Continuum Mechanics

CT 4353 2010/2011
Continuum Mechanics
Introduction to Tensors
PART 1
Dr. ir. Niki Kringos
Assistant Professor
Group of Mechanics of Infrastructure Materials
Faculty of Civil Engineering & Geosciences
TU Delft
[email protected]
Introduction to vectors and tensors
TENSORS?
CT4353 Continuum Mechanics, 2010-2011 [email protected]
1
Tensors are the words of continuum mechanics
Higher order tensor operations

E: dv = x [d i v + f a ] dv = 0
v v
Balance of moment of momentum
Du Wint = d : C : dv + : ( u) ( v) dv
T

v v
Linearized Internal Virtual Work
1 1 1
C 1
v = I I I + II

2D 3 9 V
Part of an Evolution Law
2
0th order tensor: SCALAR
A physical quantity, completely described by 1 (real) number

Value does not change upon reorientation of basis
1st order tensor: VECTOR
Something with length and direction

Value does not change upon reorientation of basis
Values of components do change upon reorientation of basis
Example? temperature, density, time
Example? velocity, force, electric field
3
u?

u

x2
x1
a

x = x u =

2 b

b
u

a x1

4
x2
x1
a

x = x u =

2 b

b
u

1
0

e1 =

e2 =

0 1

e2
1
0

u = a +b

0
1
e1 x1
a
u = a e1 + b e2

x2 w=?
d u = a e1 + b e2
b
v u
v = c e1 + d e2

w=u-v

e2 a - c

w =
b - d

e1 c a x1 w = (a - c) e1 + (b - d ) e2

w = (a - c) e1 (d - b ) e2

5
vi transformation:

vf v f = f (v i )

v1f

v1i e2 f
v = v2 = v1f + v2f + v3f
f
v2f

f
v2i e1

v3

v1f
= v1f e1 + v2f e2 + v3f e3

direct notation basis notation

e2
Vector is base invariant e1 vf = v1f e1 + v2f e2 + v3f e3

Orthonormal right-handed system (ORH)
Rectangular Cartesian System (RCS) {e1, e2 , e3 }
x2
(Cylindrical basis:
e {e , e , e } non-homogeneous )
2 r z
e x
1 1
e
3 s s s
x3 s = e1 + e2 + e2

x1 x2 x2
s 1 s s
Structure changed! s = er + e + ez
r r z
6
vi

vf v f = f (v i )

v1i e2 v1f 1 0 0 v1i

v2f
v f = 0 1 0 v i
v2i e1

2

2

v1f v f 0 0 1 v i
3 3
v2i e3 v f i
1 1 0 0 v1
v3f f i
v3i e2

v2 = 0 1 0 v2

v2f v f 0 0 1 v i
3 3
vi

vf v f = f (v i )

v1i e2 v1f 1 0 0 v1i

v2f
v f = 0 1 0 v i
v2i e1

2

2

v1f v f 0 0 1 v i
3 3
2nd order tensor: TENSOR

7
Magnitude of w:
x2
w= w 2
d w
(d - b ) w (e2 )
(d - b ) cos = =
b w w w

e2 (a - c)
(d - b ) ?
e1 x1 0
(a - c) (b - d ) = (b - d )
c a 1

a - c (d - b ) = w (e2 )

w =b - d w = (a - c) e1 (d - b ) e2

w (e2 ) = w cos

w cos = ?
x2

wn
= w co s

n u
w n
=
n
e2 u
u
e1 x
1
u
w = w cos
u
w (e2 ) = w cos
w
w u = w u cos

w

8
dot product
wu scalar >>
inner >>
Geometrical interpretation PROJECTION

wu
= w cos u
w = cos

w w
u
=
u uw

u

wu

u u

dot product
wu scalar >>
inner >>

wu = w cos w u
= cos

e2 e1 e2 = ? 0
e1 e2 = 1 0 = 0
1
e2 e1 = ?

e1 e1 = ? orthogonality, zero dot product
e

1 projection of a vector on itself is itself
9
u1 dot product
u = u uv scalar >>
2 inner >>
v1
v = v v1
2 u v = u1 u2 v = (u1v1 + u2 v2 )
2
Index notation: Or in index notation:

2 u v = (u1e1 + u2 e2 ) (v1 e1 + v2 e2 )
u = u1e1 + u2 e2 = ui ei
i=1 = (u1e1 ) (v1e1 + v2 e2 )
2
v = v1e1 + v2 e2 = vi ei + (u2 e2 ) (v1e1 + v2 e2 )
2
i=1
= (u1v1 + u2 v2 ) = uivi
i=1
u1 dot product
u = u uv scalar >>
2 inner >>
v1
v = v 2
2 u v = uivi
i=1
Index notation:
2 vu = u v
u = u1e1 + u2 e2 = ui ei
commutative
i=1
2
v = v1e1 + v2 e2 = vi ei ?
v v= 0
i=1
10
TENSOR notations MATRIX notations
Direct notation: u u
1

{u} = u
2

u3
Basis notation: ui ei

ui
{u}T = u = u1 u2 u3
Index notation:
u T = {u}
Right handed system 1

0
0

e2 Orthonormal system

e1 = 0 ; e2 = 1; e3 = 0

Basis in a Cartesian system
0

0

1

Cartesian basis vectors
e u1 ,u2 , u3
1
e Cartesian components
3
Any vector u is uniquely represented by a linear combination of the

basis vectors:
u = u1e1 + u2 e2 + u3 e3

11
e
2 u = u1e1 + u2 e2 + u3 e3
u
3
u1 u2 u = ui ei = ui ei = ui
e i=1
1
e u3
3
Repeated index = implicit summation rule

3
ui pi = ui pi = u1p1 + u2 p2 + u3 p3
i=1
6 3
ui pi ui pi ui p j ui p j

i=1 i=1
e2 e1 e2 = 0 e
2
e2 e1 = 0

e1 e1 = 1
e e
1 1

e
3
ei e j = 0 when i j

e e = e e = e e =0 e e =
1 2 1 3 2 3 i j ei e j = 1 when i = j

e e = e e = e e =1
1 1 2 2 3 3
Kronecker delta ij e e = ij
i j
12
ei e j = 0 when i j
i j ei e j = 1 when i = j

3
ii = ? ii = ii = 11 + 22 + 33 = 3
i =1
3
ij ui = ? ij ui = ij ui = 1j u1 + 2j u2 + 3j u3 = u j
i =1
3 3
u ej = ? u e j = ui ei e j = ui ij = u j
i=1 i=1
ei e j = 0 when i j

3
ij ui = ij ui = 1j u1 + 2j u2 + 3j u3 = u j ij ui = u j
i =1
aibj vk ei (e j ek ) = aibj vk ei jk = aibj vk ei jk

same ?

scalar, vector = aibj v j ei (= aibk vk ei )

or tensor? = bj v jai ei = bj v jai

basis indicial aibj vk ei (e j ek )
aibj vk (ei e j )ek = aibj vk ij ek = a jbj vk ek = a jbj vk
p

13
ei e j = 0 when i j

3
ij ui = ij ui = 1j u1 + 2j u2 + 3j u3 = u j ij ui = u j
i =1
Tqms pk vq km id = Tqmsvq pm id = Tqpsvq id
mk Amk = Akk
mk mk = kk = 3
m =k m k
1 0
direct notation index notation matrix notation
s = {a }T {b}
s = a b s = aibi
= a {b}
ai T = meaningless!!
a = Av
a = Av ai = Aij v j {a } = [A]{v}

Aij T = is possible!!
14
xi x = x i ei = x i
=?
x j
x i ? x 1 x1 x1 x i x1 x 1 x1
= + + + = or or
x j x 1 x 2 x3 x j x1 x 2 x 3
x2 x2 x2 x2 x2 x2
+ + + or or
x 1 x 2 x3 x1 x 2 x3
x3 x3 x3 x3 x3 x 3
+ + or or
x1 x 2 x3 x1 x 2 x 3
xi
= ij
x j
Aij A12 A11 A12

=? =0 =0 =1
Amn A34 A22 A12
Aij
if i = m and j = n =1
Aij
Aij
= im jn
Amn
15
si = Bpis Asp + Aimvm
Determine C ijk such that si = C ijk Ajk
Ajk
Bpis Asp Bkij Ajk
Aimvm Aik vk Ajk vk ij Aj k vk ij
si = Bkij Ajk + Ajk vk ij

= (Bkij + vk ij )Ajk
si = C ijk Ajk C ijk = Bkij + vk ij

si = Bpis Asp + Aimvm
Determine C ijk such that si = C ijk Ajk
C ijk Ajk = Bpis Asp + Aimvm r.h.s linear w.r.t. [A]

Ajk Asp Aim
C ijk = Bpis + vm Aij
A A A = im jn
Amn
C ijk j k = Bpis s p + i m vm
Ajk Asp Aim
C i = B i + iC vijk = Bpis + vm ???
Ajk Ajk Ajk
C ijk = Bkij + ij vk
16
Indicial notation rules

1. No index can appear 3 or more times in a single term
2. An index that occurs twice is a DUMMY index and is
understood to be summed from 1 to 3
3. An index that occurs once is a FREE index and must appear
once in every other term
4. No vector or tensor component may be raised to a power 2
(A )ij
CHECK: Tij = aibj Aij2
ai Aijbi
Tij = ai + bj
Tkj + ai Aijbk 6
s - aibi uk2bk ui pi
i=1
s = aibi ?
& s = Aij v jbi
ai = Aij v j
?
& s = Aij v j Bij w j
bi = Bij w j
?
bi = Bik wk s = Aij v j Bik wk
17
1
4 4 4 4
= 3 =
akbk or akbk 3 akbk
?
akbk 3
1
Aimbm 3
A b Aimbm
Aimbm
= im m = m=1
3
wm wm m=1 wm wm wm wm
wm wm
or ?
m=1
3
f f f f f
= kjor = 3
?
kk k=1 kk kjkk
kk
k=1
Indicial precedence rules

1. differentiation, transposition, inversion etc. go before indices
2. summation symbol must be inserted as closely as possible to
repeated dummy sum indices
3. implied summations go before applying scalar operator
w = b11 + b22 + b33
Aij2 = (A2 )
ij w
w =
Aij 1 = (A 1 ) ? (b11 + b22 + b33 )
ij
bkk
w w w
AijT = (AT ) = + +
ij b11 b22 b33
18
exceptions to tensor rules
1p1 (p1 a ) + 2p2 (p2 a ) + 3p3 (p3 a )

scalar, vector or tensor?
3
k pk (pk a)
k =1
[A]{p}i = i {p}i (no sum on i )
i-th eigenvector i-th eigenvalue
a jk aki aij
Yijk = + +
xi x j xk
Yjim = a?? + a?? + a??

x j xi xm
aim amj a ji
Yjim = + +
x j xi xm
19
U ijkl = Tikjl +Wiljk
U pqim
U1234 = T1324 +W1423
U pqim = Tpiqm +Wpmqi
uskm = Asi Bmik

uijk = A B
il klj
What have we seen so far?
- tensor notation vs matrix notation

- base vectors, ORH-systems
- tensor rules
TENSORS?
- played with indices
- use of Kronecker delta
- some tensor algebra
20
CT 4353 2010/2011
Continuum Mechanics
Introduction to Tensors
PART 2
Assistant Professor
TU Delft
[email protected]
dot product
wu scalar >>
inner >>

wu
= w cos u
w = cos
w w
u uw

=
u
u

wu

u u
1
1. a way to multiply 2 vectors in the same plane to get
Cross product a 3th vector perpendicular to the plane
(Vector product) u w = n

uw 2. magnitude of cross product represents the area
between the 2 vectors
A = uw
uw w area = u w
n
u
A = uw
uw w
h
sin = h = w sin
n w
h
A= u h = u w sin
u u w = u w sin
e e3 u w = n

1 e2
u w = det u1 u2 u3

w1 w2 w3
u w = u w sin n

2
A = uw
uw w
h
sin = h = w sin
n w
h
A= u h = u w sin
u u w = u w sin
?
u w = w u commutative u w = n

uw = wu
u w = (w u) u w = u w sin n


e2
e1 e2 = n
= e3 e2 e1 = - n
= e3
u w = u w sin n

e2 e3 = n
= e1 e3 e2 = - n
= e1
u w = (w u)
e3 e1 = n
= e2 e1 e3 = -n
= e2
e1
e3 ei ei = 0
u w = (u1e1 + u2 e2 + u3 e3 )(w1e1 + w2 e2 + w3 e3 )
u w = (u2 w3 u3 w2 ) e1 (u1w3 u3 w1 ) e2 + (u1w2 u2 w1 ) e3
e e3
1 e2
u w = det u1 u2 u3

w1 w2 w3

3
e1 e1 = 0 e1 e2 = e3 e1 e3 = -e2
3
e2 e1 = -e3 e2 e2 = 0 e2 e3 = e1 ei e j = ijk ek
k=1
e3 e1 = e2 e3 e2 = -e1 e3 e3 = 0
ijk 1 for even permutations 2 2

Permutation symbol -1 1 +1 1
-1 for odd permutations 3 3
Alternating >>
Levi-Civita >> 0 for repeated index odd even
3
e1 e1 = 11k ek = 111e1 + 112 e2 + 113 e3 = 0
k=1
3
e2 e1 = 21k ek = 211e1 + 212 e2 + 213 e3 = e3
k=1
e e e3 1 1
1 2 3

u v = det u1 u2 u3
+1 2 -1 2 ei e j = ijk ek
3 3
k=1
v1 v2 v3
even odd
u2 u3 u1 u3 u1 u2
u v = e1 det v e2 det
v3
+ e3 det
v1 v3

v1 v2
2

= (u2 v3 - u3 v2 ) e1 (u1v3 - u3 v1 ) e2 + (u1v2 - u2 v1 ) e3
= u2 v3 e1 + u3 v1e2 + u1v2 e3 u3 v2 e1 u1v3 e2 u2 v1e3
3 3 3 3
= ijk uiv j ek = ijk uiv j ek = ijk uiv j ek
i=1 j=1 k=1 i,j,k=1 u v = ijk uiv j ek
4
uw
v w
h
n
Volume?
h = v cos
A = u w = u w sin
u
V = Ah = u w v cos triple scalar product

V = (u w) v
= u w sin v cos box product
u w = n


uw
u3
u1 u2 3
(u w) v = det w1 w2 w3 = ijk uiw j vk

v1 v2 v3 i,j,k=1
v w
e e3
1 e2
uw = wu u w = det u1 u2 u3

u w = (w u) u w1 w2 w3

V = (u w) v V = (u w) v = (w v) u = (v u) w
triple scalar product = v (u w) = u (w v) = w (v u)
or box product
5
cross product:
ei e j = ijk ek e e e3
1 2
u v = (ui ei )(v j e j ) u v = det u1 u2 u3

= uiv j (ei e j ) (u v)i = ijk u j vk v1 v2 v3

= uiv j ijk ek (u v)k = uiv j ijk u v = ijk u j vk ei
triple scalar product:

? u u
(ei e j ) em = ijk km = ijm 1 2 u3

(u v) w = det v1 v2 v3
(u v) w = (uiv j ijk ek ) (wm em )
w1 w2 w3

= uiv j wm ijk (ek em )
= uiv j wm ijk km = uiv j wk ijk
e e e3 1 1
1 2 3

u v = det u1 u2 u3
+1 2 -1 2 ei e j = ijk ek
3 3
k=1
v1 v2 v3
even odd
ijk = jki = kij = jik = ikj = kji

3
The identity: ijn kln = ik jl il jk
n =1
imk pin = ?? ?? ?? ??
i mk p i n mknp 1234 move cyclically forward from summed index
1 ? 2 ? 1 ? 2?
13 24 1423 = mn kp mp kn
imk pin = mn kp mp kn
6
3 1 1
ijn kln = ik jl il jk 13 24 1423
3
+1 2
3
-1 2
n =1
even odd
ijk nkp = ?? ?? ?? ?? ijpn

= ip jn in jp 13 24 1423
kln lij = ?? ?? ?? ?? nkij

= ni kj nj ki
ijn kjn = ?? ?? ?? ?? nink ijkj ijn ijn = 6
= nn ik nk in = ik jj ij jk
= 3ik ik = 2ik = 3ik ik = 2ik ijn kjn = 2ik
PROOF THAT: (a b)c = (a c)b - (b c)a (left triple cross product)
(a b)c (u v) = ijk u j vk
i i
a b = d
[d c ]i = ijkd jck
d j = (a b)j
= ijk (a b)j ck (a b)j = jmnambn
= ijk jmnambnck
= (km in kn im )ambnck kimn
ijk jmn = ?? ?? ?? ??
= km inambnck kn imambnck
= 13 24 1423
= akbick aibkck = (a c)bi - (b c)ai = km in kn im
7
u, v 1st order tensors (vectors)
dot product 0th order tensors (scalars)

3
u v = v u = viui
i=1
cross product 3 1st order tensors (vectors)
u v = - (v u) = ijk uiv j ek
i,j,k=1
2nd order tensors (tensors)

A = Aij ei e j
2nd order tensors (tensors) A = Aij ei e j
a 3 by 3 matrix
a linear transformation (mapping) from vectors to vectors
ordered set of numbers that change upon a change of basis
dyads and dyadics
8
tensor product
direct >>
uv matrix >>
outer >>
dyadic >>
u1 u v u v u1v3
1 1 1 2
u v = u2 v1 v2
v3 = u2 v1 u2 v2 u2 v3

u
3 u3 v1 u3 v2 u3 v3

A
11 A12 A13

A = u v = A21 A22 A23

A A A
31 32 33
1 1 0 0

A A13
11 A12 0
e1 e1 = 1 0 0 = 0 0 0

A = u v = A21 A22 A23 0 0 0 0

A A33
31 A32
1 0 1 0

e1 e2 = 0 0 1 0 = 0 0 0

0 0 0 0
3
A= Aij ei e j 1 0 0 1
i,j=1

0
= Aij ei e j e1 e3 = 0 0 1 = 0 0 0

0 0 0 0

9
x2
v Linear transformation
e2
e1 x1 A13 v1
A11 A12
e3
u = A v = A21 A22 A23 v2
x3
u A A33 v3
31 A32
-1
v = A u
3 3
u1 = A11v1 + A12 v2 + A13 v3 = A1j v j ui = Aij v j = Aij v j
j=1 j=1
3
u2 = A21v1 + A22 v2 + A23 v3 = A2j v j
3 3
j=1
3 u = ui ei = Aij v j ei
u3 = A31v1 + A32 v2 + A33 v3 = A3j v j i=1 i,j=1
j=1
u, v 1st order tensors (vectors)
dot product 0th order tensors (scalars)

3
u v = v u = viui
i=1
cross product 3 1st order tensors (vectors)

u v = - (v u) = ijk uiv j ek
i,j,k=1
dyadic product 2nd order tensors (tensors)

T
u v = ( v u) = A
10
A, B 2nd order tensors (tensors)
dot product ???
cross product
???
dyadic product ???
A, B 3rd order tensors
dot product ???
cross product
???
dyadic product ???
11
A, B 4th order tensors
dot product ???
cross product
???
dyadic product ???

3
A A13
11 A12

B
11 B12

B13

A= Aij ei e j
i,j=1
A = A21 A22 A23 B = B21 B22 B23 3

A
31
A32 A33

B31 B32

B33

B= Bkl ek el
k,l=1
3
dot product A B = Aij Bkl ei e j ek el
i,j,k,l=1
3 3
u v = {u}T {v} = uivi = Aij Bkl ei jk el
i=1 i,j,k,l=1
3
= Aij Bkl jk ei el
i,j,k,l=1
3
= Aik Bkl ei el
i,k,l=1
12
3
A A13
11 A12

B
11 B12

B13

A= Aij ei e j
i,j=1
A = A21 A22 A23 B = B21 B22 B23 3

A
31
A32 A33

B31 B32

B33

B= Bkl ek el
k,l=1
3
dot product A B = Aij Bkl ei e j ek el
i,j,k,l=1
3
The dot product of two 2nd = Aij Bkl ei jk el
order tensors gives i,j,k,l=1
3
= Aij Bkl jk ei el
a second order tensor i,j,k,l=1
3
= Aik Bkl ei el
i,k,l=1

3
A A13
11 A12

B
11 B12

B13

A= Aij ei e j
i,j=1
A = A21 A22 A23 B = B21 B22 B23 3

A
31
A32 A33

B31 B32

B33

B= Bkl ek el
k,l=1
dot product Linear transformation

A13 v1
A11 A12

The dot product of two 2nd u = A v = A21 A22 A23 v2

order tensors gives A33 v3
A31 A32
a second order tensor A = ? B
A = C B A=C:B
13
3
A A13
11 A12

B
11 B12

B13

A= Aij ei e j
i,j=1
A = A21 A22 A23 B = B21 B22 B23 3

A
31
A32 A33

B31 B32

B33

B= Bkl ek el
k,l=1
double dot product dyadic product
3 3
A:B = Aij Bij A B = Aij Bkl ei e j ek el
i,j=1 i,j,k,l=1
The double dot product of The dyadic product of two

two 2nd order tensors gives 2nd order tensors gives
a scalar a 4th order tensor
TENSOR NOTATIONS ( 3 engineers 5 opinions-arrangement)
A B?
A = Aij ei e j
AB = Aij Bkl ei e j ek el = Aij Bkl
or
A = Aij ei e j A B = Aij Bkl ei e j ek el
u = A v A B
or or AB = Aik Bkl ei el = Aik Bkl
AB A B = Aik Bkl ei el
u = Av
14
(a b + c d ) v = a (b v) + c (d v)
(a b + c d ) v = (a b) v + (c d ) v
= a (b v) + c (d v)
a b = ab
(a b + c d ) v = (ab + cd ) v
= (ab) v + (cd ) v
= a (b v) + c (d v)
(a b)
? ?
(a b)= a b (a b)=a b
= a b = a b
a b +a p
?
a b + a p=a (b + p )
(a b + a p )
? ?
= a b + a p = a b + a p
15
uv vu
uv = vu u AT v = v A u
But if v = u then
u v = ui ei v j e j uv = vu ui AijTv j = v j Ajiui
= uiv j ei e j u v = u u ui AijTv j v j Ajiui = 0

= uiv j ij = ui ei u j e j uiv j (AijT Aji ) = 0
= viui = uiu j ei e j
AijT = Aji
= vu = viu j ei e j
= vi ei u j e j Aij = Aji symmetric!
= vu
Aij = Aji skew-
symmetric!
T
(A + B) = AT + BT
T
AT = (Aij ei e j )
AT = (AT ) ei e j
T
ij
AT = Aij (ei e j )
AT = Aji ei e j AT = Aij (e j ei )
16
A v = v AT (a b) v = a (b v)
A v = (Aij ei e j ) v (a b) v = (ai ei bj e j ) vk ek
= Aij ei (e j v) = aibj vk (ei e j ) ek
= Aij v j ei = aibj vk ei (e j ek )
= Ajivi e j = aibj vk ei jk
= AijTvi e j = aibj v j ei
= vi AijT e j = ai eibj v j
= v AT = a (b v)

Tensor operation Indicial notation Direct notation Other
Linear transformation of a
vi = Fij x j v = F x (v = Fx)
vector x into a new vector v
Composition of two tensors C = A B
C ij = Aik Bkj (C = AB)
A and B
Inner product of two tensors
s = Aij Bij s =A:B (s = AiB)
A and B
Dot product of a vector w s = wi Fij x j s = w F x (s = wiFx)

into a linear transformation
Vector dyadic multiplication Tij = aibj T=ab (T = ab)
Tensor dyadic multiplication Yijkl = Aij Bkl Y=AB (T = AB)

17
Time fordeformation measures
TENSORS?
18
CT 4353 2010/2011
Continuum Mechanics
Deformation measures
Assistant Professor
TU Delft
[email protected]
x2 u = u1e1 + u2 e2 + u3 e3
u
e 3
2 e x1 u2 u = ui ei = ui ei
u1
1 i=1
e3
x3 u3

x2 x2 x2

e e e
2 e x1 2 e x1 2 e x1
1 1 1
e3 e3 e3
x3 t0 x3 t1 x3 t3

1
X2 x2

E2 e
2
E1 X1 e x
1 1
E3 e
3
X3 x3

t=0 t = t1
reference configuration current configuration
UNDEFORMED DEFORMED
U =U1 E1 +U 2 E2 +U3 E3 u = u1e1 + u2 e2 + u3 e3

U u
x2
X2
V v
E2 e
2
E1 X1 e x
1 1
E3 e
3
X3 x3

t=0 t = t1
2
U =U1 E1 +U 2 E2 +U3 E3 u = u1e1 + u2 e2 + u3 e3

U u
x2
X2
V v
E2 e
2
E1 X1 e x
1 1
E3 e
t=0 3 t = t1
X3 x3

Uu Xx x = A X
x = ?X
Vv xX X = A-1 x

U =U1 E1 +U 2 E2 +U3 E3 u = u1e1 + u2 e2 + u3 e3

U u
x2
X2
V v
E2 e
2
E1 X1 e x
1 1
E3 e
t=0 3 t = t1
X3 x3

Deformation change in length
3
U =U1 E1 +U 2 E2 +U3 E3 u = u1e1 + u2 e2 + u3 e3

dX dx
U u
x2
X2
V v
E2 e
2
E1 X1 e x
1 1
E3 e
t=0 3 t = t1
X3 x3

dX dx dx = F dX
Deformation change in length
dx dX d X = F-1 d x

(X, t ) x
dx = dX = dX
x = (X, t ) X X

3 3
dX xi
d x = d x i ei d xi = dX j
i=1 j=1 X j
dx 3
xi
dx = d X j ei
i,j=1 X j
dX j = E j dX
3
X2 , x2 xi

E2 , e2
dx = (E j d X )ei
i,j=1 X j

E1, e1 X1, x1
E3 , e3 (b v) a = (a b) v
X3 , x3 3
xi
dx = (ei E j ) d X
i,j=1 X j
4
(X, t ) x
dx = dX = dX
x = (X, t ) X X

3 3
dX xi
d x = d x i ei d xi = dX j
i=1 j=1 X j
dx 3
xi
3
x dx = d X j ei
Xi (ei E j ) i,j=1 X j
dX j = E j dX
i,j=1 j
x1
x1 x1 3
xi

X1 X2 X3
dx = (E j d X )ei
i,j=1 X j

x x2 x 2
2 =F (b v) a = (a b) v
X X3
1 X2 3
x xi
3
X
x3 x3 dx = (ei E j ) d X
i,j=1 X j
1 X2 X3
Deformation
Example: measures
Deformation gradient
X2 , x2
(-1,1) (1,1) (1,-1) (-1,-1)
(5,5) (7,5)
1
x1 = (18 + 4X1 + 6X2 )
4
(2,2) (4,2) 1
(-1,1) (1,1) x 2 = (14 + 6X2 )
4
X1, x1

(5,5) (7,5) (4,2) (2,2)
(-1,-1) (1,-1)
5
Deformation
Example: measures
1
x1 = (18 + 4X1 + 6X2 )
4
X2 , x2 1
x 2 = (14 + 6X2 )
(5,5) (7,5) 4
x1 x1

X1 X2
F=
x 2 x2
(2,2) (4,2)

(-1,1) (1,1)
X1 X2
X1, x1

x1 1 x1 1
(-1,-1) (1,-1) = (4) = 1 = (6 ) = 1 . 5
X1 4 X2 4
1 1. 5 1 -1
F =
F-1 =

x2
=0
x2 1
= (6 ) = 1. 5
0 1. 5
0 2 X1 X2 4
3
Deformation
Example: measures
X2 , x2

deformed
X1, x1
undeformed
1 1. 5 1 -1

F=
F =
-1

0 1. 5
0 2
3
6
Deformation
Example: measures
x = (X, t ) 1 0
E1 = E2 =
0 1
X2 , x2
1 1 . 5
1
1
[E2 ] =
(E1 ) = F E 1 = 0
0 1 . 5

0

[E1 ] 1 1 . 5
0 1 . 5
(E2 ) = F E 2 = =
0 1 . 5 1 1 . 5

E2

X1, x1

E1

1 1. 5 1 -1
F =
F-1 =

0 1. 5
0 2
3
Deformation
Example: measures
x = (X, t ) 1 0
E1 = E2 =
0 1
X2 , x2 X = 1 (x, t )
1 1. 5 1 1
=
e2 (E1 ) = F E 1 = 0 0

0 1. 5

e 1 1 . 5 0 1. 5
1 (E2 ) = F E 2 = =
0 1 . 5 1 1. 5

-1 [e2 ]
1 0
-1 [e1 ] X1, x1 e1 = e2 =

0 1

1 - 1 1 1

1 (e1 ) = F e1 = 2 0 = 0
-1

0 3

1 1. 5 1 -1 1 - 1 0 - 1
F =
F =
-1

-1
1 (e2 ) = F e2 = =
0 1. 5 0 2 2 1 2
3
0 3 3
7
dx = F dX dv = ? dV

dX

dv
dx
dV
X2 , x2 X2 , x2

E2 , e2 E2 , e2

E1, e1 X1, x1 E1, e1 X1, x1
E3 , e3 E3 , e3
X3 , x3 X3 , x3

Length Volume
dV = dX1dX2dX3 dv = dx1dx 2dx3
dX3
dx3
dX1
dx1
dX2 dv = ? dV dx 2
in reference configuration in current configuration
u w1 v1 d X1 d x1 (d X1 )
1
V = (u w) v = det u2 w2 v2 d X = d X2 d x = d x 2 (d X2 )

u3 w3 v3 d X3 d x (d X )
3 3
8
x1 x1 x1 x1 x1
x1 d X1 + d X2 + d X3
X X2 X3
X1 X2 X3
d X 1
1
x x2 x2
x x x
dx = 2
d X2 = 2 d X1 + 2 d X2 + 2 d X2
X1 X2 X3
X1 X2 X3
x
d X3
3 x3 x3 x3 x x
X X2 X3
d X1 + 3 d X2 + 3 d X3
1 X1

X2 X3

x1 x x1

d X1 1 d X2 d X3

d x1
X1 X2 X3

3 x x
d x = d x 2
= d x i ei dx = 2
d X1 + 2
d X2 +

x 2
d X2

i=1 X1 X2 X3

d x3

x3 d X x3 d X
x3
d X3
X 1
X2
2

1 X3

x
x1 x1 x1 1 d X + x1 d X + x1 d X
X 1 2 3

X3 X2 X3
X1 X2 d X 1
1
x x2 x2
x x x
dx = 2
d X2 = 2 d X1 + 2 d X2 + 2 d X2
X1 X2
X3 X1 X2 X3
x d X
3 x3 x3
3
x3 x3 x3
X
d X1 + d X2 + d X3
1 X2 X3
X1

X2 X3

x1 x x1

d X1 1 d X2 d X3

u w1 v1 X1 X2 X3

1 x
x
x
v = det u2 w2 v2 dx = 2
d X1 + 2
d X2 + 2
d X2

X1 X2 X
u3 w3 v3

3

x3 d X x3 d X x3
d X3
X 1 2
1 X2
X3

9
x1
x1
x1

X1
X2 X3
u w v

x

x x 1 1 1
dx = 2 d X1 +
2 d X2 + 2 d X3 v = det u w v
X1 X2 X3 2 2 2

x3 x3 x u
3 w 3 v3

3

X1
X2

X3
=u+w+v

x1 x1 x1
dv = ? dV
X1 X2 X3

x x 2 x 2
d v = d x1d x 2 d x3 = det 2 d X1 d X2 d X3 d v = det F d V
X X X

1 2 3
x
3 x3 x3 dv = JdV
X X X3
1 2
current
det F = J configuration

dv = JdV
reference dm = 0 dV = dv = J dV
configuration
isochoric
J = 1?
volume preserving
10
dx = F dX dv = det F dV da = ?dA

dX
da
dv
dx dA
dV
X2 , x2 X2 , x2 X2 , x2

E2 , e2 E2 , e2 E2 , e2

E1, e1 X1, x1 E1, e1 X1, x1 E1, e1 X1, x1
E3 , e3 E3 , e3 E3 , e3
X3 , x3 X3 , x3 X3 , x3

Length Volume Area
dv = JdV da dx = JdA dX
da
dx dx = F dX da F dX = JdA dX

dX
u A v = v AT u T
dX F da = JdA dX

dA d X FT d a J d A d X = 0

T
d a = da n

d A = dA N uv = vu F da dX JdA dX = 0

dv = da dx dV = dA dX
(F T d a J d A ) d X = 0

FT d a J d A = 0
T

da = JF dA
FT FT d a = J FT d A
Nansons formula

AT A T = I T
da = JF dA

11
dx = F dX dv = det F dV da n
= J FT d A N

dX
da
dv
dx dA
dV
X2 , x2 X2 , x2 X2 , x2

E2 , e2 E2 , e2 E2 , e2

E1, e1 X1, x1 E1, e1 X1, x1 E1, e1 X1, x1
E3 , e3 E3 , e3 E3 , e3
X3 , x3 X3 , x3 X3 , x3

Length Volume Area
F important measure of material deformation

U R
F = RU

dX dx
F

F = vR
R

V

J = det F = det U = det v Strains

12
Measure of change of length

2 2
dX dx dX

dx = dx dx - dX dX dx = F dX

X2 , x2 = F dX F dX - dX dX A v = v AT

E2 , e2

E1, e1 X1, x1 = d X FT F d X - d X d X
E3 , e3
X3 , x3

= d X (FT F - I) d X = d X (C - I) d X

(Right) Cauchy-Green deformation tensor C = FT F

1
Lagrangian-Green strain tensor E = (FT F - I)
2
Measure of change of length

2 2
dX dx dX

dx = dx dx - dX dX d X = F-1 d x

T
X2 , x2 = d x d x - F -1 d x F -1 d x A v = vA

E2 , e2 T -1
= dx dx - dx F F dx
E1, e1 X1, x1
E3 , e3
X3 , x3 = d x (I - FT F -1 ) d x = d x (I - b 1 ) d x

(Left) Cauchy-Green deformation tensor b = F FT

1
Euler-Almansi strain tensor = (I - b -1 )
2
13
x (X + u) u
x=X+u F= F= =I+
X X X

u T
u u
C = I + I +
C = FT F
X
X X

X2 , x2 x

T
u u
= I T + I +
E2 , e2 X X

E1, e1 X1, x1 T T
E3 , e3 u
+ + u
u u
= I +

X3 , x3
X X X X

1
1 u u
T T
u u
E = (C - I) = + + finite strains

2 2 X X X X

1 u u
T
= + small strains
2 X X

Reference configuration Current configuration
C = FT F b = F FT

(R) Cauchy-Green deformation tensor (L) Cauchy-Green deformation tensor
1 1
E = (C - I) ? = (I - b -1 )

2
2
Lagrangian-Green strain tensor Euler-Almansi strain tensor
14
1 1
= (I - FT F -1 ) E = (FT F - I)
2 2
1 1
= (FT FT ) (I - FT F -1 ) (F F -1 ) = (FT F-T ) (FT F - I) (F 1 F)
2
2
1 1
= FT FT (I - FT F -1 ) F F -1 = FT F -T (FT F - I) F 1 F

2

2

1 1
= FT (FT F - I) F -1 = FT (I - F -T F 1 ) F

2
2

= FT E F -1 = FT F

= * (E) = *-1 ()

Reference configuration Current configuration
C = FT F b = F FT

(R) Cauchy-Green deformation tensor (L) Cauchy-Green deformation tensor
1 1
E = (C - I) = (I - b -1 )

2
2
Lagrangian-Green strain tensor Euler-Almansi strain tensor
* () = FT () F -1 push-forward = FT E F -1

*-1 () = FT () F pull-back E = FT F

15
Time to move on to.
- Stress tensors, stress rates

- Mechanical balance laws
- Virtual work, Clausius Planck inequality
- Linearization, directional derivative
- Hyperelasticity
TENSORS?
- Plasticity
- Elasto-visco-plasticity
- .
16
CT 4353 2010/2011
Continuum Mechanics
Higher order tensors
Assistant Professor
TU Delft
[email protected]

Tensor operation Indicial notation Direct notation
Linear transformation of a
vi = Fij x j v = F x (v = Fx)
vector x into a new vector v
Composition of two tensors C = A B
C ij = Aik Bkj (C = AB)
A and B
Inner product of two tensors
s = Aij Bij s =A:B (s = A : B)
A and B
Dot product of a vector w s = wi Fij x j s = w F x (s = w Fx)

into a linear transformation
Vector dyadic multiplication Tij = aibj T = a b (T = a b)
Tensor dyadic multiplication Yijkl = Aij Bkl Y = A B (T = A B)

1
u A v = ui Aij v j
remember:
= ui AjiTv j
T
u Av = v A u = v j AjiTui
= vi AijTu j
= v AT u
u Av = A : u v u A v = ui Aij v j
= u v :A = uiv j Aij = Aij uiv j
= u v :A = A :u v

E: dv = x [d i v + f a ] dv = 0
v v
Balance of moment of momentum (Notes: Fundamentals)
Du Wint = d : C : dv + : ( u) ( v) dv
T

v v
Linearized Internal Virtual Work (Notes: Fundamentals)
1 1 1
C 1
v = I I I + II

2D 3 9 V
Part of an Evolution Law (Notes: Large Strains)
2
3
4th order tensors!!
C= C ijkl ei e j ek el
M=C:L i,j,k,l=1
2nd order tensors 2nd order tensors

M = Hx 3
H= H ijk ei e j ek
u = H :Z i,j,k=1
vector vector
3
v = Ax A= Aij ei e j
i,j=1

3
2nd order tensors
A= Aij ei e j
i,j=1
a a b a b a b c d c d c d
1 11 1 2 1 3 1 1 1 2 1 3

a b = a2 b1 b2 b3 = a2b1 a2b2 a2b3 ; c d = c2d1 c2d2 c2d3

a3 a3b1 a3b2 a3b3 c3d1 c3 d2 c3d3

a b : c d = a (b : c) d a b c d = a (b c) d
or? = (b c) a d
= (a b) : (c d)
3
3
2nd order tensors
A= Aij ei e j
i,j=1
1 11 1 2 1 3 1 1 1 2 1 3


a3 a b a b a b c d c d c d
3 1 3 2 3 3 3 1 3 2 3 3
a b : c d = a1b1c1d1 + a1b2c1d2 + ... + a3b3c3d3 = anbmcndm

3
2nd order tensors
A= Aij ei e j
i,j=1
1 11 1 2 1 3 1 1 1 2 1 3


a3 a3b1 a3b2 a3b3 c3d1 c3 d2 c3d3

a b : c d = a1b1c1d1 + a1b2c1d2 + ... + a3b3c3d3 = anbmcndm
= ancnbmdm = (a c)(b d) a b : c d = (a c)(b d)
ei e j : ek el = (ei ek )(e j el ) = ik jl
4
2nd order tensors A : u v = u A v
A : u v = Aij ei e j : uk ek vl el
= Aij uk vl ei e j : ek el
= Aij uk vl ik jl ei e j : ek el = (ei ek )(e j el ) = ik jl

= ui Aij v j
= uA v
3rd order tensors 3

H= H ijk ei e j ek
i,j,k=1
3
ei e j = ijk ek = ijk ek E : A = ijk ei e j ek : Alm el em
k=1
= ijk Alm ei e j ek : el em
permutation symbol
= ijk Alm ei (e j ek ) : (el em )
3
E= ijk ei e j ek = ijk Alm ei jl km
i,j,k=1
= ijk Ajk ei
permutation tensor
vector
5
3rd order tensors 3

H= H ijk ei e j ek
i,j,k=1
3 E : u v = ijk ei e j ek : ul el vm em
ei e j = ijk ek = ijk ek = ijk ul vm ei e j ek : el em
k=1
permutation symbol = ijk ul vm ei jl km

= ijk u j vk ei
3
E= ijk ei e j ek = kij uiv j ek
i,j,k=1
= ijk ui v j ek = u v
permutation tensor
E : u v = uv

3
4th order tensors
C= C ijkl ei e j ek el
i,j,k,l=1
AB=C (u v) (w x) = u v w x = C
M = C : L = C ijkl ei e j ek el : Lmn em en
= C ijkl Lmn ei e j (ek el ) : (em en )
= C ijkl Lmn ei e j km ln
= C ijkl Lkl ei e j
2nd order tensor
6
4th order tensors

A : B C = (A : B) C = C (A : B)
A : B C = Aij ei e j : Bkl ek el C mn em en
= Aij BklC mn (ei e j ) : (ek el ) em en
= Aij BklC mn ik jl em en
= Aij BijC mn em en = C mn em en Aij Bij
= (A : B) C = C (A : B) 2nd order tensor
4th order tensors

A B : C D = (B : C)(A D) = (A D)(B : C)
A B : C D = Aij ei e j Bkl ek el : C mn em en Dpq e p eq
= Aij BklC mn Dpq ei e j (ek el ) : (em en ) e p eq
= Aij BklC mn Dpq kl mn ei e j e p eq
= Aij BklC kl Dpq ei e j e p eq
= BklC kl Aij Dpq ei e j e p eq
= (B : C)(A D) = (A D)(B : C)
4th order tensor
7
nth order tensors 3

C i1 ,i2 ,...,in ei1 ei2 ... ein
i1 ,i2 ,...,in =1
Unit tensors
2nd order I = ij ei e j = e j e j
4th order I = I = ik jl ei e j ek el = ei e j ei e j
Iu = u A-1 A = I v = A u
A v = A-1 A u
-1
I:A=A A-1 A u = u A-1 v = I u
A-1 v = u

3

Gradient grad = ek = ek =
k=1 x k xk

0 = E1 + E2 + E3 = e1 + e2 + e3
X1 X2 X3 x1 x2 x3
dX dx
U u
X2 x2
V v
E2 e2
E1 X1 e1 x1
E3 e3
t=0 x3 t = t1
X3
8
Gradient ui
grad u = ei e j
x j
ui ui Ui
u = ei e j 0 u = ei E j 0 U = Ei E j
x j X j X j
3
x
dx = Xi (ei E j ) d X = F d X
i,j=1 j
dX x1
x1 x1
X1 X2 X3

dx x x2 x 2
F= 2
X X2 X3
1
x x3 x3
3
X X2 X3
1
Gradient ui
u = ei e j
x j
ui ui Ui
x j X j X j
3
x=X+u x
dx = Xi (ei E j ) d X = F d X
i,j=1 j
u
X F = 0 x = 0 (X + u) = 0 X + 0 u
X2 , x2 x Xi ui
E2 , e2
= Ei E j + ei E j
X j X j
E1, e1 X1, x1
E3 , e3 = I + 0 u I = ij ei e j = e j e j
X3 , x3
9
Gradient ui
u = ei e j
x j
ui ui Ui
x j X j X j
x=X+u Tjm
T = e j em ei
xi
u
Tjm
X 0T = e j em Ei
X2 , x2 x Xi
E2 , e2
E1, e1 X1, x1 u j vm
E3 , e3 (u v) = e j em ei
X3 , x3 xi

3
Divergence uk uk u u u
div u = u = = = 1+ 2+ 3
k=1 x k xk x1 x 2 x3
u1
u1 u1 u = tr ( u) = u : I
x1 x2 x3

ui u u2 u2
u = ei e j = 2
x j x x2 x3
1
u u3 u3
3
x x2 x3
1 Trace
10
3
Divergence uk uk u u u
div u = u = = = 1+ 2+ 3
k=1 x k xk x1 x 2 x3
u1
u1 u1 u = tr ( u) = u : I
x1 x2 x3
ui
ui u u2 u2 u : I = ei e j : ek ek
u = ei e j = 2 x
x j x x2 x3 j
1 u
u u3 u3 = i (ei e j ) : (ek ek )
3 x j
x x2 x3
1 ui u
= ik jk = k
x j xk
3
Tmk Tmk
div T = em = em T = tr ( T) = T : I
k,m=1 x k xk
Forward and backward gradient operations

ui u j
grad u = ei e j grad u = ei e j
x j xi
u1 u1 u1 u1 u2 u3

x1 x2 x3 x1 x1 x1

u u2 u2 u u2 u3
grad u = 2 grad u = 1
x x2 x3 x x2 x 2
1 2
u u3 u3 u u2 u3
3 1
x x2 x3 x x3 x3
1 3
= u = u
Backward gradient Forward gradient
T
u = ( u)
11

Tjm Tjm
grad T = e j em ei = T grad T = ei e j em = T
xi xi
T = T
uk uk
div u = = u div u = = u
xk xk
u = u
Tmk Tkm
div T = em = T div T = em = T
xk xk
T = TT

Tmk Tkm
div T = em = T div T = em = T
xk xk
div T = T : I div T = I : T
Tij Tjm
T : I = ei e j ek : el el I : T = el el : ei e j em
xk xi
Tij Tjm
= ei (e j ek ) : (el el ) = (el el ) : (ei e j ) em
xk xi
Tij Til Tjm Tlm

= jl kl ei = ei = li lj em = em
xk xl xi xl
12
div (v) = div v + v grad scalar
(v )i
div (v) =
xi
(vi )
=
xi
vi
= + vi
xi xi
= div v + v grad
div A = div A + A grad vector
(A)ij
div (A) = ei
x j
(Aij )
= ei
x j
Aij
= ei + Aij ei
x j x j
( )
= div A + A
13
div (AT u) = A : u + div A u scalar
div (AT u) = AT div (u) + div (AT ) u

Aji
AT div (u) = AijT ei e j
uk div (AT ) u = e j uk ek
x k div u = u : I xi
u Aji
= AijT ei e j k ek el : e p e p = uk jk
xl xi
uk
= AijT ei (e j ek ) el : e p e p Aji
xl = uj
T uk xi
= Aij jk (ei el ) : (e p e p )
xl Aij
uk u j = ui = div A u
= Aji jk ip lp = Ajp x j
xl x p
u j ui
= Aji = Aij = A : u
xi x j
Tensor decompositions B = sym B + skw B
1
symmetric tensor A = AT Aij = Aji sym B =
2
(B + BT )
1
skew-symmetric tensor A = AT Aij = Aji skw B = (B - BT )
2
s s12 s13
B=S+W 11 S : W =0
S = s12 s22 s23

s13 s23 s33
S = ST -w23 = 1 W
0 2
W = W T w12 w13 = w13
1
W = -w12

0 w23
-w12 = ijkWij ek
2
-w13 -w23 0

14
1
2
(B + BT )
1
2
spatial velocity gradient tensor l = F F1
d = 21 (l + l T ) d = d
T
rate of deformation tensor
w = 21 (l l T ) w = w spin tensor
T
1
2
(B + BT )
1
2
1
Du E = (0 v) 0 u + (0 u) 0 v
T T
2
= sym (0 u) 0 v
T

15
Tensor decompositions A = iso A + dev A
Deviatoric strain tensor:
B 'deviatoric' tr B = 0
B 'isotropic' B = I
Isotropic strain tensor:
Tensor decompositions A = iso A + dev A
1 1
iso A = (tr A) I dev A = A - (tr A) I
3 3
1 ( )
tr A = I : A
spherical = A - (I : A) I
3
A = I + P : A 1
= A - I (I : A)
3
1
1 =A- II:A
=
3
(I : A) 3 I:A=A
1
1 =I:A- II:A
P=I- II 3
3 1
= I - I I : A
3
16
A
(Characteristic) invariants Ik 11 A12 A13

A = A21 A22 A23
I1 = A11 + A22 + A33
A A A
31 32 33
A11 A12 A11 A13 A22 A23

I2 = det
+ det + det

A21 A22 A31 A33 A32 A33
A
11 A12 A13

I3 = det A21 A22 A23

A A A
31 32 33
Spectral analysis A p = p
eigenvector eigenvalue
= principle direction = principle value
An
= n

T
A=A
An
- n
=0
(A - I) n = 0
n = 0 det A - I = 0

trivial solution 3 2
i 1i + 2i 3 = 0
eigenpairs
characteristic equation
(i ,pi )
17
For a symmetric matrix it is ALWAYS possible to find a
triad of orthonormal eigenvectors! A p = p
A = Aij Ei E j A* = ?? EIGENBASIS
I = E1 E2 E3 Q = p1 p2 p3 Principle direction cosine matrix
E2 p2
Qmn = (pn ) pi i p j = ij
E1 p1 m
E3 p3
1 0 0
Aij = Ei i Ai E j
= 0 2 0
A*mn = pm i Ai pn pi i p j = ij
& 0 0 3
A*mn = n mn Ai pn = n pn (no sum on n )
A* = n mn pm pn = n (pn pn ) [A] = [Q][][Q]T

Positive definite v Bv >0 for all v

1
sym B =
2
(B + BT ) v sym B v > 0 for all v
all I k of the symB > 0
all k of the symB > 0
Orthogonal tensors
rotations and reflection
T T Q-1 = QT
QQ = Q Q = I
u = Q v u=v
18
Tensor decompositions Orthogonal tensors
w
Q

Q w
u
w u = w u cos Q u
( )( )
w u = Q w Q u QT Q = I QT = Q-1
w u = (w Q ) (Q u)
T
( ) ()
2
det QT Q = det Q det Q = 1
w u = w (Q Q) u
T
Gausss/Divergence theorem
u n d s = div u d V
s V
A n d s = div A d V
s V
n d s = grad d V
s V
19

E: dv = x [d i v + f a ] dv = 0
v v
Balance of moment of momentum
Du Wint = d : C : dv + : ( u) ( v) dv
T

v v
Linearized Internal Virtual Work
1 1 1
C 1
v = I I I + II

2D 3 9 V
Part of an Evolution Law
Time forstress measures
TENSORS?
20
CT 4353 2010/2011
Continuum Mechanics
Hyper-Elasticity
Assistant Professor
TU Delft
[email protected]
Clausius-Planck inequality
The internal dissipation should always be non-negative:
D = P : F 0
P : F = Rate of int. mech. work = Helmholtz free energy
First Piola-
Kirchhoff D = 0 Mechanical process is reversible
stress tensor elastic materials
( )
= F (X) strain energy or stored energy
D > 0 Mechanical process is irreversible

dissipative materials internal variables
(
= F (X), 1, 2,... n )
Clausius-Planck inequality: dissipative materials
D = P : F 0
n

(
= F (X) ) Strain energy

= P
F
: F : i 0
i=1 i
(
= F (X), 1, 2,... n )
Evolution laws:
Helmholtz free energy
(
i = F (X), 1, 2,... n )
Coleman-Noll procedure :
elastic range inelastic range
n

i = 0 P= i 0 : i 0
F i=1 i

Clausius-Planck inequality: elastic materials
D=0 Mechanical process is reversible
(F)

D = P : F = P :F=0
F
(F) HYPERELASTIC
P=
F materials
P : F = Work per unit ref. volume
t
Strain energy function corresponds to
= P : F dt
work done by stresses from initial to
t0
final configuration

Hyperelasticity
Objectivity implies: (F) = (QF) = (F

)
Since Q is arbitrary: Q = R T ( ) (
(F) = R T F = R T RU = (U) )
Objectivity condition
(F) (C) C = U2 C = FT F
=
F
: = :F = : C = 2 F : F
F t F C t C

: F = 2F

:F S = F 1 P S = F 1 2 F =2
F C C C
T
P= = 2F P = J FT = 2J 1 F F
F C C
Isotropic hyperelastic response
Restrictions on strain energy function: isotropy = Q dX
dX
dX
dx = F =F
Q dX = F dX
x = (X) = F QT
F
dx
dX (F) = (F
) = FQT ( )
t = t1
t=0
(X
)
Isotropic
x=
= Q dX
dX

dX (C) = (C
) = F(
T F ) (
= QFTFQT )
= (QCQT )
t = 0
Set Q = R : (C) = (RCR T ) = (b)
(F) = (QF) Objectivity! (L) Cauchy-Green deformation tensor
Representation theorem for invariants
Isotropic scalar valued tensor function,

invariant under a rotation
(
(C) = QCQT )
Fulfills:
Can be expressed in terms of the

principal invariants of its arguments
(C) = (I1 (C), I2 (C), I3 (C)) = (I1 (b), I2 (b), I3 (b))
I1 (C) = C : I = 12 + 22 + 23 = b : I = I1 (b)
I2 (C) = 21 I12 C : C = 1222 + 2223 + 1223 = 21 I12 b : b = I2 (b)

I3 (C) = det C = J2 = 12 22 23 = det b = I3 (b)
i i = 1,...3 principal stretches

Strain energy functions
Strain energy function for hyperelastic material: (C)
T
S=2 = 2J 1 F F
C C
Restrictions on strain energy function: isotropy (

(C) = QCQT )
Restrictions on strain energy function: incompressibility J =1
1 T = (F) p (J 1)
S = pC + 2 = p I + F
C F

Incompressibility S = p C 1
+2 + I1 I - 2 C
I I
I
& isotropy 1 2 2
-1
= p I + 2 b -2 b
I1 I2
Strain energy functions
1 I2 1
(I1, I2, I3 ) = c (I1 3) + (I3 1) + (1 c) 3 + (I3 1)
Blatz-Ko 2 2 I3

material = = ; c 0, 1 ; with & : Lam material constants
2 1 2
(I3=1 incompressibility) : Poisson ratio
1
Neo-Hookean c = 1 (I1, I2, I3 ) =
2
(I1 3) +

(
I3 1 )

(I3=J2)
(C) 3
Ii I1 (I : C)
S=2 = = =I
C C I C C C
i =1 i
I2 1 (C : C)

= 2 2 (I : C) I
C C

S = 2 + I1 I C + I3
C
1 = I1I C

I1 I2 I2 I3 (
det C )
= J2 C 1
= s1 I + s2 C + s3 C 1 C

General stress relation for isotropic hyperelastic materials
( )
S = 2 I1 + I1 I2 I I2 C + I3 I3 C 1

= s1 I + s2 C + s3 C 1
Blatz-Ko Neo-Hookean
I
s1 c + (1 c) 1

I3

(1 c)
s2
I3 0
I
s3
c I3 + (1 c) I3
2 I
3
I3
= J 1 F S FT

Comparison Strain Energy Functions
2
( I1, I 2 , I3 ) = ( I1 3) +
2
(
I3 1

)
1 +1
2 +1
( )
1
S1 = 1 4 12
1

-6000
Kirchhoff Stress
14000 Analytical
-5000 CAPA Old N.H.
12000 CAPA Old N.H.
Kirchhoff stress
10000 Analytical -4000
8000 -3000
6000 -2000
tension
4000 compression
-1000
2000
0 0
0.00 0.10 0.20 0.30 0.40 0.50 0 -5 -10 -15
Euler Almansi strain Euler Almansi Strain

2
( ) ( ( )) ( )
1 2
( 1, 2 , 3 ) = 1 + 2 + 3 3 + log 1 2 3
2 2
log 1 2 3
2 2

J = 1 1 log ( J )

-6000
30000
25000 CAPA New N.H. -5000
Kirchhoff Stress
Kirchhoff stress
20000 Analytical -4000

15000 -3000 Analytical
10000 CAPA New N.H.
-2000
5000
-1000
0
0.00 0.10 0.20 0.30 0.40 0.50 0
Euler Almansi strain 0 -10 -20 -30 -40
Euler Almansi Strain

2
( I1, I 2 , I3 ) =
2
(1 )
I 3 +

I 3 1

( )
-25000
150000 Analytical New N.H.
Kirchhoff Stress
Analytical New N.H Analytical Old N.H.
-20000
Kirchhoff Stress
Analytical old N.H

100000
-15000
50000 tension -10000
-5000
0 compression
0.00 0.20 0.40 0.60 0
Euler Almansi strain 0 -500 -1000 -1500

( ) ( ( )) ( )
1 2
( 1, 2 , 3 ) = 12 + 22 + 32 3 + log 1 2 3 log 1 2 3
2 2
Comparison stress tensors versus
(a) Euler Almansi Strain (b) stretch in tension
30000 30000
Stress
Second Piola Second Piola
25000 25000
Piola Kirchhoff
Piola Kirchhoff
Cauchy
20000 Cauchy 20000
Stress
15000 15000
10000 10000
5000 5000
0 0
0 0.1 0.2 0.3 0.4 0.5 0.0 1.0 2.0 3.0 4.0
Stretch lamda 1
(a) (b)

Comparison stress response versus
(a) Euler Almansi Strain (b) stretch in compression
-100000 -100000
-90000 Piola Kirchhoff -90000
Piola Kirchhoff
-80000 Cauchy -80000
Stress
Cauchy
-70000 -70000
Stress
-60000 -60000
-50000 -50000
-40000 -40000
-30000 -30000
-20000 -20000
-10000 -10000
0 0
0 -500 -1000 -1500 0 0.2 0.4 0.6 0.8 1
Euler Almansi Strain Stretch
(a) (b)

CT 4353 2009/2010
Continuum Mechanics
Time derivatives
Assistant Professor
TU Delft
[email protected]
Time derivatives
t=0 t = t1
U u
x2
X2
v
V
E2 e2

E1 X1 e1 x1

E3 e3

X3 x3


Time derivatives
t=0 t = t1
U u
x2
X2
v
V
E2 e2

E1 X1 e1 x1

E3 e3

X3 x3

(x, t ) (x, t ) = D (x,t ) (x, t ) (x, t ) d x

= +
Dt t x dt

(X, t ) (X, t ) = D (X ,t ) (X, t ) (X, t ) d X
= +
Dt t X dt

Time derivatives
Local time derivative of
(x, t ) = D (x,t ) (x, t ) (x, t ) d x

= +
Dt t x dt

Material time derivative of Convective rate of change of

Changing position of a particle
D (x, t ) (x,t ) (x, t ) * [X, t ] x = * [X,t ]

(x, t ) = = +
Dt t x t
x t X=
*
1
[x,t ]

(x, t )
= + v
t t X= 1[x,t ]
*
x

Time derivatives
Local time derivative of
(x, t ) = D (x,t ) (x, t ) (x, t ) d x

= +
Dt t x dt

Material time derivative of Convective rate of change of

Changing position of a particle
D (X, t ) (X, t )
(X, t ) = =
Dt t
Material time derivative of Local time derivative of

Time derivatives
(x, t ) = D (x,t ) (x, t ) (x, t ) d x

= +
Dt t x dt

Material time derivative of a scalar valued tensor function
Material time derivative of a vector valued tensor function

D v(x, t )
v (x, t ) = = v (x, t ) v(x, t ) d x
+
Dt t x dt
v(x, t )
= + v v
t
Spatial acceleration field velocity gradient
D (X, t ) V(X, t ) V(X, t ) d X V(X, t )

V
V(X,t ) = = + =
Dt t X dt t

Time derivatives
vq = v p + d v v (x, t )
dv = dx = v dx = l dx
vq x

q v v X x X
dx l = = = = F F-1
d v x X x X x

p
X2 , x2 vp Spatial velocity gradient

E ,e
2 2
E1 , e1 X1, x1 D F (X,t ) d x d x
E3 , e3 F = = = X dt

Dt dt X
X3 , x3
v
=
X
v x
= =l F
F = l F l = F F-1 x X

gradient
Material velocity
Reynolds transport theorem
I (t) = (x, t ) d v Material time derivative?

I (t) = D (x,t ) d v D
v

Dt v Dt
(x, t ) d v v (t )
v
Transform to reference configuration! x = (X, t ) d v = J (X
,t ) d V

I (t) = D ( (X, t ), t ) J (X, t ) d V = D dV
Dt
D t ( (X

, t ), t ) J (X

, t )
V V
I (t) = ( (X, t ), t ) J (X, t ) + ( (X, t ), t ) J (X, t ) d V

V
J (X, t )

= ( (X, t ), t ) + ( (X, t ), t ) J (X, t ) d V x = (X, t )
V
J (X, t )
d v = J (X, t ) d V
J (X , t )
= (x, t ) + (x, t ) d v
J (X, t )
v
Reynolds transport theorem
J (X, t ) J
I (t) = (x, t ) + (x, t ) d v J = d et F J = :F
J (X , t )
F
v

d et A
J div v

=
(x, t ) + (x, t )
dv A J -T
J = J F
v F
= (x, t ) + (x, t ) d i v v d v
= d et A

(A

-T
)

v
J = J F-T : F F = l F
= + v + (x, t ) d i v v d v
t
v J = J F-T : l F = J F-T FT : l
d i v ( v)

= d v + d iv ( v) d v J = J (F-T FT ) : l = J I : grad v
v
t v

D = J tr (grad v) = J d iv v
Dt t
d v = d v + v n d s J

v v s
Example: Velocity Gradient
1 0 x1
v1 = = X2 = x 2
xi t
Fij = F = 0 1 0
X j
v2 = 0
0 0 1
v3 = 0 0 0

vi
x1 (X, t ) = X1 + (t )X2 Lij = L = 0 0 0
x j

X2 , x2 x 2 (X, t ) = X2 0 0 0

x3 (X, t ) = X3
L = D + W

0 0 0
1 2 0 2

D = 0 0 W = 0 0
2 2

X1, x1 0 0 0 0 0 0


Rate of Deformation
D d x2 = D [ d x d x ]
Dt Dt
dX D
dx = F dX = F d X F d X
dx Dt

X2 , x2 T = D d X F T F d X
A v = vA Dt
E2 , e2

E1 , e1 X1, x1 D
E3 , e3 = d X
Dt
(F T
F) d X

X3 , x3
Dd X D T
Dt
=0 = dX (
Dt
F F) d X

C = FT F (R) Cauchy-Green

deformation tensor
C = F T F + FT F = d X (F T F + FT F ) d X

1
E = (FT F - I) Lagrangian-Green = d X C d X
2 strain tensor
1
E = (F T F + FT F ) = d X 2 E d X

2
Rate of Deformation
D d x2 = 2 d X E d X

Dt
dX
d X = F -1 d x = 2 F -1 d x E F -1 d x
dx
= 2 d x (F -T E F -1 ) d x
X2 , x2

E2 , e2

E1 , e1 X1, x1 = 2 dx d dx

E3 , e3
X3 , x3

C = FT F

d = F -T E F -1
C = F T F + FT F

1
E = (FT F - I) rate of deformation tensor
2
1
E = (F T F + FT F )
2
Material time derivative & Directional derivative
Directional derivative of F, with respect to v:
d
Dv F(X, t) =
(
x + v
)

=
d x
+
v
=
v
d X d X X X
0 0
D F(X, t) x x v
= F (X, t) = = =
Dt t X X t X
D F (X, t)
= F (X) = Dv F (X)
Dt

Rate of deformation tensor:
symmetric part of spatial velocity gradient
Lagrangian E = 21 C = 21 (F T F + FT F ) Material velocity Spatial velocity

material strain gradient gradient
T
( )
rate
E = 21 C = 21 l F F + FTl F
F = l F
( )
E = 21 C = 21 FTlT F + FTl F = 21 FT lT + l F ( )
Rate of deformation d = FT E F 1
(
1 = 1 FT FT lT + l FF 1 = 1 lT + l
d = FT EF 2 2 ) ( )
l =d+w
symmetric part: d = 21 (l + l T ) dT = d 0
3 2

w = 3 1
anti-symmetric part: w = 21 (l l T ) w T = w
0

0
2 1

Lie time derivative
t=0 t=t1
G (X, t) =
*
1
[g (x, t)] 1 g = g (x, t)
(X, t)
G L (g) = * G
(X, t)

X3, x3
E3, e3
E2, e2 X2, x2
E1, e1
X1, x1
1. Pull back (X, t) = D G (X, t)

G v
2. Compute material derivative

L (g) = * G (X, t) = D G (X, t)
3. Push forward * v

Continuum Mechanics
Stress tensors
Tom Scarpas
The traction vector
q
n q
q a

t = lim
a
a 0
2 / 11 Continuum Mechanics
The traction vector
q
n tn
q a

t = lim Newtons 3rd law :
a
a 0
tn tn = tn
Cauchy stress tensor
e3 e3
23 t2e2
da1 t
da2 22
n 21
t1e1
t2e 2 e2 e2
e2
da e1
e1
da3
t3 - e 3
t2e2 = 21 e1 + 22 e2 + 2 3 e3
3
t je j = ij ei
i=1
Cauchy stress tensor
e3 Equilibrium:
3
t d a + t je j d a j = 0
j=1
da1 t 3 daj
da2 t = t je j daj ? d a j = (n e j ) d a
n da
t1e1 j=1
t2e 2 3
e2 3
= t je j (n e j ) t je j = ij ei
i=1
j=1
e1 da
da3 3 3
t3 - e 3 = ij e j n ei ( )
j=1 i=1
(e j n) ei = (ei e j ) n
Cauchy 3 3
stress tensor = ij ei e j ( ) n =n

j=1 i=1
First Piola-Kirchhoff stress tensor
n dq
da
dq

t = lim
da
da 0
t= n
Cauchy stress
n dq dq
N
da dA
dq dq

t = lim T = lim
da dA
da 0 dA0
t= n T=?N
Cauchy stress
n dq dq dq = T dA = t da = n da
N
da dA Area transformation:
d a = J FT dA
da = n da ; dA = N dA
dq dq d a = n da = J FT N dA

t = lim T = lim
da dA
da 0 dA0 d q = T dA = n da
t= n T=
T =P
?NN = J FT N dA
First T = J FT N

Cauchy stress Piola-Kirchhoff
= P N
stress
Second Piola-Kirchhoff stress tensor
dQ
n dq
N
dA da
d Q =
*
1
[d q ]
= F 1 d q
(
d Q = F 1 J FT dA ) dq = n da
= J F 1 FT dA n da = J FT N dA

= S dA
d q = n da = J FT dA
Continuum Mechanics
Stress tensors
Tom Scarpas
Projections
n e2 = ?cos
X2 e2
B

n n
e2
e1

A e1 C X1 da
e3 da2
AC = AB cos
( )
da j = n e j da
(
= AB n e2 )
Continuum Mechanics
Balance Laws
Tom Scarpas
Balance of Momentum
Momentum Balance of Momentum Principle
m = v dv Dm
=F
v Dt
t X3 D

Dt v
v dv = t da + f dv
a v
n

?
X2
f
X1
Balance of Momentum
=0
D D D Dv

Dt v
v dv =
(
Dt v dv

) =

v
v
( ) ( )
Dt dv + Dt dv

v
t X3 =
Dv
dv
v
Dt
n

X2
f
X1
Balance of Momentum
m = v dv Dm
=F
v Dt
t X3 D

Dt v
v dv = t da + f dv
a v
n
?
X2
f
X1
Balance of Momentum
t da = n da = div dv
a a v
t X3 divergence theorem
X2
f
X1
Balance of Momentum
Dm
m = v dv =F
Dt
v
D
t X3

Dt v
v dv = t da + f dv
a v
n
(div + f a) dv = 0
v
X2
f Residual
Local spatial equilibrium i v++f
forces :r =ddi v f a =
a00
X1
Balance of Moment of Momentum
Moment of Momentum Balance of M Principle
M = x v dv ( ) DM
=M
v Dt
Moment M = x t da + x f dv
a v
X3
t
X2
f
X1
M = x v dv ( ) DM
=M
v Dt
Moment M = x t da + x f dv
a v
X3
t
D M D
Dt
=
Dt
( )
x v dv

?
v

= x tda + x f dv
X2
a v
f
X1

D M D
= ( )
x v dv =

Dt Dt
v
D D
x v dv + x D v dv
=
Dt
( )
x v dv = Dt (
Dt
)

v v

D Dv
= v v dv + xv (
Dt
)
dv + x (
Dt
)
dv
v
Dv
= x dv
Dt
v
M = x ( v ) dv DM
=M
v Dt
Moment M = x tda + x fdv

a v
X3
t x a dv = x t da + x f dv
v a v
X2 Explore nature of stress tensor

f
X1
x a dv = x t da + x f dv
v a v
1 if cyclic permutation
E = -1 if reverse cyclic permutation
Prove : E : x t = x t
0 otherwise
( ) (
Definition: ei e j ek : (el em ) = ei e j ek em el

)
3 3
E:xt= Eijk ei e j ek : x l tm el em
i,j,k=1 l,m=1
3
= Eijk x l tm (ei e j ek ) em el
i,j,k,l,m=1
3
E: x t = ( )
Eijk t m ei e j ek em x l el

i, j,k,l,m =1
( )
3
ei e j e k e m e l =
= ( )
Eijk t m ( ek em ) e j el x l ei
i, j,k,l,m =1
( e m e k ) ( ei e j ) e l = 3
( e m e k ) ( e l e j ) ei
= Eijk t m km jl x l ei
i, j,k,l,m =1
3
= ( Eijk t k x j ) ei
i, j,k =1
3
E: x t = ( )

i, j,k,l,m =1
3
= ( )
i, j,k,l,m =1
3
i, j,k,l,m =1
3
i, j,k =1
= ( E t ) x = Wt x Wt
definition
3
E: x t = ( )

i, j,k,l,m =1
3
= ( )
i, j,k,l,m =1
3
i, j,k,l,m =1
3
i, j,k =1
= ( E t ) x = Wt x = x t Wt
definition
finally
x t da = E : x t da
a a
finally
3 3

x t da = E : x t da = Eijk x j tk da ei
i=1
j,k=1
a a a
3 3

= Eijk x j kl nl da ei
i=1
j,k,l=1
a
i = 1...3
3 3

j,k,l Eijk x j kl nl da = (
x Eijk x j kl dv = )
a =1 v j,k,l=1 l
3
vi 3 3
div v = v : I = kl
i=1 x i j,k=1 Eijk kj dv + Eijk x j x dv
j,k,l=1 l
v v i=1...3
finally :
3 3 similar to x t

x t da =
Eijk x j kl nl da ei
a i=1 a j,k,l=1
= E : Tdv + (x div ) dv
v v
x a dv = x div dv + E : dv + x f dv
v v v v
E : T dv + x [d i v + f a ] dv = 0
v v
equilibrium
E : T dv = 0 E : T = 0
v
Cauchy stress tensor symmetry
3
T
E: = Ei j k k j ei = (E1 2 33 2 + E1 3 22 3 ) e1 +
i, j,k =1
(E2 3 11 3 + E2 1 33 1 ) e2 + (E3 1 22 1 + E3 2 11 2 ) e3
= (3 2 2 3 ) e1 + (1 3 3 1 ) e2 + (2 1 1 2 ) e3
=0

32 23

1 3 3 1 = 0

2 1 1 2
Continuum Mechanics
Balance Laws
Tom Scarpas
x a dv = x t da + x f dv
v a v
0 t t
3 2

skew tensor Wt = t3 0 t1 : Wt x = t x

t2 t1 0

alternating tensor E = e1 e2 e3 + e2 e3 e1 + e3 e1 e2
e3 e2 e1 e1 e3 e2 e2 e1 e3
( )
x t = Wt x = E t x = E : x t
2
1
Plasticity
- an introduction
Tom Scarpas
Section of Structural Mechanics
Delft University of Technology
The Netherlands
1/37
aspects of soil experimental response
0.6 4
[m/m]
0.5
kPa
[m/m 3
[Mp
0.4
* 100
strain
[MPa]
stress
strain
0.3 2
radial
0.2
stress
1
radial
0.1
0 0
0 0.1 0.2 0.3
axial strain [m/m]
2/37
concepts of elasto-plastic response
E
p
v
= e + p e

3/37
1
H F WCHF = ( Y ) P
2
Y
C D
WOHFA = P
1
WOCFA = P ( Y ) P
A B 2

O
p e = P + q
1
with: q = ( Y ) ; = P
2
4/37
notion of yield surface
v
pc1
pc2
pc3
v
v
ultimate
response
yield
stress
pc
5/37
triaxial response simulation
I1
I1 = x + y + z m =

3
1
b 2 2
g c h b
J2 = x m + y m + z m g
2
+
2
+ +
2 2 2
( )
f I1, J2 ,cos3, =0
xy yz xz
3 3 J3
cos 3 = 3/ 2
3 J2
6/37
f (I1,J2, cos3, ) = 0
triaxial response
simulation
J2
= constant
I1
Plasticity theory fundamentals
J2
f ( ,) 0
I1
J2
f ( ,) > 0
stress reduction f ( ,) = 0
I1
10/37
2 f ( ,) = 0
y

1

Perfectly plastic
2 f ( ,) = 0

Current yield
surface

y
1 0

Initial yield
surface
Isotropic hardening 11/37
f ( , ) = 0
Initial yield Loading

surface
kinematic hardening
Current yield
f ( ,) = 0
surface
mixed hardening
12/37
more general formulations:
(
f I1, J2 ,cos3,,,T = 0 )

1 2
2 1
3 2 1 > 2 > 3
1 > 2 > 3 3
13/37
Plasticity response surfaces
space diagonal
von Mises 1 = 2 = 3
1 3
Tresca
2
3
deviatoric
plane

1 2
14/37
Plasticity response surfaces
3 Drucker-Prager
Mohr-Coulomb
3
c cot
2
0
1

1 2
15/37
the Desai response surface
J2 I1 + R
n
I1 + R
q
(1 cos3 ) = 0
1/2
f = 2 +
pa pa pa

2
elastic
limit
16/37
the Desai response J2
surface
hardening phase R I1
J2 I1 +R
n q
I1 + R 1/2
f = 2 + (1 cos3) = 0
pa pa pa

J2
I1
17/37
J2 I1 +R
n
I1 + R
q
1/2
f = 2 + (1 cos3) = 0
pa pa pa

18/37
2
hardening response
e
T, , Wp j T = 25 C displ. rate = 8.5 mm/s
0.08
0.06
. 0.67 Wp
01 0.04

= analytical vs..
1 + 250 Wp 0.02 experimental
0
0 0.05 0.1 0.15
Wp
19/37
softening response:
- degradation
T = 25 OC
0.1
e j
2
0.075 = 1max
2 32
k Wpf
8.5 mm/sec
0.05
e j
o 0.32
k T = 25 C = 0.24 + 2
0.025
1 mm/sec
5 mm/sec
0
0 0.25 0.5 0.75 1
Wpf
20/37
softening phase
J2 I1+R
n
I1+R
q
f = 2 + =0 J2
pa pa pa

=0
= (T ,Wpf ) I1
z p
Wpf = d pf
21/37
J2
f ( , ) = 0

f f
df = d + d = 0

I1
f
f ( , ) f

f 2
p = 2
f
0 1
1
22/37
softening response:
- cracking

s
t
s t
2
+ q e 2 2
s + t j e2
= ft p
T, , cr j
23/37
3D cracking:
- orthogonal stress/strain compatibility
xy = yx
y xz = zx
...... .......
x
...... .......
24/37
3D cracking simulation:
- 3 orthogonal cracking planes
25/37
The internal dissipation should always be non-negative:

0
D = :
: = Work per unit ref. volume = Helmholtz free energy
D = 0 Mechanical process is reversible

elastic materials
= ( ) strain energy or stored energy
D > 0 Mechanical process is irreversible

dissipative materials internal variables
= ( , 1 , 2 ,... n )
26/37
Clausius-Planck inequality: dissipative materials
0
D = :
= ( ) Strain energy
n

=

:

: i 0
i =1 i
= ( , 1 , 2 ,... n )
Evolution laws:
i = ( , 1 , 2 ,... n )
n

i = 0 = i 0 : i 0
i=1 i
27/37
Elementary model of Elastoplastic response
E
y

e p
= (e, )

e
=
e t t
p
= +
e t e t t

= + p 0
e
e
28/37

+ 0
=

e
p
e
Coleman-Noll procedure

=
e

p + q 0 q =

29/37
Maximum plastic dissipation

principle
{ ( ) }
,q # F , q 0
Given : ( p , ) {, q } ( )
# F , q 0
{, q }
(
max p + q )
30/37
(
minimize p + q )
subject to F (,q ) 0
31/37
Notions of Constrained Minimization
minimize ( )
g , q (
g , q ) + f ( , q ) = 0

subject to f ( , q ) 0
(
g , q ) + f ( , q ) = 0
q q
( )
f , q = 0
Lagrange multiplier 0
( )
Since 0 and f , q 0 for ( )
f , q = 0 :
if f 0 f ( , q ) = 0 plastic deformation
if f = 0 f ( , q ) < 0 elastic deformation
Kuhn-Tucker conditions
32/37
(
minimize p + q )
subject to F (,q ) 0
F (,q )
p + =0

F (,q )
+ =0
q
F (,q ) = 0
0
Kuhn-Tucker conditions
33/37
1
H F WCHF = ( Y ) P
2
Y
C D
WOHFA = P
1
WOCFA = P ( Y ) P
A B 2

O
p e = P + q
1
with: q = ( Y ) ; = P
2
34/37
E
y
= p + q 0
e
e p
= ; q =
e

1
q q = ( Y ) ; = P
Y 2
F (, q ) = = Y q 0

O
p e
35/37
F (, q ) = = Y q 0
= Y + k
internal variable

k q =k =

Y
= + q
p
O e

=?
36/37
F (, q ) = = Y q 0
= Y + k

k q =k =

Y
= + q
p
O e

1 2 1 2
= E e + k
2 2
37/37
2
1 Plasticity
- an introduction
Tom Scarpas
The Netherlands
38/37
E
Viscoelasticity
E1 1
E2 2 - an introduction
EN N
Tom Scarpas
The Netherlands
1/20
concepts of visco-elastic response
Maxwell model

EM

v ( t ) = e

t
e ( t ) = v ( t ) = v dt
EM t
( t ) = e + v ( t ) = E ( v ) = E e
2/20
Maxwell model

EM
v el

el
v ( t ) =

tu t

el ( t ) = v ( t u ) = v dt
EM t
( t ) = e + v ( t ) ( t ) = v ( t u ) unloaded
3/20
Elementary model of Viscoelastic response
EM
1
= e ( ) = E M 2e
2
Clausius-Planck inequality Coleman-Noll procedure
e
= = = E M e
e t e
v v 0
= +
e t e t
Definition

= + v 0
e e v ( t ) =

4/20
E
Kelvin model

E
v
v ( t ) =

v ( t ) = ( t ) = E t
= v ( t ) + ( t )
= + E = E

t
+ = = 1 exp
E E

5/20
E
Kelvin model

tu
v ( t ) =

d ( t ) = ( t ) = E tu t
= d ( t ) + ( t )
0= + E = E
t tu

+ =0 = t u exp
unloaded
6/20
E
Kelvin model

tu
v ( t ) =

d ( t ) = ( t ) = E tu t
= d ( t ) + ( t )
0= + E = E
t tu

+ =0 = t u exp
unloaded
7/20
E = E = + v
1
( )
= + d = E 2 + 0
2

v = v ( )
= =
t
(
= E )
Definition
= ( + v ) ( E ) v
= ( E + v ) ( E ) v ( t ) =

= ( v ) 0
8/20
Standard solid model
E ( t ) = E
EM

..
9/20
E ( t ) = E
EM

= EM
M ( t ) = v ( t ) = E M ( t ) v ( t )
v ( t ) =
1

( ( t ) v ( t ) )
10/20
E
v ( t ) =
1

( ( t ) v ( t ) )
EM

Laplace transform
t ( t ) t ( t )
v ( t ) = ( t ) e ( ) d ( t ) v ( t ) = e ( ) d
0 0
Numerical solution
..
11/20
E
EM

( t ) = ???
12/20
E
EM

tu
t ( t ) tu t
( t ) v ( t ) = e ( ) d
0
13/20
E
EM
E
tu
t ( t )
( t ) v ( t ) = e ( ) d tu
0
t
14/20
E
EM
E
tu
t ( t )
( t ) v ( t ) = e ( ) d tu
0
t
( t ) = E ( t ) + E M ( t ) v ( t )
t ( t )
= E ( t ) + v ( t ) = E ( t ) + E M e ( ) d
0
15/20
E
v ( t ) =
1

( ( t ) v ( t ) )
EM

Laplace transform
t ( t ) t ( t )
v ( t ) = ( t ) e ( ) d ( t ) v ( t ) = e ( ) d
0 0
( t ) = E ( t ) + E M ( t ) v ( t ) ..
16/20
E
EM

u
( t )
t t
u
( t ) = E ( ) d + E M e ( ) d
0 0
t
( t )
= E + E M e
( ) d
0
t
relaxation function
= G ( t ) ( ) d
0
17/20
E
E
EM tu
tu
( t ) = ??? t
18/20
E = ( ) + M ( e )
= E
1 1
= E 2 + E M e2
2 2
EM
M e
= +
t e t
M = v M v
= +
t e t t
M M
= + v
Clausius-Planck inequality e e
M M
= + + v
e e
M M ..
= + v 0
e e
19/20
E = ( ) + M ( e )
= E
1 1
= E + E M e2
2
2 2
EM

M M
M = v = E M e + v 0
e e
M M
= + v 0
e e
v ( t ) = ( ( t ) v ( t ) )
= + M 1
20/20
Composite models of Viscoelastic response
E
1 1
= ( ) + M ( e ) = E 2 + E M e2
2 2
EM

v ( t ) = ( ( t ) v ( t ) )
1

= ( ) + 1M ( e ) + 2M ( e ) +
E1 1
1v ( t ) =
1
1
( ( t ) 1v ( t ) )
E2 2
2v ( t ) =
1
2
( ( t ) 2v ( t ) )
EN N
21/20
Composite models of Inelastic response
visco-elastic elasto-visco-plastic
E E
y
E1 1 E1 1
E2 2 E2 2
EN N EN N
= ( ) + = ( e ) +
1M ( e ) + 2M ( e ) + 1M ( e ) + 2M ( e ) +
22/20
E
Viscoelasticity
E1 1
E2 2 - an introduction
EN N
Tom Scarpas
The Netherlands
23/20
Continuum Mechanics
Large strain
viscoelastic modelling
Tom Scarpas
VE Modelling

L L
stretch: = Lr Le
L0
Lr L L
r = ; e = = e r =
L0 Lr L0
2/28 Continuum Mechanics

VE Modelling
dx = F dX dx = Fe dx r
v
dx r = Fr dX
dx = Fe Fr dX F = Fe Fr
Y
Fr
Lr Le L dxr
dX Fe
X 3, x 3
F dx
= e r
X2 , x 2
X1, x1

VE Modelling
Helmholtz free energy function
v Fv
dxp
dX Fe
F
Y
Define : Fr = Fv dx
Lr Le L = v (Ce )
S : 21 C 0
v

1C
: Ce 0
S: 2
Ce
VE Modelling
v

1C
: Ce 0
S: 2
Ce ?
C = FT F ; F = Fe Fv
= FvT FeT Fe Fv
= FvT Ce Fv Ce = FvTC Fv 1
C e = F vT C Fv 1 + FvT C Fv 1 + FvT C F v 1

VE Modelling
v
S: 1C : C e 0
2 Ce

v v
: Ce = : F vT C Fv 1 + FvT C Fv 1 + FvT C F v 1
Ce Ce

VE Modelling
v
S: 1C : C e 0
2 Ce

v v
Ce Ce
v T 1 1 p 1
: Fv C Fv = Fv : C Fv
Ce Ce
A : (BC) = (BT A) : C

VE Modelling
v
S: 1C : C e 0
2 Ce

v v
Ce Ce
v T 1 1 p 1 1 p T
: Fv C Fv = Fv : C Fv = Fv Fv : C
Ce Ce Ce
A : (BC) = (ACT ) : B

VE Modelling
v
S: 1C : C e 0
2 Ce

v v
Ce Ce
F vT C Fv 1 + FvT C F v 1 =
F v 1 = Fv 1 F v Fv 1 ; F vT = FvT F vT FvT

VE Modelling
v
S: 1C : C e 0
2 Ce

v v
Ce Ce
F vT C Fv 1 + FvT C F v 1 = FvT F vT FvT C Fv 1 + FvT C Fv 1 F v Fv 1


VE Modelling
v
S: 1C : C e 0
2 Ce

v v
Ce Ce

= l vT Ce + Ce l v

l v = F v Fv 1

VE Modelling
v
S: 1C : C e 0
2 Ce

v v
Ce Ce

= l vT Ce + Ce l v

v v
: F vT C Fv 1 + FvT C F v 1 = : l vT Ce + Ce l v
Ce Ce .

VE Modelling
v v
: F vT C Fv 1 + FvT C F v 1 = : l vT Ce + Ce l v
Ce Ce
v
= 2 : Ce l v ?
Ce
1 T
S: (
2
A + A =S:A)

VE Modelling
v 1 v
T 1 T
: Fv C Fv + Fv C Fv = : l vT Ce + Ce l v
Ce Ce
v
= 2 : Ce l v
Ce
v
= 2 Fe 1 Fe :C l
Ce e v
trick

VE Modelling
v 1 v
T 1 T
: Fv C Fv + Fv C Fv = : l vT Ce + Ce l v
Ce Ce
v
= 2 : Ce l v
Ce
v
= 2 Fe 1 Fe :C l
Ce e v
v T
(B A) : C = A : (BC)
T = 2 Fe : Fe Ce l v
Ce
v
= 2 Fe : Fe l v
Ce

VE Modelling
v
1
: Ce 0
S: 2C
Ce

S 2 Fv 1 v FvT : 21 C
Ce
Coleman-Noll
v T T procedure
+ 2F F F : Fe l v 0
e Ce e e

Stress Dissipated energy

v T v T T
S = 2 Fv 1 Fv 2 Fe Fe Fe : Fe l v 0
Ce Ce
VE Modelling
v T
S = 2 Fv 1 Fv
Ce
v T T
2 Fe Fe Fe : Fe l v 0
Ce
1 v T T
e = J = FSF = F 2 Fv
T
Fv F
Ce

v T
= 2 Fe Fe
Ce

VE Modelling
v T T
Ce
1 v T T
e = J = FSF = F 2 Fv
T
Fv F
Ce

v T
= 2 Fe Fe
Ce
e FeT : Fe l v = e FeT : Fe l v FeT FeT

= e FeT Fe 1 : Fe l v FeT
A : (BC) = (ACT ) : B
= e b 1 T
e : Fe l v Fe

VE Modelling
v T T
. 2 Fe Fe Fe : Fe l v = e b
e
1
: F l F
e v e
T
Ce
Fe l v FeT = sym Fe l v FeT + skw Fe l v FeT

= Fe sym l v FeT + skw Fe l v FeT


VE Modelling
v T T
. 2 Fe Fe Fe : Fe l v = e b
e
1
: F l F
e v e
T
Ce

d = Fe sym l v FeT + skw Fe l v FeT

L (be ) = 1 (be )
dt
be = Fe FeT
Fe = F Fv 1 FeT = FvT FT
be = F Fv 1 FvT FT = FCv 1 FT

VE Modelling
v T T
2 Fe Fe Fe : Fe l v = e b
e
1
: F l F
e v e
T
Ce


L (be ) = 1 (be )
dt
be = Fe FeT
d Fe = F Fv 1 FeT = FvT FT
dt
(
= F 1 FC
v
1 T
F F )
T

be = F Fv 1 FvT FT = FC 1 T
v F

VE Modelling
v T T
e
1
: F l F
e v e
T
Ce


L (be ) = 1 (be )
dt

d
= F
dt

(
1
)
1 T T
FCv F F

d

= F (Cv ) F1 T
dt

VE Modelling
v T T
e
1
: F l F
e v e
T
Ce


L (be ) = 1 (be )
dt 1
C v v =I
C
d C v C 1
+ Cv C 1
=0
( )
v v
= F 1 1 T T
FCv F F
dt C 1 = C 1 C C 1
v v v v
d
1 T
= F (Cv ) F = F C 1
C 1 T
v v v F
C
dt

VE Modelling
v T T
e
1
: F l F
e v e
T
Ce


L (be ) = 1 (be )
dt

d

= F
dt

(1
)
1 T T
FCv F F

Cv = FvT Fv
C = F T F + F T F
v v v v v
d
1 T
= F (Cv ) F = F C 1
C 1 T
v v v F
C
dt

VE Modelling
v T T
e
1
: F l F
e v e
T
Ce


L (be ) = 1 (be )
dt

d

= F
dt

1
( 1 T
)
FCv F F T

d
= F (Cv ) FT = F C
1
v
1
C 1 T
v v F
C
dt
= F Fv 1 FvT F vT + F v Fv 1 FvT FT

= 2 Fe sym l v FeT

VE Modelling
v T T
e
1
: F l F
e v e
T
Ce


L (be ) = 1 (be )
dt

d

= F
dt

1
( 1 T
)
FCv F F T

d
= F (Cv ) FT = F C
1
v
1
C 1 T
v v F
C
dt
= F Fv 1 FvT F vT + F v Fv 1 FvT FT

= 2 Fe sym l v FeT

VE Modelling
Dissipated energy inequality
v T T
. 2 Fe
Ce
Fe Fe : Fe l v = e b
e
1 1
:
2 L ( b e ) + skw F l (
F T
e v e

)
(
= e : 21 L (be ) + skw Fe l v FeT

) be 1
= e : 21 L (be ) b
e
1
0

.

VE Modelling
Visco component integration procedure
e : 12 L ( b e ) be1 0

VE Modelling
e : 12 L ( b e ) be1 0
12 L ( b e ) = C v1 : e b e e : C v1 : e 0
1 1 1
C v1 = I I I + II
2D 3 9V
D = D ( b e ) > 0 ; V = V ( b e ) > 0
.

VE Modelling
e : 12 L ( b e ) be1 0
12 L ( b e ) = C v1 : e b e 1 1 1
C v1 = I I I + II
2D 3 9V
D = D ( b e ) > 0 ; V = V ( b e ) > 0

( )
12 L ( b e ) = 12 F Cv1 F T = C v1 : e b e = C v1 : e F Cv1 F T
t
be

VE Modelling
e : 12 L ( b e ) b e1 0
12 L ( b e ) = C v1 : e b e 1 1 1
C v1 = I I I + II
2D 3 9V
D = D ( b e ) > 0 ; V = V ( b e ) > 0

t

( )
12 L ( b e ) = 12 F Cv1 F T = C v1 : e b e = C v1 : e F Cv1 F T
1
t
( )
C v = 2 F 1 C v1 : e F Cv1

VE Modelling
Field compaction

t
dx
det Fv t+ t
dx
Field compaction
Newly laid layer
Pre-compacted layer
Impedance layers
around top layers
Rolling direction
Base layer
Impedance layers
around Base layers
First Pass
Vertical Displacement, mm
Second Pass
Time Steps
t
First Pass dx
Percent of Compaction
det Fv t+ t
dx
Second Pass
Time Steps
A section in SH 21 in Texas with density
measurements at three locations under the
drum
C B A
2.5 2.0 3.0 1.5 2.5

0.5
18.00
16.00
Point A
Percent Air Voids %
14.00
12.00
10.00 Measurement
8.00 Model
6.00
4.00
2.00
0.00
0 1 2 3 4
Pass Number
C B A
2.5 2.0 3.0 1.5

0.5 2.5
18.00
16.00
Percent Air Voids %
14.00
Point B
12.00
10.00 Measurement
8.00 Model
6.00
4.00
2.00
0.00
0 1 2 3 4
Pass Number
C B A
2.5 2.0 3.0 1.5

0.5 2.5
18.00
16.00
Percent Air Voids %
14.00
12.00 Point C
10.00 Measurement
8.00 Model
6.00
4.00
2.00
0.00
0 1 2 3 4
Pass Number
C B A
2.5 2.0 3.0 1.5

0.5 2.5
Continuum Mechanics
Large strain
viscoelastic modelling
Tom Scarpas
Parameter determination
Mirella M Villani, MSc

Program of Mechanics of Infrastructure Materials
Room 6.50
CT4353 Continuum Mechanics, 2010-2011 1

Material properties

Example 1

Example 2
HEART ANIMATIONS COMPUTED BY THE IMMERSED BOUNDARY METHOD by D. McQueen, C. Peskin, and E. Wolfson
http://math.nyu.edu/faculty/peskin/

Parameters and models
Constitutive model
Parameter
Strain energy function

Phenomenological
(, )
approach
Theoretical approach

From experimental data to input
parameters
d
stretch Stress
d
time stretch
Parameters+
Boundary
conditions

Case of study: Relaxation test
1. Constitutive model
2. Appropriate strain energy function

( I1 , I 2 , I3 ) =
2
( I1 3 ) + (
1

I 3 1)

Courtesy of Nottingham University
3. Parameters determination

Characteristics of the Zener model
1 = 1 = M 1
M 11 = M 11e = 11v
M M

Basic concept

+ M
3.5 t
3.0
2.5
P11 (MPa)
2.0
experimental data
1.5
1.0

0.5
0.0

0 1000 2000 3000 4000 5000
M t (s)
M

Basic concept

+ M
d
EM

M M

Stretch
Li
i =
( Li ) 0 1.03
stretch in the axial direction
1 1.025
1.02 total stretch
stretch 1
1.015
1.01
1.005
2
stretch
1
0 2000 4000 6000 8000 10000 12000
t (s)
1 0 0 12 0 0 P11 0 0

F = 0 2 0 C = 0 22 0 P = 0 P22 0
0 0 2
2 0 0 2
0 0 P22
time

Strain Energy function in principal
stretches
Neo-Hookean material
3

I1 = 12 + 2 22 = 2j =
I 1 2
(
)
1 j=1
( I1 , I 2 , I3 ) = ( I1 3) + I3 1
2
3
= 2j

I3 = 12 42 = I 1
j=1 I 3 2 3
j
2

Pi = = 2 = 2 2 i
i i i 3 2 3 2
2 j 2 j
j
= +
i I 1 i j=1 I 3 i j=1
2

1 3
Pi = 2 2 i = i 1 2 2j 3 2 1
i j=1
i
= + j 2
I 1 I 3 j=1 i

1 1
P11 = 1 1
P22 = 2 1
2
(
1 1 2
2 4
)
2
(
2 1 2
2 4
)

Non linear spring component
P
=
For t=t 11
fin
1 1
P11 = 1 1 1
( )
2 2 4
1 ( 12 24 )

1 1 2

1 P 1
P22 = 2 1 P22 = 11
( )
2 ( 1 2 )

2 4
2
2 2 4 1
2 1 2 1
1 ( 12 24 )

Pii
( Pii ) + M( Pii )
Pii = Pii

( Pii )
t
tn tn+1
M M
P
=
11
For t=t 1
P11 = 1 1
fin 1
2 ( 2 4 ) 1 (1)
1 1 2 1 ( 1 2 )
2 4

P 1
1 P22 = 11
P22 = 2 1
( 2 )

2 4
2
1
2 ( 2 4 )

1
1 ( 12 24 )
2 1
2 1 2
The following steps are needed to find an analytical expression for :

1 1
2 P22 12 = P 2
( 1 2 ) ( 1 2 )
1 11
2 4 2 4
2

( ) 2 P22 1 P11
ln ( )
2 4
= ln
1 2
2
2
2 1 22 1 2 11
P P
1 2 ( 2 P11 1 P22 )
= ln
ln ( 1224 ) (2)
1 P11 2 P22

M M
P
=
For t=t 11
fin
1 1
P11 = 1 1 1
( )
2 2 4
1 ( 12 24 )

1 1 2

1 P 1
P22 = 2 1 P22 = 11
( )
2 ( 1 2 )

2 4
2
2 2 4 1
2 1 2 1
1 ( 12 24 )

The following steps are needed to find an analytical expression for :

0.20

1 1
2 P22 12 = P 2

( 1 2 )
2 4 1 11 2 ( 2 4 )
1 2

0.15
(
ln ( 12 24 )

) 2 P22 1 P11
= ln
2 1 P22 1 2 P11
2 2
P11 (MPa) 0.10
experimental data
hyperelastic model
1 2 ( 2 P11 1 P22 )
0.05
= ln
ln ( 1224 )
1 P11 2 P22
0.00
0 1000 2000 3000 4000 5000
t (s)

Instantaneous response of hyper-elastic
component of the Maxwell element
P
M P = Fv
P F T
M v
P11
2
( P11 ) = P11 P11 =
P M 1
M P11

M M 11
M 1e
at
at t n 0+
M 1e = M 1 = 1
( P11 )
L

M P11 = M P11 = M M 1 1
1

M 1
2
(
M 1
2

M 2
4
) M

P = P = 1 1
M M 2
( ) M
M 22 M 22
M 2 M 1 M 2
2 2 4

For t=t o M and M can be computed as before M M

In-time response of Maxwell
hyper-elastic component
From M Pii ( t ) determine : M i ( t ) e
( M i )e
M i = ( M i )e ( M i )v i = 1,...3
= 1.0 @ tn 0
1.0 = M i @ tn + 0
tn tn
0 +0 t
M M
From M Pii ( t ) determine : M i ( t ) e Solve for 1e and 2e

2
M 1 1 1
M P11 = P =
M M 1e
4 M
( )

M 11
M 1e
2
2
M 1e M 1e M 2e
2
M 2 1 1
P = =
M 22 M M 2e
P
( )

M 22
M 2e
M 2e M 1e M 2e
2 2 4 M

M M
Solve for M 1e and M 2e
2 M 1 1/ 4 M
M M 1e
M 2 M P22
2
M 2e
3

M 2e =
M 32e 0
M 1e M M 1 M P11
3
M
M 2e
2
( M 1e
2
M 2e
4
) M

Pii stretch in the axial direction
( Pii ) + M( Pii ) 1.02
1.015
1.01 elastic stretch

stretch 1
1.005
( Pii )
0.995
t 0.99
tn tn+1 0 1000 2000 3000 4000 5000
t (s)
M ( P11 ) = P11 ( P11 )

viscous component
From M i ( t ) e determine M i ( t ) v
M i = ( M i )e ( M i )v i = 1,...3

1.02
1.015
1.01

stretch 1
1.005
viscous stretch
0.995
0.99
0 1000 2000 3000 4000 5000
t (s)
M M
Stretch comparison
1.015
1.01
1.005
stretch 1
0.995
total stretch
viscous stretch
elastic stretch
0.99
0 1000 2000 3000 4000 5000
t (s)

Stress comparison
2.5
2.0
P11 (MPa)
1.5 experimental data

maxwell
single spring
1.0
0.5
0.0
0 1000 2000 3000 4000 5000
t (s)

Time response of Maxwell viscous
component
D = P : F

T T
= P : F 2 Fe Fv : F + 2 Ce Fv : Fv
Ce Ce
T
= P 2 Fe Fv : F + 2 Ce :Lv 0
C e C e
T Dissipated
P = 2 Fe Fv = 2 Ce = Ce S e :Lv 0
Ce Ce energy
Stress
Evolution law L v = C 1
v :
1
M = M Ce M Se = M Ce M Fe M Pe

Considering the first component In terms of principal stretches Lv =
1

M 11 = M e11 M e11 M Pe11
C F
1
= M e1 M e1 M P11e =
2
M e1 M P11e

viscous component
i M
= ( M i )v i = 1,...3 ( M i )v
( M i )e
1.025
numerical
fitting
1.02
lambda1v
curve fit tn t n +1 t
experimental data
1.015
1.01
0 300 600 900 1200
time (s)

viscous component
From i ( t ) e determine i ( t ) v
M M
( M i )v
= 1.0 @ tn 0
total
= M i @ tn + 0
external
numerical
i = ( M i )e ( M i )v fitting
M i = 1,...3
Compute time derivative
tn t n +1 t
( M i )v

M M
tn t n +1 t

Time response of Maxwell viscous
component
1
Postulate : M Lv = M
e1 M P11e
M
= M 1v
1v
300000
viscosity coefficient (MPa s) .
250000 1.4
200000
1.2
150000
1.0
100000

50000
P11 (MPa)
model results
0 0.6
0 0.5 1 1.5 2 2.5 3 3.5
Mandel/b (MPa)
0.4
0.2
0.0
3450 3650 3850 4050 4250 4450 4650 4850 5050
t (s)

Stress comparison for all the cycles
3.5
3.0
2.5
P11 (MPa)

maxwell
1.5 single spring
1.0
0.5
0.0
0 2000 4000 6000 8000 10000 12000
t (s)

Continuum Mechanics
Large strain
elasto-plastic modelling I
Tom Scarpas
Elasto-Plastic Modelling

L L
stretch: = Lr Le
L0
Lr L L
r = ; e = = e r =
L0 Lr L0

dx = F dX dx = Fe dx r
v
dx r = Fr dX
dx = Fe Fr dX F = Fe Fr
Y
Fr
Lr Le L dxr
dX Fe
X 3, x 3
F dx
= e r
X2 , x 2
X1, x1

Helmholtz free energy function
v Fp
dxp
dX F
F
Y Define : Fe = F
dx
Fr = Fp
Lr Le L = p (C , p )
S : 21 C 0
p p
S: 1C
: C + p 0
2
C p
p p

1C 0
S: 2 : C + p
C p
?
C = FT F ; F = F Fp
= FpT FT F Fp
= FpT C Fp C = FpTC Fp 1
C = F pT C Fp 1 + FpT C Fp 1 + FpT C F p 1

p
: C + p 0
1C p
S: 2
C p
p p T

: C = : Fp C Fp 1 + FpT C Fp 1 + FpT C F p 1
C C
p p 1 1 p T
: FpT C Fp 1 = Fp 1 : C Fp = Fp Fp : C
C C C
A : (BC) = (BT A) : C
A : (BC) = (ACT ) : B

p
: C + p 0
1C p
S: 2
C p
p p T

C C
F pT C Fp 1 + FpT C F p 1 =
F p 1 = Fp 1 F p Fp 1 ; F pT = FpT F pT FpT

p
: C + p 0
1C p
S: 2
C p
p p T

C C
F pT C Fp 1 + FpT C F p 1 = FpT F pT FpT C Fp 1 + FpT C Fp 1 F p Fp 1


p
: C + p 0
1C p
S: 2
C p
p p T

C C

= l pT C + Cl p

l p = F p Fp 1

p
: C + p 0
1C p
S: 2
C p
p p T

C C

= l pT C + Cl p

p T 1 p
1 T
: Fp C Fp + Fp C Fp = : l pT C + C l p .
C C


p T 1 p T
1 T
: Fp C Fp + Fp C Fp = : l p C + C l p
C C
p
= 2 : Cl p
C
1 p
= 2 F F :C l
C p
p
(B A) : C = A : (BC)
T
= 2 F : FT Cl p
C
p
= 2 F : F l p
C

p
S : 21 C : C + p 0
p
C p
p T 1
S 2 Fp 1
Fp : 2 C
C
Coleman-Noll
p T T p
+ 2 F F F : F l p p 0

procedure
C p
Stress Dissipated energy

p p T p
S = 2 Fp 1 FpT 2 F T
F F : F l p p 0
C C p
Plastic component
p
S = 2 Fp 1 FpT
C
p T T p
2 F F F : F l p p 0
C p
1 p T T
= J = FS F = F 2 Fp
T
Fp F
C

p T
= 2 F F
C

Plastic component
p T T p
2 F F F : F l p p 0
C p
1 p T T
= J = FS F = F 2 Fp
T
Fp F
C

p T
= 2 F F
C
FT : F l p = FT : F l p FT FT

= FT F 1 : F l p FT
A : (BC) = (ACT ) : B
= b 1 T
: F l p F

p
2 F FT FT : F l p = b

1
: F l F
p
T
C
F l p FT = sym F l p FT + skw F l p FT

= F sym l p FT + skw F l p FT


p

1
: F l F
p
T
C

d = F sym l p FT + skw F l p FT

L (b ) = (b )
1
dt
b = F FT
F = F Fp 1 FT = FpT FT
b = F Fp 1 FpT FT = FC 1 T
p F

p

1
: F l F
p
T
C


L (b ) = (b )
1
dt
b = F FT
d F = F Fp 1 FT = FpT FT
dt
(
= F 1 FC 1 T
p F )
FT
b = F Fp 1 FpT FT = FC 1 T
p F


p

1
: F l F
p
T
C


L (b ) = (b )
1
dt

d

= F
dt
1
( 1 T
)
T
FCp F F

d

= F (Cp ) F1 T
dt

p

1
: F l F
p
T
C


L (b ) = (b )
1
dt 1
C p C p =I
d C p C 1
+ Cp C 1
=0
( )
p p

= F 1 1 T T
FCp F F
dt 1 = C 1 C C 1
C
p p p p
d
1 T
= F (Cp ) F = F C 1
C 1 T
p p p F
C
dt

p

1
: F l F
p
T
C


L (b ) = (b )
1
dt

d
Cp = FpT Fp

= F
dt

1
( 1 T
) T
FCp F F

= F T F + F T F
C p p p p p
d
= F (Cp ) FT = F C
1
p
1
C 1 T
p p F
C
dt

p

1
: F l F
p
T
C


L (b ) = (b )
1
dt

d

= F
dt
1
( 1 T
) T
FCp F F

d
1 T
= F (Cp ) F = F C 1
C C1 T
F
p p p

dt
= F Fp 1 FpT F pT + F p Fp 1 FpT FT

= 2 F sym l p FT

p

1
: F l F
p
T
C


L (b ) = (b )
1
dt

d

= F
dt
1
( 1 T
) T
FCp F F

d
1 T
= F (Cp ) F = F C 1
C C1 T
F
p p p

dt
= F Fp 1 FpT F pT + F p Fp 1 FpT FT

= 2 F sym l p FT

p
2 F
C
FT FT : F l p = b 1 1
( T
: 2 L (b ) + skw F l p F
)
( )
= : 21 L (b ) + skw F l p FT b

1
= : 21 L (b ) b 1

Dissipated energy inequality

p T T p
2 F F F : F l p p 0
C p
: 21 L (b ) b 1
+ q : p 0

p
q =
p

Plastic component integration procedure
: 21 L (b ) b 1
+ q : p 0

Maximum plastic dissipation
Given : ( 1 L (b ) b 1 ,
2
p )
{ , q } (
# f , q 0)
{ , q }
(
max : 21 L (b ) b

1
+ q : p )

Plastic component integration procedure
: 21 L (b ) b 1
+ q : p 0

Maximum plastic dissipation Constrained minimization problem
Given : ( 1 L (b ) b 1 ,
2
p ) minimize :
{ , q } (
# f , q 0) (
: 21 L (b ) b

1
+ q : p )
( )
# f , q 0
{ , q }
(
max : 21 L (b ) b

1
+ q : p )

Constrained minimization problem
1 T
21 L (b ) = 21 F (C p ) F
minimize : t
(
: 21 L (b ) b 1
) ( )
= f b
+ q : p

= ( f ) FC 1 T
p F
( )
# f , q 0

1
t
( (
Cp ) = 2 F 1 f FC
p
1
)
f
21 L (b ) = b

(
N = FT f F ) pull-back
= f
q
1
0 ; f ( ,q ) 0 ; f ( ,q )= 0
t
(Cp ) = 2 C 1 NC
p
1

Continuum Mechanics
Large strain
elasto-plastic modelling
Tom Scarpas
Continuum Mechanics
Large strain
elasto-visco-plastic modelling
Tom Scarpas
EVP Modelling
E E1 EN
E
EM

1 N
E E
y
E1 1 E1 1
E2 2 E2 2
EN N EN N
Continuum Mechanics
EVP Modelling
E
y
E1 1
Lr Le L
X 3, x 3
X2 , x 2
X1, x1
Continuum Mechanics
EVP Modelling
E
y
dx p = Fp dX
Fp F = F Fp
dxp
dX F
F
Y dx = F dx r
dx
Lr Le L dx = F dX
X 3, x 3
X2 , x 2
X1, x1
Continuum Mechanics
EVP Modelling
E1 1
Fp F = F Fp
dxp
dX F
F
Y
Fv
dx
dxv Fe
Lr Le L dx = F dX
dx = Fe dx r
X 3, x 3 dx v = Fv dX
F = Fe Fv
X2 , x 2
X1, x1
Continuum Mechanics
EVP Modelling
Helmholtz free energy function E
y
Fp
dxp E1 1
dX F
F
Fv
dx
dxv Fe = v (Ce ) + p (C , p )
S : 21 C 0
v p
: Ce
p 0

1C p
S: : C +
2
Ce C p
Continuum Mechanics
EVP Modelling
v p
: Ce
p 0

1C p
S: : C +
2
Ce C p
p T 1
S 2 Fv 1 v T
Fv 2 Fp 1
Fp : 2 C
Ce C

Coleman-Noll
+ 2 Fe v
Fe Fe : Fe l v
T T
Ce procedure

+ 2 F p FT FT : F l p 0
C p

p
p
Continuum Mechanics
EVP Modelling
Coleman-Noll procedure
Fp
Stress dxp
dX F
1 v T 1 p T F
S = 2 Fv Fv + 2 Fp Fp
Ce C
Fv
= Se + S dx
dxv Fe
additive decomposition
Dissipated energy
E
y
v T T
Ce
E1 1
p T T p
2 F F F : F l p p 0
C p
Continuum Mechanics
EVP Modelling
Strain-axial [-]
0.0008
CAPA-3D
E
y 0.0006 Analytical
0.0004
E1 1
0.0002
< Y t [s]
0.0000
0 5 10 15 20
Stress-axial [MPa]
0.08
k 0.06 CAPA-3D
Y Analytical
0.04
0.02
t [s]
0.00
-0.02 0 5 10 15 20
F (,q ) = Y + k
-0.04
-0.06
Continuum Mechanics
EVP Modelling
0.0020
CAPA-3D
Strain-axial [-]
E Analytical
y 0.0015
0.0010
E1 1
> Y
0.0005
t [s]
0.0000
0 5 10 15 20

0.0020
VE spring
Strain-axial [-]
k EP spring
Y
0.0015
Slider
Dashpot
0.0010
Analytical
0.0005
F (,q ) = Y + k 0.0000
0 5 10 15
t [s]
-0.0005
Continuum Mechanics
EVP Modelling
Example
f ( ,i , q ) = 3
2 (
s ,i y0 q ) s ,i = ( s ,i s ,i )
1
2

= dev ( ,i ) = I I I
1
s ,i : ,i
3
2

3 3 2
p ( ,i , ) = ,i + ,i +
2 i =1 i =1 10
( )
(
)
8
1 ( )
y y0 + exp 1
= y + y0 y exp
= 0.05
2
= 0.10

= 0.20
0

0 20 40 60 80
Continuum Mechanics
EVP Modelling
-1.80E-04
Maxwell material model
Standard Linear Solid I
-1.50E-04
Elasto-visco-plasticity
I
-1.20E-04
Stress (MPa)
-9.00E-05
II
II
-6.00E-05
III
-3.00E-05 III
0.00E+00
0 -0.0002 -0.0004 -0.0006 -0.0008 -0.001 -0.0012 -0.0014
strain
Continuum Mechanics
EVP Modelling
-1.40E-04
Strainrate=0.000125
-1.20E-04 Strainrate=0.000250
Strainrate=0.00125
-1.00E-04
-8.00E-05
Stress (MPa)
-6.00E-05
-4.00E-05
-2.00E-05
0.00E+00
0 2 4 6 8 10 12 14 16
Time (sec)
Continuum Mechanics
EVP Modelling
E
y
Ee

Continuum Mechanics
EVP Modelling
E
y
Ee

dX
VX
dx = F dX F
vx
dx
Continuum Mechanics
EVP Modelling
E
y
Ee V
dx e
F Fp
VX
dX
F
dx = F dX Vx
dx
dxe = F dX ; dx = Fp dx e
= Fp F dX
Continuum Mechanics
EVP Modelling
E
dx v
Fe
y
Ee Fv V
dx e
F Fp
dX
VX
F
dx = F dX Vx
dx
dxe = F dX ; dx = Fp dx e
= Fp F dX
dxe = Fe dx v ; dx v = Fv dX dxe = Fe Fv dX
Continuum Mechanics
EVP Modelling
E
dx v
Fe
y
Ee Fv V
dx e
F Fp
dX
VX
F
dx = F dX Vx
dx
dx e = F dX ; dx = Fp dx e
= Fp F dX
dxe = Fe dx v ; dx v = Fv dX dx e = Fe Fv dX
dx = Fp Fe Fv dX
Continuum Mechanics
EVP Modelling
Continuum Mechanics
Large strain
elasto-visco-plastic modelling
Tom Scarpas
Continuum Mechanics
Continuum Mechanics
Strong discontinuities
Tom Scarpas
multiaxial material response
1
multiaxial material response
1
compressive vs. tensile response
15OC & 0.1%/sec

6
[N/mm ] 2
0
-25 -20 -15 -10 -5 0 5 10 15 20 25
urad [mm] uax [mm]
2
Gf = ?
material response determination
L= L
F L
L= L
L
L+L
L= L
L
L+L
L= L
L
L+L
Lu= L/2 Lu= L/2

F
L
L+L
Lu= L/2 Lu= L/2

F
L
L+L
Lu= L/2 Lu= L/2

F
L
L+L
Lu= L/2 Lu= L/2

F
L
L+L
150 mm
50 mm
DBM Mixture
ACRe Mixture
Tensile response phases (summary)
Lu w
L
cohesive crack model

ft
ft
Gf
w
no loading slope !!!
L
w
crack response interpretation
(
= f t T, , w )

Lu w
L
classical fracture mechanics theory

r

r
ft B

r
ft B

r
A
initiation
propagation
ft B

r
A
initiation
propagation
+ d
B
r
A
initiation
propagation
+ d
B
r
A
+ d
B
r
A
propagation
ft B
A r
propagation
ft B
A r
r
initiation
propagation
n
ft
s w
Dnn 0 0
w D = 0 Dtt 0
u v 0 0 Dss
X2
X3 X1
iff : w 0
cohesive crack implementation
Quadratic elements network Interface elements network

straight crack simulation
straight crack simulation
tangential crack displacements

s ft
t
Dnn 0 0
w D = 0 Dtt 0
u
0 0 Dss
v
X2
X3 X1 iff : w 0
tangential crack displacements
u v
X2
X3 X1
Dnn 0 0
D = 0 Dtt 0
0 0 Dss
iff : w 0
cohesive crack model
n
s
t
t
s ( ) (
2 + q s2 + 2t = f T, , w )

+
s t
cohesive crack implementation

n
s
t
s t
s
large scale computations
Delft embedded cracking model
2
6 5
7
12 11
19 18 17
4
8
1
20 16
1 2 3
0
9
1,
10
3
0
1,
13 14 15
1,0 1,0
2
6 5
7
12 11
19 18 17
J2 8
4
1
20 16
1 2 3
0
9
1,
10
3
0
1,
I1 13 14 15
1,0 1,0
2
6 5
7
12 11
19 18 17
4
8
1
20 16
1 2 3
0
9
1,
10
J2 3
0
1,
13 14 15
1,0 1,0

s
t
I1
s t
Discontinuous simulation of cracking
V V +
F
u w
w
1 if X V +
u Hd =

0 if X V
u ( X ) = H d u + ( X ) + (1 H d ) u ( X )
(
F ( u ) = F Hd u+ ) + F (1 H d ) u ?
= H d u + + H d F u + ( ) ( )
H d u + (1 H d ) F u
1 if X V + V V+
Hd =

0 if X V
if X
H d = N : = +
0 if X V V w
1 if X V + V V+
Hd = Dirac delta

0 if X V
if X
H d = N : = +
0 if X V V w
1 if X V + V V+
Hd =

0 if X V
if X
H d = N : = +
0 if X V V w
(
F ( u ) = F Hd u+ ) + F (1 H d ) u
( ) H u + (1 H ) F (u )
= H d u + + H d F u + d

d

= u N + H F ( u ) u N + (1 H ) F ( u )

+
d
+

d

= H F ( u ) + (1 H ) F ( u ) + ( u u ) N
d
+
d

+

= F + a u b N
1 if X V + V V+
Hd =

0 if X V
if X
H d = N : = +
0 if X V V w
(
F ( u ) = F Hd u+ ) + F (1 H d ) u
( ) H u + (1 H ) F (u )
= H d u + + H d F u + d

d

= u N + H F ( u ) u N + (1 H ) F ( u )

+
d
+

d

= H F ( u ) + (1 H ) F ( u ) + ( u u ) N
d
+
d

+

= F + a u b N
F ( u ) = F + a u b N V V+
(

)
= I + a u b N F 1 F
= Fd F w
F ( u ) = F + a u b N V V+

(
= I + a u b N F 1 F

)
= Fd F w
F Fd
F = Fd F dX
F
Constitutive Formulation
D = P : F
0
n

= P

:F
F
: i 0
= ( F ( X ) , 1 , 2 ,... n )
i =1 i
( F ( X ) , 1 , 2 ,... n ) Helmholtz free energy / unit volume

P=
F
V
V + (
= F (X), 1, 2,... n , au b )
free energy free energy unit volume

d = = =h
unit surface unit volume unit surface
w

d = = =h V V+
unit surface unit volume unit surface

(
F ( u ) = I + a u b N F 1 F

)
w
1 if X
= : = +
h 0 if X V V
lim h 0
1
F (u ) = I + a u b N F F
h
= Fd F

d = = =h V
unit surface unit volume unit surface V+
d d ( F ( X ) , 1 , 2 ,... n , au b )
=
au b au b
w
( h ) ( ) ( ) ( F )
= =h =h = T = P N
au b au b F au b
dq
N
dq

T = lim dA
dA
dA0
identify
d V V+
d = d (au b) : = T
au b
F w
F
u w Gf
w
u
identify V V+
d
d = d (au b) : = T
au b
F
w
ft
( )
d cde u fgh = Gf

{ ( t ) }
1 exp f G c u f
f de gh

Gf
cd u fg
e h
discontinuous simulation of cracking
t
s
n 1
3
u ( x, t ) = u ( x, t ) + H d w ( x, t )
1 if x V + n
Hd = w = Nk w k

0 if x V k =1

Continuum Mechanics

Uploaded by

Copyright:

Available Formats

Continuum 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

Continuum Mechanics

Uploaded by

Copyright:

Available Formats

CT 4353 2010/2011

Introduction to vectors and tensors

CT4353 Continuum Mechanics, 2010-2011 [email protected]

Tensors are the words of continuum mechanics

CT4353 Continuum Mechanics, 2010-2011 [email protected]

Higher order tensor operations

Linearized Internal Virtual Work

CT4353 Continuum Mechanics, 2010-2011 [email protected]

0th order tensor: SCALAR

A physical quantity, completely described by 1 (real) number

1st order tensor: VECTOR

Something with length and direction

CT4353 Continuum Mechanics, 2010-2011 [email protected]

Introduction to vectors and tensors

0th order tensor: SCALAR

Example? temperature, density, time

1st order tensor: VECTOR

Example? velocity, force, electric field

CT4353 Continuum Mechanics, 2010-2011 [email protected]

CT4353 Continuum Mechanics, 2010-2011 [email protected]

Introduction to vectors and tensors

CT4353 Continuum Mechanics, 2010-2011 [email protected]

Introduction to vectors and tensors

direct notation basis notation

Introduction to vectors and tensors

Orthonormal right-handed system (ORH)

Rectangular Cartesian System (RCS) {e1, e2 , e3 }

v1i e2 v1f 1 0 0 v1i

CT4353 Continuum Mechanics, 2010-2011 [email protected]

Introduction to vectors and tensors

v1i e2 v1f 1 0 0 v1i

0th order tensor: SCALAR

1st order tensor: VECTOR

2nd order tensor: TENSOR

Introduction to vectors and tensors

Geometrical interpretation PROJECTION

Introduction to vectors and tensors

Geometrical interpretation PROJECTION

CT4353 Continuum Mechanics, 2010-2011 [email protected]

Index notation: Or in index notation:

Introduction to vectors and tensors

CT4353 Continuum Mechanics, 2010-2011 [email protected]

TENSOR notations MATRIX notations

CT4353 Continuum Mechanics, 2010-2011 [email protected]

Introduction to vectors and tensors

Right handed system 1

CT4353 Continuum Mechanics, 2010-2011 [email protected]

Repeated index = implicit summation rule

Introduction to vectors and tensors

CT4353 Continuum Mechanics, 2010-2011 [email protected]

Introduction to vectors and tensors

aibj vk ei (e j ek ) = aibj vk ei jk = aibj vk ei jk

Tqms pk vq km id = Tqmsvq pm id = Tqpsvq id

Introduction to vectors and tensors

direct notation index notation matrix notation

Introduction to vectors and tensors

Aij A12 A11 A12

CT4353 Continuum Mechanics, 2010-2011 [email protected]

si = Bpis Asp + Aimvm