Solid Mechanics

Lecture Notes on
Solid Mechanics
Dr.Ing. JensUwe Bohrnsen
Institute of Applied Mechanics
Spielmann Str. 11
38104 Braunschweig
www.infam.tu-braunschweig.de
October 2010
Contents
1 Introduction and mathematical preliminaries 3
1.1 Vectors and matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Indical Notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Rules for matrices and vectors . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.4 Coordinate transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.5 Tensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.6 Scalar, vector and tensor elds . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.7 Divergence theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.8 Summary of chapter 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.9 Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2 Traction, stress and equilibrium 17
2.1 State of stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1.1 Traction and couplestress vectors . . . . . . . . . . . . . . . . . . . 17
2.1.2 Components of stress . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.1.3 Stress at a point . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.1.4 Stress on a normal plane . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2.1 Physical principles . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2.2 Linear momentum . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.2.3 Angular momentum . . . . . . . . . . . . . . . . . . . . . . . . . . 23
I
2.3 Principal stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.3.1 Maximum normal stress . . . . . . . . . . . . . . . . . . . . . . . . 24
2.3.2 Stress invariants and special stress tensors . . . . . . . . . . . . . . 27
2.5 Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3 Deformation 33
3.1 Position vector and displacement vector . . . . . . . . . . . . . . . . . . . . 33
3.2 Strain tensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.3 Stretch rationite strains . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.4 Linear theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.5 Properties of the strain tensor . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.5.1 Principal strain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.5.2 Volume and shape changes . . . . . . . . . . . . . . . . . . . . . . . 38
3.6 Compatibility equations for linear strain . . . . . . . . . . . . . . . . . . . 40
3.8 Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4 Material behavior 44
4.1 Uniaxial behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.2 Generalized Hookes law . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.2.1 General anisotropic case . . . . . . . . . . . . . . . . . . . . . . . . 45
4.2.2 Planes of symmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.2.3 Isotropic elastic constitutive law . . . . . . . . . . . . . . . . . . . . 48
4.2.4 Thermal strains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.3 Elastostatic/elastodynamic problems . . . . . . . . . . . . . . . . . . . . . 50
4.3.1 Displacement formulation . . . . . . . . . . . . . . . . . . . . . . . 50
4.3.2 Stress formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.5 Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
II
5 Twodimensional elasticity 55
5.1 Plane stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
5.2 Plane strain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
5.3 Airys stress function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
5.5 Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
6 Energy principles 63
6.1 Work theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
6.2 Principles of virtual work . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
6.2.1 Statement of the problem . . . . . . . . . . . . . . . . . . . . . . . 65
6.2.2 Principle of virtual displacements . . . . . . . . . . . . . . . . . . . 66
6.2.3 Principle of virtual forces . . . . . . . . . . . . . . . . . . . . . . . . 68
6.3 Approximative solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
6.3.1 Application: FEM for beam . . . . . . . . . . . . . . . . . . . . . . 72
6.5 Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
A Solutions 80
A.1 Chapter 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
A.2 Chapter 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
A.3 Chapter 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
A.4 Chapter 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
A.5 Chapter 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
A.6 Chapter 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
III
1
These lecture notes are based on Introduction to Linear Elasticity
by P.L. Gould (see bibliography).
Here:
Mostly linear theory with exception of denition of strain.
(Nonlinear theory see Introduction to continuum mechanics.)
Prerequisites:
Statics, strength of materials, mathematics
Additional reading:
see bibliography
Bibliography
[1] W. Becker and D. Gross. Mechanik elastischer Korper und Strukturen. Springer, 2002.
[2] P.L. Gould. Introduction to Linear Elasticity. Springer, 1999.
[3] D. Gross, W. Hauger, and P. Wriggers. Technische Mechanik IV. Springer, 2002.
[4] G.E. Mase. Continuum Mechanics. Schaums outlines, 1970.
2
1 Introduction and mathematical
preliminaries
1.1 Vectors and matrices
A vector is a directed line segment. In a cartesian coordinate system it looks like
depicted in gure 1.1,
z
y
x
P
a
e
z
e
y
e
x
a
x
a
y
a
z
P
a
x
3
x
2
x
1
e
3
e
2
e
1
a
1
a
2
a
3
Figure 1.1: Vector in a cartesian coordinate system
e. g., it can mean the location of a point P or a force. So a vector connects direction
and norm of a quantity. For representation in a coordinate system unit basis vectors
e
x
, e
y
and e
z
are used with [e
x
[ = [e
y
[ = [e
z
[ = 1.
[ [ denotes the norm, i. e., the length.
Now the vector a is
a = a
x
e
x
+a
y
e
y
+a
z
e
z
(1.1)
3
4 CHAPTER 1. INTRODUCTION AND MATHEMATICAL PRELIMINARIES
with the coordinates (a
x
, a
y
, a
z
) = values/length in the direction of the basis vec-
tors/coordinate direction.
More usual in continuum mechanics is denoting the axis with e
1
, e
2
and e
3
a = a
1
e
1
+a
2
e
2
+a
3
e
3
(1.2)
Dierent representations of a vector are
a =
_
_
_
a
1
a
2
a
3
_
_
_
= (a
1
, a
2
, a
3
) (1.3)
with the length/norm (Euclidian norm)
[a[ =
_
a
2
1
+a
2
2
+a
2
3
. (1.4)
A matrix is a collection of several numbers
A =
_
_
_
_
_
_
A
11
A
12
A
13
. . . A
1n
A
21
A
22
A
23
. . . A
2n
.
.
.
.
.
.
.
.
.
A
m1
A
m2
A
m3
. . . A
mn
_
_
_
_
_
_
(1.5)
with n columns and m rows, i.e., a (mn) matrix. In the following mostly quadratic
matrixes n m are used.
A vector is a one column matrix.
Graphical representation as for a vector is not possible. However, a physical inter-
pretation is often given, then tensors are introduced.
Special cases:
Zero vector or matrix: all elements are zero, e.g., a =
_
0
0
0
_
and A =
_
0 0 0
0 0 0
0 0 0
_
Symmetric matrix A = A
T
with A
T
is the transposed matrix, i.e., all elements
at the same place above and below the main diagonal are identical, e.g., A =
_
1 5 4
5 2 6
4 6 3
_
1.2. INDICAL NOTATION 5
1.2 Indical Notation
Indical notation is a convenient notation in mechanics for vectors and matrices/tensors.
Letter indices as subscripts are appended to the generic letter representing the tensor
quantity of interest. Using a coordinate system with (e
1
, e
2
, e
3
) the components of
a vector a are a
i
(eq. 1.7) and of a matrix A are A
ij
with i = 1, 2, . . . , m and j =
1, 2, . . . , n (eq. 1.6). When an index appears twice in a term, that index is understood
to take on all the values of its range, and the resulting terms summed. In this so-called
Einstein summation, repeated indices are often referred to as dummy indices, since their
replacement by any other letter not appearing as a free index does not change the meaning
of the term in which they occur. In ordinary physical space, the range of the indices is
1, 2, 3.
A
ii
=
m
i=1
A
ii
= A
11
+A
22
+A
33
+. . . +A
mm
(1.6)
and
a
i
b
i
= a
1
b
1
+a
2
b
2
+. . . +a
m
b
m
. (1.7)
However, it is not summed up in an addition or subtraction symbol, i.e., if a
i
+b
i
or a
i
b
i
.
A
ij
b
j
=A
i1
b
1
+A
i2
b
2
+. . . +A
ik
b
k
(1.8)

free dummy
Further notation:
i=1
a
i
= a
1
a
2
a
3
(1.9)
a
i
x
j
= a
i,j
with a
i,i
=
a
1
x
1
+
a
2
x
2
+. . . (1.10)
or
A
ij
x
j
=
A
i1
x
1
+
A
i2
x
2
+. . . = A
ij,j
(1.11)
This is sometimes called comma convention!
1.3 Rules for matrices and vectors
Addition and subtraction
AB = C C
ij
= A
ij
B
ij
(1.12)
component by component, vector similar.
Multiplication
Vector with vector
Scalar (inner) product:
c = a b = a
i
b
i
(1.13)
Cross (outer) product:
c = a b =
e
1
e
2
e
3
a
1
a
2
a
3
b
1
b
2
b
3
=
_
_
_
a
2
b
3
a
3
b
2
a
3
b
1
a
1
b
3
a
1
b
2
a
2
b
1
_
_
_
(1.14)
Cross product is not commutative.
Using indical notation
c
i
=
ijk
a
j
b
k
(1.15)
with permutations symbol / alternating tensor
ijk
=
_
_
1 i, j, k even permutation (e.g. 231)
1 i, j, k odd permutation (e.g. 321)
0 i, j, k no permutation, i.e.
two or more indices have the same value
. (1.16)
Dyadic product:
C = a b (1.17)
Matrix with matrix Inner product:
C = AB (1.18)
C
ik
= A
ij
B
jk
(1.19)
Inner product of two matrices can be done with Falk scheme (g. 1.2(a)). To
get one component C
ij
of C, you have to do a scalar product of two vectors a
i
1.3. RULES FOR MATRICES AND VECTORS 7
and b
j
, which are marked in gure 1.2 with a dotted line. It is also valid for
the special case of onecolumn matrix (vector) (g. 1.2(b))
c = Ab c
i
= A
ij
b
j
. (1.20)
B
ij
A
ij
C
ij
(a) Product of matrix with matrix
b
j
A
ij
c
i
(b) Product of matrix with vector
Figure 1.2: Falk scheme
Remarks on special matrices:
Permutation symbol (see 1.16)
ijk
=
1
2
(i j)(j k)(k i) (1.21)
Kronecker delta
ij
=
_
_
_
1 if i = j
0 if i ,= j
(1.22)
so
ij

_
0 0
0 0
0 0
_
for i, j = 1, 2, 3 (1.23)
ij
a
i
= a
j

ij
D
jk
= D
ik
(1.24)
Product of two unit vectors
e
i
e
j
=
ij
(orthogonal basis) (1.25)
Decomposition of a matrix
A
ij
=
1
2
(A
ij
+A
ji
)
. .
symmetric
+
1
2
(A
ij
A
ji
)
. .
anti-symmetric/skrew symmetric
(1.26)
1.4 Coordinate transformation
Assumption:
2 coordinate systems in one origin rotated against each other (g. 1.3).
x
1
x
1
x
2
x
2
x
3 x
3
Figure 1.3: Initial (x
1
, x
2
, x
3
) and rotated (x
1
, x
2
, x
3
) axes of transformed coordinate sys-
tem
The coordinates can be transformed
x
1
=
11
x
1
+
12
x
2
+
13
x
3
=
1j
x
j
(1.27)
x
2
=
2j
x
j
(1.28)
x
3
=
3j
x
j
(1.29)
x
i
=
ij
x
j
(1.30)
with the constant (only constant for cartesian system) coecients
ij
= cos(x
i
, x
j
)
. .
direction cosine
=
x
j
x
i
= cos(e
i
, e
j
) = e
i
e
j
. (1.31)
In matrix notation we have
x
= R
..
rotation matrix
x. (1.32)
R
ij
= x
i,j
(1.33)
So the primed coordinates can be expressed as a function of the unprimed ones
x
i
= x
i
(x
i
) x
= x
(x). (1.34)
1.5. TENSORS 9
If J = [R[ does not vanish this transformation possesses a unique inverse
x
i
= x
i
(x
i
) x = x(x
). (1.35)
J is called the Jacobian of the transformation.
1.5 Tensors
Denition:
A tensor of order n is a set of N
n
quantities which transform from one coordinate system
x
i
to another x
i
by
n order transformation rule
0 scalar a a(x
i
) = a(x
i
)
1 vector x
i
x
i
=
ij
x
j
2 tensor T
ij
T
ij
=
ik
jl
T
kl
with the
ij
as given in chapter 1.4 (
ij
= x
i,j
). So a vector is a tensor of rst order which
can be transformed following the rules above.
Mostly the following statement is o.k.:
A tensor is a matrix with physical meaning. The values of this matrix are depending on
the given coordinate system.
It can be shown that
A
= RAR
T
. (1.36)
Further, a vector is transformed by
x
i
=
ij
x
j
or x
j
=
ij
x
i
(1.37)
so
x
j
=
ij
i
x
(1.38)
which is only valid if
ij
i
=
j
. (1.39)
This is the orthogonality condition of the direction cosines. Therefore, any transformation
which satises this condition is said to be an orthogonal transformation. Tensors satisfying
orthogonal transformation are called cartesian tensors.
Another proof of orthogonality: Basis vectors in an orthogonal system give
ij
= e
i
e
j
(1.40)
= (
ik
e
k
) (
j
e
) (1.41)
=
ik
j
e
k
e
(1.42)
=
ik
k
(1.43)
=
ik
jk
(1.44)
= e
i
e
j
(1.45)
= (
ki
e
k
) (
j
e
) (1.46)
=
ki
k
(1.47)
=
ki
kj
(1.48)
.
1.6 Scalar, vector and tensor elds
A tensor eld assigns a tensor T(x, t) to every pair (x, t) where the position vector x
varies over a particular region of space and t varies over a particular interval of time.
The tensor eld is said to be continuous (or dierentiable) if the components of T(x, t)
are continuous (or dierentiable) functions of x and t. If the tensor T does not depend
on time the tensor eld is said to be steady (T(x)).
1. Scalar eld: = (x
i
, t) = (x, t)
2. Vector eld: v
i
= v
i
(x
i
, t) v = v(x, t)
3. Tensor eld: T
ij
= T
ij
(x
i
, t) T = T(x, t)
Introduction of the dierential operator : It is a vector called del or NablaOperator,
dened by
= e
i
x
i
and
2
=
..
Laplacian operator
= =

x
i

x
i
. (1.49)
A few dierential operators on vectors or scalar:
grad = =
,i
e
i
(result: vector) (1.50)
div v = v = v
i,i
(result: scalar) (1.51)
curl v = v =
ijk
v
k,j
(result: vector) (1.52)
1.7. DIVERGENCE THEOREM 11
Similar rules are available for tensors/vectors.
1.7 Divergence theorem
For a domain V with boundary A the following integral transformation holds for a rst-
order tensor g
_
V
divgdV =
_
V
gdV =
_
A
n gdA (1.53)
_
V
g
i,i
dV =
_
A
g
i
n
i
dA (1.54)
and for a second-order tensor
_
V
ji,j
dV =
_
A
ji
n
j
dA (1.55)
_
V
divdV =
_
V
dV =
_
A
ndA. (1.56)
Here, n = n
i
e
i
denotes the outward normal vector to the boundary A.
1.8 Summary of chapter 1
Vectors
a =
_
_
_
a
1
a
2
a
3
_
_
_
= a
1
e
1
+a
2
e
2
+a
3
e
3
= a
1
_
_
_
1
0
0
_
_
_
+a
2
_
_
_
0
1
0
_
_
_
+a
3
_
_
_
0
0
1
_
_
_
Magnitude of a:
[a[ =
_
a
2
1
+a
2
2
+a
2
3
is the length of a
Vector addition:
_
_
_
a
1
a
2
a
3
_
_
_
+
_
_
_
b
1
b
2
b
3
_
_
_
=
_
_
_
a
1
+b
1
a
2
+b
2
a
3
+b
3
_
_
_
Multiplication with a scalar:
c
_
_
_
a
1
a
2
a
3
_
_
_
=
_
_
_
c a
1
c a
2
c a
3
_
_
_
Scalar (inner, dot) product:
a b = [a[[b[ cos = a
1
b
1
+a
2
b
2
+a
3
b
3
Vector (outer, cross) product:
a b =
e
1
e
2
e
3
a
1
a
2
a
3
b
1
b
2
b
3
= e
1
a
2
a
3
b
2
b
3
e
2
a
1
a
3
b
1
b
3
+e
3
a
1
a
2
b
1
b
2
=
_
_
_
a
2
b
3
a
3
b
2
a
3
b
1
a
1
b
3
a
1
b
2
a
2
b
1
_
_
_
Rules for the vector product:
a b = (b a)
(c a) b = a (c b) = c(a b)
(a +b) c = a c +b c
a (b c) = (a c) b (a b) c
Matrices
A =
_
_
_
_
_
_
A
11
A
12
A
13
... A
1n
A
21
A
22
A
23
... A
2n
.
.
.
.
.
.
.
.
.
.
.
.
A
m1
A
m2
A
m3
... A
mn
_
_
_
_
_
_
= A
ik
Multiplication of a matrix with a scalar:
c A = A c = c A
ik
e.g.: c
_
A
11
A
12
A
21
A
22
_
=
_
c A
11
c A
12
c A
21
c A
22
_
1.8. SUMMARY OF CHAPTER 1 13
Addition of two matrices:
A+B = B+A = (A
ik
) + (B
ik
) = (A
ik
+B
ik
)
e.g.:
_
A
11
A
12
A
21
A
22
_
+
_
B
11
B
12
B
21
B
22
_
=
_
A
11
+B
11
A
12
+B
12
A
21
+B
21
A
22
+B
22
_
Rules for addition of matrices:
(A+B) +C = A+ (B+C) = A+B+C
Multiplication of two matrices:
C
ik
= A
i1
B
1k
+A
i2
B
2k
+... +A
il
B
lk
=
l
j=1
A
ij
B
jk
i = 1, ..., m k = 1, ..., n
e.g.:
_
B
11
B
12
B
21
B
22
_
_
A
11
A
12
A
21
A
22
_ _
A
11
B
11
+A
12
B
12
A
11
B
12
+A
12
B
22
A
21
B
11
+A
22
B
21
A
21
B
12
+A
22
B
22
_
Rules for multiplication of two matrices:
A(BC) = (AB)C = ABC
AB ,= BA
Distributive law:
(A+B) C = A C+B C
Dierential operators for vector analysis
Gradient of a scalar eld f(x, y, z)
gradf(x
1
, x
2
, x
3
) =
_
_
_
f(x
1
,x
2
,x
3
)
x
1
f(x
1
,x
2
,x
3
)
x
2
f(x
1
,x
2
,x
3
)
x
3
_
_
_
Derivative into a certain direction:
f
a
(x
1
, x
2
, x
3
) =
a
[a[
gradf(x
1
, x
2
, x
3
)
Divergence of a vector eld
div v(X(x
1
, x
2
, x
3
), Y (x
1
, x
2
, x
3
), Z(x
1
, x
2
, x
3
)) =
X
x
1
+
Y
x
2
+
Z
x
3
Curl of a vector eld
curl v(X(x
1
, x
2
, x
3
), Y (x
1
, x
2
, x
3
), Z(x
1
, x
2
, x
3
)) =
_
_
_
Z
x
2
Y
x
3
X
x
3
Z
x
1
Y
x
1
X
x
2
_
_
_
Nabla (del) Operator
=
_
_
_
x
1
x
2
x
3
_
_
_
f(x
1
, x
2
, x
3
) =
_
_
_
f
x
1
f
x
2
f
x
3
_
_
_
= gradf(x
1
, x
2
, x
3
)
v(X(x
1
, x
2
, x
3
), Y (x
1
, x
2
, x
3
), Z(x
1
, x
2
, x
3
)) =
X
x
1
+
Y
x
2
+
Z
x
3
= div v
v =
e
1
e
2
e
3
x
1
x
2
x
3
X Y Z
= curl v
Laplacian operator
u = = div gradu =

2
u
x
2
1
+

2
u
x
2
2
+

2
u
x
2
3
Indical Notation Summation convention
A subscript appearing twice is summed from 1 to 3.
e.g.:
a
i
b
i
=
3
i=1
a
i
b
i
= a
1
b
1
+a
2
b
2
+a
3
b
3
D
jj
= D
11
+D
22
+D
33
1.9. EXERCISE 15
Comma-subscript convention
The partial derivative with respect to the variable x
i
is represented by the so-called
comma-subscript convention e.g.:
x
i
= ,
i
= grad
v
i
x
i
= v
i,i
= divv
v
i
x
j
= v
i,j
2
v
i
x
j
x
k
= v
i,jk
1.9 Exercise
1. given: scalar eld
f(x
1
, x
2
, x
3
) = 3x
1
+x
1
e
x
2
+x
1
x
2
e
x
3
(a)
gradf(x
1
, x
2
, x
3
) =?
(b)
gradf(3, 1, 0) =?
2. given: scalar eld
f(x
1
, x
2
, x
3
) = x
2
1
+
3
2
x
2
2
Find the derivative of f in point/position vector
_
5
2
8
_
in the direction of a
_
3
0
4
_
.
3. given: vector eld
V =
_
_
_
x
1
+x
2
2
e
x
1
x
3
+ sin x
2
x
1
x
2
x
3
_
_
_
(a)
divV(X(x
1
, x
2
, x
3
), Y (x
1
, x
2
, x
3
), Z(x
1
, x
2
, x
3
)) =?
(b)
divV(1, , 2) =?
4. given: vector eld
V =
_
_
_
x
1
+x
2
e
x
1
+x
2
+x
3
x
3
+ sin x
1
_
_
_
(a)
curlV(x
1
, x
2
, x
3
) =?
(b)
curlV(0, 8, 1) =?
5. Expand and, if possible, simplify the expression D
ij
x
i
x
j
for
(a) D
ij
= D
ji
(b) D
ij
= D
ji
.
6. Determine the component f
2
for the given vector expressions
(a) f
i
= c
i,j
b
j
c
j,i
b
j
(b) f
i
= B
ij
f
j
7. If r
2
= x
i
x
i
and f(r) is an arbitrary function of r, show that
(a) (f(r)) =
f
(r)x
r
(b)
2
(f(r)) = f
(r) +
2f
(r)
r
,
where primes denote derivatives with respect to r.
2 Traction, stress and equilibrium
2.1 State of stress
Derivation of stress at any distinct point of a body.
2.1.1 Traction and couplestress vectors
M
n
n
F
n
A
n
Figure 2.1: Deformable body under loading
Assumption: Deformable body
Possible loads:
surface forces: loads from exterior
body forces: loads distributed within the interior, e.g., gravity force
17
18 CHAPTER 2. TRACTION, STRESS AND EQUILIBRIUM
At any element A
n
in or on the body (n indicates the orientation of this area) a resultant
force F
n
and/or moment M
n
produces stress.
lim
A
n
0
F
n
A
n
=
dF
n
dA
n
= t
n
stress vector/traction (2.1)
lim
A
n
0
M
n
A
n
=
dM
n
dA
n
= C
n
couple stress vector (2.2)
The limit A
n
0 expresses that every particle has its own tractions or, more precise,
the traction vector varies with position x. In usual continuum mechanics we assume
C
n
0 at any point x. As a consequence of this assumption every particle can have
only translatory degrees of freedom. The traction vector represents the stress intensity at
a distinct point x for the particular orientation n of the area element A. A complete
description at the point requires that the state of stress has to be known for all directions.
So t
n
itself is necessary but not sucient.
Remark:
Continua where the couple stress vector is not set equal to zero can be dened. They are
called CosseratContinua. In this case each particle has additionally to the translatory
degrees of freedom also rotary ones.
2.1.2 Components of stress
Assumption:
Cartesian coordinate system with unit vectors e
i
innitesimal rectangular parallelepiped;
t
i
are not parallel to e
i
whereas the surfaces are perpendicular to the e
i
, respectively (g.
2.2). So, all e
i
represents here the normal n
i
of the surfaces.
Each traction is separated in components in each coordinate direction
t
i
=
i1
e
1
+
i2
e
2
+
i3
e
3
(2.3)
t
i
=
ij
e
j
. (2.4)
With these coecients
ij
a stress tensor can be dened
=
_
_
_
11

12

13
21

22

23
31

32

33
_
_
_
=
ij
, (2.5)
with the following signconvention:
2.1. STATE OF STRESS 19
x
1
x
2
x
3
e
1
e
2
e
3
t
1
t
2
t
3
11
12
13
Figure 2.2: Tractions t
i
and their components
ij
on the rectangular parallelepiped sur-
faces of an innitesimal body
1. The rst subscript i refers to the normal e
i
which denotes the face on which t
i
acts.
2. The second subscript j corresponds to the direction e
j
in which the stress acts.
3.
ii
(no summation) are positive (negative) if they produce tension (compression).
They are called normal components or normal stress
ij
(i ,= j) are positive if coordinate direction x
j
and normal e
i
are both positive
or negative. If both dier in sign,
ij
(i ,= j) is negative. They are called shear
components or shear stress.
2.1.3 Stress at a point
Purpose is to show that the stress tensor describes the stress at a point completely.
In g. 2.3, f is a body force per unit volume and
dA
i
= dA
n
cos(n, e
i
) = dA
n
n e
i
(2.6)
dA
n
=
dA
i
n e
i
=
dA
i
n
i
(2.7)
with n e
i
= n
j
e
j
e
i
= n
j
ij
!
= n
i
. (2.8)
Equilibrium of forces at tetrahedron (g. 2.3):
t
n
dA
n
t
i
dA
i
+f
_
1
3
hdA
n
_
. .
volume of the tetrahedron
= 0 (2.9)
t
1
t
2
t
3
t
n
n
x
1
x
2
x
3
dA
1
dA
2
dA
3
dA
n
f
Figure 2.3: Tractions of a tetrahedron
_
t
n
n
i
t
i
+f
h
3
_
dA
n
= 0 (2.10)
Now, taking the limit dA
n
0, i.e., h 0 reduces the tetrahedron to a point which gives
t
n
= t
i
n
i
=
ji
e
i
n
j
. (2.11)
Resolving t
n
into cartesian components t
n
= t
i
e
i
yields the Cauchy theorem
t
i
e
i
=
ji
e
i
n
j
t
i
=
ji
n
j
(2.12)
with the magnitude of the stress vector
[t
n
[ =
_
(t
i
t
i
). (2.13)
Therefore, the knowledge of t
i
=
ji
n
j
is sucient to specify the state of stress at a point
in a particular cartesian coordinate system. As is a tensor of 2. order the stress tensor
can be transformed to every rotated system by
ji
=
ik
jl
kl
(2.14)
with the direction cosines
ij
= cos(x
i
, x
j
).
2.2. EQUILIBRIUM 21
dA
n
s
t
n
n
Figure 2.4: Normal and tangential component of t
n
2.1.4 Stress on a normal plane
Interest is in the normal and tangential component of t
n
(g. 2.4).
Normalvector: n = n
i
e
i
Tangentialvector: s = s
i
e
i
(two possibilities in 3-D)
Normal component of stress tensor with respect to plane dA
n
:
nn
= t
n
n =
ij
n
i
e
j
n
k
e
k
=
ij
n
i
n
k
jk
=
ij
n
j
n
i
(2.15)
Tangential component:
ns
= t
n
s =
ij
n
i
e
j
s
k
e
k
=
ij
n
i
s
j
(2.16)
2.2 Equilibrium
2.2.1 Physical principles
Consider an arbitrary body V with boundary A (surface) (g. 2.5).
x
1
x
2
x
3
f
r
t
P
Figure 2.5: Body V under loading f with traction t acting normal to the boundary of the
body
In a 3-d body the following 2 axioms are given:
1. The principle of linear momentum is
_
V
f dV +
_
A
t dA =
_
V
d
2
dt
2
udV (2.17)
with displacement vector u and density .
2. The principle of angular momentum (moment of momentum)
_
V
(r f) dV +
_
A
(r t) dA =
_
V
(r u) dV (2.18)
Considering the position vector r to point P(x)
r = x
j
e
j
(2.19)
and further
r f =
ijk
x
j
f
k
e
i
(2.20)
r t =
ijk
x
j
t
k
e
i
(2.21)
The two principles, (2.17) and (2.18), are in indical notation
_
V
f
i
dV +
_
A
ji
n
j
dA =
_
V
u
i
dV
_
note, that

( ) =
d
2
dt
2
()
_
(2.22)
_
V
ijk
x
j
f
k
dV +
_
A
ijk
x
j
lk
n
l
dA =
_
V
ijk
x
j
u
k
dV, (2.23)
where the Cauchy theorem (2.12) has been used. In the static case, the inertia terms on
the right hand side, vanish.
2.2.2 Linear momentum
Linear momentum is also called balance of momentum or force equilibrium. With the
assumption of a C
1
continuous stress tensor we have
_
V
(f + )dV =
_
V
udV (2.24)
2.2. EQUILIBRIUM 23
or
_
V
(f
i
+
ji,j
)dV =
_
V
u
i
dV (2.25)
using the divergence theorem (1.56). The above equation must be valid for every element
in V , i.e., the dynamic equilibrium is fullled. Consequently, because V is arbitrary the
integrand vanishes. Therefore,
+f = u (2.26)
ji,j
+f
i
= u
i
(2.27)
has to be fullled for every point in the domain V . These equations are the linear
momentum.
2.2.3 Angular momentum
Angular momentum is also called balance of moment of momentum or momentum equi-
librium. We start in indical notation by applying the divergence theorem (1.55) to
_
V
[
ijk
x
j
f
k
+ (
ijk
x
j
lk
)
,l
]dV =
_
V
ijk
x
j
u
k
dV . (2.28)
With the product rule
(
ijk
x
j
lk
)
,l
=
ijk
[x
j,l
lk
+x
j
lk,l
] (2.29)
and the property x
j,l
=
jl
, i.e., the position coordinate derivated by the position coordi-
nate vanishes if it is not the same direction, yields
_
V
ijk
[x
j
f
k
+
jl
lk
+x
j
lk,l
]dV =
_
V
ijk
x
j
u
k
dV . (2.30)
Applying the linear momentum (2.25)
ijk
x
j
(f
k
+
lk,l
u
k
) = 0 (2.31)
the above equation is reduced to
_
V
ijk
jl
lk
dV = 0 (2.32)
which is satised for any region dV if
ijk
jk
= 0 (2.33)
holds.
Now, if the last equation is evaluated for i = 1, 2, 3 and using the properties of the
permutation symbol, it is found the condition
ij
=
ji
=
T
(2.34)
fullls (2.33).
This statement is the symmetry of the stress tensor. This implies that has only six
independent components instead of nine components. With this important property of
the stress tensor the linear momentum in indical notation can be rewritten
ij,j
+f
i
= u
i
, (2.35)
and also Cauchys theorem
t
i
=
ij
n
j
. (2.36)
This is essentially a boundary condition for forces/tractions. The linear momentum are
three equations for six unknowns, and, therefore, indeterminate. In chapter 3 and 4 the
missing equations will be given.
2.3 Principal stress
2.3.1 Maximum normal stress
Question: Is there a plane in any body at any particular point where no shear stress
exists?
Answer: Yes
For such a plane the stress tensor must have the form
=
_
_
_
(1)
0 0
0
(2)
0
0 0
(3)
_
_
_
(2.37)
2.3. PRINCIPAL STRESS 25
with three independent directions n
(k)
where the three principal stress components act.
Each plane given by these principal axes n
(k)
is called principal plane. So, it can be
dened a stress vector acting on each of these planes
t =
(k)
n (2.38)
where the tangential stress vector vanishes. To nd these principal stresses and planes
(k = 1, 2, 3)
ij
n
j
(k)
n
i
!
= 0 (2.39)
must be fullled. Using the Kronecker delta yields
(
ij
(k)
ij
)n
j
!
= 0 (2.40)
This equation is a set of three homogeneous algebraic equations in four unknowns (n
i
with i = 1, 2, 3 and
(k)
). This eigenvalue problem can be solved if
[
ij
(k)
ij
[ = 0 (2.41)
holds, which results in the eigenvalues
(k)
, the principal stresses. The corresponding
orientation of the principal plane n
(k)
is found by inserting
(k)
back in equation (2.40)
and solving the equation system. As this system is linearly dependent (cf. equation
(2.40)) an additional relationship is necessary. The length of the normal vectors
n
(k)
i
n
(k)
i
= 1 (2.42)
is to unify and used as additional equation. The above procedure for determining the
principal stress and, subsequently, the corresponding principal plane is performed for
each eigenvalue
(k)
(k = 1, 2, 3).
The three principal stresses are usually ordered as
(1)

(2)

(3)
. (2.43)
Further, the calculated n
(k)
are orthogonal. This fact can be concluded from the following.
Considering the traction vector for k = 1 and k = 2
ij
n
(1)
j
=
(1)
n
(1)
i

ij
n
(2)
j
=
(2)
n
(2)
i
(2.44)
and multiplying with n
(2)
i
and n
(1)
i
, respectively, yields
ij
n
(1)
j
n
(2)
i
=
(1)
n
(1)
i
n
(2)
i
ij
n
(2)
j
n
(1)
i
=
(2)
n
(2)
i
n
(1)
i
.
Using the symmetry of the stress tensor and exchanging the dummy indices i and j, the
left hand side of both equations is obviously equal. So, dividing both equations results in
0 = (
(1)
(2)
)n
(1)
i
n
(2)
i
. (2.45)
Now, if
(1)
,=
(2)
the orthogonality of n
(1)
and n
(2)
follows. The same is valid for other
combinations of n
(k)
.
To show that the principal stress exists at every point, the eigenvalues
(k)
(the principal
stresses) are examined. To represent a physically correct solution
(k)
must be real-valued.
Equation (2.41) is a polynomial of third order, therefore, three zeros exist which are not
necessarily dierent. Furthermore, in maximum two of them can be complex because
zeros exist only in pairs (conjugate complex). Let us assume that the real one is
(1)
and
the n
(1)
direction is equal to the x
1
direction. This yields the representation
=
_
_
_
(1)
0 0
0
22

23
0
23

33
_
_
_
(2.46)
of the stress tensor and, subsequently, equation (2.41) is given as
(
(1)
(k)
)(
22
(k)
)(
33
(k)
)
2
23
= 0. (2.47)
The two solutions of the curly bracket are
(
(k)
)
2
(
22
+
33
)
(k)
+ (
22
33
2
23
) = 0 (2.48)
(2,3)
=
1
2
_
(
22
+
33
)
_
(
22
+
33
)
2
4(
22
33
2
23
)
_
. (2.49)
For a real-valued result the square-root must be real yielding
(
22
+
33
)
2
4(
22
33
2
23
) = (
22
33
)
2
+ 4
2
23
!
0. (2.50)
With equation (2.50) it is shown that for any arbitrary stress tensor three real eigenvalues
exist and, therefore, three principal values.
2.3. PRINCIPAL STRESS 27
2.3.2 Stress invariants and special stress tensors
In general, the stress tensor at a distinct point dier for dierent coordinate systems.
However, there are three values, combinations of
ij
, which are the same in every co-
ordinate system. These are called stress invariants. They can be found in performing
equation (2.41)
[
ij
(k)
ij
[ = (
(k)
)
3
I
1
(
(k)
)
2
+I
2
(k)
I
3
!
= 0 (2.51)
with
I
1
=
ii
= tr (2.52)
I
2
=
1
2
(
ii
jj
ij
ij
) (2.53)
I
3
= [
ij
[ = det (2.54)
and represented in principal stresses
I
1
=
(1)
+
(2)
+
(3)
(2.55)
I
2
= (
(1)
(2)
+
(2)
(3)
+
(3)
(1)
) (2.56)
I
3
=
(1)
(2)
(3)
, (2.57)
the rst, second, and third stress invariant is given. The invariance is obvious because
all indices are dummy indices and, therefore, the values are scalars independent of the
coordinate system.
The special case of a stress tensor, e.g., pressure in a uid,
=
0
_
_
_
1 0 0
0 1 0
0 0 1
_
_
_

ij
=
0
ij
(2.58)
is called hydrostatic stress state. If one assumes
0
=

ii
3
=
m
of a general stress state,
where
m
is the mean normal stress state, the deviatoric stress state can be dened
S =
m
I =
_
_
_
11
m

12

13
12

22
m

23
13

23

33
m
_
_
_
. (2.59)
In indical notation (I =
ij
: itendity-matrix (3x3)):
s
ij
=
ij
ij
kk
3
(2.60)
where
kk
/3 are the components of the hydrostatic stress tensor and s
ij
the components
of the deviatoric stress tensor.
The principal directions of the deviatoric stress tensor S are the same as those of the
stress tensor because the hydrostatic stress tensor has no principal direction, i.e., any
direction is a principal plane. The rst two invariants of the deviatoric stress tensor are
J
1
= s
ii
= (
11
m
) + (
22
m
) + (
33
m
) = 0 (2.61)
J
2
=
1
2
s
ij
s
ij
=
1
6
[(
(1)
(2)
)
2
+ (
(2)
(3)
)
2
+ (
(3)
(1)
)
2
], (2.62)
where the latter is often used in plasticity.
Remark: The elements on the main diagonal of the deviatoric stress tensor are mostly
not zero, contrary to the trace of s.
Stress
Tractions
t
i
=
ij
e
j
t
1
t
2
t
3
11
12
13
21
22
23
31
32
33
Stress Tensor
=
_
_
_
11

12

13
21

22

23
31

32

33
_
_
_
11
,
22
,
33
: normal components
12
,
13
,
23
: shear components
21
,
31
,
32
Stress at a point
t
i
=
ji
n
j
Transformation in another cartesian coordinate system
ij
=
ik
jl
kl
=
ik
kl
lj
with direction cosine:
ij
= cos (x
i
, x
j
)
Stress in a normal plane
Normal component of stress tensor:
nn
=
ij
n
j
n
i
Tangential component of stress tensor:
ns
=
ij
n
i
s
j
=
_
t
i
t
i
2
nn
Equilibrium
ij
=
ji
=
T

ij
,
j
+f
i
= 0 (static case)
with boundary condition: t
i
=
ij
n
j
Principal Stress
In the principal plane given by the principal axes n
(k)
no shear stress exists.
Stress tensor referring to principal stress directions:
=
_
_
_
(1)
0 0
0
(2)
0
0 0
(3)
_
_
_
with
(1)

(2)

(3)
Determination of principal stresses
(k)
with:
[
ij
(k)
ij
[
!
= 0
11
(k)
12

13
21

22
(k)
23
31

32

33
(k)
!
= 0
with the Kronecker delta :
ij
=
_
1 if i = j
0 if i ,= j
Principal stress directions n
(k)
:
(
ij
(k)
ij
) n
(k)
j
= 0
(
11
(k
) n
(k)
1
+
12
n
(k)
2
+
13
n
(k)
3
= 0
21
n
(k)
1
+(
22
(k)
) n
(k)
2
+
23
n
(k)
3
= 0
31
n
(k)
1
+
32
n
(k)
2
+ (
33
(k)
) n
(k)
3
= 0
Stress invariants
The rst, second, and third stress invariant is independent of the coordinate system:
I
1
=
ii
= tr =
11
+
22
+
33
I
2
=
1
2
(
ii
jj
ij
ij
)
=
11
22
+
22
33
+
33
11
12
12
23
23
31
31
I
3
= [
ij
[ = det
Hydrostatic and deviatoric stress tensors
A stress tensor
ij
can be split into two component tensors, the hydrostatic stress tensor
M
=
M
_
_
_
1 0 0
0 1 0
0 0 1
_
_
_

M
ij
=
M
ij
with
M
=

kk
3
and the deviatoric tensor
S =
M
I =
_
_
_
11
M

12

13
21

22
M

23
31

32

33
M
_
_
_

ij
=
ij
kk
3
+S
ij
.
2.5. EXERCISE 31
2.5 Exercise
1. The state of stress at a point P in a structure is given by
11
= 20000
22
= 15000
33
= 3000
12
= 2000
23
= 2000
31
= 1000 .
(a) Compute the scalar components t
1
, t
2
and t
3
of the traction t on the plane
passing through P whose outward normal vector n makes equal angles with
the coordinate axes x
1
, x
2
and x
3
.
(b) Compute the normal and tangential components of stress on this plane.
2. Determine the body forces for which the following stress eld describes a state of
equilibrium in the static case:
11
= 2x
2
1
3x
2
2
5x
3
22
= 2x
2
2
+ 7
33
= 4x
1
+x
2
+ 3x
3
5
12
= x
3
+ 4x
1
x
2
6
13
= 3x
1
+ 2x
2
+ 1
23
= 0
3. The state of stress at a point is given with respect to the Cartesian axes x
1
, x
2
and
x
3
by
ij
=
_
_
_
2 2 0
2
2 0
0 0
2
_
_
_
.
Determine the stress tensor
ij
for the rotated axes x
1
, x
2
and x
3
related to the
unprimed axes by the transformation tensor
ik
=
_
_
_
0
1
2
1
2
1
2
1
2

1
2
2
1
2

1
2
_
_
_
.
4. In a continuum, the stress eld is given by the tensor
ij
=
_
_
_
x
2
1
x
2
(1 x
2
2
)x
1
0
(1 x
2
2
)x
1
(x
3
2
3x
2
)
3
0
0 0 2x
2
3
_
_
_
.
Determine the principal stress values at the point P(a, 0, 2
a) and the correspond-

ing principal directions.
5. Evaluate the invariants of the stress tensors
ij
and
ij
, given in example 3 of chapter
2.
6. Decompose the stress tensor
ij
=
_
_
_
3 10 0
10 0 30
0 30 27
_
_
_
into its hydrostatic and deviator parts and determine the principal deviator stresses!
7. Determine the principal stress values for
(a)
ij
=
_
_
_
0 1 1
1 0 1
1 1 0
_
_
_
and
(b)
ij
=
_
_
_
2 1 1
1 2 1
1 1 2
_
_
_
and show that both have the same principal directions.
3 Deformation
3.1 Position vector and displacement vector
Consider the undeformed and the deformed conguration of an elastic body at time t = 0
and t = t, respectively (g. 3.1).
x
1
X
1
x
2
X
2
x
3
X
3
x
P(x)
u
p(X)
X
undeformed
deformed
t = 0 t = t
Figure 3.1: Deformation of an elastic body
It is convenient to designate two sets of Cartesian coordinates x and X, called material
(initial) coordinates and spatial (nal) coordinates, respectively, that denote the unde-
formed and deformed position of the body. Now, the location of a point can be given in
material coordinates (Lagrangian description)
P = P(x, t) (3.1)
and in spatial coordinates (Eulerian description)
p = p(X, t). (3.2)
Mostly, in solid mechanics the material coordinates and in uid mechanics the spatial
coordinates are used. In general, every point is given in both
X = X(x, t) (3.3)
33
34 CHAPTER 3. DEFORMATION
or
x = x(X, t) (3.4)
where the mapping from one system to the other is given if the Jacobian
J =
X
i
x
j
= [X
i,j
[ (3.5)
exists.
So, a distance dierential is
dX
i
=
X
i
x
j
dx
j
(3.6)
where ( )
denotes a xed but free distance. From gure 3.1 it is obvious to dene the
displacement vector by
u = Xx u
i
= X
i
x
i
. (3.7)
Remark: The Lagrangian or material formulation describes the movement of a particle,
where the Eulerian or spatial formulation describes the particle moving at a location.
3.2 Strain tensor
Consider two neighboring points p(X) and q(X) or P(x) and Q(x) (g. 3.2) in both
congurations (undeformed/deformed)
x
1
X
1
x
2
X
2
x
3
X
3
ds
Q(x +dx)
P(x)
u +du
q(X+dX)
p(X)
u
dS
Figure 3.2: Deformation of two neighboring points of a body
which are separated by dierential distances ds and dS, respectively. The squared length
of them is given by
[ds[
2
= dx
i
dx
i
(3.8)
[dS[
2
= dX
i
dX
i
. (3.9)
3.2. STRAIN TENSOR 35
With the Jacobian of the mapping from one coordinate representation to the other these
distances can be expressed by
[ds[
2
= dx
i
dx
i
=
x
i
X
j
x
i
X
k
dX
j
dX
k
(3.10)
[dS[
2
= dX
i
dX
i
=
X
i
x
j
X
i
x
k
dx
j
dx
k
. (3.11)
To dene the strain we want to express the relative change of the distance between the
point P and Q in the undeformed and deformed body. From gure 3.2 it is obvious that
ds +u +du dS u = 0
du = dS ds.
(3.12)
Taking the squared distances in material coordinates yield to
[dS[
2
[ds[
2
= X
i,j
X
i,k
dx
j
dx
k
dx
i
dx
i
= (X
i,j
X
i,k
jk
)
. .
=2
L
jk
dx
j
dx
k (3.13)
with the Green or Lagrangian strain tensor
L
jk
, or in spatial coordinates
[dS[
2
[ds[
2
= dX
i
dX
i
x
i
X
j
x
i
X
k
dX
j
dX
k
= (
jk
x
i
X
j
x
i
X
k
)
. .
=2
E
jk
dX
j
dX
k
(3.14)
with the Euler or Almansi strain tensor
E
jk
.
Beware that in general (especially for large displacements)
x
i
X
j
,= x
i,j
Taking into account that
u
i
x
k
=
X
i
x
k
x
i
x
k
= X
i,k
ik
X
i,k
= u
i,k
+
ik
(3.15)
or
u
i
X
k
=
X
i
X
k
x
i
X
k
=
ik
x
i
X
k
x
i
X
k
=
ik
u
i
X
k
(3.16)
the Green strain tensor is
L
jk
=
1
2
[(u
i,j
+
ij
)(u
i,k
+
ik
)
jk
]
=
1
2
[u
i,j
u
i,k
+u
i,j
ik
+
ij
u
i,k
+
jk
jk
]
=
1
2
[u
k,j
+u
j,k
+u
i,j
u
i,k
]
with u
i,j
=
u
i
x
j
(3.17)
and the Almansi tensor is
E
jk
=
1
2
_
jk
(
ij
u
i
X
j
)(
ik
u
i
X
k
)
_
=
1
2
_
u
k
X
j
+
u
j
X
k
u
i
X
j
u
i
X
k
_ (3.18)
3.3 Stretch rationite strains
The relative change of deformation, the unit extension e, corresponds to the strain in a
particular direction. The denition in the undeformed conguration is
[dS[ [ds[
[ds[
=: (e) (3.19)
with the direction e =
ds
|ds|
=
dx
|dx|
. The strain tensor was dened by the absolute distance
[dS[
2
[ds[
2
. Relating them to either the undeformed or deformed conguration yields
[dS[
2
[ds[
2
[ds[
2
= 2
dx
j
[dx
i
[
L
jk
dx
k
[dx
i
[
= 2e
T
E
L
e (3.20)
or
[dS[
2
[ds[
2
[dS[
2
= 2
dX
j
[dX
i
[
E
jk
dX
k
[dX
i
[
= 2e
T
E
E
e . (3.21)
Now with the trick
[dS[
2
[ds[
2
[ds[
2
=
_
[dS[ [ds[
[ds[
_ _
[dS[ +[ds[
[ds[
_
. .
|dS||ds|
|ds|
+2
|ds|
|ds|
=
|dS|+|ds|
|ds|
= ( + 2) (3.22)
the unit extension is given as root of
2
+ 2 2e
T
E
L
e = 0 (3.23)
3.4. LINEAR THEORY 37
i.e.,
(e) = 1
+
()
_
1 + 2e
T
E
L
e (3.24)
where the minus sign is physically nonsense as there are no negative extensions. An
analogous calculation for the deformed conguration gives
(e) = 1
_
1 2e
T
E
E
e . (3.25)
3.4 Linear theory
If small displacement gradients are assumed, i.e.
u
i,j
u
k,l
u
i,j
(3.26)
the non-linear parts can be omitted:
L
ij
=
1
2
(u
i,j
+u
j,i
) (3.27)
E
ij
=
1
2
(
u
i
X
j
+
u
j
X
i
) (3.28)
Furthermore,
u
i,j
<< 1 u
i,j

u
i
X
j
(3.29)
and the strain tensors of both congurations are equal.
ij
=
L
ij
=
E
ij
=
1
2
(u
i,j
+u
j,i
) (3.30)
ij
is called linear or innitesimal strain tensor. This is equivalent to the assumption of
small unit extensions
2
, yielding
2(e) = 2e
T
E
L
e = 2e
T
E
E
e . (3.31)
With both assumptions the linear theory is established and no distinction between the
congurations respective coordinate system is necessary. The components on the main
diagonal are called normal strain and all other are the shear strains. The shear strains
here
ij
=
1
2
(u
i,j
+u
j,i
) =
1
2
ij
(3.32)
are equal to onehalf of the familiar engineering shear strains
ij
. However, only with
the denitions above the strain tensor
=
_
_
_
11

12

13
12

22

23
13

23

33
_
_
_
(3.33)
has tensor properties. By the denition of the strains the symmetry of the strain tensor
is obvious.
3.5 Properties of the strain tensor
3.5.1 Principal strain
Besides the general tensor properties (transformation rules) the strain tensor has as the
stress tensor principal axes. The principal strains
(k)
are determined from the character-
istic equation
[
ij
(k)
ij
[ = 0 k = 1, 2, 3 (3.34)
analogous to the stress. The three eigenvalues
(k)
are the principal strains. The corre-
sponding eigenvectors designate the direction associated with each of the principal strains
given by
(
ij
(k)
ij
)n
(k)
i
= 0 (3.35)
These directions n
(k)
for each principal strain
(k)
are mutually perpendicular and, for
isotropic elastic materials (see chapter 4), coincide with the direction of the principal
stresses.
3.5.2 Volume and shape changes
It is sometimes convenient to separate the components of strain into those that cause
changes in the volume and those that cause changes in the shape of a dierential element.
Consider a volume element V (a b c) oriented with the principal directions (g. 3.3),
then the principal strains are
(1)
=
a
a

(2)
=
b
b

(3)
=
c
c
(3.36)
3.5. PROPERTIES OF THE STRAIN TENSOR 39
a
b
c
3
2
1
Figure 3.3: Volume V oriented with the principal directions
under the assumption of volume change in all three directions.
The volume change can be calculated by
V + V = (a + a)(b + b)(c + c)
= abc
_
1 +
a
a
+
b
b
+
c
c
_
+O(
2
)
= V + (
(1)
+
(2)
+
(3)
)V +O(
2
).
(3.37)
With the assumptions of small changes , nally,
V
V
=
(1)
+
(2)
+
(3)
=
ii
(3.38)
and is called dilatation. Obviously, from the calculation this is a simple volume change
without any shear. It is valid for any coordinate system. The dilatation is also the rst
invariant of the strain tensor, and also equal to the divergence of the displacement vector:
u = u
i,i
=
ii
(3.39)
Analogous to the stress tensor, the strain tensor can be divided in a hydrostatic part
M
=
_
M
0 0
0
M
0
0 0
M
_
_

M
=

ii
3
(3.40)
and a deviatoric part
D
=
_
11
M

12

13
12

22
M

23
13

23

33
M
_
_
. (3.41)
The mean normal strain
M
corresponds to a state of equal elongation in all directions
for an element at a given point. The element would remain similar to the original shape
but changes volume. The deviatoric strain characterizes a change in shape of an element
with no change in volume. This can be seen by calculating the dilatation of
D
:
tr
D
= (
11
M
) + (
22
M
) + (
33
M
) = 0 (3.42)
3.6 Compatibility equations for linear strain
If the strain components
ij
are given explicitly as functions of the coordinates, the six
independent equations (symmetry of )
ij
=
1
2
(u
i,j
+u
j,i
)
are six equations to determine the three displacement components u
i
. The system is
overdetermined and will not, in general, possess a solution for an arbitrary choice of the
strain components
ij
. Therefore, if the displacement components u
i
are singlevalued and
continuous, some conditions must be imposed upon the strain components. The necessary
and sucient conditions for such a displacement eld are expressed by the equations (for
derivation see [2] )
ij,km
+
km,ij
ik,jm
jm,ik
= 0. (3.43)
These are 81 equations in all but only six are distinct
1.

2
11
x
2
2
+

2
22
x
2
1
= 2

2
12
x
1
x
2
2.

2
22
x
2
3
+

2
33
x
2
2
= 2

2
23
x
2
x
3
3.

2
33
x
2
1
+

2
11
x
2
3
= 2

2
31
x
3
x
1
4.

x
1
_
23
x
1
+

31
x
2
+

12
x
3
_
=

2
11
x
2
x
3
5.

x
2
_
23
x
1

31
x
2
+

12
x
3
_
=

2
22
x
3
x
1
6.

x
3
_
23
x
1
+

31
x
2

12
x
3
_
=

2
33
x
1
x
2
_
_
or
x
E = 0. (3.44)
The six equations written in symbolic form appear as
E = 0 (3.45)
Even though we have the compatibility equations, the formulation is still incomplete in
that there is no connection between the equilibrium equations (three equations in six
unknowns
ij
), and the kinematic equations (six equations in nine unknowns
ij
and u
i
).
We will seek the connection between equilibrium and kinematic equations in the laws of
physics governing material behavior, considered in the next chapter.
Remark on 2D:
For plane strain parallel to the x
1
x
2
plane, the six equations reduce to the single
equation
11,22
+
22,11
= 2
12,12
(3.46)
or symbolic
E = 0. (3.47)
For plane stress parallel to the x
1
x
2
plane, the same condition as in case of plain strain
is used, however, this is only an approximative assumption.
Deformations
Linear (innitesimal) strain tensor :
L
ij
=
E
ij
=
ij
=
1
2
(u
i,j
+u
j,i
)
=
_
_
_
u
1,1
1
2
(u
1,2
+u
2,1
)
1
2
(u
1,3
+u
3,1
)
1
2
(u
1,2
+u
2,1
) u
2,2
1
2
(u
2,3
+u
3,2
)
1
2
(u
1,3
+u
3,1
)
1
2
(u
2,3
+u
3,2
) u
3,3
_
_
_
=
_
_
_
11
1
2

12
1
2

13
1
2

21

22
1
2

23
1
2

31
1
2

32

33
_
_
_
Principal strain values
(k)
:
[
ij
(k)
ij
[
!
= 0
Principal strain directions n
(k)
:
(
ij
(k)
ij
) n
(k)
j
= 0
Hydrostatic and deviatoric strain tensors
A stress tensor
ij
can be split into two component tensors, the hydrostatic stain tensor
M
=
M
_
_
_
1 0 0
0 1 0
0 0 1
_
_
_

M
ij
=
M
ij
with
M
=

kk
3
and the deviatoric strain tensor
(D)
=
M
I =
_
_
_
11
M

12

13
21

22
M

23
31

32

33
M
_
_
_
Compatibility:
lm,ln
+
ln,lm
mn,ll
=
ll,mn

11,22
+
22,11
= 2
12,12
=
12,12
22,33
+
33,22
= 2
23,23
=
23,23
33,11
+
11,33
= 2
31,31
=
31,31
12,13
+
13,12
23,11
=
11,23
23,21
+
21,23
31,22
=
22,31
31,32
+
32,31
12,33
=
33,12
3.8 Exercise
1. The displacement eld of a continuum body is given by
X
1
= x
1
X
2
= x
2
+Ax
3
X
3
= x
3
+Ax
2
where A is a constant. Determine the displacement vector components in both the
material and spatial form.
3.8. EXERCISE 43
2. A continuum body undergoes the displacement
u =
_
_
_
3x
2
4x
3
2x
1
x
3
4x
2
x
1
_
_
_
.
Determine the displaced position of the vector joining particles A(1, 0, 3) and
B(3, 6, 6).
3. A displacement eld is given by u
1
= 3x
1
x
2
2
, u
2
= 2x
3
x
1
and u
3
= x
2
3
x
1
x
2
. De-
termine the strain tensor
ij
and check whether or not the compatibility conditions
are satised.
4. A rectangular loaded plate is clamped along the x
1
- and x
2
-axis (see g. 3.4). On
the basis of measurements, the approaches
11
= a(x
2
1
x
2
+ x
3
2
);
22
= bx
1
x
2
2
are
suggested.
x
2
, u
2
x
1
, u
1
Figure 3.4: Rectangular plate
(a) Check for compatibility!
(b) Find the displacement eld and
(c) compute shear strain
12
.
4 Material behavior
4.1 Uniaxial behavior
Constitutive equations relate the strain to the stresses. The most elementary description
is Hookes law, which refers to a onedimensional extension test
11
= E
11
(4.1)
where E is called the modulus of elasticity, or Youngs modulus.
Looking on an extension test with loading and unloading a dierent behavior is found
(g. 4.1).
Figure 4.1: - diagram of an extension test

There is the linear area governed by Hookes law. In yielding occure which must be
governed by ow rules. is the unloading part where also in pressure yielding exist .
Finally, a new loading path with linear behavior starts. The region given by this curve is
known as hysteresis loop and is a measure of the energy dissipated through one loading
and unloading circle.
44
4.2. GENERALIZED HOOKES LAW 45
Nonlinear elastic theory is also possible. Then path is curved but in loading and
unloading the same path is given.
4.2 Generalized Hookes law
4.2.1 General anisotropic case
As a prerequisite to the postulation of a linear relationship between each component of
stress and strain, it is necessary to establish the existence of a strain energy density W
that is a homogeneous quadratic function of the strain components. The density function
should have coecients such that W 0 in order to insure the stability of the body, with
W(0) = 0 corresponding to a natural or zero energy state. For Hookes law it is
W =
1
2
C
ijkm
ij
km
. (4.2)
The constitutive equation, i.e., the stressstrain relation, is a obtained by
ij
=
W
ij
(4.3)
yielding the generalized Hookes law
ij
= C
ijkm
km
. (4.4)
There, C
ijkm
is the fourthorder material tensor with 81 coecients. These 81 coecients
are reduced to 36 distinct elastic constants taking the symmetry of the stress and the strain
tensor into account. Introducing the notation
= (
11

22

33

12

23

31
)
T
(4.5)
and
= (
11

22

33
2
12
2
23
2
31
)
T
(4.6)
Hookes law is
K
= C
KM
M
K, M = 1, 2, . . . , 6 (4.7)
and K and M represent the respective double indices:
1 = 11, 2 = 22, 3 = 33, 4 = 12, 5 = 23, 6 = 31.
From the strain energy density the symmetry of the materialtensor
C
ijkm
= C
kmij
or C
KM
= C
MK
(4.8)
is obvious yielding only 21 distinct material constants in the general case. Such a material
is called anisotropic.
46 CHAPTER 4. MATERIAL BEHAVIOR
4.2.2 Planes of symmetry
Most engineering materials possess properties about one or more axes, i.e., these axes
can be reversed without changing the material. If, e.g., one plane of symmetry is the
x
2
x
3
plane the x
1
axis can be reversed (g. 4.2),
x
1
x
2
x
3
(a) Original coordinate system
x
1
x
2
x
3
(b) Onesymmetry plane
x
2
x
1
x
3
(c) Twosymmetry planes

Figure 4.2: Coordinate systems for dierent kinds of symmetry
yielding a transformation
x =
_
_
1 0 0
0 1 0
0 0 1
_
_
x
. (4.9)
With the transformation property of tensors
ij
=
ik
jl
kl
(4.10)
and
ij
=
ik
jl
kl
(4.11)
it is
_
_
_
_
_
_
_
_
_
_
11
22
33
12
23
31
_
_
_
_
_
_
_
_
_
_
=
_
_
_
_
_
_
_
_
_
_
11
22
33
12
23
31
_
_
_
_
_
_
_
_
_
_
= C
_
_
_
_
_
_
_
_
_
_
11
22
33
2
12
2
23
2
31
_
_
_
_
_
_
_
_
_
_
= C
_
_
_
_
_
_
_
_
_
_
11
22
33
2
12
2
23
2
31
_
_
_
_
_
_
_
_
_
_
. (4.12)
The above can be rewritten
=
_
_
C
11
C
12
C
13
C
14
C
15
C
16
C
22
C
23
C
24
C
25
C
26
C
33
C
34
C
35
C
36
C
44
C
45
C
46
sym. C
55
C
56
C
66
_
_
(4.13)
but, since the constants do not change with the transformation, C
14
, C
16
, C
24
, C
26
, C
34
,
C
36
, C
45
, C
56
!
= 0 leaving 21 8 = 13 constants. Such a material is called monocline.
The case of three symmetry planes yields an orthotropic material written explicitly as
=
_
_
C
11
C
12
C
13
0 0 0
C
22
C
23
0 0 0
C
33
0 0 0
C
44
0 0
sym. C
55
0
C
66
_
_
(4.14)
with only 9 constants. Further simplications are achieved if directional independence,
i.e., axes can be interchanged, and rotational independence is given. This reduces the
numbers of constants to two, producing the familiar isotropic material. The number of
constants for various types of materials may be listed as follows:
21 constants for general anisotropic materials;
9 constants for orthotropic materials;
2 constants for isotropic materials.
We now summarize the elastic constant stiness coecient matrices for a few selected
materials.
Orthotropic: 9 constants
C
11
C
12
C
13
0 0 0
C
22
C
23
0 0 0
C
33
0 0 0
C
44
0 0
sym. C
55
0
C
66
(4.15)
Isotropic: 2 constants
C
11
C
12
C
12
0 0 0
C
11
C
12
0 0 0
C
11
0 0 0
1
2
(C
11
C
12
) 0 0
sym.
1
2
(C
11
C
12
) 0
1
2
(C
11
C
12
)
(4.16)
A number of eective modulus theories are available to reduce an inhomogeneous multi-
layered composite material to a single homogeneous anisotropic layer for wave propagation
and strength considerations.
4.2.3 Isotropic elastic constitutive law
Using the Lame constants , the stress strain relationship is
=
_
_
2 + 0 0 0
2 + 0 0 0
2 + 0 0 0
2 0 0
sym. 2 0
2
_
_
_
11
22
33
12
23
31
_
_
(4.17)
or in indical notation using the stress and strain tensors
ij
= 2
ij
+
ij
kk
(4.18)
or vice versa
ij
=

ij
2

ij
kk
2(2 + 3)
. (4.19)
Other choices of 2 constants are possible with
the shear modulus
= G =
E
2(1 +)
, (4.20)
=
E
(1 +)(1 2)
, (4.21)
Youngs modulus
E =
(2 + 3)
+
, (4.22)
Poissons ratio
=

2( +)
, (4.23)
bulk modulus
K =
E
3(1 2)
=
3 + 2
3
. (4.24)
From equation (4.21) it is obvious 1 < < 0.5 if remains nite. This is, however, true
only in isotropic elastic materials. With the denition of Poissons ratio
=
22
11
=
33
11
(4.25)
a negative value produces a material which becomes thicker under tension. These mate-
rials can be produced in reality.
The other limit = 0.5 can be discussed as: Taking the 1principal axes as
(1)
= then
both other are
(2)
=
(3)
= (see equation (4.25)). This yields the volume change
V
V
=
ii
= (1 2) (4.26)
Now, = 0.5 gives a vanishing volume change and the material is said to be incompress-
ible. Rubberlike materials exhibit this type of behavior.
Finally, using the compression/bulk modulus K and the shear modulus G and further
the decomposition of the stress and strain tensor into deviatoric and hydrostatic part,
Hookes law is a given (e
ij
are the components of
D
)
kk
= 3K
kk
s
ij
= 2Ge
ij
. (4.27)
4.2.4 Thermal strains
In the preceeding an isothermal behavior was assumed. For temperature change, it is
reasonable to assume a linear relationship between the temperature dierence and the
strain
ij
(T) = (T T
0
)
ij
(4.28)
with the reference temperature T
0
and the constant assumed coecient of thermal expan-
sion. So, Hookes law becomes
ij
=
1
E
[(1 +)
ij
ij
kk
] +
ij
(T T
0
) (4.29)
or in stresses
ij
= 2
ij
+
kk
ij
ij
(3 + 2)(T T
0
). (4.30)
Here, it is assumed that the other material constants, e.g., E and , are independent of
temperature which is valid only in a small range.
4.3 Elastostatic/elastodynamic problems
In an elastodynamic problem of a homogenous isotropic body, certain eld equations,
namely
1. Equilibrium
ij,j
+f
i
= u
i
+f = u (4.31)
2. Hookes law
ij
=
ij
kk
+ 2
ij
= I3
M
+ 2E (4.32)
3. Straindisplacement relations
ij
=
1
2
(u
i,j
+u
j,i
) E =
1
2
(u
T
+u
T
) (4.33)
must be satised at all interior points of the body. Also, prescribed conditions on stress
and/or displacements must be satised on the surface of the body. In case of elastody-
namics also initial conditions must be specied. The case of elastostatic is given when
u
i
can be neglected.
4.3.1 Displacement formulation
With a view towards retaining only the displacements u
i
the strains are eliminated in
Hookes law by using the straindisplacement relations
ij
=
ij
u
k,k
+(u
i,j
+u
j,i
) . (4.34)
Taking the divergence of
ij
(=
ij,j
) the equilibrium is given in displacements
u
k,ki
+(u
i,jj
+u
j,ij
) +f
i
= u
i
. (4.35)
Rearranging with respect to dierent operators yields
u
i,jj
+ ( +)u
j,ij
+f
i
= u
i
(4.36)
or
2
u + ( +) u +f = u . (4.37)
These equations governing the displacements of a body are called Lame/Navier equations.
If the displacement eld u
i
(x
i
) is continuous and dierentiable, the corresponding strain
eld
ij
always satisfy the compatibility constrains.
4.3.2 Stress formulation
An alternative representation is to synthesize the equation in terms of the stresses. Com-
bining the compatibility constraints with Hookes law and inserting them in the static
equilibrium produce the governing equations
ij,kk
+

kk,ij
1 +
+f
i,j
+f
j,i
+

1
ij
f
k,k
= 0 (4.38)
which are called the BeltramiMichell equations of compatibility. To achieve the above
six equations from the 81 of the compatibility constrains several operations are necessary
using the equilibrium and its divergence. Any stress state fullling this equation and the
boundary conditions
t = n (4.39)
is a solution for the stress state of a body loaded by the forces f.
Material behavior
Generalized Hookes Law
ij
= C
ijkm
km

K
= C
KM

M
with K, M = 1, 2, ..., 6 and K, M represent the respective double indices:
1 =11, 2 =22, 3 =33, 4 =12, 4 =23, 6 =31
_
_
_
_
_
_
_
_
_
_
11
22
33
12
23
31
_
_
_
_
_
_
_
_
_
_
=
_
_
_
_
_
_
_
_
_
_
C
11
C
12
C
13
C
14
C
15
C
16
C
12
C
22
C
23
C
24
C
25
C
26
C
13
C
23
C
33
C
34
C
35
C
36
C
14
C
24
C
34
C
44
C
45
C
46
C
15
C
25
C
35
C
45
C
55
C
56
C
16
C
26
C
36
C
46
C
56
C
66
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
11
22
33
12
23
31
_
_
_
_
_
_
_
_
_
_
Orthotropic material
_
_
_
_
_
_
_
_
_
_
11
22
33
12
23
31
_
_
_
_
_
_
_
_
_
_
=
_
_
_
_
_
_
_
_
_
_
C
11
C
12
C
13
0 0 0
C
12
C
22
C
23
0 0 0
C
13
C
23
C
33
0 0 0
0 0 0 C
44
0 0
0 0 0 0 C
55
0
0 0 0 0 0 C
66
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
11
22
33
12
23
31
_
_
_
_
_
_
_
_
_
_
Isotropic material
_
_
_
_
_
_
_
_
_
_
11
22
33
12
23
31
_
_
_
_
_
_
_
_
_
_
=
_
_
_
_
_
_
_
_
_
_
2 + 0 0 0
2 + 0 0 0
2 + 0 0 0
0 0 0 2 0 0
0 0 0 0 2 0
0 0 0 0 0 2
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
11
22
33
12
23
31
_
_
_
_
_
_
_
_
_
_
4.5. EXERCISE 53
Relation between Lame constants , and engineering constants:
= G =
E
2(1 +)
E =
(2 + 3)
+
=
E
(1 +)(1 2)
=

2( +)
K =
E
3(1 2)
=
3 + 2
3
Thermal strains:
ij
(T) = (T T
0
)
ij
ij
= 2
ij
+
kk
ij
ij
(3 + 2)(T T
0
)
4.5 Exercise
1. Determine the constitutive relations governing the material behavior of a point hav-
ing the properties described below. Would the material be classied as anisotropic,
orthotropic or isotropic?
(a) state of stress:
11
= 10.8;
22
= 3.4;
33
= 3.0;
12
=
13
=
23
= 0
corresponding strain components:
11
= 10 10
4
;
22
= 2 10
4
;
33
= 2 10
4
;
12
=
23
=
31
= 0
(b) state of stress:
11
= 10;
22
= 2;
33
= 2;
12
=
23
=
31
= 0
corresponding strains:
11
= 10 10
4
;
22
=
33
=
12
=
23
=
31
= 0
(c) state of stress:
When subjected to a shearing stress
12
,
13
or
23
of 10, the material develops
no strain except the corresponding shearing strain, with tensor component
12
,
13
or
23
, of 20 10
4
.
2. A linear elastic, isotropic cuboid is loaded by a homogeneous temperature change.
Determine the stresses and strains of the cuboid, if
(a) expansion in x
1
and x
2
-direction is prevented totally and if there is no preven-
tion in x
3
-direction.
(b) only in x
1
-direction, the expansion is prevented totally.
3. For steel E = 30 10
6
and G = 12 10
6
. The components of strain at a point within
this material are given by
=
_
_
_
0.004 0.001 0
0.001 0.006 0.004
0 0.004 0.001
_
_
_
.
Compute the corresponding components of the stress tensor
ij
.
5 Twodimensional elasticity
Many problems in elasticity may be treated satisfactory by a twodimensional, or plane
theory of elasticity. In general, two cases exists.
1. The geometry of the body is essentially that of a plate, i.e., one dimension is much
smaller than the others and the applied load is uniformly over the thickness dis-
tributed and act in that plane. This case is called plane stress (g. 5.1).
x
1
x
2
x
2
x
3
Figure 5.1: Plane stress: Geometry and loading
2. The geometry of the body is essentially that of a prismatic cylinder with one dimen-
sion much larger than the others. The loads are uniformly distributed with respect
to the large dimension and act perpendicular to it. This case is called plane strain
(g. 5.2).
55
56 CHAPTER 5. TWODIMENSIONAL ELASTICITY
x
1
x
2
x
3
Figure 5.2: Plane strain: Geometry and loading
5.1 Plane stress
Under the assumptions given above the stress components in x
3
direction vanish
33
=
13
=
23
!
= 0 (5.1)
and the others are = (x
1
, x
2
) only. Accordingly, the eld equations for plane stress
are
ij,j
+f
i
= u
i
i, j = 1, 2 (5.2)
and
f
3
!
= 0 . (5.3)
Hookes law is under the condition of
i3
= 0
ij
=
1 +
E

ij
ij
kk
i, j, k = 1, 2 (5.4)
and
33
=
kk
. (5.5)
This result is found by simply inserting the vanishing stress components in the generalized
Hookes law (4.19). So, the stress and strain tensors are
=
_
_
_
11

12
0
12

22
0
0 0 0
_
_
_
(5.6)
5.2. PLANE STRAIN 57
=
_
_
_
11

12
0
12

22
0
0 0
33
_
_
_
. (5.7)
The
i3
, i = 1, 2, are only zero in case of isotropic materials. In terms of the displacement
components u
i
, the eld equations may be combined to give the governing equation
E
2(1 +)
u
i,jj
+
E
2(1 )
u
j,ji
+f
i
= u
i
i, j = 1, 2 . (5.8)
Due to the particular form of the strain tensor, the six compatibility constraints would
lead to a linear function
33
and, subsequent t
0
, a parabolic distribution of the stress over
the thickness. This is a too strong requirement. Normally, only,
11,22
+
22,11
= 2
12,12
(5.9)
is required as an approximation.
5.2 Plane strain
In case of plane strain, no displacements and also no strains in x
3
direction can appear
due to the long extension,
u =
_
_
_
u
1
(x
1
, x
2
)
u
2
(x
1
, x
2
)
0
_
_
_
(5.10)
33
=
13
=
23
= 0. (5.11)
This yields the eld equations
ij,j
+f
i
= u
i
i, j = 1, 2 (5.12)
and
f
3
!
= 0 . (5.13)
Hookes law is then
ij
=
ij
kk
+ 2
ij
i, j, k = 1, 2 (5.14)
and
33
=
kk
, k = 1, 2 (5.15)
where the last condition is concluded from the fact
33
= 0. This is inserted to Hookes
law (4.19) and taken to express
33
. The tensors look
=
_
_
_
11

12
0
12

22
0
0 0
33
_
_
_
(5.16)
=
_
_
_
11

12
0
12

22
0
0 0 0
_
_
_
. (5.17)
The zerovalued shear forces
13
=
23
= 0 are a consequence of zero shear strains
23
=
13
= 0. Hookes law can also be expressed in strains
ij
=
1 +
E

ij
(1 +)
E

ij
kk
. (5.18)
Subsequent, for plane strain problems the Navier equation reads
E
2(1 +)
u
i,jj
+
E
2(1 +)(1 2)
u
j,ji
+f
i
= u
i
. (5.19)
From this equation, or more obvious from Hookes law, it is seen that exchanging
E
1
2
with
E and

1
with in plane strain or plane stress problems, respectively, allows to treat
both problems by the same equations. Contrary to plane stress, here, all compatibility
constraints are fullled, and only
11,22
+
22,11
= 2
12,12
(5.20)
remains to be required.
5.3 Airys stress function
First, assuming plane stress equation and introducing a potential function
f = V f
i
= V
,i
. (5.21)
Forces which can be expressed in the above way are called conservative forces. Introducing
further some scalar function (x) with
11
=
,22
+V (5.22a)
22
=
,11
+V (5.22b)
12
=
,12
(5.22c)
5.3. AIRYS STRESS FUNCTION 59
then Hookes law is
11
=
1
E
[(
,22
,11
) + (1 )V ]
22
=
1
E
[(
,11
,22
) + (1 )V ]
12
=
1
2G
,12
33
=
E
[(
,11
+
,22
) + 2V ] .
(5.23)
Inserting these strain representations in the compatibility constraints yields
1
E
[
,2222
,1122
+ (1 )V
,22
+
,1111
,1122
+ (1 )V
,11
] =
1
G
,1212
. (5.24)
Rearranging and using
E
G
= 2(1 +) the equation
4
= (1 )
2
V (plane stress) (5.25)
with
4
( ) = ( )
,1111
+ 2( )
,1122
+ ( )
,2222
(5.26)
is achieved. In the case of plane strain, the corresponding equation is
4
=
(1 2)
(1 )

2
V (plane strain). (5.27)
For vanishing potentials V , i.e., vanishing or constant body forces, equations (5.25) and
(5.27) are identical to
4
= 0 (5.28)
.
Further, are called Airy stress function. These functions satisfy the equilibrium and the
compatibility constraints. The solution to the biharmonic problem in Cartesian coordi-
nates is most directly written in terms of polynomials having the general form
=
n
C
mn
x
m
1
x
n
2
. (5.29)
This function has then to be applied to the boundary conditions.
Plane stress
33
=
13
=
23
!
= 0
=
_
_
_
11

12
0
12

22
0
0 0 0
_
_
_
=
_
_
_
11

12
0
12

22
0
0 0
33
_
_
_
Plane strain
u
3
= 0
33
=
13
=
23
!
= 0
=
_
_
_
11

12
0
12

22
0
0 0
33
_
_
_
=
_
_
_
11

12
0
12

22
0
0 0 0
_
_
_
Airys stress function
The solution of an elastic problem is found, if the Airys stress function is known, which
fullls the biharmonic equation
=
4
=

4
x
4
1
+ 2

4
x
2
1
x
2
2
+

4
x
4
2
= 0
fullls the boundary conditions of the problem
Stresses:
11
=

2
x
2
2
=
,22

22
=

2
x
2
1
=
,11

12
=

2
x
1
x
2
=
,12
Boundary conditions:
You have to distinguish between stress boundary conditions and displacement boundary
conditions.
Generally, at every boundary you can give either a statement about stresses or displace-
ments.
i.e.:
At a free boundary all stresses are known ( = 0), the displacements are not known
a priori.
At a clamped boundary the displacements are known (u = 0), the stresses have to
be evaluated.
Surface tractions t at the boundary:
ij
n
j
= t
n
i
11
n
1
+
12
n
2
= t
1
12
n
1
+
22
n
2
= t
2
5.5 Exercise
1. Which problem can be described by the stress function
=
F
d
3
x
1
x
2
2
(3d 2x
2
)
with the limits 0 x
1
5d, 0 x
2
d?
2. A disc (g. 5.3) is loaded by forces F and P. The following parameters are known:
l, h, thickness t of the disc which yields t l, h
2h
l
x
1
x
3
x
2
F
P
Figure 5.3: Clamped disc under loading
(a) Determine the stress boundary conditions for all boundaries.
(b) Determine the stress eld of the disc using the Airys stress function.
6 Energy principles
Energy principles are another representation of the equilibrium and boundary conditions
of a continuum. They are mostly used for developing numerical methods as, e.g., the
FEM.
6.1 Work theorem
Starting from the strain energy density of linear elastic material W =
1
2
ij
ij
integrated
over the volume leads to the total strain energy in the mixed form
E
S
=
1
2
_
V
ij
ij
dV . (6.1)
Introducing Hookes law (4.4) yields the representation in strains
E
S
=
_
V
WdV =
1
2
_
V
ij
C
ijkm
km
dV (see equation (4.2)) (6.2)
or with the inverse of the material tensor
E
S
=
1
2
_
V
ij
C
1
ijkl
kl
dV (6.3)
represented in stresses. The equivalence of equation (6.2) and equation (6.3) is valid only
for Hookes law. Assuming a linear straindisplacement relation
ij
=
1
2
(u
i,j
+ u
j,i
) the
total strain energy can be reformulated
2E
S
=
_
V
ij
1
2
(u
i,j
+u
j,i
)dV =
_
V
ij
u
i,j
dV . (6.4)
63
64 CHAPTER 6. ENERGY PRINCIPLES
In the last integral the symmetry of the stress tensor and interchanging of the indices
ij
u
i,j
=
ji
u
j,i
is used. Next, partial integration and the Gaussian integral theorem
yields
_
V
[
ij,j
u
i
+
ij
u
i,j
]dV =
_
V
(
ij
u
i
)
,j
dV =
_
A
ij
u
i
n
j
dA . (6.5)
Introducing the boundary condition
ij
n
j
= t
i
and the static equilibrium
ij,j
= f
i
it
reads
2E
S
=
_
A
ij
u
i
n
j
dA
_
V
ij,j
u
i
dV
=
_
A
t
i
u
i
dA +
_
V
f
i
u
i
dV .
(6.6)
This expression is called work theorem which is in words:
Twice the total strain energy is equal to the work of the inner force, i.e., the
body force f, and of the outer force, i.e., the surface traction, t, on the
displacements.
Assuming now an elastic body loaded by two dierent surface tractions or body forces
t
(1)
and t
(2)
or f
(1)
and f
(2)
, respectively, results in two states of deformation:
t
(1)
i
and f
(1)
i

(1)
ij
,
(1)
ij
, u
(1)
i
(6.7)
t
(2)
i
and f
(2)
i

(2)
ij
,
(2)
ij
, u
(2)
i
(6.8)
Dening the interaction energy of such a body with the stresses due to the rst loading
and the strains of the second loading:
W
12
=
_
V
(1)
ij

(2)
ij
dV =
_
A
t
(1)
i
u
(2)
i
dA +
_
V
f
(1)
i
u
(2)
i
dV (6.9)
With Hookes law it is obvious
(1)
ij

(2)
ij
=
(1)
kl
C
klij
(2)
ij
=
(1)
kl

(2)
lk
(6.10)
taking the symmetry of the material tensor into account. So, concluding from this the
interaction energy is
W
12
=
_
V
(1)
ij

(2)
ij
dV =
_
V
(2)
ij

(1)
ij
dV = W
21
(6.11)
6.2. PRINCIPLES OF VIRTUAL WORK 65
t
A
1
A
2
u
Figure 6.1: Elastic body with two dierent boundary conditions
and, subsequently, it holds
_
A
t
(1)
i
u
(2)
i
dA +
_
V
f
(1)
i
u
(2)
i
dV =
_
A
t
(2)
i
u
(1)
i
dA +
_
V
f
(2)
i
u
(1)
i
dV , (6.12)
i.e., the work of the surface forces and body forces of state 1 on the displacements
of state 2 is equal to the work of the surface forces and body forces of state 2 on the
displacements of state 1. This is called the Theorem of Betti or Reciprocal work theorem.
6.2 Principles of virtual work
6.2.1 Statement of the problem
An elastic body V with boundary A = A
1
+A
2
(see g. 6.1) is governed by the boundary
value problem
1. the equilibrium static conditions
ij,j
= f
i
in V (6.13)
and the boundary conditions
ij
n
j
= t
i
on A
1
(6.14)
and
2. the compatibility constraints geometric conditions
E = 0 in V (6.15)
with the boundary conditions
u
i
= u
i
on A
2
(6.16)
but these are identically satised when the strains are desired by dierentiation from the
displacements, i.e.,
ij
=
1
2
(u
i,j
+u
j,i
) (6.17)
when the dierentiability of the displacements is given.
In the above the bar

() denotes given values.
For approximations which satisfy only
1. the geometric condition are called geometrical admissible approximations u
i
u
i
= u
exact
i
+u
i
(6.18)
with u
i
small virtual displacement satisfying u
i
!
= 0 on A
2
.
2. the static conditions are called statically admissible approximations
ij
ij
=
exact
ij
+
ij
(6.19)
with
ij
n
j
= 0 on A
1
The virtual changes of the displacements or stresses are small, i.e., innitesimal, and real
but possible. Based on these preliminaries the principles of virtual work can be dened
either by assuming virtual displacements or virtual forces inserted in the work theorem.
6.2.2 Principle of virtual displacements
The virtual work due to a virtual displacement is given by
W =
_
A
t
i
u
i
dA +
_
V
f
i
u
i
dV . (6.20)
Using the property u
i,j
= (u
i
)
,j
and the denition of surface loads t
i
=
ij
n
j
the rst
integral reads
_
A
t
i
u
i
dA =
_
A
ij
u
i
n
j
dA =
_
V
(
ij
u
i
)
,j
dV (6.21)
=
_
V
ij,j
u
i
dV +
_
V
ij
u
i,j
dV . (6.22)
Substituting this result in the virtual work expression yields
W =
_
V
(
ij,j
+f
i
)u
i
dV +
_
V
ij
u
i,j
dV . (6.23)
Since, the equilibrium
ij,j
+f
i
= 0 is valid one nds
_
A
t
i
u
i
dA +
_
V
f
i
u
i
dV
. .
W
external
=
_
V
ij
u
i,j
dV
. .
W
internal
(6.24)
the equivalence of the virtual work of external forces W
external
and the virtual work of
internal forces W
internal
. The virtual work of internal forces are found to be
_
V
ij
u
i,j
dV =
_
V
ij
ij
dV =
_
V
kl
C
klij
ij
dV (6.25)
=
_
V
_
1
2
kl
C
klij
ij
_
dV = W
internal
. (6.26)
In the last rearrangement it is used
_
1
2
kl
C
klij
ij
_
=
1
2
kl
C
klij
ij
+
1
2
kl
C
klij
ij
(6.27)
=
kl
C
klij
ij
(6.28)
based on the product rule and the symmetry of the material tensor. Implementing in the
equivalence W
external
= W
internal
the displacement boundary condition, i.e., assuming
admissible virtual displacements the surface integral over A in W
external
is reduced to an
integral over A
1
yielding
_
V
_
1
2
kl
C
klij
ij
_
dV
_
A
1
t
i
u
i
dA
_
V
f
i
u
i
dV = 0 (6.29)
or in a complete displacement description with
ij
=
1
2
(u
i,j
+u
j,i
)
_
V
_
1
2
u
k,l
C
klij
u
i,j
_
dV
_
A
1
t
i
u
i
dA
_
V
f
i
u
i
dV = 0 . (6.30)
The above given integral equation is called the Principle of virtual displacements where
with the bar the given values are indicated. Taking into account that the volume V and
also the surface A of the elastic body will not change due to the virtual displacement,
and, also, that the prescribed boundary traction t
i
as well as the body forces f
i
will not
change, the variation, i.e., the sign , can be shifted out of the integrals, resulting in
_
_
_
V
_
1
2
u
k,l
C
klij
u
i,j
_
dV
_
A
1
t
i
u
i
dA
_
V
f
i
u
i
dV
_
_
. .
(u
i
)
= 0 . (6.31)
The expression between the brackets is called total potential energy (u
i
). The condition
above
(u
i
) = 0 (stationary potential energy) (6.32)
is a variational equation stating that the exact solution u
exact
i
gives the total potential
energy an extremum. It can be proven to be a minimum. If u
h
i
denotes an approximative
solution where h means the discretization it must be valid
(u
h
2
i
) < (u
h
1
i
) if h
2
< h
1
(6.33)
i.e., the principle of minimum total potential energy.
6.2.3 Principle of virtual forces
Next the complementary principle of the above is given. Instead of varying the displace-
ments, i.e., the geometric conditions, the forces, i.e., the statical condition, are varied. As
dened, virtual forces
t
i
=
ij
n
j
(6.34)
have to satisfy
t
i
= 0 on A
1
(6.35)
to be admissible, and, further, due to the equilibrium
ij,j
= 0 in V . (6.36)
This is caused by the fact that the prescribed body forces f
i
are not varied f
i
0.
Inserting these preliminaries into the work theorem (6.6) and performing the variation
t
i
, it holds
W
external
=
_
A
2
u
i
t
i
dA . (6.37)
The internal virtual work produced by virtual stresses
ij
is
W
internal
=
_
V
ij
ij
dV . (6.38)
Expressing
ij
with Hookes law by stresses and the inverse material tensor yields
W
internal
=
_
V
kl
C
1
klij
ij
dV . (6.39)
As before, the variation can be extracted from the integral
_
V
kl
C
1
klij
ij
dV =
_
V
_
1
2
kl
C
1
klij
ij
+
1
2
kl
C
1
klij
ij
_
dV (6.40)
=
_
V
_
1
2
kl
C
1
klij
ij
_
dV = W
internal
. (6.41)
Now, with the equivalence W
internal
= W
external
and the same argumentation as given
for virtual displacement formulation it is
_
_
_
V
1
2
kl
C
1
klij
ij
dV
_
A
2
u
i
t
i
dA
_
_
. .
c
(
ij
)
= 0 (6.42)
where
c
(
ij
) is called complementary total potential energy. Note, in (6.42) it is dened
with a negative sign. The variational equation
c
(
ij
) = 0 (6.43)
is as in case of virtual displacements an extremum of the total potential energy. Contrary
to there, here, due to the negative sign in the denition of
c
it can be proven to be a
maximum. So, for static admissible approximations
ij
it holds
c
(
ij
)
c
(
exact
ij
) . (6.44)
Clearly, for the exact solutions u
exact
i
and
exact
ij
it is valid
(u
exact
i
) =
c
(
exact
ij
) . (6.45)
Inserting the expression of both energies
_
V
1
2

kl
C
klij
. .
ij
ij
dV
_
A
1
t
i
u
i
dA
_
V
f
i
u
i
dV =
_
V
1
2

kl
C
1
klij
. .
ij
ij
dV +
_
A
2
u
i
t
i
dA (6.46)
yields the general theorem
_
V
ij
ij
dV =
_
V
f
i
u
i
dV +
_
A
1
t
i
u
i
dA +
_
A
2
u
i
t
i
dA , (6.47)
i.e., the work of the external and internal forces at the displacements is equal to twice the
strain energy.
6.3 Approximative solutions
To solve problems with Energy principles for realistic geometries require mostly approx-
imative solutions. Trial functions for the unknowns are dened often by polynomials. If
these functions are admissible such an approximation is called Ritz approximation, e.g.,
for the displacements
u(x) = c , (6.48)
i.e.,
u
i
(x
i
) =
i1
(x
i
)c
1
+
i2
(x
i
)c
2
+. . . +
in
(x
i
)c
n
(6.49)
for i = 1, 2, 3 in 3D or i = 1, 2 in 2D or i = 1 in 1D. The number n can be chosen
arbitrarily, however, it must be checked whether u(x) is admissible, i.e., the geometrical
boundary conditions must be fullled. Using a symbolic notation and with the dierential
operator matrix D from
= Du (6.50)
( see chapter 4) the total potential energy is
(u) =
1
2
_
V
(Du)
T
C
(4)
DudV
_
A
1
t
T
udA
_
V
f
T
udV . (6.51)
Inserting there the Ritz approach yields
( u) =
1
2
c
T
_
V
(D)
T
C
(4)
(D)dV
. .
R
h
c
_
_
_
V
f
T
dV +
_
A
1
t
T
dA
_
_
. .
p
T
h
c (6.52)
=
1
2
c
T
R
h
c p
T
h
c . (6.53)
6.3. APPROXIMATIVE SOLUTIONS 71
In case of the simply supported beam (see exercise 6.5) it was
w(x) = a
0
+a
1
x +a
2
x
2
+a
3
x
3
+a
4
x
4
(6.54)
c =
_
_
_
_
_
_
_
_
a
0
a
1
a
2
a
3
a
4
_
_
_
_
_
_
_
_
(6.55)
= (1, x, x
2
, x
3
, x
4
) (6.56)
or after inserting the geometric boundary conditions
w(x) =
w
B
l
2
x
2
+a
3
(x
3
lx
2
) +a
4
(x
4
l
2
x
2
) (6.57)
= [x
2
, x
3
lx
2
, x
4
l
2
x
2
]
. .
_
w
B
l
2
a
3
a
4
_
_
. .
c
. (6.58)
To determine the unknown coecients c the principle of virtual displacements is used
=

c
c = 0 (6.59)
yielding the equation system
R
h
c = p
h
. (6.60)
Hence, by inserting this result in the total energy the approximative solution u
(x) gives
( u) =
1
2
c
T
p
h
, (6.61)
when the system of equations R
h
c = p
h
is exactly solved.
The same procedure can be introduced for the complementary total potential energy.
With the admissible trial function for the stresses
(x) = c (6.62)
and
t(x) = nc (6.63)
the complementary total potential energy in symbolic notation
c
=
1
2
_
V
T
C
1
dV +
_
A
2
u
T
tdA (6.64)
is approximated by
c
( ) =
1
2
c
T
_
V
T
C
1
dV
. .
F
h
c +
_
A
2
u
T
ndA
. .
u
T
h
c (6.65)
=
1
2
c
T
F
h
c +u
T
h
c . (6.66)
The variation following the principle of virtual forces
c
= 0 (6.67)
yields the equation system
F
h
c = u
h
(6.68)
to determine the coecients c. If this equation system is solved exactly the approximated
complementary total energy is
c
( ) =
1
2
c
T
u
h
. (6.69)
6.3.1 Application: FEM for beam
Starting point is the total potential energy for a beam
beam
=
EI
2
l
_
0
(w
(x))
2
dx
l
_
0
q(x)w(x)dx
n
i=1
F(x
i
)w(x
i
) +
m
j=1
M(x
j
)w
(x
j
) . (6.70)
The next question is on the approximation for the deection w(x). First, the beam is
divided in elements
e
= [x
e
, x
e+1
] wherein each a cubic polynomial is used for the
unknowns w
e
and w
e
, i.e., for the geometric boundary conditions. The transformation
from the global coordinate x [x
e
, x
e+1
] to the local 0 < < 1 is
=
x x
e
x
e
=
x x
e
l
e
(6.71)
with the element-length l
e
. So, the approximation in
e
is
w
(e)
(x) = w
e
1
N
1
_
x x
e
l
e
_
+w
e
1
N
2
_
x x
e
l
e
_
+w
e
2
N
3
_
x x
e
l
e
_
+w
e
2
N
4
_
x x
e
l
e
_
(6.72)
The test functions N
i
are
N
1
() = 1 3
2
+ 2
3
(6.73)
N
2
() = l
e
( 2
2
+
3
) (6.74)
N
3
() = 3
2
2
3
(6.75)
N
4
() = l
e
(
2
+
3
) (6.76)
N
i
N
1
N
2
N
3
N
4
1
45
45
= 0
x = x
e
= 1
x = x
e+1
Figure 6.2: Test functions of one element
with
N
1
( = 0) = 1 N
1
( = 1) = 0
N
2
( = 0) = 0 N
2
( = 1) = 0
N
3
( = 0) = 0 N
3
( = 1) = 1
N
4
( = 0) = 0 N
4
( = 1) = 0
N
1
( = 0) = 0 N
1
( = 1) = 0
N
2
( = 0) = 1 N
2
( = 1) = 0
N
3
( = 0) = 0 N
3
( = 1) = 0
N
4
( = 0) = 0 N
4
( = 1) = 1
Inserting them in the energy, e.g., the strain energy is
EI
2
l
_
0
(w
(x))
2
dx =
EI
2
N
e=1
x
e+1
_
x
e
_
d
2
dx
2
[w
e
1
N
1
() +w
e
1
N
2
() +w
e
2
N
3
() +w
e
2
N
4
()]
_
2
dx
(6.77)
for N elements. To nd the Nsets of nodal values w
e
1
, w
e
1
, w
e
2
and w
e
2
the variation
=

w
e
1
w
e
1
+

w
e
1
w
e
1
+

w
e
2
w
e
2
+

w
e
2
w
e
2
= 0 (6.78)
is performed taking each summand independently to zero. Taking into account
d
2
dx
2
=
1
l
2
e
d
2
d
2
(6.79)
and
x
e+1
_
x
e
_
d
2
dx
2
[. . .]
_
2
dx =
1
_
0
1
l
4
e
_
d
2
d
2
[. . .]
_
2
l
e
d (6.80)
the above variation yields for the strain energy term
w
e
1
:
EI
2
l
e
l
4
e
2
1
_
0
w
e
1
(6 + 12) +w
e
1
l
e
(4 + 6) +w
e
2
(6 12) +w
e
2
l
e
(2 + 6)
(6.81)
(6 + 12)d (6.82)
w
e
1
:
EI
l
3
e
1
_
0
. . . l
e
(4 + 6)d (6.83)
w
e
2
:
EI
l
3
e
1
_
0
. . . (6 12)d (6.84)
w
e
2
:
EI
l
3
e
1
_
0
. . . l
e
(6 2)d . (6.85)
Performing these integrals and gathering the four equations in a matrix the system
EI
l
3
e
_
_
12 12 6l
e
6l
e
12 12 6l
e
6l
e
6l
e
6l
e
4l
2
e
2l
2
e
6l
e
6l
e
2l
2
e
4l
2
e
_
_
_
_
w
e
1
w
e
2
w
e
1
w
e
2
_
_
= K
e
w
e
(6.86)
is obtained with the element stiness matrix K
e
. Equation (6.86) can be reordered in
such a manner that degrees of freedom of each node are consecutive
EI
l
3
e
_
_
12 6l
e
12 6l
e
6l
e
4l
2
e
6l
e
2l
2
e
12 6l
e
12 6l
e
6l
e
2l
2
e
6l
e
4l
2
e
_
_
_
_
w
e
1
w
e
1
w
e
2
w
e
2
_
_
=
_
K
11
K
12
K
21
K
22
_
= K
e
w
e
. (6.87)
The right hand side, i.e., the loading term in the total potential energy is found similar
l
_
0
q(x)w(x)dx =
N
e=1
x
e+1
_
x
e
q(x)w
e
(x)dx . (6.88)
After variation and integration one obtains
w
e
1
:
q
0
l
e
2
(6.89)
w
e
1
:
q
0
l
2
e
12
(6.90)
w
e
2
:
q
0
l
e
2
(6.91)
w
e
2
:
q
0
l
2
e
12
(6.92)
Now, collecting all N elements in one system and taking into account that at adjacent
elements transition conditions (g. 6.3) holds.
e

e+1
w
e
1
w
e
2
= w
e+1
1
and
w
e
2
= w
e+1
1
w
e+1
2
Figure 6.3: Transition conditions at adjacent elements
Further, for simplication, all elements will have the same length l
e
it is
EI
l
3
e
_
_
12 12 0 0 6l
e
6l
e
0 0
12 12 + 12 12 0 6l
e
6l
e
+ 6l
e
6l
e
0
0 12 24 12 0 6l
e
0 6l
e

0 0 12
.
.
.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
4l
2
e
2l
2
e
0
2l
2
e
4l
2
e
+ 4l
2
e
2l
2
e

sym. 0 2l
2
e
8l
2
e
2l
2
e
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
_
_
. .
K
_
_
w
e
1
w
e
2
w
e
3
.
.
.
w
e
N
w
e
1
w
e
2
w
e
3
.
.
.
w
e
N
_
_
. .
w
h
= q
0
l
e
_
_
1
2
1
2
+
1
2
1
.
.
.
1
2
l
e
12
l
e
12
+
l
e
12
0
.
.
.
l
e
12
_
w
e
1
w
2
2
=

w
e+1
1
.
.
.
w
N
2
1
1
1
2
=

w
2
1
N
2
. (6.93)
If we reorder the element matrices K
e
i
like in (6.87) we obtain
K =
_
_
K
11
1
K
12
1
0 0
K
21
1
K
22
1
+K
11
2
K
12
2
0
0 K
21
2
K
22
2
+K
11
3
K
12
3
0 0 K
21
3
K
22
3
_
_
(6.94)
Energy Principles
Work theorem
2E =
_
A
ij
u
i
n
j
dA
_
V
ij,j
u
i
dV
=
_
A
t
i
u
i
dA +
_
A
f
i
u
i
dV
Geometric admissible approximations: u
i
= u
exact
i
+u
i
Statically admissible approximations:
ij
=
exact
ij
+
ij
Principle of virtual displacements
W
external
= W
internal
_
A
t
i
u
i
dA +
_
V
f
i
u
i
dV =
_
V
ij
u
i,j
=
_
V
ij
ij
dV
Principle of virtual forces
W
external
= W
internal

_
A
u
i
t
i
dA =
_
V
ij
ij
dV
Variational Principles
Principle of minimum total potential energy:
(u
i
)
!
= 0
__
V
1
2
u
k,l
C
klij
u
i,j
dV
_
A
t
i
u
i
dA
_
V
f
i
u
i
dV
_
!
= 0
1D:
Strain energy:
1
2
u
k,l
C
klij
u
i,j
=
1
2
xx
xx
=
E
2
(u
,x
)
2
=
E
2
(z(x))
2
=
E
2
z
2
(w
(x))
2
with dV = Adx = b dz dx:
_
V
1
2
u
k,l
C
klij
u
i,j
dV =
_ h
2
z=
h
2
_

x=0
E
2
z
2
(w
(x))
2
bdxdz
=
Eb
2
_ h
2
h
2
z
2
dz
_

x=0
(w
(x))
2
dx =
Ebh
3
2 12
_

0
(w
(x))
2
dx
=
EI
y
2
_

0
(w
(x))
2
dx
Loading:
p(x) =
q(x)
hb

_
V
p(x)w(x)dV =
_

0
q(x)
A
w(x)Adx =
_

0
q(x)w(x)dx
single forces

F(x
i
) :
n
i=0
F(x
i
)w(x
i
)
single moments

M(x
j
) : +
m
j=0
M(x
j
)w
(x
j
)
beam
(w) =
EI
y
2
_

0
(w
(x))
2
dx
_

0
q(x)w(x)dx
i=0
F(x
i
)w(x
i
) +
m
j=0
M(x
j
)w
(x
j
)
Principle of minimum complementary energy:
(
ij
)
!
= 0
__
V
1
2

kl
C
1
klij

ij
dV
_
A
u
i
t
i
dA
_
!
= 0
1D:
_
V
1
2

kl
C
1
klij

ij
dV =
_

x=0
_ h
2
z=
h
2
1
2
xx
1
E
xx
bdzdx
with
xx
=
M
y
(x)
I
y
z :
=
_

x=0
1
2
M
2
y
(x)
EI
2
y
dx
_ h
2
z=
h
2
z
2
dA =
1
2EI
y
_

x=0
M
2
y
(x)dx
6.5. EXERCISE 79
prescr. boundary bending w(x
i
):
n
i=0
F(x
i
) w(x
i
)
prescr. boundary incline w
(x
j
) : +
m
j=0
M(x
j
) w
(x
j
)
beam
(M) =
1
2EI
_

x=0
M
2
y
(x)dx
i=0
F(x
i
) w(x
i
) +
m
j=0
M(x
j
) w
(x
j
)
6.5 Exercise
1. Consider a beam under constant loading q(x) = q
0
, which is clamped at x = 0 and
simply supported at x = l, where this support is moved in z-direction for a certain
value, i.e. w(x = l) = w
B
(g. 6.4)!
z
x
w
B
q(x) = q
0
Figure 6.4: Beam under loading
(a) Solve via Principle of minimum total potential energy!
(b) Solve via Principle of minimum complementary energy!
A Solutions
A.1 Chapter 1
1. (a)
gradf(x
1
, x
2
, x
3
) =
_
_
_
f(x
1
,x
2
,x
3
)
x
1
f(x
1
,x
2
,x
3
)
x
2
f(x
!
,x
2
,x
3
)
x
3
_
_
_
=
_
_
_
3 +e
x
2
+x
2
e
x
3
x
1
e
x
2
+x
1
e
x
3
x
1
x
2
e
x
3
_
_
_
(b)
gradf(3, 1, 0) =
_
_
_
3 +e
1
+ 1e
0
3e
1
+ 3e
0
3 1 e
0
_
_
_
=
_
_
_
4 +e
3 +e
3
_
_
_
2. general:
f
a
(p
1
, p
2
, p
3
) =
a
[a[
gradf(p
1
, p
2
, p
3
)
with magnitude [a[ =
_
a
2
1
+a
2
2
+a
2
3
here:
gradf(x
1
, x
2
, x
3
) =
_
_
_
2x
1
3x
2
0
_
_
_
gradf(5, 2, 8) =
_
_
_
10
6
0
_
_
_
_
3
0
4
_
3
2
+ 0
2
+ 4
2
_
_
_
10
6
0
_
_
_
=
1
5
_
_
_
3
0
4
_
_
_
_
_
_
10
6
0
_
_
_
=
1
5
(30 + 0 + 0) = 6
80
A.1. CHAPTER 1 81
3. (a)
div
_
_
_
x
1
+x
2
2
e
x
1
x
3
+ sin x
2
x
1
x
2
x
3
_
_
_
= 1 + cos x
2
+x
1
x
2
(b)
div
_
_
_
X(1, , 2)
Y (1, , 2)
Z(1, , 2)
_
_
_
= 1 + cos + 1 =
4. (a)
curl
_
_
_
x
1
+x
2
e
x
1
+x
2
+x
3
x
3
+ sin x
1
_
_
_
=
_
_
_
x
1
+x
2
e
x
1
+x
2
+x
3
x
3
+ sin x
1
_
_
_
=
_
_
_
x
2
(x
3
+ sin x
1
)

x
3
(e
x
1
+x
2
+x
3
)
x
3
(x
1
+x
2
)

x
1
(x
3
+ sin x
1
)
x
1
(e
x
1
+x
2
+x
3
)

x
2
(x
1
+x
2
)
_
_
_
=
_
_
_
1
cos x
1
e
x
1
+x
2
1
_
_
_
(b)
curl
_
_
_
X(0, 8, 1)
Y (0, 8, 1)
Z(0, 8, 1)
_
_
_
=
_
_
_
1
1
e
8
1
_
_
_
5. expansion of D
ij
x
i
x
j
:
D
ij
x
i
x
j
= D
1j
x
1
x
j
+D
2j
x
2
x
j
+D
3j
x
3
x
j
= D
11
x
1
x
1
+D
12
x
1
x
2
+D
13
x
1
x
3
+D
21
x
2
x
1
+D
22
x
2
x
2
+D
23
x
2
x
3
+D
31
x
3
x
1
+D
32
x
3
x
2
+D
33
x
3
x
3
simplifying:
(a) D
ij
= D
ji
D
ij
x
i
x
j
= D
11
(x
1
)
2
+D
22
(x
2
)
2
+D
33
(x
3
)
2
+ 2D
12
x
1
x
2
+ 2D
13
x
1
x
3
+ 2D
23
x
2
x
3
(b) D
ij
= D
ji
D
ij
x
i
x
j
= D
11
(x
1
)
2
+D
22
(x
2
)
2
+D
33
(x
3
)
2
= 0
because D
12
= D
21
, D
13
= D
31
, D
23
= D
32
and D
11
= D
11
, D
22
= D
22
,
D
33
= D
33
82 APPENDIX A. SOLUTIONS
6. (a)
f
2
= c
2,1
b
1
c
1,2
b
1
+c
2,2
b
2
c
2,2
b
2
+c
2,3
b
3
c
3,2
b
3
= (c
2,1
c
1,2
)b
1
+ (c
2,3
c
3,2
)b
3
(b)
f
2
= B
21
f
1
+B
22
f
2
+B
23
f
3
7. (a)
f = f
,i
=
f
x
i
=
f
r

r
x
i
with
r
x
i
=

x
i
(x
i
x
i
)
1
2
=
1
2
(x
i
x
i
)
1
2
(
x
i
x
i
x
i
+x
i
x
i
) =
1
2
(x
i
x
i
)
1
2
2x
i
=
x
i
x
i
x
i
=
x
i
r
follows
f =
f
r

x
i
r
=
f
(r)x
r
or expanded:
r
2
= x
2
1
+x
2
2
+x
2
3
f =
f
r

r
x
i
with
r
x
i
=
_
_
_
r
,1
r
,2
r
,3
_
_
_
=
_
_
_
_
(x
2
1
+x
2
2
+x
2
3
)
1
2
,1
(x
2
1
+x
2
2
+x
2
3
)
1
2
,2
(x
2
1
+x
2
2
+x
2
3
)
1
2
,3
_
_
_
_
=
_
_
_
_
1
2
(x
2
1
+x
2
2
+x
2
3
)
1
2
,1
1
2
(x
2
1
+x
2
2
+x
2
3
)
1
2
,2
1
2
(x
2
1
+x
2
2
+x
2
3
)
1
2
,3
_
_
_
_
=
_
_
_
x
1
r
x
2
r
x
3
r
_
_
_
follows
f =
f
r

x
r
A.1. CHAPTER 1 83
(b)
2
f = f
,ii
= (f
,i
)
,i
=
_
f
(r) x
i
r
_
,i
=
(f
(r) x
i
)
,i
r f
(r) x
i
r
,i
r
2
=
f
(r)
,i
x
i
r +f
(r) x
i,i
r f
(r) x
i
r
,i
r
2
=
f
(r)
r
x
i
x
i
r +f
(r) x
i,i
r f
(r) x
i
r
x
i
r
2
=
f
(r)
x
i
r
x
i
r +f
(r) x
i,i
r f
(r) x
i
x
i
r
r
2
= f
(r) + 3f
(r)
1
r
f
(r)
1
r
= f
(r) +
2
r
f
(r)
A.2 Chapter 2
1. (a)
t
i
=
ji
e
j
=
ij
e
j
(
ij
=
ji
)
t =
_
_
_
20000 2000 1000
2000 15000 2000
1000 2000 3000
_
_
_
3
_
_
_
1
1
1
_
_
_
=
1
3
_
_
_
23000
11000
6000
_
_
_
(b) normal component:
nn
=
ij
n
j
= t
i
n
i
nn
= t
i
n
i
=
1
3
_
_
_
23000
11000
6000
_
_
_
3
_
_
_
1
1
1
_
_
_
=
1
3
(23000 11000 + 6000) = 6000
tangential component:
ns
=
ij
n
i
s
j
=
_
t
i
t
i
2
nn
ns
=
_
1
3
_
_
_
23000
11000
6000
_
_
_
1
3
_
_
_
23000
11000
6000
_
_
_
6000
2
=
_
1
3
[23000
2
+ (11000)
2
+ 6000
2
] 6000
2
= 13880.44
2. static problem
ij,j
+f
i
= 0
ij
=
_
_
_
2x
2
1
3x
2
2
5x
3
x
3
+ 4x
1
x
2
6 3x
1
+ 2x
2
+ 1
x
3
+ 4x
1
x
2
6 2x
2
2
+ 7 0
3x
1
+ 2x
2
+ 1 0 4x
1
+x
2
+ 3x
3
5
_
_
_
ij,j
+f
i
= 0
i = 1 :
11,1
+
12,2
+
13,3
+f
1
= 0
i = 2 :
21,1
+
22,2
+
23,3
+f
2
= 0
i = 3 :
31,1
+
32,2
+
33,3
+f
3
= 0
4x
1
+ 4x
1
+ 0 +f
1
= 0 f
1
= 0
4x
2
4x
2
+ 0 +f
2
= 0 f
2
= 0
3 + 0 + 3 +f
3
= 0 f
3
= 0
f = 0
A.2. CHAPTER 2 85
3.
ij
=
ik

jl

kl
=
ik

kl

jl
calculation of
using Falk scheme:

T

2 2 0 0
1
2

1
2
2
2 0
1
2
1
2
1
2
0 0
2
1
2

1
2

1
2
0
1
2
1
2

2
2
1 1 0 0 2
1
2
1
2

1
2
2
2
1
2
2
+
2
2
2
2
0 1
2 1
2
1
2

1
2

2
2
1
2
2
+
2
2
2
2
2 1 1 +
2
4. stress tensor at point P:
ij
=
_
_
_
0 a 0
a 0 0
0 0 8a
_
_
_
principal stress values:[
ij
(k)
ij
[
!
= 0
0
(k)
a 0
a 0
(k)
0
0 0 8a
(k)
=
(k)
(k)
0
0 8a
(k)
a 0
0 8a
(k)
+ 0
=
(k)2
8a
(k)3
8a
3
+a
2
(k)
!
= 0
(1)
= a
(
(k)2
+
(k)
7a + 8a
2
)
. .
=0
(2)
;
(3)
(
(k)
a)
. .
=0
(1)
=a
!
= 0
(k)2
(k)
7a 8a
2
!
= 0
(2,3)
= +
7a
2

_
7a
2
_
2
+ 8a
2
(2)
= a
(3)
= 8a
principal direction cor. to
(1)
= a:
(
ij
(k)
ij
)n
(k)
j
= 0
(
11
(1)
)n
(1)
1
+
12
n
(1)
2
+
13
n
(1)
3
= 0
21
n
(1)
1
+ (
22
(1)
)n
(1)
2
+
23
n
(1)
3
= 0
31
n
(1)
1
+
32
n
(1)
2
+ (
33
(1)
)n
(1)
3
= 0
(0 a)n
(1)
1
+an
(1)
2
+ 0 = 0 (1)
an
(1)
1
+ (0 a)n
(1)
2
+ 0 = 0 (2)
0 + 0 + (8a a)n
(1)
3
= 0 (3)
(3) : n
(1)
3
= 0
(2) : n
(1)
1
= n
(1)
2
(1) : n
(1)
1
= n
(1)
2
n
(1)
=
_
_
_
1
2
1
2
0
_
_
_
(2)
= a:
(0 +a)n
(2)
1
+an
(2)
2
+ 0 = 0 (1)
an
(2)
1
+an
(2)
2
+ 0 = 0 (2)
0 + 0 + (8a +a)n
(2)
3
= 0 (3)
(3) : n
(2)
3
= 0
(2) : n
(2)
1
= n
(2)
2
(1) : n
(2)
1
= n
(2)
2
n
(2)
=
_
_
_
1
2
0
_
_
_
(3)
= 8a:
8an
(3)
1
+an
(3)
2
+ 0 = 0 (1)
an
(3)
1
8an
(3)
2
+ 0 = 0 (2)
0 + 0 + 0n
(3)
3
= 0 (3)
A.2. CHAPTER 2 87
(3) : n
(3)
3
= arbitrary
(2) : 8n
(3)
2
= n
(3)
1
(1) : n
(3)
2
= 8n
(3)
1
n
(3)
=
_
_
_
0
0
1
_
_
_
5.
ij
=
_
_
_
2 2 0
2
2 0
0 0
2
_
_
_
I
1
=
ii
=
11
+
22
+
33
= 2 +
2 = 2
I
2
=
1
2
(
ii
jj
ij
ij
)
=
11
22
+
22
33
+
33
11
12
12
23
23
31
31
= 6
I
3
=
2 2 0
2
2 0
0 0
= 4 + 4
ij
=
_
_
_
0 0 2
0 1
2 1
2 1 1 +
2
_
_
_
I
1
= 0 + 1
2 + 1 +
2 = 2
I
2
=
11
22
+
22
33
+
33
11
12
12
23
23
31
31
= 6
I
3
=
0 0 2
0 1
2 1
2 1 1 +
= 4 + 4
2
6.
11
=

11
+
22
+
33
3
= 8
hydrostatic stress tensor:
(h)
=
_
_
_
8 0 0
0 8 0
0 0 8
_
_
_
deviatoric stress tensor
S =
_
_
_
3
M
10 0
10 0
M
30
0 30 27
M
_
_
_
=
_
_
_
11 10 0
10 8 30
0 30 19
_
_
_
ij
=
_
_
_
8 0 0
0 8 0
0 0 8
_
_
_
+
_
_
_
11 10 0
10 8 30
0 30 19
_
_
_
control: tr(S)
!
= 0 = 11 + 8 19
principal deviatoric stress:
[S
ij
S
(k)
ij
[
!
= 0
11 S
(k)
10 0
10 8 S
(k)
30
0 30 19 S
(k)
=(11 S
(k)
)
8 S
(k)
30
30 19 S
(k)
+ 10
10 30
0 19 S
(k)
+ 0
=S
(k)3
+ 1273S
(k)
9672
=(S
(k)
31)(S
(k)
8)(S
(k)
+ 39)
A.2. CHAPTER 2 89
S
(1)
= 39
S
(2)
= 8
S
(3)
= 31
7. principal stresses:
(a)
[
ij
(k)
ij
[
!
= 0
0
(k)
1 1
1 0
(k)
1
1 1 0
(k)
=
(k)3
+ 3
(k)
+ 2
!
= 0
(1)
= 1
(
(k)3
+ 3
(k)
+ 2) = (
(k)
+ 1) (
(k)2
+
(k)
+ 2)
!
= 0
(k)2
+
(k)
+ 2
!
= 0
(k)2
(k)
2 = 0
(2,3)
=
1
2

_
1
2
_
2
(2)
(2)
= 1
(3)
= 2
(b)
2
(k)
1 1
1 2
(k)
1
1 1 2
(k)
=(2
(k)
)
2
(k)
1
1 2
(k)
1 1
1 2
(k)
1 2
(k)
1 1
=
(k)3
+ 6
(k)2
9
(k)
+ 4
!
= 0
(1)
= 1
(
(k)2
5
(k)
+ 4) (
(1)
1)
!
= 0
(
(k)2
5
(k)
+ 4)
!
= 0
(2,3)
=
5
2

_
5
2
_
2
4
=
5
2

_
9
4
(2)
= 1
(3)
= 4
principal directions:
(a) principal direction corresponding with
(1)
:
(
11
(1)
)n
(1)
1
+
12
n
(1)
2
+
13
n
(1)
3
= 0
21
n
(1)
1
+ (
22
(1)
)n
(1)
2
+
23
n
(1)
3
= 0
31
n
(1)
1
+
32
n
(1)
2
+ (
33
(1)
)n
(1)
3
= 0
1n
(1)
1
+ 1n
(1)
2
+ 1n
(1)
3
= 0 (1)
1n
(1)
1
+ 1n
(1)
2
+ 1n
(1)
3
= 0 (2)
1n
(1)
1
+ 1n
(1)
2
+ 1n
(1)
3
= 0 (3)
n
(1)
=
_
_
_
1
2
0
_
_
_
n
(2)
=
_
_
_
2
1
2
0
_
_
_
2n
(3)
1
+ 1n
(3)
2
+ 1n
(3)
3
= 0
1n
(3)
1
2n
(3)
2
+ 1n
(3)
3
= 0
1n
(3)
1
+ 1n
(3)
2
2n
(3)
3
= 0
n
(3)
=
_
_
_
1
3
1
3
1
3
_
_
_
A.3. CHAPTER 3 91
A.3 Chapter 3
1.
u = Xx
material description: u(x)
u
1
= X
1
x
1
= x
1
x
1
= 0
u
2
= X
2
x
2
= x
2
+Ax
3
x
2
= Ax
3
u
3
= X
3
x
3
= x
3
+Ax
2
x
3
= Ax
2
spatial description: u(X)
inverting given displacement relations
(1) x
1
= X
1
(2) x
2
= X
2
Ax
3
(3) x
3
= X
3
Ax
2
(3) in (2):
x
2
= X
2
AX
3
+A
2
x
2
x
2
=
X
2
AX
3
1 A
2
in (3):
x
3
= X
3
A
X
2
AX
3
1 A
2
x
3
=
X
3
AX
2
1 A
2
displacement vector:
u
1
= X
1
x
1
= X
1
X
1
= 0
u
2
= X
2
x
2
= X
2
X
2
AX
3
1 A
2
=
X
2
A
2
+AX
3
1 A
2
u
3
= X
3
x
3
= X
3
X
3
AX
2
1 A
2
=
X
3
A
2
+AX
2
1 A
2
2.
u = Xx
_
_
_
3x
2
4x
3
2x
1
x
3
4x
2
x
1
_
_
_
=
_
_
_
X
1
X
2
X
3
_
_
_
_
_
_
x
1
x
2
x
3
_
_
_
+
_
_
_
x
1
x
2
x
3
_
_
_
_
_
_
X
1
X
2
X
3
_
_
_
=
_
_
_
x
1
+ 3x
2
4x
3
2x
1
+x
2
x
3
x
1
+ 4x
2
+x
3
_
_
_
B
= X(B) =
_
_
_
3 + 3 6 4 6
2 3 + 6 6
3 + 4 6 + 6
_
_
_
=
_
_
_
3
6
27
_
_
_
= B
=
_
_
_
3
6
27
_
_
_
_
_
_
11
1
2
_
_
_
=
_
_
_
8
7
25
_
_
_
3. strain tensor:
ij
=
_
_
_
11
1
2
12
1
2
13
1
2
21

22
1
2
23
1
2
31
1
2
32

33
_
_
_
=
_
_
_
u
1,1
1
2
(u
1,2
+u
2,1
)
1
2
(u
1,3
+u
3,1
)
1
2
(u
1,2
+u
2,1
) u
2,2
1
2
(u
2,3
+u
3,2
)
1
2
(u
1,3
+u
3,1
)
1
2
(u
2,3
+u
3,2
) u
3,3
_
_
_
=
_
_
_
3x
2
2
3x
1
x
2
+x
3

1
2
x
2
3x
1
x
2
+x
3
0
1
2
x
1
1
2
x
2
1
2
x
1
2x
3
_
_
_
compatibility:
11,22
+
22,11
= 2
12,12
6 + 0 = 2 3

A.3. CHAPTER 3 93
22,33
+
33,22
= 2
23,23
0 + 0 = 0

33,11
+
11,33
= 2
31,31
0 + 0 = 0

12,13
+
13,12
23,11
=
11,23
0 + 0 0 = 0

23,21
+
21,23
31,22
=
22,31
0 + 0 0 = 0

31,32
+
32,31
12,33
=
33,12
0 + 0 0 = 0

4. given:
(1)
11
= u
1,1
= a(x
2
1
x
2
+x
3
2
)
(2)
22
= u
2,2
= bx
1
x
2
2
integration displacement eld: (
11
= u
1,1
;
22
= u
2,2
)
(1) :
_

11
dx
1
=
_
u
1,1
dx
1
u
1
=
1
3
ax
3
1
x
2
+ax
1
x
3
2
+f(x
2
)
(2) :
_

22
dx
2
=
_
u
2,2
dx
1
u
2
=
1
3
bx
1
x
3
2
+g(x
1
)
determine integration constants f(x
2
), g(x
1
) by applying boundary conditions:
u
1
(x
1
= 0, x
2
) = u
1
(x
1
, x
2
= 0)
!
= 0
in (1) : 0 +f(x
2
) = 0 +f(x
2
) = 0 f(x
2
) = 0
u
2
(x
1
= 0, x
2
) = u
2
(x
1
, x
2
= 0)
!
= 0
in (2) : 0 +g(x
1
) = 0 +g(x
1
) = 0 g(x
1
) = 0
shear strain
12
12
= (u
1,2
+u
2,1
)
=
1
3
ax
3
1
+ 3ax
1
x
2
2
+
1
3
bx
3
2
check compatibility:
2D :
11,22
+
22,11
!
= 2
12,12
=
12,12
6ax
2
+ 0 = 6ax
2
a = a
A.4. CHAPTER 4 95
A.4 Chapter 4
1.
K
= C
KM

M
(a)
_
_
_
_
_
_
_
_
_
_
10, 8
3, 4
3, 0
0
0
0
_
_
_
_
_
_
_
_
_
_
=
_
_
_
_
_
_
_
_
_
_
C
11
C
12
C
13
0 0 0
C
12
C
22
C
23
0 0 0
C
13
C
23
C
33
0 0 0
0 0 0
0 0 0 ?
0 0 0
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
10 10
4
2 10
4
2 10
4
0
0
0
_
_
_
_
_
_
_
_
_
_
(b)
_
_
_
_
_
_
_
_
_
_
10
2
2
0
0
0
_
_
_
_
_
_
_
_
_
_
=
_
_
_
_
_
_
_
_
_
_
10
4
0, 2 10
4
0, 2 10
4
0 0 0
0, 2 10
4
0, 2 10
4
0
0
0
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
10 10
4
0
0
0
0
0
_
_
_
_
_
_
_
_
_
_
(c)
_
_
_
_
_
_
_
_
_
_
10
10
10
_
_
_
_
_
_
_
_
_
_
=
_
_
_
_
_
_
_
_
_
_
0, 5 10
4
0 0
0 0, 5 10
4
0
0 0 0, 5 10
4
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
20 10
4
20 10
4
20 10
4
_
_
_
_
_
_
_
_
_
_
(a) with (b)/(c)
C =
_
_
_
_
_
_
_
_
_
_
10
4
0, 2 10
4
0, 2 10
4
0 0 0
0, 2 10
4
C
22
C
23
C
24
C
25
C
26
0, 2 10
4
C
32
C
33
C
34
C
35
C
36
0 C
42
C
43
0, 5 10
4
0 0
0 C
52
C
53
0 0, 5 10
4
0
0 C
62
C
63
0 0 0, 5 10
4
_
_
_
_
_
_
_
_
_
_
(a)
10, 8 = 10
4
10 10
4
+ 0, 2 10
4
2 10
4
+ 0, 2 10
4
2 10
4
3, 4 = 0, 2 10
4
10 10
4
+C
22
2 10
4
+C
23
2 10
4
3, 0 = 0, 2 10
4
10 10
4
+C
23
2 10
4
+C
33
2 10
4
10, 8 = 10, 8
1, 4 = C
22
2 10
4
+C
23
2 10
4
1, 0 = C
23
2 10
4
+C
33
2 10
4
3 unknowns /2 equations!
(b) in (a): C
24
= C
34
; C
25
= C
35
; C
26
= C
36
; 0 = C
24
2 10
4
+C
34
2 10
4
; ...
suggestion: material isotropic?
C =
_
_
_
_
_
_
_
_
_
_
2 + 0 0 0
2 + 0 0 0
2 + 0 0 0
0 0 0 2 0 0
0 0 0 0 2 0
0 0 0 0 0 2
_
_
_
_
_
_
_
_
_
_
check:
C
11
= 10
4
= 2 +
C
12
= 0, 2 10
4
=
C
44
= 0, 5 10
4
= 2
_
2 + = 0, 7 10
4
,= 10
4
not isotropic!
It is an orthotropic material! (C
22
and C
33
are still unknown!)
2. (a)
_
11
22
0
12
23
31
_
_
=
_
_
2 + 0 0 0
2 + 0 0 0
2 + 0 0 0
0 0 0 2 0 0
0 0 0 0 2 0
0 0 0 0 0 2
_
_
_
_
0
0
33
0
0
0
_
_
(3 + 2)(T T
0
)
(3 + 2)(T T
0
)
(3 + 2)(T T
0
)
0
0
0
_
_
A.4. CHAPTER 4 97
11
=
33
(3 + 2)(T T
0
) (1)
22
=
33
(3 + 2)(T T
0
) (2)
0 = (2 +)
33
(3 + 2)(T T
0
) (3)
(3) :
33
=
(3 + 2) (T T
0
)
(2 +)
= (T T
0
)
E
(1 2)

1
E
(1)
+
E
(1+)(12)
= (T T
0
)
E
(1 2)

(1 2)(1 +)
E(1 2) +E
= (T T
0
)
(1 +)
(1 )
in (1) :
11
=
(3 + 2)(T T
0
)
2 +
(3 + 2)(T T
0
)
=
E
(1 +)(1 2)
(T T
0
)
(1 +)
(1 )
(T T
0
)
E
(1 2)
= (T T
0
)
_
E E(1 )
(1 2)(1 )
_
= (T T
0
)
E(1 + 2)
(1 2)(1 )
= (T T
0
)
E
1
(2) : =
22
12
=
23
=
31
= 0
(b)
_
11
0
0
12
23
31
_
_
=
_
_
2 + 0 0 0
2 + 0 0 0
2 + 0 0 0
0 0 0 2 0 0
0 0 0 0 2 0
0 0 0 0 0 2
_
_
_
_
0
22
33
12
23
31
_
_
(3 + 2)(T T
0
)
(3 + 2)(T T
0
)
(3 + 2)(T T
0
)
0
0
0
_
_
11
= (
22
+
33
) (3 + 2)(T T
0
) (1)
0 = (2 +)
22
+
33
(3 + 2)(T T
0
) (2)
0 =
22
+ (2 +)
33
(3 + 2)(T T
0
) (3)
(2) (3) :
22
(2 + ) +
33
( 2 ) = 0
22
=
33
in (2) : (2 + +)
22
(3 + 2)(T T
0
) = 0
22
= (T T
0
)
(3 + 2)
2 + 2
= (T T
0
)
E
1 2

= (T T
0
)
E
1 2

(1 +)(1 2)
E
= (T T
0
)(1 +)
=
33
11
= 2(T T
0
)
3 + 2
2 + 2
(T T
0
) (3 + 2)
=
E
(1 +)(1 2)
2(T T
0
)(1 +) (T T
0
)
E
(1 2)
= (T T
0
)
_
E(2 1)
(1 2)
_
= E(T T
0
)
12
=
23
=
13
= 0
12
=
23
=
31
= 0
3.
ij
= 2
ij
+
ij
kk
=
E
(1 +)(1 2)
= G =
E
2(1 +)
A.4. CHAPTER 4 99
=
E
2G
1 =
30 10
6
2 12 10
6
1 = 0.25
=
0.25 30 10
6
(1 + 0.25)(1 2 0.25)
= 12 10
6
11
= 2
11
+
11
(
11
+
22
+
33
) = 228000
22
= 2
22
+
22
(
11
+
22
+
33
) = 276000
33
= 2
33
+
33
(
11
+
22
+
33
) = 156000
12
= 2
12
= 24000
13
= 2
13
= 0
23
= 2
23
= 96000
=
_
_
_
228000 24000 0
24000 276000 96000
0 96000 156000
_
_
_
A.5 Chapter 5
1. (a) biharmonic equation:
x
4
1
+ 2

4
x
2
1
x
2
2
+

4
x
4
2
= 0
x
1
=
F
d
3
x
2
2
(3d 2x
2
)
x
2
1
=

3
x
3
1
=

4
x
4
1
= 0
x
1
=
6F
d
3
x
1
dx
2
+
6F
d
3
x
1
x
2
2
x
2
2
=
6F
d
3
x
1
d +
12F
d
3
x
1
x
2
x
3
2
=
12F
d
3
x
1
x
4
2
= 0
0 + 2 0 + 0 = 0 biharmonic equation is fullled
(b) stresses:
11
=
11
=

2
x
2
2
=
F
d
3
(6dx
1
12x
1
x
2
)
22
=
22
=

2
x
2
1
= 0
12
=
12
=

2
x
1
x
2
=
F
d
3
(6dx
2
6x
2
2
) =
6Fx
2
d
3
(d x
2
)
A.5. CHAPTER 5 101
3F
2d
t
2
n
4
= (
1
0
)
0
d
x
2
n
1
= (
0
1
)
n
3
= (
0
1
)
5d
n
2
= (
1
0
)
x
1
30F
d
30F
d t
1
3F
2d
t
2
boundary 1: x
2
= 0; 0 x
1
5d; n
1
= (
0
1
)
t
n
1
=
11
n
1
+
12
n
2
=
11
0 +
12
(1) t
n
1
=
12
(x
1
, 0) = 0
t
n
2
=
21
n
1
+
2
n
2
=
12
0 +
22
(1) t
n
2
=
22
(x
1
, 0) = 0
boundary 2: x
1
= 5d; 0 x
2
d; n
2
= (
1
0
)
t
n
1
=
11
1 +
12
0 =
11
(5d, x
2
) =
F
d
3
(6d 5d 12 5d x
2
) =
30F
d
2
(d 2x
2
)
t
n
2
=
21
1 +
22
0 =
21
(5d, x
2
) =
6Fy
d
3
(d x
2
)
boundary 3: x
2
= d; 0 x
1
5d; n
3
= (
0
1
)
t
n
1
=
11
0 +
12
1 =
12
(x
1
, d) =
6Fd
d
3
(d d) = 0
t
n
2
=
21
0 +
22
1 = 0
boundary 4: x
1
= 0; 0 x
2
d; n
4
= (
1
0
)
t
n
1
=
11
(1) +
12
0 =
11
(0, x
2
) = 0
t
n
2
=
21
(1) +
22
0 =
21
=
6Fx
2
d
3
(d +x
2
)
(c) deformations and displacements (assumption: plane stress)
(1)
11
=
1
E
(
11
22
) =
6F
Ed
3
x(d 2x
2
) =
u
1
x
1
(2)
22
=
1
E
(
22
11
) =
6F
Ed
3
x
1
(d 2x
2
) =
u
2
x
2
(3)
12
=
2(1 +)
E

12
=
2(1 +)
E
6Fx
2
d
3
(d x
2
) =
u
1
x
2
+
u
2
x
1
integration:
(1)
6F
Ed
3
(d 2x
2
)
x
2
1
2
+C
1
(x
2
) = u
1
(2)
6F
Ed
3
x(dx
2
2x
2
2
2
) +C
2
(x
1
) = u
2
in (3)
u
1
x
2
=
6F
Ed
3
x
2
1
+
C
1
(x
2
)
x
2
;
u
2
x
1
=
6F
Ed
3
(dx
2
2x
2
2
2
) +
C
2
(x
1
)
x
1
2(1 +)
E
6Fx
2
d
3
(d x
2
)
!
=
_
6F
Ed
3
x
2
1
+
C
1
(x
2
)
x
2
_
+
_
6F
Ed
3
_
dx
2
2x
2
2
2
_
+
C
2
(x
1
)
x
1
_
F
Ed
3
6x
2
1
+
C
2
(x
1
)
x
1
. .
f(x
1
):=K
1
=
2 +
E
F
d
3
(6dx
2
6x
2
2
)
C
1
(x
2
)
x
2
. .
f(x
2
):=K
2
C
2
(x
1
)
x
1
= K
2
F
Ed
3
6x
2
1
C
1
(x
2
)
x
2
= K
1
+
2 +
E
F
d
3
(6dx
2
6x
2
2
)
integration:
C
1
(x
2
) =
2 +
E
F
d
3
(6
dx
2
2
2
6
x
3
2
3
) K
1
x
2
+K
2
u
1
C
2
(x
1
) =
F
Ed
3
6
3
x
3
1
+K
1
x
1
+K
3
u
2
K
1
, K
2
, K
3
and K
4
can be determined by evaluating the geometric bound-
ary conditions (not given here). These boundary conditions are necessary to
prevent rigid body displacement.
A.5. CHAPTER 5 103
2. (a) thin disc + loading in x
1
x
2
-plane plane stress can be assumed

33
=
13
=
23
= 0
2h
l
x
1
x
3
x
2
F
P
3
boundary
1
: (x
2
= h)
22
(x
2
= h) = 0 (1)
12
(x
2
= h) = 0 (2)
11
22
12
11
boundary
2
: (x
2
= h)
22
(x
2
= h) = 0 (3)
12
(x
2
= h) = 0 (4)

11
22
12
11
boundary
3
: (x
1
= 0)
P =
_
A
11
(x
1
= 0) dA = t
h
_
x
2
=h
11
(x
1
= 0) dx
2
(5)
F =
_
A
12
(x
1
= 0) dA = t
h
_
x
2
=h
12
(x
1
= 0) dx
2
(6)
11
F
12
P
boundary
4
: (x
1
= l) (clamped boundary)
u
1
(x
1
= l) = 0
u
2
(x
1
= l) = 0
(7) F = t
h
_
h
12
(x
1
= l) dx
2
(8) P = t
h
_
h
11
(x
1
= l)dx
2
(9) F l = t
h
_
h
11
(x
1
= l)x
2
dx
2
11
M = F l
F

12
P
(b) estimate a admissible stress function
loading P normal stress
x
1
x
2
22
,
12
= 0
N
11
loading F bending stress
x
1
x
2
B
11
(x
1
= 0) = 0
B
11
M(x
1
) = F x
1

B
11
grows linearly with x
1
shearing stress
x
1
x
2
12

12
22
= 0
12
does not vary with x
1
(linear Moment M constant shearing force)
A.5. CHAPTER 5 105
Ansatz:
11
=
N
11
+
B
11
= a
1
+a
2
x
1
x
2
=
11
,22
11
=
a
1
2
x
2
2
+
a
2
6
x
1
x
3
2
12
= a
3
x
2
2
+a
4
=
12
,12
12
=
a
3
3
x
1
x
3
2
+a
4
x
1
x
2
=
11
+
12
= b
1
x
2
2
+b
2
x
1
x
3
2
+b
3
x
1
x
2
check:
11
=
,22
= 2b
1
+ 6b
2
x
1
x
2
22
=
,11
= 0
12
=
,12
= 3b
2
x
2
2
b
3
compatibility:
= 0
0 =

4
x
4
1
+ 2

4
x
2
1
x
2
2
+

4
x
4
2
= 0 + 0 + 0 = 0
With boundary conditions (1) - (9):
[ = b
1
x
2
2
+b
2
x
1
x
3
2
+b
3
x
1
x
2
]
Constants b
1
, b
2
, b
3
from boundary conditions:
(1)
22
(x
2
= h) = 0 validate assumption
22
(x
1
, x
2
) = 0
(2)
12
(x
2
= h) = 3b
2
h
2
b
3
b
3
= 3b
2
h
2
(3) see (1)
(4) see (2)
(5) P = t
h
_
x
2
=h
11
(x
1
= 0) dx
2
= t
h
_
h
2b
1
dx
2
= t [2b
1
x
2
]
h
h
= t[2b
1
h + 2b
1
h] = 4tb
1
h
b
1
=
P
4th
(6) F = t
h
_
x
2
=h
12
(x
1
= 0) dx
2
= t
h
_
h
(3b
2
x
2
2
b
3
) dx
2
= t[b
2
x
3
2
b
3
x
2
]
h
h
= t(b
2
h
3
b
3
h (b
2
h
3
+b
3
h)) = 2b
2
th
3
+ 2b
3
th
mit (2) F = 2b
2
th
3
+ 2 (3b
2
h
2
)th = 4b
2
th
3
b
2
=
F
4th
3
b
3
=
3F
4th
(7) F = t
h
_
h
12
(x
1
= l) dx
2
= t
h
_
x
2
=h
(3b
2
x
2
2
b
3
) dx
2
see (6)
(8) P = t
h
_
h
11
(x
1
= l) dx
2
= t
h
_
h
(2b
1
+ 6b
2
lx
2
) dx
2
= [2b
1
x
2
+ 3b
2
lx
2
2
]
h
h
= t[2b
1
h + 3b
2
lh
2
(2b
1
h + 3b
2
lh
2
)] = t4b
1
h see (5)
(9) F l = t
h
_
h
11
(x
1
= l)x
2
dx
2
= t
h
_
h
(2b
1
+ 6b
2
lx
2
) x
2
dx
2
= t[b
1
x
2
2
+ 2b
2
lx
3
2
]
h
h
= t(b
1
h
2
+ 2b
2
lh
3
(b
1
h
2
2b
2
lh
3
)) = t4b
2
lh
3
b
2
=
F
4th
3
see (6)
A.5. CHAPTER 5 107
The stress boundary conditions validate the Ansatz functions and give the
constants b
1
, b
2
, b
3
.
Stress functions:
11
=
P
2th

3
2
F
th
3
x
1
x
2
22
= 0
12
=
3
4
F
th
3
x
2
2
3
4
F
th
=
3
4
F
th
3
(x
2
2
h
2
)
A.6 Chapter 6
1-D-Beam:
z
u
tan
xx
= E
xx
= E u
,x
u = z tan z (x)

xx
= Ez
(x)
xx
= Ezw
(x)
M
y
=
_
A
z
xx
dA = Ew
(x)
_
A
z
2
dA
. .
:=I
y
(moment of inertia)
M
y
= EI
y
w
(x)
Principle of minimum total potential energy:
(w)
beam
=
EI
y
2
l
_
0
(w
(x))
2
dx
l
_
0
q(x)w(x)dA
admissible ansatz has to satisfy all geometrical conditions of the problem
w(x = 0) = 0 (1)
w
(x = 0) = 0 (2)
w(x = l) = w
B
(3)
A.6. CHAPTER 6 109
chosen global polynomial:
w(x) = a
0
+a
1
x +a
2
x
2
+a
3
x
3
+a
4
x
4
with (1) : w(x = 0) = 0 = a
0
w
(x) = a
1
+a
2
2x + 3a
3
x
2
+ 4a
4
x
3
with (2) : w
(x = 0) = 0 = a
1
with (3) : w(x = l) = w
B
= a
2
l
2
+a
3
l
3
+a
4
l
4
a
2
=
w
B
l
2
a
3
l a
4
l
2
w(x) =
w
B
l
2
x
2
+a
3
(x
3
lx
2
) +a
4
(x
4
l
2
x
2
)
The condition for the potential energy to reach an extremum for the exact solutions:
( w) =
( w)
a
3
a
3
+
( w)
a
4
a
4
= 0
( w)
a
3
= 0 and
( w)
a
4
= 0
The explicit form of ( w) is here:
with:
w
(x) = 2x
w
B
l
2
+a
3
(3x
2
2lx) +a
4
(4x
3
2l
2
x)
w
(x) = 2
w
B
l
2
+a
3
(6x 2l) +a
4
(12x
2
2l
2
)
( w) =
EI
2
l
_
0
_
2
w
B
l
2
+a
3
(6x 2l) +a
4
(12x
2
2l
2
)
_
2
dx
q
0
l
_
0
_
w
B
l
2
x
2
+a
3
(x
3
lx
2
) +a
4
(x
4
l
2
x
2
)
_
dx
(1)
( w)
a
3
!
= 0 =
EI
2
2
l
_
0
_
2
w
2
B
l
2
+a
3
(6x 2l) +a
4
(12x
2
2l
2
)
_
(6x 2l)dx
q
0
l
_
0
(x
3
lx
2
)dx
(2)
( w)
a
4
!
= 0 =
EI
2
2
l
_
0
_
2
w
2
B
l
2
+a
3
(6x 2l) +a
4
(12x
2
2l
2
)
_
(12x
2
2l
2
)dx
q
0
l
_
0
(x
4
l
2
x
2
)dx
explicit integration for (1):
EI
l
_
0
_
2
w
B
l
2
(6x 2l) +a
3
(36x
2
24lx + 4l
2
) +a
4
(72x
3
24lx
2
12l
2
x + 4l
3
)
_
dx
= q
0
l
_
0
(x
3
lx
2
)dx
EI
_
6
w
B
l
2
x
2
4
w
B
l
x + 12a
3
x
3
12a
3
x
2
+ 4l
2
x + 18a
4
x
4
8a
4
lx
3
6a
4
l
2
x
2
+ 4a
4
l
3
x
_
l
0
= q
0
_
1
4
x
4
l
3
x
3
_
l
0
EI(6w
B
4w
B
+ 12a
3
l
3
12a
3
l
3
+ 4a
3
l
3
+ 18a
4
l
4
8a
4
l
4
6a
4
l
4
+ 4a
4
l
4
)
= q
0
1
4
l
4
1
3
q
0
l
4
(1) a
3
+ 2la
4
=
q
0
l
48EI

w
B
2l
3
(2) a
3
+
21
10
la
4
=
q
0
l
60EI

w
B
2l
3
A.6. CHAPTER 6 111
a
4
1(2
21
10
) =
q
0
l
EI
_
1
48
+
1
60
_
a
4
=
q
0
24EI
in (1): a
3
=
q
0
l
48EI

w
B
2l
3
2l
q
0
24EI
a
3
=
5
48
q
0
l
EI

w
B
2l
3
the approximative solution is:
w(x) =
w
B
l
3
x
2
_
5
48
q
0
l
EI

w
B
2l
3
_
(x
3
lx
2
) +
q
0
l
24EI
(x
4
l
2
x
2
)
= w
B
x
2
2l
3
(3l x) +
q
0
x
2
48EI
(2x
2
+ 3l
2
5xl)
Principle of minimum complementary energy:
(M, B) =
1
2EI
l
_
x=0
M
2
y
(x)dx w
B
Q(x = l)
where Q(x = l) = M
(x = l)
test function for the bending moment M(x):
q(x) is constant so the test function

M(x) has to be a polynominal of second order
M(x) = a
0
+a
1
x +a
2
x
2
(x) = Q(x) = a
1
+ 2a
2
x
(x) = 2a
2
By comparison of the coecients one obtains:
2a
2
= q
0
s
2
=
q
0
2
statically admissible approximation has to satisfy the equilibrium equation

M
y
(x) = a
0
+a
1
x q
0
x
2
2
test function has to satisfy static boundary conditions:
M(x = l) = 0
0 = a
0
+a
1
l
q
0
l
2
2
a
0
=
q
0
l
2
2
a
1
l

M
y
(x) = a
1
(x l)
q
0
2
(x
2
l
2
)
Q
y
(x) =

M
y
(x) = a
1
q
0
x
Q
y
(x = l) = a
1
q
0
l

(

M
y
) =
1
2EI
l
_
x=0
_
a
1
(x l)
q
0
2
(x
2
l
2
)
_
2
dx w
B
(a
1
q
0
l)
(M
y
) =

(M
y
)
a
1
a
1
!
= 0
A.6. CHAPTER 6 113
1
2EI
l
_
x=0
2
_
a
1
(x l)
q
0
2
(x
2
l
2
)
_
(x l)dx w
B
= 0
a
1
l
_
x=0
(x l)
2
dx
q
0
2
l
_
0
(x
2
l
2
)(x l)dx = EI w
B
a
1
l
_
x=0
(x
2
2xl +l
2
)dx
q
0
2
l
_
0
(x
3
x
2
l l
2
+l
3
)dx = EI w
B
a
1
_
x
3
3
x
2
l +l
2
x
_
l
0
q
0
2
_
x
4
4

x
3
3
l l
2
x
2
2
+l
3
x
_
l
0
= EI w
B
a
1
l
3
3

5
24
q
0
l
4
= EI w
B
a
1
=
3EI
l
3
w
B
+
5
8
q
0
l

M
y
(x) =
3EI
l
3
w
B
(x l) +
5
8
q
0
l(x l)
q
0
2
(x
2
l
2
)
M
y
(x) =
3EI
l
3
w
B
(x l) +
q
0
(l 4x)
8
(x l)

Solid Mechanics

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Solid Mechanics

Uploaded by

Copyright:

Available Formats

Lecture Notes on

a) and the correspond-

Figure 4.1: - diagram of an extension test

(c) Twosymmetry planes

using Falk scheme:

You might also like