Numerical Solutions of Three Classes of Nonlinear Parabolic Integro Differential Equations 1St Edition Jangveladze Full Chapter

Numerical Solutions of Three Classes
of Nonlinear Parabolic
Integro-Differential Equations 1st
Edition Jangveladze
Visit to download the full and correct content document:
https://ebookmass.com/product/numerical-solutions-of-three-classes-of-nonlinear-par
abolic-integro-differential-equations-1st-edition-jangveladze/
Numerical Solutions of Three Classes of Nonlinear
Parabolic Integro-Differential Equations
Numerical Solutions of Three Classes
of Nonlinear Parabolic Integro-Differential
Equations
Temur Jangveladze
Ilia Vekua Institute of Applied Mathematics of
Ivane Javakhishvili Tbilisi State University
Tbilisi, Georgia
&
Georgian Technical University
Tbilisi, Georgia
Zurab Kiguradze
Ilia Vekua Institute of Applied Mathematics of
Ivane Javakhishvili Tbilisi State University
Tbilisi, Georgia
Beny Neta
Naval Postgraduate School
Department of Applied Mathematics
Monterey, CA, U.S.A.
Academic Press is an imprint of Elsevier
32 Jamestown Road, London NW1 7BY, UK
225 Wyman Street, Waltham, MA 02451, USA
525 B Street, Suite 1800, San Diego, CA 92101-4495, USA
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK
Copyright © 2016 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, further 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.
ISBN: 978-0-12-804628-9
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
For information on all Academic Press Publications

visit our website at http://store.elsevier.com/
PREFACE
This book is concerned with the numerical solutions of some

classes of nonlinear integro-di erential models. Some properties
of the solutions for investigated equations are also given. Three
types of nonlinear integro-di erential models are considered. Al-
gorithms of nding approximate solutions are constructed and
investigated. Results of numerical experiments with graphical
illustrations and their analysis are given. The book consists of
four Chapters.
In the rst Chapter three models (Model I, Model II and
Model III, that will be detailed further) to be discussed are
introduced and a brief history of integro-di erential equations
is given.
In the second Chapter, mathematical modeling of a process
of penetration of an electromagnetic eld into a substance by
integro-di erential models is described. A short description of
the integro-di erential equation that is a special model for one-
dimensional heat ow in materials with memory is also given
in the second Chapter. This model arises in the theory of one-
dimensional viscoelasticity as well. This Chapter closes with
some concluding remarks of the three investigated models.
The third Chapter is devoted to the numerical solution of
the initial-boundary value problems for models stated in the
previous Chapter. Semi-discrete schemes and nite-di erence
approximations, as well as nite elements are discussed. The
mathematical substantiation of all these questions for initial-
boundary value problems is given.
The questions of the realizations of algorithms investigated
in the third Chapter are discussed in the fourth Chapter. Re-
sults of the many numerical experiments with graphical illustra-
tions and their analysis are also given in this Chapter.
viii PREFACE
At the end of the book a list of the quoted literature and

indexes are given. The list of references is not intended to be
an exhaustive bibliography on the subject, but it is nevertheless
detailed enough to enable further independent work.
Each Chapter is concluded with a detailed section, entitled
"Comments and bibliographical notes," containing references to
the principal results treated, as well as information on important
topics related to, but sometimes not included in the body of the
text.
The authors believe that the book will be useful to scientists
working in the eld of nonlinear integro-di erential models. In
the opinion of the authors, the material presented in the book
is helpful for a wide range of readers engaged in mathematical
physics, in problems of applied and numerical mathematics, and
also MS and PhD students of the appropriate specializations.
Temur Jangveladze,
Zurab Kiguradze,
Beny Neta
ACKNOWLEDGMENTS
The rst author thanks Fulbright Visiting Scholar Program for

giving him the opportunity to visit U.S.A. and the Naval Post-
graduate School in Monterey, CA, U.S.A. for hosting him during
the nine months of his tenure in 2012-2013. The second author
thanks Shota Rustaveli National Scienti c Foundation of Re-
public of Georgia for giving him opportunity to visit U.S.A.
and the Naval Postgraduate School in Monterey, CA, U.S.A. for
hosting him during the four months of his tenure in 2013.
Abstract
This book is concerned with the numerical solutions of some

classes of nonlinear integro-di erential models. Some proper-
ties of the solutions of the corresponding initial-boundary value
problems studied in the monograph equations are given. Three
types of nonlinear integro-di erential models are considered. Al-
gorithms of nding approximate solutions are constructed and
investigated. Results of numerical experiments with tables and
graphical illustrations and their analysis are given. The book
consists of four chapters. At the end of the book a list of the
quoted literature and indexes are given. Each chapter is con-
cluded with a detailed section, entitled "Comments and biblio-
graphical notes," containing references to the principal results
treated, as well as information on important topics related to,
but sometimes not included in the body of the text.
Numerical Solutions of Three Classes of Nonlinear Parabolic Integro-Differential Equation.

http://dx.doi.org/10.1016/B978-0-12-804628-9.50008-9
© 2016 Elsevier Ltd. All rights reserved. 1
2 ABSTRACT
Key words: Electromagnetic eld penetration, Maxwell’s

equations, integro-di erential models, existence and uniqueness,
asymptotic behavior, semi-discrete and nite di erence schemes,
Galerkin’s method, nite element approximation, error estimate,
stability and convergence.
Chapter 1
Introduction
Abstract
The description of various kinds of integro-di erential equa-
tions and a brief history of their origin and applications are
given. The importance of investigations of integro-di erential
models is pointed out as well. Classi cation of integro-di eren-
tial equation is given. The main attention is paid on parabolic
type integro-di erential models. In particular, three types of
integro-di erential equations are considered. Two of them are
based on Maxwell’s equations describing electromagnetic eld
penetration into a substance. The third one is obtained by
simulation of heat ow. At the end of the chapter, as at the
end of each chapter, the comments and bibliographical notes is
given, which consists of description of references concerning to
the topic considered.
Key words: Electromagnetic eld penetration, Maxwell’s

system, heat ow equation, integro-di erential models.

http://dx.doi.org/10.1016/B978-0-12-804628-9.50001-6
© 2016 Elsevier Inc. All rights reserved. 3
4 CHAPTER 1. INTRODUCTION
In mathematical modeling of applied tasks di erential, inte-

gral, and integro-di erential (I-D, for short) equations appear
very often. There are numerous scienti c works devoted to the
investigation of di erential equations. There is a vast literature
in the eld of integral and integro-di erential models as well.
The di erential equations are connecting unknown functions,
their derivatives, and independent variables. On the other hand,
integral equations contain the unknown functions under an in-
tegral as well.
The term integro-di erential equation in the literature is
used in the case when the equation contains unknown function
together with its derivatives and when either unknown function,
or its derivatives, or both appear under an integral.
Let us recall the general classi cation of integro-di erential
equations. If the equation contains derivatives of unknown func-
tion of one variable then the integro-di erential equation is called
ordinary integro-di erential equation. The order of an equation
is the same as the highest-order derivative of the unknown func-
tion in the equation.
The integro-di erential equations often encountered in math-
ematics and physics contain derivatives of various variables;
therefore, these equations are called integro-di erential equa-
tions with partial derivatives or partial integro-di erential equa-
tions.
In the applications very often there are integro-di erential
equations with partial derivatives and multiple integrals as well,
for example, Boltzmann equation [66] and Kolmogorov-Feller
equation [288].
Volterra is one of the founders of the theory of integral and
integro-di erential equations. His works, especially in the inte-
gral and integro-di erential equations, are often cited till today.
The classical book by Volterra [469] is widely quoted in the liter-
ature. In 1884 Volterra [465] began his research in the theory of
CHAPTER 1. INTRODUCTION 5
integral equations devoted to distribution of an electrical charge

on a spherical patch. This work led to the equation, which in
the modern literature is called the integral equation of the rst
kind with symmetric kernel.
The work on the theory of elasticity became the beginning
research of Volterra leading to the theory of partial integro-
di erential equations. In 1909 Volterra [466] has studied a par-
ticular type of such equations and has shown that this integro-
di erential equation is equivalent to a system consisting of three
linear integral equations and a second order partial di erential
equations.
The rst examples of integro-di erential equations with par-
tial derivatives investigated in the beginning of the twentieth
century were in Schlesinger’s works [417], [418], where the fol-
lowing equation is investigated:
Z b
@U (x; y)
= f (x; y; s)U (x; s)ds:
@x a
Numerous works in the beginning of the twentieth century
were devoted to research of integro-di erential equations of vari-
ous kinds. The excellent bibliography in this case is given in the
classical book by Volterra [469]. In addition, Kerimov [271], the
editor of the Russian translation of this book, has updated (up
to 1970s) the list of references on integral and integro-di erential
equations.
Let us describe some classes of mathematical models of sec-
ond order promoting intensive research on partial integro-di er-
ential equations.
When we take into account hereditary phenomena, the ques-
tions of physics and mechanics lead to integro-di erential equa-
tions. A hereditary phenomenon occurs in a system when the
phenomenon does not depend only on the actual state of the
system but on all the preceding states through which the sys-
tem has passed; that is to say, it depends on the history of the

system and may therefore be called hereditary.
One of the important representatives of an integro-di er-
ential equations of elliptic type is the following equation con-
nected with the hereditary phenomenon [467]
Z tX3
@ 2 U (x; )
U (x; t) + fi (t; )d = 0;
0 i=1 @x2i
where x = (x1 ; x2 ; x3 ); is a classical three-dimensional Laplace

operator
@ 2 U (x; t) @ 2 U (x; t) @ 2 U (x; t)

U (x; t) = + + ;
@x21 @x22 @x23
and fi are known functions of their arguments. Let us note that
here and below everywhere instead of x; y; z we use x1 ; x2 ; x3 for
the designation of Cartesian coordinates.
The mathematical modeling of the vibrations of an elastic
chord in the case of linear hereditary process gives a hyperbolic
type integro-di erential equation [467]
t
@ 2 U (x; t) @ 2 U (x; t) @ 2 U (x; )
Z
= + (t; )d ; (1.1)
@t2 @x2 0 @x2
where is a known function of its arguments.

One of the most important representatives of integro-di er-
ential models is the following nonlinear equation describing string
vibration obtained by Kirchho [286] in 1876
" #
2
@ 2 U (x; t) @ 2 U (x; t)
Z
2 @U (x; t)
+ dx = 0; (1.2)
@t2 0 @x @x2
where is a known constant. Many authors investigated equa-

tion (1.2) and its natural generalizations:
"Z #
2
@ 2 U (x; t) @U (x; t) @ 2 U (x; t)
a dx =0 (1.3)
@t2 0 @x @x2
and
@ 2 U (x; t) 1
a A 2 U (t) AU (t) = 0; (1.4)
@t2
where a(S) a0 = const > 0 is a known function of its argu-
ment and A is a self-adjoint positive operator, i.e., A = A > 0.
The norm used in (1.4) is the one de ned on the range of the
operator A.
In investigating (1.3) and (1.4) type models it is su cient
to mention the following publications: [9], [24], [36], [37], [55],
[61], [65], [130], [309], [327], [328], [349], [355], [356], [357], [373],
[380], [383], [390], [394], [395] though this list is not complete.
Let us note that equations (1.3) and (1.4) are also called Kirch-
ho equations. They, along with some similar equations, de-
scribe important physical processes, among which are linear and
nonlinear dynamics of di erent dimensional bodies (see, for ex-
ample, [36], [355], [373] and [471]).
In other questions connected with hereditary phenomena,
one nds the integro-di erential equations of a parabolic type,
which were investigated by Evans [160]. These equations look
like
Z t 2
@U (x; t) @ 2 U (x; t) @ U (x; )
2
+ A(t; )d = 0; (1.5)
@t @x t0 @x2
where A is a known function of its arguments.

Integro-di erential equations of parabolic type arise in the

study of various problems in physics, chemistry, technology,
economics, etc. One very important problem of applied type
is generated by mathematical modeling of processes of electro-
magnetic eld penetration into a substance and is described by
the well-known Maxwell’s equations [300]. In the works [187]
and [188], complex system corresponding to nonlinear partial-
di erential equations was reduced to integro-di erential form. If
the coe cient of thermal heat capacity and electroconductivity
of the substance are highly dependent on temperature, then the
Maxwell’s system can be rewritten in the following form (see
[187] and [188]):
@W (x; t)
+r [a(S(x; t))r W (x; t)] = 0;
@t (1.6)
r W (x; t) = 0;
where Z t
S(x; t) = jr W (x; )j2 d : (1.7)
0
In system (1.6), r W and r W are the usual vector oper-
ators with respect to the variables x = (x1 ; x2 ; x3 ): Even one-
dimensional scalar version of this model is very complicated and
its investigation has been possible yet only for special cases. The
one-dimensional scalar case of the model (1.6), (1.7) has the fol-
lowing form
" Z ! #
t 2
@U (x; t) @ @U (x; ) @U (x; t)
= a d ; (1.8)
@t @x 0 @x @x
where a(S) a0 = const > 0 is again a known function of its

argument. Investigation of (1.6), (1.7), and (1.8) type models
began in the works [138], [187], and [188]. Since then many
scienti c publications were devoted to the investigation of exis-
tence and uniqueness of their solutions under various types of
initial and boundary conditions. In this respect, especially sig-
ni cant are the works [49], [50], [141], [146], [147], [219], [220],
[238], [254], [256], [302], [303], [304], [305], [306], [307], [322],
[333], [334], and reference therein. Authors of this book have
also made contribution in this direction, for example, see [137],
[138], [139], [140], [141], [145], [146], [147], [186], [187], [188],
[219], [220], [223], [238], [247], [248], [253], [254], [256], [261],
and reference therein.
Making certain physical assumptions in mathematical de-
scription of the above-mentioned process of penetration of elec-
tromagnetic eld into a substance, Laptev [306] has constructed
a new integro-di erential model, which represents a general-
ization of the system introduced in [187] and [188]. Founded
on Maxwell’s system the following parabolic integro-di erential
model is obtained
Z tZ
@W (x; t)
=a jr W (x; )j2 dxd W (x; t): (1.9)
@t 0
In the works [303], [305], and [306] for conditionally closed

operators, an operator scheme is constructed. This scheme is
used for (1.6), (1.7) type models to prove existence and unique-
ness of solution of initial-boundary value problems. In the above-
mentioned work [306] Laptev points out that for these so-called
averaged integro-di erential models (1.9), it is necessary to de-
velop a di erent, special approach.
In systems (1.6), (1.7), and (1.9), W = (W1 ; W2 ; W3 ) denotes
a vector, which is connected by a vector of a magnetic eld H =
(H1 ; H2 ; H3 ) and is the function of the variables (x1 ; x2 ; x3 ; t);
which we will shorten to (x; t).
The multidimensional scalar analogues of systems (1.6), (1.7),

and (1.9) have the following forms:
Z t
@U (x; t)
=r a jrU (x; )j2 d rU (x; t) (1.10)
@t 0
and
Z tZ
@U (x; t)
=a jrU (x; )j2 dxd U (x; t); (1.11)
@t 0
respectively.
In equations (1.10) and (1.11) we have x = (x1 ; : : : ; xn ) and
the vector operator rU is given by
@U (x; t) @U (x; t)
rU = gradU = ;:::; = (D1 U; : : : ; Dn U ) :
@x1 @xn
Some generalizations of the models (1.6)-(1.11) are given in
the works [137], [139], [145], [219], [305], and [306]. One kind of
these models has the forms:
n
@U (x; t) X
Di a(S(x; t)) jrU (x; t)jq 2
Di U (x; t)
@t i=1 (1.12)
= f (x; t)
and
n
@U (x; t) X
a(S(t)) Di jrU (x; t)jq 2
Di U (x; t)
@t i=1 (1.13)
= f (x; t);
where Z t
S(x; t) = jrU (x; )jq d
0
and Z tZ
S(t) = jrU (x; )jq d :
0
The models of type (1.6)-(1.13) are complex and have been
intensively studied by many authors. The existence and unique-
ness of global solutions of initial-boundary value problems for
equations and systems of type (1.6) - (1.13) were studied in
[49], [50], [137], [138], [139], [140], [141], [145], [186], [187],[188],
[219], [223], [238], [247], [248], [253], [261], [302], [303], [304],
[305], [306], [307], [322], [333], [334], and in a number of other
works as well.
The existence theorems that are proved in [137], [138], [139],
[140], [141], [145], [187], [188], and [219] are based on a-priori
estimates, modi ed Galerkin’s method and compactness argu-
ments as in [327], [328], [461], and [462] for nonlinear elliptic
and parabolic equations.
For equation (1.8) with nonhomogeneous right-hand side and
a(S) = 1 + S, or for equation (1.13) in the one-dimensional
case with q = 2, such theorem for rst initial-boundary value
problem is proved in section 3.6.
The asymptotic behavior as t ! 1 of the solutions of such
models has been the object of intensive research in recent years,
see [28], [29], [31], [32], [33], [34], [35], [139], [141], [145], [146],
[147], [217], [218], [219], [220], [223], [227], [228], [229], [231],
[232], [233], [234], [235], [236], [237], [238], [239], [240], [241],
[242], [243], [245], [246], [247], [248], [251], [252], [253], [254],
[256], [257], [261], [276], [277], [278], [280], [282], [283], and ref-
erence therein.
Another model considered in this book and studied by one
of the authors of this monograph [376] is
Z t
@U (x; t) @ @U (x; )
= a(t ) d + f (x; t): (1.14)
@t 0 @x @x
This equation serves as a very special model for one-dimens-

ional heat ow in material with memory [196], [345]. It also
arises in the theory of viscoelasticity [109], [344], [346]. It is an
example in the general theory of equations of the form
Z t
U_ (t) + a(t )g(U ( ))d = f (t) (1.15)
0
on a Hilbert space H with g a nonlinear bounded operator [51].

The dot above U means time derivative.
MacCamy [345] and later Sta ans [438] have discussed the
existence, uniqueness, boundedness and asymptotic behavior of
solutions of initial-boundary value problems for (1.14).
Many scienti c publications are dedicated to the investiga-
tion of characteristics of (1.14) and (1.15) type equations (see,
for example, [67], [70], [80], [110], [111], [162], [206], [433], and
reference therein).
Numerous scienti c works are devoted to construction of al-
gorithms for the numerical solution of initial-boundary value
problems for the above-stated models (see, for example, [20],
[30], [32], [35], [80], [96], [125], [139], [145], [148], [220], [223],
[227], [239], [241], [243], [244], [247], [248], [249], [250], [251],
[252], [253], [255], [256], [257], [258], [261], [274], [276], [277],
[279], [280], [281], [282], [283], [284], [318], [333], [376], [378],
[423], [424], and [501]).
The purpose of the present monograph is the continuation
of study and uniform description of results developed for the
numerical solution of integro-di erential models (1.6)-(1.14) and
their generalizations. It should be noted that main attention is
paid on theorems with proofs for corresponding one-dimensional

equations.
The book consists of four chapters. In the next chapter
mathematical modeling of a process of penetration of an elec-
tromagnetic eld into a substance by integro-di erential models
and the conclusion of investigated models such as (1.6)-(1.13)
is brie y given. We also give the mathematical model for heat
conduction in materials with memory (1.14). Some features of
this model are also given.
The third chapter is devoted to the numerical solution of
the initial-boundary value problems for the models stated in
the previous chapter. The semi-discrete schemes, nite di er-
ence approximations, Galerkin’s method schemes, and nite el-
ements are discussed. The mathematical substantiation of all
these questions for various types of initial-boundary value prob-
lems is given.
Results of the various numerical experiments with graphical
illustrations and their analyses are given in the fourth chapter.
At the end of the book a list of the quoted literature and
indexes are given.
The list of references is not intended to be an exhaustive bib-
liography on the subject, but it is nevertheless detailed enough
to enable further independent work.
Each chapter is concluded with a detailed section, entitled
"Comments and bibliographical notes," containing references to
the principal results treated, as well as information on important
topics related to, but sometimes not included in the body of the
text.
The authors believe that the book will be useful to the ex-
perts working in the eld of nonlinear integro-di erential mod-
els. In the opinion of the authors, the material presented in the
book is helpful for a wide range of readers engaged in mathemat-
ical physics, in problems of applied and numerical mathematics,
and also MS and PhD students of the appropriate specializa-

tions.
1.1 Comments and bibliographical

notes
Mathematical models of many natural phenomena and processes
can be described by the initial-boundary value problems posed
for nonstationary partial di erential and integro-di erential equa-
tions and systems of such equations. Investigation and numer-
ical solution of these problems are the actual sphere of mathe-
matical physics and numerical analysis. One such partial integro-
di erential model describes the process of electromagnetic eld
penetration into a substance. In the quasi-stationary approxi-
mation, the corresponding system of Maxwell’s partial di eren-
tial equations can be rewritten in integro-di erential form (1.6),
(1.7) (see [187] and [188]).
Mathematical models describing electromagnetic processes
and many relative phenomena are given in many scienti c pa-
pers and books, see, for example, [92], [136], [300], [359], [360],
[399], [419], [473], and references therein.
It is well known that electromagnetic eld di usion pro-
cesses and many other important practical processes are simu-
lated by Maxwell’s systems of partial di erential equations and
Maxwell’s-type systems as well (see, for example, [195], [197],
[290], [297], [298], [300], [360], [426], [485], [486], [487], [489],
[491], and references therein).
Many scienti c works are devoted to investigation of various
problems for Maxwell’s and Maxwell’s-type systems, see [4], [5],
[56], [60], [101], [106], [120], [121], [124], [131], [139], [144], [145],
[152], [158], [194], [195], [197], [208], [224], [225], [226], [247],
[253], [275], [276], [290], [362], [365], [367], [397], [409], [427],
1.1. COMMNETS AND BIBLIOGRAPHICAL NOTES 15
[452], [453], [454], [470], [475], [476], [477], [485], [486], [487],
[489], [491], and references therein.
As we have already pointed out, by using Maxwell’s system
[300] for mathematical simulation of the process of electromag-
netic eld penetrating into a substance, new class of integro-
di erential models (1.6), (1.7), (1.8), and (1.10) arises (see [187]
and [188]).
Integro-di erential equations arise in many other practical
processes as well, see, for example, [11], [21], [22], [36], [37],
[55], [59], [74], [84], [119], [127], [128], [164], [185], [187], [188],
[295], [298], [306], [310], [311], [312], [355], [356], [357], [363],
[373], [380], [384], [405], [425], [459], [463], [465], [466], [467],
[468], [469], [471], [484], and in a number of other works as well.
The motivation for studying integro-di erential problems
comes from the many physical models in such elds as elec-
tromagnetic wave propagation, heat transfer, nuclear reactor
dynamics, and thermoelasticity. Besides the integro-di erential
equations arise in many spheres of human activity as well. For
example, the second order fully nonlinear integro-di erential
equations are derived from the pricing problem of nancial deriva-
tives and optimal portfolio selection problem in a market [59].
In [84] nonlinear integro-di erential equations that arise from
stochastic control problems with purely jump Levy processes
are considered.
Many problems of modern science and engineering can be
described by partial integro-di erential equations. Since quite
a lot of these problems are time-dependent, most of them are
evolution equations and especially nonlinear evolution parabolic
equations, see [93], [177], [179], [180], [191], [216], [310], [311],
[312], [322], [332], [345], [346], [358], [451], and references therein).
Many scienti c works are devoted to investigation and nu-
merical solution of parabolic integro-di erential models, see [6],
[18], [45], [53], [62], [63], [69], [74], [83], [86], [87], [93], [98], [103],
[108], [115], [118], [128], [137], [139], [145], [155], [156], [163],
[175], [177], [181], [182], [191], [198], [247], [253], [263], [265],
[266], [267], [276], [279], [281], [306], [308], [310], [311], [312],
[313], [319], [320], [322], [323], [327], [353], [358], [370], [376],
[378], [389], [393], [427], [430], [432], [433], [447], [448], [449],
[450], [451], [456], [479], [483], [488], [492], [497], and references
therein.
Study of the models of type (1.6), (1.7), and (1.8) has begun
in the works [138], [187] and [188]. In these works, in particu-
lar, the theorems of existence of solution of the initial-boundary
value problem (with rst (Dirichlet) boundary conditions) for
scalar equation with one-dimensional space variable are proved.
Investigations of higher space dimensions for model (1.10) car-
ried out initially in [137] and [140].
In [306] some generalization of the system of type (1.6), (1.7)
was proposed. In particular, assuming that the temperature
of the considered body depends on the time variable, but in-
dependent of the space coordinates, then the same process of
penetration of a magnetic eld into the material is simulated
by the averaged (as the author of [306] has named it) integro-
di erential models (1.9) and (1.11).
Study of the models of type (1.9) and (1.11) has started in
the works [217] and [219].
One must note that some works were devoted to the study
of modeling of physical process of electromagnetic eld pene-
tration in the case of cylindrical conductors. In this case, the
above-mentioned integro-di erential model (1.6), (1.7), written
in cylindrical coordinates, was given in [148]. The work [333]
is devoted to the investigation of periodic problem for one-
dimensional (1.8) type model in cylindrical coordinates.
Interest in the above-mentioned integro-di erential (1.6) -
(1.11) models is increasing. Some generalizations of (1.10) and
(1.11) models, which have the forms (1.12) and (1.13) corre-
1.1. COMMNETS AND BIBLIOGRAPHICAL NOTES 17
spondingly, are given and studied in [219], [305], and in a num-

ber of other works as well. Equation (1.12) is investigated, for
example, in [145], [219], [305], and [306]. Equation (1.13) was
investigated in [145] and [219]. In the scienti c literature some
more general models have also appeared, see [137], [139], [145],
[219], [305], [306], and [322].
Many di erent kinds of initial-boundary value problems with
a variety of boundary and initial conditions are considered for
the above-mentioned integro-di erential equations. In the works
[246], [280], [281] investigation and numerical approximation of
problems with mixed boundary conditions, for (1.10) and (1.11)
type one-dimensional scalar models, are studied.
Let us also note that rst kind initial-boundary value prob-
lems with nonhomogeneous boundary condition on one side of
lateral boundary are also considered and studied in many works.
This type of the problem statement is dictated by mathematical
simulation of the physical processes, see, for example, [148] and
problem (2.64), (2.65) given in "Comments and bibliographical
notes" section in Chapter 2.
The theorems proved for investigating the asymptotic behav-
ior as time tends to in nity in some cases show the di erence be-
tween stabilization character of solutions with homogeneous and
nonhomogeneous boundary conditions of the rst kind initial-
boundary value problems. More precisely, in homogeneous case
stabilization has an exponential character, whereas in nonho-
mogeneous case it has power-like form.
The works [49], [50] are also worth mentioning, where in-
vestigation of inverse problems for multidimensional models of
(1.10) type is carried out.
Another integro-di erential model studied in this monograph
is (1.14). This model describes heat ow in material with mem-
ory [196], [345]. It also arises in the theory of viscoelasticity
[109], [344], and [346].
As a rule we cannot nd exact solutions of the considered

nonlinear di erential and integro-di erential models. Therefore,
particular attention should be paid to construction of numerical
solutions and to their importance for integro-di erential models.
The rst steps in this direction, for the models studied in this
monograph, are made in the works [139], [220], [274], [376], and
[378]. Now the research in this direction has intensi ed.
Let us note that the models being considered in this mono-
graph have arisen from practical tasks. But they can be con-
sidered as models, generalizing known nonlinear parabolic equa-
tions, which are studied in many known scienti c papers, books,
and monographs, see [76], [133], [134], [169], [199], [298], [327],
[332], [381], [385], [461], and references therein.
Many scienti c researches are devoted to numerical solution
of partial di erential and nonlinear parabolic equations as well,
see [23], [43], [104], [167], [171], [183], [193], [210], [213], [225],
[226], [301], [327], [400], [404], [407], [411], [415], [435], [439],
[440], [447], and the references listed in these papers and books.
As we have already mentioned the main part of integro-
di erential structures considered here has arisen from Maxwell’s
systems of partial di erential equations. There are many scien-
ti c papers and books on the numerical solution of Maxwell’s
systems and Maxwell’s-type systems as well, see [4], [5], [15],
[26], [27], [54], [68], [139], [142], [145], [153], [212], [224], [247],
[253], [260], [276], [314], [315], [316], [317], [336], [341], [369],
[434], [442], [472], [480], [493], [498], and references therein.
The results of these researches very often are used in construc-
tion and investigation of the numerical algorithms for the cor-
responding integro-di erential models.
The detailed description of investigation and numerical so-
lution of the above-mentioned models is given in Chapters 2-4.
More complete references and comments are given in a section
entitled "Comments and bibliographical notes" in each chapter.
Chapter 2
Mathematical Modeling
Abstract
The chapter consists of six sections. The chapter concerns
mathematical modeling of the investigated in the monograph
equations. Some mathematical features of these models are
studies as well. In the rst section general statement of di usion
process is given. Sections two and three are dedicated to a re-
duction of Maxwell’s equations to the integro-di erential mod-
els. Consequently, two types of integro-di erential equations
are obtained which are called Model I and Model II accordingly.
Both models in di erent physical assumptions describe process
of penetration of an electromagnetic eld into a substance. In
the third section Model III is considered, which represents a
special model for one-dimensional heat ow in materials with
memory. This model arises in the theory of one-dimensional
viscoelasticity as well. Next two sections are devoted to some
mathematical features of all three models above. The existence
and uniqueness properties of the solutions as well as asymp-
totic behavior of solutions of the appropriate initial-boundary
value problems are presented. This chapter closes with some
concluding remarks and bibliographical overview of the three

http://dx.doi.org/10.1016/B978-0-12-804628-9.50002-8
© 2016 Elsevier Inc. All rights reserved. 19
20 CHAPTER 2. MATHEMATICAL MODELING
investigated models.
Key words: Mathematical modeling, electromagnetic eld

penetration, Maxwell’s equations, heat ow in materials with
memory, one-dimensional viscoelasticity, integro-di erential mod-
els, existence and uniqueness, asymptotic behavior.
2.1 Electromagnetic di usion process

2.1.1 General statement of di usion process
Let us consider the phenomena which occur in a substance
placed in an external variable electromagnetic eld.
Suppose the electromagnetic eld and the currents satisfy
quasi-stationary conditions, see, e.g., [300] and [419]. Maxwell’s
system of equations is given by:
1 @( H)
= r E; (2.1)
c @t
r ( H) = 0; (2.2)
4
E = r H; (2.3)
c
where E = (E1 ; E2 ; E3 ) and H = (H1 ; H2 ; H3 ) are vectors of
electrical and magnetic elds, respectively, is the magnetic
permeability, is the conductivity of substance, and c is the
speed of light in vacuum.
In (2.3), following an assumption of quasi-stationarity, the
term proportional to @E=@t is omitted and Ohm’s law is used,
relating the vector E to a vector of density of a current J by
the following relation
J = E: (2.4)
2.1. ELECTROMAGNETIC DIFFUSION PROCESS 21
If at some moment of time the magnetic eld is solenoidal,

i.e., r ( H) = 0, then (2.2) is always satis ed. From (2.1) we
have
@
r ( H) = 0:
@t
We shall assume that the substance is isotropic with = 1
and the coe cient of electroconductivity depends on tempera-
ture, i.e., = ( ). In applications, the form of is power-
1
like, for example, for metals ; for homogeneous plasma
3=2
, and so on.
For de nition of temperature it is necessary to use the equa-
tion of balance of heat. Let us make the assumption that the
characteristic time of resistive di usion is much less than that of
heat transfer. Then, neglecting the e ect of heat conductivity,
the change in temperature is de ned only by Joule’s heating and
taking into account (2.4), we have
@
cv = EJ = E 2 ; (2.5)
@t
where cv is the speci c heat capacity of the environment. Thus
the factor of heat capacity can also depend on temperature (as
a rule, in a power-like form).
Equations (2.1), (2.3), (2.5) form the closed system for def-
inition of an electromagnetic eld and temperature under an
appropriate initial and boundary conditions.
The questions of existence and uniqueness of the solutions of
linear di erential problems ( = (x)), in general enough state-
ment, are considered in [136], [297], and in a number of other
works. Thus, in [297] the movement of environment is supposed
within the framework of magnetic-hydrodynamical approach. In
these works the transition from classical statement to the gener-
alized one is made. The requirement that the functions satisfy
equation (2.2) and boundary conditions is replaced with the re-
quirement of their belonging to special functional spaces. The

problem for equations (2.1), (2.3) is formulated in the variational
form, in particular, in terms of variational inequalities.
The study of system (2.1)-(2.3) can be made also on the
basis of the equations which have been written down only for
vectors E or H. As an example of similar research, see [504], in
which the nonlinearity caused by dependence of the magnetic
permeability on H is entered.
2.1.2 A reduction of system of nonlinear dif-

ferential equations to the integro-di er-
ential model (Model I)
We assume that a massive body is placed in a variable magnetic
eld. It is necessary to describe the distribution of the eld
inside the body.
Inserting the resistance = 1= in (2.4) we shall express
E = J and substitute in the penultimate equation (2.3), we
have
c
E= r H:
4
Substituting this expression in (2.1), we receive
@H c2
+ r ( r H) = 0: (2.6)
@t 4
A variable magnetic eld penetrating into a substance in-
duces a variable electrical eld, which causes occurrence of cur-
rents. The currents result in heating of the substance and in-
crease of its temperature, which in uences its resistance. From
reasonings given in [399], it follows that one order change of
temperature results in change of resistance on several orders,
so at large uctuations of temperature the dependence = ( )
must be taken into account. Last essential restriction, which
will be made, is connected to the assumption that the change in

body temperature under action of a current J submits to Joule’s
law
@
0 cv = J 2: (2.7)
@t
Here is density of the substance, 0 is the density at time t = 0,
and cv (as was already mentioned in 2.1.1) is its speci c heat
capacity. Generally they also depend on temperature. Equation
(2.7) does not take into account transfer of heat due to heat
conductivity and radiation. Number of other physical e ects
are not considered. However, from the mathematical point of
view system (2.6), (2.7) is complex enough.
Now we shall begin the reduction of system (2.6), (2.7) to a
system of nonlinear integro-di erential equations.
Let us rewrite equation (2.7) in the following form
cv @
0( ) = J 2:
( ) @t
Introducing the function

Z
cv
S( ) = 0( ) d ;
0
( )
we have
@S
= J 2:
@t
Let us assume that the process begins at the moment of time
t = 0 in which the temperature 0 of the substance is constant.
Integrating this equation on a segment [0; t]; we shall nd
Z t
S( (x; t)) S( 0 ) = J 2d :
0
The functions 0 ; , and cv are positive in physical sense;

therefore, the function S( ) is monotonically increasing. From
here follows that the inverse function = (S) is uniquely de-
termined, connected with S( ) by parity (S( )) = : Thus, it
is possible to write down
Z t
(x; t) = J 2d :
0
From (2.3), (2.4) we have
c
J= r H;
4
so, Z t
c 2
(x; t) = r H d :
0 4
Substituting this expression in (2.6) as argument of function
= ( ), we receive
t
c2
Z
@H c 2
+ r r H d r H = 0;
@t 4 0 4
r H = 0:
Let us introduce the notations
c2 c
a(S) = ( (S)); W = H
4 4
and rewrite the system in the following form:
Z t
@W
+r a jr W j2 d r W = 0;
@t 0 (2.8)
r W = 0:
If we consider a at eld of the form W = (0; 0; U ), where

the function U = U (x1 ; x2 ; t) depends on two spatial variables,
then
@U @U
r W = ; ;0
@x2 @x1
and (2.8) takes the form
Z t
@U
=r a jrU j2 d rU : (2.9)
@t 0
Let us note again that equations such as (2.8), (2.9) for the
rst time have arisen in [187] and [188]. In these works, as well
as in [137], [138], [139], [140], [141], [145], [186], [219], [223],
[238], [247], [253], and [261] together with other questions, the
uniqueness of the solutions of the initial-boundary value prob-
lems for equations (2.8) and (2.9) is given, under a rather general
assumptions on the function a = a(S).
Assume that 0 and cv are constants. Let us give examples
of function ( ), inducing a(S), see, e.g., [145], [304], and [306].
If ( ) = , > 1, then
a(S) = C1 (C0 S) 1;
where C0 and C1 are some positive constants. Thus, the power
growth of the resistance ( ) gives a coe cient determined only
on a nite interval. Let us notice that the physical substances
do not have such property.
If ( ) = , then
a(S) = C1 eC0 S ;
with positive constants C0 and C1 , i.e., the linear growth of the
function ( ) gives exponential function a(S). The linear growth
of resistance with temperature is characteristic of metals.

If ( ) = , 0 < < 1, then
a(S) = C1 (C0 + S) 1 ;
with positive constants C0 and C1 . Thus, under-linear growth
of the function ( ) gives sedate growth of the coe cient a(S).
If ( ) = , < 0, then
a(S) = C1 (C0 + S) 1 + > 0;

with positive constants C0 and C1 , i.e., a decreasing function
( ) gives decreasing functions a(S). This is a general rule in
case 0 and cv are constants. Suppose that function ( ) is
di erentiable for 0 . Then according to the de nition we
have
da c2 d d c2 d =d c2
= = = ( ) 0 ( ):
dS 4 d dS 4 dS=d 4 cv
From here it is obvious that the function a(S) decreases or

grows simultaneously with the function ( ). Let us note that
the decrease of resistance ( ) with growth of temperature is
characteristic of semiconductors in solid, gaseous, and plasma
phases, for which, as it was already remarked in 2.1.1, the the-
oretical formula ( ) = K 3=2 is valid.
2.2 On the averaged Model II

We shall begin now a derivation of the averaged equation [306]
describing again the process of penetration of an electromagnetic
eld into a substance.
2.2. ON THE AVERAGED MODEL II 27
Following [419] and as in section 2.1, we shall consider the

system of the Maxwell’s equations, describing the interaction of
an electromagnetic eld into a substance:
1 @H
= r E; (2.10)
c @t
r H = 0; (2.11)
4
J = r H; (2.12)
c
J = E: (2.13)
Joule’s law (2.7) is a localization of the law of allocation of
heat [419]
Z
dQ = JEdxdt: (2.14)
Here dQ is in ow of a thermal energy absorbed by a body in an

electromagnetic eld in time dt on a substance, occupying area
R3 . Joule’s law does not take into account the process of
transfer of heat inside a body, which is fair if one to consider
the temperature time-dependent but independent of the spatial
coordinates. Thus, in this case it is possible to write that =
(t) and dQ = mcv d , where m is the weight of the substance.
Therefore, (2.14) becomes
Z
d
mcv ( ) = JEdx:
dt
Using Ohm’s law E = ( )J and repeating the process of
exception, we shall receive the following analogue of system (2.8)
in the same designations:
Z tZ
@W 1
+r a jr W j2 dxd r W = 0;
@t j j 0
r W = 0:
Here j j is the volume of the substance. Due to averaging the
coe cient of this equation depends only on the variable t, and
consequently the equation can be rewritten in the form
Z tZ
@W
=a jr W j2 dxd W: (2.15)
@t 0
Here we used the known vector identity
r r W = W + r(r W ):
For a at eld W = (0; 0; U ), where U = U (x1 ; x2 ; t) is a
function of two spatial variables, system (2.15) becomes
Z tZ
@U
=a jrU j2 dxd U: (2.16)
@t 0
Laptev [306] remarked that the research of the so-called aver-

aged integro-di erential models (2.15) and (2.16) requires inde-
pendent theory than he applied for investigation of (2.8), (2.9),
and such type equations.
2.3 Mathematical Model III

The classical linear theory of heat conduction for homogeneous
and isotropic media is given by
@U @ 2U
=k 2;
@t @x
where U (x; t) is the absolute temperature and k > 0 is a con-
stant called thermal di usivity. This parabolic equation allows
a thermal disturbance at any point in the medium to be in-
stantly felt at other points, see [196]. Gurtin and Pipkin [196]
2.3. MATHEMATICAL MODEL III 29
have suggested a heat ow model based on a memory e ect in

the medium. The linear one-dimensional version of their the-
ory assumes that the internal energy, (x; t); and the heat ux,
q(x; t), satisfy the following relations:
Z 1
(x; t) = bU (x; t) + B( )U (x; t )d ; (2.17)
0
Z 1
@U (x; t )
q(x; t) = c( ) d ; (2.18)
0 @x
where b is the instantaneous speci c heat, B is an energy re-
laxation function, and c is the thermal conductivity (see [345]).
Thus is a functional of the history of the temperature and q
is a functional of the history of the temperature gradient. Mac-
Camy [345] treated a partially nonlinear version of the above
model by replacing (2.18) by
Z 1
@U (x; t )
q(x; t) = K( ) d : (2.19)
0 @x
Now assume that the material is at zero temperature and in-
ternal energy up to time t = 0. This is not restrictive since
we can incorporate nonzero initial values with the forcing term.
Incorporating (2.17), (2.19) with the balance of heat equation,
we have
Z t
@U (x; t) @U (x; )
b + B(t ) d =
@t 0 @
Z t (2.20)
@ @U (x; )
K(t ) d + r(x; t);
0 @x @x
where r(x; t) is the heat source. Let R(x; t) be the resolvant for
B, i.e., the function U given by
Z t
1
U (x; t) = (x; t) + R(t ) (x; )d
b 0
solves the equation

Z t
(x; t) = bU (x; t) + B(t )U (x; )d :
0
Now de ne a(t) and f (x; t) by

Z t
1
a(t) = K(t) + R(t )K( )d ;
b 0
Z t
1
f (x; t) = r(x; t) + R(t )r(x; )d ;
b 0
then (2.20) becomes

Z t
@U (x; t) @ @U (x; )
= a(t ) d + f (x; t); (2.21)
@t 0 @x @x
which is (1.14).
2.4 Some features of Models I and II

2.4.1 Existence and uniqueness of the
solutions
In the present section we give some features of those types of
equations considered in sections 2.1 and 2.2.
As models we shall consider the following two nonlinear par-
tial integro-di erential equations:
" Z ! #
t 2
@U @ @U @U
a d = f (x; t) (2.22)
@t @x 0 @x @x
and
2.4. SOME FEATURES OF MODELS I AND II 31
!
Z tZ 1 2
@U @U @ 2U
a dxd = f (x; t): (2.23)
@t 0 0 @x @x2
These equations are one-dimensional one-component case of

the equations arising at mathematical modeling of process of
penetration of an electromagnetic eld into a substance with
temperature-dependent coe cient of electroconductivity. As we
noted in section 2.1, equation (2.22) given here was o ered in
[187] and [188] at rst and then was generalized in numerous
works [49], [50], [137], [139], [141], [145], [148], [219], [303], [304],
[305], [306], [333], and [334]. Equation (2.23) describes the same
process as (2.22) and is rst given in [306].
Under certain conditions on the initial data for these equa-
tions the existence and uniqueness of the initial-boundary value
problems with rst kind boundary conditions are given.
For investigations of initial-boundary value problems dis-
cussed in the book, we use usual Sobolev spaces Wpk ( ), W kp ( ),
Lp ( ), C k ( ), Lq 0; T ; Wpk ( ) , Lq 0; T ; W kp ( ) and their
properties, see, for example, [7], [136], [175], [327], and refer-
ences therein. Some description of these spaces will be given in
section 3.6 as well.
In the domain QT = (0; T ) of the variables (x; t); where
= (0; 1) and T is a xed positive constant for equations (2.22)
and (2.23) we shall consider the following initial-boundary value
problem:
U (0; t) = U (1; t) = 0; t 2 [0; T ]; (2.24)

U (x; 0) = 0; x 2 : (2.25)
We shall search the solution of problem (2.22), (2.24), (2.25)
which satis es the identity
Z Z
@U
+ A(U ) V dxdt = f V dxdt; (2.26)
QT @t QT
where V is any function from the appropriate space (this space

for case a(S) = 1 + S will be speci ed in section 3.6), and
" Z ! #
t 2
@ @U @U
A(U ) a d :
@x 0 @x @x
The basic characteristics of equations such as (2.22) is that

higher derivatives of the nonlinear term depend on time inte-
gral. The coe cients a = a(S) in equation (2.22) contains inte-
gral determining nonlocal operator. In the theory of di erential
equations
Z one often uses the fact that the Volterra’s operator of
t
a kind U d is increasing [175]. However, in a combination
0
with derivative on spatial variable this property might be lost.
For example, the equation
Z t
@U @ @U
= a d
@t @x 0 @x
Z t
is parabolic but if we replace U d = V then the nonlinear
0
equation becomes hyperbolic
@ 2V @ @V
2
= a ;
@t @x @x
which generally does not possess global solutions.
The following existence and uniqueness statement takes place
for problem (2.22), (2.24), (2.25) (see [138]).
Theorem 2.1 If
a(S) = (1 + S)p ; 0 < p 1;
and f 2 W21 (QT ); then there exists a unique solution U of prob-

lem (2.22), (2.24), (2.25) satisfying the following properties:
U 2 L2p+2 (0; T ; W 12p+2 ( )); @U=@t 2 L2 (QT );
@ 2U
2 L2 (QT );
@x2
p @ 2U
T t 2 L2 (QT ):
@t@x
Theorem 2.1 and some of its generalizations for (2.22), (2.23)

type equations and for their multidimensional variants are proved
in [137], [138], [139], [141], [145], [187], [188], and [219], with ap-
plication of Galerkin’s and compactness methods [327], [328],
and [461].
It should be noted that Theorem 2.1 in case p = 1 will be
proved completely in section 3.6.
Below we discuss the asymptotic behavior of the solutions of
the initial-boundary value problems. The existence result can
also be proved based on the a-priori estimates given in the inves-
tigation of the asymptotic behavior of the problem for Models I
and II.
In [305] and [306] it was shown that equation (2.22) gives rise
to coercive monotone operators in the space W 1p (0; 1; L2 (0; T ))
but not in the traditional L2 (0; T ; W 1p (0; 1)). In mentioned works
by the same scheme the analogous results are obtained for (1.12)
type models.
2.4.2 Asymptotic behavior of the solutions as

time tends to in nity
This section is devoted to longtime behavior of the solutions
of initial-boundary value problems for the nonlinear integro-
di erential models (2.22) and (2.23). These equations arise at
mathematical modeling of process of penetration of an electro-
magnetic eld into a substance, with temperature dependent
coe cient of electroconductivity. These equations were derived
in sections 2.1 and 2.2.
The initial-boundary value problems with rst-type bound-
ary conditions are stated. Investigations, which are made in
[217], [218], [228], [231] are given. Let’s formulate the state-
ment for asymptotic behavior of the solutions of the problem
with homogeneous boundary conditions.
At rst for one nonlinear integro-di erential problem (2.22)
we shall give a-priori estimates of the solutions independent of
t. From these estimates the stabilization of the solution follows
as t ! 1.
In the cylinder Q = (0; 1); we consider problem (2.22),
(2.24), (2.25).
Let us assume that
a(S) a0 = const > 0:
If
2a0
;
C
where C here and below is a positive constant independent of t,
the following estimate takes place
t=2
jjU ( ; t)jj e jjU0 jj :
Here jj jj is a norm of the space L2 ( ), i.e.,

Z 1=2
jjU ( ; t)jj = U 2 (x; t)dx :
Sometimes, we will use subscript to indicate di erent kinds of

norms.
Thus, for a problem (2.22), (2.24), (2.25) as t ! 1 the sta-
bility of the solution has decaying exponential rate in the norm
of the space L2 ( ). The asymptotic behavior of the solution of
problem (2.22), (2.24), (2.25) as t ! 1 can also be proved in
the norm of the space W21 ( ).
Using a technique found in [270], in which the stabilization of
the solution of an initial-boundary value problem for the equa-
tions of a barotropic viscous liquid is established, the following
statement is proved in [217].
Let us assume that a = a(S) satis es a(S) a0 = const > 0
and the two additional restrictions: a0 (S) 0 and a00 (S) 0:
In this case the following a-priori estimates are true for all t
[218]:
Z tZ 2
@U
dxd C;
0 @x
Z t Z 2
d @U
dx d C:
0 d @x
Therefore,
Z 1 Z 2
@U
dxd C;
0 @x
Z 1 Z 2
d @U
dx d C:
0 d @x
Now, using well-known fact (see, for example, [25], [270], and
[296]) it is clear that
jjU ( ; t)jjW 1 ( ) ! 0; t ! 1:
2
Thus, stabilization of the solution of problem (2.22) is proved

in the norm of the space W21 ( ) under the following conditions
on coe cient a = a(S):
a(S) a0 = const > 0; a0 (S) 0; a00 (S) 0:

Note that such type of result is true for problem (2.23)-(2.25)
as well.
It is interesting to get the stabilization character in a stronger
norm as well. Let us also note that some results about exis-
tence, uniqueness, and asymptotic behavior of the solutions for
above-discussed integro-di erential models and that generaliza-
tions with several kinds of initial and boundary conditions are
studied in [28], [29], [31], [32], [33], [34], [35], [49], [50], [137],
[138], [139], [140], [141], [145], [146], [147], [186], [187], [188],
[217], [218], [219], [220], [222], [223], [227], [228], [229], [231],
[232], [233], [234], [235], [236], [237], [238], [239], [240], [241],
[242], [243], [245], [246], [247], [248], [249], [251], [252], [253],
[254], [256], [257], [261], [276], [277], [278], [280], [282], [283],
[302], [303], [304], [305], [306], [322], [333], and [334] among oth-
ers.
2.4.3 Rate of the asymptotic behavior of

solutions of Model I
The relations in subsection 2.4.2 give the stabilization of a non-
stationary solution with the norm of the space W21 ( ), but usu-
ally do not provide any information on the convergence rate.
Here we rectify this de ciency and give asymptotic behavior to

be also con rmed by the numerical experiments in Chapter 4.
Consider the following initial-boundary value problem:
@U @ @U
= a(S) ; (x; t) 2 Q; (2.27)
@t @x @x
U (0; t) = U (1; t) = 0; t 0; (2.28)

U (x; 0) = U0 (x); x 2 [0; 1]; (2.29)
where Z t 2
@U
S(x; t) = d : (2.30)
0 @x
Here a = a(S) and U0 = U0 (x) are given functions of their
arguments.
Let us estimate the rate of asymptotic behavior of the solu-
tion. In particular, let us show that the convergence occurring
in the relations in subsection 2.4.2 is also exponential.
Note that, in this section, we derive asymptotics with indi-
cation of order for k@U=@tk as well.
Let us note once again that below in this and in the next
sections C denote positive constants independent of t.
Theorem 2.2 If a(S) a0 = const > 0; a0 (S) 0; a00 (S)

0, and U0 2 W22 (0; 1) \ W k2 (0; 1); then the solution of problem
(2.27)-(2.30) satis es the relation
@U @U a0
+ C exp t :
@x @t 2
Proof. We multiply (2.27) by @U=@t and integrate the resulting

relation over the interval [0; 1]: By using the formula of integra-
tion by parts and taking into account the boundary conditions
(2.28), we obtain
2 Z 1 2
@U 1 @ @U
+ a(S) dx = 0: (2.31)
@t 2 0 @t @x
Let us di erentiate relation (2.27) with respect to t:
@ 2U @ @a(S) @U @ 2U
+ a(S) = 0: (2.32)
@t2 @x @t @x @t @x
We multiply (2.32) by U and integrate the resulting relation
over the interval [0; 1]:
Z 1 2 Z 1 2
@ U @a(S) @U
2
U dx + dx
0 @t 0 @t @x
(2.33)
Z 1 2
1 @ @U
+ a(S) dx = 0:
2 0 @t @x
It follows from (2.31) and (2.33) that
Z 1 2 2 Z 1 2
@ U @U @ @U
2
U dx + + a(S) dx
0 @t @t 0 @t @x
Z 1 2
@a(S) @U
+ dx = 0;
0 @t @x
which, in view of the relation
Z 1 2 2
@ U @U 1 d2
2
U dx + = kU k2 ;
0 @t @t 2 dt2
acquires the form
1 2
1 d2
Z
d @U
2
kU k2 + a(S) dx = 0: (2.34)
2 dt dt 0 @x
Let us multiply (2.32) by @U=@t and integrate the resulting

relation over the interval [0; 1]. By using the formula of integra-
tion by parts, the boundary conditions (2.28), and Poincare’s
inequality
@U @ 2U
;
@t @t @x
we obtain
2 2 Z 1 4
d @U @U 1 @ @U
+ 2a0 + a0 (S) dx 0: (2.35)
dt @t @t 2 0 @t @x
Multiplying (2.27) by U and integrating the resulting rela-
tion over the interval [0; 1]; we obtain
Z 1 2
1 d 2 @U
kU k + a(S) dx = 0; (2.36)
2 dt 0 @x
which, together with Poincare’s inequality and the condition
a(S) a0 ; implies that
1d
kU k2 + a0 kU k2 0: (2.37)
2 dt
We multiply relations (2.34), (2.36), and (2.37) by ; , and
, respectively, where ; , and are positive constants. By
summing the resulting relations and inequality (2.35), we obtain
Z 1 2
+ d 2 @U
kU k + a(S) dx + a0 kU k2
2 dt 0 @x
1 2 2
d2
Z
d @U d @U
+ 2
kU k2 + a(S) dx +
2 dt dt 0 @x dt @t
2 Z 1 4
@U 1 0 @ @U
+2a0 + a (S) dx 0;
@t 2 0 @t @x
Another random document with
no related content on Scribd:
The cramming processes which are resorted to in order to force
children at fixed times from the lower to the higher grades of public
schools, and more especially from grammar to normal or high
schools, is a fruitful source of evil in this direction. It is not always so
much hard study as it is the badly-arranged and too numerous
subjects of study that make the strain. I have spoken of this in
another connection as follows:42 “Our children are too largely in the
hands of those educationalists to whom Clouston refers,43 who go on
the theory that there is an unlimited capacity in every individual brain
for education to any extent and in any direction. Children varying in
age and original capacity, in previous preparation, and in home-
surroundings are forced into the same moulds and grooves. The
slow must keep pace with the fleet, the frail with the sturdy, the
children of toil and deprivation with the sons and daughters of wealth
and luxury. A child is always liable to suffer from mental overwork
when the effort is made to force its education beyond its receptive
powers. Education is not individualized enough. The mind of the
child is often confused by a multitude of ill-assorted studies.
Recreation is neglected and unhealthy emulation is too much
cultivated. In many communities admissions to various grades of
public schools are regulated entirely by the averages obtained at
examinations, instead of on the general record of the pupils in
connection with proper but not too severe examinations. In
consequence often, after the campaign of overwork and confusion
called an examination, we see children developing serious
disturbances of health or even organic disease—paroxysmal fever,
loss of appetite, headache or neckache, disturbed sleep, temporary
albuminuria, chorea, hysteria, and hystero-epilepsy.”
42 “Toner Lecture on Mental Overwork and Premature Disease among Public and
Professional Men.” delivered March 19, 1884, Washington, Smithsonian Inst.,
January, 1885.
43 Clinical Lectures on Mental Diseases.
The term students' hysteria has been applied to the neuromimetic

disorders from which medical students frequently suffer during their
attendance upon lectures. Some years since, when engaged in
examining students upon the lectures upon the practice of medicine
delivered by Professor DaCosta, I saw many illustrations of this
affection, some of which were very amusing. In a paper on hysteria
which received the prize at the Physical Society of Guy's Hospital, P.
Horrocks44 writes that during the fortnight following the death of the
late Napoleon, Sir James Paget was consulted for stone in the
bladder by no less than four gentlemen who had nothing the matter
with them. “How many students,” says Horrocks, “are there, of one
year's standing or more, who have not imagined and really became
convinced that they were suffering from some disease, generally a
fatal disease?” In such cases we have a combination of true
psychical influence with overwork and the unhygienic surroundings
for which our medical colleges are notorious.
44 Med. and Surg. Reporter, vol. xxxvii., Nov. 24, 1877.
It has been my personal experience that comparatively few cases of

hysteria occur among female medical students. Not long since a
thesis was presented at graduation by a woman medical student45
on the curative effects of professional training in neurasthenic and
hysterical women. In this she shows that there are certain relations
of mind over body which enable it to modify bodily conditions and
ward off disease when other remedies appear almost powerless.
She illustrates the therapeutic power of mental impressions and
occupations by two cases in which a judicious and careful course of
study acted to cure severe nervous and uterine troubles. One of
these women, who had suffered with neurasthenia and general
debility, severe nervous headaches, and other symptoms, was able
during her last year at college to attend fifteen lectures a week,
besides clinics, prepare for examination on five subjects, and was
seldom troubled with even headache. She afterward was employed
in hospital work, and could walk five miles a day without discomfort.
That women medical students know when and how to take care of
themselves during the menstrual period, and that they can, if they
see fit, cease work or lighten their labors at that time, would partly
account for their escaping from nervous break-down.
45 “The Therapeutic Value of Mental Occupation,” by Hannah M. Thompson, M.D.,
Medical and Surgical Reporter, November, 1883.
That any form of irritation in a patient predisposed to hysteria may

act as an exciting cause in this affection has led Laségue to apply
the term peripheral hysteria to certain cases. One of his cases was a
girl fourteen years old, who, having suffered for a few hours with her
eyes because of some sand thrown into one of them by a playmate,
awoke the next morning with a spasm of the eyelid on that side,
which rendered it impossible for her to open that eye; and it
remained closed during four months. He considered that the irritation
produced by the sand was not the immediate cause of the spasm,
but that its long duration was an hysterical phenomenon. The patient
afterward became the subject of hysterical manifestations. In another
complete hysterical aphonia came on after a slight bronchitis.
Another, after an attack of indigestion, refused to touch either food or
drink for twenty-four hours, and later was troubled with regurgitation
from constriction of the pharynx or œsophagus which lasted for
some weeks.
Anæmia and chlorosis are frequent exciting causes of hysteria in

children, particularly in girls.
Disorders of menstruation play a prominent part as exciting causes

of special hysterical manifestations. The period of the establishment
of the menstrual function is one that is particularly fertile in the
production of hysteria, much more so than acquired disorders of
menstruation occurring later. Menorrhagia, dysmenorrhœa, and
certain local utero-vaginal disorders may act upon those predisposed
to hysteria as exciting causes. These conditions themselves are, on
the other hand, sometimes caused by nervous, hysterical states in
the individual.
With reference to the very common assertions that continence on the

one hand, and sexual over-indulgence on the other, are the most
prolific causes of hysteria, the true stand to take is that neither of
these positions is philosophically correct; for, as Briquet has shown,
nuns on the one hand and prostitutes on the other are frequent
victims of this protean disorder.
As affirmed by Jolly, sexual over-irritation, particularly that induced

by onanism, more frequently causes hysteria than sexual abstinence
or deprivation.
The occurrence of hysteria and hysterical choreas among pregnant

women has long been recognized. Scanzoni, quoted by Jolly,46
states that of 217 patients whom he had treated, 165, or 76 per cent.
had been puerperal, and that of the latter not less than 65 per cent.
had borne children more than three times. Cases of grave hysteria or
hystero-epilepsy have been aggravated by pregnancy and have led
to premature labor.
46 Op. cit.
Chrobak attacks the etiological problem of hysteria by referring its

causation to movable kidneys! He observed 19 such patients in
Vienna, 16 being in Oppolzer's clinic.47 Three times no subjective
symptoms accompanied the anomaly; eight times the trouble could
be referred either to the dislocation of the kidneys or to disease of
the same; and eight times the evidence of hysteria was
unmistakable. Among the latter eight cases neither vaginal, uterine,
nor ovarian conditions were recognized. He concludes that there is a
direct nervous connection between the kidneys and the genital
organs, and between both and hysteria.
47 Medizinische-Chirurgische Rundschan, quoted by Boston Medical and Surgical
Journal, 1870, lxxxiii. 430-432.
In brief, the truth is that frequent or severe local irritation in any part
of the body in an individual of the hysterical diathesis may act as the
exciting cause of an hysterical paroxysm or of special hysterical
manifestations. Irritation or disease of the uterus or ovaries may
result in hysteria, as may the bite of a dog, a tumor of the brain, a
polypus in the nose, a phymosis, an irritated clitoris, a gastric ulcer, a
stenosis of the larynx, a foreign body in the eye, a toe-nail ulcer, or a
movable kidney.
Whatever tends to exhaust the nervous system will also cause

hysteria, but only in those who have some inherited predisposition to
the disorder. C. Handfield Jones48 mentions heatstroke and severe
physical labor as such causes. One of the sequelæ of heatstroke
enumerated by Sir R. Martin, and quoted by Jones, is a distressing
hysterical state of the nervous system, with an absence of self-
control in laughing and crying, the paroxysm being followed by great
prostration of nervous power.
48 Op. cit.
The effect of imitation in the production of hysteria has been known

in all ages. Most of the epidemics and endemics of nervous
disorders which have from time to time prevailed in various parts of
the world have either been hysterical in character or have had in
them a large element of hysteria. While it is impossible in a practical
work to devote much space to this branch of the subject, a
discussion of hysteria in its etiological relations would be imperfect
without some reference to these outbreaks. In ancient times, in the
Middle Ages, and within comparatively recent periods extraordinary
epidemics have occurred. No country within the range of medical
observation has been entirely free from them. Communities civilized
and semi-civilized, Christian and Mohammedan, Protestant and
Catholic, have had a fair share of the visitations. Some of them
constitute epochs in history, and, as Hecker,49 their greatest
historian, has remarked, their study affords a deep insight into the
work of the human mind in certain states of society. “They are,” he
says, “a portion of history, and will never return in the way in which
they are recorded; but they expose a vulnerable part of man—the
instinct of imitation—and are therefore very nearly connected with
human life in the aggregate.”
49 The Epidemics of the Middle Ages, from the German of J. F. C. Hecker, M.D.,
Professor at the Frederick William University at Berlin, etc., translated by B. G.
Babington, M.D., F. R. S., etc.; 3d ed., London, 1859.
Some authors under hysteria, others under catalepsy, others under
ecstasy, still others under chorea, have discussed these epidemics—
a fact which serves to emphasize the truth that while these affections
have points of difference, they have also an easily-traced bond of
union. They are but variations of the same discordant tune. Briquet
in an admirable manner sketches their history from the age of
Pausanias and Plutarch to the time of Mesmer. Of American writers,
James J. Levick50 of Philadelphia has furnished one of the most
valuable contributions to this subject.
50 “An Historical Sketch of the Dance of St. Vitus, with Notices of some Kindred
Disorders,” Med. and Surg. Reporter, vol. vii., Dec. 21 and 28, 1861, p. 276, and Jan.
4 and 11, 1862, p. 322.
In the year 1237 a hundred children or more were suddenly seized

with the dancing mania at Erfurt; another outbreak occurred at
Utrecht in 1278.
As early as the year 1374 large assemblages of men and women

were seen at Aix-la-Chapelle affected with a “dancing mania.” They
formed circles and danced for hours in wild delirium. Attacks of
insensibility, of convulsions, and of ecstasy occurred. The disease
spread from Germany to the Netherlands. In a few months it broke
out in Cologne, and about the same time at Metz. “Peasants,” we are
told, “left their ploughs, mechanics their workshops, housewives their
domestic duties, to join the wild revelry, and this rich commercial city
became the scene of the most ruinous disorder.”
The festival of St. John the Baptist was one celebrated in strange
wild ways in these early days. Fanatical rites, often cruel and
senseless, were performed on these occasions. Hecker supposes
that the wild revels of St. John's Day in 1374 may have had
something to do with the outbreak of the frightful dancing mania
soon after this celebration. It at least brought to a crisis a malady
which had been long impending.
The Flagellants afford another illustration of an early religio-nervous

craze. Self-flagellation was indulged in for generations before the
fourteenth century, but it then became epidemic. A brotherhood of
Flagellants was formed; they marched in processions carrying
scourges, with which they violently lashed and scourged themselves.
As late as 1843, Flagellant processions, but without the whips and
scourging, were continued in Lisbon on Good Friday.
Strasburg was visited by the dancing plague in 1418. Those afflicted

were conducted to the chapel of St. Vitus, where priests attempted to
relieve them by religious ceremonies. The name St. Vitus's dance,
still so common as a synonym for chorea, has come down to us
because of the alleged wonderful doings of this saint in behalf of
those affected during some of the dancing epidemics. Both Hecker
and Madden51 give interesting details of the personal history of St.
Vitus, who was a Sicilian, born in the time of Diocletian, and even in
childhood is said to have worked great miracles, and was delivered
from many sufferings and torments. He died about the year 303. His
body was moved to Apulia, afterward to St. Denys in France, and still
later to the abbey of Corvey in Saxony. A legend was invented that
St. Vitus, just before he bent his neck to the sword, prayed to God
that he might protect from the dancing mania all those who should
solemnize the day of his commemoration and fast upon its eve.
51 Phantasmata; or, Illusions and Fanaticisms, etc., by R. R. Madden, F. R. S.,
London, 1857.
Another strange disorder called tarantismus derived its name from

the fact that it was supposed to be caused by the bite of the
tarantula, a ground-spider common in Apulia, Italy. According to
Hecker, the word tarantula is the same as terrantola, a name given
by the Italians to a poisonous lizard of extraordinary endowments.
The fear of the insect was so general that its bite was much oftener
imagined than actually received. The disorder was probably in
existence long before the fifteenth century, although the first account
of it, that of Nicholas Perotti, refers to its occurrence in this century.
Many symptoms followed the bite or supposed bite: the individuals
became melancholy, stupefied, lost their senses, and, above all,
were irresistibly impelled to dance until exhausted and almost
lifeless. It was believed that the results of the bite could be cured, or
at least much benefited, by dancing to a certain kind of music.
Tarantism was at its height in the seventeenth century. To this day, in
some parts of Italy, dances called tarantellas are performed with
intricate figures to marked time.
Abyssinia was visited by a dancing mania called the tigretier, which,

according to Hecker, resembled the original mania of the St. John
dancers. It exhibited a similar ecstasy. Those affected with it were
cured by dancing to the music of trumpeters, drummers, fifers, etc.
Levick says that the dancing mania of the fifteenth century is still
kept in popular remembrance in some places by an annual festival,
especially at Echtermarch, a small town in Luxembourg, where a
jumping procession occurs annually on Whit Tuesday. In the year
1812, 12,678 dancers were in the procession.
The Anabaptists, a religious sect of the sixteenth century, exhibited

many of the wild and grotesque phenomena of hysteria or hystero-
epilepsy.
The French Calvinists or Camisards, who appeared near the close of

the seventeenth century, were also the subjects of ecstasy and of
peculiar fits of trembling. These trembleurs experienced convulsive
shocks in the head, the shoulders, the legs, and the arms, and were
sometimes thrown violently down.
About 1731 and later great crowds frequented the tomb of Deacon
François de Paris, an advocate of the doctrines of Jansenius. It was
reported that miracles were performed at his tomb: the sick were
brought there, and often were seized with convulsions and pains,
through which they were healed. The subjects of these attacks are
sometimes spoken of as the Jansenist Convulsionnaires. The tomb
was in the cemetery of St. Médard, and hence those who visited the
place were also termed the Convulsionnaires of St. Médard. This
disorder increased, multiplied, and disseminated, lasting with more
or less intensity for fifty-nine years. Great immorality prevailed in the
secret meetings of the believers.
Hecker gives some remarkable instances of the effect of sympathy
or imitation exhibited on a smaller scale than in the epidemics of the
Middle Ages. One is of a series of cases of fits in a Lancashire
factory, the first one brought on by a girl putting a mouse into the
bosom of another. In Charité Hospital in Berlin in 1801 a patient fell
into strong convulsions, and immediately afterward six other patients
were affected in the same way; by degrees eight more were
attacked. At Redruth, England, a man cried out in a chapel, “What
shall I do to be saved?” Others followed his example, and shortly
after suffered most excruciating bodily pain. The occurrence soon
became public; hundreds came, and many of them were affected in
the same way. The affection spread from town to town. Four
thousand people were said in a short time to be affected with this
malady, which included convulsions.
Hecker in the edition of his work referred to has also a treatise on

child pilgrimages.52 These pilgrimages, like the dancing mania,
occurred in the Middle Ages. The greatest was the boy crusade in
the year 1212. The passion to repossess the Holy Land then had its
grip on Catholic Europe. The first impulse to the child pilgrimages
was given by a shepherd-boy, who had revelations and ecstatic
seizures, and held himself to be an ambassador of the Lord. Soon
thirty thousand souls came to partake of his revelations; new child-
prophets and miracle-workers arose; the children of rich and poor
flocked together from all quarters; parents were unable to restrain
them, and some even began to urge them. A host of boys, armed
and unarmed, assembled at Vendôme, and started for Jerusalem
with a boy-prophet at their head. They got to Marseilles, and
embarked on seven large ships. Two ships were wrecked, and not a
soul was saved. The other five ships reached Bougia and
Alexandria, and the young crusaders were all sold as slaves to the
Saracens. In Germany child-prophets arose, especially in the Rhine
countries and far eastward. An army of them gathered together,
crossed the Alps, and reached Genoa. They were soon scattered;
many perished; many were retained as servants in foreign lands;
some reached Rome. A second child's pilgrimage occurred twenty-
five years later. It was confined to the city of Erfurt. One thousand
children wandered, dancing and leaping, to Armstadt, and were
brought back in carts. Another child's pilgrimage from Halle, in
Suabia, to Mount St. Michel in Normandy, occurred in 1458.
52 Translated by Robert H. Cooke, M. R. C. S.
In the convent of Yvertet in the territory of Liège, in 1550, the

inmates were seized with a leaping and jumping malady. The
disorder began with a single individual, and was soon propagated.
Sometimes the convulsive disorders of early days, especially those

occurring in convents, were associated with the strange delusion that
the subjects of them were changed into lower animals. Various
names have been given to disorders of this kind, such as
lycanthropia or wolf madness, zoomania or animal madness, etc.
Burton in the Anatomy of Melancholy gives an interesting summary
of these disorders, which are also discussed by Levick.
In 1760 a religious sect known as the Jumpers prevailed in Great

Britain. They were affected with religious frenzy, and jumped
continuously for hours. Other jumping epidemics have appeared at
different times, both in Great Britain and in this country.
The New England witchcraft episode is of historical interest in

connection with this subject of epidemic hysteria. This excitement
occurred during the latter part of the seventeenth century. Adults and
children were its subjects. The Rev. Cotton Mather records many
cases, some of which illustrate almost every phase of hysteria.
Individuals who were seized with attacks, which would now be
regarded as hysterical or hystero-epileptic, were supposed to have
become possessed through the machinations of others. Those who
were supposed to be possessed were tried, condemned, and
executed in great numbers. Many accused themselves of converse
with the devil. The epidemic spread with such rapidity, and so many
were executed, that finally the good sense of the people came to the
rescue.
The nervous epidemics, nearly all religious, which have occurred in
this country have usually been during the pioneer periods, and have
therefore appeared at different eras as one part of the country after
another has been developed. Kentucky, Tennessee, Virginia, and
neighboring States were visited time and again. Even to-day we
occasionally hear of outbreaks of this kind in remote or primitive
localities, whether it be in the far South-west or in the woods of
Maine.
David W. Yandell53 has published a valuable paper on “Epidemic

Convulsions,” the larger part of the materials of which were collected
by his father for a medical history of Kentucky. From this it would
appear the convulsions were first noticed in the revivals from 1735 to
1742. Many instances are related of fainting, falling, trance,
numbness, outcries, and spasms. The epidemic of Kentucky spread
widely, reappeared for years, and involved a district from Ohio to the
mountains of Tennessee, and even to the old settlements in the
Carolinas. Wonderful displays took place at the camp-meetings. At
one of these, where twenty thousand people were present, sobs,
shrieks, and shouts were heard; sudden spasms seized upon scores
and dashed them to the ground. Preachers went around in ecstasy,
singing, shouting, and shaking hands. Sometimes a little boy or girl
would be seen passionately exhorting the multitude, reminding one
of the part taken by the children in the epidemics of the Middle Ages.
A sense of pins and needles was complained of by many; others felt
a numbness and lost all control of their muscles. Some subjects
were cataleptic; others were overcome with general convulsions.
53 Brain, vol. iv., Oct., 1881, p. 339 et seq.
The term jerks was properly applied to one of the forms of

convulsion. Sometimes the jerking affected a single limb or part. The
Rev. Richard McNemar has given a graphic description of this
jerking exercise in a History of the Kentucky Revival. The head
would fly backward and forward or from side to side; the subject was
dashed to the ground, or would bounce from place to place like a
football, or hop around with head, limbs, and trunk twitching and
jolting in every direction. Curiously, few were hurt. Interesting
descriptions of the jerks can be found in various American
autobiographical and historical religious works. In such books as the
Autobiography of Peter Cartwright, a Western Methodist, for
instance, striking accounts of some of the phases of these epidemics
are to be found. Lorenzo Dow in his Journal, published in
Philadelphia in 1815, has also recorded them.
Hysterical laughter was a grotesque manifestation often witnessed.

The holy laugh began to be a part of religious worship. Dancing,
barking, and otherwise acting like dogs, were still other
manifestations. It is remarkable that, according to Yandell, no
instance is recorded in which permanent insanity resulted from these
terrible excitements.
The absurd and extraordinary exhibitions witnessed among the

Shakers belong to the same category, and have been well described
by Hammond and others.
In a History of the Revival in Ireland in 1859, by the Rev. William

Gibson, instances of excitement that fairly rivalled those which
occurred in our Western States are given. Cases of ecstasy are
described.
The religious sect known as the Salvation Army, which has in very
recent years excited so much attention, curiosity, and comment both
in America and England, has much in common with the Jumpers, the
Jerkers, and the Convulsionnaires. The frenzied excitement at their
meetings, with their tambourine-playing, dancing, shouting, and
improvising are simply the same phases of religio-hysterical disorder,
modified by differences in the age and environment.
In 1878, in the district of Tolmezo, Italy, an epidemic of hysteria

which recalls the epidemics of the Middle Ages occurred. It has been
described by M. Léon Colin.54 It was reported to the prefect of
Undine that for three months some forty females living in the
commune of Verzeguis had been attacked by religious mania. “From
the report it appears that the first was in the person of a woman
named Marguerite Vidusson, who had been the subject of simple
hysteria for about eight years. In January, 1878, she began to suffer
from convulsive attacks, accompanied by cries and lamentations.
She was regarded as the subject of demoniacal possession, and on
the first Sunday in May was publicly exorcised. Her affection,
however, increased in severity; the attacks were more frequent and
more intense, and were especially provoked by the sound of the
church-bells and by the sight of priests. Seven months later three
other hysterical girls became subject to convulsive and clamorous
attacks. Here, again, an attempt was made to get rid of the
supposed demon. A solemn mass was said in the presence of the
sufferers, but was followed only by a fresh outbreak. At the time of
the visit of the delegates eighteen were suffering, aged from sixteen
to twenty-six years, except three, whose ages were respectively
forty-five, fifty-five, and sixty-three years. Similar symptoms had also
appeared in a young soldier on leave in the village.” During the
attacks the patients talked of the demon which possessed them,
stated the date on which they were seized by it, and the names of
the persons who were possessed before them. Some boasted of
being prophetesses and clairvoyants and of having the gift of
tongues. In all, the sound of church-bells caused attacks, and
religious ceremonies appeared not only to aggravate the disease in
the sufferers, but also to cause its extension to those not previously
attacked. M. Colin points out that the soil is particularly favorable for
the development of an epidemic of this nature. The people of
Verzeguis are backward in education and most superstitious.
Functional nervous diseases are common among them. The
inhabitants of the village are largely cut off from intercourse with the
adjacent country in consequence of comparative inaccessibility and
the frequent interruption of communications by storms and floods.
Craniometric observations on twelve of the inhabitants seemed to
show that the brachycephalic form of skull predominated, and that
the development of the cranium was slightly below the average. The
epidemic proved extremely obstinate.
54 Annales d'Hygiène, quoted in Lancet, Oct. 16, 1880.
In Norway and New Caledonia similar hysterical outbreaks have
been observed in recent times.
An endemic of hysteria from imitation occurred in Philadelphia in

1880. Some of the cases fell under my own observation. A brief
account of them is given by Mitchell in his Lectures. The outbreak
occurred in a Church Home for Children, to which Dr. S. S. Stryker
was physician. The Home contained ninety-five girls and six boys; all
of them were well nourished and in good condition. The epidemic
began by a girl having slight convulsive twitchings of the extremities,
with a little numbness. Attacks returned daily; respiration became
loud and crowing. She soon had all the phenomena of convulsive
hysteria. Many of her comrades began to imitate her bark. Soon
another girl of ten was attacked with harsh, gasping breathing, with
crowing, speechlessness, clutching at her throat, and the whole
series of phenomena exhibited by the first girl attacked. Nine or ten
others were affected in like manner, and many of the remaining
children had similar symptoms in a slight degree. At first convulsions
occurred irregularly; after a while they appeared every evening; later,
both morning and evening. The presence of visitors would excite
them. Many interesting hysterical phases occurred among the
children. One night some of them took to walking about on their
hands and knees; others described visions. The girls often spoke of
being surrounded by wild beasts, and one child would adopt the
fiction which another related in her hearing. The cases were
scattered about in different hospitals, and made good recoveries in
from one to two months.
The Jumpers or Jumping Frenchmen of Maine and Northern New

Hampshire were described by Beard in 1880.55 They presented
nervous phenomena in some phases allied to hysteria. In June,
1880, Beard visited Moosehead Lake and experimented with some
of them. Whatever order was given them was at once obeyed. One
of the Jumpers, who was sitting in a chair with a knife in his hand,
was told to throw it, and he threw it quickly so that it struck in a beam
opposite; at the same time he repeated the order to throw it with a
cry of alarm. They were tried with Latin and Greek quotations, and
repeated or echoed the sound as it came to them. They could not
help repeating any word or sound that came from the person that
ordered them. Any sudden or unexpected noise, as the report of a
gun, the slamming of a door, etc., would cause them to exhibit some
phenomena. It was dangerous to startle them where they could
injure themselves, or if they had an axe, knife, or other weapon in
their hands. Since the time of Beard's observation accounts of their
doings have now and then found their way into newspapers. One
recent account tells of one of these peculiar people jumping from a
raft into the Penobscot River on an order to jump.
55 Journal of Nervous and Mental Diseases, vol. vii., 1880, p. 487.
Hammond56 has described under the name Miryachit an affection

which seems to be essentially the same disorder as that of which the
Jumpers are the victims. He quotes from a report of a journey from
the Pacific Ocean through Asia to Europe by Lieutenant B. H.
Buckingham and Ensigns Geo. C. Foulk and Walter McLean of the
United States Navy, an account of this disease. The party made their
first observations on this affection while on the Ussuri River in
Siberia. The captain of the general staff approached the steward of
the boat suddenly, and without any apparent reason or remark
clapped his hand before his face; instantly the steward clapped his
hand in the same manner, put on an angry look, and passed on.
When the captain slapped the paddle-box suddenly, the steward
instantly gave it a similar thump. Some of the passengers imitated
pigs grunting or called out absurd names, etc.; the poor steward
would be compelled to echo them all. The United States naval
officers were informed that the affection was not uncommon in
Siberia, and that it was commonest about Yakutsk, where the winter
cold is extreme. Both sexes were subject to it, but men much less
than women. It was known to Russians by the name of Miryachit.
56 New York Med. Journ., Feb. 16, 1884.
In both these classes of cases a suggestion of some kind was

required, and then the act took place independently of the will.
“There is another analogous condition known by the Germans as
Schlaftrunkenheit, and to English and American neurologists as
somnolentia or sleep-drunkenness. In this state an individual on
being suddenly awakened commits some incongruous act of
violence, ofttimes a murder. Sometimes this appears to be a dream,
but in others no such cause could be discovered.” Curious instances
are mentioned by Hammond of this disorder.
The phenomena of automatism at command in hypnotized subjects

have much similarity to the phenomena of these affections, and the
same explanation to a certain extent will answer for both.
Paget57 has ably discussed the subject of neuromimesis in general,

and Mitchell58 devotes two lectures to its consideration. As already
stated when discussing the synonyms of hysteria, the mistake must
not be made of supposing all cases of hysteria to be instances of
neuromimesis; but, as Mitchell remarks, the hysterical state,
however produced, is a fruitful source of mimicry of disease in its
every form, from the mildest of pains up to the most complete and
carefully-devised frauds. “Its sensitiveness and mobility, its timidity
and emotionalness, its greed of attention, of sympathy, and of power
in all shapes, supply both motive and help, so that while we must be
careful not to see mimicry in every hysteric symptom, we must in
people of this temperament be more than usually watchful for this
form of trouble, and at least reasonably suspicious of every peculiar
or unusual phenomenon.”
57 Op. cit.
58 Op. cit.
SYMPTOMATOLOGY.—At the outset of the discussion of the

symptomatology of hysteria, hysterical cases should be divided into
four classes—viz. (1) Cases in which the symptoms are involuntary;
(2) cases in which the symptoms are artificially induced and become
involuntary; (3) cases in which the symptoms are acted or simulated,
but in which the patient, because of impaired mental power, is
irresistibly impelled to their performance; (4) cases in which the
symptoms are purely acts of deception which are under the control
of the patient.
Keeping in mind these different classes, we will always be able to

link to the phenomena of hysteria the psychical element which is
present in all genuine cases of this disorder. To comprehend the
existence of the psychical element in the first class, in which the
manifestations are absolutely involuntary, may offer difficulties. In
these cases, at a period more or less recent or remote, psychical
stimuli may have acted to produce the hysterical phenomena, and,
once produced, these have been repeated and intensified by habit,
and continue independently both of volition and consciousness. The
experiments of Dercum and Parker show how hysterical symptoms
may be artificially induced and may get beyond the patient's control.
The difference between induced and simulated manifestations must
always be clearly borne in mind. To induce a set of phenomena a
certain mechanism must be set in action, and this, through rational,
explicable processes, leads to certain results. The psychical element
enters here both positively and negatively—positively, in the
determination to produce a certain train of events; negatively, in the
condition of mental concentration or abstraction which is a part of the
procedure. In the third class of cases acting or simulation is
dependent upon the irresistible inclinations of the patient. This may
seem to some an uncertain and even dangerous ground to take. I
am convinced, however, after observing many hysterical cases, that
acts clearly purposive, so far as the particular performance is
concerned, are sometimes the result of a general unstable mental
condition. Some at least of these patients are as irresistibly impelled
to swallow blood and vomit, to scream and gesticulate, etc., as is the
monomaniac to commit arson, to ravish, or to kill. In the fourth class,
the cases of pure, unmitigated, uncontrollable deception, the
psychical element is very evident, although some may question
whether such cases should be ranged under the banner of hysteria,
where it is both convenient and customary to place them.
The symptoms of hysteria may develop in any order or after any

fashion. The graver hysterical phenomena, such as convulsions,
paralysis, and anæsthesia, often seem to come on suddenly, but
usually this suddenness of onset is apparent rather than real. Minor
hysterical symptoms, such as general nervous irritability, pains,
aches, and discomforts, and mental peculiarities, have usually been
present for a long time. These minor evidences of the hysterical
constitution are sometimes the only phenomena ever presented.
Todd59 has described an expression of countenance which he

designates as the facies hysterica. The characteristics of this
expression are a remarkable depth and prominent fulness, with more
or less thickness, of the upper lip, and a peculiar drooping of the
upper eyelids. It would be absurd to assert that all hysterical patients
presented this cast of countenance, but an appearance which
approaches closely to this description is presented in a fair
percentage of cases. It has seemed to me that male hysterics were
more likely to have this peculiar facies than hysterical females.
59 Reynolds's System of Medicine, vol. ii. p. 656.
The psychical peculiarities or mental disorders of hysteria form a

large and important part of its phenomena. We have to deal not only
with peculiar and diverse psychical manifestations, but to one form of
mental disorder it is clinically convenient and correct to apply the
designation hysterical insanity.
In the mildest cases of ordinary hysteria conditions of mental

irritability and mobility are sometimes the only striking features.
“Patients,” says Jolly,60 “are timid, easily overcome by any
unexpected occurrence, sentimental, and sensitive. Every trifle
annoys and upsets them; and there is this peculiarity—that a more
recent stimulus may often effect a diversion in an exactly opposite
direction.”
60 Op. cit.
As bearing upon the question of the mental state in hysteria, the

confessions obtained by Mitchell from several patients are of great
interest. One patient, who had learned to notice and dwell upon any
little symptom, vomited daily and aroused much sympathy. She took
little or no food. Spasms came on, and she confessed that every
new symptom caused new anxiety, and that somehow she rather
liked it all. She gradually lost all her symptoms except vomiting, and
overcame this by desperate efforts. Another patient confessed to
having played a game upon her doctor for a long time by pretending
she took no food. She would get out of bed at night, but remain there
all day; she filled up a vessel with water to make others believe she
passed large quantities of urine, etc. Another patient, a girl of
nineteen, who came on a litter from a Western State, after a time
regained her feet. In her confession she stated that what she lacked
was courage. She believed that she would have overcome her
difficulties if any one had told her that nothing was the matter. “In
looking back over the year with the light of the present,” she says, “I
can only say that I believe that there was really nothing the matter
with me; only it seemed to me as if there was, and because of these
sensations I carried on a sort of starvation process physical and
mental.”
The older and some of the more recent classifications of insanity

recognize hysterical insanity as a distinct form of mental disease.
Morel and Skae, however, in their etiological classifications, and
Hammond, Spitzka, Mann, and Clouston in their recently-published
works, give it a “local habitation and a name.” Krafft-Ebing not only
recognizes hysterical insanity as a distinct form of mental disease,
but, after the German fashion, subdivides it quite minutely, as
follows: First, transitory forms: a. with fright; b. hystero-epileptic
deliria; c. ecstatic visionary forms; d. moria-like conditions. Second,
chronic forms: a. hystero-melancholia; b. hystero-mania; c.
degenerative states with hysterical basis.
Spitzka61 speaks of chronic hysterical insanity as an intensification of

the hysterical character, to which “a silly mendacity is frequently
added, and develops pari passu with advancing deterioration.” At the
State Hospital for the Insane at Norristown and at the department for
the insane of the Philadelphia Hospital cases of chronic hysterical
insanity have come under my observation. Hammond under
hysterical mania includes several different and somewhat distinct
mental disorders.
61 Insanity, its Classification, Diagnosis, and Treatment, by E. C. Spitzka, M.D., New
York, 1883.
With regard to the occurrence of hysterical manifestations amongst

patients suffering from some well-recognized non-hysterical forms of
insanity, a tour through any large asylum will afford abundant
evidence. Cases of tremor closely simulating cerebro-spinal
sclerosis have been observed frequently among the insane.
Paralysis, contracture, hysterical joints, hysterical neuralgias,
convulsions, and cataleptoid phenomena are among other hysterical
manifestations which have fallen under personal observation among
the insane of various classes.
A remarkable case of hysterical motor paralysis was observed at the

State Hospital for the Insane at Norristown. This patient was an
intelligent single woman about thirty-five years of age, of good family,
well educated; she had been a teacher and writer, and became
insane through family and business troubles. When only eight years
of age she was paralyzed for two years and a half, and had had at
times during her life, before becoming insane, attacks of partial or
complete unconsciousness. Prior to coming under observation she
had been an inmate of an English private asylum. She was sick on
shipboard coming to this country, and on her arrival was in a state of
delirium and insomnia, with attacks of loss of sight. Four months
later she developed mania with suicidal inclinations. Just before the
development of this maniacal condition her lower limbs became
comparatively helpless, and soon after she entirely lost their use. I
found her in this condition, and examination showed no change in
knee-jerk, electrical reactions, nutrition, nor genito-urinary conditions,
which led me to diagnosticate the absence of any organic spinal
trouble. The case was pronounced one of hysterical paralysis, and it
was prophesied that she would eventually completely recover,
probably suddenly. For one year her paralysis remained, her mental
condition varying very greatly during this time—sometimes in a

Numerical Solutions of Three Classes of Nonlinear Parabolic Integro Differential Equations 1St Edition Jangveladze Full Chapter

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

Numerical Solutions of Three Classes of Nonlinear Parabolic Integro Differential Equations 1St Edition Jangveladze Full Chapter

Uploaded by

Copyright:

Available Formats

Numerical Solutions of Three Classes

Copyright © 2016 Elsevier Inc. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any

British Library Cataloguing in Publication Data

Library of Congress Cataloging-in-Publication Data

For information on all Academic Press Publications

This book is concerned with the numerical solutions of some

At the end of the book a list of the quoted literature and

The rst author thanks Fulbright Visiting Scholar Program for

This book is concerned with the numerical solutions of some

Numerical Solutions of Three Classes of Nonlinear Parabolic Integro-Differential Equation.

Key words: Electromagnetic eld penetration, Maxwell’s

Key words: Electromagnetic eld penetration, Maxwell’s

Numerical Solutions of Three Classes of Nonlinear Parabolic Integro-Differential Equation.

In mathematical modeling of applied tasks di erential, inte-

integral equations devoted to distribution of an electrical charge

tem has passed; that is to say, it depends on the history of the

where x = (x1 ; x2 ; x3 ); is a classical three-dimensional Laplace

@ 2 U (x; t) @ 2 U (x; t) @ 2 U (x; t)

where is a known function of its arguments.

where is a known constant. Many authors investigated equa-

where A is a known function of its arguments.

Integro-di erential equations of parabolic type arise in the

where a(S) a0 = const > 0 is again a known function of its

In the works [303], [305], and [306] for conditionally closed

The multidimensional scalar analogues of systems (1.6), (1.7),

This equation serves as a very special model for one-dimens-

on a Hilbert space H with g a nonlinear bounded operator [51].

paid on theorems with proofs for corresponding one-dimensional

and also MS and PhD students of the appropriate specializa-

1.1 Comments and bibliographical

spondingly, are given and studied in [219], [305], and in a num-

As a rule we cannot nd exact solutions of the considered

Numerical Solutions of Three Classes of Nonlinear Parabolic Integro-Differential Equation.

Key words: Mathematical modeling, electromagnetic eld

2.1 Electromagnetic di usion process

If at some moment of time the magnetic eld is solenoidal,

quirement of their belonging to special functional spaces. The

2.1.2 A reduction of system of nonlinear dif-

will be made, is connected to the assumption that the change in

Introducing the function

The functions 0 ; , and cv are positive in physical sense;

If we consider a at eld of the form W = (0; 0; U ), where

of resistance with temperature is characteristic of metals.

a(S) = C1 (C0 + S) 1 + > 0;

From here it is obvious that the function a(S) decreases or

2.2 On the averaged Model II

Following [419] and as in section 2.1, we shall consider the

Here dQ is in ow of a thermal energy absorbed by a body in an

Here we used the known vector identity

Laptev [306] remarked that the research of the so-called aver-

2.3 Mathematical Model III

have suggested a heat ow model based on a memory e ect in

solves the equation

Now de ne a(t) and f (x; t) by

then (2.20) becomes

2.4 Some features of Models I and II

These equations are one-dimensional one-component case of

U (0; t) = U (1; t) = 0; t 2 [0; T ]; (2.24)

where V is any function from the appropriate space (this space

The basic characteristics of equations such as (2.22) is that

a(S) = (1 + S)p ; 0 < p 1;

and f 2 W21 (QT ); then there exists a unique solution U of prob-