Structural Overlay Design Using NDT Methods

Zhong Wu, Ph.D., P.E. Louisiana Transportation Research Center 2007 Transportation Engineering Conference Baton Rouge, February 11-14, 2007

Outline

Background Objective Overview of Overlay Design Methods Research Projects Summary Recommendations

Background

Asphalt overlay has been considered as

simplest and fastest means of restoring the distressed surfaces of rigid and flexible pavements

A quality overlay design

improve the roadways rideability restore the pavement structural stability.

Existing subgrade condition Existing pavement strength Engineering judgment

Structural overlay thickness design requests

Background (cont)

Current DOTD overlay thickness design

the 1993 AASHTO Pavement design guide software (DARWin)

One required design input, subgrade resilient modulus value, Mr,

the pre-assigned parish-map values not base on in-situ soil properties

Uses component analysis (layer co-efficients) method to determine the existing pavements structure number SNeff = aihi Such method can lead to design errors (over- or underestimated overlay thickness)

since design values do not represent actual field conditions

Overlay Design Example

150
Typical pavement section Hot mix asphalt D1=102 mm, a1=0.0165/mm Base course D2=241 mm, a2=0.0063/mm Subbase D3=457 mm, a3=0.0040/mm Subgrade soil

100
Difference in overlay thickness (mm)

Mr = 2,000 psi => Underestimated AC thickness of 1.5 in

50

0 -30 -20 -10 -50

10

20

30

40 50 60 70 80 Change in Mr (MPa)

Mr = -2,000 psi => Overestimated AC thickness of 2.0 in

-100

W18=5,000,000 ESALs R=95 % PSI=1.9 S0=0.35 Design Mr=34.5 MPa Design SN=5

-150 -4000 -2000

2000

4000

6000 8000 10000 Change in Mr (psi)

Objective

to establish a methodology for mechanistic pavement overlay design, based on

in-situ pavement conditions, and utilizing non destructive test (NDT) methods, specifically the FWD and/or Dynaflect.

Dynatest 8002 model Falling Weight Deflectometer

Dynaflect

Overview of Overlay Design Methods

Effective Thickness (ET) Approach

1993 AASHTO Pavement Design Guide Asphalt Institute (AI) ET Method (MS-17) AI Benkelman Beam Deflection Method (MS-17) Caltran Flexible Pavement Rehabilitation EVERPAVE (WsDOT) New M-E Pavement Design Guide

Deflection-based Approach

Mechanistic-Empirical (M-E) Approach

Effective Thickness (ET) Approach - Asphalt Institute (MS-17)

Thickness of Overlay = Tn - Te
Tn, new pavement thickness, determined from AI Design Chart for Full-depth Asphalt Concrete, using ESALd and Mr Te, effective thickness of existing pavement structure Te= Cihi

where, hi=thickness of the ith layer of the existing pavement; Ci=conversion factor associated with the ith existing layer

Effective Thickness (ET) Approach -1993 AASHTO Pavement Design Guide

Overlay Thickness Equation

hOL SNOL aOL SN f SNeff aOL

SNf from AASHTO pavement design equation where Mr is a required input, which can be determined from

Laboratory Testing Backcalculation from NDT measurements Approximate relationships (used by DOTD) NDT method Component analysis method (used by DOTD) Remaining life method

The effective structure number of existing pavement, SNeff

Deflection-Based Approach - Asphalt Institute Benkelman Beam (MS-17)

Pavement is modeled as a two-layer system

Layer 1: AC Overlay Layer 2: Existing pavement

Overlay thickness is determined as following steps:

Determine Representative Rebound Deflection (RRD)

Layer 2s elastic modulus is determined from RRD.

Compute Design Rebound Deflection (DRD) based on the allowable ESAL: DRD=1.0363 (ESAL)-0.2438

Other Deflections (e.g. FWD, Dynaflect) can be converted into Benkelman beam deflections, such as
Benkelman Beam = 1.61 * FWD Benkelman Beam = 20.63 * Dynaflect

Mechanistic-Empirical (M-E) Approach

Modeled pavement as multi-layered elastic or visco-elastic structure Pavement materials described by their stiffness and strengths at different times of the year Determine the critical stress, strain, or deflection by mechanistic methods Predict resulting damages by empirical failure criteria, e.g. fatigue cracking, rutting.

Mechanistic-Empirical (M-E) Approach -EVERPAVE

Developed by Washington DOT Steps:

Backcalculate layer moduli using FWD data Analyze and determine the two failure criteria parameters.

Fatigue cracking Rutting

log N f 14.82 3.291log( t ) 0.854log(Eac )

log N f 1.077 1018 ( V ) 4.4843

Compute allowable repetitions to failure at each season Compute damage at each season and sum the seasonal damage ratio. Determine the overlay thickness based on the sum of the damage ratio is less than or equal to one.

EVERPAVE Design Input

New M-E Pavement Design Guide

Developed under NCHRP 1-37A New Traffic input Enhanced Integrated Climatic Model (EICM)

Season variations

New Distress Models Need Calibration

Dr. Matthew W. Witczak (2003)

NDT Overlay Design Survey

State Arkansas Mississippi Alabama Maryland Virginia California North Carolina South Carolina Idaho Minnesota Oregon Texas Washington Method Equivalent Thickness Equivalent Thickness 1993 AASHTO 1993 AASHTO 1993 AASHTO Deflection-based AI Deflection-based Deflection-based M-E M-E M-E M-E M-E ELMOD5 Spreadsheet program Spreadsheet program (VDOT) Spreadsheet program Design Manual Spreadsheet program Spreadsheet program WinFlex MNPAVE / FPS-19W EVERPAVE Software ROADHOG

Project Selection

Four in-service pavements

I-12 (ESALd=24,400,000, life=15yrs) LA28 (ESALd=1,513,000, life=10yrs) LA74 (ESALd=700,590, life=10yrs) LA44 (ESALd=353,256, life=10yrs)

Each project about 3 to 5 miles long

Design Plan

Based on current DOTD overlay design method (Mr-parish map, SNeff-estimated)

I-12 4.5 AC overlay + 2 cold planning LA28 4.5 AC Overlay + 2 cold planning LA44 3.5 AC overlay + 2 cold planning LA74 3.5 AC overlay + 2 cold planning

NDT Tests

FWD and Dynaflect tests were performed on each project site

at 0.1 mile interval on both traffic directions

FWD Load
8 4 6 6 12 12 12 12

d1

d2

d3

d4

d5

d6

d7

d8

d9

Dynaflect Deflection Analysis

Kinchen and Temple (1980) developed a Pavement Evaluation Chart for Louisiana

SN of existing pavements Subgrade Modulus

Routinely use in pavement research projects

Pavement Evaluation Chart

FWD Deflection (mils)

10

20

30

40

50

0
0.1 0.3 0.5 0.7 0.9 1.11 1.3 1.5 1.7 1.9 2.1 2.3 2.5 2.7 2.9 3.1

LA 74

FWD Analysis (D0 & D9)

Station (miles)
FWD Deflection (mils)
D1 (EB) D0 (W B) D9 (EB) D9 (W B)

10

15

20

25

0
1.0 1.2 1.4 1.6 1.8 2.0 3.1 3.3 3.5 3.7 3.9 6.0 6.2 6.4 6.6 6.8 7.0

LA 44

Station (miles)

D9 (SB)

D0 (SB)

D9 (NB)

D0 (NB)

Dynaflect (SN)

Overlay Thickness
Deflection-Based Approach (AI Method)
8
Overlay Thickness (in)

6 4.5" 4 3" 3" 2" 2 0 I-12 LA28 Project 4.5" 4.5" 3.5" 4"

Current Plan AI (NB/EB) AI (SB/WB)

3.5" 3" 2" 2"

LA74

LA44

Equivalent Thickness Method (Arkansas ROADHOG)

8
Overlay Thickness (in)
Current Plan ROADHOG(NB/EB) ROADHOG(SB/WB)

6 4

4.5"

4.5" 4.2" 3.5" 2.7" 2.9" 2.2"

3.5" 2.8" 2.6"

2" 2" 2 0 I-12

LA28 Project

LA74

LA44

Based on M-E Design Approach

Only 0 or 1 overlay thickness required for all four projects. Possible explanations:

Backcalculated modulus too high Default values used in distress models (no Calibrated) Not fully understand how to choose a representative design value

Summary (Deflection-based method)

Simple to use (e.g. AI method) Needs to verify and calibrate the relationship between FWD (or Dynaflect) measured deflections and BB rebound deflections Relationship between allowable rebound deflection and ESALd also needs to be verified and calibrated

Summary (Equivalent Thickness method)

Simple to use (e.g. AASHTO and ROADHOG) 1993 AASHTO NDT-based method generally underestimate the overlay thickness, due to overestimate the existing pavement SN. ROADHOG uses its own relationship in estimation of SNeff.

Such relationships between SNeff and delta(D) may or may not be applicable to Louisiana condition

Summary (M-E design method)

Complicate to use. M-E-based overlay design method needs sophisticate inputs, which usually are not available directly from in-situ NDT tests The fatigue and rutting models used in any ME base design software must be verified and calibrated before any locally implementation.

Proposed NDT-based Overlay Design Procedure for Louisiana

Use Effective Thickness approach The future Structure Number (SNfuture) determined from 1993 AASHTO design equation

Mr determined from in-situ tests (DCP, FWD or Dynaflect) If FWD used, SNeff (FWD) needs to be scaled down to SNeff (Dynaflect) for Louisiana Condition

SNeff determined from FWD or Dynaflect test

Overlay thickness = (SNfuture-SNeff)/aAC

Future Works

Further validate the Dynaflect deflection determined SNeff The proposed overlay design procedure will be automated into a EXCEL spreadsheetbased program