Solution Manual For Continuum Mechanics For En3e R.N. Mase, B. Smelser, M. Rossmann

SOLUTIONS MANUAL FOR
Continuum
Mechanics for Engineers,
Third Edition
by
Thomas Mase
Ronald E. Smelser
SOLUTIONS MANUAL FOR
Continuum
Mechanics for Engineers,
Third Edition
by
Thomas Mase
Ronald E. Smelser
Boca Raton London New York
CRC Press is an imprint of the

Taylor & Francis Group, an informa business
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2010 by Taylor and Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works
Printed in the United States of America on acid-free paper

10 9 8 7 6 5 4 3 2 1
International Standard Book Number: 978-1-4398-0898-6 (Paperback)
This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and
information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and
publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission
to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any
future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic,
mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or
retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact
the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides
licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment
has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation
without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Chapter 2 Solutions
Problem 2.1
Let v = a × b, or in indicial notation,
vi e ^j × bk e
î = aj e ^k = εijk aj bk e
î
Using indicial notation, show that,

(a) v · v = a2 b2 sin2 θ ,
(b) a × b · a = 0 ,
(c) a × b · b = 0 .
Solution
(a) For the given vector, we have
v · v = εijk aj bk e
î · εpqs aq bs e
^p = εijk aj bk εpqs aq bs δip = εijk aj bk εiqs aq bs
= (δjq δks − δjs δkq ) aj bk aq bs = aj aj bk bk − aj bk ak bj
= (a · a) (b · b) − (a · b) (a · b) = a2 b2 − (ab cos θ)2
= a2 b2 1 − cos2 θ = a2 b2 sin2 θ

(b) Again, we find

a × b · a = v · a = (εijk aj bk e
î ) · aq e
^q = εijk aj bk aq δiq = εijk aj bk ai = 0
This is zero by symmetry in i and j.

(c) This is
a × b · b = v · b = (εijk aj bk e
î ) · bq e
^q = εijk aj bk bq δiq = εijk aj bk bi = 0
Again, this is zero by symmetry in k and and i.
Problem 2.2
With respect to the triad of base vectors u1 , u2 , and u3 (not necessarily unit vectors), the
triad u1 ,u2 , and u3 is said to be a reciprocal basis if ui · uj = δij (i, j = 1, 2, 3). Show that
to satisfy these conditions,
u2 × u3 u3 × u1 u1 × u2
u1 = ; u2 = ; u3 =
[u1 , u2 , u3 ] [u1 , u2 , u3 ] [u1 , u2 , u3 ]
and determine the reciprocal basis for the specific base vectors
u1 ^2 ,
e1 + e
= 2^
u2 ^3 ,
e2 − e
= 2^
u3 ^1 + e
= e ^2 + e^3 .
1
2 Continuum Mechanics for Engineers
Answer
1
u1 = 5
(3^ ^2 − 2^
e1 − e e3 )
1
u2 = 5 e1 + 2^
(−^ ^3 )
e2 − e
1
u3 = 5 e1 + 2^
(−^ e2 + 4^
e3 )
Solution
For the bases, we have
u2 × u3 u3 × u1 u1 × u2
u1 ·u1 = u1 · = 1; u2 ·u2 = u2 · = 1; u3 ·u3 = u3 · =1
[u1 , u2 , u3 ] [u1 , u2 , u3 ] [u1 , u2 , u3 ]
since the triple scalar product is insensitive to the order of the operations. Now
u2 × u3
u2 · u1 = u2 · =0
[u1 , u2 , u3 ]
since u2 ·u2 ×u3 = 0 from Pb 2.1. Similarly, u3 ·u1 = u1 ·u2 = u3 ·u2 = u1 ·u3 = u2 ·u3 = 0.
For the given vectors, we have

2 1 0

[u1 , u2 , u3 ] = 0 2 −1 = 5
1 1 1
and

^1 e^2 e^3
e
1
u2 × u3 = 0 2 −1 = 3^ ^2 − 2^
e1 − e e3 ; u1 = ^2 − 2^
e1 − e
(3^ e3 )
1 1 1 5

^1 e ^2 e^3
e
1
u3 × u1 = 1 1 1 = −ê1 + 2^ ^3 ;
e2 − e u2 = e1 + 2^
(−^ ^3 )
e2 − e
2 1 0 5

^1 e ^2 e ^3
e
1
u1 × u2 = 2 1 0 = −ê1 + 2^
e2 + 4^
e3 ; u3 = e1 + 2^
(−^ e2 + 4^
e3 )
0 2 −1 5
Problem 2.3
î , and let b = bi e
Let the position vector of an arbitrary point P (x1 x2 x3 ) be x = xi e î be
a constant vector. Show that (x − b) · x = 0 is the vector equation of a spherical surface
having its center at x = 21 b with a radius of 21 b.
Solution
For
(x − b) · x = (xi e î ) · xj e
î − bi e ^j = (xi xj − bi xj ) δij = xi xi − bi xi =
= x21 + x22 + x23 − b1 x1 − b2 x2 − b3 x3 = 0
Now
2 2 2
1 1 1 1 2 1
x1 − b1 + x2 − b2 + x3 − b3 = b1 + b22 + b23 = b2
2 2 2 4 4
Chapter 2 Solutions 3
This is the equation of a sphere with the desired properties.
Problem 2.4
Using the notations A(ij) = 21 (Aij + Aji ) and A[ij] = 21 (Aij − Aji ) show that
(a) the tensor A having components Aij can always be decomposed into a sum of
its symmetric A(ij) and skew-symmetric A[ij] parts, respectively, by the decom-
position,
Aij = A(ij) + A[ij] ,
(b) the trace of A is expressed in terms of A(ij) by
Aii = A(ii) ,
(c) for arbitrary tensors A and B,
Aij Bij = A(ij) B(ij) + A[ij] B[ij] .
Solution
(a) The components can be written as

Aij + Aji Aij − Aji
Aij = + = A(ij) + A[ij]
2 2
(b) The trace of A is

Aii + Aii
A(ii) = = Aii
2
(c) For two arbitrary tensors, we have

Aij Bij = A(ij) + A[ij] B(ij) + B[ij] = A(ij) B(ij) + A[ij] B(ij) + A(ij) B[ij] + A[ij] B[ij]
= A(ij) B(ij) + A[ij] B[ij]
since the product of a symmetric and skew-symmetric tensor is zero

Aij + Aji Bij − Bji 1
A(ij) B[ij] = = (Aij Bij + Aji Bij − Aij Bji − Aji Bji )
2 2 4
1
= (Aij Bij + Aji Bij − Aji Bij − Aij Bij ) = 0
4
We have changed the dummy indices on the last two terms.
Problem 2.5
Expand the following expressions involving Kronecker deltas, and simplify where possible.
(a) δij δij , (b) δij δjk δki , (c) δij δjk , (d) δij Aik
Answer
(a) 3, (b) 3, (c) δik , (d) Ajk
Solution
(a) Contracting on i or j, we have
δij δij = δjj = δii = δ11 + δ22 + δ33 = 1 + 1 + 1 = 3
(b) Contracting on k and then j gives
δij δjk δki = δij δji = δii = 3
(c) Contracting on j yields

δij δjk = δik
(d) Contracting on i gives
δij Aik = Ajk
Note: It may be helpful for beginning students to write out all terms.
Problem 2.6
If ai = εijk bj ck and bi = εijk gj hk , substitute bj into the expression for ai to show that
a i = g k ck h i − h k ck g i ,
or in symbolic notation, a = (c · g)h − (c · h)g.
Solution
We begin by changing the dummy indices for bj = εjmn gm hn and
ai = εijk bj ck = εijk εjmn gm hn ck = − (εjik εjmn gm hn ck ) = − (δim δkn − δin δkm ) gm hn ck

= −gi hk ck + gk hi ck = gk ck hi − hk ck gi
where we have used the anti-symmetry of εijk = −εjik and the ε−δ identity. Symbolically
a = (c · g)h − (c · h)g
Problem 2.7
By summing on the repeated subscripts determine the simplest form of
(a) ε3jk aj ak , (b) εijk δkj , (c) ε1jk a2 Tkj , (d) ε1jk δ3j vk .
Answer
(a) 0, (b) 0, (c) a2 (T32 − T23 ), (d) −v2
Solution
(a) Summing gives
ε3jk aj ak = ε31k a1 ak + ε32k a2 ak = ε312 a1 a2 + ε321 a2 a1 = a1 a2 − a2 a1 = 0

(b)
εijk δkj = εij1 δ1j + εij2 δ2j + εij3 δ3j
= εi21 δ12 + εi31 δ13 + εi12 δ21 + εi32 δ23 + εi13 δ31 + εi23 δ32 = 0
(c)
ε1jk a2 Tkj = ε12k a2 Tk2 + ε13k a2 Tk3
= ε123 a2 T32 + ε132 a2 T23 = a2 T32 − a2 T23 = a2 (T32 − T23 )
(d)
ε1jk δ3j vk = ε12k δ32 vk + ε13k δ33 vk = 0 + ε132 δ33 v2 = −v2
Problem 2.8
Consider the tensor Bik = εijk vj .
(a) Show that Bik is skew-symmetric.
(b) Let Bij be skew-symmetric, and consider the vector defined by vi = εijk Bjk
(often called the dual vector of the tensor B). Show that Bmq = 12 εmqi vi .
Solution
(a) For a tensor to be skew-symmetric, one has Aij = −Aji . For the given tensor
Bik = εijk vj = −εkji vj = −Bki
(b) For the dual vector of the tensor B, we have
εmqi vi = εmqi εijk Bjk = (δmj δqk − δmk δqj ) Bjk = Bmq − Bqm = [Bmq − (−Bmq )]
= 2Bmq
since B is skew-symmetric.
Problem 2.9
Use indicial notation to show that
Ami εmjk + Amj εimk + Amk εijm = Amm εijk
where A is any tensor and εijk is the permutation symbol.
Solution
Multiply both sides by εijk and simplify
Amm εijk εijk = 6Amm = Ami εmjk εijk + Amj εimk εijk + Amk εijm εijk
= Ami 2δmi + Amj 2δmj + Amk 2δmk = 6Amm
Problem 2.10
If Aij = δij Bkk + 3Bij , determine Bkk and using that solve for Bij in terms of Aij and its
first invariant, Aii .
Answer
Bkk = 61 Akk ; Bij = 13 Aij − 1

18 δij Akk
Solution
Taking the trace of Aij gives
Aii = δii Bkk + 3Bii = 3Bkk + 3Bii = 6Bkk
since i and k are dummy indices. This gives
1
Bkk = Akk
6
Substituting for Bkk and solving for Bij gives
1 1 1
3Bij = Aij − δij Akk or Bij = Aij − δij Akk
6 3 18
Problem 2.11
Show that the value of the quadratic form Tij xi xj is unchanged if Tij is replaced by its
symmetric part, 21 (Tij + Tji ).
Solution
The quadratic form becomes
1 1 1
Tij xi xj = (Tij + Tji )xi xj = (Tij xi xj + Tji xi xj ) = (Tij xi xj + Tij xj xi ) = Tij xi xj
2 2 2
since i and j are dummy indices and multiplication commutes.
Problem 2.12
With the aid of Eq 2.7, show that any skew symmetric tensor W may be written in terms
of an axial vector ωi given by
1
ωi = − εijk wjk
2
where wjk are the components of W.
Solution
Multiply by εimn
εimn ωi = − 21 εimn εijk wjk
= − 12 (δmj δnk − δmk δnj ) wjk
= − 12 (wmn − wnm ) = wnm ,
or,
εmni ωi = wnm
Problem 2.13
Show by direct expansion (or otherwise) that the box product λ = εijk ai bj ck is equal to
the determinant
a1 a2 a3

b1 b2 b3 .

c1 c2 c3
Thus, by substituting A1i for ai , A2j for bj and A3k for ck , derive Eq 2.42 in the form
det A = εijk A1i A2j A3k where Aij are the elements of A.
Solution
Direct expansion gives
λ = εijk ai bj ck = ε1jk a1 bj ck + ε2jk a2 bj ck + ε3jk a3 bj ck

= ε12k a1 b2 ck + ε13k a1 b3 ck + ε21k a2 b1 ck + ε23k a2 b3 ck + ε31k a3 b1 ck + ε32k a3 b2 ck
= ε123 a1 b2 c3 + ε132 a1 b3 c2 + ε213 a2 b1 c3 + ε231 a2 b3 c1 + ε312 a3 b1 c2 + ε321 a3 b2 c1
= a1 b2 c3 − a1 b3 c2 − a2 b1 c3 + a2 b3 c1 + a3 b1 c2 − a3 b2 c1
and

a1 a2 a3

b1 b2 b3 = a1 b2 c3 + a2 b3 c1 + a3 b1 c2 − a1 b3 c2 − a2 b1 c3 − a3 b2 c1 = λ

c1 c2 c3
Using the suggested substitutions for ai , bi , ci , we have A3
λ = εijk A1i A2j A3k = ε1jk A11 A2j A3k + ε2jk A12 A2j A3k + ε3jk A13 A2j A3k
= ε12k A11 A22 A3k + ε13k A11 A23 A3k + ε21k A12 A21 A3k + ε23k A12 A23 A3k
+ε31k A13 A21 A3k + ε32k A13 A22 A3k
= ε123 A11 A22 A33 + ε132 A11 A23 A32 + ε213 A12 A21 A33 + ε231 A12 A23 A31
+ε312 A13 A21 A32 + ε321 A13 A22 A31
= A11 A22 A33 − A11 A23 A32 − A12 A21 A33 + A12 A23 A31 + A13 A21 A32 − A13 A22 A31
and
A11 A12 A13

A21 A22 A23 = A11 A22 A33 − A11 A23 A32 + A12 A23 A31 − A12 A21 A33

A31 A32 A33

+A13 A21 A32 − A13 A22 A31 = λ
Problem 2.14
Starting with Eq 2.42 of the text in the form
det A = εijk Ai1 Aj2 Ak3
show that by an arbitrary number of interchanges of columns of Aij we obtain
εqmn det A = εijk Aiq Ajm Akn
which is Eq 2.43. Further, multiply this equation by the appropriate permutation symbol
to derive the formula
6 det A = εqmn εijk Aiq Ajm Akn .
Solution
Each row or column change introduces a minus sign. After an arbitrary number of row
and column changes, we have
εqmn det A = εijk Aiq Ajm Akn

Multiplying by εqmn gives

εqmn εqmn det A = (δmm δnn − δmn δnm ) det A = (3 · 3 − δnn ) det A
= (9 − 3) det A = εqmn εijk Aiq Ajm Akn
from the ε − δ identity.
Problem 2.15
Let the determinant of the tensor Aij be given by
A11 A12 A13

det A = A21 A22 A23

.
A31 A32 A33

Since the interchange of any two rows or any two columns causes a sign change in the value
of the determinant, show that after an arbitrary number of row and column interchanges
Amq Amr Ams

Anq Anr Ans = εmnp εqrs det A .

Apq Apr Aps

Now let Aij = δij in the above determinant which results in det A = 1 and, upon expansion,
yields
εmnp εqrs = δmq (δnr δps − δns δpr ) − δmr (δnq δps − δns δpq ) + δms (δnq δpr − δnr δpq ) .
Thus, by setting p = q, establish Eq 2.7 in the form
εmnq εqrs = δmr δns − δms δnr .
Solution
Letting Aij = δij in the determinant gives

δmq δmr δms

δnq δnr δns

δpq δpr δps
= δmq (δnr δps − δns δpr ) − δmr (δnq δps − δns δpq ) + δms (δnq δpr − δnr δpq )
and
εmnp εqrs = δmq (δnr δps − δns δpr ) − δmr (δnq δps − δns δpq ) + δms (δnq δpr − δnr δpq )
since
δ11 δ12 δ13 1 0 0
= 1 = ε123 ε123 det A

δ21 δ22 δ23 = 0 1 0

δ31 δ32 δ33 0 0 1
Setting p = q gives

δmp δmr δms

δnp δnr δns

δpp δpr δps
= δmp (δnr δps − δns δpr ) − δmr (δnp δps − δns δpp ) + δms (δnp δpr − δnr δpp )
= δnr δms − δns δmr − δmr (δns − 3δns ) + δms (δnr − 3δnr )
= δnr δms − δns δmr + 2δmr δns − 2δms δnr = δmr δns − δms δnr
= εpmn εprs
Problem 2.16
Show that the square matrices
 
1 0 0
5 2
[Bij ] =  0 −1 0  and [Cij ] =
−12 −5
0 0 1
are both square roots of the identity matrix.
Solution
The product of the matrix with itself should be the identity matrix for it to be a square
root. Thus     
1 0 0 1 0 0 1 0 0
 0 −1 0   0 −1 0  =  0 1 0 
0 0 1 0 0 1 0 0 1
and
5 2 5 2 25 − 24 10 − 10 1 0
= =
−12 −5 −12 −5 −60 + 60 −24 + 25 0 1
Problem 2.17
Using the square matrices below, demonstrate
(a) that the transpose of the square of a matrix is equal to the square of its transpose
(Eq 2.36 with n = 2),
(b) that (AB)T = BT AT as was proven in Example 2.33
   
3 0 1 1 3 1
[Aij ] =  0 2 4  , [Bij ] =  2 2 5  .
5 1 2 4 0 3
Solution
(a) For the matrix A, we have
    
3 0 1 3 0 1 14 1 5
2
[Aij ] =  0 2 4  0 2 4  =  20 8 16 
5 1 2 5 1 2 25 4 13
and     
3 0 5 3 0 5 14 20 25
T 2
Aij

= 0 2 1  0 2 1 = 1 8 4 
1 4 2 1 4 2 5 16 13
T 2
This shows that A2 = AT . Similarly for B, we have
    
1 3 1 1 3 1 11 9 19
[Bij ]2 =  2 2 5   2 2 5  =  26 10 27 
4 0 3 4 0 3 16 12 13
and     
1 2 4 1 2 4 11 26 16
2
BTij =  3

2 0  3 2 0 = 9 10 12 
1 5 3 1 5 3 19 27 13
(b) For (AB)T = BT AT , we have

    
3 0 1 1 3 1 7 9 6
[Aij ] [Bij ] =  0 2 4  2 2 5  =  20 4 22 
5 1 2 4 0 3 15 17 16
and     
1 2 4 3 0 5 7 20 15
BTij ATij =  3

2 0  0 2 1 = 9 4 17 
1 5 3 1 4 2 6 22 16
The result is demonstrated.
Problem 2.18
Let A be any orthogonal matrix, i.e., AAT = AA−1 = I, where I is the identity matrix.
Thus, by using the results in Examples 2.9 and 2.10, show that det A = ±1.
Solution
From Example 2.9
det AAT = det A det AT

and from Example 2.10,

det A = det AT
Then
det AAT = det A det AT = det A det A = (det A)2 = det I = 1

and
(det A) = ±1
Problem 2.19
A tensor is called isotropic if its components have the same set of values in every Cartesian
coordinate system at a point. Assume that T is an isotropic tensor of rank two with
0 0 0
components tij relative to axes Ox1 x2 x3 . Let axes Ox1 x2 x3 be obtained with respect
√ to
Ox1 x2 x3 by a righthand rotation of 120◦ about the axis along n ^ = (ê+e ^+e ^) / 3. Show
by the transformation between these axes that t11 = t22 = t33 , as well as other relationships.
00 00 00
Further, let axes Ox1 x2 x3 be obtained with respect to Ox1 x2 x3 by a right-hand rotation
◦
of 90 about x3 . Thus, show by the additional considerations of this transformation that if
T is any isotropic tensor of second order, it can be written as λI where λ is a scalar and I
is the identity tensor.
Solution √
For a 120◦ rotation about the axis n
^ = (^ ^3 ) / 3, the transformation matrix is
^2 + e
e1 + e
 
0 1 0
[aij ] =  0 0 1 
1 0 0
with det A = 1. The transformation is Tij

0
= aiq ajm Tqm or
 0
t11 t012 t013
    
0 1 0 t11 t12 t13 0 0 1
 t021 t022 t023  =  0 0 1   t21 t22 t23   1 0 0 
t031 t032 t033 1 0 0 t31 t32 t33 0 1 0
 
t22 t23 t21
=  t32 t33 t31 
t12 t13 t11
Thus if T is isotropic, the transformation will not distinguish between the primed and
unprimed coordinates. This gives
t11 = t22 ; t22 = t33 ; t33 = t11 or t11 = t22 = t33
and
t12 = t23 ; t13 = t21 ; t21 = t32 ; t23 = t31 ; t31 = t12 ; t32 = t13
This results in
t12 = t23 = t31 and t13 = t21 = t32
For a 90 rotation about the x3 axis, the transformation matrix is
◦
 
0 1 0
[aij ] =  −1 0 0 
0 0 1
with det A = 1. The resulting transformation is

 0
t012 t013
    
t11 0 1 0 t11 t12 t13 0 −1 0
 t021 t022 t023  =  −1 0 0   t21 t22 t23   1 0 0 
t031 t032 t033 0 0 1 t31 t32 t33 0 0 1
 
t22 −t21 t23
=  −t12 t33 −t13 
t32 −t31 t11
Again
t11 = t22 = t33 = λ
and
t12 = −t21 ; t13 = t23 ; t21 = −t12 ; t23 = −t13 ; t31 = t32 ; t32 = −t31
This results in
t12 = −t12 = −t21 = 0; t13 = −t13 = t23 = 0; t32 = −t31 = t31 = 0
For a tensor to be isotropic, T = λI
Problem 2.20
0 0 0
For a proper orthogonal transformation between axes Ox1 x2 x3 and Ox1 x2 x3 show the
invariance of δij and εijk . That is, show that
0
(a) δij = δij ,
0
(b) εijk = εijk .
0
Hint: For part (b) let εijk = aiq ajm akn εqmn and make use of Eq 2.43.
Solution
(a) The transformation is
δ0ij = aiq ajm δqm = aim ajm = δij
(b) The transformation is

ε0ijk = aiq ajm akn εqmn
From 2.43, we have
ε0ijk = εijk det A =εijk (+1)
since the transformation is proper orthogonal.
Problem 2.21
0 0 0
The angles between the respective axes of the Ox1 x2 x3 and the Ox1 x2 x3 Cartesian systems
are given by the table below
x1 x2 x3
x10 45◦ 90◦ 45◦
x20 60◦ 45◦ 120◦
x30 120◦ 45◦ 60◦
Determine
(a) the transformation matrix between the two sets of axes, and show that it is a
proper orthogonal transformation,
√
(b) the equation of the plane x1 + x2 + x3 = 1/ 2 in its primed axes form, that is,
0 0 0
in the form b1 x1 + b2 x2 + b3 x3 = b.
Answer
 
√1 0 √1
 2 2 
(a) [aij ] =  √1 √1 − √12
 

 2 2 
1 √1 √1
− 2
√
2 2
0 0 0
(b) 2x1 + x2 + x3 = 1
Solution
(a) The transformation matrix is
  √1 √1

cos 45◦ cos 90◦ cos 45◦ 0

 2 2
[aij ] =  cos 60◦ cos 45◦ cos 120◦  =  12 √1 − 12 
2 
cos 120◦ cos 45◦ cos 60◦ − 12 √1 1
2
2
and for a proper orthogonal matrix

1 1 1 1 1 1
det A = √ √ + √ +√ √ + √ =1
2 2 2 2 2 2 2 2 2 2
(b) The transformation is
x0 = Ax
or
x = AT x0
This gives
  √1 1
− 12

x0
 
x1 2 2
√1 √1   10  1
 0
 x2  = 
2 2  x2 = √
√1 x03 2
x3
2
− 21 1
2
and
x0 x02 x03
 
  √1 + −
x1 2 2 2

 x2  =  x0 x0 

√2
2
+ √3
2


x3 x0 x02 x03
√1 − +
2 2 2
Thus
x0 x0 x0 x0 x0 x0 x0 x0

1
x1 + x2 + x3 = √1 + 2 − 3 + √2 + √3 + √1 − 2 + 3 =√
2 2 2 2 2 2 2 2 2
and
2x01 + x02 + x03 = 1
Problem 2.22
Making use of Eq 2.42 of the text in the form det A = εijk A1i A2j A3k write Eq 2.71 as
|tij − λδij | = εijk (t1i − δ1i ) (t2j − δ2j ) (t3k − δ3j ) = 0
and show by expansion of this equation that

1
λ3 − tii λ2 + (tii tjj − tij tji ) λ − εijk t1i t2j t3k = 0
2
to verify Eq 2.73 of the text.
Solution
Expansion of the equation gives
|tij − λδij | = εijk (t1i − λδ1i ) (t2j − λδ2j ) (t3k − λδ3k )

= εijk t1i t2j − λ (t2j δ1i + t1i δ2j ) + λ2 δ1i δ2j (t3k − λδ3k )
= εijk [t1i t2j t3k − λ (t3k t2j δ1i + t3k t1i δ2j ) − λδ3k t1i t2j
+λ2 δ3k (t2j δ1i + t1i δ2j ) + λ2 δ1i δ2j t3k − λ3 δ1i δ2j δ3j
= εijk t1i t2j t3k − λεijk (t3k t2j δ1i + t3k t1i δ2j + δ3k t1i t2j )
+λ2 εijk (t2j δ1i δ3k + t1j δ2j δ3k + δ1i δ2j t3k ) − λ3 εijk δ1i δ2j δ3j
= εijk t1i t2j t3k − λ (ε1jk t3k t2j + εi2k t3k t1j + εij3 t1i t2j )
+λ2 (ε1j3 t2j + εi23 t1i + ε12k t3k ) − λ3 ε123
= εijk t1i t2j t3k − λ (t33 t22 − t32 t23 + t33 t11 − t31 t13 + t11 t22 − t12 t21 )
+λ2 (t22 + t11 + t33 ) − λ3 ε123
This is the desired result since
tii tjj − tij tji = (t11 + t22 + t33 ) (t11 + t22 + t33 ) − (t1 jtj1 + t2j tj2 + t1j tj3 )
and
(t11 + t22 + t33 ) (t11 + t22 + t33 ) = t211 + t222 + t233 + 2 (t11 t22 + t22 t33 + t11 t33 )
t1 jtj1 + t2j tj2 + t1j tj3 = t211 + t12 t21 + t13 t31 + t21 t12 + t222
+ t23 t32 + t31 t13 + t32 t23 + t233
= t211 + t222 + t233 + 2 (t21 t12 + t13 t31 + t23 t32 )
or
tii tjj − tij tji = 2 (t11 t22 + t22 t33 + t11 t33 ) − 2 (t21 t12 + t13 t31 + t23 t32 )
Problem 2.23
For the matrix representation of tensor B shown below,
 
17 0 0
[bij ] =  0 −23 28 
0 28 10
determine the principal values (eigenvalues) and the principal directions (eigenvectors) of
the tensor.
Answer
λ1 = 17, λ2 = 26, λ3 = −39

√ √
^ (1) = e
n ^ (2) = (4^
^1 , n ^ (3) = (−7^
e3 ) / 65, n
e2 + 7^ e2 + 4^
e3 ) / 65
Solution
The eigenvalues are given by

17 − λ 0 0
det (B − λI) =

0 −23 − λ 28
0 28 10 − λ
= (17 − λ) (−23 − λ) (10 − λ) − 282 = 0

= (17 − λ) λ2 + 13λ − 230 − 784 = (17 − λ) (λ + 39) (λ − 26) = 0

For λ = 17, the eigenvector is found from

    
17 − 17 0 0 n1 0
 0 −23 − 17 28   n2  =  0 
0 28 10 − 17 n3 0
or
−40n2 + 28n3 = 0
28n2 − 7n3 = 0
The result is
n2 = n3 = 0
together with the unit vector relation
n21 + n22 + n23 = 1
yields
^ (1) = e
n ^1
For λ = 26, the eigenvector is found from
    
17 − 26 0 0 n1 0
 0 −23 − 26 28   n2  =  0 
0 28 10 − 26 n3 0
or
−9n1 = 0
−49n2 + 28n3 = 0
28n2 − 16n3 = 0
The result is
n1 = 0
7n2 = 4n3
n21 + n22 + n23 = 1
yields
4 7 e2 + 7^
4^ e3
n2 = √ ; n3 = √ or n
^ (2) = √
65 65 65
Similarly, for λ = −39, the eigenvector is found from
    
17 + 39 0 0 n1 0
 0 −23 + 39 28   n2  =  0 
0 28 10 + 39 n3 0
or
56n1 = 0
16n2 + 28n3 = 0
28n2 + 49n3 = 0
The result is
n1 = 0
4n2 = −7n3
n21 + n22 + n23 = 1

yields
7 4 e2 + 4^
−7^ e3
n2 = − √ ; n3 = √ or n
^ (3) = √
65 65 65
Problem 2.24
Consider the symmetrical matrix
 
5 3
4 0 2
 
[Bij ] =  0 0  .
 
4
 
3 5
2 0 2
(a) Show that a multiplicity of two occurs among the principal values of this matrix.
(b) Let λ1 be the unique principal value and show that the transformation matrix
 
√1 0 − √1
 2 2 
[aij ] =  0 1
 
0 
 
√1 0 √1
2 2
gives B∗ according to B∗ij = aiq ajm Bqm .

h i
(c) Taking the square root of B∗ij and transforming back to Ox1 x2 x3 axes show
that  
3 1
0
hp i  2 2 
Bij =  0 2 0  .
 
 
1 3
2 0 2
(d) Verify that the matrix  

− 12 0 − 32
 
[Cij ] = 
 
0 2 0 
 
− 32 0 − 12
is also a square root of [Bij ].
Solution
(a) The principal values are found from
 5 3

2 −λ 0 2
det[Bij − λδij ] = det  0 4−λ 0 =0
3 5
0 − λ
"2 2
2 2 #
5 3
= (4 − λ) −λ −
2 2
= (4 − λ) λ2 − 5λ + 4 = (4 − λ)2 (λ − 1) = 0

(b) For λ = 1, the principal vector is given by

5 3
    
2 −1 0 2 n1 0
 0 4−1 0   n2  =  0 
3 5
2 0 2 −1 n3 0
or
n1 = −n3 ; n2 = 0
together with the unit vector relation yields
1 1
^ (1) = √ e
n ^1 − √ e^3
2 2
Select n
^ (2) to be perpendicular to n
^ (1) for the λ2 = λ3 = 4 roots.
^ (2) = e
n ^2
and n
^ (3) = n
^ (1) × n
^ (2)
1 1
^ (3) = √ e
n ^1 + √ e^3
2 2
The transformation matrix is
 
√1 0 − √12
2
[aij ] =  0 1 0 
 
√1 0 √21
2
and B∗ = ABAT .
   
√1 0 − √12 5 3 √1 0 √1
 
2 0 2 2 1 0 0
 2 2
B∗ij =  0

1 0  0 4 0  0 1 0 = 0 4 0 
  
√1 0 √12 3
0 5
− √12 0 √1 0 0 4
2 2 2 2
(c) Taking the square root yields

 
hq i 1 0 0
B∗ij = 0
 2 0 
0 0 2
√ √
and transforming back gives B = AT B∗ A
    √1  
√1 0 √1 0 − √12 3 1

2 2 1 0 0 0
 2   2 2
hp i
Bij =  0 1 0  0 2 0  0 1 0 = 0 2 0 
 
− √1 0 √1 2
0 0 2 √1
2 2
0 √1
2
1
2 0 3
2
(d) For [Cij ] to be a square root of [Bij ], we must have
− 21 − 32
 1
− 32
   5 3

0 −2 0 2 0 2
[Cij ] [Cij ] =  0 2 0  0 2 0 = 0 4 0  = [Bij ]
− 32 0 − 12 − 32 0 − 12 3
2 0 5
2
Problem 2.25
Determine the principal values of the matrix
 
4 0 √0
[Kij ] =  0 √11 − 3
 ,
0 − 3 9
and show that the principal axes Ox∗1 x∗2 x∗3 are obtained from Ox1 x2 x3 by a rotation of 60◦
about the x1 axis.
Answer
λ1 = 4, λ2 = 8, λ3 = 12.
Solution
The principal values are found from

4−λ 0 0 √ 2
√

det [K − λI] = 0

11 √
− λ − 3 = (4 − λ) (11 − λ) (9 − λ) + 3 =0
0 − 3 9−λ
2
= (4 − λ) λ − 20λ − 96 = 0
= (4 − λ) (λ − 8) (λ − 12) = 0
The eigenvalues are
λ1 = 4; λ2 = 8; λ3 = 12
The associated principal vectors are given by Eq 2.57. For λ1 = 4 we have
√
7n2 − 3n3 = 0
√
− 3n2 + 5n3 = 0
or
n2 = n3 = 0
so that
^ (1) = e
n ^1
For λ2 = 8, we have
−4n1 = 0
√
3n2 − 3n3 = 0
√
− 3n2 + n3 = 0
or √
n3 = 3n2 and n21 + n22 + n33 = 1
so that √
(2) 1 3
^
n = e^2 + ^3
e
2 2
−8n1 = 0
√
−n2 − 3n3 = 0
√
− 3n2 − 3n3 = 0
or √
n2 = − 3n3 and n21 + n22 + n33 = 1
so that √
(3) 3 1
^
n =− ^2 + e
e ^3
2 2
 
1 0 0 cos 0◦ cos 90◦ cos 90◦
 
√
[aij ] =  0 1 3
 = cos 90◦ cos 60◦ cos 30◦ 
  
2
√ 2
0 − 23 1 cos 90◦ cos 120◦ cos 60◦
2
This is the desired result.
Problem 2.26
^ (q) (q = 1, 2, 3)
Determine the principal values λ(q) (q = 1, 2, 3) and principal directions n
for the symmetric matrix
 
3 − √12 √12
1 
[Tij ] =  − √1 9 3 

2 2 2 

2
√1 3 9
2 2 2
Answer
λ(1) = 1, λ(2) = 2 λ(3) = 3
√
^ (1)
n = 12 e1 + e
2^ ^2 − e
^3
√
^ (2)
n = 12 e1 − e
2^ ^2 + e
^3
√
^ (3)
n = − (^ ^3 ) / 2
e2 + e
Solution
3 1 1

−λ − 2√ √
2 2 2 2
1 9 3
det [T − λI] = − 2√ 4 −λ =0

2 4
1 3 9
√
2 2 4 4 − λ
= (λ − 3) (λ − 2) (λ − 1) = 0
The eigenvalues are

λ1 = 1; λ2 = 2; λ3 = 3
The associated principal vectors are determined from Eq 2.70. For λ1 = 1 we have
1 1 1
n1 − √ n2 + √ n3 = 0
2 2 2 2 2
1 5 3
− √ n1 + n2 + n3 = 0
2 2 4 4
1 3 5
√ n1 + n2 + n3 = 0
2 2 4 4
Only two of these are independent. This gives

√
n1 = 2n2 and n2 = −n3
and together with

n21 + n22 + n33 = 1
we have
1 √
^ (1) =
n ^2 − e
e1 + e
2^ ^3
2
1 1 1
− n1 − √ n2 + √ n3 = 0
2 2 2 2 2
1 1 3
− √ n1 + n2 + n3 = 0
2 2 4 4
1 3 1
√ n1 + n2 + n3 = 0
2 2 4 4

√
n1 = − 2n2 and n2 = −n3
and together with

n21 + n22 + n33 = 1
we have
1 √
^ (2) =
n ^2 + e
e1 − e
2^ ^3
2
3 1 1
− n1 − √ n2 + √ n3 = 0
2 2 2 2 2
1 3 3
− √ n1 − n2 + n3 = 0
2 2 4 4
1 3 3
√ n1 + n2 − n3 = 0
2 2 4 4
n1 = 0 and n2 = n3
and together with

n21 + n22 + n33 = 1
we have
1
^ (3) = √ (^
n ^3 )
e2 + e
2
Problem 2.27
For the second-order tensor Cij = ui vj :
(a) Calculate the principal invariants IC , IIC , and IIIC . Reduce your answer to its
simplest form, but leave it in index notation. State the reasons for simplification.

(b) If u = 1 2 3 and v = 5 10 4 write out the matrix form for Cij .
(c) Use the numbers of (b) to validate the answer of (a).
Solution
(a) The invariants are

IC = ui vi
1
IIC = 2
[(ui vi )(uj vj ) − (ui vj )(uj vi )]
1
= 2
[ui uj vi vj − ui uj vj vi ] = 0
IIIC = det ui vj = εijk u1 u2 u3 vi vj vk = 0
|{z} | {z }
skew sym
(b)  
5 10 4
[Cij ] = 10 20 8 
15 30 12
(c)
?
5 + 20 + 12 = 1 2 3 · 5 10 4
Here is a Matlabr check of the numbers:
>> C = [5 10 4; 10 20 8; 15 30 12];
>> I_C = trace(C)
I_C = 37
>> II_C = 0.5*((trace(C))^2 - trace(C*C))
II_C = 0
>> III_C = det(C)
III_C = 0
>>
Problem 2.28
Let D be a constant tensor whose components do not depend upon the coordinates. Show
that
∇ (x · D) = D
î is the position vector.
where x = x i e
Solution
Using indicial notation, we have
(xi dij ) ,k = xi,k dij + xi dij,k = δik dij + 0 = dkj
Problem 2.29
î having a magnitude squared x2 = x21 + x22 + x33 . Determine
Consider the vector x = x i e
(a) grad x ,
(b) grad (x−n ) ,
(c) ∇2 (1/x) ,
(d) div (xn x) ,
(e) curl (xn x), where n is a positive integer.
Answer
(a) xi /x, (b) −nxi /x(n+2) , (c) 0, (d) xn (n + 3), (e) 0.
Solution
(a) Using indicial notation, we have
(grad x)i = x,i
Now
∂ x2 ∂x
= 2x = 2xx,i
∂xi ∂xi
and
∂ (xj xj )
= 2δij xj = 2xi
∂xi
Equating these expressions gives
xi
2xx,i = 2xi or x,i =
x
(b) We have
x,i −n xi −nxi
grad x−n = x−n = −nx−n−1 x,i = −n

i ,i
= = n+2
xn+1 xn+1 x x
(c) For this calculation
∇2 (1/x) = x−1

,ii
Now
∂x−1

∂ ∂ −xi
=
∂xi ∂xi ∂xi x3
from (b). Repeating this process gives
x2

∂ −xi xi,i −3xi δii xi xi 3
=− − xi =− +3 5 =− 3 +3 5 =0
∂xi x3 x3 x5 x3 x x x
(d) The expression div(xn x) is
∂ nxn−1 xi xi
div(xn x) = (xn xi ) = nxn−1 x,i xi + xn xi,i = + δii xn
∂xi x
2
n x
= nx + 3xn = (n + 3) xn
x2
(e) For curl(xn x), where n is a positive integer, we have

xj xk
εijk (xn xk ),j = εijk nxn−1 x,j xk + xn xk,j = εijk nxn−1 + xn δkj

x
= εijk nxn−2 xj xk + xn δkj = 0

by symmetry and skew symmetry
Problem 2.30
If λ and ϕ are scalar functions of the coordinates xi , verify the following vector identities.
Transcribe the left-hand side of the equations into indicial notation and, following the
indicated operations, show that the result is the right-hand side.
(a) v × (∇ × v) = 12 ∇(v · v) − (v · ∇)v

(b) v · u × w = v × u · w
(c) ∇ × (∇ × v) = ∇(∇ · v) − ∇2 v
(d) ∇ · (λ∇ϕ) = λ∇2 ϕ + ∇λ · ∇ϕ
(e) ∇2 (λϕ) = λ∇2 ϕ + 2(∇λ) · (∇ϕ) + ϕ∇2 λ
(f) ∇ · (u × v) = (∇ × u) · v − u · (∇ × v)
Solution
We note that ∇(v · v) is (vi vi ),j = 2vi,j vi in indicial notation.
(a) Now v × (∇ × v) is
εpqi vq εijk vk,j = (δpj δqk − δpk δqj ) vq vk,j = vk vk,p − vq vp,q
or
1
v × (∇ × v) = ∇(v · v) − (v · ∇)v
2
(b) For v · u × w = v × u · w, we have for this scalar quantity
vi εijk uj wk = wk εkij uj vi
or
v·u×w=w·v×u
(c) For ∇ × (∇ × v) = ∇(∇ · v) − ∇2 v, we have for this vector operation
εpqi (εijk vk,j ),q = (δpj δqk − δpk δqj ) vk,jq = vk,pk − vp,qq
or
∇ × (∇ × v) = ∇(∇ · v) − ∇2 v
(d) For this scalar operation ∇ · (λ∇ϕ) = λ∇2 ϕ + ∇λ · ∇ϕ, we have
(λϕ,i ),i = λ,i ϕ,i + λϕ,ii
or
∇ · (λ∇ϕ) = λ∇2 ϕ + ∇λ · ∇ϕ
(e) For this scalar operation ∇2 (λϕ) = λ∇2 ϕ + 2(∇λ) · (∇ϕ) + ϕ∇2 λ, we have
(λϕ),ii = (λ,i ϕ + λϕ,i ),i = (λ,ii ϕ + λ,i ϕ,i + λ,i ϕ,i + λϕ,ii )
= (λϕ,ii + 2λ,i ϕ,i + λ,ii ϕ)
or
∇2 (λϕ) = λ∇2 ϕ + 2(∇λ) · (∇ϕ) + ϕ∇2 λ
(f) For the scalar ∇ · (u × v) = (∇ × u) · v − u · (∇ × v), we have
(εijk uj vk ),i = εijk (uj,i vk + uj vk,i ) = εkij uj,i vk − εjik vk,i uj
or
∇ · (u × v) = (∇ × u) · v − u · (∇ × v)
Problem 2.31
Let the vector v = b × x be one for which b does not depend upon the coordinates. Use
indicial notation to show that
(a) curl v = 2b ,
(b) div v = 0 .
Solution
(a) The expression in indicial notation is
(curl v)i = εijk vk,j = εijk (εkpq bp xq ),j = (δip δjq − δiq δjp ) bp xq,j
= (δip δjq − δiq δjp ) bp δqj = bi δjj − δij bj = (3 − 1) bi = 2bi
(b) The divergence is
div v = vi,i = (εijk bk xj ),i = εijk bk xj,i = εijk bk δji = 0
since two arguments for the permutation symbol are equal.
Problem 2.32
Transcribe the left-hand side of the following equations into indicial notation and verify that
the indicated operations result in the expressions on the right-hand side of the equations
for the scalar ϕ, and vectors u and v.
(a) div (ϕv) = ϕ div v + v · grad ϕ

(b) u × curl v + v × curl u = −(u · grad)v − (v · grad)u + grad(u · v)
(c) div (u × v) = v · curl u − u · curl v
(d) curl(u × v) = (v · grad)u − (u · grad)v + u div v − v div u
(e) curl(curl u) = grad(div u) − ∇2 u
Solution
(a) In indicial notation
div(ϕv) = (ϕvi ),i = ϕ,i vi + ϕvi,i = v · grad ϕ + ϕ div v
(b) Using indicial notation
(u × curl v + v × curl u)i = εijk uj (εkpq vq,p ) + εijk vj (εkpq uq,p )

= (δip δjq − δiq δjp ) (uj vq,p + vj uq,p )
= uq vq,i − up vi,p + vq uq,i − vp ui,p
= [−(u · grad)v − (v · grad)u + grad(u · v)]i
since grad(u · v)i = up vp,i + vp up,i .

(c) We have
div(u × v) = (εijk uj vk ),i = εijk uj,i vk + εijk uj vk,i = εkij uj,i vk − εjik uj vk,i
= v · curl u − u · curl v
(d) For curl(u × v), we have
[curl(u × v)]p = εpqi (εijk uj vk ),q = εipq εijk (uj,q vk + uj vk,q )

= (δpj δqk − δpk δqj ) (uj,q vk + uj vk,q )
= up,q vq − uj,j vp + up vk,k − uq vp,q
= (v · grad)u − v div u + u div v − (u · grad)v
(e) We have
[curl(curl u)]p = εpqi (εijk uk,j ),q = εipq εijk uk,jq = (δpj δqk − δpk δqj ) uk,jq
= uq,pq − up,qq = uq,qp − up,qq = grad(div u)p − ∇2 u p

Problem 2.33
Let the volume V have a bounding surface S with an outward unit normal ni . Let xi be
the position vector to any point in the volume or on its surface. Show that
Z
(a) xi nj dV = δij S ,
S
Z
(b) ∇ (x · x) · n
^ dS = 6V ,
S
Z Z
(c) λw · n^ dS = w · grad λdV, where w = curl v and λ = λ(x) ,
S V
Z
(d) ei × x, e
[^ ^j , n
^ ] dS = 2Vδij where e
î and e
^j are coordinate base vectors.
S
Hint: Write the box product
ei × x, e
[^ ^j , n ei × x) · (^
^ ] = (^ ej × n
^)
and transcribe into indicial notation.
Solution
(a) Applying the divergence theorem gives
Z Z Z
xi nj dS = xi,j dV = δij dV = δij V
S V
(b) Using indicial notation and the divergence theorem, we have

Z Z Z Z Z
(xj xj ),i ni dS = (xj xj ),ii dV = (2xj,i xj ),i dV = 2δij xj,i dV = 2δij δij dV
S V V V V
Z
= 2δii dV = (2) (3) V
V
(c) Again using indicial notation and the divergence theorem, we have
Z Z Z Z
λw · n
^ dS = λwi ni dS = (λwi ),i dV = (λεijk vk,j ),i dV
S
ZS V V
= (λ,i εijk vk,j + λεijk vk,ji ) dV

V
The second term in parentheses is zero by symmetry arguments, and

Z Z Z
λw · n
^ dS = (λ,i εijk vk,j ) dV = w · grad λ dV
S V V
(d) Using the hint, we find

ei × x, e
[^ ^j , n ei × x) · (^
^ ] = (^ ej × n ^p · εqjr nr e
^ ) = εpik xk e ^q = εpik xk εqjr nr δpq
= εpik xk εpjr nr
and
Z Z Z Z
ei × x, e
[^ ^j , n
^ ] dS = εpik xk εpjr nr dS = (εpik εpjr xk ),r dV = εpik εpjr xk,r dV
S
ZS Z V Z V
= εpik εpjr δkr dV = εpir εpjr δkr dV = (δij δrr − δir δjr ) dV
ZV V V
= (3δij − δij ) dV = 2Vδij

V
Problem 2.34
Use Stokes’ theorem to show that upon integrating around the space curve C having a
differential tangential vector dxi that for ϕ(x).
I
ϕ,i dxi = 0
C
Solution
Applying Stokes’ theorem gives
I Z
ϕ,i dxi = εpqi np ϕ,iq dS = 0
C S
by symmetry of the indices.
Problem 2.35
For the position vector xi having a magnitude x, show that x,j = xj /x and therefore,
δij xi xj
(a) x,ij = − 3 ,
x x
−1
3xi xj δij
(b) x ,ij
= − 3 ,
x5 x
2
(c) x,ii = .
x
Solution
xj
From Prob 2.29, we have x,j = where x = xj xj .
x
(a) We have
x
i xi,j xj δij xi xj
x,ij = = − xi x−2 = − 3
x ,j x x x x
(b) For the given expression, we have
xi xi,j xi xj δij
x−1 = −x−2 x,i = −x−2 = 3x−4 x,j xi − x−2

,ij ,j
=3 5 − 3
x ,j x x x
(c) The expression is
x xi,i xi xi 3 x 3 2
i
x,ii = (x,i ),i = = −x−2 x,i xi + =− 2 + =− 2 + =
x ,i x x x x x x
Problem 2.36
Show that for arbitrary tensors A and B, and arbitrary vectors a and b,
(a) (A · a) · (B · b) = a · (AT · B) · b ,
(b) b × a = 21 (B − BT ) · a, if 2bi = εijk Bkj ,
(c) a · A · b = b · AT · a .
Solution
(a) Using indicial notation, we have
Aij aj Bik bk = aj Aij Bik bk
or
(A · a) · (B · b) = a · (AT · B) · b
(b) We can write the cross product of two vectors as

1 1 1
(b × a)i = εijk bj ak = εijk εjpq Bqp ak = (δkp δqi − δkq δpi ) Bqp ak = (Bik ak − Bki ak )
2 2 2
or
1
b×a= (B − BT ) · a
2
(c) We have
ai Aij bj = bj Aij aj
or
a · A · b = b · AT · a
Problem 2.37
Use Eqs 2.42 and 2.43 as necessary to prove the identities
(a) [Aa, Ab, Ac] = (det A)[a, b, c] ,

(b) AT · (Aa × Ab) = (det A)(a × b) ,
for arbitrary vectors a, b, c, and tensor A.
Solution
(a) The triple scalar product is
[Aa, Ab, Ac] = εijk Aip ap Ajq bq Akr cr
From Eq 2.43
εpqr det A =εijk Aip Ajq Akr
and
[Aa, Ab, Ac] = εijk Aip ap Ajq bq Akr cr = εpqr det Aap bq cr = (det A) [a, b, c]
(b) The r component of the vector quantity is

T
A · (Aa × Ab) r = Air εijk Ajp ap Akq bq = εijk Air Ajp Akq ap bq
= εpqr det Aap bq = det A (a × b)r
Problem 2.38
Let ϕ = ϕ(xi ) and ψ = ψ(xi ) be scalar functions of the coordinates. Recall that in the
indicial notation ϕ,i represents ∇ϕ and ϕ,ii represents ∇2 ϕ. Now apply the divergence
theorem, Eq 2.93, to the field ϕψ,i to obtain
Z Z
ϕψ,i ni dS = (ϕ,i ψ,i + ϕψ,ii ) dV .
S V
Transcribe this result into symbolic notation as

Z Z Z
∂ψ
∇ϕ · ∇ψ + ϕ∇2 ψ dV

ϕ∇ψ · n^ dS = ϕ dS =
S S ∂n V
which is known as Green’s first identity. Show also by the divergence theorem that
Z Z
(ϕψ,i − ψϕ,i ) ni dS = (ϕψ,ii − ψϕ,ii ) dV ,
S V
and transcribe into symbolic notation as

Z Z
∂ψ ∂ϕ
ϕ∇2 ψ − ψ∇2 ϕ dV

ϕ −ψ dS =
S ∂n ∂n V
which is known as Green’s second identity.
Solution
The divergence theorem applied to the field is
Z Z Z
ϕψ,i ni dS = (ϕψ,i ),i dV = (ϕ,i ψ,i + ϕψ,ii ) dV
S V V
Symbolically, this is
Z Z Z
∂ψ
ϕψ,i ni dS = ϕ∇ψ · n
^ dS = ϕ dS
S S S ∂n
and Z Z
∇ϕ · ∇ψ + ϕ∇2 ψ dV

(ϕ,i ψ,i + ϕψ,ii ) dV =
V V
The result is Z Z
∂ψ
∇ϕ · ∇ψ + ϕ∇2 ψ dV

ϕ dS =
S ∂n V
Again applying the divergence theorem gives

Z Z
(ϕψ,i − ψϕ,i ) ni dS = (ϕψ,i − ψϕ,i ),i dV
S V
Z
= (ϕ,i ψ,i + ϕψ,ii − ψ,i ϕ,i − ψϕ,ii ) dV
ZV Z
ϕ∇2 ψ − ψ∇2 ϕ dV

= (ϕψ,ii − ψϕ,ii ) dV =
V V
and
Z Z Z
(ϕψ,i − ψϕ,i ) ni dS = (ϕψ,i ni − ψϕ,i ni ) dS = (ϕ∇ψ · n
^ + ψ∇ϕ · n
^ ) dS
S S S
Z
∂ψ ∂ϕ
= ϕ +ψ dS
S ∂n ∂n
This gives Z Z
∂ψ ∂ϕ
ϕ∇2 ψ − ψ∇2 ϕ dV

ϕ +ψ dS =
S ∂n ∂n V
Chapter 3 Solutions
Problem 3.1
At a point P, the stress tensor relative to axes Px1 x2 x3 has components tij . On the area
n1 )
^ 1 , the stress vector is t(^
element dS(1) having the unit normal n , and on area element
(2) n
^ 2 the stress vector is t 2 . Show that the component of t(^
(^ ) n1 )
dS with normal n in the
n
(^ )
^ 2 is equal to the component of t 2 in the direction of n ^1.
direction of n
Solution
For the traction on the plane with normal n
^ 2 in the direction n
^1
n2 )
(^ (1) (2) (1) (2) (1) (2) (1) n1 )
(^ (2)
ti ni = tji nj ni = tij ni nj = tji ni nj = ti ni
where the symmetry of the stress and dummy indices substitution was used.
Problem 3.2
Verify the result established in Problem 3.1 for the area elements having normals
1
^1 =
n e1 + 3^
(2^ e2 + 6^
e3 )
7
1
^ 2 = (3^
n e1 − 6^
e2 + 2^
e3 )
7
if the stress matrix at P is given with respect to axes Px1 x2 x3 by
 
35 0 21
[tij ] =  0 49 0  .
21 0 14
Solution
For t(^
n1 )
, we have
 n )
(^
 
t1 1
    
35 0 21 2 28
 (^n )  1
 t2 1  =  0 49 0   3  =  21 
n )
(^
t3 1 21 0 14 7 6 18
and
n1 ) 1 6
t(^ ·n
^ 2 = (26^
e1 + 21^ e3 ) ·
e2 + 18^ e1 − 6^
(3^ e2 + 2^
e3 ) = −
7 7
For t(^
n2 )
, we have
 n )
(^
 
t1 2
    
35 0 21 3 21
 (^n )  1
 t2 2  =  0 49 0   −6  =  −42 
n )
(^
t3 2 21 0 14 7 2 13
31
and
n2 ) 1 6
t(^ ·n
^ 1 = (21^
e1 − 42^ e3 ) ·
e2 + 13^ e1 + 3^
(2^ e2 + 6e3 ) = −
7 7
Problem 3.3
The stress tensor at P relative to axes Px1 x2 x3 has components in MPa given by the matrix
representation  
t11 2 1
[tij ] =  2 0 2 
1 2 0
where t11 is unspecified. Determine a direction n ^ at P for which the plane perpendicular to
n)
^ will be stress-free, that is, for which t(^
n = 0 on that plane. What is the required value
of t11 for this condition?
Answer
1
^=
n ^2 − 2^
e1 − e
(2^ e3 ), t11 = 2 MPa
3
Solution
For t(^
n)
= 0, we have
      
t11 2 1 n1 t11 n1 + 2n2 + n3 0
      
 2 0 2   n2  =  2n1 + 2n3 = 0 
      
1 2 0 n3 n1 + 2n2 0
This gives
n1 = −n3 and n1 = −2n2
and since 2
1
ni ni = n21 + − n1 + (−n1 )2 = 1
2
we have
2 1 2
n1 = ± n2 = ∓ n3 = ∓
3 3 3
Finally, we find
2 1 2
t11 −2 − =0
3 3 3
and
t11 = 2
Problem 3.4
The stress tensor has components at point P in ksi units as specified by the matrix
 
−9 3 −6
[tij ] =  3 6 9  .
−6 9 −6
Determine:
(a) the stress vector on the plane at P whose normal vector is
1
^=
n e1 + 4^
(^ e2 + 8^
e3 ) ,
9
(b) the magnitude of this stress vector,

(c) the component of the stress vector in the direction of the normal,
(d) the angle in degrees between the stress vector and the normal.
Answer
n)
(a) t(^ e + 11^
= −5^ e2 − 2^
e3
n)
(^
√ 1
(b) t = 150
23
(c)
9
(d) 77.96◦
Solution
(a) The traction vector is
     
−9 3 −6 1 −5
 3 1
6 9   4  =  11 
−6 9 −6 9 8 −2
(b) The magnitude of the stress vector is

2
t n) = t(^
n) n)
= (−5)2 + (11)2 + (−2)2 = 150
(^
· t(^

and
t n) = 12.25
(^
(c) The component of the stress vector in the direction of the normal is
n) 1 23
t(^ ·n
^ = (−5^
e1 + 11^ e3 ) ·
e2 − 2^ e1 + 4^
(^ e2 + 8^
e3 ) =
9 9
(d) The angle between the stress vector and the normal is
23
n)
|^
n| cos θ = (12.25) (1) cos θ =
n)
t(^ ^ = t(^
·n
9
and
cos θ = 0.209 or θ = 77.96◦
Problem 3.5
Let the stress tensor components at a point be given by tij = ±σ0 ni nj where σ0 is a
positive constant. Show that this represents a uniaxial state of stress having a magnitude
±σ0 and acting in the direction of ni .
Solution
The stress vector is
n)
(^
ti = tij nj = ±σ0 ni nj nj = ±σ0 ni
since nj nj = 1
Problem 3.6
Show that the sum of squares of the magnitudes of the stress vectors on the coordinate
planes is independent of the orientation of the coordinate axes, that is, show that the sum
e1 ) (^
(^ e ) e2 ) (^
(^ e ) e3 ) (^
(^ e )
ti ti 1 + ti ti 2 + ti ti 3
is an invariant.
x3 ^3 )
(e
ti
^2 )
(e
ti
^ )
(e
ti 1
x2
x1
Solution
The sum is
e1 ) (^
(^ e ) e2 ) (^
(^ e ) e3 ) (^
(^ e )
ti ti 1 + ti ti 2 + ti ti 3 = ti1 n1 ti1 n1 + ti2 n2 ti2 n2 + ti3 n3 ti3 n3
= ti1 ti1 + ti2 ti2 + ti3 ti3 = tij tij
since n1 n1 = n2 n2 = n3 n3 = 1
Problem 3.7
With respect to axes Ox1 x2 x3 the stress state is given in terms of the coordinates by the
matrix
x22
 
x1 x2 0
[tij ] =  x22 x2 x3 x23 
2
0 x3 x3 x1
Determine
(a) the body force components as functions of the coordinates if the equilibrium
equations are to be satisfied everywhere
(b) the stress vector at point P(1, 2, 3) on the plane whose outward unit normal
makes equal angles with the positive coordinate axes.
Answer:
−3x2 −3x3 −x1
(a) b1 = , b2 = , b3 =
ρ ρ ρ
n) e1 + 19^
(6^ e2 + 12ê3 )
(b) t(^ = √
3
Solution
(a) The equilibrium equations are tij,j + ρbi = 0
For i = 1
t11,1 + t12,2 + t13,3 + ρb1 = 0
x2 + 2x2 + 0 + ρb1 = 0
and
3x2
b1 = −
ρ
For i = 2
t21,1 + t22,2 + t23,3 + ρb2 = 0
0 + x3 + 2x3 + ρb2 = 0
and
3x3
b2 = −
ρ
For i = 3
t31,1 + t32,2 + t33,3 + ρb3 = 0
0 + 0 + x1 + ρb3 = 0
and
x1
b3 = −
ρ
(b) The stress at the point P (1, 2, 3) is
 
2 4 0
[tij ] =  4 6 9 
0 9 3
The direction is n
^ = √1 (^
3
^3 ). The stress vector is
^2 + e
e1 + e
       
n)
(^
t1 2 4 0 1 6
 t n)  =  4 6 9  √1  1  = √1  19 
 (^       
 2    3  3  
n)
(^
t3 0 9 3 1 12
Problem 3.8
Relative to the Cartesian axes Ox1 x2 x3 a stress field is given by the matrix
 
2
2 3 2

 1 − x1 x2 + 3 x2 − 4 − x2 x1 0 

[tij ] =  2
1 3

 .
− 4 − x2 x1 − x2 − 12x2 0
3
 
 
0 0 3 − x21 x2
(a) Show that the equilibrium equations are satisfied everywhere for zero body forces.
(b) Determine the stress vector at the point P(2, −1, 6) of the plane whose equation
is 3x1 + 6x2 + 2x3 = 12.
Answer
n) 1
(b) t(^ = e1 − 40^
(−29^ e2 + 2^
e3 )
7
Solution
(a) For zero body force bi = 0, the equilibrium equations are
tij,j = 0
For i = 1
t11,1 + t12,2 + t13,3 + ρb1 = 0
−2x1 x2 + 2x1 x2 + 0 + 0 = 0
For i = 2
t21,1 + t22,2 + t23,3 + ρb2 = 0
− 4 − x22 + −x22 + 4 + 0 + 0 = 0

For i = 3
t31,1 + t32,2 + t33,3 + ρb3 = 0
0+0+0+0=0
(b) At the point P (2, −1, 6), the stress is
 
7
−6 0 
 3



[tij ] =  11 
 −6 − 0
3

 
 
0 0 1
The normal to the plane is n

^ = 17 (3^
e1 + 6^ e3 ) and the stress vector is
e2 + 2^
       
n)
(^ 7
 t1   3 −6 0   3   −29 
       
   1  1 
n)
 t(^  =  −6 − 11 0   6  =  −40 
 2   3 7  7 
       
       
n)
(^
t3 0 0 1 2 2
Problem 3.9
The stress components in a circular cylinder of length L and radius r are given by
 
Ax2 + Bx3 Cx3 −Cx2
[tij ] =  Cx3 0 0  .
−C2 0 0
(a) Verify that in the absence of body forces the equilibrium equations are satisfied.
(b) Show that the stress vector vanishes at all points on the curved surface of the
cylinder.
x3
L
x2
r
x1
Solution
(a) The equilibrium equations are
tij,j = 0
For i = 1
t11,1 + t12,2 + t13,3 = 0
0+0+0=0
For i = 2
t21,1 + t22,2 + t23,3 = 0
0+0+0=0
For i = 3
t31,1 + t32,2 + t33,3 = 0
0+0+0=0
x x
2
(b) The normal to the cylindrical surface is n
^= ^2 + 3 e
e ^3 , the stress vector is
r r
1
      
n)
(^
t1 Ax2 + Bx3 Cx3 −Cx2 0 (Cx3 x2 − Cx2 x3 )
    x   r 
 2  
 t2n)  = 
 (^
=
  
Cx3 0 0  0
 r  

x3
   
n)
(^
t3 −C2 0 0 0
r
Problem 3.10
Rotated axes Px01 x02 x03 are obtained from axes Px1 x2 x3 by a right-handed rotation about the
line PQ that makes equal angles with respect to the Px1 x2 x3 axes (see sketch). Determine
the primed stress components for the stress tensor in (MPa)
 
3 0 6
[tij ] =  0 0 0 
6 0 −3
if the angle of rotation is (a) 120◦ , or (b) 60◦ .

x3
Q
β
β
P
x1 β x2
Answer
   
h i 0 0 0 h i 1 −5 10 10
(a) t0ij =  0 −3 6 MPa, (b) t0ij =  10 −11 −2 MPa
0 6 3 3 10 −2 16
Solution
(a) The stress is
t0ij = aiq ajm tqm
The transformation matrix for 120◦ about n^ = √1 (^
3
^3 ) is
^2 + e
e1 + e
 
0 1 0
[aij ] =  0 0 1 
1 0 0
This gives
     
0 1 0 3 0 6 0 0 1 0 0 0
t0ij

= 0 0 1  0 0 0  1 0 0 = 0 −3 6 
1 0 0 6 0 −3 0 1 0 0 6 3
(b) For a rotation of 60◦ , the rotation matrix is
 
2 2 −1
1
[aij ] =  −1 2 2 
3 2 −1 2
This gives
    
2 2 −1 3 0 6 2 −1 2
1 1
t0ij

= −1 2 2  0 0 0   2 2 −1 
3 2 −1 2 6 0 −3 3 −1 2 2
   
−15 30 30 −5 10 10
1 1
= 30 −33 −6  = 10 −11 −2 
9 30 −6 48 3 10 −2 16
Problem 3.11
At the point P, rotated axes Px01 x02 x03 are related to the axes Px1 x2 x3 by the transformation
matrix  √ √ 
a√ 1 − 3 1 + √3
1
[aij ] = 1 + √3 b√ 1− 3 
3
1− 3 1+ 3 c
h i
where a, b, and c are to be determined. Determine t0ij if the stress matrix relative to
axes Px1 x2 x3 is given in MPa by
 
1 0 1
[tij ] =  0 1 0  .
1 0 1
Answer
 √ √ 
h i 1 11 + 2√ 3 5+ 3 −1√
t0ij =  5 + 3 5√ 5− √ 3  MPa
9
−1 5− 3 11 − 2 3
Solution
From the orthogonality condition,
aij aik = δjk or aji aki = δjk
This results in
 √ √  √ √   
a√ 1− 3 1 + √3 a√ 1+ 3 1 − √3 1 0 0
1
1 + √3 b√ 1 − 3   1 − √3 b√ 1+ 3 = 0 1 0 
9 0 0 1
1− 3 1+ 3 c 1+ 3 1− 3 c
Thus
√ 2 √ 2
a2 + 1 − 3 + 1 + 3 = 9
√ 2 √ 2
1 + 3 + b2 + 1 − 3 = 9
√ 2 √ 2
1 − 3 + 1 + 3 + c2 = 9
and
a2 = b2 = c2 = 1 or a = b = c = ±1
Also
√ √ √ √
a 1+ 3 +b 1− 3 + 1+ 3 1− 3 =0
√ √ √ √
a 1− 3 +b 1+ 3 + 1− 3 1+ 3 =0
or
a+b=2
Thus
a=b=1
Finally √ √ √ √
1− 3 1+ 3 +b 1+ 3 +c 1− 3 =0
and
c=1
The stress is t0ij = aiq ajm tqm

 √ √    √ √ 
1√ 1 − 3 1 + √3 1 0 1 1√ 1+ 3 1 − √3
1 1
t0ij

=  1 + √3 1√ 1 − 3  0 1 0   1 − √3 1√ 1+ 3 
3 1 0 1 3
1− 3 1+ 3 1 1+ 3 1− 3 1
 √ √ 
11 + 2√ 3 5 + 3 −1√
1
=  5+ 3 5√ 5− √ 3 
9
−1 5 − 3 11 − 2 3
Problem 3.12
The stress matrix referred to axes Px1 x2 x3 is given in ksi by
 
14 0 21
[tij ] =  0 21 0 
21 0 7
Let rotated axes Px01 x02 x03 be defined with respect to axes Px1 x2 x3 by the table of base
vectors
^1
e ^2
e ^3
e
^01
e 2/7 3/7 6/7
^02
e 3/7 −6/7 2/7
^03
e 6/7 2/7 −3/7
(a) Determine the stress vectors on planes at P perpendicular to the primed axes;
e01 ) e02 ) e03 )
determine t(^ , t(^ , and t(^ ^1 , e
in terms of base vectors e ^2 , and e
^3 .
(b) Project each of the stress vectors
h iobtained in (a) onto the primed axes to deter-
mine the nine components of t0ij .
(c) Verify the result obtained in (b) by a direct application of Eq 3.33 of the text.
Answer
 
h i 1 143 36 114
t0ij =  36 166 3  ksi
7 114 3 −15
Solution
(a) The stress vectors are
(e^0 )
       
t1 1 14 0 21 2 22
1
(e^0 )
     
t2 1 = 0   3 = 9
       
0 21
7
 
     
(e^0 )
t3 1 21 0 7 6 12
0
or t(e^1 ) = 22^
e1 + 9^
e2 + 12^ e3 ;
 ^0  
(e )
    
t 2 14 0 21 3 12
 1 0   1   
 (e^2 )  
= 0 =
    
 t2 21 0 −6 −18 
7
   
 0
    
(e^ )
t3 2 21 0 7 2 11
0
or t(e^2 ) = 12^
e1 − 18^
e2 + 11ê3 ;
 ^0  
(e )
    
t 3 14 0 21 6 3
 1 0   1   
 (e^3 )  
= 0 =
    
 t2 21 0   2 6 
 0
  7   
( )
^
e
t3 3 21 0 7 −3 15
0
or t(e^3 ) = 3^
e1 + 6^ e3 ;
e2 + 15^
(b) Projecting these stress vectors onto the e0i axes gives
0
t(e^1 ) = 22ê1 + 9^ e2 + 12^ e3
22 9 12
e01 + 3ê02 + 6ê03 + e01 − 6^
e02 + 2^
e03 + e01 + 2^
e02 − 3v03

= 2^ 3^ 6^
7 7 7
1 0 0 0

= e1 + 36^
143^ e2 + 114ê3
7
0
t(e^2 ) = 12ê1 − 18ê2 + 11ê3
12 18 11
e01 + 3^
e02 + 6^
e03 − e01 − 6^
e02 + 2^
e03 + e01 + 2^
e02 − 3^
e03

= 2^ 3^ 6^
7 7 7
1 0 0 0

= e1 + 166^
36^ e2 + 3ê3
7
0
t(e^3 ) = 3^
e1 + 6ê2 + 15^ e3
3 6 15
e01 + 3^
e02 + 6ê03 + e01 − 6^
e02 + 2^
e03 + e01 + 2^
e02 − 3^
e03

= 2^ 3^ 6^
7 7 7
1 0 0 0

= 114ê1 + 3ê2 − 15^ e3
7
Thus  
143 36 114
1
t0ij

= 36 166 3 
7 114 3 −15
(c) Transforming the stresses by t0ij = aiq ajm tqm , we obtain
    
2 3 6 14 0 21 2 3 6
1 1
t0ij

=  3 −6 2   0 21 0   3 −6 2 
7 6 2 −3 21 0 7 7 6 2 −3
 
143 36 114
1
=  36 166 3 
7 114 3 −15
The results are identical.
Problem 3.13
At point P, the stress matrix is given in MPa with respect to axes Px1 x2 x3 by
   
6 4 0 2 1 1
Case 1: [tij ] =  4 6 0  Case 2: [tij ] =  1 2 1 
0 0 −2 1 1 2
Determine for each case
(a) the principal stress values,

(b) the principal stress directions.
Answer
(a) Case 1: σ(1) = 10 MPa, σ(2) = 2 MPa, σ(3) = −2 MPa

Case 2: σ(1) = 4 MPa, σ(2) = σ(3) = 1 MPa
^ +e
e ^ ^ −e
e ^
(b) Case 1: n ^ (2) = ± 1 √ 2 , n
^ (1) = ± 1 √ 2 , n ^ (3) = ∓^ e3
2 2
^ +e
e ^ +e ^3 (2) −^ ^
e1 + e −^ ^ + 2^
e1 − e e3
^ (1) = 1 √2
Case 2: n ,n^ = √ 2, n ^ (3) = √2
3 2 6
Solution
Case 1: The principal values are found from
 
6−σ 4 0 6−σ 4 0
det

 4 6−σ 0  = 4 6−σ 0 =0

0 0 −2 − σ 0 0 −2 − σ
= (−2 − σ) σ2 − 12σ + 20 = 0

= (−2 − σ) (σ − 10) (σ − 2) = 0
The results are

σ(1) = 10, σ(2) = 2, σ(3) = −2
For σ(1) = 10, the principal direction is found from
   (1)   
6 − 10 4 0 n1 0
(1) 
 4 6 − 10 0  n2  =  0 
0 0 −2 − 10 (1) 0
n3
^ (1) · n
and n ^ (1) = 1. The result is
(1) (1)
−4n1 + 4n2 = 0
(1) (1)
+4n1 − 4n2 = 0
(1)
−12n3 = 0
or
(1) (1) (1)
n1 = n2 and n3 = 0
Using the unit vector relation

2 2 2 2 2
(1) (1) (1) (1) (1)
n1 + n2 = n1 + n1 = 2 n1 =1
and
(1) (1) 1 1 1
n1 = n2 = ± √ ^ (1) = ± √ e
n ^1 ± √ e^2
2 2 2
    
(2)
6−2 4 0 n 0
  1   
  (2)
= 0 
   
 4 6−2 0   n2
    
(2)
0 0 −2 − 2 n3 0
^ (2) · n
(2) (2)
4n1 + 4n2 = 0
(2) (2)
4n1 + 4n2 = 0
(2)
−4n3 = 0
or
(2) (2) (2)
n1 = −n2 and n3 = 0
2 2 2 2 2
(2) (2) (2) (2) (2)
n1 + n2 = n1 + −n1 = 2 n1 =1
and
(2) (2) 1 1 1
n1 = −n2 = ± √ ^ (2) = ± √ e
n ^1 ∓ √ e^2
2 2 2
For σ(3) = −2, the principal direction is found from
    
(3)
6+2 4 0 n 0
  1   
  (3)
= 0 
   
 4 6+2 0   n2
    
(3)
0 0 −2 + 2 n3 0
^ (3) · n
(3) (3)
8n1 + 4n2 = 0
(3) (3)
4n1 + 8n2 = 0
(3)
0n3 = 0
or
(3) (3)
n1 = n2 = 0
2
(3)
n3 = ±1
and
(3)
n3 = ±1 ^ (3) = ±^
n e3
Case 2: The principal values are found from
 
2−σ 1 1 2−σ 1 1
det

 1 2−σ 1  = 1 2−σ 1 =0

1 1 2−σ 1 1 2−σ
= (σ − 1) −σ2 + 5σ − 4 = 0

= (σ − 1) (σ − 1) (σ − 4) = 0
The results are

σ(1) = 4, σ(2) = σ(3) = 1
    
(1)
2−4 1 1 n 0
  1   
  (1)
= 0 
   
 1 2−4 1   n2
    
(1)
1 1 2−4 n3 0
^ (1) · n
(1) (1) (3)
−2n1 + 1n2 + n3 = 0
(1) (1) (1)
n1 − 2n2 + n3 = 0
(1) (1) (1)
n1 + n2 − 2n3 = 0
or
(1) (1) (1)
n1 = n2 = n3
2 2 2 2 2 2 2
(1) (1) (1) (1) (1) (1) (1)
n1 + n2 + n3 = n1 + n1 + n1 = 3 n1 =1
and
(1) (1) (1) 1 1 1 1
n1 = n2 = n3 = ± √ ^ (1) = ± √ e
n ^1 ± √ e^2 ± √ e^3
3 3 3 3
    
(2)
2−1 1 1 n 0
  1   
  (2)
= 0 
   
 1 2−1 1   n2
    
(2)
1 1 2−1 n3 0
^ (2) · n
(2) (2) (2)
n1 + n2 + n3 = 0
(2) (2) (2)
n1 + n2 + n3 = 0
(2) (2) (2)
n1 + n2 + n3 = 0
This is one equation in three unknowns. Select n ^ (2) to be perpendicular to n(1) Using the
unit vector relation and
^ (2) · n
n ^ (1) = 0
we can choose
(2) (2) (2)
n1 = −n2 and n3 = 0
and 2 2 2 2 2
(2) (2) (2) (2) (2)
n1 + n2 = n1 + −n1 = 2 n1 =1
This gives
(2) (2) 1 1 1
n1 = −n2 = ± √ ^ (2) = ∓ √ e
n ^1 ± √ e^2
2 2 2
For σ(3) = σ(2) = 1, the multiple root requires that we find n^ (3) that is perpendicular to
(1) (2)
the principal directions n^ and n ^ . This is found from

e ^1 ^2 e
e ^3

1
^ (3) = n
n ^ (1) × n
^ (2) = √1 √1 √1 = √ (−^ e1 − 2ê2 + 2ê3 )

3 3 3 6
− √1 √1

2 2
0
Problem 3.14
When referred to principal axes at P, the stress matrix in ksi units is
 
∗ 2 0 0
tij =  0 7 0 
0 0 12
If the transformation matrix between the principal axes and axes Px1 x2 x3 is
− 53 − 54
 
1
1 
[aij ] = √ a21 a22 a23 
2 − 35 −1 − 54
where a21 , a22 , and a23 are to be determined, calculate [tij ].
Answer
 
7 3 0
[tij ] =  3 7 4  ksi
0 4 7
Solution
The orthogonality condition requires
aij aik = δjk or aji aki = δjk
or     
− 35 1 − 45 − 35 a21 − 53 1 0 0
1    
−1  =  0
    
 a a22 a23   1 a22 1 0 
2  21    
− 53 −1 − 45 − 45 a23 − 54 0 0 1
This gives
3 4
− a21 + a22 − a23 = 0
5 5
3 4
− a21 − a22 − a23 = 0
5 5
a21 + a22 + a223 = 2
2 2
The first two equations give
a22 = 0, 3a21 = −4a22
This together with the final equation yields

2 √ √
3 4 2 3 2
a221 + − a21 =2 or a21 =± and a23 =∓
4 5 5

 
− 53 1 − 45
1  
[aij ] = √  8
− 65 
 
5 0
2 
− 53 −1 − 45
The stress is
[tij ] = [aij ]T t∗ij [aij ]

    
− 35 8
5 − 35 2 0 0 − 35 1 − 54
1    1  
=√  1 0  √  85 − 56 
    
0 −1   0 7 0
2   2 
− 45 − 65 4
−5 0 0 12 − 35 −1 −5 4
Problem 3.15
The stress matrix in MPa when referred to axes Px1 x2 x3 is
 
3 −10 0
[tij ] =  −10 0 30  .
0 30 −27
Determine
(a) the principal stresses, IT , IIT , IIIT ,

(b) the principal stress directions.
Answer
(a) IT = 23 MPa, IIT = 0 MPa, IIIT = −47 MPa

^ (1) = −0.394e
(b) n ^1 + 0.788e^2 + 0.473e^3
^ (2) = 0.931e
n ^1 + 0.274e^2 + 0.304e^3
^ (3) = 0.110e
n ^1 + 0.551e^2 − 0.827e^3
Solution
 
3 − σ −10 0 3−σ −10 0
det

 −10 0 − σ 30  = −10 0−σ 30 =0

0 30 −27 − σ 0 30 −27 − σ
= σ3 + 24σ2 − 1081σ = 0
= σ (σ − 23) (σ + 47) = 0
The results are

σ(1) = 23, σ(2) = 0, σ(3) = −47
(b) For σ(1) = 23, the principal direction is found from
    
(1)
3 − 23 −10 0 n 0
  1   
  (1)
= 0 
   
 −10 0 − 23 30   n2
    
(1)
0 30 −27 − 23 n3 0
^ (1) · n
(1) (1)
−20n1 − 10n2 = 0
(1) (1) (1)
−10n1 − 23n2 + 30n3 = 0
(1) (1)
30n2 − 50n3 = 0
or
(1) 1 (1) (1) 3 (1)
n1 = − n2 and n3 = n
2 5 2
2 2 2 1 2 2 3 2
(1) (1) (1) (1) (1) (1)
n1 + n2 + n3 = − n2 + n2 + n2 =1
2 5
and
(1) (1) (1)
n2 = ±0.788 n1 = ∓0.394 n3 = ±0.473
^ (1) = ∓0.394^
n e1 ± 0.788^
e2 ± 0.473^
e3

    
(2)
3 − 0 −10 0 n 0
  1   
  (2)
= 0 
   
 −10 0 − 0 30   n2
    
(2)
0 30 −27 − 0 n3 0
^ (2) · n
(2) (2)
3n1 − 10n2 = 0
(2) (2)
−10n1 + 30n3 = 0
(2) (2)
30n2 − 27n3 = 0
or
(2) (2) (2) (2)
10n2 = 3n1 and 3n3 = n1
2 2 2 2 2 2
(2) (2) (2) (2) (2) (2)
n1 + n2 + n3 = n1 + 0.3n1 + 3n1 =1
and
(2) (2) (2)
n1 = ±0.913 n2 = ±0.274 n3 = ±0.304
^ (2) = ±0.913^
n e1 ± 0.274^
e2 ± 0.304^
e3
    
(3)
3 − (−47) −10 0 n1 0
    
  (3)
= 0 
   
 −10 0 − (−47) 30   n2
    
(3)
0 30 −27 − (−47) n3 0
^ (3) · n
(3) (3)
50n1 − 10n2 = 0
(3) (3) (3)
−10n1 + 47n2 + 30n3 = 0
(3) (3)
30n2 + 20n3 = 0
or
(3) 1 (3) (3) 3 (3)
n1 = n and n3 = − n2
5 2 2
2 2 2 1 2 2 3 2
(3) (3) (3) (3) (3) (3)
n1 + n2 + n3 = n2 + n2 + − n2 =1
5 2
and
(3) (3) (3)
n2 = ±0.110 n1 = ±0.551 n3 = ∓0.827
^ (3) = ±0.110^
n e1 ± 0.551^
e2 ∓ 0.827^
e3
Problem 3.16
At point P, the stress matrix relative to axes Px1 x2 x3 is given in MPa by
 
5 a −a
[tij ] =  a 0 b 
−a b 0
where a and b are unspecified. At the same point relative to axes Px∗1 x∗2 x∗3 the matrix is
 
∗ σ(1) 0 0
tij =  0 2 0  .
0 0 σ(2)
If the magnitude of the maximum shear stress at P is 5.5 MPa, determine σI and σIII .
Answer
σ(1) = 7 MPa, σ(3) = −4 MPa
Solution
The invariance of the trace gives
tr T = tii = 5 + 0 + 0 = σ(1) + 2 + σ(3)
or
σ(1) + σ(3) = 3
The maximum shear stress is
σ(1) − σ(3)
= 5.5
2
or
σ(1) − σ(3) = 11
The result is
σ(1) = 7 and σ(3) = −4
Problem 3.17
The state of stress at point P is given in ksi with respect to axes Px1 x2 x3 by the matrix
 
1 0 2
[tij ] =  0 1 0  .
2 0 −2
Determine
(a) the principal stress values and principal stress directions at P,

(b) the maximum shear stress value at P,
^ = ni e
(c) the normal n î to the plane at P on which the maximum shear stress acts.
Answer
(a) σ(1) = 2 ksi, σ(2) = 1 ksi, σ(3) = –3 ksi

e1 + e
2^ ^ e1 + 2^
−^ e3
^ (1) =
n √ 3, n ^ (2) = e ^ (3) =
^2 , n √
5 5
(b) (σS )max = ± 2.5 ksi
^ + 3^
e e3
^ = 1√
(c) n
10
Solution
 
1−σ 0 2 1−σ 0 2
det 

0 1−σ 0 =
0 1−σ 0 =0

2 0 −2 − σ 2 0 −2 − σ
= (1 − σ) σ2 + σ − 6 = 0

= (1 − σ) (σ + 3) (σ − 2) = 0
The results are

σ(1) = 2, σ(2) = 1, σ(3) = −3
    
(1)
1−2 0 2 n 0
  1   
  (1)
= 0 
   
 0 1−2 0   n2
    
(1)
2 0 −2 − 2 n3 0
^ (1) · n
(1) (1)
−n1 + 2n3 = 0
(1)
−n2 = 0
(1) (1)
2n1 − 4n3 = 0
or
(1) (1) (1)
n1 = 2n3 and n2 = 0
2 2 2 2 2
(1) (1) (1) (1) (1)
n1 + n3 = 2n3 + n3 = 5n3 =1
and
(1) 1 (1) 2 (1)
n3 = ± √ n1 = ± √ n2 = 0
5 5
or
2 1
^ (1) = ± √ e
n ^1 ± √ e^3
5 5
    
(2)
1−1 0 2 n 0
  1   
  (2)
= 0 
   
 0 1−1 0   n2
    
(2)
2 0 −2 − 1 n3 0
^ (2) · n
(2)
2n3 = 0
(2) (2)
2n1 − 3n3 = 0
or
(2) (2)
n3 = n1 = 0
2 2 2 2
(2) (2) (2) (2)
n1 + n2 + n3 = n2 =1
and
(2) (2) (2)
n1 = 0 n2 = ±1 n3 = 0
or
^ (2) = ±^
n e2
    
(3)
1 − (−3) 0 2 n 0
  1   
  (3)
= 0 
   
 0 1 − (−3) 0   n2
    
(3)
2 0 −2 − (−3) n3 0
^ (3) · n
(3) (3)
4n1 + 2n3 = 0
(3)
4n2 = 0
(3) (3)
2n1 + n3 = 0
or
(3) (3) (3)
n2 = 0 and n3 = −2n1
2 2 2 2 2
(3) (3) (3) (3) (3)
n1 + n2 + n3 = n1 + −2n1 =1
and
(3) 1 (3) (3) 2
n1 = ± √ n2 = 0 n3 = ∓ √ ,
5 5
or
1 2
^ (3) = ± √ e1 ∓ √ e3
n
5 5
^ (1) and n
The result can be found through the cross product of n ^ (2) . This gives the negative
of the result for a right-hand coordinate system.
(b) The maximum shear stress is
σ(1) − σ(3) 2 − (−3)
(σS )max = = = 2.5
2 2
(c) The plane of maximum shear stress is
^ (1) + n
n ^ (3) ^3 −^
e1 + e
2^ e1 + 2^
e3 ^1 + 2^
e e3
^=
n √ = √ + √ = √
2 10 10 10
^ (3) for a right-hand coordinate system was used.

Here n
Problem 3.18
The stress tensor at P is given with respect to Ox1 x2 x3 in matrix form with units of MPa
by  
4 b b
[tij ] =  b 7 2 
b 2 4
where b is unspecified. If σ(3) = 3 MPa and σ(1) = 2σII , determine
(a) the principal stress values,

(b) the value of b,
(c) the principal stress direction of σ(2) .
Answer
(a) σ(1) = 8 MPa, σ(2) = 4 MPa, σ(3) = 3 MPa
^ (2) = e
(b) b = 0, (c) n ^1
Solution
(a) The principal stress values are found from the invariance of the trace of the stress
tensor.
tr T = t11 + t22 + t33 = 4 + 7 + 4 = 15 = σ(1) + σ(2) + σ(3) = 2σ(2) + σ(2) + 3
and
σ(1) = 8 σ(2) = 4 σ(3) = 3
(b) To determine the value of b, we can use the invariance of the determinate.

4 b b
det T = b 7 2 = σ(1) σ(2) σ(3) = 8 · 4 · 3 = 96

b 2 4
or
4 (24) − b (4b − 2b) + b (2b − 7b) = 96 − 2b2 − 5b2 = 96
This gives b = 0.
(c) To find the principal direction for σ(2) = 4, we have
    
(2)
4−4 0 0 n 0
  1   
  (2)
= 0 
   
 0 7−4 2   n2
    
(2)
0 2 4−4 n3 0
This gives
(2) (2)
3n2 + 2n3 = 0
(2)
2n2 = 0
and
(2) (2)
n2 = n3 = 0
The unit vector condition gives
2
(2) (2)
n1 =1 or n1 = ±1
and
^ (2) = ±^
n e1
Problem 3.19
The state of stress at P, when referred to axes Px1 x2 x3 is given in ksi units by the matrix
 
9 3 0
[tij ] =  3 9 0 
0 0 18
Determine
(a) the principal stress values at P,

^ ∗ = ni e
(b) the unit normal n ^∗i of the plane on which σN = 12 ksi and σS = 3 ksi.
Answer
(a) σ(1) = 18 ksi, σ(2) = 12 ksi, σ(3) = 6 ksi

√ ∗
∗
^∗1 + 6^
e e +e ^∗3
^ =
(b) n √2
2 2
Solution
(a) The principal stresses are found by
 
9−σ 3 0 9−σ 3 0
det

 3 9−σ 0  = 3 9−σ 0 =0

0 0 18 − σ 0 0 18 − σ
= (18 − σ) σ2 − 18σ + 72 = 0

= (18 − σ) (σ − 12) (σ − 6) = 0
The principal values are

σ(1) = 18 σ(2) = 12 σ(3) = 6
(b) With respect to the principal axes
∗
n
t(^ )
=T ·n
or     
18 0 0 n1 18n1
    
  n2  =  12n2 
 0 12 0     

0 0 6 n3 6n3
This gives
∗
n
t(^ )
^∗1 + 12n2 e
= 18n1 e ^∗2 + 6n3 e
^∗3
and
∗
n
σN = t(^ )
^ = 18n21 + 12n22 + 6n23 = 12
·n
Thus
3n21 + 2n22 + n23 = 2
Now
∗ ∗
n n
σ2S = t(^ )
· t(^ )
− σ2N = 324n21 + 144n22 + 36n23 − 144 = 9
or
9n21 + 4n22 + n23 = 4.25
Finally
n21 + n22 + n23 = 1
These three equations can be solved to give
√
1 3 1
n1 = √ , n2 = , n3 = √
2 2 2 2 2
and
1 ∗ √ ∗
^∗ = √ e
n ^1 + 6^ ^∗3
e2 + e
2 2
Problem 3.20
Verify the result listed for Problem 3.19b above by use of Eq 3.63.
Solution
From Eq 3.63, we have
(σN − σII ) (σN − σIII ) + σ2S (12 − 12) (12 − 6) + 32 9 1

n21 = = = =
(σI − σII ) (σI − σIII ) (18 − 12) (18 − 6) (6) (12) 8
(σN − σIII ) (σN − σI ) + σ2S (12 − 6) (12 − 18) + 32 −36 + 9 3

n22 = = = =
(σII − σIII ) (σII − σI ) (12 − 6) (12 − 18) −36 4
(σN − σI ) (σN − σII ) + σ2S (12 − 18) (12 − 12) + 32 9 1

n22 = = = =
(σIII − σI ) (σIII − σII ) (6 − 18) (6 − 12) (−12) (−6) 8
Problem 3.21
Sketch the Mohr’s circles for the various stress states shown on the cube which is oriented
along the coordinate axes.
σ0 σ0
σ0 σ0
σ0
σ0
(a) (b)
σ0 σ0
σ0 2σ0
σ0 2σ0
σ0 σ0
(c) (d)
Problem 3.22
The state of stress referred to axes Px1 x2 x3 is given in MPa by the matrix
 
9 12 0
[tij ] =  12 −9 0  .
0 0 5
Determine
(a) the normal and shear components, σN and σS , respectively, on the plane at P
whose unit normal is
1
^ = (4^
n e1 + 3^
e2 ) ,
5
(b) verify the result determined in (a) by a Mohr’s circle construction similar to that
shown in the figure from Example 3.5.
Answer
σN = 14.04 MPa, σS = 5.28 MPa
Solution
(a) To find σN , we have
n)
t(^ =T ·n
^
or       
n)
(^ 4 72
t1 9 12 0 5 5
      
 t2n)
 (^
 =  12 0   35 = 21
     
−9 5

      
n)
(^
t3 0 0 5 0 0
The stress vector is
n) 72 21
t(^ = ^1 + e
e ^2
5 5
The normal stress component is

n)
(^ 72 21 4 3 288 + 63
σN = t ·n
^= ^1 + e
e ^2 · ^1 + 3^
e e2 = = 14.04
5 5 5 5 25
The shear stress component is
(72)2 + (21)2
n)
σ2S = t(^ n)
· t(^ − σ2N = − (14.04)2 = (5.28)2
25
(b) Verify this by construction of the Mohr’s circle.
Problem 3.23
Sketch the Mohr’s circles for the simple states of stress given by
 
σ0 0 σ0
(a) [tij ] =  0 σ0 0  ,
σ0 0 σ0
 
σ0 0 0
(b) [tij ] =  0 2σ0 0  ,
0 0 −σ0
and determine the maximum shear stress in each case.
Answer
(a) (σS )max = σ0

(b) (σS )max = 32 σ0
Solution
(a) The principal values are
 
σ0 − σ 0 σ0 σ0 − σ 0 σ0
det 

0 σ0 − σ 0  = 0 σ0 − σ 0 =0

σ0 0 σ0 − σ σ0 0 σ0 − σ
= (σ0 − σ) σ2 − 2σ0 σ

= (σ0 − σ) σ (σ − 2σ0 ) = 0
and
σI − σIII 2σ0 − 0
(σS )max = = = σ0
2 2
(b) The principal values are σI = 2σ0 , σII = σ0 , σIII = −σ0 , and
σI − σIII 2σ0 − (−σ0 ) 3
(σS )max = = = σ0
2 2 2
Problem 3.24
Relative to axes Ox1 x2 x3 , the state of stress at O is represented by the matrix
 
6 −3 0
[tij ] =  −3 6 0  [ksi] .
0 0 0
Show that, relative to principal axes Ox∗1 x∗2 x∗3 , the stress matrix is
 
3 0 0
[tij ] =  0 9 0  [ksi] ,
0 0 0
and that these axes result from a rotation of 45◦ about the x3 axis. Verify these results by
Eq 3.75.
Solution
The principal stresses are found from
 
6−σ −3 0 6−σ −3 0
det  −3

6−σ 0  = −3
6−σ 0 =0

0 0 0−σ 0 0 0−σ
= −σ σ2 − 12σ + 27 = 0

= −σ (σ − 9) (σ − 3) = 0
and
σ(1) = 9 σ(2) = 3 σ(3) = 0
For a counterclockwise rotation about the x3 axis, we have the transformation matrix
 
√1 √1 0
 2 2 
[aij ] =  − √1 √1
 
2 2
0 
 
0 0 1
t0ij = aiq ajm tqm gives

     
√1 √1 0 6 −3 0 √1 − √12 0 3 0 0
2 2 2
     
t0ij = 

− 12 √1 √1 √1 0 = 0
     
√
2
0   −3 6 0  2 2
9 0 
     
0 0 1 0 0 0 0 0 1 0 0 0

t11 + t22 t11 − t22
t011 = + cos 2θ + t12 sin 2θ
2 2
(6 + 6) (6 − 6) π π
= + cos + (−3) sin = 3
2 2 2 2
t11 + t22 t11 − t22

t022 = − cos 2θ − t12 sin 2θ
2 2
(6 + 6) (6 − 6) π π
= − cos − (−3) sin = 9
2 2 2 2
t11 − t22
t012 = − sin 2θ + t12 cos 2θ
2
(6 − 6) π π
=− sin + (−3) cos = 0
2 2 2
Problem 3.25
The stress matrix representation at P is given by
 
29 0 0
[tij ] =  0 −26 6  [ksi] .
0 6 9
Decompose this matrix into its spherical and deviator parts, and determine the principal
deviator stress values.
Answer
SI = 25 ksi, SII = 6 ksi, SIII = −31 ksi
Solution
The spherical stress is
1 1 1
σM = tr T = tii = (29 − 26 + 9) = 4
3 3 3
The deviatoric stress is
1
Sij = tij − δij tkk
3
Thus the deviatoric stress is
     
29 0 0 4 0 0 25 0 0
[Sij ] =  0 −26 6  −  0 4 0 = 0 −30 6 
0 6 9 0 0 4 0 6 5
The principal deviatoric stress is found from

 
25 − S 0 0 25 − S 0 0
det 

0 −30 − S 6  = 0 −30 − S 6 = 0
0 6 5−S 0 65−S
= (25 − S) S2 + 25S − 186 = 0

= (25 − S) (S − 6) (S + 31) = 0
The principal deviatoric stresses are
SI = 25, SII = 6, SIII = −31
Problem 3.26
Let the second invariant of the stress deviator be expressed in terms of its principal values,
that is, by
IIS = SI SII + SII SIII + SIII SI .
Show that this sum is the negative of two-thirds the sum of squares of the principal shear
stresses, as given by Eq 3.60.
Solution
The second invariant of the stress deviator is
IIS = SI SII + SII SIII + SIII SI
where
σI + σII + σIII
SI = σI − = σI − σM
3
similarly for the other values. Then

IIS = (σI − σM ) (σII − σM ) + (σII − σM ) (σIII − σM ) + (σIII − σM ) (σI − σM )
= σI σII + σII σIII + σIII σI − 2σM (σI + σII + σIII ) + 3σ2M
= σI σII + σII σIII + σIII σI − 3σ2M
2
(σI σII + σII σIII + σIII σI ) σI + σII + σIII
=9 −3
9 3

σ2I + σ2II + σ2III − σI σII − σII σIII − σIII σI
=−
3
σI − σIII σII − σIII σIII − σI
σS1 = σS2 = σS3 =
2 2 2
Therefore,

σ2 − 2σI σIII + σ2III + σ2II − 2σII σIII + σ2III + σ2III − 2σIII σI + σ2I
σ2S1 + σ2S2 + σ2S3 = I
4
σ2I + σ2II + σ2III − σI σII − σII σIII − σIII σI
=
2
Finally, we find
2 2
σS1 + σ2S2 + σ2S3 = SI SII + SII SIII + SIII SI

−
3
Problem 3.27
Verify the results presented in Eqs 3.89 and 3.90 for the octahedral shear stress.
Solution
Equation 3.88 is
1 2 1
σ2oct = σI + σ2II + σ2III − (σI + σII + σIII )2
3 9
3 2 2 2
1 2
σI + σ2II + σ2III + 2 (σI σII + σII σIII + σIII σI )

= σI + σII + σIII −
9 9
2 2 2 2

= σI + σII + σIII − 2 (σI σII + σII σIII + σIII σI )
9
1h i
= (σI − σII )2 + (σII − σIII )2 + (σIII − σI )2
9
and i1/2
1h
σoct = (σI − σII )2 + (σII − σIII )2 + (σIII − σI )2
3
This is Eq 3.89.
Now σI = SI + σM for each principal stress. This results in
1h i
σ2oct = (SI + σM − SII − σM )2 + (SII + σM − SIII − σM )2 + (SIII + σM − SI − σM )2
9
1h i
= (SI − SII )2 + (SII − SIII )2 + (SIII − SI )2
9
1
2 S2I + S2II + S2III − 2 (SI SII + SII SIII + SIII SI )

=
9
Also (SI + SII + SIII )2 = 0 and S2I + S2II + S2III = −2 (SI SII + SII SIII + SIII SI ). This gives

2 1 2 2 2
1 2 2 2
2
σoct = 3 SI + SII + SIII = S + SII + SIII = − (SI SII + SII SIII + SIII SI )
9 3 I 3
and from Problem 3.26 1/2

2
σoct = − IIS
3
This is Eq 3.90.
Problem 3.28
At point P in a continuum body, the stress tensor components are given in MPa with respect
to axes Px1 x2 x3 by the matrix
 √ 
1 −3 √2
[tij ] =  √
−3 1
√ − 2  .
2 − 2 4
Determine
(a) the principal stress values σI , σII , and σIII , together with the corresponding prin-
cipal stress directions,
(b) the stress invariants IT , IIT , and IIIT ,
(c) the maximum shear stress value and the normal to the plane on which it acts,
(d) the principal deviator stress values,
(e) the stress vector on the octahedral plane together with its normal and shear
components,
(f) the stress matrix for axes rotated 60◦ counterclockwise with respect to the axis
PQ, which makes equal angles relative to the coordinate axes Px1 x2 x3 .
Answer
(a) σI = 6 MPa, σII = 2 MPa, σIII = −2 MPa

(1) 1
√
^ =2 e
n ^1 − e^2 + 2^ e3
√
^ (2) = 21 e
n ^1 − e^2 − 2^ e3
^ +e
e ^
^ (3) = 1√ 2
n
2
(b) IT = 6 MPa, IIT = −4 MPa, IIIT = −24 MPa
√ √ √
1+ 2 e ^1 − 1 − 2 e ^2 + 2^
e3
(c) (σS )max = 4 MPa n ^ max = √
2 2
(d) SI = 4 MPa, SII = 0 MPa, SIII = −4 MPa
6ê∗1 + 2^
e∗ − 2ê∗3 q
(e) ^ n)
t(^ = √2 , σN = 2, σoct = 32 3
3 √ √ 

h i 1 −12 √ −12 + 3√ 2 −12 − 3 2
(f) t0ij =  −12 + 3√2 33 − 12 2 −3 √  MPa
9
−12 − 3 2 −3 33 + 12 2
Solution
 √  √
1−σ −3 √2
1−σ −3
√2

det  √

−3 1− σ − 2  = −3 − 2 = 0
1−
√ √ √σ
2 − 2 4−σ 2 − 2
4−σ
= (1 − σ) σ2 − 5σ + 2 − 3 (10 − 3σ) + 2 (2 + σ) = 0

= σ3 − 6σ2 − 4σ + 24
= (σ − 6) (σ − 2) (σ + 2) = 0
The results are

σ(1) = 6, σ(2) = 2, σ(3) = −2
 √ 
(1)
  
1−6 −3 2 n1 0
 √   (1)
    
= 0 
   
 −3 1 − 6 − 2   n2
 √ √    
(1)
2 − 2 4−6 n3 0
^ (1) · n
(1)
√ (1)
(1)
−5n1 − 3n2 + 2n3 = 0
(1) (1)
√ (1)
−3n1 − 5n2 − 2n3 = 0
√ (1) √ (1) (1)
2n1 − 2n2 − 2n3 = 0
From the first two equations, we have

(1) (1)
n1 = −n2
and this result together with the third equation gives

(1)
√ (1)
n3 = 2n1

2 2 2 2 2 √ 2
(1) (1) (1) (1) (1) (1)
n1 + n2 + n3 = n1 + −n1 + 2n2 =1
and
√
(1) 1 (1) 1 (1) 2
n1 =± n2 = ∓ n3 = ±
2 2√ 2
1 1 2
^ (1)
n =± e ^1 ∓ e ^2 ± ^3
e
2 2 2
 √ 
(2)
  
1−2 −3 2 n 0
 √  1   
  (2)
= 0 
   
 −3 1 − 2 − 2   n2
 √ √    
(2)
2 − 2 4−2 n3 0
^ (2) · n
(2)
√ (2) (2)
−1n1 − 3n2 + 2n3 = 0
(2) (2)
√ (2)
−3n1 − n2 − 2n3 = 0
√ (2) √ (2) (2)
2n1 − 2n2 + 2n3 = 0
or √ (2)
(2) (2) (2)
n2 = −n1 and n3 = − 2n1
2 2 2 2 2 √ 2
(2) (2) (2) (2) (2) (2)
n1 + n2 + n3 = n1 + −n1 + − 2n1 =1
and
√
(2) 1 (2) 1 (2) 2
n1 =± n2 = ∓ n =∓
2 2√ 3 2
1 1 2
^ (2)
n =± e ^1 ∓ e ^2 ∓ ^3
e
2 2 2
 √ 
(3)
  
1 − (−2) −3 2 n1 0
 √  
  (3)
  
= 0 
   
 −3 1 − (−2) − 2   n2
 √ √ 
(3)
  
2 − 2 4 − (−2) n3 0
^ (3) · n
(3)
√ (3) (3)
3n1 − 3n2 + 2n3 = 0
(3) (3)
√ (3)
−3n1 + 3n2 − 2n3 = 0
√ (3) √ (3) (3)
2n1 − 2n2 + 6n3 = 0
or from the second and third equation
(3) (3) (3)
n1 = n2 and n3 = 0
2 2 2 2 2
(3) (3) (3) (3) (3)
n1 + n2 + n3 = n2 + n2 =1
and
(3) 1 (3) 1 (3)
n2 = ± √ n1 = ± √ n3 = 0
2 2
1 1
n(3) = ± √ e1 ± √ e2
2 2
(b) The invariants of T are
IT = σI + σII + σIII = 6 + 2 − 2 = 6
IIT = σI σII + σII σIII + σIII σI = 6 (2) + 2 (−2) − 2 (6) = −4
IIIT = σI σII σIII = 6 (2) (−2) = −24
(c) The maximum shear stress is
σI − σIII 6 − (−2)
(σS )max = = =4
2 2
The normal to the plane of maximum shear stress is
" √ ! #
^∗1 + e
e ^∗3 1 1 1 2 1
^ max = √
n =√ e1 − e2 + e3 + √ (e1 + e2 )
2 2 2 2 2 2
1 h √ √ √ i
= √ 1 + 2 e1 − 1 − 2 e2 + 2e∗3
2 2
(d) The principal stress deviators are given that the mean stress is σM = 1
3
(σI + σII + σIII ) =
1
3
(6 + 2 − 2) = 2
SI = σI − σM = 6 − 2 = 4
SII = σII − σM = 2 − 2 = 0
SIII = σIII − σM = −2 − 2 = −4
(e) The stress vector on the octahedral plane with normal n

^ = √1 (^
3
^2 + e
e1 + e ^3 )is
t(n) = T · n
^
or      
6 0 0 1 6
1 1
 0 2 0 √  1 = √  2 
0 0 −2 3 1 3 −2
This gives
n) 1
t(^ = √ (6^
e1 + 2^
e2 − 2^
e3 )
3
The normal stress on this plane is
1 1
σN = t(n) · n
^ = √ (6^
e1 + 2^ e3 ) · √ (^
e2 − 2^ ^2 + e
e1 + e ^3 ) = 2
3 3
and the octahedral stress is
n) n) 36 + 4 + 4 32
σ2oct = t(^ · t(^ − σ2N = −4=
3 3
This gives
r
32
σoct =
3
(f) The stress on the plane with a normal making equal angles to the coordinate axes and
a 60◦ counterclockwise rotation is given by t0ij = aiq ajm tqm . The transformation matrix
is (Problem 3.10)  
2 2 −1
1
[aij ] =  −1 2 2 
3 2 −1 2
and
  √   
0 1 2 2 −1 1 −3 √2 1 2 −1 2
tij =  −1 2 2   −3
√ √ 1 − 2   2 2 −1 
3 2 −1 2 3 −1 2 2
2 − 2 4
 √ √ 
−12 √ −12 + 3√ 2 −12 − 3 2
1
=  −12 + 3√2 33 − 12 2 −3 √ 
9
−12 − 3 2 −3 33 + 12 2
Problem 3.29
In a continuum, the stress field relative to axes Ox1 x2 x3 is given by
 
x21 x2 x1 1 − x22 0
 1 2 
[tij ] =  x1 1 − x22  .
 
3 x2 − 3x 2 0
 
0 0 2x23
Determine
(a) the body force distribution if the equilibrium equations are to be satisfied through-
out the field,
√
(b) the principal stresses at P(a, 0, 2 a),
(c) the maximum shear stress at P,
(d) the principal deviator stresses at P.
Answer
4x3
(a) b1 = b2 = 0, b3 = −
ρ
(b) σI = 8a, σII = a, σIII = −a
(c) (σS )max = ±4.5a
16 5 11
(d) SI = a, SII = − a, SIII = − a
3 3 3
Solution
(a) The body force distribution is found from the equilibrium equations
tij,j + ρbi = 0
For i = 1
t11,1 + t12,1 + t13,3 + ρb1 = 2x1 x2 − 2x1 x2 + 0 + ρb1 = 0
or
b1 = 0
For i = 2
1
t21,1 + t22,2 + t23,3 + ρb2 = 1 − x22 + 3x22 − 1 + 0 + ρb2 = 0

3
and
b2 = 0
For i = 3
t31,1 + t32,2 + t33,3 + ρb3 = 4x3 + ρb3 = 0
and
4x3
b3 = −
ρ
√
(b) At the point P a, 0, 2a the stress is
 
0 a 0
[tij ] =  a 0 0 
0 0 8a
The principal stress is found from the eigen-problem

 
0−σ a 0 0−σ a 0
det  a

0−σ 0 = a
0−σ 0 =0

0 0 8a − σ 0 0 8a − σ
2 2

= (8a − σ) σ − a = 0
and
σI = 8a σII = a σIII = −a
(c) The maximum shear stress at P is
σI − σIII 8a − (−a)
(σS )max = = = 4.5a
2 2
(d) The principal deviatoric stresses are
8a 16a
SI = σI − σM = 8a − =
3 3
8a 5a
SII = σII − σM = a − =−
3 3
8a 11a
SIII = σIII − σM = −a − =−
3 3
given σM = 1
3
(σI + σII + σIII ) = 1
3
(8a + a − a) = 8a/3.
Problem 3.30
In describing the yield surface in plasticity the second invariant of the deviator stress, often
denoted by J2 , plays an important role. Starting with the second invariant of the deviator
stress
J2 = IIS = − S(1) S(2) + S(2) S(3) + S(1) S(3)
derive the formula
3 2
J2 = σ .
2 oct
Solution
Substitute S(i) = σ(i) − σM and carryout some algebra.
J2 = IIS = −(S(1) S(2) + S(2) S(3) + S(1) S(3) )
= −((σ(1) − σM )(σ(2) − σM ) + (σ(2) − σM )(σ(3) − σM ) + (σ(1) − σM )(σ(3) − σM ))
= −(σ(1) σ(2) + σ(2) σ(3) + σ(1) σ(3) − 2(σ(1) + σ(2) + σ(3) )σM + 3σ2M )
= −(σ(1) σ(2) + σ(2) σ(3) + σ(1) σ(3) − 3σ2M )

2(σ(1) σ(2) + σ(2) σ(3) + σ(1) σ(3) ) σ(1) + σ(2) + σ(3)
= − −3
2 9
!
(σ(1) + σ(2) + σ(3) )2 − (σ2(1) + σ2(2) + σ2(3) )
σ(1) + σ(2) + σ(3)
= − −
2 3
!
3(σ2(1) + σ2(2) + σ2(3) ) 3(σ(1) + σ(2) + σ(3) )2
= − − +
2·3 2·3·3
!
(σ2(1) + σ2(2) + σ2(3) ) σ(1) + σ(2) + σ(3) 2

3
= 2 −
3 3
3 2
= 2 σOCT
Problem 3.31
Show that
1 2
J2 = IIS = −(S(1) S(2) + S(2) S(3) + S(1) S(3) ) = S(1) + S2(2) + S2(3)
2
where J2 is the second invariant of the deviator stress and S(1) , S(2) , S(3) are its principal
values.
Solution
Use the fact that IS = 0 and substitute
S(1) = −S(2) − S(3)

S(2) = −S(1) − S(3)
S(3) = −S(1) − S(2)
J2 = S(1) (S(2) + S(3) ) + S(2) (S(1) + S(2) ) + S(3) (S(2) + S(3) )

= S2(1) + S2(2) + S2(3) − J2
Problem 3.32
Let the stress tensor components tij be derivable from the symmetric tensor field ϕij by the
equation tij = εiqk εjpm ϕkm,qp . Show that, in the absence of body forces, the equilibrium
equations are satisfied. Recall from Problem 2.15 that

δji δjq δjk

εiqk εjpm = δpi δpq δpk .
δmi δmq δmk
Solution The equilibrium equations without body forces are
tij,j = 0 = εiqk εjpm ϕkm,qpj
For

δji δjq δjk

εiqk εjpm = δpi δpq δpk

δmi δmq δmk
= δji (δpq δmk − δpk δmq ) − δjq (δpi δmk − δpk δmi ) + δjk (δpi δmq − δpq δmi )
then for i the free index
tij,j = εiqk εjpm ϕkm,qpj

= [δji (δpq δmk − δpk δmq ) − δjq (δpi δmk − δpk δmi ) + δjk (δpi δmq − δpq δmi )] ϕkm,qpj
= ϕkk,ppi − ϕkq,qki − ϕkk,qiq + ϕki,jkj + ϕkq,qik − ϕki,ppk = 0
Problem 3.33
Verify that ∂tij /∂tmn = δim δjn and use this result (or otherwise) to show that
∂IIS
= −Sij .
∂tij
that is, the derivative of the second invariant of the deviatoric stress with respect to the
stress components is equal to the negative of the corresponding component of the deviatoric
stress.
Chapter 4 Solutions
Problem 4.1
The motion of a continuous medium is specified by the component equations
1
x1 = 2 (X1 + X2 )et + 12 (X1 − X2 )e−t ,
1
x2 = 2 (X1 + X2 )et − 12 (X1 − X2 )e−t ,
x3 = X3 .
(a) Show that the Jacobian determinant J does not vanish, and solve for the inverse
equations X = X(x, t).
(b) Calculate the velocity and acceleration components in terms of the material
coordinates.
(c) Using the inverse equations developed in part (a), express the velocity and ac-
celeration components in terms of spatial coordinates.
Answer
(a) J = cosh2 t − sinh2 t = 1

X1 = 12 (x1 + x2 )e−t + 21 (x1 − x2 )et
X2 = 12 (x1 + x2 )e−t − 21 (x1 − x2 )et
X3 = x3
(b) v1 = 12 (X1 + X2 )et − 21 (X1 − X2 )e−t
v2 = 12 (X1 + X2 )et + 21 (X1 − X2 )e−t
v3 = 0
a1 = 12 (X1 + X2 )et + 12 (X1 − X2 )e−t
a2 = 12 (X1 + X2 )et − 12 (X1 − X2 )e−t
a3 = 0
(c) v1 = x2 , v2 = x1 , v3 = 0
a1 = x1 , a2 = x2 , a3 = 0
Solution
(a) For the motion to be invertible, the Jacobian of the deformation gradient is to be
non-zero.
t −t t −t

∂xi 1 et + e−t et − e−t 0

J = det F = |FiA | =
= e −e e +e 0
∂XA 2 0 0 2
2 2
et + e−t et − e−t

= − = cosh2 t − sinh2 t = 1 6= 0
2 2
69
Now
x1 + x2 = (X1 + X2 ) et
x1 − x2 = (X1 − X2 ) e−t
and
X1 + X2 = (x1 + x2 ) e−t
X1 − X2 = (x1 − x2 ) et
This yields
(x1 + x2 ) e−t + (x1 − x2 ) et

X1 =
2
(x1 + x2 ) e−t − (x1 − x2 ) et
X2 =
2
X3 = x3

dx
(b) The velocity is v = or
dt X
1 1
v1 = (X1 + X2 )et − (X1 − X2 )e−t
2 2
1 1
v2 = (X1 + X2 )et + (X1 − X2 )e−t
2 2
v3 = 0

dv
The acceleration is a = or
dt X
1 1
a1 = (X1 + X2 )et + (X1 − X2 )e−t
2 2
1 1
a2 = (X1 + X2 )et − (X1 − X2 )e−t
2 2
a3 = 0
(c) From part (a), we have the sum and differences X1 + X2 and X1 − X2 to find the spatial
forms
1 1
(x1 + x2 ) e−t et − (x1 − x2 ) et e−t = x2

v1 =
2 2
1 1
(x1 + x2 ) e−t et + (x1 − x2 ) et e−t = x1

v2 =
2 2
v3 = 0
The acceleration in spatial form is
1 1
(x1 + x2 ) e−t et + (x1 − x2 ) et e−t = x1

a1 =
2 2
1 1
(x1 + x2 ) e−t et − (x1 − x2 ) et e−t = x2

a2 =
2 2
a3 = 0
Problem 4.2
Let the motion of a continuum be given in component form by the equations
x1 = X1 + X2 t + X3 t2 ,
x2 = X2 + X3 t + X1 t2 ,
x3 = X3 + X1 t + X2 t2 .
(a) Show that J 6= 0, and solve for the inverse equations.

(b) Determine the velocity and acceleration
(1) at time t = 1 s for the particle which was at point (2.75, 3.75, 4.00) when
t = 0.5 s.
(2) at time t = 2 s for the particle which was at point (1, 2, −1) when t = 0.
Answer
(a) J = (1 − t3 )2
X1 = (x1 − x2 t)/(1 − t3 )
X2 = (x2 − x3 t)/(1 − t3 )
X3 = (x3 − x1 t)/(1 − t3 )
(b) (1) v = 8ê1 + 5^
e2 + 5^
e3 , a = 6ê1 + 2^
e2 + 4^
e3
(2) v = −2^ e1 + 3^
e2 + 9ê3 , a = −2^
e1 + 2^
e2 + 4^
e3
Solution
(a) The Jacobian is

1 t t2
J = det F = |FiA | = t2 = 1 − t3 2

1 t
t t2 1
The inverse is using Cramer’s rule

x1 t t2

x2 1 t

x3 t2 1 x1 (1 − t3 ) − x2 (t − t4 ) x1 − x2 t
X1 = = 2
=
J 3
(1 − t ) (1 − t3 )

1 x1 t2
2
t x2 t

t x3 1 x2 (1 − t3 ) − x3 (t − t4 ) x2 − x3 t
X2 = = 2
=
J 3
(1 − t ) (1 − t3 )

1 t x1
2
t 1 x2

t t2 x3 x3 (1 − t3 ) − x1 (t − t4 ) x3 − x1 t
X3 = = =
J (1 − t3 )2 (1 − t3 )

dx
(b) The velocity is v = or
dt X
v1 = X2 + 2X3 t; v2 = X3 + 2X1 t; v3 = X1 + 2X2 t
Using the inverse equations from part (a), we have the spatial forms
x2 + x3 t − 2x1 t2 x3 + x1 t − 2x2 t2 x1 + x2 t − 2x3 t2

v1 = ; v2 = ; v3 =
(1 − t3 )2 (1 − t3 )2 (1 − t3 )2

dv
dt X
a1 = 2X3 ; a2 = 2X1 ; a3 = 2X2
and the spatial form is
2 (x3 − x1 t) 2 (x1 − x2 t) 2 (x2 − x3 t)

a1 = 2
; a2 = 2
; a3 =
(1 − t3 ) (1 − t3 ) (1 − t3 )2
Problem 4.3
A continuum body has a motion defined by the equations
x1 = X1 + 2X2 t2 ,
x2 = X2 + 2X1 t2 ,
x3 = X3 .
(a) Determine the velocity components at t = 1.5 s of the particle which occupied
the point (2, 3, 4) when t = 1.0 s.
(b) Determine the equation of the path along which the particle designated in part
(a) moves.
(c) Calculate the acceleration components of the same particle at time t = 2 s.
Answer
(a) v1 = 2, v2 = 8, v3 = 0
(b) 4x1 − x2 = 5 in the plane x3 = 4
(c) a1 = 4/3, a2 = 16/3, a3 = 0.
Solution
dx
(a) The velocity is v = or
dt X
v1 = 4X2 t; v2 = 4X1 t; v3 = 0
When t = 1.0 s, the initial location of the particle is found from
2 = X1 + 2X2
3 = X2 + 2X1
4 = X3
The solution is
4 1
X1 = ; X2 = ; X3 = 4
3 3
The velocity components at t = 1.5 s are

1 3 4 3
v1 = 4X2 t = 4 = 2; v2 = 4X1 t = 4 = 8; v3 = 0
3 2 3 2
(b) The path for the particle at the position in (a) is given by

2 4 1 2
x1 = X1 + 2X2 t = +2 t
3 3

1 4 2
x2 = X2 + 2X1 t2 = +2 t
3 3
x3 = X3 = 4
Eliminating t gives
12x1 − 16 = 3x2 − 1
or
4x1 − x2 = 5 in the plane x3 = 4

dv
(c) The acceleration is a = or
dt X
a1 = 4X2 ; a2 = 4X1 ; a3 = 0
or
1 4 4 16
a1 = 4X2 = 4 = ; a2 = 4X1 = 4 = ; a3 = 0
3 3 3 3
Problem 4.4
If the motion x = x(X, t) is given in component form by the equations
x1 = X1 (1 + t), x2 = X2 (1 + t)2 , x3 = X3 (1 + t2 ) ,
determine expressions for the velocity and acceleration components in terms of both La-
grangian and Eulerian coordinates.
Answer
v1 = X1 = x1 /(1 + t)
v2 = 2X2 (1 + t) = 2x2 /(1 + t)
v3 = 2X3 t = 2x3 t/(1 + t2 )
a1 = 0
a2 = 2X2 = 2x2 /(1 + t)2
a3 = 2X3 = 2x3 /(1 + t2 )
Solution
dx
The velocity is v = or
dt X
v1 = X1 ; v2 = 2X2 (1 + t) ; v3 = 2X3 t
To establish the Eulerian form we must invert the position equations to get
x1 x2 x3
X1 = ; X2 = 2
; X3 =
1+t (1 + t) (1 + t2 )
and
x1 2x2 2x2 t
v1 = X1 = ; v2 = 2X2 (1 + t) = ; v3 = 2X3 t =
1+t 1+t (1 + t2 )

dv
dt X
a1 = 0; a2 = 2X2 ; a3 = 2X3
or
2x2 2x3
a1 = 0; a2 = ; a3 =
(1 + t)2 (1 + t2 )
Problem 4.5
The Lagrangian description of a continuum motion is given by
x1 = X1 e−t + X3 (e−t − 1) ,
x2 = X2 et − X3 (1 − e−t ) ,
x3 = X3 et .
Show that these equations are invertible and determine the Eulerian description of the
motion.
Answer
X1 = x1 et − x3 (et − 1)
X2 = x2 e−t + x3 (e−2t − e−3t )
X3 = x3 e−t
Solution
The Jacobian must be non-zero for the equations be invertible.
−t
e 0 e−t − 1
J = det F = [FiA ] = 0

et − (1 − e−t ) = e−t e2t = et 6= 0

0 0 et
Inverting the equation for x3 gives
X3 = x3 e−t
Then
x1 = X1 e−t + X3 (e−t − 1) = X1 e−t + x3 e−t (e−t − 1)

x2 = X2 et − X3 (1 − e−t ) = X2 et − x3 e−t (1 − e−t )
X3 = x3 e−t
The Eulerian form of the motion is
X1 = x1 et x3 (e−t − 1)
X2 = x2 e−t + x3 e−2t (1 − e−t )
X3 = x3 e−t
Also by using Cramer’s rule, we have

−t
x1 0 e−t − 1 e x1 e−t − 1
−t

X2 = e−t 0
t

X1 = e x2 e − (1 − e−t ) ;
x2 − (1 − e−t )

x3 0 et 0 x3 et

e−t 0 x1
−t

X3 = e 0 et x2
0 0 x3
Problem 4.6
A velocity field is given in Lagrangian form by
v1 = 2t + X1 , v2 = X2 et , v3 = X3 − t .
Integrate these equations to obtain x = x(X, t) with x = X at t = 0, and using that result
compute the velocity and acceleration components in the Eulerian (spatial) form.
Answer
v1 = (x1 + 2t + t2 )/(1 + t)
v2 = x2
v3 = (2x3 − 2t − t2 )/2(1 + t)
a1 = 2, a2 = x2 , a3 = −1
Solution
Integrating the velocity equations gives
t2
x1 = t2 + X1 t + C1 ; x2 = X2 et + C2 ; x3 = X3 t − + C3
2
The initial conditions are x = X at t = 0. This gives
x1 = X1 = C1 ; x2 = X2 = X2 + C2 and C2 = 0; x3 = X3 = C3
This results in
x1 = t2 + X1 t + X1
x2 = X2 et
t2
x3 = X3 t − + X3
2
Inverting directly gives
x1 − t2 x3 + 12 t2
X1 = ; X2 = x2 e−t ; X3 =
1+t 1+t
Substituting these positions gives the Eulerian form
2t (1 + t) + x1 − t2 2x3 − 2t − t2
v1 = , v2 = x2 , v3 =
1+t 2 (1 + t)

dv
dt X
a1 = 2; a2 = X2 et = x2 , a3 = −1
Problem 4.7
If the motion of a continuous medium is given by
x1 = X1 et − X3 (et − 1) ,
x2 = X2 e−t + X3 (1 − e−t ) ,
x3 = X3 ,
determine the displacement field in both material and spatial descriptions.
Answer
u1 = (X1 − X3 )(et − 1) = (x1 − x3 )(1 − e−t )

u2 = (X2 − X3 )(e−t − 1) = (x2 − x3 )(1 − et )
u3 = 0
Solution
The displacement field is u = x − X. For the given motion, we have
u1 = x1 − X1 = X1 et − X3 (et − 1) − X1 = (X1 − X3 ) et − 1

u2 = x2 − X2 = X2 e−t + X3 (1 − e−t ) − X2 = (X3 − X2 ) 1 − e−t

u3 = x3 − X3 = X3 − X3 = 0
This is the material form for the displacement field. To find the spatial form, the mo-
tion must be inverted. This can be accomplished directly without the calculation of the
Jacobian.
x1 = X1 et − X3 (et − 1) = X1 et − x3 (et − 1)
x2 = X2 e−t + X3 (1 − e−t ) = X2 e−t + x3 (1 − e−t )
x3 = X3
and
X1 = x1 e−t + x3 1 − e−t

X2 = x2 et − x3 et − 1

X3 = x3
The spatial form of the displacement field is
u1 = x1 − X1 = (X1 − X3 ) et − 1 = x1 e−t + x3 1 − e−t − x3 et − 1

= (x1 − x3 ) 1 − e−t

u2 = x2 − X2 = (X3 − X2 ) 1 − e−t = x3 − x2 et − x3 et − 1 1 − e−t

= (x3 − x2 ) et − 1 = (x2 − x3 ) 1 − et

and
u3 = x3 − X3 = x3 − x3 = 0
Problem 4.8
The temperature field in a continuum is given by the expression
θ = e−3t /x2 where x2 = x21 + x22 + x23 .
The velocity field of the medium has components
v1 = x2 + 2x3 , v2 = x3 − x1 , v3 = x1 + 3x2 .
Determine the material derivative dθ/dt of the temperature field.
Answer
dθ/dt = −e−3t (3x2 + 6x1 x3 + 8x2 x3 )/x4
Solution
The material time derivative of the temperature field is
e−3t e−3t

dθ ∂θ ∂θ ∂
= + vi = −3 2 + vi
dt ∂t ∂xi x ∂xi x2
Now, using the chain rule
e−3t

∂ −3t −3 ∂x
=e −2x
∂xi x2 ∂xi
The last term in the parentheses is
∂ (xj xj )1/2

∂x 1 −1/2 ∂xj 1 xi
= = (xk xk ) 2xj = xj δij =
∂xi ∂xi 2 ∂xi x x
So that
e−3t

∂ −3t −3 ∂x
xi x
i
=e −2x = e−3t −2x−3 = e−3t −2 4
∂xi x2 ∂xi x x
This gives
dθ e−3t x
i e−3t
= −3 2 − 2e−3t vi 4 = − 4 3x2 + 2vi xi

dt x x x
e−3t
= − 4 3x2 + 2 [x1 (x2 + 2x3 ) + x2 (x3 − x1 ) + x3 (x1 + 3x2 )]

x
e−3t e−3t
= − 4 3x2 + 2 [3x1 x3 + 4x2 x3 ] = − 4 3x2 + 6x1 x3 + 8x2 x3

x x
Problem 4.9
In a certain region of a fluid the flow velocity has components
v1 = A(x31 + x1 x22 )e−kt , v2 = A(x21 x2 + x33 )e−kt , v3 = 0
where A and k are constants. Use the (spatial) material derivative operator to determine
the acceleration components at the point (1, 1, 0) when t = 0.
Answer
a1 = −2A(k − 5A), a2 = −A(k − 5A), a3 = 0
Solution The acceleration is

dv ∂v
a= = + v · grad v
dt ∂t
or
∂vi ∂vi
ai = + vk
∂t ∂xk
So that
a1 = −kA(x31 + x1 x22 )e−kt + A(x31 + x1 x22 )e−kt 3Ax21 e−kt

+ A(x21 x2 + x33 )e−kt 2Ax1 x2 e−kt

a2 = −kA(x21 x2 + x33 )e−kt + A(x31 + x1 x22 )e−kt (2Ax1 x2 ) e−kt

+ A(x21 x2 + x33 )e−kt Ax21 e−kt

and
a3 = 0
At the point (1, 1, 0) and t = 0, we have
a1 |(1,1,0),t=0 = −kA (2) + A (2) (3A) + A (1) (2A) = −2A (k − 4A)

a2 |(1,1,0),t=0 = −kA (2) + A (2) (2A) + A (1) (A) = −A (k − 5A)
a3 |(1,1,0),t=0 = 0
Problem 4.10
A displacement field is given in terms of the spatial variables and time by the equations
u1 = x2 t2 , u2 = x3 t, u3 = x1 t .
Using the (spatial) material derivative operator, determine the velocity components.
Answer
v1 = (2x2 t + x3 t2 + x1 t3 )/(1 − t4 )
v2 = (x3 + x1 t + 2x2 t3 )/(1 − t4 )
v3 = (x1 + 2x2 t2 + x3 t3 )/(1 − t4 )
Solution
The velocity is
du ∂u
v= = + v · grad u
dt ∂t
or
dui ∂ui ∂ui
vi = = + vk
dt ∂t ∂xk
This is an implicit equation for the velocity field. For the given displacement field
v1 = 2x2 t + v1 (0) + v2 t2 + v3 (0) = 2x2 t + v2 t2

v2 = x3 + v1 (0) + v2 (0) + v3 (t) = x3 + v3 (t)

v3 = x1 + v1 (t) + v2 (0) + v3 (0) = x1 + v1 (t)
Now, we have
v1 = 2x2 t + t2 (x3 + tv3 ) = 2x2 t + t2 (x3 + t [x1 + tv1 ])
and
2x2 t + x3 t2 + x1 t3
v1 =
1 − t4
Similarly
v2 = x3 + t (x1 + tv1 ) = x3 + t x1 + t 2x2 t + t2 v2

and
x3 + x1 t + 2x2 t3
v2 =
1 − t4
Finally
v3 = x1 + t 2x2 t + t2 v2 = x1 + t 2x2 t + t2 [x3 + tv3 ]

and
x1 + 2x2 t2 + x3 t3
v3 =
1 − t4
Alternatively, the problem can be solved using Cramer’s rule. We write the system of
equations as
v1 −t2 v2 = 2x2 t
v2 −tv3 = x3
−tv1 +v3 = x1
The determinant is
1
−t2 0

0 1 −t = 1 − t4

−t 0 1
and
2x2 t −t2 0 1 2x2 t 0
1 1
0

v1 = x3 1 −t ; v2 = x3 −t
1 − t4
x1
1 − t4
−t x1 1

0 1

1 −t2 2x2 t
1
v3 = 0 1 x3
1 − t4
−t 0 x1
Problem 4.11
For the motion given by the equations
x1 = X1 cos ωt + X2 sin ωt ,
x2 = −X1 sin ωt + X2 cos ωt ,
x3 = (1 + kt)X3 ,
where ω and k are constants, determine the displacement field in Eulerian form.
Answer
u1 = x1 (1 − cos ωt) + x2 sin ωt

u2 = −x1 sin ωt + x2 (1 − cos ωt)
u3 = x3 kt/(1 + kt)
Solution
The Jacobian is
cos ωt sin ωt

0
J = det F = |FiA | = − sin ωt cos ωt = cos2 ωt + sin2 ωt (1 + kt)

0
0 0 1 + kt
which is not zero. Multiplying the first equation in the motion by sin ωt and the second
equation by cos ωt yields
x1 sin ωt = X1 sin ωt cos ωt + X2 sin2 ωt

x2 cos ωt = −X1 cos ωt sin ωt + X2 cos2 ωt
and adding gives

x1 sin ωt + x2 cos ωt = X2
Also
x1 cos ωt = X1 cos2 ωt + X2 sin ωt cos ωt

x2 sin ωt = −X1 sin2 ωt + X2 cos ωt sin ωt
and subtracting yields

x1 cos ωt − x2 sin ωt = X1
The displacements are
u1 = x1 − X1 = x1 − x1 cos ωt + x2 sin ωt = x1 (1 − cos ωt) + x2 sin ωt

u2 = x2 − X2 = x2 − x1 sin ωt − x2 cos ωt = −x1 sin ωt + x2 (1 − cos ωt)

1 x3 kt
u3 = x3 − X3 = x3 − x3 =
1 + kt 1 + kt
Problem 4.12
Show that the displacement field for the motion analyzed in Problem 4.1 has the Eulerian
form
u1 = x1 − (x1 + x2 )e−t /2 − (x1 − x2 )et /2 ,

u2 = −x2 − (x1 + x2 )e−t /2 + (x1 − x2 )et /2 ,
and by using the material derivative operator (dui /dt = ∂ui /∂t + vj ∂ui /∂xj ), verify the
velocity and acceleration components calculated in Problem 4.1.
Solution
From Problem 4.1 the inverted motion is
(x1 + x2 ) e−t + (x1 − x2 ) et
X1 =
2
(x1 + x2 ) e−t − (x1 − x2 ) et
X2 =
2
X3 = x3
The displacement field is
(x1 + x2 ) e−t + (x1 − x2 ) et

u1 = x1 − X1 = x1 −
2
(x1 + x2 ) e−t − (x1 − x2 ) et
u2 = x2 − X2 = x2 −
2
u3 = x3 − X3 = x3 − x3 = 0
The velocity field is
du1 ∂u1 ∂u1

v1 = = + vk
dt ∂t ∂xk
(x1 + x2 ) e−t (x1 − x2 ) et e−t et
−t
et

e
= − + v1 1 − − − v2 −
2 2 2 2 2 2
and
du2 ∂u2 ∂u2
v2 = = + vk
dt ∂t ∂xk
(x1 + x2 ) e−t (x1 − x2 ) et
−t
et e−t et

e
= + − v1 − + v2 1 − −
2 2 2 2 2 2
Adding and subtracting these equations, yields
v1 + v2 = x1 + x2
v1 − v2 = −x1 + x2
and
v1 = x2 ; v2 = x1 ; v3 = 0
The material description is
(X1 + X2 ) et (X1 − X2 ) e−t

v1 = x2 = −
2 2
and
(X1 + X2 ) et (X1 − X2 ) e−t
v2 = x1 = +
2 2
The accelerations are
a1 = x1 ; a2 = x1 ; a3 = 0
Problem 4.13
The Lagrangian description of a deformation is given by
x1 = X1 + X3 (e2 − e−2 ) ,
x2 = X2 − X3 (e2 − 1) ,
x3 = X3 e2 .
Determine the components of the deformation matrix FiA and from it show that the Jaco-
bian J does not vanish. Invert the mapping equations to obtain the Eulerian description of
the deformation.
Answer
J = e2
X1 = x1 − x3 (1 − e−4 )
X2 = x2 + x3 (1 − e−2 )
X3 = x3 e−2
Solution
The deformation gradient is
e2 − e−2
 
1 0
∂xi
[FiA ] = = 0 1 e2 − 1 
∂XA
0 0 e2
The Jacobian is
1 0 e2 − e−2
J = det F = 0 1

e2 − 1 = e2

0 0 e2
and is non-vanishing. From the third equation of the motion, we have
X3 = x3 e−2
Then we have
x1 = X1 + x3 e−2 (e2 − e−2 ); x2 = X2 − x3 e−2 (e2 − 1)
and
X1 = x1 − x3 (1 − e−4 ); X2 = x2 + x3 (1 − e−2 ); X3 = x3 e−2
Problem 4.14
A homogeneous deformation has been described as one for which all of the deformation and
strain tensors are independent of the coordinates, and may therefore be expressed in general
by the displacement field ui = Aij Xj where the Aij are constants (or in the case of a motion,
functions of time). Show that for a homogeneous deformation with the Aij constant:
(a) plane material surfaces remain plane,
(b) straight line particle elements remain straight,
(c) material surfaces which are spherical in the reference configuration become ellip-
soidal surfaces in the deformed configuration.
Solution
(a) The displacement field is
ui = xi − Xi = Aij Xj
The spatial points are given by
xi = Xi + Aij Xj = (δij + Aij ) Xj
Let a material plane be given by

βi Xi + α = 0
and let
Xi = (δij + Bij ) xj
be the inverse of the spatial points. Then
βi (δij + Bij ) xj + α = 0
is the spatial description of the deformed plane or
λj xj + α = 0
where λj = βj + βi Bij is a constant. Thus the deformed surface is a plane.

(b) Since planes deform into planes, the intersection of two planes is a straight line, and
straight lines deform into straight lines.
(c) Spherical material surfaces are given by
Xi Xi = R2
Then in the spatial or deformed configuration this becomes
(δij + Bij ) xj (δik + Bik ) xk = R2
and
(δjk + Bkj + Bjk + Bij Bik ) xj xk = R2
This gives
βjk xj xk = R2
an ellipsoid.
Problem 4.15
An infinitesimal homogeneous deformation ui = Aij Xj is one for which the constants Aij are
so small that their products may be neglected. Show that for two sequential infinitesimal
deformations the total displacement is the sum of the individual displacements regardless
of the order in which the deformations are applied.
Solution
Let the two displacements be
(1) (1) (2) (2)
ui = Aij Xj ; ui = Aij Xj
Then the spatial positions are

(1) (1) (2) (2)
xi = δij + Aij Xj ; xi = δij + Aij Xj
Successive motions give

(2) (2) (1)
xi = δij + Aij δjk + Ajk Xk

(1) (2) (2) (1)
= δik + Aik + Aik + Aij Ajk Xk
Neglecting the product terms, this gives

(2) (1) (2)
xi = δik + Aik + Aik Xk
and
(2) (1) (2) (1) (2)
xi − Xi = Aik + Aik Xk = ui + ui
Problem 4.16
For the homogeneous deformation defined by
x1 = αX1 + βX2 ,
x2 = −βX1 + αX2 ,
x3 = µX3 ,
where α, β and µ are constants, calculate the Lagrangian finite strain tensor E. Show that
if α = cos θ, β = sin θ and µ = 1 the strain is zero and the mapping corresponds to a rigid
body rotation of magnitude θ about the X3 axis.
Answer
α2 + β2 − 1
 
0 0
1 
EAB =2 0 α2 + β2 − 1 0 
0 0 µ2 − 1
Solution
The deformation gradient is  
α β 0
[FiA ] =  −β α 0 
0 0 µ
The Green’s deformation tensor is C = FT · F and
  2
α + β2
  
α −β 0 α β 0 0 0
[CAB ] =  β α 0   −β α 0  =  0 α2 + β2 0 
0 0 µ 0 0 µ 0 0 µ2
The Lagrangian finite strain tensor is E = 12 (C − I) or

 2
α + β2 − 1

0 0
1
[EAB ] =  0 α2 + β2 − 1 0 
2
0 0 µ2 − 1
For α = cos θ, β = sin θ, and µ = 1, the strain is
cos2 θ + sin2 θ − 1
   
0 0 0 0 0
1
[EAB ] =  0 cos2 θ + sin2 θ − 1 0 = 0 0 0 
2
0 0 12 − 1 0 0 0
The mapping is
x1 = cos θX1 + sin θX2
x2 = − sin θX1 + cos θX2
x3 = X3
This is a rotation in the plane.
Problem 4.17
Given the deformation defined by
1
x1 = X1 , x2 = X2 + X23 , x3 = X3
2
(a) Sketch the deformed shape of the unit square OABC in the plane X1 = 0.
(b) Determine the differential vectors dx(2) and dx(3) which are the deformed vectors
resulting from dX(2) = dX(2)^ I2 and dX(3) = dX(3)^ I3 , respectively, that were
originally at corner C.
(c) Calculate the dot product dx(2) · dx(3) , and from it determine the change in the
original right angle between dX(2) and dX(3) at C due to the deformation.
(d) Compute the stretch Λ at B in the direction of the unit normal
√
N^ = ^ I2 + Î3 / 2 .
X3
C B
O
A X2
X1
Unit square OABC in the reference configuration.
Answer
(b) dx(2) = dX(2) e
^2 , dx(3) = dX(3) (^ ^3 )
e2 + e
◦
(c) ∆θ = −45√
(d) Λ(N
^) = 2.5
Solution
(a) The points move to
Point O (0, 0, 0) → O (0, 0, 0) Point A (0, 1, 0) → a (0, 1, 0)
Point B (0, 1, 1) → b (0, 01.5, 1) Point C (0, 0, 1) → c (0, 0.5, 1)
(b) The differential vectors dx(2) and dx(3) result from dX(2) = dX(2)^
I2 and dX(3) =
(3)^
dX I3 . This is determined from the deformation gradient through dx = F · dX. The
deformation gradient is  
1 0 0
[FiA ] =  0 1 1 
0 0 1
Now     
1 0 0 0 0
 0 1 1   dX(2)  =  dX(2) 
0 0 1 0 0
and
dx(2) = dX(2) e
^2
Also     
1 0 0 0 0
 0 1 1   0  =  dX(3) 
0 0 1 dX(3) dX(3)
and
dx(3) = dX(3) (^ ^3 )
e2 + e
(c) The dot product is
q
2
(2) (3)
= dx dx cos θ = dX (2)
2 dX(3) cos θ
(2) (3)
dx · dx
√
= 2dX(2) dX(3) cos θ
and
dX(2) · dX(3) dX(2) dX(3) 1
cos θ = = √ =√
dx(2) · dx(3) (2)
2dX dX (3) 2
This gives
θ = 45◦
(d) The stretch is found from C = FT · F and
    
1 0 0 1 0 0 1 0 0
[CAB ] =  0 1 0   0 1 1 = 0 1 1 
0 1 1 0 0 1 0 1 2
Now  
0

h i 1 0 0
√1 √1  1 
Λ2(0,1,1) = 0 2 2
 0 1 1   √2  = 2.5
0 1 2 √1
2
Problem 4.18
Given the deformation expressed by
x1 = X1 + AX22 , x2 = X2 , x3 = X3 − AX22
where A is a constant (not necessarily small), determine the finite strain tensors E and e,
and show that if the displacements are small so that x ≈ X, and if squares of A may be
neglected, both tensors reduce to the infinitesimal strain tensor .
Answer
 
0 Ax2 0
[ij ] =  Ax2 0 −Ax2 
0 −Ax2 0
Solution
The finite strain tensors E is E = 1
(C − I) = 21 FT · F − I . The deformation gradient is

2
 
1 2AX2 0
[FiA ] =  0 1 0 
0 −2AX2 1
and
  
1 0 0 1 2AX2 0
[CAB ] =  2AX2 1 −2AX2   0 1 0 
0 0 1 0 −2AX2 1
 
1 2AX2 0
=  2AX2 1 + 8A2 X22 −2AX2 
0 −2AX2 1
The finite strain tensor E is
 
0 2AX2 0
1
[EAB ] = 2AX2 8A2 X22 −2AX2 
2
0 −2AX2 0
Neglecting squares of A we have
 
0 AX2 0
[ij ] =  AX2 0 −AX2 
0 −AX2 0
T
The finite strain tensors e is e = 12 (I − c) = 12 I − F−1 · F−1 . Inverting the defor-
mation
X1 = x1 − Ax22 , X2 = x2 , X3 = x3 + Ax22
Then, the inverse deformation gradient is
 
−1 1 −2Ax2 0
FiA =  0 1 0 
0 2Ax2 1
and
  
1 0 0 1 −2Ax2 0
[cAB ] =  −2Ax2 1 2Ax2   0 1 0 
0 0 1 0 2Ax2 1
 
1 −2Ax2 0
=  −2Ax2 1 + 8A2 x22 2Ax2 
0 2Ax2 1
The finite strain tensor e is
 
0 −2Ax2 0
1
[eAB ] =  −2Ax2 8A2 x22 2Ax2 
2
0 2Ax2 0
Neglecting squares of A we have

 
0 −Ax2 0
[ij ] =  −Ax2 0 2Ax2 
0 Ax2 0
Now if x ≈ X the infinitesimal strain measures are the same.
Problem 4.19
For the infinitesimal homogeneous deformation xi = Xi + Aij Xj where the constants Aij
, and
are very small, determine the small strain tensor from
√
it the longitudinal (normal)
^ ^ ^
strain in the direction of the unit vector N = I1 − I3 / 2.
Answer
2e(N
^ ) = A11 − A13 − A31 + A33
Solution
The infinitesimal strain tensor is ij = 21 (ui,j + uj,i ). The displacement is ui = xi − Xi =
Aij Xj , and since xi ≈ Xi , we have
 
2A11 A12 + A21 A13 + A31
1
[ij ] = A21 + A12 2A22 A23 + A32 
2 A +A A +A 2A
31 13 32 23 33
Then e(N
^ ) in the direct given is
   √1 
h i 2A11 A12 + A21 A13 + A31 2
2e(N √1 0 − √12  A21 + A12 2A22 A23 + A32   0 
^) =
 
2
A31 + A13 A32 + A23 2A33 − √12
= A11 − A13 − A31 + A33
Problem 4.20
A deformation is defined by
x1 = X1 / X21 + X22 , x2 = X2 / X21 + X22 ,

x3 = X3 .
Determine the deformation tensor C together with its principal values.
Answer
−2
C(1) = C(2) = X21 + X22 , C(3) = 1
Solution
−X21 + X22
 
−2X1 X2 0
∂xi 1
[FiA ] = = 2  −2X1 X2 X21 − X22 0 
∂XA X21 + X22 2
0 0 X21 + X22
The deformation tensor C = FT · F or CAB = FAi FiB

−X21 + X22 −2X1 X2 −X21 + X22
  
0 −2X1 X2 0
1
[CAB ] = 4  −2X1 X2 X21 − X22 0  −2X1 X2 X21 − X22 0 
X21 + X22 0 0 2 2 2
X1 + X2

0 0 2 2 2
X1 + X2

2
 
X21 + X22 0 0
1 2 2 2

=
 
4  0 X1 + X 2 0 
X21 + X22 4
0 0 X1 + X22
2
Since the deformation tensor is diagonal, the principal values are the diagonal values.
−2
C(1) = C(2) = X21 + X22 , C(3) = 1
Problem 4.21
For the deformation field given by
x1 = X1 + αX2 , x2 = X2 − αX1 , x3 = X3
where α is a constant, determine the matrix form of the tensors E and e, and show that
the circle of particles X21 + X22 = 1 deforms into the circle x21 + x22 = 1 + α2 .
Answer
α2
 
0 0
1 
[Eij ] = 2
0 α2 0 
0 0 0
−α2
 
0 0
1
[eij ] = 2
 0 −α2 0 
2 (1 + α2 ) 0 0 0
Solution
The finite strain tensor E requires the deformation gradient F since E = 1
2
(C − I) =
1 T

2 F ·F−I  
1 α 0
[FiA ] =  −a 1 0 
0 0 1
and
    
1 −α 0 1 α 0 1 0 0
1 
[EAB ] = a 1 0   −a 1 0 − 0 1 0 
2 0 0 1 0 0 1 0 0 1
 2 
α 0 0
1
=  0 α2 0 
2
0 0 1
To compute e, we need to invert the deformation. The Jacobian is

1 α 0
J = det F = −a 1 0 = 1 + α2

0 0 1
This is non-zero, and

x1 − αx2 αx1 + x2
X1 = ; X2 = ; X3 = x3
1 + α2 1 + α2
This gives  
1 −α 0
1 
F−1

iA = a 1 0 
1 + α2
0 0 1 + α2
T
and e = 12 (I − c) = 12 I − F−1 · F−1
    
1 0 0 1 α 0 1 −α 0
1 1  −a 1
[eAB ] =  0 1 0  − 0  a 1 0 
2 0 0 1 (1 + α2 )2 0 0 1 + α2 0 0 1 + α2
 2 
α 0 0
1  0 α2 0 
=
2 (1 + α2 )
0 0 0
The circle X21 + X22 = 1 gives
2 2
x1 − αx2 αx1 + x2
+ =1
1 + α2 1 + α2
and 2
(x1 − αx2 )2 + (αx1 + x2 )2 = 1 + α2
Simplifying, we have
x21 + x22 = 1 + α2
This is a circle.
Problem 4.22
Let the deformation of a continuum be given by the equations
x1 = X1 + kX22 , x2 = X2 − kX21 , x3 = X3
where k is a constant. Determine the Lagrangian finite strain tensor E, and from it, assuming
k is very small, deduce the infinitesimal strain tensor . Verify this by calculating the
displacement field and using the definition 2ij = ui,j + uj,i for the infinitesimal theory.
Solution
The finite strain tensor is E = 1
FT · F − I . The deformation gradient F is
1

2
(C − I) = 2
 
1 2kX2 0
[FiA ] =  −2kX1 1 0 
0 0 1
and
    
1 −2kX1 0 1 2kX2 0 1 0 0
1 
[EAB ] = 2kX2 1 0   −2kX1 1 0 − 0 1 0 
2 0 0 1 0 0 1 0 0 1
2 2
 
4k X1 2kX2 − 2kX1 0
1
=  2kX2 − 2kX1 4k2 X22 0 
2
0 0 0
Neglecting squares of k, gives the linear strain tensor

 
0 kX2 − kX1 0
[ij ] =  kX2 − kX1 0 0 
0 0 0
u1 = x1 − X1 = kX22 , u2 = x2 = X2 = −kX21 , u3 = x3 − X3 = 0
The infinitesimal strain tensor
 
0 kX2 − kX1 0
1
[ij ] = [ui,j + uj,i ] =  kX2 − kX1 0 0 
2 0 0 0
Problem 4.23
Given the displacement field
u1 = AX2 X3 , u2 = AX23 , u3 = AX21
where A is a very small constant, determine
(a) the components of the infinitesimal strain tensor , and the infinitesimal rotation
tensor ω.
(b) the principal values of , at the point (1, 1, 0).
Answer
(a) 11 = 22 = 33 = 0, 12 = AX3 , 13 = 12 A(X2 + 2X1 ) and 23 = AX3
ω11 = ω22 = ω33 = 0, ω12 = −ω21 = 21 AX3 ,
ω13 = −ω31 = 12 AX1 − AX1 and ω23 = −ω32 = AX3
(b) (I) = 32 A, (II) = 0, (III) = − 23 A
Solution
(a) The infinitesimal strain tensor is ij = 12 (ui,j + uj,i ) and
 
0 AX3 AX2 + 2AX1
1
[ij ] = AX3 0 2AX3 
2 AX + 2AX 2AX 0
2 1 3
The infinitesimal rotation is ωij = 12 (ui,j − uj,i ) and

 
0 AX3 AX2 − 2AX1
1
[ωij ] =  −AX3 0 2AX3 
2 AX − 2AX −2AX 0
2 1 3
(b) At (1, 1, 0), the infinitesimal strain tensor is

 
0 0 3A
1
[ij ](1,1,0) =  0 0 0 
2 3A 0 0
The principal values are given by

3

− 0 2A
" 2 #
3
det ( − I) = 0
2
− 0 = − −
A =0
3A 2
2 0 −
and
3 3
(I) = A; (II) = 0; (III) = − A
2 2
Problem 4.24
A 45◦ strain rosette measures longitudinal strains along the X1 , X2 , and X01 axes shown
below. At point O the strains recorded are
11 = 6 × 10−4 , 22 = 4 × 10−4 and 011 = 8 × 10−4

Determine the shear strain γ12 at O, together with ε022 , and verify that ε11 +ε22 = ε011 +ε022 .
See Eq 4.82.
X2
o X1‛
45
o
45
X2‛
O
X1
Solution
The transformation equation is
11 + 22 11 − 22 γ12
011 = + cos 2θ + sin 2θ
2 2 2
This gives
6+4 6−2 γ12
8= + cos 90◦ + sin 90◦
2 2 2
and solving
γ12 = 6 × 10−4
The second principal strain is
11 + 22 11 − 22 γ12
022 = + cos 2θ + sin 2θ
2 2 2
and
6+4 6−4 3
022 = + cos 270◦ + sin 270◦
2 2 2
Solving gives
022 = 2
We have
11 + 22 = 6 + 4 = 011 + 022 = 8 + 2
Problem 4.25
By a direct expansion of Eq 4.87, 2ωi = εijk ωkj , show that ω1 = ω32 = −ω23 , etc. Also,
show that only if A is a very small constant does the mapping
x1 = X1 − AX2 + AX3 ,
x2 = X2 − AX3 + AX1 ,
x3 = X3 − AX1 + AX2 ,
represent a rigid body rotation (E ≡ 0). Additionally, determine the infinitesimal rotation
tensor ωij in this case; from it, using the result proven above, deduce the rotation vector
ωi .
Answer
^1 + e
ω = A(e ^2 + e
^3 )
Solution
Expanding 2ωi = ijk ωkj , gives
2ω1 = 12k ωk2 + 13k ωk3 = 123 ω32 + 132 ω23 = ω32 − ω23 = 2ω32 = −2ω23
2ω2 = 23k ωk3 + 21k ωk1 = 231 ω13 + 213 ω31 = ω13 − ω31 = 2ω13 = −2ω31
2ω3 = 31k ωk1 + 32k ωk2 = 312 ω21 + 321 ω12 = ω21 − ω12 = 2ω21 = −2ω12
To compute the finite strain tensor E = 12 FT · F − I . the deformation gradient is

 
1 −A A
[FiA ] =  A 1 −A 
−A A 1
and
    
1 A −A 1 −A A 1 0 0
1 
[EAB ] = −A 1 A  A 1 −A  −  0 1 0 
2 A −A 1 −A A 1 0 0 1
2 2 2
 
2A −A −A
1
=  −A2 2A2 −A2 
2
−A2 −A2 2A2
This is only zero when A is small.
u1 = x1 − X1 = −AX2 + AX3
u2 = x2 − X2 = −AX3 + AX1
u3 = x3 − X3 = −AX1 + AX2
h i
The infinitesimal rotation is ω = 21 (grad u) − (grad u)T and
 
0 −A A
[grad u] = [ui,j ] =  A 0 −A 
−A A 0
The infinitesimal rotation is

     
0 −A A 0 A −A 0 −A A
1 
[ωij ] = A 0 −A  −  −A 0 A  =  A 0 −A 
2 −A A 0 A −A 0 −A A 0
This gives
ω1 = ω23 = A; ω2 = ω13 = A; ω3 = ω21 = A
The rotation vector is
ω = A (^ ^2 + e
e1 + e ^3 )
Problem 4.26
For the displacement field
u1 = kX1 X2 , u2 = kX1 X2 , u3 = 2k(X1 + X2 )X3
where k is a very small constant, determine the rotation tensor ω, and show that it has
only one real principal value at the point (0, 0, 1).
Answer
√ √
ω(1) = 0, ω(2) = −ω(3) = ik 2, where i = −1
Solution
The rotation tensor ω requires the displacement gradient grad u. This is
 
kX2 kX1 0
[ui,j ] =  kX2 kX1 0 
2kX3 2kX3 2k (X1 + X2 )
and
   
kX2 kX1 0 kX2 kX2 2kX3
1 
[ωij ] = kX2 kX1 0  −  kX1 kX1 2kX3 
2 2kX3 2kX3 2k (X1 + X2 ) 0 0 2k (X1 + X2 )
k (X1 − X2 )
 
0 −kX3

 k (X1 − X2 ) 2 

= − 0 −kX3 
2

 
kX3 kX3 0
The rotation tensor at (0, 0, 1) is

 
0 0 −k
[ωij ](0,0,1) = 0 0 −k 
k k 0
The principal value is found from

−λ 0 −k
det (ω − λI) = 0 −λ −k = −λ λ2 + k2 − k2 λ = λ λ2 + 2k2 = 0

k k −λ
and √ √
λI = 0; λII , λIII = ±ik 2 or ω(I) = 0; ω(II) = −ω(III) = ik 2
Problem 4.27
Let the deformation
x1 = X1 + AX2 X3 ,
x2 = X2 + AX23 ,
x3 = X3 + AX21 ,
where A is a constant be applied to a continuum body. For the unit square of material line
elements OBCD as shown, calculate at point C
(a) the stretch and unit elongation for the element in the direction of diagonal OC,
(b) the change in the right angle at C if A = 1; if A = 0.1.
X3
D C
O
1 B X2
X1
Unit square OBCD in the reference configuration.
Answer
√
(a) Λ2(OC) = 1 + 2A + 4A2 , e(OC) = 1 + 2A + 4A2 − 1
(b) ∆θ(A=1) = 60◦ , ∆θ(A=0.1) = 11.77◦
Solution
(a) The deformation gradient is
 
1 AX3 AX2
∂xi
[FiA ] = = 0 1 2AX3 
∂XA 2AX1 0 1
The Green’s deformation tensor is

  
1 0 2AX1 1 AX3 AX2
[CAB ] = [FAi ] [FiB ] =  AX3 1 0  0 1 2AX3 
AX2 2AX3 1 2AX1 0 1
2 2
 
1 + 4A X1 AX3 AX2 + 2AX1
= AX3 1 + A2 X23 A2 X2 X3 + 2AX3 
AX2 + 2AX1 A X2 X3 + 2AX3 1 + A2 X22 + 4A2 X23
2
At the point C, we have (0, 1, 1) and CAB we have

 
1 A A
[CAB ](0,1,1) =  A 1 + A2 A2 + 2A 
A A2 + 2A 2
1 + A + 4A 2

^ · C · N.
Now Λ2 = N ^ = √1 ^
^ Along the diagonal N I2 + ^
I3 and
2
  
1 A A 0
h i  
Λ2(0,1,1) = √1 √1 2 2 √1
  
0 2 2
 A 1+A A + 2A 
2

  
A A2 + 2A 1 + A2 + 4A2 √1
2
This gives p
e( N
^) = Λ − 1 = 1 + A2 + 4A2 − 1
^ 1 and N
(b) The angle between N ^ 2 is
N^1 · C · N
^
cos θ = p p 2
^1 · C · N
N ^1 N ^2 · C · N
^2
For A = 1, we have  
1 1 1
[CAB ](0,1,1) = 1 2 3 
1 3 6
and   
h i 1 1 1 0
^1 · C · N
^2 = 0

N 1 0  1 2 3  0 =3
1 3 6 1
  
h i 1 1 1 0
N^1 · C · N
^1 = 0 1 0  1 2 3  1 =2
1 3 6 0
  
h i 1 1 1 0
N^2 · C · N
^2 = 0 0 1  1 2 3  0 =6
1 3 6 1
So that √
3 3
cos θ = √ √ =
2 6 2
and
θ = 30◦ and ∆θA=1 = 60◦
For A = 0.1, we have  

1 0.1 0.1
[CAB ](0,1,1) =  0.1 1.01 0.21 
0.1 0.21 1.05
and h i h i
^1 · C · N
N ^ 2 = 0.21; N^1 · C · N
^ 1 = 1.01; N^2 · C · N
^ 2 = 1.05
This gives
0.21 0.21
cos θ = √ √ =
1.01 1.05 1.03
and
θ = 78.24◦ and ∆θA=0.1 = 11.76◦
Problem 4.28
For the homogeneous deformation expressed by the equations
√
√ 3 2
x1 = 2X1 + X2 ,
4 √
3 2
x2 = −X1 + X2 + X3 ,
4 √4
3 2
x3 = X1 − X2 + X3 ,
4 4
determine
^ for the line element originally in the direction of
normal n
(a) the unit
√
^ ^ ^ ^
N = I1 − I2 + I3 / 3.
(b) the stretch Λ(N
^ ) of this element.
(c) the maximum and minimum stretches at the point X1 = 1, X2 = 0, X3 = −2 in

the reference configuration.
Answer
√ √ √
2e^1 + 2−7 e^2 + 2+7 e^3
(a) ^=
n √
104
(b) Λ( N
^) = 1.472
(c) Λ(max) = 2; Λ(min) = 0.5
Solution
√
√
 
2 3
 2 4 √
0 
 
[FiA ] =  −1
 3 2 

 4 √4 
3 2
 
1 −
4 4
^ = B. This gives
Now F · N
 √    √ 
√ 2 3 1 2
 2 0  √   √ 
 4 √   3   4 3√ 
1 −7 + 2
   

[Bij ] =  3 2 
  −√  =  √
   
 −1 
4 √4  3   4 √ 3
   
  
 3 2  1   7 + 2 
1 − √ √
4 4 3 4 3
and n
^ = B/B √
104 104
B·B=B = 2
and B = √
48 4 3
Then √ √ √
e1 + −7 + 2 e
2^ ^2 + 7 + 2 e^3
n^= √
104
^ ·C·N
(b) Square of the stretch is Λ2 = N ^ and
  √ 
√ √ 2 3
2 −1 1  2 0
4
 
 √  √ 

[CAB ] = [FAi ] [FiB ] =  2 3 3 3   3 2 
−    −1
 
4 √4 √4  4 √4
 
  
 2 2  3 2 
0 1 −
  4 4 4 4
 4 0 0 
 

= 9 
 0 0 
4

 
 1 
0 0
4
Then   
1
 4 0 0  √ 
  3 
 13
1 1 1 1

Λ2 =
 9 
=
√ −√ √ √

 0 0  −
6

3 3 3  4 3 
 
 
 1  1 
0 0 √
4 3
and
Λ = 1.47
√
(c) Since C in (b) is C∗ then Λ(max) = 4 = 2 and Λ(min) = 1/4 = 12 .
p
Problem 4.29
Let the deformation
x1 = a1 (X1 + 2X2 ), x2 = a2 X2 , x3 = a3 X3
where a1 , a2 , and a3 are constants be applied to the unit cube of material shown in the
sketch. Determine
(a) the deformed length l of diagonal OC,

(b) the angle between edges OA and OG after deformation,
(c) the conditions which the constants must satisfy for the deformation to be possible
if
(1) the material is incompressible,
(2) the angle between elements OC and OB is to remain unchanged.
X3 , x3
E
F
D
C
O
G
A
X2 , x2
B
X1 , x1
Unit cube having diagonal OC.
Answer
(a) l2 = 9a21 + a22 + a23

2a1
(b) cos θ = q
4a21 + a22
(c) (1) a1 a2 a3 = 1,
(2) 9a21 + a22 = 2a23
Solution
 
a1 2a1 0
[FiA ] =  0 a2 0 
0 0 a3
^ =^
The diagonal is N I1 + ^
I2 + ^
I3 . The deformed diagonal is n ^
^ =F·N
    
a1 2a1 0 1 3a1
n] =  0
[^ a2 0   1  =  a2 
0 0 a3 1 a3
The length is
l2 = (3a1 )2 + a22 + a23 = 9a21 + a22 + a23
(b) The angle between OA and OG
N^ OA · C · N
^
cos θ = p p OG
N^ OA · C · N
^ OA N ^ OG · C · N
^ OG
where C = FT · F
  
a1 0 0 a1 2a1 0
[CAB ] = [FAi ] [FiB ] =  2a1 a2 0   0 a2 0 
0 0 a3 0 0 a3
 2 2

a1 2a1 0
=  2a21 4a21 + a22 0 
0 0 a23
Now
a21 2a21
  
h i 0 0
^ OA · C · N
^ OG = 1 0 0  2a21 4a21 + a22 0   1  = 2a21

N
0 0 a23 0
 2
2a21
 
h i a1 0 1
^ OA · C · N
^ OA = 1 0 0  2a2 4a2 + a2 0   0  = a21

N 1 1 2
0 0 a23 0
 2 2
 
h i a1 2a1 0 0
^ OG · C · N
^ OG = 0 1 0  2a2 4a2 + a2 1  = 4a21 + a22

N 1 1 2 0  
2
0 0 a3 0
This gives
2a21 2a1
cos θ = q =q
a1 4a21 + a22 4a21 + a22
(c) (1) For the material to be incompressible J = 1. This gives

a1 2a1 0
J = det [FiA ] = 0

a2 0 = a1 a2 a3 = 1
0 0 a3
(2) The angle between OB and OC is
^
I1 + ^
I ^
I +^
I +Î3 2
cos β = √ 2 · 1 √2 =√
2 3 6
The angle between OB and OC after deformation is
^ OB · C · N
N ^
cos θ = p p OC
N^ OB · C · N
^ OB N ^ OC · C · N
^ OC
where
^ +^ ^ ^
I2 + ^
^ OB = I1√
N
I2 ^ OC = I1 + √
and N
I3
2 3
Now
2 2
^ OC = 9a1√+ a2
^ OB · C · N
N
6
2 2
N ^ OB = 9a1 + a2
^ OB · C · N
2
2 2 2
N ^ OC = 9a1 + a2 + a3
^ OC · C · N
3
Equating the cosine of the two angles gives
9a21 + a22
√
2
cos θ = r r6 = cos β = √
2 2 2 2 2 6
9a1 + a2 9a1 + a2 + a3
2 3
and simplifying
9a21 + a22 = 2a23
Problem 4.30
A homogeneous deformation is defined by
x1 = αX1 + βX2 , x2 = −αX1 + βX2 , x3 = µX3
where α, β and µ are constants. Determine

(a) the magnitudes and directions of the principal stretches,
(b) the matrix representation of the rotation tensor R,
(c) the direction of the axis of the rotation vector, and the magnitude of the angle
of rotation.
Answer
(a) Λ2(1) = Λ2(e^1 ) = 2a2 , Λ2(2) = Λ2(e^2 ) = 2β2 , Λ2(3) = Λ2(e^3 ) = µ2

 
√1 √1 0
 2 2 
(b) [Rij ] =  − √1 √1
 
2 2
0 
 
0 0 1
^ = Î3 ; Φ = 45◦
(c) n
Solution
 
α β 0
∂ui
[FiA ] = =  −α β 0 
∂XA 0 0 µ
The Green’s deformation tensor is C = FT · F or

2α2
    
α −α 0 α β 0 0 0
[CAB ] = [FAi ] [FiB ] =  β β 0   −α β 0 = 0 2β2 0 
0 0 µ 0 0 µ 0 0 µ2
Since this is diagonal, we have
Λ2(1) = Λ2(^ 2
e1 ) = 2a ; Λ2(2) = Λ2(^ 2
e2 ) = 2β ; Λ2(3) = Λ2(^
e3 ) = µ
2
(b) The right stretch tensor is U = C1/2 so that

 √ 
2α √0 0
[UAB ] =  0 2β 0 
0 0 µ
and R = F · U−1 . The inverse of the stretch tensor is

 
√1 0 0
2α
−1  
UAB =  0 √1 0 
 
 2β 
1
0 0 µ
and     
α β 0 √1 0 0 √1 √1 0
  2α   2 2 
[RAB ] =  −α β 0 √1 0 = − √12 √1
    
0  2β 2
0 
    
1
0 0 µ 0 0 µ
0 0 1
(c) The angle of rotation is
1
1 + 2 cos Φ = Rii = 1 + 2 √
2
and
1
cos Φ = √ or Φ = 45◦
2
The direction is found from
    
√1 √1 0 n1 n1
 2 2    
− √12 √1 0   n2  =  n2 
    

 2    
0 0 1 n3 n3
and
√
n1 + n2 =2n1
√
−n1 + n2 = 2n2
n3 = n3
The first two equations give

n1 = n2 = 0
and since
n21 + n22 + n23 = 1
we have
n3 = ±1
The direction of the rotation is

n ^3 = ^
^=e I3
Problem 4.31
Consider the deformation field
x1 = X1 − AX2 + AX3 ,
x2 = X2 − AX3 + AX1 ,
x3 = X3 − AX1 + AX2 ,
where A is a constant. Show that the principal values of the right stretch tensor have a
√
^ = Î1 + Î2 + Î3 / 3.
multiplicity of two, and that the axis of the rotation tensor is along N
Determine the matrix of the rotation vector together with the angle of rotation φ.
Answer
√
Λ(1) = 1, Λ(2) = Λ(3) = 1 + 3A2 = β
 
β+2 β − 1 − 3A β − 1 + 3A
1 
[Rij ] = β − 1 + 3A β+2 β − 1 − 3A 
3β β − 1 − 3A β − 1 + 3A β+2
φ = cos−1 (1/β)
Solution
 
1 −A A
∂xi
[FiA ] = =  A 1 −A 
∂XA −A A 1
The right stretch tensor is

  
1 A −A 1 −A A
[CAB ] = [FAi ] [FiB ] =  −A 1 A  A 1 −A 
A −A 1 −A A 1
2 2 2
 
1 + 2A −A −A
=  −A2 1 + 2A2 −A2 
−A2 −A2 1 + 2A2

1 + 2A2 − λ −A2 −A2
det (C − λI) =
2 2

−A 1 + 2A − λ −A2 =0

2 2
−A −A 1 + 2A2 − λ
= 1 + 3A2 − λ (1 − λ) 1 + 3A2 − λ = 0

The roots are

Λ(1) = 1; Λ(2) = Λ(3) = 1 + 3A2
The principal stretch tensor is

   
1 √ 0 0 1 0 0
[U∗AB ] = [C∗AB ]1/2 = 0 1 + 3A2 √ 0 = 0 β 0 
0 0 1 + 3A2 0 0 β
and   
1 0 0

−1/2 1 0 0
[U∗AB ]−1 = 0 1 + 3A2 0 = 0 1/β 0 
  

2 −1/2 0 0 1/β

0 0 1 + 3A
The rotation tensor is R = F · U−1 . The inverse stretch tensor can be found from the
eigenvectors associated with the principal stretches and inverse principal stretch tensor.
For Λ(1) = 1. We have
    
(1)
2A2 −A2 −A2 n1 0
    
  (1)
 −A2 2A2 −A2   n2 = 0 
   
    
(1)
−A2 −A2 2A2 n3 0
and
(1) (1) (1)
2n1 − n2 − n3 = 0
(1) (1) (1)
−n1 + 2n2 − n3 = 0
(1) (1) (1)
−n1 − n2 + 2n3 = 0
This gives
(1) (1) (1) 1
n1 = n2 = n3 = √
3

^ = √1 ^
This is the axis of rotation N I1 + ^
I2 + ^
I3 since the other two roots are equal.
3
For Λ(2) = Λ(3) = 1 + 3A2 , we have to select two vectors that are perpendicular to the
above vector. One choice is
1
n(2) = √ (^ ^2 )
e1 − e
2
and

(3) (1) (2) 1 1
n =n ×n = √ (^
e1 + e ^3 ) × √ (^
^2 + e ^2 )
e1 − e
3 2
1
= √ (^ ^2 − 2^
e1 + e e3 )
6

 
√1 √1 √1
 3 3 3 
[aij ] =  √1 − √12
 
2
0 
 
√1 √1 − √26
6 6
The inverse stretch tensor is U−1 = AT · (U∗ )−1 · A or

   
√1 √1 √1 1 0 0 √1 √1 √1
3 2 6  3 3 3
−1   1
 
UAB =  √ − √12 √1 √1 − 12
   
  0 1/β 0  √ 0 
 3 6   2 
1 2 √1 √1
√
3
0 − √6 0 0 1/β 6 6
− √26
  
√1 √1 √1 √1 √1 √1
 3 2 6  3 3 3 
= √1 − √12 √1 √1 − √12β 0
  
 2β

 3 6  
√1 0 − 26
√ √1 √1 − √26β
3 6β 6β
 
β+2 β−1 β−1
1 
=  β−1 β+2 β−1 

3β
β−1 β−1 β+2
The rotation tensor is

   
1 −A A β+2 β−1 β−1
1
[RiB ] = [FiA ] U−1

 A 1 −A  β−1 β+2 β−1 
AB =

−A A 1 3β β − 1 β − 1 β + 2
 
β+2 β − 1 − 3A β − 1 + 3A
1 
= β − 1 + 3A β+2 β − 1 − 3A 
3β β − 1 − 3A β − 1 + 3A β+2
The angle of rotation is
3 (β + 2) 2
1 + 2 cos Φ = Rii = =1+
3β β
or
1
cos Φ =
β
Problem 4.32
For the deformation field
√
x1 = 3X1 + X2 ,
x2 = 2X2 ,
x3 = X3 .
determine
(a) the matrix representation of the rotation tensor R,

(b) the right stretch tensor U and the left stretch tensor V, then show that the
principal values of U and V are equal,
(c) the direction of the axis of rotation and the magnitude of the angle of rotation.
Answer
 √ √ 
3+1 3−1 0
1  √ √
(a) [Rij ] = √  − 3 + 1 3+1 0 

2 2 √
0 0 2 2
√ √
(b) Λ(1) = 6, Λ(2) = 2, Λ(3) = 1
(c) N^ = Î3 ; φ = 15◦
Solution
 √ 

∂xi
3 1 0
[FiA ] = = 0 2 0 
∂XA
0 0 1
The right stretch tensor is

 √  √   √ 
3 0 0 3 1 0 √3 3 0
[CAB ] = [FAi ] [FiB ] =  1 2 0  0 2 0 = 3 5 0 
0 0 1 0 0 1 0 0 1
The principal stretches are

√
3−λ 3 0
√

det (C − λI) =

3 5−λ 0 =0

0 0 1−λ
= (1 − λ) (λ − 6) (λ − 2) = 0
The principal stretch tensors are

 
6 0 0
[C∗AB ] =  0 2 0 
0 0 1
and  √ 
6 √0 0
[U∗AB ] =  0 2 0 
0 0 1
The inverse principal stretch is
√
 
1/ 6 0√ 0
[U∗AB ]−1 = 0 1/ 2 0 
0 0 1
The inverse stretch is determined from U−1 = AT · (U∗ )−1 · A where the A is found from
the eigenvectors for the principal stretches.
 √ 
1 3
0
 2√ 2 
[aij ] =  − 23 1
 
2 0 
 
0 0 1
This gives

1
√
3
 √ 
1
√
3

2 − 2 0 1/ 6 0 0 2 2 0
 √  √  √ 
U−1

3 1
AB =  − 23 1
   
2 0  0
2 1/ 2 0  2 0 
   
0 0 1 0 0 1 0 0 1
 √ √ 
1+3 3 3−√3 0
1  √
= √ 3−3 3+ 3 0 
√
4 6
0 0 4 6
Since F = R · U, we have R = F · U−1 , we find

 √   √ √ 
3 1 0 1√+ 3 3 3−√3 0
1
[RiA ] =  0 2 0  √  3−3 3+ 3 0 
√
0 0 1 4 6
0 0 4 6
 √ √ 
1 + √3 −1 +√ 3 0
1
= √  1− 3 1+ 3 0
√

2 2
0 0 2 2
(b) The square of the left stretch tensor is V 2 = F · FT or

 √  √   
2 3 1 0 3 0 0 4 2 0
V =  0 2 0  1 2 0  =  2 4 0 
0 0 1 0 0 1 0 0 1
The principal values are

4−λ 2 0 h i
= 0 = (1 − λ) (4 − λ)2 − 4 = (1 − λ) (λ − 6) (λ − 2)

2 4 − λ 0

0 0 1−λ
These are the same as the principal values of the square of the right stretch tensor.
^ = ni^
(c) Let the axis of rotation be N Ii then
 √ √    
1 + √3 −1 +√ 3 0 n1 n1
1 
√ 1− 3 1+ 3 0   n2  =  n2 
2 2 √
0 0 2 2 n3 n3
and
√ √ √
1 + 3 − 2 2 n1 + 3 − 1 n2 = 0
√ √ √
1 − 3 n1 + 1 + 3 − 2 2 n2 = 0
Solving gives n1 = n2 = 0 and for a unit vector n3 = 1. Thus N ^ =^

I3 . The angle of
rotation can be found from n ^ ^
^ = R · N where N = √ (^
1
e 1 + ^2 ) and
e
2
 √ √    √   
1+3 −1 + 3 0 √1 3
n1
2 2
1 
 √ √      
√  1− 3 √1 = 1  =  n2
     
1+ 3 0  2 2

2 2 √      
0 0 2 2 0 0 n3
This gives √
3 1
^=
n ^1 + e
e ^2
2 2
The unit vector is at Φ = 60◦ and the angle change is ∆Φ = 15◦ clockwise.
Problem 4.33
Let a displacement field be given by
1 1 1
u1 = (X3 − X2 ), u2 = (X1 − X3 ), u3 = (X2 − X1 ) .
4 4 4
Determine
(a) the volume ratio dV/dV0 ,
(b) the change inthe right angle between line elements originally along the unit
√ √
vectors N1 = 3I1 − 2I2 − I3 / 14 and N2 = I1 + 4Î2 − 5Î3 / 42. Explain
^ ^ ^ ^ ^ ^
your answer.
Answer
(a) dV/dV0 = 1.1875

(b) ∆θ = 0◦
Solution
(a) The Jacobian of the deformation gradient gives the volume ratio. The deformations
are
1 1 1
x1 = X1 + (X3 − X2 ), x2 = X2 + (X1 − X3 ), x3 = X3 + (X2 − X1 )
4 4 4
This gives  
4 −1 1
∂x1 1
[FiA ] = =  1 4 −1 
∂XA 4 −1 1 4
and
4 −1 1
1 19 dV 19
J = det F = 1 4 −1 =
16 ; 0
= = 1.1875
4 −1 1 4 dV 16
(b) The angle is by definition
N^1 · C · N
^
cos θ = p p 2
^1 · C · N
N ^1 N ^2 · C · N
^2
The right deformation tensor is C = FT · F and

  
4 1 −1 4 −1 1
1 
[CAB ] = [FAi ] [FiB ] = −1 4 1  1 4 −1 
16 1 −1 4 −1 1 4
 
18 −1 −1
1 
= −1 18 −1 
16 −1 −1 18
Now
  
18 −1 −1 √1
h i 1   42 
^1 · C · N
N ^2 = √3 − √214 − √114

 −1 18

−1   √4

16 

14  42 
−1 −1 18 − √542
57 − 152 + 95
= √ √ =0
14 42(16)
and
cos θ = 0; and θ = 90◦
^ 1 and N
Since N ^ 2 are perpendicular initially, ∆θ = 0.
Problem 4.34
Consider again the deformation given in Example 4.15, namely
x1 = 2(X1 − X2 ), x2 = X1 + X2 , x3 = X3 .
Determine
(a) the left stretch tensor V,
(b) the direction normals of the principal stretches of V.
Answer
 √ 
2 2 √0 0
(a) [VAB ] =  0 2 0 
0 0 1
^ ^ ^ ^ ^ 3 = Î3
(b) N1 = I1 , N2 = I2 , N
Solution
 
2 −2 0
∂xi
[FiA ] = = 1 1 0 
∂XA 0 0 1
The square of the left deformation tensor is V 2 = F · FT

    
2 2 −2 0 2 1 0 8 0 0
V =  1 1 0   −2 1 0 = 0 2 0 
0 0 1 0 0 1 0 0 1
Since this is diagonal, √

 
2 2 √0 0
[VAB ] =  0 2 0 
0 0 1
(b) The stretch tensor is diagonal, V = V ∗ and
^1 =^
N I1 , ^2 =^
N I2 , ^3 =^
N I3
Problem 4.35
A deformation field is expressed by
x1 = µ(X1 cos βX3 + X2 sin βX3 ) ,

x2 = µ(−X1 sin βX3 + X2 cos βX3 ) ,
x3 = νX3 ,
where µ, β, and ν are constants.

(a) Determine the relationship between these constants if the deformation is to be a
possible one for an incompressible medium.
(b) If the above deformation is applied to the circular cylinder shown by the sketch,
determine
(1) the deformed length l in terms of L, the dimension a, and the constants µ,
β, and ν of an element of the lateral surface which has unit length and is
parallel to the cylinder axis in the reference configuration, and
(2) the initial length L of a line element on the lateral surface which has unit
length and is parallel to the cylinder axis after deformation.
X3
X2
X1
Answer
(a) µ2 ν = 1
p
(b) (1) l = µ2 β2 a2 + ν2
1p 2 2
(2) L = β a +1
ν
Solution
µ cos βX3 µ sin βX3 −µβX1 sin βX3 + µβX2 cos βX3
 

∂xi
[FiA ] = =  −µ sin βX3 µ cos cos βX3 −µβX1 cos βX3 − µβX2 cos βX3 
∂XA 0 0 ν
For an incompressible material, the Jacobian of the deformation gradient must be one.
Thus

J = det F = −µ sin βX3 µ cos βX3 −µβX1 cos βX3 − µβX2 cos βX3

0 0 ν
= ν µ cos βX3 + µ sin βX3 = νµ = 1
2 2 2 2 2

(b) (1) The deformed length is

   
l1 0
 l2  = [FiA ]  0 
l3 1
  
0
=  −µ sin βX3 µ cos βX3 −µβX1 cos βX3 − µβX2 cos βX3   0 
0 0 ν 1
−µβX1 sin βX3 + µβX2 cos βX3
 
=  −µβX1 cos βX3 − µβX2 cos βX3 

ν
and
l2 = l21 + l22 + l23 = µ2 β2 X21 + X22 + ν2

Since X21 + X22 = a2 , the result is

p
l= µ2 β2 a2 + ν2
(2) To calculate the initial length, we need the inverse of the deformation gradient. This is
µν cos βX3 −µν sin βX3 µ2 βX1 sin βX3 − µ2 βX2 cos βX3
 
1 
µν sin βX3 µν cos βX3 −µ2 βX1 cos βX3 − µ2 βX2 cos βX3 
−1
FiA =
νµ2
0 0 µ2
The initial length is

   
L1 −1 0
 L2  = F  0 
iA
L3 1
µν cos βX3 −µν sin βX3 µ2 βX1 sin βX3 − µ2 βX2 cos βX3
  
0
1 
= µν sin βX3 µν cos βX3 −µ2 βX1 cos βX3 − µ2 βX2 cos βX3   0 
νµ2
0 0 µ2 1
µ βX1 sin βX3 − µ βX2 cos βX3
2 2
 
1 
= −µ 2
βX1 cos βX3 − µ2 βX2 cos βX3 
νµ2
µ2
The initial length squared is
1 2 2
L2 = L21 + L22 + L23 = β X1 + X22 + 1

ν2
and
1p 2 2
L= β a +1
ν
Problem 4.36
A velocity field is defined in terms of the spatial coordinates and time by the equations,
v1 = 2tx1 sin x3 , v2 = 2tx2 cos x3 , v3 = 0 .
At the point (1, −1, 0) at time t = 1, determine

(a) the rate of deformation tensor and the vorticity tensor,
√
^ = (^
(b) the stretch rate per unit length in the direction of the normal n ^2 + e
e1 + e ^) / 3,
(c) the maximum stretch rate per unit length and the direction in which it occurs,
(d) the maximum shear strain rate.
Answer
   
0 0 1 0 0 1
(a) [dij ] =  0 2 0 , [wij ] =  0 0 0 
1 0 0 −1 0 0
(b) Λ̇/Λ = 4/3

(c) Λ̇/Λ max = 2, ^ =e
n ^2
(d) γ̇max = 1.5
Solution
(a) The velocity gradient is
2t sin x3 2tx1 cos x3

 
0
∂vi
= 0 2t cos x3 −2tx2 sin x3 
∂xj 0 0 0
and  
0 2 0
∂vi
= 0
 0  2
∂xj (1,−1,0),t=1 0 0 0

The rate of deformation tensor is D = 12 grad v + (grad v)T or
     
0 0 2 0 0 0 0 0 1
1 
[dij ] = 0 2 0  +  0 2 0  =  0 2 0 
2 0 0 0 2 0 0 1 0 0

The vorticity tensor is W = 12 grad v − (grad v)T
     
0 0 2 0 00 0 0 1
1 
[wij ] = 0 2 0 − 0 2
0  =  0 0 0 
2 0 0 0 2 00 −1 0 0
√
(b) The rate of stretching along n^ = (^ ^) / 3 is
^2 + e
e1 + e
  
0 0 1 √1
Λ̇ h i  3  4
= dij ni nj = √1 √1 √1  0 2 0   √1  =
  
Λ 3 3 3   3  3
1 0 0 √1
3
(c) The principal rate of stretching is

−λ 0 1

0 2 − λ 0 = (−λ) (2 − λ) (−λ) − (2 − λ) = 0

1 0 −λ
= (2 − λ) λ2 − 1 = 0

and
λ(1) = 2; λ(2) = 1; λ(3) = −1
The maximum rate of stretching is

Λ̇
=2
Λ max
The direction is the eigenvector associated with the eigenvalue 2. This is

    
−2 0 1 n1 0
 0 0 0   n2  =  0 
1 0 −2 n3 0
or
−2n1 + n3 = 0
n1 − 2n3 = 0
Solving gives
n1 = n3 = 0
or
^ (1) = e
n ^2
(d) The maximum shear strain is found from γ̇max ^ 2 where the vectors n
^1 · D · n
=n ^ 1 and
^ 2 are perpendicular. Thus
n
  
2 0 0 √1
h i  2  3
γ̇max = √ 1 1  
=

0 √2  0 1 0  0
2    2
0 0 −1 − √12
Problem 4.37
Let NA and ni denote direction cosines of a material line element in the reference and
current configurations, respectively. Beginning with Eq 4.140, ni Λ = xi,A NA , and using
the indicial notation throughout, show that
(a) Λ̇/Λ = dij ni nj ,
(b) Λ̈/Λ = Qij ni nj + ṅi nj where Qij = 21 (ai,j + aj,i ) with ai being the components
of acceleration.
Solution
(a) Taking the time derivative gives
ṅi Λ + ni Λ̇ = ẋi,A NA = vi,A NA

since NA is fixed. Taking the dot product with ni gives
ṅi ni Λ + ni ni Λ̇ = vi,A ni NA
Now, ni ni = 1 and 2ṅi ni = 0. This gives

∂vi ∂vi ∂xj ∂vi
Λ̇ = ni NA = ni NA = ni nj Λ
∂XA ∂xj ∂XA ∂xj
where Eq 4.140 was used. Thus
Λ̇ ∂vi
= ni nj = (dij + wij ) ni nj = dij ni nj
Λ ∂xj
since wij ni nj = 0.
(b) We begin by taking the time derivative of
ṅi Λ + ni Λ̇ = ẋi,A NA = vi,A NA
to obtain
∂ai ∂xj
n̈i Λ + 2 ṅi Λ̇ + ni Λ̈ = v̇i,A NA = ai,A NA = NA
∂xj ∂XA
Taking the dot product with ni gives
∂ai ∂xj
n̈i ni Λ + 2 ṅi ni Λ̇ + ni ni Λ̈ = ni NA
∂xj ∂XA
or
∂ai ∂xj
n̈i ni Λ + Λ̈ = NA
∂xj ∂XA
Now
d
(ni ṅi ) = ṅi ṅi + ni n̈i = 0
dt
or
ni n̈i = −ṅi ṅi
This gives
∂ai ∂xj ∂ai
−ṅi ṅi Λ + Λ̈ = ni NA = ni nj Λ
∂xj ∂XA ∂xj
and

Λ̈ ∂ai ai,j + aj,i ai,j − aj,i
= ni nj + ṅi ṅi = + ni nj + ṅi ṅi
Λ ∂xj 2 2

ai,j + aj,i
= ni nj + ṅi ṅi = Qij ni nj + ṅi ṅi
2
The second term in parentheses with ni nj is zero since it is antisymmetric.
Problem 4.38
In a certain region of flow the velocity components are
v1 = x31 + x1 x22 e−kt , v2 = x32 − x21 x2 e−kt ,

v3 = 0
where k is a constant, and t is time in s. Determine at the point (1, 1, 1) when t = 0,

(a) the components of acceleration,

(b) the principal values of the rate of deformation tensor,
(c) the maximum shear rate of deformation.
Answer
(a) a1 = 2(4 − k), a2 = −4, a3 = 0

(b) D(1) = 4, D(2) = 2, D(3) = 0
(c) γ̇max = ±2
Solution
(a) The acceleration is
dvi ∂vi ∂vi
ai = = + vj
dt ∂t ∂xj
Now
a1 = −k x31 + x1 x22 e−kt + 3x21 + x22 e−kt x31 + x1 x22 e−kt

+ (2x1 x2 ) e−kt x31 − x1 x22 e−kt

and
a1 |(1,1,1),t=0 = 2 (4 − k)
Also
a2 = −k x32 − x21 x2 e−kt + (−2x1 x2 ) e−kt x32 + x21 x2 e−kt

+ 3x22 − x21 e−kt x32 − x21 x2 e−kt

and
a2 |(1,1,1),t=0 = −4
Finally,
a3 = 0

(b) The rate of deformation tensor is D = 12 grad v + (grad v)T

3x21 + x22 e−kt (2x1 x2 ) e−kt
 
0
∂vi
=  (−2x1 x2 ) e−kt 3x22 − x21 e−kt 0 
∂xj
0 0 0
At the given point and time, we have

 
4 2 0
∂vi
=  −2 2 0 
∂xj (1,1,1),t=0 0 0 0
The rate of deformation tensor is

 
4 0 0
[dij ] =  0 2 0 
0 0 0
and   
4 0 0 √1
h i  2 
γ̇max = √1 √1  = ±2
  
2
0 2
 0 2 0  0
  
0 0 0 − √12
where the ± results from the selection of the orthogonal vectors.
Problem 4.39
For the motion
x1 = X1 ,
x2 = X2 et + X1 (et − 1) ,
x3 = X1 (et − e−t ) + X3 ,
determine the velocity field vi = vi (x), and show that for this motion
(a) L = Ḟ · F−1 ,
(b) D = ˙ at t = 0.
Solution
(a) Inverting the equations of motion gives
X1 = x1
X2 = x2 e−t − x1 1 − e−t

X3 = x3 − x1 et − e−t


 
1 0 0
∂xi
[FiA ] = =  et − 1 et 0 
∂XA et − e−t 0 1
and  
1 0 0
∂XA
F−1 −t
e−t

Ai = =  e −1 0 
∂xi − (et − e−t ) 0 1
Also  
0 0 0
et et

ḞiA = 0 
et + e−t 0 0
The velocity is

dx1
v1 = =0
dt X

dx2
v2 = = x1 + x2
dt X

dx3
= x1 et + e−t

v3 =
dt X
and  
0 0 0
[Lij ] = [vi,j ] =  1 1 0 
et + e−t 0 0
Also, we have
  
h −1 i 0 0 0 1 0 0
et et e−t − 1 e−t

[Lij ] = ḞiA FAj =  0  0 
et + e−t 0 0 − (et − e−t ) 0 1
 
0 0 0
= 1 1 0 
et + e−t 0 0
(b) The rate of deformation tensor is D = 1
2 grad v + (grad v)T . This is at t = 0
et + e−t
   
0 0 0 0 1
1 
[dij ] = 1 1 0 + 0 1 0 
2 et + e−t 0 0 0 0 0
t −t
 
0 1 e +e
1
=  1 2 0 
2 t
e +e −t
0 0
At t = 0  
0 1 2
1
[dij ] = 1 2 0 
2 2 0 0
u1 = x1 − X1 = 0
u2 = x2 − X2 = (X1 + X2 ) et − 1 = (x1 + x2 ) 1 − e−t

u3 = x3 − X3 = X1 et − e−t = x1 et − e−t

and = 12 grad u + (grad u)T = 21 (ui,j + uj,i )
0 1 − e−t et − e−t
   
0 0 0
1  −t −t
[ij ] = 1−e 1−e 0 + 0 1 − e−t
  0 
2 t
e −e −t
0 0 0 0 0
−t t −t
 
0 1−e e −e
1
=  1 − e−t 2 (1 − e−t ) 0 
2 t
e −e −t
0 0
The time derivative is
e−t et + e−t
 
0
1
[˙ ij ] = e−t 2e−t 0 
2 e + e−t
t
0 0
At t = 0  
0 1 2
1
[˙ ij ]t=0 =  1 2 0 
2 2 0 0
This is identical to the rate of deformation D.
Problem 4.40
Determine an expression for the material derivative d (ln dx) /dt in terms of the rate of
^ = dx/dx.
deformation tensor D and the unit normal n
Answer
^ ·D·n
d (ln dx) /dt = n ^
Solution
The derivative is
.
d (ln dx) dx
=
dt dx
and
.

d dx dx
Λ̇ = =
dt dX dX
So that
.
d (ln dx) dx dX Λ̇
= = Λ̇ = ^·D·n
=n ^
dt dx dx Λ
from Problem 4.37.
Problem 4.41
A velocity field is given in spatial form by
v1 = x1 x3 , v2 = x22 t, v3 = x2 x3 t .
(a) Determine the vorticity tensor W and the vorticity vector w.

(b) Verify the equation εpqi wi = wqp for the results of part (a).
(c) Show that at the point (1, 0, 1) when t = 1, the vorticity tensor has only one real
root.
Answer
(a) w1 = 21 x3 t, w2 = 21 x1 , w3 = 0
√ √
(c) W(1) = 0, W(2) = −W(3) = ı/ 2, where ı = −1
Solution
(a) The vorticity tensor is W = 1
2 grad v − (grad v)T where (grad v)ij = ∂xj .
∂vi
This gives
 
x3 0 x1
[vi,j ] =  0 2x2 t 0 
0 x3 t x2 t
This gives
   
x 0 x1 x3 0 0
1  3
[wij ] = 0 2x2 t 0 − 0 2x2 t x3 t  
2 0 x3 t x2 t x1 0 x2 t
 
0 0 x1
1
= 0 0 −x3 t 
2 −x x3 t 0
1
The vorticity vector is

1
w= curl v or wi = εijk vk,j
2
So that
1 1
w1 = (v3,2 − v2,3 ) = (x3 t)
2 2
1 1
w2 = (v1,3 − v3,1 ) = (x1 )
2 2
1
w3 = (v2,1 − v1,2 ) = 0
2
(b) For εpqi wi = wqp , we have
1 1
ε123 w3 = w21 = 0; ε231 w1 = w32 = (x3 t) ; ε312 w2 = w13 = x1
2 2
and
1 1
ε213 w3 = −w3 = w12 = 0; ε132 w2 = −w2 = w31 = − x1 ; ε321 w1 = −w1 = w23 = − (x3 t)
2 2
(c) At the point (1, 0, 1) and t = 1, the vorticity tensor is
 
0 0 1
1
[wij ](1,0,1),t=1 =  0 0 −1 
2 −1 1 0
The eigenvalues are

−λ 0 1
2
1 1
det (wij − λδij ) = 0 −λ − 12 = −λ λ2 + −λ =0

4 4
1 1
−2 2 −λ

1
= −λ λ2 + =0
2
or
1
λ(1) = 0; λ(2) , λ(3) = ±i √
2
Problem 4.42
Consider the velocity field
v1 = ex3 −ct cos ωt, v2 = ex3 −kt sin ωt, v3 = c
where c and ω are constants.
(a) Show that the speed of every particle is constant.

(b) Determine the acceleration components ai .
(c) Calculate the logarithmic stretching, Λ̇/Λ = d (ln Λ) /dt,√for the element in the
^ = (^
current configuration in the direction of n ^2 ) / 2 at x = 0.
e1 + e
Answer
(b) a1 = −ωv2 , a2 = ωv1 , a3 = 0

(c) 12 e−ct cos ωt
Solution
The square of the speed of a particle is
v2 = v21 + v22 + v23 = e2(x3 −ct) cos2 ωt + e2(x3 −ct) sin2 ωt + c2

= e2(x3 −ct) + c2
If the particle has constant speed, the material time derivative of the square of the speed
should be zero. This is

d v2 ∂ v2 ∂ v2

∂v ∂v
= + vi = 2v + vi
dt ∂t ∂xi ∂t ∂xi

= −2c e2(x3 −ct) + 0 + 0 + c 2e2(x3 −ct) = 0
(b) The acceleration components are

dvi ∂vi ∂vi
ai = = + vk
dt ∂t ∂xk
Then
∂v1 ∂v1
a1 = + vk = −cex3 −ct cos ωt − ωex3 −ct sin ωt + 0 + 0 + cex3 −ct cos ωt
∂t ∂xk
= −ωex3 −ct sin ωt = −ωv2
∂v2 ∂v2
a2 = + vk = −cex3 −ct sin ωt + ωex3 −ct cos ωt + 0 + 0 + cex3 −ct sin ωt
∂t ∂xk
= ωex3 −ct cos ωt = ωv1
and
∂v3 ∂v3
a3 =
+ vk =0
∂t ∂xk
(c) From Problem 4.40 the logarithmic stretching is
d (ln Λ) Λ̇
= ^·D·n
=n ^
dt Λ
Now dij = 1
2
(vi,j + vj,i ) where
ex3 −ct cos ωt

 
0 0
∂vi
[vi,j ] = = 0 0 ex3 −ct sin ωt 
∂xj 0 0 0
and
ex3 −ct cos ωt
 
0 0
1
[dij ] = 0 0 ex3 −ct sin ωt 
2 ex3 −ct cos ωt ex3 −ct sin ωt 0
At x = 0, we have
  √1 
e−ct cos ωt

0 0
Λ̇  2 
h i1
= √1 0 √1  0 0 e−ct sin ωt   0 
Λ 2 2 2 e−ct cos ωt e−ct sin ωt 0 √1
2
e−ct cos ωt
=
2
Problem 4.43
Show that the velocity field
v1 = 1.5x3 − 3x2 , v2 = 3x1 − x3 , v3 = x2 − 1.5x1
corresponds to a rigid body rotation, and determine the axis of spin (the vorticity vector).
Answer
^1 + 1.5e
w=e ^ 2 + 3e
^3
Solution
The velocity gradient is
 
0 −3 1.5
∂vi
[Lij ] = = 3 0 −1 
∂xj −1.5 1 0
The rate of deformation tensor is identically zero, D = 0. This is a rigid motion. Since
the velocity gradient is skew symmetric,
W=L
The vorticity vector is
w1 = w32 = 1; w2 = w13 = 1.5; w3 = w21 = 3
or
^1 + 1.5^
w=e e2 + 3^
e3
Problem 4.44
For the steady velocity field
v1 = x21 x2 , v2 = 2x22 x3 , v3 = 3x1 x2 x3
^3 ) /5.
e1 − 3 e
determine the rate of extension at (2, 0, 1) in the direction of the unit vector (4^
Answer
48
Λ̇/Λ = −
25
Solution
x21
 

2x1 x2 0
∂vi
[Lij ] = = 0 4x2 x3 2x22 
∂xj
3x2 x3 3x1 x3 3x1 x2
and
x21
 
4x1 x2 3x2 x3
1 ∂vi ∂vj 1
[dij ] = + =  x21 8x2 x3 3x1 x3 + 2x22 
2 ∂xj ∂xi 2
3x2 x3 3x1 x3 + 2x22 6x1 x2
At the point (2, 0, 1), we have

 
0 2 0
[dij ](2,0,1) = 2 0 3 
0 3 0
and   4 
0 2 0
Λ̇ 4
5 48
− 35

= 5 0  2 0 3   − 53  = −
Λ 0 3 0 25
0
Problem 4.45
Prove that d(ln J) /dt = div v and, in particular, verify that this relationship is satisfied for
the motion
x1 = X1 + ktX3 , x2 = X2 + ktX3 , x3 = X3 − kt(X1 + X2 )
where k is a constant.
Answer
J̇/J = div v = 4k2 t/(1 + 2k2 t2 )
Solution
The material time derivative is .
J̇ det F
=
J J
and from Eq 4.160
.
det F = det F tr Ḟ · F−1 = J tr L = J div v

This gives
.
J̇ det F 1
= = (J div v) = div v
J J J
For the motion, the velocity field is
v1 = kX3 ; v2 = kX3 ; v3 = −k (X1 + X2 )


 
1 0 kt
∂xi
[FiA ] = = 0 1 kt 
∂XA −kt −kt 1
and
1 0 kt
J = det F = 0

1 kt = 1 + 2k2 t2

−kt −kt 1
This gives
J̇ = 4k2 t
Inverting the motion, we have

1 + k2 t2 x1 − k2 t2 x2 − ktx3
X1 =
1 + 2k2 t2

−k t x1 + 1 + k2 t2 x2 − ktx3
2 2
X2 =
1 + 2k2 t2
ktx1 + ktx2 + ktx3
X3 =
1 + 2k2 t2
Substituting into the velocity field gives
k (ktx1 + ktx2 + ktx3 ) k (ktx1 + ktx2 + ktx3 ) −k (x1 + x2 − 2ktx3 )

v1 = ; v2 = ; v3 =
1 + 2k2 t2 1 + 2k2 t2 1 + 2k2 t2
Then
k2 t k2 t 2k2 t
vi,i = v1,1 + v2,2 + v3,3 = 2 2
+ 2 2
+
1 + 2k t 1 + 2k t 1 + 2k2 t2
4k2 t J̇
= 2 2
=
1 + 2k t J
Problem 4.46
Prove the identity
xi,A
=0, J = det(xi,A )
J ,i
Solution Computing the derivative gives

xi,A (xi,A ),i xi,A J,i
= −
J ,i J J2
∂xi,A ∂XC
(xi,A ),i = = xi,AC XC,i
∂XC ∂xi
Use
∂J
= XC,j J
∂xj,C
to get
∂J ∂xj,C ∂xj,C
J,i = = XC,j J = XC,j Jxj,CB XB,i
∂xj,C ∂xi ∂xi
Substitution gives

xi,A xi,AC XC,i xi,A XC,j Jxj,CB XB,i
= −
J ,i J J2
xi,AC XC,i − δAB XC,j xj,CB

=
J
xi,AC XC,i − XC,j xj,CA

=
J
Assume xi are smooth functions of XA such that order of partial differentiation does not
matter.
xi,A xi,AC XC,i − XC,j xj,AC
= =0
J ,i J
Problem 4.47
Equation 4.151 gives the material derivative of dx2 in terms of dij . Using that equation as
the starting point, show that d2 (dx2 )/dt2 is given in terms of dij and its time derivative
by
d2 (dx2 )/dt2 = 2(ḋij + vk,i dkj + vk,j dik )dxi dxj .
Solution
Equation 4.151 is
d dx2
= dx · 2D · dx = 2dij dxi dxj
dt
and !
d2 dx2 d d dx2 . .
2
= = 2ḋij dxi dxj + 2dij dxi dxj + 2dij dxi dxj
dt dt dt
.
Now dxi = vi,j dxj . This gives

d2 dx2
= 2ḋij dxi dxj + 2dij vi,k dxk dxj + 2dij vj,k dxk dxi
dt2
= 2 ḋij + dkj vk,i + dik vk,j dxi dxj
where we have changed the dummy indices.
Problem 4.48
A continuum body in the form of the unit cube shown by the sketch undergoes the homo-
geneous deformation
x1 = λ1 X1 , x2 = λ2 X2 , x3 = λ3 X3
where λ1 , λ2 , and λ3 are constants.
Determine the relationships among λ1 , λ2 , and λ3 if
(a) the length of diagonal OC remains unchanged,

(b) the rectangular area ABFE remains unchanged,
(c) the triangular area ACE remains unchanged.
X3 , x3 X3 , x3
E E
F F
D D
C C
O O
G G
A A
X2 , x2 X2 , x2
B B
X1 , x1 X1 , x1
(a) (b)
X3 , x3
E
F
D
C
O
G
A
X2 , x2
B
X1 , x1
(c)
Answer
(a) λ21 + λ22 + λ23 = 3

(b) λ22 λ21 + λ23 = 2
(c) λ21 λ22 + λ22 λ23 + λ23 λ21 = 3
Solution
(a) The deformation gradient, Jacobian and right deformation tensor are
λ21
   
λ1 0 0 0 0
[FiA ] =  0 λ2 0 , J = det F = λ1 λ2 λ3 , T
C = F · F, [CAB ] =  0 λ22 0 
0 0 λ3 0 0 λ23
The length OC is
  
λ21 0 0 √1
h i  3 
Λ2 = 1 = √1 √1 √1 λ22 √1
  
 0 0  
3 3 3   3 
0 0 λ23 √1
3
1 2
λ1 + λ22 + λ33 = 1

=
3
(b) Equation 4.159 gives
dSq = XA,q JdS0A
Now  
1
 λ1 0 0
∂XA 1
 h
−1
i
[XA,q ] = = 0 0  = FAq
 
∂xq  λ2 
1
0 0 λ3
This gives
 
1
λ1 0 0
h i h i 
 0
√1 √1 dS = λ1 λ2 λ3 √1 √1  0 λ1 0  dS

2
0 2 2
0 2 2
 
1
0 0 λ3
h i
= λ1 λ2 λ3 √1 0 √1 dS0
2λ1 2λ3
and
(λ1 λ2 λ3 )2

1 1 2
dS2 = 2
+ 2 dS0
2 λ1 λ2
2
Since dS2 = dS0 , we have
2 = λ22 λ23 + λ21 λ23 = λ23 λ22 + λ21

(c) For the triangular area,

 
1
λ1 0 0
h i h i 
 0
√1 − √13 √1 dS = λ1 λ2 λ3 √1 − √13 √1 1
0  dS

 0 λ2
3 3 3 3  
1
0 0 λ3
h i
= λ√
2 λ3
− λ√
1 λ3 λ√
1 λ2 dS0
3 3 3
and
1 2 2 2
dS2 = λ2 λ3 + λ21 λ23 + λ21 λ22 dS0

3
Now dS/dS0 = 1, and
λ22 λ23 + λ21 λ23 + λ21 λ22 = 3
Problem 4.49
Let the unit cube shown in Problem 4.47 be given the motion
1 1
x1 = X1 + t2 X2 , x2 = X2 + t2 X1 , x3 = X3
2 2
Determine, at time t,
(a) the rate-of-change of area ABFE,
(b) the volume of the body.
Answer
.
^1 /(1 − 14 t4 ) − te
(a) dS = t3 e ^ 2 − t3 e
^3
(b) V = (1 − 4 t )
1 4
Solution
(a) The time rate of change of the area is , Eq 4.162,
.
dSi = vk,k dSi − vi,j dSj

v1 = X2 t; v2 = X1 t; v3 = 0
Inverting the motion gives
x1 − 12 t2 x2 x2 − 12 t2 x1
X1 = 1 4
; X2 = ; X3 = x3
1− 4t 1 − 14 t4
t x2 − 12 t2 x1 t x1 − 21 t2 x2

v1 = ; v2 = ; v3 = 0
1 − 14 t4 1 − 14 t4

 t3 
−2 t 0
1 3
[vi,j ] =  t − t2 0 
1 − 41 t4
0 0 0
and
t3
vk,k = −
1 − 14 t4
This gives
.
dSi = vk,k dSi − vi,j dSj
3
− t2
 
t 0
t3 h
√1 √1
i 1 h
√1 √1
i
3
=− 0 dS − 0  t − t2 0  dS
1 − 14 t4 2 2
1 − 41 t4 2 2
0 0 0
1 h
t3 3
i
= −√ √t t
−√ dS
1 − 14 t4 2 2 2
(b) The current volume is given by the Jacobian,

1 t2 0

2 2 1 2
J = det F = t2

1 0 = 1 − t
0 4
0 1
Problem 4.50
For the homogeneous deformation
x1 = X1 + αX2 + αβX3 ,
x2 = αβX1 + X2 + β2 X3 ,
x3 = X1 + X2 + X3 ,
where α and β are constants, determine the relationship between these constants if the
deformation is isochoric.
Answer
β = (α2 + α)/(α2 + α + 1)
Solution
For an isochoric motion J = det F = 1. The deformation gradient is
 
1 α αβ
∂xi
[FiA ] = =  αβ 1 β2 
∂Xa
1 1 1
and
1 α αβ

J = αβ 1 β2 = 1 + αβ2 + α2 β2 − αβ − β2 − α2 β = 1
1 1 1
This gives
β αβ + α2 β − α − β − α2 = 0

or
β α + α2 − 1 − α − α2 = 0

The relation is
α2 + α
β=
α2 + α − 1
Problem 4.51
Show that for any velocity field v derived from a vector potential ψ by v = curl ψ, the flow
is isochoric. Also, for the velocity field
v1 = ax1 x3 − 2x3 , v2 = −bx2 x3 , v3 = 2x1 x2
determine the relationship between the constants a and b if the flow is isochoric.
Answer
a=b
Solution
v = curl ψ or vi = εijk ψk,j
For an isochoric motion, div v = 0. This gives
vi,i = εijk ψk,ji = 0
since the permutation symbol is skew symmetric in ji. For the velocity field,
vi,i = ax3 − bx3 + 0 = 0
and
a=b
Chapter 5 Solutions
Problem 5.1
Determine the material derivative of the flux of any vector property Q∗i through the spatial
area S. Specifically, show that
Z Z
d
Q∗i ni dS = Q̇∗i + Q∗i vk,k − Q∗k vi,k ni dS

dt S S
in agreement with Eq 5.5.
Solution
Transcribing to symbolic notation
Z Z Z
d d d
Q∗i ni dS = Q∗i dSi = Q∗ · dS
dt S dt S dt S
Transforming to the reference configuration

Z Z Z
d d d
Q∗ · dS = Q∗ · JdS0 · F−1 = Q∗ · JdS0 · F−1 Eq 4.159

dt S dt 0 S0 dt
Z S .
= Q̇∗ · JdS0 · F−1 + Q∗ · J̇dS0 · F−1 + Q∗ · JdS0 · (F−1 )
S0
Now . . .
(F · F−1 ) = (I) = 0 = Ḟ · F−1 + F · (F−1 )
or .
(F−1 ) = −F−1 · Ḟ · F−1 = −L · F−1 and J̇ = J div v
so that R R
RQ̇ +∗Q · div
0
Q∗ · dS
∗ ∗
d
v − Q∗ · L · JdS

dt S = S0 · F−1
= S Q̇ + Q · div v − Q · L · dS
∗ ∗

In indicial form Z Z
d
Q∗ · dS = Q̇∗i + Q∗i vk,k − Q∗k vi,k ni dS

dt S S
Problem 5.2
∗
Let the property Pij... in Eq 5.1 be the scalar 1 so that the integral in that equation
represents the instantaneous volume V. Show that in this case
Z Z
d
Ṗij... = dV = vi,i dV .
dt V V
131
Solution
From
Z Z
d d
dV = JdV0 (Eq 4.163)
dt V dt V0
Z Z Z Z
d 0
0 0
= JdV = J̇dV = Jvi,i dV = vi,i dV
V0 dt V0 V0 V
Problem 5.3
Verify the identity
εijk ak,j = 2 (ẇi + wi wj,j − wj vi,j ) ,
and by using this identity as well as the result of Problem 5.1, prove that the material
derivative of the vorticity flux equals one half the flux of the curl of the acceleration; that
is, show that Z Z
d 1
wi ni dS = εijk ak,j ni dS .
dt S 2 S
Solution
The acceleration is
dvi ∂vi ∂vi
ai = = + vk
dt ∂t ∂xk
∂vi ∂vi ∂vk ∂vk ∂vi ∂vk
= + vk − vk + vk = + vk (2Wik ) + vk
∂t ∂xk ∂xi ∂xi ∂t ∂xi
But
Wik = εkip wp (Eq 4.154)
This gives
∂vi ∂vk ∂vi

ai = + vk (2εkip wp ) + vk = + 2vk εipk wp + vk vk,i
∂t ∂xi ∂t
Taking the curl of a

∂vi
εpqi ai,q = εpqi + 2εpqi εijk (vk wp ),q + εpqi (vk vk,i ),q
∂t ,q
∂
= (εpqi vi,q ) + 2 (δpj δqk − δpk δqj ) (vk,q wp + vk wp,q )
∂t
+εpqi (vk,q vk,i + vk vk,iq )
The last term is zero because of the skew symmetric - symmetric product. Now

∂wp
εpqi ai,q = 2 + vk wp,k +2vk,k wp −2wq vp,q −2vp wj,j = 2 (ẇp + wp vk,k − wk vp,k )
∂t
From Problem 5.1

Z Z Z
d 1
wi ni dS = (ẇi + wi vk,k − wk vi,k ) ni dS = εijk ak,j ni dS
dt S S S 2
Problem 5.4
Making use of the divergence theorem of Gauss together with the identity
∂wi 1
= εijk ak,j − εijk εkmq (wm vq ),j
∂t 2
show that Z Z
∂ 1
wi dV = εijk ak + wj vi − wi vj nj dS .
∂t V S 2
Solution
From Problem 5.3
∂wi
εijk ak,j = 2 + 2εijk εkpq (wp vq ),j
∂t
Z Z Z
∂ ∂wi 1
wi dV = dV = εijk ak,j − εijk εkpq (wp vq ),j dV
∂t V V ∂t V 2
Z
1
= εijk ak − (δip δjq − δiq δjp ) (wp vq ) nj dS
S 2
Z
1
= εijk ak − wi vj + wj vi nj dS
S 2
Problem 5.5
Show that the material derivative of the vorticity of the material contained in a volume V
is given by Z Z
d 1
wi dV = εijk ak + wj vi nj dS .
dt V S 2
Solution
From Eq 5.3
Z Z Z
d ∂wi
wi dV = dV + vk wi nk dS
dt V ∂t
ZV S
Z
1
= εijk ak − wi vj + wj vi nj dS + vk wi nk dS (Prob. 5.4)
S 2 S
Z
1
= εijk ak + wj vi nj dS
S 2
Problem 5.6
Given the velocity field
q
v1 = ax1 − bx2 , v2 = bx1 + ax2 , v3 = c x21 + x22
where a, b, and c are constants, determine

(a) whether or not the continuity equation is satisfied,
(b) whether the motion is isochoric.
Answer
(a) only when ρ = ρ0 e−2at (b) only if a = 0
Solution
(a) Continuity requires
ρ̇ + ρ div v = ρ̇ + ρvi,i = 0
Now
div v = v1,1 + v2,2 + v3,3 = a + a + 0
and
ρ̇
ρ̇ + ρ (2a) = 0 or = −2a
ρ
Integrating gives
ln ρ = −2at + C
If at t = 0, ρ = ρ0 then C = ln ρ0 and
ρ = ρ0 e−2at
(b) For an isochoric motion div v = vi,i = 2a = 0. The motion is isochoric only if a = 0.
Problem 5.7
For a certain continuum at rest, the stress is given by
tij = −p0 δij
where p0 is a constant. Use the continuity equation to show that for this case the stress
power may be expressed as
p0 ρ̇
tij dij = .
ρ
Solution
The stress power is
tij dij = −p0 δij dij = −p0 dii = −p0 vi,i
From the continuity equation
ρ̇
ρ̇ + ρvi,i = 0 or vi,i = −
ρ
and
p0 ρ̇
tij dij =
ρ
Problem 5.8
Consider the motion xi = (1 + t/k)Xi where k is a constant. From the conservation of mass
and the initial condition ρ = ρ0 at t = 0, determine ρ as a function of ρ0 , t, and k.
Answer
ρ0 k 3
ρ=
(k + t)3
Solution
The motion xi = (1 + t/k) Xi inverts to
xi
Xi =
t
1+
k
The velocity is

dxi Xi x xi
vi = = = i =
dt Xi k t (k + t)
k 1+
k
The divergence of the velocity field is
xi,i δii 3
vi,i = = =
(k + t) (k + t) (k + t)
Continuity requires
3 dρ 3
ρ̇ + ρvi,i = ρ̇ + ρ =0 or =− dt
(k + t) ρ (k + t)
Integration gives
ln ρ = − ln (k + t)3 + ln C where C = ρ0 k3
and
ρ0 k3
ρ=
(k + t)3
Problem 5.9
By combining Eqs 5.17b and 5.13, verify the result presented in Eq 4.161.
Solution
Eqs 5.17b and 5.13 are
.
(ρJ) = ρ̇J + ρJ̇ = 0
and
ρ̇ + ρvi,i = 0
Combining
−ρvi,i J + ρJ̇ = 0 or J̇ = vi,i J
the desired result, Eq 4.161.
Problem 5.10
Using the identity
εijk ak,j = 2 (ẇi + wi vj,j − wj vi,j )
as well as the continuity equation, show that

d wi εijk ak,j + 2wj vi,j
= .
dt ρ 2ρ
Solution
Differentiating

d wi ẇi wi ẇi wi
= − 2 ρ̇ = + vk,k
dt ρ ρ ρ ρ ρ
where the continuity equation ρ̇ + ρvk,k = 0 was used. Now

d wi 1 1
= εijk ak,j − wi vj,j + wj vi,j + wi vj,j
dt ρ ρ 2
1
= (εijk ak,j + 2wj vi,j )
2ρ
Problem 5.11
State the equations of motion and from them show by the use of the material derivative
∂vi
v̇i = + vj vi,j ,
∂t
and the continuity equation that
∂ (ρvi )
= (tij − ρvi vj ),j + ρbi .
∂t
Solution
The equations of motion are
∂vi
tij,j + ρbi = ρv̇i = ρ + ρvj vi,j
∂t
Rearranging
∂vi
tij,j + ρbi − ρvj vi,j − ρvi vj,j − ρ,j vi vj = ρ − ρvi vj,j − ρ,j vi vj
∂t
or
∂ (ρvi ) ∂ρ
(tij − ρvi vj ),j + ρbi = − vi − ρvi vj,j − ρ,j vi vj
∂t ∂t
∂ (ρvi ) ∂ρ
= − vi + ρ,j vj − ρvi vj,j
∂t ∂t
∂ (ρvi )
= − vi (ρ̇ + ρvj,j )
∂t
The last term is zero from the continuity equation, ρ̇ + ρvj,j = 0.
Problem 5.12
Determine the form which the equations of motion take if the stress components are given
by tij = −pδij where p = p(x, t).
Answer
ρai = −p,i + ρbi

Solution
ρv̇i = tij,j + ρbi
So that
ρv̇i = (−pδij ),j + ρbi = −p,j δij + ρbi
or
ρai = −p,i + ρbi
Problem 5.13
Let a material continuum have the constitutive equation
tij = αδij dkk + 2βdij
where α and β are constants. Determine the form which the equations of motion take in
terms of the velocity gradients for this material.
Answer
ρv̇i = ρbi + (α + β)vj,ij + βvi,jj
Solution
ρv̇i = tij,j + ρbi
Now
tij,j = (αδij dkk + 2βdij ),j = αδij dkk,j + 2βdij,j

= αdkk,i + 2βdij,j = αvk,ki + β (vi,jj + vj,ij )
= (α + β) vj,ji + βvi,jj
and
ρv̇i = (α + β) vj,ji + βvi,jj + ρbi
Problem 5.14
Assume that distributed body moments mi act throughout a continuum in motion. Show
that the equations of motion are still valid in the form of Eq 5.22, but that the angular
momentum principle now requires
εijk tjk + mi = 0
implying that the stress tensor can no longer be taken as symmetric.
Solution
The balance of linear momentum, Eq 5.22, is unaffected by the distributed body moments.
The balance of angular momentum, Eq 5.50 is modified by adding an additional term
Z . Z Z
εijk (xj vk )ρ dV = εijk xj tqk nq dS + (εijk xj ρbk + mi ) dV
V
Z ZS h V
i
εijk (vj vk + xj ak ) ρ dV = εijk (xj tqk ),q + (εijk xj ρbk + mi ) dV
V S
Since εijk vj vk = 0,
Z
[εijk xj ρak − εijk (xj,q tqk + xj tqk,q ) − (εijk xj ρbk + mi )] dV = 0
V
Recognizing xj,q = δjq and rearranging

Z
[εijk xj (ρak − ρbk − tqk,q ) − (εijk tjk + mi )] dV = 0
V
The first term is zero from the equations of motion leaving
εijk tjk + mi = 0
Problem 5.15
For a rigid body rotation about the origin, the velocity field may be expressed by vi =
εijk ωj xk where ωj is the angular velocity vector. Show that for this situation the angular
momentum principle is given by
.
Mi = (ωj Iij )
where Mi is the total moment about the origin of all surface and body forces, and Iij is the
moment of inertia of the body defined by the tensor
Z
Iij = ρ (δij xk xk − xi xj ) dV .
V
Solution
The angular momentum principal is
Z Z
d d
Mi = εijk xj ρvk dV = εijk xj ρεkmn ωm xn dV
dt V dt V
Z
d
= (δim δjn − δin δjm ) xj ρωm xn dV
dt V
Z
d
= (xj ρωi xj + xj ρωj xi ) dV
dt V
Z .
d
= ωj ρ (δij xk xk + xj xi ) dV = (ωj Iij )
dt V
Problem 5.16
Determine expressions for the stress power tij dij in terms of
(a) the first Piola-Kirchoff stress tensor,

(b) the second Piola-Kirchoff stress tensor.
Answer
(a) tij dij = ρḞiA PiA /ρ0 (b) tij dij = ρsAB ĊAB /2ρ0
Solution
(a) From Eq 5.44
Jtji = PAi xj,A
or
ρxi,A PAj
tij =
ρ0
since J = ρ0 /ρ. Now
ρvj,i xi,A PAj
tij dij = tij vj,i =
ρ0
but L · F = Ḟ so
ρḞjA PAj
tij dij = tij vi,j =
ρ0
The summation makes this equivalent to the answer in the text.
(b) The second Piola-Kirchoff stress tensor gives
ρxi,A xj,B dij sAB

tij dij =
ρ0
From Eq 4.149 xi,A xj,B dij = ĖAB and 2EAB = CAB + δAB from Eq 4.45, so that
ĊAB
ĖAB =
2
and
ρsAB ĊAB
tij dij =
2ρ0
Problem 5.17
Show that, for a rigid body rotation about the origin, the kinetic energy integral Eq 5.53
reduces to the form given in rigid body dynamics, that is,
1
K= ωi ωj Iij
2
where Iij is the inertia tensor defined in Problem 5.15.
Solution
The kinetic energy is
Z
1
K (t) = ρvi vi dV
2 V
The rigid velocity field is vi = εijk ωj xk and

Z Z
1 1
K (t) = ρεijk ωj xk εipq ωp xq dV = ρ (δjp δkq − δjq δkp ) ωj xk ωp xq dV
2 V 2 V
Z Z
1 1
= ρ (ωj ωk δjk xk xk − ωj ωk xj xk ) dV = ωj ωk ρ (δjk xq xq − xj xk ) dV
2 V 2 V
1
= ωj ωk Ijk
2
Problem 5.18
Show that one way to express the rate of change of kinetic energy of the material currently
occupying the volume V is by the equation
Z Z Z
n)
(^
K̇ = ρbi vi dV − tij vi,j dV + vi ti dS
V V S
and give an interpretation of each of the above integrals.
Solution
The rate of change of the kinetic energy is
Z Z Z
d 1
K̇ (t) = ρvi vi dV = ρv̇i vi dV = vi (tij,j + ρbi ) dV
dt 2
Z hV V
i
V
Z Z
= (vi tij ),j − vi,j tij + ρvi bi dV = (ρvi bi − vi,j tij ) dV + vi tij nj dS
ZV Z V S
n)
(^
= (ρvi bi − vi,j tij ) dV + vi ti ndS
V S
The first term is the rate of work of the body forces. The second term is the stress
power, and the final term is the rate of surface force working.
Problem 5.19
Consider a continuum for which the stress is tij = −p0 δij and which obeys the heat con-
duction law qi = −κθ,i . Show that for this medium the energy equation takes the form
ρu̇ = −p0 vi,i + ρr + κθ,ii .
Solution
From Eq 5.65, the energy equation is
ρu̇ − tij dij − ρr + qi,i = 0
The stress power is
tij dij = −p0 δij vi,j = −p0 vi,i
This gives
ρu̇ = −p0 vi,i + ρr + κθ,ii
Problem 5.20
If mechanical energy only is considered, the energy balance can be derived from the equa-
tions of motion. Thus, by forming the scalar product of each term of Eq 5.22 with the
velocity vi and integrating the resulting equation term-by-term over the volume V, we
obtain the energy equation. Verify that one form of the result is
.
1
ρ(v · v) + tr (T · D) − ρb · v − div (T · v) = 0 .
2
Solution
The equation of motion multiplied by the velocity and integrated over the volume is
Z Z
(tij,j vi + ρbi vi ) dV = (ρv̇i vi ) dV
V V
. . .
but (vi vi ) = 2v̇i vi and v̇i vi = 21 (vi vi ) = 12 (v · v). Now
(tij vi ),j = tij,j vi + tij dij or tij,j vi = div (T · v) − tr (T · D)
Therefore Z .
1
div (T · v) − tr (T · D) + ρb · v − ρ(v · v) dV = 0
V 2
The result follows from localization.
Problem 5.21
Show that for a continuum body experiencing volumetric growth
Z Z
d
ρf dV = ρḟ + ρcf dV .
dt P P
Solution
Begin by transferring to the reference configuration
Z Z Z
d
ρf dV =
d
[ρ (X, t) f (X, t)] JdV0 = ˙ dV0
(ρfJ)
dt P dt P0 P0
Z Z
ρ̇fJ + ρḟJ + ρfJ̇ dV0 = ρ̇f + ρḟ + ρfvk,k JdV0

=
P0 P0
Transferring back to the current configuration gives

Z Z
d
ρf dV = ρḟ + (ρ̇ + ρvk,k ) f dV
dt P P0
Using the mass conservation equation ρ̇ + ρvk,k = ρc, the result is obtained.
Problem 5.22
Using the local form of the balance of mass, Eq 5.90, show that the balance of linear
momentum for a material with volumetric growth is Eq 5.93
dvi
ρ = tji,j + ρbi .
dt
Solution
The balance of linear momentum for a growing body is
Z Z Z Z
d n)
(^
ρvi dV = ti dS + ρbi dV + ρcvi dV
dt P ∂P P P
Using the result for the derivative of an integral with a growing mass, we have
Z Z Z Z Z
d n)
(^
ρvi dV = ρv̇i + ρcvi dV = ti dS + ρbi dV + ρcvi dV
dt P P ∂P P P
This leaves Z Z Z
n)
(^
ρv̇i dV = ti dS + ρbi dV
P ∂P P
n)
(^
and using ti = tji nj together with the divergence theorem yields the desired result.
Problem 5.23
Using the local form of the balance of mass, Eq 5.90, and linear momentum, Eq 5.93, show
that the stress in a volumetrically growing body is symmetric.
Solution
The balance of moment of momentum is
Z Z Z
d n)
(^
εijk xj ρvk dV = εijk xj tk dS + εijk xj ρ (bk + cvk ) dV
dt P ∂P P
The derivative of the first integral is

Z Z Z
d
εijk xj ρvk dV = εijk ρ (ẋj vk + xj v̇k + cxj vk ) dV = εijk ρ (xj v̇k + cxj vk ) dV
dt P P P
The first term is the cross product of a vector with itself. The remaining terms are
Z Z Z
n)
(^
εijk ρ (xj v̇k + cxj vk ) dV = εijk xj tk dS + εijk xj ρ (bk + cvk ) dV
P ∂P P
The remaining integral is

Z Z Z
n)
(^
εijk ρxj v̇k dV = εijk xj tk dS + εijk xj ρbk dV
P ∂P P
This is the balance of angular momentum and symmetry follows directly.
Problem 5.24
If a continuum has the constitutive equation
tij = −pδij + αdij + βdik dkj
where p, α and β are constants, and if the material is incompressible (dii = 0), show that
tii = −3p − 2βIID
where IID is the second invariant of the rate of deformation tensor.
Solution
Computing the trace of the constitutive equation gives
tii = −pδii + αdii + βdik dki and tii = −3p + βdik dki
Now IID = 1
2
(dii djj − dik dki ) = 1
2
(−dik dki ) for an incompressible material. Thus
tii = −3p − 2βIID
Problem 5.25
Starting with Eq 5.173 for isotropic elastic behavior, show that
tii = (3λ + 2µ)ii ,

and using this result, deduce that

1 λ
ij = tij − δij tkk .
2µ 3λ + 2µ
Solution
Equation 5.173 gives
tij = λδij kk + 2µij
for a linear isotropic elastic material. Taking the trace gives
tii = λδii kk + 2µii = (3λ + 2µ) ii
Solving and substituting for ii yields
λ
tij = δij tkk + 2µij
3λ + 2µ
or
1 λ
ij = tij − δij tkk
2µ 3λ + 2µ
Problem 5.26
For a Newtonian fluid, the constitutive equation is given by
tij = −pδij + τij

= −pδij + λ∗ δij dkk + 2µ∗ dij (see Eq 5.175) .
By substituting this constitutive equation into the equations of motion, derive the equation
ρv̇i = ρbi − p,i + (λ∗ + µ∗ )vj,ji + µ∗ vi,jj .
Solution
The equation of motion is
tij,j + ρbi = ρv̇i
Substituting the constitutive equation yields
ρv̇i = (−pδij + λ∗ δij dkk + 2µ∗ dij ) ,j +ρbi

= −p,j δij + λ∗ δij dkk,j + 2µ∗ dij,j + ρbi
= −p,i + λ∗ dkk,i + 2µ∗ dij,j + ρbi
= −p,i + (λ∗ + µ∗ ) vj,ji + µ∗ vi,jj
Problem 5.27
Combine Eqs 5.177a and Eq 5.177b into a single viscoelastic constitutive equation having
the form
tij = δij {R}kk + {S}ij
where the linear time operators {R} and {S} are given, respectively,
3K {P} − 2 {Q} 2 {Q}

{R} = and {S} = .
3 {P} {P}
Solution
Equations 5.177a and 5.177b are
{P}Sij = 2{Q}ηij
tii = 3Kii
Substituting for the stress and strain deviators gives

1 1
{P} tij − δij tkk = 2 {Q} ij − δij kk
3 3
This gives

1
{P} (tij − δij Kkk ) = 2 {Q} ij − δij kk
3
or
1
{P} (tij ) = 2 {Q} ij − δij kk + {P} δij Kkk
3
Rearranging
2 {Q} 2 {Q}

tij = ij + δij K − kk
{P} 3 {P}
Problem 5.28
Assume a viscoelastic medium is governed by the constitutive equations Eq 5.177. Let a
slender bar of such material be subjected to the axial tension stress t11 = σ0 f(t) where σ0 is
a constant and f(t) some function of time. Assuming that t22 = t33 = t13 = t23 = t12 = 0,
determine 11 , 22 , and 33 as functions of {P}, {Q}, and f(t).
Answer
3K {P} + {Q}
11 = σ0 f (t)
9K {Q}
2 {Q} − 3K {P}
22 = 33 = σ0 f (t)
18K {Q}
Solution
From the constitutive equations, Eq 5.177, and the given data
tii = t11 + t22 + t33 = t11 and so t11 = 3Kkk
and
1 1 1
{P} tij − δij t11 = 2 {Q} ij − δij kk = 2 {Q} ij − δij t11
3 3 9K
For i = j = 1,
1 1
{P} t11 − δ11 t11 = 2 {Q} 11 − δ11 t11
3 9K
or
2 {P} 2 {Q} 3K {P} + {Q}

2 {Q} 11 = + t11 or 11 = t11
3 9K 9K {Q}
For i = j = 2,
1 1
{P} − δ22 t11 = 2 {Q} 22 − δ22 t11
3 9K
or
2 {Q} {P} 2 {Q} − 3K {P}

2 {Q} 22 = − t11 or 22 = t11
9K 3 18K {Q}
For i = j = 3
1 1
{P} − δ33 t11 = 2 {Q} 33 − δ33 t11
3 9K
or
2 {Q} {P} 2 {Q} − 3K {P}

2 {Q} 33 = − t11 or 33 = t11
9K 3 18K {Q}
where t11 = σ0 f (t)
Problem 5.29
Use the definition of the free energy along with the reduced form of the Clausius-Duhem
equation to derive the local dissipation inequality.
Solution
The free energy is ψ = u − ηθ and the reduced form of the Clausius-Duhem equation is,
Eq 5.84
1
ρθη̇ − ρu̇ + dij tij − qi gi > 0
θ
Taking the time derivative of the free energy gives, ψ̇ = u̇ − η̇θ − ηθ̇ or u̇ = ψ̇ + η̇θ + ηθ̇.
Substituting for u̇ gives
1
ρθη̇ − ρ ψ̇ + η̇θ + ηθ̇ + dij tij − qi gi > 0
θ
or
1
−ρ ψ̇ + ηθ̇ + dij tij − qi gi > 0
θ
(See A. E. Green and N. Laws, On the Formulation of Constitutive Equations in Thermo-
dynamical Theories of Continua, Quarterly Journal of Mechanics and Applied Mathematics,
Vol 20, 1967, Eq 2.11)
Problem 5.30
The constitutive model for a compressible, viscous, and heat-conducting material is defined
by

ψ = ψ̃ θ, gk , FiA , ḞiA ,

η = η̃ θ, gk , FiA , ḞiA ,

tij = t̃ij θ, gk , FiA , ḞiA ,

qi = q̃i θ, gk , FiA , ḞiA .
Deduce the following restrictions on these constitutive response functions:

∂ψ̃ θ, gk , FkB , ḞkB
(a) = 0,
∂gi

∂ψ̃ θ, gk , FkB , ḞkB
(b) = 0,
∂ḞiA
∂ψ̃ (θ, FkB )
(c) η̃ (θ, FiA ) = − ,
∂θ
∂ψ̃ (θ, FkB )
(d) t̃ij = ρFjA ,
∂FiA
1
(e) − q̃i (θ, gk , FiA , 0) gi > 0.
θ
Solution
Beginning with the free energy
∂ψ̃ ∂ψ̃ ∂ψ̃ ∂ψ̃

ψ̇ = θ̇ + ġk + ḞiA + F̈iA
∂θ ∂gk ∂FiA ∂ḞiA
Substituting into the reduced Clausius-Duhem inequality, Eq 5.85 and collecting terms,
gives

∂ψ̃ ∂ψ̃ ∂ψ̃ ∂ψ̃ 1
−ρ + η̃ θ̇ + tij − ρ ḞiA − ρ ġk − F̈iA − qi gi > 0
∂θ ∂FiA ∂gk ∂ḞiA θ
The quantities θ, gi ,FiA , ḞiA , and F̈iA can be chosen arbitrarily by Eq 5.103. Select a
process with FiA and θ = θ0 , a constant. Also, assume a uniform temperature field so
that gi = 0. Then
∂ψ̃
− F̈iA > 0
∂ḞiA
Since F̈iA can be chosen arbitrarily, this inequality can be violated unless
∂ψ̃
− = 0 which implies ψ = ψ̃ (θ, FiA , gi )
∂ḞiA
Now proceed as in the text from Eq 5.99. (See Coleman and Mizel, 1964.)
Problem 5.31
Assume the constitutive relationships
u = ũ (CAB , η) ,
θ = θ̃ (CAB , η) ,
tij = t̃ij (CAB , η) ,
qi = q̃i (CAB , η, gk ) ,
for an elastic material. Use the Clausius-Duhem inequality to show

∂ũ
θ= ,
∂η
u = ũ (CAB , η) ,

1 ∂ũ ∂ũ
tij = ρFiA + FjB ,
2 ∂CAB ∂CBA
q̃i gi > 0 .
Solution
Begin with the Clausius-Duhem inequality, Eq 5.84
1
ρθη̇ − ρu̇ + dij tij − q i gi > 0
θ
Now
∂ũ ∂ũ
u̇ = ĊAB + η̇
∂CAB ∂η
Combining
∂ũ ∂ũ 1
ρ θ− η̇ − ρ ĊAB + dij tij − qi gi > 0
∂η ∂CAB θ
Now
ĊAB = ḞiA FiB + FiA ḞiB = Lij FjA FiB + FiA Lij FjB
= 2FiA FjB Lij = 2FiA FjB dij
where dummy indices were switched and dij replaces Lij since only the symmetric part
is retained in the product.

∂ũ ∂ũ 1
ρ θ− η̇ + tij − 2ρFiA FjB Lij − q̃i gi > 0
∂η ∂CAB θ
We know η̇, Lij , and gi are independent and can be chosen independently. Choose a
thermodynamic process in which η̇ = 0 and Lij = 0 to conclude
1
q̃i gi > 0 or q̃i gi > 0
θ
since the temperature is always positive.
Now choose a process where η̇ = 0 and gi = 0. Then

∂ũ
tij − 2ρFiA FjB Lij > 0
∂CAB
Since this must hold for all process and Lij may be chosen independently, for this in-
equality to always hold
∂ũ
tij = 2ρFiA FjB
∂CAB
where it is assumed that
∂ũ 1 ∂ũ ∂ũ
= +
∂CAB 2 ∂CAB ∂CBA
Finally choose Lij = 0 and gi = 0 resulting in

∂ũ
ρ θ− η̇ > 0
∂η
which must hold for all processes. Since η̇ is arbitrary, it coefficient must be zero for the
inequality to hold. This gives
∂ũ
θ=
∂η
Problem 5.32
Use the basic kinematic result of superposed rigid body motion given in Eq 5.121b to show
the following:
∂x+i
(a) = Qik ,
∂xk
(b) F+
iA = Qij FjA .
Solution
a. From Eq 5.121b
x+
i = ai + Qim xm
differentiating gives
∂x+
i ∂ai ∂xm
= + Qim = Qim δmj = Qij
∂xj ∂xj ∂xj
b. The deformation gradient is
F+ +
iA = xi,A = (ai + Qim xm ),A = Qim xm,A = Qij FjA
where dummy indices have been exchanged.
Problem 5.33
Show that the Jacobian transforms as follows under a superposed rigid body motion:
J+ = J .
Solution
By definition J = det FiA and J+ = det F+
iA = det (Qij FjA ) = det Qij det FjA = +1 det FjA =
J where the results form Problem 5.32b. was used.
Problem 5.34
Utilize the result of Problem 5.33 along with the law of conservation of mass to show that
ρ+ = ρ.
Solution
The continuity equation gives
ρ+ = ρ0 det J+ = ρ0 det J = ρ
Problem 5.35
Show that the gradient of the stress components transforms under superposed rigid body
motion as follows:
∂t+ij ∂tmn
+ = Qim .
∂xj ∂xn
Solution
The divergence of t+
ij is
∂t+
ij ∂t+
ij ∂xn ∂xn ∂
= = + (Qim Qjq tmq )
∂x+
j ∂xn ∂x+j ∂xj ∂xn
where we have used the chain rule and the transformation equation for stress. Now
from x+
k = ak + Qkq xq , Qkn Qkq xq = δnq xq = xn = Qkn xk − ak since Q is a proper
+

orthogonal tensor. Then
∂xn ∂x+
k
= Q kn = Qkn δkj = Qjn
∂x+
j ∂x+
j
and
∂t+

ij ∂tmq ∂tmq ∂tmn
= Qjn Qim Qjq = δnq Qim = Qim
∂x+
j ∂xn ∂xn ∂xn
Problem 5.36
Use the superposed rigid body motion definitions to show the following relationships:
(a) C+
AB = CAB ,
(b) U+
AB = UAB ,
(c) R+
iA = Qij RjA ,
(d) B+
ij = Qim Bmk Qkj .
Solution
From Problem 5.32
(a) C+ + +
AB = FiA FiB = Qij FjA Qik FkB = δjk FjA FkB = FjA FjB = CAB
(b) U+ + +
AB = FiA RiB = Qij FjA Qik RkB = δjk FjA RkB = FkB RkB = UAB
(c) F+
iA = RiB UAB = RiB UAB from result b. Now FiA = Qij FjA = RiB UAB or
+ + + + +
+ −1
RiB = Qij FjA UAB = Qij RjB
(d) B+ + +
ij = FiA FjA = Qim FmA Qjn FnA = Qim Qjn Bmn
Problem 5.37
In the context of rigid body dynamics, consider the motion defined by
x (X, t) = X
along with
x = QT (t) x+ − a (t)

show that
v+ = ȧ (t) + ω (t) × x+ − a (t)

where ω is the angular velocity of the body.
Solution
The velocity is
v+ = ẋ+ = ȧ + Q̇x + Qẋ
but for a rigid body
ẋ = Ẋ = 0
and
v+ = ȧ + Q̇x = ȧ + ΩQx = ȧ + ΩQ QT x+ − a

= ȧ + Ω x+ − a = ȧ + ω × x+ − a

where ω is the axial vector of the skew tensor Ω. Eq 5.133 was used in the final step
î = −εijk ωk bj e
Ωij bj e î = εikj ωk bk e
î
or
Ωb = ω × b
Chapter 6 Solutions
Problem 6.1
In general, the strain energy density W may be expressed in the form
W = C∗αβ α β (α, β = 1, . . . , 6)
where C∗αβ is not necessarily symmetric. Show that this equation may be rearranged to
appear in the form
1
W = Cαβ α β
2
where Cαβ is symmetric, so that now
∂W
= Cαβ β = tβ
∂β
in agreement with Eq 6.8.
Solution
Let
C∗αβ + C∗βα C∗αβ − C∗βα
C∗αβ = +
2 2
Now

C∗αβ + C∗βα C∗αβ − C∗βα
W= α β + α β
2 2
Cαβ + C∗βα
∗
1
= α β = Cαβ α β
2 2
and
∂W 1 1 1
= Cαβ δγα β + Cαβ α δγβ = (Cγβ β + Cαγ α )
∂γ 2 2 2
But
Caβ = Cβα
and
∂W 1
= (Cβγ β + Cαγ α ) = Cαγ α
∂γ 2
Problem 6.2
Let the stress and strain tensors be decomposed into their respective spherical and deviator
components. Determine an expression for the strain energy density W as the sum of a
dilatation energy density W(1) and a distortion energy density W(2) .
151
Answer
1 1
W = W(1) + W(2) = tii jj + Sij ηij
6 2
Solution
The strain energy density is

1 1 1 1
W = tij ij = Sij + δij tkk ηij + δij kk
2 2 3 3

1 1 1 1
= Sij ηij + ηii tkk + Sii kk + δii tkk jj .
2 3 3 9
But Sii = ηii = 0 and

1 1
W= tkk jj + Sij ηij
6 2
since δii = 3.
Problem 6.3
If the strain energy density W is generalized in the sense that it is assumed to be a function
of the deformation gradient components instead of the small strain components, that is, if
W = W(FiA ), make use of the energy equation and the continuity equation to show that in
this case Eq 6.16 is replaced by
∂W
Jtij = FjA .
∂FiA
Solution
The strain energy density is
W = ρ0 u
and
Ẇ = ρ0 u̇
Now from the energy equation
1 1
u̇ = tij dij = tij vi,j
ρ ρ
and
ρ0
Ẇ = tij vi,j
ρ
Now we also have for W (FiA )
∂W (FiA ) ∂W (FiA )
Ẇ (FiA ) = ḞiK = vj,q FqK
∂FjK ∂FjK
Then from the expressions for Ẇ, we have
ρ0 ∂W (FqA )
Ẇ = tij vi,j = vi,j FjK
ρ ∂FiK
and
∂W (FqK )
Jtij = FjA
∂FiA
Problem 6.4
For an isotropic elastic medium as defined by Eq 6.23, express the strain energy density in
terms of
(a) the components of ij ,
(b) the components of tij ,
(c) the invariants of ij .
Answer
1
(a) W = (λii jj + 2µij ij ),
2
(3λ + 2µ) tij tij − λtii tjj
(b) W = ,
4µ (3λ + 2µ)
(c) W = 12 λ + µ (I )2 − 2µII

Solution
(a) For the isotropic linear material, tij = 2µij + λδij kk and
1 1 1
W= tij ij = ij (2µij + λδij kk ) = (2µij ij + λii kk )
2 2 2

1 λ
(b) From Eq 6.23, tii = (3λ + 2µ) ii and ij = tij − δij tkk and
2µ (3λ + 2µ)

1 1 λ
W = tij ij = tij tij − tjj tkk
2 4µ (3λ + 2µ)
(c) I = tr = ii and II = (ii jj − ij ij ) or ij ij = I2 − 2II . From (a)
1
2

1 2 λ
λI + 2µ I2 − 2II =

W= + µ I − 2µII
2 2
Problem 6.5
Let tij be any second-order isotropic tensor such that
t0ij = aim ajn tmn = tij
for any proper orthogonal transformation aij . Show that by successive applications of the
transformations
   
0 0 −1 0 0 1
[aij ] =  −1 0 0  and [aij ] =  −1 0 0 
0 1 0 0 −1 0
every second-order isotropic tensor is a scalar multiple of the Kronecker delta, δij .
Solution
For the first transformation
     
t11 t12 t13 0 0 −1 t11 t12 t13 0 −1 0
 t21 t22 t23  =  −1 0 0   t21 t22 t23   0 0 1 
t31 t32 t33 0 1 0 t31 t32 t33 −1 0 0
 
t33 t31 −t32
=  t13 t11 −t12 
−t23 −t21 t22
Thus t11 = t22 = t33 , t12 = −t31 , t13 = −t32 , t21 = −t13 , t23 = −t12 , t31 = −t23 ,
t32 = −t21 , or we can write t12 = t31 = −t23 ; t21 = t13 = −t32 (a)
For the second transformation
     
t11 t12 t13 0 0 1 t11 t12 t13 0 −1 0
 t21 t22 t23  =  −1 0 0   t21 t22 t23   0 0 −1 
t31 t32 t33 0 −1 0 t31 t32 t33 1 0 0
 
t33 −t31 −t32
=  −t13 t11 t12 
−t23 t21 t22
Here, we have t11 = t22 = t33 , t12 = −t31 , t13 = −t32 , t21 = −t13 , t23 = t12 , t31 = −t23 ,
t32 = t21 , or we can write t12 = −t31 = t23 ; t21 = −t13 = −t32 (b) Combining (a) and (b)
gives t12 = t31 = t23 = t21 = t13 = t32 = 0. Now let t11 = t22 = t33 = λ, then tij = λδij .
Problem 6.6
Verify that Eqs 6.33a and 6.33b when combined result in Eq 6.28 when Eqs 6.29 are used.
Solution
From Eqs 6.33a and 6.33b

1 1 1
tij − δij tkk = 2G ij − δij kk = 2G ij − δij tkk
3 3 9K
or

1 2G − 3K 1+ν −3ν
ij = tij + δij tkk = tij + δij tkk
2G 9K E 3E
1+ν ν
= tij − δij tkk
E E
Problem 6.7
For an elastic medium, use Eq 6.33 to express the result obtained in Problem 6.2 in terms
of the engineering elastic constants K and G.
Answer
W = 12 Kii jj + G ij ij − 13 ii jj

Solution
From Problem 6.2
tii jj Sij ηij 3Kii jj 2Gηij ηij

W= + = +
6 2 6 2
K 1 1
= ii jj + G ij − δij kk ij − δij mm
2 3 3

K 1
= ii jj + G ij ij − ii kk
2 3
Problem 6.8
Show that the distortion energy density W(2) (see Problem 6.2) for a linear elastic medium
may be expressed in terms of (a) the principal stresses, σ(1) , σ(2) , σ(3) and (b) the principal
strains, (1) , (2) , (3) in the form
2 2 2
σ(1) − σ(2) + σ(2) − σ(3) + σ(3) − σ(1)
(a) W(2) = ,
12G
1h 2 2 2 i
(b) W(2) = (1) − (2) + (2) − (3) + (3) − (1) G.
3
Solution
(a) For the stresses,

1 1 1 1
W(2) = Sij Sij = tij − δij tkk tij − δij tkk
4G 4G 3 3

1 1
= tij tij − tii tjj
4G 3
Note that tii = σ(1) + σ(2) + σ(3) in terms of the principal stresses, σ(i) , and

tij tij = σ(1) + σ(2) + σ(3) σ(1) + σ(2) + σ(3) − 2 σ(1) σ(2) + σ(2) σ(3) + σ(3) σ(1)
Then

1 2 2
W(2) = −2 σ(1) σ(2) + σ(2) σ(3) + σ(3) σ(1) + σ(1) + σ(2) + σ(3)
4G 3
1 h 2 i
= −6 σ(1) σ(2) + σ(2) σ(3) + σ(3) σ(1) + 2 σ(1) + σ(2) + σ(3)
12G
1
= −6 σ(1) σ(2) + σ(2) σ(3) + σ(3) σ(1)
12G
i
+2 σ2(1) + σ2(2) + σ2(3) + 2σ(1) σ(2) + 2σ(2) σ(3) + 2σ(3) σ(1)
1 h 2 2 2 i
= σ(1) − σ(2) + σ(2) − σ(3) + σ(3) − σ(1)
12G
(b) Similarly for the strains,

1 1 1
W(2) = Gηij ηij = G ij − δij kk ij − δij kk = G ij ij − ii jj
3 3 3

2 2
= G −2 (1) (2) + (2)(3) + (3)(1) + (1) + (2) + (3)
3
G
= −6 (1) (2) + (2) (3) + (3)(1)
3 i
+2 2(1) + 2(2) + 2(3) + 2(1) (2) + 2(2) (3) + 2(3)(1)
Gh 2 2 2 i
= (1) − (2) + (2) − (3) + (3) − (1)
3
Problem 6.9
Beginning with the definition for W (Eq 6.21a), show for a linear elastic material represented
by Eq 6.23 or by Eq 6.28 that ∂W/∂ij = tij and ∂W/∂tij = ij (Note that ∂ij /∂mn =
δim δjn .)
Solution
From Eq 6.21a, we have using Eq 6.23
1 1
W= tij ij = (2µij + λδij kk ) ij
2 2
Now
∂W 1
= [4µδim δjn ij + (λδij δkm δkn ij + λδij kk δim δjn )]
∂mn 2
1
= [4µmn + (λδmn ii + λδmn kk )] = 2µmn + λδmn kk
2
= tmn

1 1
W= tij ij = [(1 + ν) tij − νδij tkk ] tij
2 2E
and
∂W 1
= [2 (1 + ν) δim δjn tij − ν (δij δkm δkn tij + δij δim δjn tkk )]
∂tmn 2E
1
= [2 (1 + ν) tmn − ν (δmn tjj + δmn tkk )]
2E
1
= [(1 + ν) tmn − νδmn tkk ] = mn
E
Problem 6.10
For an isotropic, linear elastic solid, the principal axes of stress and strain coincide, as was
shown in Example 6.1. Show that, in terms of engineering constants E and ν this result is
given by
(1 + ν) σ((q) − ν σ((1) + σ((2) + σ((3)
(q) = , (q = 1, 2, 3).
E
Thus, let E = 106 psi and ν = 0.25, and determine the principal strains for a body subjected
to the stress field (in ksi)  
12 0 4
[tij ] =  0 0 0  .
4 0 6
Answer
(1) = −4.5 × 10−6 , (2) = 0.5 × 10−6 , (3) = 13 × 10−6
Solution
Consider
1
[(1 + ν) tij − νδij tkk ]
ij =
E
Let n
^ (q) , (q = 1, 2, 3) be the eigenvectors corresponding to σ(1) , σ(2) , σ(3) . Then
(q) 1 (q)
ij nj = [(1 + ν) tij − νδij tkk ] nj
E
but (q)
tij − δij σ(q) nj = 0
This gives
(q) 1h (q) (q)

i
ij nj = (1 + ν) δij σ(q) nj − νδij tkk nj
E
1 (q)
= (1 + ν) σ(q) − νtkk ni
E
For principal strains
(q)
ij − δij (q) nj = 0
or
(q) (q) 1 (q)
ij nj = (q) ni = (1 + ν) σ(q) − νtkk ni
E
and
1
(q) = (1 + ν) σ(q) − ν σ(1) + σ(2) + σ(3)
E
since tkk is aninvariant.   
12 0 4 0 0 0
For [tij ] =  0 0 0  the matrix of principal stresses is σ(q) =  0

4 0 
4 0 6 0 0 14
Now
1
ij = 6 [(1.25) tij − (0.25) δij (18)]
10
and  
10.5 0 5
[ij ] =  0 −4.5 0  10−6
5 0 3
This gives  
−4.5 0 0
0  10−6

(q) = 0 0.5
0 0 13
From
1
(q) = (1 + ν) σ(q) − ν σ(1) + σ(2) + σ(3)
E
and
(1) = −4.5 × 10−6 (2) = 0.5 × 10−6 (3) = 13 × 10−6
Problem 6.11
Show for an isotropic elastic medium that
1 2 (λ + µ) ν λ 2µν 3Kν
(a) = , (b) = , (c) = ,
1+ν 3λ + 2µ 1+ν λ + 2µ 1 − 2ν 1+ν
(d) 2µ(1 + ν) = 3K(1 − 2ν).
Solution
(a)
1 1 2 (λ + µ)
= =
1+ν λ 3λ + 2µ
1+
2 (λ + µ)
(b)
λ
ν 2 (λ + µ) λ
= =
1−ν λ λ + 2µ
1−
2 (λ + µ)
(c)
3κ (1 − 2ν)
2ν
2µν 2 (1 + ν) 3Kν
= =
1 − 2ν 1 − 2ν 1+ν
(d)
3κ (1 − 2ν)
2µ(1 + ν) = 2 (1 + ν) = 3K(1 − 2ν)
2 (1 + ν)
Problem 6.12
Let the x1 x3 plane be a plane of elastic symmetry such that the transformation matrix
between Ox1 x2 x3 and Ox01 x02 x03 is
 
1 0 0
[aij ] =  0 −1 0  .
0 0 1
Show that, as the text asserts, this additional symmetry does not result in a further reduc-
tion in the elastic constant matrix, Eq 6.40.
Solution
The transformation is
     
1 0 0 t1 t6 t5 1 0 0 t1 −t6 t5
 0 −1 0   t6 t2 t4   0 −1 0  =  −t6 t2 −t4 
0 0 1 t5 t4 t3 0 0 1 t5 −t4 t3
Thus t0α = tα for (α = 1, 2, 3, 5) and t0a = −ta for (α = 4, 6). Similarly 0α = α for
(α = 1, 2, 3, 5) and 0a = −a for (α = 4, 6). These conditions are identical to those satis-
fied in deriving Eq 6.40.
Problem 6.13
Let the x1 axis be an axis of elastic symmetry of order N = 2. Determine the form of the
elastic constant matrix Cαβ , assuming Cαβ = Cβα .
Answer
 
C11 C12 C13 C14 0 0

 C12 C22 C23 C34 0 0 

 C13 C23 C33 C34 0 0 
[Cαβ ] =  

 C14 C24 C34 C44 0 0 

 0 0 0 0 C55 C56 
0 0 0 0 C56 C66
Solution
2π
For N = 2, we have θ = π since N = . Thus
θ
 
1 0 0
[aij ] =  0 −1 0 
0 0 −1
and
     
1 0 0 t1 t6 t5 1 0 0 t1 −t6 −t5
 0 −1 0   t6 t2 t4   0 −1 0  =  −t6 t2 t4 
0 0 −1 t5 t4 t3 0 0 −1 −t5 t4 t3
Thus t0α = tα for (α = 1, 2, 3, 4) and t0a = −ta for (α = 5, 6). Similarly 0α = α for
(α = 1, 2, 3, 4) and 0a = −a for (α = 5, 6). This gives
t01 = C11 01 + C12 02 + C13 03 + C14 04 + C15 05 + C16 060
and
t1 = C11 1 + C12 2 + C13 3 + C14 4 + C15 5 + C16 6
or
t1 = C11 01 + C12 02 + C13 03 + C14 04 − C15 05 − C16 06
The result is C16 = C61 = C15 = C51 = 0. Likewise

from t02 = t2 , C25 = C52 = C26 = C62 = 0
from t03 = t3 , C35 = C53 = C36 = C63 = 0
from t04 = t4 , C45 = C54 = C46 = C64 = 0, also
t05 = C55 05 + C56 06 = −C55 5 − C56 6 = −t5

t06 = C56 05 + C66 06 = −C56 5 − C66 6 = −t6
Problem 6.14
Assume that, by the arguments of elastic symmetry, the elastic constant matrix for an
isotropic body has been reduced to the form
 
C11 C12 C12 0 0 0
 C12 C11 C12 0 0 0 
 
 C12 C12 C11 0 0 0 
[Cαβ ] = 
 0
 .
 0 0 C 44 0 0 

 0 0 0 0 C44 0 
0 0 0 0 0 C44
Show that, if the x1 axis is taken as an axis of elastic symmetry of any order (θ is arbitrary),
C11 = C12 + 2C44 . (Hint: Expand t023 = a2q a3m tqm and 023 = a2q a3m qm .)
Solution
 
1 0 0
[aij ] =  0 cos θ sin θ 
0 − sin θ cos θ
Now from
t023 = a2q a3m tqm = − sin θ cos θt22 + cos2 θ − sin2 θ t23 + sin θ cos θt33

or in single subscript notation
t04 = − sin θ cos θt2 + cos2 θ − sin2 θ t4 + sin θ cos θt3

Similarly,
023 = a2q a3m qm = − sin θ cos θ22 + cos2 θ − sin2 θ 23 + sin θ cos θ33

or
04 = −2 sin θ cos θ2 + cos2 θ − sin2 θ 4 + 2 sin θ cos θ3

Now
t04 = C44 044 = 2C44 sin θ cos θ (3 − 2 ) + C44 cos2 θ − sin2 θ 4

Also from the stress transformation and the constitutive equation, we have
t04 = − sin θ cos θ (C12 1 + C11 2 + C12 3 ) + cos2 θ − sin2 θ C44 4

+ sin θ cos θ (C12 1 + C12 2 + C11 3 )

= sin θ cos θ (C11 − C22 ) (3 − 2 ) + cos2 θ − sin2 θ C44 4

Comparing the expressions for t04 , we find
2C44 = C11 − C22
Problem 6.15
If the axis which makes equal angles with the coordinate axes is an axis of elastic symmetry
of order N = 3, show that there are twelve independent elastic constants and that the elastic
matrix has the form

 
C11 C12 C13 C14 C15 C16

 C13 C11 C12 C16 C14 C15 

 C12 C13 C11 C15 C16 C14 
[Cαβ ] =   .

 C41 C42 C43 C44 C45 C46 

 C43 C41 C42 C46 C44 C45 
C42 C43 C41 C45 C46 C44
Solution
2π
For N = 3, we have N = = 3 and θ = 120◦ . The transformation matrix is
θ
 
0 1 0
[aij ] =  0 0 1 
1 0 0
This gives
     
0 1 0 t1 t6 t5 0 0 1 t2 t4 t6
 0 0 1   t6 t2 t4   1 0 0  =  t4 t3 t5 
1 0 0 t5 t4 t3 0 1 0 t6 t5 t1
Thus we have
t01 = t2 , t02 = t3 , t03 = t1 , t04 = t5 , t05 = t6 , t06 = t4
and
01 = 2 , 02 = 3 , 03 = 1 , 04 = 5 , 05 = 6 , 06 = 4
Now
t01 = C11 01 + C12 02 + C13 03 + C14 04 + C15 05 + C16 06
and
t2 = C21 1 + C22 2 + C23 3 + C24 4 + C25 5 + C26 6
Using the strain transformations, we have
t2 = C21 03 + C22 01 + C23 02 + C24 06 + C25 04 + C26 05
Equating t01 = t2 yields
C11 = C22 , C12 = C23 , C13 = C21 , C14 = C25 , C15 = C26 , C16 = C24
Likewise from t02 = t3
C21 = C32 , C22 = C33 , C23 = C31 , C24 = C35 , C25 = C36 , C26 = C34
from t03 = t1
C31 = C12 , C32 = C13 , C33 = C11 , C34 = C15 , C35 = C16 , C36 = C14
from t04 = t5
C41 = C52 , C42 = C53 , C43 = C51 , C44 = C55 , C45 = C56 , C46 = C54
from t05 = t6
C51 = C62 , C52 = C63 , C53 = C61 , C54 = C65 , C55 = C66 , C56 = C64
from t06 = t4
C61 = C42 , C62 = C43 , C63 = C41 , C64 = C45 , C65 = C46 , C66 = C44
Problem 6.16
For an elastic body whose x3 axis is an axis of elastic symmetry of order N = 6, show that
the nonzero elastic constants are C11 = C22 , C33 , C55 = C44 , C66 = 12 (C11 − C12 ), and
C13 = C23 .
Solution
2π
For N = 6, we have N = = 6 and θ = 60◦ . The transformation matrix for the x3 as a
θ
symmetry axis is  √ 
1 2
0
 2√ 2
[aij ] =  − 2 1

2 2 0 
0 0 1
and T 0 = AT AT . This yields

√ √ √ √
t01 t06 t05
   
√ + 2 3t6
t1 + 3t2√ − 3t1 + 3t2√− 2t6 2 3t4 +√2t5
 t06 1
t02 t04  =  − 3t√1+ 3t2 − 2t6 3t1 + t2 −√2 3t6 2t4 − 2 3t5 
t05 t04 t03 4
2 3t4 + 2t5 2t4 − 2 3t5 4t3
In a similar manner 0 = AAT . This gives

√ √ √ √
01 06 /2 05 /2
   
2
√ 1 + 62 √
+ 2 36 −2 31 + 2 3√ 2 − 26 2 34 +√25
 06 /2 1
02 04 /2  =  −2 3√ 1 + 2 32 − 26 61 + 22 −
√ 2 36 24 − 2 35 
05 /2 04 /2 03 8
2 34 + 25 24 − 2 35 83
Now
t01 = C11 01 + C12 02 + C13 03 + C14 04 + C15 05 + C16 06
1 √ 1 √
= C11 21 + 62 + 2 36 + C12 61 + 22 − 2 36 + C13 3
8 8
1 √ 1 √ 1 √ √
+ C14 24 − 2 35 + C15 2 34 + 25 + C16 −2 31 + 2 32 − 26
4 4 4
Also
1 √
t01 = t1 + 3t2 + 2 3t6
4
1
= (C11 1 + C12 2 + C13 3 + C14 4 + C15 5 + C16 6 )
4
3
+ (C21 1 + C22 2 + C23 3 + C24 4 + C25 5 + C26 6 )
4
√
3
+ (C61 1 + C62 2 + C63 3 + C64 4 + C65 5 + C66 6 )
2
Equating coefficients on i with Cij = Cji , we find

√ √
for 1 2√3C16 + 3C12 = 3C12 − 2√3C16
for 2 2 3C26 + 3C22√= 3C11 + 2 3C16
for 3 √ −3C13 + 2 3C36 = 3C√23
for 4 2√ 3C46 + 3C24 = C14
√ + 2 3C15
for 5 2 3C√56 + 3C25 = −2 √3C14 + C√15
for 6 3C16 + 2 3C66 + 3C26 = 3C11 − 3C12
Continue in the same way for t02 = t2 , t03 = t3 , etc. to obtain 36 equations (not all
independent) in the 21 unknown Cij s. Solve to verify that all coefficients are zero except
1
C11 = C22 , C13 = C21 , C33 , C44 = C55 , C66 = (C11 − C12 )
2
Problem 6.17
Develop a formula in terms of the strain components for the strain energy density W for
the case of an orthotropic elastic medium.
Answer
1 1
W = 2
(C11 1 + 2C12 2 + 2C13 3 ) 1 + 2
(C22 2 + 2C23 3 ) 2
1

+ 2 C33 23 + C44 24 + C55 25 + 2
C66 6
Solution
The strain energy is
tα α 1
W= = [(C11 1 + C12 2 + C13 3 ) 1 + (C21 1 + C22 2 + C23 3 ) 2
2 2
+ (C31 1 + C32 2 + C33 3 ) 3 + C44 24 + C55 25 + C66 26

where the constitutive relation for an orthotropic material was used. Collecting terms
gives
1
W= {[C11 1 + (C12 + C21 ) 2 + (C13 + C31 ) 3 ] 1
2
+ [C22 2 + (C23 + C32 ) 3 ] 2 + C33 23 + C44 24 + C55 25 + C66 26
Since Cij = Cji , we have
1 1
W= (C11 1 + 2C12 2 + 2C13 3 ) 1 + (C22 2 + 2C23 3 ) 2
2 2
1 2 2 2 2

+ C33 3 + C44 4 + C55 5 + C66 6
2
Problem 6.18
Show that, for an elastic continuum having x1 as an axis of elastic symmetry of order N = 2,
the strain energy density has the same form as for a continuum which has an x2 x3 plane of
elastic symmetry.
Answer
2W = (C11 1 + 2C12 2 + 2C13 3 + 2C14 4 ) 1 + (C22 2 + 2C23 3 + 2C24 4 ) 2
+ (C33 3 + 2C34 4 ) 3 + C44 24 + (C55 5 + 2C56 6 ) 5 + C66 26
Solution
2π
For N = = 2, we have θ = π and
θ
 
1 0 0
[aij ] =  0 −1 0 
0 0 −1
From Problem 6.13,
t0α = tα for (α = 1, 2, 3, 4) and t0α = −tα for (α = 5, 6)
and
0α = α for (α = 1, 2, 3, 4) and 0α = −α for (α = 5, 6)
For the x2 x3 plane of symmetry, we have
 
−1 0 0
[aij ] =  0 1 0 
0 0 −1
Thus we can show
t0α = tα for (α = 1, 2, 3, 4) and t0α = −tα for (α = 5, 6)
and
0α = α for (α = 1, 2, 3, 4) and 0α = −α for (α = 5, 6)
In each case, we have C15 = C16 = C25 = C26 = C35 = C45 = C46 = 0 with Cαβ = Cβα
(See Problem 6.13) This gives
1
W= {[C11 1 + 2C12 2 + 2C13 3 + 2C14 4 ] 1 + [C22 2 + 2C23 3 + 2C24 4 ] 2
2
+ [C33 3 + 2C34 4 ] 3 + C44 24 + [C55 5 + 2C56 6 ] 5 + C66 26
Problem 6.19
Let the stress field for a continuum be given by
 
x1 + x2 t12 0
[tij ] =  t12 x1 − x2 0 
0 0 x2
where t12 is a function of x1 and x2 . If the equilibrium equations are satisfied in the
absence of body forces and if the stress vector on the plane x1 = 1 is given by t(e^1 ) =
^2 , determine t12 as a function of x1 and x2 .
^1 + (6 − x2 ) e
(1 + x2 ) e
Answer
t12 = x1 − x2 + 5
Solution
The equilibrium equations are tij,j = 0
t11,1 + t12,2 + t13,3 = 0 for i=1

t12,1 + t22,2 + t23,3 = 0 for i=2
t13,1 + t23,2 + t33,3 = 0 for i=3
The resulting equations are

∂t12
1+ +0=0 and t12 = −x2 + f1 (x1 ) + C1
∂x2
∂t12
−1+0=0 and t12 = x1 + f2 (x2 ) + C2
∂x1
0+0+0=0
Now we find
∂t12 df2
1+ =1+0+ + 0 = 0 and f2 = −x2
∂x2 dx2
∂t12 df1
− 1 = −1 + 0 + + 0 = 0 and f1 = x1
∂x1 dx1
This gives
t12 = x1 − x2 + C
On the plane x1 = 1, we have
      
1 + x2 1 − x2 + C 0 1 1 + x2 1 + x2
 1 − x2 + C 1 − x2 0   0  =  1 − x2 + C  =  6 − x2 
0 0 x2 0 0 0
Equating, we have
1 − x2 + C = 6 − x2
and C = 5.
Problem 6.20
Invert Eq 6.45b to obtain Hooke’s law in the form

ν
tij = 2G ij + kk δij
1 − 2ν
which, upon combination with Eqs 6.44 and 6.43, leads to the Navier equation

1
G ui,jj + uj,ij + ρbi = 0 .
1 − 2ν
This equation is clearly indeterminate for ν = 0.5. However, show that in this case Hooke’s
law and the equilibrium equations yield the result
1
Gui,jj + tjj,i + ρbi = 0 .
3
Solution
Eq 6.45b is
Eij = (1 + ν) tij − νδij tkk
and
Eii = (1 − 2ν) tkk
This gives

E ν ν
tij = ij + δij kk = 2G ij + δij kk
1+ν 1 − 2ν 1 − 2ν
Now
ν
tij,j + ρbi = 2G ij,j + δij kk,j + ρbi = 0
1 − 2ν
We have
2ij,j = ui,jj + uj,ij and kk,j = uk,kj
and the Navier equations are

2ν 1
G ui,jj + uj,ij + uj,ji + ρbi = G ui,jj + uj,ji + ρbi = 0
1 − 2ν 1 − 2ν
For ν = 0.5, this result is indeterminate. However, from Eii = (1 − 2ν) tkk we find
ii = ui,i = 0. Hooke’s law gives
E
Eij,j = (ui,jj + uj,ij ) = (1 + ν) tij,j − νδij tkk,j
2
From the equilibrium equations and ui,i = 0, we have
E ν
ui,jj = −ρbi − tjj,i
2 (1 + ν) 1+ν
and
ν 1
Gui,jj + tjj,i + ρbi = Gui,jj + tjj,i + ρbi = 0
1+ν 3
v 0.5 1
since = = for ν = 0.5.
1+ν 1.5 3
Problem 6.21
Let the displacement field be given in terms of some vector qi by the equation
2 (1 − ν) qi,jj − qj,ji
ui = .
G
Show that the Navier equation (Eq 6.49) is satisfied providing bi ≡ 0 and qi is bi-harmonic
so that qi,jjkk = 0. If q1 = x2 /r and q2 = −x1 /r where r2 = xi xi , determine the resulting
stress field.
Answer
t11 = −t22 = 6QGx1 x2 /r5 ; t33 = 0
t12 = t21 = 3QG x22 − x21 /r5
t13 = t31 = 3QGx2 x3 /r5 ; t23 = t32 = −3QGx1 x2 /r5 ;
where Q = 4(1 − ν)/G
Solution
The Navier equation is (Eq 6.49)
µui,kk + (λ + µ) uk,ki + ρbi = 0
For the given displacement field, the Navier equation is
[2 (1 − ν) qj,kkji − qk,kjji ]
2 (1 − ν) qi,jjkk − qj,jikk + (λ + µ) + ρbi = 0
µ
(λ + 2µ)
since G = µ. Since 2 (1 − ν) = and if qi,jjkk = 0 for a bi-harmonic function
(λ + µ)
(−µ + λ + 2µ − λ − µ)
−qi,jikk + ρbi = 0
µ
for zero body forces. Thus the Navier equation is satisfied.
x2 x1 2 (1 − ν) qi,jj − qj,ji
If q1 = and q2 = − where r2 = xi xi = x21 +x22 +x23 and ui = ,
r r G
then
2 (1 − ν) qi,kkj − qk,kij
ui,j =
G
To compute the stresses, we use the strains and Hooke’s law. Now

2 (1 − ν) 6x1 x2 2 (1 − ν) 2r−3 − 6x21 r−5
u1,1 = u2,1 = u3,1 = 0
Gr5 −3 2 −5
G
2 (1 − ν) −2r + 6x2 r −2 (1 − ν) 6x1 x2
u1,2 = u2,2 = u3,2 = 0
G Gr5
−2 (1 − ν) 6x2 x3 2 (1 − ν) 6x1 x3
u1,3 = u2,3 = u3,3 = 0
Gr5 Gr5
Hooke’s law is tij = 2µij + λδij kk = G (ui,j + uj,i ) + λδij tkk . The stresses are
24 (1 − ν) x1 x2 6GQx1 x2
t11 = 2Gu1,1 + λuk,k = 2Gu1,1 = 5
=
r r5
−4 (1 − ν) 6x1 x2 −6GQx1 x2
t22 = 2Gu2,2 + λuk,k = 2Gu2,2 = =
Gr5 r5
t33 = 2Gu3,3 + λuk,k = 2Gu3,3 = 0
12G (1 − ν) x22 − x21 3GQ x22 − x21
t12 = G (u1,2 + u2,1 ) = t21 = =
Gr5 r5
−12G (1 − ν) x2 x3 −3GQx2 x3
t13 = G (u1,3 + u3,1 ) = t31 = =
Gr5 r5
12G (1 − ν) x1 x3 3GQx1 x3
t23 = G (u2,3 + u3,2 ) = t32 = u2,3 = =
Gr5 Gr5
4 (1 − ν)
where Q = .
G
Problem 6.22
If body forces are zero, show that the elastodynamic Navier equation (Eq 6.55) will be
satisfied by the displacement field
ui = φ,i + εijk ψk,j
provided the potential functions φ and ψk satisfy the three-dimensional wave equation.
Solution
The elastodynamic Navier equation (Eq 6.55) is
µui,kk + (λ + µ) uk,ki = ρüi
with zero body forces. Then
ui,jj = φ,ijj + εiqk ψk,qjj

uj,ji = φ,jji + εiqk ψk,qji
and
üi = φ̈,i + εiqk ψ̈k,q
Combining, we obtain

µ (φ,ijj + εiqk ψk,qjj ) + (λ + µ) φ,jji = ρ φ̈,i + εiqk ψ̈k,q
since εiqk ψk,qji = 0. Now

(λ + 2µ) φ,jj − ρφ̈ ,i + εiqk µψk,jj − ρψ̈k ,q = 0
This is true if each term is zero individually. This is the three-dimensional wave equation.
Problem 6.23
Show that, for plane stress, Hooke’s law Eq 6.120a and Eq 6.121 may be expressed in terms
of the Lamé constants λ and µ by

1 λ
ij = tij − δij tkk (i, j, k = 1, 2) ,
2µ 3λ + 2µ
λ
kk = − tii (i = 1, 2) .
2µ (3λ + 2µ)
Solution
Equation 6.120a can be written as
1
ij = [(1 − ν) tij + νδij tkk ] (i, j, k = 1, 2)
E
(1 − ν) 1 ν λ
Now = and = , then
E 2µ E 2µ (3λ + 2µ)

1 λ
ij = tij + δij tkk (i, j, k = 1, 2)
2µ 3λ + 2µ
and
ν λ
33 = − tkk = − tkk (k = 1, 2)
E 2µ (3λ + 2µ)
Problem 6.24
For the case of plane stress, let the stress components be defined in terms of the function
φ = φ(x1 , x2 ), known as the Airy stress function, by the relationships,
t11 = φ,22 , t22 = φ,11 , t12 = −φ,12 .

Show that φ must satisfy the biharmonic equation ∇4 φ = 0 and that, in the absence of
body forces, the equilibrium equations are satisfied identically by these stress components.
If φ = Ax31 x22 − Bx51 where A and B are constants, determine the relationship between A
and B for this to be a valid stress function.
Answer
A = 5B
Solution
The compatibility equation (Eq 6.119) is
11,22 + 22,11 = 212,12
Using Eq 6.120a, this can be written
1
[t11,22 − νt22,22 + t22,11 − νt11,11 − 2 (1 + ν) t12,12 ] = 0
E
Substituting the stress function gives
φ,2222 − νφ,1122 + φ,1111 − νφ,2211 − 2 (1 + ν) (−φ,1212 ) = 0
or
φ,1111 + φ,2222 + 2φ,1212 − ν (φ,1122 + φ,2211 − φ,1212 ) = 0
This gives
φ,1111 + φ,2222 + 2φ,1212 = ∇4 φ = 0
the bi-harmonic equation since the second term in parentheses is zero.
The equilibrium equations are
tij,j = 0 (i, j = 1, 2)
and
t11,1 + t12,2 = φ,221 − φ,122 = 0

t21,1 + t22,2 = −φ,211 + φ,112 = 0
For φ = Ax31 x22 − Bx51 , we find
φ,1111 = −120Bx1
φ,2222 = 0
φ,1212 = 12Ax1
and
φ,1111 + φ,2222 + 2φ,1212 = −120Bx1 + 0 + 24Ax1 = 0
Thus
A = 5B
if the bi-harmonic equation is satisfied.
Problem 6.25
Develop an expression for the strain energy density, W, for an elastic medium in (a) plane
stress and (b) plane strain.
Answer

(a) W = t211 + t222 − 2νt11 t22 + 2 (1 + ν) t212 /2E
(b) W = µ + 12 λ 211 + 222 + λ11 22 + 2µ212

Solution
(a) For plane stress (Eq 6.120a), the strain energy is
1 1
W= tij ij = (t11 11 + t22 22 + 2t12 12 )
2 2
1 2
t − νt11 t22 + t222 − νt22 t11 + 2 (1 + ν) t212

=
2E 11
(b) For plane strain (Eq 6.23), the strain energy is
1
2µ211 + λ (11 + 22 ) 11 + 2µ222 + λ (11 + 22 ) 22 + 4µ212

W=
2
1
(2µ + λ) 211 + 222 + 2λ11 22 + 4µ212

=
2
Problem 6.26
Show that φ = x41 x2 + 4x21 x32 − x52 is a valid Airy stress function, that is, that ∇4 φ = 0,
and compute the stress tensor for this case assuming a state of plane strain with ν = 0.25.
Answer
24x21 x2 − 20x32 −4x31 − 24x1 x22
 
0
[tij ] = −4x31 − 24x1 x22
 12x21 x2 + 8x32 0 
1 2 3

0 0 4 9x1 x2 − 3x2
Solution
The Airy stress function must satisfy the bi-harmonic equation
∇4 φ = φ,1111 + φ,2222 + 2φ,1212 = 0
Now
φ,1111 = 24x2
φ,2222 = −120x2
φ,1212 = 48x2
and
∇4 φ = 24x2 − 120x2 + 2 (48x2 ) = 0
The stresses are
t11 = φ,22 = 24x21 x2 − 20x32
t22 = φ,11 = 12x21 x2 + 8x32
t12 = −φ,12 = −4x31 − 24x1 x22
= ν (t11 + t22 ) = 14 9x21 x2 − 3x32

t33
Problem 6.27
Verify the inversion of Eq 6.147 into Eq 6.148. Also, show that the two equations of Eq 6.149
may be combined to produce Eq 6.148.
Solution
Equation 6.147 is
1+ν ν
ij = tij − δij tkk + α (θ − θ0 ) δij
E E
and
1+ν ν
ii = tii − 3 tkk + 3α (θ − θ0 )
E E
This gives
E
tkk = [ii − 3α (θ − θ0 )]
1 − 2ν
Then
E
tij = [(1 − 2ν) ij + νδij kk − (1 + ν) δij α (θ − θ0 )]
(1 + ν) (1 − 2ν)
This is the desired result.
Equation 6.149a is
E
Sij = ηij
(1 + ν)
or
1 E E 1
tij − δij tkk = ηij = ij − δij kk
3 (1 + ν) (1 + ν) 3
This gives
E 1 1
tij = ij − δij kk + δij tkk
(1 + ν) 3 3
From Eq 6.149b, we have
E
tkk = [kk − 3α (θ − θ0 )]
(1 − 2ν)
or

E 1 1 E
tij = ij − δij kk + δij [kk − 3α (θ − θ0 )]
(1 + ν) 3 3 (1 − 2ν)
E
= [(1 − 2ν) ij + νδij kk − (1 + ν) δij α (θ − θ0 )]
(1 + ν) (1 − 2ν)
Problem 6.28
Develop appropriate constitutive equations for thermoelasticity in the case of (a) plane
stress and (b) plane strain.
Answer
(a) ij = [(1 + ν)tij − νδij tkk ]/E + δij (θ − θ0 )α (i, j, k = 1, 2)

33 = νtii /E + α(θ − θ0 ) (i = 1, 2)
(b) tij = λδij kk + 2µij − δij (3λ + 2µ)α(θ − θ0 ) (i, j, k = 1, 2)
t33 = νtii − αE(θ − θ0 ) = λii − (3λ + 2µ)α(θ − θ0 ) (i = 1, 2)
Solution
(a) From Eq 6.147, we have
1+ν ν
ij = tij − δij tkk + α (θ − θ0 ) δij (i, j, k = 1, 2)
E E
and
ν
(t11 + t22 ) + α (θ − θ0 )
33 =
E
(b) From Eq 6.23 and Eq 6.147, we have
tij = 2µij + λδij kk − (3λ + 2µ) δij α (θ − θ0 ) (i, j, k = 1, 2)
and
t33 = ν (t11 + t22 ) − α (θ − θ0 ) E = λ (11 + 22 ) − (3λ + 2µ) α (θ − θ0 )
Problem 6.29
Consider the Airy stress function
φ5 = D5 x21 x32 + F5 x52 .
(a) Show that for this to be valid stress function, F5 = −D5 /5.
(b) Construct the composite stress function
φ = φ5 + φ3 + φ2
where
1 1 1
φ = D5 x21 x32 − x52 + B3 x21 x2 + A2 x21 .
5 2 2
For this stress function show that the stress components are
t11 = D5 6x21 x2 − 4x32 ,

t22 = 2D5 x32 + B3 x2 + A2 ,

t12 = −6D5 x1 x22 − B3 x1 .
Solution
(a) Take derivatives and substitute into the biharmonic
φ,1111 = 0
φ,1122 = 120F5 x2
φ,1122 = 12D5 x2
φ,1111 + 2φ,1122 + φ,1122 = 120F5 + 24D5 = 0
(b) Simply take partial derivatives t11 = φ,22 , t22 = φ,11 and t12 = −φ,12 .
Problem 6.30
A rectangular beam of width unity and length 2L carries a uniformly distributed load of
q lb/ft as shown. Shear forces V support the beam at both ends. List the six boundary
conditions for this beam the stresses must satisfy using stresses determined in Problem 6.29.
x2
q [lb/ft]
V V
c
O x1
c
L L
Answer
1. t22 = −q at x2 = +c
2. t22 = 0 at x2 = −c
3. t12 = 0 at x2 = ±c
R+c
4. −c t12 dx2 = qL at x1 = ±L
R+c
5. −c t11 dx2 = 0 at x1 = ±L
R+c
6. −c t11 x2 dx2 = 0 at x1 = ±L
Solution
See Answer.
Problem 6.31
Using boundary conditions 1, 2, and 3 listed in Problem 6.30, show that the stresses in
Problem 6.29 require that
q 3q q
A2 = − , B3 = − , D5 = .
2 4c 8c3
Thus, for the beam shown the stresses are

q 2 2 3
t11 = x1 x2 − x2 ,
2I 3

q 1 3 2 2 3
t22 = x − c x2 − c ,
2I 3 2 3
q
x1 x22 − c2 x1 ,

t12 = −
2I
where I = 23 c3 is the plane moment of inertia of the beam cross section.
Solution
Take partial derivatives t11 = φ,22 , t22 = φ,11 and t12 = −φ,12 to get
2D3 c3 + B3 c + A2 = −q
−2D3 c3 − B3 c + A2 = 0
6D3 c2 + B3 = 0
Solve to get solution.
Problem 6.32
Show that, using the stresses calculated in Problem 6.31, the boundary conditions 4 and 5
are satisfied, but boundary condition 6 is not satisfied.
Solution
Problem 6.33
Continuing Problems 6.31 and 6.32, in order for boundary condition 6 to be satisfied an
additional term is added to the stress function, namely
φ3 = D3 x32 .
Show that, from boundary condition 6,
L2

3q 1
D3 = − 2 ,
4c 15 6c
so that finally
q 2 6
t11 = x21 − x22 + c2 − L2 x2 .
21 3 15
Solution
Problem 6.34
Show that for the shaft having a cross section in the form of an equilateral triangle the
warping function is
ψ (x1 , x2 ) = λ x32 − 3x21 x2 .

Determine
(a) the constant λ in terms of the shaft dimension,
(b) the torsional rigidity K,
(c) the maximum shearing stress.
x2
2a a
O
x1
Mt
Answer
1
(a) λ = −
6a√
9G 3a4
(b) K =
5
3Mt a
(c) t23 |max = at x1 = a, x2 = 0
2K
Solution
Write out Φ and differentiate
√ √
Φ = k(x1 − a)(x1 − x2 3 + 2a)(x1 + x2 3 + 2a)
h i
= k(x1 − a) (x1 + 2a)2 − 3x22
Differentiate to find k from ∇2 Φ = −2

h i
Φ,1 = k (x1 + 2a)2 − 3x22 + k(x1 − a)2(x1 + 2a)
Φ,11 = 2k(x1 + 2a) + 2k(x1 + 2a) + 2k(x1 − a)
Φ,11 = 6kx1 + 6ka
Φ,2 = k(x1 − a)(−6x2 ) = −6kx2 (x1 − a)
Φ,22 = −6kx1 + 6ka
So
∇2 Φ = Φ,11 + Φ,22 = 6kx1 + 6ka − 6kx1 + 6ka = 12ka = −2Gα
giving
Gα
k=−
6a
The moment is found from
Z
Mt = 2 ΦdA
A
Za Z (x1 +2a)/√3
h 1 i
= 2Gα · 2 (x1 − a) (x1 + 2a)2 − 3x22 dx2 dx1
−
−2a 0 6a
Za " 3 #
1 2 x1 + 2a x1 + 2a
= 4Gα − (x1 − a) (x1 + 2a) √ − √ dx1
−2a 6a 3 3
This form is easily evaluated using Matlab below.
>> syms a x
>> f = -(x-a)*((x+2*a)^2*((x+2*a)/sqrt(3))-(x+2*a)^3/(3*sqrt(3)))/(6*a)
f = 1/27*(-x+a)*(x+2*a)^3*3^(1/2)/a
>> m = int(f,-2*a,a)
m = 9/20*3^(1/2)*a^4
So √
9 3Gαa4
Mt =
5
and √
9 3Ga4
K = Mt /α =
5
Compute the shear stresses
x2
t31 = Φ2 = −Gα (x1 − a)
a
Gα
h i
t32 = −Φ,1 = − (x1 + 2a)2 − 3x22 + (x1 − a)2(x1 + 2a)
6a
Gα 2
=− 3x1 + 6ax1 − 3x22
6a
Maximum shear stress occurs at three locations nearest the centroid. The easiest to com-
pute is at x1 = a , x2 = 0. Here, t31 = 0 and t32 = 9Ga2 /6a. So
3Gαa 3GMt a 3GMt a 5 15Mt
τmax = = = √ = √
2 2K 2 9 3Ga4 18 3a3
Problem 6.35
Consider the Galerkin vector that is the sum of three double forces, that is, let
x3 ^r
e
^θ
e
^ψ
e
ψ
r
O
x2
θ
x1
x x2 x
1
F=B e^2 + 3 e
^1 +
e ^3
r r r
2
where B is a constant and r = xi xi . Show that the displacement components are given by
2B (1 − 2ν) xi
ui = − .
r3
Using the sketch above, (spherical coordinates) observe that the radial displacement ur
(subscript r not a summed indice, rather indicating the radial component of displacement)
ui xi
ur = ,
r
and show that ur = −2B (1 − 2ν) r−2 . Also, show that uψ = uθ = 0. Thus
∂ur 4B (1 − 2ν) 2B (1 − 2ν)

r = = and ψ = θ = − ,
∂r r3 r3
so that the cubical dilatation is r + ψ + θ = 0. From Hooke’s law
tij = λδij uk,k + 2µ (ui,j + uj,i )
which reduces here to tij = 2µij so that
8BG (1 − 2ν)
trr = ,
r3
and
4BG (1 − 2ν)
tψψ = tθθ = − .
r3
Chapter 7 Solutions
Problem 7.1
Introduce the stress deviator Sij and the viscous stress deviator τ∗ij = τij − 31 δij τkk into
Eq 7.5 to prove that Sij = τ∗ij .
Solution
Eq 7.5 is
tij = −pδij + τij
and
tkk = −3p + τkk
The stress deviator is
1
Sij = tij − δij tkk
3
Now
1 δij τkk
Sij + δij tkk = tij = −pδij + τij = −pδij + τ∗ij +
3 3
or
tkk τkk
Sij + δij +p− = τ∗ij
3 3
The term in parentheses is zero, and the result is shown
Problem 7.2
Determine an expression for the stress power (a) tij dij and (b) τij dij for a Newtonian fluid.
First, show that
2 ∗
∗
τij = κ − µ δij dkk + 2µ∗ dij .
3
Answer
(a) tij dij = −pdii + κ∗ dii djj + 2µ∗ βij βij

(b) τij dij = κ∗ (tr D)2 + 2µ∗ βij βij
Solution
(a) The stress is from Eq 7.10
tij = −pδij + λ∗ δij dkk + 2µ∗ dij
with (Eq 7.12a)

1
κ∗ = (3λ∗ + 2µ∗ )
3
179
gives

2
tij dij = −pδij dij + κ∗ − µ∗ δij dkk dij + 2µ∗ dij dij
3

1
= −pdjj + κ∗ dkk djj + 2µ∗ dij dij − dkk djj
3
∗ ∗
= −pdjj + κ dkk djj + 2µ βij βij
where βij = dij − 13 δij dkk

(b) From Eq 7.5 tij = −pδij + τij and Eq 7.10 and Eq 7.12a yields

2 ∗
τij = κ − µ δij dkk + 2µ∗ dij
∗
3
Now

2
τij dij = κ∗ − µ∗ δij dkk dij + 2µ∗ dij dij
3

∗ ∗ 1
= κ dkk dii + 2µ dij dij − dkk djj
3
= κ∗ dkk dii + 2µ∗ βij βij = κ∗ tr (D)2 + 2µ∗ βij βij
Problem 7.3
Determine the constitutive equation for a Newtonian fluid for which Stokes condition holds,
that is, for κ∗ = 0.
Answer
tij = −pδij + 2µ∗ βij
Solution
From Eq 7.10
with (Eq 7.12a)
1
κ∗ = (3λ∗ + 2µ∗ )
3
Now

2
tij = −pδij + κ∗ − µ∗ δij dkk + 2µ∗ dij
3

1
= −pδij + κ∗ δij dkk + 2µ∗ dij − δij dkk
3
∗
= −pδij + 2µ βij
if κ∗ = 0
Problem 7.4
Develop an expression of the energy equation for a Newtonian fluid assuming the heat
conduction follows Fourier’s law.
Answer
ρu̇ = −pvi,i + λ∗ vi,i vj,j + 21 µ∗ (vi,j + vj,i ) + κ∗ θ,ii + ρr
Solution From Eq 5.65

ρu̇ = tij dij + ρr − qi,i
From Eq 7.10
ρu̇ = (−pδij + λ∗ δij dkk + 2µ∗ dij ) dij + ρr + κθ,ii

= −pdii + λ∗ dkk dii + 2µ∗ dij dij + ρr + κθ, ii
but dij = (vi,j + vj,i ) /2 and so
µ∗
ρu̇ = −pdii + λ∗ vk,k vi,i + (vi,j + vj,i )2 + κθ,ii
2
Problem 7.5
The dissipation potential Ψ for a Newtonian fluid is defined as a function of D and β by
1 ∗ 2
Ψ= κ djj dii + µ∗ βij βij , where κ∗ = λ∗ + µ∗ .
2 3
Show that ∂Ψ/∂dij = τij .
Solution The derivative of Ψ is

∂Ψ κ∗ ∂ ∂
= (dii djj ) + µ∗ (βij βij )
∂dpq 2 ∂dpq ∂dpq
κ∗

∂βij
= (δip δiq djj + δjp δjq dii ) + 2µ∗ βij
2 ∂dpq
Now

∂βij ∂ 1
βij = βij dij − δij dkk
∂dpq ∂dpq 3

δij δpq
= βij δip δjq − δpk δqk = βpq − βii
3 3
= βpq
since βii = 0. This gives
∂Ψ
= κ∗ δpq djj + 2µ∗ βpq
∂dpq

δpq
= κ∗ δpq djj + 2µ∗ dpq − djj
3

2
= κ∗ − µ∗ δpq djj + 2µ∗ dpq = τpq
3
This last result follows from Problem 7.2

Problem 7.6
Verify the derivation of the Navier-Stokes equations for a Newtonian fluid as given by
Eq 7.25.
Solution Eq 7.25 is
ρv̇i = ρbi − p,i + (λ∗ + µ∗ ) vj,ji + µ∗ vi,jj
The constitutive equation is, Eq 7.10
and the equation of motion is Eq 7.17
tij,j + ρbi = ρv̇i
Computing the divergence of the stress
tij,j = −p,j δij + λ∗ δij dkk,j + 2µ∗ dij,j = −p,i + λ∗ vk,ki + µ∗ (vi,ji + vj,ii )
Changing dummy indices and collecting terms yield the desired result
−p,i + (λ∗ + µ∗ ) vk,ki + µ∗ vj,ii + ρbi = ρv̇i
Problem 7.7
Consider a two-dimensional flow parallel to the x2 x3 plane so that v1 = 0 throughout the
fluid. Assuming that an incompressible, Newtonian fluid undergoes this flow, develop a
Navier-Stokes equation and a continuity equation for the fluid.
Answer
(Navier-Stokes) ρv̇i = ρbi − p,i + µ∗ vi,jj (i, j = 2, 3)

(Continuity) vi,i = 0 (i = 2, 3)
Solution
From Eq 7.28 the Navier-Stokes equation is
ρv̇i = ρbi − p,i + µ∗ vi,jj
Then
p,1 = ρb1 for i = 1

ρv̇i = ρbi − p,i + µ∗ vi,jj for i, j = 2, 3
The continuity equation is, Eq 7.16
ρ̇ + ρvi,i = 0
where ρ̇ = 0 for an incompresible material, and
vi,i = 0 for i = 2, 3
Problem 7.8
Consider a barotropic, inviscid fluid under the action of conservative body forces. Show
that the material derivative of the vorticity of the fluid in the current volume V is
Z Z
d
wi dV = vi wj nj dS .
dt V S
Solution
From Problem 5.5 Z Z
d 1
wi dV = εijk ak + wj vi nj dS
dt V S 2
The equation of motion for an inviscid fluid is,Eq 7.29
ρv̇i = ρbi − p,i
1
where from Eq 7.34,bi = −Ω,i 7.35, P,i = p,i and the equation of motion is
ρ
ρak = −ρ (Ω + P),k
Now Z Z Z
1 1 ρ
εijk ak nj dS = εijk ak,j dV = − εijk (Ω + P),kj dV = 0
S 2 V 2 V 2
from the skew symmetry of εijk . Then
Z Z
d
wi dV = vi wj nj dS
dt V S
Problem 7.9
Show that for an incompressible, inviscid fluid the stress power vanishes identically as one
would expect.
Solution
The constitutive equation for an incompressible, inviscid fluid is tij = −pδij . The stress
power is
1
tij dij = − pδij (vi,j + vj,i ) = −pvi,i = 0
2
since vi,i = 0. An alternate solution exploits the symmetry of the stress tensor
1
tij dij = tij (vi,j + vj,i ) = tij vi,j = −pδij vi,j = −pvi,i = 0
2
Problem 7.10
Show that the vorticity and velocity of a barotropic fluid of constant density moving under
conservative body forces are related through the equation ẇi = wj vi,j . Deduce that for a
steady flow of this fluid vj wi,j = wj vi,j .
Solution
From Problem 5.10
d wi εijk ak,j + 2wj vi,j
=
dt ρ 2ρ
For a constant density

1
ẇi = εijk ak,j + wj vi,j
2
For a barotropic fluid with conservative body forces, Eq 7.43 gives
ρv̇i = (Ω + P),i
and hence
1
εijk ak,j = εijk (Ω + P),kj = 0
ρ
and
ẇi = wj vi,j
Expanding ẇj gives
∂wi
+ vj wi,j = wj vi,j
∂t
and for steady flow
vj wi,j = wj vi,j
Problem 7.11
In terms of the vorticity vector w, the Navier-Stokes equations for an incompressible fluid
may be written as
ρv̇i = ρbi − p,i − 2µ∗ εijk wk,j .
Show that, for an irrotational motion, this equation reduces to the Euler equation
ρv̇i = ρbi − p,i .
Solution
The equation of motion is, Eq 7.28
ρv̇i = ρbi − p,i + µ∗ vi,jj
From Eq 4.153
vi,jj = vj,ji − 2εijk wk,j
and
ρv̇i = ρbi − p,i + µ∗ (vj,ji − 2εijk wk,j )

= ρbi − p,i − 2µ∗ εijk wk,j
since vj,ji = 0 from incompressibility. Now for irrotational flow wi = 0 and
ρv̇i = ρbi − p,i
This is Eq 7.29.
Problem 7.12
Carry out the derivation of Eq 7.41 by combining the Euler equation with the continuity
equation, as suggested in the text.
Solution
The Euler equation is
∂vi
ρv̇i = ρ + vj vi,j = ρbi − p,i
∂t
and the continuity equation is
∂ρ
ρ̇ + ρvi,i = + vi p,i + ρvi,i = 0
∂t
∂vi ∂ρ
For steady flow = = 0 and bi = 0. This gives
∂t ∂t
ρvj vi,j = −p,i (Euler) and vi p,i + ρvi,i = 0 (continuity)
For a barotropic flow

dp
p,i = ρ,i = c2 ρ,i
dρ
Substituting into the Euler equation gives
ρvj vi,j + c2 ρ,i = 0
Multiplying by vi gives
ρvi vj vi,j + c2 vi ρ,i = ρvi vj vi,j − c2 ρvi,i = 0
where the continuity equation was used. This gives
ρ vi vj − c2 ρδij vi,j = 0

This is Eq 7.41.
Problem 7.13
Consider the velocity potential φ = x2 x3 /r2 where r2 = x21 + x22 . Show that this satisfies
the Laplace equation φ,ii = 0. Derive the velocity field and show that this flow is both
incompressible and irrotational.
Solution
The individual terms of the Laplacian are

∂ ∂ x2 x3 ∂ x2 x3 ∂r
φ,11 = = −2 3
∂x1 ∂x1 r2 ∂x1 r ∂x1
Now
∂r ∂ h i 1 h i x
i
= (xj xj )1/2 = (2δij xj ) (xj xj )−1/2 =
∂xi ∂xi 2 r
Thus
∂ h x2 x3 x1 i h x2 x3 x2 x3 x1 x1 i
φ,11 = −2 3 = −2 4 + 8
∂x r r r r5 r
12
= 8x1 x2 x3 − 2x2 x3 x1 + x2 r = 6x1 − 2x32 x3 r−6
2 2
−6 2
Similarly
φ,22 = −6x21 + 2x32 x3 r−6 and φ,33 = 0

so that
φ,11 + φ,22 + φ33 = 0
Now the velocity is

−2x1 x2 x3 x3 x21 − x22 x2
v1 = φ,1 = ; v2 = φ,2 = ; v3 = φ,3 =
r4 r4 r2
and the continuity equation is
vi,i = v1,1 + v2,2 = v3,3 = φ,11 + φ,22 + φ,33 = φ,ii = 0
This is an incompressible flow. For irrotational motion
curl v = εijk vk,j e

î = εijk φ,kj e
î = 0
Problem 7.14
If the equation of state of a barotropic fluid has the form p = λρk where k and λ are
constants, the flow is termed isentropic. Show that the Bernoulli equation for a steady
motion in this case becomes
kp 1
Ω+ + v2 = constant .
(k + 1) ρ 2
Also, show that for isothermal flow the Bernoulli equation takes the form
p ln ρ 1 2
Ω+ + v = constant .
ρ 2
Solution
The Bernoulli equation, Eq 7.44, is
Z x2
∂vi v2
dxi + + Ω + P = G (t)
x1 ∂t 2
and for steady flow becomes

v2
+ Ω + P = G0
2
Now
dp
= λkρk−1
dρ
and Zp
dp k p p0 kp
P= = − = + C1
p0 ρ k−1 ρ ρ0 ρ (k − 1)
Thus
v2 kp
+Ω+ = constant
2 ρ (k − 1)
For an isothermal flow dp = λdρ and
Z Z
dp λ p
P= = dρ = λ ln ρ + C2 = ln ρ + C2
ρ ρ ρ
and
v2 p ln ρ
+Ω+ = constant
2 ρ
Problem 7.15
Derive Eq 7.44 by taking the scalar product of dxi (the differential displacement along a
streamline) with Eq 7.43 and integrating along the streamline, that is, by the integration of
Z x2
(v̇i + Ω,i + P,i ) dxi .
x1
Solution
Eq 7.43 is
∂vi
v̇i + Ω,i + P,i = + vj vi,j + Ω,i + P,i = 0
∂t
Taking the scalar product with dxi and integrating gives
Z x2
∂vi
+ vj vi,j + Ω,i + P,i dxi = G (t)
x1 ∂t
or Z x2
∂vi
+ vj vi,j dxi + Ω + P = G (t)
x1 ∂t
v v
i i
Along a streamline, dxi = ds and so vj vi,j dxi = vj vi,j ds = vi vi,j dxj = vi dvi
v v
and Z x2
∂vi v2
dxi + + Ω + P = G (t)
x1 ∂t 2
Problem 7.16
Verify that Eq 7.47 is the material derivative of Eq 7.46. Also, show that for a barotropic,
inviscid fluid subjected to conservative body forces the rate of change of the circulation is
zero (See Eq 7.47).
Solution
The material derivative of Eq 7.46 is
I I
d
Γ̇C = vi dxi = (v̇i dxi + vi,q vi dxq )
dt
from Eq 5.7. Now vi,q dxq = dvi and v̇i = − (Ω + P),i for a barotropic, inviscid fluid. So
that
I I I x
v2 1

Γ̇C = (v̇i dxi + vi dxi ) = − (Ω + P),i dxi + vi dvi = − (Ω + P) + =0
2 x1
Problem 7.17
Determine the circulation ΓC around the square in the x2 x3 plane shown in the figure if the
velocity field is given by
v = x3 − x22 e

^2 + (x3 + x2 ) e
^3 .
x3
(−1,1) (1,1)
x2
O
(−1,−1) (1,−1)
Answer
ΓC = 0
Solution
Beginning at the lower right-hand corner, the circulation, ΓC , is
I Z1 Z −1 Z −1 Z1
ΓC = vi dxi = v3 (1, x3 ) dx3 + v2 (x2 , 1) dx2 + v3 (−1, x3 ) dx3 + v2 (x2, − 1) dx2
−1 1 1 −1
Z1 Z −1 Z −1 Z1
1 − x22 , dx2 + −1 − x22, dx2

= (x3 + 1) dx3 + (x3 − 1) dx3 +
−1 1 1 −1
4 8
=2− +2− =0
3 3
Alternatively, from Eq 7.46 with ni = n1
Z ZZ
ΓC = ε1jk vk,j n1 dS = (ε123 v3,2 + ε132 v2,3 ) dx2 dx3
S
ZZ
= (1 − 1) dx2 dx3
Chapter 8 Solutions
Problem 8.1
Referred to principal axes, the invariants of the Green deformation tensor CAB are
I1 = λ21 + λ22 + λ23 ,
I2 = λ21 λ22 + λ21 λ23 + λ22 λ23 ,
I3 = λ21 λ22 λ23 .
For an isotropic, incompressible material, show that
1
I1 = λ21 + λ22 + ,
λ21 λ22
1 1
I2 = 2
+ 2 + λ21 λ22 .
λ1 λ2
Solution
For an incompressible material,
1
I3 = 1 = λ21 λ22 λ23 or λ23 =
λ21 λ22
Then
1
I1 = λ21 + λ22 +
λ21 λ22
and
1 1 1
I2 = λ21 λ22 + λ21 + λ22 + 2 + λ21 λ22

=
λ21 λ22 2
λ1 λ2
Problem 8.2
Derive the following relationships between invariants I1 , I2 , and I3 , and the deformation
gradient, CAB :
∂I1
(a) = δAB ,
∂CAB
∂I2
(b) = CAB − I1 δAB ,
∂CAB
∂I3
(c) = I3 C−1
AB .
∂CAB
Solution
189
(a) For I1 = CMM and

∂I1 ∂CMM
= = δMA δMB = δAB
∂CAB ∂CAB
(b) For I2 = 1
2
(CMN CMN − CMM CNN ) and

∂I2 ∂ 1
= (CMN CMN − CMM CNN )
∂CAB ∂CAB 2
1
= CMN δAM δBN − [CNN δMA δMB + CMM δNA δNB ]
2
1
= CAB − (I1 δAB + I1 δAB ) = CAB − I1 δAB
2
(c) For I3 = 16 εLMN εPQR CLP CMQ CNR and differentiating
∂I3 1 1
= εLMN εPQR δAL δBP CMQ CNR + εLMN εPQR CLP δAM δBQ CNR
∂CAB 6 6
1
+ εLMN εPQR CLP CMQ δAN δBR
6
1 1 1
= εAMN εBQR CMQ CNR + εLAN εPBR CLP CNR + εLMA εPQB CLP CMQ
6 6 6
1
= εAMN εBQR CMQ CNR
2
where the property of εPOR and dummy indices have been used in the last
step. This is the cofactor of CAB and, so from Eq 2.46, we have
∂I3
= (det CAB ) C−1 −1
AB = I3 CAB
∂CAB
Problem 8.3
Use the definitions of I1 and I2 in terms of the principal stretches λ1 , λ2 , and λ3 to show

∂W 2 2 2
∂W 2 ∂W
(a) = λ − λ3 + λ2 ,
∂λ1 λ1 1 ∂I1 ∂I2

∂W 2 ∂W ∂W
(b) = λ2 − λ23 + λ21 .
∂λ2 λ2 2 ∂I1 ∂I2
Solution
(a) For an incompressible material W (I1 , I2 ), differentiating gives

∂W ∂W ∂I1 ∂W ∂I2
= +
∂λ1 ∂I1 ∂λ1 ∂I2 ∂λ1

∂W 2 ∂W 2
= 2λ1 − 3 2 + − 3 + 2λ1 λ22
∂I1 λ1 λ2 ∂I2 λ
1
2 ∂W ∂W
λ2 − λ33 + λ22

=
λ1 1 ∂I1 ∂I2
where results from Problem 8.1 were used noting that λ21 λ22 λ23 = 1
(b) Now
∂W ∂W ∂I1 ∂W ∂I2
= +
∂λ2 ∂I1 ∂λ2 ∂I2 ∂λ2

∂W 2 ∂W 2
= 2λ2 − 2 3 + − 3 + 2λ21 λ2
∂I1 λ1 λ2 ∂I2 λ
2
2 ∂W ∂W
λ2 − λ33 + λ21

=
λ2 2 ∂I1 ∂I2
Problem 8.4
Let W(I1 , I2 , I3 ) be the strain energy per unit volume for a homogeneous, isotropic material.
Show that the first Piola-Kirchhoff stress components may be written as follows:
∂W ∂W ∂W ∂W −1
PiA ≡ =2 FiA + 2 (Bij − I1 δij ) FjA + 2I3 F .
∂FiA ∂I1 ∂I2 ∂I3 iA
Solution
Using the chain rule gives
∂W ∂W ∂I1 ∂W ∂I2 ∂W ∂I3

PiA ≡ = + +
∂FiA ∂I1 ∂FiA ∂I2 ∂FiA ∂I3 ∂FiA
Each term must be evaluated in order. From Problem 8.2, the invariants are given in
terms of CAB . We first calculate
∂ ∂ ∂CMN ∂ ∂
= = (FjM FjN )
∂FiA ∂CMN ∂FiA ∂CMN ∂FiA
∂
= (δij δAM FjN + FjM δij δAN )
∂CMN
∂
= (δAM FiN + FiM δAN )
∂CMN
From Problem 8.2, we have

∂I1 ∂I1 ∂CMN
= = δMN (δAM FiN + FiM δAN )
∂FiA ∂CMN ∂FiA
= FiA + FiA = 2FiA
∂I2 ∂I2 ∂CMN

= = (CMN − I1 δMN ) (δAM FiN + FiM δAN )
= CAN FiN + CMA FiM − I1 FiA − I1 FiA
= FjA FjN FiN + FjM FjA FiM − 2I1 FiA = 2 (Bij − I1 δij ) FjA
and
∂I3 ∂I3 ∂CMN
= = I3 C−1
MN (δAM FiN + FiM δAN )
= I3 C−1 −1 −1
AN FiN + I3 CMA FiM = 2I3 CAM FiM
−1
since C−1
AM = CMA . Now CAM FiM = (FiA FiM )
−1 −1
iM FiA FiM = FiM δAM = FiA ,
FiM = F−1 −1 −1 −1
so
∂I3
= 2I3 F−1
iA
∂FiA
This gives
∂W ∂W ∂I1 ∂W ∂I2 ∂W ∂I3

PiA = = + +
∂FiA ∂I1 ∂FiA ∂I2 ∂FiA ∂I3 ∂FiA
∂W ∂W ∂W −1
=2 FiA + 2 (Bij − I1 δij ) FjA + 2I3 F
∂I1 ∂I2 ∂I3 iA
Problem 8.5
The Cauchy stress is given by
1
tij = FjA PiA .
J
Start with the result of Problem 8.4 to show that

2 ∂W ∂W ∂W
tij = Bij + (Bik − I1 δik ) Bjk + I3 δij .
J ∂I1 ∂I2 ∂I3
Solution
Forming the indicated product with the result of Problem 8.4 gives

1 2 ∂W ∂W ∂W
tij = FjA PiA = FjA FiA + (Bik − I1 δik ) FjA FkA + I3 FjA F−1
iA
J J ∂I1 ∂I2 ∂I3

2 ∂W ∂W ∂W
= Bij + (Bik − I1 δik ) Bjk + I3 δij
J ∂I1 ∂I2 ∂I3
Problem 8.6
Assuming a strain energy of the form

1
W = w (λ1 ) + w (λ2 ) + w
λ 1 λ2
for an isotropic, incompressible material, show that
∂3 W ∂3 W
λ1 = λ 2 .
∂λ21 ∂λ2 ∂λ22 ∂λ1
Solution
−1
Let g (λ1 , λ2 ) = λ21 λ22 for an incompressible material, then
W = w (λ1 ) + w (λ2 ) + w (g (λ1, λ2 ))

Now differentiating W gives

∂3 W ∂2 ∂
= [w (λ1 ) + w (λ2 ) + w (g (λ1, λ2 ))]
∂λ21 ∂λ2 ∂λ21 ∂λ2

∂2

dw (λ2 ) ∂
= 2 + w (g (λ1, λ2 ))
∂λ1 dλ2 ∂λ2

∂ ∂ dw (λ2 ) ∂
= + w (g (λ1, λ2 ))
∂λ1 ∂λ1 dλ2 ∂λ2
∂3 w (g (λ1 , λ2 ))
=
∂λ21 ∂λ2
Using the chain rule gives
∂3 w (g (λ1 , λ2 )) ∂2 ∂w ∂g ∂ ∂2 w ∂g ∂g ∂w ∂2 g

= = +
∂λ21 ∂λ2 ∂λ21 ∂g ∂λ2 ∂λ1 ∂g2 ∂λ1 ∂λ2 ∂g ∂λ1 ∂λ2
2
∂3 w ∂g ∂2 w ∂g ∂2 g ∂2 w ∂2 g ∂g ∂w ∂3 g

∂g
= +2 2 + 2
+
∂g 3 ∂λ1 ∂λ2 ∂g ∂λ1 ∂λ1 ∂λ2 2
∂g ∂λ1 ∂λ2 ∂g ∂λ21 ∂λ2
Similarly
∂3 W ∂3 w (g (λ1 , λ2 ))
2
=
∂λ1 ∂λ2 ∂λ1 ∂λ22
and
2
∂3 w (g (λ1 , λ2 )) ∂3 w ∂2 w ∂g ∂2 g ∂2 w ∂2 g ∂g ∂w ∂3 g

∂g ∂g
2
= +2 2 + 2
+
∂λ1 ∂λ2 ∂g3 ∂λ2 ∂λ1 ∂g ∂λ2 ∂λ2 ∂λ1 2
∂g ∂λ2 ∂λ1 ∂g ∂λ22 ∂λ1
Now
∂g ∂ −2 −2 ∂g ∂ −2 −2
= λ λ = −2λ−3 −2
1 λ2 and = λ λ = −2λ−2 −3
1 λ2
∂λ1 ∂λ1 1 2 ∂λ2 ∂λ2 1 2
∂2 g ∂2 g
= = 4λ−3 −3
1 λ2
∂λ2 ∂λ1 ∂λ1 ∂λ2
and finally
∂3 g ∂3 g
= −12λ−4 −3
1 λ2 and = −12λ−3 −4
1 λ2
∂λ21 ∂λ2 ∂λ1 ∂λ22
Substituting yields
∂3 W ∂3 w ∂2 w ∂w
−8λ−8 −7
−28λ−6 −5
−12λ−4 −3

2
= 3 1 λ2 + 2 1 λ2 + 1 λ2
∂λ1 ∂λ2 ∂g ∂g ∂g
and
∂3 W ∂3 w −7 −8
∂2 w ∂w
−28λ−5 −6
−12λ−3 −4

2
= 3
−8λ 1 λ 2 + 2 1 λ2 + 1 λ2
∂λ1 ∂λ2 ∂g ∂g ∂g
Then
∂3 W ∂3 W
λ1 = λ 2
∂λ21 ∂λ2 ∂λ1 ∂λ22
Problem 8.7
For biaxial loading of a thin vulcanized rubber sheet the strain energy may be written as
W = C1 (I1 − 3) + C2 (I2 − 3) + C3 (I2 − 3)2
(Rivlin and Saunders, 1951).
(a) Use the definitions of invariants I1 , I2 , and I3 in terms of stretches λ1 , λ2 , and

λ3 to show
1 1 1
(I2 − 3)2 = 4 + 4 + 4 + 2I1 − 6I2 + 9 .
λ1 λ2 λ3
(b) Substitute the results from (a) into the strain energy above to obtain
w (λ1 ) = (C1 + 2C3 ) λ21 + (C2 − 6C3 ) λ−2 −4

1 + C3 λ3 − (C1 + C2 − 3C3 )
where
1
W = w (λ1 ) + w (λ2 ) + w .
λ1 λ2
Solution
(a) We note that for an incompressible material, λ21 λ22 λ23 = 1 and
(I2 − 3)2 = I22 − 6I2 + 9

2
1 1 1 1 1 1
= + + −6 + + +9
λ21 λ22 λ23 λ21 λ22 λ23

1 1 1 1 1 1
= 4 + 4 + 4 +2 + +
λ1 λ2 λ3 λ21 λ22 λ22 λ23 λ23 λ21

1 1 1
−6 + + +9
λ21 λ22 λ23
1 1 1
= 4 + 4 + 4 + 2 λ23 + λ21 + λ22

λ1 λ2 λ3

1 1 1
−6 + + +9
λ21 λ22 λ23
This gives
1 1 1
(I3 − 3)2 = 4
+ 4 + 4 + 2I1 − 6I2 + 9
λ1 λ2 λ3
(b) Now

1 1 1
W = C1 λ21 + λ22 + λ23 − 3 + C2

+ + − 3
λ21 λ22 λ23

1 1 1
+ C3 + + + 2I1 − 6I2 + 9
λ41 λ42 λ43
Collecting terms yields

C2 1 1
W = C1 λ21 + + C 3 + 2λ 2
1 − 6 + 3 − C1 − C2
λ21 λ41 λ21

C2 1 1
+ C1 λ22 + 2 + C3 + 2λ 2
2 − 6 + 3 − C1 − C2
λ2 λ42 λ22

C2 1 1
+ C1 λ23 + 2 + C3 + 2λ 2
3 − 6 + 3 − C1 − C2
λ3 λ43 λ23
where the constant terms were distributed equally across the three terms. Now
let
w (λ) = (C1 + 2C3 ) λ2 + (C2 − 6C3 ) λ−2 − (C1 + C2 − 3C3 )
Then

1
W = w (λ1 ) + w (λ2 ) + w (λ3 ) = w (λ1 ) + w (λ2 ) + w
λ1 λ 2
where incompressibility was used to eliminate λ3 .
Problem 8.8
Consider a material having reference configuration coordinates (R, Θ, Z) and current con-
figuration coordinates (r, θ, z). Assume a motion defined by
R
r= , θ = f (Θ) , z = Z .
g (Θ)
Determine FiA and Bij in terms of g, g0 , and f0 .
Answer
0 2
g0 f0 g0
   
1 1 g
− 0  + − 3 0 
 g
 g2 
 g2
 g 2 g 
FiA =
 f0 ,

Bij = 

f0 g0
0 2
f

 0 0 

g
 − 0 
   g3 g 
0 0 1 0 0 1
Solution
The physical components of the deformation gradient are (Malvern, Appendix II).
∂r 1 ∂r ∂r
 
g0
 
1
∂R R ∂Θ ∂Z − 0 
  g g2
  

∂θ r ∂θ ∂θ

f0
  
[FiA ] =  r r = 
 ∂R R ∂Θ ∂Z   0 0 

   g 
 ∂z 1 ∂z ∂z 
0 0 1
∂R R ∂Θ ∂Z
Now B = F · FT or Bij = FiA FjA in indicial notation. In matrix form
g0
  
1 1
 g − g2 0   g 0 0 
  
f0   g0 f0
[Bij ] = [FiA ] [FiA ]T = 
 
 0 0  − 2 0 

g  g g

 
0 0 1 0 0 1
 0 2 
1 g f0 g 0
 2+ 2
− 3 0 
 g g g 
 0 2 
= 0 0
fg f 
 − 3 0 
g g
 
 
0 0 1
Note: For further discussion or student projects see Rajagopal and Tao, “On an inho-
mogeneous deformation of a generalized Neo-Hookean material,” Journal of Elasticity, 28,
165-184, 1992.
Problem 8.9
Show
∂W 2 2 ∂W
t11 = 2 λ21 − λ−2 −2
− 2 λ−2

1 λ2 1 − λ 1 λ2 ,
∂I1 ∂I2
−2 ∂W 2 2 ∂W
= 2 λ22 − λ−2 − 2 λ−2

t22 1 λ2 2 − λ 1 λ2 ,
∂I1 ∂I2
are the nonzero Cauchy stress components for biaxial tensile loading of a homogeneous,
isotropic, incompressible rubber-like material by solving for the indeterminant pressure p
from the fact that t33 = 0 condition.
Solution
Consider the principal axes where
λ21
   
λ1 0 0 0 0
[FiA ] =  0 λ2 0  and [Bij ] =  0 λ22 0 
0 0 λ3 0 0 λ23
From Eq 8.32a
   2   −2 
1 0 0 λ1 0 0 λ1 0 0
∂W ∂W
[tij ] = p  0 1 0 +  0 λ22 0 +  0 λ−2
2 0 
∂I1 ∂I2
0 0 1 0 0 λ23 0 0 λ−2
3
Now
∂W 2 ∂W −2
t33 = p + λ + λ =0
∂I1 3 ∂I2 3
and
∂W 2 ∂W −2 ∂W 1 ∂W 2 2
p=− λ − λ =− − λ λ
∂I1 3 ∂I2 3 ∂I1 λ21 λ22 ∂I2 1 2
Substituting for p, The stresses are
∂W −2 −2 ∂W 2 2 ∂W 2 ∂W −2
t11 = − λ λ − λ λ + λ + λ
∂I1 1 2 ∂I2 1 2 ∂I1 1 ∂I2 1
∂W 2 ∂W −2
λ1 − λ−2 −2
λ1 − λ21 λ22

= 1 λ2 +
∂I1 ∂I2
and
∂W −2 −2 ∂W 2 2 ∂W 2 ∂W −2
t22 = − λ λ − λ λ + λ + λ
∂I1 1 2 ∂I2 1 2 ∂I1 2 ∂I2 2
∂W 2 ∂W −2
λ2 − λ−2 −2
λ2 − λ21 λ22

= 1 λ2 +
∂I1 ∂I2
Problem 8.10
The following compression force-deflection data was obtained for a highly filled, polybuta-
diene rubber having initial gauge length of 0.490 in and an undeformed cross-section area
of 1 in2
Displ. (in) Force (lbs)

–8.95E–04 –7.33E+00
–5.45E–03 –4.15E+01
–8.06E–03 –7.57E+01
–1.13E–02 –1.29E+02
–1.44E–02 –1.83E+02
–1.80E–02 –2.52E+02
–2.55E–02 –3.93E+02
–2.94E–02 –4.62E+02
–3.28E–02 –5.35E+02
–4.00E–02 –6.72E+02
–4.39E–02 –7.40E+02
–4.75E–02 –8.13E+02
–5.46E–02 –9.45E+02
–5.80E–02 –1.01E+03
–6.58E–02 –1.15E+03
–7.30E–02 –1.28E+03
–7.69E–02 –1.35E+03
–8.05E–02 –1.42E+03
–8.80E–02 –1.56E+03
–9.19E–02 –1.63E+03
–9.55E–02 –1.70E+03
–1.03E–01 –1.86E+03
–1.06E–01 –1.93E+03
–1.17E–01 –2.20E+03
–1.28E–01 –2.50E+03
–1.38E–01 –2.85E+03
–1.50E–01 –3.26E+03
–1.60E–01 –3.73E+03
–1.71E–01 –4.22E+03
–1.77E–01 –4.42E+03
–1.83E–01 –4.48E+03
(a) Generate a stress-stretch plot.

(b) Using a spreadsheet, or other appropriate tool, show that simple Mooney-Rivlin
constants C1 = 1, 550 and C2 = −500 represent the material for the range and
type of loading given.
(c) Generate a set of significantly different constants C1 and C2 which might equally
well model the material for this range and type of loading, showing that C1 and
C2 are not unique.
Chapter 9 Solutions
Problem 9.1
By substituting Sij = tij − 31 δij tkk and ηij = ij − 13 δij kk into Eq 9.7 and combining those
two equations, determine expressions in operator form for
(a) the Lamé constant, λ,

(b) Young’s modulus, E,
(c) Poisson’s ratio, ν.
Answer
(a) {λ} = K − 2{Q}/3{P}

(b) {E} = 9K{Q}/(3K{P} + {Q})
(c) {ν} = (3K{P} − 2{Q})/(6K{P} + 2{Q})
Solution From Eqs 9.7
{P} Sij = 2 {Q} ηij

tii = 3Kii
(a) Now
1 1
{P} tij − δij tkk = 2 {Q} ij − δij kk
3 3
and
2 {Q} 2 {Q}

tij = ij + δij K − kk
{P} 3 {P}
2 {Q}
{λ} = K −
3 {P}
(b) and (c) Inverting we have
{P} 3K {P} − 2 {Q}

ij = tij − δij tkk
2 {Q} 18K {Q}
From Eq6.28, we have
1 + {ν} {P} {ν} 3K {P} − 2 {Q}

= =
{E} 2 {Q} {E} 18K {Q}
199
Solving simultaneously gives
9K {Q} 3K {P} − 2 {Q}

{E} = {ν} =
3K {P} + {Q} 6K {P} + 2 {Q}
Problem 9.2
Compliances are reciprocals of moduli. Thus, in elasticity theory D = 1/E, J = 1/G, and
B = 1/K. Show from the stress-strain equations of a simple one-dimensional tension that
1 1
D= J+ B .
3 9
Solution
From Eq 6.27a,
1 λ+µ
D= =
E µ (3λ + 2µ)
By partial fractions
A C λ+µ
+ =
µ 3λ + 2µ µ (3λ + 2µ)
and
3A = 1 and 2A + C = 1
or
1 1
A= and C=
3 3
This gives
1/µ 3 J B
D= + = +
3 9 (2λ + 3µ) 3 9
See Table 6.1
Problem 9.3
The four-parameter model shown consists of a Kelvin unit in series with a Maxwell unit.
Knowing that γMODEL = γKELVIN +γMAXWELL , together with the operator equations Eqs 9.13
and 9.14, determine the constitutive equation for this model.
G1
G2 η2
T T
η1
Answer
G2 η1 γ̈ + G1 G2 γ̇ = η1 T̈ + (G1 + G2 + η1 /τ2 ) Ṫ + (G1 /τ2 ) T
Solution
The model and Eqs 9.13 and 9.14, give
T (∂t + 1/τ2 )
γ= + T
(G1 + η1 ∂t ) (G2 ∂t )
or
[(G1 + η1 ∂t ) (G2 ∂t )] γ = (G2 ∂t ) T + [(G1 + η1 ∂t ) (∂t + 1/τ2 )] T
Expanding, we find
G1 η1
G1 G2 γ̇ + G2 η1 γ̈ = G2 Ṫ + G1 Ṫ + η1 T̈ + T+ Ṫ
τ2 τ2
Rearranging gives

η1 G1
G2 η1 γ̈ + G1 G2 γ̇ = η1 T̈ + G1 + G2 + Ṫ + + T
τ2 τ2
Problem 9.4
Develop the constitutive equations for the three-parameter models shown.
G1 G2 η2
η2 T
T T T
η1 G1
(a) (b)
G2 η2
T T
η1
(c)
Answer
(a) γ̈ + γ̇/τ1 = [(η1 + η2 ) /η1 η2 ] Ṫ + (1/τ1 η2 ) T

(b) Ṫ + T/τ2 = (G2 + G1 ) γ̇ + (G1 /τ2 ) γ
(c) Ṫ + T/τ2 = η1 γ̈ + (G2 + η1 /τ2 ) γ̇
Solution (a) For this model

Ṫ T
γ̇ = +
(G1 + η1 ∂t ) η2
This gives
G1 η2 γ̇ + η1 η2 γ̈ = η2 Ṫ + G1 T + η1 Ṫ
and
γ̇ η1 + η2 1
γ̈ + = Ṫ + T
τ1 η1 η2 η2 τ1
(b) For this model
G2
T= γ̇ + G1 γ
(∂t + 1/τ2 )
This gives
1 G1 G1
Ṫ + T = G2 γ̇ + G1 γ̇ + γ = (G1 + G2 ) γ̇ + γ
τ2 τ2 τ2
(c) For this model
G2
T= γ̇ + η1 γ̇
(∂t + 1/τ2 )
Then
1 η1 η1
Ṫ + T = G2 γ̇ + η1 γ̈ + γ̇ = G2 + γ̇ + η1 γ̈
τ2 τ2 τ2
Problem 9.5
A proposed model consists of a Kelvin unit in parallel with a Maxwell unit. Determine the
constitutive equation for this model.
G2 η2
T G1 T
η1
Answer
Ṫ + T/τ2 = η1 γ̈ + (G1 + G2 + η1 /τ2 ) γ̇ + (G1 /τ2 ) γ
Solution This model gives
G2
T = (G1 + η1 ∂t ) γ + γ̇
(∂t + 1/τ2 )
Expanding gives
1 G1 η1
Ṫ + T = G1 γ̇ + η1 γ̈ + γ+ γ̇ + G2 γ̇
τ2 τ2 τ2
and collecting terms

1 η1 G1
Ṫ + T = η1 γ̈ + G1 + G2 + γ̇ + γ
τ2 τ2 τ2
Problem 9.6
For the four-parameter model shown, determine
(a) the constitutive equation,

(b) the relaxation function, G(t). [Note that G(t) is the sum of the G(t)’s of the
parallel joined units.]
G1
η2 T
T
G2
η3
Answer
(a) Ṫ + T/τ2 = η3 γ̈ + (G1 + G2 + η3 /τ2 ) γ̇ + (G1 /τ2 ) γ
(b) G(t) = G1 + G2 e−t/τ2 + η3 δ (t)
Solution (a) The model yields

G2
T = G1 γ + γ̇ + η3 γ̇
(∂t + 1/τ2 )
Expanding gives
1 G1 η3
Ṫ + T = G1 γ̇ + γ + G2 γ̇ + η3 γ̈ + γ̇
τ2 τ2 τ2
Collecting terms, we find

1 η3 G1
Ṫ + T = η3 γ̈ + G1 + G2 + γ̇ + γ
τ2 τ2 τ2
(b) To compute the relaxation function, we apply a step change in strain, γ = γ0 U (t).
Now γ̇ = γ0 δ (t), and γ̈ = γ0 δ̇ (t). This gives

1 η3 G1
Ṫ + T = η3 γ0 δ̇ (t) + G1 + G2 + γ0 δ (t) + γ0 U (t)
τ2 τ2 τ2
Integrating gives
Zt Zt
t0 /τ2 0 0 η3 0
Te t/τ2
= η3 γ0 e δ̇ (t ) dt + G1 + G2 + γ0 et /τ2 δ (t0 ) dt0
0 τ2 0
Zt
G1 0
+ γ0 et /τ2 U (t0 ) dt0
τ2 0
or

t/τ2 t/τ2 U (t) η3
Te = η3 γ0 e δ (t) − + G1 + G2 + γ0 U (t)
τ2 τ2

+ G1 γ0 et/τ2 − 1 U (t)
and h i
T = γ0 η3 δ (t) + G2 e−t/τ2 U (t) + G1 U (t)
Problem 9.7
For the model shown the stress history is given by the accompanying diagram. Determine
the strain γ(t) for this loading during the intervals
(a) 0 6 t/τ 6 2,
(b) 0 6 t/τ 6 4.
Use superposition to obtain answer (b).
t
G
T G T
T0
η 0 1 2 3 4 5 t/τ
Answer

(a) γ(t) = T0 J 2 − e−t/τ U (t)

(b) γ(t) = T0 J 2 − e−t/τ U (t) − T0 J 2 − e−(t−2τ)/τ U (t − 2τ)
Solution The constitutive equation is
1 1 2
γ̇ + γ = Ṫ + T
τ G η
(a) For T = T0 U (t) as shown in the diagram, we have
1 1 2
γ̇ + γ = T0 δ (t) + T0 U (t)
τ G η
Integrating gives
Z Z
T0 t t0 /τ 2T0 t t0 /τ
γet/τ = e δ (t0 ) dt0 + e U (t0 ) dt0
G 0 η 0
T0 2T0 t/τ
= U (t) + e − 1 U (t)
G G
and for 0 6 t 6 2τ
e−t/τ

2
1 − e−t/τ U (t) = T0 J 2 − e−tτ U (t)

γ (t) = T0 +
G G
(b) By superposition for 0 6 t 6 4τ, we have

γ (t) = T0 J 2 − e−tτ U (t) + T0 J 2 − e−(t−2τ)/τ U (t − 2τ)

Problem 9.8
For the model shown determine
(a) the constitutive equation,

(b) the relaxation function, G(t),
(c) the stress, T (t) for 0 6 t 6 t1 , when the strain is given by the accompanying
graph.
G γ
G η
T T
λ
1
η 0 t
Answer
(a) Ṫ + T/τ = ηγ̈ + 3Gγ̇ + (G/τ) γ

(b) G(t) = ηδ (t) + G 1 + e−t/τ2

(c) T (t) = λ 2η − 2e−t/τ Gt U (t)
Solution (a) The model’s constitutive equation gives
Gγ̇
T = Gγ + + ηγ̇
1
∂t +
τ
Expanding
1 G
Ṫ + T = ηγ̈ + (3G) γ̇ + γ
τ τ
(b) See the solution to Problem 9.6 with G1 = G2 = G and η2 = η3 = η, then

G (t) = ηδ (t) + G 1 + e−t/τ
(c) From the graph,
γ (t) = tλU (t) , γ̇ (t) = λU (t) γ̈ (t) = λδ (t)
and from (a)

1 λG
Ṫ + T = ηλδ (t) + (3G) λU (t) + tU (t)
τ τ
Integrating, we have
Zt Zt Zt
t0 /τ 0 0 t0 /τ 0 0λG 0
Te t/τ
= ηλ e δ (t ) dt + (3G) λ e U (t ) dt + t0 et /τ U (t0 ) dt0
0 0 τ 0
and
h i
T et/τ = ηλ + 3ηλ et/τ − 1 + λG et/τ (t − τ) + τ U (t)
or
T (t) = λ 2η − ηe−t/τ + Gt U (t)
Problem 9.9
For the model shown in Problem 9.8, determine T (t) when γ(t) is given by the diagram
shown here.
λ
1
γ0
0 t
Answer

T (t) = γ0 ηδ (t) + G 1 + e−t/τ U (t) + λ η 2 − e−t/τ + Gt U (t)
Solution For the γ0 portion from answer (b), Problem 9.8, we have
h i
T (t) = γ0 ηδ (t) + G 1 + e−t/τ U (t)
For the ramp portion from answer (c), Problem 9.8, we have
h i
T (t) = λ η 2 − e−t/τ + Gt U (t)
Combining we have the final result

h i h i
T (t) = γ0 ηδ (t) + G 1 + e−t/τ U (t) + λ η 2 − e−t/τ + Gt U (t)
Problem 9.10
The three-parameter model shown is subjected to the strain history pictured in the graph.
Use superposition to obtain T (t) for the t > t1 , from T (t) for t 6 t1 . Let γ0 /t1 = λ.
G2 γ
T T γ0
G1 η1 0 t1 t
Answer

For t 6 t1 ; T (t) = λ η1 1 − e−t/τ1 + G2 t U(t)

For t > t1 ; T (t) = λ η1 1 − e−t/τ1 + G2 t U(t)

−λ η1 1 − e−(t−t1 )/τ1 + G2 (t − t1 ) U(t − t1 )
Solution The constitutive equation for the model is
1 G2
Ṫ + T = (G1 + G2 ) γ̇ + γ
τ τ1
For t 6 t1 , γ = λtU (t), γ̇ = λU (t), and
1 G2
Ṫ + T = (G1 + G2 ) λU (t) + λtU (t)
τ τ1
Integrating, we find
Zt Zt
t0 /τ 0 0 G2 0
Te t/τ
= (G1 + G2 ) λ e U (t ) dt + λ et /τ t0 U (t0 ) dt0
0 τ1 0
and h i
T (t) = λ η1 1 − e−t/τ1 + G2 t U (t)
For t > t1 by superposition of adding a segment with a negative slope at t1 to give a

constant strain history beyond t1 , we have
h i
T (t) = λ η1 1 − e−t/τ1 + G2 t U (t)
h i
− λ η1 1 − e−(t−t1 )/τ1 + G2 (t − t1 ) U (t − t1 )
Problem 9.11
For the model shown determine the stress, T (t) at (a) t = t1 ; (b) t = 2t1 ; and (c) t = 3t1 ,
if the applied strain is given by the diagram. Use superposition for (b) and (c).
η γ
G
T T γ0
η 0 t1 2t1 3t1
Answer

(a) T (t1 ) = (γ0 η/t1 ) 2 − e−t1 /τ

(b) T (2t1 ) = (γ0 η/t1 ) −2 + 2e−t1 /τ − e−2t1 /τ

(c) T (3t1 ) = (γ0 η/t1 ) −e−t1 /τ + 2e−2t1 /τ − 3e−3t1 /τ
Solution
(a) The constitutive equation is
1
Ṫ + T = ηγ̈ + 2Gγ̇
τ
γ0 γ0 γ0
For t 6 t1 , we have γ (t) = tU (t); γ̇ (t) = U (t); and γ̇ (t) = δ (t). Substituting
t1 t1 t1
into the constitutive model gives
1 γ0 γ0
Ṫ + T = η δ (t) + 2G U (t)
τ t1 t1
Integrating we obtain,
Zt Zt
γ0 0 γ0 0
T et/τ = η et /τ δ (t0 ) dt0 + 2G et /τ U (t0 ) dt0
t1 0 t1 0
and
ηγ0
T (t) = 2 − e−t/τ U (t)
t1
For t = t1 , we have
ηγ0
T (t1 ) = 2 − e−t1 /τ
t1
(b) Next for t > t1 , we have by superposing a ramp with a slope with −2 γt10 tU (t). This
gives
ηγ0
h i h i
T (t) = 2 − e−t/τ U (t) − 2 2 − e−(t−t1 )/τ U (t − t1 )
t1
For t = 2t1 , we have
ηγ0 h i
T (2t1 ) = −2 + 2e−t1 /τ − e−2t1 /τ
t1
(c) Now for 2t1 6 t 6 3t1 , we superpose a term with the original positive slope to have a
zero strain history
ηγ0
h i h i
T (t) = 2 − e−t/τ U (t) − 2 2 − e−(t−t1 )/τ U (t − t1 )
t
h 1 i
+ 2 − e−(t−2t1 )/τ U (t − 2t1 )
and for t = 3t1 , we have

ηγ0 h −3t1 /τ i
T (3t1 ) = −e + e−2t1 /τ − e−t1 /τ
t1
Problem 9.12
If the model shown in the sketch is subjected to the strain history γ (t) = (σ0 /2G) 2 − e−t/2τ U (t),
as pictured in the time diagram, determine the stress, T (t).
γ
G η
T0  t

T T γ (t) =
2G
2 − e− 2τ
T0
2G
G
0 t
Answer

T (t) = T0 e−t/τ + U(t)
Solution The constitutive equation for the model is

1 G
Ṫ + T = 2Gγ̇ + γ
τ τ
For the given strain history,

T0 1 −t/2τ h i
γ̇ (t) = e U (t) + 2 − e−t/2τ δ (t)
2G 2τ
and substituting into the constitutive equation gives

1 1 −t/2τ h
−t/2τ
i T0 h i
Ṫ + T = T0 e U (t) + 2 − e δ (t) + 2 − e−t/2τ U (t)
τ 2τ 2τ
Zt Z
t0 /τ
h
−t0 /2τ
i
0 0 T0 t t0 /τ
Te t/τ
= T0 e 2−e δ (t ) dt + e U (t0 ) dt0
0 τ 0
This gives
T et/τ = T0 (2 − 1) + T0 et/τ U (t) ,
or
T et/τ = T0 e−t/τ + T0 U (t) ,
Problem 9.13
For the hereditary integral, Eq 9.38
Zt
γ (t) = J (t − t0 ) (dT (t0 ) /dt0 ) dt0
−∞
assume T (t) = est where s is a constant. Let τ = t − t0 be the “elapsed time” of the load
application and show that γ(t) = sest J̄ (s) where J̄ (s) is the Laplace transform of J(t).
Solution For the hereditary integral

Zt
dT (t0 ) 0
γ (t) = J (t − t0 ) dt
−∞ dt0
0
With T (t) = est , we have T (t0 ) = est , and
dT (t0 ) 0
0
= sest
dt
For τ = t − t0 , dτ = −dt0 , and when t0 = t, τ = 0 and when t0 = −∞, τ = ∞. Then
Z0 Z∞ Z∞
γ (t) = − J (τ) ses(t−τ) dτ = J (τ) sest e−sτ dτ = sest J (τ) e−sτ dτ
∞ 0 0
and
γ (t) = sest J̄ (s)
Problem 9.14
Using T (t) = est as in Problem 9.13, together with the hereditary integral Eq 9.41a
Zt
T (t) = G (t − t0 ) (dγ (t0 ) /dt0 ) dt0 ,
−∞
and the result of Problem 9.13 show that Ḡ (s) J̄ (s) = 1/s2 where Ḡ (s) is the Laplace
transform of G(t). Assume s is real.
Solution The stress is

Zt
T (t) = G (t − t0 ) (dγ (t0 ) /dt0 ) dt0
−∞
0
Now from Problem 9.13, γ (t0 ) = sest J̄ (s), then
dγ (t0 ) 0
0
= s2 est J̄ (s) = s2 es(t−τ) J̄ (s)
dt
Thus
Z0 Z∞
T (t) = − G (τ) s2 es(t−τ) J̄ (s) dτ = s2 est G (τ) e−sτ J̄ (s) dτ = est
∞ 0
from Problem 9.13. Then

s2 Ḡ (s) J̄ (s) = 1
or
1
Ḡ (s) J̄ (s) =
s2
Problem 9.15
Taking the hereditary integrals for viscoelastic behavior in the form Eq 9.40
Zt
γ (t) = J0 T (t) + T (t0 ) [dJ (t − t0 ) /d (t − t0 )] dt0
0
and Eq 9.41c
Zt
T (t) = G0 γ (t) + γ (t0 ) [dG (t − t0 ) /d (t − t0 )] dt0 ,
0
show that for the stress loading T (t) = est and with τ = t − t0 , the expression
G0 Ā (s) + J0 B̄ (s) + Ā (s) B̄ (s) = 0
results, where here

Z∞
Ā (s) = e−sτ (dJ/dτ) dτ
0
and Z∞
B̄ (s) = e−sτ (dG/dτ) dτ .
0
Solution For T (t) = est , consider the integral

Zt
dJ (t − t0 ) 0
T (t0 ) dt
0 d (t − t0 )
Let τ = t − t0 , then when t0 = 0, τ = t and dt0 = −dτ. The above integral becomes
Z0 Zt Zt
0 dJ (τ) dJ (τ) dJ (τ)
− est dτ = es(t−τ) dτ = est e−sτ dτ = est Ā (s)
t dτ 0 dτ 0 dτ
and from Eq 9.40

γ (t) = J0 est + est Ā (s)
From Eq 9.41c
Zt
dG (t − t0 )

st0
st st
dt0

e = G0 J0 + Ā (s) e + J0 + Ā (s) e
0 d (t − t0 )
Zt
dG (τ)
= G0 J0 + Ā (s) est + J0 + Ā (s) es(t−τ)

dτ
0 dτ
Zt
dG (τ)
= G0 J0 + Ā (s) est + J0 + Ā (s) est e−sτ

dτ
0 dτ
Now
est = G0 J0 + Ā (s) est + J0 + Ā (s) est B̄ (s)

and

1 = G0 J0 + Ā (s) + J0 + Ā (s) B̄ (s)
= 1 + G0 Ā (s) + J0 B̄ (s) + Ā (s) B̄ (s)
This
G0 Ā (s) + J0 B̄ (s) + Ā (s) B̄ (s) = 0
Problem 9.16
Let the stress relaxation function be given as G(t) = a(b/t)m where a, b, and m are con- m
1
stants and t is time. Show that the creep function for this material is J (t) = amπ sin mx bt
with m < 1. Use the identity Ḡ (s) J̄ (s) = 1/s2 where barred quantities are Laplace trans-
forms.
Solution Let
G (t) = abm t−m = abm tk−1
Since m < 1, k > 0. Then
abm Γ (k)
Ḡ (s) =
sk
where Γ (k) is the gamma function. Then
1 sk sk−2
J̄ (s) = = m 2
=
Ḡ (s) s 2 ab Γ (k) s abm Γ (k)
π
But Γ (k) Γ (1 − k) = , and so
sin πk
(sin πk) Γ (1 − k)
J̄ (s) =
abm πs2−k
By definition Γ (k + 1) = kΓ (k) and (1 − k) Γ (1 − k) = Γ (2 − k), this gives
(sin πk) (1 − k) Γ (1 − k) (sin πk) Γ (2 − k)

J̄ (s) = m 2−k
=
ab πs (1 − k) abm πs2−k (1 − k)
and since m = 1 − k, we have
[sin (1 − m) π] Γ (2 − k)
J̄ (s) =
abm mπs2−k
Inverting gives
m
sin (mπ) t2−k−1 sin (mπ) tm sin (mπ) t
J (t) = m
= m
=
ab mπ ab mπ amπ b
Problem 9.17
A three-parameter solid has the model shown. Derive the constitutive equation for this
model and from it determine (a) the relaxation function, and (b) the creep function for the
model.
G1 η1
T T
G2
Answer
(a) G(t) = G2 + G1 e−t/τ1

∗ ∗
(b) J (t) = (1/ (G1 + G2 )) e−t/τ1 + (1/G2 ) 1 − e−t/τ1
where τ∗1 = (G1 + G2 ) τ1 /G2
Solution
The constitutive equation is
G1 γ̇
T= + G2 γ
1
∂t +
τ1
and expanding gives
1 G2
Ṫ + T = (G1 + G2 ) γ̇ + γ
τ1 τ1
(a) The relaxation function is given by applying a step function in strain, γ = γ0 U (t); and
γ̇ = γ0 δ (t). This gives a differential equation for the stress as
1 G2
Ṫ + T = (G1 + G2 ) γ0 δ (t) + γ0 U (t)
τ1 τ1
Zt Zt
t0 /τ1 G20 0 0
Te t/τ1
= (G1 + G2 ) γ0 e δ (t ) dt + γ0 et /τ1 U (t0 ) dt0
0 τ1 0
G2 t/τ1
= (G1 + G2 ) γ0 U (t) + γ0 e − 1 U (t)
τ1
or h i
T (t) = γ0 (G1 + G2 ) e−t/τ1 + G2 1 − e−t/τ1 U (t)
This gives
G (t) = G2 + G1 e−t/τ1
(b) The relaxation function results from applying a step function in stress, T = T0 U (t),
and Ṫ = T0 δ (t). The constitutive equation can be rewritten as
G2 1 1
γ̇ + γ= Ṫ + T
(G1 + G2 ) τ1 G1 + G2 (G1 + G2 ) τ1
or
1 T0 T0
γ̇ + γ= δ (t) + U (t)
τ∗1 G1 + G2 G2 τ∗1
(G1 + G2 ) τ1
where τ∗1 = . Integrating, we have
G2
Zt Zt
∗ T0 0 ∗ T0 0 ∗
γet/τ1 = et /τ1 δ (t0 ) dt0 + et /τ1 U (t0 ) dt0
G1 + G2 0 G2 τ∗1 0
and
1 1 t/τ∗1
t/τ∗
γe 1 = T0 + e − 1 U (t)
G1 + G2 G2
This gives
1 ∗ 1 ∗

γ (t) = T0 e−t/τ1 + 1 − e−t/τ1 U (t)
G1 + G2 G2
and the relaxation function is
1 ∗ 1 ∗

J (t) = e−t/τ1 + 1 − e−t/τ1
G1 + G2 G2
Problem 9.18
A material is modeled as shown by the sketch. (a) For this model determine the relaxation
function, G(t). (b) If a ramp function strain as shown by the diagram is imposed on the
model, determine the stress, using the appropriate hereditary integral involving G(t).
G γ(t)
T 2G η T
λ
1
G 2η 0 t
Answer
(a) G(t) = G + 2Ge−2t/τ + Ge−t/2τ

(b) T (t) = Gλ t + 3τ − τe−2t/τ − 2τe−t/2τ U(t)
Solution
(a) From Eq 9.35, G (t) = Ge−t/τ for a Maxwell unit, and G (t) = G for a spring. Elements
in parallel are additive for relaxation functions
G (t) = G + 2Ge−2t/τ + Ge−t/2τ

(b) Using the hereditary integral, Eq 9.41a in the form

Zt
dγ (t0 )
T (t) = G (t − t0 ) dt0
0 dt0
With γ (t) = λtU (t); γ̇ (t) = λU (t) + λtδ (t) this gives
Zt h i
0 0
T (t) = [λU (t0 ) + λt0 δ (t0 )] G + 2Ge−2(t−t )/τ + Ge−(t−t )/2τ dt0
0
and h i
T (t) = Gλ t + 3τ − τe−2t/τ − 2τe−t/2τ U (t)
Problem 9.19
Determine the complex modulus, G∗ (iω) for the model shown using the substitution iω
for ∂t in the constitutive equation.
G1 η1
T T
G2
Answer
G∗ (iω) = G2 + (G1 + G2 ) τ21 ω2 + iG1 τ1 ω / 1 + τ21 ω2

Solution
The constitutive equation for the model is
G1 γ̇
T= + G2 γ
1
∂t +
τ1
and expanding
1 G2
Ṫ + T = (G1 + G2 ) γ̇ + γ
τ1 τ1
Now, we can substitute iω for ∂t in the above or rewrite the constitutive equation as
iω
T = G1 γ + G2 γ
1
iω +
τ1
and
(G1 + G2 ) iω + (G2 /τ1 ) τ1 (G1 + G2 ) iω + G2
T= γ= γ
1 (1 + iωτ1 )
iω +
τ1
Rationalizing the denominator gives

T τ1 (G1 + G2 ) iω + G2 1 − iωτ1 G2 + (G1 + G2 ) ω2 τ21 + iG1 τ1 ω
= =
γ (1 + iωτ1 ) 1 − iωτ1 1 + ω2 τ21
and
∗ G2 + (G1 + G2 ) ω2 τ21 + iG1 τ1 ω
G (iω) =
1 + ω2 τ21
Problem 9.20
Show that, in general, J0 = 1/G0 (1 + tan 2δ) and verify that G0 and J0 for the Kelvin model
satisfies this identity. (Hint: Begin with G∗ J∗ = 1.)
Solution Now we have

G0 + iG00 J0 − iJ00 1
G∗ J∗ = 1 or 0 0
= 0 0
G J GJ
But
G00 J00
= = tan δ
G0 J0
and so
1
(1 + i tan δ) (1 − i tan δ) =
G0 J0
This gives
1
J0 =
1 + tan2 δ

G0
For the Kelvin model, T = [G + η∂t ] γ or replacing ∂t with iω, we have
γ 1 G − iωη G − iωη 1 + iωτ

= J∗ = = 2 =
T G + iωη G − iωη G + ω2 η2 G (1 + ω2 τ2 )
This gives
1 1
J0 = = 0
G 1 + tan2 δ
2 2

G (1 + ω τ )
since G0 = G for the Kelvin model and J00 /J0 = tan δ = ωτ
Problem 9.21
Let the complex viscosity (denoted here by η∗ (iω)) be defined through the equation
T0 eiωt = η∗ iωγ0 eiωt .

Determine η∗ (iω) in terms of G∗ (iω) (see Eq 9.49) and calculate η∗ (iω) for the model
shown below.
G η
T T
η
Answer
η∗ (iω) = η 2 + ω2 τ2 − iτηω / 1 + ω2 τ2

Solution
From the definition of the complex viscosity
T0 eiωt = η∗ iωγ0 eiωt

and Eq 9.49 G∗ = T0 /γ0 , we find

T0
= G∗ = η∗ iω
γ0
Thus
G∗
η∗ =
iω
For the model, the constitutive equation is
1
Ṫ + T = ηγ̈ + 2Gγ̇
τ
or
1 h i
iω + T = η (iω)2 + 2Giω γ
τ
The complex relaxation modulus is
h i
2
T η (iω) + 2Giω
G∗ = =
iω + τ1

γ
and the complex viscosity is

h i
2
G ∗ η (iω) + 2Giω −τηω2 + 2τGiω
η∗ = = =
iω + τ1 iω

iω iω − ω2 τ
This simplifies to
∗ 2 + ω2 τ2 η − iηωτ
η (iω) =
1 + ω2 τ2
Problem 9.22
From Eq 9.55a in which G0 = T0 cos δ/γ0 and 9.55b in which G00 = (T0 sin δ)/γ0 show that
J0 = (γ0 cos δ)/T0 and that J00 = (γ0 sin δ)/T0 . Use G∗ J∗ = 1.
Solution
From the identity
G∗ J∗ = (G0 + iG00 ) (J0 − iJ00 ) = (G0 J0 + G00 J00 ) + i (G00 J0 − G0 J00 ) = 1
Collecting the real and imaginary parts gives
G0 J0 + G00 J00 = 1 (a)

G00 J0 − G0 J00 = 0 (b)
From (b) and Eq 9.55b, we have

T0 sin δ 0 T0 cos δ 00
J = J
γ0 γ0
and so from (a)

T0 cos δ 0 T0 sin δ

J + tan δ J0 = 1
γ0 γ0
Solving
γ0 cos δ γ0 sin δ
J0 = and J00 =
T0 T0
Problem 9.23
Show that the energy dissipated
R per cycle is related directly to the loss compliance, J00
by evaluating the integral T dγ over one complete cycle assuming T (t) = T0 sin ωt. (See
Fig. 9.10.)
Answer
Z
T dγ = T02 πJ00
Solution
For a complete cycle,
Z 2π/ω Z 2π/ω
dγ
T dt = T0 (sin ωt) γ0 ω cos (ωt − δ) dt
0 dt 0
and expanding
Z 2π/ω Z 2π/ω
dγ
T dt = T0 γ0 ω sin ωt (cos ωt cos δ + sin ωt sin δ) dt
0 dt 0
From the result of Problem 9.22

Z 2π/ω " Z Z 2π/ω #
dγ 2π/ω
sin 2ωt
T 2
dt = T0 ω J 0
dt + J00
sin ωtdt = T02 πJ00
2
0 dt 0 2 0
Problem 9.24
For the rather complicated model shown here, determine the constitutive equation and from
it G∗ (iω). Sketch a few points on the curve G00 vs. ln(ωτ).
T 2G η T
G 2η
Answer

T̈ + (5/2τ) Ṫ + 1/τ2 T = 4Gγ̈ + (11G/2τ) γ̇ + G/τ2 γ
G∗ (iω) = G0 + iG00 where
G0 = G 1 + (35/4) τ2 ω2 + 4τ4 ω4 / 1 + (17/4) τ2 ω2 + τ4 ω4

G00 = G 3τω + (9/2) τ3 ω3 / 1 + (17/4) τ2 ω2 + τ4 ω4

For ln(ωτ) = 0, G00 = 1.2G

For ln(ωτ) = 1, G00 = 1.13G
For ln(ωτ) = 2, G00 = 0.572G
For ln(ωτ) = ∞, G00 = 0
Solution
The constitutive equation for this model is
G 2G
T= 1
γ̇ + γ̇ + Gγ
∂t + 2τ ∂t + τ2
or upon expansion

5 1 1 G
∂2t + ∂t + 2 T = 4G∂2t + ∂t + 2 γ
2τ τ 2Gτ τ
Let ∂kt = (iω)k and

2 5 1 2 1 G
(iω) + (iω) + 2 T = 4G (iω) + (iω) + 2 γ
2τ τ 2Gτ τ
Now T/γ = G∗ and
35 2 2
+ 4τ4 ω4 + i 3τω + 92 τ3 ω3

∗ G 1+ 4 τ ω
G = 17 2 2
1+ 4 τ ω + τ4 ω4
For ln(ωτ) = 0, G00 = 1.2G
For ln(ωτ) = 1, G00 = 1.13G
For ln(ωτ) = 2, G00 = 0.572G
For ln(ωτ) = ∞, G00 = 0
Problem 9.25
A block of viscoelastic material in the shape of a cube fits snugly into a rigid container. A
uniformly distributed load p = −p0 U(t) is applied to the top surface of the cube. If the
material is Maxwell in shear and elastic in dilatation determine the stress component t11 (t)
using Eq 9.71. Evaluate t11 (0) and t11 (∞).
x2 p = −p0 U (t)
x1
Answer
t11 (t) = −p0 [1 − (6G/(3K + 4G))]e−(3K/(3K+4G)τ)t ]U(t)

t11 (0) = −p0 [(3K − 2G)/(3K + 4G)]
t11 (∞) = −p0
Solution
In compact operator notation
{P} Sij = 2 {Q} ηij
and tkk = 3Kkk
for a Maxwell shear behavior {P} = ∂t + and {Q} = (G∂t ). Now from symmetry,
1

τ
t11 = t33 , and from the loading condition = −p0 U (t). Also, 11 = 33 = 0. From
t22
Eq 9.71a, we have

1 1 1
∂t + t11 − (2t11 + t22 ) = 2 (G∂t ) − 22
τ 3 3
or
1 3K − 2G 1
ṫ11 + t11 = ṫ22 + t22
τ1 3K + 4G τ
3K + 4G
where τ1 = τ. Inserting t22 = −p0 U (t), and ṫ22 = −p0 δ (t) integrating and
3K
simplifying, gives
6G
t11 (t) = −p0 1 − e−t/τ U (t)
3K + 4G
3K − 2G
When t = 0, t11 = −p0 , and when t → ∞, t11 = −p0.
3K + 4G
Problem 9.26
A slender viscoelastic bar is loaded in simple tension with the stress T (t) = t11 (t) = T0 U(t).
The material may be modeled as a standard linear solid in shear having the model shown,
and as elastic in dilatation. Using the hereditary integrals, Eq 9.42, determine the axial
strain 11 (t) and the lateral strain 22 (t).
G
T G T t11 t11
η
Answer
11 (t) = T0 {[(6K + G)/3K − e−t/τ ]/3G}U(t)

22 (t) = T0 [(1/9K) − (2 − e−t/τ )/6G)]U(t)
Solution
The creep function for the model is the sum of the creep functions of the individual
components.
2 − e−t/τ
J (t) =
G
From Eq 9.42a and tkk = 3Kkk where tkk = T0 U (t) and Sij = 23 T0 U (t)
Zt 0
2 − e−(t−t )/τ 0 2 2 − e−t/τ

1 2
2 11 − kk = T0 δ (t0 ) dt = T0 U (t)
3 0 3 G 3 G
Thus
T0 6K + G −t/τ
11 = −e U (t)
G 3K
and since 22 = 33 by symmetry, from tkk = 3Kkk , T0 U (t) = 3K (11 + 222 ), we find
2 − e−t/τ

1
22 (t) = T0 − U (t)
9K 6G
Problem 9.27
A cylinder of viscoelastic material fits snugly into a rigid container so that 11 = 22 =
rr = 0 (no radial strain). The body is elastic in dilatation and has a creep compliance
JS = J0 (1 + t) with J0 a constant. Determine t33 (t) if ˙ 33 = A (a constant).
x3
x2
x1
r
Answer
t33 (t) = {A[Kt + 4(1 − e−t )/3J0 ]}U(t)
Solution
For JS = J0 (1 + t), the Laplace transform is

1 1 1+s
J̄S = J0 + 2 = J0
s s s2
1
and from J̄S ḠS = , we find
s2
J0
ḠS (s) =
1+s
Inverting, we have
e−t
G (t) =
J0
From Eq 9.43a
Zt 0
2Ae−(t−t )

1 2 4A
dt0 = 1 − e−t U (t)

t33 − tii =
3 0 3 J0 3J0
and since tii = 3Kii = 3K33 , then ṫii = 3KA and 13 tii = KAt. So
4 (1 − e−t )

t33 (t) = A Kt + U (t)
3J0
Problem 9.28
A viscoelastic body in the form of a block is elastic in dilatation and obeys the Maxwell
law in distortion. The block is subjected to a pressure impulse t11 = −p0 δ(t) distributed
uniformly over the x1 face. If the block is constrained so that 22 = 33 = 0, determine
11 (t) and t22 (t).
t33
t22
t11= −poδ(t)
Answer
t22 (t) = p0 [(2G − 3K)δ(t)/(3K + 4G) − (6G/(3K + 4G))e[−3K/(3K+4G)τ]t ]U(t)

11 (t) = p0 [3δ(t)/(3K + 4G) − [4G/(3K + 4G)K]e[−3K/(3K+4G)τ]t ]U(t)
Solution
Given t11 = −p0 δ (t); 11 = 33 = 0; this yields
tii = −p0 δ (t) + t22 + t33 and ii = 11 (t)
From the elastic response
(a) − p0 δ (t) + t22 + t33 = 3K11 (t)
and Eq 9.71

1 1
(b) {P} −p0 δ (t) + (p0 δ (t) − t22 − t33 ) = 2 {Q} 11 − 11
3 3

1 1
(c) {P} t22 + (p0 δ (t) − t22 − t33 ) = 2 {Q} − 11
3 3

1 1
(d) {P} t33 + (p0 δ (t) − t22 − t33 ) = 2 {Q} − 11
3 3
From (c) and (d), t22 = t33 , and from (a) and (c)

1 p0 δ (t) − 2t22
{P} t22 + (p0 δ (t) − 2t22 ) = 2 {Q}
3 9K
Then
(3 {P} K + 4 {Q}) t22 = (2 {Q} − 3K) p0 δ (t)
For a Maxwell material {P} = ∂t + τ1 and {Q} = (G∂t )

3K 3K
(3K + 4G) ṫ22 + t22 = (2G − 3K) p0 δ̇ (t) − p0 δ (t)
τ τ
Taking the Laplace transform gives
3K 3K
(3K + 4G) st̄22 + t̄22 = (2G − 3K) p0 s − p0
τ τ
or
2G − 3K
(s + λ) t̄22 = p0 s − λp0
3K + 4G
3K
where λ =
3K + 4G
2G − 3K s λ
t̄22 = p0 − p0
3K + 4G s + λ s + λ
Inverting gives

2G − 3K 6K
t22 = p0 δ (t) − p0 e−λt U (t)
3K + 4G 3K + 4G
To find 11 (t) use −p0 δ (t) + 2t22 = 3K11 (t)

3δ (t) 4G −λt
11 (t) = p0 − e U (t)
3K + 4G (3K + 4G) K
Problem 9.29
A viscoelastic cylinder is inserted into a snug fitting cavity of a rigid container. A flat,
smooth plunger is applied to the surface x1 = 0 of the cylinder and forced downward at
a constant strain rate ˙ 11 = 0 . If the material is modeled as the three-parameter solid
shown in shear and as elastic in dilatation, determine t11 (t) and t22 (t) during the downward
motion of the plunger.
G x2
T T
G η x3
x1
Answer
t11 (t) = −0 [(4Gτ/3)(1 − e−t/τ ) + (K + 4G/3)t]U(t)

t22 (t) = 0 [(2Gτ/3)(1 − e−t/τ ) + (−K + 2G/3)t]U(t)
Solution
The constitutive equation is
1 G
Ṫ + T = 2Gγ̇ + γ
τ τ
Thus {P} = ∂t + τ and {Q} = 2G∂t + τ . Here ˙ 11 = 0 and 11 =
1 G

−0 t, ii = −0 t,
t22 = t33 , so that tii = t11 + 2t22 . From Eq 9.71

1 1 G 1
∂t + t11 − (t11 + 2t22 ) = 2 2G∂t + −0 t + 0 t
τ 3 τ 3
and
tii = t11 + 2t22 = 3K11 = −3K0 t
Thus
1 8G 4G
ṫ11 + t11 = −K0 − K0 t − 0 − 0 t
τ 3 3τ
Integrating and simplifying

4Gτ −t/τ 4G
t11 (t) = −0 1−e + K+ t U (t)
3 3
From t11 + 2t22 = −3K0 t, we have t22 = − 12 t11 − 32 K0 t and

2
−t/τ 2
t22 (t) = 0 Gτ 1 − e + −K + G t U (t)
3 3
Problem 9.30
For a thick-walled elastic cylinder under internal pressure p0 , the stresses are trr = A−B/r2 ;
tθθ = A + B/r2 and the radial displacement is given by ur = 1+ν E [A(1 − 2ν)r + B/r] where A
and B are constants involving p0 , E is Young’s modulus, and ν is Poisson’s ratio. Determine
trr , tθθ , and ur for a viscoelastic cylinder of the same dimensions that is Kelvin in shear
and elastic dilatation if p = p0 U(t).
trr
tθθ
trr
r
tθθ
p0
Answer
trr = same as elastic solution but with p0 now p0 U(t)

tθθ = same as elastic solution but with p0 now p0 U(t)
ur (t) = (3Ar/(6K + 2G))(1 − e−(3K+G)t/Gτ) )U(t) + (B/2Gr)(1 − e−t/τ )U(t)
Solution
Because there are no elastic constants in the expressions for the stresses, the stresses will
be the same for the viscoelastic tube but the pressure will be replaced by p0 U (t). For
the displacement u, the operator for (1 + ν) /E is {P} /2 {Q} and for (1 + ν) (1 − 2ν) /E
is [3P/ (6KP + 2Q)]. For a Kelvin material, {P} = 1 and {Q} = (G + η∂t ). Taking A =
A1 p0 U (t) and B = B1 p0 U (t)

3 1 B1 p0
u (t) = A1 rp0 U (t) + U (t)
6K + 2G + 2η∂t 2G + 2η∂t r
Taking the Laplace transform, we have
3 B1 p0 1
ū (s) = A1 rp0 +
(6K + 2G + 2ηs) s r (2G + 2ηs) s
Inverting, we find
3Ar B
u (t) = 1 − e−(3K+G)t/Gτ U (t) + 1 − e−t/τ U (t)
6K + 2G 2Gr
Problem 9.31
A viscoelastic half-space is modeled as Kelvin in shear, elastic in dilatation. If the point
force P = P0 e−t is applied at the origin of a stress free material at time t = 0, determine
trr (t) knowing that for an elastic half-space the radial stress is
trr = P0 [(1 − 2ν)A − B]/2π
where A and B are functions of the coordinates only.
P = P0e−t
x2
θ
r
x1
x3
Answer
trr = (P0 /2π)[(3(G − η)e−t + 9Ke−(3K+G)t/η ]A/(3K + G − η) − e−t B
Solution
Given P = P0 e−t , we have
P0 e−t
trr = [(1 − 2ν) A − B]
π
The operator form of (1 − 2ν) = 3 {Q} / (3K {P} + {Q}). for a Kelvin material {P} = 1 and
{Q} = (G + η∂t ). Therefore, the transform is

P0 3 (1/τ + s) B
t̄rr (s) = A−
2π (1 + s) (3K + G) s/η 1+s
Inverting using partial fractions
P0 3(G − η)e−t + 9Ke−(3K+G)t/η

trr (t) = A − e−t B U (t)
2π 3K + G − η
Problem 9.32
The deflection at x = L for an end-loaded cantilever elastic beam is w = P0 L3 /3EI. Deter-
mine the deflection w(L, t) for a viscoelastic beam of the same dimensions if P = P0 U(t)
assuming (a) one-dimensional analysis based on Kelvin material, and (b) three-dimensional
analysis with the beam material Kelvin in shear, elastic in dilatation. Check w(L, ∞) in
each case.
P0
Answer
(a) w(L, t) = (P0 L3 /3EI)(1 − e−t/τE )U(t), τE = η/E

w(L, ∞) = (P0 L3 /3EI), the elastic deflection.
(b) w(L, t) = (P0 L3 /3EI)[(1 − e−t/τ ) + (1/9K)e−t/τ ]U(t)
w(L, ∞) = (P0 L3 /3EI), the elastic deflection.
Solution
For an elastic beam w (L) = PL3 /3EI. For a viscoelastic beam, w (L, t) = P (t) L3 /3 {E} I
(a) in the one-dimensional theory, the Kelvin constitutive model is T = (E + η∂t ) . Thus
P0 U (t) L3
w (L, t) =
3 (E + η∂t ) I
or
1 P0 L3
ẇ + w= U (t)
τE 3ηI
Integrating, we have
Zt
P0 L3 0 P0 L3 t/τe
we1/τE = et /τe U (t0 ) dt0 = e − 1 U (t)
3ηI 0 3ηI
and
P0 L3
w (L, t) = 1 − e−t/τe U (t)
3EI
where τe = η/E. Thus

P0 L3
w (L, ∞) =
3EI
the elastic response.
(b) In 3-D theory, for a Kelvin material {P} = 1 and {Q} = (G + ∂t ) so that{E} = 9K (G + η∂t ) / (3K + G + ∂t ).
Thus
P0 (3K + G + ∂t ) U (t) L3
w (L, t) =
27K (G + η∂t ) I
and
1 P0 L3
ẇ + w = [(3K + G) U (t) + ηδ (t)]
τ 27KηI
Integrating
Z t Zt
P0 L3

1/τ t0 /τ 0 0 t0 /τ 0 0
we = (3K + G) e U (t ) dt + ηe δ (t ) dt
27KηI 0 0
continuing and simplifying
P0 L3 1 − e−t/τ e−t/τ

w (L, t) = +
3I E 9K
Thus
P0 L3
w (L, ∞) =
3EI
the elastic response.
Problem 9.33
A simply-supported viscoelastic beam is subjected to the time-dependent loading f(x, t) =
q0 t where q0 is a constant and t is time. Determine the beam deflection w(x, t) in terms
of the elastic beam shape X(x) if the beam material is assumed to be (a) one-dimensional
Kelvin, and (b) three-dimensional Kelvin in shear, and elastic in dilatation. Compare the
results.
q0t
L
Answer

(a) w(x, t) = X(x) t − τE (1 − e−t/τE ) U(t)

(b) w(x, t) = X(x) t − (3τE / (3K + G)) (1 − e−t/τE ) U(t)
where τE = η/E
Solution
(a) For a Kelvin material, T = (E + ∂t ) . The beam deflection differential equation is
d4 w
{EI} = f (x, t)
dx4
Assume a separable solution w (x, t) = X (x) T (t). Then
EIXIV = q0 and {E} T = t
and
1
Ṫ + T = Et/η
τE
where τE = η/E. Integrating
Zt
E 0
T et/τe = et /τE t0 dt0
η 0
and h i
T (t) = t − τE 1 − e−t/τE U (t)
9K{Q}/{P}
(b) With {E} = 3K+{Q}/{P} and {P} = 1 and {Q} = (E + ∂t ) for a Kelvin material. From
{E} T = Et, we have
1 E
Ṫ + T = [(3K + G) t + η]
τ 9Kη
Integrating
Z t Zt
t/τ E t0 /τ 0 0 t0 /τ 0 0
Te = (3K + G) e t dt + ηe U (t ) dt
9Kη 0 0
and
3KτE
T (t) = t − 1 − e−t/τ U (t)
3K + G
If the material is incompressible, K → ∞ and 3/ (3 + GK) → 1 so that (a) and (b) are equal
if τE = τ.
K10443
6000 Broken Sound Parkway, NW

Suite 300, Boca Raton, FL 33487
270 Madison Avenue
an informa business New York, NY 10016
www.taylorandfrancisgroup.com 2 Park Square, Milton Park
Abingdon, Oxon OX14 4RN, UK

Solution Manual For Continuum Mechanics For En3e R.N. Mase, B. Smelser, M. Rossmann

Uploaded by

Copyright:

Available Formats

Solution Manual For Continuum Mechanics For En3e R.N. Mase, B. Smelser, M. Rossmann

Uploaded by

Document Information

Original Title

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

Solution Manual For Continuum Mechanics For En3e R.N. Mase, B. Smelser, M. Rossmann

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What publisher published the book?

What publisher published the book?

SOLUTIONS MANUAL FOR

Boca Raton London New York

CRC Press is an imprint of the

No claim to original U.S. Government works

Printed in the United States of America on acid-free paper

International Standard Book Number: 978-1-4398-0898-6 (Paperback)

Using indicial notation, show that,

(b) Again, we find

This is zero by symmetry in i and j.

Again, this is zero by symmetry in k and and i.

This is the equation of a sphere with the desired properties.

(c) for arbitrary tensors A and B,

Aij Bij = A(ij) B(ij) + A[ij] B[ij] .

(b) The trace of A is  

since the product of a symmetric and skew-symmetric tensor is zero

(a) 3, (b) 3, (c) δik , (d) Ajk

δij δij = δjj = δii = δ11 + δ22 + δ33 = 1 + 1 + 1 = 3

(b) Contracting on k and then j gives

δij δjk δki = δij δji = δii = 3

(c) Contracting on j yields

or in symbolic notation, a = (c · g)h − (c · h)g.

ai = εijk bj ck = εijk εjmn gm hn ck = − (εjik εjmn gm hn ck ) = − (δim δkn − δin δkm ) gm hn ck

(a) 0, (b) 0, (c) a2 (T32 − T23 ), (d) −v2

ε3jk aj ak = ε31k a1 ak + ε32k a2 ak = ε312 a1 a2 + ε321 a2 a1 = a1 a2 − a2 a1 = 0

= Ami 2δmi + Amj 2δmj + Amk 2δmk = 6Amm

Bkk = 61 Akk ; Bij = 13 Aij − 1

λ = εijk ai bj ck = ε1jk a1 bj ck + ε2jk a2 bj ck + ε3jk a3 bj ck

Using the suggested substitutions for ai , bi , ci , we have A3

det A = εijk Ai1 Aj2 Ak3

show that by an arbitrary number of interchanges of columns of Aij we obtain

εqmn det A = εijk Aiq Ajm Akn

εqmn det A = εijk Aiq Ajm Akn

Multiplying by εqmn gives

(b) For (AB)T = BT AT , we have

The result is demonstrated.

and from Example 2.10,

with det A = 1. The transformation is Tij

t11 = t22 ; t22 = t33 ; t33 = t11 or t11 = t22 = t33

with det A = 1. The resulting transformation is

t12 = −t12 = −t21 = 0; t13 = −t13 = t23 = 0; t32 = −t31 = t31 = 0

For a tensor to be isotropic, T = λI

δ0ij = aiq ajm δqm = aim ajm = δij

(b) The transformation is

and for a proper orthogonal matrix

(b) The transformation is

|tij − λδij | = εijk (t1i − δ1i ) (t2j − δ2j ) (t3k − δ3j ) = 0

and show by expansion of this equation that

|tij − λδij | = εijk (t1i − λδ1i ) (t2j − λδ2j ) (t3k − λδ3k )

This is the desired result since

λ1 = 17, λ2 = 26, λ3 = −39

= (17 − λ) λ2 + 13λ − 230 − 784 = (17 − λ) (λ + 39) (λ − 26) = 0

For λ = 17, the eigenvector is found from

n21 + n22 + n23 = 1

together with the unit vector relation

n21 + n22 + n23 = 1

(b) The trace of A is

>> II_C = 0.5((trace(C))^2 - trace(CC))