Full Download PDF of Advanced Differential Equations 1st Edition Youssef Raffoul - Ebook PDF All Chapter

Advanced Differential Equations 1st
Edition Youssef Raffoul - eBook PDF

Go to download the full and correct content document:
https://ebooksecure.com/download/advanced-differential-equations-ebook-pdf/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
(eBook PDF) Differential Equations 4th Edition
http://ebooksecure.com/product/ebook-pdf-differential-
equations-4th-edition/
Linear Algebra and Partial Differential Equations 1st

edition- eBook PDF
https://ebooksecure.com/download/linear-algebra-and-partial-
differential-equations-ebook-pdf/
(eBook PDF) Elementary Differential Equations 10th

Edition
http://ebooksecure.com/product/ebook-pdf-elementary-differential-
equations-10th-edition/
Schaum's Outline of Differential Equations - eBook PDF
https://ebooksecure.com/download/schaums-outline-of-differential-
equations-ebook-pdf/
First Course in differential equations (11ed) /
Differential Equations and Boundary Value Problems
(9ed) Solutions manual 9th Edition - eBook PDF
https://ebooksecure.com/download/first-course-in-differential-
equations-11ed-differential-equations-and-boundary-value-
problems-9ed-solutions-manual-ebook-pdf/
(eBook PDF) Fundamentals of Differential Equations 9th

Edition
http://ebooksecure.com/product/ebook-pdf-fundamentals-of-
differential-equations-9th-edition/
Fundamentals of Differential Equations 8th Edition

(eBook PDF)
http://ebooksecure.com/product/fundamentals-of-differential-
equations-8th-edition-ebook-pdf/
(eBook PDF) Differential Equations and Linear Algebra

3rd Edition
http://ebooksecure.com/product/ebook-pdf-differential-equations-
and-linear-algebra-3rd-edition/
(eBook PDF) Differential Equations and Linear Algebra

4th Edition
http://ebooksecure.com/product/ebook-pdf-differential-equations-
and-linear-algebra-4th-edition-2/
Advanced Differential
Equations
Youssef N. Raffoul
Professor of Mathematics
University of Dayton
Dayton, OH, United States
Copyright Elsevier 2022

Academic Press is an imprint of Elsevier
125 London Wall, London EC2Y 5AS, United Kingdom
525 B Street, Suite 1650, San Diego, CA 92101, United States
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
Copyright © 2023 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means, electronic
or mechanical, including photocopying, recording, or any information storage and retrieval system,
without permission in writing from the publisher. Details on how to seek permission, further
information about the Publisher’s permissions policies and our arrangements with organizations such
as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website:
www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the
Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this ﬁeld are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical treatment
may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating
and using any information, methods, compounds, or experiments described herein. In using such
information or methods they should be mindful of their own safety and the safety of others, including
parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume
any liability for any injury and/or damage to persons or property as a matter of products liability,
negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas
contained in the material herein.
Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress
British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library
ISBN: 978-0-323-99280-0
For information on all Academic Press publications

visit our website at https://www.elsevier.com/books-and-journals
Publisher: Katey Birtcher

Editorial Project Manager: Rafael G. Trombaco
Production Project Manager: Prem Kumar Kaliamoorthi
Designer: Margaret Reid
Typeset by VTeX

To my wonderful wife Nancy, who is
the wind beneath my wings.

Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
CHAPTER 1 Preliminaries and Banach spaces . . . . . . . . . . . . . . . . 1
1.1 Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Escape velocity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3 Applications to epidemics . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4 Metrics and Banach spaces . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.5 Variation of parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.5.1 RC circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.6 Special differential equations . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
CHAPTER 2 Existence and uniqueness . . . . . . . . . . . . . . . . . . . . . . . 27
2.1 Existence and uniqueness of solutions . . . . . . . . . . . . . . . . . . 27
2.2 Existence on Banach spaces . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.3 Existence theorem for linear equations . . . . . . . . . . . . . . . . . . 44
2.4 Continuation of solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.5 Dependence on initial conditions . . . . . . . . . . . . . . . . . . . . . . 48
2.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
CHAPTER 3 Systems of ordinary differential equations . . . . . . . . 57
3.1 Existence and uniqueness . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.2 x = A(t)x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.2.1 Fundamental matrix . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.2.2 x = Ax . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.2.3 Exponential matrix eAt . . . . . . . . . . . . . . . . . . . . . . . . 82
3.3 x = A(t)x + g(t) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
3.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
CHAPTER 4 Stability of linear systems . . . . . . . . . . . . . . . . . . . . . . . 103
4.1 Definitions and examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
4.2 x = A(t)x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
4.3 Floquet theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
4.3.1 Mathieu’s equation . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
4.3.2 Applications to Mathieu’s equation . . . . . . . . . . . . . . . 125
4.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
CHAPTER 5 Qualitative analysis of linear systems . . . . . . . . . . . . 129
5.1 Preliminary theorems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
5.2 Near-constant systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
5.3 Perturbed linear systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
vii

viii Contents
5.4 Autonomous systems in the plane . . . . . . . . . . . . . . . . . . . . . 139

5.5 Hamiltonian and gradient systems . . . . . . . . . . . . . . . . . . . . . 145
5.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
CHAPTER 6 Nonlinear systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
6.1 Bifurcations in scalar systems . . . . . . . . . . . . . . . . . . . . . . . . 159
6.2 Stability of systems by linearization . . . . . . . . . . . . . . . . . . . . 168
6.3 An SIR epidemic model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
6.4 Limit cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
6.5 Lotka–Volterra competition model . . . . . . . . . . . . . . . . . . . . . 185
6.6 Bifurcation in planar systems . . . . . . . . . . . . . . . . . . . . . . . . . 192
6.7 Manifolds and Hartman–Grobman theorem . . . . . . . . . . . . . . 198
6.7.1 The stable manifold theorem . . . . . . . . . . . . . . . . . . . . 201
6.7.2 Global manifolds . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
6.7.3 Center manifold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
6.7.4 Center manifold and reduced systems . . . . . . . . . . . . . 213
6.7.5 Hartman–Grobman theorem . . . . . . . . . . . . . . . . . . . . 218
6.8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
CHAPTER 7 Lyapunov functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
7.1 Lyapunov method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
7.1.1 Stability of autonomous systems . . . . . . . . . . . . . . . . . 238
7.1.2 Time-varying systems; non-autonomous . . . . . . . . . . . 249
7.2 Global asymptotic stability . . . . . . . . . . . . . . . . . . . . . . . . . . 256
7.3 Instability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
7.4 ω-limit set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
7.5 Connection between eigenvalues and Lyapunov functions . . . 267
7.6 Exponential stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
7.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
CHAPTER 8 Delay differential equations . . . . . . . . . . . . . . . . . . . . . 287
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
8.2 Method of steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
8.3 Existence and uniqueness . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
8.4 Stability using Lyapunov functions . . . . . . . . . . . . . . . . . . . . 294
8.5 Stability using fixed point theory . . . . . . . . . . . . . . . . . . . . . . 299
8.5.1 Neutral differential equations . . . . . . . . . . . . . . . . . . . 300
8.5.2 Neutral Volterra integro-differential equations . . . . . . . 306
8.6 Exponential stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
8.6.1 Instability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
8.7 Existence of positive periodic solutions . . . . . . . . . . . . . . . . . 319
8.8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
CHAPTER 9 New variation of parameters . . . . . . . . . . . . . . . . . . . . . 329
9.1 Applications to ordinary differential equations . . . . . . . . . . . . 329
9.1.1 Periodic solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334

Contents ix
9.2 Applications to delay differential equations . . . . . . . . . . . . . . 335

9.2.1 The main inversion . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
9.2.2 Variable time delay . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
9.3 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
Student Resources
For the Partial Solutions Manual, visit the companion site:
https://www.elsevier.com/books-and-journals/book-companion/9780323992800

Preface
Differential equations are widely used by mathematicians, physicists, engineers, biol-

ogists, chemists, and scientists who work in relevant fields. They encounter the use of
differential equations in the study of Newton’s law of cooling, Maxwell’s equations,
Newton’s laws of motion, fluid dynamics equations, equations in plasma dynamics,
equations in stellar dynamics, Hook’s law, Schrödinger’s equation, acoustic wave
equation, equations in chemical kinetics, equations in thermodynamics, Einstein’s
equations for general relativity, population models, epidemics, and so on. For many
years, the author has been encouraged by the graduate students at the University of
Dayton to write a concise and reader-friendly book on the subject of advanced dif-
ferential equations. So this book grew out of lecture notes that the author has been
constantly revising and using for a graduate course in differential equations. The
book should serve as a two-semester graduate textbook in exploring the theory and
applications of ordinary differential equations and differential equations with delays.
It is intended for students who have basic knowledge of ordinary differential equa-
tions and real analysis. While writing this book, the author tried to balance rigor
and presenting the most difficult material in an elementary format by adopting easier
and friendlier notations that make the book accessible to a wide audience. It was the
author’s main intention to provide many examples to illustrate the theory conveyed
in the theorems. The author made every effort to include contemporary topics such
as the use of fixed point theory in several places to prove the existence and unique-
ness, various notions of stability, and the existence of positive periodic solutions on
Banach spaces. What makes the book appealing and distinguished from other books
is the addition of Chapters 8 and 9 on delay differential equations with advanced
topics. The author is convinced that any student who completes the whole book, es-
pecially Chapters 8 and 9, should be ready to carry on with meaningful research in
delay differential systems.
Much of the pedagogical and mathematical development of this book is influ-
enced by the author’s style of presentation. The literature on differential equations is
vast and well established, and some of the ideas found their way into this book.
Since stability is the central part of this book, namely by the Lyapunov method,
we must mention some history. Lyapunov functions are named after Alexander
Lyapunov, a Russian mathematician, who in 1892 published his book The General
Problem of Stability of Motion. Lyapunov was the first to consider the modifications
necessary in nonlinear systems to the linear theory of stability based on lineariz-
ing near a point of equilibrium. His work, initially published in Russian and then
translated to French, received little attention for many years. Interest in Lyapunov
stability started suddenly during the Cold War period when his method was found to
be applicable to the stability of aerospace guidance systems, which typically contain
xi

xii Preface
strong nonlinearities not treatable by other methods. More recently, the concept of
the Lyapunov exponent related to Lyapunov’s first method of discussing stability has
received wide interest in connection with chaos theory.
Chapter 1 deals with various introductory topics, including variation of parame-
ters formula, metric spaces, and Banach spaces.
In Chapter 2 we introduce Gronwall’s inequality that we make use of to prove the
uniqueness of solutions. We introduce theorems on the existence and uniqueness of
solutions, their dependence on initial data, and their continuation on maximal inter-
val.
In Chapter 3 we introduce systems of differential equations. We briefly discuss
how the existence and uniqueness theorems of Chapter 2 are extended to suit systems.
Then we develop the notion of the fundamental matrix as a solution and utilize it to
write solutions of non-homogeneous systems so that they can be analyzed.
Stability theory is the central part of this book. Chapters 4–8 are totally devoted
to stability. In Chapter 4 we are mainly concerned with the stability of linear systems
via the variation of parameters. The chapter also includes a nice section of Floquet
theory with its application to Mathieu’s equation.
Chapters 5 and 6 are deeply devoted to the study of the stability of linear systems,
near-linear systems, perturbed systems, autonomous systems in the plane, and stabil-
ity by linearization. Chapter 5 is ended with the study of Hamiltonians and gradient
systems. We begin Chapter 6 by looking at stability diagrams in scalar equations and
move into the study of bifurcations as it naturally arises while looking at stability.
Bifurcation occurs when the dynamics abruptly change as certain parameters move
across certain values. We end Chapter 6 by considering stable and unstable man-
ifolds, which then delves into the Hartman–Grobman theorem. The theorem says
that the behavior of a dynamical system in the domain near a hyperbolic equilibrium
point is qualitatively the same as the behavior of its linearization near this equilibrium
point, where hyperbolicity means that no eigenvalue of the linearization has a zero
real part.
Chapter 7 delves deeply into the stability of general systems using Lyapunov
functions. We prove general theorems regarding the stability of autonomous and non-
autonomous systems by assuming the existence of such Lyapunov function. We touch
on the notion of ω-limit set and its correlation to Lyapunov functions. The chapter is
concluded with a detailed discussion on exponential stability.
Chapter 8 is solely devoted to the study of delay differential equations. It con-
tains recent development in the research of delay differential equations. We begin
the chapter by pointing out how basic results from ordinary differential equations
are easily extended to delay differential equations. We introduce the method of steps
and show how to piece together a solution. The transition of moving from ordinary
differential equations to delay differential equations was made simple through the
extension of Lyapunov functions to Lyapunov functionals. Then we move on to

Preface xiii
a whole new concept, fixed point theory. The use of fixed point theory alleviates
some of the difficulties that arise from the use of Lyapunov functionals when study-
ing stability. Later on, we apply fixed point theory to the study of stability and the
existence of positive periodic solutions of neutral differential equations and neutral
Volterra integro-differential equations, respectively. We end the chapter with the use
of Lyapunov functionals to obtain necessary conditions for the exponential stability
of Volterra integro-differential equations with finite delay.
Chapter 9 deals with current research concerning the use of a new variation of
parameters formula. The objective is to introduce a new method for inverting first-
order ordinary differential equations with time-delay terms to obtain a new variation
of parameters formula that we use to study the stability, boundedness, and periodicity
of general equations in ordinary and delay differential equations.
A combination of Chapters 1–3, 6, and 7 can be used to deliver a course on non-
linear systems for engineers.
The author has not attempted to give the historical origin of the theory, except
in very rare cases. This resulted in the situation that not every reference listed in
References is mentioned in the text or the body of the book.
Exercises play an essential learning tool of the course and accompany each chap-
ter. They range from routine calculations to solving more difficult problems to open-
ended ones. Students must read the relevant material before attempting to do the
exercises.
I am indebted to Dr. Mohamed Aburakhis, who fully developed all the codes
for all the figures in the book. I like to thank the hundreds of graduate students
at the University of Dayton whom the author taught for the last 20 years and who
helped the polishing and refining of the lecture notes, most of which have become
this book.
A heartfelt appreciation to Jeff Hemmelgarn from the University of Dayton for
carefully reading the whole book and pointing out many typos.
Youssef N. Raffoul
University of Dayton
Dayton, Ohio
June 2021

CHAPTER
Preliminaries and Banach

spaces
1
We briefly discuss basic topics of ordinary differential equations and provide exam-
ples that illustrate the need for a comprehensive and systematic theory of differential
equations. In addition, we introduce metrics and Banach spaces, which we will use
throughout the book.
1.1 Preliminaries
Let R denote the set of real numbers, let I be an interval in R, and consider a function
x : I → R. We say the function x has a derivative at a point t ∗ ∈ I if the limit
x(t) − x(t ∗ )
lim
t∈I, t→t ∗ t − t∗
exists as a finite number. In this case, we adopt the notation
x(t) − x(t ∗ )
x (t ∗ ) = lim ,
t∈I, t→t ∗ t − t∗
where x (t ∗ ) is the instantaneous rate of change of the function x at t ∗ . If t ∗ is one of

the endpoints of the interval I , then the above definition of derivative becomes that
of a one-sided derivative. If x (t ∗ ) exists at every point t ∗ ∈ I , then we say that x is
dx
differentiable on I and write x (t). Throughout the book, we might use to indicate
dt

x (t). Similarly, if x (t) has a derivative function, then we call it the second derivative
of the function x(t) and denote it by x (t). For higher-order derivatives, we use the
notations
d nx
x (t), x (4) (t), . . . , x (n) (t), or .
dt n
The order of a differential equation is defined by the highest derivative present in the
equation. An nth-order ordinary differential equation is a functional relation of the
form
dx d 2 x d 3 x d nx
F t, x, , 2 , 3 , ..., n = 0, t ∈ R (1.1.1)
dt dt dt dt
Advanced Differential Equations. https://doi.org/10.1016/B978-0-32-399280-0.00007-3
1

2 CHAPTER 1 Preliminaries and Banach spaces
between the independent variable t and the dependent variable x, and its derivatives
dx d 2 x d 3 x d nx
, 2 , 3 , ..., n .
dt dt dt dt
Loosely speaking, by a solution of (1.1.1) on an interval I , we mean a function x(t) =
ϕ(t) such that

F t, ϕ(t), ϕ (t), . . . , ϕ (n) (t)
is defined for all t ∈ I and

F t, ϕ(t), ϕ (t), . . . , ϕ (n) (t) = 0
for all t ∈ I . If we require, for some initial time t0 ∈ R, a solution x(t) to satisfy the
initial conditions
dx d 2x d n−1 x
x(t0 ) = a0 , (t0 ) = a1 , (t 0 ) = a 2 , . . . , (t0 ) = an−1 , (1.1.2)
dt dt 2 dt n−1
for constants ai , i = 0, 1, 2, ..., n−1, then (1.1.1) along with (1.1.2) is called an initial
value problem (IVP).
Following the notation of (1.1.1), a first-order differential equation takes the form
dx
F t, x, = 0.
dt
dx
Hence, if we assume that we can solve for , then we have
dt
x (t) = f (t, x)
for some function f that satisfies certain continuity conditions.

Let x : I → R be a function. A differentiable function z : I → R is called an
antiderivative of the function x on the interval I if
z (t) = x(t) for all t ∈ I.
The set of all antiderivatives of x is denoted by

x(t)dt
and called the indefinite integral of the function x. When we calculate the indefi-
nite integral t 2 dt, we in fact solve the first-order differential equation x (t) = t 2 .
The family of its solutions is given by t 3 /3 + c, where c is any constant. Thus we
may write the solution as x(t) = t 3 /3 + c, which is a one-parameter family of so-
lutions, the same as the family of all the antiderivatives of t 3 . Now, if we impose
an initial condition on the differential equation, say x(t0 ) = x0 , for some initial time

1.1 Preliminaries 3
t0 and real number x0 , then the constant c is uniquely determined by the relation
x0 = t03 /3 + c. In this case the differential equation has the unique solution given by
x(t) = t 3 /3 + x0 − t03 /3. However, without imposing the condition x(t0 ) = x0 , the
differential equation would have infinitely many solutions given by x(t) = t 3 /3 + c.
Differential equations play an important role in modeling the behavior of physical
systems such as falling bodies, vibration of a mass on a spring, and swinging pen-
dulum. To illustrate the need for the theoretical study of differential equations and in
particular nonlinear ones, we examine a few examples.
Consider the first-order differential equation
x (t) = h(t)g(x), x(t0 ) = x0 , t ≥ t0 ,
where h, g : R → R are continuous.

1. If g(x0 ) = 0, then x(t) = x0 is a solution.
2. In a region where g(x) = 0, we can divide by g(x) so that
x (t)
= h(t).
g(x(t))
Separating the variables and then integrating both sides from t0 to t give

t x (s)ds t
= h(s)ds.
t0 g(x(s)) t0
Using the transformation u = x(s) with x(t0 ) = x0 , we arrive at

x(t) t
du
= h(s)ds.
x0 g(u) t0
dG 1
If for some function G, we have = , then the above expression implies that
dx g
t
G(x(t)) − G(x(t0 )) = h(s)ds
t0
or
t
x(t) = G−1 G(x0 ) + h(s)ds ,
t0
provided that the inverse of G exists. Note that the right side of the above expres-
sion depends on the initial time t0 and the initial value x0 . Therefore to emphasize
the dependence of solutions on the initial data, we may write a solution x(t) in
the form x(t) = x(t, t0 , x0 ).

FIGURE 1.1
This example shows infinitely many solutions.
In the next example, we illustrate the existence of more than one solution. Consider
the differential equation
3
x (t) = x 1/3 (t), x(0) = 0, t ∈ R.
2
It is clear that x(t) = 0 is a solution. Hence we may consider a solution x1 (t) = 0 and
let

0 for t ≤ 0,
x2 (t) =
t 3/2 for t > 0,
which is also a continuous and differentiable solution. Likewise, for t1 > 0, we have

0 for t ≤ t1 ,
x3 (t) =
(t − t1 )3/2 for t > t1 .
Continuing in this way, we see that the differential equation has infinitely many solu-
tions. Similarly, if x is a solution, then −x is also a solution (see Fig. 1.1).
In the next example, we show that solutions may escape (become unbounded) in
finite time. The differential equation
x (t) = x 3 (t), x(t0 ) = x0 > 0, t ≥ t0
has the solution

1
x(t) = .
1
+ 2(t0 − t)
x02

1.1 Preliminaries 5
1
We can easily see that the solution is only valid for t < t0 + and becomes un-
2x02
bounded (escapes) as t approaches t0 + 1
from the left.
2x02
The most important application in engineering problems is Newton’s law
d 2 x(t) dx(t)
m = F [t, x(t), ]
dt 2 dt
for the position x(t) of a particle with mass m acted on by a force F , which may be
a function to time t, the position x(t), and the velocity dx(t)
dt . For example, if the force
is only due to gravity, then we have the second-order differential equation
d 2 x(t)
m = −mg,
dt 2
which has the solution
1
x(t) = − gt 2 + c1 t + c2 ,
2
where c1 and c2 are constants that can be uniquely determined by specifying the
position and velocity of the particle at some instant of time.
In the next example, we consider the problem of leaky bucket.
Example 1.1. We have a bucket with no water flowing into it and having a hole in the
bottom. If Q is the volumetric flow rate, then Qin − Qout = Qstored . Since Qin = 0,
we obtain Qstored = −Qout . Let h(t) be the height of the water in the bucket at time t,
and assume that the initial height at time t = 0 is h0 . Let v(t) be the velocity of
the leaked water (flow velocity). The volumetric flow rate Qstored can be calculated
by multiplying the velocity by the area of the bucket, that is, Qstored = Abucket dh(t)dt .
Similarly, Qout is the flow velocity multiplied by the area of the hole Ahole . It follows
that Qout = Ahole v(t).√For fluids of height h(t), the velocity of water coming out at
the bottom is v(t) = 2gh(t). By rearranging the terms, we arrive at the first-order
differential equation in h given by
dh(t) √
= −k h, h(0) = h0 , (1.1.3)
dt
Ahole
where k = 2g > 0. By separating the variables in (1.1.3) and then integrat-
Abucket
ing, we arrive at the solution
kt 2
h(t) = h0 − . (1.1.4)
2
Note that the solution
√
h(t) given by (1.1.4) decreases from the initial height h0 . More-
over, at time t ∗ = 2 kh0 , the water is completely drained (by gravity), h(t ∗ ) = 0, and

√
the bucket will remain empty or the height will remain zero after t ∗ = 2 h0
k . There-
fore we may write the solution as
⎧
⎪ √
⎨ √h − kt 2 , 0 ≤ t ≤ 2 h0 ,
0
h(t, 0, h0 ) = 2
√
k
⎪
⎩ 0, 2 h0
t> k .
Another interesting application is the projectile problem that we analyze in the

next section.
1.2 Escape velocity

Let M and R be the mass and radius of the Earth, respectively. We are interested
in finding the smallest initial velocity for a mass m to exit the Earth’s gravitational
field, the so-called escape velocity. We assume that no external forces are acting on
the system other than the gravitational force. Newton’s universal gravitational law
states that the force between two massive bodies is proportional to the product of the
masses and inversely proportional to the square of the distance between them, where
the mass of each body can be considered as concentrated at its center. For a mass m
with position x above the surface of the Earth, the force F on the mass is given by
Mm
F = −G ,
(R + x)2
where G is the proportionality constant in the universal gravitational law. The mi-
nus sign means the force on the mass m points in the direction of decreasing x. By
2
Newton’s second law, force F = m ddt x2 (mass times acceleration). When x = 0, that
is, at the Earth’s surface, the gravitational force equals −mg, where g is the gravita-
tional constant. Therefore
GM
= g, g ≈ 9.8 m/s2 .
R2
We have the second-order differential equation
d 2x M
= −G
dt 2 (R + x)2
g
=− , (1.2.1)
(1 + Rx )2
where the radius of the Earth is known to be R ≈ 6350 km. We transform Eq. (1.2.1)
2
into a first-order differential equation in terms of the velocity v by noting that ddt x2 =
dt . If we write v(t) = v(x(t)), that is, considering the velocity of the mass m as a
dv

1.2 Escape velocity 7
function of its distance above the Earth. Then using the chain rule, we have
dv dv dx
=
dt dx dt
dv
=v ,
dx
since v = dx
dt . As a consequence, (1.2.1) becomes the first-order differential equation
dv g 1
=− .
dx (1 + Rx )2 v
Suppose the mass is shot vertically from the Earth’s surface with initial velocity
v(x = 0) = v0 . Separating the variables and integrating both sides, we obtain

dx
vdv = −g ,
(1 + Rx )2
which gives
v2 gR 2
= +c
2 R+x
for some constant c. Using the initial velocity condition, we find
v02
c = −gR + .
2
Substituting c back into the solution and simplifying, we arrive at the solution
gRx
v 2 = v02 − .
R+x
The escape velocity is defined as the minimum initial velocity v0 , such that the mass
can escape to infinity. Therefore v0 = vescape when v → 0 as x → ∞. Taking the
limit, we have
gRx
0 = v02 − lim = v02 − 2gR,
x→∞ R + x
or
2
vescape = 2gR.
√
With R ≈ 6350 km and g = 127,008 km/h2 , we get vescape = 2gR ≈ 40,000 km/h.
In contrast, the muzzle velocity of a modern high-performance rifle is 4300 km/h,
which is not enough for a bullet shot into the sky to escape from Earth’s gravity.
Now we formally attempt to qualitatively analyze differential equations.

Definition 1.2.1. Let D be an open subset of R2 , and let f : D → R be a continuous

function. Let (t0 , x0 ) ∈ D. We say that x(t) = x(t, t0 , x0 ) is a solution of
x = f (t, x), x(t0 ) = x0 , (1.2.2)
on an interval I if t0 ∈ I , x : I → R is differentiable, (t, x(t)) ∈ D for t ∈ I , x (t) =

f (t, x(t)) for t ∈ I , and x(t0 ) = x0 .
Definition 1.2.2. A solution x(t) of (1.2.2) is said to be bounded on the interval
I = [0, ∞) if for any t0 ∈ [0, ∞) and r > 0, there exists a positive number α(t0 , r)
depending on t0 and r such that |x(t, t0 , x0 )| ≤ α(t0 , r) for all t ≥ t0 and x0 such that
|x0 | < r. It is uniformly bounded if α is independent of the initial time t0 .
Definition 1.2.3. Let x(t) and y(t) be solutions of (1.2.2) with respect to initial
conditions x0 and y0 , respectively. The solution x(t) is then said to be stable if for
every ε > 0, there exists δ = δ(ε, t0 ) > 0 such that
|x(t) − y(t)| < ε whenever |x0 − y0 | < δ.
Consider the linear differential equation
x (t) = 1, x(t0 ) = x0 . (1.2.3)
It is easy to check that x(t) = x0 + (t − t0 ) is the solution of (1.2.3). If y(t) is another

solution with y(t0 ) = y0 , then we have y(t) = y0 + (t − t0 ). For any ε > 0, let δ = ε.
Then
|x(t) − y(t)| = |x0 + (t − t0 ) − y0 − (t − t0 )| = |x0 − y0 | < ε
whenever |x0 − y0 | < δ. Hence the solution x(t) is stable but clearly unbounded.
This simple example shows that the notion of a solution being unbounded does not
automatically imply that the same solution is unstable with respect to another solution
starting at a different initial point. In Chapters 6 and 7 we will discuss boundedness
and stability in more detail. The previous examples illustrate the need for a coherent
theory for addressing the following issues:
1. Existence and uniqueness.
2. Boundedness and stability.
3. The dependence of solutions on the initial data.
1.3 Applications to epidemics

The law of mass action is a useful concept that describes the behavior of a system
that consists of many interacting parts, such as molecules, that react with each other,
or viruses that are passed along from a population of infected individuals to non-
immune ones. The law of mass action was first derived for chemical systems but

1.3 Applications to epidemics 9
subsequently found wide use in epidemiology and ecology. To describe the law of
mass action, we assume that m substances s1 , s2 , . . . , sm together form a product
with concentration p. Then the law of mass action states that dp dt is proportional to
the product of the m concentrations si , i = 1, . . . , m, that is,
dp
= ks1 s2 . . . sm .
dt
Suppose we have a homogeneous population of fixed size divided into two groups.
Those who have the disease are called infective, and those who do not have the dis-
ease are called susceptible. Let S = S(t) be the susceptible portion of the population,
and let I = I (t) be the infective portion. Then by assumption, we may normalize the
population and have S + I = 1. We further assume that the dynamics of this epidemic
satisfies the law of mass action. Hence, for positive constant λ, we have the nonlinear
differential equation
I (t) = λSI. (1.3.1)
Let I (0) = I0 , 0 < I (0) < 1, be a given initial condition. By substituting S = 1 − I
into (1.3.1) it follows that
I (t) = λI (1 − I ), I (0) = I0 . (1.3.2)
If we can solve (1.3.2) for I (t), then S(t) can be found from the relation I + S = 1.
We separate the variables in (1.3.2) and obtain
dI
= λdt.
I (1 − I )
Using partial fraction on the left side of the equation and then integrating both sides
yield
Ln|I | − Ln|1 − I | = λt + c,
or for positive constant c1 , we have
c1 eλt
I (t) = .
1 + c1 eλt
Applying I (0) = I0 gives the solution
I0 eλt
I (t) = . (1.3.3)
1 − I0 + I0 eλt
Now for 0 < I (0) < 1, the solution given by (1.3.3) is increasing with time as ex-
pected. Moreover, using L’Hospital’s rule, we have
I0 eλt
lim I (t) = lim = 1.
t→∞ t→∞ 1 − I0 + I0 eλt

Hence the infection will grow, and everyone in the population will eventually get
infected.
1.4 Metrics and Banach spaces

This section is devoted to introductory materials related to Cauchy sequences, metric
spaces, contraction, compactness, contraction mapping principle, and Banach spaces.
Materials in this section will be of use in several places of the book, especially in
Chapters 2, 8, and 9. Throughout the book, by C(I, Rn ), we denote the space of all
continuous functions f : I → Rn on an interval I , possibly infinite.
Definition 1.4.1. A pair (E, ρ) is a metric space if E is a set and ρ : E × E → [0, ∞)
such that for all y, z, and u in E, we have
(a) ρ(y, z) ≥ 0, ρ(y, y) = 0, and ρ(y, z) = 0 implies y = z;
(b) ρ(y, z) = ρ(z, y); and
(c) ρ(y, z) ≤ ρ(y, u) + ρ(u, z).
The next definition is concerned with Cauchy sequences.
Definition 1.4.2. (Cauchy sequence) A sequence {xn } ⊆ E is a Cauchy sequence if
for each ε > 0, there exists N ∈ N such that n, m > N =⇒ ρ(xn , xm ) < ε.
Complete metric spaces play a major role when showing that a fixed point belongs
to the metric space of interest.
Definition 1.4.3. (Completeness of metric space) A metric space (E, ρ) is said to be
complete if every Cauchy sequence in E converges to a point in E.
Definition 1.4.4. A set L in a metric space (E, ρ) is compact if each sequence in L
has a subsequence with a limit in L.
Definition 1.4.5. Let {fn } be a sequence of real functions with fn : [a, b] → R.
1. {fn } is uniformly bounded on [a, b] if there exists M > 0 such that |fn (t)| ≤ M
for all n ∈ N and t ∈ [a, b].
2. {fn } is equicontinuous at t0 if for each ε > 0, there exists δ > 0 such that for
all n ∈ N, if t ∈ [a, b] and |t0 − t| < δ, then |fn (t0 ) − fn (t)| < ε. Also, {fn } is
equicontinuous if {fn } is equicontinuous at each t0 ∈ [a, b].
3. {fn } is uniformly equicontinuous if for each ε > 0, there exists δ > 0 such that for
all n ∈ N, if t1 , t2 ∈ [a, b] and |t1 − t2 | < δ, then |fn (t1 ) − fn (t2 )| < ε.
It is easy to see that {fn } = {x n } is not an equicontinuous sequence of functions
on [0, 1] but each fn is uniformly continuous.
Proposition 1.1. [Cauchy criterion for uniform convergence] If {Fn } is a sequence
of bounded functions that is Cauchy in the uniform norm, then {Fn } converges uni-
formly.

1.4 Metrics and Banach spaces 11
Definition 1.4.6. A real-valued function f defined on E ⊆ R is said to be Lipschitz

continuous with Lipschitz constant K if |f (x) − f (y)| ≤ K|x − y| for all x, y ∈ E.
It is easy to see that the function f (x) = x 2 is not Lipschitz on R. This is due to the
fact that for any x and y in R, we have that f (x) − f (y) = |x 2 − y 2 | = |x + y||x − y|,
and so there is no constant K such that |x 2 − y 2 | ≤ K|x − y|. Definition 1.4.6 implies
that f is globally Lipschitz since the constant K is uniform for all x and y in R.
Remark 1.1. It is an easy exercise that a Lipschitz continuous function is uniformly

continuous. Also, if each fn in a sequence of functions {fn } has the same Lipschitz
constant, then the sequence is uniformly equicontinuous.
Lemma 1.1. If {fn } is an equicontinuous sequence of functions on a closed bounded

interval, then {fn } is uniformly equicontinuous.
Proof. Suppose {fn } is equicontinuous on [a, b]. Let ε > 0. For each x ∈ K, let
δx > 0 be such that |y − x| < δx =⇒ |fn (x) − fn (y)| < ε/2 for all n ∈ N. The
collection {B(x, δx /2) : x ∈ [a, b]} is an open cover of [a, b], so it has a finite
subcover {B(xi , δxi /2) : i = 1, . . . , k}. Let δ = min{δxi /2 : i = 1, . . . , k}. Then, if
x, y ∈ [a, b] with |x − y| < δ, then there is some i with x ∈ B(xi , δxi /2). Since
|x − y| < δ ≤ δxi /2, we have |xi − y| ≤ |xi − x| + |x − y| < δxi /2 + δxi /2 = δxi .
Hence |xi − y| < δxi and |xi − x| < δxi . So, for any n ∈ N, we have |fn (x) − fn (y)| ≤
|fn (x) − fn (xi )| + |fn (xi ) − fn (y)| < ε/2 + ε/2 = ε. So {fn } is uniformly equicon-
tinuous.
The next theorem gives us the main method of proving compactness in the spaces
we are interested in.
Theorem 1.4.1. [Ascoli–Arzelà] If {fn (t)} is a uniformly bounded and equicontin-

uous sequence of real-valued functions on an interval [a, b], then there is a subse-
quence that converges uniformly on [a, b] to a continuous function.
Proof. Since {fn (t)} is equicontinuous on [a, b], by Lemma (1.1) {fn (t)} is uni-
formly equicontinuous. Let t1 , t2 , . . . be a listing of the rational numbers in [a, b]
(note that the set of rational numbers is countable, so this enumeration is possi-
ble). The sequence {fn (t1 )}∞n=1 is a bounded sequence of real numbers (since {fn }
is uniformly bounded), so it has a subsequence {fnk (t1 )} converging to a num-
ber, which we denote φ(t1 ). It will be more convenient to represent this subse-
quence without subsubscripts, so we write fk1 for fnk and switch the index from
k to n. So the subsequence is written as {fn1 (t1 )}∞ n=1 . Now, the sequence {fn (t2 )} is
1
bounded, so it has a convergent subsequence, say {fn (t2 )}, with limit φ(t2 ). We con-
2
tinue in this way obtaining a sequence of sequences {fnm (t)}∞ n=1 (one sequence for
each m), each of which is a subsequence of the previous one. Furthermore, we have
fnm (tm ) → φ(tm ) as n → ∞ for each m ∈ N. Now, consider the “diagonal” functions
defined Fk (t) = fkk (t). Since fnm (tm ) → φ(tm ), it follows that Fr (tm ) → φ(tm ) as
r → ∞ for each m ∈ N (in other words, the sequence {Fr (t)} converges pointwise at

each tm ). We now show that {Fk (t)} converges uniformly on [a, b] by showing that
it is Cauchy in the uniform norm. Let ε > 0, and let δ > 0 be as in the definition of
uniformly equicontinuous for {fn (t)} applied with ε/3. Divide [a, b] into p intervals,
where p > b−a δ . Let ξj be a rational number in the j th interval for j = 1, . . . , p. Re-
call that {Fr (t)} converges at each of the points ξj , since they are rational numbers.
So, for each j , there is Mj ∈ N such that |Fr (ξj )−Fs (ξj )| < ε/3 whenever r, s > Mj .
Let M = max{Mj : j = 1, . . . , p}. If t ∈ [a, b], then it is in one of the p intervals, say
the j th. So |t − ξj | < δ, and thus |frr (t) − frr (ξj )| = |Fr (t) − Fr (ξj )| < ε/3 for ev-
ery r. Also, if r, s > M, then |Fr (ξj ) − Fs (ξj )| < ε/3 (since M is the maximum of
the Mi ). So for r, s > M, we have
|Fr (t) − Fs (t)| = |Fr (t) − Fr (ξj ) + Fr (ξj ) − Fs (ξj ) + Fs (ξj ) − Fs (t)|
≤ |Fr (t) − Fr (ξj )| + |Fr (ξj ) − Fs (ξj )| + |Fs (ξj ) − Fs (t)|
ε ε ε
≤ + + = ε.
3 3 3
By the Cauchy criterion for convergence, the sequence {Fr (t)} converges uniformly
on [a, b]. Since each Fr (t) is continuous, the limit function φ(t) is also continu-
ous.
Remark 1.2. The Ascoli–Arzelà theorem can be generalized to a sequence of func-

tions from [a, b] to Rn . Apply the Ascoli–Arzelà theorem to the first coordinate
function to get a uniformly convergent subsequence. Then apply the theorem again,
this time to the corresponding subsequence of functions restricted to the second co-
ordinate, getting a subsubsequence, and so on.
The next criterion, known as the Weierstrass M-test plays an important role in
showing the existence of solutions.
Lemma 1.2. (Weierstrass M-test) Let {fn } be a sequence of functions defined on a

set E. Suppose that for all n = 1, . . . , there is a constant Mn such that |fn (t)| ≤ Mn
for all t ∈ E. If
∞
∞

Mn < ∞, then fn (t)
n=1 n=1
converges absolutely and uniformly on the E.
We remark that the Weierstrass M-test can be easily generalized if the domain of
the sequence of functions is a subset of Banach space endowed with an appropriate
norm.
Here is an example of the Weierstrass M-test.
Example 1.2. For n = 1, 2, . . . , define the sequence of functions {fn } on R by

1 1 1
fn (t) = 2 . Then |fn (t)| = | 2 | ≤ 2 := Mn for all t ∈ R and n ≥ 1. Since
t + n2 t + n2 n

∞ ∞
1 1
the series converges, by the Weierstrass M-test the series con-
n 2 t + n2
2
n=1 n=1
verges uniformly on R. Moreover, as each term of the series is continuous and the
convergence is uniform, the sum function is also continuous. (As the uniform limit
of continuous functions is continuous.)
Here is another example with a simple twist to it.
Example 1.3. We prove that the series
∞ 2
n + x4
n4 + x 2
n=1
converges to a continuous function f : R → R.

Let c be a positive constant. Then for all x ∈ [−c, c], we have that
n2 + x 4 n2 + x 4 1 c4
≤ ≤ + := Mn .
n4 + x 2 n4 n2 n4
On the other hand, the series
∞
∞
∞

1 1
Mn = 2
+ c 4
n n4
n=1 n=1 n=1
converges, so Weierstrass M-test implies that the series converges uniformly to

a function f on the bounded interval [−c, c]. Each term in the series is continuous
and since the uniform limit of continuous functions is continuous, the limit func-
tion f is continuous on [−c, c] for every c > 0. Now since every x ∈ R lies in such
an interval for sufficiently large c, it follows that f is continuous on R. Note that the
series does not converge uniformly on R, so we cannot use the argument that the sum
is continuous on R because the series converges uniformly on R.
Banach spaces form an important class of metric spaces. We now define Banach
spaces in several steps.
Definition 1.4.7. A triple (V , +, ·) is said to be a linear (or vector) space over
a field F if V is a set and the following are true.
1. Properties of +
a. + is a function from V × V to V . Outputs are denoted x + y.
b. for all x, y ∈ V , x + y = y + x (+ is commutative).
c. for all x, y, w ∈ V , x + (y + w) = (x + y) + w (+ is associative).
d. there is a unique element of V , which we denote 0, such that for all x ∈ V ,
0 + x = x + 0 = x (additive identity).
e. for each x ∈ V , there is a unique element of V , which we denote −x, such
that x + (−x) = −x + x = 0 (additive inverse).

2. Scalar multiplication
a. · is a function from F × V to V . Outputs are denoted α · x or αx.
b. for all α, β ∈ F and x ∈ V , α(βx) = (αβ)x.
c. for all x ∈ V , 1 · x = x.
d. for all α, β ∈ F and x ∈ V , (α + β)x = αx + βx.
e. for all α ∈ F and x, y ∈ V , α(x + y) = αx + αy.
Commonly, the real numbers and complex numbers are fields in the above defini-
tion. For our purposes, we only consider the field of real numbers F = R.
Definition 1.4.8. (Normed spaces) A vector space (V , +, ·) is a normed space if for
each x ∈ V , there is a nonnegative real number x, called the norm of x, such that
for all x, y ∈ V and α ∈ R,
1. x = 0 if and only if x = 0,
2. αx = |α|x,
3. x + y ≤ x + y.
Remark 1.3. A norm on a vector space always defines a metric ρ(x, y) = x − y on
the vector space. Given a metric ρ defined on a vector space, it is tempting to define
v = ρ(v, 0). But this is not always a norm.
Definition 1.4.9. A Banach space is a complete normed vector space, that is, a vector
space (X, +, ·) with norm · for which the metric ρ(x, y) = x − y is complete.
Example 1.4. The space (Rn , +, ·) over the field R is a vector space (with the usual
vector addition + and scalar multiplication ·), and there are many suitable norms for
it. For example, if x = (x1 , x2 , . . . , xn ), then
1. x = max |xi |,
1≤i≤n

n
2. x = xi2 ,
i=1

n
3. x = |xi |,
i=1

n 1/p
4. xp = |xi |p ,p≥1
i=1
are all suitable norms. Norm 2 is the Euclidean norm: the norm of a vector is its
Euclidean distance to the zero vector, and the metric defined from this norm is the
usual Euclidean metric. Norm 3 generates the “taxi-cab” metric on R2 , and Norm 4
is the l p norm.
Throughout the book, it should cause no confusion to use | · | instead of || · || to
denote a particular norm.
Remark 1.4. Consider the vector space (Rn , +, ·) as a metric space with its metric
defined by ρ(x, y) = x − y, where · is any of the norms in Example 1.4. The

completeness of this metric space comes directly from the completeness of R, and
hence (Rn , · ) is a Banach space.
Remark 1.5. In the Euclidean space Rn , compactness is equivalent to closedness
and boundedness (Heine–Borel theorem). In fact, the metrics generated from any of
the norms in Example 1.4 are equivalent in the sense that they generate the same
topologies. Moreover, compactness is equivalent to closedness and boundedness in
each of those metrics.
Example 1.5. Let C([a, b], Rn ) denote the space of all continuous functions
f : [a, b] → Rn .
1. C([a, b], Rn ) is a vector space over R.
2. If f = max |f (t)|, where | · | is a norm on Rn , then (C([a, b], Rn ), · ) is
a≤t≤b
a Banach space.
3. Let M and K be two positive constants and define
L = {f ∈ C([a, b], Rn ) : f ≤ M; |f (u) − f (v)| ≤ K|u − v|}.
Then L is compact.
Proof (of part 3). Let {fn } be any sequence in L. The functions are uniformly
bounded by M and have the same Lipschitz constant, K. So the sequence is uni-
formly equicontinuous. By the Ascoli–Arzelà theorem there is a subsequence {fnk }
that converges uniformly to a continuous function f : [a, b] → Rn . We now show that
f ∈ L. Well, |fn (t)| ≤ M for each t ∈ [a, b], so |f (t)| ≤ M for each t ∈ [a, b] and
hence f ≤ M. Now fix u, v ∈ [a, b] and ε > 0. Since {fnk } converges uniformly
to f , there is N ∈ N such that |fnk (t) − f (t)| < ε/2 for all t ∈ [a, b] and k ≥ N . So
fixing any k ≥ N , we have
|f (u) − f (v)| = |f (u) − fnk (u) + fnk (u) − fnk (v) + fnk (v) − f (v)|
≤ |f (u) − fnk (u)| + |fnk (u) − fnk (v)| + |fnk (v) − f (v)|
< ε/2 + K|u − v| + ε/2 = K|u − v| + ε.
Since ε > 0 was arbitrary, |f (u) − f (v)| ≤ K|u − v|. Hence f ∈ L. We have demon-
strated that {fn } has a subsequence converging to an element of L. Hence L is
compact.
Example 1.6. Consider R as a vector space over R and define the metric d(x, y) =
|x − y|
. For each x ∈ R, we define x = d(x, 0). Explain why · is not a
1 + |x − y|
norm on R.
Example 1.7. Let φ : [a, b] → Rn be continuous, and let S be the set of continuous
functions f : [a, c] → Rn with c > b and f (t) = φ(t) for a ≤ t ≤ b. Define ρ(f, g) =
f − g = sup |f (t) − g(t)| for f, g ∈ S. Then (S, ρ) is a complete metric space
a≤t≤c<
but not a Banach space since f + g is not in S.

Example 1.8. Let (S, ρ) be the space of continuous bounded functions f : (−∞, 0] →
R with ρ(f, g) = f − g = sup |f (t) − g(t)|. Then:
−∞<t≤0
1. (S, ρ) is a Banach space.

2. The set L = {f ∈ S : f ≤ 1, |f (u)f (v)| ≤ |u − v|} is not compact in (S, ρ).
Proof (of 2). Consider the sequence of functions defined

0 if t ≤ −n,
fn (t) =
n + 1 if −n < t ≤ 0.
t
Then the sequence converges pointwise to f = 1, but ρ(fn , f ) = 1 for all n ∈ N.

So there is no subsequence of {fn } converging in the norm · (i.e., converging
uniformly) to f .
Example 1.9. Let (S, ρ) be the space of continuous functions f : (−∞, 0] → Rn

with
∞

ρ(f, g) = 2−n ρn (f, g)/{1 + ρn (f, g)},
n=1
where
ρn (f, g) = max |f (s) − g(s)|,
−n≤s≤0
and | · | is the Euclidean norm on Rn . Then:

1. (S, ρ) is a complete metric space. The distance between all functions is bounded
by 1.
2. (S, +, ·) is a vector space over R.
3. (S, ρ) is not a Banach space because ρ does not define a norm, since ρ(x, 0) =
x does not satisfy αx = |α|x.
4. Let M and K be given positive constants. Then the set
L = {f ∈ S : f ≤ M on (−∞, 0], |f (u) − f (v)| ≤ K|u − v|}
is compact in (S, ρ).
Proof (of 4). Let {fn } be a sequence in L. It is clear that if fn → f uniformly on

compact subsets of (−∞, 0], then we have ρ(fn , f ) → 0 as n → ∞. Let us be-
gin by considering {fn } on [−1, 0]. Then the sequence is uniformly bounded and
equicontinuous, and so there is a subsequence, say {fn1 }, converging uniformly to
some continuous f on [−1, 0]. Moreover, the argument of Example 1.5 shows that
|f (t)| ≤ M and |f (u) − f (v)| ≤ K|u − v|. Next, we consider {fn1 } on [−2, 0].
Then the sequence is uniformly bounded and equicontinuous, and so there is a sub-
sequence, say {fn2 }, converging uniformly, say, to some continuous f on [−2, 0].

Continuing this way, we arrive at Fn = fnn , which has a subsequence of {fn } con-
verging uniformly on compact subsets of (−∞, 0] to a function f ∈ L. This proves
that L is compact.
We leave the proof of next result to the reader.
Theorem 1.4.2. Let g : (−∞, 0] → [1, ∞) be a continuous strictly decreasing func-

tion with g(0) = 1 and g(r) → ∞ as r → −∞. Let (S, | · |g ) be the space of
continuous functions f : (−∞, 0] → Rn for which
|f (t)|
|f |g := sup
−∞<t≤0 |g(t)|
exists. Then:
1. (S, | · |g ) is a Banach space.
2. Let M and K be given positive constants. Then the set
L = {f ∈ S : f ≤ M on (−∞, 0], |f (u) − f (v)| ≤ K|u − v|}
is compact in (S, ρ).
Definition 1.4.10. Let (E, ρ) be a metric space, and let D : E → E. The operator or
mapping D is a contraction if there exists α ∈ (0, 1) such that

ρ D(x), D(y) ≤ αρ(x, y).
The next theorem is known by the name of Caccioppoli theorem, or the Banach
contraction mapping principle [9]. A proof can be found in many places, such as
Burton [15] or Smart [58].
Theorem 1.4.3. (Contraction mapping principle) Let (E, ρ) be a complete metric

space, and let D : E → E be a contraction operator. Then there exists a unique
φ ∈ E such that D(φ) = φ. Moreover, if ψ ∈ E and if {ψn } is defined inductively by
ψ1 = D(ψ) and ψn+1 = D(ψn ), then ψn → φ, the unique fixed point.
Proof. Let y0 ∈ E and define the sequence {yn } in E by y1 = Dy0 , y2 = Dy1 =

D(Dy0 ) = D 2 y0 , . . . , yn = Dyn−1 = D n y0 . Next, we show that {yn } is a Cauchy
sequence. Indeed, if m > n, then
ρ(yn , ym ) = ρ(D n y0 , D m y0 )
≤ αρ(D n−1 y0 , D m−1 y0 )
..
.
≤ α n ρ(y0 , ym−1 )


≤ α n ρ(y0 , y1 ) + ρ(y1 , y2 ) + . . . + ρ(ym−n−1 , ym−n )

≤ α n ρ(y0 , y1 ) + αρ(y0 , y1 ) + . . . + α m−n−1 ρ(y0 , y1 )

≤ α n ρ(y0 , y1 ) 1 + α + . . . + α m−n−1
1
≤ α n ρ(y0 , y1 ) .
1−α
Thus, since α ∈ (0, 1), we have that
ρ(yn , ym ) → 0 as n → ∞.
This shows that the sequence {yn } is Cauchy. Since (E, ρ) is a complete metric space,
{yn } has a limit, say y in E. Since the mapping D is continuous, we have that
D(y) = D( lim yn ) = lim D(yn ) = lim yn+1 = y,

n→∞ n→∞ n→∞
and y is a fixed point. It remains to show that y is unique. Let x, y ∈ E be such that
D(x) = x and D(y) = y. Then

0 ≤ ρ(x, y) = ρ D(x), D(y) ≤ αρ(x, y),
which implies that

0 ≤ (1 − α)ρ(x, y) ≤ 0.
Since 1 − α = 0, we must have ρ(x, y) = 0, and hence x = y. This completes the
proof.
Another form of the contraction mapping principle:
Theorem 1.4.4. (Contraction mapping principle, Banach fixed point theorem) Let
(E, ρ) be a complete metric space, and let P : E → E be such that P m is a con-
traction for some fixed positive integer m. Then there is a unique x ∈ E such that
P (x) = x.
1.5 Variation of parameters

In this section we develop the variation of parameters formula, which we use, in one
form or another, throughout this book. Consider the nonhomogeneous differential
equation
x (t) = a(t)x(t) + g(t, x(t)), x(t0 ) = x0 , t ≥ t0 ≥ 0, (1.5.1)
where g ∈ C([0, ∞) × R, R) and a ∈ C([0, ∞), R).

1.5 Variation of parameters 19
t
− t0 a(u)du
Multiplying both sides of (1.5.1) by the integrating factor e and observ-
ing that
d t
− a(u)du
t
− a(u)du
t
− a(u)du
x(t)e t0 = x (t)e t0 − a(t)x(t)e t0 ,
dt
we arrive at
d
− t a(u)du

− t a(u)du
x(t)e t0 = g(t, x(t))e t0 .
dt
Integration of the above expression from t0 to t and using x(t0 ) = x0 yield
t
− t a(u)du − s a(u)du
x(t)e t0 = x0 + g(s, x(s))e t0 ds,
t0
from which we get

t t t
a(u)du
x(t) = x0 e t0
+ g(s, x(s))e s a(u)du
ds, t ≥ t0 ≥ 0. (1.5.2)
t0
It can be easily shown that if x(t) satisfies (1.5.2), then it satisfies (1.5.1). Expression
(1.5.2) is known as the variation of parameters formula. Note that (1.5.2) is a func-
tional equation in x since the integrand is a function of x. If we replace the function
g with a function h ∈ C([0, ∞), R), then (1.5.2) takes the special form
t t t
t0 a(u)du
x(t) = x0 e + h(s)e s a(u)du ds, t ≥ t0 ≥ 0. (1.5.3)
t0
Another special form of (1.5.2) is that if the function a(t) is constant for all t ≥ t0
and g is replaced with h as before, then from (1.5.3), we have that
t
x(t) = x0 e a(t−t0 )
+ ea(t−s) h(s)ds, t ≥ t0 ≥ 0. (1.5.4)
t0
It is easy to compute, using (1.5.4), that the differential equation
x (t) = 2x(t) + t, x(0) = 3,
has the solution

13 2t t 1
x(t) = e − − .
4 2 4
Remark 1.6. The variation of parameters formula given by (1.5.2) is valid for
g ∈ C(R × R, R) and a ∈ C(R, R).
Next, we consider the following application regarding the variation of parameters.

i
(a)
(b) R
+
η C
−
FIGURE 1.2
RC circuit.
1.5.1 RC circuit
Fig. 1.2 shows a resistor R and capacitor C connected in a series. The battery con-
nected to this circuit by a switch provides an electromotive force, or emf force η.
Initially, there is no charge on the capacitor. When the switch is flipped to (a), the
battery connects, and the capacitor charges. Similarly, when the switch is flipped
to (b), the battery disconnects, and the capacitor discharges with energy dissipated
in the resistor. Our aim is to determine the voltage drop across the capacitor during
charging and discharging.
We start by observing that the voltage drops across a capacitor VC and resistor VR
are given by
q
VC = , VR = iR,
C
respectively, where C is the capacitance and R is the resistance. The current i and the
charge q are related by the relation
dq
i= .
dt
From the first equation, we have
dVC dq 1 i 1 VR
= . = = ,
dt dt C C C R
which implies that

dVC
VR = RC .
dt

1.6 Special differential equations 21
Kirchhoff’s voltage law states that the emf η in any closed loop is equal to the sum
of the voltage drops in that loop. When the switch is thrown to (a), this gives
VR + VC = η.
Substituting VR into VR + VC = η gives the linear differential equation

dVC 1 1
+ VC =η , VC (0) = 0.
dt RC RC
An application of the variation of parameters formula gives the solution
t
η s/RC
VC (t) = e−t/RC e ds,
0 RC
or

VC (t) = η 1 − e−t/RC .
Thus the voltage starts at zero and rises slowly but exponentially to η with character-
istic time scale given by RC. In other words,
VC (t) → η as t → ∞.
Now suppose the switch is thrown to (b). Then by Kirchhoff’s voltage law
VR + VC = 0,
which results in the differential equation

dVC 1
+ VC = 0, VC (0) = 0.
dt RC
Then the solution during the discharge phase is given by
VC (t) = ηe−t/RC ,
and the voltage decays exponentially to zero with the characteristic time scale given
by RC.
1.6 Special differential equations

Next, we consider second-order nonlinear differential equations of the form
x = f (t, x, x ), (1.6.1)
where f (t, x, y) is a function of three variables defined in the region D = {(t, x, y) :

a1 < t < a2 , b1 < x < b2 , c1 < y < c2 } with f , ∂f ∂f
x , and ∂y continuous in D. Little is

known about how to solve (1.6.1) except in certain particular cases. Two interesting
and useful cases occur when (1.6.1) has one of the following forms:
F (t, x , x ) = 0 (1.6.2)
or
F (x, x , x ) = 0. (1.6.3)
x
For (1.6.2), the transformation = u transforms the original second-order differ-
ential equation into a first-order differential equation F (t, u, u ) = 0, which can be
solved. To see this, we consider the nonlinear second-order differential equation
tx + 2x + t = 1, t > 0.
For x = u, we have the first-order differential equation tu + 2u + t = 1. Using the

variation of parameters formula given by (1.5.2), we arrive at
c1 1 t
u= + −
t2 2 3

for some constant c1 . Since x(t) = u(t)dt, we arrive at the general solution
c1 t t2
x(t) = − + − + c2
t 2 6
for some constant c2 .
Equations of the form (1.6.3) can be reduced to first-order equations by the substi-
tution x = u. Eq. (1.6.3) is now replaced by the equivalent system of two differential
equations
x = u, F (x, u, u ) = 0.
By using x as the independent variable we have
du du dx du
x = = = u.
dt dx dt dx
Hence (1.6.3) reduces to the first-order differential equation
du
F (x, u, u ) = 0.
dx
As an application, we consider the differential equation
x + (x )2 x + x x = 0, x(0) = 0, x (0) = −1.
Then, under the mentioned substitution, we have the new first-order differential equa-
tion
du
u + u2 x + ux = 0
dx

1.7 Exercises 23
or
du
+ xu = −x.
dx
By the variation of parameters, we arrive at the solution
x2
u = −1 + ce 2 .
Using u = −1 when x = 0, we have c = 0. Now using the transformation x = u, it

follows that
dx
= −1.
dt
Using x(0) = 0, we arrive at the solution
x(t) = −t.
1.7 Exercises
Exercise 1.1. Show that the differential equation
3
x (t) = x 1/3 (t), x(0) = 0, t ∈ R
2
has infinitely many solutions.
Exercise 1.2. Show that the differential equation
x (t) = x α (t), x(0) = 0, t ∈ R
has infinitely many solutions for α ∈ (0, 1).
Exercise 1.3. Find the escape time of the solutions for the differential equation
x (t) = x 2 (t), x(t0 ) = x0 = 0, t ∈ R.
Exercise 1.4. Show that the solution x(t, t0 , x0 ) of
x (t) = t, x(t0 ) = x0 = 0, t ≥ 0
is stable but unbounded.
Exercise 1.5. Let x = (x1 , x2 , . . . , xn ), xi ∈ R, i = 1, 2, . . . , n.

n
Show that x = xi2 does not define a norm.
i=1

Exercise 1.6. Consider the following power series L(x), which is also known as
Euler’s dilogarithm function:
∞ n
x
L(x) = .
n2
n=1
(a) Compute the domain of convergence for L(x). Be sure to give a full analysis of
the endpoints.
(b) Show that L(x) uniformly converges on its entire domain of convergence. (Use
Weierstrass M-test.)
(c) Explain why the L(x) is continuous on its domain of convergence.
Exercise 1.7. Consider the sequence {fn } of functions fn : R → R defined by

nx
fn (x) = √ .
1 + n2 x 2
Find the pointwise limit of this sequence as n → ∞. Does the sequence converge
uniformly on R?
Exercise 1.8. Show that the space C(I, R), where I is any compact subset of R,
endowed with the supremum norm
||f ||∞ = sup |f (t)|

t∈I
is a Banach space.
Exercise 1.9. Show that the space T of all continuous T -periodic functions given
by
T = {f ∈ C(R, R) : f (t + T ) = f (T ) for all t ∈ R}
with the maximum norm
||f || = max |f (t)|
t∈[0,T ]
is a Banach space.
Exercise 1.10. Show that the space C([0, 1], R) endowed with the L1 norm
1
||f ||1 = |f (t)|dt
0
is a metric space but is not complete.
Exercise 1.11. For 1 ≤ p < ∞, let

lp = {a = (a1 , a2 , . . .) : ak ∈ R, |ak |p < ∞}
k

1.7 Exercises 25
and
1/p
||a||p = |ak |p .
k
Show that lp endowed with || · ||p is a Banach space.
Exercise 1.12. Prove Remark 1.1.
Exercise 1.13. Find a continuous solution satisfying
x (t) + x(t) = f (t), x(0) = 0,

1, 0 ≤ t ≤ 1,
where f (t) =
0, t > 1.
Is the solution differentiable at t = 1?
Exercise 1.14. Find a continuous solution satisfying
x (t) + 2tx(t) = f (t), x(0) = 2,

t, 0 ≤ t < 1,
where f (t) =
0, t ≥ 1.
Is the solution differentiable at t = 1?
Exercise 1.15. For the SI epidemic model of Section 1.3, use Eq. (1.3.1) and the
relation I + S = 1 to obtain a first-order differential equation in terms of S and S .
Assume a positive condition S(0) = 1 − I (0) := S0 and find the solution S(t) and
lim S(t).
t→∞
Exercise 1.16. Let d > 0, and let f : [0, d] → R be continuous. Let k be any constant
and suppose x(t) solves the differential inequality
x (t) ≤ kx(t) + f (t), t ∈ [0, d].
Show that
t
x(t) ≤ x(0)ekt + ek(t−s) f (s)ds, t ∈ [0, d].
0
Hint: There is a continuous and nonnegative function g(t) such that
x (t) + g(t) = kx(t) + f (t), t ∈ [0, d].
Exercise 1.17. Suppose a function f : [0, d] × R → R is continuous and that f (t, x)

is bounded for all (t, x) ∈ [0, d] × R, where d is a positive constant. Suppose x = ϕ(t)
solves the differential equation
x (t) = xf (t, x), x(0) = 1, t ∈ [0, d].
Show that there is a constant K such that ϕ(t) ≤ eKt for all t ∈ [0, d].

Exercise 1.18. In Problems 1–5 solve the differential equation. If no initial condi-
tions are given, find the general solution.
1. t 2 x + t = 1, t > 0.
2. t 2 x + 2tx = 1, t > 0.
3. xx − 2(x )2 + 4x 2 = 0, x(1) = 1, x (1) = 2.
4. sin(x )x = sin(x), x(1) = 2, x (1) = 1.
5. tx − x = t 2 et , t > 0.

CHAPTER
Existence and
uniqueness
2
In Chapter 1 we laid out the basics we need to use in this and the following chap-
ters. This chapter is devoted to the existence and uniqueness and the continuation of
solutions. We limit our discussions and proofs to scalar differential equations, and in
Chapter 3 we indicate how the theory can be naturally extended to vector equations.
2.1 Existence and uniqueness of solutions

This section is devoted to the existence and uniqueness of solutions of the initial value
problem (IVP)
x = f (t, x), x(t0 ) = x0 , (2.1.1)
where we assume that f : D → R is continuous and D is a subset of R × R. In the
case the differential equation (2.1.1) is linear, a solution can be found. However, in
general, this approach is not feasible when the differential equation is not linear, and
hence another approach must be indirectly adopted that establishes the existence of
a solution of (2.1.1). For the development of the existence theory, we need a broader
definition of Lipschitz condition.
Definition 2.1.1. The function f : D → R is said to satisfy the global Lipschitz
condition in x if there exists a Lipschitz constant k > 0 such that
|f (t, x) − f (t, y)| ≤ k|x − y| for (t, x), (t, y) ∈ D. (2.1.2)
Definition 2.1.2. The function f : D → R is said to satisfy a local Lipschitz con-

dition in x if for any (t1 , x1 ) ∈ D, there exists a domain D1 ⊂ D such that f (t, x)
satisfies a Lipschitz condition in x on D1 , that is, there exists a positive constant k1
such that
|f (t, x) − f (t, y)| ≤ K1 |x − y|, for (t, x), (t, y) ∈ D1 . (2.1.3)
Definition 2.1.1 can be easily extended to functions f : D → Rn , where D ⊂

R × Rn under a proper norm. Let R = {(t, x) : |t| ≤ a, |x| ≤ b} be any rectangle
in D. If f and ∂f
∂x are continuous on R, which is the case in this chapter, then f and
∂f
∂x are bounded on R. Therefore there exist positive constants M and K such that
∂f
|f (t, x)| ≤ M and | |≤K (2.1.4)
∂x
Advanced Differential Equations. https://doi.org/10.1016/B978-0-32-399280-0.00008-5
27

28 CHAPTER 2 Existence and uniqueness
for all points (t, x) in R. Now for any two points (t, x1 ), (t, x2 ) in R, by the mean
value theorem there exists a constant η ∈ (x1 , x2 ) such that
∂f
f (t, x1 ) − f (t, x2 ) = (t, η)(x1 − x2 ),
∂x
from which it follows that
∂f
|f (t, x1 ) − f (t, x2 ≤ | (t, η)||x1 − x2 |
∂x
≤ K |x1 − x2 |. (2.1.5)
∂f
We have shown that if f and ∂y are continuous on R, then f satisfies a global Lips-
chitz condition on R.
Example 2.1. Consider f (t, x) = x 2/3 in the rectangle R = {(t, x) : |t| ≤ 1, |x| ≤ 1}.
We claim f does not satisfy the Lipschitz condition on R. Consider the pair of points
(t, x1 ) and (t, 0) in R where x1 > 0. Then
|f (t, x1 ) − f (t, 0) −1/3
| = x1 → ∞ as x1 → 0+ ,
x1 − 0
and hence there exists no K such that (2.1.5) holds.
Remark 2.1. Consider (2.1.1) on the rectangle D ⊂ R × R defined by
D = {(t, x) : |t − t0 | ≤ a, |x − x0 | ≤ b},
where a and b are positive constants. Let
τ = t − t0 , u = x − x0 , and g(τ, u) = f (τ + t0 , u + x0 ).
Then the function g is defined on the rectangle
D ∗ = {(τ, u) : −a ≤ τ ≤ a, −b ≤ u ≤ b},
and the IVP (2.1.1) is equivalent to the IVP problem
u = g(τ, u), u(0) = 0.
For emphasis, we restate the following definition.

Definition 2.1.3. We say that x is a solution of (2.1.1) on an interval I if x : I → R
is differentiable, (t, x(t)) ∈ D for t ∈ I , x (t) = f (t, x(t)) for t ∈ I , and x(t0 ) = x0
for (t0 , x0 ) ∈ D.
In preparation for the next theorem, we observe that the IVP (2.1.1) is equivalent
to
t
x(t) = x0 + f (s, x(s))ds. (2.1.6)
t0

2.1 Existence and uniqueness of solutions 29
Relation (2.1.6) is an integral equation since it contains an integral of the unknown

function x. This integral is not a formula for the solution, but rather it provides an-
other relation satisfied by a solution of (2.1.1). We have the following definition.
Definition 2.1.4. We say x : I → R is a solution of the integral equation (2.1.6) on
an interval I if t0 ∈ I , x is continuous on I , (t, x(t)) ∈ D for t ∈ I , and (2.1.6) is
satisfied for t ∈ I .
The next theorem is fundamental for the proof of the existence theorems.
Theorem 2.1.1. Let D be an open subset of R2 , and let (t0 , x0 ) ∈ D. Then x is
a solution of (2.1.1) on an interval I if and only if x satisfies the integral equation
given by (2.1.6) on I .
Proof. Let x(t) be a solution of (2.1.1) on an interval I . Then t0 ∈ I , x is differen-
tiable on I , and hence x is continuous on I . Moreover, (t, x(t)) ∈ D for t ∈ I , x(t0 ) =
x0 , and x (t) = f (t, x(t)) for t ∈ I . Now an integration of x (t) = f (t, x(t)) from t0
to t gives (2.1.6) for t ∈ I . For the converse, if x satisfies (2.1.6) for t ∈ I , then t0 ∈ I ,
and x is continuous on I . Moreover, (t, x(t)) ∈ D for t ∈ I , and (2.1.6) is satisfied
for t ∈ I . Thus x(t) is differentiable on I . By differentiating (2.1.6)
t with respect to t,
we arrive at x (t) = f (t, x(t)) for all t ∈ I and x(t0 ) = x0 + t00 f (s, x(s))ds = x0 .
This completes the proof.
A graduate textbook in differential equations would be incomplete without the
statement and proof of Picard’s local existence and uniqueness using successive ap-
proximations. In the next section we prove similar theorems on Banach spaces using
fixed point theory.
We note that Picard’s local existence and uniqueness theorem and Cauchy–Peano
existence theorem are widely used, and variant proofs of the theorems can be found
in [17], [29], [31], [54], and [70].
The heart of proving our next results is the construction of a sequence of functions
that converge to a limit function that satisfies the IVP (2.1.1), although the members
of the sequence individually do not.
Theorem 2.1.2 addresses the three basic issues:
(1) Members of the constructed sequence {xn } exist for all time.
(2) The sequence converges, and the limiting function satisfies the integral equation
(2.1.6) and hence the IVP (2.1.1).
(3) The limiting function, which is a solution of (2.1.1), is unique.
Theorem 2.1.2. (Picard’s local existence and uniqueness) Let D ⊂ R × R be defined
as
D = {(t, x) : |t − t0 | ≤ a, |x − x0 | ≤ b},
where a and b are positive constants. Assume that f ∈ C(D, R) and f satisfies the
Lipschitz condition (2.1.2). Let
M = max |f (t, x)| (2.1.7)

(t,x)∈D

and
b
h = min{a, }. (2.1.8)
M
Then the IVP (2.1.1) has a unique solution, denoted by x(t, t0 , x0 ), on the interval
|t − t0 | ≤ h and passing through (t0 , x0 ). Furthermore,
|x(t) − x0 | ≤ b for |t − t0 | ≤ h.
Proof. Before we begin the proof, we recommend to look at Fig. 2.1.
FIGURE 2.1
Interval of existence.
First, we note that since f is continuous on D, condition (2.1.7) is automatically

implied. The idea of the proof is to construct a sequence of functions {xn } that con-
verges uniformly to a unique x on the interval |t − t0 | ≤ h and is a solution of (2.1.1).
We begin by successively defining a sequence of functions {xn } for |t − t0 | ≤ h by
setting
x0 (t) = x0 ,
t
x1 (t) = x0 + f (s, x0 (s))ds,
t0
..
.
t
xn (t) = x0 + f (s, xn−1 (s))ds. (2.1.9)
t0
According to (2.1.7) and (2.1.8), we have that
|x1 (t) − x0 | ≤ hM ≤ b for |t − t0 | ≤ h.

t
Therefore f (s, x1 (s))ds can be defined for |t − t0 | ≤ h, and hence
t0
|x2 (t) − x0 | ≤ hM ≤ b.
Similarly, we may show that x3 (t), . . . , xn (t) are well defined on |t − t0 | ≤ h. Thus
t
|xn (t) − x0 | = | f (s, xn−1 (s))ds| ≤ M|t − t0 |. (2.1.10)
t0
Next, we show that {xn } converges uniformly to a function, say x, on |t − t0 | ≤ h.

Note that
xn (t) = x0 (t) + [x1 (t) − x0 (t)] + [x2 (t) − x1 (t)] + · · · + [xn (t) − xn−1 (t)],
or

n−1
xn (t) = x0 (t) + [xj +1 (t) − xj (t)].
j =0
Taking the limit if it exists, we have

n−1
lim xn (t) = x0 (t) + lim [xj +1 (t) − xj (t)]. (2.1.11)
n→∞ n→∞
j =0
Our next task is to compute |xj +1 (t) − xj (t)|. Using (2.1.9) and then (2.1.2) followed
by (2.1.10) give
t
|x2 (t) − x1 (t)| ≤ f (s, x1 (s)) − f (s, x0 (s))ds
t0
t
≤K x1 (s) − x0 (s)ds
t0
t t
≤ KM |s − t0 |ds = KM (s − t0 )ds
t0 t0
(t − t0 )2
= KM , |t − t0 | ≤ a.
2
In a similar fashion, we obtain
t
|x3 (t) − x2 (t)| ≤ f (s, x2 (s)) − f (s, x1 (s))ds
t0
t
≤K x2 (s) − x1 (s)ds
t0

t (s − t0 )2
≤K KM ds
t0 2
(t − t0 )3
= MK 2 , |t − t0 | ≤ a.
3!
Continuing this way, we arrive at
(t − t0 )j
|xj (t) − xj −1 (t)| ≤ MK j −1 . (2.1.12)
j!
To complete the induction argument, we assume that (2.1.12) holds for j and show
that it holds for j + 1. Using (2.1.12), we arrive at
t
|xj +1 (t) − xj (t)| ≤ f (s, xj (s)) − f (s, xj −1 (s))ds
t0
t
≤K xj (s) − xj −1 (s)ds
t0
t (s − t0 )j
≤K MK j −1 ds
t0 j!
(t − t0 )j +1 (t − t0 )j +1
= MK j = MK j , |t − t0 | ≤ a.
j !(j + 1) (j + 1)!
Next, we substitute (2.1.12) into (2.1.11) and get
M |K(t − t0 )|j +1
n−1
lim xn (t) = x0 + lim
n→∞ K n→∞ (j + 1)!
j =0
M
n−1
(Kh)j +1
≤ x0 + lim
K n→∞ (j + 1)!
j =0
M Kh
≤ x0 + e − 1 < ∞.
K
Hence, by the Weierstrass M-test, {xn (t)} converges uniformly, say to a function x(t).
Next, we show that x(t) is a solution of (2.1.1) on D. We have
|x(t) − x0 | = |x(t) − xn (t) + xn (t) − x0 |

≤ |x(t) − xn (t)| + |xn (t) − x0 |
≤ |x(t) − xn (t)| + M|t − t0 |
for every fixed t ∈ (t0 − h, t0 + h). As a result,
|x(t) − x0 | = lim |x(t) − xn (t)| + M|t − t0 |

x→∞

= 0 + M|t − t0 |
≤ Mh ≤ b
for t ∈ (t0 − h, t0 + h). So x(t) ∈ D. If the sequence {xn (t)} converges uniformly and
{xn (t)} is continuous on the interval |t − t0 | ≤ h, then
t t
lim f (s, xn (s))ds = lim f (s, xn (s))ds
n→∞ t t0 n→∞
0
by the uniform continuity of f . As a consequence,
x(t) = lim xn+1 (t)

n→∞
t
= x0 + lim f (s, xn (s))ds
n→∞ t
0
t
= x0 + lim f (s, xn (s))ds
t0 n→∞
t
= x0 + f (s, lim xn (s))ds
t0 n→∞
t
= x0 + f (s, x(s))ds,
t0
that is,
t
x(t) = x0 + f (s, x(s))ds, |t − t0 | ≤ h. (2.1.13)
t0
The integrand f (s, x(s)) in (2.1.13) is a continuous function, and hence x(t) is dif-
ferentiable with respect to t, and its derivative is equal to f (t, x(t)). So the proof of
the existence is complete. It remains to show that the solution x(t) on D is unique.
Let y(t) be another solution of (2.1.1) with y(t0 ) = x0 . Then
t
y(t) = x0 + f (s, y(s))ds.
t0
Let
N = sup |x(t) − y(t)|.
|t−t0 |≤h
Then by the Lipschitz condition, we have

t
|x(t) − y(t)| ≤ K x(s) − y(s)ds (2.1.14)
t0
or
|x(t) − y(t)| ≤ KN |t − t0 |.

The substitution of this estimate into (2.1.14) yields
|x(t) − y(t)| ≤ KN |t − t0 |2 /(2!) for |t − t0 | ≤ h.
Repeating this substitution, we obtain

|t − t0 |m
|x(t) − y(t)| ≤ KN , m = 1, 2, . . . ,
m!
for |t − t0 | ≤ h. Since the right side of this inequality tends to zero as m → ∞, we
have that
N = sup |x(t) − y(t)| = 0.
|t−t0 |≤h
This completes the proof.

We will give another proof of the uniqueness once Gronwall’s inequality is intro-
duced. To illustrate the above procedure, we provide the following examples.
Example 2.2. We consider the IVP
x (t) = 2t (1 + x), x(0) = 0.
Set x0 (t) = 0. Then for any n ≥ 1, we have the recurrent formula

t
xn (t) = 2s(1 + xn−1 (s))ds.
0
For n = 1, we have
t
x1 (t) = 2sds = t 2 ,
0
and for n = 2,
t t4
x2 (t) = 2s(1 + s 2 )ds = t 2 + .
0 2
We leave it to the reader to verify that
t 2n (t 2 )i
n
t4 t6
xn (t) = t + + + . . . ,
2
= − 1.
2 3! n! i!
i=0
Definition 2.1.5. A sequence {xn } of functions in C([a, b], R) converges uniformly

to x ∈ C([a, b], R) if lim ||xn − x|| = 0.
n→∞
It was established in Example 2.2 that

n
(t 2 )i
xn (t) = −1
i!
i=0

2
and the true solution of the IVP is given by x(t) = et − 1 on the interval [0, 1]. Thus
∞

n
(t 2 )i 1
max − 1 − (et − 1) = lim
2
lim ||xn − x|| = lim = 0.
n→∞ n→∞ t∈[0,1] i! n→∞ i!
i=0 i=n+1
Therefore

n
(t 2 )i 2
lim xn (t) = lim − 1 = et − 1.
n→∞ n→∞ i!
i=0
Example 2.3. Next, we use Theorem 2.1.2 to find the interval of existence of the
unique solution of
x (t) = 1 + x 2 , x(0) = 1.
Let
D = {(t, x) : |t| ≤ a, |x − 1| ≤ b},
where a and b are constants. It is clear that
M = max |f (t, x)| = 1 + (1 + b)2

(t,x)∈D
and
b b
h = min{a, } = min{a, }.
M 1 + (1 + b)2
b
So we must find the maximum of the function g(b) = . Using calculus,
1 + (1 + b)2
√ √
we see
√
that the function g has its maximum at b = 2 and is given by g( 2) =
2√
2
. Thus the interval of existence and uniqueness is
1+(1+ 2)
√ √ √
2 2 2
|t| ≤ √ , or t ∈ − √ , √ .
1 + (1 + 2)2 1 + (1 + 2) 1 + (1 + 2)2
2
Now suppose in Example 2.3 we chose D = {(t, x) : |t| ≤ 10, |x − 1| ≤ b}, that
is, a = 10. Then the interval of existence found above is much smaller or included in
(−10, 10). Recall that the interval of existence is included in D.
The next corollary is an immediate consequence of Theorem 2.1.2 and the discus-
sion leading to (2.1.5).
Corollary 2.1. Suppose D ⊂ R × R, f ∈ C(D, R), and ∂f ∂x is continuous on D. Then
for any (t0 , x0 ) ∈ D, the IVP (2.1.1) has a unique solution on an interval containing t0
in its domain.
As an example, consider
x (t) = tx 1/2 .

Another random document with
no related content on Scribd:
FIG. 2. EXTERIOR GRAMMAR SCHOOL BUILDING FOR MANUAL TRAINING AND
DOMESTIC SCIENCE, EVANSTON, ILLINOIS.
FIG. 3. FLOOR PLANS OF BUILDING, EVANSTON, ILLINOIS.

The proper placing of centers in a community will depend upon the number of
pupils to be cared for, the distance they must travel to get to the center, and the
site available.
2. Division or Allotment of Time. Two divisions of time are common in
grammar school shopwork, the one-fourth and the one-half day period once a
week. In some cities manual training is given in sixth, seventh and eighth grades
of the grammar schools. In others it is given in seventh and eighth grades only.
In the former case, to the best of the author’s information, the period never
exceeds one-fourth day each week. In the latter it very frequently occupies one-
half day a week. The outline for drawing and manual training as given in this
book presupposes the one-half day period. In favor of this period of time are the
following: The pupils go and come to manual training on time out of school
hours. This is a very decided gain and permits the placing of centers so as to
accommodate schools of widely differing locations. Second, more and better
work is accomplished in a one-half day period of one year than in a one-fourth
day period for two years. In the one-fourth day period the pupil hardly gets his
tools set and adjusted when the bell signals him to begin to “clean up,” resulting
in much unprofitable effort. Our college administrators, who are responsible for
originating the short and infrequent period spread over a long period of months
or years, have long since found that better work and more of it is obtained
where the study is given a more intensive view, the total number of hours for the
course remaining the same but being condensed into less calendar time.
The chief objection offered against the one-half day period is that the pupil
becomes tired, exhausted, and therefore disinterested and troublesome before
the close of the period. Where the full two hours and a half are devoted entirely
to shopwork, especially if the shopwork is of such a nature as to make little
appeal to the interest of the pupil, this argument is valid. If, however, each
period has its recitation on assigned study and its demonstration on the new
work to be presented there remains but two hours of work requiring the student
to be on his feet and, if the interest is what it should be, very few boys will
complain of fatigue. The writer makes it a custom to give, in the place of the
conventional recess, a short five minute rest period. Boys are permitted to talk
and move about the shop but he has found that as many boys prefer to continue
their woodwork as prefer to rest.
If the one-fourth day period is to be used, it will be necessary to give
recitations and demonstrations on alternate days, and will necessitate
introducing the work lower than the seventh grade. It is hardly profitable to begin
serious, systematic work lower than the seventh grade, and when it is begun in
seventh grade it is hardly possible to make it serious with a time allotment of
less than one-half day each week.
There is not the same need for recess in shopwork as in academic work. A
five minute rest period is sufficient to permit pupils to make known to each other
their wishes or information. In this way it is possible to dismiss the pupils ten
minutes earlier than they otherwise would be, thus allowing the morning class
extra time for reaching home.
In the high school the time allotment is generally permitted to be governed by
the periods arranged for the academic subjects. The common arrangement is to
give two consecutive periods equal to two of the recitation periods of the
academic subjects for shopwork and another for drawing each day thruout the
week. If the periods are one hour each, which is unusual in high schools tho
common in colleges, but one period is given to the shop.
Where manual training has been given serious consideration in the seventh
and eighth grades of the grammar schools under competent instructors it ought
to be possible to cover the necessary benchwork in wood in the first half of the
freshman year of the high school. This will leave the second half for turning or
for benchwork in metal, preferably the latter.
To mechanical drawing the first half of the freshman year of one period each
day should be devoted, followed in the second half by freehand drawing,
perspective and design.
The mechanical drawing of the grammar schools, it will be noted in the lesson
outlines, takes the first twelve weeks or lessons of each year. Mechanical
drawing in grammar schools is usually presented in one of three ways. First, by
having the pupil make his drawing then immediately make the object drawn in
wood, carrying on woodwork and drawing side by side thruout the year. Second,
by having the pupil make the object in wood first, followed by the drawing. Third,
by taking the first ten or twelve weeks of the year for making up all the drawings
of that year, following this with a continuous application in wood.
After experimenting thru a number of years the writer finds the third practice
possesses many marked advantages. Among other things that make it more
satisfactory are the following: It permits concentration of the pupil’s attention
upon one thing at a time. Where woodwork and drawing are carried on side by
side or even where they alternate the pupil’s attention and interest are divided.
So much more interesting do the pupils find the woodwork with its freer activity
that the drawing suffers immeasurably, it being almost impossible to get
anything like the proper attitude toward the technique of drawing when the
young pupil is allowed to see the immediate application in wood all around him.
The instructor’s struggles for neatness and accuracy in the drawings are no
match for the barbarous haste of the beginner in his desire to get thru with the
drawing and get at the woodwork. It is impossible to get concentration on
drawing in a woodshop with tools all about and the knowledge on the part of the
pupil that only the drawing separates him from the tools.
The ideal way would be to have a separate drawing-room and equipment as
in high school. This, however, is impracticable in most grammar schools. The
woodworking teacher being the drawing teacher makes it impossible to utilize
both shop and separate drawing-room to advantage. The fitting up and heating
of rooms that are to be used only part of the school time makes a separate
drawing-room an unwarranted expense in grade schools. A satisfactory
substitute is to utilize the woodshop benches for drawing benches but to remove
all tools, having it distinctly understood that ten or twelve weeks are for drawing,
and that, no matter how many drawings are produced by a pupil, he will begin
no woodwork until the time allotted to drawing is up. It becomes possible to
secure the right attitude toward the drawing. By this concentration of attention
both drawing and woodwork are the gainers.
Second, it enables the shop instructor to tell what supplies are going to be
needed for the woodwork and to get them delivered in time without returning
from his summer’s vacation several weeks before school begins. In the twelve
weeks of drawing the woodworking tools and equipment can be looked over and
put in order in plenty of time without breaking into the summer months that
belong to the instructor. Where the woodwork begins at the beginning of school
in September the instructor must either take the fore part of his vacation at the
close of school to put his tools in shape or, if he has them simply cleaned and
vaselined by the pupils and stores them for the summer, he must come back
several weeks before school. This is true whether he does his own sharpening
or has it done, and the advantage in having woodwork begin some weeks later
than school is very manifest.
Third, this latter arrangement gives the pupil an intelligent preview of the
whole year’s work in wood thru the drawings he makes in the first ten or twelve
weeks.
Mechanical drawing, even in the grades, has a right to a clean, quiet, well
lighted room without unnecessary distractions either to the eye or ear. This, with
a definite understanding on the pupil’s part that drawing technique is the major
and the utility of the drawing the minor consideration, should put the pupil in the
right attitude toward his drawing work and make it possible to secure the best
drawings he is capable of producing. No one, not even a finished draftsman,
could produce good drawings surrounded by the noise and dust of neighboring
woodworkers. Under the alternating system there are always slow pupils who, if
they finish their drawings before they make the application, must do it while the
others are working in wood. Add to the noise and dust this pupil’s feeling that he
too ought to be at his woodwork and the limit of unfavorable conditions for
producing a drawing are reached. Making the year’s drawings the first twelve
weeks of the year enables one to avoid these unfavorable conditions.
Fourth, this arrangement makes possible a graduated transition from the
quietness and restrictedness of the academic class room to the noise and
greater freedom of the woodshop.
When beginning pupils come to the grammar school manual training shop for
the first time at the beginning of school in September, it is with an overplus of
energy and noise. To reduce these sufficiently to permit of getting anything like
satisfactory results in shopwork, the instructor is placed at once squarely before
a large problem in discipline. This problem is very greatly simplified by
introducing the pupil to ten or twelve weeks or lessons in mechanical drawing
before beginning the woodwork.
Conditions surrounding a pupil in mechanical drawing classes are very similar
to those he finds in his regular academic classes and he can readily be brought
to understand that quietness, and orderliness with seriousness of purpose are
as necessary a part of his manual training as of his academic work. After this
attitude has been fixed in the pupil’s mind in connection with his manual training
thru the mechanical drawing when the transition to woodwork is made, where
more freedom must be allowed, the pupil will be better able to distinguish
between legitimate noise and noise that is entirely unnecessary, and between
freedom and license.
3. Informational and Related Matter Pertaining to Woodwork and
Mechanical Drawing.
Closely related to any subject is a vast fund of informational matter. If the
student is to have an intelligent understanding of the subject matter, he must be
given opportunity to become acquainted with at least the most important of this
related information.
In the seventh grade the necessary study of tools and processes occupies the
pupil’s time fully. In the eighth grade opportunity offers itself for introducing such
subjects as wood structure, tree growth, lumbering, and milling. In high school,
the pupil should be made familiar with the most common woods, their
classification, characteristics, and uses.
High school pupils should be assigned outside readings on forestry. They
should secure and classify specimens of the more common woods and should
be able to recognize the tree by leaf, fruit, bark, wood and tree form. See Figs.
4, 5, and 6.
In the grammar grades, mounted specimens should be prepared illustrative of
tree structure, shrinkage, defects, etc. As in the high school, pupils should be
encouraged to seek and prepare specimens illustrative of the subjects under
consideration.
It is now possible to rent or purchase very excellent lantern slides on forestry,
lumbering, milling, etc. Add the use of these to that of the mounted specimens if
at all possible.
The detailed lesson outlines indicate definitely where these subjects are to be
given attention in the course. The pages of the text are also indicated. The high
school library should be provided with the very excellent bulletins of the United
States Department of Agriculture, Division of Forestry, most of which are for free
distribution.
4. Woodfinishing. The subject of woodfinishing is treated in a manner
quite similar to that of woodworking. No pieces of woodwork that should have a
finish are ever sent from the shop until they have been treated to a finish
calculated to make them fit for immediate placing in their future surroundings.
While the general outline of the course in woodwork makes no mention of
woodfinishing, the lesson outline indicates the gradual introduction of the
subject, beginning with the simplest finishes first and terminating in high school
in the rubbed copal varnishes.
FIG. 4. CHART ILLUSTRATING WOOD STRUCTURE.
By T. B. Kidner, October, 1908 Manual Training Magazine
In woodfinishing, as in woodworking, the aim has been to have the pupil treat
the subject in a serious and workmanlike manner. In seventh grade little
woodfinishing is done. The woodworking processes need the centering of the
pupil’s attention, in the first place. Second, the simple pieces which the beginner
is able to make require no finish as a rule. In one group stain and wax is used.
This is the group in which decorative design is emphasized. In the eighth grade
the woodfinishing problem becomes important. Almost all of the pieces require a
finish.
FIG. 5. CHART ILLUSTRATING TIMBER DEFECTS.

FIG. 6. CHART ILLUSTRATING PROPERTIES OF TREES.
The greatest obstacle to proper woodfinishing lies in the desire of the pupil to
take his piece home as soon as the woodwork is completed. Unless a definite
understanding is had with the class beforehand, proper woodfinishing is difficult
to obtain. Most boys are subject to reason, so that it is not at all necessary to
have woodfinishing slighted or to resort to makeshifts. The writer makes it a
practice to take plenty of time when the subject of woodfinishing comes up for
its first discussion to explain in detail the commercial methods of finishing fine
furniture, a piano for illustration, counting the different operations and coatings it
will receive and the labor and time expended upon the finish. A comparison is
then made between a finely rubbed finish and the cheap, sticky, unrubbed
finishes of cheap furniture.
Having established in the minds of the pupils the fact that woodfinishing is an
art second to none and that it requires time to do it well, there is not that
impatience that breeds sullen looks when the woodfinishing is to be begun after
the woodwork has been completed. The pupil will take the woodfinishing as a
matter of course and goes about it in a cheerful and manly spirit.
In grammar schools, woodfinishing has been made as simple as is consistent
with good work. Coming as the boys do but once a week and each finishing
application requiring over night for drying or hardening, the total time is quite
long even with the simple finish of filler, shellac, and wax. If the pupil wishes a
very dark finish, a stain which requires one or more periods must precede his
filler.
In high school, pupils come every day thus permitting the application of
rubbed varnish finishes, either shellac or copal, without unnecessary loss of
time. Here special finishing rooms are necessary.
5. Structural and Decorative Design. Among other requirements for a
course in woodwork and drawing as stated in the foreword is this: “At least a few
problems should be given which involve invention or design or both, thereby
stimulating individual initiative on the part of the pupils.” The present outlines in
woodwork and drawing have been planned with this in mind. In the seventh
grade the pupil is given little opportunity to exercise his initiative in either
woodwork or drawing. The reason for this, as has been previously stated, is a
firm belief that initiative in any subject to be of value must be based upon a fair
knowledge of the subject matter dealt with, its limitations and its possibilities. In
other words, that appreciation must precede invention or initiative.
With the limited time allowed manual training, at most one-half day each week
in the general educational scheme, a seventh grade beginner has about all he
can well manage in becoming familiar with his subject matter, with learning to
handle his tools and work his material.
But one group in the seventh grade will admit of decorative design. These
problems, Group VI, have purposely been made simple as to woodwork that the
pupil may give most of his attention to the design. In eighth grade, modifications
of outline and dimensions of any project are permitted where a fair degree of
merit is shown. Modifications of joints or fastenings are not to be made,
however, unless a pupil wishes to transfer a project from some other group into
the group in which the class is working.
In high school the pupil is expected to “work up” in his drawing class projects
original in so far as his ability will permit, subject to limitations mentioned
hereafter.
Eighth grade boys are expected to make at least one application of decorative
design to the pieces of woodwork made. The projects made by the high school
boys are, as a rule, not so well calculated to take decorative design. Their
efforts at decorative design will come later in connection with the metalwork of
the first year.
In high school the design is to be taught by special drawing teachers who
have informed themselves of the limitations of the shop methods when it comes
to applying these designs. It is for the shop instructor to specify the kind of joint
or joints that are to be used and the material, also the limitations as to
decoration. Present methods of organization in high schools hardly permit of the
teaching of shopwork and design and by the same instructor, which is the ideal
way providing, of course, that the instructor is expert in both. This is a
combination difficult to find. It is gratifying, however, to know that some schools
are insisting that their shop men become informed in design as well as
shopwork.
While these drawings are being worked up in the drafting room the pupil’s
shop periods are given over to the making of the exercise joints and mastering
the principles involved in their making. By the time these exercises are
completed, the working drawing will be completed ready for use in the shop.
The proper correlation of design and shopwork is not a problem beyond
solution, because of the direct relation of the two departments, providing there is
a strong administrative head able to secure proper esprit de corps. In the
grammar schools, however, the problem becomes less satisfactory of solution
by correlation.
The first objection lies in the fact that the regular grade teacher has both boys
and girls to teach and the problems must therefore be the same for the whole
room. The second objection lies in the fact that the problem in design has to
pass thru too many hands before it reaches the boy. If design is to be taught to
the best advantage, it must have the interest of the teacher and she must have
an intelligent understanding not only of the subject of design but of the particular
problem that is to be presented. The difficulties in the way are not
insurmountable where the drawing supervisor herself presents the problem to
the pupils. Even here, however, one frequently finds the drawing supervisor so
much more interested in the freehand drawing that her dislike for the design
makes her unfitted for such correlation work.
When, however, as is the case in cities, the drawing supervisor must reach
the pupils thru the regular teacher, correlation becomes in most every instance a
farce. The teaching of design is another imposition on an already overburdened
grade teacher. Very seldom does she understand the problem and it becomes a
distasteful subject to be got over in the easiest way possible. Department
teaching in the upper grammar grades would do much to aid in the correlation of
drawing and shop. Until this is made possible, we can hope for little in the way
of results from grammar school correlation, unless it be in a small system where
the supervisor teaches the children directly.
The whole subject of design as it relates to woodworking is a constant source
of discussion among manual training shop men. Many good teachers insist that
design has no place at all in a course in woodworking. Others admit that it ought
to have a place but feel that the results obtained do not justify the time spent
upon it. Still others approach the whole field of woodworking from the side of
design, tool processes and organized woodworking subject matter being mere
incidents to the problem in design.
Like every extreme position each of these points of view has good in it, but
there is sufficient error accompanying each to impair the validity of the
conclusions and to make the resulting applications unhappy as related to
ordinary public school conditions.
The whole subject of design as it relates to the manual training shop is one
that has demanded thought on the part of the author. It is one of those places
where teaching theory failed to bring efficiency either in the results obtained in
design or in the reaction upon the boy. He has been forced to the opinion, from
his own experience and from his observation of the efforts of others to teach
design to grammar school pupils, that the cause for dissatisfaction and
discouragement is due to our insistence upon one and only one method of
presentation—the inductive or synthetic.
In judging results we must consider the results obtained from every member
of a class and the good each boy has got out of his experience. This efficiency
test most effectively excludes the exhibition of a few “accidentals” as evidence
that our method is the correct one. There is no reason why design should seek
justification on any ground other than that offered by other subjects.
Inductive or synthetic teaching of design has its place; so also has the
deductive or analytic. Happily those educators who insist on the use of one
method or the other only are becoming few. In other subjects we are finding that
the teaching results which demand the respect and approval of educators of
safe and sane judgment are obtained by the use of both methods
interchangeably. There is no formal notice when one is to be used or the other—
whichever method fits the occasion is used without apology. This is right; to do
otherwise is to sacrifice the boy or girl for the sake of the method. We are all
agreed that the child is the more important consideration. In fact, some
psychologists tell us that induction and deduction are one and the same
process, the difference being merely a matter of emphasis. It is this difference in
placing the emphasis that we seek to discuss.
Our methods in the high school have made much of the inductive. This is
right. Pupils of high school and college age are ready for this method, tho our
high school pupils often would profit by having a little less of this with more of
the deductive.
However, when it comes to grammar school teaching, the maximum of use
has to be made of the deductive or analytic method. This is acknowledged in the
academic subjects. Woodworking when taught so as to meet the efficiency test
that is applied to academic teaching also makes use of this method mostly. Our
design, however, has always been taught by the inductive or synthetic method,
no one seeming to have the temerity to make use of any other. As a result we
find the views of design in the grammar school as stated above. Those who
advocate it urge the “accidentals” as sufficient justification. Those who reject it
base their argument on the fact that results based on a few accidentals will not
satisfy the same efficiency test that is applied to other subjects.
Experience has shown, at least to the author’s satisfaction, that the deductive
or analytic method when given maximum emphasis with beginners in design is
all that is needed to bring the results up to a standard equal to that of other
subjects. It is the rational method of presenting any subject to beginners.
The terms deductive and inductive have such wide application that it may be
well to specify more particularly just what we mean. A concrete illustration will
suffice to show the distinction we seek to make between what we choose to
designate the deductive or analytic and the inductive or synthetic methods.
Suppose we wish to have a class, with little or no information about the
subject, design a booklet to meet certain specified conditions. Three distinct
stages of progress manifest themselves in what we shall call the complete
method. First, the pupils must be given information bearing upon the problem.
Second, they must be given experience in handling problems of that type. Third,
they will utilize this information and experience in designing the booklet to meet
given conditions.
The first step will be the taking of a type form and analyzing it. Either the
instructor will demonstrate or, better, each pupil may be given a booklet of type
form and required to take it apart and put it together again. Any way to give the
pupil the information in a form that will cause it “to stick.” In woodwork, it would
be done by means of the traditional shop demonstration—a wise practice, since
psychology teaches us that sight percepts are among the strongest.
Second, the pupils must acquire experience. Let them make a booklet
according to definite specifications provided them by the instructor.
The process thus far is mainly deductive or analytic. So far there has been no
invention or design, but the pupils are now prepared for it. Using the information
and experience now available, let them design a booklet to meet certain
conditions. This latter part we would call the inductive or synthetic process.
We should have two aims in our teaching of design: (1) Appreciation, (2)
Development of the creative faculty. Since all must be able to appreciate good
line and good form when they get out into life while only a few will ever become
designers in a creative sense, it is essential, as it is also rational, that attention
should be paid first to appreciation. Past efforts show how hopeless is the
problem when we strive to give to the pupils appreciation of and feeling for line
and form by demanding original forms in the very beginning. The beginner’s
efforts at creation are abortive and the appreciation that he derives is nil. By our
insistence on this method we have given to our pupils the idea that design
means making something out of nothing. He is not far wrong if we demand of
him original designs before we have given him anything tangible upon which to
work. We say tangible as distinguished from academic principles or rules of
design. If nothing tangible is given the pupil he must get it outside of his school
experience. This explains the superabundance of “wienerwurst” forms, bouquets
tied with ribbons, circles, etc., etc.
It is possible to create unknown out of what is seemingly unknown. When we
stop to analyze the process, however, we find that we have made use of
information, appreciation, and feeling that are known. Sometimes we make
ourselves believe that our pupils are creating unknown out of unknown without
these requisites. Analysis will show that our continued suggestions to him,
drawn from our own fund of known are the causes, and not the pupil’s faculty.
This method of teaching is the kind we have been used to in design. It works
pretty well with small classes and individual instruction. Try it on large classes of
beginners and it is not possible to bring results that stand for class efficiency.
And why should this particular method be insisted upon exclusively with
beginners? Why should not design, like mechanical drawing and woodwork and
other subjects be developed upon a substructure consisting of information and
appreciation secured by allowing or even insisting that the boy handle good
design until he becomes saturated with a feeling for good line and good form?
Of course, if any pupil comes to a beginning class with this information and
feeling, due allowance should and can readily be made. It is highly probable that
there would be less inclination on the part of our pupils to insist that designers
are born not made were more use made of the deductive method. When the
boys no longer see their efforts result in crudities and are enabled to acquire the
necessary feeling and information as their work proceeds, then you find a happy
and interested class that as a whole takes design as a matter of course and not
as something intended only for the few.
Whatever the method of teaching design in the regular classroom, lack of time
demands the most direct treatment of shop design. A grammar school boy is not
inclined to listen very patiently to anything that smacks of the academic. (1) Give
the boy something definite with which to work and (2) keep him working, or
“playing,” as one has fittingly designated it, until he has made a conscientious
effort to “make it a part of himself,” that is, until he succeeds in changing the
form until it no longer resembles the original but still possesses the pleasing
appearance of the original.
If he succeeds in doing this, he is well on the way to creative effort. Not all
boys are of equal ability in other lines of endeavor, neither are they in this. By
this method of attack, however, even the stupidest—usually stupid only in the
matter of design—is not without compensation for his effort. He has learned
somewhat of the principles that govern good design by hearing them explained
and seeing them illustrated in a piece of good design. He will have developed
some feeling for line and form thru having played with good line and form. He
can at the very least fit the form given him to an outline made by himself after
suggestions of good line placed upon the board. To this extent, at least, you
have benefited him, whereas, by the usual method he—and there are many like
him—would have simply sat idle in discouragement—if he were not more
mischievously occupied.
If our old art schools were to be criticised because they made too much use of
the imitative method when they strove to give to their students information and
appreciation and feeling for form and line thru copying historic ornament, it
would seem, from results obtained, both tangible and in the effect upon the
pupils, that our modern schools are open to criticism when they seek to force
originality upon immature minds before they have given these minds any
information or feeling.
Of course grammar school boys are not interested in historic ornament, at
least not in America. This is the weakness of the imitative method and helped to
bring in the movement which now seems to have swung to the opposite extreme
—it lacks vitality for young pupils. Instead of giving the boy historic fragments,
give him a form that is vitally interesting to him because he sees its immediate
application in the thing that is to be made in wood. Let him play with this form
combining imitation and modification and creation just as far as he is able.
Make the problem concrete, stating the principles you have to state in a
language the boy can understand. There will not be time to bring out every
principle that might be involved in design. There must be time to bring out those
involved in the particular problem under discussion. Balance and symmetry, for
illustration, are pretty well understood by the boy in the simple form in which he
will have occasion to use them.
Take as an illustration the bookrack, Fig. 7. To present such a problem we
would place upon the blackboard the blank forms as shown, also the decorative
form as shown.
The lesson immediately divides itself into two parts for consideration: (1) The
Construction, (2) The Decoration. Under the subject of Construction our normal
school notes would suggest the following points to be brought up: Use,
Construction, Decoration; Requirements of Utility; Limitations of Materials and
Processes; Proportions of Parts and Details; Harmony of Parts and Details;
Points of Force; Construction as Decoration. (According to Payne.) Under
Decoration: Supporting Outline; Center of Interest; Symmetry; Repetition;
Radiation; Rhythm; Contrast; Proportion in Curves; Proportion in Spaces; Unity;
Subordinate Centers of Interest; Balance.
FIG. 7. TEACHING DESIGN IN THE PROBLEM OF THE BOOK-
RACK.
CONSTRUCTION
DECORATION
OUTLINES
MOTIVE GIVEN ADAPTATION
APPLICATION—
BY A PUPIL
Taking these in their natural order, but without making much ado about the
“framework,” the shop man who has made some study of the principles involved
can call the boys’ attention to the most important points:
(1) The construction. Since the shopwork is to be carried on by class
instruction and not individually, it will be necessary to limit the joint or joints used
to those specified for the Group in which the project is to be worked out. Joints
of previous Groups may be used also. The book rack will be made in Group VII.
Some form of the groove joint is to be used, none other.
Here we call attention to the difference between the designer and the shop
man in their handling of the problem. The discussion of construction gives the
designer an opportunity to display the possibilities of his subject. He enumerates
all the joints that may be used with propriety in making such a piece as the
bookrack, and the pupils are encouraged to make use of as many varieties as
possible. He is totally oblivious of the fact that, while this is good teaching in
design, it is making the applications impossible except with individual instruction
—a method of instruction that may be used in small school systems but not in
cities.
(2) The manner of placing the members and the use to which the rack is to be
put will together determine the proportions of the members.
(3) For decoration, we might depend entirely upon the good form of the
outline and the stain and grain of the wood. With this particular piece, however,
we shall make use of a decorative form which will be outlined or incised and
colored with a dye.
(4) Since the design is to be made in wood and wood splits easily along the
grain, we must be careful in making an outline not to get sharp points. Also, in
making a decorative design we must avoid thin parts that will bring incised lines
close together. Also, we must take into account in planning the members the
facts of shrinkage or swelling and the strength of the wood. The grain on the
vertical members must extend vertically and that of the horizontal member must
extend from vertical member to vertical member. This to be illustrated by
referring to some similar construction.
(5) In striving for pleasing outlines, or decorative forms either, strive to avoid a
sameness made by using many lines or forms of the same size. “Large, medium
and small” is a key that unlocks many a puzzle as to what causes unpleasant
feelings in both outline and decoration. Long sweeping curves with short snappy
ones, rather than a series made with a compass. Make a special point of the
fact, which almost every boy overlooks, that the simple forms of outline are
invariably the more pleasing. To the beginner design means making something
unlike anything that was ever seen on the earth below or heaven above—hence
the freakish, fussy forms that are usually offered. Try telling the class you are
going to place an excellent form on the board then draw a well proportioned
oblong and watch the expressions on their faces. Yet a well proportioned oblong
with appropriate decorative form is one of the most pleasing of forms. There will
be no need to urge them to make “unique” forms. Their inexperience and their
zeal will produce a sufficient number. Rather urge, or insist that they postpone
search for “unique” form until they have more information.
Illustrate with blackboard sketches as you go along each of these points.
Keep the boys “playing” with outline forms until you have assured yourself they
have done their best. With them, pick out three of the best and place these in
permanent form for keeping—put them on another sheet of paper. Next, start
them on the decoration. The development of a decorative form will come much
harder than the outline. Here again the beginner will want to exhibit “unique”
forms—unique only in that they are founded upon his ignorance. Unless the boy
is not a beginner, it will be necessary in about twenty-four out of every twenty-
five cases to insist that he start with the form you have placed upon the board
for his use. If you were dealing with a few pupils, you might take his “original”
form and step by step get him to work it into a good form. With large classes this
is not possible, nor is it necessary. Simply insist that he place the form given him
in his outline and in so doing he will acquire enough feeling for line and form to
enable him to proceed of his own accord.
(6) Have the boy put on a supporting outline, that is, tell him to draw a line
around his outline and parallel to it. Show the class on the blackboard how this
is to be done.
(7) Put in the main mass and break it up explaining as you do so that you are
seeking to get large, medium and small forms-proportion of parts. Call attention
to the efforts made to keep the lines in harmony.
(8) Call attention to the center of interest you have created. It is unfortunate
that lack of time forbids the boy’s placing colors on these designs. Very
frequently a touch of color is used to create a center of interest, the form for this
in black and white not giving the proper significance at all. A design which in
outline seems to be fussy because of too many parts will, by a proper selection
and placing of colors, be made most pleasing. On the other hand, a design in
outline that seems agreeable may, when in color, not be agreeable because the
colors make certain parts stand out too prominently. A study of the color plate in
Projects in Beginning Woodwork and Mechanical Drawing will make this clear.
(9) If the form proposed happens to illustrate repetition, radiation, symmetry,
or if some boy develops a form that does, take time to say a word about them.
While you will not have time to “teach design” in the few lessons, a word here
and there may serve to awaken further interest on the part of some boy.
After all is said, we recognize that the time is short, that not much can be
done. On the other hand, what little can be done is worth doing and doing well;
its possible significance can not be overestimated.
6. Shop Excursions. In the grammar schools, and more especially in the
high schools, plans should be made for several excursions to near by shops in
which the pupils may get an insight into the workings of related industries. The
saw-mill, lumber yards, planing mills, furniture factories, architectural or drafting-
rooms and, in fact, anything relating to the industrial employment of men and
machinery may be visited.
That the trip may be one of profit the instructor should see to it that the pupils
are prepared for the trip by previous talks on what is to be seen and by after
talks on the meaning of what they saw.
In every case it will be necessary, or at least advisable, to have a time
arranged with the superintendent of the factory to be visited. Pupils should be
given to understand that they are being privileged and must act the part of
gentlemen, refraining from asking needless questions of the workmen or
handling the equipment. In many factories no talking to the men at all is desired.
The questions of young pupils are often impertinent and embarassing without
their intending them so to be. The better plan is, as has just been suggested, to
have the pupils prepared by preliminary talks then take them thru the shop with
eyes and ears only open, clinching the lessons of observation afterward.
Pupils should keep together in solid lines and, should any accident occur, the
instructor should see that any loss to the factory owner or workmen is “made
good.” Usually the class will voluntarily make recompense. It is safer and less
likely to cause embarassment if it is understood beforehand that all members of
the class who go will be expected to help repay the instructor for any money so
expended.
One might think the company well able to stand such loss. It is, but it is not
always the company’s loss. Even if it were, their courtesy ought not to be
abused. We have in mind a mold for an intricate piece of casting representing a
day’s labor for two men ruined by a student’s accidentally brushing against it
with his overcoat. As the men were on “piece work” it meant no loss to the
company, except delay in getting out the finished article. It did mean a loss to
the two men, who could ill afford it. The instructor quietly settled for the damage
or loss and the pupils reimbursed him upon reaching school. This probably
prevented the factory from excluding succeeding classes as undesirables. In
woodworking shops there is little chance for such accidents. Nevertheless
workmen there do not wish their tools or work handled. Each class should bear
constantly in mind, while on the shop excursions, that it is making succeeding
classes welcome or unwelcome in that shop.
7. Stock Bills. Every piece of woodwork made by a pupil consisting of
more than one member should have in addition to the working drawing a
carefully made stock bill. The reason is two-fold: It not only prevents the pupil’s
cutting out stock wrongly thru misreading the drawing, but it saves time for the
pupil. It is a practice that he will have to master later in life if he follows any of
the mechanical trades and is just as essential a part of his shopwork as is the
drawing or woodwork. Where the drawings are made by referring to plates,
experience has shown that many a boy will be able to make a good drawing
without fully interpreting its meaning. The making of the stock bill will show him
his weakness, also it will show the instructor. No boy can make out his stock bill
without being able to read his drawing. After the drawing has been made and
then its stock bill, the boy will have become so conversant with the plans of the
thing he is to make that few mistakes are made in working the wood, that is,
mistakes due to ignorance of the drawing.
STOCK BILL (Form)
Name ________________________ Article ________________________
Kind of
Grade ________________________ ________________________
Wood
Finished Sizes Cutting Sizes

Pieces Thickness Width Length Pieces Thickness Width Length
1 3⁄8 3 51⁄2 1 3⁄8 13⁄4 6
1 1⁄2 11⁄2 41⁄2 1 1⁄2 13⁄4 5
2 1⁄2 11⁄2 9 2 1⁄2 31⁄4 91⁄2
1 1⁄2 51⁄2 12 1 1⁄2 53⁄4 121⁄2
INSTRUCTIONS
All articles in seventh grade will be made of White Pine or Yellow Poplar; those in
eighth grade of Chestnut.
Stock bills are not needed for articles composed of one piece of material only.
Finished sizes are the sizes to which the pieces are to be planed. Your drawing will
tell you these sizes.
Pieces of irregular shape are to be figured at their widest and longest dimensions.
Cutting sizes are obtained from the finished sizes by adding 1⁄4″ to the width and
1⁄2″ to the length. Cutting sizes are the sizes to which you work in sawing out the stock
preparatory to planing it.
All stock will be mill-planed on two surfaces to the correct thickness except that for
the ring toss, spool holder, game-board, and laundry register. Thickness of mill-planed
stock will be the same whether for finished sizes or cutting sizes. On rough stock, or
stock that has not been mill-planed, if the finished size is 3⁄4″ thick the cutting size will
be 1″ thick.
Sometimes it is possible to save material by combining two irregular pieces. The
finished stock sizes will indicate the number of pieces while the cutting size will indicate
the size of the single piece from which they are to be cut.
Remember that length always means “along the grain of the wood,” and that a
piece may be wider than long. Under the word “Pieces” put the number of pieces that
are of the same size.
In the elementary schools the form of stock bill used should be as simple and
explicit as is possible. The appended form is one that has proven satisfactory.
That it may be in convenient form for student use, it has been included with

Full Download PDF of Advanced Differential Equations 1st Edition Youssef Raffoul - Ebook PDF All Chapter

Uploaded by

Copyright:

Available Formats

Full Download PDF of Advanced Differential Equations 1st Edition Youssef Raffoul - Ebook PDF All Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Full Download PDF of Advanced Differential Equations 1st Edition Youssef Raffoul - Ebook PDF All Chapter

Uploaded by

Copyright:

Available Formats

Advanced Differential Equations 1st

Edition Youssef Raffoul - eBook PDF

(eBook PDF) Differential Equations 4th Edition

Linear Algebra and Partial Differential Equations 1st

(eBook PDF) Elementary Differential Equations 10th

Schaum's Outline of Differential Equations - eBook PDF

(eBook PDF) Fundamentals of Differential Equations 9th

Fundamentals of Differential Equations 8th Edition

(eBook PDF) Differential Equations and Linear Algebra

(eBook PDF) Differential Equations and Linear Algebra

Copyright Elsevier 2022

Library of Congress Cataloging-in-Publication Data

British Library Cataloguing-in-Publication Data

For information on all Academic Press publications

Publisher: Katey Birtcher

Copyright Elsevier 2022

Copyright Elsevier 2022

Copyright Elsevier 2022

5.4 Autonomous systems in the plane . . . . . . . . . . . . . . . . . . . . . 139

Copyright Elsevier 2022

9.2 Applications to delay differential equations . . . . . . . . . . . . . . 335

Copyright Elsevier 2022

Differential equations are widely used by mathematicians, physicists, engineers, biol-

Copyright Elsevier 2022

Copyright Elsevier 2022

Copyright Elsevier 2022

Preliminaries and Banach

exists as a finite number. In this case, we adopt the notation

where x  (t ∗ ) is the instantaneous rate of change of the function x at t ∗ . If t ∗ is one of

Copyright Elsevier 2022

for some function f that satisfies certain continuity conditions.

z (t) = x(t) for all t ∈ I.

The set of all antiderivatives of x is denoted by

Copyright Elsevier 2022

x  (t) = h(t)g(x), x(t0 ) = x0 , t ≥ t0 ,

where h, g : R → R are continuous.

Using the transformation u = x(s) with x(t0 ) = x0 , we arrive at

Copyright Elsevier 2022

x  (t) = x 3 (t), x(t0 ) = x0 > 0, t ≥ t0

has the solution

Copyright Elsevier 2022

Copyright Elsevier 2022

Another interesting application is the projectile problem that we analyze in the

1.2 Escape velocity

Copyright Elsevier 2022

Copyright Elsevier 2022

Definition 1.2.1. Let D be an open subset of R2 , and let f : D → R be a continuous

x  = f (t, x), x(t0 ) = x0 , (1.2.2)

on an interval I if t0 ∈ I , x : I → R is differentiable, (t, x(t)) ∈ D for t ∈ I , x  (t) =

|x(t) − y(t)| < ε whenever |x0 − y0 | < δ.

Consider the linear differential equation

x  (t) = 1, x(t0 ) = x0 . (1.2.3)

It is easy to check that x(t) = x0 + (t − t0 ) is the solution of (1.2.3). If y(t) is another

1.3 Applications to epidemics

Copyright Elsevier 2022

I  (t) = λI (1 − I ), I (0) = I0 . (1.3.2)

Copyright Elsevier 2022

1.4 Metrics and Banach spaces

Copyright Elsevier 2022

where x (t ∗ ) is the instantaneous rate of change of the function x at t ∗ . If t ∗ is one of

z (t) = x(t) for all t ∈ I.

x (t) = h(t)g(x), x(t0 ) = x0 , t ≥ t0 ,

x (t) = x 3 (t), x(t0 ) = x0 > 0, t ≥ t0

x = f (t, x), x(t0 ) = x0 , (1.2.2)

on an interval I if t0 ∈ I , x : I → R is differentiable, (t, x(t)) ∈ D for t ∈ I , x (t) =

x (t) = 1, x(t0 ) = x0 . (1.2.3)

I (t) = λI (1 − I ), I (0) = I0 . (1.3.2)

L = {f ∈ C([a, b], Rn ) : f ≤ M; |f (u) − f (v)| ≤ K|u − v|}.

L = {f ∈ S : f ≤ M on (−∞, 0], |f (u) − f (v)| ≤ K|u − v|}

L = {f ∈ S : f ≤ M on (−∞, 0], |f (u) − f (v)| ≤ K|u − v|}

x (t) = 2x(t) + t, x(0) = 3,

For x = u, we have the first-order differential equation tu + 2u + t = 1. Using the

x + (x )2 x + x x = 0, x(0) = 0, x (0) = −1.

Using u = −1 when x = 0, we have c = 0. Now using the transformation x = u, it

x (t) = x α (t), x(0) = 0, t ∈ R

x (t) = x 2 (t), x(t0 ) = x0 = 0, t ∈ R.