Asymptotic Perturbation Methods For Nonlinear Differential Equations in Physics Attilio Maccari Full Chapter

Asymptotic Perturbation Methods: For
Nonlinear Differential Equations in

Physics Attilio Maccari
Visit to download the full and correct content document:
https://ebookmass.com/product/asymptotic-perturbation-methods-for-nonlinear-differe
ntial-equations-in-physics-attilio-maccari/
Asymptotic Perturbation Methods
Asymptotic Perturbation Methods
For Nonlinear Differential Equations in Physics
Attilio Maccari
Author All books published by WILEY-VCH are carefully
produced. Nevertheless, authors, editors, and
Prof. Attilio Maccari publisher do not warrant the information
Via Alfredo Casella 3 contained in these books, including this book,
00013 Rome to be free of errors. Readers are advised to keep
Italy in mind that statements, data, illustrations,
procedural details or other items may
Cover Image: © Callista Images/Getty inadvertently be inaccurate.
Images
Library of Congress Card No.: applied for
British Library Cataloguing-in-Publication Data

A catalogue record for this book is available
from the British Library.
Bibliographic information published by

the Deutsche Nationalbibliothek
The Deutsche Nationalbibliothek lists
this publication in the Deutsche
Nationalbibliografie; detailed bibliographic
data are available on the Internet at
<http://dnb.d-nb.de>.
© 2023 WILEY-VCH GmbH, Boschstraße 12,

69469 Weinheim, Germany
All rights reserved (including those of

translation into other languages). No part of
this book may be reproduced in any form – by
photoprinting, microfilm, or any other
means – nor transmitted or translated into a
machine language without written permission
from the publishers. Registered names,
trademarks, etc. used in this book, even when
not specifically marked as such, are not to be
considered unprotected by law.
Print ISBN: 978-3-527-41421-5

ePDF ISBN: 978-3-527-84172-1
ePub ISBN: 978-3-527-84173-8
oBook ISBN: 978-3-527-84174-5
Typesetting Straive, Chennai, India

v
Contents
About the Author xi

Foreword xiii
Introduction xv
1 The Asymptotic Perturbation Method for Nonlinear

Oscillators 1
1.1 Introduction 1
1.2 Nonlinear Dynamical Systems 3
1.3 The Approximate Solution 5
1.4 Comparison with the Results of the Numerical Integration 10
1.5 External Excitation in Resonance with the Oscillator 11
1.6 Conclusion 16
2 The Asymptotic Perturbation Method for Remarkable

Nonlinear Systems 19
2.1 Introduction 19
2.2 Periodic Solutions and Their Stability 21
2.3 Global Analysis of the Model System 27
2.4 Infinite-period Symmetric Homoclinic Bifurcation 35
2.5 A Few Considerations 41
2.6 A Peculiar Quasiperiodic Attractor 42
2.7 Building an Approximate Solution 44
2.8 Results from Numerical Simulation 46
2.9 Conclusion 52
3 The Asymptotic Perturbation Method for Vibration Control

with Time-delay State Feedback 53
3.1 Introduction 53
3.2 Time-delay State Feedback 53
3.3 The Perturbation Method 56
3.4 Stability Analysis and Parametric Resonance Control 59
3.4.1 The Frequency–Response Curve Is 62
vi Contents
3.5 Suppression of the Two-period Quasiperiodic Motion 63

3.6 Vibration Control for Other Nonlinear Systems 68
4 The Asymptotic Perturbation Method for Vibration Control by

Nonlocal Dynamics 69
4.1 Introduction 69
4.2 Vibration Control for the van der Pol Equation 72
4.3 Stability Analysis and Parametric Resonance Control 74
4.4 Suppression of the Two-period Quasiperiodic Motion 79
4.5 Conclusion 82

Continuous Systems 83
5.1 Introduction 83
5.2 The Approximate Solution for the Primary Resonance of the nth
Mode 86
5.3 The Approximate Solution for the Subharmonic Resonance of Order
One-half of the nth Mode 91
5.4 Conclusion 93
6 The Asymptotic Perturbation Method for Dispersive Nonlinear

Partial Differential Equations 95
6.1 Introduction 95
6.2 Model Nonlinear PDES Obtained from the Kadomtsev–Petviashvili
Equation 97
6.3 The Lax Pair for the Model Nonlinear PDE 98
6.4 A Few Remarks 100
6.5 A Generalized Hirota Equation in 2 + 1 Dimensions 100
6.6 Model Nonlinear PDEs Obtained from the KP Equation 101
6.7 The Lax Pair for the Hirota–Maccari Equation 103
6.8 Conclusion 105
7 The Asymptotic Perturbation Method for Physics

Problems 107
7.1 Introduction 107
7.2 Derivation of the Model System 108
7.3 Integrability of the Model System of Equations 111
7.4 Exact Solutions for the C-integrable Model Equation 112
7.4.1 Nonlinear Wave 112
7.4.2 Solitons 112
7.4.3 Dromions 113
7.4.4 Lumps 116
7.4.5 Ring Solitons 116
7.4.6 Instantons 117
Contents vii
7.4.7 Moving Breather-Like Structures 117

7.5 Conclusion 120
8 The Asymptotic Perturbation Model for Elementary Particle

Physics 121
8.3 Integrability of the Model System of Equations 124
8.4 Exact Solutions for the C-integrable Model Equation 125
8.4.2 Solitons 126
8.4.3 Dromions 126
8.4.4 Lumps 127
8.4.7 Moving Breather-like Structures 129
8.5 A Few Considerations 130
8.6 Hidden Symmetry Models 130
8.8 Coherent Solutions 138
8.8.2 Solitons 138
8.8.3 Dromions 139
8.8.4 Lumps 139
8.8.7 Moving Breather-like Structures 142
8.9 Chaotic and Fractal Solutions 143
8.9.1 Chaotic–Chaotic and Chaotic–Periodic Patterns 143
8.9.2 Chaotic Line Soliton Solutions 145
8.9.3 Chaotic Dromion and Lump Patterns 145
8.9.4 Nonlocal Fractal Solutions 147
8.9.5 Fractal Dromion and Lump Solutions 147
8.9.6 Stochastic Fractal Dromion and Lump Excitations 148
8.10 Conclusion 150
9 The Asymptotic Perturbation Method for Rogue Waves 151

9.2 The Mathematical Framework 153
9.3 The Maccari System 154
9.4 Rogue Wave Physical Explanation According to the Maccari System and
Blowing Solutions 156
9.5 Conclusion 158
viii Contents
10 The Asymptotic Perturbation Method for Fractal and Chaotic

Solutions 159
10.2 A New Integrable System from the Dispersive Long-wave Equation 161
10.3 Nonlinear Coherent Solutions 165
10.3.2 Solitons 165
10.3.3 Dromions 166
10.3.4 Lumps 166
10.3.7 Moving Breather-Like Structures 168
10.4 Chaotic and Fractal Solutions 168
10.4.1 Chaotic–Chaotic and Chaotic–Periodic Patterns 168
10.4.2 Chaotic Line Soliton Solutions 168
10.4.3 Chaotic Dromion and Lump Patterns 169
10.4.4 Nonlocal Fractal Solutions 169
10.4.5 Fractal Dromion and Lump Solutions 169
10.4.6 Stochastic Fractal Excitations 170
10.4.7 Stochastic Fractal Dromion and Lump Excitations 170
10.5 Conclusion 171

Relativistic and Quantum Physics 173
11.2 The NLS Equation for a1 > 0 174
11.3 The NLS Equation for a1 < 0 176
11.4 A Possible Extension 178
11.5 The Nonrelativistic Case 180
11.6 The Relativistic Case 183
11.7 Conclusion 185
12 Cosmology 187
12.2 A New Field Equation 188
12.3 Exact Solution in the Robertson–Walker Metrics 191
12.4 Entropy Production 195
12.5 Conclusion 197
13 Conﬁnement and Asymptotic Freedom in a Purely Geometric

Framework 199
13.2 The Uncertainty Principle 201
13.3 Confinement and Asymptotic Freedom for the Strong Interaction 203
Contents ix
13.4 The Motion of a Light Ray Into a Hadron 207

13.5 Conclusion 208
14 The Asymptotic Perturbation Method for a Reverse

Inﬁnite-Period Bifurcation in the Nonlinear Schrodinger
Equation 209
14.2 Building an Approximate Solution 210
14.3 A Reverse Infinite-Period Bifurcation 212
14.4 Conclusion 215
Conclusion 217
References 219
Index 235
xi
About the Author
Attilio Maccari is a free-lance researcher in nonlinear physics. He received his PhD

from La Sapienza Rome University in 1990. He has published about one hundred
papers, mainly on coupled oscillators and nonlinear partial differential equations in
physics and engineering.
Among his most important accomplishments, we recall the Maccari system for
rogue waves as well as the Hirota–Maccari equation for nonlinear systems. Both
equations are integrable and with remarkable nonlinear waves. His recent work
has been devoted to vibration control, and he found two new methods in order to
perform this very important task, time-delay state feedback control and nonlocal
feedback control.
xiii
Foreword
This textbook is devoted to nonlinear physics.

The asymptotic perturbation method is used as a mathematical tool and is
explained in some detail, and the theory is developed systematically, starting with
nonlinear oscillators, limit cycles and their bifurcations, followed by iterated non-
linear maps, continuous systems, nonlinear partial differential equations (NPDEs),
and culminating with infinite-period bifurcation in the nonlinear Schrodinger
equation and fractal and chaotic solutions in NPDEs.
A remarkable feature of the book is the emphasis on applications. There are sev-
eral examples, and the scientific background is explained at an elementary level and
closely integrated with the mathematical theory.
This book is ideal for an introductory course at the senior or first-year graduate
level. It is also advisable for a scientist who has not a deep knowledge about nonlin-
ear physics but now wants to begin a complete study.
The prerequisites are multivariable calculus and introductory physics.
xv
Introduction
Nonlinear systems are paramount in engineering and science. Many perturbation

methods can be used to study these systems in order to predict remarkable bifurca-
tions (a qualitative change in their behavior). In this book, we will use the asymp-
totic perturbation (AP) method both for nonlinear ordinary differential equations
(NODEs) and nonlinear partial differential equations (NPDEs).
In Chapters 1–4, we will study NODEs and will derive a suitable model system
to find the most important nonlinear system characteristics. The main finding is
that a nonlinear model system of equations describes their behavior. In particular, in
Chapter 2, we will describe an infinite-period bifurcation for a parametrically excited
Liènard system and find a peculiar attractor for a weakly nonlinear oscillator with a
two-period quasiperiodic forcing.
In Chapter 3, we consider vibration control with time-delay state feedback
and perform a successful control strategy. In Chapter 4, we illustrate another
vibration-control method based on nonlocal dynamics. Numerical simulation
confirms our method’s validity.
In Chapter 5, we enlarge our perspective and study nonlinear continuous systems,
in particular the vibrations of an Euler–Bernoulli beam resting on a nonlinear elas-
tic foundation and with an external periodic excitation. Frequency–response and
external force–response curves can be easily found and compared with numerical
simulation.
In Chapter 6, the AP method is used to study NPDEs, and we are able to
find two new and integrable nonlinear equations, the Maccari system and the
Hirota–Maccari equation.
At the same time, the AP method can be used in order to find approximate solu-
tions to relevant physics problems. In Chapter 7, we study the Zakharov–Kusnetsov
equation and show the existence of interacting localized solutions because the
ZK equation can be described through a C-integrable (solvable via an appropriate
change of variables) system of nonlinear evolution equations. Dromions, lumps,
ring solitons, and breathers exist for this remarkable nonlinear equation.
In Chapter 8, we study the connection between the AP method and elementary
particle physics.
xvi Introduction
In Chapter 9, we try to explain the rogue waves appearance in nonlinear

systems.
In Chapter 10, we arrive at one of the most important findings in this book,
fractal and chaotic solutions are possible for nonlinear systems and perhaps at a
very fundamental level we must let the particle concept (i.e. a coherent solution)
down because we can state that in general solutions have fractal and chaotic
properties.
In Chapter 11, we use the AP method in order to arrive at nonlinear quantum
mechanics and achieve the Einstein–de Brogle soliton-particle concept by study-
ing the weakly nonlinear Klein–Gordon equation for a particle confined in a box.
In Chapter 12, we illustrate how to modify the Einstein equation so as to explain
the accelerating and irreversible evolution of the universe. According to Prigogine’s
ideas, the entropy increase is connected with matter production.
In Chapter 13, this new theory is used to find how confinement and asymptotic
freedom can be explained in a framework where particles are like small black holes.
Finally, Chapter 14 is devoted to a reverse infinite-period bifurcation for the nonlin-
ear Schrodinger equation in 2 + 1 dimensions.
Many teaching years allowed me writing this book, and I would like to thank
my students at Foligno in Perugia University, Italy, for their helpful and valuable
suggestions.
1
The Asymptotic Perturbation Method for Nonlinear

Oscillators
1.1 Introduction
Oscillations are a fundamental topic in physics. When a system is near its equilib-
rium point, it begins to oscillate, but if the displacement increases, then the nonlin-
ear terms are not negligible. The starting point is the differential equation for the
harmonic oscillator
d2 X
+ 𝜔2 X(t) = 0 (1.1)
dt
where X(t) is the displacement and ω the circular frequency. The most general solu-
tion is
X(t) = 2𝜌 cos(−𝜔t + 𝜃) (1.2)
where 𝜌 and 𝜃 are fixed by the initial conditions (the Cauchy problem)
if X(0) = X 0 for the displacement

̇
and X(0) = Ẋ 0 for the initial velocity
then we easily get

√(
√
√ ( ̇ )2 )
√ X0
2𝜌 = (X0 )2 + (1.3)
𝜔
and
( ̇ )
X0
tan𝜃 = (1.4)
𝜔X0
Now, we can consider a weakly nonlinear part in the differential Eq. (1.1) or, on the
contrary, a strongly nonlinear part but with small solutions. The first consequence is
that the amplitude and the phase are slowly varying with time, so we can introduce
another slow time
𝜏 = 𝜀q t (1.5)
where 𝜀 is a bookkeeping device and q is a rational number that will be chosen after-
wards. If we want to study the asymptotic solution behavior (t → ∞) and 𝜀 → 0, then
Asymptotic Perturbation Methods: For Nonlinear Differential Equations in Physics,
First Edition. Attilio Maccari.
© 2023 WILEY-VCH GmbH. Published 2023 by WILEY-VCH GmbH.
2 1 The Asymptotic Perturbation Method for Nonlinear Oscillators
𝜏 must assume finite values. So, we assume that an approximate solution is given by
X(t) = 2𝜌(𝜏) cos(−𝜔t + 𝜃(𝜏)) = (𝜌(𝜏) exp(−i𝜔t + i𝜃) + c.c.) (1.6)
or better
X(t) = 𝜀(1+r) 𝛹0 + (𝜀𝛹1 exp(−i𝜔t) + 𝜀2 𝛹2 exp(−2𝜔t) + 𝜀3 𝛹3 exp(−3i𝜔t) + c.c. + h.o.t.)

(1.7)
where c. c. stands for complex conjugate and h. o. t. for higher order terms.
Following this path, we are mixing the most important features of two well-known
perturbation methods, the harmonic balance and the multiple scale methods (for
more details about these two perturbation methods, see Refs. [202, 203, 249]).
If we consider a weakly nonlinear differential equation
d2 X
+ 𝜔2 X(t) = NL (1.8)
dt
where NL stands for the nonlinear part, for instance,
aX(t)2 + bX(t)3 (1.9)
we can insert the solution (1.7) in the nonlinear Eq. (1.8) and with some algebra
manipulation, we get for n = 0
𝜔2 𝜀(1+r) 𝛹0 = 2a𝜀2 |𝛹 |2 (1.10)
then r = 1, for n = 2
−3𝜔2 𝜀2 𝛹2 = a𝜀2 𝛹 2 (1.11)
and for n = 1
−2i𝜔𝜀q 𝜓𝜏 = 2a(𝜀82 + r)𝛹0 𝛹 + 𝜀2 𝛹2 (c.c.𝛹 ) + 3b𝜀2 |𝛹 |2 𝛹 (1.12)
then, q = 2 for the proper nonlinear term balance and with some algebra
manipulation
d𝛹 iA
= |𝛹 |2 𝛹 (1.13)
d𝜏 2𝜔
where
10a2
A= +b (1.14)
3𝜔2
d𝜌 d𝜃 A 2
=0 = 𝜌 (1.15)
d𝜏 d𝜏 2𝜔
We observe that the variable change (1.5) implies that
d d
→ −in𝜔 + 𝜀q (1.16)
dt d𝜏
when the temporal differential operator acts on the function
𝛹n (𝜏) exp(−in𝜔t) (1.17)
From Eq. (1.10), we can see that the approximate solution is always periodic, the
amplitude is constant, but the period changes and becomes
1.2 Nonlinear Dynamical Systems 3
2𝜋 A 2
T= where 𝛺 = 𝜔 − 𝜌 (1.18)
𝛺 2𝜔
However, if
( )
10a2
b=− (1.19)
3𝜔2
the period does not change and is equal to the linear case period.
In this chapter, we want to extend this method and study a generalized Van der
Pol–Duffing oscillator in resonance with a periodic excitation
̈ + X(t) + f2 X 2 (t) + f3 X 3 (t) = g0 X(t)

X(t) ̇ ̇
+ g1 X(t)X(t) ̇
+ g2 X 2 (t)X(t) + 2F cos(𝛺t)
(1.20)
We use the asymptotic perturbation (AP) method based on Fourier expansion and
time rescaling (see above) and demonstrate through a second-order perturbation
analysis the existence of one or two limit cycles. Moreover, we identify a sufficient
condition to obtain a doubly periodic motion when a second low frequency appears,
in addition to the forcing frequency. The comparison with the solution obtained by
the numerical integration confirms the validity of our analysis.
1.2 Nonlinear Dynamical Systems

The study of nonlinear dynamical systems has interested many researchers, and var-
ious methods have been used. Historically, the AP method was first applied in order
to study the most important characteristics of a nonlocal oscillator [112, 113, 118].
We now devote our attention to the following type of nonlinear equation
̈ + f (X(t)) = g (X(t), X(t))
X(t) ̇ (1.21)
where the dot denotes differentiation with respect to the time and the functions f (x)
and g(x, y) are supposed to be analytic.
The limit cycles of the modified Van der Pol equation
X(t) ̇
̈ + X(t) + X 3 (t) = 𝜀(1 − X 2 (t))X(t) (1.22)
have been studied in Ref. [23] by means of a time transformation method.
Phase portraits and dynamical properties of the equation
X(t) ̇
̈ + (𝛼 + 𝛽X2 (t))X(t) + 𝛾X(t) + 𝛿X3 (t) = 0 (1.23)
have been investigated with the methods of differentiable dynamics [74] and the
equation
̈ + X(t) = 𝜀f(X(t), X(t))
X(t) ̇ (1.24)
with the method of averaging, the KBM method, the method of multiple scales, and
the Poincaré–Lindstedt method [202, 203].
Note that Eqs. (1.22)–(1.24) belong to the general class (1.21) and are characterized
by the fact that f (x) is an odd function of x.
We restrict our study to the following particular case of Eq. (1.21)

̈ + X(t) + f2 X 2 (t) + f3 X 3 (t) = g0 X(t)
X(t) ̇ ̇
+ g1 X(t)X(t) ̇
+ g2 X 2 (t)X(t) (1.25)
Eq. (1.5) can be considered a generalized Van der Pol–Duffing equation because it
includes as particular cases the Van der Pol oscillator (f 2 , f 3 , g1 = 0 and g0 = − g2 ≠ 0)
and the Duffing equation (f 2 = g1 = g2 = 0 and g0 = f 3 ≠ 0). Many authors have stud-
ied the problem of approximating the limit cycle of the Van der Pol equation. Stokes
[249] used the nonlinear Galerkin method and developed a series representation;
Deprit and Schmidt [47] utilized the Poincaré–Lindstedt method to find the ampli-
tude and frequency of the limit cycle; and Garcia-Magallo and Bejarano [57] consid-
ered a generalized Van der Pol equation by means of the harmonic balance method.
The steady-state behavior of the Van der Pol oscillator has also been studied by inte-
gral manifold methods and symbolic manipulation packages by Gilsinn [59, 61].
Mehri and Ghorashi [195] considered the periodically forced Duffing equation in
order to establish sufficient conditions to have a periodic solution, and Qaisi [233]
studied a similar problem using an analytical approach based on the power series
method. In a series of papers [69–71], Hassan used the higher order method of mul-
tiple scales with reconstitution and the harmonic balance method to determine the
periodic state response of the Duffing oscillator.
In our treatment of Eq. (1.25), no conditions are imposed on the coefficients f 2 , f 3 ,
g1 , and g2 , which can be of order 1. Only the dissipative coefficient g0 is supposed to
be of order e2 . Eq. (1.25) transforms into
̈ + X(t) + f2 X 2 (t) + f3 X 3 (t) = 𝜀2 g0 X(t)
X(t) ̇ ̇
+ g1 X(t)X(t) ̇
+ g2 X 2 (t)X(t) (1.26)
In the second section, we calculate the approximate solution good to the order
of 𝜀4 and construct accurate expressions for the limit cycle of Eq. (1.26). Moreover,
we demonstrate that, in the first approximation, the behavior of the solution can be
described by means of a model system of differential equations, which represents
the characteristics of Eq. (1.26) by means of a reduced set of parameters.
Usually, perturbation analysis is carried out only to the first order because, in many
cases, a second order-calculation does not change the qualitative behavior of the
solution. However, in Section 1.2, we demonstrate that if the parameters are appro-
priately chosen, we can find two limit cycles and can calculate their positions only
by a second-order perturbation analysis.
In Section 1.3, a comparison with the results of the numerical integration permits
discussion of the validity of the AP method.
In Section 1.4, we treat an extension of Eq. (1.26) that is a nonlinear oscillator
forced by a small periodic excitation, of order e2 , in resonance with the natural fre-
quency of the oscillator
X(t) ̇
̈ + X(t) + f2 X 2 (t) + f3 X 3 (t) = 𝜀2 g0 X(t) ̇
+ g1 X(t)X(t) ̇
+ g2 X 2 (t)X(t) + 2𝜀2 f cos(t)
(1.27)
We demonstrate that, under appropriate conditions, a stable limit cycle appears
and calculate the relative approximate solution. Moreover, we derive sufficient
conditions for the existence of a doubly periodic motion when the fundamental
oscillation is subjected to a slight modulation, with an amplitude proportional to

the magnitude of the periodic excitation.
Finally, in the last section, we briefly recapitulate the most important results and
indicate some possible generalizations of the present study.
1.3 The Approximate Solution
The AP method we use to calculate the approximate solution was first developed in
Refs. [1, 2], and then in this section, we sketch the main steps of this perturbation
technique.
First of all, we now introduce a rational number
q = rational number (1.28)
the temporal rescaling
t = eq t (1.29)
where the rational number q will be fixed afterwards because it establishes to what
extent we can push the temporal asymptotic limit in such a way that the nonlinear
effects become consistent and not negligible. If t → ∞, then 𝜀 → 0, when 𝜏 assumes
a finite value.
If we take 𝜀 = 0 in Eqs. (1.26) and neglect nonlinear terms, we see that it admits
simple harmonic solutions X(t) = A exp(−it) + c. c., where A is a constant depend-
ing on initial conditions and c. c. stands for complex conjugate. Nonlinear effects
induce a modulation of the amplitude A and the appearance of higher harmonics.
The modulation is best described in terms of the rescaled variable t that accounts for
the need to look on larger time scales, to obtain a nonnegligible contribution from
the nonlinear term.
The assumed solution X(t) of (1.26) can be expressed by means of a power series
in the expansion parameter 𝜀, we formally write
∑
+∞
X(t) = 𝜀𝛾n 𝛹n (t, 𝜀) exp(−int) (1.30a)
n=−∞
with 𝛾 n = |n| for n ≠ 0, and 𝛾 0 = r is a positive number, which will be fixed later on;
in consequence of the reality of (1.30a)
𝛹n (t, 𝜀) = c.c.(𝛹(−n) (t, 𝜀)) (1.30b)
The assumed solution (1.30a) can be considered a combination of the different har-
monics, solutions of the linear equation, i.e. of the equation obtained after neglecting
all the nonlinear terms, and the coefficients of this combination depend on 𝜏 and 𝜀.
Eq. (1.30a) can be written more explicitly
X(t) = 𝜀r 𝛹0 (t; 𝜀) + 𝜀𝛹1 (t; 𝜀) exp(−it) + 𝜀2 𝛹2 (t; 𝜀) exp(−2it) + 𝜀3 𝛹3 (t; 𝜀) exp(−3it)o(𝜀4 )

(1.30c)
The functions 𝛹 n (t, 𝜀) depend on the parameter 𝜀, and we suppose that 𝛹 n ’s limit
for 𝜀 → 0 exists and is finite and, moreover, they can be expanded in power series of
𝜀, i.e.
∑
∞
𝛹n (𝜏; 𝜀) = 𝜀i 𝛹n(i) (𝜏) (1.31)
i=0
In the following, for simplicity, we use the abbreviations 𝛹n(0) = 𝛹n for n ≠ 1 and
𝛹1(0)
= 𝛹 for n = 1.
Note that the variable change (1.29) implies that
( )
(𝜕𝜓n exp(−int)) 𝜕𝜓
= −in𝜓n + 𝜀q n exp(−int) (1.32)
𝜕t 𝜕t
After inserting this expansion into Eq. (1.26), we obtain equations for every har-
monic and for a fixed order of approximation, which are right for the purpose of
determining the coefficients.
For n = 0, we obtain
𝜀r 𝛹0 + 2f2 |𝛹 |2 𝜀2 + o(𝜀4 , 𝜀r+2 ) = 0 (1.33a)
A correct balance of terms shows r = 2, and then we derive the following relation
𝜀2 𝛹0 = −2𝜀2 f2 (|𝛹 |2 ) + O(𝜀4 ) (1.33b)
For n = 2, taking into account Eq. (1.32), we have
−3𝜀2 𝛹2 + f2 𝛹 2 𝜀2 = −ig1 𝜀2 𝛹 2 + o(𝜀4 , 𝜀2+q ) (1.34a)
and then
f2 + ig1 2 2
𝜀2 𝛹 2 = 𝜀 𝛹 + o(𝜀4 ) (1.34b)
3
For n = 1, Eq. (1.26) yields for the right-hand side
d𝛹 1+q
−2i 𝜀 + 2f2 (𝛹0 𝛹 𝜀3 + 𝛹2 (c.c.(𝛹 ))𝜀3 ) + 3f3 |𝛹 |2 𝛹 𝜀3 (1.35a)
dt
and for the left-hand side
i𝜀3 g0 𝛹 − ig1 (𝛹0 𝛹 𝜀3 + 𝛹 𝛹2 (c.c.𝛹 )𝜀3 ) − i𝜀3 g2 |𝛹 |2 𝛹 + o(𝜀5 , 𝜀1+2q ) (1.35b)
If q = 2, the first term has the same magnitude order of nonlinear terms.
Taking into account Eqs. (1.33b) and (1.34b), we can derive a differential equation,
which involves only 𝛹 ,
d𝛹
= 𝛼1 𝛹 + (𝛽1 + i𝛽2 )|𝛹 |2 𝛹 (1.36)
dt
with
g0
𝛼1 = (1.37)
2
g2 g1 f2
𝛽1 = − (1.38)
2 2
g12 3 5
𝛽2 = − f3 + f22 (1.39)
6 2 3
Substituting the polar form

𝛹 (𝜏) = 𝜌(𝜏) exp(i𝜃(𝜏)) (1.40)
into Eq. (1.36), and separating real and imaginary parts, we arrive at the following
model system:
d𝜌
= 𝛼1 𝜌 + 𝛽1 𝜌3 (1.41)
dt
dJ
= 𝛽2 𝜌2 (1.42)
dt
As we can see from Eqs. (1.30c), (1.31), and (1.40), the approximate solution of
Eq. (1.26) can be written as a sum of a contribution of order 𝜀 and a contribution of
order 𝜀2
X(t) = 𝜀X1 (t) + 𝜀2 X2 (t) + o(𝜀3 ),
X1 (t) = 2𝜌(𝜏) cos(−t + 𝜃(𝜏)),
2 2
X2 (t) = −2f2 𝜌2 (𝜏) + f2 𝜌2 (𝜏) cos(−2t + 2𝜃(𝜏)) + g1 𝜌2 (𝜏) sin(−2t + 2𝜃(𝜏))
3 3
(1.43)
By inspection of Eq. (1.41), which can be easily integrated, we conclude that a
stable steady-state response is possible if 𝛼 1 > 0 and 𝛽 1 < 0. In this case, we obtain a
stable equilibrium point, which corresponds to a stable limit cycle for Eq. (1.26), and
its approximate expression is given by (1.43), with
√
𝛼
𝜌(t) = 𝜌E = − 1 = constant, 𝜃(t) = 𝛽2 𝜌E t (1.44)
𝛽1
The natural frequency of the oscillator is subject to a slight modification and
becomes
𝜔E = 𝜔 − 𝛽2 𝜌E (1.45)
If we want to improve the validity of the approximate solution, we must include
higher order terms. However, we can easily conclude that 𝛹0(1) = 𝛹1(1) = 𝛹2(1) = 0
(for their definition, see Eq. (1.31)). Indeed, we consider Eq. (1.26) for n = 0 and
Eqs. (1.33b) and (1.34a) for n = 0 and n = 2 in such a way to obtain
( ( ))
2
𝛹0(1) = −2f2 𝛹1(1) (c.c.𝛹 ) + 𝛹 c.c. 𝛹1(1) , 𝛹2(1) = ( f2 + ig1 )𝛹1(1) 𝛹 (1.46)
3
After inserting (1.26b) into (1.26a), we see that the resulting equation is satisfied
if 𝛹1(1) = 0. Recall that we can always assume that the initial condition is 𝛹1(1) (0) = 0,
because the initial conditions associated with equation (1.25), X(0) = X 0 and
̇
X(0) = Ẋ 0 , can be used to determine 𝛹 (0) = 𝜌(0) exp(i𝜃(0)).
A valid higher order approximation can be derived only if we take into account
𝛹1(2) , 𝛹2(2) , 𝛹0(2) .
For n = 0, we derive the following relation
( (
𝜀2 𝛹0 + 𝜀4 𝛹0(2) = (−2f2 𝜀2 + A1 𝜀4 )|𝛹 |2 − 2f2 𝜀4 𝛹1(2) (c.c.𝛹 )
( ) ) )
+ c.c.𝛹1(2) 𝛹 + A2 𝜀4 |𝛹 |4 + h.o.t. (1.47a)
where h. o. t = higher order terms and

38 3 11 2
A1 = g0 g1 , A2 = 10f3 f2 − f + g1 g2 − fg (1.47b)
9 2 9 2 1
The obvious conclusion is
𝛹0(2) = (A1 |𝛹 |2 + A2 |𝛹 |4 ) − 2f2 (c.c.𝛹 )𝛹1(2) + 𝛹 (c.c.𝛹 )(2)
1 (1.47c)
In a similar way, for n = 2, we obtain
𝛹2(2) = (B1 + iB ̃2 )|𝛹 |2 𝛹 2 + 2 (f2 + ig1 )𝛹 𝛹 (2)

̃1 )𝛹 2 + (B2 + iB (1.48)
1
3
with
( )
7 ̃1 = 2 f2 g0
B1 = − g0 g1 , B (1.49a)
18 9
7 5 19 7 21
B2 = − f 32 − f2 f3 − gg + g f2 (1.49b)
18 4 36 1 2 18
̃2 = − 19 g3 − 29 f2 g2 + 15 f3 g1 − 15 f 2 g1
B (1.49c)
288 1 12 4 4 2
If we neglect only terms of order 𝜀6 or higher, Eq. (1.33a–c) transforms into
(
d𝛹 (2) ( )
+ 𝜀 −2i 1 + (2f2 + ig1 ) 𝛹0 𝛹1(2) + 𝛹2 𝛹1(2) + (6f3 + 2ig2 )|𝛹 |2 𝛹1(2)
5
dt
)
+ (3f3 + ig2 )𝛹 2 𝛹1(2) + ig0 𝛹1(2)
( 2 ( ))
d𝛹 (2) (2)
+ 𝜀5 + (2f 2 + ig1 ) 𝛹 0 𝛹 + 𝛹 2
(c.c.𝛹 ) + (2f2 + ig1 )𝛹 3 (c.c.𝛹 2 )
d𝜏 2
( )
+ 𝜀5 (6f3 + 2ig2 )|𝛹2 |2 + (3f3 + ig2 )𝛹02 𝛹
( )
d𝛹 g2 d|𝛹 |2 𝛹
+ 𝜀5 (6f3 + 2ig2 )𝛹2 𝛹0 (c.c.𝛹 ) − g0 −
d𝜏 3 d𝜏
( ( ))
g 1 (d(𝛹 0 𝛹 + 𝛹 2 (c.c.𝛹 )))
+ 𝜀5 + o(𝜀7 ) (1.50)
2 (d𝜏)
The term ddt𝛹2 in Eq. (1.50) can be eliminated taking into account that if we differ-
2
entiate Eq. (1.36), we have

d2 𝛹 ( )
2
= 𝛼12 𝛹 + 4𝛼1 (𝛽1 + i𝛽2 )|𝛹 |2 𝛹 + 3𝛽12 + 4i𝛽1 𝛽2 − 𝛽22 |𝛹 |4 𝛹 (1.51)
dt
Moreover, from (1.50), we see that it is necessary to consider Eq. (1.26) for n = 3
𝛹3 = (C1 + iC2 )𝛹 3 (1.52a)
where
3f3 + 2f22 − 24g12 (3g2 + 5f2 g1 )
C1 = , C2 = (1.52b)
24 24
If we use the abbreviation
𝛹m = 𝛹 + e2 𝛹1(2) (1.53a)
and Eq. (1.47c),(1.48), and (1.52a), then the final form of Eq. (1.46) is
d𝛹m
𝛼2 )𝛹m + (𝛽1 + 𝛽̃1 + i𝛽2 + i𝛽̃2 )|𝛹m |2 𝛹m + (̃
= (𝛼1 + ĩ g2 )|𝛹m |4 𝛹m
g1 + ĩ
dt
(1.53b)
with
( )
𝛼12g0 𝛼1
𝛼
̃2 = − + (1.54)
2 2
g1 g 1 3 5
𝛽̃1 = A1 + f2 B
̃1 + 1 B − g g2 − g f + g f 2 (1.55)
2 2 24 0 1 4 0 3 6 0 2
g1 g0 g2 3
𝛽̃2 = B ̃ −f B −f A − − g0 g1 f2 (1.56)
2 1 2 1 2 1 2 8
g1 g
̃
𝛾1 = ̃2 + 1 B2 + 11 g2 g12 + 1 f2 g3 + 63 g2 f 2 − 35 g1 f 3
A2 + f2 B
2 2 72 36 1 18 2 36 2
5 21
− f3 g2 + g1 f2 f3 (1.57)
4 24
g1 g2 g g f
̃
𝛾2 = ̃2 − f2 B2 − f2 A2 − 2 − 2 1 2 + 9 f 3 + 25 f 4 − 53 f3 f 2
B
2 8 8 8 2 18 2 6 2
11 2 7 2 2
− f3 g1 + g1 f2 (1.58)
24 18
and we arrive at the following modified system model
d𝜌
= 𝛼1 𝜌 + (𝛽1 + 𝛽̃1 )𝜌3 + ̃
𝛾1 𝜌5 (1.59)
dt
d𝜃
= (𝛽2 + 𝛽̃2 )𝜌2 + ̃
𝛾2 𝜌4 (1.60)
dt
The approximate solution up to the o(e4 )-th order is
X(t) = 𝜀X1 (t) + 𝜀2 X2 (t) + 𝜀3 X3 (t) + 𝜀4 X4 (t) + o(𝜀5 ),
X1 (t) = 2𝜌(t) cos(−t + 𝜃(t)),
2 2
X2 (t) = − 2f2 𝜌2 (t) + f2 𝜌2 (t) cos(−2t + 2𝜃(t)) + g1 𝜌2 (t) sin(−2t + 2𝜃(t)),
3 3
X3 (t) = 2C1 𝜌3 (t) cos(−3t + 3𝜃(t)) + 2C2 𝜌3 (t) sin(−3t + 3𝜃(t))
X4 (t) = 2D1 𝜌4 (t) cos(−4t + 4𝜃(t)) + 2D2 𝜌4 (t) sin(−4t + 4𝜃(t)). (1.61a)
where
( )
1 5 3 5 31 11
D1 = f2 + f2 f3 − g12 f2 − g1 g2 (1.61b)
15 18 4 9 6
( )
1 43 3 19 3 38
D2 = g1 f2 + f2 g2 + f3 g1 − g13 (1.61c)
15 36 12 2 9
We can calculate a more accurate expression for the limit cycle, and equations
(1.44) and (1.45) become
( )
𝛾1 𝛼1 − 𝛽1 𝛽̃1
̃
𝜌 E = 𝜌E 1 +
M
(1.62)
2𝛽12
( )2
𝜔M ̃2 + 𝛽̃2 𝜌M
E = 𝜔E + 𝛼 E + 𝛾̃2 𝜌4E (1.63)
If we examine the Eq. (1.59) carefully, we can easily understand that there is the
possibility of two limit cycles and then of a qualitative change in the behavior of the
oscillator with respect to the previsions of the first-order approximation. In fact, we
suppose now that the dissipative coefficient g0 is of order e4 and g2 is chosen in such
a way that g2 − g1 f 2 is of order 𝜀, then 𝛼 1 (see (1.37)) and 𝛽 1 (see (1.38)) are of order
𝜀4 and 𝜀2 , respectively. For example, we can take g0 = 0.0001, g1 = 1, g2 = 1.01, and
f 2 = 1.
Taking into account that 𝜌 must be of order e, then all terms in (39) have the same
magnitude order and we can obtain for the equilibrium values of 𝜌 two positive roots
of order e and then two different limit cycles. Depending on the parameters, the
larger limit cycle and the origin can be asymptotically stable and the smaller unstable
or vice-versa.
By means of the variable change
√
g1 |
|̃ |𝛽 |
| |
t → 2 t, 𝜌 → | 1 |𝜌 (1.64)
𝛽 1 |̃ |
| g1 |
which implies
|̃
𝛾 |𝛼
𝛼1 → 1 2 1 (1.65)
𝛽 1
we can always set (𝛽1 + 𝛽̃1 ) = ±1, ̃ 𝛾1 = ±1.
There are four distinct cases:
(i) (𝛽1 + 𝛽̃1 ) = 1, ̃
𝛾1 = 1: for 𝛼 1 < 0, there is an unstable limit cycle, and for 𝛼 1 > 0,
no limit cycle appears;
(ii) (𝛽1 + 𝛽̃1 ) = −1, ̃𝛾1 = −1: for 𝛼 1 < 0, there is no limit cycle, and for 𝛼 1 > 0, only a
stable limit cycle appears. Note that cases (i) and (ii) are connected by a simple
temporal inversion, followed by the change of the sign of 𝛼 1 .
(iii) (𝛽1 + 𝛽̃1 ) = −1, ̃𝛾1 = 1: for 𝛼 1 < 0, we have a larger stable limit cycle and a
smaller unstable limit cycle; for 𝛼 1 > 0, only the stable limit cycle is present
(Figure 1.1).
(iv) (𝛽1 + 𝛽̃1 ) = 1, ̃𝛾1 = −1: for 𝛼 1 < 0, there is only an unstable limit cycle and for
𝛼 1 > 0, a larger unstable limit cycle and a smaller stable limit cycle are present.
Note that the last cases are also connected by a temporal inversion.
In all these cases, the origin is stable for 𝛼 1 < 0 and unstable for 𝛼 1 > 0.
1.4 Comparison with the Results of the Numerical

Integration
To verify our analysis, we have computed a numerical solution of (1.26), by means of
the standard Runge–Kutta method. A computer search was conducted to find these
solutions, and their periodicity was verified. We have chosen the following set of
parameters
f2 = 2, f3 = 1, g0 = 0.01, g1 = 2, g2 = −2
Amplitude
0.6
0.3
Dissipation coefficient
0.5 0.0 0.5
Figure 1.1 Dissipation (𝛼 1 )–response (𝜌E ) space. Rectangles are stable limit cycles, and
crosses represent unstable limit cycles.
We expect the appearance of a stable limit cycle with amplitude 𝜌M E

= 0.035 (see
equation (42) and case (iii) of the precedent section). In Figure 1.2, we show a com-
parison between the approximate solution (41) and the numerical solution: crosses
represent the approximate solution and circles represent the numerical solution.
Only a cycle is represented, as the solution repeats itself one cycle after another.
The agreement of the results appears to be excellent because the maximum differ-
ence is 6 ⋅ 10−5 and the medium difference is 2 ⋅ 10−5 , i.e. of order e5 as expected.
In Figure 1.3, we have increased the dissipation coefficient of an order of magni-
tude (g0 = 0.1), and even in this case, we expect the appearance of a stable limit cycle
with amplitude 𝜌M E
= 0.141 (see Eq. (1.42) and case (iii) of the precedent section). The
maximum difference between the approximate solution (1.61a–c) and the numerical
solution is now 5 ⋅ 10−2 and the medium difference is 2 ⋅ 10−3 .
The AP method is then a valid tool to approximate solutions of nonlinear oscilla-
tors with small dissipation coefficients.
1.5 External Excitation in Resonance with the Oscillator

In this section, we consider a nonlinear oscillator in resonance with an external force
(see (1.27)). The AP method can be applied by following the same direction as in the
preceding section. We obtain
d𝛹 f
= 𝛼1 𝛹 + (𝛽1 + i𝛽2 )|𝛹 |2 𝛹 + i (1.66)
dt 2
Y
0.10
0.05
–0.10 –0.05 0.05 0.10 X
–0.05
–0.10
̇
Figure 1.2 Phase space diagram (X(t), Y(t)) with Y(t) = X(t) with f 2 = 2, f 3 = 1, g0 = 0.01,
g1 = 2, and g2 = 2. Circles are the numerical solution, and crosses represent the
approximate solution.
Y 0.4
0.2
–0.4 –0.2 0.2 0.4 X
–0.2
–0.4
̇
Figure 1.3 Phase space diagram (X(t), Y(t)) with Y(t) = X(t) and f 2 = 2, f 3 = 1, g0 = 0.1,
g1 = 2, and g2 = −2. Circles are the numerical solution and crosses represent the
approximate solution.
and then
d𝜌 f
= 𝛼1 𝜌 + 𝛽1 𝜌3 + sin𝜃 (1.67)
dt 2
dJ f
𝜌 = 𝛽2 𝜌3 + cos 𝜃 (1.68)
dt 2
We can easily determine the equilibrium points and their possible stability. By
means of the variable change
√
1 |𝛼 |
t→ t, 𝜌 → || 1 ||𝜌 (1.69)
|𝛼1 | | 𝛽1 |
which implies
𝛽2 → |𝛽1 |𝛽2 , 𝛼2 → |𝛼1 |𝛼2 (1.70)
we can always set 𝛼 1 = ± 1, 𝛽 1 = ± 1.

The equilibrium points 𝜌0 must satisfy the equation
f2
Br3 + 2sr2 + r − = 0 r = 𝜌20 (1.71a)
4
( )
𝛼1 + 𝛽1 r
𝜃0 = arctan (1.71b)
𝛽2 r
where
B = 1 + 𝛽22 s = 𝛼1 𝛽1 = ±1 (1.71c)
Every equilibrium point of (1.71a–c) corresponds to a periodic solution of the start-

ing system (1.27), and it is easily calculated by means of the standard formulas for
the roots of a third-order equation.
The standard linearization method permits the computation of the Lyapunov
exponents relative to each equilibrium point. They are
√ √
𝜆1 = 𝛼1 + r(2𝛽1 + 𝛥), 𝜆2 = 𝛼1 + r(2𝛽1 − 𝛥) (1.72a)
where
𝛥 = 1 − 3𝛽22 (1.72b)
We can now follow the stationary solutions of (1.71a–c) as the amplitude of

the external force is varied, and at each step, along the curve that develops in
external force–response space, the stability of the solution of (51) is determined by
Eq. (52).
If 3B > 4, we have only an equilibrium point (Figure 1.4). If 3B < 4, s > 0, there is
only an equilibrium point (Figure 1.5).If 3B < 4, s < 0, we have one, two, or three
equilibrium points (Figure 1.6); we observe a saddle–node fold (or cyclic fold) bifur-
cation, and the solution jumps up to a larger stable orbit as the amplitude of the
external force is increased. A fold bifurcation corresponds to a vertical tangency in
the external force–response space, where the derivative of the response with respect
to the control parameter is infinite.
Amplitude
1.0
0.5
External force
1.0 2.0
Figure 1.4 External force–response space. Circles are sources.
Amplitude
2.0
1.0
External force
0.5 1.0
Figure 1.5 External force–response space. Rectangles are sinks.

Amplitude
2.0
1.0
External force
0.5 1.0
Figure 1.6 External force–response space. Rectangles are sinks, crosses represent saddle
points, and circles stand for sources.
We now consider the system (1.67)–(1.68) when 𝜌(t) is near (1.44). If the external
excitation is sufficiently small, we obtain
d𝜌 f
= −2𝛼1 (𝜌 − 𝜌2 ) + sin(𝛺t + 𝜃0 ) (1.73)
dt 2
with
𝜃(t) = 𝛺t + 𝜃0 (1.74a)
𝛼2 𝛽1 − 𝛽2 𝛼1
𝛺= (1.74b)
𝛽1
where 𝜃 0 depends on the initial conditions.
Equation (1.85) can be easily resolved and we get
𝜌(t) = 𝜌2 + (𝜌0 − 𝜌2 ) exp(−2𝛼1 t)

2𝛼 sin(𝛺t + 𝜃0 )−𝛺 cos(𝛺t + 𝜃0 )+(−2𝛼1 sin(𝜃0 )+𝛺 cos(𝜃0 )) exp(−2𝛼1 t)
+𝜈 1
4𝛼 21 + 𝛺2
(1.75)
where 𝜌0 depends on the initial conditions.

We conclude that the resulting motion is quasiperiodic, with the frequencies of 1
and Ω.
If we insert (1.74) and (1.75) into (1.43), we obtain the approximate solution up to
the O(e3 )-th order.
Y
0.20
0.10
–0.20 –0.10 0.10 0.20 X
–0.10
–0.20
Figure 1.7 Associated map of the nonautonomous Eq. (1.27) with f 2 = −1, f 3 = −1,
g0 = 0.02, g1 = 1, g2 = −3, and f = 3.5 ⋅ 10−4 . Crosses are the approximate solution, and
circles represent the numerical solution.
In Figure 1.7, we show a comparison between the approximate solution (1.43)

and the numerical solution. We represent the associated map of the nonau-
tonomous Eq. (1.27), which is obtained with the values (X(0),Y (0)), (X(T),Y (T)),
(X(2T),Y (2T)), …, where T is the period of the external excitation. Crosses represent
the approximate solution and circles represent the numerical solution.
The closed curves reveal that the motion is quasiperiodic because of the presence
of the frequency (1.74a). The agreement of the results is excellent because the max-
imum difference is 0.0075 and the medium difference is 0.0058, i.e. of order 𝜀3 as
expected.
1.6 Conclusion
We have demonstrated the power of the AP method and how it produces useful
approximate solutions. In particular, we have treated a class of strongly nonlinear
oscillators subject to an external periodic force in resonance with the natural fre-
quency of the oscillator.
We have found bifurcations and limit cycles, which are influenced by the presence
of external excitation. In addition, we can observe a quasiperiodic motion, charac-
terized by the combination of the natural frequency with a low frequency connected
to the external excitation.
1.6 Conclusion 17
We indicate two possible extensions of the present chapter:

(i) study of subharmonic and superharmonic resonances for the system (6), with
the observation of period-doubling bifurcations;
(ii) study of more complicated dynamical systems, such as three-dimensional sys-
tems or coupled oscillators, eventually subject to external excitations.
19
The Asymptotic Perturbation Method for Remarkable

Nonlinear Systems
2.1 Introduction
A parametrically excited Liénard system is investigated by an asymptotic per-
turbation method based on Fourier expansion and time rescaling. Two coupled
equations for the amplitude and the phase of solutions are derived. Their fixed
points correspond to limit cycles for the Liénard system, and we determine the
stability of steady-state response as well as response–parametric excitation and
response–frequency curves. We use the Poincarè–Bendixson theorem, the Dulac’s
criterion, and energy considerations to study the existence and characteristics of
limit cycles of the two coupled equations. A limit cycle corresponds to a mod-
ulated motion in the Liénard system. We show that modulated motion can also
be obtained for very low values of the parametric excitation and construct an
approximate analytic solution. Moreover, we observe an unusual infinite-period
homoclinic bifurcation, as in certain cases, due to the symmetry of the two coupled
equations, two stable limit cycles approach a saddle point and merge to form a
greater stable limit cycle. Subsequently, this limit cycle and another unstable limit
cycle coalesce and annihilate through a saddle–node bifurcation. The comparison
with the solution obtained by the numerical integration confirms the validity of our
analysis. It is well known that many mechanical systems and oscillating circuits
can be modeled by the Liénard system [65, 84, 226]
̈ + f (X(t)) = g(X(t))X(t)
X(t) ̇ (2.1)
where the dot denotes differentiation with respect to the time and the arbitrary func-
tions f (x) and g(x) are supposed to be analytic. Obviously, Eq. (2.1) can be interpreted
mechanically as the equation of motion for a unit mass subject to a nonlinear damp-
ing force g(x)ẋ and a nonlinear restoring force −f (x).
We restrict our study to the following particular case of Eq. (2.1)
X(t) ̇
̈ + 𝜔20 X(t) + f2 X 2 (t) + f3 X 3 (t) = g0 X(t) ̇
+ g1 X(t)X(t) ̇
+ g2 X 2 (t)X(t) (2.2a)
corresponding to the choice
f(x) = 𝜔20 x + f2 x2 + f3 x3 , g(x) = g0 + g1 x + g2 x2 (2.2b)
Asymptotic Perturbation Methods: For Nonlinear Differential Equations in Physics,

First Edition. Attilio Maccari.
© 2023 WILEY-VCH GmbH. Published 2023 by WILEY-VCH GmbH.
20 2 The Asymptotic Perturbation Method for Remarkable Nonlinear Systems
Equation (2.2a,b) can be also considered a generalized Van der Pol–Duffing sys-
tem, because it includes as particular cases the Van der Pol oscillator ( f 2 , f 3 , g1 = 0
and g0 = −g2 ≠ 0) and the Duffing equation (f 2 = g1 = g2 = 0 and g0 , f 3 ≠ 0). Through
a second-order perturbation analysis, we have investigated Eq. (2.2a,b) in [119], and
the existence of one or two limit cycles has been demonstrated. Moreover, a resonant
external excitation has been added to Eq. (2.2a,b) and, by a lower order perturbation
analysis, we have derived a sufficient condition to obtain a two-period quasiperi-
odic motion when a second low frequency appears in addition to the forcing fre-
quency. Finally, analytic approximate solutions have been checked by numerical
integration.
In this chapter, we consider the Liénard oscillator in the case of parametric exci-
tation. In particular, we introduce a principal parametric excitation 2fX(t) cos(2𝛺t)
into Eq. (2.2a,b) and obtain
X(t) ̇
̈ + 𝜔20 X(t) + f2 X 2 (t) + f3 X 3 (t) = g0 X(t) ̇
+ g1 X(t)X(t)
2 ̇
+ g2 X (t)X(t) + 2fX(t) cos(2𝛺t) (2.3)
Principal parametric resonance or 1/2-subharmonic external resonance has been
studied in many papers and systems similar to Eq. (2.3) have been considered in
[13, 204].
In Section 2.2, we use the AP method to calculate an approximate analytic
solution. We derive a model system of two coupled differential equations in the
phase and amplitude of solutions. Stable (unstable) fixed points of the model system
correspond to stable (unstable) limit cycles of the parametrically excited Liénard
system (2.3). Subsequently, we compare frequency–response and parametric
excitation–response curves with the results of the numerical integration.
In Section 2.3, we perform a global analysis of the model system by the
Poincarè–Bendixson theorem, the Dulac’s criterion, and energy considerations.
We derive several conditions in order to exclude or permit the existence of limit
cycles, which correspond to the modulated motions for the Liénard systems (2.3).
Moreover, we study the modulated motion that appears for very low values of the
parametric excitation. In this case, the two-period quasiperiodic motion is charac-
terized by a slight modulation of the fundamental oscillation with a modulation
amplitude proportional to the magnitude of the parametric excitation.
In Section 2.4, we demonstrate the presence of an unusual infinite-period homo-
clinic bifurcation. In certain cases, the symmetry of the model system gives rise to
the appearance of two stable limit cycles, a saddle point, and a surrounding unstable
limit cycle. If we increase the parametric excitation, the two limit cycles approach
the saddle point and subsequently merge to form a greater stable limit cycle. Finally,
with a further increase in the parametric excitation, this limit cycle and the unstable
limit cycle coalesce and disappear through a saddle–node bifurcation.
Note that in the usual homoclinic bifurcation, a limit cycle approaches a saddle
point, and at the bifurcation point, the limit cycle and a branch of both the stable and
unstable manifolds of the saddle point coincide, forming a homoclinic connection,
which is essentially a cycle limit with an infinite period. As the control parame-
ter increases, the limit cycle suddenly vanishes (“blue sky catastrophe”). On the
contrary, in the model system of the parametrically excited Liénard oscillator, we

observe two limit cycles that at the same time touch the saddle point, merge, and
give rise to a great limit cycle surrounding the saddle point.
In Section 2.5, we briefly recapitulate the most important results and indicate some
possible generalizations of the present study.
2.2 Periodic Solutions and Their Stability

The AP method (see Chapter 1 and also [112, 118, 120]) can be considered as a tech-
nique that links together the most useful characteristics of the harmonic balancing
and the multiple scales method [202], but its origin can be found in the theory of
weakly nonlinear partial differential equations ([114, 122] and Chapter 6). We intro-
duce a detuning parameter s defined by 𝜔0 = 𝛺 + 𝜀2 𝜎 and in order to balance the
effect of the nonlinearity and damping with the parametric excitation, we scale the
parametric excitation coefficient f and the dissipative coefficient g0 as 𝜀2 f , 𝜀2 g0 .
No conditions are imposed on the coefficients f 2 , f 3 , g1 , g2 , which can be of order 1.
Equation (2.3) yields
̈ + 𝛺2 X(t) + 2𝜀2 𝜎𝛺X(t) + 𝜀4 𝜎 2 X(t) + f2 X 2 (t) + f3 X 3 (t)
X(t)
̇
= g0 X(t) ̇
+ g1 X(t)X(t) ̇
+ g2 X 2 (t)X(t) + 2𝜀2 fX(t) cos(2𝛺t) (2.4)
We now introduce the slow time
t = eq t (2.5)
where q is a rational positive number, which will be fixed afterward. The value of q
fixes the magnitude order of the temporal asymptotic limit in such a way that the
nonlinear effects become consistent but not negligible. If t → + ∞, then 𝜀 → 0, when
𝜏 assumes a finite value.
If we take 𝜀 = 0 in Eq. (2.4) and neglect nonlinear terms, we see that it admits sim-
ple harmonic solutions X(t) = A exp(−i𝛺t) + c. c., where A is a constant depending
on initial conditions and c. c. stands for complex conjugate. We expect that nonlinear
effects will induce a modulation of the amplitude A and the appearance of higher
harmonics. By means of the rescaled time (5), we can describe the slow modulation
due to the nonlinear terms.
The assumed solution X(t) of Eq. (2.4) can be expressed by means of a power series
in the expansion parameter 𝜀, we formally write
∑
∞
X(t) = 𝜀𝛾n 𝜓n (𝜏, 𝜀) exp(−in𝛺t) (2.6)
n=−∞
with 𝛾 n = |n| for n ≠ 0, and 𝛾 0 = r is a positive number, which will be fixed later on;
being the assumed solution X(t) is real, we have
𝜓n (t, 𝜀) = 𝜓−n
∗
(t, 𝜀) (2.7)
The assumed solution (2.6) can be written more explicitly
X(t) = 𝜀r 𝜓0 (𝜏; 𝜀) + 𝜀𝜓1 (𝜏; 𝜀) exp(−i𝛺t) + 𝜀2 𝜓2 (𝜏; 𝜀) exp(−2𝛺it) + O(𝜀3 ) (2.8)
and we see that it can be considered a combination of the various harmonics with
coefficients depending on 𝜏 and 𝜀.
We suppose that the functions 𝜓 n (𝜏, 𝜀)s can be expanded in power series of, i.e.
∑
∞
𝜓n (𝜏; 𝜀) = 𝜀i 𝜓n(i) (𝜏) (2.9)
i=0
We have assumed in Eq. (2.9) that the limit of the 𝜓 n (𝜏; 𝜀)’s for e → 0 exists and is
finite. In the following, for simplicity, we use the abbreviations 𝜓n(0) = 𝜓n for n ≠ 1
and 𝜓1(0) = 𝜓 for n = 1.
Note that the introduction of the slow time (2.5) implies that
( )
d q d𝜓n
(𝜓 exp(−in𝛺t)) = −in𝛺𝜓n + 𝜀 exp(−in𝛺t) (2.10)
dt n d𝜏
In order to determine the coefficients 𝜓 n (𝜏; 𝜀), we insert the assumed solution
(2.8) into Eq. (2.4), and obtain various equations for each harmonic n and for a fixed
order of approximation on the perturbation parameter 𝜀.
For n = 0, we get
2f2
𝜀r 𝜓0 + |𝜓|2 𝜀2 + O(𝜀4 , 𝜀r+2 ) = 0 (2.11)
𝛺2
A correct balance of terms shows that r = 2 and then we derive the following rela-
tion
2f
𝜀2 𝜓0 = −𝜀2 22 (|𝜓|2 ) + O(𝜀4 ) (2.12)
𝛺
For n = 2, taking into account Eq. (2.10), we have
−3𝜀2 𝛺2 𝜓2 + f2 𝜀2 𝜓 2 = −ig1 𝛺𝜀2 𝜓 2 + O(𝜀4 , 𝜀2+q ) (2.13)
and then
f2 + ig1 𝛺 2 2
𝜀2 𝜓2 = 𝜀 𝜓 + O(𝜀4 ) (2.14)
3𝛺2
For n = 1, Eq. (2.4) yields
d𝜓 1+q
2i𝛺 𝜀 − 2𝜀3 𝜎𝛺𝜓 − 2𝜀3 f2 (𝜓0 𝜓 + 𝜓2 𝜓 ∗ ) − 3𝜀3 f3 |𝜓|2 𝜓 − i𝜀3 𝛺g0 𝜓
d𝜏
− i𝜀3 g1 𝛺(𝜓0 𝜓 + 𝜓2 𝜓 ∗ ) − i𝜀3 g2 𝛺|𝜓|2 𝜓 + f 𝜓 ∗ + O(𝜀5 , 𝜀1+2q ) = 0 (2.15)
If q = 2, the first term has the same magnitude order of all the other terms.
As we can see from Eqs. (2.12) and (2.14), we can derive a differential equation for
the evolution of the complex amplitude,
d𝜓
= (𝛼1 + i𝛼2 )𝜓 + (𝛽1 + i𝛽2 )|𝜓|2 𝜓 + i𝛾𝜓 ∗ (2.16)
d𝜏
with
g0 f
𝛼1 = , 𝛼 = −𝜎, g = (2.17)
2 2 2𝛺
( )
g gf 1 g21 3f3 5f 2
𝛽1 = 2 − 1 22 , 𝛽2 = − + (2.18)
2 2𝛺 𝛺 6 2 3𝛺2
Substituting the polar form

𝜓(t) = 𝜌(𝜏) exp(i𝜃(𝜏)) (2.19)
into Eq. (2.16), and separating real and imaginary parts, we arrive at the following
model system
d𝜌
= 𝛼1 𝜌 + 𝛽1 𝜌3 + 𝛾𝜌 sin2𝜃 (2.20)
d𝜏
d𝜃
= 𝛼2 + 𝛽2 𝜌2 + 𝛾 cos2𝜃 (2.21)
d𝜏
Taking into account Eqs. (2.12) (2.14) and (2.19), the approximate solution of
Eq. (2.4) can be written as a sum of a contribution of order 𝜀 and a contribution of
order 𝜀2
X(t) = 𝜀X1 (t) + 𝜀2 X2 (t) + O(𝜀3 )
X1 (t) = 2𝜌(t) cos(𝛺t − 𝜃(t))
2f 2f 2g
X2 (t) = − 22 𝜌2 (t) + 22 𝜌2 (t) cos(2𝛺t − 2𝜃(t)) + 1 𝜌2 (t) sin(2𝛺t − 2𝜃(t))
𝛺 3𝛺 3𝛺
(2.22)
The validity of the approximate solution should be expected(to )be restricted to
bounded intervals of the t-variable and then on time-scale t = O 𝜀12 . If one wishes
( )
to construct approximate solutions on larger intervals such that 𝜏 = O 1𝜀 , then the
higher terms will in general affect the solution and must be included. Moreover, the
approximate solution (2.22) will be within O(𝜀3 ) of the true solution on bounded
intervals of the t-variable, and if the solution is periodic, for all t.
By means of the variable change
√
1 |𝛼 |
t→ t, 𝜌 → || 1 ||𝜌 (2.23)
|𝛼1 | | 𝛽1 |
which implies
g → g|𝛼1 |, 𝛽2 → |𝛽1 |𝛽2 , 𝛼2 → |𝛼1 |𝛼2 (2.24)
we can always set 𝛼 1 = ±1, 𝛽 1 = ±1 and then the model system (2.20) and (2.21) has
only three independent parameters: 𝛼 2 , 𝛽 2 , and 𝛾.
Phase-locked periodic solutions of the Liénard system (2.4) correspond to fixed
points of the model system (2.20) and (2.21), i.e. to the solutions of the equations
d𝜌/d𝜏 = d𝜃/d𝜏 = 0.
Note that the model system (2.20) and (2.21) is invariant under the transformation
(𝜌, 𝜃) → (𝜌, 𝜃 + 𝜋) (2.25)
and hence possesses the corresponding symmetry. Thus, if there is an equilib-
rium point at (𝜌0 , 𝜃 0 ), then there is also one at (𝜌0 , 𝜃 0 + 𝜋). In order to simplify
the following discussion, we consider only half of the system. If we state that
the model system contains an equilibrium point, then it actually contains two
equilibrium points, the other one being located at the symmetrical position under
the transformation (2.25).
The trivial solution (𝜌 = 0) is possible, but also a steady-state finite-amplitude

response (𝜌0 , 𝜃 0 ) exists and is given by
√
−(𝛼1 𝛽1 + 𝛼2 𝛽2 ) ± 𝛥
r= , r = 𝜌20 (2.26a)
1 + 𝛽22
( )
1 𝛼1 + 𝛽1 r
𝜃0 = arctan (2.26b)
2 𝛼2 + 𝛽2 r
where
( )
𝛥 = 𝛾 2 1 + 𝛽22 − (𝛼1 𝛽2 − 𝛼2 𝛽1 )2 (2.27)
The standard linearization method permits the computation of the eigenvalues of

the Jacobian matrix relative to each equilibrium point. They are
√ √
𝜆1 = 𝛼1 + 2𝛽1 r + 1 − 4𝛽2 r(𝛼2 + 𝛽2 r), 𝜆2 = 𝛼1 + 2𝛽1 r − 1 − 4𝛽2 r(𝛼2 + 𝛽2 r)
(2.28)
We can determine the parametric excitation–response curve because as the ampli-
tude of the parametric excitation is varied, the stability of the solution is determined
by Eq. (2.28). Moreover, we have checked the validity of our analysis by numeri-
cal integration of Eq. (2.4) with the standard fifth-order Runge–Kutta–Fehlberg
method.
In Figure 2.1, we show the results of the stability analysis for a typical case.
A jump phenomenon is clearly observable, because for increasing values of the
parametric excitation, we observe a discontinued transition from the trivial solution
to a finite steady-state periodic response. Moreover, we observe that this stable
periodic solution comes from a saddle–node fold (or cyclic fold) bifurcation. A fold
bifurcation corresponds to a vertical tangency in the parametric excitation–response
space, where the derivative of the response with respect to the control parameter is
infinite.
If f < f A , only the stable trivial solution is possible. When f A < f < f B , there are
three possible solutions: the stable trivial solution and two nontrivial solutions, the
large of which is stable. When f = f B , a subcritical pitchfork bifurcation occurs and
then if f > f B , there are two possible solutions: the unstable trivial solution and a
nontrivial solution, which is stable. We see that the parametric resonance is not
excited when f < f A , while for f A < f < f B , we observe the subcritical instability as
the parametric resonance may or may not be excited, depending on the initial con-
ditions. Finally, it is always excited if f > f B .
If we want to observe the jump phenomenon, we must vary f from a point below
f A and the response will be trivial until f reaches f B . A small increase in f beyond
f B causes an upward jump from zero to point B. As f increases further, B continues
to increase slowly toward point C. Obviously, if one starts at point C and gradually
decreases the parametric excitation, the response will slowly decrease until point A
is reached. Finally, a further decrease determines a downward jump to zero and the
response becomes trivial.
0.2
0.18
C
0.16
0.14
0.12
B
0.1
0.08 A
0.06
0.04
0.02
0 fA 0.02 fB 0.04 0.06 0.08 f 0.1
Figure 2.1 Response–parametric excitation curve, i.e. amplitude of the response (𝜌) as
function of the excitation (f ). Solid lines stand for stable and dashed lines for unstable
solutions, while boxes correspond to the numerical exact solution.
The frequency–response curve is

√
f2
𝜎 = 𝛽2 r ± − 1 − r 2 − 2𝛼1 𝛽1 r (2.29)
4𝛺2
Equation (2.29) presents a threshold level f c for the activation of the parametric
resonance, given by
√
fc = 2𝛺 (1 + r 2 + 2𝛼1 𝛽1 r) (2.30)
In Figure 2.2, we show a typical frequency–response curve. Depending on the

value of 𝜎, one, two, or three solutions are possible. If 𝜎 < 𝜎 A , only the trivial solution
0.2
ρ
0.18
0.16
0.14
0.12
0.1
0.08
0.06
0.04
0.02
–0.1 –0.05 σA 0 σB 0.05 σC σ 0.1
Figure 2.2 Frequency–response curve, i.e. the response (𝜌) as function of the detuning (𝜎).
Solid lines stand for stable and dashed lines for unstable solutions.
is possible, which is stable. If 𝜎 A < 𝜎 < 𝜎 B , there are two possible solutions: a nontriv-
ial solution, which is stable, and a trivial solution, which is unstable. If 𝜎 > 𝜎 B , there
are three possible solutions, the trivial solution, which is stable; and two nontrivial
solutions, the larger of which is stable. The subcritical instability appears because,
in the latter case, the response depends on the initial conditions and the system
tends asymptotically to the trivial solution or to the steady-state periodic response.
If 𝜎 > 𝜎 C , only the stable trivial solution is possible.
In order to verify if the AP method is a valid tool for approximate solutions of
nonlinear oscillators with small dissipation coefficients, we compare the analytic
approximate solution with the numerical integration in the following two examples.
We have chosen the following set of parameters
f2 = 1.0, f3 = −0.11, g0 = −0.01, g1 = 1.0, g2 = −1.0, f = 0.01, 𝜎 = 0.01 (2.31)
2.3 Global Analysis of the Model System 27
Y
0.2
0.1
–0.2 –0.1 0.1 0.2 X
–0.1
–0.2
Figure 2.3 Phase-space diagram (X(t), Y(t)) with Y(t) = X(t) ̇ with f 2 = 1.0, f 3 = −0.11,
g0 = −0.01, g1 = 1.0, g2 = −1.0, s = 0.01, and f = 0.01. Circles are the numerical solution, and
crosses represent the approximate solution.
We expect the appearance of a stable limit cycle with amplitude 𝜌0 = 0.085 (see
̇
Eq. (2.25)). In Figure 2.3, we show in the phase space (X(t), Y (t) = X(t)), a compar-
ison between the analytic approximate solution (2.22) and the numerical solution:
circles represent the numerical solution and crosses represent the approximate solu-
tion. As the solution repeats itself one cycle after another, we have represented only
a cycle. The agreement of the results appears to be excellent, because the maxi-
mum difference is 3 × 10−3 and the medium difference is 1 × 10−3 , i.e. of order e3
as expected.
In Figure 2.4, we have increased the parametric excitation by an order of mag-
nitude (f = 0.1). According to the model system (2.20) and (2.21) and Eq. (2.25), in
this case, we have a stable limit cycle with amplitude 𝜌0 = 0.165. The maximum dif-
ference between the approximate solution (2.22) and the numerical solution is now
0.04 and the medium difference is 0.02, i.e. of order e3 as expected.
2.3 Global Analysis of the Model System
In this section, we perform a global analysis of the model system (2.20) and (2.21)
and determine several conditions for the existence of limit cycles. They correspond
to two-period quasiperiodic solutions of the Liénard system (2.4).
Another random document with
no related content on Scribd:
TOM.
I now come to the very prince of pets, the one of all I ever had the
most noble and most dear,—Tom, a Newfoundland setter, the
favorite dog of my brother Albert. He has been a member of our
family for five or six years past. We brought him from the city to our
pleasant village home in Pennsylvania, where we now live.
Tom is a dog of extraordinary beauty, sagacity, and good feeling. He
is very large, and, with the exception of his feet and breast, jet black,
with a thick coat of fine hair, which lies in short curls, glossy and
silken. He has a well-formed head, and a handsome, dark eye, full of
kindness and intelligence. His limbs are small, and his feet
particularly delicate. He is, I am sorry to say, rather indolent in his
habits, always prefers to take a carriage to the hunting-ground, when
he goes sporting with his master, and he sleeps rather too soundly at
night to be a good watch-dog. We make him useful in various ways,
however, such as carrying baskets and bundles, and sometimes we
send him to the post-office with and for letters and papers. These he
always takes the most faithful care of, never allowing any one to look
at them on the way. He is a remarkably gentlemanly dog in his
manner, never making free with people, or seeming too fond at first
sight; but if you speak to him pleasantly, he will offer you a friendly
paw in a quiet way, and seem happy to make your acquaintance. He
never fawns, nor whines, nor skulks about, but is dignified, easy, and
perfectly at home in polite society. He is a sad aristocrat, treats all
well-dressed comers most courteously, but with shabby people he
will have nothing to do. Tom knows how to take and carry on a joke. I
recollect one evening, when we had visitors, and he was in the
parlour, I put on him a gay-colored sack of my own, and a large
gypsy hat, which I tied under his throat. Instead of looking ashamed
and trying to get these off, as most dogs would have done, he
crossed the room and sprang on to the sofa, where he sat upright,
looking very wise and grave, like some old colored woman at church.
The illustrious General Tom Thumb once travelled with my brother
and this dog, and, falling very much in love with his namesake,
offered any price for him. Of course, my brother would not think for a
moment of selling his faithful friend, and even had he felt differently, I
doubt very much whether Tom, who had been used to looking up to
full-grown men, would have shown much obedience or respect, for
such a funny little fellow as the General. It was amusing to observe
the dog’s manner toward his small, new acquaintance. He was kind
and condescending, though he sometimes seemed to think that the
General was a little too much inclined to take liberties with his
superiors in age and size,—rather more forward and familiar than
was quite becoming in a child.
Two or three years ago, Tom was the beloved playfellow of my
brother Frederic’s youngest daughter,—our little Jane. She always
seemed to me like a fairy-child, she was so small and delicate, with
such bright golden curls falling about her face,—the sweetest face in
the world. It was beautiful to see her at play with that great, black
dog, who was very tender with her, for he seemed to know that she
was not strong. One evening she left her play earlier than usual, and
went and laid her head in her mother’s lap, and said, “Little Jane is
tired.” That night she sickened, and in a few, a very few days she
died. When she was hid away in the grave, we grieved deeply that
we should see her face no more, but we had joy to know that it
would never be pale with sickness in that heavenly home to which
she had gone; and though we miss her still, we have great
happiness in the thought that she will never be “tired” any more, for
we believe her to be resting on the bosom of the Lord Jesus.
One day last spring, I remember, her mother gave me a bunch of
violets, saying, “They are from the grave of little Jane.” I suppose
they were like all other blue violets, but I thought then I had never
seen any so beautiful. It seemed to me that the sweet looks of the
child were blooming out of the flowers which had sprung up over the
place where we had laid her.
Tom seems much attached to all our family, but most devotedly so to
my brother Albert. They two have hunted very much together, and
seem equally fond of the sport. If Tom sees his master with his
hunting-dress on, and his fowling-piece in hand, he is half beside
himself with joy. But when he returns from the hunt, spent and weary,
he always comes to me to be fed and petted.
You will remember that years have passed by since this brother and I
were schoolmates and playmates together. He is now a fine young
man, while I am a full-grown woman, who have seen the world I
used to think so grand and glorious, and found it—no better than it
should be. But of my brother. He is our youngest, you know, and so
has never outgrown that peculiar fondness, that dear love, we
always give to “the baby.” While I have been writing these histories,
and recalling in almost every scene the playmate of my childhood, I
can only see him as a boy,—a little black-eyed, rosy-cheeked boy; it
is very difficult to think of him as a man, making his own way bravely
in the world. Last spring we observed that dear Albert’s bright face
had become very thoughtful and serious; we knew that something
was weighing on his mind, and finally it came out. He was about to
leave us all for a long time, it might be for ever; he was going to
California! We were very unhappy to hear this, but, as it was on
some accounts the best thing that my brother could do, we finally
consented, and all went to work as cheerfully as we could to help
him off.
It was a bright May morning when he left, but it seemed to us that
there never was a darker or sadder day. The dear fellow kept up
good courage till it came to the parting; then his heart seemed to
melt and flow out in his tears, fast dropping on the brows and necks
of his mother and sisters, as he held them for the last time to his
heaving breast. But I will not dwell on this parting, for my own eyes
grow so dim I cannot well see to write.
I remember that poor Tom seemed greatly troubled that morning; he
knew that something sad was happening, and looked anxiously in
our faces, as though he would ask what it was; and when my brother
patted him on the head, bade him good by, and passed out of the
gate, forbidding him to follow, the faithful creature whined sadly, and
looked after him wistfully, till he was out of sight.
After Albert had been gone about an hour, I remember that I went up
into his room, and sat down in his favorite seat, by the window. O,
how still and lonely and mournful it seemed there! Near me hung my
brother’s fencing-sword and mask, which he had used only the day
before,—on the floor lay the game-bag, which he had always worn in
hunting, and which he had flung out of his trunk, not having room for
it. This brought my merry brother before me more clearly than any
thing else. I took it up and held it a long time, mourning at heart, but I
could not weep. Suddenly I heard a low whine in the hall, and Tom
stole softly into the room. He came to me and laid his head in my
lap; but when he saw the game-bag there, he set up a most mournful
cry. Then I flung my arms about him, bowed my head down against
his neck, and burst into tears. I forgot that he was a poor dumb
brute, and only remembered that he loved my brother, and my
brother loved him, and that he mourned with me in my sorrow. After
this, it was very affecting to see Tom go every day, for a long while,
to the gate, out of which he had seen his master pass for the last
time, and then stand and look up the street, crying like a grieved
child.
As you will readily believe, Tom is now dearer than ever to us all; we
cannot see him without a sweet, sad thought of that beloved one so
far away. I am not now at home, but I never hear from there without
hearing of the welfare of the noble dog which my brother, in going,
bestowed upon me.
SUPPLEMENTARY STORIES.
It is twenty years since the first part of this little volume was
published. The dear children for whom those simple stories of my
childhood were told are men and women now, and wonderful
changes have taken place in all our lives and in all the world. But in
growing old I have not lost any thing of my old love of pets; and I
hope that my little readers of this time will understand and share that
feeling. I hope that you, dear boys and girls, look on all innocent
dumb creatures about you as friends, and have not only a kindly
interest in them, but respect them for all that is lovely and wonderful
in their brief existences, and as objects of the unceasing care and
tenderness of our Father in heaven. Every smallest creature that
lives represents a thought of God,—was born out of his great, deep,
infinite life.
I hope you especially like to hear about dogs and cats, birds and
chickens, for it is of them that I have a few new stories to relate, as
true as they are amusing or marvellous.
FIDO THE BRAVE.
First I must relate the somewhat tragical history of a certain little

shaggy brown-and-white spaniel belonging to some friends of ours in
the country. He was a stray dog, and came to them in a very forlorn
condition, and had evidently been vagabondizing about in the fields
and woods for some days, for he was ravenously hungry, and his
long hair was dirty, and stuck full of straws, briers, and burrs, till he
bristled like a hedgehog. The first thing that the kind lady did, after
feeding him, was to put him into a warm bath. Then she set herself
to work to rid him of his encumbrances,—sticks, straws, briers, and
burrs. It was a long time before she got down to the dog; but when at
last she laid down scissors, scrubbing-brush, and comb, and
deposited her poor protégé on the floor, he was a good deal
diminished in size, but looked really handsome, and very bright,
quaint, and droll.
He took at once to his new home, and soon became a great pet,
showing himself to be grateful, affectionate, and full of cleverness,
fun, and fire. His pluck was beyond all question. Though not
quarrelsome, he would, when in the least degree put upon, fight any
dog in the neighborhood, whatever his size and breed, and he
generally came off victorious. But he was altogether too rash and
venturesome, given to worrying cows, horses, hogs, and old
stragglers; rushing into all sorts of danger, and coming out, when he
did come out, and was not brought out, with his little eyes dancing
and his bushy tail in air, as though enjoying the risk of the thing, and
the terror of his kind mistress.
Among other sportive tricks was a way he had of running before the
locomotive when the train was coming in or going out of the station,
near by the house of my friends. Nearly every day he could be seen
frisking about it, dancing frantically up and down before it, and
barking valorously. He really seemed to take a malicious satisfaction
in defying and insulting that rumbling, puffing, snorting monster, that,
big as it was, ran away from him as fast as possible.
“The pitcher goes often to the well, but is broken at last.”
One fatal day the little spaniel miscalculated the speed of his big
enemy, and failed to get out of the way in time. He was all off the
track but one hind leg, when he was struck by the locomotive and
knocked into a ditch,—that one hind leg being pretty badly mashed,
you may believe. The poor little fellow set up a great outcry, but the
unfeeling engineer never stopped the train to attend to him, and the
railroad folks kept the accident out of the papers. Fido made his way
home all alone, dragging his mashed leg behind him. Though greatly
shocked, his mistress did not scold him, but sent for a surgeon, who,
after a careful examination, and consulting his books, decided that
an amputation was necessary. Then the good, brave lady held her
poor, dear pet on her lap while the dreadful operation was
performed. She asked a gentleman of the family to hold him, but he
had not the nerve. After the stump had been skilfully dressed, the
little dog evidently felt better, soon ceased to bemoan his loss, and
took kindly to a light supper. He rested well that night, and in the
morning the doctor pronounced him better. His kind mistress nursed
him faithfully till he was restored to perfect health. He never seemed
to fret about his maimed condition, but hopped around on three legs
as merry and active as ever. It was observed, however, that he gave
a wide berth to railway trains, and howled whenever he heard the
whistle of the engine, ever after. Still the fight wasn’t out of him. He
was as jealous of his honor and as fiery and plucky as before his
disaster.
One afternoon, while taking a quiet three-legged stroll some distance
away from home, he encountered on the highway a big, surly bull-
dog, who presumed on the spaniel’s diminutive size and crippled
condition to insult him and rail at him. Brave Fido dashed at once at
the ugly bully’s throat, and bit and hung on in the most furious and
desperate way. It was a gallant fight he made, and it did seem for a
while as though he must come off victorious, like David after his
engagement with Goliah. But at last the infuriated bull-dog tore
himself free, and then proceeded to make mince-meat of the poor
spaniel. He tore his ears half off, and his eyes half out, and mangled
his head generally, till it was disfigured to the last degree. Then he
bit and chewed the left, the only left hind leg, till one might say that
he was next to a locomotive and a whole train of cars at the
mangling business. At this desperate stage of the combat a woman
came out of a farm-house near by, drove the bull-dog away with a
poker, and took up poor Fido. As he had become insensible, she
thought him dead, and flung him down in a fence corner, out of the
way of travel, and there left him, meaning, let us hope, to have him
decently buried in the morning. But Fido was not yet ready to give up
this life. The cool evening dew revived him; brought him to his
senses, in part at least. He could not yet see, but, guided by some
mysterious instinct, he made his sure way, dragging himself by his
fore legs, which were only two you know, across the fields to his
home. His mistress was awakened in the night by hearing him
scratching and whining at the door, and made haste to arise and
take in the poor crippled, blinded, bleeding creature, who laid himself
panting and moaning at her feet. I hope I need not tell you that she
did not give him up. She prepared a soft bed for him in an old
basket, washed and dressed his wounds, and though every body,
especially the doctor, said he must die, that he was as good as dead
then, she was sure she could fetch him round, and she did fetch him
round amazingly.
But alas! Fido’s troubles were not over, even when he got so that he
could hobble about on his three legs, and see tolerably well; for one
cold morning, as he lay curled up in his basket near the kitchen
stove, he was, I grieve to say, terribly scalded by a careless cook,
who spilled a kettle of hot water over him. Even then his mistress
refused to give him up to die, but dressed his burns with sweet oil, or
applied a pain-killer, or “Dalley’s Salve,” and administered Mrs.
Winslow’s Soothing Syrup, perhaps,—anyhow she nursed him so
skilfully and faithfully that she fetched him round again. He is no
beauty nowadays, but alive, and alive like to be. It is my opinion that,
like the great Napoleon, that dog bears a charmed life.
CAT TALES.
FAITHFUL GRIMALKIN.
Many years ago, when my parents lived in old Connecticut, my
mother had a pet cat, a pretty graceful creature, frisky and arch and
gay, though clad in sober gray. She was a favorite with all the large
household, but especially attached herself to my mother, following
her about everywhere,—“up stairs, down stairs, and in my lady’s
chamber,” accompanying her in her walks, hiding behind every bush,
and prancing out upon her in a surprising, not to say startling,
manner.
At last she grew out of kittenhood, laid aside, in a measure, kittenish
things, and became the happiest, fondest, proudest feline mamma
ever beheld. She caressed and gloated over her little, blind, toddling,
mewing, miniature tigers in a perfect ecstasy of maternal delight.
Just at this interesting period of pussy’s life our family moved from
the old place to a house in the country, about a mile away. My
mother was ill, and was carried very carefully on a bed from one
sick-room to another. In the hurry, trouble, and confusion of that time,
poor pussy, who lodged with her family in an attic, was quite
forgotten. But early in the morning of the first day in the new house,
—a pleasant summer morning, when all the doors and windows
were open,—as my mother lay on her bed, in a parlor on the first
floor, she saw her cat walk into the hall and look eagerly around. The
moment the faithful creature caught sight of her beloved mistress,
she came bounding into the room, across it, and on to the bed,
where she purred and mewed in a delighted, yet reproachful way,
quite hysterical, licking my mother’s hand and rubbing up against her
cheek in a manner that said more plainly than words, “Ah! my dear
madam, didst thou think to leave thy faithful Grimalkin behind?
Where thou goest, I will go.”
She was taken into the kitchen and treated to a cup of new milk; but
after a few moments given to rest and refreshment she disappeared.
Yet she went only to come again in the course of an hour, lugging
one of her kittens, which she deposited on the bed, commended to
my mother’s care, and straightway departed. In an almost incredibly
short time she came bounding in with a second kitten. She continued
her journeys till the whole litter had been safely transported, over hill
and dale, ditches and stone-walls, through perils of unfriendly dogs
and mischievous boys, and the family flitting was complete.
After this, our noble puss was loved and respected more than ever.
She dwelt long in the land, and her kits grew up, I believe, to be
worthy of such a mother.
OBEDIENT THOMAS.
Now I want to give you an instance of filial respect and submission
in a young cat. When we first came to Washington, nearly two years
ago, I took to petting a handsome cat belonging to the relatives with
whom we then lived. I fed and caressed her, and she became very
fond of me, always running to meet me when I entered the garden
which she haunted, or the barn in which she lodged. She was rather
wild in her ways, and so stole a nest, in which she finally hid away
some kittens, that she afterwards reared to be wilder than herself.
These somehow disappeared, all but one, which, when he was
about half grown, I undertook to tame. It was a difficult, tedious job;
but I persevered, and at last found him a more affectionate, docile
pet than ever his mother had been. She had seemed fond of him in
his wild, unregenerate days, but as soon as he became
domesticated, and I began to show a partiality for him, she grew very
severe with him, scratching his face and boxing his ears whenever
she saw me caressing him. I soon noticed that when she was near
he was shy, pretending not to be on intimate terms with me; while, if
she was out of the way, I had only to call his name, to have him
come galloping up from the furthest part of the long garden, to rub
against me, to lick my hand, and show every feline fondness and
delight. Now we live at another house, and I seldom see my pets,
mother and son; but they are loving and constant still, proving that
the poet Coleridge didn’t know every thing when he talked about “the
little short memories” of cats.
Master Thomas has grown large and strong, and is accounted a
gallant young fellow by all the young pussies in the neighborhood.
But while toward cats of his own sex he is fierce and combative, he
is just as meek and deferential to his mother as he was in his tender
kittenhood. The other day I encountered him in the old garden, and
was surprised to find how stalwart he had become. I stooped to
caress him, and he seemed as susceptible to gentle overtures as
ever, arched his back, switched his tail, and purred rapturously.
Suddenly the mother cat stole out from behind a tree, and confronted
us. “Good morning, madam,” I said, for I always talk to cats and dogs
just as I talk to other people. “You have a fine son here; a handsome
young fellow, that favors you, I think.” But she wasn’t to be softened
by the compliment. She walked straight up to him, and boxed him
first on one ear and then on the other, quite in the old motherly way.
As for him he never thought of resenting the old lady’s act, or
opposing her will, but drooped his lordly tail, and hastily retreated.
Now that is what I call good family discipline.
This city of Washington is a place where the wits of people are
sharpened, if anywhere, and perhaps even cats and dogs become
uncommonly clever and knowing here. Only yesterday I was told of a
Washington cat which had just been found out in a wonderful trick.
Observing that, when the door-bell rang, the one servant of the
household was obliged to leave the kitchen, she managed to slyly
ring the bell, by jumping up against the wire, and invariably, when
her enemy, the cook, went to the door, she would slip into the
kitchen, and help herself to whatever tempting article of food was
within reach. At last some one watched, and caught her at her secret
“wire-pulling.” Poor puss retired with a drooping tail and a most
dejected aspect, evidently realizing that the game was up.
Another cat I know of was of so amiable and benevolent a
disposition that she actually adopted into her own circle of infant kits
a poor, forlorn little foundling of a rat. As her nursling he grew and
thrived, seeming quite as tame as the others; and when a
mischievous boy set a rat-terrier on him, and so finished him, cat and
kittens really seemed to mourn for their foster son and brother.
KATRINA AND KATINKA.

Once on a time—no matter when—in a certain beautiful city—no
matter where—there lived two lovely twin sisters, with the brightest
eyes, and the cunningest little roly-poly figures, and the slenderest
ears with the softest pink satin lining, and the spryest motions
imaginable. They were brunettes in complexion, with white breasts
and tail-tips, and they were kittens. Katrina and Katinka were their
names, if I remember rightly,—maybe I don’t, but anyhow they might
have had those names, which, to my thinking, are very pretty and
appropriate for kittens.
Well, these same twin pussies were singularly fond of each other,
and more singularly good to each other. They never called names, or
scratched, or spat in each other’s pretty faces, or pulled each other’s
little smellers, or quarrelled over their meals. They were so
marvellously alike that it was always difficult to tell them apart; and
when they slept, as they always did, hugged close in each other’s
arms, you couldn’t have told to save you where one kitten left off and
the other kitten began.
They not only slept, ate, and played together, but, as they grew
older, took their strolls for health and recreation and their mouse-
hunts in the same close and loving companionship. They were very
curious and wide-awake little bodies, and liked to see all they could
of the great, busy world; so every pleasant afternoon, when there
was much driving and walking up and down the fine street on which
they lived, they could be seen strolling down the long walk to the
gate, always exactly side by side,—“neck and neck,” as the horse
people say,—as even in their pace, and as perfectly matched in their
action, as ever were a pair of trained ponies in Hyde Park. Reaching
the gate, they would pause and stand quite still for a half-hour or so,
gravely gazing through the palings at the passers,—pedestrians,
equestrians, and drivers of fast horses,—like a pair of dear little
brigadiers reviewing their brigades marching by. Then, with the air of
having discharged a public duty to the entire satisfaction of the
community, they would wheel exactly together, and again, precisely
neck and neck and tail and tail, trot gently homeward.
So they lived on, in and for each other, almost as much united as if
they had been a pair of small feline female Siamese twins, amiable,
loving, and virtuous, and grew in knowledge and stature up to a
comely young cathood. At last it happened that a very interesting
event occurred to the twin sisters at precisely the same time,—they
became happy mothers, were blessed with three or four fine kittens
apiece. But alas! before the little strangers had got fairly to feel their
legs, before they had got their eyes open, all save one mysteriously
disappeared from each nest. It was one fatal morning when the twin
sisters had slipped out of their happy attic apartment for a little air,—
to take their “Constitutional” in a trot down the long gravel walk to
see how the world would look to them now they were mothers,—that
this kit-napping occurred. When they returned to their families, they
found them strangely thinned out; but they were mothers for all that,
and did not seem to fret much, or abate their maternal pride a jot.
You see the ruling power in the human household in which they were
domesticated, and who was to them as a providence, had ordered a
little hydropathy for their poor, feeble, sprawling, blind darlings,—
beginning with what is called in water-cures “the heroic treatment,” a
cold plunge; and it didn’t agree with them,—it never does with any
but the healthy and hardy patients,—so it was they never came
back. But under the blue waves they sleep well, though never a mew
or a purr comes bubbling up to the surface to tell the spot where they
lie on beds of tangled sea-grass. “Requies-cat in pace,” as old
tombstones say.
The next mournful event in this true family history was the untimely
death of Katrina’s one darling. This had proved to be but a frail
flower of kittenhood; very pretty she was,—“too sweet to live,” people
said. Her constitution was defective, her nervous system was
extremely delicate. Before she was a week old she had something
alarmingly like a fit of catalepsy. Suddenly, while imbibing
nourishment, with her fond mother purring over her, and two or three
children looking on in smiling sympathy, she gave a piteous wild
mew, rolled over on her back, and stuck up her four little legs and
laid out her little tail stiff as a poker! On the ninth day of her little life
she opened her blinking blue eyes on this great wonderful world, in
which she had as good a right to be as you or I; but she didn’t seem
to like the looks of things, for she soon closed those small eyes
again, and never opened them more. Life was evidently too hard a
conundrum for her poor, weak little brain, and she gave it up.
Of course Katrina was greatly afflicted, but she did not abandon
herself utterly to grief. Had not her sister a kitten left? and had not
they two always had every thing in common? So as soon as the
sympathetic children had buried her dead out of her sight under a
lilac-bush, she went straightway to Katinka, and, with her full
consent, began to divide with her the duties and joys of maternity. All
three just cuddled down together in one nest; from mamma or auntie
Master Catkin took nourishment, just as it suited his whim or
convenience, and, as you might suppose, he grew and thrived
astonishingly. So equal and perfect was this partnership in the kitten,
that it was impossible for a stranger to tell which of the two cats was
the real mother. One day all three were brought down to the parlor to
amuse some visitors. Both mammas seemed equally nervous about
having the baby kitten handled, and presently one of them caught it
by the neck,—the cat’s usual, immemorial way of transporting her
young,—and started with it for the attic; when, to the surprise and
immense amusement of all present, the other caught hold of the tail,
and so the two bore it away in triumph.
After this I am afraid the children gave the little kitten rather more
travelling than he liked. It was such fun to see the two anxious cats
following him, mewing, and at the first chance catching him up, and
lugging him home in that absurd manner. Generally the real certain
true mother seized on the head, but sometimes she was
magnanimous enough to yield the post of honor to the aunt, and take
to the tail herself.
So things went on for a few weeks, and then there happened to this
estimable cat-family another sad event,—for this is a tragedy I am
writing, though you may not have suspected it,—Katinka died! What
of has never yet been decided; physicians differed about it, and the
coroner could not make it out. But this much is certain, Katinka died.
The grief of Katrina was and is very affecting to behold. She mopes,
she mews, and her slender tail, which she used to carry erect with
such a jaunty air, droops dolefully. She takes no longer the
“Constitutional” trot down the walk to the front gate. Life seems to
have grown dull and wearisome to her, and the pleasures of mouse-
hunting and tree-climbing appear to have lost their zest. If she
remembers at all the halcyon period when much of her precious time
was spent in a dizzy round of gayety, in a swift pursuit of a ball of
cotton, or a futile pursuit of her own tail, it is in sad wonder that she
could ever have been so merry and thoughtless. She grows thin,
neglects her toilet, and often refuses food; but when the children
offer her catnip, she turns languidly away. If she were acquainted
with Shakespeare, she would doubtless say,—“Canst thou minister
to a mind diseased?” “Throw physic to Bose and Jowler,—I’ll none of
it!”
Friendly cat-neighbors call in occasionally, but they cannot console
her. All the petting of the household fails thus far to make her cheery
and playful as once she was. She is fed on the very “milk of human
kindness,” but grief has licked the cream off.
She seems to find her only consolation in her care and affection for
the motherless catkin, and in his fondness for her. I am sorry to say
that he does not show a very deep sense of his loss; perhaps he is
too young to realize it. His good aunt seems sufficient for all his
needs, and he thrives finely, is fat and jolly, and full of all kittenish
pranks and mischievous tricks. Poor Katrina will have a time with
him, I fear, as he is sadly petted and indulged. Such a lazy rascal as
he is too,—don’t earn the salt of his porridge, that is, if he took it
salted,—and, though quite old enough to “go on the war path,” has
never yet killed his mouse, or brought home a rat’s scalp, or a
ground-squirrel’s brush, or as much as a feather from a tomtit’s wing.
Ah! of all the darlings in the world, an aunty’s darling is the likeliest to
be spoiled.
This is all I know about this curious cat-family. I hope, dear children,
that my true story may not sadden you, for I really wish you, one and
all, the merriest of merry Christmases, and the happiest of happy
New Years.
All I can say in the way of a moral to my little story is: How beautiful
is love! even when shown in the fortunes and sorrows of cats and
kittens, how beautiful is love!
FEATHERED PETS.
OUR COUSINS THE PARROTS.

These strangely interesting birds, according to natural history,
belong to the second bird family, the Psittacidæ. I never knew how
many wonderful and splendid varieties this family contained until I
saw living varieties of all, or nearly all, in the known world, in the
Zoölogical Gardens of London, where they are kept in a great
gallery,—a beautiful parrot paradise, all by themselves. They were a
wonder to behold, but a perfect astonishment to listen to. The
confusion of tongues was something almost distracting. The Tower
of Babel, in its talkingest day, never approached it, I am sure. A large
sewing-circle of elderly ladies might come nearer the mark. The
colors of their plumage I have no words to describe. They fill my
memory with tropic splendors whenever I think of them, to this day.
’Tis strange that but one species of parrots was known to the ancient
Greeks and Romans,—the Parakeet of India,—at least up to the time
of Nero. That gentle prince, with his amiable love of pets, is said to
have sent emissaries far up the Nile to collect new varieties for the
gratification of his royal whim and dainty appetite; for, when the poor
little captives ceased to amuse him by their conversational powers,
he ate them. I hope they lay hard on his stomach, and made him talk
in his sleep!
The early Portuguese navigators found parrots at the Cape of Good
Hope, and at other points on the African coast; and the very first
creatures that welcomed Columbus to the isles of the New World
were Parakeets. The Macaws of South America are very handsome
birds, but not remarkably tractable or agreeable. They are fond of old
friends, but are fierce to strangers, and have a singular dislike to
children. The gray and scarlet parrot, called the Yaco, is a charming
bird for a pet. It is clever and docile, and learns readily to talk,
preferring to imitate the voices of children. The Cockatoos of New
Guinea are very pretty and graceful pets. They do not like to be
caged, but may be safely allowed to have the range of the premises,
as they will immediately come when called; thus setting an excellent
example to rebellious children. The green parrot, most common in
this country, is a native of Africa.
Dear old Dr. Goldsmith, whose Natural History is all out of fashion
now, except with us old folks, tells some amusing stories about
parrots. Among these is an anecdote of a famous fellow, belonging
to King Henry the Seventh, Queen Elizabeth’s grandfather. This bird,
sitting on his perch in the palace-yard at Westminster, used to hear
the talk of gentlemen who came to the river to take boats. And one
day, while overlooking the busy traffic of the Thames, he fell from a
tree into the water; and while there, floating helplessly, he cried: “A
boat! twenty pounds for a boat!” A waterman rescued him, and took
him to the king, demanding his twenty pounds. The king, who was
not remarkably generous, hesitated about giving so large a sum; but
finally agreed to leave the amount of the reward to the parrot. That
ungrateful fellow, who sat on his perch, still shaking the water from
his feathers, when appealed to, turned his head slyly on one side,
and said, “Give the knave a groat” (about fourpence). I hope,
children, you won’t doubt the truth of this story; it isn’t good to get
into sceptical habits of mind in early life.
For many years there lived in the porter’s lodge of the old
Pennsylvania Hospital a distinguished and venerable citizen,—a
parrot of rare cleverness and intelligence. This famous bird belonged
to the porter, and was one of many feathered pets, the chief favorite
and familiar. A remarkable affection and sympathy existed between
these two friends; yet I am sorry to say their relations were not
altogether pleasant and peaceful. Innumerable were their quarrels
and make-ups. The bird was very knowing, and almost
supernaturally gifted as a talker, especially, like some human orators,
in the language of railing and taunting. The old man, his master, had
one deplorable weakness,—he would occasionally drink too much
whiskey; so much that, getting quite beside himself, he would leave
his lodge and his innocent feathered family, and go off on a
desperate spree, which sometimes lasted for days. Now, Master
Paul Parrot thought this weakness, through which he suffered in
loneliness and neglect, very reprehensible and not to be winked at,
and when the fit of dissipation was coming on his master, it is said,
would remonstrate with him, in a friendly way, like a very Mentor.
When this proved in vain, and he saw the misguided old man leave
the lodge for some of his disreputable haunts, he would endeavor to
put a good face on the matter, would hop about on his perch in great
excitement, and call out to the other birds: “The old man has gone on
a spree!—on a spree! He won’t be back for a week! Let’s have a
time. Ha, ha!”
When the old porter came home, this naughty bird would be very apt
to mock and taunt him, calling out: “So you’ve come back,—have
you? O, how drunk you are! Now we’ll have a row.” And there always
was a row; for the indignant porter never failed to beat Mr. Paul, for
his impudence, soundly. Then the bird, seeking the dignified
retirement of the darkest corner of the lodge, sulked and muttered,
till, the old porter’s good-humor returning, he made friendly
overtures. The two were reconciled, and “every thing was lovely”
again.
At length the poor old porter died; and as his successor was no bird-
fancier the feathered family at the lodge was broken up and
dispersed. The clever parrot was kindly treated in a new home; but
he never seemed happy. He evidently missed his old master,—
missed his caresses and his scoldings. Or perhaps he found the
steady goings-on of a moral household too dull for his taste, for
when I went to see him, I found him as glum, stupid, and morose as
an old politician who had had his day. All he would say was, “O you
goose!”
There is another curious parrot in Philadelphia, in a store kept by a
maiden lady whose voice is so exceeding shrill and parrot-like that it
is difficult to tell when she leaves off talking and the parrot begins.
One day, as a customer was examining an article on the counter,
Miss Polly called out: “What are you doing with that? Put it down! put
it down!” The lady looked round very indignantly for the offender,
saying: “Well, ma’am, I must say you have a very impudent child.”
There is in the same city another parrot, who recites a verse of an
old song in a most distinct and triumphant manner:—
“O pretty Polly,
Don’t you cry,
For your true-love
Will come by and by.”
There is in Brooklyn, New York, a parrot that sings many of the
popular airs correctly, and with as much expression as many
fashionable singers give to them. This bird is singularly social and
affectionate, and has a horror of being alone. He will sometimes
awake in the middle of the night, and arouse the household by
crying: “O dear! I am all alone!—all alone! Somebody come to me!”
I have heard much of a clever parrot once kept by some relatives of
ours on an old place in a quiet little village. Mistress Polly had free
range of the house and yard, and throughout the town was as well
known as the oldest inhabitant. Through all the pleasant weather she
haunted the tall trees in front of the house, climbing to the highest
branches, and from there superintending the affairs of the
neighborhood, and making astronomical and meteorological
observations. In the spring and autumn she watched from these lofty
perches the flight of great flocks of pigeons and crows with intense
but decidedly unfriendly interest. She would scream and scold at
them in a most insolent and defiant manner, evidently criticising the
order of their march and all their manœuvres and evolutions, for all
the world like a newspaper editor finding fault with the conduct of
great armies. Doubtless she was astonished and disgusted to see
the great host sweep steadily on, following their leader, paying no
heed to her shrieking, railing, and evil prophecies. Yet she was never
so absorbed by her duties on the watch-tower that she failed to
come to her meals. These she took with the family, perched on the
back of a chair or the corner of the table. She was very fond of
coffee, and was always provided with a cup. She would take it up by
the handle with her claws, and drink from it without spilling a drop. A

Asymptotic Perturbation Methods For Nonlinear Differential Equations in Physics Attilio Maccari Full Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Asymptotic Perturbation Methods For Nonlinear Differential Equations in Physics Attilio Maccari Full Chapter

Uploaded by

Copyright:

Available Formats

Asymptotic Perturbation Methods: For

Nonlinear Differential Equations in

For Nonlinear Differential Equations in Physics

British Library Cataloguing-in-Publication Data

Bibliographic information published by

© 2023 WILEY-VCH GmbH, Boschstraße 12,

All rights reserved (including those of

Print ISBN: 978-3-527-41421-5

Typesetting Straive, Chennai, India

About the Author xi

1 The Asymptotic Perturbation Method for Nonlinear

2 The Asymptotic Perturbation Method for Remarkable

3 The Asymptotic Perturbation Method for Vibration Control

3.5 Suppression of the Two-period Quasiperiodic Motion 63

4 The Asymptotic Perturbation Method for Vibration Control by

5 The Asymptotic Perturbation Method for Nonlinear

6 The Asymptotic Perturbation Method for Dispersive Nonlinear

7 The Asymptotic Perturbation Method for Physics

7.4.7 Moving Breather-Like Structures 117

8 The Asymptotic Perturbation Model for Elementary Particle

9 The Asymptotic Perturbation Method for Rogue Waves 151

10 The Asymptotic Perturbation Method for Fractal and Chaotic

11 The Asymptotic Perturbation Method for Nonlinear

13 Conﬁnement and Asymptotic Freedom in a Purely Geometric

13.4 The Motion of a Light Ray Into a Hadron 207

14 The Asymptotic Perturbation Method for a Reverse

About the Author

Attilio Maccari is a free-lance researcher in nonlinear physics. He received his PhD

This textbook is devoted to nonlinear physics.

Nonlinear systems are paramount in engineering and science. Many perturbation

In Chapter 9, we try to explain the rogue waves appearance in nonlinear

The Asymptotic Perturbation Method for Nonlinear

X(t) = 2𝜌 cos(−𝜔t + 𝜃) (1.2)

if X(0) = X 0 for the displacement

then we easily get

X(t) = 𝜀(1+r) 𝛹0 + (𝜀𝛹1 exp(−i𝜔t) + 𝜀2 𝛹2 exp(−2𝜔t) + 𝜀3 𝛹3 exp(−3i𝜔t) + c.c. + h.o.t.)

̈ + X(t) + f2 X 2 (t) + f3 X 3 (t) = g0 X(t)

1.2 Nonlinear Dynamical Systems

We restrict our study to the following particular case of Eq. (1.21)

oscillation is subjected to a slight modulation, with an amplitude proportional to

1.3 The Approximate Solution

q = rational number (1.28)

the temporal rescaling

𝛹n (t, 𝜀) = c.c.(𝛹(−n) (t, 𝜀)) (1.30b)

X(t) = 𝜀r 𝛹0 (t; 𝜀) + 𝜀𝛹1 (t; 𝜀) exp(−it) + 𝜀2 𝛹2 (t; 𝜀) exp(−2it) + 𝜀3 𝛹3 (t; 𝜀) exp(−3it)o(𝜀4 )

Substituting the polar form

where h. o. t = higher order terms and

𝛹2(2) = (B1 + iB ̃2 )|𝛹 |2 𝛹 2 + 2 (f2 + ig1 )𝛹 𝛹 (2)

entiate Eq. (1.36), we have

1.4 Comparison with the Results of the Numerical

We expect the appearance of a stable limit cycle with amplitude 𝜌M E

1.5 External Excitation in Resonance with the Oscillator

–0.10 –0.05 0.05 0.10 X

–0.4 –0.2 0.2 0.4 X

𝛽2 → |𝛽1 |𝛽2 , 𝛼2 → |𝛼1 |𝛼2 (1.70)

we can always set 𝛼 1 = ± 1, 𝛽 1 = ± 1.

Every equilibrium point of (1.71a–c) corresponds to a periodic solution of the start-

We can now follow the stationary solutions of (1.71a–c) as the amplitude of

Figure 1.4 External force–response space. Circles are sources.

Figure 1.5 External force–response space. Rectangles are sinks.

𝜌(t) = 𝜌2 + (𝜌0 − 𝜌2 ) exp(−2𝛼1 t)

where 𝜌0 depends on the initial conditions.

–0.20 –0.10 0.10 0.20 X