Master of Science (Mathematics) (DDE)

Semester – III
Paper Code – 21MAT23C3

FLUID DYNAMICS

DIRECTORATE OF DISTANCE EDUCATION

MAHARSHI DAYANAND UNIVERSITY, ROHTAK
(A State University established under Haryana Act No. XXV of 1975)
NAAC 'A+’ Grade Accredited University
Material Production
Content Writer: Dr. Poonam Redhu

Copyright © 2022, Maharshi Dayanand University, ROHTAK

All Rights Reserved. No part of this publication may be reproduced or stored in a retrieval system or transmitted
in any form or by any means; electronic, mechanical, photocopying, recording or otherwise, without the written
permission of the copyright holder.
Maharshi Dayanand University
ROHTAK – 124 001

ISBN :
Price :
Publisher: Maharshi Dayanand University Press
Publication Year : 2022
 Paper Code: 21MAT23C3
Fluid Dynamics
M. Marks = 100 Term End Examination = 80
Time = 3 Hours Assignment = 20
Course Outcomes
Students would be able to:
CO1 Be familiar with continuum model of fluid flow and classify fluid/flows based on physical properties of a
fluid/flow along with Eularian and Lagrangian descriptions of fluid motion.
CO2 Derive and solve equation of continuity, equations of motion, vorticity equation, equation of the moving
boundary surface, pressure equation and equation of impulsive action for a moving inviscid fluid.
CO3 Calculate velocity fields and forces on bodies for simple steady and unsteady flow including those derived
from potentials.
CO4 Understand the concepts of velocity potential, stream function and complex potential, and their use in
solving two-dimensional flow problems applying complex-variable techniques.
CO5 Represent mathematically the potentials of source, sink and doublets in two-dimensions as well as three-
dimensions, and study their images in impermeable surfaces.
Section – I
Kinematics - Velocity at a point of a fluid,Eulerian and Lagrangian methods,Streamlines, path lines and
streak lines,Velocity potential, Irrotational and rotational motions, Vorticityand circulation, Equation of
continuity, Boundary surfaces, Acceleration at a point of afluid, Components of acceleration in
cylindrical and spherical polar co-ordiantes.
Section – II
Pressure at a point of a moving fluid, Euler equation of motion, Equations of motion incylindrical and
spherical polar co-ordinates, Bernoulli equation, Impulsive motion, Kelvincirculation theorem, Vorticity
equation, Energy equation for incompressible flow, Kineticenergy of irrotational flow, Kelvin minimum
energy theorem, Kinetic energy of infinitefluid. Uniqueness theorems.
Section – III
Axially symmetric flows, Liquid streaming part a fixed sphere, Motion of a sphere through aliquid at
rest at infinity, Equation of motion of a sphere, Kinetic energy generated byimpulsive motion, Motion of
two concentric spheres,Three-dimensional sources, sinks and doublets, Images of sources, sinks and
doublets in rigidimpermeable infinite plane and in impermeable spherical surface.
Section-IV
Two dimensional motion; Use of cylindrical polar co-ordinates, Stream function,Axisymmetric flow,
Stoke stream function, Stoke stream function of basic flows,Irrotational motion in two-dimensions,
Complex velocity potential, Milne-Thomson circletheorem, Two-dimensional sources, sinks, doublets
and their images, Blasius theorem.
Books Recommended:
1. W.H. Besaint and A.S. Ramasey, A Treatise on Hydromechanics, Part II, CBSPublishers, Delhi, 1988.
2. F. Chorlton, Text Book of Fluid Dynamics, C.B.S. Publishers, Delhi, 1985
3. O’Neill, M.E. and Chorlton, F., Ideal and Incompressible Fluid Dynamics, EllisHorwood Limited, 1986.
4. R.K. Rathy, An Introduction to Fluid Dynamics, Oxford and IBH PublishingCompany, New Delhi, 1976.
5. G.K. Batchelor, An Introduction to Fluid Mechanics, Foundation Books, NewDelhi,1994.
 Contents

Unit Title of Chapter Page No.

1 Eulerian and Lagrangian Descriptions of Fluid Motion 5-56

and Conservation of Mass

2 Pressure and Equation of Impulsive Action 57-101

3 Axially Symmetric Flow and Three-dimensional Sources, 102-133

Sinks and Double

4 Two-dimensional Motion and Stoke’s Stream Function 134-172

⚝⚝⚝
 UNIT - I
EULERIAN AND LAGRANGIAN DESCRIPTIONS OF FLUID AND
CONSERVATION OF MASS

Content
1.1 Objective
1.2 Elementary Concepts
1.3 Introduction
1.4 Kinematics - Lagrangian and Eulerian Methods
1.5 Streamlines, Pathlines and Streaklines
1.6 Velocity Potential
1.7 Vorticity
1.8 Flow and Circulation
1.9 Equation of Continuity
1.10 Boundary Surface
1.11 Acceleration at a Point of a Fluid
1.12 Check Your Progress
1.13 Summary of the Unit
1.14 Keyword
1.15 Self-Assessment Test
1.16 Suggested Reading
1.1 Objective
In this unit reader will learn about Eulerian and Lagrangian method, Conservation of Mass, Flow lines,
Velocity Potential, Rotational and Irrotational Motion, Boundary Surface and Acceleration.
1.2 Elementary Concept
Vectors are used to express physical quantities that are direction dependent, such as force, displacement,
and velocity. Other physical quantities, such as stress and strain, require more than one direction to fully
describe them, and these are referred to as tensors.
Some useful results on vectors are as follows:
 
(i) Let q = ui + vj + wk = (u, v, w) be a non-zero vector, then q = q = u 2 + v 2 + w 2
6 Partial Differential Equations

 q u v w
A unit vector q in the direction of q is given by q = = i + j + k , where
q q q q
u v w 
= cos= α l,= cos= β m, = cos= γ n are called the d.c.’s of q ; α , β , γ denote the angles which
q q q

q makes with the coordinate axes. Clearly, l2 + m 2 + n 2 = 1.
 
(ii) If a and b are two non-zero vectors, then
  
= a.b ab cos θ= , a × b ab sin θ n
   
where θ is the acute angle between a and b , n being the unit vector normal to the plane of a and b .
     
Here a and b , n form a right handed screw system and thus a × b =− b × a ( )
   
Note that a . b =0 and a × b =0 are respectively the conditions of perpendicularly and parallelism.
        
(iii) If a , b and c are nonzero vectors, then a . b × c ≡ a b c  is called scalar triple product.
           
Note that a b c  =  b c a  &  b a c  = − a b c 
  
Similarly, the vector quantity a . b × c is defined as the vector triple product which can be represented as
      
a . b × c = (a.c)b − (a.b)c

∂φ  ∂φ  ∂φ
(iv) ∇φ= i + j +k , where φ is a scalar function, is called gradient of the scalar φ
∂x ∂y ∂z
∂ ∂ ∂
∇ i + j + k
and = is a vector (operator).
∂x ∂y ∂z
The total differential of φ (x, y, z) is

∂φ ∂φ ∂φ
dφ = dx + dy + dz
∂x ∂y ∂z
 ∂φ ∂φ  ∂φ  
=  i +
 ∂x ∂y
j+
∂z 
(
k  . dxi + dyj + dzk )
In vector form, this gives
  
dφ = ∇φ .dr where  dφ = ∂φ dx + ∂φ dy + ∂φ dz
r = xi + yj + zk , dr = idx + jdy + kdz, and
∂x ∂y ∂z
∂φ  ∂φ  ∂φ 
∇φ= i+ j+ k
∂x ∂y ∂z
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 7
  ∂u ∂v ∂w 
(v) divq = ∇.q = + + ,q = (u, v, w)
∂x ∂y ∂z
 
If ∇.q = 0 , then q is said to be solenoidal vector.

∂ 2φ ∂ 2φ ∂ 2φ 2
Also, ∇.∇φ = ∇ φ = 2 + 2 + 2 , ∇ is Laplacian operator.
2

∂x ∂y ∂z

i j k
  ∂ ∂ ∂
(vi) Curlq = ∇ × q =
∂x ∂y ∂z
u v w

 ∂w ∂v    ∂u ∂w    ∂v ∂u 
= i  −  + j − + k − 
 ∂y ∂z   ∂z ∂x   ∂x ∂y 
(vii) (a) Gradient of a scalar is a vector.
(b) Divergence of a scalar and curl of a scalar are meaningless.
(c) Divergence of a vector is a scalar and curl of a vector is a vector.

(viii)=
curlgradφ 0,= divcurlq 0
     
= graddivq − ∇ 2 q i.e., =
Curlcurlq ∇ 2 q graddivq − curlcurlq

(ix) Gauss’s Divergence Theorem

  
(a) ∫ q.dS = ∫ divqdV
S V

 
(b) ∫ n × qdS =∫ curlqdV
S V

(x) Green’s Theorem

 
(a) ∫ ∇φ .∇ψ dV = ∫ φ∇ψ .dS − ∫ φ∇ ψ dV =
2
∫ψ∇φ.dS − ∫ψ∇ φ.dV
2

V S V S V

 ∂ψ ∂φ 
(b) ∫ (φ∇ 2ψ −ψ∇ 2φ=
).dV ∫  φ ∂n −ψ ∂n dS
V V

   

(xi) Stoke’s Theorem
= ∫ q.dr
C
∫=
curlq.dS ∫ curlq.ndS
S S

(xii) Orthogonal Curvilinear Co-Ordinates:

Let there be three orthogonal families of surfaces
 8 Partial Differential Equations

f1 (x, y, z) α=
= , f 2 (x, y, z) β=
, f 3 (x, y, z) γ (1)

where x, y, z are cartesian co-ordinates of a point P(x, y, z) in space. The surfaces

α=constant, β=constant, γ=constant (2)

form an orthogonal system in which every pair of surfaces is an orthogonal system. The values α , β , γ
are called orthogonal curvilinear co-ordinates. From equations in (1), we can get
= α , β , γ ), y y(
x x(= =α , β , γ ), z z(α , β , γ )
The surfaces (2) are called co-ordinate surfaces.

Let r be the position vector of the point P ( x, y, z )
 
i.e., r = xi + yj + zk = r(α , β , γ )

∂r
A tangent vector to the α -curve ( β =constant, γ =constant) at P is . A unit tangent vector is
∂α

∂r   2 2 2
 ∂α ∂ r  ∂ r  ∂x   ∂y   ∂z 
e1 =  or = h1 e1 where h 1 = =   +  + 
∂r ∂α ∂α  ∂α   ∂α   ∂α 
∂α

Similarly, e2 , e3 are unit vectors along β -curve and γ -curve, respectively such that
 
∂r  r h e
∂
= h= 2 e2 , 3 3
∂β ∂γ
  
 ∂r ∂r ∂r
Further, dr = dα + dβ + dγ = h1 e1dα + h 2 e2 dβ + h 3 e3dγ
∂α ∂β ∂γ
 
Therefore, (ds) 2 = dr.dr = h12 dα 2 + h 2 2 dβ 2 + h 32 dγ 2 where h1dα , h 2 dβ , h 3dγ are arc lengths along
α , β , γ - curves.
In orthogonal curvilinear co-ordinates, we have the following results.

 1 ∂φ 1 ∂φ 1 ∂φ 
(a) gradφ =  , , 
 h1 ∂α h 2 ∂β h 3 ∂γ 
  1  ∂ ∂ ∂ 
(b) If q = (q1 , q 2 , q 3 ) , then=
divq  ∂α (h 2 h 3q1 ) + ∂β (h1h 3q 2 ) + ∂γ (h1h 2 q 3 ) 
h 1h 2 h 3  
 
(c) If curlq= ξ= (ξ1 , ξ 2 , ξ3 ), then
 Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 9

1  ∂ ∂ 
=ξ1  ∂β (h 3q 3 ) − ∂γ (h 2 q 2 ) 
h 2h3  

1 ∂ ∂ 
=ξ2  (h1q1 ) − (h 3q 3 ) 
h 3 h1  ∂γ ∂α 

1  ∂ ∂ 
=ξ3  (h 2 q 2 ) − (h1q1 ) 
h 2 h1  ∂α ∂β 

1  ∂  h 2 h 3 ∂φ  ∂  h1h 3 ∂φ  ∂  h 2 h1 ∂φ  
=(d) ∇ 2φ   +  +  
h 1h 2 h 3  ∂α  h 1 ∂α  ∂β  h 2 ∂β  ∂γ  h 3 ∂γ  
(xiii) The Cartesian co-ordinate system (x, y, z) is the simplest of all orthogonal co-ordinate systems. In
many problems involving vector field theory, it is convenient to work with other two most common
orthogonal co-ordinates i.e. cylindrical polar co-ordinates and spherical polar co-ordinates denoted
respectively by (r, θ , z) and (r, θ ,ψ ) .

(a) Cylindrical Polar Co-ordinates (r, θ , z)

Here ( α, β, γ ) ≡ ( r, θ, z ) , related to cartesian co-ordinates by

x= r cos θ, y= r sin θ, z= z
(r ≥ 0, 0 ≤ θ ≤ 2π, −∞ < z < ∞)

h1 1,=
Scalar factors are= h 2 r,=
h 3 1 and

volume element is dxdydz

= rdrdθdz .

(b) Spherical Polar Co-ordinates (r, θ, ψ )

Here ( α, β, γ ) ≡ ( r, θ, ψ ) , related to cartesian co-ordinates by

x= r sin θ cos ψ, y= r sin θ sin ψ, z= r cos θ

(r ≥ 0, 0 ≤ θ ≤ π, 0 ≤ ψ ≤ 2π)

Scalar factors are h1 = 1, h2 = r, h3= r sin θ and volume element

is dxdydz
= r 2 sin θdrdθdψ

1.3. Introduction
1.3.1 Fluid Dynamics
The branch of science that deals with fluids in motion is known
as fluid dynamics. A substance that may flow is referred to as a
fluid; otherwise, it is referred to as a solid. There are two types
of fluids: (i) incompressible liquids (those whose volumes do not change when the pressure changes)
and (ii) compressible gases (those whose volumes do change when the pressure changes). The science of
10 Partial Differential Equations

moving incompressible fluids is often referred to as hydrodynamics. However, there are no sharp
distinctions between the three states of matter, i.e. solid, liquid and gases.
In reality, there is not much of a difference between the three states of matter. Matter is thought to be
made up of molecules in random relative motion under the influence of intermolecular forces in a
microscopic view of fluids. The distance between molecules in solids is small, and this spacing persists
even under intense molecular forces. The spacing between molecules is larger in liquids, resulting in a
less well-ordered arrangement of molecules and weaker intermolecular forces; nevertheless,
intermolecular forces are much weaker in gases, and molecules appear in disordered arrays with very
high intermolecular spacing.
If we imagine that our microscope, which has observed the molecular structure of matter, has a variable
focal length, we could change our observation of matter from a fine detailed microscopic viewpoint to a
longer range macroscopic viewpoint, in which the gaps between the molecules would not be visible and
the matter would appear to be continuously distributed. We will use this macroscopic view of fluids
throughout our research, in which physical quantities related with fluids inside a particular volume V are
assumed to be distributed continuously and uniformly within a sufficiently small volume V. The
Continuum hypothesis is the name given to this discovery. It means that at any location of a fluid, we
can specify a different velocity, pressure, density, and so on. Moreover, for a continuous or ideal fluid
we can define a fluid particle as the fluid contained within an infinitesimal volume whose size is so
small that it may be regarded as a geometrical point.
1.3.2 Stresses
A fluid element is subjected to two types of forces. Body
force is one of them, and surface force is the other. The
surface force is proportional to the mass of the body on
which it acts, whereas the body force is proportional to the

mass of the body on which it operates. Suppose F is the
surface force acting on an elementary surface area dS at a
point P of the surface S. Let F1 and F2 be resolved parts of

F in the directions of the tangent and normal at P. The
normal force per unit area is called the normal stress and is
also called pressure. The tangential force per unit area is
called the shearing stress.
1.3.3 Viscosity
The internal friction between the fluid's particles provides resistance to deformation. Tangential and
shearing forces are the source of friction (stresses). Fluids with this attribute are known as viscous or
real fluids, while those without it are known as inviscid, ideal, or perfect fluids.
Although all fluids are real, viscous effects are often ignored when the rates of variation in fluid velocity
with distance are small.
 Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 11
The body force per unit area at every point of surface of an ideal fluid acts along the normal to the
surface at that point, as defined by the definitions of body force and shearing stress. As a result, perfect
fluid has no shearing stress. Thus, we conclude that viscosity of a fluid is that property by virtue of
which itis able to offer resistance to shearing stress. It is a kind of molecular frictional resistance.
1.3.4 Velocity of Fluid at a Point

Assume that at time t , a fluid particle at the point P having position vector r and at time t + δ t the
 
same particle has reached at point Q having position vector r + δ r .
    
  (r + δ r) − r δ r dr
The particle velocity q at point P= is q Lt = Lt = , where the limit is assumed to
δ t →0 δt δ t →0 δ t dt
 
exist uniquely. Clearly q is in general dependent on both r and t , so we may write
    
=q q(r,
= t) q(x, y, z, t) , r = xi + yj + zk

Suppose, q = ui + vj + wk and since

 dr dx dy dz
q = = i + j + k
dt dt dt dt
dx dy dz
Therefore,
= u = ,v = ,w
dt dt dt

Remarks: (i) A point where q = 0 , is called a stagnation point.
(ii) When the velocity at each point is independent of time, i.e. the flow pattern is the same at each
instant, the motion is termed as steady; otherwise, it is termed as unsteady.
1.3.5 Flux across any surface


The flux i.e. the rate of flow across any surface S is defined by the integral ∫ ρ (q.n)dS
S
, where ρ is the

density, q is the velocity of the fluid and n is the outward unit normal at any point of S . Also, we have

Flux = density × normal velocity × area of the surface.

1.4 Eulerian and Lagrangian Methods (Local and Total range of Change)
We have two methods for studying the general problem of fluid dynamics.
1.4.1 Eulerian Method
In this method, we fix a point in the space occupied by the fluid and observation is made of whatever
changes of velocity, density pressure etc, take place at that point. i.e. point is fixed and fluid particles are
allowed to pass through it. If P(x, y, z) is the point under reference, then x, y, z do not dependent upon
the time parameter t, therefore x,  z do not exist (dot denotes derivative w.r.t. time t).
 y,
 12 Partial Differential Equations

Suppose f (x, y, z, t) be a scalar function associated with some property of the fluid (e.g. its density)
 
i.e. f (x, y, z, t) = f (r, t) , where r = xi + yj + zk is the position vector of the point P, then
 
∂f f (r, t + δ t) − f (r, t)
= Lt (1)
∂t δ t → 0 δt

∂f
Here, is called local time rate of change.
∂t
1.4.2 Lagrangian Method
In this approach, observations are made at each point and instant, i.e., any particle of the fluid is chosen
and its particular motion is observed and studied throughout its course. Let a fluid particle be initially at
the point (a, b, c). After lapse of time t, let the same fluid particle be at (x, y, z). It is obvious that x, y, z
are functions of t. However, after the motion is allowed, the particles that had different positions at the
beginning possess different positions. As a result, the final position's coordinates (x, y, z) are affected by
(a, b, c). Thus
=x f1=
(a, b, c, t), y f=
2 (a, b, c, t), z f 3 (a, b, c, t) .

For this case, if f (x, y, z, t) be scalar function associated with the fluid, then
  
df f (r + δ r, t + δ t) − f (r, t)
dt δLt
= (2)
t →0 δt
where x,
 y,
 z exist.

df
Here is called an individual time rate or total rate or particle rate of change.
dt
Now we established the relation between these two time rates (1) and (2).
We have f = f (x, y, z, t)

df ∂f dx ∂f dy ∂f dz ∂f
Therefore, = + + +
dt ∂x dt ∂y dt ∂z dt ∂t

df  ∂f  ∂f  ∂f    dx  dy  dz   ∂f  ∂f
=  i+ j + k  . i + j + k  + = ∇f .q +
dt  ∂x ∂y ∂z   dt dt dt  ∂t ∂t
 dx dx dx 
where q = i + j + k = (u, v, w)
dt dt dt
df ∂f 
Thus = + q.∇f
dt ∂t
Remarks: (i) The relation
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 13

df ∂f 
= + q.∇f
dt ∂t
df  ∂  
=  + q.∇  f
dt  ∂t 
d ∂  
⇒ =  + q.∇ 
dt  ∂t 
d
The operator is called Lagrangian operator or material derivative, i.e. time rate of change in
dt
Lagrangian view. Sometimes, it is called ‘differentiation following the fluid’.

(ii) Similarly, for a vector function F(x, y, z, t) associated with some property of the fluid, we can show
that
 
dF  ∂ F   
⇒ =  + q.∇ F 
dt  ∂t 
Hence the relation (3) holds for both scalar and vector functions associated with the moving fluid.
(iii) The Eulerian method is sometimes also called the flux method.
(iv) Both Lagrangian and Eulerian methods were used by Euler for studying fluid dynamics.
(v) Lagrangian method resembles very much with the dynamics of aparticle
(vi) The two methods are essentially equivalent, but depending upon the problem, one has to judge
whether Lagrangian method is more useful or the Eulerian.
Some important types of flows:
(i ) Laminar(streamline) and Turbulent Flows: Laminar flow is defined as a flow in which each fluid
particle traces a distinct curve and the curves traced by any two individual fluid particles do not
intersect. The flow is said to be turbulent if each fluid particle does not trace out a specific curve and the
curves traced out by fluid particles intersect. The fluid particles in such flows move in a zig-zag pattern.
(ii) Uniform and Non–Uniform Flows: Uniform flow is defined as a flow in which the fluid particles
have similar velocities at each section of the channel or pipe. The flow is said to be non-uniform if the
fluid particles have different velocities in each portion of the channel or pipe. For channel flows, such
flows are specified.
(iii) Steady and Unsteady Flows: A steady flow is one in which the properties and conditions
associated with the fluid's motion are independent of time, therefore the flow pattern remains constant.
The flow is considered to be unsteady if the properties and conditions associated with the fluid's motion
change with time, causing the flow pattern to change.
14 Partial Differential Equations

(iv) Rotational and Irrotational Flows: A rotational flow is one in which the fluid particles continue to
revolve around their own axes during motion, whereas an irrotational flow is one in which the fluid
particles do not rotate about their own axes during motion.
(v) Barotropic Flow: The flow is considered to be barotropic when the fluid pressure is merely a
function of density. These fluids are known to as barotropic fluids.
Remarks: A fluids is said to be isotropic w.r.t. some property such as pressure, density, etc. if that
property is the same in all directions at a point. If such property is not to be anisotropic w.r.t. that
property.
1.5 Streamline, Pathlines and Streaklines
1.5.1 Streamline
It is a curve drawn in the fluid such that the direction of the tangent to it at any point coincides with the
 
direction of the fluid velocity vector q at that point. At any time t , let q = (u, v, w) be the velocity at
each point P(x, y, z) of the fluid. The direction ratios of the tangent to the curve at P(x, y, z) are

dr = (dx, dy, dz) .
  
Since the tangent and the velocity at P have the same direction, therefore q × dr = 0 , i.e.

 × (dxi + dyj + dzk)
(ui + vj + wk)  =0

⇒ (vdz − wdy)i + (wdx − udz)j + (udy − vdx)k =
0
⇒ vdz − wdy ==
0 wdx − udz =udy − vdx
dx dy dz
⇒ = =
u v w
These are the differential equations for the streamlines and the solution of these equations give
streamlines.
 
Remarks: (i) It is noted that the streamlines are not well defined at the stagnation point where q = 0 . A
streamline may divide into two branches at s uch a point.

(ii) If we draw the streamlines through every point of a closed curve in the fluid, we obtain a stream-
Tube. A stream-tube of very small cross-section is called a stream filament.
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 15
(iii) The component of velocity at right angle to the streamline is always zero, so there is no flow across
the streamlines. Thus, if the boundary of stream tube is replaced by a rigid boundary, the flow is not
affected. The principle of conservation of mass, then gives that the flux across any cross section of the
stream tube should be the same.
1.5.2 Pathlines
When the fluid motion is steady so that the pattern of flow does not vary with time, the paths of the fluid
particles coincide with the streamlines. But in case of unsteady motion, the flow pattern varies with
time and the paths of the particles do not coincide with the streamlines. However, the streamline
through any point P does touch the pathline through P. Pathlines are the curves described by the fluid
particles during their motion i.e. these are the paths of the particles.
The differential equations for pathlines are

dr 
=q
dt
dx dy dz
i.e.= u,= v,= w (1)
dt dt dt
where ( x, y, z ) are the cartesian co-ordinates of the fluid particle and not a fixed point of space. The
equation of the pathline which passes through the point (x 0 , y 0 , z 0 ) , which is fixed in space, at time t = 0
say, is the solution of (1) which satisfy the initial condition that=x x=
0, y y=
0,z z 0 when t = 0. The
solution gives a set of equations of the form

x = x(x 0 , y 0 , z 0 , t) 

y = y(x 0 , y 0 , z 0 , t)  (2)
z = z(x 0 , y 0 , z 0 , t) 

which, as t takes all values greater than zero, will trace out the required pathline.
Remarks: Streamlines give the motion of each particle at a given instant whereas pathlines give the
motion of a given particle at each instant. We can make these observations by using a suspension of
aluminum dust in the moving liquid.
1.5.3 Streaklines
A streakline, in addition to streamlines and pathlines, is useful for observational purposes. This is the
curve formed by all fluid particles that have interacted with a specific fixed point in space at some point
in time. As a result, a streakline is the locus of various particles passing through a fixed point. The
streakline is observed when a neutrally buoyant marker fluid is continuously injected into the flow at
a fixed point of space from time τ =-∞. The marker fluid may be smoke if the main flow involves a
gas such as air, or a dye such as potassium permanganate (KMnO4) if the main flow involves a liquid
such as water.
 16 Partial Differential Equations

If the co-ordinates of a particle of marker fluid are (x, y, z) at time t and the particle coincided with the
injection point (x 0 , y 0 , z 0 ) at some time t0, where τ≤ t , then the time-history (streakline) of this particle
is obtained by solving the equations for a pathline, subject to the initial condition that
=x x=0, y y=
0,z z 0 at t = τ. As τ takes all possible values in the angle -∞≤τ≤ t , the locations of all
fluid particles on the streakline through (x0, y0, z0) are obtained. Thus, the equation of the streaklines at
time t is given by

x = x(x 0 , y 0 , z 0 , t,τ ) 

=y y(x 0 , y 0 , z 0 , t,τ )  (−∞ ≤ τ ≤ t) (3)
z = z(x 0 , y 0 , z 0 , t,τ ) 

Remark: (i) For a steady flow, streaklines also coincide with streamlines and pathlines.
(ii) Streamlines, pathlines and streaklines are termed as flowlines for a fluid.
1.6 Velocity Potential

Assume that q = ui + vj + wk is the velocity at any time t at each point P(x, y, z) of the fluid. Also
assume that the expression udx + vdy + wdz is an exact differential, say −dφ .

Then, −dφ = udx + vdy + wdz

 ∂φ ∂φ ∂φ ∂φ 
i.e. −  dx + dy + dz + dt  = udx + vdy + wdz
 ∂x ∂y ∂z ∂t 

where φ = φ (x, y, z, t) is some scalar function, uniform throughout the entire field of flow.

−∂φ −∂φ −∂φ ∂φ

Therefore,
= u = ,v = ,w =, 0
∂x ∂y ∂z ∂t

∂φ
But = 0 ⇒ φ = φ (x, y, z)
∂t
  ∂φ ∂φ  ∂φ  
Hence q = ui + vj + wk = −  i + j + k  = −∇φ
 ∂x ∂y ∂z 

where φ is termed as the velocity potential and the flow of such type is called flow of potential kind.

In the above definition, the negative sign in q = −∇φ is a convention and it ensures that flow takes
place from higher to lower potentials. The level surfaces φ (x, y, z, t) =constant, are called equipotential
or equipotential surfaces.
Theorem: At all points of the field of flow the equipotential (i.e. equipotential surfaces) are cut
orthogonally by the streamlines.

Proof: If the fluid velocity at any time t be q = (u, v, w) , then the equations of streamlines are
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 17
dx dy dz
= = (1)
u v w
 
The surfaces given by q.dr = 0 , i.e. udx + vdy + wdz =
0 (2)

are such that the velocity is at right angles to the tangent planes. The curves (1) and the surfaces (2) cut
each other orthogonally. Suppose that the expression on the left hand side of (2) is an exact differential
say, − dφ , then

dφ = udx + vdy + wdz (3)

where φ is velocity potential.

The necessary and sufficient condition for the relations

∂φ ∂φ ∂φ
u=
− ,v =
− ,w =
−
∂x ∂y ∂z
 
i.e. q = −∇φ to hold is curlq φ) 0
= curl(−∇= (4)
The solution of (2) i.e. dφ = 0 is φ (x, y, z) =constant (5)
The surfaces (5) are called equipotentials. Thus the equipotentials are cut orthogonally by the
streamlines.

Note: When curlq = 0 , the flow is said to be irrotational or of potential kind, otherwise it is rotational.

For irrrotational flow, q = −∇φ .

Example: The velocity potential of a two dimensional flow is φ = cxy . Find the streamlines.

Solution: The streamlines are given by

dx dy dz
= =
u v w

where q = (u, v, w)

For an irrotational motion, we have curlq= 0= curl(−∇φ )

i.e. q = −∇φ , where φ is the velocity potential.

From here,

 ∂φ ∂φ ∂φ 
(u, v, w) =− , ,  = −(cy, cx, 0)
 ∂x ∂y ∂z 
i.e. u = −cy, v =−cx, w =0
Therefore, streamlines are
 18 Partial Differential Equations

dx dy dz
= =
−cy −cx 0
i.e. xdx − ydy = 0, dz = 0
i.e. x 2 − y 2 = a 2 , z= K
which are rectangular hyperbola in the plane parallel to xy-plane.
Example: If the speed of fluid is everywhere the same, show that the streamlines are straight.
Solution: The streamlines are given by the differential equation
dx dy dz
= =
u v w
where u, v, w are constants.
The solutions are
constant, vz − wy =
vx − uy = constant

The intersections of these planes are certanity straight lines. Hence the result.
Example: Find the streamlines and pathlines of the particles for the two-dimensional velocity field
x
=u = , v y,=
w 0
1+ t
Solution: For streamlines, the differential equations are
dx dy dz
= =
u v w
Therefore,
dx dy dz
(1 + t) = =
x y 0

Here t =constant = t0 (at given instant), therefore the solutions are (1 + t 0 ) log x = log y + c1 , z = c 2

Or log x (1+ t 0 ) =log y + log a, z =c2

(1+ t 0 )
Or x = ay,
= z c 2 which are the required streamlines.

For pathlines, we have

dx dy dz
= u,= v,= w
dt dt dt
dx x dy dz
Therefore, = ,
= y,= 0
dt 1 + t dt dt
dx dt dy
⇒ = , = dt, dz= 0 ⇒ log .x =log(1 + t) + log a, log y =t + log b, z =c
x 1+ t y
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 19

⇒ x = a(1 + t), y = be t , z = c
x −a
⇒
= y be =
a
;z c

which are the required pathlines.

Note: In case of pathlines, t must be eliminated since these give the motion at each instant (i.e.
independent of t).
Example: Obtain the equations of the streamlines, pathlines and streaklines which pass through (l, l,
x t  y
0) at t = 0 for the two dimensional flow u = 1 +  , v = , w =0 .
t0  t0  t0

Solution: We define the dimensionless co-ordinates X, Y, Z and time T by writing

x y z t dx dy dz dt
X
= ,Y =
= ,Z = ,T such that dX
= = , dY = , dZ = , dT and
l l l t0 l l l t0
Xl Yl
u= (1 + T), v = ,w = 0
t0 t0

Streamlines are given by

dx dy dz
= =
u v w
t ldX t 0 ldY dZ
⇒ 0 = =
Xl(1 + T) Yl 0
dX dY dZ
⇒ ==
X(1 + T) Y 0
Integrating these, we get
Z=constant= C1 (1)

and log X =
(1 + T ) log Y + log C2 , where C2 is constant

C2Y (1+T )
⇒X= (2)

As variable X, Y, Z and T are independent and C1 , C2 are constants, (1) & (2) given the family of
streamlines at all times t = t 0 T . In particular, x =1 =y, z =0, t =0 ⇒ C1 =0, C 2 =1

And we get streamline as Y = X, i.e., y = x and z = o

dX dY dZ
Pathlines are given by X(1 + T),
= Y,
= 0
=
dT dT dT
Now, X, Y, Z are the dimensionless co-ordinates of a fluid particle and are functions of T.
20 Partial Differential Equations

dX  T2 
Therefore, (1 T)dT ⇒ log X =
=+  T +  + log K1
X  2 
2
K1eT + T
⇒X= /2
(3)

dY dY
=Y⇒ =dT ⇒ log Y =T + log K 2
dT Y

K 2 eT
⇒Y= (4)

dZ = 0 ⇒ Z = cons tan t = K 3 (5)

These are the parametric equations of pathlines. The pathline through P(1,1,0) i.e. X=1=Y,Z=0,T=0 is
obtained when K =1 K= 2 1, K=
3 0

T2
T+
⇒ X= e 2
, Y= eT , Z= 0

Elimination of T gives,
 T  T  T  log Y 
T 1+  1+ 
1+ 
 2
1+
 2 


⇒X
= e  2
= e  T
= Y = Y
 2 ,Z
= 0

The pathline which passes through X=Y=1, Z=0 when T= τ is given by

 1 1 
X exp. T + T 2 − τ − τ 2  ,
=
 2 2 
Y = exp(T − τ ), Z = 0
These are the parametric equations of the streakiness true for all values of T.
At T=0 the equation given as

 1 
X exp.  −τ − τ 2  ,
=
 2 
Y = exp(−τ ), Z = 0

Eliminating τ , we have,
−τ log Y ⇒=
= τ log Y
Therefore,
 τ   τ   τ   log Y 
−τ  1+  1+  1− 
−τ  1+   2  2 
⇒X
= e  2
=  e   =
2 Y = Y ,Z
= 0

Article: To obtain the differential equations for streamlines in cylindrical and spherical co-ordinates.
Solution: We know that the streamlines are obtained from the differential equations.
  
q × dr =
0 (1)
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 21
 
where q is the velocity vector and r is the position vector of a liquid particle.

If the motion is irrotational, then


q = −∇φ
Therefore, the differential equation (1) become
 
∇φ × dr =0 (2)

(i) In cylindrical co-ordinates (r, θ , z) , we have

  ∂φ 1 ∂φ ∂φ 
dr = (dr, rdθ , dz) and ∇
= φ grad
=φ  , , 
 ∂r r ∂θ ∂z 
Thus, the different equation (2) become

 ∂φ 1 ∂φ ∂φ  
 , ,  × (dr, rdθ , dz) =
0
 ∂r r ∂θ ∂z 
dr rdθ dz
⇒ = = (3)
∂φ / ∂r 1/ r.∂φ / ∂θ ∂φ / ∂z

(ii) In spherical co-ordinates (r, θ ,ψ ) , we have


dr = (dr, rdθ , r sin θ dψ )

 ∂φ 1 ∂φ 1 ∂φ 
and ∇
= φ grad
= φ  , , 
 ∂r r ∂θ r sin θ ∂ψ 

 ∂φ 1 ∂φ 1 ∂φ  
The differential equation (2) become  , ,  × (dr, rdθ , r sin θ dψ ) 0
=
 ∂r r ∂θ r sin θ ∂ψ 
dr rdθ r sin θ dψ
= = (4)
∂φ / ∂r 1/ r.∂φ / ∂θ 1 ∂φ / ∂ψ
r sin θ
Equations (3) and (4) are the required differential equations.
A
Example: Show that if the velocity potential of an irrotational fluid motion is φ = ψ cos θ , where
r2
(r, θ ,ψ ) are the spherical co-ordinates of any point, the lines of flow lie on the surface r = k sin 2 θ , k
being a constant.
Solution: The differential equations for lines of flow (streamlines) are
dr rdθ r sin θ dψ
= =
∂φ / ∂r 1/ r.∂φ / ∂θ 1 ∂φ / ∂ψ
r sin θ
 22 Partial Differential Equations

From first two members, we have

dr rdθ
=
2A 1 A 
−ψ 3 cos θ  − 2 ψ sin θ 
r r r 
dr 2rdθ dr cos θ
⇒ = ⇒ = 2 dθ
cos θ sin θ r sin θ
⇒
= log r 2 log sin θ + log=k ⇒ r k sin 2 θ
Hence the result.
Note: In cartesian co-ordinates, the above velocity potential can be written as

y
φ= A(x 2 + y 2 + z 2 ) −3/2 z.tan −1  =
 , where x r sin
= sin θ sinψ , z r cos θ are spherical polar
θ cosψ , y r=
x
substitutions.
Also, the streamlines r = k sin 2 θ can be written as r 3 = kr 2 sin 2 θ
⇒ (x 2 + y 2 + z 2 )3/2= k(x 2 + y 2 )

⇒ x 2 + y 2 +=
z 2 k 2/3 ( x 2 + y 2 ) 2/3

which are the streamlines in cartesian co-ordinates.

Example: At the point in an incompressible fluid having spherical polar co-ordinates (r, θ ,ψ ) , the
velocity components are (2Mr −3 cos θ , Mr −2 sin θ , 0) where M is a constant. Show that velocity is of
potential kind. Find the velocity potential and the equations of streamlines.

Solution: Here dr = drr + rdθθ + r sin θ dψψ

= q 2Mr −3 cos θ r + Mr −2 sin θθ
Then,
r rθ r sin θψ
 1
curlq = ∂ ∂ ∂
r sin θ
2 ∂r ∂θ ∂ψ
2Mr cos θ
−3
Mr sin θ
−2
0
 1 
curlq
= r.0 + rθ .0 + r sin θψ (−2Mr −3 sin θ + 2Mr −3 sin
= θ ) 0
r sin θ 
2 

Therefore, the flow is potential kind.

Now, using the relation
  ∂φ 1 ∂φ  1 ∂φ  
q = −∇φ = −  r + θ+ ψ
 ∂r r ∂θ r sin θ ∂ψ 
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 23
From here,
∂φ ∂φ ∂φ
= 2Mr −3 cos θ=
,− θ,
Mr −2 sin= 0
∂r ∂θ ∂ψ
Therefore,
∂φ ∂φ ∂φ
dφ = dr + dθ + dψ
∂r ∂θ ∂ψ
= (−2Mr −3 cos θ )dr + (−Mr −2 sin θ )dθ
= d(Mr −2 cos θ )

Integrating, we get

φ = Mr −2 cos θ , which is the required velocity potential.

The streamlines are given by
dr rdθ r sin θ dψ
= =
−∂φ / ∂r −1/ r.∂φ / ∂θ − 1 ∂φ / ∂ψ
r sin θ
dr rdθ r sin θ dψ
= =
2Mr cos θ Mr sin θ
−3 −3
0
From the last term, ψ = constant.

From the first two terms, we get

dr 2 cos θ
= = dθ 2 cot θ dθ
r sin θ
Integrating, we get

log r log sin 2 θ + constant

=

⇒r=A sin 2 θ ,ψ = constant

The equation ψ = constant shows that the streamlines lie in planes which pass through the axis of
symmetry θ = 0 .
1.7 Vorticity
1.7.1 Vorticity
 
If q = (u , v, w) be the velocity vector of a fluid particle, then the vector ξ defined by
  
ξ = curlq = ∇ × q
24 Partial Differential Equations

is called the vortex vector or vorticity and its components are (ξ1 , ξ 2 , ξ3 ) given by
∂w ∂v ∂u ∂w ∂v ∂u
ξ1 = − , ξ2 = − , ξ3 = −
∂y ∂z ∂z ∂x ∂x ∂y

In other words, we can say the fluid motion is said to be rotational if

 
= ξ curlq ≠ 0
  
If= = 0 , then the fluid motion is said to be irrotational or of potential kind and then q = −∇φ .
ξ curlq
1.7.2 Vortex Line
It is a curve in the fluid such that the tangent at any point on the curve has the direction of the vorticity

vector ξ .
  
The differential equations of vortex lines are given by ξ × dr = 0
dx dy dz
i.e. = =
ξ1 ξ2 ξ3

where ξ = (ξ1 , ξ 2 , ξ3 ) .

1.7.3 Vortex Tube

The locus of vortex line drawn at each point of a closed curve forms a vortex tube. A vortex tube with
small cross section is called a vortex filament.
Let S be the surface and V be the volume of a vortex tube, then by Guass divergence thorem,
  
∫
S
n̂.ξ dS = ∫ dV = ∫ ∇.(∇ × q)dV = 0
V
∇.ξ
V
(1)


Here, n̂.ξ δ S is called the strength of the vortex tube.

A vortex tube with a unit strength is called a unit vortex tube. Equation (1) shows that the total strength
of a vortex tube emerging from S is equal to that entering S. This means that vortex lines and tubes
cannot originate or terminate at internal points in a fluid. They can only form closed curves or terminate
on boundaries. However, in case of vorticity, it is not necessary to assume incompressibility of the fluid.
1.8 Flow and Circulation
1.8.1 Flow
Let A and B be two points in the fluid.
A  
Then ∫ q.dr = 0 is called the flow along any path from A to B
B

 B  B
If motion is irrotational then q = −∇φ and flow − ∫ ∇φ .dr =− ∫ dφ =φ (A) − φ (B)
A A
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 25
1.8.2 Circulation
It is the flow round a closed curve. If C be the closed curve in a moving fluid, then circulation Γ about
   
Γ ∫=  = n.
q.dr ∫ n.curlqdS  ξ dS
C is given by=
C S
∫ S
 
If the motion irrotational, then q = −∇φ and thus, Γ = − ∫ ∇φ .dr = − ∫ dφ =φ (A) − φ (A) = 0, where A is
C C

any point on the curve C. This shows that for an irrotational motion, circulation is zero.
1.8.3 Theorem
The necessary and sufficient condition such that the vortex lines are at right angles to the streamlines, is
 ∂φ ∂φ ∂φ 
(u, v, w) = µ  , , 
 ∂x ∂y ∂z 

i.e. q= µ∇φ , where µ and φ are functions of x, y, z and t.

Proof: Necessary condition: We know that the differential equation

  
q.dr = 0 is integrable if

 ∂Q ∂R 
 −=  + ... + .. 0 + Qdy + Rdz 0 is integrable if
 pdx =
 ∂z ∂y 
 
q.curlq = 0 (exactness condition)
  
i.e. =
q.ξ 0,= ξ curlq
This shows that the streamlines are at right angles to the vortex lines. Thus the streamlines and vortex
 
lines are at right angles to each other if the differential equation q.dr = 0 is integrable. However, it is
 
certainty not an exact equation, because its exactness implies that q = −∇φ , hence curlq = 0 i.e. vortex
lines cease to exist.
 
Hence, there exists an integrating factor µ (function of x, y, z, t) such that µ −1 q.dr = 0 is integrable.

If this differential equation is integrable, then we can write

 
µ −1 q.dr = dφ , where φ is a scalar function of x, y, z, t
   
⇒ µ −1 q.dr =∇φ .dr  dφ = ∇φ .dr

⇒ q = µ∇φ .
 
Sufficient condition: let us take q= µ∇φ ⇒ ∇φ = µ −1 q

Then, curlq
= curl ( µ∇φ )
26 Partial Differential Equations

ξ = ∇ × µ∇φ = µ (∇ × ∇φ ) + ∇µ × ∇φ = ∇µ × ∇φ
Therefore,
  
q.ξ = (∇µ × ∇φ ).q = ∇µ .(∇φ × q)
 
= ∇µ .( µ −1 q × q) = 0

This shows that the directions of streamlines and vortex lines are at right angles to each other.
1.9. Equation of Continuity
1.9.1 Equation of Continuity by Euler’s Method (Equation of conservation of Mass)
When the region of a fluid has no inlets (sources) or outlets (sinks) through which fluid can enter or
leave the region, than the amount of fluid within the region is conserved in accordance with the principle
of conservation of mass. We formulate this principle
mathematically by means of the so-called equation of
continuity.
Let us consider a closed geometrical surface S fixed in
space, which is drawn in a region through which fluid
flows. It is supposed that S does not include points which
are either sources or sinks. Let V be the volume enclosed
by S and let n̂ denotes the unit vector normal to an element
δ S of S drawn outwards.

Let q be the fluid velocity and ρ be the fluid density.

First, we consider the mass of fluid which leaves V by flowing across an element δ S of S in time δ t .
This quantity is exactly that which is contained in a small cylinder of cross-section

 δt.
δ S of length (q.n)

Thus the mass of the fluid is = density × Volume= ρ (q.n) δ t.δ S

Hence the rate at which fluid leaves V by flowing across the


 δS
element δ S is ρ (q.n).

Summing over all such elements, we obtain the rate of flow of

fluid coming out of V across the entire surface S. Hence, the rate at which mass flows out of the region
V is
   
S
 δ S = ( ρ q).nδ S
∫ ρ (q.n) ∫ S
∫  =
S
F.ndS ∫
V
∇.FdV by Gauss divergence


= ∫ div( ρ q)dV (1)
V
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 27
Now, the mass M of the fluid possessed by the volume V of the fluid is

M = ∫ ρ dV , where ρ = ρ (x, y, z, t) with (x, y, z) the cartesian co-ordinates of a general point of V, a

V

fixed region of space.

Since the space co-ordinates are independent of time t, therefore the rate of decrease of mass within V is

dM d  ∂ρ
− −  ∫ ρ dV  =
= −∫ dV (2) | V does not change w.r.t. time
dt dt  V  V
∂t

But the considered region is free from source and sink i.e. the mass is neither created nor destroyed, so
by law of conservation of mass , i.e. mass is neither created nor destroyed and thus from equations (1)
and (2), we get
 ∂ρ
∫V div( ρ q)dV = −∫
V
∂t
dV

 ∂ρ  
⇒ ∫  + div( ρ q)  dV = 0
V  ∂t 
Since V is arbitrary, we conclude that at any point of the fluid which is neither a source nor a sink,
∂ρ  ∂ρ 
+ div( ρ q) 0, i.e.
= =+ ∇.( ρ q) 0 (3)
∂t ∂t
Equation (3) is known as equation of continuity of fluid and holds at all points of the fluid free from
sources and sinks.
Corollary (1) We know that
  
div(=ρ q) ρ divq + q.(gradρ )
∂ρ  
Therefore (3) takes the form + ρ∇.(q) + (q.∇) ρ =0 (4)
∂t
D
Corollary (2) We know that the differential operator is given by
Dt
D ∂ 
= + (q.∇)
Dt ∂t
Dρ 
Therefore, from (4), we obtain the equation of continuity as + ρ (∇.q) =
0
Dt
Dρ 
i.e. + ρ divq =
0 (5)
Dt
1 Dρ 
Corollary (3) Equation (5) can be written as + divq =
0
ρ Dt
28 Partial Differential Equations

D 
⇒ (log ρ ) + divq =
0 (6)
Dt
∂ρ
Corollary (4) When the motion of fluid is steady, then = 0 and thus the equation of continuity (3)
∂t

becomes div( ρ q) = 0 (7)

| Here ρ is not a function of time i.e. ρ = ρ (x, y, z)

Dρ
Corollary (5) When the fluid is incompressible, then ρ =constant and thus =0
Dt
The equation of continuity becomes

div(q) = 0 (8)

which is same for homogeneous and incompressible fluid.

Corollary (6) If in addition to homogeneity and incompressibility, the flow is of potential kind such that

q = −∇φ , then the equation of continuity becomes single word

div(−∇φ ) = 0 ⇒ ∇.(∇φ ) = 0 ⇒ ∇ 2φ = 0 (9)

which is known as the Laplace equation.

1.9.2 Equation of Continuity in Cartesian Co-ordinates
Let (x, y, z) be the rectangular cartesian co-ordinates.

Let q = ui + vj + wk (1)

∂  ∂  ∂ 
and =
∇ i+ j+ k (2)
∂x ∂y ∂z

∂ρ 
Then, the equation of continuity + div( ρ q) =
0 can be written as
∂t
∂ρ ∂ ( ρ u) ∂ ( ρ v) ∂ ( ρ w)
+ + + 0
= (3)
∂t ∂x ∂y ∂z

∂ρ ∂ρ ∂ρ ∂ρ  ∂u ∂v ∂w 
+u +v +w +ρ + + =0 (4)
∂t ∂x ∂y ∂z  ∂x ∂y ∂z 
which is the required equation of continuity in Cartesian co-ordinates.
∂ρ
Corollary: (1) If the fluid motion is steady, then = 0 and the equation (3) becomes
∂t
∂ ( ρ u) ∂ ( ρ v) ∂ ( ρ w)
+ + 0
= (5)
∂x ∂y ∂z
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 29
 
Corollary: (2) If the fluid is incompressible, then ρ =constant and the equation of continuity is ∇.q =0

∂u ∂v ∂w
i.e. + + 0
= (6)
∂x ∂y ∂z
Corollary: (3) If the fluid is incompressible and of potential kind, then equation of continuity is

∇ 2φ =
0

∂ 2φ ∂ 2φ ∂ 2φ 
i.e. + + 0
= , where q = −∇φ .
∂x 2 ∂y 2 ∂z 2

1.9.3 Equation of Continuity in Orthogonal Curvilinear Co-ordinates

Let (u1 , u 2 , u 3 ) be the orthogonal curvilinear co-ordinates and e 1 , e 2 , e 3 be the unit vectors tangent to the
co-ordinate curves.

Let q = q1 e 1 + q 2 e 2 + q 3 e 3 (1)

The general equation of continuity is

∂ρ 
+ div( ρ q) =
0 (2)
∂t

We know from vector calculus that for any vector point function f = (f1 , f 2 , f 3 ) ,

 1  ∂ ∂ ∂ 
∇.f  (h 2 h 3f1 ) + (h1h 3f 2 ) + (h 2 h1f 3 )  (3)
h 1h 2 h 3  ∂u1 ∂u 2 ∂u 3 
where h1 , h 2 , h 3 are scalars.

Using (3), the equation of continuity (2) becomes

∂ρ 1  ∂ ∂ ∂ 
+  (h 2 h 3 ρ q1 ) + (h1h 3 ρ q 2 ) + (h 2 h1 ρ q 3 )  =
0
∂t h1h 2 h 3  ∂u1 ∂u 2 ∂u 3 
(4)
Corollary: (1) When motion of fluid is steady, then equation
(4) becomes
∂ ∂ ∂
(h 2 h 3 ρ q1 ) + (h1h 3 ρ q 2 ) + (h 2 h1 ρ q 3 ) =0 (5)
∂u1 ∂u 2 ∂u 3

Corollary (2) When the fluid is incompressible, the equation of continuity is ( ρ =constant)

∂ ∂ ∂
(h 2 h 3q1 ) + (h1h 3q 2 ) + (h 2 h1q 3 ) =
0 (6)
∂u1 ∂u 2 ∂u 3
30 Partial Differential Equations

Corollary: (3) When fluid is incompressible and irrotational, then ρ =constant

 1 ∂ 1 ∂ 1 ∂ 
q = −∇φ = −  , ,  φ and the equation of continuity becomes
 h1 ∂u1 h 2 ∂u 2 h 3 ∂u 3 

∂  h 2 h 3 ∂φ  ∂  h1h 3 ∂φ  ∂  h 2 h1 ∂φ 
 +  +  =0 (7)
∂u1  h1 ∂u1  ∂u 2  h 2 ∂u 2  ∂u 3  h 3 ∂u 3 

Now, we shall write equation (4) in cylindrical and spherical polar co-ordinates.
1.9.4 Equation of Continuity in Cylindrical Co-ordinates (r, θ , z)

Here,

=q (q , q
r θ , qψ ) & =
h1 1,=
h 2 r,=
h3 1

The equation of continuity becomes

∂ρ 1  ∂ ∂ ∂ 
+  (rρ q r ) + ( ρ qθ ) + (rρ q z )  =
0
∂t r  ∂r ∂θ ∂z 
∂ρ 1 ∂ 1 ∂ ∂
+ (rρ q r ) + ( ρ qθ ) + ( ρ q z ) =
0 (8)
∂t r ∂r r ∂θ ∂z
Corollary: (1) When the fluid motion is steady, then equation (8) becomes
∂ ∂ ∂
(rρ q r ) + ( ρ qθ ) + r ( ρ q z ) =
0 (9)
∂r ∂θ ∂z
Corollary: (2) For incompressible fluid, equation of continuity is
∂ ∂ ∂
(rq r ) + (qθ ) + r (q z ) =
0 (10)
∂r ∂θ ∂z
Corollary: (3) When the fluid is incompressible and is of potential kind, then equation (8) takes the
form
∂ ∂φ ∂ ∂φ ∂ ∂φ
(r ) + ( ) + (r ) =
0 (11)
∂r ∂r ∂θ r∂θ ∂z ∂z

where q = −∇φ ; ∇ is expressed in cylindrical co-ordinates.

1.9.5 Equation of Continuity in Spherical Co-ordinates (r, θ ,ψ )


=q ( q r , qθ , q z ) & =
h1 1,= h 3 r sin θ
h 2 r,=

The equation of continuity becomes

∂ρ 1 ∂ 2 ∂ ∂ 
+ 2  (r sin θρ q r ) + (r sin θρ qθ ) + (rρ qψ )  =
0
∂t r sin θ  ∂r ∂θ ∂ψ 
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 31

∂ρ 1  ∂ 2 ∂ ∂ 
⇒ + 2
∂t r sin θ sin θ ∂r (r ρ q r ) + r ∂θ (sin θρ qθ ) + r ∂ψ ( ρ qψ )  =
0 (12)
 
Corollary: (1) For steady case, equation (12) becomes
∂ 2 ∂ ∂
sin θ (r ρ qr ) + r (sin θρ qθ ) + r ( ρ qψ ) =
0 (13)
∂r ∂θ ∂ψ
Corollary: (2) For incompressible fluid, we have
∂ 2 ∂ ∂
sin θ (r q r ) + r (sin θ qθ ) + r (qψ ) =
0 (14)
∂r ∂θ ∂ψ
Corollary: (3) When the fluid is incompressible and is of potential kind, then equation (8) takes the
form
∂ 2 ∂φ ∂ ∂φ ∂ 1 ∂φ
(r sin θ ) + (sin θ )+ ( )= 0 (15)
∂r ∂r ∂θ ∂θ ∂ψ sin θ ∂ψ

where q = −∇φ ; ∇ is expressed in spherical co-ordinates.

1.9.6 Symmetrical forms of Motion and Equation of Continuity for Them

We have the following three types of symmetry which are special cases of cylindrical and spherical
polar co-ordinates.
(i) Cylindrical Symmetry: In this type of symmetry, with suitable choice of cylindrical polar co-
ordinates (r, θ, z), every physical quantity is independent of both θ and z so that
∂ ∂  
= = 0 &=
q q(r, t)
∂θ ∂z
For this case, the equation of continuity in cylindrical co-ordinates, reduces to
∂ρ 1 ∂
+ ( ρ q1r) =
0 (1)
∂r r ∂r
If the flow is steady, then equation (1) becomes
∂
( ρ q1r ) = 0 ⇒ ρ q=
1r cons tan
= t F(t) (say)
∂r
=
Further, if the fluid is incompressible then q1r constant
= G(t) .

(ii) Spherical Symmetry: In this case, the motion of fluid is symmetrical about the centre and thus
with the choice of spherical polar co-ordinates (r, θ, ψ ), every physical quantity is independent of
∂ ∂  
both θ & ψ . So that = = 0 &= q q(r, t)
∂θ ∂ψ
The equation of continuity in spherical co-ordinates, reduce to
32 Partial Differential Equations

∂ρ 1 ∂
+ 2 . ( ρ q1r 2 ) =
0 (2)
∂r r ∂r
∂
For steady motion, it becomes ( ρ q1r 2 ) = 0 ⇒ ρ q1r=
2
t F(t) (say)
cons tan=
∂r
And for incompressible fluid, it has the form
= q1r 2 cons
= tan t G(t) (say).

(iii) Axial Symmetry :- (a) In cylindrical co-ordinates (r, θ, z), axial symmetry
∂
Means that every physical quantity is independent of θ i.e = 0 and thus the equation of continuity
∂θ
∂ρ 1  ∂ ∂ 
becomes +  ( ρ q r r) + ( ρ qθ r) =
0 =
0
∂r r  ∂r ∂z 
(b) In spherical co-ordinates (r, θ, ψ ), axial symmetry means that every physical quantity is
∂
independent of ψ i.e. = 0 and the equation of continuity, for this case, reduce to
∂ψ
∂ρ 1 ∂ 1 ∂
+ 2 . (ρqr r2 ) + ( ρ qθ sin θ ) =
0
∂r r ∂r r sin θ ∂θ
Example: If σ(s) is the cross-sectional area of a stream filament, prove that the equation of
∂ ∂
continuity is ( ρσ ) + ( ρσ q) =
0 , where δ s is an element of arc of the filament and q is the fluid
∂t ∂s
speed.
Solution: Let P and Q be the points on the end sections of the stream filament.
The rate of flow of fluid out of volume of filament is
∂
( ρσ q)Q − ( ρσ q) P = ( ρσ q) P δ s (1)
∂s
where we have retained the terms upto first order only, since
δ s is infinitesimally small. Now, the fluid speed is along the
normal to the cross-section. At time t, the mass within the
segment of filament is ρσδ s and its rate of increase is

∂ ∂
= ( ρσ )δ s = ( ρσδ s ) 0 (2)
∂t ∂t
Using law of conservation of mass, we have from (1) and (2)
∂ ∂
( ρσ )δ s + ( ρ qσδ s) =
0 | Total rate=0
∂t ∂s
∂ ∂
i.e. ( ρσ ) + ( ρσ q) =
0 (3)
∂t ∂s
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 33
which is the required equation at any point P of the filament.
∂
Deduction: For steady incompressible flow, ( ρσ ) =0 and
∂t
equation (3) reduce to
∂ ∂
( ρσ q) =0 ⇒ (σ q) =0 ⇒σq =constant
∂s ∂s
which shows that for steady incompressible flow product of
velocity and cross-section of stream filament is constant. This
result means that the volume of fluid acrossing every section per
dis tan ce volume
unit time is constant ⇒ σ q =⇒
c σ c
=⇒ =c
t t
Example: A mass of fluid in such a way that each particle describe a circle in one plane about a fluid
axis, show that the equation of continuity is
∂ ∂
(ρ ) + ( ρω ) =
0
∂t ∂θ
where ω is the angular velocity of a particle whose azimuthal angle is θ at time t.
Solution: Here, the motion is in plane and the particles describe a circle,
∂ ∂
z = constant, r =constant ⇒ =0 = (1)
∂z ∂r
i.e. there is only rotation.
Now, the equation of continuity in cylindrical co-ordinates (r , θ , z )

∂ρ 1 ∂ 1 ∂ ∂
+ (rρ q r ) + ( ρ qθ ) + ( ρ q z ) =
0 (2)
∂t r ∂r r ∂θ ∂z
Using (1) in (2), we get
∂ρ 1 ∂
+ ( ρ qθ ) =
0
∂t r ∂θ
∂ρ 1 ∂
+ ( ρ rω ) = q q=
0 where = θ rω
∂t r ∂θ
∂ρ ∂
⇒ + ( ρω ) =
0
∂t ∂θ
Hence the result.
34 Partial Differential Equations

Example: A mass of fluid is in motion so that the lines of motion lie on the surface of co-axial
∂ρ 1 ∂ ∂
cylinders, show that the equation of continuity is + 0 where vθ , vz the
( ρ vθ ) + ( ρ v z ) =
∂t r ∂θ ∂z
velocities are perpendicular and parallel to z.
Solution: The equation of continuity in cylindrical co-ordinates (r , θ , z ) is

∂ρ 1 ∂ 1 ∂ ∂ 
+ ( ρ rv r ) + ( ρ vθ ) + ( ρ v z ) =
0 , where q = (v r , vθ , v z )
∂t r ∂r r ∂θ ∂z
Since the lines of motion (path lines) lie on the surface of cylinder, therefore the component of velocity
in direction of dr is zero i.e. vr = 0

∂ρ 1 ∂ ∂
Thus, the equation of continuity in the present case reduces to + ( ρ vθ ) + ( ρ v z ) =
0
∂t r ∂θ ∂z
Hence the result.
Example: The particles of a fluid move symmetrically in space with regard to a fixed centre, prove that
∂ρ ∂ρ ρ ∂ 2
the equation of continuity is +u + . (r u) = 0 , where u is the velocity at distance r.
∂t ∂r r 2 ∂r
Solution: Firstly, derive the equation of continuity in spherical polar co-ordinates. The present case is
the case of spherical symmetry as the motion is symmetrical w.r.t. a fixed centre.
∴Equation of continuity is
∂ρ 1 ∂ ∂ ∂
+ 2 . ( ρ r 2q r ) =
0, = 0=
∂t r ∂r ∂θ ∂ψ
∂ρ 1 ∂
⇒ + . ( ρ r 2 u) = 0, q r = u
∂t r 2 ∂r
∂ρ 1 ∂ρ 2 1 ∂
⇒ + 2 . .(r u) + 2 ρ (r 2 u) =
0
∂t r ∂r r ∂r
∂ρ ∂ρ ρ ∂ 2
⇒ +u + (r u) =0
∂t ∂r r 2 ∂r
Hence the result.
Example: If the lines of motion are curves on the surfaces of cones having their vertices at the origin
and the axis of z for common axis, prove that the equation of continuity is
∂ρ ∂ρ q r 2 ρ q r cos ecθ ∂
+ + + . ( ρ qψ ) =
0
∂t ∂r r r ∂ψ
Solution: The equation of continuity in spherical polar co-ordinates is

∂ρ 1  ∂ 2 ∂ ∂ 
⇒ + 2
∂t r sin θ sin θ ∂r (r ρ q1 ) + r ∂θ (sin θρ q 2 ) + r ∂ψ ( ρ q 3 )  =
0
 
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 35
In the present case, the lines of motion lie on the surfaces of cones, i.e. θ =constant, so the velocity
component in the θ -direction is zero. i.e. qθ = 0 .

The equation of continuity becomes

∂ρ 1 ∂ 2 1 ∂
⇒ + 2 (r ρ q r ) + ( ρ qψ ) =
0
∂t r ∂r r sin θ ∂ψ

∂ρ 1  2 ∂  cos ecθ ∂
⇒ + 2 r ( ρ q r ) + ρ q r .2r  + ( ρ qψ ) =
0
∂t r  ∂r  r ∂ψ
∂ρ ∂ 2 ρ q r cos ecθ ∂
⇒ + (ρqr ) + + ( ρ qψ ) =
0
∂t ∂r r r ∂ψ

Hence the result.

Example: Show that polar form of equation of continuity for a two dimensional incompressible fluid is
∂ (ru) ∂v µ cos θ  
+ 0 . If u = −
= , then find v and the magnitude of the velocity q , where q =(u, v).
∂r ∂θ r2
Solution: The equation of continuity in polar co-ordinates (r , θ ) in two dimensional (z=0) is
∂ρ 1 ∂ 1 ∂
+ ( ρ rq r ) + ( ρ qθ ) =
0
∂t r ∂r r ∂θ
 ρ ∂ ρ ∂
In the present case, ρ =constant and q =(u, v), so (ru) + (v) =
0
r ∂r r ∂θ
µ cos θ
Now, u = − , so we get
r2
∂ µ cos θ ∂ µ cos θ ∂v ∂v µ cos θ
(− .r) + (v) =
0⇒ + 0⇒
= =
−
∂r r 2
∂θ r 2
∂θ ∂θ r2
µ sin θ
⇒v=−
r2
 µ
∴q = q = u 2 + v2 =
r2
Example: Liquid flows through a pipe whose surface is the surface of revolution of the curve
x2
y= a + k about the x-axis (−a ≤ x ≤ a ) . If the liquid enters at the end x = −a of the pipe with velocity
a
V, show that the time taken by a liquid particle to transverse the entire length of the pipe from x = −a to
2a 2 1
x = a is 2
(1 + k + k 2 ) , where k is assumed to be so small that the flow remains appreciably
V(1 + k) 3 5
one-dimensional throughout.
 36 Partial Differential Equations

x2
Solution: Here, y= a + k
a

kx 2 2
is π=
Therefore, Area of the section distant x from 0= y 2 π (a + )
a
Area at x=-a is =π (a + ka) 2 =π a 2 (1 + k) 2

Applying the equation of continuity at the two sections, i.e. expressing equal rates of volumetric flow,
kx 2 2 dx
we get π a 2 (1 + k) 2 V =π (a +
) x where x = the velocity at the section distant x from 0 is. Thus,
a dt
2
1 kx
we
= have dt 2
(1 + 2 ) 2 dx
(1 + k) V a
So, the required time of flow is
a
1 kx 2 2
=T 2∫ 2
(1 + 2
) dx
0
(1 + k) V a
a
2 k2x4 kx 2
(1 + k) 2 V ∫0
= (1 + 4 + 2 2 )dx
a a
2a  2 1 2
= 2 1 + k + k 
(1 + k) V  3 5 
Hence the result.

Remark: If the velocity vector q = (u, v, w) is kinematically possible for an incompressible fluid
 ∂u ∂v ∂w
motion, then the equation of continuity must be satisfied. i.e. ∇.q = 0 ⇒ + + =0.
∂x ∂y ∂z
 c 2 (x j − yi)

Example: Test whether the motion specified by q = , c being constant, is a possible motion
x 2 + y2
for an incompressible fluid. If so determine the equations of the streamlines. Also test whether the
motion is of potential kind and if so determine the velocity potential and equipotential surfaces.
 2 −y x 
Solution: We have q c=
=  2 2
, 2 2
, 0  (u, v, w)
x +y x +y 
  ∂  −y  ∂  x     2xy 2xy 
Therefore, ∇.q = c 2   2 2 
+  2 2 
= q = c2  2 2
− 2 2 
=0
 ∂x  x + y  ∂y  x + y   x +y x +y 
Thus the equation of continuity for an incompressible fluid is satisfied and so such a motion is possible.
The equation of streamlines are
 Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 37

dx dy dz
= =
u v w
dx dy dz
⇒ 2 2 2
=2 2 2
=
−c y / (x + y ) c x / (x + y ) 0
⇒ xdx + ydy = 0, dz = 0

⇒ x 2 + y2 =
constant, z=constant (1)

which are the circles whose centers lie on the z-axis and whose planes are perpendicular to this axis.
Further, we have

i j k
 ∂ ∂ ∂
∇×q =
∂x ∂y ∂z
−c 2 y c2 x
0
x 2 + y2 x 2 + y2

  ∂  x  ∂  y  
=curlq c 2   2 2 
+  2 2 
k
 ∂x  x + y  ∂y  x + y  
 y2 − x 2 x 2 − y2  
= c2  2 2 2
+ 2 2 2
k= 0
 (x + y ) (x + y ) 

Thus, the flow is irrotational, i.e. of potential kind, so we can find φ ( x, y, z ) such that q = −∇φ

∂φ c 2 y ∂φ −c 2 x ∂φ
=−u = 2 , =− v = , =0
∂x x + y 2 ∂y x 2 + y 2 ∂z
The last equation shows that φ = φ (x, y) . The first equation is

∂φ 1 c2 x
= , ⇒ φ (x, y)
2
= c 2 tan −1 ( ) + f (y) (2)
∂x y x y
1+  
y
∂φ −c 2 x
= 2 + f ′(y) ⇒ f ′(y)= 0
∂y x + y 2

Using equation (2)

We have, f(y) = constant (3)
 38 Partial Differential Equations

Since the constant potential gives rise to no velocity, so we may choose the constant in (3) to be zero
x
and therefore from equation (2), we get φ (x, y) = tan −1   (4)
y
Now the equipotential surfaces are given by
x
φ= cons tan t ⇒ = cont. ⇒ =
y c1x (5)
y
From equation (1) and (5), we observe that the streamlines are accordingly orthogonal to the
equipotentials.

Example: Show that a fluid of a constant density can have a velocity q given by

  −2xyz (x 2 − y 2 )z y 
q= 2 2 2
, 2 2 2
, 2 2 
 (x + y ) (x + y ) (x + y ) 
Find the vorticity vector.
Solution: We have

∂u 3x 2 − y 2 ∂v y 2 − 3x 2 ∂w
= 2yz.
= , 2yz.
= , 0
∂x (x 2 + y 2 )3 ∂y (x 2 + y 2 )3 ∂z
  ∂u ∂v ∂w
∴ divq = ∇.q = + + = 0
∂x ∂y ∂z

Thus, q is a velocity vector of a possible fluid of constant density (incompressible).

Also,

∂u ∂v 3y 2 − x 2 3y 2 − x 2
= − 2xz. 2 − 2xz. = 0,
∂y ∂x (x + y 2 )3 (x 2 + y 2 )3
∂v ∂w x 2 − y2 x 2 − y2
−= − = 0,
∂z ∂y (x 2 + y 2 ) 2 (x 2 + y 2 ) 2
∂w ∂u −2xy 2xy
=− + 2 = 0
∂x ∂z (x + y ) (x + y 2 ) 2
2 2 2

   ∂v ∂w   ∂w ∂u   ∂u ∂v 
∴ξ= curlq=  −  −  j +  −  k= 0
i + 
 ∂z ∂y   ∂x ∂z   ∂y ∂x 

Vorticity vector is a zero vector and thus the motion is irrotational, i.e. q = −∇φ .

∂φ 1
=−u =2xyz. 2 =yz(2x)(x 2 + y 2 ) −2
∂x (x + y 2 ) 2
(x 2 + y 2 ) −1 − yz
=⇒ φ yz =
−1 (x + y 2 )
2
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 39
Example: If the velocity of an incompressible fluid at the point (x.y,z) is given by

 3xz 3yz 3z 2 − r 2  cos θ

 5 , 5 , 5  . Prove that the liquid motion is possible and that the velocity potential is 2 .
 r r r  r

Also find the streamlines.

Solution: We have

r 2 = x 2 + y2 + z2
∂r x ∂r y ∂r z
⇒ = , = , =
∂x r ∂y r ∂z r

∂u 3z  5 x  3z
= 10 
r − x.5r 4 .  =7 ( r 2 − 5x 2 )
∂x r  r r
∂v 3z 2 ∂w 3z 2
7 (
= r − 5y 2 ) , = 7 (
r − 5z 2 )
∂y r ∂z r
∂u ∂v ∂w 3z
∴ + + = 7 ( 5r 2 − 5(x 2 + y 2 + z= 2
)) 0
∂x ∂y ∂z r
So the liquid motion is possible.

Now, if the velocity potential exists, then q = −∇φ

∂φ 3xz
⇒ =−u =− 5
∂x r
Integrating, we get
3z
2 ∫
⇒φ =− 2x(x 2 + y 2 + z 2 ) −5/2 dx

3z  −2  z r cos θ cos θ
−   (x 2 + y 2 + z 2 ) −3/2 =3 = 3 = 2
=
2 3  r r r
where we have neglected the constant of integration.
The streamlines are
dx dy dz xdx + ydy + zdz
= = 2
= 2
3xz 3yz 3z − r 3z(x 2 + y 2 + z 2 ) − r 2 z
From 1st and 2nd members, we get x= ay (1)
From 1st and 4nd members, we obtain
dx xdx + ydy + zdz
=
3x 2r 2
40 Partial Differential Equations

udx 2xdx + 2ydy + 2zdz

= 3,
3x x 2 + y2 + z2
⇒ x 4 = (x 2 + y 2 + z 2 )3 (2)

Intersection of (1) and (2) represents the streamlines.

Example: In the steady motion of homogenous liquid if the surfaces=f1 a=
1, f2 a 2 define the
streamlines, prove that the most general value of the velocity component u, v, w are
∂ (f1 , f 2 ) ∂ (f , f ) ∂ (f , f )
F(f1 , f 2 ) , F(f1 , f 2 ) 1 2 , F(f1 , f 2 ) 1 2
∂ (y, z) ∂ (z, x) ∂ (x, y)
where F is an arbitrary function.
Solution: As the motion is steady, so the streamline are independent of t , i.e., f1 , f 2 are functions of x,
y z only.
=
Differentiating 1, f2
f1 a= a2 , we get

∂f1 ∂f ∂f
dx + 1 dy + 1 dz =
0 (1)
∂x ∂y ∂z
∂f1 ∂f ∂f
dx + 1 dy + 1 dz =
0 (2)
∂x ∂y ∂z
Solving (1) and (2), we get
dx dy dz
= =
∂f1 ∂f 2 ∂f1 ∂f 2 ∂f1 ∂f 2 ∂f1 ∂f 2 ∂f1 ∂f 2 ∂f1 ∂f 2
− − −
∂y ∂z ∂z ∂y ∂z ∂x ∂x ∂z ∂x ∂y ∂y ∂x

dx dy dz
i.e., = = (3)
∂ (f1 , f 2 ) ∂ (f1 , f 2 ) ∂ (f1 , f 2 )
∂ (y, z) ∂ (z, x) ∂ (x, y)
But the equation of streamlines are
dx dy dz
= = (4)
u v w
Comparing (3) and (4), we have
u v w
= = = F (say)
∂ (f1 , f 2 ) ∂ (f1 , f 2 ) ∂ (f1 , f 2 )
∂ (y, z) ∂ (z, x) ∂ (x, y)

∂ (f1 , f 2 ) ∂ (f1 , f 2 ) ∂ (f , f )
=u F= , v F= ,w F 1 2 (5)
∂ (y, z) ∂ (z, x) ∂ (x, y)
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 41
Now, we determine the nature of F.
For possible fluid motion, the velocity components must satisfy the equation of continuity, namely
∂u ∂v ∂w
+ + = 0
∂x ∂y ∂z

∂F ∂ (f1 , f 2 ) ∂ ∂ (f1 , f 2 ) ∂F ∂ (f1 , f 2 ) ∂ ∂ (f1 , f 2 ) ∂F ∂ (f1 , f 2 ) ∂ ∂ (f1 , f 2 )

+F + +F + +F 0 (6)
=
∂x ∂ (y, z) ∂x ∂ (y, z) ∂y ∂ (z, x) ∂y ∂ (z, x) ∂z ∂ (x, y) ∂z ∂ (x, y)
∂ ∂ (f1 , f 2 ) ∂ ∂ (f1 , f 2 ) ∂F ∂ (f1 , f 2 )
But + + 0 (By property of Jacobian)
=
∂x ∂ (y1 , z) ∂x ∂ (z1 , x) ∂y ∂ (x1 , y)

So, equation (6) becomes

∂ ∂ (f1 , f 2 ) ∂ ∂ (f1 , f 2 ) ∂F ∂ (f1 , f 2 )
+ + 0
=
∂x ∂ (y1 , z) ∂x ∂ (z1 , x) ∂y ∂ (x1 , y)

∂F ∂F ∂F
∂x ∂y ∂z
∂f ∂f1 ∂f1 ∂ (F, f1 , f 2 )
⇒ 1 = 0,i.e., =0
∂x ∂y ∂z ∂ (x, y, z)
∂f 2 ∂f 2 ∂f 2
∂x ∂y ∂z

But the vanishing Jacobian means that F , f1 , f 2 are not independent. So, F can be written as function of

f1 and f 2 , i.e., F=F( f1 , f 2 ).

Using F in (5), we get the required velocity components.

Example: Prove that the liquid motion is possible when velocity at the point (x, y, z) is given by

 3x 2 − r 2 3xy 3xz 
 5
, 5 , 5  and the streamlines are the intersection of the surfaces
 r r r 

(x + y 2 + z 2 ) = c(y 2 + z 2 ) 2 by the planes passing through OX. State whether the motion is irrotational.
2 3

∂u ∂v ∂w
Solution: + + 0 ⇒ Liquid motion is possible.
=
∂x ∂y ∂z
Streamlines are
dx dy dz
2
= 2
=
3x − r 3xy 3zx
Last two members give y = az (1)
which is a plane passing through OX.
42 Partial Differential Equations

Also,
xdx + ydy + zdz ydy + zdz
2 2
=
x(3r − r ) 3x(y 2 + z 2 )
(2xdx + 2ydy + 2zdz 2(2ydy + 2zdz)
⇒ 3. =
x 2 + y2 + z2 y2 + z2

⇒ (x 2 + y 2 + z 2 )3= c(y 2 + z 2 ) 2 (2)

The streamlines are intersections of equations (1) and (2).

Hence the result.
∂w ∂v ∂u ∂w ∂v ∂u  
We can find that − = 0, − = 0, − = 0 , so that curlq = 0 , i.e. motion is irrotational.
∂y ∂z ∂z ∂x ∂x ∂y

Example: Describe the fluid motion given by q = − w(yi − x j) , w being constant.
 
Solution: We note that the equation of continuity ∇q = 0 is satisfied, so q is possible velocity vector
for motion of an incompressible fluid.
Also,
i j k

∇×q ∂= ∂ ∂ 2wk
∂x ∂y ∂z
− wy wx 0

Thus, the flow is not of potential kind.

It can easily be shown that a rigid body rotating about the z-axis with constant vector angular velocity
wk gives the same type of motion.
The equations of streamlines are
dx dy dz
= =
− wy wx 0

⇒ x 2 + y2 =
constant, z=constant, which are the circles with centre on the z-axis.

1.10 Boundary Surfaces

Physical conditions that should be satisfied on given boundaries
of the fluid in motion are called boundary conditions. The
simplest boundary condition occurs where an ideal and
incompressible fluid is in contact with the rigid impermeable
boundary, e.g., wall of a container or the surface of a body which
is moving through the fluid.
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 43

Let P be any point on the boundary surface where the velocity of fluid is q and velocity of the

boundary surfaces is u .
The velocity at the point of contact of the boundary surface and the liquid must be tangential to the
surface otherwise the fluid will break its contact with the boundary surface. Thus, if n̂ be the unit normal
to the surface at the point of contact, then
   
(q − u).n =0 ⇒ q.n =u.n (1)

In particular, if the boundary surfaces is at rest, then u = 0 and the condition becomes

q.n = 0 (2)

Another type of boundary condition arrives at a free surface where liquid borders a vacuum e.g. the
interface between liquid and air is usually regarded as free surface. For this free surface, pressure p
satisfies
p= Π (3)

where Π denotes the pressure outside the fluid i.e. the atmospheric pressure. Equation (3) is a dynamic
boundary condition.
The third type of boundary condition occurs at the boundary between two immiscible ideal fluids in
 
which the velocities are q1 & q 2 and pressures are p1 and p 2 , respectively.

Now, we find the condition that a given surfaces satisfy to be a boundary surface.
1.9.1 Article: To obtain the differential equation satisfied by the boundary surface of a fluid in motion.
or

To find the condition that the surface
= F(r, t) F(x,
= y, z, t) 0 may represent a boundary surface.
 
Proof: If q be the velocity of fluid and u be the velocity of the boundary surfaces at a point P of
contact, then
   
(q − u).n =0 ⇒ q.n = u.n (1)
 
where (q − u) is the relative and n is a unit vector normal to the surface at P.

The equation of the given surface= is F(r, t) F(x,
= y, z, t) 0 (2)

We know that a unit vector normal to the surface (2) is given by

∇F
n =
∇F
 
Thus, from (1), we get q.∇F = u.∇F (3)
44 Partial Differential Equations

Since the boundary surface is itself in motion, therefore at time (t + δ t ) , it’s equation is given by
 
F(r + δ r, t + δ t) =
0 (4)

From (2) and (4), we have

  
F(r + δ r, t + δ t) − F(r, t) = 0
    
i.e. F(r + δ r, t + δ t) − F(r, t + δ t) + F(r, t + δ t) − F(r, t) =
0

By Taylor’s series, we can have

  ∂ 
(δ r.∇)F(r, t + δ t) + δ t {
∂t
}
F(r, t) =
0

∂F ∂F ∂F 
| F(x + δ x, y + δ y, z +
= δ z) F(x, y, z) + δ x +δ y +δz = + ... F(x, y, z) + δ r.∇F
∂x ∂y ∂z

δr   ∂F
⇒  .∇  F(r, t + δ t) + 0
=
 δ t  ∂t

Taking limit as δ t → 0 , we get


 dr   ∂F
⇒  .∇  F(r, t + δ t) + 0
=
 dt  ∂t
∂F 
⇒
∂t
( )
+ q.∇ F =
0 (5)

DF
i.e. =0
Dt
which is the required condition for any surface F to be a boundary surface.

Corollary: (1) q = (u, v, w) , then the condition (5) becomes

∂F ∂F ∂F ∂F
+u +v +w 0
=
∂t ∂x ∂y ∂z
∂F
In case, the surface is rigid and does not move with time, then = 0 and the boundary condition is
∂t
∂F ∂F ∂F 
u +v +w 0 i.e.
= (q.∇)F =
0
∂x ∂y ∂z
∂F ∂F ∂F ∂F
Corollary: (2) The boundary condition +u +v +w 0 is a linear equation and its solution
=
∂t ∂x ∂y ∂z
gives
dt dx dy dz D d
= = = | ≡ in Lagrangian view
1 u v w Dt dt
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 45

dx dy dz
⇒ = u, = v, = w
dt dt dt
which are the equations of path lines.
Hence, once a particle is in contact with the surface.
Corollary: (3) From equation (5), we have
 −∂F
(q.∇)F =
∂t
−∂F
 ∇F
⇒ q. = ∂t
∇F ∇F
−∂F

⇒ q.n =∂t
∇F
which gives the normal velocity.
Also from (1), we get
−∂F
  
⇒ u.n =∂t | q.n = u.n
∇F

which gives the normal velocity of the boundary surface.

Example: Show that the ellipsoid

x2 
n  y
2
z 
2

+ kt   +    = 1
a 2 k 2 t 2n  b   c  
is a possible form of the boundary surface of a liquid.
Solution: The surface F(x, y, z, t) = 0 can be a possible boundary surface, if it satisfies the boundary
condition,
DF ∂F ∂F ∂F ∂F
= +u +v +w =0 (1)
Dt ∂t ∂x ∂y ∂z
where u, v, w satisfy the equation of continuity
 ∂u ∂v ∂w
∇.q =0 i.e. + + 0
= (2)
∂x ∂y ∂z

x2 
n  y
2
z 
2

Here, F ( x, y, z , t ) ≡ 2 2 2n + kt   +    − 1 =0
a k t  b   c  
46 Partial Differential Equations

∂F x 2 .2n  y  2  z  2 
Therefore, − 2 2 2n +1 + nkt n −1   +   
=
∂t a k t  b   c  

∂F 2x ∂F 2kt n y ∂F 2kt n z
= 2= , = , ,
∂x a k 2 t 2n ∂y b 2 ∂z c2
Thus, from (1), we get

x 2 .2n 
n −1  y 
2
z 
2
2xu 2kt n yv 2kt n zw
− + nkt +
     + 2 2 2n
+ + 0
=
a 2 k 2 t 2n +1  b   c   a k t b2 c2

 nx  x 2  ny  2kt n y  nz  2kt n z
or  u −  2 2 2n +  v +  2 +  w +  2 = 0
 t a k t  2t  b  2t  c
which will hold, if we take

 nx   ny   nz 
 u −  = 0,  v +  = 0,  w +  = 0
 t   2t   2t 
nx ny nz
⇒ u = ,v = − ,w = −
t 2t 2t
It will be a justifiable step if equation (2) is satisfied.
n −n −n
i.e. + + 0 which is true.
=
t 2t 2t
Hence the given ellipsoid is a possible form of the boundary surface of a liquid.
Example: Obtain the condition for the surface z = η ( x, y, t ) to be the boundary of a moving fluid.

z − η (x, y, t) =
Solution: Write F(x, y, z, t) = 0 (1)

We know that for the surface F(x, y,= z, t) F(r,
= t) 0 the condition for it to be a boundary of a moving
dF ∂F 
fluid is = + (q.∇)F = 0 (2)
dt ∂t
From (1) and (2), we get
dz ∂η dx ∂η dy ∂η
− − − 0
=
dt ∂x dt ∂y dt ∂t

∂η ∂η ∂η
⇒ w−u −v − 0
=
∂x ∂y ∂t
∂η ∂η ∂η
⇒u +v + −w =
0
∂x ∂y ∂t
which is the required condition.
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 47

x2 y2
Example: Show that 2 tan t + 2 co t 2 t =
2
1 is possible form for boundary surface of a liquid. Also find
a b
an expression for the normal velocity.

x2 2 y2 2
Solution: Here, F(x, y,
= z, t) tan t + co t= t −1 0 (1)
a2 b2
∂F 2x ∂F 2y ∂F ∂F x 2 y2
⇒ = 2 tan 2 t, = 2 co t 2 t, = 0, = 2 2.tan t(sec 2 t) + 2 2.cot t(− cos ec 2 t)
∂x a ∂y b ∂z ∂t a b
The condition for (1) to be boundary surface of a liquid is
dF ∂F ∂F ∂F ∂F
= +u +v +w =0 (2)
dt ∂t ∂x ∂y ∂z

2xu 2 2yv 2 2x 2 2 2y 2
⇒ tan t + co t t + tan t(sec t) − cot t(cos ec 2 t)
a2 b2 a2 b2

2x 2  x sec 2 t  2y 2  y cos ec 2 t 
⇒ 2 tan t  u +  + co t t  v − =0
a  tan t  b 2  cot t 

which will be satisfied if we take,

x sec 2 t yco sec 2 t

u=
− ,v =
− ,w =
w(x, y, t)
tan t cot t
∂u ∂v ∂w
We note that u, v, w must satisfy the equation of continuity + + 0 (3)
=
∂x ∂y ∂z
−1 1
i.e. + +0=0 , which is true.
sin t cos t sin t cos t
Hence the given surface is a possible form for the boundary surface of a liquid.

 − ∂F − ∂F
= q.n
Normal velocity = ∂
=t ∂t
1
∇F  ∂F
( ) ( )
2
2
∂F  2
 +
 ∂x ∂y 

x2 2 y2
− 2
2.tan t(sec t) + 2
2.cot t(cos ec 2 t)
= a b
1
 2x 2 
2
 2y 2 
2
 2

 2 tan t  +  2 co t t  
 a  b  
a 2 y 2 cot t(cos ec 2 t) − b 2 x 2 tan t(sec 2 t)
= 1
 b 4 x 2 tan 4 t + a 4 y 2 co t 4 t  2
48 Partial Differential Equations

Example: Show that all necessary conditions can be satisfied by a velocity potential of the form
φ = α x 2 + β y 2 + γ z 2 and a boundary surface of the form F = ax 4 + by 4 + cz 4 − X(t) = 0 where X(t) is a
given functions of time and α , β , γ ;a, b, c are suitable functions of time.
Solution: Here, the velocity potential is given by
  ∂φ ∂φ ∂φ 
q = −∇φ = −  , ,  = (u, v, w) (1)
 ∂x ∂y ∂z 
and the equation of continuity is

∇.q = 0 ⇒ ∇ 2φ = 0
∂ 2φ ∂ 2φ ∂ 2φ
i.e. + + 0
=
∂x 2 ∂y 2 ∂z 2

⇒ 2α + 2 β + 2γ = 0 ⇒ α + β + γ = 0 (2)

Now, the boundary surface condition is

∂F ∂F ∂F ∂F
+u +v +w 0
= (3)
∂t ∂x ∂y ∂z

where F = ax 4 + by 4 + cz 4 − X(t) = 0 (4)

Using (1) and (4) in (3), we get

a ' x 4 + b ' y 4 + c' z 4 − X ' (t) − 2α x(4ax 3 ) − 2 β y(4by3 ) − 2γ z(4cz3 ) =

0
(5)
⇒ (a ' − 8α a)x 4 + (b ' − 8β b)y 4 + (c' − 8γ c)z 4 − X ' (t) =
0
Since (4) and (5) both must hold for all points (x, y, z) on the surface, so they are identical. Comparing,
a ' − 8α a b ' − 8β b c' − 8γ c X ' (t)
we get = = =
a b c X(t)
These conditions will hold if α , β , γ ;a, b, c are some suitable functions of time, where α + β + γ =0
Hence φ and F=0 satisfy the necessary condition for velocity potential and boundary surface if
α , β , γ ;a, b, c are some suitable function of time.

x2 y2 z2
Example: Determine the restriction on f1 , f 2 , f 3 if f1 (t) + f 2 (t) + f 3 (t) =
1 is a possible form of
a2 b2 c2
boundary surface of a moving liquid.

x2 y2 z2
Solution: F(x, y,=
z, t) f1 (t) + f 2 (t) + f 3 (t)
= −1 0
a2 b2 c2
The boundary surface condition is
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 49
∂F ∂F ∂F ∂F
+u +v +w 0
=
∂t ∂x ∂y ∂z
x2 ' y2 ' z2 ' 2x 2y 2z
⇒ 2
f 1 (t) + 2
f 2 (t) + 2
f 3 (t) + u 2 f1 + v 2 f 2 + w 2 f 3 =0
a b c a b c
2x  xf '  2y  yf '  2z  zf ' 
⇒ 2 f1  u + 1  + 2 f 2  v + 2  + 2 f 3  w + 3  = 0
a  2f1  b  2f 2  c  2f 3 

This will be satisfied if we take

xf ' yf ' zf '

u=
− 1 ,v = − 3 , where u, v, w must satisfied the continuity equation
− 2 ,w =
2f1 2f 2 2f 3

∂u ∂v ∂w
+ + 0
=
∂x ∂y ∂z
1f' f' f' 
⇒ −  1 + 2 + 3  =0
2  f1 f 2 f 3 

Integrating, we get
log(f1f 2 f 3 ) = constant

f1f 2 f 3 =constant

which is the required restriction.

1.11 Acceleration at a Point of a Fluid
Suppose that a fluid particle is moving along a curve C, initially it being at point A(t0=0) with

position vector r A . Let P and P' be its positions at time t and t +δt with position vectors r and r+δr
 
respectively. Therefore, δ r = PP′
The points A, P, P` are geometrical points of region occupied by fluid and they coincide with the
locations of the same fluid particle at times t0, t, t + δt, respectively.

Let f be the acceleration of the particle at time t when it coincides with P. By definition
 (change in particle velocity in time δ t)
f = Lt (1)
δ t →0 δt
  
But the particle velocity at time t is q(r, t) and at time t +δt it is q(r + δ r, t + δ t) .

Thus (1) becomes

  
 [q(r + δ r, t + δ t) − q(r, t)]
f = Lt (2)
δ t →0 δt
50 Partial Differential Equations
       
q(r + δ r, t + δ t) − q(r, t) q(r + δ r, t + δ t) − q(r, t + δ t) q(r, t + δ t) − q(r, t)
Now,
= + (3)
δt δt δt

Since r is independent of time t, therefore
   
[q(r + δ r, t + δ t) − q(r, t)] ∂ q
Lt = (4)
δ t →0 δt ∂t
Using Taylor’s expansion, we get
    
q(r + δ r, t + δ t) − q(r, t + δ =
t) (δ r.∇)q(r, t + δ t)+ ∈ (5)

where ∈ =O[(δ r) 2 ]

 ∂ ∂ ∂ 
 F(x + δ x, y + δ y, z + δ z) − F(x, y, =
z)  δ x + δ y + δ z  F(x, y, z)
 ∂x ∂y ∂z 
2
1 ∂ ∂ ∂ 
+ δ x + δ y + δ z  .F(x, y, z) + ...
2 !  ∂x ∂y ∂z 

 ∂ ∂ ∂   
and  δ x + δ y + δ z = (δ r.∇) , where δ=  ∇ ≡ ∂ i + ∂ j + ∂ k
r δ xi + δ yi + δ zi,

 ∂x ∂y ∂z  ∂x ∂y ∂z

But δ r is merely the displacement of the fluid particle in time δt, therefore,
 
δ r = q(r, t)δ t (6)

Thus, from (5), we obtain

  
[q(r + δ r, t + δ t) − q(r, t + δ t)]  
Lt = (q.∇)q (7)
δ t →0 δt

where R.H.S. of (4) and (7) are evaluated at P( r , t).
Hence, from (2), the acceleration of fluid at P in vector form is
given by

 ∂q  
f= + (q.∇)q (8)
∂t
Remark: We have obtained the acceleration, i.e. rate of change

of velocity q . The same procedure can be applied to find the rate of change of any physical property

associated with the fluid, such as density. Thus, if F ( r , t) i s any scalar or vector quantity
associated with the fluid, its rate of change at time t is given by
DF ∂F 
= + (q.∇)F
Dt ∂t
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 51
D ∂ 
The operator = + (q.∇) is Lagrangian operator.
Dt ∂t
1.11.1 Component of Acceleration in Cartesian Co-ordinates
  
Let u, v, w be the cartesian components of q and f1 , f 2 , f 3 that of f i.e. q =(u, v, w),

f =( f1 , f 2 , f 3 ).

 ∂q  
Then from equation f = + (q.∇)q (1)
∂t
We get
∂v ∂v ∂v ∂v
f1 = + u +v +w
∂t ∂x ∂y ∂z
∂u ∂u ∂u ∂u
f2 = + u +v +w
∂t ∂x ∂y ∂z
∂w ∂w ∂w ∂w
f3 = + u +v +w
∂t ∂x ∂y ∂z

which are the required cartesian components of f .
In tensor form with co-ordinates xi, and velocity components qi (i = 1, 2, 3), the above set of equations
can be written as
∂q i ∂q
fi
= + q jq i, j , where q i, j = i
∂t ∂x j

1.11.2 Component of Acceleration in Curvilinear Co-ordinates

Before obtaining the acceleration components in curvilinear co-ordinates, we obtained a more suitable
form of equation (1), as

 ∂q 1 2  
=f + ∇( q ) − q × (∇ × q)
∂t 2
 
∂q 1  2   ∂q 1 2  
= + ∇( q ) − q × = ξ + ∇( q ) + ξ × q
∂t 2 ∂t 2
  
where ξ = curlq = ∇ × q
  
   ∂q  ∂q  ∂q
We have (q.∇= )q (q.i)  
+ (q.j) 
+ (q.k) (2)
∂x ∂y ∂z
  
For ay three vectors A, B, C , we have
 52 Partial Differential Equations
        
A × (B ×= C) (A.C)B − (A.B)C
        
(A.B)C
= (A.C)B − A × (B × C)

    ∂ q
In particular, taking = A q,= B i,= C , we get
∂x
  
 ∂q  ∂q  ∂ q
 = (q. )i − q(i × )
(q.i)
∂x ∂x ∂x

∂  1  2    ∂q 
= i  q  − q ×  i ×  (3)
∂x  2   ∂x 
 
 ∂q ∂  1  2    ∂q 

(q.j) j  (4)
=
∂y  q  − q ×  j× 
∂y  2   ∂y 
 
 ∂q ∂  1 2    ∂ q 

(q.k) = k  q  − q ×  k ×  (5)
∂z ∂z  2   ∂z 

Adding (3), (4) and (5), we get


   1 2     ∂q 
(q.∇)q =∇  q  − q × ∑  j × 
2   ∂x 
 1 2   
= ∇ q  − q× ∇×q
2 
( )
Thus, from (1), we obtained
 
 dq ∂ q  1 2   
f
= =
dt ∂t
+ ∇ q  − q× ∇×q
2 
( )

∂q  1 2   
= + ∇ q  +ξ ×q (6)
∂t 2 
Now, let (u1 , u 2 , u 3 ) denote the orthogonal curvilinear co-ordinates.
  
Also,
= let q (q= 1 , q 2 , q 3 ), f (f
= , f
1 2 3, f ), ξ (ξ1 , ξ 2 , ξ3 ) , where the term have their usual meaning. We know
that the expression for the operator ∇ in curvilinear co-ordinates is

1 ∂ 1 ∂ 1 ∂ 
∇≡ , ,  , where h1 , h 2 , h 3 are scalar factors.
 h1 ∂u1 h 2 ∂u 2 h 3 ∂u 3 
 
The components of ξ = curlq n the curvilinear system are given by
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 53

1  ∂ ∂ 
=ξ1  (h 3q 3 ) − (h 2 q 2 )  
h 3h 2  ∂u 2 ∂u 3 
1  ∂ ∂  
=ξ2  (h1q1 ) − (h 3q 3 )   (7)
h 1h 3  ∂u 3 ∂u1 
1  ∂ ∂  
=ξ3  (h q
2 2 ) − (h q
1 1 )
h 1h 2  ∂u1 ∂u 2  
Using these results in (6), we find that

∂q1 1 ∂ 
f1 = + (q12 + q 2 2 + q 32 ) + (ξ 2 q 3 − ξ3q 2 ) 
∂t 2h1 ∂u1 
∂q 2 1 ∂ 
f2 = + (q12 + q 2 2 + q 32 ) + (ξ3q1 − ξ1q 3 )  (8)
∂t 2h 2 ∂u 2 
∂q 3 1 ∂ 
f3 = + (q12 + q 2 2 + q 32 ) + (ξ1q 2 − ξ 2 q1 ) 
∂t 2h 3 ∂u 3 
which are the components of acceleration in curvilinear co-ordinates.
Now, we write the components of the acceleration in cylindrical (r, θ, z) and spherical (r, θ, ψ )
co-ordinates.
1.11.3 Components of Acceleration in Cylindrical Co-ordinates (r, θ, z)
Here,

=q ( q1 , q 2 , q3 ) & =
h1 1,=
h 2 r,=
h3 1

∂ 1 ∂ ∂ 
Therefore, ∇ ≡  , , 
 ∂r r ∂θ ∂z 
and

1 ∂ ∂  1 ∂q ∂q 
ξ1 = (q 3 ) − (rq 2 )  = 3 − 2 
r  ∂θ ∂z  r ∂θ ∂z

∂ ∂  ∂q1 ∂ q 
ξ2 =  ∂z (q1 ) − (q 3 ) = −
 ∂z ∂r
3

∂r 
1 ∂ ∂  ∂q q 1 ∂q 
ξ3 =  (rq 2 ) − (q1 )  = 2
+ 2− 1

r  ∂r ∂θ  ∂r r r ∂θ 

Thus,
54 Partial Differential Equations

∂q1 1 ∂ ∂q ∂q  ∂q q 2 q ∂q 
f1 = + (q12 + q 2 2 + q 32 ) + (q 3 1 − q1 3 ) −  q 2 2 + 2 − 2 1 
∂t 2 ∂r ∂z ∂r  ∂r r r ∂θ 
∂q ∂q ∂q ∂q ∂q ∂q ∂q q 2 q ∂q
= 1 + q1 1 + q 2 2 + q 3 3 + q 3 1 − q1 3 − q 2 2 − 2 + 2 1
∂t ∂r ∂r ∂r ∂z ∂r ∂r r r ∂θ
2
∂q ∂q ∂q q q ∂q
= 1 + q1 1 + q 3 1 − 2 + 2 1
∂t ∂r ∂z r r ∂θ
If we define the differential operator
D d ∂ ∂ q ∂ ∂
≡ = + q1 + 2 + q3
Dt dt ∂t ∂r r ∂θ ∂z

Dq1 q 2 2 Du v 2 
f1 = − ≡ −
Dt r Dt r 
Dq 2 q1q 2 Dv uv 
f2 = + ≡ +  (9)
Dt r Dt r 
Dq1 Dw 
f3
= ≡ 
Dt Dt 
where (q1 , q 2 , q 3 ) = (u, v, w)

Equation (9) gives the required components of acceleration in cylindrical co-ordinates.

1.11.4 Components of Acceleration in Spherical Co-Ordinates (r, θ ,ψ )

Here, = q ( q1 , q 2 , q 3 ) & =
h1 1,= h 3 r sin θ
h 2 r,=

∂ 1 ∂ ∂ 
Therefore, ∇ ≡  , , 
 ∂r r ∂θ r sin θ∂ψ 
and

1 ∂ ∂  1  ∂q ∂q  
=ξ1  ∂θ (r sin θ q 3 ) − ∂ψ (rq = 2 )  q 3 cos θ + sin θ 3 − 2  
r sin θ
2
  r sin θ  ∂θ ∂ψ  
1  ∂ ∂  1  ∂q1 ∂q 3  
ξ2
= (q ) − (r sin θ q =) − sin θ q − r sin θ 
r sin θ  ∂ψ
3   ∂r  
1 3
∂r  r sin θ  ∂ψ
1 ∂ ∂   ∂q ∂q  
ξ3 =  (rq 2 ) − (q1 )  =q 2 + r 2 − 1  
r  ∂r ∂θ   ∂r ∂θ  

Thus,
Eulerian and Lagrangian Descriptions of Fluid and Conservation of Mass 55

∂q1 1 ∂  ∂q1 q3 ∂q 3  q 2  ∂q 2 ∂q1 

f1 = +
∂t 2 ∂r  ∂ψ − q 3 sin θ − r sin θ ∂r  − r  q 2 + r ∂r − ∂θ 
(q12 + q 2 2 + q 32 ) +
r sin θ
   
∂q ∂q ∂q ∂q q 3 ∂q1 q 32 ∂q ∂q q 2 q ∂q
= 1 + q1 1 + q 2 2 + q 3 3 + − − q3 3 − q 2 3 − 2 + 2 1
∂t ∂r ∂r ∂r r sin θ ∂ψ r ∂r ∂r r r ∂θ
If we define the differential operator
D d ∂ ∂ q ∂ q3 ∂
≡ = + q1 + 2 +
Dt dt ∂t ∂r r ∂θ r sin θ ∂ψ

Then,

Dq1 q 2 2 + q 32 Du v 2 + w 2 
f1 = − ≡ − 
Dt r Dt r 
Dq 2 q1q 2 − q 32 cot θ Dv uv − w 2 cot θ 
f2 = + ≡ + 
Dt r Dt r 
Dq 3 q1q 2 + q 2 q 3 cot θ Dw w(u + v cot θ ) 
f3 = + ≡ + 
Dt r Dt r 
Equation (10) gives the required components of acceleration in spherical co-ordinates.
1.12 Check your progress
1. Discuss the methods of Eulerian and Lagrangian.
2. Derive the equation of continuity.

3. Find the flowlines for the two dimensional velocity field
= q ( x(1 + 2t), y, 0 ) which pass through

(1, 1, 0) at time t=0.

Answer. Streamlines: y = x, z = 0
1+ log y
=
Pathlines: x y= ,z 0
1− log y
=
Streaklines: x y= ,z 0

4. What is boundary surface and derive necessary and sufficient condition for a surface to be boundary
surface.
5. What is acceleration and find components of it.
6. If the velocity of an incompressible fluid at the point (x,y,z) is given by

 3 xz 3 yz 3 xy  cos θ
 5 , 5 , 5  , prove that the liquid motion is possible and that the velocity potential is . Also
 r r r  r2
determine the streamline.
 56 Partial Differential Equations

1.13 Summary
In this section, we have studied methods to study the dynamic motion of fluids, flow lines and its types,
conservation of mass, boundary surface and acceleration.
1.14 Keyword
Streamlines, Pathlines and Streak line, Velocity, Boundary condition, and Acceleration at a point
1.15 Self-Assessment Test
1. Find the relationship between Eulerian and Lagrangian method.
2. Find whether the flow with velocity components
−2𝑥𝑥𝑦𝑦𝑦𝑦 �𝑥𝑥 2 −𝑦𝑦 2 �𝑧𝑧 𝑦𝑦
𝑢𝑢 = (𝑥𝑥 2 , 𝑣𝑣 = , 𝑧𝑧 = 𝑥𝑥 2 +𝑦𝑦 2
+𝑦𝑦 2 )2 (𝑥𝑥 2 +𝑦𝑦 2 )2

3. If the velocity components of a flow are

ay ax
u= − 2 2
, v= 2 , w=c, show that the stream lines are helices
= cos t , y a=
x a= sin t , z act ,
x +y x + y2
described on the cylindrical surface: x 2 + y 2 =
a2
4. The velocity components of a certain flow are
2 2 2 1
y + z − 2x 3kxy 3kxz
u=
k 5
,v= − 5 , where k is a constant and r=(x 2 + y 2 + z ) .
− 5 , w= 2 2

r r r
5. Find the velocity field of an irrotational spherical symmetric flow whose stream function is
3 2
Ua Ur
ψ (r ,θ )
= sin 2 θ − sin 2 θ , where U and a are constants.
2r 2
ui
6. Prove that in an unsteady flow the streamlines and the pathlines will coincide if u is independent of
i

time.
1.16 Suggested Reading
(i) F. Chorlton, Text Book of Fluid Dynamics, C.B.S. Publishers, Delhi, 1985
(ii) O’Neill, M.E. and Chorlton, F., Ideal and Incompressible Fluid Dynamics, Ellis Horwood Limited,
1986.
(iii) R.K. Rathy, An Introduction to Fluid Dynamics, Oxford and IBH Publishing Company, New
Delhi, 1976.
 UNIT - II
PRESSURE AND EQUATION OF IMPULSIVE ACTION

Content
2.1 Objective
2.2 Introduction
2.3 Euler Equation of Motion
2.4 Bernoulli Equation
2.5 Impulsive Motion
2.6 Kelvin Circulation Theorem
2.7 Vorticity Equation
2.8 Energy Equation for Incompressible Flow
2.9 Kinetic Energy of Irrotational Flow
2.10 Kelvin Minimum Energy Theorem
2.11 Kinetic Energy of Infinite Fluid
2.12 Uniqueness Theorems
2.13 Check Your Progress
2.14 Summary of the Unit
2.15 Keyword
2.16 Self-Assessment Test
2.17 Suggested Reading
2.1 Objective
In this unit reader will learn about Pressure at a point, Conservation of Momentum, Implusive action,
Bernoulli’s equation for steady and unsteady motion, Kinetic energy and theorems on energy.
2.2 Introduction
Let P represents a point in a ideal (inviscid) fluid moving with

velocity q . We insert an elementary rigid plane area δA into
this fluid at point P. This plane area also moves with the
 
velocity q of the local fluid at P. If δ F denotes the force
58 Partial Differential Equations

exerted on one side of δA by the fluid particles on the other side, then this force will act normal to δA.

δF
Further, if we assume that Lt exists uniquely, then this limit is called the (hydrodynamic) fluid
δ t →0 δ A

pressure at point P and is denoted by p. In general, the pressure is always equal in all the directions and
it will be proved in next theorem.
Theorem: Prove that the pressure p at a point P in a moving inviscid fluid is same in all directions.

Proof: Let q be the velocity of the fluid. We consider an elementary tetrahedron PQRS of the fluid at a
point P of the moving fluid. Let the edges of the tetrahedron be PQ = δx, PR = δy, PS = δz at time t,
where δx, δy, δz are taken along the co-ordinate axes OX, OY, OZ, respectively. This tetrahedron is

also moving with the velocity q of the local fluid at P. Let p be the pressure on the face QRS where
area is δA. Suppose that < l, m,n> are the d.c’s of the normal to δs drawn outwards from the tetrahedron.
Then,
lδA = projection of the area δs on yz-plane.
= area of face PRS (triangle)
1 δ yδ z
= δ y.δ z =
2 2
Similarly, mδA = area of face PQS (triangle)
1 δ xδ z
= δ x.δ z =
2 2
nδA = area of face PQR (triangle)
1 δ yδ x
= δ y.δ x =
2 2
The total force exerted by the fluid, outside the tetrahedron,
on the face QRS is

− pδ A(li + m j + nk )
=
− p (lδAsi + mδA j + nδA k )
=
p
− (δ yδ zi + δ xδ z j + δ yδ xk )
=
2
Let px, py, pz be the pressures on the faces PRS, PQS, PRQ.
The forces exertedon these faces by the exterior fluid are
1  1 p δ xδ zj, 1 p δ yδ xk respectively.
p xδ yδ zi, y z
2 2 2
Thus, the total surface force on the tetrahedron is
Pressure and Equation of Impulsive Action 59

p  + 1 p δ yδ zi + 1 p δ xδ zj + 1 p δ yδ xk

− (δ yδ zi + δ xδ zj + δ yδ xk) x y z
2 2 2 2
1
(p x − p)δ yδ zi + (p y − p)δ xδ zj + (p z − p)δ yδ xk  (1)
2 

In addition to surface force (fluid forces), the fluid may be subjected to body forces which are due to
external ca


uses such as gravity. Let F be the mean body force per unit mass within the tetrahedron. Volume of the
1 1
tetrahedron PQRS is hδ A i.e. δ xδ yδ z , where h is the perpendicular from P on the face QRS.
3 6
1 
Thus, the total force acting on the tetrahedron PQRS is = ρ Fδ xδ yδ z (2)
6
1 1
where ρ is the mean density of the fluid and volume of tetrahedron hδ A or δ xδ yδ z where h is the
3 6
perpendicular from P on the face QRS
From (1) and (2), the net force acting on the tetrahedron is
1 1 
(p x − p)δ yδ zi + (p y − p)δ xδ zj + (p z − p)δ yδ xk  + ρ Fδ xδ yδ z
2  6

Now, the acceleration of the tetrahedron is dq and the mass 1
ρδ xδ yδ z
dt 6
of fluid inside it is constant.
Thus, the equation of motion of the fluid contained in the tetrahedron is

1 1  1  dq 
  
(p x − p)δ yδ zi + (p y − p)δ xδ z j + (p z − p)δ yδ xk  + ρ Fδ xδ yδ z = ρδ xδ yδ z  
2  6 6  dt 
 
(f = ma)

   1  1  dq 
i.e. (p x − p)lδ Ai + (p y − p)mδ A j + (p z − p)nδ Ak + ρ Fhδ A = ρ hδ A  
3 3  dt 
This equation contains quantities of the second and third orders of small quantities. When we make the
edges of the tetrahedron vanishing small, then to the second order, we have
1
(p x − p)δ yδ zi + (p y − p)δ xδ zj + (p z − p)δ yδ xk  =0
2 
60 Partial Differential Equations

p x − p= 0, p y − p= 0, p z − p= 0
(3)
⇒ px = p y = pz
Since the choice of axes is arbitrary, the relation (3) establishes that at any point P of a moving ideal
fluid, the pressure p is same in all directions.
2.3 Euler’s Equations of Motion
2.3.1 Euler’s Equation of Motion of an Ideal Fluid (Equation of Conservation of Momentum)
The equation of motion of a fluid is, in its most fundamental form, a relation equating the rate of change
of momentum of a specified region of the fluid to the vector sum of all forces which act on that region,
in accordance with Newton’s second law of motion.
Consider a region V of an ideal fluid bounded by the closed surface S which consists of the same fluid

particles at all times. Let q be the velocity and ρ be the density of the fluid. Then ρdτ is an element of
mass within S and its remains constant.
The linear momentum of volume τ is
 
M = ∫ qρ dτ |mass × velocity = momentum
τ

 
dM d  dq
dt dt ∫τ ∫τ dt dτ
Rate of change of momentum =
is = qdτ (1)

The fluid within τ is acted upon by two types of forces.

The first types of forces are the surface forces which are due to the
fluid exterior to τ. Since the fluid is ideal, the surface force is simply
the pressure p directed alongthe inward normal at all point of S.

The total surface force on S is ∫ p(−n)dS  =
− ∫ pndS
S
=
S
∫ ∇pdττ
(By gauss div. theorem) (2)


The second type of forces are the body forces which are due to some external agent. Let F be the body
force per unit mass acting on the fluid. Then F ρdτ is the body force on the element of mass ρdτ and
the total body force on the mass within τ is

∫ ρ dτ
F
τ
(3)

By Newton’s second law of motion, we have

Rate of change of momentum = total force
 Pressure and Equation of Impulsive Action 61

dq 
⇒∫ ρ = dτ ∫ Fρ dτ − ∫ ∇pdτ
τ dt τ τ

 dq  
⇒ ∫  ρ − Fρ − ∇p dτ = 0
τ  dt 
Since dτ is arbitrary, we get

dq 
ρ − Fρ − ∇p =0
dt

dq  1
= F − ∇p (4)
dt ρ
which holds at every point of the fluid and is known as Euler’s dynamical equation for an ideal fluid.
Remark: The above method for obtaining the Euler’s dynamical equation of motion, is also known as
flux method.
2.3.2 Other forms of Euler’s equation of motion
d D ∂ 
(i) We know that = = + (q.∇)
dt Dt ∂t
Therefore, equation (4) becomes

∂ q    1
+ (q.∇)q = F − ∇p (5)
∂t ρ
   1 2     
But (q.∇)q =∇  q  + ξ × q, ξ =curlq
2 
Therefore, Euler’s equation of motion becomes

∂q  1 2     1
+ ∇  q  + ξ × q = F − ∇p (6)
∂t 2  ρ
Equation (6) is called Lamb’s hydrodynamical equation.
(ii) Cartesian Form
   ∂p ∂p ∂p 
Let q (u,
= = v, w), F (X, Y, Z) and ∇p = , ,  then equation (5) gives
 ∂x ∂y ∂z 
62 Partial Differential Equations

∂u ∂u ∂u ∂u 1 ∂p 
+u +v +w = X−
∂t ∂x ∂y ∂z ρ ∂x 
∂v ∂v ∂v ∂v 1 ∂p 
+u +v +w = Y−  (7)
∂t ∂x ∂y ∂z ρ ∂y 
∂w ∂w ∂w ∂w 1 ∂p 
+u +v +w =Z− 
∂t ∂x ∂y ∂z ρ ∂z 
Equation (7) are the required equations in Cartesian form.
(iii) Equation of Motion in Cylindrical Co-ordinates (r, θ, z)
Here,
   ∂p 1 ∂p ∂p 
q (q
= r , qθ , q z ), dr (dr, rdθ , dz) and ∇p = , , 
 ∂r r ∂θ ∂z 

Let F = (Fr , Fθ , Fz )

Also, the acceleration components in cylindrical co-ordinates are


dq  dq r qθ 2 dqθ q r qθ dq z 
=  − , + , 
dt  dt r dt r dt 

Thus, the equation of motion


dq  1
= F − ∇p , becomes
dt ρ

dq r qθ 2 1 ∂p 
− =Fr −
dt r ρ ∂r 
dqθ q r qθ 1 ∂p 
+ = Fθ −  (8)
dt r ρ r ∂θ 
dq z 1 ∂p 
= Fz − 
dt ρ ∂z 
(iv) Equation of Motion in spherical Co-ordinates (r , θ ,ψ )

Here,
   ∂p 1 ∂p 1 ∂p 
q (q
= r , qθ , qψ ), dr (dr, rdθ , r sin dψ ) and ∇p = , , 
 ∂r r ∂θ r sin θ ∂ψ 

Let F = (Fr , Fθ , Fψ )

Also, the acceleration components in spherical co-ordinates are

Pressure and Equation of Impulsive Action 63

dq  dq r qθ + qψ dqθ q z 2 cot θ q r qθ dq z qθ q z cot θ
2 2

=
 − , − + , + 
dt  dt r dt r r dt r 
Thus, the equation of motion

dq  1
= F − ∇p , becomes
dt ρ
2 2
du qθ + qψ 1 ∂p 
− Fr −
= 
dt r ρ ∂r 
dv q z 2 cot θ q r qθ 1 ∂p 
− + = Fθ −  (9)
dt r r ρ r ∂θ 
dw qθ q z cot θ 1 ∂p 
+ = Fψ − 
dt r ρ r sin θ ∂ψ 
Remark: The two equations, i.e. the equation of continuity and Euler’s equation of motion, comprise
the equation of motion of an ideal fluid. Thus, the equations

∂ρ  ∂ q    1
+ div( ρ q) =
0 and + (q.∇)q = F − ∇p are fundamental to any theoretical study of ideal fluid
∂t ∂t ρ
flow. These equations are solved subject to the appropriate boundary and initial conditions dictated by
the physical characteristics of the flow. With the density function known, the above equations are four
simultaneous differential equations for the three velocity components and pressure.
Example: A homogeneous liquid occupies a length 2l of a straight tube of uniform small bore and is
acted on by a body force which is such that the fluid is attracted to a fixed point of the tube, with a force
varying as the distance from that point. Discuss the motion and determine the pressure within the liquid.
Solution: We note that the uniform small bore of the tube allows us to ignore any variation of velocity
across any cross section of the tube and to suppose that the flow is unidirectional. Let u be the velocity
along the tube and p be the pressure at a general point at distance x from the centre of force O.

Also, let z be the distance of the centre of mass G of the fluid, from O,
The equations of motions of the fluid are
64 Partial Differential Equations

Equation of continuity
∂u
=0 ⇒ u = u (t) (1)
∂x
Euler’s equation
∂u ∂u 1 ∂p
+u X
=−
∂t ∂x ρ ∂x
(2)
∂u 1 ∂p
⇒ + 0 =− µ x −
∂t ρ ∂x

where − µ xi if the body force per unit mass, μ being a positive constant.

We observe that equation (2) can be written as

du 1 dp
−µ x −
= (3)
dt ρ dx
Integrating (3) w.r.t. x, we get

du x2 p
x =−µ − +c (4)
dt 2 ρ
where c is at the most a function of t only.
Let be the pressure at the free surfaces x=z - l and x=z + 1 of the liquid. Using these boundary

conditions, equation (4) gives

du (z − l) 2 Π
(z − l) −µ
= − +c
dt 2 ρ
du (z + l) 2 Π
(z + l) −µ
= − +c
dt 2 ρ
which on subtraction give
du
= −µ z (5)
dt
dz
But in the fluid motion, all fluid particle move with the same velocity u and u = , so equation (5)
dt
becomes

d2z
= −µ z (6)
dt 2
which represents a S.H.M. and its solution is
 Pressure and Equation of Impulsive Action 65

=z A cos ( mt + ε ) | (D 2 + m)z =⇒
0 D=± mi

where A and ε are constants which can be determined from initial conditions.
To find pressure, we use (5) in (3) to get
1 dp
= µ (z − x)
ρ dx
Integrating w.r.t. x, we find

p µ (z − x) 2
= +D
ρ −2
Using the boundary condition p=∏ at x=z-l,

Π µ l2
D
= +
ρ 2

p Π µ Π µ
(  ρ 2 (
z − x ) − l2  = −  z − x + l )( z − x − l ) 
2
∴ = −
ρ ρ 2

which gives the required pressure.

Example: A quantity of liquid of density ρ occupies a length 2a of a long straight tube of uniform small
cross-section and is under the action of a force kx per unit mass towards a fixed point O in the tube,
where x is the distance from O. Show that when the nearer free surface is at a distance z from O, the
 x z
pressure at distance x exceeds the atmospheric pressure Π by kρ ( x − z )  a − + 
 2 2
Solution: Similar to above (only pressure)

The Boundary conditions are p=∏ at x=z , x=z+2a

du
⇒ −k ( z + a )
= (5)
dt
Using (5) in (3), we have
66 Partial Differential Equations

dp −kρ
= ρ k ( z + a − x ) ⇒= (z + a − x) + D
2
p
dx 2
kρ a 2
p = Π at x = z ⇒ D = Π +
2
kρ a 2
kρ
(z + a − x)
2
∴p = Π + −
2 2
 x z
⇒ p=− Π kρ ( x − z )  a − + 
 2 2
Example: A straight tube ABC of small bore is bent so as to make the angle ABC at right angle and AB
equal to BC. The end C is closed and the tube is placed with the end A upwards and AB vertical and is
filled with liquid. If the end C be opened, prove that the pressure at any point of the vertical tube is
instantaneously diminished one-half. Also find the instantaneous change of pressure at any point of the
horizontal tube, the pressure at any point of the horizontal tube, the pressure of the atmosphere being
neglected.
Solution: Let AB=BC=a. When the liquid in AB has fallen through a distance z at time t, let P be any
point in the vertical column such that AM=z, BP=x, BM=a-z

∂u
If u and p be the velocity and pressure at P, then the equation of continuity is = 0 i.e. u=u(t).
∂x
∂u ∂u 1 ∂p
And the equation of motion is +u =−g −
∂t ∂x ρ ∂x
∂u 1 ∂p
i.e. =−g −
∂t ρ ∂x
Integrating w.r.t. x, we get
∂u p
x =−gx − + c (1)
∂t ρ
Using the boundary condition p=0 at x=a-z
Pressure and Equation of Impulsive Action 67
∂u
c =(a − z) + g(a − z)
∂t
∴ Equation (1) becomes
∂u p ∂u
x =−gx − + (a − z) + g(a − z)
∂t ρ ∂t

p  ∂u 
⇒ =−(x − a + z)  + g  (2)
ρ  ∂t 
Now, we take a point Q in BC, where BQ= x ' .
Let u ' , p ' be the velocity and pressure at Q, then

p'  ∂u ' 
⇒ −(x ' − a) 
=  |z=0 and g is not affecting. (3)
ρ  ∂t 

Equating the pressure at B when x=0 and x ' =0, we obtain from (2) and (3) that

 ∂u  ∂u ' ∂u
(a − z)  + g  =
a −a
=  u ' = −u
 ∂t  ∂t ∂t

 ∂u    ∂u 
Initially, when C is just opened, then z=0, t=0 and we have a   + g  = −a  
=  ∂t  t 0=
  ∂t  t 0

 ∂u  −g
⇒  = (4)
 ∂t 0 2

∴From equation (2), initially, the pressur p 0 is given by

p0  ∂u   g
=− ( x − a )   + g  =− (x − a)
ρ  ∂t 0  2

1
p0
⇒= ρ g(a − x) (5)
2
But when the end C was closed, the liquid was at rest and the hydrostatic pressure p1 at P was

p1 ρ=
= gh ρ g(a − x) (6)

|h=AP, pressure due to liquid above the point.

From (5) and (6), we obtain
1
p0 = p1
2
Thus the pressure is diminished to one-half.
68 Partial Differential Equations

Now, from equation (3), the initial pressure at Q is given by

p '0  ∂u '   ∂u  g
= − ( x' − a )  (x ' − a)   =−
 = (a x ') | u'
ρ  ∂t 0  ∂t 0 2
=u and using (4)
1
'
⇒ p= 0 ρ g(a − x ' )
2
But when the end C was closed, the hydrostatic pressure
p 2 at Q was ρ ga (Pressure at Q, B, C etc. is same)

∴ Instantaneous change of pressure at any point of the horizontal tube is

1 1
p 2 − p '0 = ρ ga − ρ g(a − x ' ) = ρ g(a + x ' ) .
2 2
Example: Homogeneous liquid is in motion in a vertical plane within a curved tube of uniform small
bore under the action of gravity. Find the period of oscillation.

B A
z

N h

β θ α
O
Solution: Let O be the lowest point of the tube, AB the equilibrium level of the liquid, h the height of
AB above O, α and β the inclinations of the tube to the horizontal at A and B respectively and θ its
inclinations at a distance s from O along the tube.
Let a, b denote the lengths of OA and OB, respectively and suppose that at time t, the liquid is displaced
a small distance z along the tube from the equilibrium position. The assumption of uniform small bore
means the flow is unidirectional along the tube. Let the velocity be u(s, t). The equation of continuity
∂u
gives = 0 , so the Euler’s equation of motion
∂s
∂u ∂u 1 ∂p
+u −g sin θ −
=
∂t ∂s ρ ∂s
becomes
Pressure and Equation of Impulsive Action 69

du 1 ∂p
−g sin θ −
=
dt ρ ∂s
du dy 1 ∂p
−g −
=
dt ds ρ ∂s

Integrating w.r.t. s, we get

du p
s =−gy − + c(t) (1)
dt ρ
The boundary conditions at the free surfaces are
p=∏ for y=h+z sinα, s = OM = a+z at M
p=∏ for y=h-z sinβ, s = ON =-( b-z) at N
Using these boundary conditions in (1), we get
du Π
(a + z) =−g(h + z sin α ) − + c(t)
dt ρ
du Π
−(b − z) =−g(h − z sin β ) − + c(t)
dt ρ
du
(a + b) −gz(sin α + sin β )
= (2)
dt

dz du d 2 z
Since u = ⇒ = , so equation (2) is
dt dt dt 2

d2z d2z
(a + b) =
− gz(sin α + sin β ) ⇒ −µ z
= (3)
dt 2 dt 2
g(sin α + sin β )
where µ =
a+b
Equation (3) represents a S.H.M and its period T of oscillation is given by
1
2π  a+b  2
T = 2π 
=  .
µ  g(sin α + sin β ) 
Remark: The above example is a model of a water clock.
Example: AB is a tube of small uniform bore forming a quadrantal arc of a circle of radius O, OA being
horizontal and OB vertical with B below O. The tube is full of liquid of density ρ, the end B being
du 2g
closed. If B is suddenly opened, show that initially = , where u=u(t) is the velocity and that the
dt π
 2θ 
pressure at a point whose angular distance from A is θ immediately drops to ρ ga  sinθ − above
 π 
 70 Partial Differential Equations

atmospheric pressure. Further prove that when the liquid remaining in the tube subtends an angle β at
the centre.

d 2 β −2g 2  β 
= sin   .
dt 2 aβ 2
Solution: In the fig, arc AP = s, arc AP` = s+δs at time t. The equation of motion of the element PP`
along the arc in the direction of s increasing is
du 1 ∂p
= g cos θ − (1)
dt ρ ∂s
dy dθ
= sin θ ,
y a= a cos θ
ds ds
But s= aθ
ds
⇒ =a
dθ
dy
∴ = cos θ
ds
 
dq ∂ q  
Also, from the relation = + (q.∇)q , we get g cos θ θ
dt ∂t
u
du ∂u ∂u ∂u ∂  1 2  
= +u = +  u  , q =us g
dt ∂t ∂s ∂t ∂s  2 
∴ Equation (1) becomes
∂u ∂  1 2  dy 1 ∂p
+  u  =− g (2)
∂t ∂s  2  ds ρ ∂s
Integrating both sides w.r.t. s, we get
∂u 1 2 p
s + u = gy − + f (t) (3)
∂t 2 ρ
∂u
Since, at any time t, is the same at all points of the liquid. Initially taking t=0 at A, we have s = 0, u
∂t
= 0, p = ∏, y = 0, so equation (3) gives
Π
0=
− + f (0) (4)
ρ
1
Also, when t=0 at B, we have s =π a, u =0, p =Π, y =a , so equation (3) becomes
2
Pressure and Equation of Impulsive Action 71

1  ∂u  Π
π a   = ga − + f (0) (5)
2  ∂t  t =0 ρ
Using (4) in (5), we get

 ∂u  2g
  = (6)
 ∂t  t =0 π

 du   ∂u ∂u  2g
Therefore,   =  +u  = |since u =0, when t =0
=  dt  t 0=
 ∂t ∂s  t 0 π

Hence, initially we have

du 2g
=
dt π
which proves the first result.
∂u 2g
Using=
s aθ ,=
y a sin θ , = ,= t 0 in equation (3), we obtained
u 0,=
∂t π
 2g  p
aθ  = + 0 ga sin θ − + f (0)
π  ρ
2gaθ p Π
⇒ = ga sin θ − + |using (4)
π ρ ρ

 2θ 
⇒ p − Π ρ ga  sin θ − 
=
 π 
which proves the second results.
Now, when the liquid inside the tube subtends an angle β at the center, then
π π 
θ = − β ,s =a  − β  , y =a cos β
2 2 
Equation (3) at the upper free surface of the liquid (p=∏), becomes

π  ∂u 1 2 Π
a −β  + = u ga cos β − + f (t) (7)
2  ∂t 2 ρ
1
At the lower end B, s = π a, y = a, p = Π , so equation (3) gives
2
π a ∂u1 Π
+ u 2 = ga − + f (t) (8)
2 ∂t 2 ρ
Subtracting (8) from (7), we obtain
72 Partial Differential Equations

∂u
−β a = ga cos β − ga
∂t
∂u β
⇒ βa =ga(1 − cos β ) =2ga sin 2
∂t 2
∂u 2g 2 β
⇒ =sin (9)
∂t β 2

Now, equation (9) takes the form

d2β 2g β
= − sin 2
dt 2
aβ 2
Hence the result.
Example: A tube AB of uniform fine bore is in the form of the arc of the cycloid s = 4a sinψ , the ends
π
A and B being given by ψ =0 and ψ = respectively. It is fixed in a vertical plane so that A is its
2
lowest point and the tangent there is horizontal. The tube, which is closed at A and opens at B, is full of
2a
uniform liquid. If the end A is opened, show that the tube empties in time π .
g

π
Solution: The cycloid is s = 4a sinψ , 0 ≤ ψ ≤ where (s, ψ ) are intrinsic co-ordinates.
2
The equation of motion in the direction of s increasing, is
du 1 ∂p
−g sinψ −
= (1)
dt ρ ∂s
s
But sinψ = and from the relation
4a
 
dq ∂ q  
= + (q.∇)q , we get
dt ∂t
du ∂u ∂u ∂u ∂  1 2  
= +u = +  u  , q =us
dt ∂t ∂s ∂t ∂s  2 
Therefore, equation (1) becomes
∂u ∂  1 2  gs 1 ∂p
+  u = − − (2)
∂t ∂s  2  4a ρ ∂s
Integrating both sides w.r.t. s, we get

∂u  1 2  gs 2 p
s + u  = − − + f (t) (3)
∂t  2  8a ρ
 Pressure and Equation of Impulsive Action 73
When the end A is opened, then at any subsequent instant, let the fluid level be at distance L from O,
where pressure is Π . At this free surface, equation (3) is

∂u  1 2  gL2 Π
L + u  = − − + f (t) (4)
∂t  2  8a ρ
Also, at the end A, s = 0(ψ = 0), p = Π , so equation (3) gives

1 2 Π
u =
− + f (t) (5)
2 ρ
Subtracting (5) from (4), we get

∂u −gL2 ∂u −gL
L = ⇒ =
∂t 8a ∂t 8a
ds 
But =
u = L , so we get
dt

 = gL  + λ 2 L = g
L − ⇒L 0, λ 2 =
8a 8a
The solution of this differential equation is
=L k1 cos λ t + k 2 sin λ t
=⇒ L λ (k cos λ t − k sin λ t)
2 1

π
Initially=t 0,= L 0 , so we get
L 4a sin= 4a,=
2
k 2 0,=
= k1 4a

Thus, L = 4a cos λ t (6)

When the tube becomes empty, then L=0 and equation (6) gives

π π π 2a
λt = ⇒t= = = π
2 2λ g g
2
8a
Hence the result.
Example: Air obeying Boyle’s law is in motion in a uniform tube of small cross section. Prove that if
ρ be the density and V be the velocity at a distance x from a fixed point at a time t, then

∂2 ρ ∂2
2  (
=  ρ V 2 + k ) ,
∂t 2
∂x 

where k is a constant.
Solution: At time t, let p be the pressure and V be the velocity at a distance x from the end of the tube.
74 Partial Differential Equations

Then, by Boyle’s law

p=kρ (1)

The equation of continuity becomes

∂ρ ∂
+ [ ρ V ] =0 (2)
∂t ∂x
and the equation of motion is
dV ∂V ∂V 1 ∂p
= +V =
− (3)
dt ∂t ∂x ρ ∂x
Differentiating equation (2) w.r.t. t , we get

∂ 2 ρ ∂  ∂ρ ∂V 
+  V+ρ 0
= (4)
∂t 2
∂x  ∂t ∂t 
From equations (1) and (3), we obtained
∂V ∂V k ∂ρ
+V =
− (5)
∂t ∂x ρ ∂x
Using equation (2) and (5) in equation (4), we have

∂ 2 ρ ∂  ∂ρ ∂V ∂ 
2
+  −k − ρ V − V ( ρ V ) =
0
∂t ∂x  ∂x ∂x ∂x 
| Eliminating derivatives w.r.t. t in 2nd term of (4)

∂ 2 ρ ∂  ∂ρ ∂ 
⇒ 2 − k + ( ρ V 2 ) =
0
∂t ∂x  ∂x ∂x 
∂2 ρ ∂  ∂ 2 
⇒ 2 −  (V + k) ρ  =
0
∂t ∂x  ∂x 
∂2 ρ ∂2 2
⇒ =
∂t 2 ∂x 2
(V + k) ρ
Hence the result.
2.4 Bernoulli’s Equation
2.4.1 For Steady Flow
Euler’s equation of motion may be thought of as a differential equation for the pressure p if the velocity

q is known. By integrating it, we shall obtain a special form of Euler’s dynamical equation in terms of
pressure, which is known as Bernoulli equation or pressure equation.
Pressure and Equation of Impulsive Action 75
We shall obtain a special form of Euler’s dynamical equation in terms of pressure. The Euler’s
dynamical equation is

dq  1
= F − ∇p (1)
dt ρ
  
where q is velocity, F is the body force, p and ρ are pressure and density, respectively. F be
conservative, so that it can be expressed in term of a body force potential function Ω as

F = −∇Ω (2)

∂q
When the flow is steady, then =0 (3)
∂t
Therefore, in case of steady motion with a conservative body force equation (1) on using (2) and (3),
gives

 1 2    1
∇  q  − q × ξ = −∇Ω − ∇p
2  ρ
 1 2  1  
⇒  q + Ω  + ∇p = q × ξ (4)
2  ρ

  
dq ∂ q  1 2    ∂q
| = + ∇  q  − q × ξ and = 0
dt ∂t 2  ∂t
Further, if we suppose that the liquid is barotropic, i.e., density is a function of pressure p only, then we
can write
1 dp
∇ (p) = ∇ ∫
ρ ρ
Using this in (4), we get

 1 2 dp   
⇒ ∇  q + Ω + ∫  = q ×ξ (5)
2 ρ

Multiplying (5) scalarly by q and noting that
     
q.(q × ξ )=(q × q ).ξ =0
  1 2 dp 
We get, ⇒ q.∇  q + Ω + ∫  = 0 (6)
2 ρ

If s is a unit vector along the streamline through general point of the fluid and s measure distance along

this streamline, then since s is parallel to q , therefore equation (6) gives
76 Partial Differential Equations

∂  1 2 dp 
⇒  q + Ω + ∫  =0
∂s  2 ρ
Hence along any particular streamline, we have
1 2 dp
q + Ω + ∫ =C (7)
2 ρ
where C is constant which takes different values for different streamlines. Equation (7) is known as
Bernoullli’s Equation which applies to steady flow of ideal barotropic fluid in which the body force are
conservative. This result applies to steady flow of ideal.
    
Remark: (i) If q × ξ = 0 i.e. if q and ξ are parallel, then streamlines and vortex lines coincide and q is
said to be Beltrami vector.

If ξ =0 the flow is irrotational.

For both of these flow patterns,

1 2 dp
q + Ω + ∫ =C, where C is same at all points of the fluid.
2 ρ
dp p
(ii) For homogenous incompressible fluids, ρ is constant and ∫ =
ρ ρ
Then Bernoulli’s Equation becomes
1 2 p
q + Ω + =C
2 ρ

So that if q is known, the pressure can be calculated.
  1 p
(iii) If body force is neglected, then F = 0 and the Bernoulii’s equation becomes q 2 + =constant
2 ρ
2.4.2 For Unsteady Irrotational Flow
Here also, we suppose that the body forces are conservative

i.e. F = −∇Ω
 
For irrotational flow, q = −∇φ ⇒ ∇ × q = 0

The equation of motion


∂q  1 2     1
∂t 2 
( )
+ ∇  q  − q × ∇ × q = F − ∇p
ρ
(1)

It becomes
Pressure and Equation of Impulsive Action 77

 ∂φ   1 2  1
−∇   + ∇  q  = −∇Ω − ∇p
 ∂t  2  ρ

 1 2 dp ∂φ 
⇒ ∇  q + Ω + ∫ −  =0 Barotopic fluid
2 ρ ∂t 
Integrating, we get

 1 2 dp ∂φ 
⇒  q + Ω + ∫ −  = f (t) (2)
2 ρ ∂t 
which is the required equation.
Bernoulli’s equation for unsteady and incompressible irrotational flow:
dp p
If the liquid is homogenous, then ∫ =and the equation (2) becomes
ρ ρ

 1 2 p ∂φ 
⇒  q + Ω + −  = f (t)
2 ρ ∂t 
Further, for steady case,
∂φ
= 0,=f (t) constant
∂t
1 2 dp
q +Ω+ = constant
2 ρ
Example: A long straight pipe of length L has a slowly tapering circular cross section. It is inclined so
that its axis make and angle α to the horizontal with its smaller cross-section downwards. The radius of
the pipe at its upper end is twice that of at its lower end and water is pumped at a steady rate through the
pipe to emerge at atmospheric pressure. Its pumping pressure is twice the atmospheric pressure, show
32  ∏
that the fluid leaves the pipe with a speed U given
= by U 2  gL sin α +  , where ∏ is atmospheric
15  ρ
pressure.
Solution: The assumption that the pipe is slowly tapering means that any variation in the velocity over
any cross-section can be ignored. Let the velocity at the wider end of the pipe be V and the emerging

velocity be U (velocity at the lower end). The only body force is that of gravity, so F = −gjˆ and
consequently Ω =gy
 ∂Ω ˆ ∂Ω ˆ ∂Ω ˆ
 F = −∇Ω ⇒ −qjˆ = −∇Ω = − i− j− k
∂x ∂y ∂z
∂Ω
⇒ −g = − ⇒ Ω = gy
∂y
78 Partial Differential Equations

p 1 2
Bernoulli’s equation, + q + Ω =C (1) [ For water ρ is const.
ρ 2

p 1 2
becomes + q + gy =
C
ρ 2
Applying this equation of the two ends of the pipe, we get

2Π 1 2 Π 1 2
+ V + gL sin α= + U (2) for lower end y=0
ρ 2 ρ 2
Let a and 2a be the radii of the lower and upper ends respectively, then by the principle of conservation
of mass
U
π(2a) 2 V =π(a) 2 U ⇒ V = (3)
4
From (2) and (3), we obtain

1  U2  1
Π + ρ  + gρL sin α= ρU 2
2  16  2
1  U2 
⇒ ρ  U2 −  = Π + gρL sin α
2  16 
15
⇒ρU 2 = Π + gρL sin α
32
32  Π
⇒ U2
=  gL sin α + 
15  ρ

Hence the result.

Example: A sphere is at rest in an infinite mass of homogeneous liquid of density ρ , the pressure at
infinity being Π . Show that, if the radius R of the sphere in any manner, the pressure at the surface of
the sphere at any time is

ρ  d 2 (R 2 )  dR 
2

Π+  +  
2  dt 2  dt  
Pressure and Equation of Impulsive Action 79
Solution: In the incompressible liquid, outside the sphere, the fluid velocity q will be radial and thus
will be a function of r, the radial distance from the centre of the sphere (the origin), and time t only. The
equation of continuity in spherical polar co-ordinates becomes
1 d 2
(r u) = 0 (1)
r 2 dr
 ∂ 
 q (u, 0,=
= 0), u u(r, t), ∇ ≡  , 0, 0 
 ∂r 
 1 ∂
∇.q =2 (r 2 u) i.e. spherical symmetry
r ∂r

r 2 u cons tan
⇒= = t f (t)

On the surface of the sphere, r=R, u= R

Therefore, f (t) = R 2 R

Thus, r 2 u = R 2 R (2)
We observe that u → 0 as n → ∞ , as required
  
From (1), it is clear that curl q = 0 and imply that the motion is irrotational and q = −∇φ

∂φ ∂φ f f
⇒u=− ⇒− = 2 (3) φ = from (2)
∂r ∂r r r
The pressure equation for irrotational non-steady fluid motion in the absence of the body forces is
p 1  2 ∂φ p 1 ∂φ
+ q − = C(t) i.e. + u 2 − C(t)
= (4)
ρ 2 ∂t ρ 2 ∂t
where C(t) is a function of time t.
f
As r → ∞, p → Π, u= → 0, φ → 0
r2
Π
So that C(t)
= ∀ t (5)
ρ
Therefore, from (2), (3), (4), and (5), we get
2
p Π ∂ f 1  R 2 R 
= + ( )−  2  (6)
ρ ρ ∂t r 2 r 

∂f d 2   2 + 2RR 2
But = (R =
R) RR
∂t dt
At the surface of the sphere, we have r=R and equation (6) gives
80 Partial Differential Equations

p Π 1  2 ) − 1 R 2
= + (2RR 2 + RR
ρ ρ R 2
(7)
p Π  − 1 R 2 = Π + 1 (3R 2 + 2RR)
⇒ = + 2R 2 + RR 
ρ ρ 2 ρ 2

d 2 (R 2 )  2 d  2 = ( 2RR
 + (R)  + 2R 2 ) + R 2 = 2RR
 + 3R 2
Now, + (R) = (2RR)
dt 2 dt
Therefore, from (7), we obtain

ρ  d 2 (R 2 )  dR 
2

p=Π+  +  
2  dt 2  dt  

Hence the result.

−7
Example: An infinite mass of ideal incompressible fluid is subjected to a force µr 3 per unit mass
directed towards the origin. If initially the fluid is at rest and there is a cavity in the form of the sphere
r=a in it, show that the cavity will be completely filled after an interval of time πa 5 3 (10µ) −1 2 .

Solution: The motion is entirely radial and consequently irrotational and the present case is the case of
spherical symmetry. The equation of continuity is
1 d 2
0 r 2u =
(r u) =⇒ constant = f(t) (1)
r 2 dr

=
On the surface of the sphere, r R,
= R v (say)

Therefore, r 2 r = f (t) = R 2 R ⇒ f (t) = R 2 R  = R 2 dv + 2Rv 2

 + R2RR

dt

f (t) dv dv dR dv
⇒ =2v 2 + R =2v 2 + R =2v 2 + Rv (2)
R dt dR dt dR
The Euler’s equation of motion, in radial direction, using r = u is
∂u ∂u 1 ∂p
+u Fr
=−
∂t ∂r ρ ∂r

∂u ∂  f (t)  f (t)
But =  = +, Fr = −µr −7 3
∂t ∂t  r 2  r 2
So, we need to integrate the Euler’s equation

f (t) ∂  1 2  −µ ∂  p 
+  u = −   (3)
r 2 ∂r  2  r 7 3 ∂r  ρ 
Pressure and Equation of Impulsive Action 81

Let us assume that the cavity has radius R at time t and its velocity then is R = v . Integrating (3) over
the whole liquid ( r = R to r = ∞ ) at time t, we obtain
∞ ∞ ∞
 −f (t) 
0
1 2  3µ  1   p 
  + u =  4  − 
 r R  2 v 4  r 3 R  ρ R

Since the fluid is at rest at infinity, u ∞ = 0 . Also=

p ∞ 0,=
p R 0 (cavity), thus we get

f (r) 1 2 3µ 1 dv 3µ 1
− v =− 4
⇒ 2Rv + 3v 2 =
− using (2)
R 2 4 R 3 dR 2 R 43

dv 3µ 2
To make it exact, we multiply by R 2 so that 2R 3 v + 3R 2 v 2 =
− R 3
dR 2

d(R 3 v 2 ) 3µ 2
⇒ − R 3
=
dR 2
9µ 5 3
Integrating, we get R 3 v 2= A − R (4)
10
9µ 5 3
When R= a, R ≡ v= 0 , which gives A = a
10
Now, we take v= R < 0 because as the cavity fills, R decreases with time. Thus (4) gives
12 12
dR  9µ   a5 3 − R5 3 
−  t =
= −∫  
dt  10   R3 
12 0 π2
 9µ  R 3 2 dR 6a 5 3 3πa 5 3
  t = −∫ 5 3 ∫
2
Therefore, = sin θdθ =
a (a − R5 3 )
12
 10  5 0
10

R 5 3 =a 5 3 sin 2 θ i.e. R =a ( sin θ )

65

πa 5 3 (10µ) −1 2
Thus, t =

Hence the result.

2.5 Impulsive Motion
When there is a sudden, but a finite change in the fluid velocity over a short interval t of time t, or a high
pressure on the boundary acting over time t, or the rapid fluctuation in the velocity of a rigid body
immersed in the fluid, impulsive motion occurs in the fluid. Impulsive behaviour is a term used to
describe these types of actions. The situation of impulsive action is effectively modeled mathematically
by letting the body force or pressure approach to infinity while δt→0 in such a way that the integral of
body force or pressure over the time interval δt remains finite in this limit. It the flow is incompressible,
82 Partial Differential Equations

infinitely rapid propagation of the effect of the impulsive action takes place, so that an impulsive
pressure is produced instantaneously throughout the fluid. Here, we consider only the incompressible

fluid with constant density ρ. The impulsive body force I and impulsive pressure P are defined as
 t +δ t 
I = Lt ∫
δ t →0 t
F dt

t + δt
P = Lt ∫
δt → 0 t
p dt (1)

where F is the body force per unit mass and p is the pressure.

We note that finite body forces such as gravity do not contribute to the impulsive body force I .
To determine the equation of impulsive motion, we consider the Euler’s equation
  
Dq dq ∂q    1
≡ ≡ + ( q. ∇ ) q =F − ∇p (2)
Dt dt ∂t ρ
Integrating equation (2) w.r.t. time t from t to t + δt and taking limit as δt→0, we get
 t +δ t 
t +δ t ∂q t +δ t  1 t +δ t
Lt ∫ dt + Lt ∫ (= q. ∇ ) dt Lt ∫ Fdt − Lt ∫ ∇ p dt
δ t →0 t ∂t δ t →0 t δ t →0 t ρ δ t →0 t (3)

Assuming that fluid is accelerated impulsively at t = 0 and since we expect a finite change in q as a
result of the impulse, we get from (2) and (3)
  1
q '− q = I − ∇ P (4)
ρ
 
where q and q′ denote respectively the fluid velocity before and after the impulsive action.
Thus, the equation of impulsive motion is
  
ρ ( q '− q=) ρ I − ∇P (5)

which holds at each point of the fluid.

In cartesian co-ordinates, (5) can be expressed as
∂P
ρ (u '− u ) = ρ X' =
∂x
∂P
ρ (v'− v) = ρ Y' −
∂y

∂P
ρ (w '− w ) = ρ Z' −
∂z
where
  
q = (u, v, w), q′ = (u', v', w'), I = (X', Y', Z')
Pressure and Equation of Impulsive Action 83
 
When there is no externally applied impulse, then I =0 and equation (5) becomes
 
– ∇P = ρ ( q '− q ) (6)
 
Further, if the motion is irrotational, then q = − ∇φ , q ' = − ∇ φ ′ , where φ and φ ' denote the velocity
potential just before and just after the impulsive action, then (6) becomes
P = ρ (φ' – φ) (7)
where we have ignored the constant of integration since an extra pressure, constant throughout the
liquid, would not affect the impulsive motion.
Corollary: If the fluid is at rest prior to the impulsive action, then the velocity q generated in the fluid
by the impulse is given by
  1    
q =I − ∇ P (8) | In (4), put q = 0 and q ' ≡ q
ρ
For this case, equation (6) can be put as

– ∇P = ρ q (9)

and equation (7) becomes

P= ρ φ (10)
Equations (7) and (10) give the relation between impulsive pressure P and the velocity potential φ.
Remark: From the above discussion, we observe that, likewise, an irrotational motion can be brought to
rest by applying an impulsive pressure -ρφ throughout the fluid.
Example: A sphere of radius a is surrounded by an infinite liquid of density ρ, the pressure at infinity
being Π. The sphere is suddenly annihilated. Show that the pressure at distance r from the center
a
immediately falls to π 1 −  . Show further that if the liquid is brought to rest by impinging on a
 r
a
concentric sphere of radius , the impulsive pressure sustained by the surface of the sphere is
2
7 ∏ ρa 2 / 6 .

Solution: Let v' be the velocity at a distance r' from the center of the sphere at any time t and p be the
pressure. The equation of continuity (case of spherical symmetry) is
1 d 2
(
r '2 dr '
)
r ' v' = 0 ⇒ r '2 v 2 = f(t) (1)

Equation of motion is
∂ v' ∂ v' − 1 ∂ p
+ v' = | No body force
∂t ∂ r' ρ ∂ r'
84 Partial Differential Equations

f (t ) ∂ v' 1 ∂p
or 2
+ v' =−
r' ∂ r' ρ ∂ r'

Integrating w.r.t. r', we get

f (t ) 1 2 p
+ v' = − + C
r' 2 ρ

Since r'→ ∞ ⇒ p = Π, v' = 0 so that C = Π/p.

− f (t ) 1 2 ∏ − P
Thus + v' = (2)
r' 2 ρ

When, sphere is suddenly annihilated i.e. r' = a, v' = 0, p = 0, then

f (t) ∏a
− = ∏ / ρ i.e. f(t) = − (3)
a ρ

The velocity v' vanishes just after annihilation, so from (2) and (3), we get
∏ a ∏ −p 2∏
= ⇒ =∏−p
ρ r' ρ r'

Thus, the pressure at the time of annihilation (r' = r) is

2∏ a
= ∏− p ⇒ p = ∏ 1 − 
r  r

which proves the first result.


Now, let P be the impulsive pressure at a distance r', then from the relation – ∇ P = ρ q ,

we get
dP
- = ρ v'
d r'

⇒dP = – ρv' dr'

From the equation of continuity, we have
r2 v = r'2 v' = f(t) (4)
So dP = – ρ v (r2/r’2) dr' (5)
where r is the radius of the inner surface and v is the velocity there. Integrating (5), we get
P = ρ v (r2/r') + C1 ,
when r' → ∞, P = 0 so that C1 = 0
Thus, P = ρ v(r2/r') (6)
Pressure and Equation of Impulsive Action 85
Equation (6) determines the impulsive pressure P at a distance r'. The velocity v at the inner surface of
the sphere (p = 0) is obtained from (2) as
f (t ) 1 2 ∏
− + v = (7)
r 2 ρ

From (4), f(t) =

d 2
dt
( )
r v = r2
dv
dt
dr
+ v. 2r = r 2
dt
dv dr
dr dt
+ 2 rv 2

f (t ) dv
⇒ =rv + 2 v2
r dr
Thus (7) becomes
dv 1 ∏
rv + 2 v 2 − v2 = −
dr 2 ρ

dv 3 2 − ∏
or rv + v =
dr 2 ρ

dv − 2∏ 2
⇒ 2 r3 v + 3v 2 r 2 = r |Multiplying by r2
dr ρ

⇒
(
d r3 v2
=−
)
2∏ 2
r
dr ρ

Integrating, we get
2∏ 3
r3 v2 = − r + C2
3ρ

2 ∏ a3
Since r = a, v = 0 so we find C2 =
3ρ

2∏ 3 3
Therefore, r3 v2 = (a – r )
3ρ

The velocity v at the surface of the sphere r = a/2, on which the liquid strikes, is

2 ∏ a 3 − (a / 2) 14 ∏
3
v2 = =
3 ρ (a / 2)3 3 ρ

From relation (6), using r = a/2, we get

ρ 14 ∏ a 2
P= . (8)
4 3 ρ r'

which determines the impulsive pressure at a distance r’ from the centre of the sphere.
Thus, the impulsive pressure at the surface of the sphere of radius a/2 is given by
86 Partial Differential Equations

ρ 14 ∏ a 2
P= = 7 ∏ρa2 / 6
4 3 ρ a/2

Hence the result.

Example: A mass of liquid surrounds a solid sphere of radius a, and its outer surface, which is a
concentric sphere of radius b, is subjected to a given constant pressure Π , no other surface being in
action on the liquid. The solid sphere suddenly shrinks into a concentric sphere. Determine the
subsequent motion, and the impulsive action on the sphere.
Solution: At time t, let p the pressure and v ' be the velocity at distance r ' from the centre of the sphere.
The equation of continuity (spherical symmetry) is
1 d
r'2 dr ' ( )
r '2 v ' =
0 ⇒ r '2 v ' =
f (t) (1)

Equation of motion is
∂v ' ∂v ' 1 ∂p
+ v' =
−
∂t ∂r ' ρ ∂r '
f ( t ) ∂  1 2  1 ∂p
⇒ +  v'  =
− ( 2)
r' 2
∂r '  2  ρ ∂r '
Let R, r be the radii of the external and internal boundaries at time, and V , v their velocities. These
quantities are functions of time t only and=V R,=  . Also p=0, p = Π on the internal and external
v r,
boundaries.
Integrating (2) w.r.t. r ' from r ' =r to r ' =R, we get
1 1  1 Π
− f ( t )  −  + ( v 2 − V 2 ) = ( 3)
r R 2 ρ
But v ' f (=
r '2 = t ) r=
2
v R 2V

dv dv
⇒ f ( t ) = 2rrv
 + r2 = 2rv 2 + r 2 v
dt dr
So, equation (3) becomes

 dv   1 1  v 2  V 2  Π
−  2rv 2 + r 2 v   −  + 1 − 2  =
 dr   r R  2  v  ρ

 dv   1 1  v 2  r4  Π
−  2rv 2 + r 2 v   −  + 1 − 4  =
 dr   r R  2  R  ρ

Multiplying both sides by 2r2 and observing that

Pressure and Equation of Impulsive Action 87

R 3 − r 3 = b 3 − a 3 = c3 ( say )
we obtain.

 dv  
1 1



1 r2 
 2Π r 2
−  4r 3 v 2 + 2r 4 v   − 3  + v2 r 4 
 − =
 dr   r ( r + c3 ) 
1/3
r
2
( r 3 + c3 ) 
1/3
ρ
  

   
2 4  1 r2  −2Π r
2
1 1  d
⇒
 −  ( v r
2 4
) − v r  −  = A
 r ( r 3 + c3 )  dr
1/3
 r 2 ( r 3 + c3 )1/3  ρ
  

Integrating, we get

 
1 1 2 4 −2Π 3
vr  − 1 
= r + c1
 ( r + c )  3ρ
r 3 3 3
 
2Π 3
Using v=0 when r=a, we find c1 = a
3ρ
∴Equation ( 4 ) is
 1 1  2Π 3 3
v2r 4  − = 
 r R  3ρ
(a − r )
 1 1  2Π  a − r 
3 3
i.e. v3  −  = 4 
 r R  3ρ  r 
which gives the expression for the velocity for subsequent motion.
Now, let P be the impulsive pressure at a distance r ' and r be the radius of the solid sphere, then from

the relation −∇P = ρq , we get

dP
− ρ v ' ⇒ dp =
= − ρ v ( r 2 / r '2 ) dr '
− ρ v 'dr ' = r '2 v ' =
r2v
dr '
Integrating, we get

=P ρ v ( r 2 / r ') + C2

As P = 0 when r ' = R, so C2 = − ρ vr 2 / R
1 1
Thus, P ρ vr 2  − 
=
 r' R 
This gives the impulsive pressure on the surface of the sphere. The whole impulse on the sphere
 88 Partial Differential Equations

4 ρ vr 3 ( R − r ) / R.
4 r 2 P =Π
=Π
and the whole momentum destroyed
R R
= ∫ 4Πr '2 ρ v 'dr ' = 4Π ρ r 2 v ( R − r ) .
4Πρ ∫ r 2 v dr ' =
r r

2.6 Kelvin’s Circulation Theorem

In addition to Bernoulli’s equation, other results of general nature follow from Euler’s equation. Now,
we establish Kelvin’s circulation theorem which prescribes conditions under which the circulation
around a closed circuit, moving with the fluid and consisting of the same fluid particles, is constant.
Statement: For an ideal fluid, the circulation around any closed circuit of fluid particles, which moves
with the fluid, remains constant provided that the body forces are conservative and the pressure is a
function of density only
Proof: We know that
 
Γ =∫ q.dr
C

 
where Γ is the circulation, q is the fluid velocity and r is the position vector of a fluid particle at any
time t. Time derivative of Γ following the motion of fluid is
dΓ d   d  
dt dt ∫C
= q ⋅ dr
= ∫ dt ( q ⋅ dr )
C

  
 dq   d    dq     d   dr  
= ∫  ⋅ dr + q ⋅ ( dr )=
 ∫C  dt ⋅ dr + q.dq  (1) ( dr=) d  = dq
C 
dt dt dt  dt 

Since the system of body forces is conservative, therefore F = −∇Ω , where Ω is a potential function.
Euler’s equation of motion is

dq  1 1
F
=− ∇p = −∇Ω − ∇p ( 2)
dt ρ ρ
Multiplying both sides of (2) scalarly by dr , we get
Pressure and Equation of Impulsive Action 89

 dq  1 
dr ⋅ = −dr ⋅∇Ω − dr ⋅∇p
dt ρ

dq  dp 
⇒ ⋅ dr = −dΩ − ( 3) dr ⋅∇ ≡ d
dt ρ
Using ( 3) in (1) , we get
dΓ  dp  
=
dt ∫ −dΩ − ρ
C
+ q ⋅ dq 

 1  dp 
∫ d  2 q
2
= − Ω − 
C  ρ
1  dp
= ∫ d  2 q
2
− Ω  − ∫ ( 4)
C  C ρ
The first integral is (4) in zero. If p is a function of ρ only, then the second integral in (4) also vanishes.
Hence, we get
dΓ
= 0 ⇒ Γ = constant for all times.
dt
Corollary: (1) In a closed circuit C of fluid particles moving under the same conditions as in the
theorem,
   
∫ curlq.ds = ∫ ⋅ ds = Constant
ξ ________ ( 5 )
s s

where S is any open surface whose rim is C. To establish (5), we note that, by Stoke’s Theorem
  
∫ curlq.ds =∫ q ⋅ dr = Γ = constant
s S

This shows that the product of the cross-section and angular velocity at any point on a vortex filament is
constant all along the vortex filament and for all times.
Corollary: (2) Under the conditions of the theorem, vortex lines move with the fluid.
Proof: Let C be any closed curve drawn on the surface of a vortex tube but not embracing it. Let S be
the portion of the vortex tube rimmed by C. By definition, vortex lines lie on S. Thus
   
=0 ∫ curlq.ds
= ∫ q ⋅ dr circulation is zero on the surface
0 C

Let C be a material curve and S be a material surface, then

d 
= ∫
dt S
( 
)
n.curlq dS ∫=
dt
S
(d 
n.curlq ) ds 0
90 Partial Differential Equations

Thus n ⋅ curlq remains zero, so that S remains a surface composed of vortex lines. Consequently, vortex
lines are tubes move with the fluid, i.e., vortex filaments are composed of the same fluid particles. This
explains why smoke rings maintain their forms for long periods of time.
Corollary: (3) Under the conditions of the theorem, if the flow is irrotational in a material region of the
fluid at some particular time, say t = 0, the flow is always irrotational in that material region thereafter,
i.e. if the motion of an ideal fluid is once irrotational it remains irrotational for afterwards provided the
external forces are conservative and the liquid is barotropic.
 
Proof: Suppose that at some instant the fluid on the material surface S is irrotational, then ri = 0 for all
points of S. Let C be the boundary of S, then

 
Γ ∫ q ⋅ dr
=
C S
( 
)
= ∫ n ⋅ curlq dS= ∫ n ⋅ ξ dS
S
( )
= 0

But by Kelvin’s circulation theorem, Γ is constant for all times. Hence Γ is zero for all subsequent

∫(
 ξ dS = 0
times. At any later time, n.
S
)
  
If we don't take S to be non-zero infinitesimal element, say ∆S then n  ⋅ ξ∆S = 0 ⇒ ξ = 0 at all
points of S for all times and hence the motion is irrotational permanently. This proves the permanency of
irrotational motion.
Remarks: (1) The above three corollaries are properties of vortex filaments.
(2) The Kelvin’s theorem is true whether the motion be rotational or irrotational. In case of irrotational
 
motion ξ = 0 and thus Γ =0
(3) From the results of the theorem, we conclude that vortex filaments must either form closed curves or
have their ends on the boundary surface of the fluid. a vortex in an ideal fluid is therefore permanent.
2.7 Vorticity Equation
Euler’s equation of motion for an ideal fluid under the action of a conservative body force with potential
Ω per unit mass is
 
dq ∂q 1    1
= + ∇  q 2  − q × ξ = −∇Ω − ∇p (1)
dt dt 2  ρ
  
where the vorticity is ξ =curlq=∇ × q
If the fluid has constant density, then on taking curl of equation (1), we get

∂q 1     1 
∂t 2 
( 
)
∇ × + ∇ × ∇  q 2  − ∇ × q × ξ = ∇ ×  −∇Ω − ∇p 
ρ 
Pressure and Equation of Impulsive Action 91

∂q  
⇒ ∇ × − ∇ × q ×ξ = 0
∂t
( )
∂   
⇒ (∇ × q ) − ∇ × q × ξ = 0
∂t
( )
∂     
⇒
∂t
( )
ξ = ∇ × q ×ξ = ( ) (ξ .∇ ) q − ( q ⋅∇ ) ξ

∂ξ  
⇒
∂t
+ ( q ⋅∇ ) ξ= ξ ⋅∇ ( ) q

dξ  
⇒ = ξ ⋅∇ q
∂t
( ) ( 2)
which is the required vortices equation Equation (2) is called Helmholtz’s vortices equation.

For two-dimensional motion, the vorticity vector ξ is perpendicular to the velocity vector q and the

R.H.S. of (2) is identically zero (ξ ⋅∇ ≡ 0). Thus, for two-dimensional motion of an ideal fluid, vorticity
is constant.
In the case when the body force is non-conservative, equation (2) becomes

dξ    
dt
( )
= ξ ⋅∇ q + curlF , where F is body force per unit mass.

In the case when the body force is non-conservative equation (2) becomes

dξ
dt
( )
= ξ .∇ q + curl F

Example: A motion of inviscid incompressible fluid of uniform density is symmetrical about axis r=0
where ( r, θ , z ) are cylindrical polar co-ordinates. The cylindrical polar resolute of velocity are
q r ( r, z ) , 0, q z (r, z)  .

Show that if a fluid particle has vorticity of magnitude ξ0 when r = r0 , its vorticity when at general

distance r from the axis of symmetry has magnitude ξ =  ξ0 r  r if any body forces acting are
 0 
conservative.

Solution: The vorticity vector ξ satisfies the vorticity equation

dξ 
dt
(
= ξ ⋅∇ q )
 92 Partial Differential Equations

r̂ r θ z
  1 ∂ ∂ ∂
Now,
= r curlq
=
r ∂r ∂θ ∂z
q r ( r, z ) 0 q z ( r, z )

1  ∂ ∂  
=  r ( 0 ) + r θ  q r ( r, z ) − q z ( r, z )  + z ( 0 ) 
r  ∂z ∂r  
 ∂qr ∂qz 
θ 
= −  ( 2)
 ∂z ∂r 
Therefore, we have
  ∂q ∂q   ∂ 1 ∂  ∂ 
ξ ⋅∇ = θˆ  r − z  r + θ +z 
 ∂z ∂r   ∂r r ∂θ ∂z 
1  ∂q ∂q  ∂ θ ⋅ r = θ ⋅ z
=  r− z
r  ∂z ∂r  ∂θ θ ⋅ θ =1

Thus, ξ =(  1  ∂q r ∂q z  ∂
⋅∇ q )
r  ∂z
− 
∂r  ∂θ
q r r + q z z ( )
∂ z
=0
q  ∂q ∂q  ∂ r ∂θ
= r r− z
r  ∂z ∂r  ∂θ ∂ r 
=θ
∂θ
q r  ∂q r ∂q z 
=  − θ
r  ∂z ∂r 
  q 
( )
Hence ξ ⋅∇ q = r ξ
r
( 3) | using ( 2 )
∴ From equation (1) and ( 3) , we get

dξ q r 
= ξ ( 4)
dt r

Now, q r = q ⋅ r, so equation ( 4 ) becomes

dξ 
( )

r = q ⋅ r ξ _______ ( 5 )
dt

  dr dr
Since r 2 = r 2 ⇒ r ⋅ = r
dt dt
  
r dr dr dr dr  dr
⇒ = ⇒ r ⋅ = ⇒ r ⋅ q =
r dt dt dt dt dt
Pressure and Equation of Impulsive Action 93

So, equation ( 5 ) takes the form

 
dξ dr  dξ  dr
r = ξ ⇒r −ξ =0
dt dt dt dt

dξ  dr 
r =ξ  
dt dt = d ξ
⇒ 2
0⇒   = 0
r dt  r 
 
ξ ξ ξ
=
⇒ cons tan t ⇒ = constant= 0
r r r0

ξ 
⇒ξ  0 r Hence proved .
 r0 
2.8 Energy Equation for Incompressible Flow
Let us consider the motion of an incompressible fluid under the action of conservative body force with

potential Ω per unit mass, so that F = −∇Ω .
The principal of energy states that the change in energy is equal to work done by the external forces. The
∂Ω
potential due to external forces is supposed to be independent of time, so that =0
∂t
d ∂ 
Also, ≡ + q.∇
dt ∂t
dΩ ∂Ω  
⇒ ≡ + (q.∇)Ω= (q.∇)Ω (1)
dt ∂t
Now, Euler’s equation of motion is

dq  1 1
⇒ F
=− ∇p = −∇Ω − ∇p (2)
dt ρ ρ

Multiplying (2) scalarly by q , we get

 dq  
ρ q = − ρ q.(∇Ω) − q.(∇p)
dt
1 d 2  
⇒ ρ ( q ) + ρ q.(∇Ω) = −q.(∇p)
2 dt
1 d 2 dΩ 
⇒ ρ (q ) + ρ =−q.(∇p) |using (1)
2 dt dt
1 d 2 
⇒ ρ ( q + Ω ) = −q.(∇p) (3)
2 dt
94 Partial Differential Equations

Integrating (3) over a region τ , we get

1 d 2 
∫τ 2 dt
ρ ( q + Ω ) dτ = − ∫τ ∇p)dτ
q.(

d  1 2  
⇒  ∫ ρ q dτ + ∫ ρΩdτ  =
− ∫ q.(∇p)dτ (4)
dt  τ 2 τ  τ

1
Now, ∫τ 2 ρ q dτ= T, ∫ ρΩdτ= V
2

where T and V represent the kinetic and potential energy of the liquid.
Also, using Gauss divergence theorem,
  
∫ (q.
τ
∇ p)dτ =∫ div(pq)dτ
τ
| ∇.q =0

  

= ∫=
S
(pq).dS ∫ (pq).ndS
S

where S is the surface enclosing volume τ .

Hence equation (4) becomes
d   
− ∫ (pq).dS =
(T + V) = 
− ∫ (pq).ndS (5)
dt S S

Equation (5) expresses that the rate of change of total energy of any region of the fluid as it moves is
equal to the rate of work done by the pressure over the boundary surface.

Cor. If the fluid moves tangentially over the surface S, then T+V = constant ( q.n =0). On the other hand
if T+V = constant, then using Gauss divergence theorem and the equation of continuity, we obtain
 
∫τ
(q.∇p)dτ =0. But the choice of τ is arbitrary, so it follows that at all points of the flow region, q is

orthogonal to p, i.e. streamlines are orthogonal to isobars (surface of constant pressure).

2.9 Kinetic Energy (K.E.) of Irrotational Flow
Definition: Acyclic and Cyclic Irrotational Motion
An irrotational motion is called acyclic if the velocity potential φ is a single valued function i.e. when at
every field point, a unique velocity potential exists, otherwise the irrotational motion is said to be cyclic.
Clearly, only acyclic irrotational motion is possible in a simply connected region.
For a possible fluid motion, even if φ is multivalued at a particular point, the velocity at that point must
be single-valued. Hence if we obtain two different values of φ , these values can only differ by a
constant. At present, we restrict ourself to acyclic irrotational motion for which we prove a number of
results related to φ .
Pressure and Equation of Impulsive Action 95
ρ ∂φ
2 ∫ ∂x
We shall prove that K.E. is given by T = φ dS , where φ is the velocity potential. We know that if
S

τ be the finite region occupied by the fluid, then the K.E. is given by
1 2 1  1 
2 ∫τ 2 ∫τ 2 ∫τ
T
= ρ q d
=τ ρ (q.q)d
=τ ρ (∇ φ .∇ φ )dτ | q = −∇φ

If the fluid density is constant, then

ρ
T=
2 ∫τ (∇φ.∇φ )dτ (1)

Now, div(φ∇φ ) =∇.φ∇φ =∇φ .∇φ + φ (∇.∇φ ) =∇φ .∇φ + φ∇ 2φ

=∇φ .∇φ ∇ 2φ =
0 (2)

From (1) and (2), we get

ρ ρ 
2 ∫τ
T= div(φ∇φ )dτ = ∫ (φ∇φ ).ndS |Guass divergence
2 S

theorem
ρ  ρ ∂φ
2∫
= φ (∇φ .n)dS = ∫φ dS
S
2 ∂n S

where S = S0 + S1 + S2 + ... + Sn denotes the sum of the outer surface S0 and the inner surfaces
ρ ∂φ
S1 ,S2 ,...,Sn and n is unit normal to S drawn out of the fluid on each boundary. Also, T = − ∫ φ dS ,
2 S ∂n

where n is unit normal to S drawn inside the fluid on each boundary.

2.10 Kelvin’s Minimum Energy Theorem
The kinetic energy of irrotational motion of a liquid occupying a finite simply connected region is less
than that of any other motion of the liquid which is consistent with the same normal velocity of the
boundary.

Proof: Let T be the K.E., q be the fluid velocity and φ be the velocity potential of the given irrotational
motion. Let τ be the region occupied by the fluid and S be the surface of this region, then
ρ 2 ρ ρ ∂φ
( −∇φ ) =
2

2 ∫τ 2 ∫τ 2 ∫ ∂n
T
= dτ
q= dτ φ dS (1)
S


Let T1 be the K.E. and q1 be the velocity of any other motion of the fluid consistent with the same
normal velocity of the boundary S (or consistent with the same kinematic boundary condition).
For both the motions, the continuity equation is satisfied, i.e.
 96 Partial Differential Equations
 
∇.q =0=∇.q1 (2)

The boundaries have the same normal velocity, i.e.

   
q.n = q1.n ⇒ (q1 − q).n = 0 (3)

Now, let us consider

ρ ρ       2
T1=
−T
2τ ∫ ( q1 − q =
2 2
) dτ ∫ 2q(q
2 τ 
1 − q) + q1 −q dτ ( ) 
ρ    ρ   2
= ∫
2τ
2q(q1 − q)dτ +
2 ∫τ ( q1 − q dτ )
  ρ  
∫ ( q − q ) dτ
2
− ρ ∫ ∇φ .(q1 − q)dτ +
= 1 (4)
τ 2 τ

From vector calculus, we have

     
=∇. φ (q1 − q)  =∇φ .(q1 − q) + φ∇.(q1 − q)
  |using (2)
=∇φ .(q1 − q)
Therefore, equation (4) becomes
  ρ   2
T1 − T = − ρ ∫ ∇. φ (q1 − q)  dτ + ∫ q1 − q dτ
2τ
( )
τ
   
 + ρ q − q 2 dτ
= − ∫ φ (q1 − q).ndS
2τ∫ ( 1 ) |using (3)
S

ρ   2
= (
2 ∫τ
)
q1 − q dτ > 0

⇒ T1 > T

Hence the result.

2.11 Kinetic Energy of Infinite Liquid
2.11.1 Theorem: An infinite liquid is in irrotational motion which is at rest at infinity and is bounded
internally by solid surface (s) S. Show that the K.E. of the moving fluid is
ρ ∂φ
2 ∫ ∂n
T= φ dS
S

where S = S1 + S2 + … SN denotes the sum of the inner boundaries S1, S2, …, SN and n normal to S
drawn out of the fluid on each boundary.
Proof: Let Σ be a large surface enclosing the surface (s) S and τ be the region bounded by S
internally and by Σ externally.
 Pressure and Equation of Impulsive Action 97

Using the result of K.E. for finite liquids, we find that the K.E. T* for finite region τ is given by
ρ ∂φ ρ ∂φ
=T* ∫ φ dS + ∫ φ dS
2 S ∂n 2 Σ ∂n
(1)


Now, divq = ∇ 2φ =0 through out τ and the Gauss divergence theorem gives
 

∫ divqdτ =
τ
0 ⇒ ∫ n.qdS =0
S∪Σ

 ∇φ dS =⇒ ∂φ
⇒ ∫
S∪Σ
n. 0 ∫
S∪Σ
∂n
dS =0 S1 S2

∂φ ∂φ S3
⇒∫ dS + ∫ dS =
0 (2)
S
∂n Σ
∂n SN

Since the surface S is solid, therefore no flow takes place across it, hence
∂φ
∫ ∂n dS = 0
S
|the flux across S is zero (3)

From equation (2), we get

∂φ
∫ ∂n dS = 0
Σ
(4)

For the surface Σ , as Σ goes to infinity, the liquid is at rest

 
⇒ q = 0 ⇒ ∇φ = 0 ⇒ φ =constant = c (5)

Hence as Σ goes to infinity, the K.E. of the liquid is

ρ ∂φ ρ ∂φ
2 ∫ ∂n
T* → T
= φ dS + c ∫ dS |using (5)
S
2 ∂n Σ

ρ ∂φ
2 ∫ ∂n
= φ dS |using (4)
S

Hence the result.

Remark: We note that the K.E. for finite and infinite liquid has the same expression, only difference
being in the structure of S.
2.11.2 Theorem: Show that acyclic irrotational motion is impossible in a finite volume of fluid bounded
by rigid surfaces at rest or in infinite fluid at rest at infinity and bounded internally by rigid bodies at
rest.
Proof: If possible suppose that acyclic irrotational motion is possible and let φ be the velocity
potential. Then, K.E. of the fluid is
98 Partial Differential Equations

ρ 2 ρ ∂φ
T = ∫ (∇φ ) dτ = ∫ φ dτ (1)
2τ 2 τ ∂n

where S is the sum of all the rigid boundaries when τ is finite or the sum of internal rigid
boundaries when τ is infinite.
Now, since the boundaries are rigid, then at every point of S, the normal velocity is zero liquid is
at rest.
∂φ
=0 at each point of S. (2)
∂n
From (1) and (2), we get
2 2  
τ
∫ q dτ = 0 ⇒ q = 0 ⇒ q = 0 at each point of τ
So, there is no motion of fluid. Hence acyclic irrotational motion is impossible.
Corollary: If the solid boundaries in motion are instantaneously brought to rest, show that the motion of
the fluid will instantaneously case to be irrotational.
Proof: If possible, assume that the motion remains irrotational, then the K.E. is given by (1). When the
 
surface S (Solid boundary) is brought to rest instantaneously, then q = 0 at each point of S.

⇒ φ is constant at each point of S

∂φ 
⇒ = 0 at each point of S ⇒ q =0 in τ implies there is no motion. Thus the motion is no longer
∂n
irrotational.

Remark: When the boundaries are brought to rest, the ensuring motion is rotational with curlq ≠ 0 .

2.12 Uniqueness Theorems

Now we prove that the velocity potential for acyclic irrotational motion is unique, subject to certain
restrictions.
2.12.1 Theorem 1: If the region occupied by the fluid is finite, then only one irrotational motion of the
fluid exists when the boundaries have prescribed velocities.
OR
Show that there cannot be two different forms of acyclic irrotational motion of a given liquid whose
boundaries have prescribed velocities.
Proof. If possible, let φ1 and φ2 be two different velocity potentials representing two motions, then
which shows that the motions are the same.

∇ 2φ1 ==
0 ∇ 2φ2 (1)
Pressure and Equation of Impulsive Action 99
Since the kinetic conditions at the boundaries are satisfied by both flows, therefore at each point of S,
∂φ1 ∂φ2
= (2)
∂n ∂n
Let φ= φ1 − φ2

∂φ ∂φ1 ∂φ2
∇ 2φ = ∇ 2φ1 − ∇ 2φ2 = 0 at each point of fluid and = − = 0 at each point of S.
∂n ∂n ∂n
So, φ represents a possible irrotational motion.

Also, the K.E. of the fluid is given by

ρ  ρ ∂φ ∂φ
=∫
2τ
qdτ = ∫ φ
2 S ∂n
0 |
∂n
=0

 
⇒q=0 at each point of fluid.
⇒ ∇φ =0 at each point of fluid.

⇒ ∇φ1 − ∇φ2 = 0 ⇒ ∇φ1 = ∇φ2 which show that the motions are the same.

Moreover φ is unique apart from an additive constant which gives rise to no velocity and thus can be
taken as zero.
2.12.2 Theorem II. If the region occupied by the fluid is infinite and fluid is at rest at infinity, prove
that only one irrotational motion is possible when internalboundaries have prescribed velocities.
Proof. If possible, let there be two irrotational motions given by two different velocity potentials φ1 and
φ2 The conditions on boundaries are

∂φ1 ∂φ2
= (1)
∂n ∂n
 
and q1= 0= q 2 at infinity (2)

Let us write φ= φ1 − φ2 (3)

⇒ ∇ 2φ = ∇ 2φ1 − ∇ 2φ2 = 0 − 0 = 0

⇒ motion given by φ is also irrotational.

∂φ ∂φ1 ∂φ2
Also, = − =0 |using (1)
∂n ∂n ∂n
 
⇒ q.n = 0 ⇒ q = 0 on the surface.
  
Further, q = −∇φ = −∇φ1 + ∇φ2 = q1 − q 2 = 0 at infinity |using (2)
100 Partial Differential Equations
 
Therefore, q = 0 everywhere on the surface and also at infinity. Hence, we get φ = constant

⇒ φ1 − φ2 =constant (4)

Without loss of generality, we can take the constant on R.H.S. of (4) to be zero (it gives no motion) and
thus we get φ1 = φ2

Remark: The above two uniqueness theorems are useful in finding solutions of ∇ 2φ =
0 subject to
prescribed boundary conditions, as the value of φ , obtained anyway is unique.

2.13 Check Your Progress

1. Define Conservative filed.
2. What is Euler’s dynamical equations of motion and derive it for cylindrical co-ordinates.
3. What is the difference between kinetic energy for a finite and infinite liquid.
4. Can there be two velocity potential for acyclic motion. Justify it.
−3
5. An infinite mass of incompressible fluid is acted upon by a force mr 2 per unit mass directed to the
origin. If initially the fluid is at rest and there is a cavity in the form of the sphere r = a in it, show that

( )
1 5
the cavity will be filled up completely after an interval of time 2 m
2
a 4
.
5
6. Show that if the velocity field

B( x 2 − y 2 ) 2 Bxy
u ( x, y )
= 2 2 2
, v ( x, y ) = , w( x, y ) 0 satisfies the equations of motion for inviscid
(x + y ) ( x + y 2 )2
2

incompressible flow, then determine the pressure associated with this velocity field, B being a constant.
2.14 Summary
In this section, we have studied about Pressure at a point, Euler’s dynamical equation, Bernoullis’s
equation for steady and unsteady state, Impulsive action, Vorticity equation, Kinetic energy theorem for
irrotational flow and theorems related to energy equation.
2.15 Keyword
Pressure, Force, Impulsive Action, Potential Flow, Conservative Body Force, Circulation and Kinetic
Energy.
2.16 Self-Assessment Test
dp 1 2
1. Prove that for steady motion of an inviscid isotropic
= flow p f ( p ), ∫ρ +=
2
q + Ω constant over a

surface containing the streamlines and vortex lines.

2. Derive the most general form of Bernoulli’s equation.
Pressure and Equation of Impulsive Action 101
3. A perfect incompressible fluid is moving steadily around the outside of a fixed cylinder of radius a
a
and vertical axis oz being . Show that the motion is irrotational. If the surface of the fluid is open to
r
the atmosphere, and the origin O is chosen so that on the free surface z = 0 when r = a, prove, by means
a2
of Bernoull’s equation or otherwise, that the equation of the free surface is 2gz = 1 − .
r2
4. Stream is rushing from a boiler through a conical pipe, the diameters of the ends of which are D and
d. If V and v be the corresponding velocities of the stream and if the motion be supposed to be that of
divergence from the vertex of the cone, prove that

v D 2 (v2 − V2 )/2K
= e
V d2
where K is the pressure divided by the density, and supposed constant.
5. A stream in a horizontal pipe, after passing a contraction in the pipe at which its sectional area is A, is
delivered at atmospheric pressure at a place, where the sectional area is B. Show that if a side tube is
connected with the pipe at the former place, water will be sucked up through it into the pipe from a
s2  1 1 
reservoir at a depth  2 − 2  below the pipe, s being the delivery per second.
2g  A B 
6. Liquid is contained between two parallel planes, the free surface is a circular cylinder of radius ‘a’
whose axis is perpendicular to the planes. All the liquid within a concentric circular cylinder of radius b
is suddenly annihilated. Prove that if Π be the pressure at the outer surface, the initial pressure at any
point of the liquid, distance r from the centre is Π (log r − log b) / (log a − log b) .

7. An infinite mass of homogeneous fluid is at rest subject to a uniform pressure Π , and contains a
spherical cavity of radius a, field with gas at a pressure m Π . Prove that if the inertia of the gas be
neglected, and Boyle’s law be supposed to hold throughout the ensuing, the radius of the sphere will
oscillate between the values a and na, where n is determined by the equation 1 + 3m log n − n 3 =
0 . If m
be nearly equal to 1, show that the time of oscillation will be 2π a 2ρ / 3Π , ρ being the density of the
fluid.
2.17 Suggested Reading
1. O’Neill, M.E. and Chorlton, F., Ideal and Incompressible Fluid Dynamics, Ellis Horwood Limited,
1986.
2. R.K. Rathy, An Introduction to Fluid Dynamics, Oxford and IBH Publishing Company, New
Delhi, 1976.
 UNIT - III
AXIALLY SYMMETRIC FLOW AND THREE-DIMENSIONAL SOURCES,
SINKS AND DOUBLETS

Content
3.1 Objective
3.2 Introduction
3.3 Axially Symmetric Flows
3.4 Sphere at Rest in a Uniform Stream
3.5 Motion of a Sphere through a Liquid at Rest at Infinity
3.6 Equation of Motion of a Sphere
3.7 Kinetic Energy Generated by Impulsive Motion
3.8 Motion of Two Concentric Spheres
3.9 Three-dimensional Sources, Sinks And Doublets
3.10 Hydrodynamical Images for Three-Dimensional Flow
3.11 Check Your Progress
3.12 Summary of the Unit
3.13 Keyword
3.14 Self Assessment Test
3.15 Suggested Reading
3.1 Objective
In this unit reader will learn about Axially symmetric flow,
Uniform flow, Sphere at rest, Motion of sphere through liquid at
rest, Kinetic energy, Motion of two concentric circles, Three-
dimensional source, Sink and doublet and their velocity potential
and images of three-dimensional flows.
3.2 Introduction
We propose to investigate irrotational motion in three
dimensions, with a focus on the motion of a sphere. We shall
analyse the motion of a sphere at rest in a uniform flow and
through liquid at rest at infinity, in particular.
Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 103
3.3 Axially Symmetric Flows
Let (r,θ,ψ) be the spherical coordinates w.r.t. an origin O. A potential flow that is axially symmetric
about the axis θ = 0, p (i.e., z-axis is taken as the axis of symmetry) has the property that at any point P,
all the scalar and vector quantities associated with the flow are independent of azimuthal angle ψ such
∂
that = 0 , where (r,θ,ψ) are spherical polar coordinates.
∂ψ

The equation of continuity div q = 0 for steady flow of an incompressible fluid becomes

1 ∂ 2 1 ∂
2
(r q r ) + (sin θq θ ) =
0 (1)
r ∂r r sin θ ∂θ

For irrotational motion q = −∇φ , where φ velocity potential and thus

∂φ 1 ∂φ
qr − ,q =
−
∂r r ∂θ
From equation (1), we have
1 ∂  2 ∂φ  1 ∂  ∂φ 
2
= r  2  sin
= θ  0 (2)
r ∂r  ∂r  r sin θ ∂θ  ∂θ 
Let a solution of (2) in separable variables r, θ has the form
φ = −R(r)Θ(θ) (3)
Using (3) in (2), we get
∂  2 ∂  1 ∂  ∂ 
 r (RΘ)  + [ sin θ (RΘ)  =0
∂r  ∂r  sin θ ∂θ  ∂θ 
∂  ∂R  R ∂  ∂Θ 
→ Θ  r2 +  sin θ =0
∂r  ∂r  sin θ ∂θ  ∂θ 
1 d  2 dR  1 d  dΘ 
→ r =−  sin θ  (4)
R dr  dr  Θ sin θ dθ  dθ 
The L.H.S. of (4) is a function of r only while the R.H.S. is a function of θ only. The equation can
therefore be satisfied if and only if either side is a constant, say n(n+1) and thus we get
1 d  2 dR 
r =  n(n + 1) (5)
R dr  dr 
and
d  dΘ 
 sin θ  + n(n + 1)Θ sin θ =0 (6)
dθ  dθ 
104 Partial Differential Equations

To solve (5), we put

dR
R =r m ⇒ =mr m −1
dr
1 d 2 m −1
Thus (5) ⇒
r m dr
( r mr ) = n(n + 1)

d
⇒ m ( r 2 mr m −1 ) =
r m n(n + 1)
dr
Therefore, solution of (5) can be written as
= A n r n + Bn r − (n +1)
R(r) (7)
Therefore, equation (6) becomes.

d  dΘ 
− sin θ  sin θ ( − sin θ ) + n(n + 1)Θ sin θ =
0
dµ  dµ 

d  2 dΘ 
⇒  sin θ  + n(n + 1)Θ =0
dµ  dµ 
d  dΘ 
⇒ 
dµ 
(1 − cos 2 θ )
dµ 
+ n(n + 1)Θ =0

d  dΘ 
⇒ 
dµ 
(1 − µ2 )
dµ 
+ n(n + 1)Θ = 0 (8)

Equation (8) is a Legendre’s Equation and possesses a solution known as Legendre Function of the first
kind Pn( µ )

Therefore,
Θ(θ=
) Pn (µ=
) Pn (cos θ)

Hence the general solution of (3) is of the form

φ(r, θ) =−  A n r n + Bn r − (n +1)  Pn ( cos θ ) (9)

The most general solution is the sum of all such solutions and thus
∞
φ(r, θ) =−∑  A n r n + Bn r − (n +1)  Pn ( cos θ ) (10)
n =0

3.1. Uniform Flow: Consider the flow which corresponds to a potential given by (9) with

An = 0 (n =
Uδ1n , Bn = 0,1, 2,.....)
where U is a constant and δij is Kronecker delta ( δij = 1, δij = 0 for i ≠ j )
Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 105

Since P1 ( cos θ )= cos θ, equation (9) becomes

φ(r, θ) = − Ur cos θ = − Uz | z = r cos θ

Thus
∂φ 
q = −∇φ = − k = U k
dz
which is a uniform streaming motion of the fluid with speed U along the direction of z-axis or the axis
θ =0 .
3.4 Sphere at Rest in a Uniform Stream
Consider an impermeable solid sphere of radius ‘a’ at rest with its center at the pole of a system of
spherical polar co-ordinates (r, θ, ψ). The sphere is immersed in an infinite homogeneous liquid with
constant density ρ, which in the absence of the sphere, would be flowing as a uniform stream with speed
U along the direction θ = 0.
The presence of the sphere will produce a local perturbation of the uniform streaming motion such that
the disturbance diminishes with increasing distance r from center of sphere. We say that the perturbation
of the uniform stream tends to zero as r→∞.
In this problem z-axis is the axis of symmetry. The undisturbed velocity of incompressible fluid is
Uk i.e. q = Ukˆ and the velocity potential φ0 due to such a uniform flow would be

φ0 =− Uz =− Ur cos θ

When the sphere is inserted, the undisturbed potential −Ur cosθ of uniform stream has to be modified by
“perturbation potential” due to the presence of the sphere. This must have the following properties.
(i) It must satisfy Laplace equation for the case of axial symmetry
(ii) It must tend to zero at large distances from the sphere

∂φ1 ∂φ
⇒ = + U cos θ (r = a, a ≤ θ ≤ π) = 0 at r = a i.e.normal
∂r ∂r
velocity at the sphere is zero
and
| ∇φ1 | → 0 as r → ∞.
Hence a suitable form of function φ1 (in view of (9)) is

φ1 =− Br −2 cos θ

which, on using the above boundary condition gives

2B 1 3
3
cos
= θ U cos θ ⇒
= B Ua
a 2
106 Partial Differential Equations

Hence, the solution φ is given by

Ua 3
φ(r, θ) = − Ur cos θ − cos θ
2r 2

 a3 
=− U  r + 2  cos θ, r ≥ a (2)
 2r 

Now, the uniqueness theorem II infer that the velocity potential in (2) is unique.
The velocity components at P(r, θ, ψ ), r ≥ a , are

∂φ  a3 
qr =
− U 1 − 3
=  cos θ
∂r  r 
1 ∂φ  a3 
qθ =
− = − U  1 + 3  sin θ ( 3)
1 ∂θ  2r 
1 ∂φ
qψ =
− = 0.
r sin θ ∂ψ
Varoius terms related to motion are obtained as follows:
(i) Stagnation Points: Stagnation points are those points in the flow where the velocity vanishes
i.e., q = 0 . Thus, these points are obtained by solving the equations

 a3  
U 1 − 3  cos θ = 0 
 r  

and  (4)

 a3 
U  1 + 3  sin θ =0 
 2r  

which are satisfied only by r= a,sin θ= 0,i.e., r= a, θ= 0, π . Thus, the stagnation points are

( r= a, θ= 0 ) and ( r= a, θ= π ) on the sphere. These are referred to respectively as the rear and forward
stagnation points.
(ii) Streamlines: The equations of streamlines

dr rdθ r sin θψ
= =
qr qθ qψ

for the present case, become

Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 107

dr rdθ r sin θdψ

= = ,r ≥ a
 a3   a3  0
U 1 − 3  cos θ − U 1 + 3  sin θ
 r   2r 
⇒ dψ = 0 ⇒ψ = constant.
and
1  2r 3 + a 3 
⇒  3 3  dr =−2 cot θdθ
r  r −a 
 2r + a 3 r −2 
⇒  3 3  dr =−2 cot θdθ
 r −a 
Integrating, we get
log(r 2 − a 3 r −1 ) = −2 log sin θ + log C
 r3 − a 3  2
⇒ log   = − log sin θ + log C
 r 
Cr
⇒ sin 2 θ = 3 3 , where C ≥ 0
r −a
For each value of C, the above equation gives a streamline in the plane ψ = constant. The choice of C =
0 corresponds to the sphere and the axis of symmetry.
(iii) Pressure at any Point: The pressure at any point of the fluid is obtained by applying Bernoulli’s
equation along the streamline through that point, taking the pressure at ∞ to be of constant value p ∞ .
Thus, in the absence of body force, the Bernoulli’s equation for homogeneous steady flow is
p 1
+ ( −∇φ ) = C1
2

ρ 2
 we get
at infinity, p = p and − ∇φ = Uk,
∞

p∞ 1 2
C
=1 + U
ρ 2
Thus
1 1
p p ∞ + ρU 2 − ρ ( ∇φ )
2
=
2 2

1  2  a3 
2 2

1 2 a3 
p p ∞ + ρU − ρ  U 1 − 3  cos θ + U 1 + 3  sin 2 θ 
⇒= 2 2

2 2   r   2r  

| ∇φ = − q

 2

1 2  a3   a3  2
⇒ 2
p p ∞ − ρU 1 − 3  cos θ + 1 + 3  sin θ − 1
= ( 5)
2  r   2r  
108 Partial Differential Equations

which gives the pressure at any point of the fluid. Of particular interest is the distribution of pressure on
the boundary of the sphere. It is obtained by putting r = a in (5) and thus

 2

1 2  a3  2
p p ∞ − ρU 1 + 3  sin θ − 1
=
2  2r  
1 9  1 2
= p ∞ − ρU 2  sin 2 θ − 1= 2
 p ∞ + ρU (4 − 9 sin θ)
2 4  8
1
= p ∞ + ρU 2 ( 9 cos 2 θ − 5 )
8
The maximum pressure occurs at the stagnation ponts, where θ = 0 or π. Thus
1
p max = p ∞ + ρU 2
2

which is called stagnation pressure.

π
The minimum pressure occurs along the equatorial circle of the sphere where θ = .
2
Therefore,
5
p min. = p ∞ − ρU 2
8
A fluid is presumed to be incapable of sustaining a negative pressure, thus

8p ∞
p min. =0 ⇒ U =
5ρ
At this stage the fluid will tend to break away from the surface of the sphere and cavitation is said to
occur. i.e., a vacuum is formed.
(iv) Thrust on the Hemisphere: Now, we find the thrust
(force) on the hemisphere on which the liquid impinges, r = a,
0 ≤ θ ≤ π/2. Let δS be a small element at P0 (a, θ, ψ) of the
hemisphere bounded by circles at r = a and at angular
distances θ and θ + δθ from axis of symmetry (i.e., z-axis).
The component of thrust on δS is pcosθ δS. Hence the total
thrust on the hemisphere is a long Z′O and is given by
 Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 109

dS =( 2πa sin θ )( adθ )

=F ∫
hemisphere
p cos θ dS

π /2
= ∫ p cos θ ( 2πa sin θ )( adθ )
0
π /2
 1 
∫ ( 2πa ) sin θ cos θ p + ρU 2 ( 9 cos 2 θ − 5 )  dθ
2
= ∞
0
8 
( using value of p at boundary )
 1 
πa 2  p ∞ − ρU 2 
=
 16 
3.5 Motion of a Sphere through a Liquid at Rest at Infinity

Let a solid sphere of radius ‘a’ centered at O be moving with uniform velocity − Uk in incompressible
fluid of infinite extent, which is at rest at infinity. Z-axis is the axis of symmetry and k is a unit vector
in this direction. (As the sphere is moving with velocity − Uk ⇒ the relative velocity of fluid if the
sphere be considered to be at rest is Uk .)
The boundary value problem for φ is now to solve

∇ 2 φ =0 (1)
−∂φ
such that − U cos θ, ( r =
= a) (2)
∂r
and
| ∇φ |→ 0, (r → ∞) ( 3)
The present case is also a problem with axial symmetry about the axis θ = 0, π,
So, φ = φ(r, θ)

Also, since P1 ( cos θ )= cos θ and the boundary condition (2) implies that the dependence of φ on θ must
be like cosθ, therefore φ has the form
110 Partial Differential Equations

 B  B
−  Ar + 2  P1 ( cos θ ) =
φ= −  Ar + 2  cos θ
 r   r 
1
However, to satisfy (3), it is necessary that A = 0, and then from (2), we get B = Ua 3
2
Thus, the solution for φ is

− Ua 3
φ ( r, θ )
= cos θ (4)
2r 2
From here, the velocity components are obtained to be

∂φ − Ua 3 −1 ∂φ − Ua 3
qr =
− = 3 cos θ, q θ = = 3 sin θ, q ψ =
0,
∂r r r ∂θ 2r
where (r, θ, ψ) are spherical polar co-ordinates. The various terms of particular importance related to
this motion are obtained as follows.
Streamlines: The differential equations for streamlines are
dr rdθ r sin θψ
= =
qr q θ qψ
dr rdθ r sin θdψ
i.e. = 3
= 3
− Ua Ua 0
3
cos θ − 3 sin θ
2r 2r
⇒ dψ = 0 = ψ = Constant.
and
dr
= 2 cot θ dθ ⇒ log
= r 2 log sin θ + log C
r
⇒ r = Csin 2 θ

K.E. of the Liquid: Let S be the surface of sphere and ρ be the density of liquid, then K.E. is given by
ρ ∂φ
2 ∫s ∂n
T
=1 φ dS (5)

where n is the outwards unit normal. But for the sphere n is along radius vector

 ∂φ   ∂φ 
so,  φ  =  −φ 
 ∂n s  ∂r r =a
1 
=  Ua cos θ  ( U cos θ )
2 
1 2
= U a cos 2 θ
2
Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 111
Therefore,
ρ 1 2 ρaU 2 π 0≤θ≤π
cos 2 θ ( 2πa sin θ )( adθ )
2 ∫S 2 ∫
2
=T1 U a=cos θdS
4 0 0 ≤ ψ ≤ 2π
πρa 3 U 2 π
=
2 ∫0
cos 2 θ sin θ dθ

π
πρa 3 U 2  − cos3 θ  1 2 2 4 3  U 
2
=  3  = πρa U =  πρ a  
2  0 3 3  4 
1
= M1U 2
4
4
where M1= πρa 3 is the mass of the liquid displaced by the sphere.
3
Also, K.E. of the sphere moving with speed U is given by
1
T2 = MU 2 (7)
2
4
where M= πσa 3 is the mass of the sphere, σ being the density of the material of the sphere.
3
Therefore, from (6) and (7), total K.E. T is given by
1 M' 2
T = T1 + T2 =  M + U (8)
2 2 

M'
The quantity M+ is called the virtual mass of the sphere.
2
Accelerating Sphere Moving in a Fluid at Rest at Infinity: The solution derived above for φ is
applicable when the sphere translates unsteadily along a straight line. In the present case, we take U =
U(t) and get the velocity potential as

− U(t)a 3
φ = φ ( r, θ, t ) = cos θ (1)
2r 2
The instantaneous values of velocity components and K.E. at time t are given by

−U ( t ) a 3 −U ( t ) a 3
=qr = cos θ, q θ = sin θ, qψ 0 | similar to steady case
r3 2r 3
1 1  2
and =T  M + M ' U ( t ) (2)
2 2 
The pressure at any point of the fluid is obtained by using Bernoulli’s equation for unsteady flow of a
homogeneous liquid, in the absence of body force, as
112 Partial Differential Equations

p 1 2 ∂φ
+ U − f (t)
= (3)
ρ 2 ∂t
where f(t) is a function of time t only.
p∞
Let p ∞ be the pressure at infinity where the fluid is at rest, then f ( t ) = and thus
ρ
p p ∞ 1 2 ∂φ
= − U + (4)
ρ ρ 2 ∂t
∂φ
To find we proceed as follows:
∂t
Since U = − Uk =− U ( t ) k is the velocity of the sphere, the velocity potential given in (1) can be
expressed in the form

1 a (U ⋅ r )
3

φ= (5)
2 r3
Since r is the position vector of a fixed-point P of the fluid relative to the moving centre O of the
sphere, it follows that
∂
U
= (− r ) (6)
∂t
∂r ∂r
Also, since r 2 =r ⋅ r ⇒ r =r ⋅ =− r ⋅ U |using (6)
∂t ∂t

( ) ( )
= ( − r ) ⋅ − Uk = rU r ⋅ k | r = r r
= r U cos θ

∂r
⇒ = U cos θ (7)
∂t
Differentiating (5) w.r.t. time t and using (6) & (7), we get
∂φ 1 3  U 2 cos θ ∂U 3U 2 
= a − 3 − 2 + 3 cos 2 θ 
∂t 2  r r ∂t r 
a U
3  cos θ U 3U cos θ 
2 2 2
∂U
= −  + 3 −  =
U
2  r2 r r3 
 ∂t
Aslo,
2 U 2a 6 U 2a 6
U= q 2r + q θ=
6
2
cos 2
θ + 6
sin 2 θ
r 4r
2 6
Ua  2 1 2 
= 6  cos θ + sin θ 
r  4 
Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 113
Thus, the pressure at any point of the fluid can be obtained from equation (4). In particular, at a point on
the sphere r = a, we have
∂φ −1 
=  Ua cos θ + U 2 − 3U 2 cos 2 θ 
∂t 2
U2
and =U
4
( 4 cos 2 θ + sin 2 θ )

and the corresponding pressure is given by

p p 1  1
= ∞ − Ua cos θ + U 2 ( 9 cos 2 − 5 ) (8)
ρ ρ 2 8

The force (thrust) acting on the sphere is given by

π
F = ∫ p cos θ ( 2πa sin θ )( adθ ) k
0

π 1 . 1 
= 2πa 2 k ∫  p ∞ − ρ U a cos θ + ρU 2 ( 9 cos 2 θ − 5 )  cos θ sin θ dθ
0
 2 8 
2  = 1  4 πa 3ρ  Uk
  1 M ' Uk

= πρa 3 Uk   =
3 23  2

4 3
where M ' = πa ρ is mass of the liquid displaced by the sphere. This shows that the force acts on an
3
accelerating sphere in the direction opposing the sphere’s motion.
3.6 Equation of Motion of the Sphere
Let R be the external force per unit mass in the direction of motion of the sphere. Let us use the result
that the rate of doing work is equal to the rate of increase in K.E.
Thus,

dT 1 d  M' 2   M '  dU
RU = =  M +  U ( t ) =
M + U
dt 2 dt  2    2  dt

dU 1 dU
⇒ M R − M'
= (9)
dt 2 dt
If the liquid is not there, then M′ = 0 and the equation of motion of the sphere is
dU
M =R (10)
dt
Comparing equation (9) & (10), we note that the presence of the liquid offers a resistance of the amount
1 dU
M' to the motion of the sphere. Let R′ be the external force per unit mass on the sphere when
2 dt
there is no liquid, then
114 Partial Differential Equations

MR = external force on the sphere in the presence of the liquid.

= ( M − M ') R '
MR '− M ' R ' =
4πσa 3 4πρa 3
Since, M= , M' =
3 3
M − M' ' σ + ρ '
=R = R R (11)
M σ
From equations (9) & (11), we find
dU  σ − ρ  1 dU
M
=   R '− M '
dt  σ  2 dt
 M '  dU  σ − ρ   M −M'
or M +  =   R ='  R '
 2  dt  σ   M 

   
dU  M − M '   σ−ρ 
∴ M = R ' =
 R' (12)
dt  M + M '  1
σ+ 
 2   2
This is the required equation of motion of the sphere in a liquid at rest at infinity. From equations (10) &
(12), we note that the effect of the presence of the liquid reduces the external force in the ration
ρ σ
σ − ρ : σ + . If s = is the specific gravity of the sphere compressed with the liquid, then the above
2 ρ
1
ratio is expressed as s − 1: s + .
2
Remark: We have already studied the impulsive actions in Unit-II, where, we had derived the relation
between the impulsive pressure P and the velocity potential φ as P = ρφ. Here, we derive the expression
for K.E. generated due to impulsive action.
3.7 Kinetic Energy Generated by Impulsive Motion
Let us consider incompressible fluid, initially at rest, which is set in motion by the application of
impulse I1 , I 2 ,.......I m to rigid boundaries S1 ,S2 ,.....,Sm respectively. The fluid may be of finite or
infinite extent. We know that the K.E. of the irrotational motion generated in the fluid is given by
ρ ∂φ
T
= ∫ φ dS
2 S ∂n
(1)

where S = S1 ,S2 ,.....,Sm , n is outwards unit normal on each Si . Let the velocity given to Si be U i (i = 1,
2,…, m), then on Si , we have

∂φ 
− −n.U
= i (2) | q = −∇φ
∂n
Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 115
Using (2) in (1), we have

ρ m
− ∑ U i ∫ nφdS
T= ( 3)
2 i =1 Si

But the impulsive force exerted by the fluid on Si is R i , where

 = ρ∫ nφdS
Ri = ∫
Si
nPdS
Si
| P = ρφ (4)

Thus from (3) and (4), we get

1 m
− ∑ U1⋅ ⋅ R i
T= (5)
2 i =1
which is the required expression for K.E. generated by impulsive motion.
Problem of Initial Motion (Motion of Two Concentric Spheres): Here we consider two concentric
spheres of radii a and b (a < b), the intervening space being filled with liquid of density ρ . The motion
is started by impulses and is consequently irrotational such that the equation of continuity is ∇ 2 φ =0 , φ
being velocity potential. The motion is discussed at the instant of start, as long as the spheres are
concentric.
Example: Incompressible liquid of constant density ρ is contained within a region bounded by two
concentric rigid spherical surfaces of radii a, b (a < b). The fluid is initially at rest. If the inner boundary
is suddenly given a velocity Uk , show that the outer surface experiences the impulsive force

2πρUa 3 b3 
k
b3 − a 3
Also calculate the corresponding K.E. generated by the impulsive motion.
Solution: The motion generated in the fluid is irrotational so that q = −∇φ ⇒ ∇ 2 φ = 0 is the equation of
continuity. The boundary conditions which φ must satisfy, are
∂φ
− = u cos θ (=
r a) (1)
∂r
and
∂φ
− = 0 (r = b) ( 2)
∂r
with (r, θ, ψ) spherical polar co-ordinates and with θ = 0

along the direction of k̂ . The form of boundary conditions suggests a solution of ∇ 2 φ =0 ,of the form
 116 Partial Differential Equations

φ = − ( Ar + Br 2 ) cos θ (3)

which satisfy (1) and (2) if

2B 2B
A− 3
= U, A − 3 = 0
a b
3
− Ua − Ua 3 b3
⇒=A = , B
b3 − a 3 2 ( b3 − a 3 )

Thus, the solution of the problem is

Ua 3  b3 
=φ r +  cos θ (4)
b3 − a 3  2r 3 

Impulsive force acting on the outer boundary in the direction of k̂ is

=F ∫ (P) r =b
cos θdSk
Sb

where,
ρUa 3  b3 
( P ) = ( ρφ ) =  b +  cos θ
=r b=r b
b3 − a 3  2b 2 
3 ρUa 3 b cos θ
=
2 b3 − a 3
and for the outer sphere r = b
2π ( b sin θ )( bdθ ) , 0 ≤ θ ≤ π
dS =

3 ρa 3 b
∫0 2 b3 − a 3 cos θ ( 2πb sin θ ) dθk
π
2 2
Thus,impulsive
= force, F

3πρUa 3 b3 k π 2
b3 − a 3 ∫0
= cos θ sin θ dθ

2πρUa 3 b3 
= k
b3 − a 3
Hence the result.
Now, if U1 , U 2 denote the velocity of spheres of radii a & b respectively, and R1 , R 2 be the
corresponding impulsive forces exerted by the fluid, then
Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 117

 2πρUa 3 b3 
U1= Uk, U 2= 0, R 2= F= k
b3 − a 3
1
∴ K.E., T = − ∑ U i .R i
2
1 1 1
=−  U1 ⋅ R1 + U 2 ⋅ R 2  =
− U1 ⋅ R1 = − Uk ⋅ R1 ( 5)
2 2 2
Also,
ρ∫ nφdS ⇒ R1 =
R1 = ρ∫ nφdS
S1 S1

⇒ R1 ⋅ k =ρ∫ n ⋅ k ( φ )r =a dS
Sa

 Ua 3  b3  
= −ρ∫ cos θ  3 3  a + 2  cos θ  2π ( a sin θ )( a dθ )
b −a  2a 
Sa


ρUa 3 2a 3 + b3 π
∫
2
=− 3 3 2
.2 πa cos 2 θ sin θ dθ
b −a 2a 0

−2 πρUa ( 2a + b )
3 3 3

=
3 b3 − a 3
Thus, form equation (5), we get

1 πρU a ( 2a + b )
2 3 3 3

T=
3 b3 − a 3
Deduction: If we let b→∞, then it becomes the case of a sphere of radius ‘a’ moving in an infinite liquid
at rest at infinity and we get

 2a 3 
πρU 2 a 3 1 + 3 
1  b  1
T= Lt 3
= πρU 2 a 3
b →∞ 3 a 3
1− 3
b
14 3 1
=  πρa = U2 M ' U2
43  4

4
where M =' πρa 3 is the mass of liquid displaced by the sphere by the sphere r = a.
3
3.8 Motion of Two Concentric Spheres
The space between two spheres is filled with incompressible fluid. The spheres have radii a, b (a < b)
and move with constant speeds U, V respectively along the line of centres. Show that at the instant when
the spheres are concentric, the velocity potential is given by
118 Partial Differential Equations

 3 1 3 3 −2 
( a U − b V ) r + 2 ( U − V ) a b r  cos θ
3

φ=
b3 − a 3
Also determine the impulse which is required to produce the velocity U to the inner sphere, when outer
sphere is at rest.
Solution: Let ρ be the density of the liquid.
2
We are to solve ∇ φ =0 under the boundary conditions
∂φ
− = cos θ, r = a (1)
∂r
∂φ
and − = V cos θ= ,r b (2)
∂r
where U & V are taken in the same direction.
The solution of the Laplace equation is of the form
φ = − ( Ar + Br −2 ) cos θ
∂φ  2B 
⇒ − =  A − 3  cos θ
∂r  r 
and thus, the boundary conditions give
2B 2B
A− = U, A − 3 = V
a3 b
Solving for A & B, we find
1 (U − V) a b
3 3
a 3 U − b3 V
=B = , A
2 a 3 − b3 a 3 − b3
Thus, the velocity potential for this motion is

( a U − b3V ) r + 12 ( U − V ) a 3b3r −2  cos θ

 3
φ= (3)
b3 − a 3
Hence the result
Impulse: When outer sphere is at rest, then V = 0 and from equation (3), we get
Ua 3  b3 
=φ  r +  cos θ ( 4)
b3 − a 3  2r 2 
4
Let M= πa 3σ be the mass of inner sphere
3
4
and M ' = πa 3ρ is the mass of liquid displaced by the inner sphere.
3
If I be the impulse, then by the principle of linear momentum, we have
Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 119
=I MU + Totalimpulsive Pr essure
i.e. MU + ∫ ( P )r =a cos θdS
I=
Sa

π Ua 3  b3  2
= MU + ρ∫  a +  cos θ 2π ( a sin θ )( a dθ )
0 b3 − a 3  2a 2 
| P = ρφ

πρUa 3 ( 2a 3 + b3 ) π
i.e. I MU +
= 3
b −a 3 ∫0
cos 2 θ sin θ dθ

2 πρa U ( 2a + b )
3 3 3

= MU +
3 b3 − a 3
1 M ' U ( 2a + b )
3 3

= MU +
2 b3 − a 3
Deduction: If b →∞, then it will be the case of a solid sphere moving in an infinite liquid and

M'  M'
I=MU + U=M + U
2  2 
Remark: The problem in which we solve the Laplace equation ∇ 2 φ =0 when the normal derivative of
∂φ
φ i.e., is given on the boundary, then such type of problem is called a Neumann problem, whereas
∂n
the solution of ∇ 2 φ =0 when the value of φ is given on the boundary, is termed as Dirichlet problem.

3.9 Three-dimensional Sources, Sinks and Doublets Source

An outward symmetrical radial flow of fluid in all directions is termed as a
three-dimensional source or a point source or a simple source. Thus, a
source is a point at which fluid is continuously created and distributed e.g.
an expanding bubble of gas pushing away the surrounding fluid. If the
volume of fluid per unit time which is emitted from a simple source at O is
constant and equal to 4πm, then m is termed as strength of the source.
Sink: A negative source is called a sink. At such points, the fluid is
constantly moving radically inwards from all directions. Thus, a simple sink
of strength m is a simple source of strength −m.
3.9.1 Velocity Potential due to a Simple Source of Strength m: Let there be a source of strength m at
a point O. With O as the centre, we draw a sphere of radius r around O. The flow across the sphere per

∫
unit volume is given by q.ndS
S

In case of a source there is only the radial velocity i.e., q has only radial component q r
120 Partial Differential Equations

Therefore, the flow is

= ∫=
q dS | q.n
r qr
S

= q r ( 4πr 2 ) .

But the volumetric rate of emission of fluid from O is 4πm per unit
time.
Thus, we get
4πm= q r ( rπr 2 )
m ∂ m
⇒ qr = 2
=
−   (1)
r ∂r  r 

It is observed that curl q = 0 (except at r = 0), therefore for irrotational flow,

∂φ
qr = − | q = −∇φ ( 2)
∂r
From (1) & (2), we find
m
φ= , ignoring the constant of integration
r
which is the required expression for the velocity potential for a source.
Remarks: (a) For a simple sink of stre ngth m, the velocity
m
potential is φ = − .
r
(b) A source or sink implies the creation or annihilation of
liquid at a point. Both are points at which the velocity potential
(and stream function for two-dimensional case) become infinite
and therefore, they require special analysis.
∂φ m m
− rˆ =
(c) Velocity at a point due to a source is q = 2
rˆ = r
∂r r r3
.
3.9.2 A simple Source in Uniform Stream: Let us consider a simple source of strength m at O in a
 k be the unit vector along z-axis which is taken as the
uniform stream having undisturbed velocity Uk,
axis of symmetry of the flow. We find the velocity potential at any point P (z, θ, ψ). From P, draw ⊥ on
OZ. Let OP = r, POZ = θ ; OM = z

We observe that the velocity potential of the uniform stream in the absence of source is
Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 121

∂φ
q = −∇φ ⇒ Uk = − k
∂z
φ1 =− Uz =− Ur cos θ (1)
∂φ
⇒ = − U ⇒ φ = − Uz
∂z
and the velocity potential of the simple source is
m
φ2 = (2)
r
Thus, the total velocity potential of the combination is
m
φ = φ1 + φ2 = − Ur cos θ + (3)
r
The velocity components at P ( r, θ, ψ ) are

∂φ m
qr = − = U cos θ + 2
∂r r r ≥ 0, 0 ≤ θ ≤ π
1 ∂φ
qθ =
− = − U sin θ ∂
r ∂θ 0 ≤ ψ ≤ 2π, ≡0
∂ψ
1 ∂φ
qψ =
− = 0
r sin θ ∂ψ

The stagnation points are given by q = 0, i.e.,

m
U cos θ + = 0,sin θ = 0 ⇒ θ = 0 or π
r2

m
But θ = 0 gives r to be imaginary ⇒ θ = π and r =
U

 m 
Thus, there is only one stagnation point  , π, 0  .
 U
 
3.9.3 Doublet (Dipole): The combination of a source and a sink of equal strength, at a small distance
apart, is called a doublet.
To Find the Velocity Potential of Doublet: Suppose that
there is a simple source of strength m at O1 and a simple
sink of strength m at O 2 . Origin O is taken as the mid-point
of O1O 2 . It is also assumed that there is no other source or
sink. Let P be a fixed point within the fluid and
122 Partial Differential Equations

OP = r , O1P = r1 , O 2 P = r2 | POO1 = θ,
OO1 =
h, OO 2 =
− h, h =
h

The velocity potential at P due to the combination of source and sink at O1 and O 2 is

m m mr2 − mr1
φ= − =
r1 r2 r1r2
m ( 2h ) . ( 2 r ) 4mh.r
Thus =φ =
rr2 ( r1 + r2 ) r1r2 ( r1 + r2 )
2µ.r
= = , where µ 2mh (1)
r1r2 ( r1 + r2 )

In equation (1), let us first keep µ a finite constant and non-zero vector, so that µ = | µ | is a finite
constant and non-zero scalar. Let h → 0 along O1O.

Then m→∞ in such a way that µ remains the same finite non-zero constant vector. In that case, both
r1 , r2 → r and thus under this limiting process, (1) results in

2µ.r µr cos θ µ cos θ

=φ = = ( 2)
2r 3 r3 r2
The limiting source sink combination obtained at O when we keep the direction of h fixed but let h→0
and m→∞ with µ = 2mh remaining a finite non-zero constant, is called a three-dimensional doublet (or
dipole). The scalar quantity µ is called the moment or strength of the doublet. The vector quantity
µ = µµ is called the vector moment of the doublet & µ̂ (unit vector from 02 to 01) determines the
direction of the axis of the doublet from sink to source. From (2), the velocity components are given by
∂φ 2µ cos θ
qr =− = 3
∂r r
1 ∂φ µ sin θ
qθ = − =3
r ∂θ r
qψ = 0

The streamlines due to the doublet are given by

dr rdθ r sin θdψ
= =
2µ cos θ µ sin θ 0
3 3
r r
dr
⇒ = 2 cot θ dθ, d= ψ 0
r
⇒ = r A sin 2 θ=
, ψ constant
Thus, the streamlines due to the doublet lie in planes through the axis of the doublet.
Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 123
3.9.4 Doublet in a Uniform Stream: Let there be a do ublet of
vector moment u = µkˆ at O in a uniform stream whose velocity in
the absence of the doublet is − Ukˆ everywhere (U = constant).
Let P be a point in the fluid having spherical polar co-ordinates (r,
θ, ψ), the direction OZ of the doublet’s axis being the line θ = 0.
PM is the perpendicular on OZ and OM = z= r cos θ . We shall find
the resultant velocity potential due to the combination of the uniform stream and the doublet. We know
that the velocity potential due to the uniform stream is
φ1= Uz= Ur cos θ (1)
and the velocity potential due to a doublet at O, is
µ cos θ
φ2 = 2 ( 2)
r
Thus, the resultant velocity potential at P due to the combination, is

φ = φ1 + φ2 = ( Ur + µr −2 ) cos θ

The absence of ψ in the form of φ is due to the axial symmetry of the flow about OZ.
The velocity components of the flow are
∂φ  2µ 
q r =− =−  − U − 3  cos θ
∂r  r 
1 ∂φ  µ
qθ =
− = U + 3  sin θ
r θ  r 
1 ∂φ
=qψ = 0
r sin θ ∂ψ

Stagnation points ( q = 0 ) are obtained by solving

 2µ   µ
 U − 3  cos
= θ 0,  U + 3  sin
= θ 0
 r   r 
1/3
 2µ 
which are satisfied when sinθ = 0 and r =  
U
Thus, we have the two stagnation points.

  2µ 1/3    2µ 1/3 
   , 0  and    , π

 U    U  
which lie on the axis of symmetry.
124 Partial Differential Equations
1/3
 2µ  1 3
If we write
= r a= i.e. a   = i.e. µ Ua , then for the region r ≥ a, we obtain the same velocity
U 2
potential as for a uniform flow past a fixed impermeable sphere of radius a and centre O. Thus, for r ≥ a,
1
the effect of the sphere is that of a doublet of strength µ = Ua 3 situated at its centre, its axis pointing
2
upstream. So, the sphere can be represented by a suitably chosen
singularity at its centre.
3.9.5 Line Distribution of Sources: Let us consider a uniform line
source AB of strength m per unit length. This means that the
elemental section of AB at a distance x from A and of length δx is a
point source of strength mδx.
Let P be a point in the fluid at a distance r from this element, then the
mδx
velocity potential at P due to the point source is .
r
The total velocity potential at P due to the entire line distribution AB (= 2l) is
21
dx
φ =∫ (1)
0
r

Let AM = x1 , BM = x 2 , where AM is the orthogonal projection of AP on AB. Also, let PM = d, AP = r1 ,

( x1 − x )
2
BP = r2 . Since r 2 = + d 2 + r12 − x12 , therefore, from (1), we get

2/ dx
φ =m ∫
( x1 − x ) + ( r12 − x12 )
0 2

{ )}
 log x − x + 
2/ β 1
( 1 ) ( x1 − x ) + (r − x ∫ dx
2 2 2
 1 1  α
x2 + a2
= m 
−1
( )
β
  = log x + x 2 +a 2 
 0   α

{ }
0

= m log ( x1 − x ) + ( x1 − x ) + ( r12 − x12 ) 

2

  2l

 {
= m log(x1 + r1 ) − log x 2 + x 22 + r12 − x12 
 }
| x1 − 2l = x1 − AB = x 2

 x +r 
= m log  1 1  , where r12 − x12 = d 2 = r22 − x 22 .
 x 2 + r2 
Again, the relation r12 − x12 = r22 − x 22
Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 125

r1 + x1 r2 − x 2 r1 + r2 + x1 − x 2
⇒ = =
r2 + x 2 r1 − x1 r1 + r2 + x 2 − x1
r1 + r2 + 2l
=
r1 + r2 − 2l

Thus,
 r + r + 2l 
φ =m log  1 2  ( 2)
 r1 + r2 − 2l 
a+l
= m log  
 a −l
where 2a is the length of major axis of the ellipsoid of revolution through P having A and B as foci since
constant. It follows from here that the equipotential surfaces φ = constant
for such an ellipsoid r1 + r2 =
are precisely the family of confocal ellipsoid r1 + r2 =
2a obtained when a is allowed to vary.

 ∂φ 
Expression for Velocity: The velocity at P is given by q = −∇φ = −   n (3)
 ∂n 
Let P be any point on the ellipsoid specified by parameter a and P′ the neighboring point on the
ellipsoid specified by parameter a + δa, where

PP ' = δnn
∂  2 + l 
Thus q = −m  log n
∂n  a − l 
 1 1  ∂a 
=−m  − n
 a + l a − l  ∂n
2lm ∂a 
= 2 2 n ( 4)
a − l ∂n
The normal at P to the a-surface bisects the angle 2α between the focal radii AP, BP.
Now,

( r1 + δr1 ) = r12 + ( δn ) − 2r1δn cos (180 − α )

2 2

a 2 + b 2 − c3
cos C =
= r12 + ( δn ) + 2r1δn cos α
2
2ab
⇒ c = a + b 2 − 2ab cos C
2 3
126 Partial Differential Equations

⇒ 2r1δr1= 2r1δn cos α + ( δn ) − ( δr1 )

2 2

| ( δr1 ) = ( δn )
2 2
⇒ δr1 = δn cos a
∂r1
⇒ = cos α
∂n
∂r2
Similarly, = cos α
∂n
Since, 2a=r1 + r2
∂a ∂r1 ∂r2
⇒ 2 = + =. cos α + cos α
= 2 cos α
∂n ∂n ∂n
∂a
⇒ =cos α
∂n
and thus, from equation (4), the velocity of fluid at P is given by

 2lm cos α 
q =  2 2  n
 a −l 
3.10 Hydrodynamical Images for Three-Dimensional Flows
Let us consider a fluid containing a distribution of sources, sinks and doublets. If a surface S can be
drawn in the fluid across which there is no flow, then any system of sources, sinks and doublets on
opposite sides of this surface S may be said to be images of one another w.r.t. to the surface. Further, if
the surface S be considered as a rigid boundary and the liquid removed from one side of it, the motion
on the other side will remain unaltered.
Remark: It is clear that the surface S must be covered by streamlines. Thus, the fluid flow is along S as
the velocity of fluid at any point normal to S must be zero.
3.10.1 Image of a Source in a Plane: Consider a simple source of strength m situated at A(a, 0, 0) at a
distance a from an infinite plane YY′.
We shall show that the appropriate image system for this is
an equal source of strength m at A′(−a, 0, 0), the reflection of
A in the plane. To prove this, we consider two equal sources
of strength m at A(a, 0, 0) & A′ (−a, 0, 0) with no rigid
boundary. Let P0 be any
point on the plane YY′. Then the fluid velocity at P0
due to the two sources is
m m m m
q= AP0 + A ' P0 q= r= r
( AP0 ) ( A ' P0 )
3 3
r2 r3
Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 127

m
⇒q
=
( AP0 )
3 ( AP + A ' P )
0 0

m 2m
= =
( AP0 )
3
2OP0 ( ) ( AP0 )
3 ( OP )
0

 AP0 + A ' P0
(
= AO + oP0 + A 0 ' + OP0) ( )
= 2OP0

This shows that at any point P0 of the plane YY′, the fluid flows tangentially to the plane x = 0 and so
there is no transport of fluid across this plane. Let φ denotes the velocity potential of the flow, then at all
∂φ
points P0 on x = 0, we have = 0 for the region of flow x ≥ 0 . Hence, the image of a source of strength
∂n
m at A in the rigid plane YY' is a source of strength m at A' , as required.
3.10.2 Image of Doublet in a Plane: Consider a pair of
sources −m at A and m at B, taken close together and on one
side of the rigid plane YY′. The image system is −m at A′, m
at B′, where A′ & B′ are respectively the reflections of A and
B in the plane YY′. In the limiting case, when B→A along
BA in such a way as to form a doublet at A, we find that the
image of a doublet in an infinite impermeable rigid plane is a
doublet of equal strength and symmetrically disposed to the
other
w.r.t the plane (antiparallel to each other).
Example: A three-dimensional doublet of strength µ

whose axis is in the direction OZ is distance a from the rigid plane z = 0 which is the sole boundary of
liquid of constant density ρ, infinite in extent. If p ∞ be the pressure at ∞, show that the pressure on the
a 5
plane is least at a distance from the doublet.
2
Solution: Let there be a doublet of strength µ at the point A
with OA = a and YY′ (i.e., z = 0) be the infinite plane. Then
the image system is an equal doublet of strength µ at A′, the
reflection of A in the plane z = 0, and the axis along ZO . The
line OZ is taken as the initial line θ = 0 and plane z = 0 is
θ =π/2. So that P (r, θ, ψ)
 128 Partial Differential Equations

is confined to the region 0 ≤ θ ≤ π/2. Let= , A ' P r2 and α1 , α 2 be the angles which these lines
AP r1=
make with the axis of the doublets as shown in the figure. Then, the velocity potential at P is
µ cos α1 µ cos α
=φ + (1)
r12 r22

r12 = r 2 + a 2 − 2ra cos θ 

where 2  ( 2)
r2 = r 2 + a 2 + 2ra cos θ 

AM OM − OA r cos θ − a
But cos α
=1 = =
r1 r1 r1

A ' M A 'O + OM a + r cos θ

and cos (180 −=
α2 ) = =
r2 r2 r2

⇒ cos α 2 =−
( a + r cos θ )
r2

Using the relation in (1) , we get

µ  r cos θ  µ  − ( a + r cos θ )   r cos θ − a r cos θ + a 
=φ  +  =   −  ( 3)
r12  r1  r22  r2   r1
3
r23 

Further from ( 2 ) we have

∂r1 ∂r r − a cos θ
2r1 = 2r − 2a cos θ ⇒ 1 =
∂r ∂r r1

∂r2 r + a cos θ ∂r1 ra sin θ

Similarly,
= = ,
∂r r2 ∂θ r1
∂r2 ra sin θ
= − .
∂θ r2
Thus from ( 3) , the velocity components are given by
∂φ  cos θ  ∂r  1 cos θ  ∂r  1 
qr = − = µ  3 − 3  2  4 ( r cos θ + a ) − 3 + 3  1  4 ( r cos θ − a ) 
∂r  r2  ∂r  r2 r1  ∂r  r1 
 cos θ
µ 3 −3
=
( r + a cos θ )( r cos θ + a ) − cos θ + 3 ( r − a cos θ )( r cos θ − a ) 

 r2 r25 r13 r15 
 Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 129

  ∂r   ∂r  

1 ∂φ µ r sin θ
( r cos θ − a )  1  ( r cos θ + a )  2  
qθ = − =  3 +3  ∂θ  − r sin θ − 3  ∂θ  
4 3 4
r ∂θ r  r1 r1 r2 r2 
 
µ  r sin θ 3ra sin θ ( r cos θ − a ) r sin θ ra sin θ ( r cos θ + a ) 
=  3 + − + 3 
r  r1 r15 r23 r25 
qψ = 0
2 2
When the point P lies on the plane YY′ i.e., θ = π/2, we have r1= r2= r 2 + a 2 and so at (r, π/2, ψ), the
velocity components are

q r =−6µ ra / ( r 2 + a 2 )
5/2
, q θ =0, q ψ =0 .

Along the streamline through this point, Bernoulli’s equation is

p 1 2 p∞
+ =q cons tan
= t .
ρ 2 ρ
where q=0 at infinity.
Thus, the pressure at any point on the plane YY′ is given by
 
1  36µ 2 a 2 r 2 
p = p∞ − ρ
2  ( r 2 + a 2 )5 
 
2 2 2
18µ a r
i.e. p ( r=
) p∞ − 2 2 5
(r + a )
Now,
dp
= 36ρµ 2 a 2 r ( 4r 2 − a 2 ) / ( r 2 + a 2 )
6
p ' ( r=
)
dr
1
which gives p ' ( r ) 0=
= when r a
2
Also
a  a 
p '  −  < 0, p '  +  > 0
2  2 
a
i.e. p' ( r ) changes sign from negative to positive when r passes through
2
a
⇒ p is minimum at r= θ = π / 2
2
a 
i.e. at the point  , π / 2, ψ 
2 
130 Partial Differential Equations

The distance P0 A is given by

2
a 2 5
  +a = a
2 2

5
Hence p is least at a distance a from the doublet and the minimum value is
2
5
9 4 1
= p ∞ − ρµ 2   6
p min.
2 5 a

Theorem (Weiss’s Sphere Theorem): Let φ ( r, θ, ψ ) be the velocity potential at a point P having
spherical polar co-ordinate ( r, θ, ψ ) in an incompressible fluid having irrotational motion and no rigid
boundaries. Also suppose φ has no singularities within the region r ≤ a . Then if a solid impermeable
sphere of radius a is inserted into the flow with its centre at the origin of co-ordinates, the new velocity
potential at P in the fluid is
a2
a  a2  1 r
φ ( r, θ, ψ ) + φ  , θ, ψ  − ∫ φ ( R, θ, ψ )dR (r > a)
r  r  a 0

3.10.3 Images in Impermeable Spherical Surfaces: We have already studied the effect of placing a
solid impermeable sphere in a uniform stream of incompressible fluid, taking the case of axial
symmetry. Here, we discuss the disturbance produced when a sphere is placed in more general flow.
We shall make use of Weiss’s Sphere Theorem which states as follows:
“Let φ(r, θ, ψ) be the velocity potential at a point P having spherical polar co-ordinates (r, θ, ψ) in an
incompressible fluid having irrotational motion and no rigid boundaries. Also suppose that φ has no
singularities within the region r ≤ a. Then, if a solid impermeable sphere of radius a is introduced into
the flow with its centre at the origin of co-ordinates, the new velocity potential at P in the fluid is

a  a2  1 a 2 /r
φ ( r, θ, ψ ) + φ  , θ, ψ  − ∫ φ ( R, θ, ψ ) dR, ( r > a )
r  r  a 0

a2
where r and are the inverse points w.r.t the sphere of radius a.”
r
Here, the last two terms refer to perturbation potential due to
the presence of the sphere.
3.10.4 Image of a Source in a Sphere: Suppose a source of
strength m is situated at point A at a distance f (> a) from the
centre of the sphere of radius a.
Let B be the inverse point of A w.r.t. the sphere, then OB=a2/f
 Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 131

The velocity potential at P (r, θ, ψ) in the fluid due to a

simple source of strength m at A (f, 0, 0) is
m
φ(r, θ) =
AP
( OP ) + ( OA ) − ( AP ) r 2 + f 2 − ( AP )
2 2 2 2

From ∆OPA, cos θ

= =
2 ( OP )( OA ) 2rf
⇒ AP = r 2 + f 2 − 2rf cos θ
Thus, the velocity potential is
θ ) m ( r 2 + f 2 − 2rf cos θ )
−1/2
φ ( r,= (1)
Inserting a solid sphere in the region r ≤ a, where a < f, we obtain on using Weiss’s sphere theorem, a
perturbation potential

a  a 2  1 a 2 /r
φ ( R, θ ) dR
r  r  a ∫0
φ ,θ −

−1/2
am  a 4 a2  m a 2r 2 2 −1/2
∫
2
i.e.  2
+ f − 2 f cos θ  −  R + f − 2Rf cos θ  dR
r r r  r 0
( ma / f ) m a 2 /r dR
i.e. −
a ∫
r 2 − 2r ( a 2 / f ) cos θ + ( a 2 / f )
2 0
R − 2Rf cos θ + f 2
2

This shows that the image system of a point source of strength m placed at distance f(> a) from the
ma a2
centre of solid sphere consists of a source of strength at the inverse point in the sphere, together
f f
m
with a continuous line distribution of sinks of uniform strength per unit length extending from the
a
centre to the inverse point.
3.10.5 Image of a doublet in a solid sphere: Let us consider a
doublet AB with its axis BA pointing towards the centre O
of a sphere of radius a. Let OA = f, OB = f + δf. Let A′, B′
be the inverse points of A & B in the sphere so that
OB' a 2 / ( f + δf )
OA ' a 2 / f ,=
=

At A, B we associate simple sources of strengths m and −m so that the strength of the doublet is µ =
mδf, where µ is to remain a finite non-zero constant as m→∞ and δf→0 simultaneously.
132 Partial Differential Equations
−1
a2 a2 a 2 a 2  δf 
B' A ' =OA '− OB' = − = − 1 + 
f f + δf f f  f 

a 2 a 2 a 2 δf
= − + to the first order
f f f f
a2
= 2 δf to the first order
f
ma
Now, from the case of “Image of source in a sphere”, the image of m at A consists of at A′ together
f
m
with a continuous line distribution from O to A′ of sinks of strength per unit length and the image of
a
−ma
−m at B consists of at B′ together with a continuous line distribution from O to B′ of sources
(f + δ f )
m
of strength per unit length.
a
The line distribution of sinks and sources from O to B′ cancel each other leaving behind a line
m
distribution of sinks of strength per unit length from B′ to A′ i.e. sink of strength
a
m m  a2  a µa
2 ( ) 2 at B'. The source at B′ is of strength
B'A='  2 δf=
 mδf=
a a f  f f
−1
−ma −ma  δf  ma  δf 
=1 +  =− 1 − 
f + δf f  f  f  f 
to the first order terms
−ma ma −ma µa
= f
+ 2 δ= + 2
f f f f
ma µa
which is equivalent to a sink at B’ and a source 2 at B’
f f
µa
As there is already a sink at B’, therefore source and sink at B’ neutralize.
f2
ma ma
Finally, we are left with source at A′ and a sink at B′. Thus, to the first order, we obtain a
f f
doublet at A′ of strength
ma ma a 2
( B=
' A ') δf
f f f2
ma 3 µa 3
= δf
= .
f3 f3
Axially Symmetric Flow and Three-Dimensional Sources, Sinks and Doublets 133

Hence in the limiting case as δf→0, m→∞, we obtain a doublet at A of strength µ with its axis towards
µa 3
O, together with a doublet at the inverse point A′ of strength 3 with its axis away from O.
f
3.11 Check Your Progress
1. Determine the velocity potential and stream functions, if a sphere is moving in a liquid at rest at
infinity. Hence obtain the equations to the lines of flow.
2. Find the motion of sphere through a liquid at rest at infinity.
3. Find the velocity potential due to the simple source of strength µ.
4. The space between a uniform rigid sphere of radius a and density σ , and a fixed concentric spherical
envelope of radius b is filled with liquid of constant density ρ . An impulse is applied to the rigid
sphere and the system is set into motion from rest. Show that just after the impulse, the K.E. of the
1  2σ b3 + 2a 3 
liquid and sphere is πρa 3 U 2  + 3 3  , where U is the velocity of the sphere.
3  ρ b −a 

5. If r n Sn (θ, ψ ) is harmonic function. Then prove that the function r − (n +1)Sn (θ, ψ ) is also harmonic.
∞
6. Theorem (Statement): If φ ( r, θ, ψ ) has an expression of the form φ ( r, θ, ψ )= ∑ α r S ( θ, ψ ) , the
n
n
n
n =0

series on the right being uniformly convergent w.r.t. r, then for constant λ,
r
1 ∞
α n r n Sn (θ, ψ )
r λ ∫0
R λ−1
φ ( R, θ, ψ ) dR =∑
n =0 n+λ
3.12 Summary
In this section, we have studied about axially symmetric flow, liquid steaming past a fixed sphere,
Equation of motion of sphere and its motion study at rest at infinity, Kinetic energy generated by
impulsive motion, problem on motion of two concentric circles and three-dimensional sources, sinks,
doublets and their images.
3.13 Keyword
Symmetric flow, motion of sphere, sources, sinks, and doublets
3.14 Self-Assessment Test
1. Find the velocity potential when a sphere is moving with constant velocity in a liquid which is
otherwise at rest.
2. Prove that the image with regard to a sphere of a doublet whose axis passes through the center is a
doublet at the inverse point.
3. What is the basic difference between source and line source.
3.15 Suggested Reading
1. O’Neill, M.E. and Chorlton, F., Ideal and Incompressible Fluid Dynamics, Ellis Horwood Limited,
1986.
2. R.K. Rathy, An Introduction to Fluid Dynamics, Oxford and IBH Publishing Company, New Delhi,
1976.
3. G.K. Batchelor, An Introduction to Fluid Mechanics, Foundation Books, New Delhi, 1994.
 UNIT - IV
TWO-DIMENSIONAL MOTION AND STOKE’S STREAM FUNCTION

Content
4.1 Objective
4.2 Introduction
4.3 Use of Cylindrical Polar Co-ordinates
4.4 Stream Function
4.5 Axisymmetric Flow and Stoke Stream Function
4.6 Irrotational Motion in Two-Dimensions
4.7 Complex Velocity Potential
4.8 Images in Two Dimensions
4.9 Blasius Theorem
4.10 Check Your Progress
4.11 Summary of the Unit
4.12 Keyword
4.13 Self-Assessment Test
4.14 Suggested Reading
4.1 Objective
In this unit reader will learn about two-dimensional motion, Stream function, Stoke stream function,
Irrotational motion in two-dimensions, Complex velocity potential, Milne-Thomson circle theorem,
Two-dimensional sources, sinks, doublets and their images.
4.2 Introduction
Assume that a fluid travels in such a way that the flow pattern in one plane within the fluid is the same
as the flow pattern in all other parallel planes within the fluid at any given time. The flow is therefore
described as two-dimensional flow or plane flow at the considered instant. A plane of flow is defined as
any of the parallel planes.
If we consider the plane of flow to be z = 0, then all physical characteristics associated with the fluid,
such as velocity, density, and pressure, are independent of z at any point in the fluid with cartesian
coordinates (x, y, z).
Thus, q� = q� (x, y, t) , ρ = ρ(x ,y, t) etc.
Two-dimensional Motion and Stoke’s Stream Function 135
The pattern will change from instant to instant in an unsteady plane flow, but in a steady situation, the
flow will be plane for all t and the physical quantities will be atmost function x and y.
Planar flow, as described above, is impossible to achieve in practice, although in some important
instances, a close approach to planar flow may occur. The smooth flow of a uniformly deep river with
wide-spaced banks can be considered roughly planar.
Remarks: The uniqueness theorem for three-dimensional irrotational flow hold for two-dimensional
flow also.
4.3 Use of Cylindrical Polar Co-ordinates
2
For an incompressible irrotational flow of uniform density, the equation of continuity ∇ φ = 0 for the
velocity potential φ (r, θ , z ) in cylindrical polar co-ordinates (r, θ , z ) is

1 ∂  ∂φ  1 ∂ 2φ ∂ 2φ
r + + 0
= (1)
r ∂r  ∂r  r 2 ∂θ2 ∂z 2
If the flow is two-dimensional and the coordinate axes are set in such a way that all physical quantities
related to the fluid are independent of z, then φ = φ (r, θ ) and equation(1) simplifies to
1 ∂  ∂φ  1 ∂ 2φ
r + 0
= (2)
r ∂r  ∂r  r 2 ∂θ2
Let us seek solutions of equation (2) by putting
φ (r, θ )=-f(r)g( θ ) (3)

In equation (2) for separation of variables. Thus, we get

1d 1
g(θ)  rf ' (r)  + 2 f(r)g"(θ) = 0
r dr r
d '
r [rf (r)]
dr g" (θ)
i.e. = (4)
f(r) g(θ)
Thus L.H.S. of equation (4) is a function of r only and R.H.S. is a function of θ only. As r and θ are
independent variables, so each side of (4) is a constant λ (say). Thus, we have

r 2f '' (r) + rf ' (r) g '' (θ)

= λ; = -λ
f(r) g(θ)

i.e. r 2f '' ( r ) + rf ' ( r ) − λf ( r ) = 0 (5)

and g (θ) + λ g (θ) = 0

"
(6)
136 Partial Differential Equations

Equation (6) has a periodic solution when λ > 0. Normally the physical problem requires that
2
g (θ + 2π)= g (θ) and this is satisfied when λ = n for n = 1, 2, 3, ….
Thus the basic solution of equation (6) is
g (θ) = acos(nθ) + bsin(nθ) (7)

Now, (5) is the Euler -Homogenous type and it is reduced to a linear differential equation of constant co-
dt 1
efficients by putting r = e t i.e., t = log r ⇒ =
dr r
Also
df df dt 1 df
f ' (r) = = . =
dr dt dr r dt
d 2f d  df  d  1 df 
and f " (r) = 2
=  =  
dr dr  dr  dr  r dt 

1 d  df  df  1 
=   + - 2 
r dr  dt  dt  r 

1  d  df  dt  1 df
=    - 2
r  dt  dt  dr  r dt

1 d 2f 1 df
= 2 2- 2
r dt r dt

2 d 2f df
⇒ r f ''(r) =2 −
dt dt
d 2f df df 2 d 2f 2
Therefore, the equation (5) reduces to - + - n f = 0 ⇒ -n f = 0
dt 2 dt dt dt 2
±nt
It’s a solution is f = exp(±nt) = e = (e t ) ±n = r ±n , n=1,2,3,…

i.e. f = cr n + dr -n (8)
A special solution of (2) is obtained by linear superposition of the form (7) and (8) to given

φ(r, θ) =−f (r)g(θ) =−(A n r n + Bn r − n )(Cn cos nθ + D n sin nθ) (9)

The most general solution of the form

∞
φ(r, θ) =−∑ (A n r n + Bn r − n )(Cn cos nθ + D n sin(nθ) (10)
n =1
Two-dimensional Motion and Stoke’s Stream Function 137

Particular cases. (i) For n=0, we have f = k1+ k2t = k1+ k2 logr and g = k3+ k4 θ so that another solution
of equation (2) is
φ( r, θ) = - ( k1+ k2 logr)( k3+ k4 θ )
(ii) For n=1, we get a special solution as φ = - r cos θ , φ = -r sin θ , φ = -r-1 cos θ , φ = -r-1 sin θ
Example: Discuss the uniform flow past an infinitely long circular cylinder.
Solution: Let P be the point with cylindrical polar co-ordinates (r, θ , z) in the flow region of an
unbounded incompressible fluid of uniform density moving irrotationally with uniform velocity − Ui at
infinity past the fixed solid cylinder r ≤ a
The presence of the cylinder produces a perturbation which is such as to satisfy the Laplace equation
and to become vanishingly small for large r. This suggests taking the velocity potential for
r > a,0 ≤ θ ≤ 2π in the form

(r, θ) Ur cos θ − Ar −1 cos θ ,

φ= (1)

Ux Ur cos θ and due to perturbation, it is

where the velocity potential of the uniform stream is=
− Ar cos θ which tends to zero as r → ∞ and gives rise to a velocity pattern which is symmetrical
−1

about θ= 0, π (the term r −1 sin θ is not there since it does not give symmetric flow). As there is no flow
across r = a, so the boundary condition on the surface is

∂φ
= 0 , when r = a (2)
∂r
Using (2) in (1), we get A=-Ua2 for all satisfying 0 ≤ θ ≤ 2π .
Thus, the velocity potential for a uniform flow past a fixed infinite cylinder is

 a2 
φ(r,
= θ) U cos θ  r +  , r > a,0 ≤ θ ≤ 2π (3)
 r 

The cylindrical components of velocity are (q = −∇φ)

∂φ  a2 
qr =− =− U cos θ 1 − 2  ,
∂r  r 

1 ∂φ  a2 
qθ =− =U sin θ 1 + 2  ,
r ∂θ  r 

∂φ
qz =
− = 0
∂z
138 Partial Differential Equations

We note that as r → ∞ , q r =
− U cos θ,q θ =
U sin θ which are consistent with the velocity at infinity
− Ui of the uniform stream.
Example: A cylinder of infinite length and nearly circular section moves through an infinite volume of
liquid with velocity U at right-angles to its axis and in the direction of the positive x-axis. If the section
is specified by the equation
r a(1+ ∈ cos nθ)
=
where n is positive integer and ∈ is small, show that the approximate value of the velocity potential of
the fluid is

a a
n +1
a
n −1

Ua  cos θ+ ∈   cos(n + 1)θ− ∈   cos(n − 1)θ 
 r r r 

Solution: Let the tangent at a point P on the plane of the

cylinder in the plane of flow makes angels α, (π − α) with the
radial line OP draw from 0 as shown in the figure.
At large radial distances r from OZ, the fluid velocity becomes
vanishingly small.
Let us assume the velocity potential φ(r, θ) of the form
−r − k cos 1, 2,...) .
sin kθ(k =

Thus , we seek a solution of the form

∞
φ ( r, θ ) = −∑ r − k (A k cos kθ + Bk sin kθ)
k =1

(1)
(If we take k=0, this would add on to φ an arbitrary constant
A0 )

At θ = 0 and θ = π on the boundary, q θ = 0 which is satisfied by taking=

Bk 0,
= (k 1, 2,...)

Thus, the velocity potential simplified to the form

∞
φ ( r, θ ) =−∑ A k r − k cos kθ (2)
k =1

which approximately remains unaltered on replacing θ by 2π − θ .

At any point ( r, θ, z ) of the fluid, the cylindrical polar velocity components are (𝑞𝑞� = −∇𝜙𝜙)

∂φ ∞
qr =
− −∑ k.A k .r − (k +1) cos kθ
=
∂r k =1
Two-dimensional Motion and Stoke’s Stream Function 139

1 ∂φ ∞
qθ =
− −∑ k.A k .r − (k +1) sin kθ
=
r ∂θ k =1

∂φ
qz =
− 0
=
∂z
At P on the boundary, since (π − α) is the angle between
the tangent and the radius vector OP, therefore
1 dr d
cot(π −=
α) = (log r)
r dθ d θ
d
− cot α
⇒= [log a(1+ ∈ cos nθ)]
dθ
1
(−a ∈ n sin nθ)
a(1+ ∈ cos nθ)

∈ n sin nθ
⇒ cot α =
1+ ∈ cos nθ
∈ n sin nθ
⇒ cot α = (3)
1+ ∈ cos nθ
The normal component of velocity U N of the boundary at P is

𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑
cos(𝜋𝜋 − 𝛼𝛼) = , sin(𝜋𝜋 − 𝛼𝛼) = 𝑟𝑟
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑
U N U sin(α − θ)
=

= U(sin α cos θ − cos α sin θ)

U cos θ (1+ ∈ cos nθ ) − sin θ ∈ n sin nθ 

i.e. U N = (4)
(1+ ∈ cos nθ )
2 2 2 2
+ ∈ n sin nθ

As there is no transport of fluid across the surface and no break away from it, so U N is also the normal
velocity component of the fluid.
Thus,
U N q r sin α + q θ cos α
=

 ∞   ∞ 
 ∑ ( )  −∑ k.A k .r sin kθ  (∈ n sin nθ )
− (k +1) − (k +1)
− k.A k .r cos kθ  1+ ∈ cos nθ
 k 1=  + k1 
(1+ ∈ cos nθ ) + (∈ n sin nθ ) (1+ ∈ cos nθ ) + (∈ n sin nθ )
2 2 2 2
 140 Partial Differential Equations
∞
−∑ k.A k .a − (k +1) (1+ ∈ cos nθ ) cos kθ (1+ ∈ cos nθ ) + sin kθ ∈ n sin nθ 
− (k +1)

= k =1
(5)
(1+ ∈ cos nθ ) + (∈ n sin nθ )
2 2

From equations (4) and (5) we have

∞
−∑ k.A k .a − (k +1) (1+ ∈ cos nθ ) cos kθ (1+ ∈ cos nθ ) + sin kθ ∈ n sin nθ 
− (k +1)

k =1

= U cos θ (1+ ∈ cos nθ ) − sin θ ∈ n sin nθ 

We further simplify equation (6) for the terms upto 1st order in ∈ .
L.H.S. of (6)
∞
=− �𝑘𝑘=1 𝑘𝑘. 𝐴𝐴𝑘𝑘 . 𝑎𝑎−(𝑘𝑘+1) (1−∈ (𝑘𝑘 + 1)cos𝑛𝑛𝑛𝑛) [cos𝑘𝑘𝑘𝑘+∈ cos𝑘𝑘𝑘𝑘cos𝑛𝑛𝑛𝑛+∈ 𝑛𝑛sin𝑘𝑘𝑘𝑘sin𝑛𝑛𝑛𝑛]
∞

= − � 𝑘𝑘. 𝐴𝐴𝑘𝑘 . 𝑎𝑎−(𝑘𝑘+1) [cos𝑘𝑘𝑘𝑘−∈ (𝑘𝑘 + 1)cos𝑘𝑘𝑘𝑘cos𝑛𝑛𝑛𝑛+∈ cos𝑘𝑘𝑘𝑘cos𝑛𝑛𝑛𝑛 + sin𝑘𝑘𝑘𝑘 ∈ 𝑛𝑛sin𝑛𝑛𝑛𝑛]

𝑘𝑘=1
∞
= −∑ k.A k .a − (k +1) [ cos kθ− ∈ k cos kθ cos nθ + sin kθ ∈ n sin nθ]
k =1

∞
 ∈k ∈n
= −∑ k.A k .a − (k +1) cos kθ − {cos(n + k)θ + cos(n − k)θ} + {cos(n − k)θ − cos(n + k)θ}
k =1  2 2 
∞
 ∈ ∈ 
= −∑ k.A k .a − (k +1) cos kθ − (n + k) cos(n + k)θ + (n − k) cos(n − k)θ  (7)
k =1  2 2 
R.H.S. of (6)

 ∈ ∈n
= U cos θ + {cos(n + 1)θ + cos(n − 1)θ} − {cos(n − 1)θ − cos(n + 1)θ}
 2 2 

 ∈ ∈ 
= U cos θ + (n + 1) cos(n + 1)θ + (1 − n) cos(n − 1)θ  (8)
 2 2 
Correct to the first order of approximation, from equations (6), (7) and (8), comparing coefficients of
cos θ, cos(n − 1)θ, cos(n + 1)θ , we get

−A
− Ua 2
U =2 1 ⇒ A1 = (9)
a
where n + k ≠ 1
 Two-dimensional Motion and Stoke’s Stream Function 141

In (7) cos kθ → cos(n − 1)θ

 1  1
− ( n − 1) A (n −1) a − n + A1a −2 ∈ (n − 1)  = U ∈ (n − 1) cos(n − k)θ → cos(n − 1)θ
 2  2
similarly for n + 1

 1  1
and − ( n + 1) A (n +1) a − (n + 2) − A1a −2 ∈ (n + 1)  = U ∈ (n + 1)
 2  2

⇒ A n −1 =U ∈ a n , A n +1 =− U ∈ a n + 2

All A k other than A1 , A n −1 , A n +1 are zero. Putting the value of these three non-zero co-efficients in (2),
we get

φ(r, θ) = −  A1r −1 cos θ + A (n −1) r − (n −1) cos(n − 1)θ + A (n +1) r − (n +1) cos(n + 1)θ

a a
n +1
a
n −1

= Ua  cos θ+ ∈   cos(n + 1)θ− ∈   cos(n − 1)θ 
 r r r 

Hence the result.

4.4 Stream Function
When all of the lines of motion are parallel to a fixed plane, such as the xy-plane, and the velocity at all
of the corresponding locations of all planes parallel to the xy-plane has the same magnitude and
direction, the motion is said to be two-dimensional.

For two-dimensional motion of incompressible fluid in xy-plane, the velocity q is a function of x,y,t
only and the differential equation of the streamline is
dx dy
= ⇒ vdx − udy =0 (1)
u v
and the corresponding equation of continuity is
∂u ∂v
+ 0
= (2)
∂x ∂y
We note that equation (2) is the condition of exactness of the equation (1), it follows that L.H.S. of (1)
must be an exact differential, say dψ . Thus

∂ψ ∂ψ
vdx − udy = dψ = dx + dy
∂x ∂y

∂ψ ∂ψ
So that u =
− ,v =
∂y ∂x

The function ψ is called the stream function or lagrange stream function or the current function.

Obviously the streams are given by the solution of equation (1), i.e. dψ =0
142 Partial Differential Equations

i.e. ψ (x, y, t) =
f (t) , for unsteady flow and

ψ (x, y) =
constant, for steady flow.

Thus, the stream function is constant along a streamline, which justifies the name given to this function.
From the above discussion, it is clear that the existence of stream function is merely a consequence of
the continuity and incompressibility of the fluid moving in two dimensions. The stream function always
exists in all type of two-dimensional motion whether rotational or irrotational. However, it should be
noted that the velocity potential exists only for irrotational motion whether two-dimensional or three-
dimensional. Thus, two-dimensional irrotational motion, both functions always exist.
Physical Interpretation of Lagrange Stream Function
Consider a curve C in the xy-plane. If the tangent at any point P of the element ds of C makes an angle
θ with x-axis, then the direction cosines of the normal at P (directed from right to left) are
< co(90 + θ), cos θ, 0 >

i.e., < − sin θ, cos θ, 0 >

The flow Q per unit depth of the fluid across the curve C from right to left is

Q = ∫ q.nˆ ds
C

where n̂ = -sinθiˆ + cos θjˆ , q= uiˆ + vjˆ

∴ Q= ∫ ( −u sin θ + v cos θ ) ds
C

𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕

= �� sin𝜃𝜃 + cos𝜃𝜃� 𝑑𝑑𝑑𝑑 , 𝑤𝑤ℎ𝑒𝑒𝑒𝑒𝑒𝑒, 𝑢𝑢 = , 𝑣𝑣 =
𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕 𝜕𝜕𝜕𝜕
𝐶𝐶

 ∂ψ dy ∂ψ dx  dx dy
= ∫  ∂y ds + ∂x
C
 ds
ds 
where, =
cos θ
ds
,sin θ
=
ds

 ∂ψ ∂ψ 
= ∫  ∂y dy + ∂x dx 
C

= ∫ dψ = ψ B − ψ A
C

where ψ A and ψ B are values of ψ at the initial and final points of the curve C. Thus the difference of
the values of the stream function at the end points of a curve equals the flow across that curve.
Deduction: If we suppose that the curve C be the streamline, then no fluid crosses its boundary, then
ψB − ψA = 0 ⇒ ψB = ψA

i.e. ψ is constant along C.

Two-dimensional Motion and Stoke’s Stream Function 143
i.e. ψ = constant along the streamline.

4.5 Axisymmetric Flow-Stoke’s Stream Function

Because the determination of all fluid velocity is reduced to finding a single scalar function, the
inclusion of a stream function gives a consistent approach to solving all two-dimensional flow issues,
whether rotational and irrotational. Unfortunately, this unified manner of approach is not practicable in
the case of general three-dimensional motion. However, there exist a number of classes of flow, which
can be uniquely characterized by means of a single scalar function. Each of these classes of flow
involves a certain mode of symmetry. The most important of these is axisymmetric flow, in which the
streamlines in all the plane passing through a given axis are the same. We have already considered such
flow for irrotational motion in spherical polar co-ordinate (r, θ, ψ ) in which the line θ =0 is the axis of
symmetry. If (r, θ, Ζ) denote cylindrical polar coordinates and z-axis be taken as axis of symmetry then
q θ = 0 and the fluid motion is the same in every plane θ = constant (meridian plane). Thus, all the
quantities associated with the flow are independent of θ . The equation of the continuity becomes
∂ ∂
(rq r ) + (rq z ) =
0 (1)
∂r ∂z
∂ ∂
i.e. (rq r ) = − (rq z ) (2)
∂r ∂z
This is the condition of exactness of the differential equation rq r dz − rq z dr =
0

dr dz
= → stream line
qr qz

⇒ rq r dz − rq z dr =
0

This means that equation(2) is an exact differential equation and let L.H.S. be an exact differential
equation say dψ .

∂ψ ∂ψ
∴ rq r dz − rq z dr = dψ = dr + dz
∂r ∂z
∂ψ ∂ψ
⇒ −rq z ,
= rq r
= (3)
∂r ∂z
The function ψ in equation (3) is called Stoke’s stream function.

The equation of streamline in the meridian plane θ = constant (at a fixed time t for unsteady flow) is
dr dz
=
qr qz

⇒ q r dz − q z dr =
0
144 Partial Differential Equations

0 , where r is the I.F.

⇒ rq r dz − rq z dr =

⇒ dψ= 0 ⇒ ψ= constant .

4.5.1 Stoke’s Stream Function in Spherical Polar Co-ordinates (r, θ, ψ )

We consider the axisymmetric flow in r, θ plane such that q ψ = 0 . The equation of continuity in
spherical polar co-ordinates becomes
1 ∂ 2 1 ∂
2
(r q r ) + 2 (r sin θq θ ) =
0
r ∂r r sin θ ∂θ
∂ 2 ∂
i.e. (r sin θq r ) = (−r sin θq θ ) (1)
∂r ∂θ
This is the condition of exactness for the differential equation
r sin θq θ dr − r 2 sin θq r dθ =0 (2)

Thus, the expression on L.H.S. of (2) is an exact differential, say dψ such that

∂ψ ∂ψ
r sin θq θ dr − r 2 sin θq r dθ = dψ = dr + dθ
∂r ∂θ
∂ψ ∂ψ
⇒ r sin θ qθ ,
= −r 2 sin θ qr
=
∂r ∂θ
dr rdθ
Streamline is = ⇒ q θ dr − rq r =
dθ 0
qr qθ

⇒ r sin θq θ dr − r 2 sin θq r dθ =0
⇒ dψ =0
⇒ ψ =constant

Remark: In the above cases, the motion need not irrotational, i.e. velocity potential may not exist.
Further, the function ψ is not harmonic, i.e it does not satisfy Laplace equation.

4.5.2 Stoke’s Stream Function for basic flows

(1) Uniform Stream: Let a uniform stream with velocity U be advancing in the direction of z-axis such
ˆ . Then, from the relations
that q = UK

1 ∂ψ 1 ∂ψ
qz =
− , qr =
r ∂r r ∂z
1 ∂ψ 1 ∂ψ
We get U =
− , 0
=
r ∂r r ∂z
 Two-dimensional Motion and Stoke’s Stream Function 145

∂ψ ∂ψ
− Ur,
= 0
=
∂r ∂z
Integrating, we obtain
1
ψ = − Ur 2 , where the constant of integration is zero.
2
In spherical polar co-ordinates, we find
1 1
ψ = − U(r sin θ) 2 = − Ur 2 sin 2 θ
2 2

(2) Simple Source at Origin: For a simple source, we have q = f (r)rˆ

If the simple source is of strength m and placed at origin, then we know that
m m
=q =2
rˆ ( 3 r), r > 0,
r r
In spherical polar co-ordinates
m
i.e. ( qr , qθ ) = rˆ (1)
r2
Also, in spherical polar co-ordinates, we have
1 ∂ψ 1 ∂ψ
qr =
− 2 , qθ = (2)
r sin θ ∂θ r sin θ ∂r
From equations(1) and (2), we get
m 1 ∂ψ ∂ψ
2
=
− 2 , 0
=
r r sin θ ∂θ ∂r
∂ψ ∂ψ
⇒ −m sin θ,
= 0
=
∂θ ∂r
⇒
= ψ m cos θ , constant of integration is zero.

A constant may be added to this solution and this is usually done to make ψ =0 along the axis of
symmetry θ =0 . In such case,
= ψ m(cos θ − 1)
For a sink of strength m at the origin, Stoke’s Stream Function is
= ψ m(cos θ − 1)
(3) Doublet at Origin: We assume that the flow is due to only a doublet of strength m at the origin O.
Taking the axis θ =0 of the system of spherical co-ordinates to coincide with the axis of the doublet, we
know that the velocity potential at P(r, θ, ψ ) is

m cos θ
=φ , r>0 (1)
r2
 146 Partial Differential Equations

∂φ 2m cos θ 1 ∂φ m sin θ
qr =
− = 3 , qθ =
− = 3 , qψ =
0 (2)
∂r r r ∂θ r
But the relation between the velocity components and the Stoke’s Stream Function ψ are

1 ∂ψ 1 ∂ψ
qr =− 2 , qθ =
r sin θ ∂θ r sin θ ∂r
(3)
From (1) and (2), we get

∂ψ −2m sin θ cos θ ∂ψ m sin 2 θ

= ,
∂θ r ∂r r2
Integrating, we obtain

− m sin 2 θ
ψ= , constant of integration is zero.
r
(4) Uniform Line Source: Let a uniform line source of
fluid extends along the straight line AB of length l.
Consider an element QQ' of length δz at a distance
z(=AQ) from A.
Thus, we have a simple source of strength m δz , where m is the constant source strength per unit length
of the distribution along AB.
Let QP=r, ∠PQM= Q, PM= d

The Stoke’s Stream function δψ at P for the simple source of strength m δz at Q is mδz ( cos θ − 1) ,
where we have adopted the convention that ψ =0 on the axis of symmetry through AB.

Then, the value of Stoke’s Stream function ψ at P due to entire line source AB is given by
l l
=ψ m ∫ cos θdz − m ∫ dz
0 0

l l +b− z
m∫ dz − ml
d 2 + (l + b − z )
0 2

 
 In ∆PQM QM QB + BM l − z + b l−z=b 
cos
= θ = = =
 PQ PQ r d 2 + (l + b − z )
2 
 
Putting l+b-z=x ⇒ dz=-dx
When z=0, x=l+b
When z=l, x=b
 Two-dimensional Motion and Stoke’s Stream Function 147
b − xdx
=∴ ψ m∫ − ml
l +b
d 2 + x2
−1
m l +b
2 b
(
= ∫ d 2 + x2 ) 2
(2 x)dx − ml

m 2
d + ( l + b ) − d 2 + b 2  − ml
2
=

2 

= m(AP-BP)-mAB
= m(AP-BP-AB)
As P is the only variable point in the above expression, the simpler form m(AP-BP) can be taken are ψ
=constant, i.e., AP-BP=constant.
These are cofocal hyperboloids of revolution about AB, with A and B foci. We have earlier that the
equipotential are confocal ellipsoids of revolution about AB with the same force. Als, it is well known
result that the two families of confocals interesect orthogonally.
4.5.3 Stoke’s Stream function for a Doublet in a Uniform Stream

Let a doublet of vector moment mkˆ be situated at the origin O in a uniform stream whose undisturbed
velocity is − Ukˆ . In spherical polar co-ordinates (r, θ, ψ ) , the Stoke’s stream functions for each separate
distribution are respectively.
1 2 2
=ψ1 Ur sin θ , for uniform stream, q − =Ukˆ
2
−m 2
=
ψ2 sin θ , for doublet at origin.
r
Here the Stoke’s Stream Function for the combination is

1 m
ψ (=
r, θ )  Ur 2 −  sin 2 θ
2 r 
The equations of stream surfaces are ψ ( r, θ ) =constant. In particular, the stream surfaces for which
ψ =0 are given by

1 2 m 2
 Ur −  sin θ =0
2 r 

1 2 m
⇒ sin θ 0 or =
= Ur − 0
2 r
1
 2m  3
⇒ θ= 0, π , i.e., the z-axis or r =  
 U 
 148 Partial Differential Equations
1
 2m  3
The surface of the sphere with centre O and radius  
 U 

Example: A point source of strength Ua 2 is introduced at O in a uniform stream whose undisturbed

θ
velocity is Uiˆ . Show that over the surface of revolution r sin θ =2a cos there is no flow, the system of
2
spherical polar co-ordinates (r, θ, ψ ) being used with θ =0 taken as x-axis. If a rigid surface of
revolution having above equation is introduced into the flow of velocity Uiˆ after removal of the point
surface, explain why the two models are hydrodynamically equivalent for corresponding points in the
field of flow?
1
Solution: Stoke’s Stream function for uniform stream is − Ur 2 sin 2 θ and for a point source at O, it is
2
Ua (cos θ − 1) . Thus, the total stream function for the combination is
2

1
ψ (r, θ) = − Ur 2 sin 2 θ + Ua 2 (cos θ − 1)
2
 1 
= U a 2 (cos θ − 1) − r 2 sin 2 θ  ; r > 0, 0 ≤ θ ≤ π
 2 
The stream surfaces are given by ψ =constant

 1 
i.e. U a 2 (cos θ − 1) − r 2 sin 2 θ  = constant
 2 
1
⇒ a 2 (1 − cos θ) + r 2 sin 2 θ =constant
2
θ 1 2 2
⇒ a 2 .2sin 2 + r sin θ =constant
2 2
θ 1
⇒ 2a 2 (1 − cos 2 ) + r 2 sin 2 θ =constant
2 2
θ 1 2 2
⇒ −2a 2 cos 2 + r sin θ =constant
2 2
The particular, taking the constant on R.H.S. to be zero, the corresponding stream surfaces is
θ
r 2 sin 2 θ =4a 2 cos 2
2
θ
i.e. r sin θ =2a cos 2 (1)
2
which is the required surface of revolution.
 Two-dimensional Motion and Stoke’s Stream Function 149
We note that equation (1) is satisfied by
θ
(i) cos = 0, i.e., θ = π
2
θ
(ii) r = a cosec .
2
θ
As no flow takes place over the surface of revolution r = acosec , we may introduce a rigid boundary
2
over the surface, excluding the fluid and source within its interior. Then the hydrodynamical image of
the external flow Uiˆ in the surface is the point source Ua 2 at O. Hence the two models are
hydrodynamically equivalent for corresponding points in the field of flow.
Also, we have
1 ∂ψ 1 ∂ψ
qr =
− 2 , q θ == , qψ 0
r sin θ ∂θ r sin θ ∂r
a
q r = U(cos θ + ), q θ = − U sin θ, q ψ = 0
r2
∴ Fluid speed q at a point P is
2 4
2 2 2 a a
q
= q + q + q= U 1 + 2   cos θ +  
r θ ψ
r r
θ
On the surface of revolution r = a cos ec , the fluid velocity is tangential to the surface and
2
U
=q 5 + 2 cos θ − 3cos 2 θ
2
The only stagnation point on the surface is given by cos θ = −1 , corresponding to θ = π .
4.6 Irrotational Motion in Two-dimensions
We know that both scalar functions, namely the velocity potential φ and the stream function ψ , exist
for a two-dimensional irrotational motion of an incompressible fluid.
The velocity potential φ is given by

  ∂φ ∂φ 
q = −∇φ = −  , 
 ∂x ∂y 
∂φ ∂φ
i.e. v = - , u = − (1)
∂y ∂x

and the stream function ψ is given by

150 Partial Differential Equations

∂ψ ∂ψ
u=- ,v = (2)
∂y ∂x

From equations (1) and (2), we get

∂φ ∂ψ ∂φ ∂ψ
= , = − (3) | using CR equations
∂x ∂y ∂y ∂x

Equation (3) implies that

∂2 ∂2
∇ 2 φ = 0 and ∇ 2 ψ = 0, ∇ 2 ≡ +
∂x 2 ∂y 2

i.e. φ and ψ are harmonic functions.

Again, from, (3) we have,

  
∇φ = gradφ = −q = −(u i + vj)
 ∂ψ  ∂ψ    ∂ψ  ∂ψ  
=-  − i+ j= i− j
 ∂y ∂x   ∂y ∂x 

∂ψ   ∂ψ  
= ( j × k) + (i × k)
∂y ∂x
 ∂ψ  ∂ψ    
= j+ i  × k = ∇ψ × k (4)
 ∂y ∂x 

Also, from (3), we note that

∂φ ∂ψ ∂ψ  ∂φ 
= − 
∂x ∂x ∂y  ∂y 

∂φ ∂ψ ∂φ ∂ψ
⇒ + 0
=
∂x ∂x ∂y ∂y
⇒ ∇φ.∇ψ = 0 (5)

Thus, for irrotational incompressible two-dimensional flow (steady or unsteady), φ(x, y) and ψ (x,y) are
harmonic functions and the family of curves φ(x, y) =
constant(equipotential)
and ψ (x,y)= constant(streamline) intersect orthogonally.

4.7 Complex Velocity Potential

In this chapter, we introduce the concept of a function of a complex variable, which can be used to solve
a variety of two-dimensional flow problems. We focus on irrotational plane flows of incompressible
fluid of uniform density that have a velocity potential and a stream function, where (x, y) denotes two-
dimensional cartesian coordinates in the flow plane, i.e. the xy-plane.
Suppose that z= x + iy and that
Two-dimensional Motion and Stoke’s Stream Function 151
W =f (z) =φ(x,y) + iψ (x,y) (1)

where x, y, φ, ψ are all real and i= −1 , we write

= φ Re(W),
= ψ Im(W)
We assume that all four first-order partial derivatives of w.r.t. x and y exist and are continuous over the

flow plane. Now the velocity q = (u, v) has components satisfying

∂φ ∂ψ ∂φ ∂ψ
u=
− =
− ,v =
− = (2)
∂x ∂y ∂y ∂x

Thus, φ and ψ satisfy the CR equations and so W=f(z) must be an analytic function of z=x+iy.

The function W =f (z) =φ(x,y) + iψ (x,y) is called the complex velocity potential or complex potential
of the plane flow.
Differentiating W = φ + iψ partially w.r.t. x, we get

∂W ∂φ ∂ψ ∂φ ∂φ
= +i = − i = −u + iv
∂x ∂x ∂x ∂x ∂y

∂W dW ∂z dW ∂z
But
= = . = 1
∂x dz ∂x dz ∂x
Thus
dW
W= =−u + iv
dz
dW
⇒− = u − iv= q cos θ − iq sin θ
dz

=q(cosθ-isinθ)=qe − iθ , q = q (3)

where the local fluid velocity q at a point P(x,y) in the plane of flow is inclined at an angle θ top the
positive x-axis. The combination u-iv is known as complex velocity.
dW dW
We have, q = − = u 2 + v 2 , θ = arg W '(z) and for stagnation points =0.
dz dz
Thus, a detailed analysis of an incompressible and irroational plane flow requires finding its complex
potential W(z).

Example: Discuss the flow for which the complex potential is W = z 2 .

Solution: We have

W = φ + iψ = z 2 = (x + iy) 2 = x 2 − y 2 + i2xy
⇒ φ(x, y)= x 2 − y 2 , ψ (x, y)= 2xy
152 Partial Differential Equations

The equipotentials, φ =constant are the rectangular hyperbolae x 2 − y 2 =constant having asymptotes
y=±x. The streamlines ψ = constant, are also the rectangular hyperbolae xy=constant having the axes
x=0, y=0 as asymptotes.
The two families of rectangular hyperbolae cut orthogonally in accordance with general theory.
dW
Also, =−u + iv =2 z ,
dz
so the only stagnation point is the origin.
4.7.1 Complex Potential for a Uniform Stream
Let a uniform stream advances with a velocity having constant magnitude U and being inclined at angle
dW
α to the positive direction of x-axis, then we have u = U cos α, v = sin α and thus − =u − iv =Ue − iα
dz
The simplest form for W, ignoring the constant of integration, is W = − Uze − iα
i.e., φ+iψ =-U(x+iy)(cosα-isinα)
=-U(xcosα + ysinα)-iU(ycosα-xsinα)

⇒ φ=-U(xcosα + ysinα)
ψ =-U(ycosα-xsinα)

Thus the equations of equipotentials are

xcosα + ysinα =constant (1)

ycosα-xsinα = constant (2)

Equations (1) and (2) represent two families of parallel straight lines inclined at angle α to the positive
y-axis and positive x-axis respectively. The two families of straight lines intersect orthogonally in
accordance with general theory.
4.7.2 Basic Hydrodynamical Singularities in Two-dimensional Flows
Two- dimensional source, sink and doublet are analysis of the three dimensional simple source, sink
and doublet discussed earlier.
1. Source (Line Source): If the two dimensional motion of a liquid consists of symmetrically
distributed outward radial flow from a point, then the point is called a simple source. But we must
remember that our two-dimensional motion is the motion of liquid occupying three dimensions, the
simple source is necessarily a line source and it may be regards as a straight axis of unit length between
two fixed planes, which emits fluid radially and symmetrically in the plane of flow. If O is the point at
which a line source lies, then we may take it as an outward radial flow through a small circle of radius r
centered at O.
Suppose the line source emits fluid at the rate 2πmρ units of mass per unit length of the source per unit
time, in all directions in the plane of flow, then m is called the strength of the line source.
Two-dimensional Motion and Stoke’s Stream Function 153
An example of a line source is a long straight hose, with perforations along its length commonly used
for watering for a long period of time.
2. Line Sink: A line source of negative strength, i.e. inward symmetrical radial flow in a plane is called
a line sink. Thus, a line source of strength m is called a line sink.
3. Line Doublet: A line doublet or dipole is formed when a line source and a line sink of equal strength
are placed infinitesimally near to each other.
Remarks: 1. A circular vertex and a line vortex are also two-dimensional hydrodynamical singularities,
which will be studied later on.
2. In a two-dimensional flow, a source, sink and doublet, vortex mean a line source, line sink, line
doublet, line vortex respectively and strength means strength per unit length.
4.7.3 Complex Potential for a Line Source
Let us consider a line source of strength m per unit length at the origin O (z=0). Since the flow is radial,
the velocity has the radial component q r only. Then the flow across a small circle of radius r centered at
O( by law of conservation of mass) is
2πrq r ρ = 2πmρ
m
⇒ qr =
r
The complex potential is obtained from the relation
−dW
= u − iv= q r cos θ − iq r sin θ
dz
m
=q r (cos θ − i sin θ) = e − iθ
r
dW m m m
⇒ − e − iθ =
= iθ
=
−
dz r re z
Integrating we get
W = −m log z (1)

where constant of integration has been ignored.

Equation (1) gives the required complex potential, where logz is analytic.
We can write (1) as

φ + iψ = −m log(reiθ ) = −m log r − imθ

⇒ φ = −m log r, ψ = −mθ
The equipotential and streamlines are
 154 Partial Differential Equations

=r constant, θ=constant
i.e., x 2 + y 2 = c1 , y = c 2 x

which are orthogonal.

If the line source is at z = z 0 instead of z=0, then

W=
−m log(z − z 0 )

For a line sink of strength m per unit length at z = z 0 , the complex potential is

=W m log(z − z 0 )

For a number of sources at z = z1 , z 2 ,..., z n with respective strengths m1 , m 2 ,..., m n the complex potential
is
W=
− m1 log(z − z1 ) − m 2 log(z − z 2 ) − ...m n log(z − z n )

4.7.4 Complex Potential For a line Doublet

Let us take a line source of strength m per unit length at the point P(z = aeiα ) and a line sink of strength
m per unit length at the origin (z=0). The complex potential due to the combination is

−m log(z − aeiα ) + m log(z − 0)

W=
 z − aeiα   aeiα 
=-m log   =-m log 1 − 
 z   z 

 aeiα a 2 e 2iα 
=m  + 2
+ ...
 z 2z 
We note that OP=a is the distance between the source and the sink. As a → 0, m → ∞ so ma → µ and P
tends to coincide with O along OP, the resultant is a doublet of strength µ at O in the direction OP (axis
of the doublet ) with complex potential

µeiα
W= (1)
z
If the sink is situated at z0, then the complex potential is

µeiα
W=
z − z0

If α =0, i.e. the doublet is along x-axis, then

µ
W=
z − z0
 Two-dimensional Motion and Stoke’s Stream Function 155
If there are several doublets each of strength µ1 , µ 2 ,..., µ n with sinks at points z1 , z 2 ,..., z n and their axes
inclined at angles α1 , α 2 ,..., α n with +ve x-axis, then

µ1eiα1 µ 2 eiα2 µ e iα n
W
= + + ... + n
z − z1 z − z 2 z − zn

Example: Discuss the flow due to a line doublet at origin of strength µ per unit length with its axis
along the x-axis.
Solution: We know that the complex potential for a doublet is

µeiα
W=
z − z0

In the present case,=

α 0,=
z 0 0, so

µ µ µ(x − iy)
w= = =
z x + iy x 2 + y2

µx µy
φ + iψ
⇒= 2
−i 2
2
x +y x + y2
µx µy
=
⇒φ 2
= 2
,ψ
x +y x + y2
2

Thus, the equipotentials φ =constant, are the coaxial circles

x 2 + y2 =
2k1x (1)

and the streamlines ψ =constant, are the coaxial circles

2k 2 x (2)
x 2 + y2 =

Family (1) have centers (k1, 0) and radii k1 and family (2) have centers (0, k2) and radii k2. The two
families are orthogonal.

 a2 
Example: What arrangement of source and sinks will give rise to the function
= W log  z −  ? Also
 z 
prove that two of the streamlines are a circle r=a and x=0.;
Solution: We have
a2 z2 − a 2
W = log(z − ) = log( )
z z
iψ log(z − a) + log(z + a) − log z
⇒ φ += (1)

which represents a line source at z=0 and two line sinks at z=±a, each of strength unity per unit length.
We can write (1) as
156 Partial Differential Equations

φ + iψ= log(x − a + iy) + log(x + a + iy) − log(x + iy)

y y y
⇒ ψ tan −1 + tan −1 − tan −1
x −a x+a x
 y y 
 +  y
= tan −1  x − a x2 + a  − tan −1
 1− y  x
 2 2 
 x −a 
 2xy  y
= tan −1  2 2 2 
− tan −1
 x − y −a  x
 x 2 + y 2 + a 2   y  
−1
= tan  2 2 2   (2)
 x + y − a   x  
Streamlines ψ =constant are given by

(
y x 2 + y2 + a 2 = ) (x 2
)
+ y 2 − a 2 x tan α , where α is a constant.

π
In particular, if we take α = , then we get the streamlines as
2

(x 2
)
+ y 2 − a 2 x =0

i.e. x 2 + y 2 = a 2 = 0, x = 0

i.e., r=a, x=0

Hence the result.
Example: Two sources, each of strength m, are placed at the points (±a, 0) and a sink of strength 2m is
placed at the origin. Show that the streamlines are the curves
(x 2 + y
= ) a 2 (x 2 − y 2 + λxy) , λ being a parameter. Show also that the fluid speed at any point is
2 2

2ma 2
,where r1 , r2 , r3 are respectively, the distance of the point from the sources and the sink.
r1r2 r3

Solution: The complex potential of the given system is

W=−m log(z − a) − m log(z + a) + 2m log z
W= m log z 2 − log(z 2 − a 2 ) 
+ iψ m log(x 2 − y 2 + i2xy) − log(x 2 − y 2 − a 2 + i2xy) 
⇒ φ= (1)
 Two-dimensional Motion and Stoke’s Stream Function 157

  2xy   2xy 
ψ m  tan −1  2 2 
− tan −1  2 2 2 
 x −y   x − y − a 
 
−2a 2 xy
=m tan −1   (2)
( )
 x 2 + y 2 2 − (x 2 − y 2 )a 2 
 
Streamlines are given by

 −2 
= = m tan −1   , say
ψ constant (3)
 λ 
From equations (2) and (3), we obtain

2 2a 2 xy
= 2
λ (x + y 2 ) 2 − (x 2 − y 2 )a 2
⇒ (x 2 + y 2 =
) 2 (x 2 − y 2 + λxy)a 2
Hence the result.
Now the fluid velocity is given by

dW  2m m m   2ma 2 
u − iv =− =− − − =
  
dz  z z − a z + a   z(z − a)(z + a) 
So, the fluid speed is given by

dW  2ma 2   2ma 2   2ma 2 

q=u − iv =− = =  = 
dz  z(z − a)(z + a)   z (z − a) (z + a)   r1r2 r3 
where
z + a = r1 , z − a = r2 , z = r3 .

Hence the result.

Example: Parallel line sources (perpendicular to the xy-plane) of equal strength m per unit length are
placed at the points z =ina, where n=-∞,....,-2,-1,0,1,2,...∞, prove that the complex potential is
 πz 
W = −m log sinh   . Hence show that the complex potential for line doublets with their axes parallel
 a 
to the x-axis, of strength µ per unit at the same points, is given by

πz mπ
W1 = −µ coth( ), µ = .
a a
Solution: Firstly, we take 2n+1 sources with origin at the middle one and then take n → ∞ . The
complex potential of these 2n+1 sources is
158 Partial Differential Equations
n n
−m log(z − 0) − m∑ log(z − ipa) − m∑ log(z + ipa)
W=
=p 1=p 1

∞ 
=-mlogz − m  ∑ log(z − ipa)(z + ipa) 
 p =1 

=-mlogz(z 2 +a 2 )(z 2 + 22 a 2 )...(z 2 + n 2 a 2 )

πz  z 2   z2   z2 
=-mlog 1 + 2  1 + 2 2  ... 1 + 2 2  + constant (1)
a  a  2 a   n a 

πz a πz a  z2 
log =
z log . = log + log , log(z 2 + a 2=
) log a 2 1 + 2  + log a 2 etc
a π a π  a 
πz
Now putting θ = in the infinite product of sinh θ , i.e.
a

 θ2  θ2   θ2 
sinh θ = θ 1 + 2 1 + 2 2  ... 1 + 2 2  ... and taking limit of (1) when n → ∞ and the constant term
 π  2 π   n π 
is neglected, we get

 πz 
W = −m log sinh   (2)
 a 
Hence the result.
Now, we observe that the complex potential for the doublets at the same points and parallel to x-axis, is
∂W
the negative derivative of (2), i.e. W1 = −
∂z
This is because the sources at z=a has the complex potential W1 =
−m log(z − a) and the doublet at z=a,
meiθ m ∂W
parallel to x-axis, has the complex potential=
W1 = = − .
z−a z−a ∂z
∂W ∂  πz 
∴ W1 =− =−  −m log sinh 
∂z ∂z  a 
 πz  π πm πz πz πm
=m  coth  = coth = µ coth , where µ =
 a a a a a a

Hence the result.

4.8 Images in Two Dimensions
If the motion of the fluid in the xy-plane is due to a system of sources, sinks, doublets, vortices etc., and
if there is a curve C in the plane such that there is no flow across it, the system of sources, sinks,
doublets(vortices) etc., is said to be images of sources, sinks, doublets etc., on other side of C.
Obviously, the curve C must be a streamline as there is no flow across it and so if we introduce a rigid
Two-dimensional Motion and Stoke’s Stream Function 159
barrier coincident with C, the fluid motion will not be affected. To discuss the images in two
dimensions, we make use of complex potential.
4.8.1 Image of a Line Source in a Plane (w.r.t. a Straight Line)
Without loss of generality, we take the rigid impermeable plane to be x=0 and perpendicular to the plane
of flow (xy-plane). Thus we are to determine the image of a line source of strength m per unit length
which runs through the point A(a,0) parallel to the rigid plane, w.r.t. the straight line OY. Let us place a
line source of strength m per unit length at A ' (-a,0).
The complex potential at a point P due to the system of line sources
is
W=
−m log(z − a) − m log(z + a) =
−m log(z − a)(z + a)
=-mlog[(r1eiθ1 )(r2 eiθ2 )]
=mlog[r1r2 ei( θ1 +θ2 ) ]

⇒ φ+iψ =-mlog(r1r2 )-im(θ1 + θ2 )

⇒ ψ = −m(θ1 + θ2 )

If P lies on y-axis, then PA = P A ' ⇒ ∠PAA ' =

∠PA 'A
i.e. π − θ1 = θ2 ⇒ θ1 + θ2 = π

Thus, ψ = − mπ = constant which shows that y-axis is a

streamline i.e. there is no flow acreoss anlong OY. In other words,
image of a line source w.r.t. a plane (a streamline) is a line source
of equal strength situated on the opposite side of the plane (stream
line) at an equal distance.
4.8.2 Image of a Line Doublet in a Plane
Let us consider the rigid impermeable plane to be x = 0 and perpendicular to the plane of flow (xy-
plane). Thus we have to determine the image of a line doublet w.r.t. the streamline OY. Let there be line
sources at the points A and B, taken very close together, of strengths − m and m per unit length. Their

respective images in OY are − m at A ' , m at B ' , are the reflections of A, B in OY. The AB makes
  
angle α with OX . Thus A 'B' makes angle (π − α) with OX . In the limiting case, as
m → ∞, AB → 0 , we have equal line doublets at A and A’ with their axis inclined at α , (π − α) to

OX . Hence either of the line doublet is the hydrodynamical image of the other in the infinite rigid
impermeable plane (stream line) x = 0.

Example: A two-dimensional doublet of strength µi per unit length is at a point z = ia in a stream of
velocity − Vi in a semi- infinite liquid of constant density occupying the half plane
160 Partial Differential Equations

y > 0 and having y = 0 as a rigid impermeable boundary, i being unit vector in the positive x-axis.
2 2
Show that the complex potential of the motion is W= Vz + 2µz / (z + a )
2
Also show that for 0 < µ < 4a V , there are no stagnation points on the boundary and that the pressure
on it is a minimum at the origin and maximum at the
points (± a 3,0).

Solution: We know that the image of the line doublet µi

at point A ' (0,a) is a line doublet µi at point A ' (0,a).
The velocity potential due to uniform stream -V i is
Vx=vRe(z) and so its complex potential is the analytic
function Vz.
Therefore, the complex potential of the system is

µ µ
W=
Vz + +
z − ia z + ia
2µz
= Vz + 2 2
= Vz + 2µz(z 2 + a 2 ) −1
z +a
dW
From here, we get = V + 2µ(a 2 − z 2 )(a 2 + z 2 ) −2
dz
On the boundary y = 0 and thus z = x ,

dW dW
therefore q =
− =
− V + 2µ(a 2 − x 2 )(a 2 + x 2 ) −2
=
dz dx
dW
For stagnation points =0
dx
⇒ Vx 4 + 2x 2 (Va 2 − µ) + Va 4 + 2µa 2 = 0 (1)

which is quadratic in x2 whose discriment is

∆ = 4 ( Va 2 − µ ) − V(Va 2 + 2µa 2 )  = 4µ(µ − 4a 2 V)

2

 
2
From here, ∆ < 0 when 0 < µ < 4a V , showing that the quadratic equation (1) has no real root.
Therefore there is no stagnation points on the boundary y=0.
Applying Bernoulli’s equation along the streamline y = 0, we have
Two-dimensional Motion and Stoke’s Stream Function 161
2
p 1 a2 − x2  p 1 2
+ V + 2µ 2 =constant + q =constant
e 2  (a + x 2 ) 2  ρ 2
2
 a2 − x2 
Thus, p is maximum when X=  V + 2µ 2 is minimum and conversely.
 (a + x 2 ) 2 
12
From here, we get X = V + 2µ(a 2 − x 2 )(a 2 + x 2 ) −2
Differentiating w.r.t. x, we get
1 −1 2 dX
X .X ' =−4µx(3a 2 − x 2 )(a 2 + x 2 ) −3 X ' =
2 dx
For extreme value of X, we have X’ = 0 which gives x=
0, ± a 3 .
We observe that X’ changes sign from positive to negative
when x passes through zero and thus X is maximum at x = 0
⇒ p is minimum at x= 0 i.e. at (0,0) i.e. the origin.

Similarly X’ changes sign from negative to positive as x passes through ± a 3 showing that X is
minimum at x = ± a 3 and thus p is maximum at ( ± a 3 ,0).
Theorem: Milne-Thomson Circle Theorem
We construct a theorem due to L.M. Milne theorem to acquire the pictures of specified flow
distributions are impermeable circular cylindrical boundary whose generators extend to infinity in either
direction and are perpendicular to the considered plane of flow.
Statement: Let f(z) be the complex potential for a flow having no rigid boundaries and such that there
are no singularities within the circle z = a . Then on introducing the solid circular cylinder z = a , with
impermeable boundary, into the flow, the new complex potential for the fluid outside the cylinder is
f (z) + f (a 2 / z), z ≥ a
given by W =

Proof: Let C be the cross-section of the cylinder with equation z = 1 .

2 2
Therefore, on the circle C, z = a ⇒ zz = a ⇒ z = a / z

where z is the image (inverse point) of the point z w.r.t. the circle. If z is outside the circle, then
z = a 2 / z is inside the circle. Further, all the singularities of f(z) lie outside C and the singularities of
f (a 2 / z) and therefore, those of f (a 2 / z) lie inside C. Therefore, f (a 2 / z) introduces no singularities
outside the cylinder. Thus, the function f(z) and f(z)+ f (a 2 / z) both have the same analytical
singularities outside C. Therefore, the conditions satisfied by f(z) in the absence of the cylinder are
 162 Partial Differential Equations

satisfied by f(z)+ f (a 2 / z) in the presence of the circular cylinder. Further, the complex potential, after
2
W=
f (z) + f (a / z) =
f (z) + f (z)
insertion of the cylinder z = a , is
= f (z) + f (z)
=a purely real quantity
But we know that W = φ + iψ, it follows that ψ =0 on C,

i.e., C is a streamlines. Hence z = a is a possible boundary for the flow specified the complex
f (z) + f (a 2 / z), z ≥ a.
potential W =

Remark: We note that f (a 2 / z) represents the perturbation to the complex potential f(z) caused by the
insertion of the cylinder z = a . It also gives the distribution of the image system within the cylinders
interior z < a for the actual distribution in the external fluid region z > a .

Exercise: Describe the uniform flow past a fixed infinite circular cylinder.
Solution: We have already dealt with this problem using cylindrical polar co-ordinates. Here, we use the
concept of complex potential.

The velocity potential due to an undisturbed uniform stream having velocity − Ui (U is real) is
Ux = U Re(z). Since z is an analytic function, the corresponding complex potential is
f(z) = Uz

Thus f (z)
= f (z)
= Uz
= Uz
= Uz

 a2   a2 
⇒ f  = u 
z  z 

When the cylinder z = a present, by circle theorem, the complex potential, for the liquid region z ≥ a
, is

W f (z) + f (a 2 / z)
=

a2
i.e. iψ U(z +
φ += )
z
Taking z = reiθ , where r ≥ a , equating real and imaginary parts, we get

 a2 
=φ Re(W)
= U cos θ  r + 
 r 
Two-dimensional Motion and Stoke’s Stream Function 163

 a2 
ψ Im(W)
= = U sin θ  r − 
 r 

2 a2
The perturbation term f (a / z) = U gives the image of the flow in the cylinder. This image
z
represents a uniform line doublet of strength Ua2 per unit length and axis in the direction i .
4.8.2 Image of Line Source in a Circular Cylinder (or in a circle)
Let a line source of strength m per unit length be
present at a point z = d in the fluid, d > a. Let us
then insert a circular cylinder z = a in the fluid.
The complex potential in the absence of cylinder is
− m log(z − d) and after the insertion of cylinder, by
circle theorem , we get

φ + iψ = W = −m log(z − d) − m log(a 2 / z) − d

 −d   a 2 
=− m log(z − d) − m log    − + z  
 z   d 

=−m log(z − d) − m log(z − a 2 / d) + m log z + constant (1)

Ignoring the constant term, we observe from(1) that the complex potential represents a line source at z =
d, another line source at the inverse point z = a 2 / d and an equal line sink at the centre of the circle.
Thus the image of a line source of strength m per unit length at z=d in a cylinder is an equal line source
at the inverse point z = a 2 / d together with an equal line sink at the centre z = 0 of the circle. Equation
(1) can be written as

 y 
φ + iψ = −m[log{(x − d) 2 + y 2 }1 2 + i tan −1  ]
 x −d

  a 2  2 
12
y   y
−1  

−m log  x −  + y  + i tan 
  d 2 
 x − a / d  
+ m 

log ( x 2
+ y )
2 12
+ i tan −1 
x
Here,

y  y  −1  y  −1 y
( )
12
r
= x 2 + y 2 ,= θ tan −1 ⇒ ψ = − m tan −1   − m tan   + m tan
x  x −d
2
 x −a /d x
 y y 
 + 
2 y
− m tan −1  x − d x − a / d  + m tan −1   =
= 0 x 2 + y2 = a 2 ⇒ a 2 − y2 = x 2
y y x
1− 2

 x −d x −a /d 
 164 Partial Differential Equations

y y
−m tan −1
= + m tan −1 =0
x x
Thus, the circular cylinder is a streamline i.e., there is no flow of fluid across the cylinder. Hence the
image of a line source of strength m per unit length at z = d in a circular cylinder is an equal line source
a2
at the inverse point z = w.r.t. the circle z = a together with an equal sink at the centre z = 0 of the
d
circle.
4.8.3 Image of a Line Doublet in a Circular Cylinder (or in a Circle)
Let there be a line doublet of strength
µ per unit length at the point z = d, its axis being inclined at an
angle α with the x-axis. The line doublet is assumed to be
perpendicular to the plane of flow i.e. parallel to the axis of
cylinder. The complex potential in the absence of the cylinder, is
µeie
z−d
When the cylinder z = a is inserted, the complex potential, by
circle theorem, becomes

µeiα µe − iα
W
= +
z − d (a 2 / z) − d

µeiα µe − iα z
= −
z − d d ( z − (a 2 / d) )

µeiα µei( π−α ) µa 2 ei( π−α ) µeiα µzei( π−α )

= + + 2 = + (1)
z−d d d z − (a 2 / d) z − d d ( z − (a 2 / d) )
If the constant term (second term) in (1) is neglected, then the complex potential in (1) is due to a line
doublet of strength µ per unit length at z = d, inclined at an angle α with x-axis and another line
µa 2
doublet of strength 2
2
per unit length at the inverse point z = (a / d) inclined at angle π − α with x-
d
axis. Thus, the image of a line doublet of strength µ per unit length at z = d inclined at an angle α
µa 2 2
with x-axis is a line doublet of strength 2
per unit length at the inverse point z = (a / d) inclined at
d
an angle π−α with x-axix.
Remark: The above three cases, that is flow past a fixed circular cylinder and images in circular
cylinder are applications of Milne-Thomson Circle theorem.
Two-dimensional Motion and Stoke’s Stream Function 165
Extension of the Circle Theorem:
We have an extension of the Milne-Thomson Circle theorem, which finds its applications in problems in
which two dimensional distributions are prescribed within a circular cylindrical boundary as follows:
Let f(z) be the complex potential for a flow having no rigid boundaries and such that there are no
singularities outside the circle z = a . Then on introducing the solid cylinder z = a into the flow and
leaving the distribution of singularities within its interior unchanged, the new complex potential for the
fluid motion within the boundary becomes

f (z) + f (a 2 / z), for z ≤ a

W=
The proof of this result is similar to that of the main theorem.

 a 
Example: A line source and a line sink of equal strength are placed at the points  ± , 0  within a fixed
 2 
circular boundary x 2 + y 2 =
a 2 . Show that the streamlines are given by

 2 a2 2
 r −  ( r − 4a ) − 4a y = ky ( r − a )
2 2 2 2 2

 4 

where x 2 + y 2 =
r 2 and k is constant.

a 
Solution: For a line source of strength m per unit length at  , 0  and a line sink of the same strength
2 
 a 
at  − , 0  , The complex potential in the absence of the circular boundary is
 2 

 a  a
f (z) =
−m log  z −  + m log  z + 
 2  2
 a2   a2 a   a2 a 
⇒f  =
− m log  −  + m log  + 
 z   z 2  z 2

where f (z) = f (z) with m and a is real .

When the circular boundary x 2 + y2 =

a 2 i.e. z = a is inserted the complex potential at points interior
to it by circle theorem becomes

 a2 
w f (z) + f  
=
 z 

  a   a  
+ iψ m log  x + + iy  − log  x − + iy  + log ( 2a + x + iy ) − log ( 2a − x + iy ) 
⇒ φ=
  2   2  
 166 Partial Differential Equations

     
 −1  y   y   y   y 
=⇒ ψ m  tan   − tan −1   + tan −1   + tan −1  
 a a  2a + x   2a − x  
x+  x− 
  2  2 

  y y  
  −  
 y y 
  x−a x+a   − 
m  tan −1  2 2  + tan −1  2a + x 2a − x  
  y2   1− y
2

  1+   
  a2   2 2
4a − x  
x2 −
  4  
  
 −1  4ay   ay 
m  tan  2 2  − tan −1  2  , x 2 + y2 =
r2
  4a − r   r 2 − a  
  4  

 
 
−1  5ay(r 2 − a 2 ) 
= m tan
  a 2
 
 ( 4a − r )  r −  + 4a y 
2 2 2 2 2

  4 
The streamlines are given=
by ψ constant
= c

y(r 2 − a 2 ) 1 c 1
i.e. 2
= tan = −
a 5a m k
(4a 2 − r 2 )(r 2 − ) + 4a 2 y 2
4

2 2 2 a2
i.e (r − 4a )(r − ) + 4a 2 y 2 = ky(r 2 − a 2 )
4
Hence the result.
Example: Within a rigid boundary in the form of the circle (x + α) 2 + (y − 4α) 2 = 8α 2 , there is a liquid
motion due to a line doublet of strength m per unit length at the point (0, 3α) with its axis along axis of y
. Show that the velocity potential is
 4(x − 3α) y − 3α 
φ =µ 2 2
+ 2 2
 (x − 3α) + y x + (y − 3α)  .

Solution: The figure shows the circle with centre C.

if (CA)(CB)=(radius)2 = 8α 2 (1)

Then B is the inverse point of A ( and vice-versa ) w.r.t. the circle of radius 8α 2= 2 2α .
 Two-dimensional Motion and Stoke’s Stream Function 167
Equation of line CA is
4α − 3α
y − 3α = (x − 0) ⇒ y − 3α = − x
−α − 0
⇒ x + y =3α (2)

CA meets x-axis where y =0 ⇒ x =3α

i.e. CA meets x-axis at ⇒ x + y =3α .

Also
( CA )( CB) = α 2 + α 2 . (4α) 2 + (4α) 2 = 2α.4α 2 = 8α 2

Hence B is the inverse point of A w.r.t. the given circle,

and the line doublet at A along y-axis has image as a line
doublet at B. The line doublet at A subtends an angle 1350
with AB and the image line doublet at B subtends an angle
π − 1350 = 450 with AB, so that the line doublet at B is
along x-axis.
The complex potential due to the line doublets at A and B
is
iπ
µe 2 4µei0
=w +
z − 3iα z − 3α
Since the object and image line doublets make supplementary angles with the line CB, therefore, for the
image line doublet

a2
' 8α 2 µ8α 2
µ =µ 2 =µ = 2 =4µ
d (CA) 2 2α
iπ
iµ 4µ
Thus, W = + , e2 =i
x + i(y − 3α) (x − 3α) + iy

iµ [ x − i(y − 3α) ] 4µ(x − 3α) − iy

=φ + iψ +
x 2 + (y − 3α) 2 (x − 3α) 2 + y 2

 µ(y − 3α) 4µ(x − 3α) 

φ  x 2 + (y − 3α) 2 + (x − 3α) 2 + y 2 
 
Hence the result.

Example: Between the fixed boundaries θ = π and θ = − π there is a two-dimensional motion due to a
6 6
source at the point (r = c, θ = α) and a sink at the origin, absorbing water at the same rate at the source
 168 Partial Differential Equations

produced it. Find the stream function and show that one of streamlines is a part of the curve
r 3 sin
= 3α c3 sin 3θ .
Solution: Let the source and sink be of strength m each. To make the geometry of the problem simple,
we make use of conformal mapping.

Let the transformation be t = z3 , where

= iθ
z re= , t Reiλ

To get π insted of π .
2 6

Thus Reiλ = (reiθ )3

⇒ R = r 3 , λ = 3θ

So, the boundary θ = ± π of the z-plane becomes λ = ± π of the t-plane. Accordingly, the point (c, α) of
6 2
z-plane becomes (c ,3α) in the t-plane.
3

We note that the mapping under consideration is conformal. Also, image of the source of strength m at
A (c3 ,3α) is the source of strength m at A ' (c3 , π − 3α) and the image of the sink of strength m at (0,0) is a
sink of strength m at (0,0).
The complex potential of the system is

−m log ( t − c3e3iα ) + m log t − m log ( t − c3ei( π−3α ) ) + m log t

W=

= 2m log t − m log ( t − c3e3iα ) − m log ( t + c3e −3iα )

= 2m log z3 − m log ( z3 − c3e3iα ) − m log ( z3 + c3e −3iα )

= 6m log(reiθ ) − m log ( r 3e3iθ − c3e3iα ) − m log ( r 3e3iθ + c3e −3iα )

φ + iψ 6m(log r + iθ) − m log ( r 3e3iθ − c3e3iα )( r 3e3iθ + c3e −3iα ) 

⇒=

= 6m(log r + iθ) − m log ( r 6 cos 6θ − c6 + 2c3r 3 sin 3α sin 3θ ) + i ( r 6 sin 6θ − c6 − 2c3r 3 sin 3α cos3θ ) 

 ( r 6 sin 6θ − c6 + 2c3 r 3 sin 3α cos 3θ ) 

−1
= 6mθ − m tan  6
ψ 
 ( r cos 6θ − c + 2c r sin 3α sin 3θ ) 
6 3 3

which is the required stream function.

For the required streamline, we take ψ =0 to get

 ( r 6 sin 6θ − c6 + 2c3 r 3 sin 3α cos 3θ ) 

−1
6θ − tan  6  =0
 ( r cos 6θ − c + 2c r sin 3α sin 3θ ) 
6 3 3
Two-dimensional Motion and Stoke’s Stream Function 169

sin 6θ r 6 sin 6θ − 2c3r 3 sin 3α cos 3θ

⇒ =
cos 6θ r 6 cos 6θ − c6 + 2c3r 3 sin 3α sin 3θ
6
⇒ c= sin 6θ 2c3 r 3 sin 3α(cos 6θ cos 3θ + sin 6θ sin 3θ)

⇒ 2c6 sin 3=
θ cos 3θ 2c3r 3 sin 3α cos 3θ
⇒ 2c3 cos 3θ(c3 sin 3θ − r 3 sin 3α) = 0

π π
But cos 3θ = 0 ⇒ 3θ = ± i.e. θ = ±
2 6
i.e. the boundary itself.
so, c3 sin 3θ − r 3 sin 3α = 0

c3 sin
= 3θ r 3 sin 3α

Hence the result.

4.9 Blasius Theorem
When a cylinder with any cross-section and infinite length is arranged with its generators perpendicular
to the plane of flow of a two-dimensional fluid motion, the fluid motion causes the cylinder to translate
and rotate. We now establish a theorem due to Blasius which enables us to evaluate the resultant force
and couple exerted on unit length of the cylinder by the moving fluid.
Statement: In a steady two-dimensional irrotational flow given by the complex potential
W = f(z) if the pressure forces on the fixed cylindrical surface having cross-section C are represented by
a force (X,Y) per unit length and a couple of moment M per unit length about the origin of co-ordinates,
then neglecting the external forces,

iρ  dW 2  ρ  dW 2 
X −iY = ∫   dz M = Real part of  − ∫ z
2 C  dz   dz 
 2 C  dz  
 

where ρ is the density of the fluid

Proof: Let ds be an element of arc of C at a point P(x,y) and the tangent at P makes an angle θ with the
x-axis. If p denotes the fluid pressure at P, the force of unit length of the section s of the cylinder is pds
normal to C(inwards).The components of this force along the co-ordinate axes are pds cos(90+ θ ), pds
cos θ
i.e. -pds sin θ , pds cos θ
The force per unit length on the element ds is
 170 Partial Differential Equations

pds sin along negative x − axis

dF =+
dX idY =
−psin θds + ip cos θds
−pds sin along postive x − axis

= ip(cos θ + isin θ)ds

 dx dy  dx dy
= ip  + i  ds cos θ
{= ,sin
= θ }
 ds ds  ds ds
= ip(dx+idy)=ipdz (1)
p 1 2
The pressure equation, in the absence of external forces, is + q =constant
ρ 2

1
⇒ p =− ρq 2 + k (2)
2
Further

dW
= −u + iv = −q cos θ + iq sin θ = −q(cos θ − sin θ) = −qe − iθ (3)
dz
 dx dy 
and dz
= dx + idy
=  + i  ds= (cos θ + i sin θ)ds= eiθds (4)
 ds ds 
The force per unit length on the cylinder is obtained by integrating (1). Therefore,

F =X + iY =∫ ipdz
C

1
∫ i(k − 2 ρq )dz
2
=
C

ip 2
2 ∫C
= − q dz {  ∫ dz = 0 }
C

ip 2 iθ
2 ∫C
= − q e ds

ip
From here; X − iY =∫ q 2 e − iθds
2C

ip
=
2C∫ (q 2e −2iθ )eiθds

2
ip  dW  ( using (3) and (4) )
= ∫  dz
2 C  dz 
 Two-dimensional Motion and Stoke’s Stream Function 171

The moment M of the cylinder per unit length is M = ∫ r × dF = ∫ ( pdssin θ ) y + (pds cos θ)x 
C C

i j k
  dy   dx    
∫C p  ds  yds + p  ds  xds  r × dF = x y 0
− pds sin θ pds cos θ 1

= ∫ p(xdx + ydy)
C

 1 
∫  k − 2 ρq
2
=  (xdx + ydy)
C 

1  ρ
= k ∫ d  (x 2 + y 2 )  − ∫ q 2 (xdx + ydy)
C 2  2C

ρ
− ∫ q 2 (xdx + ydy)
= 1st integral vanishes.
2C

−ρ 2 dx
= cos θds
2 ∫C
= q (x cos θ + ysin θ)ds
dy
= sin θds

 −ρ 
Re  ∫ q 2 (x + iy)(cos θ − isin θ)ds 
 2 C 
 ρ 
Re  − ∫ q 2 ze − iθds 
 2 C


 −e 
= Re  ∫ z(q 2e −2iθ )eiθds 
2 C 
 −e  dW  2 
= Re  ∫ z   dz 
 2 C  dz  
Hence the theorem.
4.10 Check Your progress
1. Derive the expression for stream function.
2. Find the complex velocity potential due to doublet.
3. State and prove Milne-Thomson Circle Theorem.
172 Partial Differential Equations

4. Discuss about images of source, sink and doublet in three-dimension.

5. Show that the velocity components given by

 b2 ( x2 − y 2 )   2 
u= U 1 − 2
ay
+ , v =
U  ax + 2b xy  represents a possible fluid motion in
 x + y 2 ( x 2 + y 2 )2   x 2 + y 2 ( x 2 + y 2 )2 
   
two dimension.
6. A source and sink of equal strength are placed at the point (a,0), (-a,0) respectively, within a fixed
circular boundary z = 2a . Show that the streamlines are given by

16a 2 y 2 + ky(r 2 − 4a 2 ) =(r 2 − 16a 2 )(r 2 − a 2 ) , k being constant.

4.11 Summary
In this section, we have studied about two-dimensional motion, stream function and discuss it for source,
sink and doublet, irrotational motion in two-dimension, complex potential and derive expression for it in
various cases, Milne-Thomson Circle, Blasius and Theorems
4.12 Keyword
Two-dimensional motion, stream function, complex potential, images, sources, sinks, and doublets
4.13 Self-Assessment Test
1. Find the stream function ψ ( x, y, t ) for the given velocity field
= u Ut
= , v x.

2. Show that u= 2cxy, v= c(a 2 + x 2 − y 2 ) are the velocity components of a possible fluid motion.
Determine the stream function.
3. Show that the stream function and velocity potential for a two-dimensional irrotational motion satisfy
Laplace’s equation.
4. Show how the two-dimensional irrotational motion of a liquid may be described by a complex
potential w=f(z). Find the velocity in terms of W.
4.14 Suggested Reading
1. F. Chorlton, Text Book of Fluid Dynamics, C.B.S. Publishers, Delhi, 1985
2. O’Neill, M.E. and Chorlton, F., Ideal and Incompressible Fluid Dynamics, Ellis Horwood Limited,
1986.
░⍟⍟⍟░