Chapter-5: Formulation and Solution of Elasticity Problems

Example -1:
Determine the appropriate stress function for:
(i) Uni-axial state of stress, (ii) Pure shear, (iii) Pure bending

Solution
(i) Uni-axial State of Stress
𝑦
𝑎 𝑙

𝑑 𝜎 𝜎
𝜎 𝜎

𝑦
𝑥 𝑏
𝑥

Let (𝑥, 𝑦) = 𝑐 . 𝑦  2nd degree polynomial

which satisfies ∇  = 0
Hence, stress components are:

𝜎 = = 2𝑐 = 𝜎  𝑐 =
 
𝜎 = =0, 𝜎 =− .
=0
therefore,
(𝑥, 𝑦) = 𝑐 . 𝑦 = .𝑦
(ii) Pure Shear
𝑦
𝑥
𝜏

𝜎
𝜏
𝜎 𝜏
𝑦
𝜏
𝑥

Let (𝑥, 𝑦) = 𝑏 . 𝑥. 𝑦  2nd degree polynomial

which satisfies ∇  = 0
Hence, stress components of element are:
 
𝜎 = =0, 𝜎 = =0

𝜎 =− .
= −𝑏 = 𝜏  𝑏 = −𝜏

therefore,
(𝑥, 𝑦) = 𝑏 . 𝑥. 𝑦 = −𝜏 . 𝑥. 𝑦

1
(iii) Pure Bending
𝑦

𝑀 +𝑐 𝜎 𝑀
𝑦 𝑥
−𝑐
𝑧

Let (𝑥, 𝑦) = 𝑑 . 𝑦  3rd degree polynomial

which satisfies ∇  = 0
Hence, stress components of beam are:

𝜎 = = 6𝑑 . 𝑦
 
𝜎 = =0, 𝜎 =− .
=0

-Moment of beam from summation of moment caused by 𝜎 about neutral-axis (N.A.),

𝑀 =∫ 𝜎 . 𝑦. 𝑑𝐴
-substituting 𝜎 = 6𝑑 . 𝑦
 𝑀 = 6𝑑 . ∫ 𝑦 . 𝑑𝐴

Noting the ∫ 𝑦 . 𝑑𝐴 = 𝐼 –moment of inertia of beam section

 𝑀 = 6𝑑 . 𝐼  𝑑 =
therefore, the normal stress of beam is
𝜎 = 6𝑑 . 𝑦 = .𝑦  the same as bending stress from elementary analysis
and, stress function of beam subjected to pure bending is
(𝑥, 𝑦) = 𝑑 . 𝑦 = .𝑦

2
Example -2: Bending of a Cantilever Beam Loaded at End
Consider a cantilever beam having a narrow rectangular cross-section of unit width bent by a
force 𝑃 applied at the end as shown below. The upper and lower edges are free from load and
shearing forces, having a resultant 𝑃 distributed along the end at 𝑥 = 0.
1 𝑙

𝑐 𝐴
2𝑐 𝑥
𝑐

Solution
-Stress function satisfying ∇  = 0 and boundary condition of cantilever is obtained by
superposition of second and fourth degree polynomials retaining one term each as:
 =  +  = 𝑏 . 𝑥. 𝑦 + 𝑑 . 𝑥. 𝑦
then, the stress components of cantilever are
𝜕2  𝜕2 
𝜎 = = 6𝑑 . 𝑥. 𝑦, 𝜎 = =0
𝜕𝑦2 𝜕𝑥2
𝜕2 
𝜎 =− = (−𝑏 − 3𝑑 . 𝑦 )
𝜕𝑥.𝜕𝑦

-To have the longitudinal sides at 𝑦 = ±𝑐 free from forces (stress boundary conditions)
𝜎 = (−𝑏 − 3𝑑 . 𝑐 ) = 0
±

 𝑑 =− and 𝜎 = −𝑏 + .𝑦
-To satisfy the stress boundary condition on the loaded end, the sum of the shearing forces
distributed over this end must be equal to 𝑃. Hence,
.
− ∫ 𝜎 . 𝑑𝑦 = ∫ 𝑏 − .𝑦 . 𝑑𝑦 = 𝑏 . 𝑦 − . = 2𝑏 . 𝑐 − =𝑃

 𝑏 = then 𝑑 =−
Substituting values of 𝑏 and 𝑑 into stress components of cantilever,
𝜎 =− . 𝑥. 𝑦, 𝜎 =0
𝜎 = − + .𝑦 =− . (𝑐 − 𝑦 )

Noting that 𝑐 = 𝐼 (moment of inertia of the cross-section), we have

. .
𝜎 =− , 𝜎 =0
𝜎 = − × (𝑐 − 𝑦 )
These equations of stresses coincide completely with the elementary solution as given in
Strength of Materials.

3
Let us consider now the displacement corresponding to the stresses. Applying Hooke’s law,
we find strain components
. . . . .
𝜀 = = =− , 𝜀 = = − . =
. .

𝛾 = + = =− . (𝑐 − 𝑦 )
.

-To determine the components 𝑢 and 𝑣 of the displacement, integrating 𝜀 and 𝜀 with
respect to 𝑥 and 𝑦, we find
. . . .
𝑢 = ∫ 𝜀 . 𝑑𝑥 = ∫ . 𝑑𝑥 = − + 𝑓 (𝑦 ) ,
. .
. . . . . .
𝑣 = ∫ 𝜀 . 𝑑𝑦 = ∫ . 𝑑𝑦 = + 𝑔(𝑥 )
. .
-Substituting u and v in equation of 𝛾 , we find
. [ ( )] . . [ ( )]
+ + + = . (𝑐 − 𝑦 )
. . .
. [ ( )] [ ( )] . . . .
Let 𝐹 (𝑥 ) = + , 𝐺 (𝑦 ) = + − and 𝐾=
. . . .

𝑑 𝑒
then, the above equation may be written
𝐹 (𝑥) + 𝐺 (𝑦) = 𝐾  constant!
-Such equation means that 𝐹(𝑥) must be some constant ‘𝑑’ and 𝐺(𝑦) some constant of ‘𝑒’.
.
Thus, (𝑑 + 𝑒) = ---(1)
.
[ ( )] . [ ( )] . . .
and, = + 𝑑, = + +𝑒
. . .

-Integrating these equations, functions 𝑓(𝑦) and 𝑔(𝑥) are then

. . . . . .
𝑓(𝑦) = ∫ + + 𝑒 . 𝑑𝑦 = + + 𝑒. 𝑦 + 𝑔
. . . .
. .
𝑔 (𝑥 ) = ∫ + 𝑑 . 𝑑𝑥 = + 𝑑. 𝑥 + ℎ
. .
-Substituting in the expression of 𝑢 and 𝑣, we find
. . . . .
𝑢= − + + 𝑒. 𝑦 + 𝑔
. . .
. . . .
𝑣= + + 𝑑. 𝑥 + ℎ ---(2)
. .
-The constants 𝑑, 𝑒, 𝑔, ℎ may now be determined from Eq.(1) and from the three conditions
of constraints (displacement boundary conditions) which are necessary to prevent the beam
from moving as a rigid body in the 𝑥-𝑦 plane. Assume that point 𝐴, the centroid of the end
cross-section is fixed. Then 𝑢 and 𝑣 are zero for 𝑥 = 𝑙, 𝑦 = 0, and we find from Eqs.(2),
.
𝑔 = 0, ℎ= − 𝑑. 𝑙
.
. .
(𝑣 ) = − − 𝑑. (𝑙 − 𝑥)  at center-line of beam!
. .
. .
=0  𝑑=− and ℎ=
, . .

.
=0  𝑒=
, .

4
Substituting all the constants into Eqs.(2), we find
. . . . . .
𝑢= − + + .𝑦
. . . .
. . . . . . .
𝑣= + − +
. . . .
-The equation of deflection curve at center-line of beam is
. . . .
(𝑣 ) = − +
. . .
.
which gives for the deflection at the loaded end (𝑥 = 0) the value . This value coincides
.
with the value derived in the elementary books on the Strength of Materials.

Example -3: Bending of a Simply Supported Beam Loaded by Uniform Load

Consider a beam of narrow rectangular cross-section of unit width, supported at the ends, be
bent by a uniformly distributed load of intensity ‘𝑞’ as shown below. The stress boundary
conditions at the upper and lower edges of beam are,
𝜎 = 0, 𝜎 = 0, 𝜎 = −𝑞 (1)
±

The stress boundary conditions at the ends 𝑥 = ±𝑙 are:

∫ 𝜎 . 𝑑𝑦 = ∓𝑞. 𝑙, ∫ 𝜎 . 𝑑𝑦 = 0, ∫ 𝜎 . 𝑦. 𝑑𝑦 = 0 (2)

𝒒 1 𝑞

𝑐 𝑥
2𝑐
𝒒. 𝒍 𝑐 𝒒. 𝒍
𝑞 ⁄2
𝑙 𝑙
Distribution of 𝝈𝒚𝒚 (comp.)
𝑦

Solution
-Stress function of beam is obtained by superposition of second, third and fifth degrees
polynomials retaining some of the terms as:
 =  +  +  = 𝑎 .𝑥 + 𝑏 .𝑥 .𝑦 + 𝑑 .𝑦 + 𝑑 .𝑥 .𝑦 + 𝑓 .𝑦
with 𝑎 = 𝑏 = 𝑐 = 𝑒 = 0, 𝑏 = −(5𝑓 + 𝑑 ) = 0  𝑓 = − 𝑑 ⁄5
  = 𝑎 .𝑥 + 𝑏 .𝑥 .𝑦 + 𝑑 .𝑦 + 𝑑 .𝑥 .𝑦 − 𝑑 .𝑦
then, the stress components of beam are:
𝜕2  𝜕2 
𝜎 = = 6𝑑 . 𝑥 . 𝑦 − 4𝑑 . 𝑦 + 6𝑑 . 𝑦, 𝜎 = = 2𝑑 . 𝑦 + 2𝑏 . 𝑦 + 2𝑎
𝜕𝑦2 𝜕𝑥2
𝜕2 
𝜎 =− = −6𝑑 . 𝑥. 𝑦 − 2𝑏 . 𝑥
𝜕𝑥.𝜕𝑦

From condition (1), we find

−6𝑑 . 𝑐 − 2𝑏 = 0
2𝑑 . 𝑐 + 2𝑏 . 𝑐 + 2𝑎 = 0
−2𝑑 . 𝑐 − 2𝑏 . 𝑐 + 2𝑎 = −𝑞

5
from which
𝑎 =− , 𝑏 = , 𝑑 =−
-Substituting in equation of stress components of beam
𝜎 =− .𝑥 .𝑦 + . 𝑦 + 6𝑑 . 𝑦 = − . 𝑥 .𝑦− 𝑦 + 6𝑑 . 𝑦

𝜎 =− .𝑦 + .𝑦+− = − . 𝑦 − 𝑐 .𝑦 + 𝑐

𝜎 = . 𝑥. 𝑦 − .𝑥 = − . (−𝑦 + 𝑐 ). 𝑥

-Noting that 𝑐 = 𝐼, we find

𝜎 =− . 𝑥 .𝑦 − 𝑦 + 6𝑑 . 𝑦  ‘d3’ has to be determined from other condition

𝜎 =− . 𝑦 − 𝑐 .𝑦 + 𝑐

𝜎 =− . (𝑐 − 𝑦 ). 𝑥
Distribution of 𝜎 (which is compressive) over depth of beam are −𝑞 at top, −𝑞/2 at neutral
axis and zero at bottom. But in elementary analysis, 𝜎 is zero. Value of shear stress
component 𝜎 coincides with that given by elementary theory.
It can be checked that these stress components satisfy the first two conditions of (2) at ends.
To determine 𝑑 from the condition at 𝑥 = ±𝑙 as (third condition of 2):
∫ 𝜎 . 𝑦. 𝑑𝑦 = ∫ − .𝑥 .𝑦 + . 𝑦 + 6𝑑 . 𝑦 . 𝑦. 𝑑𝑦 = 0

∫ − .𝑥 .𝑦 + . 𝑦 + 6𝑑 . 𝑦 . 𝑑𝑦 = − .𝑥 .𝑦 + . + 2𝑑 . 𝑦 =0

@ 𝑥 = ±𝑙,  − . 𝑙 × 2𝑐 + × + 2𝑑 × 2𝑐 =0
. .
 − + + 4𝑑 . 𝑐 =0

 𝑑 = . −
Hence, finally
𝜎 =− .𝑥 .𝑦 + .𝑦 + . − .𝑦

 𝜎 = . (𝑙 − 𝑥 ). 𝑦 + . 𝑦 − 𝑐 .𝑦

-with 𝐼 = ,
 𝜎 = . (𝑙 − 𝑥 ). 𝑦 + . 𝑦 − 𝑐 .𝑦
Elementary analysis Necessary correction
-The first term in this expression represents the stresses given in elementary theory of
bending, and the second term gives the necessary correction. This correction does not depend
on ‘𝑥’ and is small in comparison with the maximum bending stress, provided the span of the
beam is large in comparison with its depth. For such beams, elementary theory of bending
gives sufficiently accurate values of the stress 𝜎 .

6
-The displacement 𝑢 and 𝑣 from integration of 𝜀 and 𝜀 obtained from Hooke’s law as:
𝑢= . 𝑙 .𝑥 − 𝑥 . 𝑦 + 𝑥. 𝑦 − 𝑐 . 𝑦 + . 𝑥. 𝑦 − 𝑐 .𝑦 + 𝑐 + 𝑓 (𝑦)

𝑣=− . 𝑦 − 𝑐 . 𝑦 + 𝑐 . 𝑦 + . (𝑙 − 𝑥 ). 𝑦 + 𝑦 − 𝑐 . 𝑦

− . 𝑙 .𝑥 − 𝑥 − 𝑐 .𝑥 + 1 + .𝑐 .𝑥 +
where 𝜹 is deflection at center of beam
Assuming that at the centroid of the middle cross-section ( 𝑥 = 0 , 𝑦 = 0 ) the horizontal
displacement is zero and vertical displacement is equal to the deflection 𝜹 . But due to
. .
presence of compressive stress at center line 𝜎 = − , (𝑢) = not equal to zero.
-Equation of deflection curve at center line of beam,

(𝑣 ) = 𝛿− . 𝑙 .𝑥 − 𝑥 − 𝑐 .𝑥 + 1+ .𝑐 .𝑥
Assuming that the deflection is zero at the ends (𝑥 = ±𝑙) of the center line, we have
. 
𝛿= . 1+ . + = (𝑣 ) ,

Correction
The factor before the brackets is the deflection from elementary analysis and the second term
in the brackets represents the correction usually called the effect of shearing forces.
By differentiating 𝑣 twice with respect to 𝑥 , we find the following expression for the
curvature of beam:

= . +𝑐 . +
Correction

It is seen from above equation that the curvature is not exactly proportional to the bending
moment, 𝑀 = 𝑞. (𝑙 − 𝑥 )⁄2. The additional term in the brackets represents the necessary
correction to the usual elementary analysis.

7
Example -4: Bar under Own Weight – 3D Problem
Consider a prismatic bar hanging under its own weight as shown below. Determine the
distribution of stress, strain and displacement in the bar due to own weight. The body force
acting on the bar is given by: 𝑧
𝑏 = +𝛾. 𝑘⃗ = . 𝑘⃗
where 𝛾 - unit weight of material of bar (𝑁⁄𝑚 )
𝑙
then, body force potential by 𝑏 with respect 𝑧 𝑎 𝑏
𝑉 = ∫ 𝛾. 𝑑𝑧 = 𝛾. 𝑧
therefore, the stress distribution of bar is 𝑦
𝑂
𝜎 =𝜎 =𝜎 =𝜎 =𝜎 =0
𝑥
𝜎 = 𝛾. 𝑧
Boundary conditions are:
(𝑆 ) = 0, (𝑆 ) = 𝛾. 𝑙
where 𝑆 = 𝜎 . 𝑛 = 𝜎 and 𝑛=1
and, the resultant of uniform stress at support is
∫ (𝑆 ) . 𝑑𝐴 = ∫ (𝛾. 𝑙 ). 𝑑𝐴 = 𝛾. 𝑎. 𝑏. 𝑙
To determine the displacement components, we have the following strain distributions of bar
from generalized Hooke’s law
 .
𝜀 = = − .𝜎 =− .𝑧 , 𝛾 = + =0
 .
𝜀 = = − .𝜎 =− .𝑧 , 𝛾 = + =0

𝜀 = = = .𝑧 , 𝛾 = + =0
By integrating and employing boundary conditions, we can get the following displacement
components as:
. .
𝑢 = ∫ 𝜀 . 𝑑𝑥 = − . 𝑥. 𝑧 + 𝑓(𝑦, 𝑧) = − . 𝑥. 𝑧
. .
𝑣 = ∫ 𝜀 . 𝑑𝑦 = − . 𝑦. 𝑧 + 𝑔(𝑥, 𝑧) = − . 𝑦. 𝑧
.
𝑤 = ∫ 𝜀 . 𝑑𝑧 = . 𝑧 + ℎ(𝑥, 𝑦) = .𝑧 + . (𝑥 + 𝑦 ) − .𝑙
.
where 𝑓 (𝑦, 𝑧) = 0 , 𝑔(𝑥, 𝑧) = 0 and ℎ(𝑥, 𝑦) = . (𝑥 + 𝑦 ) − .𝑙 are obtained
from boundary conditions.

8
Example -5: Pure Bending of Curved Bar
Consider a rectangular bar on inner radius ‘𝑎’ and outer radius ‘𝑏’ with cross-section of unit
thickness subjected to bending moment ‘𝑀’ applied at ends as shown below.

(𝑏 − 𝑎)
𝑀 𝑀
𝑎 𝑟
𝑏 1
𝜃

𝑂
Fig.: Bending of Curved Bar

Solution
The following boundary condition must be satisfied:
(i) (𝜎 ) , = 0 (stress free surfaces)
(ii) ∫ 𝜎 . 𝑟. 𝑑𝑟 = −𝑀 (ends of bar) ---(1)
Here components of stresses are independent of the variable 𝜃 . Hence, the bi-harmonic
equation becomes:
     
∇ = + . . + . = + . − . + . =0 ---(2)
and, stress components from Airy’s stress function becomes:
 
𝜎 = . , 𝜎 = and 𝜎 =0 ---(3)
The general solution of bi-harmonic equation is:
 = 𝐴. log 𝑟 + 𝐵. 𝑟 . log 𝑟 + 𝐶. 𝑟 + 𝐷 ---(4)
and, satisfy compatibility condition ∇  = 0.
Therefore, the stress components are

𝜎 = . = + 𝐵. (1 + 2 log 𝑟) + 2𝐶

𝜎 = =− + 𝐵. (3 + 2 log 𝑟) + 2𝐶 ---(5)
𝜎 =0
Substituting Eq.(5) in boundary conditions [Eqs.(1)], we get
(𝜎 ) =0 + 𝐵. (1 + 2 log(𝑎)) + 2𝐶 = 0

(𝜎 ) =0 + 𝐵. (1 + 2 log(𝑏)) + 2𝐶 = 0 ---(6)

   
and, ∫ 𝜎 . 𝑟. 𝑑𝑟 = ∫ . 𝑟. 𝑑𝑟 = ∫ 𝑟. . 𝑑𝑟 = 𝑟. −∫ . 𝑑𝑟 = −𝑀

On account of Eq.(6),

𝑟. =0, ∫ 𝜎 . 𝑟. 𝑑𝑟 = 0 − || = −𝑀

 𝑀 = || = 𝐴. log(𝑏⁄𝑎 ) + 𝐵. (𝑏 . log 𝑏 − 𝑎 . log 𝑎) + 𝐶. (𝑏 − 𝑎 ) ---(7)

9
Solving Eqs.(6) and Eq.(7) simultaneously, we get
𝐴=− . 𝑎 . 𝑏 . log , 𝐵=− . (𝑏 − 𝑎 )
and 𝐶= . [𝑏 − 𝑎 + 2(𝑏 . log 𝑏 − 𝑎 . log 𝑎)]
where 𝐾 = (𝑏 − 𝑎 ) − 4𝑎 . 𝑏 . (log 𝑏⁄𝑎 )
therefore, substituting constants the expressions of stress components are
.
𝜎 =− . . log + 𝑏 . log + 𝑎 . log
.
𝜎 =− . − . log + 𝑏 . log + 𝑎 . log + 𝑏 − 𝑎
𝜎 =0

Exercise Problem:
A curved bar is of rectangular cross-section, 100150mm. The radius of its inner surface is
200mm. Calculate the radial and hoop stresses in the bar at the inside surface when subjected
to a bending moment of 5kN-m. 𝑎 = 200𝑚𝑚, 𝑏 = 200 + 150 = 350𝑚𝑚.
Ans.: 𝐾 = 0.668 × 10 𝑚 , 𝜎 (𝑟 = 𝑎) = 0, 𝜎 (𝑟 = 𝑏) = 0,
𝜎 (𝑟 = 𝑎) = 1,634.9 𝑘𝑁 ⁄𝑚 , 𝜎 (𝑟 = 𝑏) = −1,129.7 𝑘𝑁⁄𝑚
.
For straight beam, 𝜎 = = 13,333.33 𝑘𝑁 ⁄𝑚  It is much larger than value of curved bar!

Example -6:
Determine the constants of stress function  = 𝐴. 𝑟 + 𝐵. 𝑟 + 𝐶. 𝑟 + 𝐷 to serve as stress
function for an axi-symmetric continuum. Also obtain the expression for stress components.
Solution
-derivatives of the stress function,
   
= 3𝐴. 𝑟 + 2𝐵. 𝑟 + 𝐶 , = 6𝐴. 𝑟 + 2𝐵 , = 6𝐴 , =0
-Stress function must satisfy ∇  = 0 ,
   
∇ = + . − . + . =0
Substituting the derivatives,
 ∇  = 0 + × (6𝐴) − × (6𝐴. 𝑟 + 2𝐵 ) + × (3𝐴. 𝑟 + 2𝐵. 𝑟 + 𝐶 ) = 0

 ∇ = + =0  = 0 and =0
thus, the constants that satisfy the compatibility condition ∇  = 0 are
𝐴 = 0 and 𝐶 = 0
then  = 𝐵. 𝑟 + 𝐷
Therefore, stress components from Airy’s stress function are

𝜎 = . = . (2𝐵. 𝑟) = 2𝐵 ,

𝜎 = = 2𝐵 , and 𝜎 =𝜎 =0
The constant 𝐵 is obtained applying stress boundary conditions.

10
Example -7: Plate with Hole Subjected to Pure Tension
In a tensile field in which there is no hole 𝜎 = 0, 𝜎 = 0 and  = 𝜎. 𝑦 ⁄2 and hence

𝜎 = =𝜎

Fig-1a: Fig-1b:

This condition can be applied to the plate shown in Fig-1a where the effect of the hole will be
negligible at a large distance ‘𝑏’ from it. Assume the distance is defined by a circle radius ‘𝑏’
and let the radius of the hole be ‘𝑎’.
In polar coordinates (𝑦 = 𝑟. sin 𝜃)
.
 = 𝜎. 𝑟 . 𝑠𝑖𝑛 𝜃 = . (1 − cos 2𝜃)
Then, the stress components using Airy’s stress function are given by

𝜎 = = . (1 − cos 2𝜃)
 
𝜎 = . + . = . (1 + cos 2𝜃)
 
𝜎 = . − . = − . sin 2𝜃
.
So for the point defined by 𝑔ℎ𝑗 at the perimeter of the circle radius ‘𝑏’ the stresses are as
shown in Fig-1b.
Note that 𝜎 has two terms
𝜎 = + . cos 2𝜃
that is, at the boundary of the circle 𝑟 = 𝑏, the stresses consist of a constant radial stress
(𝜎⁄2) applied to a thick walled plate (axial symmetry) and a part which varies with cos 2𝜃.
For this a stress function has to be found. The stresses due to both parts for 𝑎 ≤ 𝑟 ≤ 𝑏 can be
superimposed to solve this problem of a plate having a hole and subjected to tension. For the
second part, assume the stress function
 = 𝑓(𝑟). cos 2𝜃
where 𝑓(𝑟) is a function of 𝑟 only.
(Note that 𝜎 and 𝜎 vary with cos 2𝜃 and 𝜎 vary with sin 2𝜃 . Thus the assumption that
 = 𝑓 (𝑟). cos 2𝜃 is quite reasonable.)
  
Use the compatibility equation + . + . . + . + . = 0 to find 𝑓(𝑟)
   ( ) ( ) ( )
+ . + . = . cos 2𝜃 + . . cos 2𝜃 − . cos 2𝜃

11
since cos 2𝜃 ≠ 0, hence solve the compatibility equation ∇  = 0
( ) ( ) ( )
+ . − + . − =0
Solving by any standard matematical technique
𝑓(𝑟) = 𝐴 + 𝐵. 𝑟 + 𝐶. 𝑟 +
so that the stress function due to second part is
 = 𝐴 + 𝐵. 𝑟 + 𝐶. 𝑟 + . cos 2𝜃
is a satisfactory stress function since it fulfils all the necessary conditions.
The stress components due to second part are, consequently

𝜎 = = 2𝐵 + 12𝐶. 𝑟 + . cos 2𝜃
 
𝜎 = . + . =− + 2𝐵 + . cos 2𝜃

𝜎 =− . . = − + 2𝐵 + 6𝐶. 𝑟 − . sin 2𝜃
To determine 𝐴, 𝐵, 𝐶 and 𝐷 put in boundary conditions at 𝑟 = 𝑏
𝜎 = . cos 2𝜃 -for all values of 𝜃
𝜎 = − . sin 2𝜃 -for all values of 𝜃
and, at 𝑟 = 𝑎 (unloaded boundary)
𝜎 =0
𝜎 =0
Hence, for 𝜎 at 𝑟 = 𝑏 and 𝑟 = 𝑎
=− + 2𝐵 +
0=− + 2𝐵 + and, for 𝜎 at 𝑟 = 𝑏 and 𝑟 = 𝑎
− = − + 2𝐵 + 6𝐶. 𝑏 −
0= − + 2𝐵 + 6𝐶. 𝑎 −
In matrix form
𝐴
⎡− −2 0 − ⎤⎧ ⎫ ⎧ ⎫
⎢ 2 0 ⎥⎪⎪𝐵 ⎪⎪ ⎪⎪ 0⎪⎪
⎢ ⎥ =
⎢− 2 6𝑏 − ⎥ ⎨ 𝐶 ⎬ ⎨− ⎬
⎢ ⎥⎪⎪ ⎪ ⎪ ⎪⎪ ⎪ ⎪
⎣− 2 6𝑎 − ⎦ ⎩𝐷 ⎭ ⎩ ⎭
0
Now make the plate very large by making 𝑏 → ∞. Then
.
𝐴= 𝐵=−
.
𝐶=0 𝐷=−
Hence, the stress components due to second part are
𝜎 =− . 1+3 . cos 2𝜃

𝜎 = . 1−4 +3 . cos 2𝜃 ---(1)

𝜎 = . 3 −2 − 1 . sin 2𝜃

12
To these stresses, the stresses due to the constant radial stress applied at 𝑟 = 𝑏 are added and
puting 𝑎⁄𝑏 = 0 by making 𝑏 → ∞, for example from theory of thick walled plates
𝜎 =𝐴− and 𝜎 =𝐴+
so that with the given boundary conditions at 𝑟 = 𝑎, 𝜎 = 0 and 𝑟 = 𝑏, 𝜎 = 𝜎⁄2 then
𝐴=
and 𝐵 = .𝑎 when 𝑎⁄𝑏 → 0.
Thus, the stress components due to constant radial stress are
𝜎 = . 1−

and 𝜎 = . 1+
when added to Eqs.(1) give for the total stresses of plate with hole subjected to pure tension
𝜎 = . 1− + . 1−4 +3 . cos 2𝜃

𝜎 = . 1+ − . 1+3 . cos 2𝜃 ---(2)

𝜎 = . 3 −2 − 1 . sin 2𝜃
Equations (2) show how local the disturbance is due to a hole in an otherwise uniform tensile
field (Fig-2).

Fig-2:

The greatest effect will be at 𝑟 = 𝑎, then

𝜎 =𝜎 =0
and 𝜎 , = 𝜎. (1 − 2 cos 2𝜃)
It will be evident that when 𝜃 = 𝜋⁄2
𝜎 , = 3𝜎
that is, there is a stress concentration of three times the nominal stress, 𝜎 , ⁄𝜎 = 3 is
known as the stress concentration factor (S.C.F.). Table-1 shows the variation of 𝜎 , as
𝑟 ⁄𝑎 is increased. It can be seen that for a value of 𝑟⁄𝑎 = 3 the stress 𝜎 , is almost equal
to the nominally applied stresses.

Table-1: Variation of 𝜎𝜃𝜃,𝑚𝑎𝑥 with 𝑟/𝑎 ratio

𝑟/𝑎 1.0 1.1 1.2 1.4 1.5 1.8 2.0 2.5 3.0 3.5 4.0
𝜎 , ⁄𝜎 3.0 2.44 2.07 1.65 1.52 1.30 1.22 1.12 1.07 1.05 1.04

13
Infinite plate with hole subjected to biaxial stress: To find the stress concentration factor for
an infinite plate with a hole which is subjected to biaxial stress, use the equation derived for
plate with a hole subjected pure tension. To discover the effect of 𝜎 on 𝜎 , simply put
𝜃 = (𝜃 + 𝜋⁄2) in the expression of 𝜎 , then at the periphery of the hole. Thus, expression
of 𝜎 , for biaxial tensile forces is,
𝜎 , = 𝜎 . (1 − 2 cos 2𝜃) + 𝜎 . 1 − 2 cos 2 𝜃 +
= (𝜎 + 𝜎 ) − 2(𝜎 − 𝜎 ). cos 2𝜃
Thus, if 𝜎 = 𝜎 = 𝜎 , then S.C.F. is 2. Whereas if 𝜎 = −𝜎 = |𝜎| a pure shear stress, then
𝜎 , = −4𝜎. cos 2𝜃 which gives a S.C.F. of 4 at 𝜃 = 𝜋⁄2.

Example -8: Concentrated Forces Acting on the Edge of a Large Plate

The problem to be solved is that of a concentrated force acting on the edge of a large (semi-
infinite) plate. The problem was first studied by Boussinesq for a three dimensional body and
the two dimensional problem was solved by Flamant and Mitchell. Referring to Fig-1a, the
concentrated force 𝑃 acts at a point on the surface 𝑚-𝑚, and Flamant and Mitchell showed
that the only stress at radius ‘𝑟’ was a radial one provided that 𝑥-axis coincide with the line of
action of 𝑃 with 𝑦-axis orthogonal to this to make a right-handed set of axes.

Fig-1a: Fig-1b:

This applies even when the load is not perpendicular to 𝑚 - 𝑚 (Fig-1b). The angle 𝜃 is
measured anticlockwise positive from 𝑥-axis. The stress function describing the problem is
 = 𝐴. 𝑟. 𝜃. sin 𝜃
where 𝐴 is a constant.
This equation can be shown to satisfy the compatibility equation ∇  = 0 and consequently
the stress components are

𝜎 = =0
 
𝜎 = . + . = . cos 𝜃

𝜎 =− . . =0
Thus, only 𝜎 has a value and to find the constant 𝐴, integration of forces around the semi-
circle of radius 𝑟 must equal the applied force 𝑃. The equation for 𝜎 denotes that as 𝑟
decreases the value of 𝜎 increases so that at the point of contact of the force the stresses are
infinite. This is not possible and it is assumed that there is small semi-circle of yielded
material in the vicinity of the force around which a pressure is distributed and equal to the

14
radial stresses at that radius. Referring to Fig-2, if ‘𝑡’ is the thickness of the plate, then
resolving forces vertically
⁄
2∫ 𝑡. 𝜎 . cos 𝜃 . 𝑟. 𝑑𝜃 = −𝑃 (compressive force)
and substituting for 𝜎
2𝑡. ∫ . 𝑐𝑜𝑠 𝜃. 𝑟. 𝑑𝜃 = −𝑃

 𝐴=−
.

Fig-2:

It is usual to let 𝑃 be the load per unit length in the 𝑧-direction (𝑡 = 1), then
𝐴=−
thus, substituting 𝐴 into stress function  and 𝜎
. .
=− . sin 𝜃
and 𝜎 =− .
. cos 𝜃 ---(1)
𝜎 =𝜎 =0
In this problem, the following transform equations are often useful.
𝜎 = 𝜎 . 𝑐𝑜𝑠 𝜃 + 𝜎 . 𝑠𝑖𝑛 𝜃 − 𝜎 . sin 2𝜃
𝜎 = 𝜎 . 𝑠𝑖𝑛 𝜃 + 𝜎 . 𝑐𝑜𝑠 𝜃 + 𝜎 . sin 2𝜃
𝜎 = . sin 2𝜃 + 𝜎 . cos 2𝜃 ---(2)
and 𝜎 = 𝜎 . 𝑐𝑜𝑠 𝜃 + 𝜎 . 𝑠𝑖𝑛 𝜃 + 𝜎 . sin 2𝜃
𝜎 = 𝜎 . 𝑠𝑖𝑛 𝜃 + 𝜎 . 𝑐𝑜𝑠 𝜃 − 𝜎 . sin 2𝜃
𝜎 = . sin 2𝜃 + 𝜎 . cos 2𝜃
Referring to Eq.(1) and Fig-3, and at point 𝐶 where 𝑟 = 𝑑 and 𝜃 = 0,
𝜎 =− .
Also at point 𝐵 on the circle of diameter 𝑑, 𝑟 = 𝑑. cos 𝜃 and again
𝜎 =− .
Thus the stress 𝜎 is constant around the circle of diameter 𝑑 and acts radially towards the
point 𝐴. In Fig-3 the stress trajectories are radial lines and the concentric semi-circles, and
the lines of constant stress difference (the iso-chromatics) are circles of diameter 𝑑 as
indicated in Fig-3.

15
 Fig-3:

Thus for a line 𝑚-𝑚 distant 𝐷 from the edge of the plate, the stresses 𝜎 , 𝜎 and 𝜎 can
be readily found using the transform Eqs.(2) which reduce to
𝜎 = 𝜎 . 𝑐𝑜𝑠 𝜃
𝜎 = 𝜎 . 𝑠𝑖𝑛 𝜃
𝜎 = . sin 2𝜃 = 𝜎 . sin 𝜃 . cos 𝜃

Now 𝜎 =− .
. cos 𝜃 and 𝑅. cos 𝜃 = 𝐷
where 𝑅 = 𝐴𝐵 from Fig-3,
therefore, stress components for a line m-m at point B
𝜎 =− . 𝑐𝑜𝑠 𝜃
.
𝜎 =− . 𝑐𝑜𝑠 𝜃
.

𝜎 =− .
. 𝑐𝑜𝑠 𝜃. 𝑠𝑖𝑛 𝜃
𝜎 =− .
. sin 𝜃 . 𝑐𝑜𝑠 𝜃
For a tangential load as shown in Fig-4, 𝜃 is measured from line of action of 𝑃 which is
coincident with 𝑥-axis. Then once again from Eq.(1)
𝜎 =− . cos 𝜃
.

Fig-4:

Again the integral of the resultant of the internal forces along the semi-circle shown in Fig-4
gives
∫ 𝜎 . cos 𝜃 . 𝑟. 𝑑𝜃 = ∫ 𝑐𝑜𝑠 𝜃. 𝑑𝜃 = . 𝜃 + sin 2𝜃 =𝑃

16
that is the internal forces balance the applied force and as before
𝜎 =− .
. cos 𝜃
𝜎 =𝜎 =0
For an inclined force the results of the two previous examples can be used.
Thus, the stress at point 𝑚(𝑟, 𝜃) shown in Fig-5 is
𝜎 =− .
. cos 𝛼 . cos 𝜃 + sin 𝛼 . cos +𝜃 =− .
. [cos(𝛼 + 𝜃)]
That is the angle (𝛼 + 𝜃) is still measured from the line of action of the force 𝑃.

Fig-5:

In Fig-6 for a couple applied to the boundary the stress function will be different. The stress
function for the forces at (1) and (2) can be superimposed to give
 = +
and since the forces are on a plane the stress function can be more conveniently obtained as a
function of 𝑥 and 𝑦. Conversion to 𝑟-𝜃 coordinate axes can then be carried out. Then
 = −(𝑥, 𝑦 + 𝑎) + (𝑥, 𝑦)

Fig-6:

If ‘𝑎’ is small then the first of the terms can be expanded using Taylor’s theorem as,

(𝑥, 𝑦 + 𝑎) = (𝑥, 𝑦) + 𝑎.
and, omitting higher orders of differential coefficients. Therefore
 
 = −(𝑥, 𝑦) − 𝑎. + (𝑥, 𝑦) = −𝑎.
  
Now = . + .

and = = sin 𝜃, = =

17
Hence
  
= . sin 𝜃 + .

Now using the stress function for a single force 𝑃 acting on a plane
.
=− . 𝜃. sin 𝜃
 .  .
and =− . sin 𝜃, =− . (𝜃. cos 𝜃 + sin 𝜃)
Then,
 . .
=− . 𝑠𝑖𝑛 𝜃 − . . (𝜃. cos 𝜃 + sin 𝜃)
. .
=− . 𝑠𝑖𝑛 𝜃 − . 𝑐𝑜𝑠 𝜃 − . sin 𝜃 . cos 𝜃 = − . 𝜃 + sin 2𝜃
and since

 = −𝑎.
.
then  = . 𝜃 + sin 2𝜃

with a sufficiently small but keeping 𝑃 finite, 𝑃. 𝑎 = 𝑀. Hence for a concentrated couple 𝑀
= . 𝜃 + sin 2𝜃
and 𝜃 is measured from the vertical from the point of application of 𝑀 as indicated in Fig-6.
𝜎 = . . (−2 sin 2𝜃) = − . sin 2𝜃, 𝜎 = 0, 𝜎 =0
.

with the usual restrictions applying to the point of application of 𝑀.

In the foregoing cases, the coordinate axes have been determined by the force and not the
plane. Simple radial stress therefore applies to the wedge shown in Fig-7a.
As before 𝜎 =𝜎 = 0 but the radial stress will have the form
.
𝜎 =− . cos 𝜃
where 𝛼 is a constant depending on the direction of 𝑃.

Fig-7a: Fig-7b:

Equating internal and external components of force in the direction of 𝑃 gives

−𝑃 = ∫ 𝜎 . 𝑟. cos 𝜃 . 𝑑𝜃 = −𝛼. 𝑃. ∫ 𝑐𝑜𝑠 𝜃. 𝑑𝜃
1 = 𝛼. 𝛽 + sin 2𝛽  𝛼=

.
and 𝜎 =−
.

18
For 𝑃 acting at right angles (Fig-7b), replace 𝜃 by (270° + 𝜃) in 𝜎 .
Equating internal and external forces gives
𝑃=∫ 𝜎 . 𝑟. sin 𝜃 . 𝑑𝜃 = −𝛼. 𝑃. ∫ cos + 𝜃 . sin 𝜃. 𝑑𝜃
or 𝛼=

. .
and 𝜎 =− . cos +𝜃 =−
.

and 𝜃 is now measured as indicated in Fig-7b.

With this equation for 𝜎 , the stresses 𝜎 and 𝜎 can be found. The solution will be
applicable to a triangle shaped beam.
(i) Displacement Components: For all cases, the usual equations to determine displacement
components 𝑢 and 𝑣 are used, that is
𝜀 =
𝜀 = + .
𝛾 = 𝑟. + − =0
and after integration the resulting constants would have to be evaluated.
To find these constants the boundary conditions must be carefully considered. For example,
for a point force acting normal to the edge of a semi-infinite plate, if it is assumed that there
is no tangential movement along the 𝑥-axis then when 𝜃 = 0, 𝑣 = 0. Note that any term
containing 𝑟 only must vanish since the condition along the 𝑥-axis must be true for all values
of 0 < 𝑟 < ∞. Another condition might then be that on the 𝑥-axis the radial displacement
𝑢 = 0 at some distance 𝑑 (Fig-7b). This will result in having a small constant depending on
the value of 𝑑.
(ii) Uniform Load: Consider a uniformly distribute load shown in Fig-8a. The stress function
for this loading is given by
 = 𝐵. 𝑟 . 𝜃
giving the stresse components
𝜎 = 2𝐵. 𝜃 = 𝜎
and 𝜎 = −𝐵

Fig-8a: Fig-8b:

The function 𝜎 = 𝐵. (𝑟 . 𝜃 − 𝑟 . 𝜃 ) applies to the plate shown in Fig-8b, or

 = 𝐵. 𝑟. (𝑟 . 𝜃 − 𝑟 . 𝜃 )
for compressive loading. (These are found by moving the origin of coordintes and changing
the sign of .)

19
 Exercise Problems
#1) Derive one of the 3-Navier equations of displacement formulation for 3D continuum:
∇ 𝑢+( )
. + =0

where 𝑒= + +

#2) Derive two of the 6-Beltrami-Michell equations of stress formulation for 3D continuum
as given below. You can refer a book “Theory of Elasticity” by S. Timoshenko, in
Chapter-9 General Theorems, pp. 228-232.
 
∇ 𝜎 +( )
. = −( )
. ∇𝑏 − 2

∇ 𝜎 +( )
. .
=− +

where =𝜎 +𝜎 +𝜎

∇𝑏 = + +

#3) Show that  = 𝛼. 𝑥. 𝑦 + 𝛽. 𝑥. 𝑦 is a valid stress function. The function applies to a

rectangular plate shown below lying on 𝑥 - 𝑦 plane. Calculate also the displacement
components 𝑢 and 𝑣 for points 𝐴, 𝐵, 𝐶 and 𝐷 given that at point 𝑂, 𝑢 = 𝑣 = = = 0.

Answer: @ point 𝐴, (𝑢, 𝑣) = 𝛼. 𝑑 . − , 0

𝐴 𝐵
𝑑 𝑥
𝑂
𝑑
𝐷 𝐶
𝑙

#4) Show that a simply supported beam of length 𝐿, depth ℎ and a unit thickness (width),
loaded by a concentrated load 𝑃 at the mid-span, the stress function satisfying the loading
condition is:
 = 𝑐 . 𝑥. 𝑦 + 𝑐 . 𝑥. 𝑦 + 𝑐 . 𝑦 − ≤𝑦≤ & − ≤𝑥≤
where 𝑐 , 𝑐 and 𝑐 are constants. The positive direction of ‘𝑦’ is being upwards and
𝑥 = 0 at mid-span. Also determine the stress components including the constants.
#5) Investigate what problem of plane stress is solved by the stress function as given by:
.
= . 𝑥. 𝑦 − + .𝑦 (−𝑐 ≤ 𝑦 ≤ 𝑐) & (0 ≤ 𝑥 ≤ 𝐿)

#6) Investigate what problem is solved by

=− . 𝑥. 𝑦 . (3𝑑 − 2𝑦) (0 ≤ 𝑦 ≤ 𝑑) & (0 ≤ 𝑥 ≤ 𝐿)
applied to the region included in 𝑦 = 0, 𝑦 = 𝑑, 𝑥 = 0, on the side ‘𝑥’ positive.
#7) Show that  = . 𝑥 . (𝑦 − 3𝑐 . 𝑦 + 2𝑐 ) − . 𝑦 (𝑦 − 2𝑐 )
is a stress function, and find what problem it solves when applied to the region included in
𝑦 = ±𝑐 , 𝑥 = 0, on the side ‘𝑥’ positive. i.e. (−𝑐 ≤ 𝑦 ≤ 𝑐) & (0 ≤ 𝑥 ≤ 𝐿)

20
#8) Bending of a vertical cantilever loaded by hydrostatic pressure:
Consider a vertical cantilever subjected to hydrostatic pressure (linearly varying load) as
shown below. The intensity of the load at a depth ‘𝑥’ from top is ‘𝑞. 𝑥’ where ‘𝑞’ is unit
weight of the fluid, the shearing force and bending moment at the same depth are
𝑉 = 𝑞. 𝑥 ⁄2 and 𝑀 = 𝑞. 𝑥 ⁄6, respectively.
Derive expressions for stresses in the vertical cantilever using the stress function by
combining 2nd, 3rd, 4th and 6th degree polynomials retaining some of the terms as:
= + + +
  = 𝑎 . 𝑥. 𝑦 + 𝑎 . 𝑥 + 𝑎 . 𝑥 . 𝑦 + 𝑎 . 𝑥. 𝑦 + 𝑎 . 𝑥 . 𝑦 + 𝑎 . 𝑥. 𝑦

𝑥 𝑐 𝑐

𝑞. 𝑥

𝑞. 𝑙
𝑥

Ans.: Compatibility condition ∇  = 0 is satisfied when 𝑎 = − 𝑎

-Constants of stress function obtained from stress boundary conditions are
.
𝑎 =− , 𝑎 =− , 𝑎 =− , 𝑎 =− and 𝑎 =−
-Stress component obtained in vertical cantilever are:
𝜎 = .𝑥 .𝑦 − . 𝑥. 𝑦 + . 𝑥. 𝑦 = .𝑥 .𝑦 + . −2𝑥. 𝑦 + 𝑐 . 𝑥. 𝑦

𝜎 = − .𝑥 + . 𝑥. 𝑦 − . 𝑥. 𝑦 = − . 𝑥 + 𝑞. 𝑥. −
.
𝜎 = + .𝑥 − .𝑦 − .𝑥 .𝑦 + .𝑦
= . 𝑥 . (𝑐 − 𝑦 ) − . (𝑐 − 𝑦 ) + × 𝑐 . (𝑐 − 𝑦 )
At top end of the beam (𝑥 = 0), normal stresses are zero.
i.e. 𝜎 = 𝜎 =0
and, shearing stress 𝜎 is not zero (from results of expression)
𝜎 (𝑥 = 0) = − . (𝑐 − 𝑦 ) + × 𝑐 . (𝑐 − 𝑦 )
but their values are very small all over the cross-section and their resultant is zero, so that
the condition approaches that of an end free from external forces.
#9) Verify the stress function given in Cartesian coordinates to polar coordinates given by:
 = 𝑥 − 𝑦 = (𝑥 + 𝑦 ). (𝑥 − 𝑦 ) = 𝑟 . cos 2𝜃

21
#10) Show that the stress function given by equation:
 = 𝐴. 𝑟 + + 𝐶. 𝑟 + 𝐷. 𝑟. ln 𝑟 . sin 𝜃
may be used to solve the quarter-circle curved beam shown below under an end-load 𝑃.
Determine also the constants 𝐴, 𝐵 and 𝐷; and the stress components of the curved beam.
𝑦

𝑏
𝑎
𝑟
𝜃
𝑥
𝑂
𝑷
-Stress boundary conditions are
[𝜎 ] & =0
and [𝜎 ] & =0
and also, the applied load 𝑃 may be taken to be the summation of the shear stress acting
over the free-end at 𝜃 = 0.
∫  𝜎 ] . 𝑑𝑟 = −𝑃
Ans.: the stress components of curved beam are
.
𝜎 = . 𝑟+ − . sin 𝜃
.
𝜎 = . 3𝑟 − − . sin 𝜃
.
𝜎 = . 𝑟+ − . cos 𝜃

where 𝐾 = (𝑎 − 𝑏 ) + (𝑎 + 𝑏 ). ln

22