sections at JK and LM. If the soils are
not good, the section might well be designed as indicated in (/). Here the
surfaces NO and PQ are sloped so that the foundation acts somewhat as a
rigid boat.
16-5. Concrete foundations on piles. When heavy machines must
be founded upon weak clays and silts, especially when they cause shocks
and overturning forces, one way to support the foundations is to make the
latter so big in bearing area that the pressure on the soil is too small to
cause harmful settlement. A better way often is the use of piles under
the foundations.
For example, Fig. 16-16 pictures the foundation required for a large
complicated machine. The equipment is heavy, but no serious shocks or
overturning need be provided for. The soil is silt, so that friction piles
have been selected to support the foundation. How should they and the
concrete work be designed? Here are some principles to bear in mind.
1. The center of gravity of the pile group should coincide with the

resultant of all the applied weights as closely as practicable.

2. The foundation should be cut loose from the surrounding floor unless
the latter is also supported on piles.
 9-9 N0I133S [0]

';.

] f
|,;*
___v &-?£> •
, . . . 1

l£J

(•+) (+) r+;

(if ®

«- o

03
h; i+: r+ji
i^
:
l
!
e 03

«
oa

677
678 MACHINERY FOUNDATIONS [Chap. 16

3. In this case, the entire bottom should be tied together and reinforced
like a thick slab.
4. The various walls should be tied together well by reinforcement near
their tops so as to prevent local cracking, and a few dowels should tie
these walls to the base.
5. If pressures may be applied to outside walls, the walls should be
reinforced to resist them, like cantilevered beams.
6. Projecting parts like ABDC, PQRMN, and GHJK can be supported
directly on piles so that differences of the levels of the bottoms need not
require the thickening shown in Fig. 16-15(/), because the piles can be
cut off at different levels. However, be sure to tie these parts into the
main base, MREFL, at both top and bottom.

FIG. 16-17. Head pulley and drive for a conveyor. [Courtesy of Phelps Dodge Corp., New York.)

7. Try to arrange the piles under projecting parts so that tipping and
twisting of the junction will be avoided.
8. happens that the vertical loads could be supported by a few
It often
very strong However, to get proper stability and support for all
piles.
parts, it may be desirable to use more piles but not such long, strong ones.
The arrangement of the piles finally adopted may be as shown in Fig.
16-16(a).
When vibrations and shocks are to be resisted, end-bearing piles will
almost always be satisfactory. This may not be so in the case of friction
piles in clays and silts. In such instances, one should be conservative
and base his judgment upon the severity and importance of the particular
case. Probably one should not design such friction piles for more than
50 per cent of the safe load which could be used on them if under static
loading. The pile pattern should be arranged so as to give the maximum
practicable moment of inertia for the group. The foundation should be
Art. 16-6] CONCRETE FOUNDATIONS ON SUPERSTRUCTURE 679

stiff enough to act as a fairly rigid body. This may be difficult to do for
large groups of piles.
When shears and overturning forces are to be resisted along with the
vertical ones, the piles should be located so as to give the maximum feasi-
ble resisting moment. The methods described in Chap. 10 can be applied
to the analysis.
16-6. Concrete foundations on superstructure. It is assumed in
this article that even though supported by some kind of superstructure,
concrete will be used for the immediate support of the machine. Steel
can be used directly in some instances without intervening concrete, as
shown in the erection picture in Fig. 16-17. However, more troublesome
cases will be considered here.
The first is a case where the machine is not on a steel frame but is still
raised above the floor. Figure 16-18 pictures the foundation for a large 1

blower to furnish compressed air for the convertors of a smelter. The

blower is located in the power plant and is run by a steam turbine, the
air being piped to the smelter building. As can be readily seen, the
foundation is not simple. Notice the following features in Fig. 16-18:
1 The blower and turbine rotate about an axis which is parallel to the
longitudinal axis of the foundation but above the table. Therefore two
end walls are used from top to bottom in order to provide strength and
stiffness in the direction in which any centrifugal forces may act.
2. The end walls are punctured by holes as shown in (b) in order to

accommodate piping, etc., but with proper reinforcement, the walls are
still very stiff.

3. Portal action of the frame is depended upon to provide longitudinal

stiffness, hence the use of the large fillets at the top as pictured in (c).
This also gives access to parts below the table.
4. A heavy mat is used for both weight and stiffness even though this

particular foundation is placed on moderately good andesitic rock.

5. The platform at the top is at the main working floor of the turbine

room; its edges merely rest on a shelf, but are not fastened to the table.
6. The height of the pedestal is determined by the platform level and

the basement floor, which, in turn, are made to suit the adjacent turbo-
generator.
7. The table is composed of heavy beams with as much connecting slab
as clearances will permit.
8. The is reinforced very strongly.
entire pedestal Since the possible
starting-up torque may
be large, and since the forces causing vibrations
are not too clearly known, the design of the reinforcement is largely a
matter of using good judgment and conservatism. All junctions of mem-
1
Sometimes such foundations are called pedestals, and the top may be called a table,
i.e., a flat top to hold the machine.
680 MACHINERY FOUNDATIONS [Chap. 16

Floor

%—Mat

Kb) END ELEVATION A-A

[a) PLAN AT TOP (TABLE)

Kd) PARTIAL SECT. C-C

Ke) PARTIAL SECT. D-D

\f) PARTIAL SECT. E-E

Kc) SIDE ELEVATION B-B

FIG. 16-18. Foundation for a converter air blower.

bers should be well tied together, and the theoretical stresses in the bars
probably should not exceed 8,000 to 10,000 p.s.i.
When heavy machinery is supported on beams, special problems may
arise because of the deflections of the beams, and these problems may be
in addition to those produced by the vibrations caused by the operation
Art. 16-6] CONCRETE FOUNDATIONS ON SUPERSTRUCTURE 681

of the equipment. For example, Fig. 16-19 pictures one such situation,
where a motor was to be direct-connected to a heavy rock crusher. The
motor was to be supported by beams JK and LM, whereas the crusher
was to rest upon beams AB and EF, as shown in Sketch (a). Imagine
that the direction of rotation of the motor is such that the reactions on its
supports cause a downward force on and an upward force on JK. LM

~kTT
- Column
I

I
— s!rf-
YW~~m
Hatch Hatch
\Motor\ Motor

V
i
I

I
I

I
-I

I Pilaster]

K- IX. ._LL__i/l_J L\L__ -1 I

_LJ
I
I

i
I ,J_
if/-,
i 0\ ,jg l

\A C D B\ A B ff\
Column - I

Crusher Crusher I

: I
G H I

L !

Column
-y A
~o,
\-
Pilaster

I
i

Stairway Da
3J LI
Stairway On.
3 _l II J

{a) PLAN (b) PLAN

""W W "=13 W
Q
Pilaster^
- Column

"
''
::
*-'2*-Q" 2-0
\A 2
''

XA?- :>
'- 4
%1 r'-io" 12'-0" 4'-6""'kl-4

[c) SECT. A-A (d) SECT. B-B

FIG. 16-19. Study of crusher and motor supports.

The result is a slight rotation of the entire motor about a horizontal axis
in or close to the plane of the tops of the beams. This angle multiplied
by the distance from the axis of rotation to the shaft of the motor causes 1

a horizontal movement of the motor shaft with respect to that of the

crusher. For convenience, assume that this movement is }-$ in. Such
a deflection will cause the shafts to bend a little, but since the shafts are
1
The motor rests upon a tall pedestal which greatly increases this lever arm.
682 MACHINERY FOUNDATIONS [Chap. 16

short and stiff, the resultant bending stresses may be appreciable. As

the shafts rotate, the bending distortion remains in the same direction
but the stresses in any given extreme fiber vary and reverse rapidly.
This may cause fatigue failure of the shafting or of the coupling unless a
suitable flexible coupling is used to join the shafts.
Besides the motions of the motor supports one must consider the effect
of the deflections of the members supporting the crusher. Furthermore,
such rotating machinery may cause vibrations which aggravate the harm-
ful effects. Under such conditions, it is desirable to minimize the deflec-
tions of the supports, and hence the bending in the shafts. In the particu-
lar case illustrated in Figs. 16-19(a) and (c), the framing was changed as
shown in (6) and (d). Notice the following revisions of the framing:
1. Shortening of the beams holding the motor

2. Introduction of a column at T, Sketch (6), so that the deflection of

TW will be very small

3. The use of pilasters such as QR to reduce the deflections of the

beams under the crusher

4. The addition of columns like OP to reduce the deflections of the

crusher supports still further

All problems like this should be investigated and solutions determined
in the early planning stages of the project, with the cooperation of the
mechanical and electrical engineers who are working on the layout of
equipment. Making changes after the completion of the general plans
is usually difficult and troublesome.
Figure 16-20 pictures the foundation for the drive end of a large fan
58 ft. above the ground floor of a power plant. The building consists of
heavy steelwork. In this instance a special structural cork cushion was
used under the foundation of the fan. Notice that it extends around the
edges of the base. Cork washers were also used under the nuts of the
anchor bolts to keep them from transmitting vibrations. The number of
cycles per minute was very large, but the probable centrifugal forces were
small, so that this cushioning seemed (and proved to be) sufficient.
Notice that the steel beams were encased in a very heavy floor slab, pur-
posely used to provide weight.
When
planning large installations high up in steel buildings, one should
consider whether or not the deflection of beams will be injurious. For
example, one end of the machine foundation should not rest on a stiff
girder while the other end is on a relatively flexible beam; a stiff girder

should not pass under the middle of the foundation where the ends are on
simply supported, oppositely deflecting beams; a column should not be
under or close to one edge or corner of the base while the rest is on flexible
beams; and the estimated deflections of steelwork should be such as to
equalize and not tilt or wrack the foundation. The different deforma-
 \

9 c

\ 5
J

f T J1 r
MO/aq /du B/i
'3Aoqo ldj>/[
'jdqSDM 1/J03 g

683
684 MACHINERY FOUNDATIONS [Chap. 16

tions of longcolumns caused by stress or temperature changes should not

be able to twist the foundation and misalign the bearings.
Vertical bracing of the structure near the source of vibrations may be
necessary to prevent setting other parts (sometimes quite far away) into
resonant vibration. Printing presses are often troublesome in this
respect.
When unexpected vibrations do occur, changing the mass, cutting down
the spans of the beams by intermediate columns or kneebraces to reduce

S Steel pi and beam

Table F/oor^
_^ / floor for machine

Mastic -
Cont. H bm. U
Cap and Clay
spring
-I
•s
-Precast base s Mastic
'" l
<i'"\!Ol

-Cork
iiiffl
'•
rr Wm
. Drain- "Rigid pistons

ia) id)

Exp.Jf.-^ 2nd floor

Corrugated
steel form

Basement
floor

FIG. 16-21. Some cases of isolated foundations.

the deflection, adding extra beams to increase the total stiffness, or weld-
ing cover plates on the beams may be helpful in avoiding resonance if
insulators cannot be used. It is obvious that changing a reinforced-con-
remedy such troubles will be even more difficult.
crete structure to
The reader should now review the comments previously made in con-
nection with Fig. 16-6.
16-7. Special problems. There are cases where some source of
annoying vibrations must be above the ground and inside a building but
special means other than insulators must be used to support it. Figure
16-21 shows a few such cases.
(a) This sketch pictures an insulator-mounted, reinforced-concrete,

tablelike structure completely isolated from the main building.

Art. 16-7] SPECIAL PROBLEMS 685

(b) This pictures a foundation which is practically floating in the clay

so that the pressure applied to the soil does not exceed that formerly there.
It is also cut away from the building.
(c) This illustrates the use of a sort of tower at an assay office. It is
isolated because it supports delicate instruments that would be affected
even by a person walking on the floor.
The following is a description of what happened in connection with the
foundation of a large compressor. Perhaps it contains a useful lesson.

- Vertical forces

-Horizontal moments
Secondary force or moment
Center
of shaft

One crankshaft revolution

<
"
>" !#•; UJ hi
.* "

\k\m\
[a) SIDE VIEW (b) END VIEW

FIG. 16-22. Outline of a compressor.

A similar compressor had been installed in another plant, where it

caused no trouble. It was placed on a heavy concrete block which
extended from the first floor clear through the basement to rock under
the basement floor. Thus its foundation had a large mass, tremendous
stiffness, and an unyielding support.
The general outline of the equipment is pictured in Fig. 16-22. The
pulsating forces apparently caused a rotational action about a center to
the left of and below point A, Sketch (a), and they acted in the plane of
the longitudinal axis of the machine. This was checked roughly by
686 MACHINERY FOUNDATIONS [Chap. 16

means of a vibrometer which enabled the engineer to obtain measure-

ments of the motions of various parts. The period of vibration was
slightly over 1,000 cycles per min.
The soil at the location of the compressor was as follows:

Topsoil IH ft.

Loose fine brown sand V/^ ft-

Loose medium brown sand and gravel 4 ft.

Loose medium brown sand 5 ft.

Loose medium brown sand and gravel 7 ft.

Loose very fine brown sand 5 ft.

The same sort of soil was later found to extend to at least a depth of 80 ft.
The foundation. pictured in Fig. 16-23(a) was built, and the equipment
erected on it. The vibrations were sufficiently harmful to crack certain
piping and to disturb some electrical gear in the control room nearby.

\^-<t of shaft y-i LD cycle

, Pre moulded
Floor filler
_^

or Drain
SIDE VIEW END VIEW

,,
w .„ v,„,, wt »w
\b) SIDE VIEW END VIEW

Mastic

r**T*

(c) SIDE VIEW END VIEW

FIG. 16-23. Successive designs for compressor foundation.

The right-hand end seemed to vibrate up and down so much as almost to

slap the floor like a beaver's tail.

It was then decided that the compressor should be dismantled and

the foundation shown in Fig. 16-23(6) constructed. This was done.
Although the vibrations were reduced, they were still very annoying and
threatened to cause fatigue of piping and perhaps trouble with the electri-
cal equipment. A further remedy was considered necessary.
Art. 16-7] SPECIAL PROBLEMS 687

Among the remedies studied were cement grouting, chemical grouting,

the use of piles or caissons alongside the present footing and tied to it, and
construction of another but heavier foundation. Although the first two
might help, the effectiveness of any grouting was uncertain and might be
uneven. The third involved driving piles or sinking pipe caissons inside
the plant during operations and clinching onto the existing foundation.
This would "drag" in the mass of a large amount of soil and remove the
local springiness of the earth, but it would be a difficult and costly process.
It was therefore decided that a new foundation isolated from the floor
should be built. The foundation pictured in Fig. 16-23(c) was finally con-
structed. In connection with this, notice the following:

y////v,v// //////////// '/////

^e Wall

KCTL k:to V
"Support

-6"4>
,I0"4>
equalizer
3G f
XJ CT
-Support

a) END VIEW (b) PLAN OF PIPING

FIG. 16-23A. Arrangements of piping.

1. A concrete box with 6-in.

reinforced walls was constructed first, with
backfilltamped around it. Then the floor was replaced.
2. The box was lined with 1 in. of cork board on all four sides but not

on the bottom. It seemed that any yielding of the bottom cork in addi-
tion to that of the ground itself would cause too much vibration of the
piping.
3. The bottom was tamped to assure firmness; then the heavy concrete
block was poured.
4. The top of the cork was sealed with mastic to keep out wash water.
5. The block was made deep enough to get a good section below the
drain. Heavy longitudinal reinforcement was used in the top and bottom
of the block. The drain could not be relocated without serious cost;
hence, it was allowed to span across the box but was kept clear of the con-
crete of the block. The area around a sleeve like this is a point of weak-
688 MACHINERY FOUNDATIONS [Chap. 16

ness in resisting the effect of vibrations. Therefore, since the tendency is

to move the right-hand end, as seen in Fig. 16-23 (c), up and down, 45°
reinforcement on all four sides seemed to be the best to use locally around
the opening.
When the compressor was put in service for the third time, the oper-
ators rearranged the piping as shown in Fig. 16-23.4 (a). When the
machine was started up, the 6-in. piping vibrated so badly, it looked as
though the pipe had %
ni of fuzz on both sides.
- Of course, this set other
pipes to vibrating. The operators were worried and disgusted. How-
ever, the engineer found from vibrometer readings that the foundation
itself was behaving well. What happened was the following:
1. As the air impulses left the machine at A, the air impinged against

the elbow at B. Then it did likewise at C, and again at D, where the

pipe ran about 8 ft. along the wall (away from point D), down a riser to K,
and back toward a reservoir through a 10-in. pipe.
2. The period of the pulsations and the speed of the air flow were such

that the "kicks" at the bends C, D, etc., were just right to reinforce each
other and to build up very serious vibrations.
The engineer discovered this and, with the agreement of the operators,
recommended the following remedial steps:
1 Take down all the former 6-in. piping except A B, Fig. 16-23^4 (a), and
the 10-in. pipe K.
2. Install a 10-in. pipe equalizer as shown by EF, Sketch (6), at the
level of B, near the floor.
3. Connect a 6-in. pipe across from G to H near the floor, entering the
10-in. line JK.
The equalizer "washed out" the pulsations, and the machine operated
satisfactorily. Although there was a slight vibration of the concrete
block, no noticeable effects were transmitted to the floor or to surrounding
equipment.
Figure 16-24 shows the design used for the central pier of a large
settling tank. The machinery operating the rakes caused the vertical
load and the torque shown. The provisions for piping underneath the
footing are not pictured. The footing was made square and large
enough so that the passive resistance of the soil (considering the surcharge
applied by the contents of the tank) around the edges could safely resist
the torque without help from friction under AD. The machinery rested
upon the ring of concrete shown by KP and QL. Below NO the shaft
was solid concrete reinforced by spirals / and g which were lapped over
corresponding dowels e. Dowels c and bars d are vertical and are to
resist overturning of the pier. The main bars a and b are like ordinary
footing reinforcement. Bars h and j primarily tie the outer lip to the
main body. Notice that the top above BC is circular and that the joints
Art. 16-7] SPECIAL PROBLEMS 689

at F and G isolate the tank floor from the base. Also, notice the use of
intermittent keys at the construction joint JM.
Heat can be a source of trouble for machinery and structural founda-
tions near such equipment as furnaces. A foundation for a furnace in a
metal-fabricating plant is partially pictured in Fig. 16-25. The planning
and design of the foundation for this furnace will illustrate some of the
problems connected with resistance to high temperatures.

Torque = 805
clockwise

Spirals

Pipes not shown

79-0 circular top

-19 '-6. "square base

or footing

|--tu i^w 21 £ ^ x- k, ,_
S\V"W"\\V " \VV"\v

FIG. 16-24. Central pier of a large settling tank.

A ventilated air space under the furnace between its bottom and the
floor prevents any harmful heating of the outer floor itself. This is a
feature which is often used in furnaces with fixed bottoms, and it may be
accomplished by resting the furnace on a series of parallel steel beams
supported on a concrete footing or mat. The natural or forced circulation
of air between the beams is depended upon to keep the concrete of the
foundation from dehydrating. In the case pictured in Fig. 16-25, the
furnaceis on stilts, thus securing satisfactory protection of the floor.

However, when the car which forms part of the bottom of the furnace is
690 MACHINERY FOUNDATIONS [Chap. 16

run out, the 1000°F. temperature within the furnace will cause radiation
below and near the opening in the
to heat the floor seriously, especially
bottom of the furnace.
The fact that the structural floor rests on the ground means that the
heat transmitted to the floor cannot escape readily; hence, the tempera-
ture of the concrete will increase and may quickly exceed the 500 to 600°F.
which will cause the concrete to disintegrate. Therefore, refractory
concrete made burned aggregates and Lumnite
of traprock or suitable
cement should be used for the portions labeled A and B in Fig. 16-25.
At ordinary room temperatures (70°F.) such concrete can be made so that
its crushing strength will be 3,000 to 3,500 p.s.L; at 400 to 1000°F. it

will have a compressive strength of approximately 1,500 p.s.i.

x^Insu/ot/on Inside of furnace of

<L

furnace
Floor of car-

pours

FIG. 16-25. Foundation for a furnace.

A flexible insulator D was used around both sides and ends of part B. 1

This insulator was made of 1 in. of rock-wool padding between two stiff
pieces of /2 -m asbestos board.
- The footings C were made of conven-
tional reinforced concrete. The stone concrete was not seriously affected
by heat because of its distance from the opening in the bottom of the
furnace and because of its accessibility to cooling air.
When Lumnite cement is used, the heat of chemical action is consider-
able. In concrete made with Lumnite cement this initial expansion
during setting offsets the subsequent effects of elongation when heated
by the furnace. The engineer who uses Lumnite cement should be
careful to obtain and follow the instructions of the manufacturer. 2
No reinforcement was used in the refractory concrete of part B. In
such a situation, embedded bars may actually be a serious hazard because
1
Both A and B were extended for a few feet in front of the furnace door.
2
Universal Atlas Cement Co., for example.
Art. 16-7] SPECIAL PROBLEMS 691

of their —
expansion when heated both increase of length and of diam-
eter— which may crack the refractory concrete. Ordinary bond to
the steel cannot be relied upon. If bars are used at all in part B, they

should be plain bars painted with asphalt or taped, so that when the
heat destroys the covering, a space for expansion and slippage will be
provided. Mechanical hooking at the ends or the use of threads and
nuts will probably be the only suitable attachment or anchorage for the
ends of the reinforcement.
The pedestals for turbogenerators at power plants are important
foundations that are subjected to vibrations of large frequency. Figure

Turbine - Generator,

Exciter
<£of shafts H--\ Table
f«-©-.4

U) LONGITUDINAL SECTION ON CENTERLINE (b) SECTION A-A

FIG. 16-26. A turbogenerator pedestal.

16-26 pictures one of these structures. A few suggestions for their

design are the following:
1. Make the thick and broad to obtain (a) strength as a rigid
mat
body, (6) weight, and (c) stability. The principal impulses will be acting
perpendicularly to the axis of the shaft; hence, the width of AB should
be generous.
2. The columns should be very thick and sturdy as much so as the —
equipment will permit.
3. Walls across the short direction are desirable but not usually possible

because of the requirements of the equipment. They should be as nearly

complete as they can be, somewhat as shown by CD.
692 MACHINERY FOUNDATIONS [Chap. 16

4. Junctions between columns and other members should be made

with strong, reinforced fillets, as shown in (a).

5. It is desirable to make the condenser supports EF and GH an integral

part of the adjacent columns in order to increase the stiffness of the

latter.
6. Construction joints will occur at such points as JK and LM. These
are sources of weakness. Central keys as indicated are of some value,
but heavy, continuous vertical reinforcement is the most effective. In
one case, a defective automatic valve failed to shut off the steam when the
load was taken off the generator. The latter raced until it blew up.
These construction joints at the generator end cracked so badly in spite
of very large vertical bars that the whole table had to be replaced.
7. Horizontal parts like NO are especially likely to vibrate vertically
like diaphragms. As used here, NO was stiffened by very strong edge
beams, and top and bottom were strongly reinforced both ways.
8. Brackets like P have been discussed in connection with Fig. 16-5.

9. Cross-walls like JQ often have to be thin and offset as shown. The

author has successfully used welded trussed reinforcing-bar units in these
walls because of the likelihood of bond failure due to fatigue.
10. The table RQS is usually punctured and slotted to accommodate
the equipment and piping. has to be as massive and
Nevertheless, it

strong as possible. It should be continuously reinforced, tied in at all

corners, and well supplied with stirrups and bent-up bars to hold every-
thing together. It should be approximately as stiff horizontally as
vertically. The entire pedestal should be completely isolated from the
working floor.

11. The allowable working unit stress in reinforcement for design

should not exceed 10,000 p.s.i. tension and 4,000 p.s.i. compression; that
in the concrete, 400 p.s.i. compression and zero tension. Concrete having
an ultimate compressive strength of 3,500 to 5,000 p.s.i. should be used.
Laps of bars should be 50 per cent greater than for ordinary structures.
12. For all static loads of pedestals and equipment, the computed
deflection of any member should be such as to avoid resonance. It is
probable that these deflections will be satisfactory if limited to less than
the following;:

Speed, r.p.m. Deflection, in

1,200 0.012
1,800 0.006
3,600 0.0015

13. The longitudinal and transverse forces for design will be specified
by the manufacturer of the equipment. The actual dead loads of the
Art. 16-7] SPECIAL PROBLEMS 693

machinery should be included, plus 50 to 100 per cent of the rotating

parts.
14. A maximum torque of four or five times the normal torque should
be provided for.

15. Temperature stresses and pipe thrusts should be provided for.

16. It is desirable to make sure that all reinforcement is separated from
anchor bolts and other metal which might bring in stray currents of
electricity, using 2 or 3 in. of concrete between them. This is to prevent
possible electrolysis. Separate grounding of the bars may be desirable.
If so, there should be a complete welded grid so as to avoid resistance
points that might destroy the reinforcement. This means that even
stirrups should be welded at one or two points in order to join them to
the grounded system.
17. When the mat under the pedestal must rest on plastic soils, or on
friction piles in plastic soils, there may be danger of unequal settlement
and resultant overstressing of the stiff members of the pedestal. Because
of these uncertainties and the great importance of the structure, it is
advisable to make the mat very stiff and to design the mat for the follow-
ing extreme cases:
a. As a continuous combined footing with the actual column loads on

top spread to the soil under them as though the mat were a series of
separate footings and the columns were not rigidly connected to it.
b. As a rigid free body with fixed ends for all members and the soil

pressures assumed to be P/A ± Mc/I under the footing as a whole.

The center of gravity of the total applied loads should coincide with that
of the pressure diagram.
c. As a rigid body acted upon by the actual loads in the columns on

the top and a modified pressure diagram under the footing. This last
diagram will depend upon the soil modulus and the yielding of the soil.
This is uncertain, but the probable tendency is a yielding under the areas
supposedly having the greatest soil pressures and a considerable tendency
to increase the tension in the top of the mat between columns. These
distortions should also be considered when designing the reinforcement
of the table and struts.
18. As a final step, look over the design drawings for any possible weak-
ness, and see that bars are provided to prevent any cracking. Then be
prepared to give thanks if everyone is satisfied.
 *

APPENDIX

Table 1. Approximate Formulas for Analysis of

Reinforced-concrete Members*

M = A s fsjd fs = -j—rj
A sja
A = s y-r-.
jsja
For rectangular and T-beams

2M
M = l
ifc kjbd
2
fc = , ., ,
2
For rectangular beams

A- = y/'Ipn + (/>'i)
2
— j9« i = 1 — o For rectangular beams

F V V For rectangular beams and

Vl = T^l VT =
(Eo)jd bjd '

bkd ends of continuous beams

A v fr = (vl — v L )bs For vertical stirrups
Approximate values of coefficients: A' = 0.3, j = 0.9.

* See List of Symbols for meaning of symbols.

Table 2. Allowable Values of Unit Stresses for Analysis and Design

of Substructures, Assuming Ultimate Compressive Stress = 3,000 P.S.I.

fc = 1,350 p.s.i.

fs = 18,000 p.s.i. for structural grade steel

fs = 20,000 p.s.i. for intermediate grade steel

u = 210 p.s.i.
vL = 90 p.s.i. (75 p.s.i. in two-way footings)
vt = 600 p.s.i.
n = 10

* See List of Symbols for meaning of symbols.

695
 05 rH CO CD CD 00 OS © rH H CM CO CO
CM 00 CD CD 00 rH io © co © co 00 rH
9 co rH O OS 00 00 f~- t» © LO LO rH TH

- © © © © ©
>s eo so tp <n eo t- -r?eo t-
© ©
i

rH t~- TP rH C0 io co eo t- IO
<t[
rH CO CO CO CM CN CN CN CM rH rH *
H O ©
N 00N LO ©
00 I— 00 rH 00 00 CM
CD OS CO iO D CO 00 CO COO N CO ©

O .5 * CO rH O O 00 NN CO © lO LO rH rH CO

© lO * © ©
<!
IO -H
eo eo cO ©
(-
t*- lO
"*
cO
CO CO
rH © ©
© © CN
IO C0
t-
CN
Tf CO CO CO IN CN CN CN CN rH fh rH FH fH

m h co c *
HH
oo lo co
O •t lO NO if) ON
CO N N
CO CO N W
if) (> rH
00 LO

.a -e-
rf« O ffia N N to CO LO LO rH rH co co
© \°0 _L_
O
© "* O
CO
O
(~-
so
©
>*
eo
rH
©
o
io
eo
rH
t-
© IO
© O CO
CO CN
© ©
© ©
©
<tf
^ CO CO CN CN CN IN fH fH FH fh fH -H ^ fH
CD N N Tf lO M O N
00 CM rH OS CO rH
o lO N C-l
-t co io oo cm oo MON HN rH rH
i.

oo
-e-
WOO 00 N CO to IO LO LO rH rH CO co co

fH t- CO
rH © CM eo SO
H * h B * io so eo IO IO
CO CN rH © © © ©
eo t-
<!
M N N © ©
O
O CO
LO
"* CM
co
co
©
© o
© © oo
io ©
rH O CM
N rf H
I-
CO CO
© © LO
© t>
pq
h OS 00 N tO CO lO lO rH rH rH CO CO CO CM

Fh
t^
o* o
>* © eo
©
© rH fH f- © © ©
CO
© © © CN
© CO IO©
© so
-O 5 «!
CN CN —
rH CO CN fh
fH © © © ©
tr
© ©
«HH

0) o CO* D
o OO
CO Tf
o
OS
N CO LO CM CD
tO rH CO N LO 00 rH LO CO
O N LO rH 00 t^ to
lo co

« 02
h O CO N CO LO LO -CH rH rH CO CO CO CM CN (M
© -H
rH
© —
t- lO
i CN
CO
t- so s© eo -h
rH o ©
eo co
io
I-
©
N *© W©W CO CO T*
rH rH
< <N fH fH fH rH fh © © © © © © © © © ©
CD rH T* (M 00
^ oo t^ oo
CO
cm
O 00 CM CM
r- cm © ©
CD rH tH rH LO
CO rH OS © CO
rH CD
rH OS

D N to iO ID -f rH CO CO CO CO CM CM CM CM rH
LQ _d
© © * © co co <* eo cm t~-
© CO © © fH t- r* FH
«<
CO rH © ©
SSI eo t- so »o 10 lO rH r* CO CO CO
H ^ fH © r-< © © © © © © © © © © © ©
IN rH 00 OS rH
N © N CO
H N
rH O
f H O)
OS rH CO © 00
© lo co © oo
© iO CM CO t-~ lO

rH
© N CO iO -f rH CO CO CO CM CM CN CM CM rH rH rH
.0
o © © © ©
iM © eo so ©
CO CO rH© (-
lO rH "Hj<
©
rH CN t- r*
cO CO CO CN CN
CN ©
CN CN
CO
1-fS,
^ © © © © © © © © ©
*r?<

© © © © © © ©
o
00 CO O) iO *
o CO N O iO
UJ CO N CO 00
H CO lO M H IN C5
© NN
00 t> LO rH
C-] © oo
(M ^
-e-

io -t tc co CO CM CM CM CM CM rH rH iH rH rH rH
co _d
© *
co co © © © r* CN © — ©
— eo io co CM H
\oo
© * eo
lOCO CO CN CN CN CN CN — — —
^ o © © © © © © © © © © © © © © © ©
oo rH CM t~ rH 00N O CO * rH LO N rH rH lo oo
e- © N h to CO O 00 N >0 rH CO CM rH O © oo n
rH CO CO CM CM ICM rH fH ^H ^H „ rH rH rr O © ©'
CO
© © - -
"* ©
CO CN -h © © 1— © lO IO
V*
«*!
CO CI CI
t- 1/5
r- ©
PH -H rH © eo
© eo
© © © © ©
© © © © © © © © © © © © © © © © ©

d \(N \N \<N S4M

l.a IM CM CO CO rH rH LO IO © CD N N 00 C: O ^-i CM

696
 © co
"* GO CO t-» CM © © os co b- eo © © LO © CO GO CO © © LO ©
CO O -t< © CO l>C0 t»4
1

HHN
Ft GO CO r~ CO CO i-H LO
©©
"* 00 CO t^ © t^
LO © © 00 co — © © 10 © -t 00 oi C-1 Fj4 1^-
CO CO CO CO rt Tj4 lO lO© t^ t^ 00 GO a> © © fh ^ co 01 co *-+
_d
- v* CO 03 -t O SO
lO iH © CO 05 CO ©n 10 © © CO t- CO © tP © LO f- sO CO © CO ©NFj< O © fh t- eo 00
CO
\-H
*£
rH SC >* © t- ON © CO -t lO t- © © fH CO lO © © ©N H eo 10 ©n © Ol CO lO © Ffi
fh m CO CO CO eo co eo co co CO CO CO CO CO
t^-

©3 00
-e- © © © © © © © © © © © 02 © © 00 00 00 00 GO 00 GO GO 00 00 GO N t^ t^ t^t^
+ CONMtDO
-H CO
CO N
-h KJ C.
Ol CO CO CO CO
CCN-hiCO)
LO 10 lO
CO Is-
© CO
H LO © CO t-
t— 00 00 ©~©
—LO 4 CO
©— ©
© © l>
rH LO
O _d
r-4 t}< tJ4 t>- l>- 1

1- <* fh Ol ON ecio
NlfltSOM © © eo Ft t~ © co 10 eo © CO © ©N FH F# t- on eo 10 © © CO© © —
rl\ **
H NMlflvO 1^ © © -H CO lO © t- ©\ © — CO F? IO © t^ ©N © — •rf CO Ft lO © ©
eo eo co eo eo eo eo co co co CO CO CO CO CO
rt ©
6- iCOONN cooort ©Ft O lO H CO CO I> CO 00 CO © r* O LO H CO rH t^ c-1 GO CO
+ COt-O'* t- — -* GO ^H lO © CO © © CO © © co t^ © •r GO rH LOGO ©
Ol LO OS CO
-f lO LO CO © © t^ l> t^ 00 00 00 © ©.©.©©
1

Ol CO CO CO CO 0-1 r)4 Tfi

Ol
_d
O O©O© ©©©©
© © © © © © © © © © © © © © © © ©© ©© © © © © © ©©©©©
\»
r-<\ «! rH CO CO ** lO so t- eo on © -* CO CO UO © t- © ON © H CO CO Ff lO •"# © t- © ON ©
co eo eo eo eo co co eo eo co
-*4 00 CO
<— CO <*
1 © t- 00 © -H CO CO
f*4 LO t^ 00 © rHCO rH LO © 00 © FH
iflMXJOH TJ4 0-1
-6-
co co 02 01 10 GO CO LO GO i-H -J4
— co 04 cm CO co co
1
1
O© t^ © co © © 01 LO © O) LO GO H Ft t^ rH
^ * "t lO LO
F}4
LO LO © CO © t^ t^ t^ 00 00 00 © ©
rJH

00 _d ©n © 1—
(-- ITS CO
©
i« t}< CO (M fH
CJn —
©N
-Fj- CO
© © © — © © -* eo © 00 © f* CO © © lO CO F- ©N 1--

<!
—^ csi co co Ft IO SO t- t- © ©N © F- FH co -^ 10 10 © t- © ON ©N © FH CO Ol CO
o-l
co co eo co eo

-6-
N iO CI O 00 iO CO © 00 10 CO © 00 iC CO © 00 10 CO © CO LO 00 © GO LO Ol © GO LO
© H -+ © © CO LO © eo
co xa 00 cd CO OS 00 *# t- CO LQOO
©© © 00
1-4 t-F r-i -J4 i> OS CO

— 1- d rH r-1 CO CO CO CO CO CO CO "^44
-V -+ -t LO LO LO© © t^ t^ t^ t^ GO
© CO
© © © © © SO © OS
© <*
©©© © © ©
© CO © © © © -J- © eo Ff © © CO ©
c-i O © CO © Ft © fJ-

O
<! OhhM« CO Ft Ft 10 SO © t- r- © ©n ©n © © — eo eo co co f* 10 m vo © t- oe
CD
SI
co t> -+ 00 CO W OS CO © © coi> © * © © co co © HfNH-COO
CO rH LO GO Ol
in -e-
co * r~ © h -f © 00 CO
HHH04N i-4 © GO © CO lO ©
r- co Ft © rH -f © ~ r- CO © 00 ©
t^-
rf LO LO LO LO © © © ©
Ol CO CO CO CO "^ "^t *i^ 1
^t !>•
©
1

c C"4

-t © eo © © * eo eo © © <* © CO © ©
-t CO CO l- CO © © lO ON *JI eo co t © ©©
Ff CO ©©
—^ ©N CO
** eo
© CO © f- sO © -t ©n co © co
t- t- t^ © © ©NON © © — -h -H eo eo co
«<
© © — — CO CO CO CO Ft
1
0-1 T? lO © ©
1ft

co 01 GO rh O
© OS 00 GO 00 00 l> t- © © CO LO l© * Tt 1
-J4 CO CO CO CO Ol — rH © © © © © 00 00
^COiONO: -HCOiONC! -< CO lO N O
© — 1co lo t^ r-4 CO LO t> © rH CO -rH CO Of)
CO CO CM CO CO co co co co co r~H fH fj4 -rt *H- LO LO LO LO LO
10 _d
© t^ © ©n © fH CO CO * 10 © t- © ©N © © I— © ©N
O CO © ©v ©
-H 0-1 CO Ft lO
SO ©N lO 00 Tf t- — l> © CO © ©n co 10 © eo 10 © F- Ft t—
< CO CO r-4 CO
•Ff"

© © ©H — —1 CO CO 0-1 CO CO CO —4* F^ F^ © © © I— l> t- © © © © ©N

H$< lO lO lO
r- -* rr 00 >o CO
-e-
lO —4 1^ O) 00 CO 00 -* © © -H t^ CO 00 © LO rH t> CO CO -t O LO —-J4

i-H CO"* ©l> 05 i-4 CO Tf4 lO r^ 00 © co LO © 00 © H COfhcONC: O Ol ^ >0 N

fh h
c-i

.d CO CO co co co co co co co co co co
C-I
FJ4
©©©
©©©©
CO Ft© © ©©© ©
co Tp © eo ©
©©f?sO© ©© ©© eo©© ©© ©© ©© Ol© Ft© SO© ©© ©© © Ft© ©© ©© ©©
CO -ji Ol
-*!
co eo co co co CO CO CO CO —? —t —^ —^ —t 10 lO lO 10 1/5 ©
GO CO -* COO © © -f © 00 © 00 CO lO t^
i-H CO U3 t^ OS
00
c-i
Tt4
© rH CO CO Ft< 00 © rH CO LO t^ ©© OI Ft
-HIMfO-^iO t» 00 OS © rH CO LO © I> TJ4 00 © rH CO CO "* ©
1> GO OS © rH CO Ft LO
co _d — CO OJ CO CO
1 Ol Ol Cni Ol O)
— CO CO Ft IO © h- eo on © fh eo co 10 -}< © t- © ©N © FH Ol CO Ft lO © l— © ON ©
«!
1

© t- © ©NfH CO CO T? lO © t- 00 ©n © eo CO Ft UO sC t- 00 © fh CO
C?N

© © © © © — — — 1— co eo co oi 0-1 eo eo co co co

00 ©
O IO CO 00 © 00 © CO 00 ©
rH — © -t CO © 00 lO CO h 02 t^
-}4 tJ4
lO CO© 00 © Hf 0-1 O N lO CO rH 00©Ft
© Ol CO CO lO © 1^ t^ 00 OS O © fH
— 1
1 Tf4 CO CO Ff © t^ t^ CO © ©
rJ4 LO -H -H co CO
Cm Ol CO Ol CO
co _d
LO © IO © 10 © 10 © 10 © lO © lO © 10 © lO © 10 © lO © lO © IO © 10 © lO ©
V* CO CO ^H ^* lO 10 © © t- t- cc cc CO CO Ft Ft lO
««?
©©©©© ©©©©© ©©©©©
CO
0=*.
^ O 8
Sh rn co co Ft lo © t^ GO © O H O) CO Tf4 LO © t^- 00 ©© rH CO CO Tt LO I-- CO
CO CO CO CO CO co 01 co co eo
© ©©

697
698 APPENDIX

0.025
i
kd\
^£-* hd

0.020 /
ft' V 2pn + (pn) z -pa

i k As

p 0.015
<r

^
9) /
A w \<i

0.010

0.005

0.002
0.20 0.25 0.30 0.35 0.40 0.45 0.50
Values of k

FIG. 1. Location of neutral axis of rectangular beams with tensile steel only.
 APPENDIX 699

0.&
w = Uniform load on slab, psf
-oto.7 ws Uniform load supported by
=
oi
<s « «e one - ft. strip in short direction
o E ^«
Wj/ Uniform load supported by a
c >o.6
one -ft. strip in lonq direction
J£
C J_
£ 0.5
3 O
5o |Q4 ui,
w
Assumed
c-e
|b "§0.2
e
S- 3
8 80.1

1.0 1.1 1.2 1.3 span

Ratio of lonq span to short span Plan of Slab

FIG. 2. Assumed distribution of uniform load on a rectangular panel supported on four sides and
reinforced in long and short directions.
700 APPENDIX

ZP-38 ZP-32

§31
ZP-27

DP-1 DP-2

•*>|to

1 J"
S 1
-is

10"
*"M

-t^Z'^9
1

< 8 32 ,|
<J
---W ''AW
r54M^ f
*5 10"
/'' L_ »/'
AP-3
—J 0>|<0

*>
|

r
/5"
H
AP-8
Sections ZP-38 and ZP-32 will interlock with each other.
Sections DP-1, DP-2, AP-3, SP-4, and SP-5 will interlock with each other.
Sections SP-6a, SP-7a, and AP-8 will interlock with each other.

FIG. 3. Data regarding steel sheet piling. (Courtesy of Befh/ehem Steel Company.)
 APPENDIX 701

SP-4 SP-5

15 15

10°
& toi

SP-6a
10°
toi"

SP-7a

Weight, lb. Section modulus,

Moment
Section Area, Width of
Per Per
no. sq. in. in. inertia,
lin. ft. sq. ft. Single Per lin. ft.

of bar of wall section of wall

ZP-38 16.77 18 57.0 38.0 421.2 70.2 46.8

ZP-32 16.47 21 56.0 32.0 385.7 67.0 38.3
ZP-27 11.91 18 40.5 27.0 276.3 45.3 30.2
DP-1 12.56 16 42.7 32.0 87.0 20.4 15.3

DP-2 10.59 16 36.0 27.0 53.0 14.3 10.7

AP-3 10 59
. 19% 36.0 22.0 26.0 8.8 5.4
SP-4 8.99 16 30.7 23.0 5.5 3.2 2.4
SP-5 10.98 16 37.3 28.0 6.0 3.3 2.5
SP-6a 10.29 15 35.0 28.0 4.6 3.0 2.4

SP-7a 11.76 15 40.0 32.0 4.6 3.0 2.4

AP-8 11.41 15 38.8 31.0 24.2 8.9 7.1

When minimum ultimate interlock strength is specified for SP-4, SP-5, SP-6a, and
SP-7a, they can be furnished with the following strengths at the published extra
charge: SP-4 and SP-5 = 12,000 lb. per in., SP-6a and SP-7a = 16,000 lb. per in.
Values listed apply only to the interlocking joints and not the entire cross section
of the pile. Sections SP-4 and SP-5, when subjected to direct tension as in cellular
construction, should be investigated for web extension which will occur before the
interlock strength is developed.

FIG. 3. [Concluded.)
702 APPENDIX

—X
Y

Area Depth Flange Axis XX Axis YY

Weight of of
Web
Sec- thick-
per ft. sec- sec- Thick-
tion
tion tion
Width ness I S r V 8' r'
ness
no.

Lb. [n.= In. In. In. In. In.' In. 3 In. In. 4 In. 3 In.

BP14 117 34.44 14.23 14.885 .805 .805 1228.5 172.6 5.97 443.1 59.5 3.59
BP14 102 30.01 14.03 14.784 .704 .704 1055.1 150.4 5.93 379.6 51.3 3.56
BP14 89 26 19
. 13.86 14.696 .616 .616 909.1 131.2 5.89 326.2 44.4 3 53 .

BP14 73 21.46 13.64 14.586 .506 .506 733 107.5 5.85 261.9 35.9 3.49
.

BP12 74 21.76 12.12 12.217 .607 .607 566.5 93.5 5.10 184.7 30 2.91 .

BP12 53 15.58 11.78 12.046 .436 .436 394.8 67.0 5.03 127.3 21.2 2.86

BP10 57 16.76 10.01 10.224 .564 .564 294.7 58.9 4.19 100.6 19.7 2.45
BP10 42 12.35 9 72 . 10.078 .418 .418 210.8 43.4 4.13 71.4 14.2 2.40

BP8 36 10.60 8.03 8.158 .446 .446 119.8 29.9 3.36 40.4 9.9 1.95

FIG. 4. Bethlehem H piles: properties for designing.

Notes: It is advisable to protect steel piles that are exposed, especially those that are
subjected to the corrosive action of salt water. Fresh water may also be very injurious
if it contains acids and other harmful chemicals; e.g., ferric sulfate leached from piles

of coal. Copper-bearing steel is advantageous when exposed to air but it seems to

have little advantage in sea water.
Mortar is a better protection than concrete. The cement should be one of the
special kinds that are made for use in sea water. The cover over any steel or rein-
forcement should be at least 4 in. The encasement should extend 3 ft. below the
permanent mud line. Corrosion-resistant metallic shells used as permanent forms
are very desirable.
 APPENDIX 703

1.0 i i i

f
rr
0.9 J
U
0.8 „ wh 2r.

0.7

£ 0.6
o /•/(£ 9
s /: 3
/:?
•'•• '*

0.5 \/
US
°0.4

Levels

02

0.1

20 25 30 , 35 40 45
Angle of internal friction 0,degrees
Coefficients for Coulomb's equation for active
earth pressure; horizontal component
cos<p -I2
CA - [
[l+ Vsin <p (sin (p - cos <p tan 6)J

FIG. 5. Data regarding active earth pressure.

704 APPENDIX

Clearance fine

{
— &—\ ni¥
1

•^
i.—
—
1_
•**
^r—
mi
Cranes of 80 tons capacity and
< V ? w over have 4 wheels at each end
Px = c.c. of wheels of truck
P^ distance between trucks

Diagram for Table 5.

 APPENDIX 705

Table 5. Dimensions of Shaw-box Cranes (See diagram on Opposite Page)

(Manning, Maxwell & Moore, Inc., Muskegon, Mich.)

Capacity, Max wheel

tons
A J P R V W load, lb.

40' 5'1" 870" 7H" 6'9" 57" 19,000

10 60'
80'
100'
57"
57"
6'0"
9'8"
1170"
14'2"
W
SH"

SH"
7'6"
8'6"
970"
6'6"
7'3"
8'5"
24,000
28,000
35,000

40' 6'5" ll'O" 9" 87" 7'6" 37,000

20 60' 6'8" 11'4" mi" 97" 770" 42,000
(5-ton aux.) 80' 6'8"
"
12'6" mi" 9'5" 8'2" 47,000
100' '
6 8 14'2" mi" 970" 87" 54,000

40' 6'10" 11 '8" ioH" 9 '2" 77" 50,000

30 60' 6'10" 12'0" mi"" 9'6" 770" 54.000
(5-ton aux.) 80' 6'10" 13'2" ny 2 lO'O" 8'6" 60,000
100' 670" 14'6" nw 10'6" ,
9 2
„
68,000

40' 8'4" 12'8" 13" 107" 87" 74,000

50 60' 8'4" 14'0" 13" 10'3" 8'9" 81,000
80' 8'4" 15'0" 1071" 9'5" 92,000
(10-ton aux.) 14H"
100' 8'4" 16'4" 141^" 11 '8" 10'3" 100,000

Pi Pi

60' 7'9" 4'8" 4'4" 12" 11 '3" 97" 52,000

80 80' 7'9" 5'0" 4'0" 12" 11'5" 9'10" '
60,000
(10-ton aux.) 90' 87" 5'2" 4'4" 12" 11'9" 10'2" 65,000
100' 87" 5'6" 4'6" 12" 12'0" 10'6" 70,000

60' 8'0" 5'0" 4'6" 13" 11 '6" 9'8" 78,000

100 80' 8'6" 5'4" 4'2" 13" 11 '8" 970" 86,000
(15-ton aux.) 90' 8'6" 5'6" 4'6" 13" 12'0" lO'l" 90,000
100' 8'6" 5'8" 4'4" 14" 12'0" lO'l" 95,000

150 60' 9'0" 5'0" 4'6" 14" 11'6" 9'8" 92,000

(25-ton aux.) 80' 9'6" 5'4" 4'2" 14" 11 '8" 9'10" 100,000
100' lO'O" 5'8" 4'4" 14" 12'0" 107" 109,000

200 60' 9'6" 5'0" 4'6" 16" 11'6" 9'8" 109,000

(25-ton aux.) 80' lO'O" 5'4" 4'8" 16" 1171" lO'O" 119,000
100' ll'O" 5'8" 4'4" 16" 12'0" 10'2" 130,000
706 APPENDIX

k 1
™°*»rJLd
A I

2
Tr _wh r ,

c
Level'^ t:iq,

1:S^ 1:3

£ 1:2
i.

20 25 30 35 40 45
Angle of internal friction <p, degrees

Coefficients for Coulomb's equation for passive

earth pressure ; horizontal cornponenf

C Jf cos<t> 12
[_1-Vsin <p (sin tp -cos (plan <f)J

FIG. 6. Data regarding passive earth pressure.

 BIBLIOGRAPHY'

General
Andersen, P.: "Substructure Analysis and Design," The Ronald Press Company
New York, 1956.
Casagrande, A., and R. E. Fadum: Application of Soil Mechanics in Designing Build-
ing Foundations, Paper 2213, Transactions of the American Society of Civil
Engineers, vol. 109, 1944.
Hough, B. K.: "Basic Soils Engineering," The Ronald Press Company, New York,
1957.
Krynine, D. P.: "Soil Mechanics, Its Principles and Structural Applications," 2d ed.,
McGraw-Hill Book Company, Inc., New York, 1947.
Taylor, D. W.: "Fundamentals of Soil Mechanics," John Wiley & Sons, Inc., New
York, 1948.
Terzaghi, K.: "Theoretical Soil Mechanics," John Wiley & Sons, Inc., New York,
1959.
, and R. B. Peck: "Soil Mechanics in Engineering Practice," John Wiley &
Sons, Inc., New York, 1948.
Tschebotarioff, G. Foundations, and Earth Structures,"
R.: "Soil Mechanics,
McGraw-Hill Book Company, Inc., New York, 1951.
U.S. Navy Bureau of Yards and Docks, "Soil Mechanics and Earth Structures,"
Technical Publication NAVDOCKS TP-PW-18, October, 1953.

Geology
Krynine, D. and W. R. Judd: "Principles of Engineering Geology and Geo-
P.,
technics," McGraw-Hill Book Company, Inc., New York, 1957.
Legget, R. F.: "Geology and Engineering," McGraw-Hill Book Company, Inc.,
New York, 1939.
Longwell, C. R., A. Knopf, and R. F. Flint: "Physical Geology," 3d ed., John Wiley &
Sons, Inc., New York, 1948.
Trefethen, J. M.: "Geology for Engineers," D. Van Nostrand Company, Inc., Prince-
ton, N.J., 1949.

Machinery and Vibrations

Barkan, D. D.: First Investigation of the Theory of Vibration of Massive Foundations
under Machines, Proceedings of the First Inter national Conference on Soil Mechanics
and Foundation Engineering, vol. II, 1936.
1
The references listed are purposely limited in number and confined to publications
which the ordinary reader may secure without undue difficulty.
707
708 BIBLIOGRAPHY

Newcomb, W. K.: Principles of Foundation Design for Engines and Compressors,

Transactions of the American Society of Mechanical Engineers, April, 1951.
Richart, F. E., Jr.: Foundation Vibrations, Proceedings of the American Society of
Civil Engineers, Paper 2564, Journal of Soil M
echanics and Foundations Division,
August, 1960.
Tschebotarioff, G. R., and E. R. Ward: The Resonance of Machine Foundations and
the Soil Coefficients Which Affect It, Proceedings of the Second International
Conference on Soil Mechanics and Foundation Engineering, vol. I, 1948.
Veletsos, A. S., and N. M. Newmark: Natural Frequencies of Continuous Flexural
Members, Paper 2859, Transactions of the American Society of Civil Engineers,
vol. 122, 1957.

Miscellaneous

Ayres, A. H.: Best Procedure for Unwatering of Davis Dam, Civil Engineering,
August, 1948.
Baker, A. L. L.: "Raft Foundations: The Soil-line Method of Design," 2d ed., Concrete
Publications, Ltd., London.
Blaine, E. S.: Practical Lessons in Caisson Sinkingfrom the Baton Rouge Bridge,
Engineering News-Record, Feb. 6, 1947.
Bretschneider, C. L.: Hurricane Design-wave Practices, Paper 2965, Transactions of
the American Society of Civil Engineers, vol. 124, 1959.
Chien, N.: The Present Status of Research on Sediment Transport, Paper 2824,
Transactions of the American Society of Civil Engineers, vol. 121, 1956.
Dickens, H. B., and D. M. Gray: Experience with a Pier-supported Building over
Permafrost, Proceedings of the American Society of Civil Engineers, Paper 2618,
Journal of Soil Mechanics and Foundations Division, October, 1960.
Draining Coarse Gravel with Wellpoints, Engineering News-Record, Nov. 14, 1946.
Gray, H.: Field Vane Shear Tests of Sensitive Cohesive Soils, Paper 2887, Transactions
of the American Society of Civil Engineers, vol. 122, 1957.
Green, N. B.: Design of Floating Slab Foundation, Journal of the American Concrete
Institute, vol. 28, no. 9, March, 1957.
Hogg, A. D.: Ice Pressure against Dams: Some Investigations in Canada, Paper 161,
Proceedings of the American Society of Civil Engineers, vol. 78, no. 161, December,
1952.
Hool, G. A., and W.
Kinne: Foundations, Abutments and Footings, 2d ed.,
S.
McGraw-Hill Book Company, Inc., New York, 1943.
Hopkins, D. A.: The Design of Piers, Jetties, and Dolphins, Paper 2846, Transactions
of the American Society of Civil Engineers, vol. 122, 1957.
Nagai, S.: Shock Pressures Exerted by Break Waves on Breakwaters, Proceedings of
the American Society of Civil Engineers, Paper 2504, Journal of the Water Works
Division, June, 1960.
Nebolsine, R. : Gound Dewatering for Construction, Engineering News-Record, Apr. 20,
1944.
Pengelley, C. D., E. J. Dower, and M. M. Lemcoe: Structural Model Studies of Con-
crete Slab Foundations, Journal of the American Concrete Institute, vol. 26, no. 10,
June, 1955.
Prentis, E. A., and L. White: "Cofferdams," Columbia University Press, New York,
1950.
, and : "Underpinning, Its Practice and Applications," Columbia Univer-
sity Press, New York, 1950.
 BIBLIOGRAPHY 709

Swiger, W. F.: Control of Ground Water in Excavations, Proceedings of the American

Society of Civil Engineers, Paper 2359, Journal of Construction Division, February,
1960.

Piles and Pile Foundations

An Estimation of the Point Resistance of a Pile, Proceedings of the American Society of

Civil Engineers, vol. 83, April, 1957.
Chellis, R. D.: "Pile Foundations," McGraw-Hill Book Company, Inc., New York,
1951.
Dodge, C. F., and W. F. Swiger: Vibration Testing of Friction Piles, Engineering
News-Record, May 13, 1948.
Dunham, J. W.: Pile Foundations for Buildings, Proceedings of the American Society
of Civil Engineers, vol. 80, no. 385, January, 1954.
Mansur, C. I., and J. A. Focht, Jr.: Pile-loading Tests, Morganza Floodway Control
Structure, Paper 2810, Transactions of the American Society of Civil Engineers,
vol. 121, 1956.
Mason, H. G., J. A. Bishop, L. A. Palmer, and P. P. Brown: Piles Subjected to Lateral
Earth Support, Symposium on Lateral Tests on Piles, Proceedings of the American
Society for Testing Materials, June, 1954.
Matlock, H., and L. C. Reese: Generalized Solutions for Laterally Loaded Piles,
Proceedings of the American Society of Civil Engineers, Paper 2626, Journal of the
Soil Mechanics and Foundations Division, October, 1960.
Meyerhof, G. G.: Compaction of Sands and Bearing Capacity of Piles, Proceedings
of the American Society of Civil Engineers, Paper 2292, Journal of the Soil Mechan-
ics and Foundations Division, December, 1959.
Pihlainen, J. A.: Pile Construction in Permafrost, Proceedings of the American Society
of Civil Engineers, Paper 2295, Journal of the Soil Mechanics and Foundations
Division, December, 1959.
Pile Foundations and Pile Structures, American Society of Civil Engineers Manual 27,
1946.
Richart, F. E., Jr.: Analysis for Sheet-pile Retaining Walls, Paper 2900, Transactions
of the American Society of Civil Engineers, 1957.
Symposium on Lateral Load Tests on Piles, American Society for Testing Materials,
Special Technical Publication, 154, 1953.
 INDEX

Abutments, 605-637 Ambassador Bridge, 572

approach slabs for, 631, 633 American Lumber and Treating Co., 656
for arches, 626-629 American Society of Civil Engineers, 321,
architecture of, 605-608, 625, 633 323, 348, 484
barn-door slabs for, 633 Anaconda Wire and Cable Co., 29, 191-
bearings on, 623-626 193
beveled-wing, 612 Anchor bolts, 217-227, 594
box, 616-618 embedment, 217-220
for continuous bridges, 608-620 expansion, 223, 224
contraction joints, 610, 611, 614-616 and pipe sleeves, 218-227
definition of, 604 shear on, 222
expansion joints, 610, 611, 614-616, 632 templates for setting, 220
behind, 605, 606, 631
fills Anchorages, for backstays, 263-267
flanking span, 618 for bulkheads, 419-425
floating, 615, 619 for caissons, 538, 539
locations, 604-608 dead-man, 423
planning, 605-608 for pipelines, 285-291
for rigid frames, 620-626 for suspension bridges, 628-631
for simply supported bridges, 608-620 Andersen, Paid, 561
straight-wing, 609-612 Anderson, A. B., 103
for suspension bridges, 628-631 Arches, foundations for, 626-629
U-shaped, 613-616 Architecture, of abutments, 605-608, 625,
waterproofing, 633 633
wingless, 618-620 of piers, 545, 569-578
Adobe, 10, 21 Artesian ground water, 43, 496, 498
Air locks, 518 Atwood, W. G., 324
Alameda Creek Bridge, 576
Allowable bearing pressures on soils,
86-92, 229, 584 Backstay anchorages, 263-267
for combined loads, 229, 584 Barrett waterproofing, 165
visualization of, 90, 402 Basement walls, 158-174, 400
Allowable friction of piles, 303 waterproofing, 163-168, 401
Allowable loads on piles, 362-364, 425- Bases, action of billets for, 227-229
440 of columns, 214-216
Allowable unit stresses, in concrete, 112, grouting, 224-227
695 Batter piles, 305, 411-416
in steel, 695 Bayonne Bridge, 227, 479, 628
711
712 INDEX

Beach deposits, 13 Caissons, false bottom, 527-529

Beams, grade, 144-148 floating, 521, 527, 528
Bearing value, of piles, 346-349, 362-364, Franki, 331
425-440 friction on, 513, 526
of soils, 21, 22, 86-92, 229, 584 guiding, 512, 536-539
Bearings, of bridges, 591-596, 623, 624, jacking, 534
626-629 jetting, 526
of columns, 96-99, 214-217, 227-229 materials for, 521-527
Beaulieu, A. G., 362, 364, 422, 424 Moran and Proctor, 529
Bedrock, 8, 15, 86-88 open, 511-515
excavation of, 16, 17 pipe, 531, 532
Bell Telephone Building caissons, 534 pneumatic, 515-521
Bentonite, 11,21 conversion to, 521
Bethlehem Steel Company, 233, 547, Powell method for, 533
572-575 righting, 539, 540
Bibliography, 707-709 rotating, 531
Blakeslee, C. W., 311, 336, 472 sinking, 512-515, 534
Bond, 112 sand island for, 534-536
Borers, marine, 324, 325 small, 531-534
Borings, 27-32 starting, 512-536
Boulders, definition of, 8, 18 telescoped, 531, 532
detection of, 18, 29, 30 Calder, G. J., 538
effecton pile driving, 18, 314, 315 Caliche, 10, 21
excavation of, 18 California Division of Highways, 576
Boussinesq equation, 54 Cananea Consolidated Copper Co., S.
Bridge, bearings, 591-596, 623, 624, America, 41, 187
626-629 Carquinez Strait Bridge caissons, 538
dead loads, 577-579 Cathodic protection of steel, 344
live loads, 578-580 Caughnawaga Bridge caissons, 537
piers (see Piers) Chellis, R. D., 322, 325, 348
wind loads, 580-582 Chemical consolidation of soil, 494, 644
(See also Abutments) Chesapeake City Bridge, 578
Brown, V. J., 527 Civil Engineering, 551, 553, 554
Bulb of pressure, 52, 53 Clay, 10, 19, 59, 69, 70
around piles, 303, 307, 425-440 Cobbles, definition of, 9
Bulkheads, 419-425 Cofferdams, 447-508
Buoyancy, 198-203 blow-out of, 454
braced, 464-477
bracing, 464-477
Cadling, L., 31 cantilevered, 456-464
Caissons, 511-540 cellular, 481-491
air locks for, 518 combined with caissons, 529, 530
anchors for, 538, 539 Cummings' method of analysis, 481-
belled bottoms, 531, 532 491
combined with cofferdams, 529, 530 current on, 491
compressed-air work in, 515-521 definition of, 451
construction problems, 534-540 details of, 483, 491-494
cutting edges, 522-526 double-wall, 477-480
definition of, 511 driving of piling, 456, 457
drilled-in, 358 floating, 480
effects of current on, 491, 536 and floods, 491
 INDEX 713

Cofferdams, friction on, 483, 486, 489 Dames and Moore, 28, 32, 89
grouting, 454 Dampproofing, 163
lengths of piling, 492 Davis, William Russell, 574
miscellaneous details, 483, 491-494 Definition of symbols, xiii

multibraced, 468-477 Des Moines River Bridge, 562

need for, 447, 451 Design, definition of, 2
Ohio River type, 479 Distribution of pressure in soils, 50-54,
pressures on, 452 72-76
unbalanced, 454, 464 Boussinesq equation for, 54
reuse of, 480, 492 bulb of pressure (assumed) in, 52
safety factor, 487, 492 large versus small areas, 52, 53, 91
sheet piling for, 456, 483, 492 for testing, 90
shoring, 448-451 Newmark 55
circle for testing,
single-braced, 464-468 around 301-307, 425-440
piles,
single-wall, 451-477 2:1 ratio for spread, 51, 52
stresses in, 481, 492 Dodge, C. F., 359
surcharge, effect of, 487, 506-508 Douglas Aircraft hangar, 233
unwatering, 454, 455, 491 Drainage of soils, around basements, 164
wales for, 452, 464-477 open pits for, 500-503
waves on, 491, 559, 560 piping for, 163
Cohesion in soils, 52, 57 pumping for, 497-503
Columns, steel bases for, 214-217, 227- wellpoints, 494-500
229 wells, 500-503
Combinations of loading, 229, 584 Dravo Corporation, 482
Combined footings, 120-137, 256-263, Dufresne Construction Co., Ltd., 564
381-385
special, 133-137
Compressed-air work in caissons, 515-521 Earth pressures, 161, 237, 271, 452, 703,
Concrete, formulas for, 96, 695 706
neutral axis of members, 695, 698 Earthquakes, 213, 582
reinforcement tables for, 696, 697 Economic considerations, 1, 4-6, 44
shrinkage, 139, 140, 153, 171, 276, 610- Electrolysis, 198, 342
616 cathodic protection, 344
unit stresses allowed in, 96, 112, 695 Embedment, 56- 59, 237, 238, 584
Cone of distribution, definition of, 52 Engineering, value of, 44-47
under footings, 51-54 Engineering A ews-Record, 314, 325, 359,
around piles, 303, 307, 425-440 401, 494, 498, 503, 514, 520, 528, 534,
Connecticut State Highway Department, 537, 538, 540
606 Erickson, E. L., 554
Consolidation of soils, 59, 67, 69-71 Erosion, 13-15, 549-559
chemical, 494, 644 Excavation, of bedrock, 16, 17
(See also Settlement) caissons, 511-540
Coulomb's formula, 271, 703, 706 chemical consolidation for, 494
Cowlitz River Bridge cofferdam, 473 cofferdams (see Cofferdams)
Crandell, F. J., 316 drainage, 494-503
Crane loads, 704, 705 freezing around, 492-494
Crosier, C. M., 101 grouting, 494
Cross, Hardy, 2, 5, 22, 182 large (see Cofferdams)
Cummings, A. E., 334 piecemeal methods, 503-506
Cummings, E. M., 484 shoring, 448-451
Current, pressure of, 491 soldier beams, 503-506
714 INDEX

Excavation, surcharge pressures, 506-508 Footings, isolated Spread footings)

(see

wellpoints, 494-500 lateral loads on, 229-242

Expansion of concrete, 140, 141, 285 lateral support of, 237-240, 250-252
Expansion bolts, 224, 657 with overturning, 213-291, 372-374,
Expansion joints, in abutments, 610, 611, 403-409
614-616, 632 pressure under, 50-56, 103-105, 229-
in bridges, 566, 593, 632 250, 584
in machinery foundations, 658 net, 231
in walls, 172, 276 settlement of (see Settlement)
Exploration of soils, 23-47 unsymmetrical, 242-250
borings for, 27-32 of walls, 143
wash, 28-31 Formulas for concrete analysis, 695
contract for, 39 Foundation walls, 139-175, 276-285
core, 29,30 Foundations, character of problems con-
deep, 27-32 cerning, 1, 4, 5
effect of, on location, 37-43, 545-549, Code of New York, 86
605-608 definition of, 2
on superstructure, 40-43, 568-572, engineering, 44-47
605-608 for machinery, 655-693
examples, 25, 26, 32-34, 37-44 materials for, properties of, 15-22
ground water in, 31, 43, 85, 281 Franki 331
piles,
importance of, 24-27, 39-43 Fraser, E. S., 427
load tests in, 34-36, 90 Freeman, G. L., 514, 537
log of boring, 29 Freezing, 62, 63, 281, 601
obstructions, 29, 30 around excavations, 492-494
planning 37-39
for, 27, 28, frost lenses, 62-64
preliminary, 27, 28, 37, 38 permafrost, 64
profiles from, 26, 32-34 Friction, assumed coefficients of, for
samples of soil, 30, 32 soils, 22, 271-273

shallow, 27 of bridge bearings, 583, 591-593

test pits, 27 on caissons, 513, 526
vane tests, 31 on interlocks of sheet piling, 483, 486
internal, for soils, 22, 58, 271-273
negative, on piles, 305
Fenders, 419, 458 skin, on piles, 301-303
Fetch, 560 Fungi, 322
Fill, compaction, 76, 78
consolidation, 81
construction of, 76 Garrison Dam, 565
as foundation, 76-81 General Electric Co., 359
hydraulic, 45, 46 Geology, 7, 8, 11
instead of superstructure, 605 George Washington Bridge anchorages,
around piles, 305 630, 631
settlement, 78, 84, 195, 402, 408, 421, Gilman, George T., 534
631 Glaciers, 12, 20
structures on, 76-81, 195, 402, 410, 615 Grade beams, 144-148
Floating foundations, 203, 401, 615, 619 Gravel, 9, 18
Floods (scour), 549-559 Greiner, J. E., 547
Footings, belled out, 312 Ground water, 31, 43, 85, 281
combined (see Combined footings) artesian, 43, 496, 498
conical, 103 drainage of, 163, 166, 188, 494-503
 INDEX 715

Groups of piles, 307-309, 320, 365-367, Koefoed, S. M., 480

425-440 Krynine, D. P., 5
Grouting, for bearings, 224-227, 596 Kuo, S. S., 662
chemical, 494, 644
column bases, 224-227
to consolidate soils, 494 Lateral loads, on footings, 229-242
to protect excavation, 494 on piles, 305, 372-374, 403-409
to seal cofferdams,454 on walls, 161, 162, 276-285
as underpinning, 644 Lateral pressures of earth, 161, 237, 271,
Gumbo, 11, 21 452, 453, 703, 706
Lateral support of footings, 237-240, 250-
252
Hardesty and Hanover, 576, 627 Laushey, L. M., 553
Hardpan, 10, 20, 87 Lazaro, A., Jr., 246
Hartford, F. D., 314, 554 Leaching, 1 1, 71

Hedman, A. F., 483 Lincoln Tunnel, approach to, 80

Helmers, N. F., 473, 535 structures of, 19, 515, 516, 519, 577,
Hidalgo-Reynosa Bridge, 550 608, 609

Howard, Needles, Tammen, and Bergen- Lion's Gate Bridge caissons, 530
doff, 565, 575 Load tests, diagrams of results, 36, 37,

Huey Long Bridge caissons, 535 60, 66

Humus, 10 of piles, 349-359
Hydraulic uplift, 198-203 90
of soil, 34-36, 60, 66,
Loads, on bridges, 577-582
of cranes, 704, 705
dead, of bridges, 577-579
Ice,560-568
live, of bridges, 578-580
Indian River Inlet Bridge, 562
wind, 580-582
Inkster Road Bridge, 620
Loam, 10, 21
Insulation, 63
Loess, 11, 21
Isolated footings (see Spread footings)
Log of boring, 29
Isolators for machinery, 663-665, 684
Lone Star Cement Corporation, 550
Luck, L. D., 323
Lumnite cement, 690
Jacking load tests, of piles, 353-359 Lupfer, Edward P., Corporation, 574
of soil, 35
Jacking piles, 312, 353
Jets, for caissons, 526
MacArthur piles, 328
for piles, 311, 394
Machinery foundations, 655-693
water, 311, 498 accessories, 658
Joints, in abutments, 610-616, 632
anchor bolts, 655, 667
expansion, at abutments, 632 on beams, 680-682
in superstructure, 78, 83, 389-392, 566,
compressor, 685-688
632 conveyor, 673-676, 678
in walls, 169-172, 389-392
cycles, 659
Joosten process, 494 damping, 663-665
Judgment, 2, 4, 5, 69, 352 embedment, 656
energy, 660
examples of difficulties, 665, 666
Keys in walls, 170-172 expansion, 658, 669
Kinney, E. E., 408 fan, 668, 679, 682, 683
716 INDEX

Machinery foundations, on floors, 657 Mats, two-way, 190

frequency, 659, 662 uniform, 177-189
formulas for, 662 Mattson, D. W., graphs, 248
furnace, 689-691 Mavis, F. J., 553
general principles, 655-658 Membrane waterproofing, 165-168
impulse, 659-663 Merriman Dam caissons, 520
insulation, 690 Merritt Parkway bridges, 621
isolation, 656, 663-665, 683, 684 Michigan State Highway Department,
isolators for, 663-665,684 620
Lumnite cement used 690
in, Mill Basin Bridge, 575
mass, effect of, 659-663 Mohr-Land circle, 246
overturning, 657, 673-676 Monotube piles, 328
on piles, 676-679 Moran and Proctor, 529
protection 658
of, Moretrench Corporation, 495, 497
for reciprocating machines, 659, 660 Muck, 11, 21, 61
for rotating machines, 659-661 Mud, 11, 21, 61
settlement, differential, 657, 669 Mud wave, 47, 60
effect of, 657, 661
vertical, 657
shock absorbers, 663-665 New Jersey State Highway Department,
on soil, 666-676 573, 574
on superstructure, 679-684 New York Code, 86
torque on, 679 Newmark circle, 54, 55
torsion on, 658, 688, 689 Niagara Falls Bridge Commission, 574,
turbo-generator, 665, 691-693 627, 629
vibration of, 658-666
formulas, 659-663
near walls, 672 Ohio Department of Highways, 341
Malkin, A. H., 462, 463 Open caissons, 511-515
Marine borers, 324, 325 Otto, A. L., 290
Marquardson, R. P. V., graphs, 248 Overburden, 13
Mats, 176-212 Overturning, anchor bolts, 217-227
analysis of, 177-189 anchors, for backstays, 263-267
diagonal tension in, 183 for pipelines, 285-291
electrolysis in, 198 in rock, 267, 281
floating, 203 base plates, action of, 214-216, 227-229
footings versus, 176, 396-403 bases of steel columns, 214-217
hydrostatic pressure on, 198-203 of boxlike structures, 281
local heating, 197, 198 for crane runways, 261-263
one-way, 190 eccentric loads, 242-250, 374-381
overturning applied to, 267-271 effect of weight of earth, 232, 254, 255
pavement versus, 188 of foundations, 213-291, 372-381, 403-
on piles, 385-389, 396-403 409, 657
planning of, 177-198, 203-209 grouting, 224-227
pressure under, 177, 178 isolated footings, 229-242, 374-381,
punching shear in, 183 403-411
reinforcement, 181-187 lateral loads on walls, 276-285
ribbed, 189-198 by pressure of earth, 161, 271
shear heads in, 188 lateral steadying, 237-240, 250-252
silos on, 205-209 lateral support by earth, 237-240, 254,
steel grillages in, 188 255, 263-267, 285-291, 403-409
 INDEX 717

Overturning, on mats, 267-271 Piers, choice of location, 545-549

and participation stresses, 259 collisions with, 582, 601
of pedestals on mats, 267-271 cost, 547, 568-572
of piers, 583-591 dead load,577-579
of pile foundations, 372-381, 403-409 drains for, 599
of pipe lines,285-291 and earthquakes, 582
pipe sleeves, 218-223 effect on, of current, 491, 549-559
of poles, 403-409 of ice, 560-568
of retaining walls, 271-285 of superstructures, 545-549, 568-572
on rock, 240, 280 of waves, 559, 561
safety factor against, 229 of wind, 580-582
by sliding, resistance to, 237, 273 embedment of, 584

on slopes, 260 fenders for, 421, 558, 598

of spread footings, 213-263, 372-381 forces oh, 572-583
pressure under footings, 230-233, freezing, 601
243-250 and 560-568
ice,

steadied laterally, 250-252 live loads,578-580

of towers, 252-263 location, 545-549
about two axes, 236, 237, 242-250 masonry facing, 563, 598
of unsymmetrical footings, 242-250 miscellaneous details, 598-602
approximate analysis, 243, 247-250 overturning, 583-591
Mattson graphical method, 248 planning, 544-549, 568-578
principal axes, 244-247 protection, 558, 598
uplift and, 252-263 reinforcement, 596-602
yield effect on, 239, 240 and scour, 549-559
for ships, 416-418
spalling, 596-602
Parsons, Brinckerhoff, Hall, and Mac- stability of, 583-591
donald, 283, 573, 578, 611 starling, 563, 567
Pea gravel, 9 streamlining, 555-559
Peace Bridge, 574 tops, 596-598
Peat, 11, 21 types, 568-572
Pebbles, 9 at water front, 416-418
Peck, R. B., 573 wave action on, 559, 560
Pedestals, 96-99 wind loads on, 580-582
as footings, 99-102 Pile drivers, 309-312, 319, 380, 457
on mats, 267-271 Pile foundations, 360-446
reinforcement of, 98, 99 allowable bearing value, 346-349, 362-
Permafrost, 64 364, 425-440
Phelps Dodge Corporation, 258, 615, 671, on batter 411-416
piles, 305,
678 for bulkheads, 419-425
Pie IX Boulevard Bridge, 564 choice of types, 360, 361
Piecemeal excavation, 503-506 combined footings, 381-385
Piers, 121,544-604 eccentric loads, 374-381
analysis of, 583-591 efficiency, 307-309, 425-440
architecture of, 545, 569-578 fenders, 419, 558
bearings, 591-596 groups of piles in, 307-309, 320, 365-
friction on, 583, 591-593 367, 425-440
grouting, 596 horizontal loads, 372-374, 403-411
544-604
for bridges, for isolated footings, 368-381
buoyancy, 417 for mats, 385-389
718 INDEX

Pile foundations, versus mats and foot- Piles, marine borers, 324, 325
396-403
ings, versus mats, 396-403
for piers, 416-418 Monotube, 328
sequence of driving, 392-396 negative friction on, 305
settlement of, 389-392, 410, 430-440 pedestal, 328
spacing of piles for, 365-367 pipe, 328, 329
Piles,300-359 poles, 407-409
action of, 301-307 precast concrete, 334-337
allowable bearing value, 346-349, 362- prestressed concrete, 334-337
364, 425-440 pull-out tests, 320, 357
areas of, Raymond, 327 Raymond, 327
auger, 330 records, 349-359
batter, 305, 411-416 sand, 70, 358
buckling, 304 screw, 305
bulk of pressure, 303, 307, 425-440 settlement of, 355, 389-392, 410, 430-

for bulkheads, 419-425 440

button-bottom, 329 sheet, 344-346
cast-in-place, 327-334 (See also Cofferdams)
choice of types, 360, 361 sizes and types, 700, 701
composite, 333, 358 skin friction on, 301-303
concrete, 327-337 above soft strata, 306
cone of distribution, 303, 307, 425-440 spacing, 365-367
creosoting, 323-326 steam hammer, 310
drilling holes for, 312 steel, 338-344, 702
driving, 309-320, 371, 392-396, 457 corrosion, 339-344
equipment, 309-312, 319, 380 driving, 314, 316, 338
formulas for, 346-349 electrolysis, 342
to refusal, 314 protection of, 342-344
vibration from, 315-320, 359, 394 cathodic, 344
vibrator-type, 310 sizes, 702
drop hammer, 309 step-taper, 327, 396
effect of, on soil, 313, 392-396 tests of, 349-359
effect of vibration on, 69, 359, 394, 678 uplift on, 320
efficiency of, 307-309, 425-440 for water-front piers, 416-418
end-bearing, 304-306 water jet, 311
failure,301 wooden, 321-326
surrounding, 305
fill creosoting, 323-326
versus footings, 396-403 driving, 321, 322, 325
formulas for, 346-349 fungi, 332
foundations on, 360-446 marine borers, 324, 325
Franki, 331 preservation of, 322-326
friction on, 301-303, 425-440 Pipe-line anchors, 285-291
function of, 300 Pipe sleeves, 218-223
group action of, 307-309, 320, 425 -440 Planning, definition of, 2
horizontal forces on, 305, 372-374, 403- effect of soil on, 37-43, 545-549, 605-
409 608
jacking, 312, 353, 356 value of, 39-43
jetting, 311 Pneumatic caissons, 515-521
lateral forces on, 305, 372-374, 403-409 Poles, 403-409
load tests, 349-359 Port of New York Authority, 26, 227,
Mac Arthur, 328, 331 516, 519, 577, 608, 609, 622, 628
 INDEX 719

Posey, C. J., 246, 551 Rock, bedrock, 8, 15, 86

Powell method of sinking caissons, 533 broken (see weathered, below)
Pressure, active, 162, 237, 271, 452, 703 excavation of, 16, 17
allowable on soil, 86-92, 229, 584 flour, 10
bulb of, 52 ledge (see bedrock, above)
of current, 491 walls against, 167, 280
distribution of, 50-56 weathered, 8, 12, 18
under footings, 50-56, 103-105, 230- Rutledge, P. C, 408
233, 243-250
of ice, 560-568
lateral, due to earth, 162, 237, 271, 452, Safety, factor of, 231, 273, 349, 352
703, 706 of structures, 544
passive, 237, 452, 706 St. Georges Bridge, 573
of saturated soils, 162, 273, 452 Samplers for soils, 30, 32
of surcharge, 487, 506-508 Samples of soils, 27, 29, 30, 73
water, 162, 163, 273, 453 San Francisco-Oakland Bridge caissons,
wave, 559-561 529
Principal axes of unsymmetrical footings, Sand, 9, 18
244-247 Sand island, 534-536
Proctor, C. 529
S., Sand piles, 70, 358
Profile, soil, 26,32-34 Santa Ana River Bridge, 550
Properties of soils, 15-22 Scour, 549-559
Pulaski Skyway, 574 Screw piles, 305
Purcell, Andrew, and Woodruff, 503, 528 Sedimentation, 14, 551-555
Selection of site, 23, 545-549, 604-608
Settlement, of clay, 59, 69, 70
Quicksand, 18, 59, 69, 85 danger signals, 61-71, 81-86
Quinn, Hugh, 86 of deep strata, 64, 74, 306
differential, 61, 62, 81-86, 389-392,
402, 621
Rainbow Bridge, 627, 629 effect on, of piles, 305, 425-440
Rankine Bridge, 283 of vibration, 69, 315-320, 359, 394
Rankine's formula for soils, 237, 273 estimate of, piles, 430-440
Raymond Concrete Pile Co., 316, 319, soil, 72-76

327, 380 examples of, abutment, 410

Raymond piles, 327, 396 boiler house, 195
sampler, 30 fill, 80, 195, 402, 410, 421
Rebound, 67, 68 tanks, 84, 85, 91, 402
Reinforced concrete, allowable stresses of fills, 76-81, 402, 408, 421
in, 96, 695 of granular soils, 59, 85
analysis of, 96, 695 load, diagrams of, 37, 65-68
formulas for, 695 and negligible unit pressures, 64
Reinforcement, data on, 695-697 of piles, 355, 389-392, 410, 430-440
multiplication tables for, 696, 697 record of, structure on mat, 195
in slabs, 696 sand, 59, 69
Relieving platforms, 419-425 silt, 59, 69
Retaining walls, 271-276 temporary load, 61
counterforted, 274-276 time versus, 37, 59, 60, 90, 195
Richart, F. E., 108 voids ratio, 65, 92
Rigid-frame bridges, foundations for, Shale, 8, 16, 87
620-626 Shark River Inlet Bridge, 575
720 INDEX

Shear, allowable stresses, 111, 695 Soils, slopes, 58

diagonal tension, 107-113, 234-236 strata, 11-15, 32-34
punching, 101, 106, 109-111, 234-236 subsoil, 10
of soil, 52, 56-59 symbols for, 8
Sheet piles, 344-346 tests, 22, 34-37, 64-69, 90
data on, 700, 701 vane, 31
details of, 344, 483 topsoil, 10
driving, 456, 457 types, 7-11
friction at interlocks, 483 voids ratio, 65, 68-70, 92
lengths, 492 Soldier beams, 503-506
stresses in, 481, 492 Spread footings, 95-138
Shelby tubes, 30 analysis of, 102-119
Shoring, 448-451 assumptions, 102-105
Shrinkage, of concrete, 139, 140, 153, 171, bending, 108
276, 610-616 bond, 112
of walls, 139, 140, 153 circular, 105-107
Silos on mats, 205-209 combined, 120-137, 381-385
Silt, 6, 10, 19 deformation of, 103-110
Skin friction on piles, 303 design of, 102-119
Soil mechanics, 7, 22, 64 diagonal tension in, 107-113, 234-236
Soils,bearing value, 21, 22, 86-92, 229, eccentric loads on, 242-250, 374-381
584 and grade beams, 144-148
characteristics, 8-11, 15-22 overturning of (see Overturning)
classification, 9-11 pedestals for, 99-102
clay, 10, 19, 59, 64, 70 on piles, 368-381, 425-440
cohesion, 52, 57 pressures under, 50-56, 103-105, 107,
compaction, 69 229-250, 584
consolidation, 59, 67, 69-71 punching shear in, 101, 106, 109-111,
chemical, 494, 644 234-236
distribution of pressure in, 50-56, 72- rectangular, 108-114, 116-119
76 reinforcement, 110-119
drainage, 164, 494-503 on rock, 102, 112, 240
exploration, 23-47 sloped, 112, 114-116
formed in situ, 12 square, 107-116
friction angle, 22, 58, 271-273 stepped, 113
laboratory tests, 22, 64 trapezoidal, 100, 101, 240, 254
lateral pressure of, 162, 271, 452 types, 100, 104-113
active, 271, 452, 703 unsymmetrical, 242-250
passive, 237, 452, 706 Starling, 563, 567
load tests, 34-36, 60, 64-69, 90 Steel piles (see Piles)
origin, 11-15 Stewart, Ralph W., 550, 554, 557
profiles, 26, 32-34 Stickel Bridge, 576
properties, 15-22 Storstrom Bridge cofferdams, 480
samplers for, 30-32 Strata, soil, 11-15, 32-34
samples, 27, 29, 30, 73 Streamlining of piers, 555-559
sand, 9, 18 Stresses, allowable unit, 112, 695
sedimentation, 14, 551-555 Subsoil, 10
settlement, 59-76, 85, 195, 306, 389- Substructure, definition of, 3
392, 402, 425-440 Subsurface exploration (see Exploration
shearing, 52, 56-59 of soils)
silt, 10, 59, 69 Superstructure, definition of, 3
 INDEX 721

Superstructure, effect of, on foundations, Vibration, some formulas for, 662

4, 40, 73, 544, 568 Voids, 65, 70
effect of soil on, 4, 24-27, 39, 568-572 Voids ratio, 65, 68, 92
Surcharge, 425, 450, 506-508
Suspension-bridge anchorages, 628-631
Susquehanna River Bridge, 547 Waddell and Hardesty, 575
Swiger, W. F., 359 Walls, 139-175
basement, 158-163
concentrated loads on, 148-158
Taeoma Narrows Bridge, 527 contraction, 139-141, 169-172
Talus, 12 counterforted, 274-276
Taylor, E. C, 233 drains, 164
Taylor, E. W., 233 expansion, 140, 141, 172
"Telltale," 31 footings, 143
Temperature effects, 140, 141 grade beams for, 144-148
Termites, 174 joints in, 169-172, 276
Terzaghi, K, 498 keys in, 170-172
Tests, laboratory, 22, 64-69, 89 lateral loads on, 161, 162, 276-285
of piles, 349-359 ordinary, 139-144
of soils, 34-36, 60, 64-69, 90 planning of, 148-151, 155-157, 169-
Texas Highway Department, 312 174, 276-285
Thomas A. Edison Bridge, 573 reinforcement, 140-144, 152-154, 159,
Till, 10, 21 173
Time-settlement curves, 37, 60 retaining, 271-276
Topsoil, 10 against rock, 166, 280
Tschebotarioff, G. R., 31 shrinkage, 139, 140, 153
Two-way slabs or mats, 699 to support mats, 185, 190, 193, 190
Types of foundation materials, 7-11 temperature, 140, 141
termite protection, 174
waterproofing, 163-169, '202, 281
Underpinning, 638-654
Water, artesian, 43, 496, 498
adjustment for, 639
drainage, 164, 188, 494-503
cantilevered, 650, 652
effect on soils, 12, 18-20, 60, 364
of columns, 641-645
ground, 31, 43, 85, 281
definition of, 638
jets, 311, 526
grout for, 644
as pumps, 498
necessity for, 638
pressure of, lateral, 162, 273, 453
piles for, 650-653
under mats, 198-203, 401
principles of, 639-641
protection of basements against, 163-
of walls, 645-653
169, 202, 281, 401
United States Rubber Co., 664
Uplift, 252-263
pumping, 494-503
on piles, 320, 357 scour caused by, 549-559
sumps, 500-503
transporting power of, 13, 549-559
Vane tests of soils, 31 waves, 559, 560
Varved clay, 10 wellpoints, 494-500
Vibration, effect of, 69, 315-320, 359, Waterproofing, abutments, 633
392-396 basements, 163-169, 202, 401
of machines, 658-666 and dampproofing, 163
from pile driving, 315-320, 392-396 mats, 202, 401
722 INDEX

Waterproofing, membrane, 165-168 Western Foundation Corporation, 329,

protection of, 165, 166 333, 380
walls, 163-168, 202, 281 Wheeler, Walter H., 187, 188
Waterstops, 172 Whitestone Bridge caissons, 514
Waves, 13, 491, 559 Wilbur Cross Parkway, 606, 621
Weathering, 8, 11, 12 Wind loads, 229, 580-582
Wellpoints, 494-500 Wood Preserving News, 324, 408, 426
Wells, 500-503 Wooden piles, 321-326
Westcott and Mapes, Inc., 310, 315, 354, Wooden sheeting, 345
355, 396, 423, 471 Wooden shoring, 448-451
Western Concrete Pile Co., 329, 359 Woodruff, G. B., 540
 UNIVERSITY OF FLORIDA

3 1262 05222 4606

HECKMAN
BINDERY INC.

JAN 94