Ebook Nonlinear Continuum Mechanics For Finite Elasticity Plasticity Multiplicative Decomposition With Subloading Surface Model PDF Full Chapter PDF
Ebook Nonlinear Continuum Mechanics For Finite Elasticity Plasticity Multiplicative Decomposition With Subloading Surface Model PDF Full Chapter PDF
Ebook Nonlinear Continuum Mechanics For Finite Elasticity Plasticity Multiplicative Decomposition With Subloading Surface Model PDF Full Chapter PDF
KOICHI HASHIGUCHI
Technical Adviser, MSC Software Ltd.
(Emeritus Professor of Kyushu University),
Tokyo, Japan
Elsevier
Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
Copyright © 2020 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, recording, or any information storage and
retrieval system, without permission in writing from the publisher. Details on how to seek
permission, urther information about the Publisher’s permissions policies and our arrangements
with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency,
can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright
by the Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical
treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in
evaluating and using any information, methods, compounds, or experiments described herein.
In using such information or methods they should be mindful of their own safety and the safety
of others, including parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors,
assume any liability for any injury and/or damage to persons or property as a matter of products
liability, negligence or otherwise, or from any use or operation of any methods, products,
instructions, or ideas contained in the material herein.
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
A catalog record for this book is available from the Library of Congress
ISBN: 978-0-12-819428-7
For Information on all Elsevier publications
visit our website at https://www.elsevier.com/books-and-journals
Preface xi
1. Mathematical fundamentals 1
1.1 Matrix algebra 1
1.1.1 Summation convention 1
1.1.2 Kronecker’s delta and alternating symbol 2
1.1.3 Matrix notation and determinant 2
1.2 Vector 6
1.2.1 Definition of vector 7
1.2.2 Operations of vector 7
1.3 Definition of tensor 15
1.4 Tensor operations 18
1.4.1 Properties of second-order tensor 18
1.4.2 Tensor components 19
1.4.3 Transposed tensor 20
1.4.4 Inverse tensor 21
1.4.5 Orthogonal tensor 22
1.4.6 Tensor decompositions 24
1.4.7 Axial vector 25
1.4.8 Determinant 27
1.4.9 Simultaneous equation for vector components 30
1.5 Representations of tensors 31
1.5.1 Notations in tensor operations 31
1.5.2 Operational tensors 32
1.5.3 Isotropic tensors 34
1.6 Eigenvalues and eigenvectors 35
1.6.1 Eigenvalues and eigenvectors of second-order tensor 35
1.6.2 Spectral representation and elementary tensor functions 37
1.6.3 Cayley Hamilton theorem 38
1.6.4 Scalar triple products with invariants 39
1.6.5 Second-order tensor functions 39
1.6.6 Positive-definite tensor and polar decomposition 40
1.6.7 Representation theorem of isotropic tensor-valued tensor function 42
1.7 Differential formulae 43
1.7.1 Partial derivatives of tensor functions 43
1.7.2 Time-derivatives in Lagrangian and Eulerian descriptions 48
1.7.3 Derivatives of tensor field 49
1.7.4 Gauss’ divergence theorem 51
1.7.5 Material-time derivative of volume integration 52
v
vi CONTENTS
xi
xii Preface
Henann and Anand (2009), Brepols et al. (2014), etc., in which constitu-
tive relations are formulated in the intermediate configuration imagined
fictitiously by the unloading to the stress-free state along the hyperelas-
tic relation, based on the isoclinic concept (Mandel, 1971). However, the
plastic flow rule with the generality unlimited to the elastic isotropy
remains unsolved and only the conventional plasticity model, named by
Drucker (1998), with the yield surface enclosing the elastic domain have
been incorporated so that only the monotonic loading behavior of elasti-
cally isotropic materials is concerned in them.
The subloading multiplicative hyperelastic based plastic model has
been formulated by the author recently (Hashiguchi, 2018c), which is
capable of describing the finite elastoplastic deformation/rotation rigor-
ously under the monotonic/cyclic loading process. Further, it has been
extended to the subloading-multiplicative hyperelastic-based viscoplas-
ticity recently, which is capable of describing the rate-dependent elasto-
plastic deformation behavior at the general rate from the static to the
impact loading. It is to be the best opportunity to review the multiplica-
tive hyperelastic based plasticity comprehensively and explain the
detailed formulation of the subloading multiplicative hyperelas-
tic based plastic model systematically. This is the first book on the sub-
loading multiplicative hyperelastic based plasticity and viscoplasticity
for the description of the general irreversible deformation/sliding
behavior.
The subloading surface model and the multiplicative hyperelas-
tic based plasticity are explained comprehensively providing the
detailed physical interpretations for all relevant concepts and the deriv-
ing processes of all equations. Further, the incorporation of the subload-
ing surface model to the multiplicative hyperelastic plastic relation is
described in detail. Further, it is extended to the description of the vis-
coplastic deformation by incorporating the concept of overstress, which
is capable of describing the general rate of deformation ranging from
the quasistatic to the impact loading behaviors (Hashiguchi, 2016a,
2017a). In addition, the exact hyperelastic based plastic and viscoplastic
constitutive equation of friction (Hashiguchi, 2018c) is formulated rigor-
ously, while the hypoelastic-based plastic constitutive equation of fric-
tion has been formulated formerly (Hashiguchi et al., 2005; Hashiguchi
and Ozaki, 2008; Hashiguchi, 2013a).
The aim of this book is to give a comprehensive explanation of the
finite elastoplasticity theory and viscoplasticity under the monotonic
and the cyclic loading processes. The incorporation of the Lagrangian
tensors is required originally in the formulation of finite elastoplasticity
and viscoplasticity, since the deformation of the material involved in
the reference configuration, which is invariant through the deformation,
is physically relevant. Therefore the necessity and the meanings of the
Preface xv
Koichi Hashiguchi
June 2020
C H A P T E R
1
Mathematical fundamentals
The basic matrix algebra with some conventions and symbols appear-
ing in the continuum mechanics are described in this section.
8
>
> X 3
>
> u v 5 ur vr 5 u1 v1 1 u2 v2 1 u3 v3 ;
>
>
r r
>
> r51
>
< X3
Trr 5 Trr 5 T11 1 T22 1 T33 (1.1)
>
>
>
> r51
>
> X 3
>
>
> Tir vr 5
: Tir vr 5 Ti1 v1 1 Ti2 v2 1 Ti3 v3 ;
r51
fulfilling
δir δrj 5 δij 5 δji ; δii 5 3 (1.4)
where ð ÞT stands for the transpose of the row and the column in the
matrix.
The quantity defined by the following equation is called the determi-
nant of T and is shown by the symbol det T, that is,
T11 T12 T13
detT 5 εijk T1i T2j T3k 5 εijk Ti1 Tj2 Tk3 5 T21 T22 T23 (1.12)
T31 T32 T33
with
detTT 5 detT; detðsTÞ 5 s3 detðTÞ (1.13)
Here, the number of permutations that the suffixes i, j, and k in εijk can
take is 3!. Therefore Eq. (1.12) can be written as
1
detT 5 εijk εpqr Tip Tjq Tkr (1.14)
3!
Eq. (1.14) is rewritten as
1 1 1
detT 5 Trs ðcof TÞrs ; detT 5 T: ðcof TÞ 5 trðTðcofTÞT Þ (1.15)
3 3 3
or
detT 5 T1s ðcofTÞ1s 5 T2s ðcofTÞ2s 5 T3s ðcofTÞ3s
(1.16)
5 Tr1 ðcofTÞr1 5 Tr1 ðcofTÞr1 5 Tr2 ðcofTÞr2 5 Tr3 ðcofTÞr3
where
1
ðcofTÞip εijk εpqr Tjq Tkr (1.17)
2!
noting
1 1 1 1
εijk εpqr Tip Tjq Tkr 5 Tip εijk εpqr Tjq Tkr 5 Tip ðcofTÞip
3! 3 2! 3
ðcofTÞij is called the cofactor for the i-column and the j-row. The cofactor
is obtained through multiplying the minor determinant lacking the
ith row and jth column components by the sign ð21Þi1j .
The following lemmas for the properties of the determinant hold.
Lemma 1.1: If the first and the second rows are same, that is, T2j 5 T1j
for instance, we have εijk T1i T1j T3k 5 εjik T1j T1i T3k 5 2 εijk T1i T1j T3k .
Therefore we have the lemma “the determinant having same lines or
rows is zero.” Therefore the following relation is obtained from
Eq. (1.16) that
Tis Δjs 5 Tri Δrj 5 δij detT (1.18)
Lemma 1.2: If the first and the second lines are exchanged, that is, 122
for instance, we have εijk T2i T1j T3k 5 εjik T1i T2j T3k 5 2 εijk T1i T2j T3k .
Therefore we have the lemma “the determinant changes only its sign by
exchanging lines (or rows).”
By multiplying εijk to both sides in Eq. (1.12), we have
εijk detT 5 εijk εpqr T1p T2q T3r 5 εpqr Tip Tjq Tkr (1.19)
The additive decomposition of the components T2j into T2j 5 A2j 1 B2j
leads to
εijk T1i ðA2j 1 B2j ÞT2k 5 εijk T1i A2j T2k 1 εijk T1i B2j T2k (1.20)
noting εijk Bpi Bqj Brk 5 εpqr detB due to Eq. (1.19), and thus one has the fol-
lowing product law of determinant.
detðABÞ 5 detAdetB (1.21)
which leads to
@detT @detT
5 cofT ; 5 ðcofTÞij (1.22)
@T @Tij
The permutation symbol in the third order, that is, εijk appears often
hereinafter. It is related to Kronecker’s delta by the determinants as
δ1i δ1j δ1k δ1i δ2i δ3i
εijk 5 δ2i δ2j δ2k 5 δ1j δ2j δ3j (1.23)
δ3i δ3j δ3k δ1k δ2k δ3k
which can be proved as follows: Note that the second side in Eq. (1.23)
is expanded as
δ1i δ1j δ1k
εijk 5 δ2i δ2j δ2k 5 δ1i δ2j δ3k 1 δ1k δ2i δ3j 1 δ1j δ2k δ3i
δ3i δ3j δ3k
2 δ1k δ2j δ3i 2 δ1i δ2k δ3j 2 δ1j δ2i δ3k
for instance. The third side in Eq. (1.23) can be confirmed as well.
The following relations are obtained from Eqs. (1.23) and (1.21).
2 3 2 3
δ1i δ2i δ3i δ1p δ1q δ1r δ1i δ2i δ3i δ1p δ1q δ1r
6 7 6 7
εijk εpqr 5 δ1j δ2j δ3j δ2p δ2q δ2r 5 4 δ1j δ2j δ3j 5 4 δ2p δ2q δ2r 5
δ1k δ2k δ3k δ3p δ3q δ3r δ1k δ2k δ3k δ3p δ3q δ3r
(1.24)
δip δiq δir
5 δjp δjq δjr
δkp δkq δkr
1.2 Vector
where v1, v2, and v3 are the components of v. Denoting the angle of the
direction of vector v from the direction of the base vector ei by θi ,
cosθi 5 n ei is called the direction cosine by which the component of v is
given as
vi 5 v ei 5 :v:n ei 5 :v:cosθi (1.30)
The magnitude of vector v and its unit direction vector n are given
from Eq. (1.30), noting cos2 θ1 1 cos2 θ2 1 cos2 θ3 5 1 as follows:
pffiffiffiffiffiffiffiffi v vr
:v: vr vr ; n 5 er (1.31)
:v: :v:
Because of a b 5 ar er bs es 5 ar bs δrs , the scalar product is expressed
by using the components as
a b 5 ar br (1.32)
The quantity obtained by the scalar product is a scalar and the fol-
lowing commutative, distributive, and associative laws hold.
a b5b a (1.34)
a ðb 1 cÞ 5 a b 1 a c (1.35)
sða bÞ 5 ðsa bÞ 5 a ðsbÞ 5 ða bÞs (1.36)
ðaa 1 bbÞ c 5 aa c 1 bb c (1.37)
a × b = an
c v = (a × b) · c =[abc]
n
(|| n|| = 1)
b
a = || a × b|| =|| a || ||b|| sin θ
θ
a
FIGURE 1.1 Vector product (surface vector) and scalar triple product (volume) with
area.
½abc 5 ða 3 bÞ c 5 εijr ai bj er ck ek 5 εijr ai bj ck δrk 5 εijk ai bj ck
a1 b1 c1 a1 a2 a3
(1.48)
5 a2 b2 c2 5 b1 b2 b3
a3 b3 c3 c1 c2 c3
fulfilling
½abc 5 ½bca 5 ½cab 5 2 ½bac 5 2 ½cba 5 2 ½acb (1.49)
noting the fact that the term in the right-hand side of this equation is 1
½v1 ; v2 ; v3 , 2½v1 ; v2 ; v3 , and 0 when indices i; j; k are even and odd per-
mutations and two of indices coincide with each other, respectively.
Here, note that the scalar triple product [v1 v2 v3] designates the volume
of the parallelopiped formed by the vectors v1, v2, v3 in this order.
Here, the following equations hold for the scalar triple product.
½ei ej ek 5 εijk (1.51)
½sa; b; c 5 ½a; sb; c 5 ½a; b; sc 5 s½abc (1.52)
½aa 1 bb; c; x 5 a½ucx 1 b½bcx (1.53)
noting
a 3 ðb 3 cÞ 5 aj ej 3 ðεkpq bp cq ek Þ 5 aj εkpq bp cq ej 3 ek 5 εkpq aj bp cq εkij ei
Then, the following relations hold for the vector triple product.
a 3 ðb 3 cÞ
5 ða cÞb 2 ða bÞc
ða 3 bÞ 3 c
5 ða cÞb 2 ðb cÞa (1.55)
Permuting the letters and taking the summation in Eq. (1.55), it fol-
lows that
a 3 ðb 3 cÞ 1 b 3 ðc 3 aÞ 1 c 3 ða 3 bÞ 5 0 (1.56)
noting
a 3 ðb 3 cÞ 1 b 3 ðc 3 aÞ 1 c 3 ða 3 bÞ
5 ½ða cÞb 2 ða bÞc 1 ½ða bÞc 2 ða cÞb 1 ½ðc bÞa 2 ðc aÞb 5 0
One has
ða 3 bÞ ðc 3 dÞ 5 ða cÞðb dÞ 2 ðb cÞða dÞ (1.57)
noting
ða 3 bÞ ðc 3 dÞ 5 ½ða 3 bÞ c d 5 fða 3 bÞ 3 cg d 5 fða cÞb 2 ðb cÞag d
by virtue of Eq. (1.55),
Setting t 5 a 3 b in t 3 ðc 3 xÞ 5 ðt dÞc 2 ðt cÞd due to Eq. (1.55)1, one
has
ða 3 bÞ 3 ðc 3 dÞ 5 ½ða 3 bÞ dc 2 ½ða 3 bÞ cd 5 ½abdc 2 ½abcd (1.58)
It follows from Eq. (1.58) setting c-b and d-c for the particular
case that
ða 3 bÞ 3 ðb 3 cÞ 5 ½abcb 2 ½abbc 5 ½abcb (1.59)
which leads to
ða 3 bÞ 3 ðb 3 cÞ ðc 3 aÞ 5 ½abcb ðc 3 aÞ
so that
½a 3 b b 3 c c 3 a 5 ½abc2 (1.60)
Because of
ða 3 bÞ ½ðc 3 xÞ 3 ðy 3 zÞ 5 ða 3 bÞ ð½cxzy 2 ½cxyzÞ
5 ½aby½cxz 2 ½abz½cxy
v 5 va a 1 vb b 1 vc c (1.62)
that is,
½bcv ½cav ½abv
va 5 ; vb 5 ; vc 5 (1.63)
½abc ½abc ½abc
(1.65)
where
b3c c3a a3b
a ; b ; c
½abc ½abc ½abc
(1.66)
b 3 c c 3 a a 3 b
a ; b ; c
½a b c ½a b c ½a b c
ða ; b ; c Þ are called the reciprocal vectors of the primary vectors ða; b; cÞ.
The lower part of Eq. (1.66) is verified as
ðc 3 aÞ 3 ða 3 bÞ ½abca a
b 3 c 5 5 5
½abc 2
½abc 2 ½abc
½a b c 5
ðb 3cÞ 3ðc3 aÞ ða 3bÞ ½bca½bca2½bcb½aca ½abc2
5 5 5
1
½abc3
½abc 3
½abc 3 ½abc
by using Eqs. (1.59) and (1.60), leading to the first equation for a, for
example. Here, the following relations hold.
a a 5 1; b b 5 1; c c 5 1
a b 5 a c 5 b c 5 b a 5 c a 5 c b 5 0 (1.67)
½abc½a b c 5 1
leading to
v a 5
a v ; v b 5
b v
; v c 5
c v
(1.70)
½abc ½abc ½abc
v5
a v
ðb 3 cÞ 1
b v
ðc 3 aÞ 1
c v
ða 3 bÞ
(1.71)
½abc ½abc ½abc
v 5 ða vÞa 1 ðb vÞb 1 ðc vÞc 5 ða vÞa 1 ðb vÞb 1 ðc vÞc (1.72)
where the third side is obtained from the second side by exchanging
a; b; c and ða ; b ; c Þ. Eq. (1.72) is rewritten as
v 5 gv (1.73)
where
g is regarded as the generalized identity tensor and will be called the met-
ric tensor in the general coordinate system in Section 2.2. In particular, g
is reduced in the normalized rectangular coordinate system as follows:
I 5 δij ei ej 5 ei ei ;
δij 5 ei ej (1.75)
ε:ðuvÞ 5 u 3 v (1.77)
noting
ðεijk ei ej ek Þ:ður er vs es Þ 5 εijk ei ur vs δjr δks 5 εijk ei uj vk 5 uj ej 3 vk ek
because of
ðai ei bj ej Þck ek 5 ai ei ðbj ck δjk Þ 5 ai ei ðbj cj Þ
8
>
>
>
xðabÞy 5 ða xÞðb yÞ
>
> ðabÞ:ðcdÞ 5 ðacÞðbdÞ
>
>
>
< ðxyÞ:ðabcdÞ 5 ða xÞðb yÞcd
(1.86)
> ðxyÞ:ðabcÞz 5 ða xÞðb yÞðc zÞ
>
>
>
>
>
>
>
:
ðabcdÞ:ðxyÞ 5 ðc xÞðd yÞab
ðwxÞ:ðabcdÞ:ðyzÞ 5 ða wÞðb xÞðc yÞðd zÞ
The vector product is represented in the direct notation of the alter-
nating tensor ε as follows:
a 3 b 5 ε:ðabÞ; ei 3 ej 5 ε:ðei ej Þ (1.87)
because of
a 3 b 5 εijk aj bk ei 5 εijk ei ar ðej er Þbs ðek es Þ 5 εijk ei ej ek :ðar er bs es Þ (1.88)
Let the set of nm functions in the coordinate system fO-xi g with the
origin O and the axes xi ði 5 1; 2; . . .; nÞ in the n-dimensional space be
designated by Tðp1 ; p2 ; . . .; pm Þ, where each of the indices p1 ; p2 ; . . .; pm
takes the number 1; 2; . . .; n. This set of functions is defined as the
mth-order tensor in the n-dimension, if the set of functions is observed
in the other coordinate system fO-xi g with the origin O and the axes xi
as follows:
@xp1 @xp2 @xpm
T ðp1 ; p2 ; . . .pm Þ 5 ? Tðq1 ; q2 ; . . .qm Þ (1.89)
@xq1 @xq2 @xqm
provided that only the directions of axes are different and the
relative motion between the axes does not exist. Here, we introduce the
notation
@xi
Qij 5 (1.90)
@xj
which fulfills
Qir Qjr 5 δij (1.91)
because of
@xi @xr
Qir Qjr 5
@xr @xj
Because of
Qp1 r1 Qp2 r2 ?Qpm rm T p1 p2 ?pm
5 Qp1 r1 Qp2 r2 ?Qpm rm Qp1 q1 Qp2 q2 ?Qpm qm T q1 q2 ?qm
5 ðQp1 r1 Qp1 q1 ÞðQp2 r2 Qp2 q2 Þ?ðQpm rm Qpm qm ÞTq1 q2 ?qm
5 δr1 q1 δr2 q2 ?δrm qm Tq1 q2 ?qm
Indices put in a tensor take the dimension of the space in which the
tensor is based. The number of indices which is equal to the number of
operators Qij is called the order of tensor. For instance, the transforma-
tion rules of the first-order tensor, that is, vector vi and the second-order
tensor Tij are given by
vi 5 Qir vr ; vi 5 Qri vr
Tij 5 Qir Qjs Trs ð 5 Qir Trs Qjs Þ; T ij 5 Qri Qsj T rs ð 5 Qri Trs Qsj Þ (1.95)
objective vectors and tensors usually; they are influenced by the relative
rate of motion between the coordinate systems, that is, the rigid-body
rotational rate of material. Constitutive equations of materials must be
formulated in terms of objective tensors, since material properties are
not influenced by the rigid-body rotation of material and therefore must
be described in the identical form independent of the coordinate
systems.
The following linearity is satisfied in tensors.
8
< Additivity: Tp1 p2 ?pm Ap1 p2 ?pl 1 Bp1 p2 ?pl
>
5 Tp1 p2 ?pm Ap1 p2 ?pl 1 T p1 p2 ?pm Bp1 p2 ?pl (1.96)
>
:
Homogeneity: Tp1 p2 ?pm sAp1 p2 ?pl 5 sT p1 p2 ?pm Ap1 p2 ?pl
for the component transformation of the general tensor, let Eqs. (1.93)
and (1.94) be represented formally by the symbolic (direct) notation for
convenience as follows:
v 5 Qv; v 5 QT v
(1.99)
T 5 QTQT ; T 5 QT T Q
Denoting the unit base vectors e1 ; e2 ; . . .; em of the coordinate axes
x1 ; x2 ; . . .; xm ,
the quantity Qij in Eq. (1.90) is represented in terms of the
base vectors as follows:
Qij 5 ei ej (1.100)
noting
@xi
Qij 5
@xj
5 ei @x
@xj
es 5 ei
s @xs
@xj
es 5 ei δjs es
noting
ei 5 ðei er Þer ;
ei 5 ðei er Þer
Various tensors and their algebra are addressed in this section, which
are used often throughout this book.
Two tensors A and B are same when they yield same transformation
of an arbitrary vector v, that is,
A 5 B when Av 5 Bv (1.105)
T ij 5 ei Tej (1.121)
is given by
Ξ ijk 5 ðei ej Þ:Ξek (1.123)
is given by
Tijkl 5 ðei ej Þ:T:ðek el Þ (1.125)
a Tb 5 b TT a (1.126)
Noting
a ðuvÞb 5 b ðvuÞa
it follows from Eq. (1.126) that
ðuvÞT 5 vu (1.127)
that is,
ðTT Þij 5 ðTÞji (1.129)
Ta b 5 a TT b 5 a bT ððTri ai Þbr 5 ai ðTri br Þ 5 ai ðbr T ri ÞÞ (1.141)
(
Tab 5 aTT b ððT ir ar Þbi 5 ðar T ir Þbi Þ
(1.142)
aTb 5 abTT ðai ðTjr br Þ 5 ai ðbr T jr ÞÞ
where the components are denoted as Tij21 ðT21 Þij for brevity. Then,
there follows:
ðABÞ21 5 B21 A21 (1.144)
because of
ðABÞðB21 A21 Þ ð 5 ABB21 A21 5 AIA21 5 AA21 Þ 5 I
ðQaÞ ðQbÞ 5 a ðQT QbÞ (1.147)
Comparing Eq. (1.147) with Eq. (1.146), the orthogonal tensor must
fulfill
QQT 5 QT Q 5 I (1.148)
leading to
QT 5 Q21 (1.149)
noting
Qij ei ej 5 ei ðei ej Þej 5 ei ei 5 ðei er Þer ei 5 Qri er ei
QQT 5 ðei ei Þðej ej ÞT 5 ei ei ej ej 5 δij ei ej
Qij 5 ei ej 5 cosðangle between ei and ej Þ
ei 5 QT ei 5 Qir er ; ei 5 Qei 5 Qri er ; (1.156)
Q 5 Qij ei ej 5 Qij ei ej 5 ei ei
that is,
Q 5 Q11 e1 e1 1 Q12 e1 e2 1 Q21 e2 e1 1 Q22 e2 e2 1 e3 e3
5 cosθe1 e1 1 sinθe1 e2 2 sinθe2 e1 1 cosθe2 e2 1 e3 e3 (1.158)
5 ðe1 e1 1 e2 e2 Þcosθ 1 ðe1 e2 2 e2 e1 Þsinθ 1 e3 e3
where
1 1
S sym½T ðT 1 TT Þ; Ω ant½T ðT 2 TT Þ (1.161)
2 2
which satisfy
ST 5 S; ΩT 5 2 Ω (1.162)
a Sb 5 b Sa
a ðΩbÞ 5 2 b ðΩaÞ
(1.163)
noting Eq. (1.126). Eq. (1.160) is called the Cartesian decomposition. Here,
it follows that
SST 5 S2 (1.164)
trðSΩÞ 5 trðSΩ Þ 5 0 T
(1.165)
Choosing a as the base vector ej and making the scalar product with
ei , we find
Ω ij 5 ei Ωej 5 ei ω 3 ej 5 ei ωr er 3 ej 5 ei ωr εrjk ek 5 ωr εrik δik 5 2 εijr ωr
(1.172)
that is,
Ω 12 5 2 ω3 ; Ω 23 5 2 ω1 ; Ω 13 5 ω2 (1.173)
that is,
2 3
0 2 ω3 ω2
Ω ij 5 4 0 2 ω1 5 (1.175)
ant: 0
because of
ei ωk εikj ej 5 ei ei 3 ωk ek 5 ei ei 3 ω
2εijk ωk ei ej 5
2 ej 3 ek ωk ej 5 ek ωk 3 ej ej 5 ω 3 ej ej
_
If we put ei 5 0 in Eq. (1.179), we have the variation of ei observed
from the coordinate system with the base fei g leads to
where
_T _
Ω Q Q 5 er er ; _ _
Ω ij 5 ðer ei Þðer ej Þ (1.181)
Ω designates the spin of the base vector ei observed from the coordinate
system with the base fei g, bearing in mind that only the direction is
changeable in the base vector because of the unit vector. ei can be _
rewritten as
1.4.8 Determinant
The determinant is studied already in Section 1.1 but it will be stud-
ied in more detail in this subsection.
The determinant is defined in Eqs. (1.12)(1.14) already as follows:
8 9
T11 T12 T13 < εpqr T1p T2q T3r =
1
detT 5 jTij j 5 T21 T22 T23 5 or 5 εabc εpqr Tap Tbq Tcr
T31 T32 T33 : εpqr Tp1 Tq2 Tr3 ; 3!
(1.184)
Eq. (1.18) is expressed in the direct notation as follows:
ðdetTÞI 5 TðcofTÞT ; ðdetTÞδij 5 Tip ðcofTÞjp
ðcofTÞT ðcofTÞji
T21 5 ; Tij21 5 (1.186)
detT detT
by virtue of
8
>
> εijk ð2T1i Þð2T2j Þð2T3k Þ5 2εijk T1i T2j T3k
>
>
< εijk sT1i sT2j sT3k 5s εijk T1i T2j T3k
> 3
1 1 1
> det T5 εabc εpqr Tap Tbq Tcr 5 εpqr εabc Tpa Tqb Trc 5 εabc εpqr Tpa Tqb Trc
>
> 3! 3! 3!
>
>
: detðABÞ5ε ðA B ÞðA B ÞðA B Þ5ε A A A B B B 5A A A ε detB
pqr 1a ap 2b bq 3c cr pqr 1a 2b 3c ap bq cr 1a 2b 3c abc
(1.192)
expðT 11 Þ 0 0
detðexpTÞ5 0 expðT22 Þ 0 5expðT11 1T 22 1T 33 Þ (1.193)
0 0 expðT 33 Þ
detðabÞ5εijk ða1 bi Þða2 bj Þða3 bk Þ5a1 a2 a3 εijk bi bj bk 5ða1 a2 a3 Þðb3bÞ b (1.194)
The following equations hold for the cofactor, noting Eq. (1.185)
together with Eq. (1.191).
noting
1 1 1
trðcofTÞ 5 εabc εaqr Tbq Tcr 5 ðδbq δcr 2 δbr δcq ÞTbq T cr 5 ðTqq Trr 2 Trq T qp Þ
2 2 2
II is the second principal invariant as will be defined in Section 1.6.1.
The vector product in Eq. (1.38) and the scalar triple product in
Eq. (1.48) are described by the determinant as follows:
e1 e2 e3
a2 a3 a3 a1 a1 a2
a 3 b 5 e 1 e 1 e 5 a a2 a3 (1.197)
b2 b3 1 b3 b1 2 b1 b2 3 1
b1 b2 b3
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
a2 a3 2 a3 a1 2 a1 a2 2
jja 3 bjj 5 1
b3 b1 1 b1 b2
(1.198)
b2 b3
a1
a2
a3 a e1 a e2 a e3
½abc 5 b1 b2
b3 5 b e1 b e2 b e3 (1.199)
c1 c2
c 3 c e1 c e2 c e3
ða 3 bÞ ðc 3 dÞ 5
a c b c
a d b d (1.200)
a1
a2 a3 p1 q2 r3 ai pi ai qi ai ri a p a q a r
½abc½pqr5 b1
b2 b3 p1 q2 r3 5 bi pi bi qi bi ri 5 b p b q b r
c1 c2 c3 p1 q2 r3 ci pi ci qi ci ri c p c q c r
(1.201)
noting Eq. (1.189)4. The following equation is derived as the special case
of Eq. (1.201) for the three vectors v1 ; v2 ; v3 .
v1 v1 v1 v2 v1 v3
v2 5 v1 v2 v3 5
2
v2 v2 v2 v3
(1.202)
sym: v3 v3
where
v 5 v1 v2 v3 5 εijk ðv1 Þi ðv2 Þj ðv2 Þk ð 5 εijk ðv1 ei Þðv2 ej Þðv3 ek ÞÞ (1.203)
Fünftes Kapitel.
Einmarsch ins Innere. Die ersten
Eindrücke.
Massassi, 20. Juli 1906.