EN0175

10 / 17 / 06

Summary of elementary strain concepts and their generalizations to 3D tensors:

1D

3D

(1) =

l0

l l0 = 1 l0

(U I )

(2 ) =

l l0 = 1 1 l

(I V )
1

l 2 l02 1 2 (3 ) = = 1 2l02 2

)
)

1 (C I ) 2 1 1 I B 2

(4 ) =

l 2 l02 1 = 1 2 2l 2 2

)
1 (C I ) = 1 I B 1 2 2 1 v Tv or = u + u 2

For small strain:

= U I = I V 1 = ij =
1 (ui, j + u j ,i ) 2

(1) = (2 ) = (3 ) = (4 ) =

6 equations:

11 =
1 equation:

u1 u 2 u , 22 = , 33 = 3 x1 x2 x3
1 u u 1 u u

u x

1 1 3 + 2 12 = , 13 = 2 x + x , 2 x x 1 1 2 3

23 =

1 u2 u3 + 2 x3 x2

For small strain and small rotation, the rigid body rotation part is analyzed as follows:

R = I + where is defined as the small rotation tensor. v RU = (I + )(I + ) I + + = F = I + u

= u = u

v 1 v v 1 v v u + T u = u T u 2 2

T = (antisymmetric tensor)

EN0175

10 / 17 / 06

Alternative way of deriving small strain tensor:

x2
u 2 + du 2

dl 2

dl1

1
u2

dl 20
dl10

u1

u1 + du1

x1
Normal strain:

11 22

dl dl10 = 1 = dl10 dl dl 20 = 2 = dl 20

(dl10 + du1 )2 + du 22
dl10

dl10

u1 u 2 = 1 + x + x 1 1 u 2 = 1 + x 2
2

u1 1 x 1
2

(dl 20 + du 2 )2 + du12
dl 20

dl 20

u1 u 2 + x 1 x 2 2

Shear strain:

12 = 1 + 2
tan 1 = tan 2 =
u u2 x1 du2 = = 2 = 1 dl10 + du1 1 + u1 x1 x1 u u1 x2 du1 = = 1 = 2 dl20 + du2 1 + u2 x2 x2 u1 u2 + = 2 12 x2 x1

12 = 1 + 2 =

EN0175

10 / 17 / 06

Volume change:

u1 u1 x1 x2 u 2 u dV = det (F ) = 1+ 2 x1 x2 dV0 u3 u3 x1 x2 1+
If

u1 x3 u 2 x3 u 1+ 3 x3

ui << 1 , x j 1 + u1 + u 2 + u 3 = 1 + kk x1 x 2 x3

u u u u dV = 1 + 1 + 2 + 3 + o i x dV0 x1 x 2 x3 j
Therefore:

dV dV0 = kk (bulk strain) dV0

v m
2
1
v v v 1 v 1 m Cm = 2 m (C I )m = 2 1 2 2 1 Note that = (C I ) , 2 v v 1 v m m = 2 1 1 (strain in m direction) 2

Strain gages aligned along different directions at a solid surface can be used to measure strain in the plane of the surface. For example, assuming we can measure the normal strain

EN0175

10 / 17 / 06

v m2

v m3

v m1
1
along the m1 , m2 , m3 directions as shown above, respectively, the components of surface strain tensor is simply

v v v v m1 m1 = 11 , m2 m2 = 22
The shear strain 12 is to be determined from the reading in the m3 direction.

Compatibility conditions of strains 3 displacement components 6 strain components

ij =

Eliminating u from the above 3 equations among strain (compatibility conditions) For example:

1 (ui, j + u j ,i ) 2

11 =

u 2 u1 1 u1 u 2 , 12 = , 22 = + x1 x2 2 x2 x1

2 11 2 22 3u1 3u 2 2 u1 u2 2 12 + = + = + = 2 2 2 x2 x12 x1x2 x2 x12 x1x2 x1x2 x2 x1

Similarly:

2 13 2 11 2 33 + = 2 2 x3 x12 x1x3 2 23 2 22 2 33 + =2 2 2 x3 x2 x2x3

Chap. 5 Mechanical behavior of solids In a general boundary value problem in solid mechanics, the number of unknown variables is

EN0175

10 / 17 / 06

ui (3), ij (6), ij (6) 15 unknown variables

Number of equations: Equilibrium equations:

&&i (3 equations) ij , j + i = u

Strain-displacement equations:

ij =

1 (ui, j + u j ,i ) (6 equations) 2

Therefore, we are still missing 6 equations to determine the 15 unknowns. The six additional equations to close the formulations can be generally expressed as:

= f ( )
This leads to the constitutive model of different material behaviors which can be categorized into the following types: Linear elastic material (Hookes law) Elastic-plastic material Visco-elastic material Visco-plastic material Linear elastic material: We have seen Hookes law in 1D:

= E
In 3D, one might generalize this in tensor form as

= C or ij = Cijkl kl
wWhere Cijkl is a 4th order tensor. For the most general case, a 4th order tensor has 3 = 81
4

independent components. However, and are both symmetric tensors that only have six independent components. Therefore, = C can be rewritten in a matrix form as

11 C11 22 C21 C 33 = 31 23 C41 13 C51 C 12 61

C12 C22 C32 C42 C52 C62

C13 C23 C33 C43 C53 C63

5

C14 C24 C34 C44 C54 C64

C15 C25 C35 C45 C55 C65

C16 11 C26 22 C36 33 C46 23 C56 13 C66 12

EN0175

10 / 17 / 06

Clearly, the number of independent components of C is reduced to 36. Furthermore, the 6 6 matrix of C is symmetric, which further reduces the number of independent components to 21 (the reason will be discussed in the next lecture).