Module 2 With Solutions

Download as pdf or txt
Download as pdf or txt
You are on page 1of 15

Module 2

Kinematics of deformation and Strain

Learning Objectives
• develop a mathematical description of the local state of deformation at a material point

• understand the tensorial character of the resulting strain tensor

• distinguish between a compatible and an incompatible strain field and understand the
mathematical requirements for strain compatibility

• describe the local state of strain from experimental strain-gage measurements

• understand the limitations of the linearized theory and discern situations where non-
linear effects need to be considered.

2.1 Local state of deformation at a material point


Readings: BC 1.4.1

Deformation described by deformation mapping:

x0 = ϕ(x) (2.1)

We seek to characterize the local state of deformation of the material in a neighborhood of


a point P . Consider two points P and Q in the undeformed:

P : x = x1 e1 + x2 e2 + x3 e3 = xi ei (2.2)
Q : x + dx = (xi + dxi )ei (2.3)

and deformed

P 0 : x0 = ϕ1 (x)e1 + ϕ2 (x)e2 + ϕ3 (x)e3 = ϕi (x)ei (2.4)

Q0 : x0 + dx0 = ϕi (x) + dϕi ei



(2.5)

29
30 MODULE 2. KINEMATICS OF DEFORMATION AND STRAIN

Q′

dx′
Q P′
dx u

P
x′
x
e3

e2
e1

Figure 2.1: Kinematics of deformable bodies

configurations. In this expression,

dx0 = dϕi ei (2.6)

Expressing the differentials dϕi in terms of the partial derivatives of the functions ϕi (xj ej ):
∂ϕ1 ∂ϕ1 ∂ϕ1
dϕ1 = dx1 + dx2 + dx3 , (2.7)
∂x1 ∂x2 ∂x3
and similarly for dϕ2 , dϕ3 , in index notation:
∂ϕi
dϕi = dxj (2.8)
∂xj
Replacing in equation (2.5):
 ∂ϕi 
Q0 : x0 + dx0 = ϕi + dxj ei (2.9)
∂xj
∂ϕi
dx0 i = dxj ei (2.10)
∂xj
−→
We now try to compute the change in length of the segment P Q which deformed into segment
−− →
P 0 Q0 . Undeformed length (to the square):

ds2 = kdxk2 = dx · dx = dxi dxi (2.11)

Deformed length (to the square):


∂ϕi ∂ϕi
(ds0 )2 = kdx0 k2 = dx0 · dx0 = dxj dxk (2.12)
∂xj ∂xk
2.1. LOCAL STATE OF DEFORMATION AT A MATERIAL POINT 31

−→
The change in length of segment P Q is then given by the difference between equations (2.12)
and (2.11):
∂ϕi ∂ϕi
(ds0 )2 − ds2 = dxj dxk − dxi dxi (2.13)
∂xj ∂xk
We want to extract as common factor the differentials. To this end we observe that:
dxi dxi = dxj dxk δjk (2.14)
Then:
∂ϕi ∂ϕi
(ds0 )2 − ds2 = dxj dxk − dxj dxk δjk
∂xj ∂xk
 ∂ϕ ∂ϕ 
i i (2.15)
= − δjk dxj dxk
∂x ∂x
| j {zk }
2jk : Green-Lagrange strain tensor
Assume that the deformation mapping ϕ(x) has the form:
ϕ(x) = x + u (2.16)
where u is the displacement field. Then,
∂ϕi ∂xi ∂ui ∂ui
= + = δij + (2.17)
∂xj ∂xj ∂xj ∂xj
and the Green-Lagrange strain tensor becomes:
 ∂um  ∂um 
2ij = δmi + δmj + − δij
∂xi ∂xj
(2.18)
∂ui ∂uj ∂um ∂um
=6 δij + + + − 6 δij
∂xj ∂xi ∂xi ∂xj

1  ∂ui ∂uj ∂um ∂um 


Green-Lagrange strain tensor : ij = + + (2.19)
2 ∂xj ∂xi ∂xi ∂xj
When the absolute values of the derivatives of the displacement field are much smaller than
1, their products (nonlinear part of the strain) are even smaller and we’ll neglect them. We
will make this assumption throughout this course (See accompanying Mathematica notebook
evaluating the limits of this assumption). Mathematically:
∂u ∂um ∂um
i
1 ⇒ ∼0 (2.20)
∂xj ∂xi ∂xj

We will define the linear part of the Green-Lagrange strain tensor as the small strain tensor :
1  ∂ui ∂uj 
ij = + (2.21)
2 ∂xj ∂xi
32 MODULE 2. KINEMATICS OF DEFORMATION AND STRAIN

Concept Question 2.1.1. Strain fields from displacements.


The purpose of this exercice is to determine strain fields from given displacements.

1. Find the linear and nonlinear strain fields associated with the following displacements

ua1 = x1 x2 (2 − x1 ) − c1 x2 + c2 x32 ,
x2
ua2 = −c3 x22 (1 − x1 ) − (3 − x1 ) 1 − c1 x1 .
3

2. Find the linear strain fields associated with the following displacements

ub1 = x31 x2 + 2c1 c32 x1 + 3c1 c22 x1 x2 − c1 x1 x32 ,


3 1 3
ub2 = −2c32 x2 − c22 x22 + x42 − c1 x21 x22 .
2 4 2
Solution: The expression to calculate the nonlinear (nl) strains in function of the
displacements is  
nl 1 ∂ui ∂uj ∂um ∂um
εij = + + . (2.22)
2 ∂xj ∂xi ∂xi ∂xj

When the derivatives of the displacement


  components are small in comparison to one,
∂um ∂um ∂um ∂um
i.e. ∂xi , ∂xj  1, the product ∂xi ∂xj can be neglected, and the previous equation
simplifies to the following linear (l) expression
 
l 1 ∂ui ∂uj
εij = + . (2.23)
2 ∂xj ∂xi

When we apply the Equation 2.23 to the field (ua1 , ua2 ), we obtain the following linear
(l) strain tensor " #
(3c2 +c3 ) 2
2x 2 (1 − x 1 ) −c1 + x
εla = 2 2
.
−c1 + (3c22+c3 ) x22 −2c3 x2 (1 − x1 )

On the other hand, the Equation 2.22 allows us to calculate the nonlinear (nl) strain
tensor for the field (ua1 , ua2 )  nl nl 
nl ε ε12
εa = 11 ,
ε12 εnl
nl
22

where
1 2
εnl
11 = 2x2 (1 − x1 ) [1 + x2 (1 − x1 )] + −c1 + c3 x22 − x1 (2 − x1 ) ,
2
1 2
εnl
22 = 2c3 x2 (1 − x1 ) [−1 + c3 x2 (1 − x1 )] + −c1 + 3c2 x22 + x1 (2 − x1 ) ,
2
(3c2 + c3 ) 2
εnl
12 = −c1 + x2
2
+x2 (1 − x1 ) x1 (2 − x1 )(1 + c3 ) + c1 (−1 + c3 ) + (3c2 − c23 )x22 .

2.2. TRANSFORMATION OF STRAIN COMPONENTS 33

The linear (l) strain tensor for the displacement field (ub1 , ub2 ) is
1 3
x1 + 23 c1 c22 x1 − 3c1 x1 x22
 2 
l 3x1 x2 + 2c1 c32 + 3c1 c22 x2 − c1 x32 2
εb = 1 3 .
x + 32 c1 c22 x1 − 3c1 x1 x22
2 1
−3c1 x21 x2 + x32 − 3c22 x2 − 2c32

2.2 Transformation of strain components


Readings: BC 1.5.1, 1.6.2, 1.5.2, 1.6.3, 1.6.4

Given: ij , ei and a new basis ẽk , determine the components of strain in the new basis ˜kl
1  ∂ ũi ∂ ũj 
˜ij = + (2.24)
2 ∂ x̃j ∂ x̃i
We want to express the quantities with tilde on the right-hand side in terms of their non-tilde
counterparts. Start by applying the chain rule of differentiation:
∂ ũi ∂ ũi ∂xk
= (2.25)
∂ x̃j ∂xk ∂ x̃j
Transform the displacement components:
u = ũm ẽm = ul el (2.26)
ũm (ẽm · ẽi ) = ul (el · ẽi ) (2.27)
ũm δmi = ul (el · ẽi ) (2.28)
ũi = ul (el · ẽi ) (2.29)
take the derivative of ũi with respect to xk , as required by equation (2.25):
∂ ũi ∂ul
= (el · ẽi ) (2.30)
∂xk ∂xk
and take the derivative of the reverse transformation of the components of the position vector
x:
x = xj ej = x̃k ẽk (2.31)
xj (ej · ei ) = x̃k (ẽk · ei ) (2.32)
xj δji = x̃k (ẽk · ei ) (2.33)
xi = x̃k (ẽk · ei ) (2.34)
∂xi ∂ x̃k
= (ẽk · ei ) = δkj (ẽk · ei ) = (ẽj · ei ) (2.35)
∂ x̃j ∂ x̃j
Replacing equations (2.30) and (2.35) in (2.25):
∂ ũi ∂ ũi ∂xk ∂ul
= = (el · ẽi )(ẽj · ek ) (2.36)
∂ x̃j ∂xk ∂ x̃j ∂xk
34 MODULE 2. KINEMATICS OF DEFORMATION AND STRAIN

Replacing in equation (2.24):

1 h ∂ul ∂ul i
˜ij = (el · ẽi )(ẽj · ek ) + (el · ẽj )(ẽi · ek ) (2.37)
2 ∂xk ∂xk

Exchange indices l and k in second term:

1 h ∂ul ∂uk i
˜ij = (el · ẽi )(ẽj · ek ) + (ek · ẽj )(ẽi · el )
2 ∂xk ∂xl
(2.38)
1  ∂ul ∂uk 
= + (el · ẽi )(ẽj · ek )
2 ∂xk ∂xl

Or, finally:

˜ij = lk (el · ẽi )(ẽj · ek ) (2.39)

Concept Question 2.2.1. 2d relations for strain tensor rotation.


In two dimensions, let us consider two basis ei and ẽk such that ẽ1 is oriented at an angle
θ with respect to the axis e1 . ij and ˜ij are, respectively, the components of a strain tensor
 expressed in the ei and ẽk bases (i.e. they correspond to the same state of deformation.
Using the following expression introduced in the class notes,

˜ij = lk (el · ẽi )(ẽj · ek )

derive the following relations:

˜11 = 11 cos2 θ + 22 sin2 θ + 12 sin 2θ


˜22 = 11 sin2 θ + 22 cos2 θ − 12 sin 2θ
11 − 22
˜12 = − sin 2θ + 12 cos 2θ
2

Note: It is also usual to find the following expressions for ˜11 and ˜22 in textbooks:

11 + 22 11 − 22


˜11 = + cos 2θ + 12 sin 2θ
2 2
11 + 22 22 − 11
˜22 = + cos 2θ − 12 sin 2θ
2 2

Solution: First, let us recall the following trigonometric relations between the
vectors of ei and ẽk :

e1 · ẽ1 = cos θ e1 · ẽ2 = − sin θ


e2 · ẽ2 = cos θ e2 · ẽ1 = sin θ
2.2. TRANSFORMATION OF STRAIN COMPONENTS 35

Using (2.39), it is possible to write the following:

˜11 = 11 (e1 · ẽ1 )2 + 22 (e2 · ẽ2 )2 + 212 (e1 · ẽ1 )(ẽ1 · e2 )
= 11 cos2 θ + 22 sin2 θ + 12 sin 2θ
˜22 = 11 (e1 · ẽ2 )2 + 22 (e2 · ẽ2 )2 + 212 (e1 · ẽ2 )(ẽ2 · e2 )
= 11 sin2 θ + 22 cos2 θ − 12 sin 2θ
˜22 = 11 (e1 · ẽ1 )(ẽ2 · e1 ) + 22 (e2 · ẽ1 )(ẽ2 · e2 )
+ 12 (e1 · ẽ1 )(ẽ2 · e2 ) + 21 (e2 · ẽ1 )(ẽ2 · e1 )
11 22
= − sin 2θ + sin 2θ + 12 (cos2 θ − sin2 θ)
2 2
11 − 22
= − sin 2θ + 12 cos 2θ
2
The expresssions given in the remark can be derived from these using the following trigono-
metric relations:
1 + cos 2θ 1 − cos 2θ
cos2 θ = sin2 θ =
2 2

Concept Question 2.2.2. Principal strains and maximum shear strain in 2d.
Using the relations introduced in Problem 2.2.1, show that given the components ij of
a 2d strain tensor in a basis ei :

1. The principal strains can be computed as follows:


s 2
11 + 22 11 − 22
1,2 = ± + 212
2 2

and the principal directions of strain for angles with respect to e1 satisfy:
212
tan 2θp =
11 − 22

2. The maximum shear strain can be computed as follows:


s 2
max 11 − 22
12 = + 212
2

and the normal of the planes of maximum shear form angles with respect to e1
11 − 22
tan 2θs = − .
212
Conclude that the direction of maximum shear is always oriented at an angle equal to
45o with respect to the principal directions of strain.
36 MODULE 2. KINEMATICS OF DEFORMATION AND STRAIN

Solution: Principal strains: The characteristic polynomial χ() corresponding to


the strain tensor components ij is:
χ() = det(ij − δij ) = (11 − )(22 − ) − 212
= 2 − (11 + 22 ) + (11 22 − 212 )
The roots of the characteristic polynomial are:
s 2
11 + 22 11 − 22
1,2 = ± + 212
2 2
To find the angle θp formed by the principal directions and the basis vecto e1 , use the fact
that the shear strains vanish in principal directions:
11 − 22 212
0=− sin 2θp + 12 cos 2θp ⇒ tan 2θp =
2 11 − 22
Maximum shear strain: The maximum shear strain can be found by simply finding the
value of the argument θ in the expression for transforming the shear strain component which
makes the derivative of 12 with respect to θ vanish:
11 − 22
max
12 = − sin 2θs + 12 cos 2θs
2 
∂12 11 − 22 s s
= −2 cos 2θ + 12 sin 2θ = 0
∂θ 2
By taking the square of the two previous equations and summing them, it is easy to show
that:  2
max 2 11 − 22
12 = + 212
2
The second equation leads directly to the angular relation:
22 − 11
tan 2θs =
212
From the trigonometric relation: tan (α + π2 ) = − tan1 α it is also easy to see that:
  π  1 11 − 22
tan 2 θp + =− = − = tan 2θs
4 tan 2θp 212
Thus, proving that θs = θp + π4 .

Concept Question 2.2.3. Strain tensor rotation.


Consider the following problem of a square of unit area subject to the following strain
components in the basis given, Figure 2.3(a). :
11 = 3.4 × 10−4 22 = 1.1 × 10−4 12 = 9.0 × 10−5
Since the square has its edge of unit length, the changes in length in the directions e1 and
e2 are directly equal to 11 and 22 , respectively. The shear strain 12 is equal to half of the
decrease in angle in A (for infinitesimal angles).
2.2. TRANSFORMATION OF STRAIN COMPONENTS 37

e2 2ǫ12 e2

ẽa2
ǫ22

1 ẽa1

1 30o
A e1
ǫ11
A e1
(a) Deformed unit square (b) different initial orientation of the unit square

Figure 2.2: Deformed unit square and oriented new initial configuration.

e2 2ǫ12 e2

ẽa2
ǫ22

1 ẽa1

1 30o
A e1
ǫ11
A e1
(a) Deformed unit square (b) different initial orientation of the unit square

Figure 2.3: Deformed unit square and oriented new initial configuration.
38 MODULE 2. KINEMATICS OF DEFORMATION AND STRAIN

1. Determine the strain components on a square initialy oriented at an angle equal to


30o to the axis e1 as shown on Figure 2.3(b). Sketch in this case, the deformed
configuration.

2. Determine the principal strains and sketch the deformed configuration.

3. Determine the maximum shear strain and sketch the deformed configuration.
Solution: For the solution of this problem, we are going to use extensively the
relations introduced in Problem 2.2.1. Let us first compute the two following ratios:
11 + 22 11 − 22
= 2.25 × 10−4 = 1.15 × 10−4
2 2
Orientation at an angle θ = 30o : The value of the strain tensor in the basis eai are as
follows:
11 + 22 11 − 22
˜a11 = + cos 2θ + 12 sin 2θ
2 2 √
−4 −4 1 −5 3
= 2.25 × 10 + 1.15 × 10 × + 9.0 × 10 × = 3.6 × 10−4
2 2
11 + 22 22 − 11
˜a22 = + cos 2θ − 12 sin 2θ
2 2 √
−4 −4 1 −5 3
= 2.25 × 10 − 1.15 × 10 × + 9.0 × 10 × = 9.0 × 10−5
2 2
11 − 22
˜a12 = − sin 2θ + 12 cos 2θ
2 √
−4 3 1
= −1.15 × 10 × + 9.0 × 10−5 × = −5.5 × 10−5
2 2
Figure 2.4(a) shows the deformed configuration corresponding to this case.
Principal strains: Using the relation introduced in Problem 2.2.2, the principal strains
are: s 2
= 3.7 × 10−4 : 1

11 + 22 11 − 22 2
1,2 = ± + 12 =
2 2 = 8.0 × 10−5 : 2
and their respective direction can be computed as:

b 212 θ1b ≈ 19o
tan 2θ = ⇒
11 − 22 θ2b ≈ 109o

In order to find which of the two angles solution of the equation above is associated with
which value of principal strain, one can test these values of θb in the expression of ˜11 given in
Problem 2.2.1. Figure 2.4(b) shows the deformed configuration corresponding to this case.
Maximum shear strain: Following the relations introduced in Problem 2.2.2, we can
compute the absolute value of the maximal shear strain as:
s 2
11 − 22
q
max
12 = + 12 = (1.15 × 10−4 )2 + (9.0 × 10−5 )2 = 1.46 × 10−4 .
2
2
2.3. COMPATIBILITY OF STRAINS 39

Using the fact that the maximum shear direction is oriented at an angle of 45o to one of the
principal strain direction, let us consider the case of maximum shear obtained for an angle
θc = 19o + 45o = 64o starting from e1 . We obtain 12 (θc ) = −max
12 and contend that for
this angle the maximum negative shear strain is obtained. Figure 2.4(c) shows the deformed
configuration corresponding to this case.

e2
2ǫ̃a12

ẽa2 ǫ̃a22

ẽa1

30o
ǫ̃a11

A e1
(a) Deformed configuration with initial orientation of 30o

e2 ǫ̃c22 e2
ẽ2b
ẽc1
b
ǫ̃22
2ǫ̃c12

ẽc2
ǫ̃c11
ẽ1b
19o 65o
b
ǫ̃11
A e1 A e1
(b) Deformed configuration of principal strain (c) Deformed configuration of maximum shear

Figure 2.4: Several deformed configuration of a unit square.

2.3 Compatibility of strains


Readings: BC 1.8

Given displacement field u, expression (2.21) allows to compute the strains components
ij . How does one answer the reverse question? Note analogy with potential-gradient field. In
this section, we will restrain ourselves to small perturbation theory where the displacements
and the rotations of a deformable solid are infinitesimal. Let us first restrict the analysis to
40 MODULE 2. KINEMATICS OF DEFORMATION AND STRAIN

two dimensions. The small strain tensor is defined as the symmetric part of the displacement
∂ui
gradient ∂xj
:
 
1 ∂ui ∂uj
ij = + (2.40)
2 ∂xj ∂xi
∂ui
We define the skew-symmetric part of ∂xj
as:
 
1 ∂ui ∂uj
ωij := − (2.41)
2 ∂xj ∂xi

Concept Question 2.3.1. Properties of ωij

1. Verify that ωji = −ωij , i.e. ωij is skew-symmetric Solution:


   
1 ∂uj ∂ui 1 ∂ui ∂uj
ωji = − =− − = ωij
2 ∂xi ∂xj 2 ∂xj ∂xi

∂ui
2. Verify that ij + ωij = ∂xj
Solution:
   
1 ∂ui ∂uj 1 ∂ui ∂uj ∂ui
ij + ωij = + + − =
2 ∂xj ∂xi 2 ∂xj ∂xi ∂xj

For the two-dimensional setting, the components are as follows:


 
1 ∂u1 ∂u2
ω11 = ω22 = 0, ω12 = −ω21 = − (2.42)
2 ∂x2 ∂x1

We have seen in a previous section of this module, that ij describes the change of length
of a vector dx due to deformation. We will now see that ωij represents the infinitesimal
rotation of the vector dx from the initial to the deformed configuration. ωij is thus named
the infinitesimal rotation tensor.
−→
Consider an infinitesimal rotation of a vector P Q in the neighborhood of a point P .
For this transformation, the strain tensor  vanishes. Such a transformation can only be a
−→ −−→
rotation of P Q into P Q0 by an angle θ ( θ  1) as depicted in the following figure:

dx′ Q′
θ
P Q
dx

Figure 2.5: infinitesimal rotation of a vector dx


2.3. COMPATIBILITY OF STRAINS 41

From Figure 2.5, it is possible to express dx0 in terms of θ and dx:


   
0 cos θ sin θ 1 θ
dx = dx ≈ dx (2.43)
− sin θ cos θ −θ 1

Altenatively, from (2.17), it is possible to express dx0 in terms of ω12 and dx:
 
0 1 ω12
dx = (δij + ωij ) dxj = dx (2.44)
−ω12 1

By identification of the transformation matrix components, we conclude that ω12 = −ω21 ≈ θ


corresponds indeed to an infinitesimal rotation in the plane of normal e3 . Similar conclusions
can be drawn on the remaining components: ω31 = −ω13 corresponds to an infinitesimal
rotation in the plane of normal e2 and ω23 = −ω32 corresponds to an infinitesimal rotation
in the plane of normal e1 .
The compatibility of strain is intricately related to the continuity of infinitesimal rota-
tions. In two dimensions, this can be readily expressed by requiring the equality of the mixed
∂ 2 ω12 ∂ 2 ω12
partials of ω12 : ∂x 1 ∂x2
= ∂x 2 ∂x1
. To this end, differentiate ω12 with respect to x1 :
 2
∂ 2 u2

∂ω12 1 ∂ u1
= − (2.45)
∂x1 2 ∂x2 ∂x1 ∂x21
 2
∂ 2 u1
 2
∂ 2 u1

1 ∂ u1 ∂ u2
= + − + (2.46)
2 ∂x2 ∂x1 ∂x2 ∂x1 ∂x21 ∂x2 ∂x1
∂11 ∂12
= − (2.47)
∂x2 ∂x1
and now with respect to x2 :

∂ 2 ω12 ∂ 2 11 ∂ 2 12


= − (2.48)
∂x1 ∂x2 ∂x2 2 ∂x1 ∂x2
Similarly, we can find that:
∂ω12 ∂12 ∂22
= − (2.49)
∂x2 ∂x2 ∂x1
which differentiated with respect to x1 gives:

∂ 2 ω12 ∂ 2 12 ∂ 2 22


= − (2.50)
∂x2 ∂x1 ∂x2 ∂x1 ∂x1 2
Equating the mixed partials in equations (2.48) and (2.50) we obtain:

∂ 2 12 ∂ 2 11 ∂ 2 22


2 = + (2.51)
∂x1 ∂x2 ∂x22 ∂x21

The following concept question generalizes this result to obtain all of the equations of
strain compatibility in three dimensions.
42 MODULE 2. KINEMATICS OF DEFORMATION AND STRAIN

Concept Question 2.3.2. Strain compatibility equation in 3d.


The purpose of this exercise is to derive the strain compatibility equations in 3d using
the approach followed in class for the 2d case.
1. Apply the equality of mixed partials to the small rotation tensor:
∂ 2 ωij ∂ 2 ωij
=
∂xk ∂xl ∂xl ∂xk
and show that the following relations hold:
∂ 2 ik ∂ 2 jk ∂ 2 il ∂ 2 jl
− = − (2.52)
∂xj ∂xl ∂xi ∂xl ∂xj ∂xk ∂xi ∂xk

2. How many relations are defined by (2.52) and how many strain compatibility equations
are required in order to ensure that a unique displacement may be computed from a
given small strain tensor?
3. Notice that for i = j or l = k, (2.52) is automatically verified. How many non-trivial
relations can be derived from (2.52)? Are all these relation independant?
Solution: Let us remind first that the small rotation tensor is defined as:
1
ωij = (ui,j − uj,i )
2
Thus, the gradient of small rotation reads:
1
ωij,k = (ui,jk − uj,ik )
2
By adding and substracting uk,ij form the right-hand side of the previous relation, it is to
express the gradient of small rotation only in terms of the derivatives of the componenent
of the small strain tensor:
1  1 
ωij,k = ui,jk + uk,ij − (uj,ik − uk,ij ) = ik,j − jk,i
2 | {z } | {z } 2
2ik,j 2jk,i

Thus, the mixed derivatives: ωij,kl and ωij,lk of the small rotation tensor have the following
expressions:
1 
ωij,kl = ik,jl − jk,il
2
1 
ωij,lk = il,jk − jl,ik
2
The equality of mixed partials implies:
ik,jl − jk,il = il,jk − jl,ik

Since i, j, k, l can take any value in {1, 2, 3} respectively, (2.52) comprises 34 = 81 relations.
It is easy to verify that the only non-trivial relations from (2.52) can be obtained for i 6= j
and k 6= l.
2.3. COMPATIBILITY OF STRAINS 43

i 6= j and k 6= l
1 2 and 1 2
2 3 and 2 3
3 1 and 2 1
1 2 and 1 3
2 3 and 2 1
3 1 and 3 2
Thus, obtaining the 6 following relations:

11,22 + 22,11 = 212,12


22,33 + 33,22 = 223,23
33,11 + 11,33 = 231,31
12,23 + 23,12 = 22,31 + 31,22
23,31 + 31,23 = 33,12 + 12,33
31,12 + 12,31 = 11,23 + 23,11

These six relations are linearly dependent and it is possible to show that if only three are
them are verifed then the remaining three are.

You might also like