Mathematical Modeling 1st Christof Eck Harald

Garcke Peter Knabner download

https://ebookbell.com/product/mathematical-modeling-1st-christof-
eck-harald-garcke-peter-knabner-6984000

Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.

Mathematical Modeling The Life Sciences Numerical Recipes In Python

And Matlab N G Cogan

https://ebookbell.com/product/mathematical-modeling-the-life-sciences-
numerical-recipes-in-python-and-matlab-n-g-cogan-44875372

Mathematical Modeling A Dynamical Systems Approach To Analyze

Practical Problems In Stem Disciplines Antonio Palacios

https://ebookbell.com/product/mathematical-modeling-a-dynamical-
systems-approach-to-analyze-practical-problems-in-stem-disciplines-
antonio-palacios-46197026

Mathematical Modeling For Genes To Collective Cell Dynamics Tetsuji

Tokihiro

https://ebookbell.com/product/mathematical-modeling-for-genes-to-
collective-cell-dynamics-tetsuji-tokihiro-46750844

Mathematical Modeling Of Physical Systems Applications Of Fields

Circuits And Signal Processing Adhir Baran Chattopadhyay

https://ebookbell.com/product/mathematical-modeling-of-physical-
systems-applications-of-fields-circuits-and-signal-processing-adhir-
baran-chattopadhyay-48250562
Mathematical Modeling Stefan Heinz

https://ebookbell.com/product/mathematical-modeling-stefan-
heinz-48654698

Mathematical Modeling In Biology A Research Methods Approach 1st

Edition Shandelle M Henson

https://ebookbell.com/product/mathematical-modeling-in-biology-a-
research-methods-approach-1st-edition-shandelle-m-henson-48713366

Mathematical Modeling For Intelligent Systems Theory Methods And

Simulation Mukesh Kumar Awasthi

https://ebookbell.com/product/mathematical-modeling-for-intelligent-
systems-theory-methods-and-simulation-mukesh-kumar-awasthi-48756144

Mathematical Modeling And Supercomputer Technologies 22nd

International Conference Mmst 2022 Nizhny Novgorod Russia November
1417 2022 Revised Selected Papers Dmitry Balandin

https://ebookbell.com/product/mathematical-modeling-and-supercomputer-
technologies-22nd-international-conference-mmst-2022-nizhny-novgorod-
russia-november-1417-2022-revised-selected-papers-dmitry-
balandin-48771422

Mathematical Modeling And Computation In Finance 1 Cornelis W

Oosterlee

https://ebookbell.com/product/mathematical-modeling-and-computation-
in-finance-1-cornelis-w-oosterlee-49046180
Mathematical Modeling For Epidemiology And Ecology Springer
Undergraduate Texts In Mathematics And Technology 2nd Ed 2nd Edition
Glenn Ledder

https://ebookbell.com/product/mathematical-modeling-for-epidemiology-
and-ecology-springer-undergraduate-texts-in-mathematics-and-
technology-2nd-ed-2nd-edition-glenn-ledder-49351600
Springer Undergraduate Mathematics Series

Christof Eck
Harald Garcke
Peter Knabner

Mathematical
Modeling
Springer Undergraduate Mathematics Series

Advisory Board
M.A.J. Chaplain, University of St. Andrews, St. Andrews, Scotland, UK
A. MacIntyre, Queen Mary University of London, London, England, UK
S. Scott, King’s College London, London, England, UK
N. Snashall, University of Leicester, Leicester, England, UK
E. Süli, University of Oxford, Oxford, England, UK
M.R. Tehranchi, University of Cambridge, Cambridge, England, UK
J.F. Toland, University of Cambridge, Cambridge, England, UK

The Springer Undergraduate Mathematics Series (SUMS) is a series designed for

undergraduates in mathematics and the sciences worldwide. From core foundational
material to final year topics, SUMS books take a fresh and modern approach.
Textual explanations are supported by a wealth of examples, problems and
fully-worked solutions, with particular attention paid to universal areas of difficulty.
These practical and concise texts are designed for a one- or two-semester course but
the self-study approach makes them ideal for independent use.
More information about this series at http://www.springer.com/series/3423
Christof Eck Harald Garcke
•

Peter Knabner

Mathematical Modeling

123
Christof Eck Peter Knabner
Universität Stuttgart Department Mathematik
Stuttgart Universität Erlangen-Nürnberg
Germany Erlangen
Germany
Harald Garcke
Fakultät für Mathematik
Universität Regensburg
Regensburg
Germany

ISSN 1615-2085 ISSN 2197-4144 (electronic)

Springer Undergraduate Mathematics Series
ISBN 978-3-319-55160-9 ISBN 978-3-319-55161-6 (eBook)
DOI 10.1007/978-3-319-55161-6
Library of Congress Control Number: 2017935386

Mathematics Subject Classification (2010): 00A71, 34-01, 35-01, 49-01, 74-01, 76-01, 80-01

© Springer International Publishing AG 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To the memory of
Prof. Dr. Christof Eck
(1968–2011)
Preface

Preface to the English Edition

Encouraged by the positive reception to our German textbook on mathematical

modeling, we were motivated to create an English version of this book. Although
there is a long tradition of English textbooks on mathematical modeling to the best
of our knowledge, most of them differ from the one presented here, as our book is
ordered according to a hierarchy of mathematical subjects with increasing com-
plexity. We start from simple models given by (linear) equations and end up with
complex mathematical models involving nonlinear partial differential equations
with free boundaries. In this way, we offer a variety of mathematically more and
more inclined subjects from which both elementary undergraduate and sophisti-
cated graduate courses can be composed.1 In this sense, we hope that our book fills
a gap and will be of use also in the curricula for mathematical modeling in
English-speaking universities.
We have taken the opportunity of the translation process to correct further
misprints and small inconsistencies which have come to our knowledge.
This English version of this book would not have come into existence without
support of the following people: Mrs. Eva Rütz, Mrs. Astrid Bigott, and
Mrs. Cornelia Weber have created the TEX version on the basis of the first trans-
lation provided by the authors, as usual with enormous exactness, speediness, and
dedication. This first draft has been considerably improved by Prof. Serge Kräutle
and Dr. Kei Fong Lam, who worked through the whole text and removed various
inconsistencies and mistakes not only in the usage of the English language but also
concerning the mathematical contents. We thank Clemens Heine of Springer-
Verlag for his continuous support and encouragement.

1
See the preface to the German edition for more detailed suggestions for the use of this textbook.

vii
viii Preface

It fills us with deep sadness that our scholar, colleague, and friend Christof Eck
could not participate in this enterprise. Extremely untimely, just being 43 years of
age, he passed away after a long illness. This volume is dedicated to his memory.

Regensburg, Germany Harald Garcke

Erlangen, Germany Peter Knabner
July 2016

Preface to the First German Edition

With the notion mathematical modeling, we denote the description of phenomena

from nature, technology, or economy by means of mathematical structures. The aim
of modeling is the derivation of a meaningful mathematical formulation from which
statements and solutions for the original problem can be derived. Here, in principle,
every branch of mathematics is “applicable.” In technological applications, quite
often mathematical problems arise with such complexity that they have to be
simplified by neglecting certain influences or they can only be solved with the help
of numerical methods.
If one simplifies mathematical models, for example by neglecting certain “small”
terms in an equation, then it is the task of mathematicians to check to which extent
the behavior of solutions has been changed. Historically, a variety of mathematical
concepts that have been developed were driven by the requirements of applications.
Having this in mind, it is neither by chance nor astonishing that mathematics and
real world “fit together.” Mathematical concepts whose development has been
triggered from a specific field of application often also proved to be usable in other
fields of applications. Furthermore, for a specific problem from applications, there
are in general a wide range of mathematical models feasible. The choice of the
model certainly depends on the degree of detail in the (temporal or spatial) level of
consideration.
Considering this, an education in mathematical modeling appears to be indis-
pensable, if a study in mathematics is also supposed to contain professional qual-
ifications. Often the classical courses of a mathematics curriculum can deliver such
an education only insufficiently. Typical courses concentrate on the development of
a mathematical theory, in which sample applications—if at all—only play the role
of an “optional” motivation. Also, the study in a minor subject from an application
field in general cannot close the gap because students in mathematics are often
overburdened: On the one hand, they have to learn the results of an application
field, and on the other hand, they have to extract mathematical structures from them
to make mathematical knowledge useful. Against this background, specific courses
in mathematical modeling have entered the curriculum of mathematical studies at
various universities. In this context, the present textbook wants to be helpful.
On the one hand, this textbook presents knowledge from an area “between” math-
ematics and the sciences (e.g., from thermodynamics and from continuum mechanics)
Preface ix

which students and lecturers of mathematics need in order to understand models for
problems in the sciences and engineering and also to derive them. On the other hand,
this book contains a variety of interesting, practically relevant examples for the
mathematical theories often only experienced at an abstract level during the study of
mathematics and thus answers the question often posed “what do I need this for?”.
While it cannot substitute any of the textbooks dealing with the underlying mathe-
matical structures such as linear systems of equations/linear algebra or ordinary or
partial differential equations, it nevertheless contains essential aspects of the analysis
of the models. One aim in particular is to illustrate the interactions between mathematics
and applications, which unfortunately are often neglected in mathematics courses.
Furthermore, this textbook also addresses students from the sciences and from
engineering and offers them an introduction into the methods of applied mathe-
matics and mechanics.
The content of this book is restricted to deterministic models with continuous
scales, as they are in the center of classical natural sciences and engineering. In
particular, stochastic models are beyond the scope of this book, and the same
applies to processes at very small scales for which particular models or models from
quantum mechanics and its approximations are feasible. Also, models from eco-
nomics are not in the focus of this book, as stochastic approaches play an important
role there.
An essential concept of this book consists in using the mathematical structures
(and the knowledge about them) as an ordering principle and not the different fields
of applications. This reflects the strength of mathematics, lying in the fact that one
concept can be used for totally different problem classes and fields of applications.
It allows dealing with examples from different fields of applications efficiently
without being forced to always repeat the same mathematical basic structures: This
line will be followed in Chapter 1 and in Chapters 2, 4, 6, and 7. In this order, one
finds embedded Chapters 3 (thermodynamics) and 5 (continuum mechanics). They
provide the necessary links to the natural sciences and engineering. Of course, these
chapters are also shaped by the application of mathematical tools.
The restriction of the subjects at the level of application corresponds to a
restriction at a mathematical level: Throughout this book, we use knowledge from
linear algebra and analysis intensively. Chapter 4 relies on the knowledge provided
by courses in analysis or in ordinary differential equations, and Chapter 5 makes use
in an essential way of the methods of the multidimensional differentiation and
integration (integral theorems) and in this way from the more advanced aspects of
analysis. In Chapter 7, the foundations of the geometry of curves and surfaces play
an important role. It is impossible to define a clear delimitation to the analysis of
partial differential equations. Knowledge from this field and also from linear
functional analysis certainly is useful for Chapter 6, but not necessary.
A discussion of mathematical results about partial differential equations takes
place in this chapter only insofar as there is a tight linkage to the model interpre-
tation. Therefore, the presentation cannot be completely rigorous, but possible gaps
and necessary consolidations are pointed out. In this way, this chapter does not
necessarily require an intense study of the analysis of partial differential equations,
x Preface

but hopefully, it stimulates to do so. There is no chapter dedicated solely to opti-

mization, but concepts of optimization are used in different chapters. This reflects
that optimization problems in modeling often appear as an equivalent formulation
of other mathematical problems when expressed in a variational formulation. We
have totally excluded the treatment of numerical methods from this book, although
they represent a central tool in the practical treatment of technical and scientific
problems. We think that this can be justified as there is already a vast amount of
excellent textbooks available.
The material of this book can be used in various ways for undergraduate and for
graduate students, both at an elementary level and at a more advanced level. The
simplest usages are two lecture courses of four hours each week, which cover the
whole spectrum of this book. Here, the first part should be placed in the second half
of undergraduate studies and the second part in the graduate studies. Alternatively,
a course with two hours of lecture per week of introductory character is also
possible at an early stage of graduate studies based on parts of Chapters 1, 2, and 4.
This can be further complemented by another two-hour-per-week course consisting
of parts of Chapters 5 and 6. If there is only room for one course, then it is also
possible to build a course of four hours of lecture per week out of parts of Chapters
1, 2, 3, 4, 5, and some aspects of 6. Alternatively, Chapters 5, 6, and 7 can be used
for a course about mathematical models and continuum mechanics or from the basic
sections of Chapter 5 (derivation of the conservation equations) and Chapters 6 and
7, a course on “applied partial differential equation” can be extracted. Furthermore,
this book can be used efficiently for self-study, at undergraduate, graduate, or
postgraduate level. Postgraduates working with a mathematical model related to
certain fields of application may find the necessary foundations here.
The aim of all the courses described can only be to close a certain gap to the
application fields, and they cannot substitute the concrete realization of modeling
projects. Intensive treatment of many exercises presented here is helpful, but finally,
to our understanding, modeling can only be learned by modeling practice.
A possible teaching concept as it has been used by the authors consists of problem
seminars in which tasks from the applications are posed without any mathematical
material, i.e., the development of feasible mathematical concepts is an essential part
of the work. But we hope that courses based on this textbook provide an essential
foundation for such a modeling practice.
Finally, this book seems to be also suitable for students of the natural sciences
(physics, chemistry) and of engineering. For these students, some of the modeling
aspects will be familiar from specific cases, but a rigorous inclusion in a mathe-
matical methodology should lead to further insight.
Of course, we are not going to judge the existing textbooks on mathematical
modeling. Certainly, there is a variety of excellent books, but in quantity much less
than in other mathematical fields, and many textbooks restrict themselves to an
elementary level, in particular also addressing high school students. A textbook
which tries to cover the whole spectrum from elementary aspects to recent research
is not known to us. Often, textbooks also follow different ordering principles.
The present textbook originated from courses which were given by the second
Preface xi

author at the University of Regensburg and by the first and third author at the
University of Erlangen several times, and therefore, it is the result of a complex
developing process. During this process, the authors received important support.
The authors express their thanks to Bernd Ammann, Luise Blank, Wolfgang
Dreyer, Michael Hinze, and Willi Merz for valuable suggestions. Sincere thanks are
given to Barbara Niethammer, who together with the second author has lectured a
course on mathematical modeling at the University of Bonn from which much
material entered the present book. For careful proofreading, we thank Martin Butz,
Daniel Depner, Günther Grün, Robert Haas, Simon Jörres, Fabian Klingbeil, David
Kwak, Boris Nowak, Andre Oppitz, Alexander Prechtel, and Björn Stinner. In TEX
writing, we were supported by Mrs. Silke Berghof and in particular by Mrs. Eva
Rütz who typed a large part of the manuscript and worked on the numerous figures
with strong dedication—to both our cordial thanks. We thank Serge Kräutle who
provided the figure at the cover—a numerical simulation of the Kármán vortex
street. We would also like to thank cordially Ulrich Weikard for providing
Figure 6.14 and James D. Murray for providing Figure 6.10.

Bielefeld, Germany Christof Eck

Regensburg, Germany Harald Garcke
Erlangen, Germany Peter Knabner
December 2007
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 1
1.1 What Do We Mean by (Mathematical) Modeling? . . . . . . . . .... 1
1.2 Aspects of Mathematical Modeling: Example of Population
Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Population Models with Restricted Resources. . . . . . . . . . . . . . . . 8
1.4 Dimensional Analysis and Scaling . . . . . . . . . . . . . . . . . . . . . . . . 12
1.5 Asymptotic Expansions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.6 Applications from Fluid Mechanics . . . . . . . . . . . . . . . . . . . . . . . 27
1.7 Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
1.8 Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2 Systems of Linear Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.1 Electrical Networks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.2 Space Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.3 Constrained Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
2.4 Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.5 Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
3 Basic Principles of Thermodynamics . . . . . . . . . . . . . . . . . . . . . . .... 75
3.1 The Model of an Ideal Gas and the Maxwell–Boltzmann
Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 76
3.2 Thermodynamic Systems and the Thermodynamic
Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 80
3.3 The First Law of Thermodynamics . . . . . . . . . . . . . . . . . . . . .... 81
3.4 The Second Law of Thermodynamics and the Notion
of Entropy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.5 Thermodynamic Potentials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
3.6 The Legendre Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
3.7 The Calculus of Differential Forms . . . . . . . . . . . . . . . . . . . . . . . . 99
3.8 Thermodynamics of Mixtures and the Chemical Potential . . . . . . 102

xiii
xiv Contents

3.9 Chemical Reactions in Multi Species Systems . . . . . . . . . . . . . . . 109

3.10 Equilibria of Chemical Reactions and the Mass Action Law . . . . 114
3.11 Kinetic Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
3.12 Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
3.13 Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
4 Ordinary Differential Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
4.1 One-Dimensional Oscillations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
4.1.1 Forced Oscillations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
4.2 The Lagrangian and Hamiltonian Form of Mechanics . . . . . . . . . 141
4.3 Examples from Population Dynamics . . . . . . . . . . . . . . . . . . . . . . 152
4.4 Qualitative Analysis, Phase Portraits . . . . . . . . . . . . . . . . . . . . . . . 155
4.5 The Principle of Linearized Stability . . . . . . . . . . . . . . . . . . . . . . 160
4.6 Stability of Linear Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
4.7 Variational Problems for Functions of One Variable . . . . . . . . . . 167
4.8 Optimal Control with Ordinary Differential Equations . . . . . . . . . 183
4.9 Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
4.10 Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
5 Continuum Mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
5.2 Classical Point Mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
5.3 From Particle Mechanics to a Continuous Medium . . . . . . . . . . . 203
5.4 Kinematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
5.5 Conservation Laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
5.6 Constitutive Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
5.7 The Second Law of Thermodynamics
in Continuum Mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
5.8 Principle of Frame Indifference . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
5.9 Constitutive Theory for Viscous Fluids . . . . . . . . . . . . . . . . . . . . 247
5.10 Modeling of Elastic Solids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
5.11 Electromagnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
5.12 Dispersion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
5.13 Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
5.14 Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
6 Partial Differential Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
6.1 Elliptic Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
6.1.1 Calculus of Variations . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
6.1.2 The Fundamental Solution . . . . . . . . . . . . . . . . . . . . . . . . 314
6.1.3 Mean Value Theorem and Maximum Principles . . . . . . . 317
6.1.4 Plane Potential Flows, the Method of Complex
Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 319
6.1.5 The Stokes Equations . . . . . . . . . . . . . . . . . . . . . . . .... 325
6.1.6 Homogenization . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 327
Contents xv

6.1.7 Optimal Control of Elliptic Differential Equations . . . . . . 332

6.1.8 Parameter Identification and Inverse Problems . . . . . . . . 336
6.1.9 Linear Elasticity Theory . . . . . . . . . . . . . . . . . . . . . . . . . 340
6.2 Parabolic Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
6.2.1 Uniqueness of Solutions, the Energy Method . . . . . . . . . 345
6.2.2 Large Time Behavior. . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
6.2.3 Separation of Variables and Eigenfunctions. . . . . . . . . . . 351
6.2.4 The Maximum Principle . . . . . . . . . . . . . . . . . . . . . . . . . 354
6.2.5 The Fundamental Solution . . . . . . . . . . . . . . . . . . . . . . . . 355
6.2.6 Diffusion Times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
6.2.7 Invariant Transformations . . . . . . . . . . . . . . . . . . . . . . . . 360
6.2.8 General Initial Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
6.2.9 Brownian Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
6.2.10 Traveling Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
6.2.11 Reaction Diffusion Equations and Traveling Waves . . . . 367
6.2.12 Turing Instability and Pattern Formation . . . . . . . . . . . . . 374
6.2.13 Cahn–Hilliard Equation and Pattern Formation
Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
6.3 Hyperbolic Conservation Laws . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
6.4 The Wave Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
6.5 The Navier–Stokes Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
6.6 Boundary Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405
6.7 Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
6.8 Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
7 Free Boundary Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
7.1 Obstacle Problems and Contact Problems . . . . . . . . . . . . . . . . . . . 428
7.2 Free Boundaries in Porous Media . . . . . . . . . . . . . . . . . . . . . . . . . 435
7.3 The Stefan Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446
7.4 Entropy Inequality for the Stefan Problem . . . . . . . . . . . . . . . . . . 453
7.5 Undercooled Liquids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454
7.6 Gibbs–Thomson Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456
7.7 Mullins–Sekerka Instability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
7.8 A Priori Estimates for the Stefan Problem with Gibbs–Thomson
Condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461
7.9 Phase Field Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464
7.10 Free Surfaces in Fluid Mechanics . . . . . . . . . . . . . . . . . . . . . . . . . 472
7.11 Thin Films and Lubrication Approximation . . . . . . . . . . . . . . . . . 476
7.12 Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479
7.13 Exercises. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
Appendix A: Function Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489
Appendix B: Curvature of Hypersurfaces . . . . . . . . . . . . . . . . . . . . . . . . . 493
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
Chapter 1
Introduction

1.1 What Do We Mean by (Mathematical) Modeling?

With (mathematical) modeling we denote the translation of a specific problem from

the natural sciences (experimental physics, chemistry, biology, geosciences) or the
social sciences, or from technology, into a well-defined mathematical problem. The
mathematical problem may range in complexity from a single equation to a system
of several equations, to an ordinary or partial differential equation or a system of
such equations, to an optimization problem, where the state is described by one
of the aforementioned equations. In more complicated cases we can also have a
combination of the problems mentioned. A mathematical problem is well-posed, if
it has a unique solution and if the solution of the problem depends continuously
on its data, where continuity has to be measured in such a way that the results are
meaningful for the application problem in mind. In general the phenomena to be
described are very complex and it is not possible or sensible to take all its aspects
into account in the process of modeling, because for example
• not all the necessary data are known,
• the model thus achieved cannot be solved anymore, meaning that its (numerical)
solution is expensive and time consuming, or it is not possible to show the well-
posedness of the model.
Therefore nearly every model is based on simplifications and modeling assumptions.
Typically the influence of unknown data are neglected, or only taken into account
in an approximative fashion. Usually complex effects with only minor influences on
the solution are neglected or strongly simplified. For example if the task consists
of the computation of the ballistic trajectory of a soccer ball then it is sensible to
use classical Newtonian mechanics without taking into account relativity theory. In
principle using the latter one would be more precise, but the difference in results for
a typical velocity of a soccer ball is negligible. In particular this holds true if one
takes into account that there are errors in the data, for example slight variations in the
size, the weight, and the kickoff velocity of the soccer ball. Typically available data

© Springer International Publishing AG 2017 1

C. Eck et al., Mathematical Modeling, Springer Undergraduate
Mathematics Series, DOI 10.1007/978-3-319-55161-6_1
2 1 Introduction

are measured and therefore afflicted with measurement errors. Furthermore in this
example certainly the gravitational force of the Earth has to be taken into account, but
its dependence on the flight altitude can be neglected. In a similar way the influence
of the rotation of the Earth can be neglected. On the other hand the influence of air
resistance cannot be neglected. The negligible effects are exactly those which make
the model equations more complex and require additional data, but do not improve
the accuracy of the results significantly.
In deriving a model one should make oneself clear what is the question to be
answered and which effects are of importance and have to be taken into account
in any case and which effects are possibly negligible. The aim of the modeling
therefore plays a decisive role. For example the model assumptions mentioned above
are sensible for the flight trajectory of a soccer ball, but certainly not for the flight
trajectory of a rocket in an orbit around the Earth. Another aspect shows the following
example from weather forecasting: An exact model to compute the future weather for
the next seven days from the data of today cannot serve for the purposes of weather
forecast if the numerical solution of this model would need nine days of computing
time of the strongest available supercomputer. Therefore often a balance between the
accuracy required for the predictions of a model and the costs to achieve a solution
is necessary. The costs can be measured for example by the time which is necessary
to achieve a solution of the model and for numerical solutions also by the necessary
computer capacities. Thus at least in industrial applications costs often mean financial
costs. Because of these reasons there can be no clear separation between correct or
false models, a given model can be sensible for certain applications and aims but not
sensible for others.
An important question in the construction of models is: Does the mathematical
structure of a model change by neglecting certain terms? For example in the initial
value problem
ε y  (x) + y(x) = 0 , y(0) = 1

with the small parameter ε one could think about omitting the term εy  . However,
this would lead to an obviously unsolvable algebraic system of equations

y(x) = 0 , y(0) = 1 .

The term neglected is decisive for the mathematical structure of the problem indepen-
dent of the smallness of parameter ε. Therefore sometimes terms which are identified
as small, cannot be neglected. Hence, constructing a good mathematical model also
means to take aspects of analysis (well-posedness) and numerics (costs) of the model
into account.
The essential ingredients of a mathematical model are

• an application problem to be described,

• a number of model assumptions,
1.1 What Do We Mean by (Mathematical) Modeling? 3

• a mathematical problem formulation, for example in the form of a mathematical

relation, specifically an equation, an inequality, or differential equation, or several
coupled relations, or an optimization problem.
The knowledge of the model assumptions is of importance to estimate the scope
of applications and the accuracy of the predictions of a model. The aim of a good
model is, starting from known but probably only estimated data and accepted laws
of nature to give an answer as good as possible for a given question in an application
field. A sensible model should only need data which are known or for which at least
plausible approximations can be used. Therefore the task consists in extracting as
much as possible information from known data.

1.2 Aspects of Mathematical Modeling: Example

of Population Dynamics

To illustrate some important aspects of modeling in this section we consider a very

simple example: A farmer has a herd of 200 cattle and he wants to increase this herd
to 500 cattle, but only by natural growth, i.e., without buying additional animals.
After a year the cattle herd has grown to 230 animals. He wants to estimate how long
it lasts till he has reached his goal.
A sensible modeling assumption is the statement that the growth of the population
depends on the size of the population, as a population of the double size should also
have twice as much offspring. The data available are
• the initial number x(t0 ) = 200 of animals at the initial time t0 ,
• the increment in time Δt = 1 year,
• a growth factor of r = 230/200 = 1.15 per animal and per time increment Δt.
If one sets tn = t0 + nΔt and if x(t) denotes the number of animals at time t, then
knowing the growth factor leads to the recursion formula

x(tn+1 ) = r x(tn ) . (1.1)

From this recursion formula one gets

x(tn ) = r n x(t0 ) .

Therefore the question can be formulated as:

Find a number n such that x(tn ) = 500.
The solution is
 
500
  ln
x(tn ) 200
n ln(r ) = ln , or n = ≈ 6.6.
x(t0 ) ln(1.15)
4 1 Introduction

Hence, the farmer has to wait for 6.6 years.

This is a simple population model which in principle can also be applied to other
problems from biology, for example the growth of other animal population, of plants
or bacteria. But it can also be used in apparently totally different fields of applications,
for example the computation of interests or the cooling of bodies (see Exercises 1.1
and 1.2). Without possibly noticing, in deriving the above model we have used several
important modeling assumptions, which are fulfilled sometimes, but which are not
fulfilled in a lot of cases. In particular the influence of the following effects has been
neglected:
• the spatial distribution of the population,
• limited resources, for example limited nutrients,
• a loss of population by natural enemies.
Further details which also have been neglected, are for example the age distribution
in the population, which has influence on the death rate and the birth rate, and the
subdivision in female and male animals. Additionally the model leads to non-integer
population quantities, which strictly speaking is not correct. The simplifications and
deficiencies do not render the model worthless but they have to be recognized and
taken into account to assess the result correctly. In particular the specific result of
6.6 years should not be taken too seriously, and an appropriate interpretation rather
is that the farmer presumably will reach his goal in the 7th year.
An aspect, which is not optimal for intrinsic mathematical reasons, is the time
increment of one year, because it is chosen arbitrarily. For the application under
consideration it has a sensible meaning, nevertheless also an increment of three
months or of two years could have been chosen. Furthermore we need two data, the
increment in time and the growth rate. Both data depend on each other, meaning
that the growth characteristics possibly can only be described by one number. As a
first approach one can conjecture that the growth rate depends linearly on the time
increment, i.e.,
r = 1 + Δt p

with a factor p still unknown. From r = 1.15 for Δt = 1 year we conclude that p =
0.15/year. Taking this for granted then for Δt = 2 years one has r = 1.3. Therefore
after 6 years, (6 = 3 times 2) the farmer has

200 · 1.33 = 439.4

cattle. But in the “old” model with Δt = 1 year he has

200 · 1.156 ≈ 462.61

animals. Therefore the assumption of a linear relation between r and Δt is wrong.

A better approach can be gained by the limiting process Δt → 0:

1.2 Aspects of Mathematical Modeling: Example of Population Dynamics 5

x(t + Δt) ≈ (1 + Δt p) x(t) for “small” Δt ,

or more precisely
x(t + Δt) − x(t)
lim = p x(t) ,
Δt→0 Δt

i.e.,
x  (t) = p x(t) . (1.2)

This is a continuous model in the form of an ordinary differential equation, which

does not contain an arbitrarily chosen time increment anymore. It possesses the exact
solution
x(t) = x(t0 ) e p(t−t0 ) .

If the data are as above, i.e., a time increment of Δt = 1 year and a growth rate
r = 1.15, this means
e p·1 year = 1.15

and therefore
p = ln(1.15)/year ≈ 0.1398/year .

This is a continuous exponent of growth.

The discrete model (1.1) can be perceived as a special numerical discretization
of the continuous model. An application of the explicit Euler method with time step
Δt to (1.2) leads to

x(ti+1 ) = x(ti ) + Δt p x(ti ) , or x(ti+1 ) = (1 + Δt p) x(ti ) ,

this is (1.1) with r = 1 + Δt p. In the case p < 0 a time increment of the size
Δt < (− p)−1 has to be chosen to achieve a sensible sequence of numbers. On the
other hand using the implicit Euler method one gets

x(ti+1 ) = x(ti ) + Δt p x(ti+1 ) , or x(ti+1 ) = (1 − Δt p)−1 x(ti ) ,

i.e., (1.1) with r = (1 − Δt p)−1 . Here for p > 0 the time step has to be chosen such
that Δt < p −1 . By Taylor series expansion one can see that  the different growth
factors coincide for small Δt “up to an error of the order O (Δt)2 ”:
 
(1 − Δt p)−1 = 1 + Δt p + O (Δt p)2 .

The connection between the continuous and the discrete model therefore can be
established by an analysis of the convergence properties of the numerical method.
For the (explicit or implicit) Euler method one gets for example

|x(ti ) − xi | ≤ C(te ) Δt ,
6 1 Introduction

where x(ti ) is the exact solution of (1.2) at time ti and xi is the approximate solution
of the numerical method, assuming that ti ≤ te , where te is the given final time for the
model. For details about the analysis of numerical methods for ordinary differential
equations we refer to the textbooks of Stoer and Bulirsch [123] and Deuflhard and
Bornemann [28].
Both models, the discrete and the continuous, have the seeming disadvantage
that they also allow non-integer solutions, which obviously are not realistic for the
considered example. The model describes — as it is true for every other model —
not the total reality but only leads to an idealized picture. For small populations
the model is not very precise, as in general population growth also depends heavily
on stochastic effects and therefore cannot be computed precisely in a deterministic
way. In addition for small populations the model assumptions are questionable, in
particular one neglects the age and the sex of the animals. In the extreme case of a
herd of two animals obviously the growth will depend heavily on the fact whether
there is a male and a female animal, or not. For large populations on the other hand
one can assume with a certain qualification that it possesses a characteristic uniform
distribution in age and in sex, such that the assumption of a growth proportional to
population size make sense.
The substitution of integer values by real numbers reflects the inaccuracy of the
model. Therefore it is not sensible to change the model such that integer values in
the solutions are enforced. This would only lead to an unrealistic perception of high
accuracy of the model. For a small population a stochastic model, which then “only”
provides statements about the probability distribution of the population size, makes
sense instead of deterministic models.
Nondimensionalization
The quantities in a mathematical model generally have a physical dimension. In the
population model (1.2) we have the units number and time. We denote the physical
dimension of a quantity f with [ f ] and abbreviate the units number of entities by A
and time by T . Therefore we have

[t] = T ,
[x(t)] = A ,
A
[x  (t)] = ,
T
1
[ p] = .
T
The specification of a physical dimension is not yet a decision about the physical unit
of measurements. As a unit of measurement for time one can use seconds, minutes,
hours, days, weeks, or years, for example. If we measure time in years, then t is
indicated in years, x(t) by a number, x  (t) in number/years and p in number/years.
To get models as simple as possible and furthermore in order to determine char-
acteristic quantities in a model, one can nondimensionalize the model equations.
1.2 Aspects of Mathematical Modeling: Example of Population Dynamics 7

For this aim one defines a characteristic value for every appearing dimension and
correspondingly a unit of measurement. Here it is not necessary to choose one of the
common units as for example seconds or hours but it is more appropriate to choose
a unit adapted to the problem. For the population model there are two dimensions,
therefore two characteristic values are needed, the characteristic number x and the
characteristic time t. These are chosen in such a way that the initial data t0 and
x0 = x(t0 ) are as simple as possible. Therefore a convenient unit of measurement
for time is given by
t − t0
τ= ,
t

where t denotes a unit of time which still has to be specified, and as a unit for number
we choose
x = x0 .

Setting
x
y=
x
and expressing y as a function of τ ,

x(tτ + t0 )
y(τ ) = ,
x
one obtains
t 
y  (τ ) = x (t)
x
and therefore the model becomes
x 
y (τ ) = p x y(τ ) .
t
This model gets its most simple form for the choice

1
t= . (1.3)
p

The model thus derived is the initial value problem

y  (τ ) = y(τ ) ,
y(0) = 1 . (1.4)

This model has the solution

y(τ ) = eτ .
8 1 Introduction

From this solution all solutions of the original model (1.2) can be achieved by using
a transformation:

x(t) = x y(τ ) = x0 y( p(t − t0 )) = x0 e p(t−t0 ) .

The advantage of the nondimensionalization therefore is the reduction of the solution

of all population models of a given type by the choice of units to one single problem.
Note that this holds true independent of the sign of p, although the behavior of the
solutions for p > 0 and p < 0 is different. For p < 0 the solution of (1.2) is given
by the solution (1.4) for the range τ < 0.
The scaling condition (1.3) also can be obtained by means of dimensional analysis.
In this procedure the characteristic time t to be determined is expressed as a product
of the other characteristic parameters in the model,

t = p n x0m with n, m ∈ Z .

By computing the dimension one obtains

 n
1
[t] = [ p] [x0 ]
n m
and therefore T = Am .
T

The only possible solution of this equation is given by n = −1, m = 0, if the number
of animals is interpreted as a dimension of its own. Thus we get exactly (1.3).
In more complex models typically the model cannot be reduced to a single problem
by nondimensionalization but the number of relevant parameters can be strongly
reduced and the characteristic parameters can be identified. This also relates to the
corresponding experiments: For instance, from the nondimensionalization of the
equations for airflows one can conclude how the circulation around an airplane can be
experimentally measured by using a (physical) model for the airplane much smaller
in scale. We will explain dimensional analysis in one of the following sections using
a more meaningful example.

1.3 Population Models with Restricted Resources

For large populations in nature a constant growth rate is not realistic anymore. A
restriction of the habitat, or the available nutrients, or other mechanisms impose
limitations on the growth. To construct a model it is feasible for such situations to
assume that there is a certain capacity x M > 0 for which the resources of the habitat
are still sufficient. For population quantities x smaller than x M the population still
can grow, but for values larger than x M the population decreases. This means that
the growth rate p now depends on the population x, p = p(x), and that
1.3 Population Models with Restricted Resources 9

p(x) > 0 for 0 < p < x M ,

p(x) < 0 for p > x M

have to hold true. The most simple functional form satisfying these conditions is
given by a linear ansatz for p, i.e.,

p(x) = q(x M − x) for all x ∈ R

with a parameter q > 0. With this ansatz we obtain the differential equation

x  (t) = q x M x(t) − q x(t)2 (1.5)

as a model. The additional term −q x(t)2 is proportional to the probability for the
number of encounters of two specimens of the population per unit of time. The
term represents the more competitive situation if the population size increases, the
so-called “social friction”. The Eq. (1.5) has been proposed by the Dutch biomath-
ematician Verhulst and is called logistic differential equation or equation of limited
growth.
Equation (1.5) also can be solved in closed form (compare Exercise 1.3). From

x
=q
x(x M − x)

we conclude using the partial fraction decomposition

 
1 1 1 1
= +
x(x M − x) xM x xM − x

and by integration

ln(x(t)) − ln |x M − x(t)| = x M qt + c1 , c1 ∈ R .

After the choice of an appropriate constant c2 ∈ R we obtain

x(t)
= c2 e x M qt ,
x M − x(t)

and
c2 x M e x M qt xM
x(t) = = .
1 + c2 e x M qt 1 + c3 e−x M qt

Incorporating the initial condition x(t0 ) = x0 we obtain

x M x0
x(t) = . (1.6)
x0 + (x M − x0 )e−x M q(t−t0 )
10 1 Introduction

From this exact solution the following properties can be easily derived:
• If x0 is positive, the solution always stays positive.
• If x0 is positive, then for t → +∞ the solution converges to the equilibrium point
x∞ = x M .
The graph of x can be sketched also without knowing the exact solution. From (1.5)
first we conclude

x  > 0 , if x < x M ,
x  < 0 , if x > x M .

Furthermore we have

x  = (x  ) = (q (x M − x) x) = q(x M − x) x  − q x x 
= q(x M − 2x)x  = q 2 (x M − 2x)(x M − x) x .

From these results we conclude

x  > 0 , if x ∈ (0, x M /2) ∪ (x M , ∞) ,

x  < 0 , if x M /2 < x < x M .

Thus the solution curves have an inflection point at x M /2 and the curves are concave
in the interval between x M /2 and x M , and convex otherwise. Solutions of the logistic
differential equation are depicted in Fig. 1.1.
Stationary Solutions
For more complex time-depending models a closed form solution often cannot be
found. Then it is useful to identify time independent solutions. Such solutions can
be computed using the time dependent model by just setting all time derivatives to
zero. For our model with restricted growth one gets

0 = q xM x − q x2 .

xM

xM
2

Fig. 1.1 Solutions of the logistic differential equation

1.3 Population Models with Restricted Resources 11

This equation has the two solutions

x0 = 0 and x1 = x M .

These are the solutions of the original model for specific initial data. Often time
independent solutions appear as so-called stationary limits of arbitrary solutions
for large times, meaning that they are solutions constant in time towards which
time dependent solutions converge for large times. Typically this only appears if the
stationary solution is stable in the following sense: If the initial data is only changed
slightly then also the solution changes only slightly. Using the exact solution (1.6)
the question of stability can be easily answered for the logistic differential equation:
The solution for the initial value
x(t0 ) = ε

with a small ε > 0 is given by

xM ε
xε (t) = ,
ε + (x M − ε)e−x M q(t−t0 )

it converges for t → +∞ towards x M , therefore the stationary solution x0 = 0 is not

stable. For
x(t0 ) = x M + ε

with a small ε = 0 the solution is given by

x M (x M + ε)
xε (t) = ,
(x M + ε) − ε e−x M q(t−t0 )

it converges for t → +∞ towards x M . From

xε (t) = q xε (t)(x M − xε (t))

one can conclude also without knowing the exact solution that the distance to x M
can only decrease for increasing time as from xε (t) > x M it follows xε (t) < 0 and
from xε (t) < x M it follows xε (t) > 0. Therefore the stationary solution x M is stable.
Stability is of importance, as in nature in general no instable stationary solution can
be observed, therefore they are irrelevant for most practical applications. For more
complex models sometimes no closed form solution for the time dependent equation
can be derived. However, there are techniques of stability analysis, with which often
the stability properties of stationary solutions can be deduced. Often this is done
by means of a linearization of the problem at the stationary solution followed by a
computation of the eigenvalues of the linearized problem. This will be explained in
more detail in Chap. 4.
12 1 Introduction

1.4 Dimensional Analysis and Scaling

Now we want to explain dimensional analysis using a slightly more significant

example. We consider a body of mass m, which is thrown bottom-up in vertical
direction with respect to the gravitational field of a planet (for example the Earth).
The motion of the body is described by Newton’s law

F
a= ,
m
where a denotes the acceleration of the body and F the force acting on the body.
This force is described by Newton’s law of gravitation
mE m
F = −G ,
(x + R)2

where G ≈ 6.674 · 10−11 N · m2 /kg2 denotes the gravitational constant, m E the mass
of the planet, R the radius of the planet and x the height of the body, measured from
the surface of the planet. We neglect the air resistance in the atmosphere and consider
the planet to be a sphere. If one defines the constant g by

Gm E
g= ,
R2
one gets
g R2m
F =− .
(x + R)2

For the Earth we have g = 9.80665 m/s2 , the gravitational acceleration. The motion
of the body then is described by the differential equation

g R2
x  (t) = − . (1.7)
(x(t) + R)2

This has to be completed by two initial conditions,

x(0) = 0 , x  (0) = v 0 ,

where v 0 denotes the initial velocity.

For trajectories expected in our application typically the term x(t) is very small
compared to the radius of the Earth and seems to be negligible in the denominator
in the right-hand side in (1.7). We want to investigate the validity of this ansatz in a
systematic fashion. To do so first we perform a nondimensionalization. As a specific
example we use the data
1.4 Dimensional Analysis and Scaling 13

g = 10 m/s2 , R = 107 m , and v 0 = 10 m/s ,

which have an order of magnitude corresponding to the application in mind.

The dimensions appearing here are L for the length and T for time. The given data
are the initial velocity v 0 with dimension [v 0 ] = L/T , the “planet acceleration” g with
dimension [g] = L/T 2 , and the radius R with dimension [R] = L. The independent
variable is the time t with dimension [t] = T , and the quantity to be computed is the
height x with dimension [x] = L. First we look for all representations of the form

Π = v 0a g b R c ,

which are either dimensionless (case (i)), or have the dimension of a length (case
(ii)), or have the dimension of a time (case (iii)). From
 a  b
L L
[Π ] = L c = L a+b+c T −a−2b
T T2

it follows:

Case (i): We have a + b + c = 0, −a − 2b = 0, therefore a = −2b, c = b, and

finally
 b
gR
Π= .
v 02

This leads to the identification of

v 02
ε= (1.8)
gR

as a characteristic dimensionless parameter. In fact all other dimension-

less parameters are powers of this specific one.
Case (ii): We have a + b + c = 1, a + 2b = 0 and therefore a = −2b,
c = 1 + b. As a specific unit for length one obtains

 = v 0−2b g b R 1+b = R ε−b ,

where b denotes a constant not yet specified.

Case (iii): We have a + b + c = 0, a + 2b = −1 and therefore a = −1 − 2b,
c = b + 1. Therefore a characteristic unit for time is given by

R −b
τ = v 0−1−2b g b R b+1 = ε .
v0

We will now try to nondimensionalize Eq. (1.7). To this purpose we consider a unit
for length x and a unit for time t and represent x(t) in the form
14 1 Introduction

x(t) = x y(t/t) .

From (1.7) we obtain

x g R2
y  (τ ) = − ,
t
2 (x y(τ ) + R)2

i.e.,
x 1
y  (τ ) = − . (1.9)
2
t g ((x/R) y(τ ) + 1)2

This equation has to be provided with the initial conditions

t
y(0) = 0 and y  (0) = v0 .
x

Now we want to choose x and t in such a way that as many of the appearing parameters
as possible equal 1. However, here we have more parameters than scaling units,
namely the three parameters

x x t
2
, , and v0 .
t g R x

Hence, only two of these parameters can be transformed to one and therefore there
are three different possibilities:

x x R
(a) 2 = 1 and = 1 are a consequence of x = R, t = , then the parameter
t g R g
t v0 √
is given by v 0 = √ = ε using ε from (1.8). Therefore the model reduces
x Rg
to
1 √
y  (τ ) = − , y(0) = 0 , y  (0) = ε . (1.10)
(y(τ ) + 1) 2

x t R x
(b) = 1 and v 0 = 1 can be deduced from x = R and t = , leading to 2 =
R x v0 t g
v 02
= ε for the third parameter. Then the dimensionless model is given by
Rg

1
ε y  (τ ) = − , y(0) = 0 , y  (0) = 1 .
(y(τ ) + 1)2

x t v0 v2
(c) = 1 and v 0 = 1 are a consequence of t = and x = 0 . Then the third
2
t g x g g
x v 02
parameter is given by = = ε. Thus the dimensionless model reads
R gR
1.4 Dimensional Analysis and Scaling 15

1
y  (τ ) = − , y(0) = 0 , y  (0) = 1 . (1.11)
(ε y(τ ) + 1)2

Let us mention that there is a fourth possibility. Besides (1.9) we may also use the
equivalent formulation

x3 1
y  (τ ) = − .
t
2
g R2 (y(τ ) + R/x)2

This leads, by setting t = g R 2 /v 03 x = g R 2 /v 02 , to a fourth possibility (d)

1
y  (τ ) = − , y(0) = 0 , y  (0) = 1
(y(τ ) + ε)2

with ε = v 02 /(g R).

Now we want to assess and compare the four dimensionless equations for the
application example displayed above. For R = 107 m, g = 10 m/s2 and v 0 = 10 m/s
the parameter ε is very small,

v 02
ε= = 10−6 .
Rg

This suggests to neglect terms of the order of ε in the equations.

The model (a) is then reduced to

1
y  (τ ) = − , y(0) = 0 , y  (0) = 0 .
(y(τ ) + 1)2

Because of y  (0) < 0 and y  (0) = 0 this model leads to negative solutions and there-
fore it is extremely inexact and of no use. The reason lies in the scaling within the
nondimensionalization: The parameters t and x here are given by

R
t= = 103 s and x = 107 m ,
g

both scales are much too large for the problem under investigation. The maximal
height to be reached and the instance of time for which it is reached are much
smaller than the scales x for length and t for time and therefore are “hardly visible”
in the nondimensionalized model.
The model (b) reduces to

1
0=− , y(0) = 0, y  (0) = 1 .
(y(τ ) + 1)2
16 1 Introduction

This problem is not well posed, as it has no solution. Also here the chosen scales for
time and length are much too large,

R
t= = 106 s and x = R = 107 m .
v0

Model (c) reduces to

y  (τ ) = −1 , y(0) = 0 , y  (0) = 1 . (1.12)

This model has the solution

1
y(τ ) = τ − τ 2
2
and thus describes a typical parabola shaped path-time curve for a throw within the
gravitational field of the Earth neglecting the air resistance. The back transformation

v 02
x(t) = x y(t/t) = y(gt/v 0 )
g

leads to
1
x(t) = v 0 t − gt 2 .
2
This corresponds to the solution of (1.7), if the term x(t) in the denominator of the
right-hand side of (1.7) is neglected. The scales in the nondimensionalization here
have reasonable values,

v0 v2
t= = 1 s , x = 0 = 10 m .
g g

Model (d) reduces to

1
y  (τ ) = − , y(0) = 0 , y  (0) = 1 .
y(τ )2

This model does not correspond to a constant acceleration force. Also, the initial
condition seems to be problematic for this differential equation, and the time and the
length scale chosen are much too large.
Hence, for the application considered the nondimensionalization in version (c)
is the “correct one”. The versions (a), (b), and (d) are equally well mathematically
correct, but there the small parameter ε cannot be neglected anymore because its
influence is amplified by the (too) large scaling parameters t and x.
1.5 Asymptotic Expansions 17

1.5 Asymptotic Expansions

Now we will introduce a technique with which the simplified model can be improved.
The basic idea is not to neglect the terms of order ε in the exact model (1.11), but
rather to do a series expansion of the solution of (1.11) with respect to ε to achieve
more precise solutions by keeping some of the terms beyond the zeroth order term.
The terms of higher order in ε are determined from equations which we get by
substituting a series expansion into (1.11).
We want to discuss this procedure which is called the method of asymptotic expan-
sion, first for a simple algebraic example. We consider the equation

x 2 + 0.002 x − 1 = 0 . (1.13)

The second summand has a small factor in front. Setting ε = 0.001 1, we obtain

x 2 + 2εx − 1 = 0 . (1.14)

Now we want to approximate solutions x of this equation by a series expansion of

the form
x0 + εα x1 + ε2α x2 + · · · with α > 0. (1.15)

Before doing so we first define in general what we mean by an asymptotic expansion.

Let x : (−ε0 , ε0 ) → R, ε0 > 0, be a given function. A sequence (φn (ε))n∈N0 is called
an asymptotic sequence if and only if

φn+1 (ε) = o(φn (ε)) as ε → 0

for each n = 0, 1, 2, 3, . . . . An example is the sequence φn (ε) = εnα used above. A


N
series φk (ε)xk is called asymptotic expansion of x(ε) of the order N ∈ N ∪ {∞}
k=0
with respect to the sequence (φn (ε))n∈N0 , if for M = 0, 1, 2, 3, . . . , N we have

M
x(ε) − φk (ε)xk = o(φ M (ε)) as ε → 0.
k=0

N
If k=0 φk (ε)xk is an asymptotic expansion of x(ε) we write

N
x∼ φk (ε)xk as ε → 0 .
k=0

If N = ∞, we write
18 1 Introduction

∞
x(ε) ∼ φk (ε)xk as ε → 0 .
k=0

In the special case φk (ε) = εk , we speak of an asymptotic expansion of x(ε) in

powers of ε. Note that asymptotic expansions of arbitrary order can exist, even if the
corresponding infinite series are divergent for every ε = 0. In particular an asymptotic
expansion can exist, although the Taylor expansion for x(ε) does not converge for
any ε = 0.
Now we substitute the asymptotic expansion (1.15) into (1.14) and obtain

x02 + 2εα x0 x1 + · · · + 2ε(x0 + εα x1 + · · · ) − 1 = 0 .

If this identity shall hold true, it must hold true in particular for small ε. Therefore
all terms which do not contain a factor ε (or εα ), must add up to zero. Such terms are
of order 1. We write O(1) or O(ε), respectively, and collect only those terms, which
are exactly of order 1 or ε, respectively. The equation of order 1 then reads

O(1) : x02 − 1 = 0 .

Its solutions are given by x0 = ±1. In particular we see that the equation of order
O(1) has exactly as many solutions as the original problem. This is a condition
necessary in order to speak of a regularly perturbed problem. Later on we will see
when to speak of regular or of singular perturbations.
Now we consider the terms of the next higher order in ε. What the next higher
order is depends on whether we have α < 1, α > 1, or α = 1. If α < 1, then we
conclude from the term of order εα that x1 = 0, and from the terms of order ε jα in
a successive fashion x j = 0 for 1 ≤ j < 1/α. If α = 1/k for some k ∈ N, then it
follows from the term of order kα = 1 that

2x0 xk + 2x0 = 0

and therefore xk = −1. Proceeding with the asymptotic expansion one sees that the
/ N always lead to x j = 0. Therefore only the terms xkn for
terms of order j with α j ∈
n ∈ N remain. For the corresponding powers εknα we have knα ∈ N. Therefore the
power series ansatz with α < 1, α = 1/k, for some k ∈ N, leads to the same result
as the ansatz α = 1, and therefore it is unnecessarily complicated. If α = 1/k for all
k ∈ N, then from the term of order ε we get

2x 0 = 0 ,

but this is in contradiction to the already computed solutions x0 = ±1. Therefore

the ansatz α < 1 is not sensible. In the case α > 1 the term of order ε also leads to
2x 0 = 0 which is impossible as we have already seen. Therefore α = 1 remains as
the only sensible choice and we obtain the equation
1.5 Asymptotic Expansions 19

O(ε) : 2x0 x1 + 2x0 = 0

for the order ε. Its only solution is given by x1 = −1.

If we also take terms of the next higher order ε2 into account, we obtain

x02 + 2εx0 x1 + ε2 x12 + 2ε2 x2 x0 + 2ε(x0 + εx1 + ε2 x2 ) − 1 = 0

and the terms of order ε2 lead to the identity

O(ε2 ) : x12 + 2x2 x0 + 2x1 = 0 .

Therefore we have
1 1
x2 = (x0 )−1 = ± .
2 2

The Eq. (1.13) corresponds to (1.14) for ε = 10−3 . Therefore we expect that the
numbers
x0 , x0 + εx1 , x0 + εx1 + ε2 x2

are good approximations of the solutions of (1.13) if we set ε = 10−3 . In fact we

have

x0 x0 + εx1 x0 + εx1 + ε2 x2 exact solutions

1 0.999 0.9990005 0.9990005 · · ·
−1 −1.001 −1.0010005 −1.0010005 · · ·

Therefore in this simple example the series expansion leads to very good approxi-
mations taking only a few terms into account.
This procedure becomes more interesting for complex problems without a closed
form solution. Now we want to discuss the method of asymptotic expansion for the
example (1.11), a throw in a gravitational field of a planet. We apply Taylor expansion
at z = 0
1
= 1 − 2z + 3z 2 − 4z 3 ± · · ·
(1 + z)2

to the right-hand side of the differential equation

1
yε (τ ) = − (1.16)
(1 + ε yε (τ ))2

and get
yε (τ ) = −1 + 2ε yε (τ ) − 3ε2 yε2 (τ ) ± · · · . (1.17)

We assume that the solution yε also possesses a series expansion of the form
20 1 Introduction

yε (τ ) = y0 (τ ) + εα y1 (τ ) + ε2α y2 (τ ) + · · · (1.18)

with coefficient functions y j (τ ) and a parameter α to be specified. This ansatz will

be substituted into (1.17) and then the coefficients of the same powers of ε will be
grouped together. The aim of this procedure is to determine a reasonable value for
a parameter α and to obtain solvable equations for the coefficient functions y j (τ ),
j = 0, 1, 2, . . .. The substitution of (1.18) into (1.17) leads to

y0 (τ ) + εα y1 (τ ) + ε2α y2 (τ ) + · · ·

 
= − 1 + 2ε y0 (τ ) + εα y1 (τ ) + ε2α y2 (τ ) + · · ·
 2
− 3ε2 y0 (τ ) + εα y1 (τ ) + ε2α y2 (τ ) + · · · ± · · · . (1.19)

In the same way the series expansion can be substituted into the initial conditions
and one obtains

y0 (0) + εα y1 (0) + ε2α y2 (0) + · · · = 0 ,

y0 (0) + εα y1 (0) + ε2α y2 (0) + · · · = 1 .

By comparing the coefficients of εkα , k ∈ N, one immediately gets

y j (0) = 0 for j ∈ N ∪ {0}, y0 (0) = 1 and y j (0) = 0 for j ∈ N . (1.20)

To compare the coefficients appearing in (1.19) for the same powers of ε on the left
and right-hand side is more complicated. The lowest appearing power of ε is ε0 = 1.
The comparison of the coefficients of ε0 leads to

y0 (τ ) = −1 .

Together with the initial conditions y0 (0) = 0 and y0 (0) = 1 we obtain the already
known problem (1.12) with its solution

1
y0 (τ ) = τ − τ 2 .
2
The next exponent to be considered depends on the choice of α. For α < 1 it is εα ,
comparison of the coefficients leads to

y1 (τ ) = 0 .

Together with the initial conditions y1 (0) = y1 (0) = 0 we have the unique solution
y1 (τ ) = 0. The term 2εy0 in (1.19) can only be compensated by a term of the form
εkα yk , k ∈ N, kα = 1. As in the case of y1 we conclude that y j ≡ 0 for 1 ≤ j ≤
k − 1. Analogously one sees that the terms yk , where kα ∈ / N, all have to be zero.
Hence, one could have started with the ansatz α = 1.
1.5 Asymptotic Expansions 21

For α > 1 the next exponent is given by ε1 , and the comparison of coefficients
leads to y0 (τ ) = 0. This is a contradiction to the solution computed above. Therefore
α > 1 is the wrong choice.
In summary, the only reasonable exponent is α = 1. Then the coefficients of ε1
are given by
y1 (τ ) = 2 y0 (τ ) = 2τ − τ 2 .

Together with the initial conditions y1 (0) = y1 (0) = 0 one obtains the unique solu-
tion
1 1
y1 (τ ) = τ 3 − τ 4 .
3 12

The coefficients of ε2 lead to the problem

2 3 1 4 3
y2 (τ ) = 2 y1 (τ ) − 3 y02 (τ ) = τ − τ − 3τ 2 + 3τ 3 − τ 4
3 6 4

together with the initial conditions y2 (0) = y2 (0) = 0. Its solution is given by

11 6 11 5 1 4
y2 (τ ) = − τ + τ − τ .
360 60 4
Correspondingly further coefficients y3 (τ ), y4 (τ ), · · · can be computed, but the effort
becomes larger and larger with increasing order. In particular the first three terms of
the series expansion are
   
1 1 3 1 1 11 11 6  
yε (τ ) = τ − τ 2 + ε τ − τ 4 + ε2 − τ 4 + τ 5 − τ + O ε3 .
2 3 12 4 60 360

Figure 1.2 shows the graph of the approximations y0 (τ ) of order 0, y0 (τ ) + ε y1 (τ ) of

order 1 and of the exact solution for ε = 0.2. One sees that visually the approximation
of order 1 can hardly be distinguished from the exact solution, but the approximation
of order 0 still contains a clearly visible error.
Now we want to use the series expansion to obtain a better approximation for
the height of the throw. For this purpose we first compute an approximation for the
instant of time τ = τε , at which this maximal height is reached, using the equation

yε (τ ) = 0 .

From yε (τ ) = y0 (τ ) + ε y1 (τ ) + ε2 y2 (τ ) + · · · it follows that

 
y0 (τ ) + ε y1 (τ ) + ε2 y2 (τ ) + O ε3 = 0

with y0 , y1 , y2 as above. Again we solve this equation in an approximative fashion

using the series ansatz
τε = τ0 + ε τ1 + ε2 τ2 + · · · .
22 1 Introduction

y 0 + ε y1

yε
y0

1 2

Fig. 1.2 Asymptotic expansion for the vertical throw with ε = 0.2

Comparing the coefficients of ε0 leads to

y0 (τ0 ) = 1 − τ0 = 0 ,

and therefore τ0 = 1. From the coefficients of ε and the expansion yi (τε ) = yi (τ0 ) +
ε yi (τ0 )τ1 + · · · , i = 0, 1, one obtains

1
y0 (τ0 )τ1 + y1 (τ0 ) = −τ1 + τ02 − τ03 = 0
3
and therefore τ1 = 2/3. Thus, the approximation of first order for τε is given by

2
1 + ε.
3
The corresponding height is
   
h ε = yε (τε ) = y0 (τ0 ) + ε y0 (τ0 )τ1 + y1 (τ0 ) + O ε2
  1 1  
= y0 (τ0 ) + ε y1 (τ0 ) + O ε2 = + ε + O ε2 .
2 4
If one takes into account that the gravitational force decreases with height, then the
maximal height of the throw becomes slightly bigger. For our original example with
ε = 10−6 this affects the results in the seventh digit.
A priori it is not clear whether a series expansion of the form (1.18) exists. In
order to develop a model in a mathematically rigorous fashion by using such a series
expansion makes it necessary to justify the results obtained, for example by the
derivation of an error estimate of the form
1.5 Asymptotic Expansions 23

N
yε (τ ) − ε j y j (τ ) ≤ C N ε N +1 . (1.21)
j=0

We show such an estimate for N = 1, i.e.,

|yε (τ ) − y0 (τ ) − ε y1 (τ )| ≤ Cε2 , (1.22)

for τ ∈ (0, T ) for an appropriate final time T , and ε small enough in the sense that
ε < ε0 for an appropriate ε0 to be specified. As a first step we construct a differential
equation for the error

z ε (τ ) = yε (τ ) − y0 (τ ) − ε y1 (τ ) .

From the differential equations for yε , y0 , and y1 we obtain

1
z ε (τ ) = yε (τ ) − y0 (τ ) − ε y1 (τ ) = − + 1 − 2ε y0 (τ ) .
(1 + ε yε (τ ))2

A Taylor expansion with a representation of the remainder leads to

1 1
= 1 − 2y + 3 y2,
(1 + y) 2 (1 + ϑy)4

where ϑ = ϑ(y) ∈ (0, 1). Hence, we obtain

1
z ε (τ ) = −1 + 2ε yε (τ ) − 3ε2 y 2 (τ ) + 1 − 2ε y0 (τ ) .
(1 + εϑ yε (τ ))4 ε

Substitution of yε (τ ) = z ε (τ ) + y0 (τ ) + ε y1 (τ ) leads to

z ε (τ ) = 2ε z ε (τ ) + ε2 Rε (τ ), (1.23)

where
3 yε2 (τ )
Rε (τ ) = − + 2y1 (τ ) .
(1 + εϑ yε (τ ))4

Additionally the initial conditions

z ε (0) = 0 and z ε (0) = 0

are valid. For an estimation of Rε (τ ) we need lower and upper bounds for yε (τ ).
These can be derived from the differential equation for yε and the representation
24 1 Introduction

τ τ  t 
yε (τ ) = yε (0) + yε (t) dt = yε (0) + yε (s) ds dt
0 0 0
τ t
=τ+ yε (s) ds dt .
0 0

We set tε := inf t t > 0, yε (t) < 0 . Obviously it follows from (1.16), that
yε (τ ) ≥ −1 for 0 < τ < tε and therefore

1
yε (τ ) ≥ τ − τ 2 .
2

As yε is continuous, in particular we have tε ≥ 2. Because of yε (τ ) ≤ 0 for τ < tε

we also have yε (τ ) ≤ tε for τ < tε . Now let T ≤ tε be chosen and fixed, for example
T = 2. Then we have

|Rε (τ )| ≤ 3|yε (τ )|2 + 2|y1 (τ )| ≤ C1

for τ < T with a constant C1 = C1 (T ). For a given C0 > 0 we define a further

instant of time τε > 0 by τε := inf t t > 0, |z ε (t)| ≥ C0 ε2 . As z ε is continuous
and z ε (0) = 0, we have τε > 0. For τ < min(T, τε ) we conclude from (1.23) that
τ t τ t
1 2
|z ε (τ )| = z ε (s) ds dt ≤ |z ε (s)| ds dt ≤ T (2C0 ε + C1 )ε2 .
0 0 0 0 2

Then for C0 > T 2 C1 there exists a ε0 > 0, such that 12 T 2 (2C0 ε0 + C1 ) = C0

2
, namely

1 C1
ε0 = 2
− .
2T 2C0

For all ε ≤ ε0 and all t ≤ min{T, τε } it holds true that

C0 2
|z ε (t)| ≤ ε .
2
As z ε is continuous, in particular we have τε ≥ T . Hence, (1.22) has been shown,
for C = C0 /2.
The procedure to determine an asymptotic expansion can also be formulated more
generally and abstractly in Banach spaces, i.e., in complete, normed vector spaces.
Let B1 , B2 be Banach spaces and

F : B1 × [0, ε0 ) → B2

be a smooth mapping, which is sufficiently differentiable for the following consid-

erations. For ε ∈ [0, ε0 ) we look for a solution yε of the equation
1.5 Asymptotic Expansions 25

F(y, ε) = 0 .

We make the ansatz

∞
yε = εi yi
i=0

and expand
⎛ ⎞
∞ ∞ ∞
F(yε , ε) = εi Fi (yε ) = εi Fi ⎝ εj yj⎠
i=0 i=0 j=0
⎛ ⎛ ⎞
∞ ∞
= εi ⎝ Fi (y0 ) + D Fi (y0 ) ⎝ εj yj⎠
i=0 j=1
⎛ ⎞ ⎞
∞ ∞
+ 21 D 2 Fi (y0 ) ⎝ εj yj, εj yj⎠ + . . . ⎠
j=1 j=1

= F0 (y0 ) + ε(F1 (y0 ) + D F0 (y0 )(y1 ))+

 
+ ε2 F2 (y0 ) + D F1 (y0 )(y1 ) + D F0 (y0 )(y2 ) + 21 D 2 F0 (y0 )(y1 , y1 )
+ ··· .

We try to solve these equations successively, with increasing order and obtain

F0 (y0 ) = 0 ,
D F0 (y0 )(y1 ) = −F1 (y0 ) ,
1 2
D F0 (y0 )(y2 ) = −F2 (y0 ) − D F1 (y0 )(y1 ) − D F0 (y0 )(y1 , y1 ) ,
2
..
.
D F0 (y0 )(yk ) = G k (y0 , . . . , yk−1 ) .

If the linear mapping D F0 (y0 ) : B1 → B2 possesses an inverse then the values

y1 , y2 , y3 , . . . can be computed successively.

Definition 1.1 If the values y0 , . . . , y N are solutions of the above displayed equa-
tions, then the series
N
yεN := εi yi
i=0

is called asymptotic expansion of order N .

26 1 Introduction

An important question now is: Are the solutions for the problems “perturbed” by a
small parameter ε a good approximation for the original problem? A positive answer
is encoded in the following definition.

Definition 1.2 (Consistency) The equations

F(y, ε) = 0, ε > 0 ,

are called consistent with

F(y, 0) = 0 ,

if for all solutions y0 of F(y0 , 0) = 0 it holds true that

lim F(y0 , ε) = 0 .
ε→0

Remarks
1. In general consistency does not imply convergence: also in a consistent situation
the solutions yε of F(y, ε) = 0 does not need to fulfill

yε − y0 → 0 in B1

(see also Exercise 1.11).

2. An important case in asymptotic analysis is characterized by the fact that the small
parameter appears as a factor in a term which is decisive for the mathematical
structure of the problem. For differential equations this term in general is the
highest order derivative of the unknown function appearing in the equation. In
this case one speaks of a singular perturbation. At the end of Chap. 6 we will
investigate corresponding examples.
Examples:
(i) The equation
ε x2 − 1 = 0

changes its order for ε → 0. In particular for ε = 0 the equation becomes insolv-
able and the solutions xε± of ε x 2 − 1 = 0 converge to infinity for ε → 0.
(ii) The initial value problem

1
ε yε = , yε (0) = 0 , yε (0) = 1
(yε + 1)2

changes its character if one sets ε = 0. For ε > 0 one has a differential equation
and for ε = 0 one obtains an insolvable algebraic equation.
1.6 Applications from Fluid Mechanics 27

1.6 Applications from Fluid Mechanics

Now we will discuss the introduced notions of dimensional analysis, asymptotic

expansion and singular perturbation for a considerably more complex example from
fluid mechanics. The models which we use will be derived systematically in the
framework of continuum mechanics in Chap. 5.
We consider the following example: a fluid, i.e., a liquid or a gas, flows past a
body K (Fig. 1.3). We are interested in the velocity field

v = v(t, x) ∈ R3 , t ∈ R, x ∈ R3 ,

of the fluid. We assume that for |x| → ∞ the velocity converges to a constant value,
i.e.,
v(t, x) → V ∈ R3 as |x| → ∞ .

From conservation principles and using certain constitutive assumptions about the
properties of the fluid the Navier–Stokes equations can be derived, see Chap. 5. For
an incompressible fluid with constant density 0 neglecting exterior forces we obtain

0 (∂t v + (v · ∇)v) = −∇ p + μΔv , (1.24)

∇ ·v = 0, (1.25)

where p denotes the pressure and μ the dynamic viscosity of the fluid. The viscosity
is caused by internal friction. It is high for honey and low for gases. Furthermore,
expressed in Cartesian coordinates we have

Fig. 1.3 Flow past an obstacle

28 1 Introduction

3
∂
∇ ·v = vi ∈ R for the divergence of a vector field v,
i=1
∂xi
3
∂2
Δv = v ∈ R3 the Laplacian operator and
i=1
∂x 2
i
 3 
(v · ∇)v = v i ∂i v j ∈ R3 .
i=1 j=1,2,3

First we discuss the dimensions of the appearing terms.

variables dimension
v velocity L/T
0 mass density M/L 3
p pressure = force/area (M · L/T 2 )/L 2 = M/(L T 2 )

Furthermore we have
[μ] = M/(L T ) .

As a consequence all terms in (1.24) have the dimension M/(L 2 T 2 ).

As an example for the potential of dimension analysis we consider the behavior
of a fluid flowing around a large ship. The goal is to perform experiments with a ship
whose size is reduced by the factor 100. Under which circumstances the results of
such experiments with a ship model can be transferred to the behavior of a large real
ship?
For this purpose the equation has to be transferred into a dimensionless form.
Here the relevant parameters are the characteristic length x, for example the length
of the ship, the velocity v of the ship, the density 0 , and the viscosity μ. We form
dimensionless quantities from combinations of these parameters, for example

x t
y= , τ= ,
x t

where t = t(x, V, 0 , μ) is a characteristic time still to be determined. Furthermore

we set
v
u(τ , y) =
|V |

and p
q(τ , y) = , where p is still to be determined.
p

Multiplication of the Navier–Stokes equations by t/(0 |V |) and using the transfor-

mation rules ∂t = 1t ∂τ and ∇x = x1 ∇ y leads to the equation
1.6 Applications from Fluid Mechanics 29

t|V | p t μ t
∂τ u + (u · ∇)u = − ∇q + Δu .
x 0 x|V | 0 (x)2

We set t = x/|V |, p = |V |2 0 and η = μ/0 — this is called the kinematic viscosity

— and obtain
1
∂τ u + (u · ∇)u = −∇q + Δu ,
Re
u(τ , y) → V /|V | as |y| → ∞ .

Here Re := x|V |/η is called the Reynolds number. For large |y| the Euclidean norm
of the nondimensionalized velocity converges to 1. Furthermore it still holds that

∇ ·u = 0.

This means that flow situations with different x lead to the same dimensionless form,
if the Reynolds number is the same for the different situations. If we reduce the size
of the ship by the factor 100, then one possibility is to enlarge the approach velocity
by the factor 10 and to reduce the kinematic viscosity by the factor 10 to obtain the
same Reynolds number.
With the help of the Reynolds number one can estimate which effects are of
importance for a flow and which effects are not. We discuss this for the example
of two different models for the flow resistance of a body, which we motivate by
means of heuristic considerations. For the case of small Reynolds numbers, i.e.,
for high viscosity or a small approach velocity, the viscous friction dominates the
flow resistance. Then the characteristic quantities are the velocity v of the obstacle
relative to the flow, a characteristic quantity x for the size of the obstacle, and the
dynamic viscosity μ of the fluid. The dimensions are given by [v] = L/T , [x] = L,
and [μ] = F T /L 2 , where F denotes the dimension of force. Then a combination of
these quantities has the dimension
 
v a x b μc = L a+b−2c T −a+c F c .

This is the dimension of force, if

a + b − 2c = 0, −a + c = 0, and c = 1

and therefore a = b = c = 1. A law for the friction resistance in a viscous fluid

therefore has to have the form

FR = −c R μ x v (1.26)

where FR is the friction force acting on the body, v the velocity of the body relative
to the flow velocity, and c R is the friction coefficient which is depending on the shape
of the body. For a sphere with radius r it can be shown that
30 1 Introduction

FR = −6πr μv

holds true, and this is called Stokes’ law.

For high Reynolds numbers, however, this part of the flow resistance is dominated
by a force which is necessary to accelerate the part of the fluid lying in the direction
of the motion of the body. The mass to be accelerated per time interval Δt may fulfill

Δm ≈  A |v| Δt ,

where  is the density of the fluid and A the cross sectional area of the body. Here
the term A |v| Δt just describes the volume replaced by the body in the time interval
Δt. This fluid volume is accelerated to velocity v. The supplied kinetic energy is

ΔE kin ≈ 21 Δm |v|2 ≈ 21  A |v|3 Δt .

The friction force is related to the kinetic energy by

|FR | |v| Δt ∝ ΔE kin

where ∝ indicates that the two sides are proportional to each other. Hence, we obtain

|FR | ∝ 21  A |v|2 .

The corresponding proportionality constant cd is called drag coefficient and we obtain

FR = − 21 cd A|v|v . (1.27)

As this force is proportional to the square of the velocity, for large velocities it
dominates the viscous frictional force (1.26), on the other hand for small velocities
it can be neglected compared to (1.26). Formula (1.27) can also be justified by a
dimensional analysis (see Exercise 1.12). In applications the drag coefficient has to
be determined by measurements since for most body shapes a theoretical derivation
as for a sphere in the case of Stokes’ law does not exist any more. In any case (1.27)
is only a relatively coarse approximation to reality, the real dependence of the flow
resistance on the velocity is considerably more complex. On the other hand Stokes’
law is a relatively good approximation, if only the velocity is sufficiently small.
In order to assess for a given application which of the two laws (1.26) or (1.27)
is reasonable, the coefficient of the two frictional forces can be considered:

|FR(1.27) | x|v|
∝ = Re .
|FR(1.26) | μ

In doing so we choose the scale x such that A = x 2 . Therefore Stokes’ law (1.26)
makes sense for Reynolds numbers Re 1. On the other hand the flow resistance
1.6 Applications from Fluid Mechanics 31

given by (1.27) dominates for Re  1. For Re ≈ 1 both effects have the same impor-
tance.
The Navier–Stokes equations being a complex model the question arises whether
certain terms can be neglected in specific situations. As we have transferred the
equation to a nondimensional form it is possible to speak of large or small independent
of the choice of units: now the number 1 can be interpreted as a medium sized quantity.
The only parameter is the Reynolds number and for many problems Re is very large.
Then ε = 1/ Re is a small term which suggests to neglect the term εΔu = Re 1
Δu. In
this way we obtain the Euler equations of fluid mechanics

∂τ u + (u · ∇)u = −∇q ,
∇ ·u = 0.

How good is the description of a real fluid by this reduced model? Later we will see
that the Euler equations do not allow for the formation of vortices (see Sect. 6.1.4).
Most specifically we have
⎛ ⎞
∂x2 u 3 − ∂x3 u 2
∇ × u(t, x) = ⎝∂x3 u 1 − ∂x1 u 3 ⎠ = 0
∂x1 u 2 − ∂x2 u 1

for t > 0 if ∇ × u(0, x) = 0. Additionally it turns out that

• the term εΔu cannot be small in the vicinity of the boundary ∂ K of the body,
and if the Euler equations describe the flow,
• a bounded body would exert no resistance to the flow, therefore no forces would
act against the flow. This is called d’Alembert’s paradox and
• no lift force would act on the body (in 3-D).
In reality we see that a flow past an obstacle induces the formation of vortices, which
sometimes separate from K . Additionally so-called boundary layers can be observed.
These are “thin” regions of the flow in the vicinity of the body in which the flow field
changes drastically.
What is the deficiency of the reduced model?
In the Navier–Stokes equations second derivatives with respect to the spatial
variables appear, but in the Euler equations only first derivatives. In the theory of
partial differential equations different types of equations are distinguished (see for
example Evans, [37]). According to this classification the Navier–Stokes equations
are parabolic, but the Euler equations are hyperbolic. The qualitative behavior of
hyperbolic and parabolic differential equations differs considerably. For example in
the solutions of the Euler equations even for arbitrary smooth data discontinuities
may occur, on the other hand for the Navier–Stokes equations smooth solutions are
to be expected in this case, even if the rigorous theoretical proof for three dimensions
is still outstanding. The small factor ε belongs to the term which is decisive for the
behavior of the solution. Therefore one speaks of singular perturbation. To obtain
32 1 Introduction

approximate solutions of the Navier–Stokes equations the method of asymptotic

expansion cannot be used in the form we have discussed so far. It breaks down in
boundary layers, where the solution changes strongly. In Chap. 6 we will develop
the singular perturbation theory, to obtain asymptotic expansions also in boundary
layers.

1.7 Literature

An extensive description and analysis of biological growth models can be found in

[105]. For further information on the subjects nondimensionalization, scaling, and
asymptotic analysis we recommend [89], Chaps. 6 and 7, for scaling and dimensional
analysis also [41], Chap. 1, is a good reference and for various aspects of asymptotic
analysis we refer to [68]. A presentation of singular perturbation theory with many
examples can be found in [77]. Parts of the presentation in this chapter are based on
the lecture notes [116].

1.8 Exercises

Exercise 1.1 A bank offers four different variants of a savings account:

variant A with monthly payment of interest and an interest rate of 0.3% per month,
variant B with a quarterly payment of interest and an interest rate of 0.9% per
quarter,
variant C with a semiannual payment of interest and an interest rate of 1.8% per
half-year,
variant D with an annual payment of interest and an interest rate of 3.6% per year.

(a) Compute and compare the effective interest rate which is obtained after a year
(reinvesting all paid interest).
(b) How must the interest rates be adjusted, such that they lead to the same yearly
interest rate of 3.6%?
(c) Develop an interest model that is continuous in time, which does not need a time
increment for the payment of interest.

Exercise 1.2 A police officer wants to determine the time of death of the victim of a
homicide. He measures the temperature of the victim at 12.36 p.m. and obtains 80◦ F.
According to Newton’s law of cooling the cooling of a body is proportional to the
difference between the body’s temperature and the ambient temperature. Unfortu-
nately the proportionality constant is unknown to the officer. Therefore he measures
the temperature at 1.06 p.m. once more and now he obtains 77◦ F. The ambient
1.8 Exercises 33

temperature is 68◦ F and it is assumed that the body’s temperature at the time of
death has been 98◦ F.
At what time the homicide took place?

Exercise 1.3 (Separation of variables, uniqueness, continuation of solutions) For

given functions f and g consider the ordinary differential equation

x  (t) = f (t) g(x(t)).

We look for solutions passing through the point (t0 , x0 ), i.e., x(t0 ) = x0 holds true.
(a) Show that in the case g(x0 ) = 0 locally a unique solution through the given point
exists.
(b) Assume that g = 0 in the interval (x− , x+ ), where g(x− ) = g(x+ ) = 0 and let g
be differentiable at x− and x+ . Show that the solution of the differential equation
through the point (t0 , x0 ), where x0 ∈ (x− , x+ ), exists globally and is unique.
Hint: Is the solution through the point (t+ , x+ ) unique?

Exercise 1.4 We consider the model of limited growth of populations

x  (t) = q x M x(t) − q x 2 (t) , x(0) = x0 .

(a) Nondimensionalize the model using appropriate units for t and x. Which possi-
bilities exist?
(b) What nondimensionalization is appropriate for x0 x M (x0 “much smaller
than” x M ) in the sense that omitting small terms leads to a reasonable model?

Exercise 1.5 (Nondimensionalization, scale analysis) A body of mass m is thrown

upwards in a vertical direction from the Earth’s surface with a velocity v. The air
resistance is supposed to be taken into account by Stokes’ law FR = −cv for the
flow resistance in viscous fluids, which is reasonable for small velocities. Here c is a
coefficient depending on the shape and the size of the body. The motion is supposed
to depend on the mass m, the velocity v, the gravitational acceleration g and the
friction coefficient c with dimension [c] = M/T .
(a) Determine the possible dimensionless parameters and reference values for height
and time.
(b) The initial value problem for the height is assumed to take the form

mx  + cx  = −mg , x(0) = 0 , x  (0) = v .

Nondimensionalize the differential equation. Again different possibilities are

available.
(c) Discuss the different possibilities of a reduced model if β := cv/(mg) is small.
34 1 Introduction

Exercise 1.6 A model for the vertical throw on the Earth taking into account the air
resistance is given by

mx  (t) = −mg − c|x  (t)|x  (t) , x(t0 ) = 0 , x  (t0 ) = v 0 .

In this model the gravitational force is approximated by F = −mg, the air resis-
tance for a given velocity v is described by −c|v|v with a proportionality constant c
depending on the shape and size of the body and the density of the air. This law is
reasonable for high velocities.
(a) Nondimensionalize the model. What possibilities exist?
(b) Compute the maximal height of the throw for the data m = 0.1 kg, g = 10 m/s2 ,
v 0 = 10 m/s, c = 0.01 kg/m and compare the result with the corresponding result
for the model without air resistance.

Exercise 1.7 (Nondimensionalization) We want to compute the power P, which is

necessary to move a body with known shape (for example a ship) in a liquid (for
example water). We assume that the power depends on the length  and the velocity
v of the ship, the density  and the kinematic viscosity η of the liquid, and the
gravitational acceleration g. The dimensions of the data are [] = L, [] = M/L 3 ,
[v] = L/T , [η] = L 2 /T , [P] = M L 2 /T 3 , and [g] = L/T 2 , where L denotes the
length, M the mass and T the time. Show that under these assumptions the power P
is given by
P
= Φ(Fr, Re)
2 v 3

with a function Φ : R2 → R and the dimensionless quantities

|v| |v|
Re = (Reynolds number) and Fr = √ (Froude number).
η g

Exercise 1.8 (Formal asymptotic expansion)

(a) For the initial value problem

x  (t) + ε x  (t) = −1 , x(0) = 0 , x  (0) = 1

compute the formal asymptotic expansion of the solution x(t) up to the second
order in ε.
(b) Compute the formal asymptotic expansion for the instance of time t ∗ > 0, for
which x(t ∗ ) = 0 holds true, up to first order in ε, by substituting the series
expansion t ∗ ∼ t0 + ε t1 + O(ε2 ) into the approximation obtained for x leading
to a determination of t0 and t1 .
1.8 Exercises 35

Exercise 1.9 A model already nondimensionalized for the vertical throw with small
air resistance is given by

x  (t) = −1 − ε(x  (t))2 , x(0) = 0, x  (0) = 1 .

The model describes the throw up to the maximal height.

(a) Compute the first two coefficients x0 (t) and x1 (t) in the asymptotic expansion

x(t) = x0 (t) + ε x1 (t) + ε2 x2 (t) + · · ·

for small ε.
(b) Compute the maximal height of the throw up to terms of order ε using asymptotic
expansion.
(c) Compare the results from (b) for the data of Exercise 1.6(b) with the exact result
and the result neglecting the air resistance.
Exercise 1.10 (Multiscale approach) The function y(t) is supposed to solve the
initial value problem

y  (t) + 2ε y  (t) + (1 + ε2 )y(t) = 0 , y(0) = 0 , y  (0) = 1 ,

for t > 0 and a small parameter ε > 0.

(a) Compute the approximation of the solution by means of formal asymptotic
expansion up to first order in ε.
(b) Compare the function obtained in (a) with the exact solution

y(t) = e−εt sin t .

For which times t the approximation from (a) is good?

(c) To get a better approximation one can try the approach

y ∼ y0 (t, τ ) + ε y1 (t, τ ) + ε2 y2 (t, τ ) + · · · ,

here τ = εt is a slow time scale.

Substitute this ansatz in the differential equation and compute y0 such that the
approximation becomes better.
Hint: The equation of lowest order does not determine y0 uniquely and coeffi-
cient functions in τ appear. Choose them in a clever way such that y1 is easily
computable.
Exercise 1.11 (Consistency versus convergence) For a parameter ε ∈ [0, ε0 ) with
ε0 > 0 we consider the family of operators

F(·, ε) : B1 := Cb2 ([0, ∞)) → B2 := Cb0 ([0, ∞)) × R2 ,

F(y, ε) = (y  + (1 + ε)y, y(0), y  (0) − 1) .
36 1 Introduction

Here Cbn ([0, ∞)) denotes the vector space of n-times differentiable functions with
bounded derivatives up to order n. The norms of the spaces B1 and B2 are given by

y B1 = sup {|y(t)| + |y  (t)| + |y  (t)|} ,

t∈(0,∞)

( f, a, b) B2 = sup {| f (t)|} + |a| + |b| .

t∈(0,∞)

(a) For the problem F(y, ε) = (0, 0, 0) compute the exact solution yε .
(b) Show: F(·, ε) is consistent with F(·, 0), but yε does not converge to y0 in B1 as
ε → 0.

Exercise 1.12 Derive the friction law for the flow resistance in the case of high
Reynolds numbers,
FR = − 21 cW A  |v|v ,

by means of a dimensional analysis. Use the assumptions that the frictional force
depends on the density  of the liquid, a characteristic quantity r of the body, and
the velocity v of the flow. As the drag coefficient depends on the shape of the body
choosing r such that A ≈ r 2 is feasible.
Exploring the Variety of Random
Documents with Different Content
 Octavia. Nostris eram in hortis cum matre, ante sex menses.
Multa de suo in me amore et studio dixit, repetitis a prima pueritia
exemplis. «Nunc vero munus concinno tibi,» subjicit subridens,
«quod omnium longe rerum superet pretium.» Gratias egi. Paucos
post dies, accersit in ædes Robertum. Jubet nobiscum prandere,
convertebat nostros in se oculos tot venustatum avidos. Mirabar; nec
scienti, nec percipienti mihi gliscebat amor. «Libere colloquere: scio
te præstare ingenio.—Quandoquidem, domina, jubes, parum a
reverentia quam tibi debeo et ab officio secedam meo, ne inobediens
videar, qui sum obsequentissimus, et esse debeo.—Ingeniose
respondet,» repono.—«Atenim,» ait, «ingenii quæ credis esse aliquid
in me, inspira ex tuo, et erit. Nam in te dextro convenit Jove
pulchritudo corporis et præstantia ingenii.—Ergo, Roberte, quæ se
suis non probat, pulchra tuis videtur oculis Octavia?» refert mater.
—«Videtur, per omnes Deos, per omnes Deas! pulcherrima, et est.
Diis amorem, et Deabus invidiam creare possit.—Quid si consentiret
amori tuo Octavia,» subjicit mater, «amares?—Amabo,» ait,
«quanquam et non consentiat. Nam nihili sum puer: qui tandem fiet
[pg 279] ut consentire posset amori meo? Sed locuples est illa et
Naturæ, et Fortunæ, et virtutum donis: qui fieri potest ut non
amem?—Gratum sane erit,» respondeo, «me abs te amari tam
amabili puero. Vis mihi alter esse maritus?—Dedecores mariti
nomen,» reponit, «mihi si indideris; servi et vernulæ relinque: studiis
erga te meis, ut potero, honestabo.» Finitis epulis rediit ad
Academicos, et sensi, cum abiit, aliquam meæ mihi animæ partem
deesse. Tunc mater ad me:—«Quid de puero judicas, nata?» Esse
respondeo ex Amoribus unum, eumque pulcherrimum, qui ex
Veneris gremio evolavit ad nos.—«Num dices,» adjicit mater, «fore
beatas quarum in amplexus veniet?—Sane ut futuræ sint infelices,»
respondeo, «non video. Hunc sibi reginæ generum velint.» Dein
cæcum alebam venis vulnus, Tullia mea. Tandem post mensem,
nascenti amori favit mater optima. Jubet venire Robertum; liberam
dat confabulandi mecum potestatem. Ingeniosus est Amor, quem
fata instillant. Puer per ætatem nec par amori, nec satis aptus,
amore arsit, cujus usum nesciebat. Interea contigit matrem diro
conflictari meam morbo. Ut bonam recepit valetudinem: «Ingrata
sim,» inquit, «Octavia mea, si, quæ ope et opera tua erepta sum
letho, tuis in me benefactis non respondeam meis in te. Tanti pretii
sum tibi pollicita munus, quod omne pretium superet; meministi?—
Memini, mater,» inquam; «et quicquid a te mihi venerit summo erit
in pretio.» At illa osculum ferens:—«Robertum,» reponit, «volo tibi
[pg 280] dono dare, quem et Juno optet sibi dari. Erubescis? Novi
abs te amari.—Non nego, mater,» refero.—«Hunc tibi educo,» ait,
«et servo; volo, dignus ut sit benevolentia tua, excoli diligenter omni
virtutum et disciplinarum genere. Hunc tuos in amplexus conjiciam.
Sed tener est; de ejus tantum pulchritudine et vita detraxeris,
quantum acrius indulseris voluptati. Id fac ne obliviscaris.—Non sum
oblitura,» dixi. Tertio post die, quæsivit, inter prandendum, an domi
reliqua die futura essem. Respondit in animo esse domi manere.
Post laudavit domesticam vestem, qua induta eram libera et fluxa, e
subtilissimo lino.—«Nec ista,» inquit, «invidet oculis tuas vestis
dotes, nec objicit. Mirifice in ea places: ne mutes.» Dixi non
mutaturam.
Tullia. Quæ cogitabas cum animo tuo, cum amore tuo?
Octavia. Non cogitabam adfuturam tam cito mihi tantam
felicitatem. Nescio quid non absolutum dat in manus Attalici operis
mater. «Acu pinge quæ desunt, quæ egregia es artifex,» ait. Recessi
in cubiculum. Is erat dies vigesimus Maii. Sedeo humili in lecto,
sericeo strato vellere. In variis cogitationum nebulis mihi mens
ludebat; dubius inter vigiliam et somnum allabitur sopor. Ecce hoc
momento intrat Manilia, puerum ad me perducens. Affixæ humeris
alas, et pendens pharetra; arcum læva, sagittam dextra gestabat.
Cupidinem dixisses, et vere erat.—«Parens me mittit Venus ad te,»
ait voce blanda puer, «quod scit a te vinci se pulchritudine: [pg 281]
vult serviam tibi Amorum reginæ.—Non servies,» repono,
«pulcherrime Amorum; esse si velis ex animo meus, regnabis.—At
enim,» subjicit Manilia, «dies fabulis vobis perdenda non est. Secede
paulum, formose Amor.» Ut secessit: «Ego et mater tua mancipio tibi
nexuique, nata,» inquit, «hunc puerum damus; sed tener est et
debili lumbo. Utere, sed parca voluptate; sin aliter feceris, defluet, ut
solent flores quos carpsit gelida hyems.—Depereo,» aio, «puerum
misera, et abunde erit voluptatis cupidæ et amanti, si sciero vere
esse meum.—Parce puero,» dicebat nutrix; «hac hora satisfactum
crede voluptati tuæ, si puero eripueris virginitatem, quod ejus sine
dolore non fiet. Nam illi mentula obducta præputio est, nec aperiet
impune tibi caput reginæ suæ. Promisit se fortiter facturum; quo
volueris habere inspira animos, venustatum tuarum et voluptatum
ostentatione.» Post alas Amori puero demebat. «Nam volo te,»
dicebat, «constantem esse; nolo ab ejus unquam evoles latere.»
Tollit pharetram et sagittas. «Nunc,» inquit, «opus sunt alia tibi tela,
aliis pugnandum armis.—Intelligo,» ait ille, «et esse telum mihi
sentiet dea mea, quo secum pugnem.» Abiit nutrix, et fores occlusit.
Tullia.
 Dicite: Io, Pæan, et Io bis dicite: Pæan!
Decidit in casses præda petita meos.

Octavia. Delibutus gaudio puer exsilit ad me. «Domina, [pg 282]

domina,» loquit, «meæ me nunc pudet ætatis. Quam turpe sit esse
puerum, vere sentio. Impar ero felicitati meæ.—Impar non eris,»
repeto, «si respondes amori meo, qui tua erit felicitas, tuo amore,
qui mea erit. Nihil ultra a te mihi opto.—Sed est,» inquit, «aliquando
præterea amatorium. Non bene incipiam tyro, nec graviter perficiam
puer.» Consedimus in lecto; tum ego osculor: incendia mittebant,
non scintillas oculi mei. Demisit ille manum in sinum.—«Quid tibi
vis?» dicebam; «liberius sane agis quam deceat.» Tractat papillas,
suaviatur, incalescit; non potui quin amplecterer, ita lepide et festive
agebat omnia. Tunc infert sub stolam dexteram.—«Euge,» inquam,
«quid hac quæris?—Uxorem meam quæro,» respondet; «isthic
habitat.» Risi:—«Sed non patiar id inhonestum,» aio, «si honesta
sim.—Tuum quicquam potest esse inhonestum?» refert. «Noli
maritum tuum repugnans fatigare. Amori cede.—Cedam,» aio, «tam
bello, tam ingenioso Amori. Tua sum, mi Amor; nec amplius mea
sum; tibi vivo.» Errabat interea puerilis manus, non docta Veneris
viam. Paulo supra genua hæsit; post femora pererravit, et nates
læves, Tullia, et molliter tumentes. Nec ultra audebat; pueriliter
omnia.—«Scis usum,» ego ad illum, «mi Cupido, corporis tui et mei?
—Videor mihi scire,» infit, «sed pro certo non habeo, an certo sciam.
Vidi in tabulis pictis Veneris histrioniam.—Dignus eras,» subjicio,
«pulchriori quam ego sum. Cape quæ voles, utere, [pg 283] fruere;
age maritum.» Educit e carceribus mentulam. «Quid de hac re tam
lepida fieri speras?» inquam.—«Quid jubebis id fiet,» reponit.
Præludebat, inflammescebat. Sensit sibi facem Veneream ardere et
turgere libidine. Genibus accidit; rogat ignoscam, si quid inepte
fecerit nec satis urbane. Læva mox resupinæ revolvit ad pectus
tunicam, et citus manum admovet pubi. Primoribus digitis tractat
lanuginem, scalpit rimam, metitur igneam.—«Dic, sodes, mi amor,»
inquam, «quid quærit hic sibi passerculus?» Nam arrigebat.
—«Puto,» respondet, «nidum quærere, quem teneo manu.—Sine,»
repono, «sine hic evolet.—Et ecce,» ait, «evolat;» et alacri se cursu
feminibus inserit meis divaricatis. Ah! ah! ah! ostendebam milvio
pullum.
Tullia. Per Genium meum! prurit insolitum quid ingeniosa hæc
narratio. Mira es nugarum artifex:

In tenui labor, at tenuis non gloria.

Octavia. Adhibet anhelans puer pugionem vaginæ.

Tullia. Mutinum crassum, longum?
Octavia. Crassum ut habes pollicem; longum sex pollices. Fidem
habe mihi: convenimus belle alter alteri.—«A regia aberras voluptatis
via,» dicebam. Manum porrigo cæco passeri, auspex et dux. Init
libidinis viam tenebricosam.—«Nunc fortiter percute, mergere, cæde,
macta,» subjicio. «Sic [pg 284] hæc res agitur.» Tunc adegit ensem
suo nisu, excepi meo; peni favebam susurro.—«Me læsum,» inquit,
«intus sentio; quid vero id est?» Pressit suspirium, et momento
subagitare ille, comprimere, concutere. Cevebant trepidanti oculi et
mens. Ego succutere, contra niti; sed et subabant mihi, intra pectus,
cordis intimæ fibræ. Adulabatur hæc cogitatio voluptati, venisse
Amorem in amplexus meos. Innectebam ulnis incubantis marmoreas
nates, et micantibus urgebam suavium alaparum velitationibus.
Audivi Hymenæum ex fervidis lumborum anfractibus cachinnantem,
et anhelantem Venerem. Resoluta sum in effluvium; nondum puer:
nec calida suavia et humida, nec blæsa murmura, nec spissas
compressiones, nec innumerabiles in una voluptate voluptates satis
tibi, mea Tullia, fando explicem. Cogita Psyches cum Cupidine
concubitum: is fuit meus cum Roberto. Illi ego pro Anima: is ipse
meus est Amor. Tandem connivere puero oculi, attolli, circumagi,
deficere deliquio; patranti subsiliebat intro cauda; nova exagitabatur
intemperie; intremiscebant comprimenti celeres lumbi; extra se
ponebatur, inopina mentis vertigine. «O domina, o regina,» inquit,
«elabor e vita! quid id novi?» aiebat. Impingit osculum et momento
colliquescit. Postquam perfecit, foveo languidum in sinu. Oscula
osculis, suspiria miscemus suspiriis.—«Fuit tibi quicquam meum
dulce, marite dulcissime?» dicebam.—«In hac voluptate,» respondet,
«omnes sensi voluptates. Vivus es, Octavia, summi boni fons.—Velim
tibi esse,» inquam, [pg 285] «sic et mihi egomet ero. Recrea vero
exhaustas vires brevi somno. Qui nos viderit, me Cypridem credet, te
Cupidinem in matris dormientem gremio.—Non opus illi somno,»
reponit reversa Manilia. «Basiorum, sine fine, odoros purpureis tuis
in labellis flores legat. Id illi opus est.» Post egregium par perducit
ad matrem. Advolat mater in complexus pudore offusæ.—«Quæ hæc
est,» ait, «mœroris species? Te pueri piget, Octavia, tam venusti?
Pudet, Roberte, libere lusisse in puella tam formosa? Infeliciter
cessere hæ nuptiæ? Abfuit Hymenæus? Non favit Cupidini Venus?—
Quin imo,» reponit Manilia, «feliciter acta sunt omnia, quæ fieri
oportuit nuptias facientibus. Bellissime præliatus est in virgine tua
Robertus meus; habes generum, et hic uxorem.—Bene est,» inquit
mater; et repetit adulans suavia. «Laudo,» subjicit, «athletas fortes
et pugnaces. Cessit utrique victoria.—Sane victum me fateor,» refert
Robertus; «Martem pugnando vinceret.—Sile, tam nugax quam
pulcher; sile,» aio, «qui prælii refers vulnera, sed honesta.—Sed ipse
proprio mucrone,» infit puer, «vulneratus sum, et tuo balsamo
oblitum et curatum vulnus.» Ridebat mater; affertur lauta gustatio.
Propinavit mater amantibus et salaciæ; Robertus Junoni reginæ suæ
Semproniæ, et Hebæ Junonis filiæ. Id dices, Tullia, a schola ductum,
sed et ductum ab amore. Rogavit mater an bene haberet.—«Roga
Octaviam,» respondet, «nam, me si ex animo amat, bene habeo.—
Bene habes,» repono, [pg 286] «plurimam places.—Per te, dea,
veræ subii voluptatis delubrum. Ingressus sum ad vitam, quæ illic
condita delitescebat.—Et sane,» ait mater, desiisti puer esse, vir
factus es: vexit te momento Octavia ad virilem ætatem. Male, male
illis sit qui Verticordiam dictam Venerem, quod ad pessima quæque
hominum convertat mentes, mentiuntur! A nugis vertit ad seria; e
puellarum et puerorum manibus bullas excutit et puppas,
concutiendo, et succutiendo.» Tunc demisit puerum ad Academicos.
«Ut in Venere ante annos venit sagacitas, sic veniunt et cani. Nolo
puerum,» mater ad me, «ante mensem videas. Nolo corrumpi et
insenescere in amplexibus tuis; post id temporis, cubabit tecum.
Gaudiis tuis tota nocte potiore. Sed contumax illi spiritus, nec regi
facilis; pulcher est, sed etiam ferus. Succensui ante hos dies, quod
terræ filius, nihili puer, auderet nobilibus, apud Academicos, pueris
æqualem putare se et gerere.—Erras, Domina, respondit; qui carus
Octaviæ sum, sum jam nobilis, jam comes, jam marchio, jam sum et
dux et princeps.—Jubeo venerari in Larario Superos:—Nullo amplius
egeo, dixit, Superum beneficio; superest nihil quod ab iis petam,
quem amat Octavia.—Hortamur incumbat assiduus et solers litteris.
Respondet scire se abs te amari, præterea velle scire nihil; qui sciat
tibi se placere satis doctum esse.» Ad hæc ego.—«Facio nocentem,
mater, et per me veniam consequatur.» Ergo advocatus est domum
præstituta die Robertus. Cœnavit nobiscum. [pg 287] Finitis epulis:
—«Cubabis,» mater ad puerum, «hac nocte cum Octavia. Sed hanc
dico legem: caste et pudice agitetis noctem. Legem accipis, Octavia?
—Accipio,» inquam; «sed lex hæc stat in ancipiti.—Interpretem
agam,» reponit Robertus.—«Et volo agas ut commodum tibi erit,» ait
mater, «dum tamen, ut fit, interpretando non corrumpas, non
aboleas legem.» Cœnavimus opipare, et magna matri cura in cœna
optimis muniri dapibus Roberti armamentarium: nam a Cerere et
Baccho Veneris arma. Procuratis corporibus, itum ad quietem.
Tullia. Item ad pugnam. Sed disputat de opportuno Veneri
tempore Cheronæus, ut ne vires infringat. Cum lungo refectæ sint
somno vires, indulgeri posse Veneri ait, experrectis anima et
sensibus. Sed stulte facere opinor barbatos illos philosophos, qui
prurientibus audent mentulis leges scribere et cunnis, uno
decumbentibus in lecto. Scribant et de sistenda cœlorum vertigine:
nam ut cæca mentula, sic et surda est. Nec illi dantur verba.
Socrates aut Zeno qui non riserit, suos si inter discipulos arrectas
viderit stare mentulas, arrectis auribus, si viderit Priapum et
Conisalum diva Philosophiæ præcepta aperto auscultare capite?
Præclarum spectaculum! Ut cyathos numerant pueris nutrices et
bibere ultra vetant, sic et inepti suas mentulis fututiones et ultra
vetant. Cur et non leges in medium posuere miseris mortalibus, et
despuendi et scabendi? Num satis infelix hominum vita sævis visa
mentibus, nisi rigidæ induerent catastæ sensus [pg 288] et animam?
Fatuis, Octavia, irascor homuncionibus. Nam si audaciam effrontem
ademeris et gravitatis falsam speciem superbientibus, nihil præterea
invenies quod lixæ non sit, et mediastini in lutulenta plebecula. Illis
nec genius nec ingenium, sed audax stoliditas et irrequieta ignavia.
Octavia. Aloisiæ maritus, suadente Pelagio, id consilii ab hinc
quatuor cepit menses, quod sanus laudarit nemo.
Tullia. Quid consilii cepit?
Octavia. Decima tantum nocte salacem init et libidinosam,
reliquas cœlibes in lecto agit; vidua prurit, cui non vivit maritus. Et
superstitiosus factus est Caviceus: mecum est infrequentior, nec
impune tulit. Sitienti si non dederit bibere, cui id muneris erat,
probro quis verterit sibi dedisse ipsum, aut a quovis cepisse dante,
sitientem?
Tullia. Marina Genovefa Pimentelia, alieno ære merso opibus
clara, Federico Mendozio nupsit. Induxit hæc in animum, annis et
specie florens, semel atque iterum corporis sui singulis facere
mensibus copiam juveni. Nec movit adolescens amore ardens puellæ
animum. Quos tulit virtus inepta fructus? Scelera peperit. Omnes,
intra mensem, uxoris ancillas quinque numero corrupit Federicus:
utero gerunt. Quid Genovefæ Lionella mater? «Tu ipsa,» dixit,
«lenam egisti infelicibus puellis; tuo periere lenocinio. Virtus hæc tua
tetrum desinit in scelus: pietatem credebas amens, lenocinium erat.»
Vim timens sibi Galliena a Luitprando marito, subligaculo corpus
nocte induebat, ita consuto, ut [pg 289] nulla pervium parte esset,
nec peti posset. Similiter domum implevit libidinibus adulteris. Age
nunc, o supina et inconsulta sapientia, illude incautæ facilitati! Quos
edis ludos pollens dolis? Sed hæc ad mores pertinent, quæ hominum
figmenta sunt. Natura ad Venerem alii proclives, alii segnes.
Quorundam vix sedarit taurus famem, ut de Milone Crotoniata
ferunt; satis aliis est mica panis. Aliis cyathus sitim exstinguit; aliis
ne quidem triental. Eodem jure, sive edant, sive bibant, omnes
conseri homines, quid injustius, quid stultius? Ita sunt quibus ad
decem dies infringit libidinis nervos coitus unus; aliorum vero ardet,
nocteque dieque, in medullis invicta libido: uno irritatur coitu, non
mansuescit. Hæc pendenda erant ea in trutina, et libranda ab hoc
examine. Ætatis et habitudinis habenda erat ab his pecudibus ratio;
sed ratione duci non solent. Medici fatentur certas nullas posse
induci leges. Naturæ tamen convenire putant primum valetudinis
curam non negligi. Bona sine valetudine, vita sepulchrum vitæ est.
Igitur damnant Epicuri sententiam, morosi hominis, ad vitam
Venerem non pertinere, nec felicitati conducere. Noxiam tamen aiunt
pueris et puellis, qui ad perfectam ætatem non pervenere, et
provectæ ætatis senibus, qui jam a vita decedunt, si sit frequentior.
Uno quoque mense et his et illis colenda semel Venus in illa
ascendente, et in hac labente ætate. At cum floret ætas, vigent
nervi, firma sunt latera, ad quartum quintumve duellum descendi,
sine ulla valetudinis læsione, potest. Apud Lacedæmonios lege
cautum, quisque ut vir ostenderet quinquies [pg 290] se per
mensem uxori virum esse, quinque militaret uxori congressibus. Nam
ducebant etiam uxores, qui non adoleverant, nec tamen ab hoc
immunes vectigali. At enim ignavus est et impius, qui altero quoque
die Veneri non facit, si nihil ultra lumbis et lateribus exspectet ab
annis. Uxorem duxisti, o tu quisquis es; debitor factus es uxori:
solve, malæ vir fidei, solve quod debes. Solve, aut cede bonis, et
bonorum potior pars uxor. Duriora excogitavit marcescentium
Philosophorum amens meditatio, sedula stultitia. Ride, Octavia: leges
tulit quibus rite et honeste, ah! ah! ah! subent et ceveant. Vetant
veternosi rapidos cieri motus; vetant novas indui figuras. Castam et
honestam negant esse posse

... ingemente quæ thoro,

Juvante verset arte mobilem natem.

O graves et sapientes legislatores! Audisti de mulierum Senatu

Romæ? Vocabant Senatulum. Claræ nobilitate et usu rerum
matronæ senatorio fungebantur munere. Conveniebant,
disceptabant de rebus nostris. Erant earum judicia ac si sancita
essent jure publico. De lumborum officio, de figurarum varietate, de
vegeta inertive libidine, consultæ venerabiles matronæ a Messalina
Claudii imperatoris et omnium uxore, sic censuerunt: [pg 291]
«Quandoquidem constarent intima et extima in homine omnia
septenario numero, et homo fingeretur coitu; si ad septimum una
nocte eatur coitum, id jure fieri, nec ultra peti posse. Qui velit
mulierem non moveri, Venerem mortuam velle: mutuos motus
Veneri esse pro anima. Figuram quæ cuique commoda esset,
probari. Id jus in omni Amoris regione, quod amor mutuus
constituisset, servari debere. Summum in amore imperium ad
voluptatis numen pertinere; illius esse leges condere, et conditas
interpretari.» Id fuit Senatus consultum, Octavia. Itaque plures, una
nocte, peragebat Messalina viros indefessa. A nullo unquam exegit
stipendium, quod sponte quisque non inferret. Die facto, quatuor
supra viginti e myrto et rosis corollas Priapo, Marsyæ et aliis ridiculis
Diis dedicabat victrix: erant singulæ parta victoriarum præmia.
Theodora, Sigismundi imperatoris uxor, Messalina altera, lumbos et
artus heroum suæ ætatis frangebat in hac palæstra: uno hippubinos
et lastauros hauriebat, vorabat succussu. Sempronia, mater tua,
duodecim cum Chrysogono confecit vehens curricula, uno spiritu.
Palæstritas tenera ego fortes confeci quatuor intra paucas horas.
Sed ita est, Octavia, hebescit et stupet, post sex septemve pugnas,
si acriter pugnatum sit, exhausta Venus. Ultra honesta non vivit
libido. Similes dicebat Victoria, consobrina mea, eas fœminas quas
nulla caperet Veneris satietas, vinosis illis ganeonibus, quos musto
non madidos, quos nulla siccos videt dies: ut his in Baccho nulla, sic
illis nulla in Venere voluptas. Pessimæ assuetudinis et hos et illas
misere vertigo agit. Ideo putres non commoventur; acri attamen
hilarescit motu Venus. Ceventis concussu, subantis successu,
eliciuntur calentes ex utriusque [pg 292] corpore igniculi; si absint,
friget Venus. Sic superest libido sibi, sic læta reddita letho.
Octavia. Cui bellum videri possit cum marmorea rem habere
Phidiæ Venere? Cui imbellis placebit Venus, et id placebit.
Tullia. Nugantur, nescio qui inepti, in umbilico sitam esse et
sedem habere libidinem. Utrique sane et viro et mulieri, in
motitatione, subagitatione, frictu, concussu et succussu, vere
consistit. Sed sua est ars et ratio his commotionibus: ut principio
lente surgant, post ferveant, demum desidant, ut, pacatis post
tempestatem ventis, commoti desinunt fluctus. Sunt hæc tua, cujus
in agilibus lumbis laus micat mirificæ mobilitatis. Nec Thais, Octavia,
hanc præripuerit tibi laudem. Sed quod ad figuras, censeo ego
quidem liberam cuique esse figuræ facultatem, quæ concinnior
videbitur. Nemo narrando dixerit figuras omnes, nec posuerit ob
oculos pingendo. Mutari amat Proteus Amor. Rabiosi illi Solones de
his sermonem haberi, tabulas pingi inhonestum latrant; damnant.
Sed de omnibus duelli et pugnæ figuris disputari non damnant.
Quibus itur ad perniciem humani generis, non stomachantur; quibus
ad procreationem, irascuntur. O feras bestias! malunt interire variis
artibus humanum genus, quam nasci. Laverniones luctum amant et
lethum; lucem odere et vitam. Quæ dementia! quæ impotentia!
Sapientiores Lesbii erant, inter Græcos ingeniosissimi. Nam Sappho
Lesbia fuit, decima Musa. Varias, etiam non suetas, in Venere figuras
insculpebant nummis, quos publice cudebant. [pg 293] Publicum id
erat authoramentum. Vidi ego Romæ, in domo Ursinæ, nummos
duos, alterum ex ære, alterum ex argento, in insula Lesbo, ut
dicebat, cusos. In altero Sappho nuda, cum nuda puella, tribadico
confligebat duello. Vir nudus in altero, dextro nixus genu, puellam
sustentabat nudam, et pilo figebat, divaricatis femoribus faventem.
Octavia. Ille, genu flexus, Veneri supplex faciebat.
Tullia. Hæc lumborum nisu petebat cœlos, sublata in aera. Per
nummum oblongum Venereum tuum! Octavia, his ex nummis ars
petita talium figurarum, ut ex appensis tabulis in Æsculapi et
Apollinis templis ars medica. Hos præ oculis, mea me si non fallit
opinio, habebant, et in manibus, Elephantis Milesia, Philænis,
Hermogenes Tarsensis, qui scripsere his de nugis eruditos libellos.
Octavia. Aiunt vero Elephantida apud Milesios probam et
pudicam fuisse matronam; sed quod ingeniosum, inter litteratos,
cepisset inimicum, hunc impudicum perscripsisse libellum, et
matronæ nihil tale cogitanti adscripsisse supposititium partum.
Tullia. Sanus non est, qui cum eruditis exercet inimicitias. Ulcisci
suas possunt injurias, per omnem ætatum æternitatem. Quas
pingunt tabulas, annos illæ ferent quos Zeuxidis et Apellis non
tulere; ab annis vires capient et dignitatem. Sed et est Italos inter
homines divini vir ingenii, qui his de ludicris amœna conscripsit
Colloquia. Demum, ut [pg 294] ex omni terrarum parte itur in cœlos,
sic ex omnibus muliebris corporis partibus et inflexionibus summam
itur ad voluptatem, Veneris cœlum: non una via itur. Labella rosea,
papillæ niveæ, manus lubricæ, nates læves et teretes via sunt ad
meliorem partem; et posticæ qui Veneri faciunt, in anticam veniunt
litatum ædem.
Octavia. Apage, apage his a spurcitiis. O cœnosam libidinem!
Tullia. Nec Oceani aquæ eluant, nec Phlegethontis ignes. Non, si
terra miscebitur Tartaro, satis supplicii scelerati ferent pædicatores.
Portentum! Rem ausus est clari vir ingenii, Joannes Casa, pulchro
libello tam fœdam commendare. O tempora! o mores! Nam Italis
fœcunda est facetiarum seges in hac libidine. «Sunt variæ rerum
species.» Alii puellam quærunt in puero; alii puerum in puella:
alterum in altero sexum. Male pereant perditi! «Vir nubit in
fœminam; Venus mutatur in alteram formam.» Scantinia jam lege id
apud Romanos turpitudinis erat damnatum. Sed pæderastas mutari
post fata aiunt Pithagoræi in scarabæos. «Honesti Colei Lanuvini;
Cliternini inhonesti.» Vetus id adagium. Male pereant qui amoris
arma malis his artibus in Amorem converterunt, qui Venerem
perdidere per Venerem! Cotytto his Dea, Opici his magistri. Omnia
inter animalia homo solus illudi corpori suo patitur; tantum non
publice prostat. Audi Plinium: «Uni animantum luctus est datus; uni
luxuria, et quidem innumerabilibus modis, ac per singula membra.»
Et alio in loco: «In hominum [pg 295] genere diverticula maribus
excogitata; omnia scelera Naturæ.»
Octavia. Et crescit novo scelere indignatio mea. Accipe insidias.
Formæ florentis difficilis tuitio.
 Tullia. De Roberto intelligis?
Octavia. Finita, ut dicebam, cœna, voluit mater Robertus et ego
libere colloqueremur. Secessit; consedimus. Tunc ille:—«Ecce me,
Dea mea, alieni spiritus afflatu contaminatum. Parum abfuit quin
reciperem me ad te uxor factus.» Vultus dicenti rubore suffunditur.
«Amat me Joannes Ludovicus Vives. Quintilianum vocant. Amant
Academici reliqui; nihil præterea dicunt in vita esse quod ament.
Varium est et dolis aptum amoris ingenium. Jacebam pronus in lecto,
nocte concubia. Invadit Ludovicus; palpabat nates, somno solvor.—O
quales optet, dicebat, mœchus Ganymedi Jupiter! Tales nec in Hyla
arsit Hercules, nec in Antinoo Adrianus. Tales erudita præferat libido
Hebes niveis mammis et pectori. O si commodentur Veneri meæ!
malim quam ipsam Venerem.—Suspirium duxit, osculum impegit.
Offerebat stuprum; repuli; minatus sum conquestum in me
intentatam injuriam.—Sed, pupule, adjicit, Margaris meos non sprevit
æstus, sodalis tui, sub his tectis, Roderici marchionis soror, pulchra,
nobilis, ingeniosa, litterarum amans, sexdecim dea annorum. Hæc
non sprevit amores meos.—Basium dedit, veniam petiit, abscessit.»
Tullia. Homo est venustus Ludovicus Vives, urbanus, doctus, nec
ætatis provectæ. Scies quid factum [pg 296] sit cum nobili virgine,
et ridebis. Qui odio indicto aversi fugiunt à puellis, velint nolint,
fugiunt ad pueros. Genuit Amori Amor homines et fœminas; genuit
sibi. Amore per Amorem sati, Amori nascimur. Infunditur Amor cum
sanguine in venas. Tolle Amorem e rebus, et res e Natura tolles.
Amamus vel inviti, ut amamur. Hinc fit, ut qui nolint amare quod
licet, ament præcipites quod amare turpe est. Nosti Justinam
Gomeziam, magni inter Vestales nominis? Amat perdite Alphonsinam
Albuquerquiam, Joannam Meneziam, et Antoniam Castrensem,
eodem in contubernio. Ut amat, amari se non negavit. «Hæc adyta,»
dicebat nudiustertius, «quæ credunt consecrata pudicitiæ, pulso
Amore, pervadit alia sub forma Amor. Obnubimur velo: obnubitur
nobiscum Amor. Fallit oculos, nec videtur. Vivit in medullis, latet in
venis. Perfluit venas sanguis, in nostra nec est potestate quin
perfluat. Incendit Amor venas, nec nostra est in potestate quin
inflammante calefiant Amore venæ. Quandoquidem ultra non licet,
amamus inter nos, et amamur. Animæ cibus Amor. Sunt qui sitim ut
sedarent, lotium ipsi suum biberunt: sunt qui, adurgente fame, sua
ipsa membra morsu laniarunt. Non dissimiliter, si fœmina viro, si vir
negatur fœminæ, ardebit vir in virum, fœmina in fœminam. Qui
Natura ferebatur Amor in alterum sexum, interclusa via suum
incestabit. Exoneranda est tibi vesica: vetant? meies tamen. Vestes
fœdabis, si desit matula. Et mulier matula est. Micturit tibi amor:
scelus est micturire? nihilo [pg 297] tibi minus meiet Amor. Quot
insaniunt in hac libidine pudicitiæ, quam avari aut erroribus ebrii
probant, Natura non probat! Venustis illis et aureis temporibus, rari
erant iique affecta ætate, qui hoc profiterentur sapientiæ genus.
Morum habebatur ratio. Rem ridiculam! A sene effœto et marcido
qui adolescentis operas et pensum exiget, non sapiet: qui juventa
velit florentem et viribus pigra mersum videri senectæ hyeme,
sapiet?» Sed perge dicere, Octavia, meliori nata sub sidere.
Octavia. «Per oculos tuos, mea sidera!» dicebat Robertus,
«cavebo diligenter ne quid admittam in me flagitii. Novam mihi nova
virtute creabo laudem. Tuus merebor videri.»
Tullia. Clamant multi nitentibus ad virtutem adolescentibus
obicem esse consuetudinem cum puellis. Næ cæcutientes errant.
Vidi quos honestarum, ut sumus, usus ad meliorem convertit
frugem, cupidinarios, aleones. Invenerunt, cum voluptate, veram
honestatem. Quod ipsa non fecisset per se virtus, hoc fecit auxilio.
Placuit virtus qua se noverant placituros. Erant corruptissimi in urbe
vicina adolescentium mores. Damnavit mulcta indicta Senatus
meretricias libidines, lenonum artes; sed facilius sine aeris quam
fœminarum usu vixerint adolescentes. Convertunt ingenuas in
puellas oculos et amores. Excipiebant, parentum jussu, læto vultu
qui ad honesta tenderent; contemptui habebant qui mutari nollent.
Ingeniosæ mulieris suadent illecebræ, quod Plato non suaderet.
Annus nondum, [pg 298] præterierat; mutatis repente moribus
evasere in masculam virtutem, Epicuri de grege porci. Prælucente
spe, quam ocyus concipit qui amat, quo non sequetur virtutem, vel
ad dumeta ducentem, vel e scopulis vocantem? Alter per alterum
mitigandus sexus, non alter semovendus ab altero. Præeuntes
sequentur viri fœminas, vi Naturæ ducti. Si sint bonæ, duces erunt
ad laudem; si malæ, auspices erunt ad turpia. Nunc sequere
orationis filum.
Octavia. «Enimvero amore emorior,» dicebat Robertus. «Cur
tandiu meam mihi invidet Sempronia felicitatem?» Audiit mater.
—«Non equidem invideo,» reponit, «sed delectatur mora Venus.
Amoris voluptates mora crescunt. Volo vobiscum liberaliter agere. Ite
cubitum, ite.» Et subrisit.
Tullia. Intelligo. «Ite ad coitum,» dixit; «ite ad vitam.»
Classicum canit ad pugnam.
Octavia. Utrumque osculata est. Dein duxit Manilia ad
pugilatum. Veste exuit, nudam collocat in lecto. Insilit Robertus. «En
mihi,» loquit amplexans, «summum bonum, omne bonum.—Et bene
sit,» ait Manilia, «beatis amantibus! Nec hos exstinguo cereos,
Roberte; triumpho non deerit tuo lux quæ debetur.—Et sane,»
reponit Robertus, «tenerum formosæ puellæ corpus currus est
triumphalis Amoris. Hoc ego tuo vectus curru, Octavia, ibo
tenebricosum istud per iter» (pubem vellicabat), «ibo ad gloriam.»
Uterum, femina, pectus pererrabat inspectans curiose. Demum
intumuit puero nervus. «Fave,» inquit, «fave, Venus [pg 299] mea.»
Suavium dedit.—«Favebo,» inquam, «et ut voles favebo. Servire tibi
instar regni erit. Qualem me cupies, talem habebis.—O nugacem!»
clamat advolans Manilia; «facto opus est, non dicto. Volo utrique
præstare operam, et ope mea voluptati vestræ accrescere delicias
novas. Belle arrigis, Roberte; agedum, candidum hoc in Octaviæ
pectus libidines et te effunde.—O mater,» respondeo, «vis huic meæ
adesse ignominiæ? Abi, amabo.—Te ineptam!» refert, «non fidis
nutrici, alumna? Pugna, Roberte, sed heroica pugna fortitudine
heroidi tuæ incomparabili.» Dicente Manilia, supervolat Robertus;
balista ferit uterum: at exerrantem et resilientem intercipit caudam
Manilia, officiosam insinuans manum. «Veni, fugitiva, tuum id in
ergastulum Dionæum: huc te vocant,» dicebat, «quas dominæ
debes operæ.» Post impacta in pueri lumbos manu, impellit:
momento voro puerum, absorbeo surgentem. Vetat commoveri me
Manilia. «Tolle id femur lævum, Octavia,» inquit, «contende
alterum.» Fit. «Tu, Roberte, leni et spissa concute amores tuos
subagitatione. At tu, Octavia, osculare, nec move.» Fit. «Cum
senseritis alter et alter,» adjicit, «e libidinis vena exsilire vobis
prurientes salaciæ spumas, tu, Octavia, mitte suspirium; tu, Roberte,
morsiunculis Octaviam pete.» Subagitat ille vegeta, sed molli et lenta
concussione. Amplector, osculor, nec moveo. O bona Venus! o bona
Tullia! sentio me remitti. Suspirium mitto. «Nunc, nunc, Roberte,»
ingeminat [pg 300] nutrix lena, «fave, fave Octaviæ. Rapidi tibi
ferveant lumbi.» Ille urgere, permolere. Mox collum appetit dente,
cutem carpsit; gemitum dedi. «Nunc, nunc,» repetit Manilia,
«crissanti fave Roberto succussu. Tolle lumbos, succute pernix. Bene
est, alumna. Non ita, puto, lumborum placuit flexibilitate et
mobilitate Lais.» Cœpit colliquescere dulcis puer, et ego intimos
amoris sentio mihi sinus liquido exundare incendio. Nec lateribus
peperci, nec animæ. Celeriori nunquam itum cursu ad Veneris
metam. Altera manu Manilia mulcebat mihi nates, altera Roberto.
Apprehensa identidem summis digitis cadurda stringebat,
comprimebat, et incubantis testiculos, molli emulgens compressione,
exprimebat. Defecit puer; et refugit nutrix, plaudens manibus, acta
fabula. Basiationes projecto propter latus meum puero ingerebam
innumerabiles.—«Ut me amas?» dicebam. «Ut meæ tibi Veneris
dona placuerunt? Nunquid piget? Nunquid mei piget?—Interroga,
domina,» respondet, «an pigeat in cœlis fuisse cum Jove et Junone,
et magnis Diis.—Nam,» repono, «ruunt homines in amplexus
puellarum præcipites, postea piget.—Non tu puella es,» inquit, «ut
puellæ sunt hæ nostræ. Dea es voluptatis. In te si me ceperit
satietas, et ceperit felicitatis in cœlis, et Deorum epulis».

. . . . . . . . . . . .

. . . . . . . . . . . .

(Desunt folia aliquot in codice m. s.)

. . . . . . . . . . . .
[pg 301] ...—«Insomnem egi noctem. Perstrepebat contignatio vi
concussa: cubiculi timui casum. Fabulam egitis, nec dubito; hæc tibi
Diana sagittifer fuit Apollo. Configebat tibi, noli erubescere, Octavia
mea, configebat tuum Pythonem. Hodie mane, cum vale abiens dixit,
junxi pectori pectus, nullo Dianæ tuæ exsurgere sinum mammarum
tumore sensi, nullo vidi. Et fœmineo suavius dedit suavium. Non is
vividus et micans oris color, qui heri erat; impallescebat nocturno
labore.—Erras, Eleonora, repono. Virum non habui in viduo thoro.
Sed lucta, et pudet fateri, lusimus Cupidinea, Sappho et Andromede.
O videres, Eleonora, nascentes mammarum orbes! arderes. O
videres non vere cunnum, sed cunniculum! fureres. Placui maritus,
placuit uxor, tribadico furore et sudore.»
Tullia. Erat Enemonda forma excellens, Fernando Portio soror. Et
Enemondæ erat amica Francisca Bellina, etiam forma præstans.
Nesciebant inter se quæ magis amaret, quæ magis amaretur.
Cubabant frequenter una in domo Fernandi. Secretis petebat, quales
amat Venus, insidiis, Franciscam Fernandus; se peti sciebat puella, et
gratulabatur suæ formæ. Surrexerat impatiens libidinis, surgente
Aurora, e lecto adolescens; frigido aeris halitu mitigabat ignes, in
pergula. Tremula strepebat argutatione, proximo in cubiculo, sororis
lectus. Patebat vero ostium: hanc amanti commodarat puellarum
negligentiam Venus favens. Intrat, nec vident libidine cæcæ, libidine
ebriæ. Superequitabat Francisca; Enemondam impellebat [pg 302]
ad cursum, nuda nudam. «Ambiunt,» dicebat Francisca, «meam
quotidie pudicitiam nobiliores et salaciores mentulæ. Pulchriorem ex
iis, amica, legam ego, sed tibi. Sic volo indulgere genio tuo, et meo.»
Dicens, subagitabat acerrime. Conjicit Fernandus se in lectum nudus.
Territæ puellæ, nec fugere ausæ sunt; amplexu Franciscam ligat
cursu fessam, osculatur.—«Audes tu, improba, vitiare sororem
meam,» ait, «tam sanctam, tam castam? Pœnas dabis. Domus
ulciscar meæ injurias. Patiere furores meos, ut illa tuos.—Frater mi,
frater mi,» respondit Enemonda, «ignosce amantibus; noli nos
traduci ludibrio.—Nemo sciet,» inquit; «faveat hæc mihi cunno,
favebo ego utrique lingua. Nemo sciet.»

. . . . . . . . . . . .

(Desunt multa in m. s.)

. . . . . . . . . . . .

Octavia... Nudos pingi vetant novi sapientes, novi Catones, qui, si e

cœlo ceciderunt, sane e cœlo Lunæ ceciderunt.
Tullia. Quo quisque est inter homines stolidior et vecordior, eo et
superbior est. In eis si judicium quæres, si eruditionem, operam
luseris. Nam Naturæ rerum parens finxit nudos. Nec sartor, nec
cerdo Deus. Propulsandis aeris et tempestatum injuriis inventus
vestium usus, non quod quicquam corporibus desit nostris ad
pulchritudinis laudem; non quod quicquam inhonestum et turpe e
tanti molitoris [pg 303] manibus exierit, quod tegi deceat. Consistit
corporum pulchritudo in recta membrorum compositione; nulla est,
Hercle! in vestium ornatu. Æternæ mentis maximum est opus corpus
humanum. Quis neget? Hoc se uno ars conditoris maxime jactat
specimine. Qui tegit, damnat. Si videri noluisset, deerat facultas?
Deerat ars? Expectasset, putas, tam mirabili perficiendo operi, artes
nostras et curas? Insanit qui sic obloquitur. Maxima pars orbis
terrarum alit nudos, ut creat. Et sunt plagæ in quibus usum vetat
vestium cœli et aeris torrens æstus. Videant hi lynces. Regnantis vis
Naturæ nudos ostentat hominum artus et lumbos. Nulla illi pudenda
membra, iis in regionibus, partes turpes nullæ; nudum esse probro
non est. Volet et eadem vis Naturæ in aliis abscondi terris, ut
propudia, et tegi, quædam ut sua scelera, nullius tenax consilii?
Volet, si videantur probro esse? Viderint hi lynces. Græci, artibus et
ingenio clari, pingebant nudos heroas, et heroidas nudas. Alexandri
vidi Romæ statuam, Praxitelis opus: Hercule satus altero sustinet
brachio leonis exuvias, cetera nudus. Qui videt picta in tabula
Carolum Imperatorem, Alexandro æmulum, vultum principis et
manus videt, nihil præterea videt, quod sit principis. Hæ sunt
vestium picturæ, non hominum. Inepte dixerit te vidisse, Octavia,
qui nunquam viderit nudam. Ita sua excidere a dignitate sculptura et
pictura. Nunc sunt sculptores et pictores scientiarum et artium
rudes, si unum aut alterum demas; nebulones, ebriosi, nullius artis
in arte sua. Pictura descivit a pictura, ars ab arte. Sed [pg 304]
periculum est, aiunt, si nudos nudasve pingant, id ne sit
præstigiosum quoddam ad flagitia incentivum. Nugæ! Qui degunt,
ex nostris hominibus, in India et America, ubi libere patent
puellarum partes, nulla tamen moventur libidine. Visu assuescunt, et
hebescit assuetudine libido. Crede mihi, Octavia. Avaris nos tuemur
ipsæ curis, oculis negamus dotes nostras sollicitæ, et id maxime
hominum corda inflammat. Majora opinione concipiunt, quam postea
vident. Postquam copiam nostri fecimus, non ita ardent. Obsolescunt
momento venustates, quas nudiustertius adorabant non visas. Vetat
lex? illecebras vetitis addit. Castiores essent plerique omnes, si
liberiores. Fluat, ut aqua in rivis, vinum: ebriosos vix investigabis,
unum alterumve. Ambulent nudæ: se tot, ut fit, non inflammabit
nequitiis Amor. Quibus sunt virgines nudæ in tabulis pictis, hoc non
tanguntur visu, non moventur, marmorei usu facti diuturno. Stulti illi
sapientes rem ex sua præjudicant conscientia. Pronæ ad terram
mentes ad quæque se sentiunt flagitia proclives. Nec studio
bonarum artium, quæ in eis nulla sunt, naturæ malignitas aut
corrigitur, aut temperatur. Bonis et doctis aliæ..

. . . . . . . . . . . .

(Desunt multa in m. s.)

. . . . . . . . . . . .

..... «Accipis?—Legem quidem accipio,» respondet subridens, «omen

non capio; nam te [pg 305] amicam, soror, non inimicam, capere
volo. Ad quam flexit me scurrilitatem amentia pueri, flagitium vocas.
Ludus est. Ut ut sit, accipio legem. Sed iram averte omnem a puero;
in me converte.—Videro,» reponit Juditha. «Consobrino natus e
concubina, venusta puella, odio mihi esse non debet; nec aheneum
mihi pectus. Id unum jubeo, jubenti ut præstes obsequium, mihi,
quicquid tandem in mentem venerit.—Præstabo,» inquit Lucia.
—«Curabo omnia,» subjicit Juditha; «nec tibi time a marito. Sylvas
amat, non nuptias. Impigro venatori Cephalo non multus, scio,
Procris suæ usus. Fibulam volo induas.» Uberes ad rem manarunt ex
puellæ oculis fletus, et mitescere sibi sensit Juditha ferum animum.
Dejicit, lymphata, lintea. Ut vero apertas vidit opimas juvenilis
corporis venustates: «O Venus aurea!» exclamat; nec plura. Post
paulo vocem recepit. «Sospitabit miseræ Juno,» ait.—«Ignosce,
soror,» dicebat Lucia, «linque meo me sexui. Quæ fœmina sum, nisi
cum fibulæ visum erit, fœmina non ero? Ignosce, bona soror.»
Abscesserat Mancia, et puer.—«Quam bella es cum tuo mœrore,
Lucia soror mea, et Hebes soror!» ait Juditha. «Vis mea esse? Si
eris, et mea non ero.—Volo,» reponit Lucia, «sed infelicitatem meam
novi et tuam duritiem.» Osculum flenti fixit Juditha.—«Molliisti,»
inquit, «saxeum pectus. Hac ex caute non igniculos procudisti, sed
amoris incendia.» Impegit osculum, et novos hausit calores. «Hoc
solum rogo,» subjicit, «ut me ames, et Joannem [pg 306] habeas
odio. Servi mihi servitutem: servitute hac tibi emeris regnum.»
Promisit Lucia se facturam quæcunque vellet fieri. «Et hac tecum
nocte cubabo,» ait Juditha. «Vir tibi ero novæ nuptæ.»
Octavia. Ah! ah! ah!
Tullia. Et tota nocte puellæ lectum suis implevit furoribus; mille
millies suavia dedit, subagitavit indefessa, et incestis vexavit tenera
membra manuum deliriis; sub lucem excessit. Sequentibus diebus,
Joanni non visa est irasci, sed misero vetitum ad dominam ire. Igitur
dolore conficiebatur, et peribat. Convenit Manciam. «O Mancia,»
inquit, «insaniebam; verum te amo.» Veniunt in amplexus, et hi via
fuere Joanni ad Luciæ complexus. «Me, Mancia, mea lux, ira enecat,
emorior. Me igitur ingrata illa et perfida pessumdederit, et peribo
inultus? Non peribo si voles, vivam tibi.—Favebo,» reponit Mancia,
«sed quid agitas consilii? quid paras?—Volo superbam habere in
potestate, calcare male in luto.—Non satis intelligo,» inquit Mancia,
«quid id sit quod excogitas. Vis forte calcare in lecto, non in luto?—
Malim,» reponit Joannes, «cum Tisiphone, cui lutea inter inguina
Stygius hiat bufo, concumbere, cum Charonte rem habere.—Jura
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.

More than just a book-buying platform, we strive to be a bridge

connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.

Join us on a journey of knowledge exploration, passion nurturing, and

personal growth every day!

ebookbell.com