Chapter 4: Stresses and Displacement Fields

4.0 Introduction

In this chapter different modes of crack formation are introduced to discuss their stress and
displacement field near crack tip. The stress and displacement field equations developed for linear
elastic brittle material in bi-axial stress field. These equations are discussed based on their historical
development such as Westergaard Equation (1939), Irwin modification (**), Kolosov-Muskhelishvili
(1957), Sanford (1979), William Eigen Function Approach (1957), Atluri and Kobayashi (**). All these
stress function uses the analytic function for their development which is also introduced.

4.1 Modes of Fracture

Three different modes of failure like opening, sliding and tearing mode are shown in Figure 2. Apart
from that, any complicated fracture state can be split into these three basic modes and taken
equivalently as the result of their combinations.

Figure.Error! No text of specified style in document..1 Three fracture modes Mode I:

displacement normal to crack plane (opening mode); mode II: displacement parallel to crack
plane and normal to crack edge (sliding mode); mode III: displacement parallel to crack plane
and to crack edge (tearing mode)

In this chapter, stress field around crack tip are defined by analytic function. As this function is used
by Wastergaard (1939) to define stress function near to the crack tip. Analytic function are
differential at point z 0 or nearby point or tailor series of this function should converge. This function
is taken in the form of a complex number as it satisfy certain condition and have some advantages
such as it can be expanded like tailor series, can be expressed interchangeably either in cartesian or
polar co-ordinate system or can be expressed in trigonometric forms. Let the complex number is:

1
 z=x +iy (1)

Its conjugate is:

z=x−iy (2)

And can also be written as:

z=r e
iθ
(3)

Let us consider a function as:

Z=u ( x , y ) +iv ( x , y ) (4)

If Z is real,. Δy=0 then the differentiation is.

lim [ u ( x + Δx , y ) +iv ( x + Δx , y ) ] −[ u ( x , y ) +iv ( x , y ) ]
dZ Δx →0
=
dz Δx

lim [ u ( x + Δx , y )−u ( x , y ) ]
Δx →0 v ( x + Δx , y )−v ( x , y )
¿ +i
Δx Δx

dZ ∂ u ∂ v (5)
= +i
dz ∂x ∂x

If Z is imaginary, then Δx=0.

lim [ u ( x , y+ Δy ) +iv ( x , y+ Δy ) ] − [ u ( x , y )+ iv ( x , y ) ]
dZ Δy →0
=
dz iΔy

lim [ u ( x , y+ Δy )−u ( x , y ) ]
Δy →0 v ( x , y+ Δy ) −v ( x , y )
¿ +i
iΔy iΔy

dZ 1 ∂u ∂ v ∂ v ∂u (6)
= + = −i
dz i ∂ y ∂ y ∂ y ∂ y

Here, imaginary part become real. In this two approach of differentiation at a point then the results
should be same i.e. real/ imaginary part of Eq. (5) and (6) should be same.

dZ ∂ u ∂ v (7)
= =
dz ∂x ∂ y

2
 dZ ∂ v −∂ u (8)
= =
dz ∂x ∂ y

The relation (7) and (8) are known as Cauchy-Riemann condition. Which can also be written in
generalized form as:

∂ ReZ ∂ ImZ ∂ ReZ (9)

= =
∂x ∂y ∂z

∂ ImZ −∂ ReZ ∂ ImZ (10)

= =
∂x ∂y ∂z

If above two conditions are satisfying then the function Z is differentiable at any point or nearby
point.

Example 4.1: Check whether the following function is analytic function for all Z and satisfies Cauchy-
Riemann condition.
2 2 2
f ( Z )=Z =x − y + 2ixy
Solution: Westergaard (1939) published several Airy’s stress function to study crack problem. One of
the Airy’s stress function is

φ=Z ( z )=ℜ Ź + yIm Z (11)

dZ ″ d Z '
; Z=∫ Zd z ; and Z=∫ Z d z
'
Where, z=x +iy ; Z = ;Z=
dz dz

Now, verify where the Eq (11) is a Airy’s stress function by satisfying biharmonic equation.

∂φ
=ℜ Z+ yIm Z
∂x
2
∂ φ '
2
=ℜ Z+ yIm Z
∂x
3
∂ φ ' ''
3
=ℜ Z + yIm Z
∂x
4
∂ φ '' '' '
4
=ℜ Z + yIm Z
∂x

∂ φ ∂ ℜ Ź ∂ yIm Z
= + =−ImZ+ yReZ+ ℑ Z = yReZ
∂y ∂y ∂y
2
∂ φ ∂ ReZ '
2
=ReZ + y =ReZ− yIm Z
∂y ∂y

3
 3
∂ φ ' ''
3
=−2 ℑ Z − yRe Z
∂y
4
∂ φ ''
4
=−3 ℜ Z + yImZ ' ' '
∂y
4
∂ φ ''
2 2
=ℜ Z − yImZ ' ' '
∂x y

Now, substitute the differentials in bi-harmonic equation.

4
∇ φ=0

The Airy’s stress function should satisfy certain boundary condition related to central crack in an
infinite plate in a bi-axial stress field shown in Figure 2 and is defined by Eq. (12). The crack surface
can be defined in deformed and undeformed condition. But for time being, let us consider
undeformed condition as deformation are small. This assumption defines the free surface along on
which stress component will be zero. So, defined boundary conditions are:

1. For a central crack length −a ≤ x ≤ a and y=0; the stresses σ yy =0 ; and τ xy =0.
2. For a point away from the crack i.e. Z → ∞; σ xx =σ yy =σ and τ xy =0.
3. At y=0 for any value of x=0 , shear stress τ xy =0 due to symmetry

σ σ

x
2a

Figure. Error! No text of specified style in document..2: Infinite plate with central crack and origin
at crack center in bi-axial loading

σz (12)
Z ( z )=
√ z −a
2 2

Irwin modified the above Airy’s stress for origin at the crack tip by substituting z=z 0 +a as:

4
 σ ( z 0 +a )
Z ( z )=
√ ( z + a ) −a
0
2 2

For near crack tip, the stress field is obtained by making the approximation a ≫ z 0, i.e. crack is much
2
bigger than z 0 and z 0 is very small, then

σ ( z 0 +a ) σa σ √a
−1
Z ( z )= = = (z ) 2

√ z +2 z
2
0 0 a+a −a
2 2
√ 0
2 z a √2 0

−1 −1
σ √a (13)
Z ( z )= ( r eiθ ) 2 = σ √ a ( cos θ+i sin θ ) 2

√2 √2 r
σ

y y

σ r σ
θ

x
2a

Figure Error! No text of specified style in document..3: Infinite plate with central crack and
origin at crack tip in bi-axial loading

The equation in further modified to express in terms of another material constant in fracture
mechanics called Stress Intensity Factor (SIF) and defined for mode I as K I =σ √ π a

−1
σ √π a K (14)
Z ( z )= ( cos θ+ isin θ ) 2 = I f ( θ )
√ 2 πr √ 2 πr
−K I
∂Z (z)
∂r
=Z ( z ) '= 3
2 √2 π r2
3θ
cos −isin
2
3θ
2 ( )

4.3 Mode – I: Stress and displacement field near crack tip

In this section, the stress and displacement fields near the crack tip are derived based on
Westergaard approach for plane stress and plane strain condition. Based on displacement field the
shape and displacement of crack is defined.

5
4.3.2 Stress field

In this section the stress field in plane stress and plane strain condition are derived.

Plane stress condition

As we know, from Airy’s stress function, the stress field can be determined using derivatives as:

∂φ
2
' (15)
σ xx = 2
=ℜZ − y ℑ Z
∂y
2
∂φ '
σ yy = 2
=ℜ Z+ y ℑ Z
∂x
2
∂ φ '
τ xy = =y ℜZ
∂x∂y

If the origin of the coordinate system is located at the crack tip, then the stress field equation (Mode
I) near the crack tip can be written in terms of the stress intensity factor as

{ }
θ 3θ (16)
1−sin sin

{}
2 2
σ xx
KI θ θ 3θ
σ yy = cos 1+sin sin
√ 2 πr 2 2 2
τ xy θ 3θ
sin cos
2 2

Figure Error! No text of specified style in document..4: Stress field in polar coordinate
system

The contour plots of stress field around crack tip is shown in Figure.Error! No text of specified style in
document..5 (a). The stress pattern are symmetric with x and y axis. In a laboratory test, the
photoelastic fringe pattern shows that the fringe are vertical closure to the crack tip and inclined in
forward direction away from crack tip, stresses are varies along the axis.

6
 Figure.Error! No text of specified style in document..5: (a) Numerical plots of
isochromatics by Westergaard equations (b) Experimental observation of fringe pattern (c)
Modified Westergaard equation by Irwin by adding a constant in σ xx term by σ ox.

Irwin has modified the Westergaard equation to match the laboratory fringe pattern by adding
constant term σ ox to the σ xx term to account for the behaviour of finite geometry specimens used in
the experiments.

{ }{ }
θ 3θ (17)
1−sin sin

{}
2 2
σ xx KI −σ ox
θ θ 3θ
σ yy = cos 1+sin sin + 0
√ 2 πr 2 2 2
0
τ xy θ 3θ
sin cos
2 2

Figure .6: Effect of positive and negative σ ox stress on fringe pattern

7
 Figure 7: One problem shows a fringe pattern is both forward and backward tilted

It satisfies two experimental conditions that the fringe pattern is tilted forward and for y = 0 shear
stress is not zero as shown in Figure.Error! No text of specified style in document..5(c). This
condition also satisfies small cracks but not large cracks. Later on σ ox is knows as T-stress in
numerical analysis. In polar Co-ordinate system, the stress function is written as:

{ }
5 θ 1 3θ (16)
cos − cos

{}
4 2 4 2
σ rr KI 3 θ 1 3θ
σ θθ = cos + cos
√2 πr 4 2 4 2
τ rθ 1 θ 1 3θ
sin + sin
4 2 4 2

The stress function in principal stress form is

[ ]
θ (19)
1+sin
{} σ1
=
K I

σ 2 √ 2 πr
cos
θ
2
1−sin
2
θ
2

4.3.2 Displacement field

For origin at central crack and origin at crack tip, the expression for Westergaard stress function for
mode I. Z I is shown respectively in cartesian co-ordinate system is:

KI (20)
Z I=
√ 2 π z0
σz
Z I=
√ z 2−a2
Plane stress (Mode I)

8
Using Hook’s law and strain displacement relationship, we can obtain the displacement in cartesian
co-ordinate system.

∂u 1 1
ε xx =
∂ x E xx E
[
= [ σ −ν ( σ yy + σ zz ) ]= ( ℜZ I − y ImZ I ) −ν ( ℜ Z I + yIm Z I )
' '
]
∂v 1 1
ε yy =
∂y E E
[
= [ σ yy −ν ( σ xx + σ zz ) ]= ( ℜ Z I + y ImZ I )−ν ( ℜ Z I − yIm Z I )
' '
]

ε xy = (
1 ∂u ∂ v −1−ν
+
2 ∂ y ∂x
=
E )
yRe Z'I

After integration of the above expression, the displacement field ( u , v ) are determined as shown
below by replacing E=2 (1+ ν ) as:

u xx =
1 1−ν
2G 1+ ν [ ]
ℜ Z I − yIm Z I + f ( y )

u yy =
1 1−ν
2 G 1+ ν[ ]
ℜ Z I + yRe Z I + g (x)

Where, f ( y ) and g(x ) are integration constant. For open fracture problem f ( y ) and g(x ) can be
equated to zero as shear stresses near crack tip will be equated to zero.

ε xy = (
2 ∂ y ∂x
=
∂y )
1 ∂u ∂ v ∂ f ( y ) ∂ g(x )
+ +
∂x
=0

∂ f ( y ) −∂ g ( x )
Gives, = =A
∂y ∂x

Where, A is a constant

Integrating above expressions

∫ ∂ f ( y )= A ∫ ∂ y
f ( y )=Ay + B

Similarly, g ( x )=− Ax+C

Where, B and C are constant of integration.

If f ( y ) and g(x ) are substituted in displacement field equation. All the points of the components
are displaced by the same distance given by u xx =B and u yy =C .

Rigid body rotation

9
 ω= (
1 ∂u ∂v
−
2 ∂ y ∂x
=A )
So, the stress field in cartesian co-ordinate system is:

u xx =
[
1 1−ν
2G 1+ ν
ℜ Z I − yIm Z I
] (21)

u yy =
1 2
[
2 G 1+ ν
ℑ Z I − yRe Z I
]
Polar Co-ordinate system

The Westergaard stress function and its integral in polar co-ordinate system is:

Z I=
KI
√ 2 π z0
=
KI
√2 πr [θ
cos −i sin
2
θ
2 ]
() ( ) K [ cos θ2 +i sin θ2 ]
1 1
2 2r
Z I= 2
K I √ z 0= 2
I
π π

Then the displacement field expression in plane stress condition for model thickness B is:

√ [ ]
KI r θ 1−ν θ (22)
u xx= cos +sin 2
G 2π 2 1+ ν 2

√ [ ]
KI r θ 2 θ
u yy = sin −cos 2
G 2π 2 1+ ν 2

u zz =− ( 2Eν K √ 2rπ cos θ2 ) B

I

In plane stress case there is z-direction displacement, u zz at crack tip called dimple.

Plane strain (Mode I)

For plane strain, the displacement field is also obtained by replacing ν by ν / (1−ν ) in Eq. (Error:
Reference source not found) and the expression is shown below. It is worth to no note that u zz =0 in
this case.

1 (23)
u xx =
2G
[ ( 1−2 ν ) ℜZ I − yIm Z I ]

10
 1
u yy =
2G
[ 2 ( 1−2 ν ) ℑ Z I − yRe Z I ]

In polar co-ordinate the displacement field expression is

√ [ ]
KI r θ θ (24)
u xx = cos 1−2 ν +sin2
G 2π 2 2

√ [ ]
KI r θ θ
u yy = sin 2−2 ν −cos2
G 2π 2 2

The combined expression for displacement field IN Cartesian coordinate system for both Plane
stress and plane strain (Mode I)

√ [ ]
KI r θ θ (25)
u xx = cos κ−1+ 2sin 2
2G 2 π 2 2

√ [ ]
KI r θ θ
u yy = sin κ+ 1−2 cos2
G 2π 2 2

3−ν
Where, κ= for plane stress condition and κ=3−4 ν for plane strain condition.
1+ ν

Exercise 1: If a central crack of length 10 mm is developed in large 2D surface acted upon by a far
field stress of 1 MPa. Find out the stress σ xx , σ yy, τ xy as well as maximum shear stress ( τ max ( ) )

distribution around the crack tip ( r , θ ) .

One can observe contours of constant principal stress difference as fringes knows as isochomatics.
The geometric feature of the fringe patterns have played a significant role in arriving at correct stress
field equations in fracture mechanics.

√
2
σ 1−σ 2 ( σ xx −σ yy ) 2
=τ +τ xy
2 4

4.4 Crack Opening Displacement (COD)

The displacement equation is used to find the variation of crack opening in a given bi-axial stress in
plane stress condition using Westergaard equation at y = 0 in u yy of Eq. (22) as:

u yy =
1 2
[
2 G 1+ ν
ℑ Z ( z)− yReZ (z)
]
11
 σz σz
Where, Z ( z )= 2, the integral of the function is Z̄ ( z )=∫ dz
√ z −a
2
√ z2−a2
Let, √ z 2−a2=m; then after differentiation 2 z d z=2 mdm, or z d z=mdm

1 σdm σ (26)
Z̄ ( z )=∫ ∫ = √ m=σ √ z −a
2 2

2 √m 2

For y=0, z → x

Z̄ ( z )=σ √ x 2−a 2=iσ √ a 2−x 2 (27)

For y=0; u yy becomes u yy =

1
[2
2 G 1+ ν
iσ √ a2 −x2 −0 = √
]
σ a 2−x 2
G ( 1+ν )

σ √ a2−x 2 2 σ √ a2 −x2 (28)

u yy = =
G ( 1+ ν ) E

4 σ √ a2−x 2 (29)
COD=2u yy =
E

The above expression is for an ellipse, and shows that the crack opens like a ellipse. For x=0 , the
4 σa
maximum opening at the origin and is equal to and for x=a , i.e at crack boundary the
E
displacement is zero.

The displacement equation is used to find the variation of crack opening in a given bi-axial stress in
plane strain condition using Westergaard equation at y = 0 of u yy in Eq. (23). Following similar steps,
the COD in case of plane strain condition is:

4 ( 1−ν 2 ) σ 2 2 (30)
COD=
E
√a −x

Inference from the above expression:

2 2
x y
+ =1
( )
2 2
 This expression can be written in the form of a equation of ellipse as: a 4σ .
E
Justifies that the crack opens like a ellipse.
4 σa
 The maximum crack opening is at the center of the crack.
E

12
  The COD for plane stress condition is more than the plane strain condition.

4.5 Energy Release Rate ( G I ) and Stress Intensity Factor ( K I )

In this section, the relationship between energy release rate based on energy approach and stress
intensity factor based on stress method of crack propagation is established by two methods:

(i) Principle of superposition

The crack opening displacement in case of plane stress condition is given by Eq. (29):

4 σ √ a2 −x2
COD=
E

Energy required to form two new surfaces of thickness dx and height dy and thickness B or energy
stored per unit volume

1 1 Δy σ 2 σ 2 √ a2−x 2 (31)
d U a= σ ε dV = σ Bdxdy= Δ yBdx= Bdx
2 2 dy 2 E

For unit thickness, B=1

2 σ 2 √ a2−x 2
[ ( )]
a
2 σ 2 x 2 2 a2 −1 x
a
π σ 2 a2 (32)
U a =∫
E
dx=
E 2
√ a −x + 2 sin a =
E
−a −a

Here, energy release rate for two surfaces of length 2 a. We know for two edged crack the SIF
relation is K I =σ √ πa, which is used to find the relation between SIF and ERR. Here, we have used
this method to solve this problem when the relationship of SIF is know. For single crack it will be
differentiated w.r.to a .

d U a π σ2 a KI
2
(15)
G I= = =
d (2 a ) E E

This relationship is for the central in an infinite plane only and may not true for different situations.

(ii) Generic approach

The above procedure is useful when we know the SIF for a given crack. Now we will try to solve the
problem of generic to find the stresses at crack tip. Energy required to close the crack tip is equal to
energy required for opening two surfaces of same crack length. Here, we take a crack of length
a+ Δa extended upto point Q . Now, if the crack closes due to removal of compression and reaches
point P, then we want to find out the energy released to close down the crack by length Δa . Let K I

13
be the SIF for the crack length a ; K I + Δ K I be SIF for crack length a+ Δa. Then locate the origin at
point Q , and apply Westergaard equation in polar co-ordinate system to define the crack opening in
y direction near vicinity of crack tip at point P is:

√ [ ]
KI r θ 2 θ
u yy = sin −cos 2
G 2π 2 1+ ν 2

Figure 8:

For opening of crack length s ahead of point P , y-displacement at crack tip (while keeping Q as
origin) is given by Eq (22), where θ=π and r =Δa−s .

√ [ ] √
K I+Δ K I π K + Δ KI (34)
u yy =
G
Δa−s
2π
sin
π 2
2 1+ν
−cos 2 = I
2 G ( )
Δa−s 2
2 π 1+ν

For extension of crack tip from r =0 to at r =s , by keeping origin at point P, the stress at point s is
given by Westergaard stress field equation by using θ=0, as:

( ){ ( ) ( )}
KI θ θ 3θ KI (35)
σ yy = cos 1+ sin sin =
√ 2 πr 2 2 2 √ 2 πs

Now, total elastic work required by σ y is closing the crack is equal to the energy release rate. When
the load are gradually applied then balancing the energy

Δa
B σ y uy (36)
G I BΔa=2∫ ds
0 2

14
Where, B is the thickness of the plate, BΔa area of crack developed. Substituting σ y and in the Eq.
(**) and taking the limit Δa →0

lim 2 (37)
K I ( K I + Δ K I ) √ Δa−s
Δa
G I=
Δa→ 0
∫
( 1+ ν ) ΔaG 0 √2 πs √2 π
ds

Δ K I is very small for limit Δa →0 .

2 Δa
(38)
( )
1
KI Δa−s
G I= ∫
( 1+ ν ) πGΔa 0 s
2
ds

Substituting s= Δa sin2 α leads to

2
π
2
(39)
KI
G I=
( 1+ ν ) πG
∫ 2cos 2
αdα
0

2G
Solving the integral and substituting E=
( 1+ ν )

KI
2
(40)
G I=
E

This relation is valid only for brittle materials.

For plane strain situation the relation is

KI 2
2
(41)
G I =( 1−ν )
E

4.6 Mode-II (Sliding Mode) - Stress and displacement field near

crack tip

The Westergaard function Z II for a crack of length 2 a in an infinite plate subjected to uniform in-
plane shear stress τ at infinity is given by Eq. (45). The stresses and displacement are obtained from
Eqs (***) Following the same procedure as in the previous case, and recognizing the general
applicability of the singular solution for all sliding-mode crack problem, we obtain the following
equations for stresses and displacement.

Airy’s stress function

15
 φ=− y ℑ Z̄ II (42)

Westergaard stress function

τz (43)
Z II =
√ z −a
2 2

Only difference with the Mode-I is that σ is replaced by τ .

Figure 9:: A crack of length 2 a in an infinite plate subjected to uniform in-plane shear
stresses τ at infinity with origin at center of the crack

Stress field for Plane strain condition with origin at center of the crack

σ xx =2 ℑ Z II − y ℜ Z II
'
(44)

'
σ yy =− y ℜ Z II
'
τ xy =ℜ Z II − y ℑ Z II

Displacement field for Plane strain condition with origin at center of the crack

1 (45)
u xx =
2G
[ 2 ( 1−ν ) ℑ¯Z II + y ℜ Z II ]
1
u yy =
2G
[− (1−2 ν ) ℜ Z II− y ℑ¯Z II ]

16
 y
(r, θ)

τ
τ x
2a

Figure.10. A crack of length 2a in an infinite plate subjected to uniform in-plane shear

stresses τ at infinity with origin at crack tip

Westergaard stress function when origin shifting by substituting z=z 0 +a

τ ( z0 + a) τ ( z 0 +a ) (46)
Z II = =
√ ( z 0 +a ) −a √ z 0 +2 z 0 a
2 2 2

Making approximating that z 0 ≪ a , then

τa τ √ πa K II (47)
Z II = = =
√ 2 z 0 a √ 2 π z 0 √2 π z 0

Now, defining K II =τ √ πa

−1
(48)
( )
1
K II K II K II K II θ θ
Z II = = z = 2
(r e ) =
iθ 2
cos −i sin
√ 2 π z 0 √2 π
0
√2 π √2 π 2 2

( )
' K II K II 3θ 3θ (49)
Z II = = cos −isin
2 τ 0 √ 2 π z0 3
2 2
2 √2 π r 2

{ }
θ 3θ (50)
−2−cos cos

{}
2 2
σ xx
K II θ θ 3θ
σ yy = sin cos cos
√ 2 πr 2 2 2
σ xy
θ θ
cot 1−sin sin
2 2
3θ
2 ( )

17
The stress distribution and magnitude are two important things and it was identified by Irwin in
1957.

Displacement field in Mode-II

Plane stress (Mode-II)

√ [ ]
K II r θ 2 θ (51)
u xx = sin +cos 2
G 2π 2 1+ν 2

√ [ ]
K II r θ −1−ν θ
u yy = cos + sin2
G 2π 2 1+ ν 2

( 2Eν K √ 2rπ sin θ2 ) B

u zz = II

ν
B is model thickness. To find the value for plane strain condition replace ν by in plane stress
( 1−ν )
condition.

Plane strain (Mode-II)

√ [ ]
K II r θ θ (52)
u xx = sin 2−2 ν +cos 2
G 2π 2 2

√ [ ]
K II r θ θ
u xx = cos −1+ 2 ν +sin2
G 2π 2 2

In plane stress case there is z-direction displacement at crack tip called dimple that can be estimate
by u zz . Combined expression of displacement filed for both plane stress and plane strain condition.

√ [ ]
K II r θ θ (53)
u xx = sin κ +1+2 cos2
2G 2 π 2 2

√ [ ]
K II r θ θ
u yy = cos 1−κ+ 2sin 2
G 2π 2 2

3−ν
Where, κ= for plane stress condition and κ=3−4 ν for plane strain condition.
1+ ν

Investigation while doing experiment for Mode-II

 Geometrical feature of theoretical fringes and experimental fringes are same.

18
 Figure 11: Mode – II stress distribution from (a) Photoelastic fringe pattern, and (b) Westergaard
solution of Mode II failure

Figure ** shows that the stress fields are same from both the solutions are matching

T-stress ( σ ox ) introduced by Irwin, for isochromatic fringes ( σ 1−σ 2 ) and isopachis ( σ 1+ σ 2 ) are +ve
and -ve respectively. For these two cases fringe pattern become forward and backward tilted for
isochromatic and isopachis fringes.

Figure 12: Fringe pattern for Mode-I and Mode-II

Top fringe are bigger than bottom. Axis along the crack tilt of top half and bottom half are different.

Exercise

1. Check that the mode II stress function is a biharmonic equation.

2. Derive the stress and displacement filed equations in case of mode II crack propagation.

19
4.7 Mode - III (Tearing Mode) - Stress and displacement field
near crack tip

In this case a large plate with a center crack is subjected to a far field shear stress which is acting in
anti-plane direction. For example, Mode III crack is visualized as a torsion is acting on a bar having
circumferential crack. For tearing or anti-plane mode of crack deformation the in-plane i.e. x-y plane
displacements u and v are zero, while the displacement in z-direction i.e. w , is a function of the in-
plane coordinates x and y , that is

u x =0; u y =0; u z=w ( x , y ) (56)

Figure 13: A crack length 2a in an infinite plate subjected to uniform out-of-plane shear stress
at infinite with origin at center

The Eq (56) suggest that the movement of the crack surfaces can be related to the wrapping action
of noncircular cylinders subjected to torsion. Eq. (56) renders

ε xx =ε yy =ε zz =γ xy =0

1 ∂w 1 ∂w
ε xz = ; ε yz=
2 ∂x 2 ∂y

Stress field is σ xx =σ yy =σ zz =τ xy =0

∂w (55)
τ xz =G
∂x

∂w (56)
τ yz=G
∂y

20
It should be remembered that in Mode III is not a case of plane stress and plane strain condition.
This condition is very simple as many components of displacement, stresses and strains are zero.
Hence, instead of Biharmonic equation this problem will be solved with displacement component w
as the dependent variable and then there is no need to worry for compatibility conditions any more.
Out of the three equilibrium equations, only the last one provides the non-trivial equations,

∂ τ xz ∂ τ yz (57)
+ =0
∂x ∂ y

Use Eqs (55) and (56) in Eq. (57) then or converting stress into displacement components called
Laplace equation.

2
∂ w ∂w
2
(58)
2
+ 2 =0
∂x ∂ y

2
Or, ∇ w=0

w in selected in the form as shown in Eq. (59) as given by Westergaard

1 (59)
w= ℑ Z III
G

Where, Z III is a complex function of variable z . Again one can show that the above expression
satisfied the governing differential equation. Also, substituting w in Eq (55) and Eq. (56), the shear
stress components are:

τ xz =ℑ Z III
'
(60)

τ yz=ℜ Z III
'
(61)

The following form of the stress function satisfies all boundary conditions.

' τz (62)
Z III =
√ z −a
2 2

'
Z III is transformed the origin to the crack tip by using z=a+ z 0 and after neglecting the small terms,
we obtain:

21
 ' τ √a (63)
Z III = 1

( 2 z0 ) 2

Using τ √ πa=K III and expressing z 0 in polar coordinates, the equation is simplified to

( )
' K III θ θ (64)
Z III = cos −isin
√2 πr 2 2

'
Substituting Z III is Eqs. (60) and (61), the stress field in mode III is

−K III θ (65)
τ xz = sin
√ 2 πr 2

K III θ (66)
τ yz= cos
√ 2 πr 2

Integrating Eq (64) and substituting τ √ πa=K III , we have

√ √
(67)
( )
1
2 2r θ θ
Z III = K ( z ) 2 =K III cos +isin
π III 0 π 2 2

Replace Z III in Eq (59), to find the displacement in z-direction as:

√
1 K 2r θ
u z=w ( x , y )= ℑ Z III = III sin
G G π 2

u x =0

u y =0

Exercise **.4: Derive the complete stress and displacement function at the crack tip in mode III crack
propagation.

4.8 Generalised Westergaard Equations by Sanford (1979)

Sanford (1979) introduced an additional stress function Y ( z ) to Westergaard stress function Z ( z ) to

explain the two behaviour i.e (i) forward tilting of fringes, and (ii) variation of shear stress along axis
y=0. The Airy’s stress function is modified as:

22
 φ=Z ( z )=ℜ Ź + yIm Z + yIm Y (68)

Where, stress function Y ( z )=f ( ψ , χ ) , Here, Y ( z ) is a function of two another analytic functions ψ
and χ .

Sanford used theory of elasticity of Kolosov-Muskhelishvili approach to solve the problems of

complex variable to explain the additional part yIm Y .

E. Goursat in 1988 found that, it is ways possible to find a complex potential to a given Airy’s stress
function as:

φ=ℜ [ z ¿ ψ ( z ) + χ (z) ] (69)

¿
Where, z is complex conjugate.

Kolosov in 1909 after getting ψ ( z ) and χ (z ) found the displacement equation for plane stress
condition is:

3−ν ¿' ¿ ¿' ¿ (70)

2 G ( u+ iv )= ψ ( z )−z ψ ( z )− χ ( z )
1+ ν

ν
*represents the complex conjugate function of x+iy, and for plane strain condition replace by ,
1+ ν
then the stress components are

σ x +σ y =2ψ ( z )+ 2ψ ( z )=4 ℜψ ( z )
¿' ¿
' '
(71)

σ y −σ x +2i τ xy =2 [ z ¿ ψ ' ' ( z )+ χ ' ' ( z ) ] (72)

Finding ψ and χ is now left, if we know then stress and displacement filed can be established. Using
Eq. (71) and Eq. (72), the stress field can be written as follows:

{ }{ }
ℜ ( 2ψ −z ψ − χ )
' ¿ ″ ″ (73)
σ xx
σ yy = ℜ ( 2 ψ ' + z ¿ ψ ″ + χ ″ )
τ xy ℑ ( z¿ ψ ″+ χ″)

¿
The expression for shear stress can be written as by replacing z =x−iy as:

23
 ″
τ xy =x ℑψ − y ℜψ + ℑ χ
″ ″
(74)

On the axis of symmetry for y=0, τ xy =0. Therefore, x ℑ ψ ″ + ℑ χ ″ =0

Sanford defined Y =z ψ ″ + χ ″. He showed that by appropriately selecting Y , modified Westergaard

will provide most general stress function. Replacing χ ″ in terms of ψ ″ and Y in Eq. (72) yields

σ y −σ x +2i τ xy =2 [ ( z ¿−z ) ψ ″ +Y ] (75)

The stress field of σ x , σ y , τ xy can be written in terms of ψ and Y represents the generalized
Westergaard equation as:

{ }{ }
σ xx 2 ℜψ ' −2 y ℑ ψ ″ −ℜ Y (76)
' ″
σ yy = 2 ℜψ +2 y ℑ ψ + ℜY
″
τ xy −2 y ℜ ψ + ℑ Y

Eq. (76) represents the generalized equation from which the Westergaard Equation, modification
given by Irwin are derived as follows.

Case A: Westergaard stress function

On the axis of symmetry, shear stress is zero. By defining Y as above, the condition reduces to
ℑY ( z )=0 on y=0 and setting 2 ψ ' =Z this will give the conventional Westergaard solution.

{ }{ }
σ xx ℜ Z− y ℑ Z ' (77)
σ yy = ℜ Z + y ℑ Z'
τ xy − y ℜ Z'

Case B: Modified Westergaard stress function

For ℑY ( z )=0 one can also set as Y as a real constant. Assuming Y = A and setting 2 ψ ' =Z− A ,
one gets the Irwin’s modification of Westergaard equations.

{ }{ }
σ xx ℜ Z− y ℑ Z ' −2 A (78)
'
σ yy = ℜ Z+ y ℑ Z
'
τ xy − yℜZ

Case C: Generalised Westergaard stress function by Sanford.

24
For ℑY ( z )=0 , set Y such that the imaginary part is zero and 2 ψ ' =Z−Y , then the following stress
field equation shows the variation of fringe order along the crack axis.

{ }{ }
σ xx ℜ Z− y ℑ Z ' − y ℑ Y ' + 2 ℜY (79)
σ yy = ℜ Z+ y ℑ Z ' + y ℑ Y '
τ xy − y ℜ Z ' − y ℜ Y ' −ℑ Y

This stress function can also be obtained from the Airy’s stress function (see Eq. (68))as proposed by
Sanford instead of deriving these equation using Klosov-Muskhelishvili approach.

The form of the Z and Y are expressed as follows by Sanford in a form of polynomial series.

J
J−
1
(80)
Z ( z )=∑ C 2 j z 2

J=0

J
(81)
Y ( z )=∑ C 2 j+1 z J
j=0

Where, C 2 j and C 2 j+1 are real coefficient of singular solution. Based on above expression, Sanford
explain and derive the maximum stress at the crack tip as

| |
J (82)
τ m= ∑ C 2 j +1 z J
j=0

This predicts a variation of fringe order along the crack axis. Here, τ xy along the x-axis is zero, but

maximum shear stress ( τ m ) is not zero.

Both Z and Y are expressed as series function. He was able to get analytical expression for τ xy
which varies along the crack axis. Now equation of stresses given by Sandford as a generalized
Westergaard equations is

{ }
(83)
( )
cos j−
1
θ
{ }
J 1 J
2
( 32 )θ sin θ +∑ C 2 cos jθ
j−
σ x =∑ C 2 j r
j
2
−sin j− 2 j+ 1 j r −sin θ sin ( j−1 ) θ+
j=0
( j− 12 ) j=0 j

25
 { }
(84)
J 1 ( )
cos j−
1
θ J

( )
1 2
( 32 ) θ sin θ +∑ C
j−
σ y =∑ C 2 j j−
j
r 2
−sin j− 2 j+1 j r sin θ sin ( j−1 ) θ
j=0 2
( j− 12 ) j =0

( ) ( ) { }
J
1 j−2 3
1 J
sin jθ (85)
τ xy =−∑ C2 j j− r cos j− θ sin θ−∑ C2 j+1 j r j cos ( j−1 ) θ sin θ+
j=0 2 2 j =0 j

The success of this expression is that, we are able to get a series solution, which explains variations
of maximum shear stress along the crack axis and this is done in the complex domain.

Still for series solution, one question is not answered i.e. How many higher order terms will be
included for stress solution ?

4.9 William Eigen Function Approach (1957)

He considered as a crack surface as a wedge in polar coordinate system as shown in Figure ** which
is unloaded. Also known as William’s eigan function approach. Here the solution will provide the
solution for the combination of mode I and mode II.

Airy’s stress function for this problem is defined as

φ=f 1 ( r ) f 2 ( θ ) =r
λ+ 1
f (θ ) (86)

Here, to find the stress function we have to determine λ and f ( θ ). In this problem the crack surface
i.e. θ=± α are the radial distance from crack tip. So the radial stress, σ r will remain their but
tangential stress i.e. perpendicular to crack surface σ θ=0 and τ rθ =0 for θ=± α . No boundary
condition about infinity.

If stress function is known then the stresses in a plane is expressed as

2
1 ∂ϕ 1 ∂ ϕ
σ rr = +
r ∂r r 2 ∂r 2

26
 2
∂ ϕ
σ θθ = 2
∂r

τ rθ = (
−∂ 1 ∂ ϕ
∂r r ∂θ )
After substituting Eq. (86), the radial stress becomes

1 λ 1 λ+1 ″ (87)
σ rr = f ( θ ) ( λ+1 ) r + 2 r f (θ )
r r

σ θθ=λ ( λ+1 ) r
λ−1
f (θ) (88)

−λ λ+1 ' (90)

τ rθ = 2
r f (θ )
r

At boundary conditions σ θ=0 and τ rθ =0 for θ=± α

 σ θ=0 implies φ=0 i.e. f ( θ )=0

d f (θ)
 τ rθ =0 implies φ=0 i.e. f ' ( θ ) =0 or =0
dθ

f ( θ ) is an eigen function, for every value of λ one gets the corresponding eigen function. Most
general solution is the sum of all these solutions.

λ and f ( θ )are determined based on substituting the Airy’s stress function (Eq. 86) is biharmonic
equation, which gives:

4
d f d f
2
2 (90)
+ 2 ( λ −1 ) 2 + ( λ −1 ) f =0
2 2
4
dθ dθ

Solving the differential Eq. (90), we get eigen values, λ and eigen function f ( θ ). The most general
solution of Eq. (90) is given as:

f ( θ ) to satisfy the above equation is

f ( θ )=C 1 cos ( λ−1 ) θ+C 2 sin ( λ−1 ) θ+C 3 cos ( λ+1 ) θ+C 4 sin ( λ+1 ) θ (91)

For every value of λ there is a corresponding coefficient C i. These coefficients are determined from
the boundary conditions. The most general solution will be the sum of all individual solutions. We
have to apply the boundary conditions defined above.

For boundary condition (i) σ θ=0 implies. f ( θ )=0 for θ=± α

27
 C 1 cos ( λ−1 ) α +C2 sin ( λ−1 ) α +C 3 cos ( λ+1 ) α + C4 sin ( λ+ 1 ) α=0 (92)

C 1 cos ( λ−1 ) α −C 2 sin ( λ−1 ) α +C 3 cos ( λ+1 ) α −C 4 sin ( λ+1 ) α =0 (93)

d f (θ)
Boundary condition (ii) τ rθ =0 implies f ' ( θ ) =0 or =0 for or θ=± α
dθ

−C 1 ( λ−1 ) sin ( λ−1 ) α +C 2 ( λ−1 ) cos ( λ−1 ) α −C3 ( λ+1 ) sin ( λ+1 ) α +C 4 ( λ+1 ) cos ( λ+1 )(94)
α =0

C 1 ( λ−1 ) sin ( λ−1 ) α +C 2 ( λ−1 ) cos ( λ−1 ) α +C 3 ( λ +1 ) sin ( λ+1 ) α +C 4 ( λ+ 1 ) cos ( λ+1 ) α =0
(95)

Now, solving the simultaneous Eqs. 92 to 95, and after algebraic manipulations and grouping into
two categories (i) involving C 1 and C 3; (ii) C 2 and C 4. Eq (92) and (94) can be written in matrix form
as

[ cos ( λ−1 ) α cos ( λ+1 ) α

]{ } { }
C1
( λ−1 ) sin ( λ−1 ) α ( λ+1 ) sin ( λ+ 1 ) α C 3
=
0
0
(96)

For homogenous and non-trivial solution, the determinant should be zero, then

( λ+ 1 ) cos ( λ−1 ) α sin ( λ+1 ) α −( λ−1 ) cos ( λ−1 ) α sin ( λ−1 ) α =0

The above Eq. can be re-written as:

λ [ cos ( λ−1 ) α sin ( λ+1 ) α −cos ( λ−1 ) α sin ( λ−1 ) α ] +cos ( λ−1 ) α sin ( λ+1 ) α + cos ( λ+1 ) α sin ( λ−1 ) α =0

Which can be simplified as:

λ sin 2 α + sin 2 λα =0 (97)

Eq (93) and Eq.(95) can be written in matrix form as

[ sin ( λ−1 ) α sin ( λ+ 1 ) α

]{ } { }
C2
( λ−1 ) cos ( λ−1 ) α ( λ +1 ) cos ( λ+1 ) α C 4
=0
0
(98)

For non-trivial solution, the determinant should be zero as

( λ+ 1 ) cos ( λ+1 ) α sin ( λ−1 ) α −( λ−1 ) cos ( λ−1 ) α sin ( λ+1 ) α =0

−λ sin 2 α + sin 2 λα=0 (99)

To simulate a crack use α → π in Eq (98) and (99)

28
 sin 2 πλ=0 (100)

Solving Eq. (100) gives eigen values, the roots are

n
λ= for all integers value of n (0, ±1, ±2, ±3, ±4…..). Some of the roots yield physically unacceptable
2
results, if so then discard.

 For n< 0 the displacements calculated would be of the form

1
ur = f 1 (θ)
|r n2| ; when r → 0, then ur → 0. This predicts unbound displacements at the origin

r =0. So discarded.
1 1
 For n=0 the stress and strain are of the form σ ij= f 2 (θ ) and ε ij= f 3 (θ ), then the Strain
r r
1
energy density dU = f (θ) f 3(θ)
2 2
r

Take a close contour surrounding the crack-tip and integrating over the closed region results
in the observation that it would be possible to store an infinite amount of strain energy in a
finite volume. But it is not possible. Hence n=0 is not a admissible root.

 Hence only feasible solution is that the roots will be positive. So, final boundary condition

n
λ= where n=1 ,2 , 3 , 4 , 5. ...
2

−n−2
 For n is odd i.e. n=1 ,3 , 5.... C 3 n= C ; C =−C 2 n
n+2 1 n 4 n
−n−2
 For n is even i.e. n=2 , 4 , 6. ...C 3 n=−C 1 n; C 4 n= C
n+ 2 2 n

Now, Airy’s stress function in most general form for the problem undertaken. Because we have
found the two unknowns λ and f ( θ ).

{ [ ( ) ( ) ( )] [ ( ) ( ) ]}
n+2(101)
{ [
∞ n ∞ n
1+
n−2 n−2 n−2 n−2 1+
φ= ∑ r 2
C 1n cos θ− cos θ + C2 n sin θ−sin θ + ∑ r 2 C1 n c
n=1 ,3 2 n+2 2 2 2 n=2 ,4

Observation from above Airy’s stress function

 This is a series function of cos and sin terms and consists of symmetric and asymmetric part.
Where, C 1n term is symmetric and this part of stress function is for Mode I loading and C 2 n
part addresses Mode II, which is anti-symmetric

29
 2
 If use n=2 in the 2nd term in Mode I loading: r C12 (1−cos 2 θ ) represents uniaxial stress
field in x-direction.
 So, for multi-parameter solution, we have to say that take as many terms that would satisfy
your problem under consideration.

Now, the Airy’s stress function is in series form. Now for any given problem depending on the kind of
situation you have to evaluate the coefficients. As we take more coefficients as possible that can
fulfil far field boundary condition. Through Westergaard solution and subsequent modifications our
attention was towards the far field stress (uniaxial or bi-axial) and zone close to crack tip. Now
present understanding shows that stress field is more dependent on the type of crack face
displacement is more important and far field condition is secondary. So, in case of multi-parameter
solution take as many terms which can satisfy your problem under consideration. If we define our
Airy’s stress function with the terms then we can define the stress field. Once φ is known, the stress
component can be obtained from

1 ∂φ 1 ∂ φ
2
(102)
σ rr = +
r ∂ r r2 ∂ r2
2
∂ φ
σ θθ= 2
∂r

τ rθ =
[ ]
−∂ 1 ∂ φ
∂r r ∂r

Exercise: Take only C 11 term then determine σ θθ or stress field at crack tip.

Solution:

For very near tip stress field equation is polar co-ordinate

{ }
θ 3θ
5 cos −cos

{}
2 2
σr
KI θ 3θ
σθ = 3 cos +cos + ....... higher order terms
4 √ 2 πr 2 2
τ rθ θ θ
sin +sin
2 2

At crack tip θ=0, compare it with Westergaard Equation, the principal stresses are equal.

{} { } {}
σr
K I 5−1 KI 4
σθ = 3+1 + .......= 4 +...... .
4 √ 2 πr 4 √ 2 πr
τ rθ 0 0

30
For Mode II

{ }
θ 3θ
−5 sin +sin

{}
σr 2 2
K II θ 3θ
σθ = −3 sin −sin +...... .higher order terms
4 √ 2 πr 2 2
τ rθ θ 3θ
cos +3 cos
2 2

For multi-parameter solution have 6 terms in polar co-ordinate system for Mode-I loading.

{
(103)

{ [ ] [ ]}
−1 1
5 θ 1 3θ 3 θ 1 5θ
+ 4 C 12 ( cos θ ) +r 3 C 13 cos + cos
2 2 2
¿ r C 11 cos − cos
4 2 4 2 4 2 4 2
¿

{ [ ] [ ]
}
−1 1
2 3 θ 1 3θ 2 15 θ 1 5θ
¿r C 11 cos + cos + 4 C 12 sin θ+r 2 C13 cos − cos +6 r C14 cos θ

{}
σr 4 2 3 2 4 2 5 2

[ ]
3
σθ = 35 2 3θ 3 7θ 2
¿−cos 3 θ+ r C15 cos − cos +12 r C 16 ( cos 2θ−cos 4 θ )
τ rθ 4 2 7 2

{ [ ] [ ]
−1 1
2 1 1 θ 1 3θ 3 1 θ 1 5θ
¿r C11 sin + sin − 2C 12 sin 2 θ+r 2 C 13 sin − sin + 2r C 14 ( sin θ−3 sin 3
2 2 2 2 2 2 2 2 2 2

[ ]
3
5 3 3θ 3 7θ 2
¿+ r 2 C 15 sin − sin + 3r C 16 ( sin 2 θ−4 sin 4 θ )
2 2 2 2 2

Important observation: Second term is constant constant i.e. 4 C12 ( cos2 θ ) as given in Generalized

Westergaard solution.

Six term solution for Mode-II loading

31
 { }
(104)

{ [ ] [ ]
}
−1 1
5 θ 3 3θ 3 θ 5 5θ
¿−r 2
C 21 sin − sin +r 3C 23 sin + sin +2 r C24 ( sin θ+sin 3 θ )
2
4 2 4 2 4 2 4 2

[ ]
3
5 3 θ 35 7θ 2
¿+r 2 C 25 sin + sin +r C 26 6 sin 4 θ
4 2 4 2

{ [ ] [ ]
}
−1

( )
1

{}
3 θ 3θ 15 3θ 5θ 1
σ rr ¿−r 2
C 21 sin +sin + r 2 C 23 sin −sin +6 r C 24 sin θ− sin 3θ
4 2 2 4 2 2 3
σ θθ =

[ ]
3
τ rθ ¿+
35 2
4
3θ
r C 25 sin −sin
2
7θ
2
2 1
+ 12r C 26 sin 2θ− sin 4 θ
2 ( )

{ [ ] [ ]
}
−1 1
1 1 θ 3 3θ 3 1 θ 5 5θ
¿r 2
C 21 cos + cos −r 2 C 23 cos − cos −2 r C 24 ( cos θ−cos 3 θ )
2 2 2 2 2 2 2 2 2 2

[ ]
3
2 5 3 3θ 7 7θ 2
¿−r C25 cos − cos −3 r C26 ( 2 cos 2 θ−2 cos 4 θ )
2 2 2 2 2

Important observation: Second term is zero as given in Generalized Westergaard solution.

Airy’s stress function for Generalized Mode II Westergaard Stress Solution

φ=ℑ Ý II − y [ ℜY II ( z )+ ℜ Z II ( z ) ] (105)

The generalized mode II stress field is obtained as

{ }{ }
(106)
ℑ Y II + y [ ℜ Y II + ℜ Z II ] +2 ℑ Z II
' '
σx
σy = ℑ Y II − y [ ℜY II +ℜ Z II ]
' '

τ xy − y [ ℑ Y II + ℑ Z II ]+ ℜ Z II
' '

Example: The expression for Mode II can be written as done for mode I

Solution

Example: Can we write a program to generate the fringe pattern based on above equations for Mode
I and Mode II loading situation.

Solution:

Multi-parameter stress field equation by Atluri and Kobayashi in two series forms as Mode I and
Mode II components as

32
 [ ] [
(107)
{ n n n
} ( ) ( ) ( )
n n
2+ (−1 ) + cos −1 θ− −1 cos −3 θ
2 2 2 2
n n
{
2− (−1 ) + sin
2 }
{}
σ xx
{ } ( ) ( ) ( ) { }
∞ n−2 ∞ n−2
n n n n n n n
σ yy =∑ A ¿ r 2 2−(−1 ) − cos −1 θ+ −1 cos −3 θ −∑ A IIn r 2 2+ (−1 ) − sin
n n

n=1 2 2 2 2 2 n=1 2 2
τ xy
{ n n n
} ( ) ( ) ( )
n n
− (−1 ) + sin −1 θ+ −1 sin −3 θ
2 2 2 2
n n
− (−1 ) − cos
2 { }
KI −K II
A I 1= ; A II 1= ; 4 A I 2 =−σ 0 x
√2 π √2 π
As the number of terms changes, the geometric feature of the fringe changes and the number of
terms are increased the geometric fringe pattern match with the experimental data.

Example 1: Prove that the Westergaard stress function Z ( z )=ℜ Z +i ℑ Z is a Airy’s stress function.

Solution: For Z to be analytical function, the derivatives must satisfy the Cauchy-Riemann condition.

for studying crack problems.

Z ( z )= ℜ¯Z + y ℑ¯Z (1)

The relevant Airy’s stress function could be constructed based on the stress function proposed by
Westergaard. The derivation and the integral of Z ( z ) as

'
dZ
=Z d Z =Z ″ Z̄=∫ Zdz Z̄ =∫ Z̄ dz
¯
'
dz dz

Mode I stress field equation

The Airy’s stress function for a center crack with origin at the center of the crack is

Stress function selected should satisfy Bi-harmonic equation

4
∇ ϕ=0

First take the derivatives of the Airy’s stress function in terms of x

∂ϕ ∂ ¯ ∂
= ( ℜ Z )+ y ( ℑ¯Z )= ℜ¯Z + y ℑ Z
∂x ∂x ∂x
2
∂ ϕ ∂ ¯ ∂
2
= ( ℜ Z )+ y ( ℑ Z )=ℜ Z+ y ℑ Z '
∂x ∂ x ∂ x
3
∂ ϕ ∂ ∂ (
ℑ Z ) =ℜ Z + y ℑ Z
' ' ″
3
= ( ℜ Z )+ y
∂x ∂ x ∂ x
33
 4
∂ ϕ ∂ ( ∂ (
ℜ Z )+ y ℑ Z )=ℜ Z + y ℑ Z
' ″ ″ ‴
4
=
∂x ∂ x ∂ x

Now, take the derivatives of the Airy’s stress function in terms of y

∂ϕ ∂ ¯ ∂y ∂
= ( ℜ Z ) + ℑ¯Z + y ( ℑ Z )=− ℑ¯Z + ℑ¯Z + y ℜ Z= y ℜ Z
∂y ∂y ∂y ∂y
2
∂ ϕ ∂y ∂
2
=ℜ Z +y ( ℜ Z )=ℜ Z− y ℑ Z'
∂y ∂ y ∂ y
3
∂ ϕ ∂ ∂ (
y ℑ Z ) =−ℑ Z −ℑ Z + y ℜ Z =−2 ℑ Z + y ℜ Z
' ' ' ″ ' ″
3
= ℜ Z−
∂y ∂ y ∂ y
4
∂ ϕ ∂ ∂ (
y ℜ Z )=−2 ℜ Z + ℜ Z
' ″ ″ ″
4
=−2 ℑZ +
∂y ∂y ∂y

Therefore the function selected is satisfying bi-harmonic equation.

Westergaard solution of stress field should satisfy the above boundary condition. The function
suggested by Westergaard is

Verification of stress function

Singularity means strength will be different for different condition. Suitability of stress field equation
in FM was studied by the technique of photoelasticity. In this, one observes contours of constant
principal stress difference as fringes known as isochromatics. These isochromatics are symmetric
about x- and y- axis.

But experimental isochromatics shows that these fringes are tilted forward and the angle of tilt
varies and approaches 900 as one goes closure to the crack tip. For small crack this variations are
small but for large crack this variations in significant.

SIF of More complex cases

Stress intensity factor (SIF) is very important to assess the fracture behaviour, it can be determined
from the stress function based on the following expression if Z ( z 0 ) is the stress function of the
problem defined w.r.to the crack tip then

K= lim √ 2 π z 0 Z ( z 0 )
z0 → 0

The SIF for a central crack in an infinite plate loaded uniaxially is determined as:

K I =σ √ πa

34
Concentrated load

Case-I: When a concentrated load, P is applied to open the crack. The stress function for this type
of problem is given as

Pa
Z1 =
πz √ z 2−a 2

Shifting the origin z=z 0 +a

Pa Pa
Z1 = =
√
π ( z 0 +a ) ( z 0 +a ) −a
2 2
π ( z 0 + a ) √ z 0 ( z 0 +a )

2a

Pa
K I =lim √ 2 π z 0 z 1 ( z 0 )= lim √ 2 π z 0
( )
0.5
z0 →0 z0 → 0 z0
π ( z 0 + a ) √ z 0 1+
2a

P
K I=
√ πa
Case-II: Symmetric stress is acting across the crack by concentrated load, P

The stress function is defined as

( )
2 2 0.5
2 Pz a −s
Z1 =
π √ z −s
2 2
2 2
z −a

P P

x
s s
P P

2a

Shift origin at the crack tip z=z 0 +a

2 P ( z 0 + a ) √ a2−s2
( )
2 P ( z 0 +a ) a −s
2 2 0.5
Z1 = =
√ π [ z 0 +a +2 a z 0−s ] √ z 20 +2 a z 0
2 2
π ( z 0+ a ) −s ( z 0 +a ) −a
2 2 2 2 2

35
From definition of SIF

√
2 P ( z 0 +a ) √a 2−s2 a
K I =lim √2 π z 0 Z 1 ( z 0 )=lim √2 π z 0 =2 P

√ π ( a −s )
2 2
z0 →0 z0 →0 z0
π [ z 0 +a +2 a z 0−s ] √ 2 a z0
2 2 2
1+
2a

P
For concentrated load at center, put s=0 in the above expression as P → .
2

Case-III: Asymmetric stress is acting across the crack by concentrated load, P

In this case the stress at both the crack tip will be different. Crack closure to the point of application
of load will experience more SIF than the other. The stress function is defined as:

( )
2 2 0.5
P a −s
Z1 =
π ( z−s ) z 2−a2

x
s
P
2a

[ ]
1
P a+ s
K AI = 2

√ πa a−s

[ ]
1
P a−s
B 2
K = I
√ πa a+ s
For concentrated load at center, put s=0 and the expression reduces to case I.

Case-IV: Triangular symmetrical load with maximum value, P0

dp= ( ) a−s
a
P0 ds

d K I =2
a−s
aπ
πa
P 0 2 2 ds
a −s √
√ √
a a
a−s πa 2 P 0 a a−s
K I =∫ d K I =∫ 2 P0 2 2 ds= ∫ ds
0 0 aπ a −s √ aπ 0 a+ s

√ [ √ ]
a a
a−s a−s π
( )
a
∫ ds=[ √ ( a+ s )( a−s ) ]0+ 2 a sin−1 =a −1
0 a+ s 2a 0 2

36
When P0 is constant

Exercises

1. In a large plate, a crack of length 2a is inclined with an angle with x-axis. The plate is loaded in y-
direction. Find (i) the stress intensity factor, (ii) for σ =80 MPa, 2 a=20mm and α =3 00 ,
Determine K I , and K II , (iii) stress at point H , if r =1 mm and θ=4 5 0.
2. Determine the critical crack length in a central crack in an infinite plate, loaded in Mode I, if the
critical intensity factor K Ic =60 MPa √ m and far field stress is 120 MPa.

lim f ( z +dz )−f ( z )

dw ' ' Δz→ 0
=w =f ( z )=
dz Δz
z=x +iy is a complex number.
w=f ( z ) =u ( x , y ) +iv ( x , y ) : consists of real part and imaginary part
For derivatives to exists no matter how Δz approaches zero, it is necessary that the limit of the
quotient be the same.
By definition,
lim Δw
dw Δz→ 0
=
dz Δx

Answers:

37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Fracture Mechanics Answers

i) The strain energy for a central crack of length a, in an infinite plate of unit thickness subjected to
uniaxial stress σ, is:
U = πaσ² / E, where E is the Young’s modulus of the material.

ii) Let f(x). For the derivative of f(x) to exist, the necessary and sufficient conditions are:
1. Continuity of f(x) at the point.
2. Differentiability at that point.

iii) For a given Airy’s stress function φ, the stress component σ_xx can be obtained by:
σ_xx = ∂²φ / ∂y².

iv) The bi-harmonic equation in expanded form in a Cartesian coordinate system is given by:
∂⁴φ / ∂x⁴ + 2∂⁴φ / ∂x²∂y² + ∂⁴φ / ∂y⁴ = 0.

v) For a plane stress problem, the necessary boundary conditions are:

1. Traction boundary conditions: Stresses σ_xx, σ_yy, and σ_xy must satisfy the applied loads on the
boundaries.
2. Displacement boundary conditions: Displacement components u_x and u_y must satisfy the
constraints on the boundary.

vi) The Airy’s stress function for a problem of crack under Mode I loading, with the center of the
crack as the origin, is:
φ = Re[K_I √(r / 2π) cos(θ / 2)].

vii) In the absence of body forces, for plane stress and plane strain problems, the compatibility
condition reduces to:
∂²ε_xx / ∂y² + ∂²ε_yy / ∂x² - 2∂²ε_xy / ∂x∂y = 0.

viii) The Airy’s stress function for a beam with uniformly distributed load is given by φ = -(q/24)x²(3Lx
- 4x²). Then q is:
The intensity of the uniformly distributed load per unit length.

ix) The boundary condition for the Mode I crack problem, with the center of the crack as the origin,
is:
1. Along x-axis (crack surface): σ_yy = 0, τ_xy = 0.
2. Far from the crack tip: σ_yy = σ_∞, τ_xy = 0.

x) The total strain energy of the system in terms of compliance is:

U = (1 / 2)P²C, where P is the applied load and C is the compliance of the system.

52
xi) The Airy’s stress function for Mode II loading is:
φ = Im[K_II √(r / 2π) sin(θ / 2)].

xii) Westergaard stress function for Mode II loading with crack center as the origin is:
φ(z) = Im[K_II / √(2πz)], where z = x + iy.

xiii) Westergaard stress function for Mode II loading with crack center as the origin, the stress field
after shifting the origin to the crack tip becomes:
φ(z') = Im[K_II / √(2π(z' - a))], where z' = x + iy and a is the crack length.

xiv) As per Westergaard, the function φ(z) and ψ(z) for Mode III is:
For Mode III, φ(z) = 0, and ψ(z) corresponds to the out-of-plane displacement fields:
ψ(z) = K_III / √(2πz), where K_III is the Mode III stress intensity factor.

53