Zeng He

Wen Jiang
Lin Wang

Solved
Problems in
Nonlinear
Oscillations
A Sourcebook for Scientists and
Engineers
Solved Problems in Nonlinear Oscillations
Zeng He · Wen Jiang · Lin Wang

Solved Problems
in Nonlinear Oscillations
A Sourcebook for Scientists and Engineers
Zeng He Wen Jiang
Department of Mechanics Department of Mechanics
Huazhong University of Science Huazhong University of Science
and Technology and Technology
Wuhan, China Wuhan, China

Lin Wang
Department of Mechanics
Huazhong University of Science
and Technology
Wuhan, China

ISBN 978-981-97-6112-8 ISBN 978-981-97-6113-5 (eBook)

https://doi.org/10.1007/978-981-97-6113-5

This work was supported by National Natural Science Foundation of China (12325201, 12072119).

© The Editor(s) (if applicable) and The Author(s) 2025. This book is an open access publication.

Open Access This book is licensed under the terms of the Creative Commons Attribution 4.0 International
License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribu-
tion and reproduction in any medium or format, as long as you give appropriate credit to the original
author(s) and the source, provide a link to the Creative Commons license and indicate if changes were
made.
The images or other third party material in this book are included in the book’s Creative Commons license,
unless indicated otherwise in a credit line to the material. If material is not included in the book’s Creative
Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted
use, you will need to obtain permission directly from the copyright holder.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore

If disposing of this product, please recycle the paper.

Preface

This handbook contains about 200 fully solved problems in analytical and numerical
methods for nonlinear oscillations. These comprise all the end-of-chapter problems
in Prof. Nayfeh and Prof. Mook’s famous book (Ali H. Nayfeh and Dean T. Mook.
Nonlinear Oscillations, Wiley-Interscience, 1979). Mathematical software is adopted
to make those solutions more accessible from a graphical point of view. This book
can be adopted as a supplement to course work study for graduates or senior under-
graduates. Since many exercise problems are adapted from scientific research papers,
this book also has a good reference value for scientists and engineers who work in
the area of nonlinear vibration.
We wish to express our appreciation to Jiabiao Yi and Yefeng Pu for their great help
in documentation editing and drafting, and to Springer Nature for the opportunity to
make available this supplement to Nonlinear Oscillations.

Wuhan, China Zeng He

December 2023 Wen Jiang
Lin Wang

v
Contents

1 Problem List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Chapter 2 Conservative Single-Degree-Of-Freedom Systems . . . . 1
1.2 Chapter 3 Nonconservative Single-Degree-Of-Freedom
Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Chapter 4 Forced Oscillations of Systems Having a Single
Degree of Freedom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4 Chapter 5 Parametrically Excited Systems . . . . . . . . . . . . . . . . . . . . 5
1.5 Chapter 6 Systems Having Finite Degrees of Freedom . . . . . . . . . . 7
1.6 Chapter 7 Continuous Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.7 Chapter 8 Traveling Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2 Conservative Single-Degree-of-Freedom Systems . . . . . . . . . . . . . . . . . . 11
2.1 Exercise 2.1 (Determine Solution Trajectories from Potential
Energy Curves of Conservative 1D Systems) . . . . . . . . . . . . . . . . . . 11
2.2 Exercise 2.2 (Solving Conservative Systems by the Method
of Multiple Scales or Straightforward Expansion Combined
with Reformulation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3 Exercise 2.3 (Show the System Transformation
and the Method of Multiple Scales) . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.4 Exercise 2.4 (Show the Method of Averaging) . . . . . . . . . . . . . . . . . 28
2.5 Exercise 2.5 (Show the Effect of a Change in the Equilibrium
Point of a Nonlinear Spring on the Frequency-Amplitude
Relationship) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.6 Exercise 2.6 (Rods Sliding on the Smooth Walls) . . . . . . . . . . . . . . 34
2.7 Exercise 2.7 (Period of the Osscillation Motion of a Single
Pendulum-Type System Along Separatrices) . . . . . . . . . . . . . . . . . . 35
2.8 Exercise 2.8 (Particle on a Rotating Parabola) . . . . . . . . . . . . . . . . . 36
2.9 Exercise 2.9 (A Single Pendulum Rotating Uniformly
Around a Plumb Axis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.10 Exercise 2.10 (A Single Pendulum Rotating Freely Around
a Plumb Axis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

vii
viii Contents

2.11 Exercise 2.11 (Nonlinear Oscillation of a Slide-Slider

System) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.12 Exercise 2.12 (Nonlinear Oscillation of a Rope-Block
System) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.13 Exercise 2.13 (Single Pendulum Attached with a
Rolling-Without-Slipping Circular Wheel I) . . . . . . . . . . . . . . . . . . . 53
2.14 Exercise 2.14 (Single Pendulum Attached with a
Rolling-Without-Slipping Circular Wheel II) . . . . . . . . . . . . . . . . . . 56
2.15 Exercise 2.15 (Single Pendulum Problem with Inelastic
Collision with an Inclined Wall) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
2.16 Exercise 2.16 (Simplified Model for Buckling Analysis
of Columns) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
2.17 Exercise 2.17 (Rods for Pure Rolling on a Fixed Cylindrical
Surface) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
2.18 Exercise 2.18 (Geometrically Nonlinear System Formed
by a Linear Spring and a Mass Block) . . . . . . . . . . . . . . . . . . . . . . . . 63
2.19 Exercise 2.19 (Solving Pure Cubic Nonlinear Systems
by Harmonic Balance Method) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
2.20 Exercise 2.20 (Solving Purely Fifth-Order Nonlinear
Systems by Harmonic Balance Method) . . . . . . . . . . . . . . . . . . . . . . 67
2.21 Exercise 2.21 (Solving Pure Cubic Nonlinear Systems
by Equivalent Linearization) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.22 Exercise 2.22 (Show Least Residual Value Method
and Galerkin Method for Solving Pure Cubic Nonlinear
Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2.23 Exercise 2.23 (Discuss the Possibility of Solving Pure Cubic
Nonlinear Systems by the Method of Multiple Scales) . . . . . . . . . . 70
2.24 Exercise 2.24 (Examine the Relationship Between the Period
and the Amplitude of Oscillation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.25 Exercise 2.25 (Comparison of Equivalent Linearization
Methods and the Method of Multiple Scales for Solving
Different Nonlinear Equations) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
2.26 Exercise 2.26 (Comparison of the Galerkin Method
with the the Method of Multiple Scales) . . . . . . . . . . . . . . . . . . . . . . 75
2.27 Exercise 2.27 (Equations Containing Second, Third
and Fourth Order Nonlinear Terms) . . . . . . . . . . . . . . . . . . . . . . . . . . 77
2.28 Exercise 2.28 (Solving Nonlinear Equations
with Segmentation Functions) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
2.29 Exercise 2.29 (Nonlinear Equations with a Single Arbitrary
Subpartial Nonlinear Term Only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
2.30 Exercise 2.30 (Higher Order Nonlinear Equations) . . . . . . . . . . . . . 84
2.31 Exercise 2.31 (Motion of the Disk Constrained by a Linear
Spring) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
2.32 Exercise 2.32 (Rolling Without Slip Rollers Restrained
by a Linear Spring) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Contents ix

2.33 Exercise 2.33 (Example of Higher-Order System) . . . . . . . . . . . . . . 89

3 Nonconservative Single-Degree-of-Freedom Systems . . . . . . . . . . . . . . . 95
3.1 Exercise 3.1 (Singularity Analyses I) . . . . . . . . . . . . . . . . . . . . . . . . . 95
3.2 Exercise 3.2 (Singularity Analyses II) . . . . . . . . . . . . . . . . . . . . . . . . 97
3.3 Exercise 3.3 (Singularity Analyses III) . . . . . . . . . . . . . . . . . . . . . . . 98
3.4 Exercise 3.4 (Conversion of Rayleigh’s Equation to Van Der
Pol’s Equation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
3.5 Exercise 3.5 (Plotting Phase Trajectories and Limit Cycle(S)
for the Van Der Pol Equation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
3.6 Exercise 3.6 (Plotting Phase Trajectories and Limit Cycle(S)
for a Given Equation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
3.7 Exercise 3.7 (Plotting Singularity and Phase Trajectories
for a Given Planar System) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
3.8 Exercise 3.8 (Plot the Trajectory of a Singularity for a Given
Planar System and Determine the Stability of the Singularity) . . . . 106
3.9 Exercise 3.9 (The Method of Multiple Scales for Solving
a Single Pendulum with a Coulomb Friction Torque) . . . . . . . . . . . 108
3.10 Exercise 3.10 (Singularities and Their Stability of a Linearly
Damped Pendulum Under a Constant Torque) . . . . . . . . . . . . . . . . . 111
3.11 Exercise 3.11 (Singularities and Their Stability for a Single
Pendulum with a Quadratic Damping) . . . . . . . . . . . . . . . . . . . . . . . . 114
3.12 Exercise 3.12 (Analysis of Rayleigh Equation Singularities,
Method of Averaging of Solution) . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
3.13 Exercise 3.13 (Singularity Analysis of the Restricted
Three-Body Exercise) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
3.14 Exercise 3.14 (Singularity Analysis of Amplitude
and Phase Equations Limiting the Resonance Response
of the Three-Body Problem) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
3.15 Exercise 3.15 (Modeling of a Single Pendulum
with a Damper Attached, Solving by the Method of Multiple
Scales) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
3.16 Exercise 3.16 (The Method of Multiple Scales for Solving
Van Der Pol Oscillators with Delay Amplitude Limiting) . . . . . . . . 127
3.17 Exercise 3.17 (Averaging to Solve for an Oscillator
with Both Coulomb and Viscous Damping) . . . . . . . . . . . . . . . . . . . 132
3.18 Exercise 3.18 (Averaging to Solve for Oscillators with Both
Coulomb and Square Damping) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
3.19 Exercise 3.19 (Averaging to Solve for Oscillators with Both
Viscous and Square Damping) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
3.20 Exercise 3.20 (Averaging to Solve for an Oscillator
with Both Viscous and Negative Coulomb Damping) . . . . . . . . . . . 140
3.21 Exercise 3.21 (Averaging to Solve for an Oscillator
with Both Squared Damping and Negative Coulomb
Damping) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
x Contents

3.22 Exercise 3.22 (Oscillators with Simultaneous Negative

Viscous Damping, Square Damping, and Cubic Nonlinearity) . . . . 143
3.23 Exercise 3.23 (Method of Averaging for Solving Oscillators
with Simultaneous Viscous, Quadratic and Cubic Damping
Nonlinearities) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
3.24 Exercise 3.24 (Averaging to Solve and Analyze Nonlinear
Vibrations of a Pendulum) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
3.25 Exercise 3.25 (Viscous Damping, Negative Stiffness,
and Cubic Nonlinear Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
3.26 Exercise 3.26 (Transformations of Variables and Three-Level
Number Expansion Methods for Solving a Cubic Nonlinear
System) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
3.27 Exercise 3.27 (Application of Elliptic Functions to Represent
Solutions of a Single Pendulum System with a General
Weak Damping Function) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
3.28 Exercise 3.28 (Masses Bounded in Definite Orbits,
Analyzing Singularities and Their Bifurcations) . . . . . . . . . . . . . . . . 161
3.29 Exercise 3.29 (Non-autonomous Systems with Slowly
Varying Functions) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Introduction to Elliptic Integrals and Jacobi Elliptic
Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
4 Forced Oscillations of Systems Having a Single Degree
of Freedom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
4.1 Exercise 4.1 (The Method of Multiple Scales for Analyzing
Primary Resonances of Coulomb-Damped Systems
with Cubic Nonlinearities) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
4.2 Exercise 4.2 (The Method of Multiple Scales for Analyzing
Primary Resonances of Cubic Nonlinear, Quadratic-Damped
Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
4.3 Exercise 4.3 (The Method of Multiple Scales for Analyzing
Primary Resonances of Linear Spring and Hysteresis
Damped Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
4.4 Exercise 4.4 (The Method of Multiple Scales for Analyzing
Primary Resonances of Linear Spring, Hysteresis
and Coulomb Damped Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
4.5 Exercise 4.5 (The Method of Multiple Scales for Analyzing
Primary Resonances of Linear Springs, Hysteresis
and Linearly Damped Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
4.6 Exercise 4.6 (The Method of Multiple Scales for Analyzing
Primary Resonances of Linear Spring, Hysteresis
and Quadratic Damped Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Contents xi

4.7 Exercise 4.7 (The Method of Multiple Scales for Analyzing

Primary Resonances of Quadraticd Spring Force
and Quadraticd Damped Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
4.8 Exercise 4.8 (Primary, Superharmonic, and Subharmonic
Resonances of Cubic Nonlinear and Cubic Damped Systems) . . . . 208
4.9 Exercise 4.9 (Method of Averaging for Analyzing Primary
Resonances of Alternating Damped Systems) . . . . . . . . . . . . . . . . . . 219
4.10 Exercise 4.10 (Primary, Superharmonic, and Subharmonic
Resonances of Quadratic and Cubic Nonlinear Systems) . . . . . . . . 224
4.11 Exercise 4.11 (Primary and Secondary Resonances of Five
Nonlinear, Linearly Damped Systems) . . . . . . . . . . . . . . . . . . . . . . . . 233
4.12 Exercise 4.12 (An Iterative Method for General Undamped
Nonlinear Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
4.13 Exercise 4.13 (The Method of Harmonic Balance to Analyze
Viscous Damping and Primary Resonances of Cubic
Nonlinear Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
4.14 Exercise 4.14 (The Method of Harmonic Balance to Analyze
the Primary Resonance of an Undamped Nonlinear System) . . . . . 240
4.15 Exercise 4.15 (Combination Resonance for Cubic Nonlinear
Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
4.16 Exercise 4.16 (The Response of Cubic Nonlinear Systems
with Primary Resonance and Non-resonance Excitation) . . . . . . . . 245
4.17 Exercise 4.17 (Primary and Combined Resonance for Cubic
Nonlinear Systems I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
4.18 Exercise 4.18 (Combined Primary, Subharmonic,
and Superharmonic Resonances of a Cubic Nonlinear
System) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
4.19 Exercise 4.19 (Primary and Combined Resonance for Cubic
Nonlinear Systems II) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
4.20 Exercise 4.20 (Primary and Combined Resonance of a Cubic
Nonlinear System) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
4.21 Exercise 4.21 (Primary Resonance of the System
with Coulomb Friction and Quadratic Nonlinearity) . . . . . . . . . . . . 263
4.22 Exercise 4.22 (Primary Resonance Problem for the System
with Quadratic Damp and Quadratic Nonlinear Terms) . . . . . . . . . . 267
4.23 Exercise 4.23 (Combined Resonance Problem for Quadratic
Nonlinear Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
4.24 Exercise 4.24 (Simultaneous Subharmonic
and Superharmonic Resonance of Quadratic Nonlinear
Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
4.25 Exercise 4.25 (Quadratic Nonlinear Systems with Both
Subharmonic and Combinatorial Resonances) . . . . . . . . . . . . . . . . . 278
4.26 Exercise 4.26 (Quadratic Nonlinear Systems with Both
Superharmonic and Combinatorial Resonances) . . . . . . . . . . . . . . . 280
xii Contents

4.27 Exercise 4.27 (Quadratic Nonlinear Systems

with Superharmonic, Subharmonic and Combinatorial
Resonances) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
4.28 Exercise 4.28 (Primary and Superharmonic Resonances
of Self-Excited Systems) Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
4.29 Exercise 4.29 (Primary and Superharmonic Resonances
of Self-Excited Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
4.30 Exercise 4.30 (Simultaneous Subharmonic
and Superharmonic Resonance of Self-Excited
Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
4.31 Exercise 4.31 (Self-Excited System with Both Primary
and Combined Resonances I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
4.32 Exercise 4.32 (Self-Excited System with Both Primary
and Combined Resonances II) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
4.33 Exercise 4.33 (Combination Resonance of Self-Excited
Systems) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
4.34 Exercise 4.34 (Self-Excited System with Simultaneous
Primary, Superharmonic, and Subharmonic Resonances) . . . . . . . . 304
4.35 Exercise 4.35 (Self-Excited System with Simultaneous
Primary, and Combinatorial Resonances) . . . . . . . . . . . . . . . . . . . . . 307
4.36 Exercise 4.36 (Simple Harmonic Response of Van Der Pol’s
Oscillator) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
4.37 Exercise 4.37 (Analysis on the Response of a Cubic
Nonlinear System by the Slowly Varying Excitation) . . . . . . . . . . . 313
4.38 Exercise 4.38 (Analysis of the Response of a Self-Excited
System Under Slowly Varying Excitation) . . . . . . . . . . . . . . . . . . . . 315
4.39 Exercise 4.39 (Analysis of the Response of a Self-Excited
Vibrating System Under Hard Excitation at Different
Frequencies) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
5 Parametrically Excited Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
5.1 Exercise 5.1 (The Method of Strained Parameters
to Determine the Transition Curves for a Hill Equation) . . . . . . . . . 319
5.2 Exercise 5.2 (The Method of Strained Parameters
to Determine the Transition Curves of the Hill’s Equation
for LRC Circuits with Sinusoidally Varying Resistance) . . . . . . . . . 322
5.3 Exercise 5.3 (Stability Analysis of a Supported Movable
Pendulum) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
5.4 Exercise 5.4 (Analyze the Linear and Nonlinear Stability
of a Rotating Pendulum Using the Method of Strained
Parameters and the Method of Multiple Scales) . . . . . . . . . . . . . . . . 329
5.5 Exercise 5.5 (Parametric Excitation of a Particle-String
System Subjected to Axial Forces at Both Ends) . . . . . . . . . . . . . . . 337
5.6 Exercise 5.6 (Linear and Nonlinear Stability Analysis
of a Pendulum with Varying Length) . . . . . . . . . . . . . . . . . . . . . . . . . 344
Contents xiii

5.7 Exercise 5.7 (Linear and Nonlinear Stability Analysis

of a Particle Sliding on a Rotating Parabola) . . . . . . . . . . . . . . . . . . . 352
5.8 Exercise 5.8 (Nonlinear Solution to the Pure Rolling
of a Cylinder on a Circular Surface) . . . . . . . . . . . . . . . . . . . . . . . . . . 359
5.9 Exercise 5.9 (The Method of Harmonic Balance to Solve
Duffing’s Equation and Stability Analysis) . . . . . . . . . . . . . . . . . . . . 367
5.10 Exercise 5.10 (The Method of Harmonic Balance for Solving
Pure Cubic Nonlinear Equations) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370
5.11 Exercise 5.11 (The Quadratic Parametric Excitation
of a Linear Viscous Damping System) . . . . . . . . . . . . . . . . . . . . . . . . 374
5.12 Exercise 5.12 (The Cubic Parametric Excitation of a Linear
Viscous Damping System) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
5.13 Exercise 5.13 (The High Order Nonlinear Parametric
Excitation of a Linear Viscous Damping System) . . . . . . . . . . . . . . 382
5.14 Exercise 5.14 (The Parametric Excitation of a System
with a Nonlinear Damping) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388
5.15 Exercise 5.15 (The Nonlinear Parametric Excitation
of a System with a Linear Viscous Damping) . . . . . . . . . . . . . . . . . . 391
5.16 Exercise 5.16 (The Nonlinear Parametric Excitation
of the System with a Nonlinear Damping) . . . . . . . . . . . . . . . . . . . . . 394
5.17 Exercise 5.17 (The Nonlinear Parametric Excitation
of the System with a Nonlinear Damping and High Order
Nonlinearities) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
5.18 Exercise 5.18 (Cubic Nonlinear System Subjected
to Combined Parametric and External Excitation) . . . . . . . . . . . . . . 405
5.19 Exercise 5.19 (Cubic Nonlinear System Subjected
to Combined Parametric and Multi-frequency External
Excitations) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
5.20 Exercise 5.20 (Cubic Nonlinear, Square-Damped System
Subjected to Combined Parametric and External Excitation) . . . . . 419
5.21 Exercise 5.21 (Analysis on a Double Pendulum
with Swinging Mass Attached to Springs and Moving
Platform) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
5.22 Exercise 5.22 (A Two-Frequency Parametric Excitation
of a Multi Degree-of-Freedom System with Distinct
Frequencies) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
5.23 Exercise 5.23 (A Single-Frequency Parametric Excitation
of a Three-Degree-of-Freedom System with Repeating
Frequencies) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
5.24 Exercise 5.24 (A Two-Frequency Parametric Excitation
of a Three-Degree-of-Freedom System with Repeating
Frequencies) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
5.25 Exercise 5.25 (Parametric Excitation Analysis of Oscillations
of a Spring Pendulum) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
xiv Contents

5.26 Exercise 5.26 (The Buckling of the Column with a Nonideal

Energy Source) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
5.27 Exercise 5.27 (Analysis of Two-Dimensional Sound
Propagation in a Pipe with Sinusoidal Walls) . . . . . . . . . . . . . . . . . . 466
5.28 Exercise 5.28 (Analysis of Two-Dimensional
Electromagnetic Wave Propagation in a Pipe with Sinusoidal
Walls) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
6 Systems Having Finite Degrees of Freedom . . . . . . . . . . . . . . . . . . . . . . . 487
6.1 Exercise 6.1 (Internal Resonance Analysis and Nonlinear
Solution of a Double Pendulum Problem) . . . . . . . . . . . . . . . . . . . . . 487
6.2 Exercise 6.2 (Internal Resonance Analysis of a Uniform
Rod Hanging from a Massless Chord) . . . . . . . . . . . . . . . . . . . . . . . . 500
6.3 Exercise 6.3 (Internal Resonance Analysis of a Disc
Pendulum) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511
6.4 Exercise 6.4 (Internal Resonance Analysis of a Spring
Pendulum) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522
6.5 Exercise 6.5 (Internal Resonance Analysis of a Uniform
Rod Hanging a Spring) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
6.6 Exercise 6.6 (Internal Resonance Analysis of the Plane
Motion of a Rigid Beam Supported by a Spring) . . . . . . . . . . . . . . . 538
6.7 Exercise 6.7 (Analytical Solution
of a Two-Degree-of-Freedom High-order Nonlinear
Equation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
6.8 Exercise 6.8 (Forced Oscillations
of a Two-Degree-of-Freedom Gyroscopic System
with Cubic Nonlinearity) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550
6.9 Exercise 6.9 (Forced Oscillations
of a Two-Degree-of-Freedom System with Quadratic
and Cubic Nonlinearities) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
6.10 Exercise 6.10 (Analysis of Forced Oscillations
of a Spring-Slider-Pendulum System) . . . . . . . . . . . . . . . . . . . . . . . . 564
6.11 Exercise 6.11 (Resonance Analysis on the Cylinder Rolling
Without Slip on the Circular Surface I) . . . . . . . . . . . . . . . . . . . . . . . 581
6.12 Exercise 6.12 (Resonance Analysis on the Cylinder Rolling
Without Slip on the Circular Surface II) . . . . . . . . . . . . . . . . . . . . . . 592
6.13 Exercise 6.13 (Parametric Excitation of a Stretched Wire
Carrying Two Particles) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
6.14 Exercise 6.14 (Internal Resonance, Parametric Excitation
and Saturation Phenomena for a Spring Pendulum
with a Moving Support) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622
6.15 Exercise 6.15 (The Response of a Ship Constrained to Pitch
and Roll Only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 631
Contents xv

6.16 Exercise 6.16 (The Method of Multiple Scales for Solving

Free Oscillations of Systems with Slowly Varying
Frequencies) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643
6.17 Exercise 6.17 (Analysis of a Two-Degree-of-Freedom
System with Simultaneous Internal, Subharmonic,
and Superharmonic Resonances) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647
6.18 Exercise 6.18 (Nonlinear Oscillation Analysis of a Spherical
Pendulum with a Moving Support) . . . . . . . . . . . . . . . . . . . . . . . . . . . 649
6.19 Exercise 6.19 (Primary Resonance Analysis of a
Two-Degree-of-Freedom Self-Excited System) . . . . . . . . . . . . . . . . 657
6.20 Exercise 6.20 (Free Oscillation Analysis
of a Two-Degree-of-Freedom Self-Excited System
with Heavy Eigenvalues) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660
6.21 Exercise 6.21 (Primary Resonance Analysis
of Three-Degree-of-Freedom Systems with Repeating
Frequencies, Internal Resonance, Saturation Phenomena) . . . . . . . 663
6.22 Exercise 6.22 (Primary Resonance Analysis
of Three-Degree-of-Freedom Systems with Internal
Resonance) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674
6.23 Exercise 6.23 (Stability Analysis of Nonlinear Forced
Oscillation of a Disk) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674
6.24 Exercise 6.24 (Forced Oscillation of a Spherical Pendulum) . . . . . 680
6.25 Exercise 6.25 (Nonlinear Oscillation Analysis of a Rolling
Reentry Body I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680
6.26 Exercise 6.26 (Nonlinear Oscillation Analysis of a Rolling
Reentry Body II) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 688
7 Continuous Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689
7.1 Exercise 7.1 (Longitudinal Oscillation Analysis
of Non-uniform, Non-linear Elastic Rods) . . . . . . . . . . . . . . . . . . . . . 689
7.2 Exercise 7.2 (Longitudinal Oscillation Analysis of Uniform,
Nonlinear Elastic Rods) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699
7.3 Exercise 7.3 (Longitudinal Oscillation Analysis of Uniform,
Nonlinear Elastic Rods) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 706
7.4 Exercise 7.4 (Nonlinear Analysis of Transverse Oscillations
of a Fixed Wire) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714
7.5 Exercise 7.5 (Nonlinear Analysis of Transverse Oscillations
in the Plane of an Elastic Tensioned String) . . . . . . . . . . . . . . . . . . . 734
7.6 Exercise 7.6 (Nonlinear Analysis of Planar Transverse
Oscillations of Elastic Tensioned Strings Under Parametric
Excitation Formed by Time-Varying Tension) . . . . . . . . . . . . . . . . . 743
7.7 Exercise 7.7 (Transverse Oscillation of a Hinged-Hinged
Beam Excited by First- and Second-Order Primary
Resonances at u = O(w 2 )) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
xvi Contents

7.8 Exercise 7.8 (Forced Response of a Hinged-Hinged Beam

with a Single Non-resonant Excitation at u = O(w 2 )) . . . . . . . . . . 752
7.9 Exercise 7.9 (Combined Resonance Analysis
of Hinged-Hinged Beams with Two Excitations
at u = O(w 2 )) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756
7.10 Exercise 7.10 (Combined Resonance Analysis
of a Hinged-Hinged Beam with Three Excitations
at u = O(w 2 )) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760
7.11 Exercise 7.11 (Parametric Resonance Analysis
on a Hinged-Hinged Beam at (u = O(w 2 )) . . . . . . . . . . . . . . . . . . . 764
7.12 Exercise 7.12 (Transverse Oscillation of a Hinged-Clamped
Beam Under Internal Resonance and Non-resonant
Excitation at u = O(w 2 )) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769
7.13 Exercise 7.13 (Transverse Oscillation of a Hinged-Clamped
Beam Under Internal Resonance and Non-resonance
Excitation at u = O(w 2 )) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 776
7.14 Exercise 7.14 (Transverse Oscillation
of a Clamped–Clamped Beam at Internal Resonance
and Resonant Excitation at u = O(w 2 )) . . . . . . . . . . . . . . . . . . . . . . 781
7.15 Exercise 7.15 (Transverse Oscillation
of a Clamped–Clamped Beam Under Internal
Resonance and Non-resonance Excitation at u = O(w 2 )) . . . . . . . 791
7.16 Exercise 7.16 (Transverse Oscillation
of a Clamped-Supported Beam Under Internal
Resonance and Non-resonance Excitation at u = O(w 2 )) . . . . . . . 798
7.17 Exercise 7.17 (Coupled Longitudinal and Transverse
Oscillation Analysis of Hinged-Hinged Beams at u = O(w)) . . . . 804
7.18 Exercise 7.18 (Analysis of Internal and Primary Resonances
of Cylindrical Shells, Saturation Phenomena) . . . . . . . . . . . . . . . . . . 810
7.19 Exercise 7.19 (Primary Resonance Analysis of Transverse
Oscillations of a Taut String) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 816
7.20 Exercise 7.20 (Oscillation Analysis of a Relief Valve
with Boundary Nonlinearities) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 826
7.21 Exercise 7.21 (First-Order Subharmonic Resonance
Analysis of a Uniform Circular Plate Clamped Along Its
Edge) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 841
7.22 Exercise 7.22 (First-Order Superharmonic Resonance
Analysis of a Uniform Circular Plate Clamped Along Its
Edge) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 848
7.23 Exercise 7.23 (Second-Order Superharmonic Resonance
Analysis of a Uniform Circular Plate Clamped Along Its
Edge) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851
7.24 Exercise 7.24 (Combined Resonance Analysis of a Uniform
Circular Plate Clamped Along Its Edge) . . . . . . . . . . . . . . . . . . . . . . 854
Contents xvii

7.25 Exercise 7.25 (Combined Resonance Analysis of a Uniform

Circular Plate Clamped Along Its Edge I) . . . . . . . . . . . . . . . . . . . . . 858
7.26 Exercise 7.26 (Combined Resonance Analysis of a Uniform
Circular Plate Clamped Along Its Edge II) . . . . . . . . . . . . . . . . . . . . 858
7.27 Exercise 7.27 (Axisymmetric Response of a Uniform
Circular Plate Clamped Along Its Edge with 1st to 3rd Order
Modes Subjecting To Corresponding Resonant Excitation III) . . . . 858
7.28 Exercise 7.28 (Derivation of Modal Discretization Equations
for Berger’s Equation for Axisymmetric Oscillation
Analysis of a Clamped Circular Plates) . . . . . . . . . . . . . . . . . . . . . . . 859
7.29 Exercise 7.29 (Berger Equation for Nonlinear Oscillation
Analysis of Simply Supported Rectangular Plates) . . . . . . . . . . . . . 863
7.30 Exercise 7.30 (Nonlinear Oscillation Analysis of Rotating
Circular Films) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 870
8 Traveling Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881
8.1 Exercise 8.1 (Determination on Dispersive Wave) . . . . . . . . . . . . . . 881
8.2 Exercise 8.2 (Direct Expansion and Reformulation Analysis
of Longitudinal Waves of Semi-infinite Uniform Rods
with Material Nonlinearity) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 882
8.3 Exercise 8.3 (Direct Expansion and Reformulation Analysis
of a Nonlinear Acoustic Equation) . . . . . . . . . . . . . . . . . . . . . . . . . . . 885
8.4 Exercise 8.4 (Multiscale Analysis on a One-Dimensional
Wave Equation with Cubic Nonlinearity) . . . . . . . . . . . . . . . . . . . . . 888
8.5 Exercise 8.5 (A Simplified Model for Wind-Induced
Oscillation of Overhead Power Lines) . . . . . . . . . . . . . . . . . . . . . . . . 890
8.6 Exercise 8.6 (Wave Propagating Along a Uniform, Initially
Undeformed Nonlinear Elastic bar with Linear Damping) . . . . . . . 893
8.7 Exercise 8.7 (Modeling and Multiscale Analysis
on the High-Frequency Oscillation of a Homogenous
Visco-Elastic Rod) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898
8.8 Exercise 8.8 (Modeling and Multiscale Analysis
on Low-Frequency Oscillation of a Homogeneous
Visco-Elastic Rod) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 905
8.9 Exercise 8.9 (Transform the Burgers’ Equation into the Heat
Equation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 907
8.10 Exercise 8.10 (Stationary Solutions of the Burgers’ Equation) . . . . 908
8.11 Exercise 8.11 (Steady-State Solution of the Burgers’
Equation and Their Stability Analysis) . . . . . . . . . . . . . . . . . . . . . . . 909
8.12 Exercise 8.12 (Constant Solutions of the KdV Equation
and Their Stability Analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 911
8.13 Exercise 8.13 (Traveling Wave Solutions of the Sine–Gordon
Equation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 916
xviii Contents

8.14 Exercise 8.14 (Characteristic Transformation

and Straightforward Expansion for Waves Propagating
Along a Uniform Elastic bar with Material Nonlinearity) . . . . . . . . 917
8.15 Exercise 8.15 (Forced Excitations of a Nonlinear Finite
Elastic bar) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 924
8.16 Exercise 8.16 (Derivation of the Lagrangian Form
of the Wave Equation for an Inviscid Isentropic Gas) . . . . . . . . . . . 929
8.17 Exercise 8.17 (Exact Solution of the One-Dimensional
Wave Equation for an Inviscid Isentropic Gas) . . . . . . . . . . . . . . . . . 930
8.18 Exercise 8.18 (Approximate Solutions
of the One-Dimensional Wave Equation for a Viscous
Isentropic Gas and Its Fourier Expansion) . . . . . . . . . . . . . . . . . . . . . 933
8.19 Exercise 8.19 (Derivation of the Eulerian Form
of the One-Dimensional Wave Equation for an Inviscid
Isentropic Gas) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935
8.20 Exercise 8.20 (Derivation of Linear Inviscid Acoustic Waves
in a Hardwalled Duct) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 939
8.21 Exercise 8.21 (Analysis on the Linear Waves Propagating
on the Surface of an Inviscid Liquid of Finite Depth) . . . . . . . . . . . 945
8.22 Exercise 8.22 (The Method of Multiple Scales for Wave
Group Propagation Governed by Klein–Gordon Equation) . . . . . . . 947
Chapter 1
Problem List

To make it easier for the reader to access exercise problems, we have listed all
problems in this chapter and added a short description for each problem. In addition, in
the following chapters, “the Book” refers to “Nonlinear Oscillations. Ali H. Nayfeh &
Dean T. Mook, John Wiley and Sons, New York, 1995, ISBN 0–471-12,142–8”). The
solutions are arranged according to the chapter names question-numbering in the
Book, and our handbook should be viewed as a supplement to the Book.

1.1 Chapter 2 Conservative Single-Degree-Of-Freedom

Systems

• Exercise 2.1 (Determine solution trajectories from potential energy curves of

conservative 1D systems)
• Exercise 2.2 (Solving conservative systems by the method of multiple scales or
straightforward expansion combined with reformulation)
• Exercise 2.3 (Show the system transformation and the method of multiple scales)
• Exercise 2.4 (Show the method of averaging)
• Exercise 2.5 (Show the effect of a change in the equilibrium point of a nonlinear
spring on the frequency-amplitude relationship)
• Exercise 2.6 (Rods sliding on the smooth walls)
• Exercise 2.7 (Period of the osscillation motion of a single pendulum-type system
along separatrices)
• Exercise 2.8 (Particle on a rotating parabola)
• Exercise 2.9 (A single pendulum rotating uniformly around a plumb axis)
• Exercise 2.10 (A single pendulum rotating freely around a plumb axis)
• Exercise 2.11 (Nonlinear oscillation of a slide-slider system)
• Exercise 2.12 (Nonlinear oscillation of a rope-block system)

© The Author(s) 2025 1

Z. He et al., Solved Problems in Nonlinear Oscillations,
https://doi.org/10.1007/978-981-97-6113-5_1
2 1 Problem List

• Exercise 2.13 (Single pendulum attached with a rolling-without-slipping circular

wheel I)
• Exercise 2.14 (Single pendulum attached with a rolling-without-slipping circular
wheel II)
• Exercise 2.15 (Single pendulum problem with inelastic collision with an inclined
wall)
• Exercise 2.16 (Simplified model for buckling analysis of columns)
• Exercise 2.17 (Rods for pure rolling on a fixed cylindrical surface)
• Exercise 2.18 (Geometrically nonlinear system formed by a linear spring and a
mass block)
• Exercise 2.19 (Solving pure cubic nonlinear systems by harmonic balance method)
• Exercise 2.20 (Solving purely fifth-order nonlinear systems by harmonic balance
method)
• Exercise 2.21 (Solving pure cubic nonlinear systems by equivalent linearization)
• Exercise 2.22 (Show least residual value method and Galerkin method for solving
pure cubic nonlinear systems)
• Exercise 2.23 (Discuss the possibility of solving pure cubic nonlinear systems by
the method of multiple scales)
• Exercise 2.24 (Examine the relationship between the period and the amplitude of
oscillation)
• Exercise 2.25 (Comparison of equivalent linearization methods and the method
of multiple scales for solving different nonlinear equations)
• Exercise 2.26 (Comparison of the Galerkin method with the the method of multiple
scales)
• Exercise 2.27 (Equations containing second, third and fourth order nonlinear
terms)
• Exercise 2.28 (Solving nonlinear equations with segmentation functions)
• Exercise 2.29 (Nonlinear equations with a single arbitrary subpartial nonlinear
term only)
• Exercise 2.30 (Higher order nonlinear equations)
• Exercise 2.31 (Motion of the disk constrained by a linear spring)
• Exercise 2.32 (Rolling without slip rollers restrained by a linear spring)
• Exercise 2.33 (Example of higher-order system)

1.2 Chapter 3 Nonconservative Single-Degree-Of-Freedom

Systems

• Exercise 3.1 (Singularity analyses I)

• Exercise 3.2 (Singularity analyses II)
• Exercise 3.3 (Singularity analyses III)
• Exercise 3.4 (Conversion of Rayleigh’s equation to van der Pol’s equation)
• Exercise 3.5 (Plotting phase trajectories and limit cycle(s) for the van der Pol
equation)
1.2 Chapter 3 Nonconservative Single-Degree-Of-Freedom Systems 3

• Exercise 3.6 (Plotting phase trajectories and limit cycle(s) for a given equation)
• Exercise 3.7 (Plotting singularity and phase trajectories for a given planar system)
• Exercise 3.8 (Plot the trajectory of a singularity for a given planar system and
determine the stability of the singularity)
• Exercise 3.9 (The method of multiple scales for solving a single pendulum with
a Coulomb friction torque)
• Exercise 3.10 (Singularities and their stability of a linearly damped pendulum
under a constant torque)
• Exercise 3.11 (Singularities and their stability for a single pendulum with a
quadratic damping)
• Exercise 3.12 (Analysis of Rayleigh equation singularities, method of averaging
of solution)
• Exercise 3.13 (Singularity analysis of the restricted three-body exercise)
• Exercise 3.14 (Singularity analysis of amplitude and phase equations limiting the
resonance response of the three-body problem)
• Exercise 3.15 (Modeling of a single pendulum with a damper attached, solving
by the method of multiple scales)
• Exercise 3.16 (The method of multiple scales for solving van der Pol oscillators
with delay amplitude limiting)
• Exercise 3.17 (Averaging to solve for an oscillator with both Coulomb and viscous
damping)
• Exercise 3.18 (Averaging to solve for oscillators with both Coulomb and square
damping)
• Exercise 3.19 (Averaging to solve for oscillators with both viscous and square
damping)
• Exercise 3.20 (Averaging to solve for an oscillator with both viscous and negative
Coulomb damping)
• Exercise 3.21 (Averaging to solve for an oscillator with both squared damping
and negative Coulomb damping)
• Exercise 3.22 (Oscillators with simultaneous negative viscous damping, square
damping, and cubic nonlinearity)
• Exercise 3.23 (Method of averaging for solving oscillators with simultaneous
viscous, quadratic and cubic damping nonlinearities)
• Exercise 3.24 (Averaging to solve and analyze nonlinear vibrations of a pendulum)
• Exercise 3.25 (Viscous damping, negative stiffness, and cubic nonlinear systems)
• Exercise 3.26 (Transformations of variables and three-level number expansion
methods for solving a cubic nonlinear system)
• Exercise 3.27 (Application of elliptic functions to represent Solutions of a single
pendulum system with a general weak damping function)
• Exercise 3.28 (Masses bounded in definite orbits, analyzing singularities and their
bifurcations)
• Exercise 3.29 (Non-autonomous systems with slowly varying functions)
4 1 Problem List

1.3 Chapter 4 Forced Oscillations of Systems Having

a Single Degree of Freedom

• Exercise 4.1 (The method of multiple scales for analyzing primary resonances of
Coulomb-damped systems with cubic nonlinearities)
• Exercise 4.2 (The method of multiple scales for analyzing primary resonances of
cubic nonlinear, quadratic-damped systems)
• Exercise 4.3 (The method of multiple scales for analyzing primary resonances of
linear spring and hysteresis damped systems)
• Exercise 4.4 (The method of multiple scales for analyzing primary resonances of
linear spring, hysteresis and Coulomb damped systems)
• Exercise 4.5 (The method of multiple scales for analyzing primary resonances of
linear springs, hysteresis and linearly damped systems)
• Exercise 4.6 (The method of multiple scales for analyzing primary resonances of
linear spring, hysteresis and quadratic damped systems)
• Exercise 4.7 (The method of multiple scales for analyzing primary resonances of
quadraticd spring force and quadraticd damped systems)
• Exercise 4.8 (Primary, superharmonic, and subharmonic resonances of cubic
nonlinear and cubic damped systems)
• Exercise 4.9 (Method of averaging for analyzing primary resonances of alternating
damped systems)
• Exercise 4.10 (Primary, superharmonic, and subharmonic resonances of quadratic
and cubic nonlinear systems)
• Exercise 4.11 (Primary and secondary resonances of five nonlinear, linearly
damped systems)
• Exercise 4.12 (An iterative method for general undamped nonlinear systems)
• Exercise 4.13 (The method of harmonic balance to analyze viscous damping and
primary resonances of cubic nonlinear systems)
• Exercise 4.14 (The method of harmonic balance to analyze the primary resonance
of an undamped nonlinear system)
• Exercise 4.15 (Combination resonance for cubic nonlinear systems)
• Exercise 4.16 (The response of cubic nonlinear systems with primary resonance
and non-resonance excitation)
• Exercise 4.17 (Primary and combined resonance for cubic nonlinear systems I)
• Exercise 4.18 (Combined primary, subharmonic, and superharmonic resonances
of a cubic nonlinear system)
• Exercise 4.19 (Primary and combined resonance for cubic nonlinear systems II)
• Exercise 4.20 (Primary and combined resonance of a cubic nonlinear system)
• Exercise 4.21 (Primary resonance of the system with Coulomb friction and
quadratic nonlinearity)
• Exercise 4.22 (Primary resonance problem for the system with quadratic damp
and quadratic nonlinear terms)
• Exercise 4.23 (Combined resonance problem for quadratic nonlinear systems)
1.4 Chapter 5 Parametrically Excited Systems 5

• Exercise 4.24 (Simultaneous subharmonic and superharmonic resonance of

quadratic nonlinear systems)
• Exercise 4.25 (Quadratic nonlinear systems with both subharmonic and combi-
natorial resonances)
• Exercise 4.26 (Quadratic nonlinear systems with both superharmonic and combi-
natorial resonances)
• Exercise 4.27 (Quadratic nonlinear systems with superharmonic, subharmonic
and combinatorial resonances)
• Exercise 4.28 (Primary and superharmonic resonances of self-excited systems)
• Exercise 4.29 (Primary and subharmonic resonances of self-excited systems)
• Exercise 4.30 (Simultaneous subharmonic and superharmonic resonance of self-
excited systems)
• Exercise 4.31 (Self-excited system with both primary and combined resonances
I)
• Exercise 4.32 (Self-excited system with both primary and combined resonances
II)
• Exercise 4.33 (Combination resonance of self-excited systems)
• Exercise 4.34 (Self-excited system with simultaneous primary, superharmonic,
and subharmonic resonances)
• Exercise 4.35 (Self-excited system with simultaneous primary and combinatorial
resonances)
• Exercise 4.36 (Simple harmonic response of van der Pol’s oscillator)
• Exercise 4.37 (Analysis on the response of a cubic nonlinear system by the slowly
varying excitation)
• Exercise 4.38 (Analysis of the response of a self-excited system under slowly
varying excitation)
• Exercise 4.39 (Analysis of the response of a self-excited vibrating system under
hard excitation at different frequencies)

1.4 Chapter 5 Parametrically Excited Systems

• Exercise 5.1 (The method of strained parameters to determine the transition curves
for a Hill equation)
• Exercise 5.2 (The method of strained parameters to determine the transition curves
of the Hill’s equation for LRC circuits with sinusoidally varying resistance)
• Exercise 5.3 (Stability analysis of a supported movable pendulum)
• Exercise 5.4 (Analyze the linear and nonlinear stability of a rotating pendulum
using the method of strained parameters and the method of multiple scales)
• Exercise 5.5 (Parametric excitation of a particle-string system subjected to axial
forces at both ends)
• Exercise 5.6 (Linear and nonlinear stability analysis of a pendulum with varying
length)
6 1 Problem List

• Exercise 5.7 (Linear and nonlinear stability analysis of a particle sliding on a

rotating parabola)
• Exercise 5.8 (Nonlinear solution to the pure rolling of a cylinder on a circular
surface)
• Exercise 5.9 (The method of harmonic balance to solve Duffing’s equation and
stability analysis)
• Exercise 5.10 (The method of harmonic balance for solving pure cubic nonlinear
equations)
• Exercise 5.11 (The quadratic parametric excitation of a linear viscous damping
system)
• Exercise 5.12 (The cubic parametric excitation of a linear viscous damping
system)
• Exercise 5.13 (The high order nonlinear parametric excitation of a linear viscous
damping system)
• Exercise 5.14 (The parametric excitation of a system with a nonlinear damping)
• Exercise 5.15 (The nonlinear parametric excitation of a system with a linear
viscous damping)
• Exercise 5.16 (The nonlinear parametric excitation of the system with a nonlinear
damping)
• Exercise 5.17 (The nonlinear parametric excitation of the system with a nonlinear
damping and high order nonlinearities)
• Exercise 5.18 (Cubic nonlinear system subjected to combined parametric and
external excitation)
• Exercise 5.19 (Cubic nonlinear system subjected to combined parametric and
multi-frequency external excitations)
• Exercise 5.20 (Cubic nonlinear, square-damped system subjected to combined
parametric and external excitation)
• Exercise 5.21 (Analysis on a double pendulum with swinging mass attached to
springs and moving platform)
• Exercise 5.22 (A two-frequency parametric excitation of a multi degree-of-
freedom system with distinct frequencies)
• Exercise 5.23 (A single-frequency parametric excitation of a three-degree-of-
freedom system with repeating frequencies)
• Exercise 5.24 (A two-frequency parametric excitation of a three-degree-of-
freedom system with repeating frequencies)
• Exercise 5.25 (Parametric excitation analysis of oscillations of a spring pendulum)
• Exercise 5.26 (The buckling of the column with a nonideal energy source)
• Exercise 5.27 (Analysis of two-dimensional sound propagation in a pipe with
sinusoidal walls)
• Exercise 5.28 (Analysis of two-dimensional electromagnetic wave propagation
in a pipe with sinusoidal walls)
1.5 Chapter 6 Systems Having Finite Degrees of Freedom 7

1.5 Chapter 6 Systems Having Finite Degrees of Freedom

• Exercise 6.1 (Internal resonance analysis and nonlinear solution of a double

pendulum problem)
• Exercise 6.2 (Internal resonance analysis of a uniform rod hanging from a massless
chord)
• Exercise 6.3 (Internal resonance analysis of a disc pendulum)
• Exercise 6.4 (Internal resonance analysis of a spring pendulum)
• Exercise 6.5 (Internal resonance analysis of a uniform rod hanging a spring)
• Exercise 6.6 (Internal resonance analysis of the plane motion of a rigid beam
supported by a spring)
• Exercise 6.7 (Analytical solution of a two-degree-of-freedom high-order
nonlinear equation)
• Exercise 6.8 (Forced oscillations of a two-degree-of-freedom gyroscopic system
with cubic nonlinearity)
• Exercise 6.9 (Forced oscillations of a two-degree-of-freedom system with
quadratic and cubic nonlinearities)
• Exercise 6.10 (Analysis of forced oscillations of a spring-slider-pendulum system)
• Exercise 6.11 (Resonance analysis on the cylinder rolling without slip on the
circular surface I)
• Exercise 6.12 (Resonance analysis on the cylinder rolling without slip on the
circular surface II)
• Exercise 6.13 (Parametric excitation of a stretched wire carrying two particles)
• Exercise 6.14 (Internal resonance, parametric excitation and saturation
phenomena for a spring pendulum with a moving support)
• Exercise 6.15 (The response of a ship constrained to pitch and roll only)
• Exercise 6.16 (The method of multiple scales for solving free oscillations of
systems with slowly varying frequencies)
• Exercise 6.17 (Analysis of a two-degree-of-freedom system with simultaneous
internal, subharmonic, and superharmonic resonances)
• Exercise 6.18 (Nonlinear oscillation analysis of a spherical pendulum with a
moving support)
• Exercise 6.19 (Primary resonance analysis of a two-degree-of-freedom self-
excited system)
• Exercise 6.20 (Free oscillation analysis of a two-degree-of-freedom self-excited
system with heavy eigenvalues)
• Exercise 6.21 (Primary resonance analysis of three-degree-of-freedom systems
with repeating frequencies, internal resonance, saturation phenomena)
• Exercise 6.22 (Primary resonance analysis of three-degree-of-freedom systems
with internal resonance)
• Exercise 6.23 (Stability analysis of nonlinear forced oscillation of a disk)
• Exercise 6.24 (Forced oscillation of a spherical pendulum)
• Exercise 6.25 (Nonlinear oscillation analysis of a rolling reentry body I)
• Exercise 6.26 (Nonlinear oscillation analysis of a rolling reentry body II)
8 1 Problem List

1.6 Chapter 7 Continuous Systems

• Exercise 7. 1 (Longitudinal oscillation analysis of non-uniform, non-linear elastic

rods)
• Exercise 7.2 (Longitudinal oscillation analysis of uniform, nonlinear elastic rods)
• Exercise 7.3 (Longitudinal oscillation analysis of uniform, nonlinear elastic rods)
• Exercise 7.4 (Nonlinear analysis of transverse oscillations of a fixed wire)
• Exercise 7.5 (Nonlinear analysis of transverse oscillations in the plane of an elastic
tensioned string)
• Exercise 7.6 (Nonlinear analysis of planar transverse oscillations of elastic
tensioned strings under parametric excitation formed by time-varying tension)
• Exercise 7.7 (Transverse oscillation of a hinged-hinged beam excited by first- and
second-order primary resonances at u = O(w2 ))
• Exercise 7.8 (Forced response of a hinged-hinged beam with a single non-resonant
excitation at u = O(w2 ))
• Exercise 7.9 (Combined resonance analysis of hinged-hinged beams with two
excitations at u = O(w2 ))
• Exercise 7.10 (Combined resonance analysis of a hinged-hinged beam with three
excitations at u = O(w2 ))
• Exercise 7.11 (Parametric resonance analysis on a hinged-hinged beam at u =
O(w2 ))
• Exercise 7.12 (Transverse oscillation of a hinged-clamped beam under internal
resonance and non-resonant excitation at u = O(w2 ))
• Exercise 7.13 (Transverse oscillation of a hinged-clamped beam under internal
resonance and non-resonance excitation at u = O(w2 ))
• Exercise 7.14 (Transverse oscillation of a clamped–clamped beam at internal
resonance and resonant excitation at u = O(w2 ))
• Exercise 7.15 (Transverse oscillation of a clamped–clamped beam under internal
resonance and non-resonance excitation at u = O(w2 ))
• Exercise 7.16 (Transverse oscillation of a clamped-supported beam under internal
resonance and non-resonance excitation at u = O(w2 ))
• Exercise 7.17 (Couple longitudinal and transverse oscillation analysis of hinged-
hinged beams at u = O(w))
• Exercise 7.18 (Analysis of internal and primary resonances of cylindrical shells,
saturation phenomena)
• Exercise 7.19 (Primary resonance analysis of transverse oscillations of a taut
string)
• Exercise 7.20 (Oscillation analysis of a relief valve with boundary nonlinearities)
• Exercise 7.21 (First-order subharmonic resonance analysis of a uniform circular
plate clamped along its edge)
• Exercise 7.22 (First-order superharmonic resonance analysis of a uniform circular
plate clamped along its edge)
• Exercise 7.23 (Second-order superharmonic resonance analysis of a uniform
circular plate clamped along its edge)
1.7 Chapter 8 Traveling Waves 9

• Exercise 7.24 (Combined resonance analysis of a uniform circular plate clamped

along its edge I)
• Exercise 7.25 (Combined resonance analysis of a uniform circular plate clamped
along its edge II)
• Exercise 7.26 (Combined resonance analysis of a uniform circular plate clamped
along its edge III)
• Exercise 7.27 (Axisymmetric response of a uniform circular plate clamped
along its edge with 1st to 3rd order modes subjecting to corresponding resonant
excitation)
• Exercise 7.28 (Derivation of modal discretization equations for Berger’s equation
for axisymmetric oscillation analysis of a clamped circular plates)
• Exercise 7.29 (Berger equation for nonlinear oscillation analysis of simply
supported rectangular plates)
• Exercise 7.30 (Nonlinear oscillation analysis of rotating circular films)

1.7 Chapter 8 Traveling Waves

• Exercise 8.1 (Determination on dispersive wave)

• Exercise 8.2 (Direct expansion and reformulation analysis of longitudinal waves
of semi-infinite uniform rods with material nonlinearity)
• Exercise 8.3 (Direct expansion and reformulation analysis of a nonlinear acoustic
equation)
• Exercise 8.4 (Multiscale analysis on a one-dimensional wave equation with cubic
nonlinearity)
• Exercise 8.5 (A simplified model for wind-induced oscillation of overhead power
lines)
• Exercise 8.6 (Wave propagating along a uniform, initially undeformed nonlinear
elastic bar with linear damping)
• Exercise 8.7 (Modeling and multiscale analysis on the high-frequency oscillation
of a homogenous visco-elastic rod)
• Exercise 8.8 (Modeling and multiscale analysis on low-frequency oscillation of
a homogeneous visco-elastic rod)
• Exercise 8.9 (Transform the Burgers’ equation into the heat equation)
• Exercise 8.10 (Stationary solutions of the Burgers’ equation)
• Exercise 8.11 (Steady-state solution of the Burgers’ equation and their stability
analysis)
• Exercise 8.12 (Constant solutions of the KdV equation and their stability analysis)
• Exercise 8.13 (Traveling wave solutions of the sine–Gordon equation)
• Exercise 8.14 (Characteristic transformation and straightforward expansion for
waves propagating along a uniform elastic bar with material nonlinearity)
• Exercise 8.15 (Forced excitations of a nonlinear finite elastic bar)
• Exercise 8.16 (Derivation of the Lagrangian form of the wave equation for an
inviscid isentropic gas)
10 1 Problem List

• Exercise 8.17 (Exact solution of the one-dimensional wave equation for an inviscid
isentropic gas)
• Exercise 8.18 (Approximate solutions of the one-dimensional wave equation for
a viscous isentropic gas and its Fourier expansion)
• Exercise 8.19 (Derivation of the Eulerian form of the one-dimensional wave
equation for an inviscid isentropic gas)
• Exercise 8.20 (Derivation of linear inviscid acoustic waves in a hardwalled duct)
• Exercise 8.21 (Analysis on the linear waves propagating on the surface of an
inviscid liquid of finite depth)
• Exercise 8.22 (The method of multiple scales for wave group propagation
governed by Klein–Gordon equation)

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give appropriate
credit to the original author(s) and the source, provide a link to the Creative Commons license and
indicate if changes were made.
The images or other third party material in this chapter are included in the chapter’s Creative
Commons license, unless indicated otherwise in a credit line to the material. If material is not
included in the chapter’s Creative Commons license and your intended use is not permitted by
statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder.
Chapter 2
Conservative Single-Degree-of-Freedom
Systems

2.1 Exercise 2.1 (Determine Solution Trajectories

from Potential Energy Curves of Conservative 1D
Systems)

Solution: (a) Let v = u̇, , then the differential equation for the solution trajectory of
(a) is given by

dv u
=− (2.1.1)
du v
The general solution of (2.1.1) is

v2 + u2 = 2E (2.1.2)

where E is the constant of integration, which is the total mechanical energy of the
mechanical system. Therefore, the phase trajectories of (a) are a set of concentric
circles centered at the origin (Fig. 2.1a).
(b) Let v = u̇, then the differential equation for the solution trajectory of (b) is

dv −u + u3
= (2.1.3)
du v
1
After integration, v2 + u2 − u4 = 2E (2.1.4)
2
The equation of the solution trajectory of (b) can be obtained from (2.1.4)

1
v = ± 2E − u2 + u4 (2.1.5)
2

© The Author(s) 2025 11

Z. He et al., Solved Problems in Nonlinear Oscillations,
https://doi.org/10.1007/978-981-97-6113-5_2
12 2 Conservative Single-Degree-of-Freedom Systems

The phase trajectories can be sketched directly or from the potential energy, V, of
the system:

u
1 2 1 4
f (u) = u − u , V (u) =
3
f (u)du = u − u (2.1.6)
2 4
0

One can obtain the singular points of the system s1 : (0,0), s2 : (1,0), s3 :
(−1,0) from Eq. (2.1.3), which, together with Eq. (2.1.5), can be adopted to sketch
the potential energy curve and phase trajectories of the system (Fig. 2.1b).

dv u − u3
(c) = (2.1.7)
du v

(a)

(b) (c)

Fig. 2.1 Potential energy curves and solution trajectories for Exercise 2.1 a–g
2.1 Exercise 2.1 (Determine Solution Trajectories from Potential Energy … 13

(d) (e)

(f) (g)

Fig. 2.1 (continued)

1
v2 − u2 + u4 = 2E (2.1.8)
2

1
v = ± 2E + u2 − u4 (2.1.9)
2
u
1 1
f (u) = −u + u , V (u) =
3
f (u)du = − u2 + u4 (2.1.10)
2 4
0

Singular points of the system:

14 2 Conservative Single-Degree-of-Freedom Systems

s1 : (0,0), s2 : (1,0), s3 : (−1,0)

The potential energy curve and phase trajectories of the system are shown in
Fig. 2.1c.

dv −u − u3
(d) = (2.1.11)
du v
1
v2 + u2 + u4 = 2E (2.1.12)
2

1
v = ± 2E − u2 − u4 (2.1.13)
2
u
1 2 1 4
f (u) = u + u3 , V (u) = f (u)du = u + u (2.1.14)
2 4
0

Singular points of the system:

s : (0,0)

The potential energy curve and phase trajectories of the system are shown in
Fig. 2.1d.

dv u + u3
(e) = (2.1.15)
du v
1
v2 − u2 − u4 = 2E (2.1.16)
2

1
v = ± 2E + u2 + u4 (2.1.17)
2
u
1 1
f (u) = −u − u , V (u) =
3
f (u)du = − u2 − u4 (2.1.18)
2 4
0

Singular points of the system:

s : (0,0)

The potential energy curve and phase trajectories of the system are shown in
Fig. 2.1e.
2.1 Exercise 2.1 (Determine Solution Trajectories from Potential Energy … 15

dv −u3
(f) = (2.1.19)
du v
1
v2 + u4 = 2E (2.1.20)
2

1
v = ± 2E − u4 (2.1.21)
2
u
1 4
f (u) = −u + u , V (u) =
3
f (u)du = u (2.1.22)
4
0

Singular points of the system:

s : (0,0)

The potential energy curve and phase trajectories of the system are shown in
Fig. 2.1f.
λ
dv −u − u−a
(g) = (2.1.23)
du v
This yields that the singular points of the system are
 √   √ 
a+ a2 − 4λ a − a2 − 4λ
s1 : , 0 , s2 : ,0 (2.1.24)
2 2

So, only whena2 ≥ 4λ (2.1.25)

singular points are real and


1 2 1 2 λln(u − a), u − a > 0
v + u + =E (2.1.26)
2 2 −λln(a − u), u − a < 0

v = ± 2E − u2 − 2λln(u − a), u − a > 0v

= ± 2E − u2 + 2λln(a − u), u − a < 0 (2.1.27)

u
λ
f (u) = u + , V (u) = f (u)du
u−a
0
1 2
u + λln(u − a), u−a >0
= 21 2 (2.1.28)
2
u − λln(a − u), u − a < 0
16 2 Conservative Single-Degree-of-Freedom Systems
√
Let a = 8, λ = 1, then
√ √
s1 : ( 2 + 1, 0), s2 : ( 2 − 1, 0) (2.1.29)
 √ √
v = ± 2E − u2 − 2ln(u − 8), u − 8 > 0
 √ √
v = ± 2E − u2 + 2ln( 8 − u), u − 8 < 0 (2.1.30)

1 √ √
u2 + ln(u
√− 8), u − √8 > 0
V (u) = 2 (2.1.31)
1 2
2
u − ln( 8 − u), u − 8 < 0

The potential energy curve and phase trajectories of the system are shown in
Fig. 2.1g.

2.2 Exercise 2.2 (Solving Conservative Systems

by the Method of Multiple Scales or Straightforward
Expansion Combined with Reformulation)

Solution: (a) Expanding the equation around the singularity u = 0 yields

u + u3 − u3 u5
ü + ω02 = ü + ω 2
(u − u 3
+ ) ≈ ü + ω02 u − ω02 u3 = 0
1 + u2 0
1 + u2

i.e ü + ω02 u − ω02 u3 = 0 (2.2.1)

We introduce a small, dimensionless parameter ε, which is the order of the

amplitude of the motion u. Therefore, we can write

u = εv

then Eq. (2.2.1) changes to

v̈ + ω02 v − ε2 ω02 v3 = 0 (2.2.2)

The method of multiple scales. We first solve the Eq. (2.2.2) using the the method
of multiple scales. Let the expansion of the solution be

v = v0 (T0 , T1 , T2 ) + εv1 (T0 , T1 , T2 ) + ε2 v2 (T0 , T1 , T2 ) + · · · (2.2.3)

2.2 Exercise 2.2 (Solving Conservative Systems by the Method of Multiple … 17

where Tn = εn t. Consider
d
dt2
= D0 + εD1 + ε2 D2 + · · ·
d
dt 2
= (D0 + εD1 + ε2 D2 + · · · )2 = D02 + 2εD0 D1 + ε2 (D12 + 2D0 D2 ) + · · ·
(2.2.4)

Substituting (2.2.3) into Eq. (2.2.1), we can obtain

0 = v̈ + ω02 v − εω02 v3
= [D02 + 2εD0 D1 + ε2 (D12 + 2D0 D2 )](v0 + εv1 + ε2 v2 )
+ ω02 (v0 + εv1 + ε2 v2 ) − ε2 ω02 (v0 + εv1 + ε2 v2 )3 + · · ·
= D02 v0 + ω02 v0 + ε(D02 v1 + ω02 v1 + 2D0 D1 v0 )
+ ε2 [D02 v2 + ω02 v2 + 2D0 D1 v1 + (D12 + 2D0 D2 )v0 − ω02 v03 ] + · · · (2.2.5)

Equating the coefficients of ε, ε 2 , and ε 3 to zero, we obtain

D02 v0 + ω02 v0 = 0 (2.2.6)

D02 v1 + ω02 v1 = −2D0 D1 v0 (2.2.7)

D02 v2 + ω02 v2 = −2D0 D1 v1 − (D12 + 2D0 D2 )v0 + ω02 v03 (2.2.8)

The solution of Eq. (2.2.6) is

v0 = A(T1 , T2 )eiω0 T0 + cc (2.2.9)

Substituing (2.2.9) into the Eq. (2.2.7), we obtain

D02 v1 + ω02 v1 = −2iω0 D1 Aeiω0 T0 + cc (2.2.10)

where cc denotes the complex conjugate of the preceding terms. In order to eliminate
the secular terms from the above equation, there must be

D1 A = 0 ⇒ A = A(T2 ) (2.2.11)

Substituing (2.2.11) into the Eq. (2.2.10), we obtain

v1 = 0 (2.2.12)

Substituting (2.2.9) and (2.2.12) into the Eq. (2.2.8), yields

vD02 v2 + ω02 v2 = −2D0 D2 v0 + ω02 v03

18 2 Conservative Single-Degree-of-Freedom Systems

= (−2iω0 D2 A + 3ω02 A2 A)eiω0 T0 + cc + NST (2.2.13)

where NST denotes terms which do not generate secular terms. In order to eliminate
the secular term from the above equation, there must be

−2iD2 A + 3ω0 A2 A = 0 (2.2.14)

1 iβ
Let A= ae (2.2.15)
2
and put (2.2.15) into (2.2.14), we obtain

3
i − a + aβ  + ω0 a3 = 0 (2.2.16)
8
Separating the result into real and imaginary parts, we obtain

3
a = 0, aβ  + ω0 a3 = 0 (2.2.17)
8
Therefore,

3
a = constant, β = − ω0 a2 t + β0 (2.2.18)
8
From the above results, the two-term approximation of the solution can be obtained
as
3
v = acos[(1 − a2 )ω0 t + β0 ] + · · · (2.2.19)
8
Therefore, the frequency-amplitude relationship of the oscillation is

3
ω = (1 − a2 )ω0 (2.2.20)
8

Straightforward Expansion Combined with Reformulation

We then use the straightforward expansion combined with reformulation to solve the
Eq. (2.2.2). Let the expansion of the solution be

v = v0 (t) + ε v1 (t) + · · · (2.2.21)

where ε = ε2 . Substitute (2.2.21) into the Eq. (2.2.2) and equate the coefficients of
the same power of ε in the result, giving

i.e. v̈0 + ω02 v0 = 0 (2.2.22)

2.2 Exercise 2.2 (Solving Conservative Systems by the Method of Multiple … 19

v̈1 + ω02 v1 = ω02 v03 (2.2.23)

The solution of Eq. (2.2.22) can be expressed as

v0 = acos(ω0 t + β) (2.2.24)

Then the Eq. (2.2.23) can be written as

v̈1 + ω02 v1 = ω02 a3 cos3 (ω0 t + β)

3 1
= ω02 a3 cos(ω0 t + β) + ω02 a3 cos(3ω0 t + 3β) (2.2.25)
4 4
whose particular solution is

3 1
v1 = ω0 a3 tsin(ω0 t + β) − a3 cos(3ω0 t + 3β) (2.2.26)
8 32
Therefore, the solution of Eq. (2.2.2) is

3 1
v = acos(ω0 t + β) + ε [ ω0 a3 tsin(ω0 t + β) − a3 cos(3ω0 t + 3β)] + · · ·
8 32
(2.2.27)

This solution is not valid when .t ≥ O(ε−1 )

Now we apply the reformulation method to make the above expansion uniformly
valid. Let

ε1
τ = t = (ω0 + ε1 + · · · )t ⇒ ω0 t = (1 − + · · · )τ (2.2.28)
ω0

put (2.2.28) into (2.2.27), and collect the result according to the power of ε:

lv = acos(τ + β)
1 a 3 1
+ ε̂[ τ sin(τ + β) + a3 τ sin(τ + β) − a3 cos(3τ + 3β)] + · · ·
ω0 8 32
(2.2.29)

We select 1 to eliminate secular terms of the Eq. (2.2.29) and this gives us

3
1 = − a2 ω0 (2.2.30)
8
Therefore, the two-term approximation of the solution of Eq. (2.2.1) is
20 2 Conservative Single-Degree-of-Freedom Systems

3
u = εacos(τ + β) + · · · = εacos[(1 − ε2 a2 )ω0 t + β] + · · · (2.2.31)
8

i.e. u = acos(ωt + β) (2.2.32)

where the frequency-amplitude relationship of the oscillation is

3
ω = (1 − a2 )ω0 (2.2.33)
8
This result is the same as that of the multiscale method.
(b) Straightforward expansion combined with reformulation. Let

u = ελ v

Then the original equation becomes

v̈ + ω02 v + αε4λ v5 = 0 (2.2.34)

If we take λ = 41 , then the coefficients of the nonlinear term are of order ε and
the Eq. (2.2.34) becomes

v̈ + ω02 v + εαv5 = 0 (2.2.35)

We assume that the solution of the above equation can be represented by an

expansion having the form

v = v0 (t) + εv1 (t) + · · · (2.2.36)

Then we substitute (2.2.36) into the Eq. (2.2.35) and set the coefficient of each
power of ε equal to zero. This leads to the following set of equations

v̈0 + ω02 v0 = 0 (2.2.37)

v̈1 + ω02 v1 = −αv05 (2.2.38)

The general solution of Eq. (2.2.37) can be written in the form

v0 = acos(ω0 t + β) (2.2.39)

therefore, (2.2.38) becomes

v̈1 + ω02 v1 = −αa5 cos5 (ω0 t + β)

2.2 Exercise 2.2 (Solving Conservative Systems by the Method of Multiple … 21

5 5
= − αa5 cos(ω0 t + β) − αa5 cos(3ω0 t + 3β)
8 16
1
− αa cos(5ω0 t + 5β)
5
(2.2.40)
16
The particular solution of the above equation is

5
u = ε1/4 acos(ω0 t + β) + ε5/4 [− αa5 tsin(ω0 t + β)
16ω0
5 1
+ αa5 cos(3ω0 t + 3β) + αa5 cos(5ω0 t + 5β)] (2.2.41)
128 384
Therefore, the solution of the original equation is

5
u = ε1/4 acos(ω0 t + β) + ε5/4 [− αa5 tsin(ω0 t + β)
16ω0
5 1
+ αa5 cos(3ω0 t + 3β) + αa5 cos(5ω0 t + 5β)] (2.2.42)
128 384

This solution is not valid when .t ≥ O(ε−1 )

Now we apply the reformulation method to make the above expansions uniformly
valid. Let
ε1
τ = t = (ω0 + ε1 + · · · )t ⇒ ω0 t = (1 − + · · · )τ (2.2.43)
ω0

Substituting (2.2.28) into (2.2.27) leads to

5 αa 5 1 a
u = ε1/4 acos(τ + β) + ε5/4 [− 16 ω2
τ sin(τ + β) + ω0
τ sin(τ + β)
0 (2.2.44)
+ 128
5
αa5 cos(3τ + 3β) + 1
384
αa5 cos(5τ + 5β) + · · · ]

To eliminate the secular terms from (2.2.29), we must put

5
1 = αa4 (2.2.45)
16ω0

Therefore, the first-order approximation of the solution is

5
u = u = ε1/4 acos(τ + β) + · · · = ε1/4 acos[(ω0 + αεa4 )t + β] + · · ·
16ω0
(2.2.46)

Let ε = 1, Eq. (2.2.46) becomes

u = acos(ωt + β) (2.2.47)
22 2 Conservative Single-Degree-of-Freedom Systems

where the relationship between the frequency and the amplitude is

5
ω = (ω0 + αa4 ) (2.2.48)
16ω0

(c) The method of multiple scales. The system described by Exercise 2.1(c) has three
singularities (0,0), (−1,0), (1,0), of which (0,0) is the saddle point and the remaining
two are the centers. The motion is oscillatory in the neighborhood of a center. Here
we discuss the oscillatory motion near the singularity (1,0). It is convenient to shift
the origin to the location of the center, u = 1. Thus, we let x = u − 1, then the
equation in Exercise 2.1(c) can be written as

ẍ + 2x + 3x2 + x3 = 0 (2.2.49)

Let

u = ελ v

Equation (2.2.49) changes to

v̈ + 2v + 3ελ v2 + ε2λ v3 = 0 (2.2.50)

If we take λ = 1 and write the Eq. (2.2.50) in the following form:

v̈ + ω02 v + εδv2 + ε2 αv3 = 0 (2.2.51)

where ω02 = 2, δ = 3, α = 1 (2.2.52)

Assume the solution of the above nonlinear equation can be represented by an

expansion having the form

v = v0 (T0 , T1 , T2 ) + εv1 (T0 , T1 , T2 ) + ε2 v2 (T0 , T1 , T2 ) + · · · (2.2.53)

where T0 = t, T1 = εt. Substitute (2.2.53) into (2.2.51), we obtain

0 = v̈ + ω02 v + εδv2 + ε2 αv3

= [D02 + 2εD0 D1 + ε2 (D12 + 2D0 D2 ) + · · · ](v0 + εv1 + ε2 v2 + · · · )
+ ω02 (v0 + εv1 + ε2 v2 + · · · ) + εδ(v0 + εv1 + ε2 v2 + · · · )2
+ αε2 (v0 + εv1 + ε2 v2 + · · · )3
= D02 v0 + ω02 v0 + ε(D02 v1 + 2D0 D1 v0 + ω02 v1 + δv02 )
+ ε2 [D02 v2 + 2D0 D1 v1 + (D12 + 2D0 D2 )v0 + ω02 v2 + 2δv0 v1 + αv03 ] + · · ·
(2.2.54)
2.2 Exercise 2.2 (Solving Conservative Systems by the Method of Multiple … 23

Set the coefficient of each power of ε equal to zero yield

D02 v0 + ω02 v0 = 0 (2.2.55)

D02 v1 + ω02 v1 = −2D0 D1 v0 − δv02 (2.2.56)

D02 v2 + ω02 v2 = −2D0 D1 v1 − (D12 + 2D0 D2 )v0 − 2δv0 v1 − αv03 (2.2.57)

The homogeneous solution of Eq. (2.2.55) is

v0 = Aeiω0 T0 + Ae−iω0 T0 (2.2.58)

where A = A(T1 , T2 ). Substituting the above equation into (2.2.56) yielding

D02 v1 + ω02 v1 = −2D0 D1 (Aeiω0 T0 + Ae−iω0 T0 ) − δ(Aeiω0 T0 + Ae−iω0 T0 )2

(2.2.59)
= −δAA − 2iω0 D1 Aeiω0 T0 − δA2 e2iω0 T0 + cc

where cc denotes the complex conjugate of the preceding terms. To eliminate the
secular terms in the above equation, we need D1 A = 0, so A = A(T2 ). The particular
solution of Eq. (2.2.59) is

δAA δA2 2iω0 T0

v1 = − + e + cc (2.2.60)
ω02 3ω02

Substituting (2.2.58) and (2.2.60) into (2.2.57), and taking A = A(T2 ) into
account, we obtain

D02 v2 + ω02 v2 = −2D0 D1 v1 − (D12 + 2D0 D2 )v0 − 2δv0 v1 − αv03

2 (2.2.61)
= [−2iω0 D2 A + ( 10δ
3ω2
− 3α)A2 A]eiω0 T0 + cc + NST
0

where cc denotes the complex conjugate of the preceding terms and NST denotes
terms which do notgenerate secular terms. In order to eliminate the secular terms in
the above equation, we need

10δ 2
2iω0 D2 A − ( − 3α)A2 A = 0 (2.2.62)
3ω02
1 iβ
Let A = ae (2.2.63)
2
Substituting (2.2.63) into (2.2.62) leads to

1 10δ 2
iω0 D2 a − ω0 aD2 β − ( 2 − 3α)a3 = 0 (2.2.64)
8 3ω0
24 2 Conservative Single-Degree-of-Freedom Systems

Separate the above equation into real and imaginary parts and yield

D2 a = 0 (2.2.65)

1 10δ 2
D2 β + ( − 3α)a2 = 0 (2.2.66)
8ω0 3ω02

Therefore, a = constant and

1 10δ 2
β=− ( − 3α)a2 T2 + β0 (2.2.67)
8ω0 3ω02

where β0 is a constant.Combining all above results, we can obtain

1 10δ 2
u = εacos[ω0 t − ( − 3α)a2 ε2 t + β0 ] + · · · (2.2.68)
8ω0 3ω02

Let ε = 1 yield

1 10δ 2
u = acos[ω0 t − ( − 3α)a2 t + β0 ] + · · · (2.2.69)
8ω0 3ω02

Therefore, the relation between the oscillation frequency ω and the amplitude a
is

1 10δ 2
ω = ω0 − ( − 3α)a2 (2.2.70)
8ω0 3ω02
3 5 2 2
or ω2 = ω02 + αa2 − δ a (2.2.71)
4 6ω02

Considering the Eq. (2.2.52), the above equation can be simplified as

ω2 = 2 − 3a2 (2.2.72)

(d) Let
v
u= √ (2.2.73)
α

Then the original equation becomes

v̈ + ω02 v(1 + v2 )−1 = 0 (2.2.74)

2.3 Exercise 2.3 (Show the System Transformation and the Method … 25

This equation is identical to the equation in subExercise (a), so the first-order

approximate solution of the equation is

3
v = acos[(1 − a2 )ω0 t + β] (2.2.75)
8
where
3
ω = (1 − a2 )ω0 (2.2.76)
8

2.3 Exercise 2.3 (Show the System Transformation

and the Method of Multiple Scales)

Solution: Since

d m0 u̇ m0 ü(1 − u̇/c2 )1/2 + 21 m0 u̇(1 − u̇/c2 )−1/2 (ü/c2 )

[ ] =
dt (1 − u̇/c2 )1/2 (1 − u̇/c2 )
m0 ü(1 − 21 u̇/c2 ) m0 üc(2c2 − u̇)
= = (2.3.1)
(1 − u̇/c2 )3/2 2(c2 − u̇)3/2

the equation can be written as

(1 − u̇/c2 )3/2
ü + ω02 u =0 (2.3.2)
1 − 21 u̇/c2
√
where ω02 = k/m0 . We make a Taylor expansion of (1 − u̇/c2 )3/2 /(1 − 21 u̇/c2 ) and
yield

(1 − u̇/c2 )3/2 u̇ 1 u̇
= 1 − 2 − ( 2 )2 + · · · (2.3.3)
1 − 2 u̇/c
1 2 c 8 c

Therefore, the equation can be written as

ω02 ω2
ü + ω02 u − 2
uu̇ − 04 uu̇2 = 0 (2.3.4)
c 8c
Let

u = ελ v
26 2 Conservative Single-Degree-of-Freedom Systems

then the Eq. (2.3.4) changes to

ω02 λ ω02 2λ 2
v̈ + ω02 v − ε vv̇ − ε vv̇ = 0 (2.3.5)
c2 8c4
Let λ = 1, then the Eq. (2.3.5) becomes

v̈ + ω02 v − εα1 vv̇ − ε2 α2 vv̇2 = 0 (2.3.6)

where α1 = ω02 /c2 , α2 = ω02 /8c4 .

Assume the solution of the above equation can be represented by an expansion
having the form

v = v0 (T0 , T1 , T2 ) + εv1 (T0 , T1 , T2 ) + ε2 v2 (T0 , T1 , T2 ) + · · · (2.3.7)

where T0 = t, T1 = εt, T2 = ε2 t. Substitute (2.3.7) into the Eq. (2.3.6), we obtain

0 = v̈ + ω02 v − εα1 vv̇ − ε2 α2 vv̇2

= [D02 + 2εD0 D1 + ε2 (D12 + 2D0 D2 ) + · · · ](v0 + εv1 + ε2 v2 + · · · )
+ ω02 (v0 + εv1 + ε2 v2 + · · · )
− α1 ε(v0 + εv1 + ε2 v2 + · · · )[(D0 + εD1 + · · · )(v0 + εv1 + ε2 v2 + · · · )]
− ε2 α2 (v0 + εv1 + ε2 v2 + · · · )[(D0 + εD1 + · · · )(v0 + εv1 + ε2 v2 + · · · )]2
= D02 v0 + ω02 v0
+ ε(D02 v1 + ω02 v1 + 2D0 D1 v0 − α1 v0 D0 v0 )
+ ε2 [D02 v2 + ω02 v2 + 2D0 D1 v1 + (D12 + 2D0 D2 )v0
− α1 v0 D0 v1 − α1 v1 D0 v0 − α2 v0 (D0 v0 )2 ] + · · · (2.3.8)

Set the coefficient of each power of ε equal to zero yield

D02 v0 + ω02 v0 = 0 (2.3.9)

D02 v1 + ω02 v1 = −2D0 D1 v0 + α1 v0 D0 v0 (2.3.10)

D02 v2 + ω02 v2 = −2D0 D1 v1 − (D12 + 2D0 D2 )v0 + α1 v0 D0 v1 + α1 v1 D0 v0 + α2 v0 (D0 v0 )2 (2.3.11)

The homogeneous solution of Eq. (2.3.9) is

v0 = Aeiω0 T0 + Ae−iω0 T0 (2.3.12)

where A = A(T1 , T2 ). Substituting the above equation into (2.3.10), we obtain

D02 v1 + ω02 v1 = −2D0 D1 (Aeiω0 T0 + Ae−iω0 T0 )

2.3 Exercise 2.3 (Show the System Transformation and the Method … 27

+ iα1 ω0 (Aeiω0 T0 + Ae−iω0 T0 )(Aeiω0 T0 − Ae−iω0 T0 )

= −2iω0 D1 Aeiω0 T0 + iα1 ω0 A2 e2iω0 T0 + cc (2.3.13)

To eliminate the secular terms in the above equation, we need D1 A = 0, so

A = A(T2 ). The particular solution of Eq. (2.3.13) is

iα1 A2 2iω0 T0
v1 = − e + cc (2.3.14)
3ω0

Substituting (2.3.12) and (2.3.14) into (2.3.11), and taking A = A(T2 ) into
account, we obtain

α12
D02 v2 + ω02 v2 = [−2iω0 D2 A + ( + α2 ω02 )A2 A]eiω0 T0 + cc + NST (2.3.15)
3
where cc denotes the complex conjugate of its preceding terms and NST denotes
terms which do not generate secular terms. In order to eliminate secular terms, we
need

α12
2iω0 D2 A − ( + α2 ω02 )A2 A = 0 (2.3.16)
3
1 iβ
Let A= ae (2.3.17)
2
Substituting (2.3.17) into (2.3.16) leads to

α12
8iω0 D2 a − 8ω0 σ D2 β − ( + α2 ω02 )σ 3 = 0 (2.3.18)
3
Separate the above equation into real and imaginary parts and yield

D2 a = 0 (2.3.19)

α12
−8ω0 aD2 β − ( + α2 ω02 )a3 = 0 (2.3.20)
3
So a is a constant and

1 α12
β=− ( + α2 ω02 )a2 T2 + β0 (2.3.21)
8ω0 3

Combining the above results, we obtain

1 α12
u = εacos{[ω0 − ( + α2 ω02 )ε2 a2 ]t + β0 } + · · · (2.3.22)
8ω0 3
28 2 Conservative Single-Degree-of-Freedom Systems

Let ε = 1, the first-order approximate solution of the original equation is

u = acos(ωt + β0 ) + · · · (2.3.23)

where

1 α12
ω = ω0 − ( + α2 ω02 )a2 (2.3.24)
8ω0 3

Substituting α1 = ω02 /c2 , α2 = ω02 /8c4 into the above equation, we obtain

11 ω03 2
ω = ω0 − a (2.3.25)
192 c4
√
where ω0 = k/m0 .

2.4 Exercise 2.4 (Show the Method of Averaging)

ü + ω02 u + u|u| = 0 (2.4.1)

Solution: We use the pending power ελ , λ > 0 of the small parameter ε to measure
the magnitude of u, i.e.,

u = ελ v (2.4.2)

then the Eq. (2.4.1) changes to

v̈ + ω02 v + ελ v|v| = 0 (2.4.3)

If we take λ = 1, then the Eq. (2.3.5) becomes

v̈ + ω02 v + εv|v| = 0 (2.4.4)

Here we use the method of averaging to solve the Eq. (2.4.4). When ε = 0,

v = acos(ω0 t + β) (2.4.5)

v̇ = −ω0 asin(ω0 t + β) (2.4.6)

here a, β is a constant.
2.4 Exercise 2.4 (Show the Method of Averaging) 29

The principle of the method of averaging: when ε = 0, the solution of (2.4.4) can
still be expressed in the form (2.4.5) and (2.4.6) provided that aandβ are considered
to be functions of time t rather than constants. Thus, the Eqs. (2.4.5) and (2.4.6) can
be viewed as a transformation of v(t) and v̇(t) into the dependent variables a(t) and
β(t). Differentiating the Eq. (2.4.5) with respect to t gives

v̇ = ȧcosϕ − ω0 asinϕ − aβ̇sinϕ, ϕ = ω0 t + β (2.4.7)

Comparing (2.4.6) and (2.4.7), we find that

ȧcosϕ − aβ̇sinϕ = 0 (2.4.8)

Differentiating (2.4.6) with respect to t yields

v̈ = −ω0 ȧsinϕ − ω02 acosϕ − ω0 aβ̇cosϕ (2.4.9)

Substituting (2.4.9) and (2.4.5) into the Eq. (2.4.4) yields

−ω0 ȧsinϕ − ω0 aβ̇cosϕ + εa2 cosϕ|cosϕ| = 0 (2.4.10)

Solving (2.4.8) and (2.4.10) for ȧ and β̇, we obtain

εa2
ȧ = sinϕcosϕ|cosϕ| (2.4.11)
ω0
εa
β̇ = cos2 ϕ|cosϕ| (2.4.12)
ω0

Averaging (2.4.11) and (2.4.12) over the period 2π/ω0 and considering a, β, ȧ and
β̇ to be constants while performing the averaging, we obtain the following equations
describing the slow variations of a and β:

2π
1 εa2
ȧ = sinφcosφ|cosφ|d φ
2π ω0
0
π/2 π
εa2
= [ sinφcos φd φ −
2
sinφcos2 φd φ]
π ω0
0 π/2

0 −1
εa2
=− [ x2 dx − x2 dx] = 0 (2.4.13)
π ω0
1 0
30 2 Conservative Single-Degree-of-Freedom Systems

2π
1 εa
β̇ = cos2 φ|cosφ|d φ
2π ω0
0
π/2 π
εa
= [ cos φd φ −
3
cos3 φd φ]
π ω0
0 π/2

1 0
εa
= [ (1 − x2 )dx − (1 − x2 )dx]
π ω0
0 1
1
2εa 4εa
= (1 − x2 )dx = (2.4.14)
π ω0 3π ω0
0

So, a is a constant and

4εa
β= t + β0 (2.4.15)
3π ω0

Thus, the first-order approximate solutions of (2.4.1) is

4εa
u = εacos[(ω0 + )t + β0 ] (2.4.16)
3π ω0

Let ε = 1, we can obtain

4a
u = acos[(ω0 + )t + β0 ] (2.4.17)
3π ω0

The frequency-amplitude relationship of the oscillation is

4a
ω = ω0 + (2.4.18)
3π ω0

2.5 Exercise 2.5 (Show the Effect of a Change

in the Equilibrium Point of a Nonlinear Spring
on the Frequency-Amplitude Relationship)

Solution: (a) Select the position of the mass m when the spring is at rest as the origin
of the coordinate, we can obtain the governing equation of the system
2.5 Exercise 2.5 (Show the Effect of a Change in the Equilibrium Point … 31

mẍ = −f (x), f (x) = k1 x + k3 x3

i.e. ẍ + ω02 x + αx3 = 0 (2.5.1)

where ω02 = k1 /m, α = k3 /m (2.5.2)

Let

x = ελ v (2.5.3)

(2.5.1) changes to

v̈ + ω02 v + αε2λ v3 = 0 (2.5.4)

If we take λ = 1, then the Eq. (2.5.4) becomes

v̈ + ω02 v + αε2 v3 = 0 (2.5.5)

We use the the method of multiple scales to solve the Eq. (2.5.5). Assume the
solution of the above equation can be represented by an expansion having the form

v = v0 (T0 , T1 , T2 ) + εv1 (T0 , T1 , T2 ) + ε2 v2 (T0 , T1 , T2 ) + · · · (2.5.6)

where T0 = t, T1 = εt. Substitute (2.5.6) into (2.5.5), we obtain

0 = v̈ + ω02 v + αε2 v3
= [D02 + 2εD0 D1 + ε2 (D12 + 2D0 D2 ) + · · · ](v0 + εv1 + ε2 v2 + · · · )
+ ω02 (v0 + εv1 + ε2 v2 + · · · ) + αε2 (v0 + εv1 + ε2 v2 + · · · )3
= D02 v0 + ω02 v0 + ε(D02 v1 + 2D0 D1 v0 + ω02 v1 )
+ ε2 [D02 v2 + (D12 + 2D0 D2 )v0 + 2D0 D1 v1 + ω02 v2 + αv03 ] + · · · (2.5.7)

Let the coefficients of the same power of ε be zero to obtain

D02 v0 + ω02 v0 = 0 (2.5.8)

D02 v1 + ω02 v1 = −2D0 D1 v0 (2.5.9)

D02 v2 + ω02 v2 = −(D12 + 2D0 D2 )v0 − 2D0 D1 v1 − αv03 (2.5.10)

The general solution of Eq. (2.5.8) is

32 2 Conservative Single-Degree-of-Freedom Systems

v0 = Aeiω0 T0 + Ae−iω0 T0 (2.5.11)

where A = A(T1 , T2 ). Substituting it into (2.5.9), we can obtain

D02 v1 + ω02 v1 = −2D0 D1 (Aeiω0 T0 + Ae−iω0 T0 )

= −2iω0 D1 Aeiω0 T0 + cc (2.5.12)

To eliminate the secular term, we need D1 A = 0, so A = A(T2 ). Then the particular

solution of Eq. (2.5.12) is

v1 = 0 (2.5.13)

Substituting (2.5.11) and (2.5.13) into (2.5.10), and taking A = A(T2 ) into
account, we can obtain

D02 v2 + ω02 v2 = (−2iω0 D2 A − 3αA2 A)eiω0 T0 + cc + NST (2.5.14)

where cc denotes the complex conjugate term of its preceding terms and NST denotes
terms which do not generate secular terms. To eliminate the secular term, we need

2iω0 D2 A + 3αA2 A = 0 (2.5.15)

1 iβ
Let A= ae (2.5.16)
2
and put (2.5.16) into (2.5.15), we can obtain

8iω0 D2 a − 8ω0 aD2 β + 3αa3 = 0 (2.5.17)

Separates the real and imaginary parts of the above equation yields

D2 a = 0 (2.5.18)

−8ω0 aD2 β + 3αa3 = 0 (2.5.19)

So a is a constant and

3αa2
β= T2 + β0 (2.5.20)
8ω0

Combining the above results, we can obtain

3αε2 a2
x = εacos(ω0 t + t + β0 ) + · · · (2.5.21)
8ω0
2.5 Exercise 2.5 (Show the Effect of a Change in the Equilibrium Point … 33

Let ε = 1, we can obtain the first-order approximation of the Exercise as the

following

3αa2
x = acos(ω0 t + t + β0 ) + · · · (2.5.22)
8ω0

and the the frequency-amplitude relationship of the oscillation is

3αa2
ω = ω0 + (2.5.23)
8ω0

where ω02 = k1 /m, α = k3 /m.

(b) Again, choose the position of the mass m when the spring is at rest as the origin
of the coordinate, we can obtain the governing equation of the system

mẍ = −f (x) + mg, f (x) = k1 x + k3 x3 (2.5.24)

The equilibrium coordinates of the mass x are determined by the following

equation:

k1 x + k3 (x)3 − mg = 0 (2.5.25)

Let x = y + x (2.5.26)

Then the Eq. (2.5.24) becomes

mÿ + [k1 + 3k3 (x)2 ]y + (3k3 x)y2 + k3 y3 = 0 (2.5.27)

∼2
i.e ÿ + ω0 y + δy2 + αy3 = 0 (2.5.28)

where ω̃02 = [k1 + 3k3 x)2 /m, δ = 3k3 x/m, α = k3 /m (2.5.29)

Equation (2.5.28) can be changed to the Eq. (2.2.51) in Exercise 2.2(c), the
Eq. (2.2.51) the frequency-amplitude relationship of the oscillation is can be obtained
from (2.2.70), i.e.,

∼ 1 4δ 2
ω = ω0 − ∼ ( − 3α)a2 (2.5.30)
∼2
8ω0 3ω 0

The difference between (2.5.30) and (2.5.23) is caused by the different equilibria
of the nonlinear spring in two cases.
34 2 Conservative Single-Degree-of-Freedom Systems

2.6 Exercise 2.6 (Rods Sliding on the Smooth Walls)

Solution: Considering the principle of angular impulse and momentum of the rod
about the axis of symmetry of the column wall, we can obtain

4l 2 
[m + m(R2 − l 2 )]θ̈ = −mg R2 − l 2 sinθ (2.6.1)
12

g(R2 − l 2 )1/2
i.e. θ̈ + sinθ = 0 (2.6.2)
R2 − 23 l 2

When the angle θ is very small, the rod oscillates slightly around the equilibrium
position and the Eq. (2.6.2) becomes linear:

g(R2 − l 2 )1/2
θ̈ + θ =0 (2.6.3)
R2 − 23 l 2

Therefore, the linear natural frequency is

g(R2 − l 2 )1/2
ω0 = (2.6.4)
R2 − 23 l 2

Expanding sinθ in (2.6.2) about θ = 0 and retaining to the third order term, we
can obtain

θ̈ + ω02 θ + αθ 3 = 0 (2.6.5)

where α = −ω02 /6. This equation is exactly the same as (2.5.1) in Exercise 2.5(a).
According to (2.5.23), the frequency-amplitude relationship of the oscillation is

3αa2 a2
ω = ω0 + = (1 − )ω0 (2.6.6)
8ω0 16

It can be seen that for a fixed amplitude a, the natural frequency ω of the nonlinear
osscillation is proportional to the natural frequency ω0 of the corresponding linear
oscillation. Therefore, the effect of the rod length l on ω can be reflected by the
ω0 ∼ l relationship, shown in Fig. 2.2. Here we rewrite (2.6.4) as

ω0 [1 − (l/R)2 ]1/2
√ = (2.6.7)
g/R 1 − 23 (l/R)2

It is shown that the natural frequency increases slowly with the increase of the
rod length l, and then decreases sharply after reaching the maximum value.
2.7 Exercise 2.7 (Period of the Osscillation Motion of a Single … 35

Fig. 2.2 Relationship

between the natural
frequency and rod length for
Exercise 2.6

2.7 Exercise 2.7 (Period of the Osscillation Motion

of a Single Pendulum-Type System Along Separatrices)

Solution: (a) The kinetic energy of the column, K, is

1 1 1 2 (R − r) 2
K= m[(R − r)θ̇]2 + mr [ θ̇ ]
2 22 r
13
= m(R − r)2 θ̇ 2 (2.7.1)
22
The kinetic energy of the cylinder, V , is

V = mg(R − r)(1 − cosθ ) (2.7.2)

Substituting them into Lagrange’s equation, we can write the differential equation
of motion of the cylinder as follows

3
m(R − r)2 θ̈ + mg(R − r)sinθ = 0
2
2g
θ̈ + sinθ = 0 (2.7.3)
3(R − r)

(b) Integrate Eq. (2.7.3) directly, we can obtain

1 2 2g
θ̇ − cosθ = E (2.7.4)
2 3(R − r)
36 2 Conservative Single-Degree-of-Freedom Systems

where E is the total energy of the cylinder. The minimum angular velocity of the
cylinder at point A is θ̇ (π ) = 0, then

2g
E= (2.7.5)
3(R − r)

and Eq. (2.7.4) becomes

1 2 2g 2g
θ̇ − cos θ = (2.7.6)
2 3(R − r) 3(R − r)

Then the angular velocity of the cylinder at the bottom (θ = 0) can be obtained
from the above equation

8g
θ̇ = (2.7.7)
3(R − r)

(c) The motion of the cylinder is controlled by the Eq. (2.7.6), i.e.,

4g
θ̇ = (1 + cosθ ) (2.7.8)
3(R − r)

Therefore, the time required for the cylinder to travel around the circular orbit, T ,
is
π π
dθ 4g
T =2 =2 [ (1 + cosθ )]−1/2 d θ
θ̇ 3(R − r)
0 0
π
3(R − r) dθ 3(R − r) θ π
= = tan
g 1 + cosθ g 2 0
0
=∞ (2.7.9)

which is infinite because the highest point A of the circular orbit is the saddle point
of the cylinder’s motion and the cylinder needs to take infinite time to approach the
saddle point.

2.8 Exercise 2.8 (Particle on a Rotating Parabola)

Solution: (a) The kinetic energy of the system T is

2.8 Exercise 2.8 (Particle on a Rotating Parabola) 37

1 1
T= m(ẋ2 + ż 2 + 2 x2 ) = m[(1 + 4p2 x2 )ẋ2 + ψ̇ 2 x2 ] (2.8.1)
2 2

where ψ̇ = . The potential energy of the system V is

V = mgz = mgpx2 (2.8.2)

The differential equation of the motion of the system can be obtained by using
Lagrange’s equation:

mψ̈x2 + 2mψ̇ ẋx = 0
m(1 + 4p2 x2 )ẍ + 4mp2 xẋ2 − mψ̇ 2 x = −2mgpx

˙
i.e.x + 2ẋ = 0 (2.8.3)

(1 + 4p2 x2 )ẍ + 4p2 xẋ2 + (2pg − 2 )x = 0 (2.8.4)

(b) Integrate (2.8.3) directly, we can obtain

√
ln + 2lnx = C ⇒ x2 = eC  H

√
x2 = H (2.8.5)

The left side of the above equation represents the angular momentum of the
system about the z-axis, so the above equation actually implies the conservation
of the angular momentum of the system. Solving  from the above equation and
substituting it into (2.8.4), we can obtain the governing equation of the system

H
(1 + 4p2 x2 )ẍ + 4p2 xẋ2 + (2pg − )x = 0 (2.8.6)
x4

(c) The Eq. (2.8.6) can be written as

H
ẍ = −[4p2 xẋ2 + (2pg − )x]/(1 + 4p2 x2 ) (2.8.7)
x4

Let v = ẋ (2.8.8)

H
then v̇ = ẍ = −[4p2 xv2 + (2pg − )x]/(1 + 4p2 x2 ) (2.8.9)
x4
from which it follows immediately that
38 2 Conservative Single-Degree-of-Freedom Systems

dv [4p2 xv2 + (2pg − xH4 )x]

=− (2.8.10)
dx v(1 + 4p2 x2 )

Make a further derivation, we can obtain

dv2 4p2 v2 + (2pg − H

x4
)
= −
dx2 1 + 4p2 x2
H
dv2 + 4p2 x2 dv2 + 4p2 v2 dx2 = −(2pg − )dx2
x4
H
dv2 + 4p2 d (x2 v2 ) = −(2pg − )dx2
x4

d (v2 + 4p2 x2 v2 ) = −d (2pgx2 + Hx−2 )

d (v2 + 4p2 x2 v2 + 2pgx2 + Hx−2 ) = 0

v2 + 4p2 x2 v2 + 2pgx2 + Hx−2 = h (2.8.11)

Now we can discuss the motion of the particle with initial condition H = H0 and
h = h0 .
(1) H0 = 0
Now Eq. (2.8.11) can be written as

h0 − 2pgx2
v2 = (2.8.12)
1 + 4p2 x2

The motion of the particle being bounded requires that

h0
h0 − 2pgx2 ≥ 0 ⇒ |x| ≤ (2.8.13)
2pg

So v2 ≤ h0 (2.8.14)

(2) H0 = 0
Now the Eq. (2.8.11) can be written as

h0 − 2pgx2 − H0 x−2
v2 = (2.8.15)
1 + 4p2 x2

The motion of the particle being bounded requires that

h0 − 2pgx2 − H0 x−2 ≥ 0 (2.8.16)

2.9 Exercise 2.9 (A Single Pendulum Rotating Uniformly Around a Plumb … 39

Fig. 2.3 Phase trajectories

of the system p = 1, g =
32.2, h = 1000, H = 12 for
Exercise 2.8

i.e.,
 
h0 − h20 − 8pgH0 h0 + h20 − 8pgH0
≤ x2 ≤ (2.8.17)
4pg 4pg

(d) Let p = 1, g = 32.2, h = 1000, H = 12, we can write the Eq. (2.8.11) as

1000 − 64.4x2 − 12x−2

v2 = (2.8.18)
1 + 4x2

According to (2.8.17), we can obtain

−3.9390 ≤ x ≤ −0.1096 and 0.1096 ≤ x ≤ 3.9390 (2.8.19)

The trajectories in the phase plane are shown in Fig. 2.3.

2.9 Exercise 2.9 (A Single Pendulum Rotating Uniformly

Around a Plumb Axis)

Solution: (a) The kinetic energy of the system T is

1
T= m(r 2 θ̇ 2 + 2 r 2 sin2 θ ) (2.9.1)
2
The potential energy of the system V is

V = −mgr cosθ (2.9.2)

40 2 Conservative Single-Degree-of-Freedom Systems

The differential equation of the motion of the pendulum can be obtained by using
Lagrange’s equation:

d 2θ
mr 2 − m2 r 2 sinθ cosθ = −mgrsinθ
dt 2

d 2θ g
i.e. − 2 sinθ cosθ + sinθ = 0 (2.9.3)
dt 2 r

Let τ = (g/r)1/2 t (2.9.4)

(2.9.3) changes to

θ̈ + (1 − cosθ )sinθ = 0 (2.9.5)

where θ̈ = ∂ 2 θ/∂τ 2 and  = 2 r/g.

(b) Integrate (2.9.5) directly, we can obtain

1 2
θ̇ + (1 − cosθ )sinθ d θ = E
2
1 2 1
i.e. θ̇ − (1 − cosθ )cosθ = E (2.9.6)
2 2

Let θ̇ |θ = 0 = θ̇0 , then

1 2 1
E= θ̇ − (1 − )
2 0 2

i.e. 21 θ̇ 2 = 21 θ̇02 − [(1 − 21 ) − (1 − 21 cosθ )cosθ ]

1 1
Let F(θ ) = 1 −  − (1 − cosθ )cosθ (2.9.7)
2 2
1 1
then θ̇ 2 = θ̇02 − F(θ ) (2.9.8)
2 2
From the Eq. (2.9.5) it is known that the equilibrium point of the system is

 ≤ 1; θ = 0, π (or − π ) (2.9.9)

 > 1 : θ = 0, π (or − π ), ±θ (2.9.10)

where θ = arccos(1/).
2.9 Exercise 2.9 (A Single Pendulum Rotating Uniformly Around a Plumb … 41

The potential energy curve and phase trajectories when  ≤ 1 are shown in
Fig. 2.4a and b. The equilibrium point θ = 0 is the center, while θ = π (or − π ) is
the saddle point. The potential curve and phase trajectories when  > 1 are shown
in Fig. 2.4c. The equilibrium points θ = 0 and θ = π (or − π ) are the saddle points,
while the equilibrium point θ = ±θ = ±arccos(1/) is the center. From above
results, it is clear that  = 1 is the bifurcation point.
(c) Assuming  > 1, the center of the system is

θ = ±θ = ±arccos(1/) (2.9.11)

(a) (b)

(c)

Fig. 2.4 a Potential energy curve and phase trajectories when  = 0.5. b Potential energy curve
and phase trajectories when  = 1 c Potential energy curve and phase trajectories when  = 3 for
Exercise 2.9
42 2 Conservative Single-Degree-of-Freedom Systems

We carry out the expansion of (2.9.5) about θ = θ = arccos(1/) to O( 3 ) and

obtain

0 = θ̈ + [1 − cos(θ + θ )]sin(θ + θ )
= θ̈ + sinθ + θ cosθ − 21 θ 2 sinθ − 16 θ 3 cosθ
− 21 [sin2θ + 2θ cos2θ − 21 (2θ )2 sin2θ − 16 (2θ )3 cos2θ ] (2.9.12)
= θ + (sin2 θ )θ + (sin2θ − 21 sinθ )θ 2
+( 23 cos2θ − 16 cosθ )θ 3
1 2 1
Let ω02 = sin2 θ , α = sin2θ − sinθ , δ = cos2θ − cosθ
2 3 6
(2.9.13)

(2.9.14) becomes

θ̈ + ω02 θ + αθ 2 + δθ 3 = 0 (2.9.14)

Let

θ = ελ v (2.9.15)

Equation (2.9.14) changes to ελ v̈ + ω02 ελ v + ε2λ δv2 + ε3λ αv3 = 0

If we take λ = 1, the above equation becomes

v̈ + ω02 v + εδv2 + ε2 αv3 = 0 (2.9.16)

Equation (2.9.16) is identical to Eq. (2.2.51), so the frequency-amplitude

relationship can be obtained from (2.2.70), i.e.

1 4δ 2
ω = ω0 − ( − 3α)a2 (2.9.17)
8ω0 3ω02

(d) Assuming  < 1, the center of the system is the origin. We carry out the expansion
of the Eq. (2.9.5) about the origin to O( 3 ) and obtain

1 1
θ̈ + [1 − (1 − θ 2 + · · · )](θ − θ 3 ) = 0
2 6

¨ + (1 − )θ + ( 2  − 1 )θ 3 = 0
i.e.θ (2.9.18)
3 6
2 1
let ω20 = 1 − , α1 = − (2.9.19)
3 6
2.10 Exercise 2.10 (A Single Pendulum Rotating Freely Around a Plumb Axis) 43

Then this equation is identical to (2.5.1) in Exercise 2.5(a), and therefore,

according to (2.5.23), the frequency-amplitude relationship is

3α1 a2
ω = ω0 + (2.9.20)
8ω0

2.10 Exercise 2.10 (A Single Pendulum Rotating Freely

Around a Plumb Axis)

Solution: (a) The kinetic energy of the system T is

1
T= m(r 2 θ̇ 2 + 2 r 2 sin2 θ ) (2.10.1)
2

where  = ψ̇. The kinetic energy of the system V is

V = −mgrcosθ (2.10.2)

The differential equations of the motion of the pendulum can be obtained by using
Lagrange’s equation:

r 2 sin2 θ 
˙ + 2r 2 sinθ cosθ θ̇ = 0
(2.10.3)
θ̈ + ( gr − 2 cosθ )sinθ = 0

The first equation can be further integrated as

√
(rsinθ )2  = H (2.10.4)
√
where H is associated with the angular momentum of the pendulum about z-axis.
So, the differential equation of motion of the system is
√
(rsinθ )2  = H (2.10.5)

g
θ̈ + ( − 2 cosθ )sinθ = 0 (2.10.6)
r

(b) Substitute (2.10.5) into (2.10.6) to obtain the equation controlling θ ,

g cotθ
θ̈ + (sinθ −  2 ) = 0 (2.10.7)
r sin θ
44 2 Conservative Single-Degree-of-Freedom Systems

where  = H /(gr 3 ).
This result is different from the one the question asks us to prove! Therefore, this
question is not solved furthermore.

2.11 Exercise 2.11 (Nonlinear Oscillation of a Slide-Slider

System)

Solution: (a) Let the instantaneous position coordinates of the mass m2 be y, then
there is a constraint relation

x2 + y2 = l 2 (2.11.1)

x
So ẏ = − ẋ (2.11.2)
y

The kinetic energy of the system T is

1 1 1 1 x2
T= m1 ẋ2 + m2 ẏ2 = m1 ẋ2 + m2 2 ẋ2
2 2 2 2 y
2
1 1 x
= m1 ẋ2 + m2 ẋ2 (2.11.3)
2 2 l 2 − x2

The potential energy of the system V is

1  1
V = m2 gy + kx2 = m2 g l 2 − x2 + kx2 (2.11.4)
2 2
Then the governing equation of the system can be obtained by using Lagrange’s
equation

m2 x2 m2 l 2 xẋ2 x
(m1 + )ẍ + + kx + m2 g =0 (2.11.5)
l 2 − x2 (l − x )
2 2 2
(l − x2 )1/2
2

(b) Let R = m2 /m1 and u = x/l, from the Eq. (2.11.5), we obtain

(m2 /m1 )(x/l)2 ẍ m2 lxẋ2 k x m2 x

[1 + ] + + + g=0
1 − (x/l) 2 l m1 l 4 (1 − x2 /l 2 )2 m1 l m1 l 2 (1 − x2 /l 2 )1/2
k Rg
(1 + Ru2 )ü + Ruu̇2 + [ + ]u = 0 (2.11.6)
m1 l(1 − u2 )1/2
2.11 Exercise 2.11 (Nonlinear Oscillation of a Slide-Slider System) 45

For |u| << 1, expanding the last term on the left side of the above equation and
retaining it to the third-order yields

Rg 3
(1 + Ru2 )ü + Ruu̇2 + ω02 u + u =0 (2.11.7)
2l
where
k Rg
ω02 = +
m1 l

(c) Let

u = εv (2.11.8)

Equation (2.11.7) changes to

Rg 3
v̈ + ω02 v + ε2 (Rv2 v̈ + Rvv̇2 + v )=0 (2.11.9)
2l
We seek an expansion of the solution with the following form

v = v0 (T0 , T1 , T2 ) + εv1 (T0 , T1 , T2 ) + ε2 v2 (T0 , T1 , T2 ) + · · · (2.11.10)

whereT0 = t, T1 = εt. Substitute (2.11.10) into (2.11.9), we obtain

Rg 3
0 = v̈ + ω02 v + ε2 (Rv2 v̈ + Rvv̇2 +
v )
2l
= [D0 + 2εD0 D1 + ε (D1 + 2D0 D2 ) + · · · ](v0 + εv1 + ε2 v2 + · · · )
2 2 2

+ ω02 (v0 + εv1 + ε2 v2 + · · · )

Rg 3
+ ε2 [Rv02 D02 v0 + Rv0 (D0 v0 )2 + v + ···]
2l 0
= D02 v0 + ω02 v0 + ε(D02 v1 + 2D0 D1 v0 + ω02 v1 )
+ ε2 [D02 v2 + 2D0 D1 v1 + (D12 + 2D0 D2 )v0 + ω02 v2
Rg 3
+ Rv02 D02 v0 + Rv0 (D0 v0 )2 + v ] + ··· (2.11.11)
2l 0
Let the coefficients of the same power of ε be zero, we can obtain

D02 v0 + ω02 v0 = 0 (2.11.12)

D02 v1 + ω02 v1 = −2D0 D1 v0 (2.11.13)

46 2 Conservative Single-Degree-of-Freedom Systems

D02 v2 + ω02 v2 = −2D0 D1 v1 − (D12 + 2D0 D2 )v0

(2.11.14)
−Rv02 D02 v0 − Rv0 (D0 v0 )2 − Rg v3
2l 0

The general solution of (2.11.12) is

v0 = Aeiω0 T0 + Ae−iω0 T0 (2.11.15)

where A = A(T1 , T2 ). Substituting the above equation into (2.11.13), we obtain

D02 v1 + ω02 v1 = −2D0 D1 (Aeiω0 T0 + Ae−iω0 T0 ) (2.11.16)

To eliminate the secular term, we need D1 A = 0. Therefore, A = A(T2 ) and the

particular solution of Eq. (2.11.13) is

v1 = 0 (2.11.17)

Substituting (2.11.15) and (2.11.17) into (2.11.14), and taking A = A(T2 ) into
account, we obtain
3Rg 2
D02 v2 + ω02 v2 = −2iω0 D2 A + 2Rω02 A2 Aeiω0 T0 − 2l
A Aeiω0 T0
(2.11.18)
+cc + NST

where cc denotes the complex conjugate term of its preceding terms and NST denotes
terms which do not generate secular terms. In order to eliminate the secular term, we
need
3g 2
2iω0 D2 A − R(2ω02 − )A A = 0 (2.11.19)
2l
1 iβ
Let A = ae (2.11.20)
2
Substitute (2.11.20) into (2.11.19), we obtain

R 3g 3
iω0 D2 a − ω0 aD2 β − (2ω02 − )a = 0 (2.11.21)
8 2l
Separates the real and imaginary parts of the above equation yields

D2 a = 0 (2.11.22)

R 3g 2
D2 β + (2ω02 − )a = 0 (2.11.23)
8ω0 2l

So a is a constant. Integrate (2.11.23) directly, we obtain

2.12 Exercise 2.12 (Nonlinear Oscillation of a Rope-Block System) 47

R 3g 2
β=− (2ω02 − )a T2 + β0 (2.11.24)
8ω0 2l

The first-order approximation of the solution is then obtained as

R 3g 2 2
u = εacos[ω0 t − (2ω02 − )a ε t + β0 ] + · · · (2.11.25)
8ω0 2l

Let ε = 1,

R 3g 2
u = acos[ω0 t − (2ω02 − )a t + β0 ] + · · · (2.11.26)
8ω0 2l

Therefore, the frequency-amplitude relationship is

R 3g 2
ω = ω0 − (2ω02 − )a (2.11.27)
8ω0 2l

2.12 Exercise 2.12 (Nonlinear Oscillation of a Rope-Block

System)

Solution: (a) Let the length of the oblique segment of the rope hanging mass m1 is
x, then there is a constraint

x2 = l 2 + y2 (2.12.1)

y
ẋ = ẏ (2.12.2)
x
The kinetic energy of the system, T , is

1 1 1 y2
T = 2 × m1 ẋ2 + m2 ẏ2 = (2m1 2 + m2 )ẏ2
2 2 2 x
1 2m1 y2 2
= (m2 + 2 )ẏ (2.12.3)
2 l + y2

Then calculate the generalized force of the system. For a given virtual displace-
ment δy, we can obtain from (2.12.1) that
y
δx = δy (2.12.4)
x
The virtual work of the system is
48 2 Conservative Single-Degree-of-Freedom Systems

2m1 gy
δW = −2m1 gδx + m2 gδy = (− + m2 g)δy
x
Therefore, the generalized force of the system is

2m1 gy 2m1 gy
Q=− + m2 g = m2 g − (2.12.5)
x (l + y2 )1/2
2

Substitute (2.12.3) and (2.12.5) into Lagrange’s equation, we can obtain the
governing equation of the system

2m1 gy 2m1 y l 2 ẏ2

m2 g − = m2 ÿ + (yÿ + ) (2.12.6)
(l 2 + y2 )1/2 l 2 + y2 l 2 + y2

(b) The equilibria of the system can be obtained from (2.12.6) by letting ẏ = ÿ = 0,
i.e.,

2m1 gy
m2 g − =0 (2.12.7)
(l 2+ y2 )1/2

then
lm2
ye =  , (assume )2m1 > m2 (2.12.8)
4m21 − m22

(c) Let u = y/l and R = m1 /m2 , from (2.12.6), we obtain

g 2m1 gy ÿ 2m1 y ẏ2

− = + (yÿ + )
l m2 l 2 (1 + y2 /l 2 )1/2 l m2 l 3 (1 + y2 /l 2 ) 1 + y2 /l 2

Finally, the differential equation of motion of the system becomes

g 2Rgu 2Ru u̇2

− = ü + (uü + ) (2.12.9)
l l(1 + u2 )1/2 1 + u2 1 + u2

(d) Let
ye ye
u= + ηor u = ue + η, where ue = (2.12.10)
l l
Substituting this into (2.12.9) and taking (2.12.7) into account, we can obtain

2Rgη 2R(ue + η) η̇2

− = η̈ + [(ue + η)η̈ + ]
l[1 + (ue + η) ]
2 1/2
1 + (ue + η)2
1 + (ue + η)2
2.12 Exercise 2.12 (Nonlinear Oscillation of a Rope-Block System) 49

The above equation can be written as

2Rg
− η[1 + (ue + η)2 ]3/2 = η̈[1 + (ue + η)2 ]2
l

+2R(ue + η){[1 + (ue + η)2 ](ue + η)η̈ + η̇2 } (2.12.11)

Expanding
 the left side of the equation for the small and finite |η|, retaining to
O η3 yields

η[1 + (ue + η)2 ]3/2

= bη(1 + ue2 + 2ue η + η2 )[1 + ue2 + 2ue η + η2 ]1/2
= bη(b2 + 2ue η + η2 )[1 + b−2 (2ue η + η2 )]1/2
1 1
= bη(b2 + 2ue η + η2 )[1 + b−2 (2ue η + η2 ) − b−4 (2ue η + η2 )2 ]
2 8
3
= [b3 η + 3bue η2 + b(1 + b−2 ue2 )η3 ]
2

where b2 = 1 + ue2 . Expanding the right-hand side of the equation and retaining to
O η3 , we obtain

η̈[1 + (ue + η)2 ]2 + 2R(ue + η){[1 + (ue + η)2 ](ue + η)η̈ + η̇2 }
= η̈(b2 + 2ue η + η2 )2 + 2R(ue + η)[(b2 + 2ue η + η2 )(ue + η)η̈ + η̇2 ]
= η̈(b4 + 4b2 ue η + 4ue2 η2 + 2b2 η2 )
+ 2R(ue + η)(b2 ue η̈ + 2ue2 ηη̈ + ue η2 η̈ + b2 ηη̈ + 2ue η2 η̈ + η̇2 )
= b4 η̈ + 4b2 ue ηη̈ + 4ue2 η2 η̈ + 2b2 η2 η̈
+ 2Rb2 ue2 η̈ + 4Rue3 ηη̈ + 2Rue2 η2 η̈ + 2Rb2 ue ηη̈ + 4Rue2 η2 η̈ + 2Rue η̇2
+ 2Rb2 ue η̈η + 4Rue2 η2 η̈ + 2Rb2 η̈η2 + 2Rη̇2 η
= (b4 + 2Rb2 ue2 )η̈ + (4b2 ue + 4Rue3 + 4Rb2 ue )η̈η + 2Rue η̇2
+ (2b2 + 4ue2 + 10Rue2 + 2Rb2 )η̈η2 + 2Rη̇2 η

Substituting these two results into (2.12.11), we obtain

2Rb3 g 6Rbue g
η̈ + η+ η2
+ 2Rb ue )
l(b4 2 2 l(b + 2Rb2 ue2 )
4

2Rue 3Rbg
+ 4 η̇2 + (1 + b−2 ue2 )η3
(b + 2Rb ue )
2 2 l(b + 2Rb2 ue2 )
4

2R (4b2 ue + 4Rue3 + 4Rb2 ue )

+ 4 η̇2 η + η̈η
(b + 2Rb ue )
2 2 (b4 + 2Rb2 ue2 )
50 2 Conservative Single-Degree-of-Freedom Systems

(2b2 + 4ue2 + 10Rue2 + 2Rb2 ) 2

+ η̈η
(b4 + 2Rb2 ue2 )
=0

Write the above equation as

η̈ + ω02 η + α1 η2 + α2 η̇2 + α3 η̈η + α4 η3 + α5 η̇2 η + α6 η̈η2 = 0 (2.12.12)

where
3
ω02 = l(b42Rb g
+2Rb2 ue2 )
6Rbue g
α1 = l(b4 +2Rb2 u2 )
e
α2 = (b4 +2Rb
2Rue
2 u2 )
e
(4b2 ue +4Rue3 +4Rb2 ue )
α3 = (b4 +2Rb2 ue2 )
(2.12.13)
α4 = 3Rbg
l(b +2Rb2 ue2 )
4 (1 + b−2 ue2 )
α5 = 2R
(b4 +2Rb2 ue2 )
(2b +4ue2 +10Rue2 +2Rb2 )
2
α6 = (b4 +2Rb2 ue2 )

Let

η = εv (2.12.14)

then Eq. (2.12.12) changes to

v̈ + ω02 v + εα1 v2 + εα2 v̇2 + εα3 v̈v + ε2 α4 v3 + ε2 α5 v̇2 v + ε2 α6 v̈v2 = 0

(2.12.15)

We seek an expansion of the following form for the solution of (2.12.15)

v = v0 (T0 , T1 , T2 ) + εv1 (T0 , T1 , T2 ) + ε2 v2 (T0 , T1 , T2 ) + · · · (2.12.16)

where T0 = t, T1 = εt. Substitute (2.12.16) into (2.12.15), we obtain

0 = v̈ + ω02 v + εα1 v2 + εα2 v̇2 + εα3 v̈v + ε2 α4 v3 + ε2 α5 v̇2 v + ε2 α6 v̈v2

= [D02 + 2εD0 D1 + ε2 (D12 + 2D0 D2 ) + · · · ](v0 + εv1 + ε2 v2 + · · · )
+ ω02 (v0 + εv1 + ε2 v2 + · · · ) + εα1 (v0 + εv1 + ε2 v2 + · · · )2
+ εα2 [(D0 + εD1 + ε2 D2 + · · · )(v0 + εv1 + ε2 v2 + · · · )]2
+ εα3 {[D02 + 2εD0 D1 + ε2 (D12 + 2D0 D2 ) + · · · ](v0 + εv1 + ε2 v2 + · · · )}
× (v0 + εv1 + ε2 v2 + · · · ) + ε2 α4 (v0 + εv1 + ε2 v2 + · · · )3
+ ε2 α5 [(D0 + εD1 + ε2 D2 + · · · )(v0 + εv1 + ε2 v2 + · · · )]2
2.12 Exercise 2.12 (Nonlinear Oscillation of a Rope-Block System) 51

× (v0 + εv1 + ε2 v2 + · · · )
+ ε2 α6 {[D02 + 2εD0 D1 + ε2 (D12 + 2D0 D2 ) + · · · ](v0 + εv1 + ε2 v2 + · · · )}
× (v0 + εv1 + ε2 v2 + · · · )2
= D02 v0 + ω02 v0
+ ε[D02 v1 + 2D0 D1 v0 + ω02 v1 + α1 v02 + α2 (D0 v0 )2 + α3 v0 D02 v0 ]
+ ε2 [D02 v2 + 2D0 D1 v1 + (D12 + 2D0 D2 )v0 + ω02 v2 + 2α1 v0 v1
+ α2 D0 v0 (D0 v1 + D1 v0 ) + α3 v0 (D12 v1 + 2D0 D2 v0 )
+ α4 v03 + α5 v0 (D0 v0 )2 + α6 v02 D02 v0 ] + · · · (2.12.17)

Let the coefficients of the same power of ε be zero to obtain

D02 v0 + ω02 v0 = 0 (2.12.18)

D02 v1 + ω02 v1 = −2D0 D1 v0 − α1 v02 − α2 (D0 v0 )2 − α3 v0 D02 v0 (2.12.19)

D02 v2 + ω02 v2 = −2D0 D1 v1 − (D12 + 2D0 D2 )v0 − 2α1 v0 v1

−α2 D0 v0 (D0 v1 + D1 v0 ) − α3 v0 (D12 v1 + 2D0 D2 v0 ) (2.12.20)
−α4 v03 + α5 v0 (D0 v0 )2 − α6 v02 D02 v0

The general solution of Eq. (2.12.18) is

v0 = Aeiω0 T0 + Ae−iω0 T0 (2.12.21)

where A = A(T1 , T2 ). Substituting the above equation into (2.12.19), we obtain

D02 v1 + ω02 v1 = −2iω0 D1 Aeiω0 T0 − (α1 − α2 ω02 − α3 ω02 )AA

+ (−α1 − α2 ω02 + α3 ω02 )A2 e2iω0 T0 + cc (2.12.22)

To eliminate secular terms, we need D1 A = 0. Therefore, A = A(T2 ) and the

particular solution of (2.12.22) is

v1 = B1 AA + B2 A2 e2iω0 T0 + cc (2.12.23)

α1 − α2 ω02 − α3 ω02 −α1 − α2 ω02 + α3 ω02

where B1 = − , B2 = − (2.12.24)
ω02 3ω02

Substituting (2.12.21) and (2.12.23) into (2.12.20) and taking A = A(T2 ) into
account, we can obtain
52 2 Conservative Single-Degree-of-Freedom Systems

D02 v2 + ω02 v2 = (−2iω0 D2 A − 2α1 B1 A2 A − 2α1 B2 A2 A − 2α2 ω02 B2 A2 A

−3α4 A2 A − α5 ω02 A2 A + 3α6 ω02 A2 A)eiω0 T0 + cc + NST
(2.12.25)

where cc denotes the complex conjugate term of its preceding terms and NST denotes
terms which do not generate secular terms. In order to eliminate secular terms, it is
necessary to have

2iω0 D2 A + (2α1 B1 + 2α1 B2 + 2α2 ω02 B2 + 3α4 + α5 ω02 − 3α6 ω02 )A2 A = 0
(2.12.26)
1 iβ
Let A= ae (2.12.27)
2
Substitute (2.12.27) into (2.12.26), we obtain

iω0 D2 a − ω0 aD2 β
1
+ (2α1 B1 + 2α1 B2 + 2α2 ω02 B2 + 3α4 + α5 ω02 − 3α6 ω02 )a3 = 0 (2.12.28)
8
Separate the real and imaginary parts of the above equation yields

D2 a = 0 (2.12.29)

1
−ω0 aD2 β + (2α1 B1 + 2α1 B2 + 2α2 ω02 B2 + 3α4 + α5 ω02 − 3α6 ω02 )a3 = 0
8
(2.12.30)

So a is a constant. Conduct direct integration of (2.12.30), we obtain

D 2
β= a T2 + β0 (2.12.31)
8ω0

where D = 2α1 B1 + 2α1 B2 + 2α2 ω02 B2 + 3α4 + α5 ω02 − 3α6 ω02 (2.12.32)

Combining the above results, the first-order approximation of the solution of the
original equation can be written as

D 2 2
η = εacos(ω0 t + a ε t + β0 ) + · · · (2.12.33)
8ω0

Let ε = 1, we can obtain

D 2
η = acos(ω0 t + a t + β0 ) + · · · (2.12.34)
8ω0
2.13 Exercise 2.13 (Single Pendulum Attached … 53

Therefore, the frequency-amplitude relationship of the oscilliation is

D 2
ω = ω0 + a (2.12.35)
8ω0

2.13 Exercise 2.13 (Single Pendulum Attached

with a Rolling-Without-Slipping Circular Wheel I)

Solution: (a) The kinetic energy of the system is

1 1
T= m(l θ̇ cosθ − θ̇ r)2 + m(l θ̇ sinθ )2
2 2
1
= m(l 2 + r 2 − 2rlcosθ )θ̇ 2 (2.13.1)
2
The potential energy of the system is

V = −mglcosθ (2.13.2)

Using Lagrange’s equation, we can obtain the differential equation of motion of

the system

(l 2 + r 2 − 2rlcosθ )θ̈ + rl θ̇ 2 sinθ + glsinθ = 0 (2.13.3)

(b) Expand the Eq. (2.13.3) and retain to θ 3 yields

gl rl rl gl
θ̈ + θ+ θ 2 θ̈ + θ̇ 2 θ − θ3 = 0 (2.13.4)
(l − r) 2
(l − r)2
(l − r) 2
6(l − r)2
gl rl
Let ω02 = , α= (2.13.5)
(l − r)2
(l − r)2
ω02 3
then θ̈ + ω02 θ + αθ 2 θ̈ + α θ̇ 2 θ − θ =0 (2.13.6)
6
Let

θ = εv (2.13.7)

Equation (2.13.6) changes to

54 2 Conservative Single-Degree-of-Freedom Systems

ω02 3
v̈ + ω02 v + ε2 αv2 v̈ + ε2 α v̇2 v − ε2 v =0 (2.13.8)
6
We seek an expansion of the following form for the solution of (2.13.8)

v = v0 (T0 , T1 , T2 ) + εv1 (T0 , T1 , T2 ) + ε2 v2 (T0 , T1 , T2 ) + · · · (2.13.9)

where T0 = t, T1 = εt. Substitute (2.13.9) into (2.13.8), we obtain

ω02 3
0 = v̈ + ω02 v + ε2 αv2 v̈ + ε2 α v̇2 v − ε2
v
6
= [D02 + 2εD0 D1 + ε2 (D12 + 2D0 D2 ) + · · · ](v0 + εv1 + ε2 v2 + · · · )
+ ω02 (v0 + εv1 + ε2 v2 + · · · )
ω02 3
+ ε2 [αv02 D02 v0 + αv0 (D0 v0 )2 − v + ···]
6 0
= D02 v0 + ω02 v0 + ε(D02 v1 + 2D0 D1 v0 + ω02 v1 )
+ ε2 [D02 v2 + 2D0 D1 v1 + (D12 + 2D0 D2 )v0 + ω02 v2
ω02 3
+ αv02 D02 v0 + αv0 (D0 v0 )2 − v ] + ··· (2.13.10)
6 0
Let the coefficients of the same power of ε be zeros to obtain

D02 v0 + ω02 v0 = 0 (2.13.11)

D02 v1 + ω02 v1 = −2D0 D1 v0 (2.13.12)

D02 v2 + ω02 v2 = −2D0 D1 v1 − (D12 + 2D0 D2 )v0

ω2 (2.13.13)
−αv02 D02 v0 − αv0 (D0 v0 )2 + 60 v03

The general solution of Eq. (2.13.11) is

v0 = Aeiω0 T0 + Ae−iω0 T0 (2.13.14)

whereA = A(T1 , T2 ). Substituting (2.13.14) into (2.13.12), we obtain

D02 v1 + ω02 v1 = −2D0 D1 (Aeiω0 T0 + Ae−iω0 T0 ) (2.13.15)

To eliminate secular terms, we need D1 A = 0, so A = A(T2 ). And the particular

solution of (2.2.59) is

v1 = 0 (2.13.16)
2.13 Exercise 2.13 (Single Pendulum Attached … 55

Substituting (2.11.15) and (2.11.17) into (2.11.14), and taking A = A(T2 ) into
account, we can obtain

D02 v2 + ω02 v2 = −2iω0 D2 Aeiω0 T0 + 2αω02 A2 Aeiω0 T0

1
+ ω02 A2 Aeiω0 T0 + cc + NST (2.13.17)
2
where cc denotes the complex conjugate term of its preceding terms and NST denotes
terms which do not generate secular terms. In order to eliminate secular terms, it is
necessary to have

1
−2iω0 D2 A + 2αω02 A2 A + ω02 A2 A = 0 (2.13.18)
2
1 iβ
Let A = ae (2.13.19)
2
Substitute (2.11.20) into (2.11.19), we obtain

1
iω0 D2 a − ω0 aD2 β − ω02 ( α + 4)a3 = 0 (2.13.20)
4
Separate the real and imaginary parts of the above equation yields

D2 a = 0 (2.13.21)

1
D2 β − ω0 ( α + 4)a2 = 0 (2.13.22)
4
So a is a constant. By integrating (2.11.23), we can obtain

1
β = ω0 ( α + 4)a2 T2 + β0 (2.13.23)
4
Combining the above results, the first-order approximate Solution is obtained as

1
u = εacos[ω0 t + ω0 ( α + 4)a2 ε2 t + β0 ] + · · · (2.13.24)
4
Let ε = 1, we can obtain

1
u = acos[ω0 t + ω0 ( α + 4)a2 t + β0 ] + · · · (2.13.25)
4
Then the frequency-amplitude relationship of the oscillation is
56 2 Conservative Single-Degree-of-Freedom Systems

1
ω = ω0 + ω0 ( α + 4)a2 (2.13.26)
4

2.14 Exercise 2.14 (Single Pendulum Attached

with a Rolling-Without-Slipping Circular Wheel II)

Solution: (a) The kinetic energy of the system is

1 1
T= m(l θ̇ cosθ − θ̇ r)2 + m(l θ̇ sinθ )2
2 2
1
= m(l 2 + r 2 − 2rlcosθ )θ̇ 2 (2.14.1)
2
The potential energy of the system is

1
V = −mglcosθ + k(rθ )2 (2.14.2)
2
Using Lagrange’s equation, we can obtain the differential equation of the motion
of the system

m(l 2 + r 2 − 2rlcosθ )θ̈ + mrl θ̇ 2 sinθ + mglsinθ + kr 2 θ = 0 (2.14.3)

(b) Expand the Eq. (2.13.3) and retain to θ 3 yields

gl + kr 2 /m rl rl gl
θ̈ + θ+ θ 2 θ̈ + θ̇ 2 θ − θ3 = 0 (2.14.4)
(l − r)2 (l − r)2 (l − r)2 6(l − r)2
gl + kr 2 /m rl
Let ω02 = , α= (2.14.5)
(l − r) 2
(l − r)2

¨ + ω02 θ + αθ 2 θ̈ + α θ̇ 2 θ − 1 ω02 θ 3 = 0
thenθ (2.14.6)
6
This equation is the same as the governing equation in Exercise 2.13, except that
the value of ω02 is taken differently. Therefore, the relationship between the oscillation
frequency ω and the amplitude a is the same as in Exercise 2.13, i.e.,

1
ω = ω0 + ω0 ( α + 4)a2 (2.14.7)
4
2.16 Exercise 2.16 (Simplified Model for Buckling Analysis of Columns) 57

2.15 Exercise 2.15 (Single Pendulum Problem

with Inelastic Collision with an Inclined Wall)

Solution: (a) When there is no collision, the whole motion of the system is controlled

by the equation θ + θ = 0, so that its energy equation is

θ ‘2 + θ 2 = 2E (2.15.1)

where E is the total energy. Therefore, the trajectories in the θ ∼ θ̇ phase plane are
a family of circles with different initial conditions (E).
(b) When there is an inelastic collision, let the coefficient of restitution be μ < 1,
then the governing equation of the system is

θ = α : θ “ + sin θ = 0
θ ‘ (α) (2.15.2)
θ = α, θ̇ (α) < 0 : −θ2 ‘ (α) = μ
1

where θ1 , θ2  associated with the angular velocity of the pendulum before and after
the collision, respectively. Then the trajectories of the system are governed by
√
θ = α : θ ‘ = ± 2(E + cos θ )
(2.15.3)
θ = α, θ̇ (α) < 0 : θ2‘ (α) = −μθ1‘ (α)

Figure 2.5a–c present the trajectories of the sysstem for (1) α > 0, (2) α = 0 and
(3) α < 0, respectively.
Figure 2.5a and b (α ≥ 0): the ball will always impact the wall with energy loss
and velocity decreasing, and finally rest against the wall.
Figure 2.5c (α < 0): the ball will impact the wall at the beginning, then the speed
will decrease and the energy will be lost, finally the speed will be zero when the ball
reaches the wall.

2.16 Exercise 2.16 (Simplified Model for Buckling Analysis

of Columns)

Solution: (a) The kinetic energy of the system is

1 2
T= mẋ (2.16.1)
2
The potential energy of the system is
58 2 Conservative Single-Degree-of-Freedom Systems

(a) (b)

(c)

Fig. 2.5 a Phase trajectories of the ball at α > 15◦ (θ0 = 30◦ , θ0  = 0, μ = 0.8). b Phase
◦
trajectories of the ball at α = 0 (θ0 = 30◦ , θ0  = 0, μ = 0.8). c Phase trajectories of the ball at
◦ ◦
α > −15 (θ0 = 30 , θ0  = 0, μ = 0.8) for Exercise 2.16

x 
V = Fspring dx − 2 × P(l − l 2 − x2 )
0
x 
= Fspring dx + 2P l 2 − x2 − 2Pl (2.16.2)
0

Substituting the kinetic energy and potential energy into the Lagrange’s equation,
we obtain
∂V 2Px
mẍ = − = −Fspring + √
∂x l 2 − x2

So mẍ + Fspring − √2Px

l 2 −x2
=0
2.16 Exercise 2.16 (Simplified Model for Buckling Analysis of Columns) 59

2Px
i.e. mẍ + k1 x + k3 x3 − √ + ··· = 0 (2.16.3)
l 2 − x2

Expanding the fourth term on the left-hand side of (2.16.4) and keeping it to O(x 3 ),
we obtain
2P P
mẍ + k1 x + k3 x3 − ( x + 3 x3 ) = 0
l l
2P P
i.e. mẍ + (k1 − )x + (k3 − 3 )x3 = 0 (2.16.4)
l l
1 2P 1 P
Let α1 = (k1 − ), α3 = (k3 − 3 ) (2.16.5)
m l m l
then (2.16.4) becomes

ẍ + α1 x + α3 x3 = 0 (2.16.6)

(b) In the Eq. (2.16.6) let

 √
x α , when α1 ≥ 0 α3 l 2
u = , ω0 = √ 1 , τ = ω0 t, α = 2 (2.16.7)
l −α1 , when α1 < 0 ω0

then (2.16.6) becomes


α1 ≥ 0 : u + u + αu3 = 0 (2.16.8)


α1 < 0 : u − u + αu3 = 0 (2.16.9)

After integration, the energy equation and the phase trajectories are obtained as
1
u‘2 + V (u) = E, where V (u) = 21 u2 + 41 αu4
α1 ≥ 0 : 2 √ (2.16.10)
u‘ = ± 2(E − V (u))
1
u‘2 + V (u) = E, where V (u) = − 21 u2 + 41 αu4
α1 < 0 : 2 √ (2.16.11)
u‘ = ± 2(E − V (u))

From (2.16.10) and (2.16.11) the potential energy curve V (u) ∼ u and the phase
trajectories can be calculated and plotted, as shown in Fig. 2.6a and b. The equilibrium
point x = 0 is the center when α1 > 0 while the saddle point when α1 < 0. Therefore,
α1 = 0 is the bifurcation point of the system, i.e., the column is buckled under this
condition. Thus, the buckling load Pb can be obtained from Eq. (2.16.5)
60 2 Conservative Single-Degree-of-Freedom Systems

(a) (b)

Fig. 2.6 a Potential energy curve and phase trajectories (α1 > 0, α = 0.2). b Potential energy
curve and phase trajectories (α1 < 0, α = 0.2) for Exercise 2.16

1
Pb = k1 l (2.16.12)
2

2.17 Exercise 2.17 (Rods for Pure Rolling on a Fixed

Cylindrical Surface)

Solution: (a) Assume that the rod is balanced at the vertex of the cylindrical surface.
The kinetic energy of the rod is

1 1 2
T= ( ml + mr 2 θ 2 )θ̇ 2 (2.17.1)
2 12
The potential energy of the rod is

V = mg(rcosθ + rθ sinθ ) (2.17.2)

Substituting the kinetic energy and potential energy into the Lagrange’s equation,
we obtain
1 2
( ml + mr 2 θ 2 )θ̈ + mr 2 θ θ̇ 2 = −mgrθ cosθ
12
That is, the differential equation of motion of the rod is
2.17 Exercise 2.17 (Rods for Pure Rolling on a Fixed Cylindrical Surface) 61

1 2
( l + r 2 θ 2 )θ̈ + r 2 θ θ̇ 2 + grθ cosθ = 0 (2.17.3)
12

(b) We expand the Eq. (2.17.3) around the equilibrium point θ = 0 and expand it,
retaining to O(θ 3 ), to obtain

1 2 1
( l + r 2 θ 2 )θ̈ + r 2 θ θ̇ 2 + grθ − grθ 3 = 0 (2.17.4)
12 2
1
¨
well sorted θ + ω02 θ + αθ 2 θ̈ + αθ θ̇ 2 − ω02 θ 3 = 0 (2.17.5)
2

12rg 12r 2
where ω02 = , α = (2.17.6)
l2 l2
Let

θ = εv (2.17.7)

Equation (2.17.4) changes to

ω02 3
v̈ + ω02 v + ε2 αv2 v̈ + ε2 α v̇2 v − ε2 v =0 (2.17.8)
2
We seek an expansion of solution with the following form near the singularity
v=0

v = v0 (T0 , T1 , T2 ) + εv1 (T0 , T1 , T2 ) + ε2 v2 (T0 , T1 , T2 ) + · · · (2.17.9)

where T0 = t, T1 = εt. Substitute (2.17.9) into (2.17.8), we obtain

ω02 3
0 = v̈ + ω02 v + ε2 αv2 v̈ + ε2 α v̇2 v − ε2
v
2
= [D02 + 2εD0 D1 + ε2 (D12 + 2D0 D2 ) + · · · ](v0 + εv1 + ε2 v2 + · · · )
+ ω02 (v0 + εv1 + ε2 v2 + · · · )
ω02 3
+ ε2 [αv02 D02 v0 + αv0 (D0 v0 )2 − v + ···]
2 0
= D02 v0 + ω02 v0 + ε(D02 v1 + 2D0 D1 v0 + ω02 v1 )
+ ε2 [D02 v2 + 2D0 D1 v1 + (D12 + 2D0 D2 )v0 + ω02 v2
ω02 3
+ αv02 D02 v0 + αv0 (D0 v0 )2 − v ] + ··· (2.17.10)
2 0

Retaining the equation to O(ε2 ) and equate the coefficients of the same power of
ε to be zero gives
62 2 Conservative Single-Degree-of-Freedom Systems

D02 v0 + ω02 v0 = 0 (2.17.11)

D02 v1 + ω02 v1 = −2D0 D1 v0 (2.17.12)

D02 v2 + ω02 v2 = −2D0 D1 v1 − (D12 + 2D0 D2 )v0

ω02 3
− αv02 D02 v0 − αv0 (D0 v0 )2 + v (2.17.13)
2 0
The solution of Eq. (2.17.11) is

v0 = Aeiω0 T0 + Ae−iω0 T0 (2.17.14)

where A = A(T1 , T2 ). Substituting the above equation into the Eq. (2.17.12), we
obtain

D02 v1 + ω02 v1 = −2D0 D1 (Aeiω0 T0 + Ae−iω0 T0 ) (2.17.15)

To eliminate the secular term, we need D1 A = 0, so A = A(T2 ). Then the solution

of (2.17.15) is

v1 = 0 (2.17.16)

Substitute (2.17.14) and (2.17.16) into (2.17.13), and take A = A(T2 ) into account,
we obtain

D02 v2 + ω02 v2 = −2iω0 D2 Aeiω0 T0 + 2αω02 A2 Aeiω0 T0

3ω02 2 iω0 T0
+ A Ae + cc + NST (2.17.17)
2
where cc denotes the complex conjugate term of its preceding terms and NST denotes
terms which do not generate secular terms. In order to eliminate the secular term, it
is necessary to have

3ω02 2
−2iω0 D2 A + 2αω02 A2 A + A A=0 (2.17.18)
2
1 iβ
Let A= ae (2.17.19)
2
Put (2.17.19) into (2.17.18), we obtain

1
iω0 D2 a − ω0 aD2 β − ω02 ( α + 12)a3 = 0 (2.17.20)
4
2.18 Exercise 2.18 (Geometrically Nonlinear System Formed by a Linear … 63

Separate the real and imaginary parts of the above equation yields

D2 a = 0 (2.17.21)

1
D2 β − ω0 ( α + 12)a2 = 0 (2.17.22)
4
So a is a constant. Integrate (2.17.22) directly, we obtain

1
β = ω0 ( α + 12)a2 T2 + β0 (2.17.23)
4
Combining the above results, the first-order approximation of the solution is
obtained as
1
u = εacos[ω0 t + ω0 ( α + 12)a2 ε2 t + β0 ] + · · · (2.17.24)
4
Let ε = 1, we can obtain

1
u = acos[ω0 t + ω0 ( α + 12)a2 t + β0 ] + · · · (2.17.25)
4
Therefore, the frequency-amplitude relationship of the oscillation is

1
ω = ω0 + ω0 ( α + 12)a2 (2.17.26)
4

2.18 Exercise 2.18 (Geometrically Nonlinear System

Formed by a Linear Spring and a Mass Block)

Solution: (a) The kinetic energy of the system is

1 2
T= mẋ (2.18.1)
2
The potential energy of the system is

1
V = k[(x2 + l 2 )1/2 − l ]2 (2.18.2)
2
Substituting the kinetic energy and potential energy into the Lagrange’s equation,
the differential equation of motion of the system is
64 2 Conservative Single-Degree-of-Freedom Systems

mẍ + kx(x2 + l 2 )−1/2 [(x2 + l 2 )1/2 − l ] = 0 (2.18.3)

The equation can be written as

ẍ k x x2 x2 l
+ (1 + 2 )−1/2 [(1 + 2 )1/2 − ] = 0
l ml l l l
Let

k x l
2ω2 = , u = , L = , τ = ωt
m l l
The equation becomes

ü + 2u(1 + u2 )−1/2 [(1 + u2 )1/2 − L] = 0 (2.18.4)

2Lu
or ü + 2u − √ =0 (2.18.5)
1 + u2

Note that the derivatives are obtained for the dimensionless time variable τ , i.e.
u̇ = du/d τ .
(b) Integrate the Eq. (2.18.5), we can obtain the energy equation

1 2 
u̇ + V (u) = E, V (u) = u2 − 2L 1 + u2 (2.18.6)
2

Sou̇ = ± 2[E − V (u)] (2.18.7)

The potential energy curve and the phase trajectories are shown in Fig. 2.7a–c.
The equilibrium point u = 0 is a center when L ≤ 1 while a saddle point when L > 1
is unstable; two new equilibrium points are created at the same time.
(c) Expand the Eq. (2.18.5), retaining to O(u3 ), to obtain

1
ü + 2u − 2Lu(1 − u2 + · · · ) = 0
2

ü + 2(1 − L)u + Lu3 = 0 (2.18.8)

For L = 1,

ü + u3 = 0 when L = 1 (2.18.9)

(d) From (2.18.9), we obtain

2.18 Exercise 2.18 (Geometrically Nonlinear System Formed by a Linear … 65

(a) (b)

(c)

Fig. 2.7 Potential energy curves and solution trajectories when a L = 0.5, b L = 1 and c L = 1.5
for Exercise 2.18(b)

d u̇ 1 2 1 4
dτ = − , u̇ + u = E (2.18.10)
u3 2 4
If the movement starts at −u0 ,

1 4 √ u3 du
E= u0 , d u̇ = − 2 
4 (u4 − u4 )
0

Substitute this into (2.18.10) and make integration of it to get

66 2 Conservative Single-Degree-of-Freedom Systems

√ 
u
du
τ= 2  (2.18.11)
−u0 (u04 − u4 )

Let u = −u0 cosϕ (2.18.12)

then the Eq. (2.18.11) becomes

√ θ θ
2 sinϕd ϕ 1 dϕ 1 1
τ=  =   F( , θ ) (2.18.13)
u0 1 − cos ϕ
4 u0 1 − 21 sin2 ϕ u0 2
0 0

where cosθ = −u/u0 ;F(κ, θ ) are the first class of elliptic integrals defined as

θ
dϕ
F(κ, θ ) = 
1 − κ 2 sin2 ϕ
0

Thus, the period of the system T is

4 1 π 4 7.4448
T= F( , ) = × 1.8612 ≈ (2.18.14)
u0 2 2 u0 u0

2.19 Exercise 2.19 (Solving Pure Cubic Nonlinear Systems

by Harmonic Balance Method)

Solution: (a) Substituting the one-term expansion

u = u0 cos(ωτ + β)  u0 cosψ

into the differential equation of motion of the system, we obtain

0 = −ω2 u0 cosψ + u03 cos3 ψ

1
= −ω2 u0 cosψ + u03 (3cosψ + cos3ψ)
4
3 1
= (−ω2 + u02 )u0 cosψ + u03 cos3ψ
4 4
Equating the coefficient of cosΨ to zero, we obtain
2.20 Exercise 2.20 (Solving Purely Fifth-Order Nonlinear Systems … 67
√
3
ω= u0
2
Therefore, the period of oscillation of the system is

2π 4π 7.2552
T= = √ ≈ (2.19.1)
ω u0 3 u0

From Eq. (2.18.14), the relative error of this result is

7.4448 − 7.2552
error = = 2.5%
7.4448

(b) By expanding the integrand integrating term by term, refer to (f) in Exercise 2.18,
we obtain
1 1 1 3 1 5 1
= 1 + ( sin2 φ) + ( sin2 φ)2 + ( sin2 φ)3 + · · ·
(1 − 1
2
sin2 φ)1/2 2 2 8 2 16 2
1 3 5
= 1 + sin2 φ + sin4 φ + sin6 φ + · · ·
4 32 128
Substituting this into the equation (f) in Exercise 2.18, the period of oscillation
of the system can be obtained as

π/2
4 1 3 5
T= (1 + sin2 ϕ + sin4 ϕ + sin6 ϕ + · · · )d ϕ (2.19.2)
u0 4 32 128
0

Integrating the first three terms of (2.19.2) gives


π/2
T= 4
u0
(1 + 41 sin2 φ + 3
32
sin4 φ)d φ
0
4 π π (2.19.3)
= ( + 41
u0 2
· 4
+ 3
32
· 3π
16
) = 297π
128u0
≈ 7.2895
u0

This result is closer to the exact value (2.18.14) than (2.19.1).

2.20 Exercise 2.20 (Solving Purely Fifth-Order Nonlinear

Systems by Harmonic Balance Method)

Solution: Substituting the one-term expansion

68 2 Conservative Single-Degree-of-Freedom Systems

u = acos(ωt + β)  acosψ

Into the differential equation of motion of the system, we obtain

0 = −ω2 acosψ + a5 cos5 ψ

1
= −ω2 acosψ + a5 (cos5ψ + 5cos3ψ + 10cosψ)
16
5
= (−ω2 + a4 )acosψ + · · ·
8
Equating the coefficient of cosΨ to zero, we obtain

5 2
ω= a (2.20.1)
8

2.21 Exercise 2.21 (Solving Pure Cubic Nonlinear Systems

by Equivalent Linearization)

Solution: The average of integrated square of the error over a time interval T is given
by

T
1
e= (λx − x3 )2 d τ  < (λx − x3 )2 ≥ λ2 x2 − 2λ x4 + x6
T
0

Minimize this error with respect to λ yields

de
= 2λ < x2 > −2 < x4 > = 0
dλ
T 
2π
cos4 (ωτ +β)d τ cos4 ψd ψ
<x4 >
λ= <x2 >
= u02 0T = u02 2π
0

cos2 (ωτ +β)d τ cos2 ψd ψ
0 0

= u02 ( 38 · 2π )/(( 21 · 2π )) = 43 u02

√ √
So ω = λ= 3
2 0
u
2.22 Exercise 2.22 (Show Least Residual Value Method and Galerkin Method … 69

2.22 Exercise 2.22 (Show Least Residual Value Method

and Galerkin Method for Solving Pure Cubic
Nonlinear Systems)

Solution: (a) Substituting the hypothetical solution (b) into (a), the residual can be
obtained as

R = −ω2 u0 cos(ωτ + β) + u03 cos3 (ωτ + β)

1
= −ω2 u0 cos(ωτ + β) + u03 [3cos(ωτ + β) + cos(3ωτ + 3β)]
4
3 3 1
= ( u0 − ω u0 )cos(ωτ + β) + u03 cos(3ωτ + 3β)
2
4 4
Thus, the average of integrated square of the error over a time interval 2π/ω is


2π/ω
 1 3 
R = {( u03 − ω2 u0 )2
2
cos2 (ωτ + β)d τ
T 4
0

2π/ω
1 3
+ u03 ( u03 − ω2 u0 ) cos(ωτ + β)cos(3ωτ + 3β)d τ
2 4
0

2π/ω
1 6
+ u cos2 (3ωτ + 3β)d τ }
16 0
0
1 3 1
= ( u03 − ω2 u0 )2 + u06 (2.22.1)
2 4 32

If < R2 > is minimized with respect to u0 (assuming u0 is not zero), we have

d <R2 >
du0
= ( 43 u03 − ω2 u0 )( 49 u02 − ω2 ) + 3 5
u
16 0
= 16 u0 − 49 ω2 u03 − 43 ω2 u03 + ω4 u0 +
27 5 3 5
u
16 0
= 15 u5 − 3ω2 u03 + ω4 u0 = 0
8 0

So

1 √
ω= ( 6 ± 6)u0
2

The root with the positive sign maximizes < R2 > and must be discarded. The
other root is not in agreement with the result obtained by the method of harmonic
balance and equivalent linearization in Exercises 2.20 & 2.21.
70 2 Conservative Single-Degree-of-Freedom Systems

(b) Instead of minimizing < R2 > with respect to u0 , if < R2 > is minimized with
respect to ω, we have

d < R2 > 3
= −2ωu0 ( u03 − ω2 ) = 0
dω 4
So
√
3
ω= u0
2

(c) Instead of minimizing < R2 >, R is made orthogonal to the assumed solution,
i.e.,

T
0 =< Ru0 cos(ωτ + β) >= 1
T
Ru0 cos(ωτ + β)d τ
0
2π 2π
= 1
[u ( 3 u3
Tω 0 4 0
− ω2 u0 ) cos2 ψd ψ + 41 u04 cosψcos3ψd ψ
0 0
= 21 ( 43 u03 − ω2 u0 )
√
So ω = 23 u0
This result is identical to that obtained by the harmonic balance method and the
equivalent linearization method shown in Exercises 2.20, 2.21. This method is the
Galerkin method.

2.23 Exercise 2.23 (Discuss the Possibility of Solving Pure

Cubic Nonlinear Systems by the Method of Multiple
Scales)

Solution: Let

u = εv

The equation u + u3 = 0 changes to

v + ε2 v3 = 0 (2.23.1)

We seek an expansion of the solution with the following form near the singularity
v=0

v = v0 (T0 , T1 , T2 ) + εv1 (T0 , T1 , T2 ) + ε2 v2 (T0 , T1 , T2 ) + · · · (2.23.2)

2.24 Exercise 2.24 (Examine the Relationship Between the Period … 71

where Tn = εn t. Substitute (2.13.9) into (2.23.1), we obtain

0 = v̈ + ε2 v3
= [D02 + 2εD0 D1 + ε2 (D12 + 2D0 D2 ) + · · · ](v0 + εv1 + ε2 v2 + · · · )
+ ε2 (v0 + εv1 + ε2 v2 + · · · )3
= D02 v0 + ε(D02 v1 + 2D0 D1 v0 )
+ ε2 [D02 v2 + 2D0 D1 v1 + (D12 + 2D0 D2 )v0 + v03 ] + · · · (2.23.3)

Retaining the above equation to the order of ε2 and equating coefficients of the
same power of ε gives

D02 v0 = 0 (2.23.4)

D02 v1 = −2D0 D1 v0 (2.23.5)

D02 v2 = −2D0 D1 v1 − (D12 + 2D0 D2 )v0 − v03 (2.23.6)

In order not to generate secular terms, this set of equations has only zero solutions.
Therefore, we do not get the expression for v. The reason is that, according
to the above scheme, the ε0 order in Eq. (2.23.4) is not an oscillational equation,
and therefore, the final oscillational solution of v cannot be formed. Therefore, the

equation u + u3 = 0 cannot be solved by the the method of multiple scales.

2.24 Exercise 2.24 (Examine the Relationship Between

the Period and the Amplitude of Oscillation)

Solution: (a) As known from Exercise 2.18, the differential equation of motion of
the system is

u + 2u(1 + u2 )−1/2 [(1 + u2 )1/2 − L] = 0

And from the answer to Exercise 2.18(c), the above equation has been expanded
to O(u3 ) as

u + 2(1 − L)u + Lu3 = 0

When L = 1/2 is used, the above equation becomes

 1 1
u + u + u3 = 0, when L = (2.24.1)
2 2
72 2 Conservative Single-Degree-of-Freedom Systems

By integrating (2.24.1), we obtain

1 ‘2 1 2 1 4
u + u + u =E
2 2 8
When the initial condition is u = −u0 , u = 0

1 2 1 4
E= u + u
2 0 8 0
 
1 1
So u =
‘
(u02 − u2 ) + (u04 − u4 ) = (u02 − u2 )(4 + u02 + u2 ) (2.24.2)
4 2

The time from the initial state of motion to any state is

u u
du du
τ= =2  (2.24.3)
u‘ (u02 − u2 )(4 + u02 + u2 )
−u0 −u0

(b) Let u = −u0 cosϕ, the Eq. (2.24.3) becomes

θ θ
u0 sinϕd ϕ dϕ
τ =2  =2 
0 u02 (1 − cos2 ϕ)(4 + u02 + u02 cos2 ϕ) 0 (4 + 2u02 − u02 sin2 ϕ)

θ
2 dϕ
= 
4 + 2u02 (1 − k 2 sin2 ϕ)
0

θ
2 dϕ
i.e. τ =   (2.24.4)
4 + 2u02 (1 − k 2 sin2 ϕ)
0

where k 2 = u02 /(4 + 2u02 ). Therefore, the period of oscillation of the system is

π/2
8 dϕ
T=  (2.24.5)
4 + 2u02 (1 − k 2 sin2 ϕ)
0

Numerical integration is performed for the Eq. (2.24.5) and T ∼ u0 curve is

plotted in Fig. 2.8.
2.25 Exercise 2.25 (Comparison of Equivalent Linearization Methods … 73

Fig. 2.8 Variation of period

T with amplitude u0 of the
oscillation for Exercise 2.24

2.25 Exercise 2.25 (Comparison of Equivalent

Linearization Methods and the Method of Multiple
Scales for Solving Different Nonlinear Equations)

Solution: (a) The residual is

R = α1 x + α3 x3 − λx

The mean square value of the residual is

T
  1
R2 = (α1 x + α3 x3 − λx)2 dt
T
0
     
= (α1 − λ)2 x2 + 2α3 (α1 − λ) x4 + α32 x6

In order to find the minimum value for < R2 >, we need

d < R2 >
= −2(α1 − λ) < x2 > −2α3 < x4 > = 0
dλ
So

α1 < x2 > +α3 < x4 >

λ=
< x2 >
Let x = acos(ωt + β), then
74 2 Conservative Single-Degree-of-Freedom Systems

T 2π
1 a2 a2
<x > =
2 a cos (ωt + β)dt =
2 2
cos2 ψd ψ =
T 2π 2
0 0

T 2π
1 a4 3a4
< x4 > = a4 cos4 (ωt + β)dt = cos4 ψd ψ =
T 2π 8
0 0

3
So λ = α1 + α3 a2 (2.25.1)
4
In Exercise 2.5(a), the Eq. (2.5.1) is the same as the nonlinear equation in this
Exercise, where the the method of multiple scales has been used to find√the frequency-
√
amplitude relationship as in the Eq. (2.5.23). In the equation, let ω = λ, ω1 = α1 ,
α = α3 , we can obtain (2.25.1), i.e., the results are the same as those obtained by the
the method of multiple scales.
(b) The residual is

R = α1 x + α2 x2 + α3 x3 − λx

The mean square value of the residual is

T
 1 
R =2
[(α1 − λ)x + α2 x2 + α3 x3 )2 dt
T
0
     
= (α1 − λ)2 x2 + 2α2 (α1 − λ) x3 + 2α3 (α1 − λ) x4
     
+ α2 α3 x5 + α22 x4 + α32 x6

In order to find the minimum value for < R2 >, we need

d < R2 >
= −2(α1 − λ) < x2 > −2α2 < x3 > −2α3 < x4 > = 0
dλ
So

α1 < x2 > +α2 < x3 > +α3 < x4 >

λ=
< x2 >
Let x = acos(ωt + β), then

T 2π
1 a2 a2
<x > =
2 a cos (ωt + β)dt =
2 2
cos2 ψd ψ =
T 2π 2
0 0
2.26 Exercise 2.26 (Comparison of the Galerkin Method with the the Method … 75

T 2π
1 a4 3a4
<x > =
4 a cos (ωt + β)dt =
4 4
cos4 ψd ψ =
T 2π 8
0 0

T
1
< x3 > = a3 cos3 (ωt + β)dt = 0
T
0

3
So λ = α1 + α3 a2 (2.25.2)
4
Let u = εv, we obtain

v̈ + ω02 v + εδv2 + ε2 αv3 = 0 (2.25.3)

where ω02 = α1 , δ = α2 , α = α3 . Equation (2.25.3) is the same as the Eq. (2.2.51)

in Exercise 2.2(c), where the the method of multiple scales has been used to find the
nonlinear frequency-amplitude relationship (Eq. (2.2.71)). From (2.2.71), the rela-
tionship between the nonlinear frequency and the amplitude of the present Exercise
can be obtained by

3 5
ω2 = α1 + α3 a2 − α1−1 α22 a2 (2.25.4)
4 6
It can be seen that the result obtained by the equivalent linearization method
(2.25.2) is different from that obtained by the the method of multiple scales (2.25.4).
The reason is that in the equivalent linearization method, the quadratic nonlinear term
has no effect on the mean square error and therefore does not affect the linearization
parameter λ; whereas in the the method of multiple scales, the quadratic nonlinear
term appears in the equation to eliminate secular terms and thus corrects for the
relationship between the frequency and amplitude in nonlinear oscillation.

2.26 Exercise 2.26 (Comparison of the Galerkin Method

with the the Method of Multiple Scales)

Solution: (a) When x = acos(ωt + β), the residual R is

R = −aω2 cos(ωt + β) + α1 acos(ωt + β)

+ α2 a2 cos2 (ωt + β) + α3 a3 cos3 (ωt + β)
3
= (−aω2 + α1 a + α3 a3 )cos(ωt + β)
4
1 1 1
+ α2 a2 + α2 a2 cos(2ωt + 2β) + α3 a3 cos(3ωt + 3β)
2 2 4
76 2 Conservative Single-Degree-of-Freedom Systems

According to the Galerkin method, there are

T
< Racos(ωt + β) > = 1
T
Racos(ωt + β)dt
0
= 21 a(−aω2 + α1 a + 43 α3 a ) = 0
3

3
So ω2 = α1 + α3 a2 (2.26.1)
4

(b) When x = acos(ωt + β) + B, the residual R is

R = −ω2 acos(ωt + β) + α1 [acos(ωt + β) + B]

+ α2 [acos(ωt + β) + B]2 + α3 [acos(ωt + β) + B]3
3
= [−aω2 + α1 a + 2α2 aB + α3 a3 + 3α3 aB3 ]cos(ωt + β)
4
1 2 3
+ α1 B + α2 ( a + B ) + α3 (B3 + a2 B)
2
2 2
1 3 1
+ [ α2 a + α3 a B]cos(2ωt + 2β) + α3 a3 cos(3ωt + 3β)
2 2
2 2 4

(c) Using the condition that

< Racos(ωt + β) > = 0

we obtain

T
< R[acos(ωt + β) + B] > = 1
T
R[acos(ωt + β) + B]dt
0
= 21 a[−aω2 + α1 a + 2α2 aB + 43 α3 a3 + 3α3 aB3 ] = 0
3
i.e. −aω2 + α1 a + 2α2 aB + α3 a3 + 3α3 aB3 = 0 (2.26.2)
4

(d) Using the condition that

< RB > = 0

we obtain

T
1
< RB > = RBdt
T
0
2.27 Exercise 2.27 (Equations Containing Second, Third and Fourth Order … 77
    
1 3
= B α1 B + α2 a2 + B2 + α3 B3 + a2 B =0
2 2
1 3
i.e. α1 B + α2 ( a2 + B2 ) + α3 (B3 + a2 B) = 0 (2.26.3)
2 2
From the Eqs. (2.26.2) and (2.26.3), we obtain

1
B = − α2 α1−1 a2 + · · ·
2
3 5
and ω2 = α1 + α3 a2 − α22 α1−1 a2 (2.26.4)
4 6

(e) Compare these results with (2.25.4), we can find that the success of the application
of Galerkin procedure depends on the hypothetical solution. In order to obtain the
effect of all nonlineat terms on the frequency-amplitude relationship of a nonlinear
oscillation, we should choose the hypothetical solution to include all nonlinear
coefficients appearing in the odd harmonic terms of the residual R.

2.27 Exercise 2.27 (Equations Containing Second, Third

and Fourth Order Nonlinear Terms)

Solution: (a) At the singularity point of the system, there is ü = u̇ = 0, which gives

−u + u4 = 0

So u = 1 is a singularity of the system and the potential energy of the system is

1 1
V (u) = − u2 + u5
2 5

Since .V (u = 1) = [−1 + 4u3 ]u=1 = 3 > 0
the singularity u = 1 is the center.
Let x = u − 1, which gives

ẍ − (x + 1) + (x + 1)4 = 0

i.e. ẍ + 3x + 6x2 + 4x3 + x4 = 0 (2.27.1)

Let
78 2 Conservative Single-Degree-of-Freedom Systems

x = εv (2.27.2)

Equation (2.27.1) changes to

v̈ + 3v + 6εv2 + 4ε2 v3 + ε3 v4 = 0 (2.27.3)

The general form of Eq. (2.27.3) is given by

v̈ + ω02 v + εα2 v2 + ε2 α3 v3 + ε3 α4 v4 = 0 (2.27.4)

(b) We seek an expansion of the following form for the above equation around the
singularity v = 0

v = v0 (T0 , T1 , T2 ) + εv1 (T0 , T1 , T2 ) + ε2 v2 (T0 , T1 , T2 ) + · · · (2.27.5)

where Tn = εn t. Substitute (2.27.5) into (2.27.4) and retain to O(ε2 ), we obtain

0 = v̈ + ω02 v + εα2 v2 + ε2 α3 v3 + ε3 α4 v4
= [D02 + 2εD0 D1 + ε2 (D12 + 2D0 D2 ) + · · · ](v0 + εv1 + ε2 v2 + · · · )
+ω02 (v0 + εv1 + ε2 v2 + · · · ) + εα2 (v0 + εv1 + ε2 v2 + · · · )2
+ε2 α3 (v0 + εv1 + ε2 v2 + · · · )3 + ε3 α4 (v0 + εv1 + ε2 v2 + · · · )4
= D02 v0 + ω02 v0 + ε(D02 v1 + 2D0 D1 v0 + ω02 v1 + α2 v02 )
+ε2 [D02 v2 + 2D0 D1 v1 + (D12 + 2D0 D2 )v0 + ω02 v2 + 2α2 v0 v1 + α3 v03 ] + · · ·
(2.27.6)

Equate the coefficients of the same power of ε gives

D02 v0 + ω02 v0 = 0 (2.27.7)

D02 v1 + ω02 v1 = −2D0 D1 v0 − α2 v02 (2.27.8)

D02 v2 + ω02 v2 = −2D0 D1 v1 − (D12 + 2D0 D2 )v0 − 2α2 v0 v1 − α3 v03 (2.27.9)

The general solution of (2.2.55) is

v0 = Aeiω0 T0 + Ae−iω0 T0 (2.27.10)

where A = A(T1 , T2 ). Substituting the above equation into (2.27.8), we obtain

D02 v1 + ω02 v1 = −2D0 D1 (Aeiω0 T0 + Ae−iω0 T0 ) − α2 (Aeiω0 T0 + Ae−iω0 T0 )2

= −2α2 AA − 2iω0 D1 Aeiω0 T0 − α2 A2 e2iω0 T0 + cc
(2.27.11)
2.27 Exercise 2.27 (Equations Containing Second, Third and Fourth Order … 79

To eliminate secular terms, we need D1 A = 0. Therefore A = A(T2 ) and the

particular solution of (2.27.11) is

2α2 AA α2 A2 2iω0 T0
v1 = − + e + cc (2.27.12)
ω02 3ω02

Substitute (2.27.10) and (2.27.12) into (2.27.9), and take A = A(T2 ) into account,
we obtain

D02 v2 + ω02 v2 = −2D0 D1 v1 − (D12 + 2D0 D2 )v0 − 2α2 v0 v1 − α3 v03

10α 2 (2.27.13)
= [−2iω0 D2 A + ( 3ω22 − 3α3 )A2 A]eiω0 T0 + cc + NST
0

where cc denotes the complex conjugate term of its preceding terms and NST denotes
terms which do not generate secular terms. In order to eliminate the secular term, we
need

10α22
2iω0 D2 A − ( − 3α3 )A2 A = 0 (2.27.14)
3ω02
1 iβ
Let A= ae (2.27.15)
2
Substitute (2.27.15) into (2.27.14), we obtain

1 10α22
iω0 D2 a − ω0 aD2 β − ( − 3α3 )a3 = 0 (2.27.16)
8 3ω02

Separate the real and imaginary parts of the above equation yields

1 10α22
D2 a = 0, D2 β + ( − 3α3 )a2 = 0 (2.27.17)
8ω0 3ω02

So a is a constant. Integrating the second equation of (2.27.17), we obtain

1 10α22
β=− ( − 3α3 )a2 T2 + β0 (2.27.18)
8ω0 3ω02

Combining the above results, the first-order approximate solution of the Exercise
can be written as

1 10α22
u = εacos[ω0 t − ( − 3α3 )a2 ε2 t + β0 ] + · · · (2.27.19)
8ω0 3ω02

Let ε = 1 yields
80 2 Conservative Single-Degree-of-Freedom Systems

1 10α22
u = acos[ω0 t − ( − 3α3 )a2 t + β0 ] + · · · (2.27.20)
8ω0 3ω02

Therefore, the relationship between the frequency and the amplitude of the
oscillation is

1 10α22
ω = ω0 − ( − 3α3 )a2 (2.27.21)
8ω0 3ω02
3 5 2 2
or ω2 = ω02 + α3 a2 − α a (2.27.22)
4 6ω02 2

For this question, ω02 = 3, α2 = 6, α3 = 4, so

ω2 = 3 − 7a2 (2.27.23)

2.28 Exercise 2.28 (Solving Nonlinear Equations

with Segmentation Functions)

Solution: Cases (a) and (b) are special cases of (c), therefore, we solve for case (c).
As seen from the Fig. 2.9a–c, F(u) is an odd function of u, so this Exercise can be
solved by harmonic balance method, equivalent linearization method and Galerkin
method. Here, we use the Galerkin method.
Let

u = acos(ωt + β)

The residual R is R = −ω2 acos(ωt + β) + F[acos(ωt + β)]

From the condition < Racos(ωt + β) > = 0, we obtain

T
1
0 = Racos(ωt + β) = Racos(ωt + β)dt
T
0
T
ω2 a2 1
=− + acos(ωt + β)F[acos(ωt + β)]dt
2 T
0
2π
ω2 a2 1
=− + acosψF[acosψ]d ψ
2 2π
0
2.28 Exercise 2.28 (Solving Nonlinear Equations with Segmentation Functions) 81

(a) (b)

(c)

Fig. 2.9 F(u) for three different cases (a)–(c) in Exercise 2.28

π/2
ω2 a2 2
=− + acosψF[acosψ]d ψ
2 π
0

√
Since d ψ = −du/(asinψ) = −du/ a2 − u2 , the above equation can be changed
to

0 =< Racos(ωt + β) >

0
ω2 a2 2 uF(u)du
=− − √
2 π a 2 − u2
a
ac a
ω2 a2 2 uF(u)du 2 uF(u)du
=− + √ + √
2 π a −u
2 2 π a 2 − u2
0 ac
ac a
ω2 a2 2 uF(u)du 2 uF(u)du
=− + √ + √ (2.28.1)
2 π a −u
2 2 π a 2 − u2
0 ac

The two integrals on the right-hand side of Eq. (2.28.1) are calculated as follows:
82 2 Conservative Single-Degree-of-Freedom Systems

ac ac ac

uF(u)du uF(u) k1 u2
√ = √ du = √ du
a 2 − u2 a 2 − u2 a 2 − u2
0 0 0

 ac 
= k1 (−u a2 − u2 a0c + a2 − u2 du)
0
 u 2 a2 u
= k1 (−u a2 − u2 a − u2 a0c + sin−1
ac
0 + ac
0 )
2 2 a
u 2 a2 u
= k1 (− a − u2 a0c + sin−1 a0c )
2 2 a
k1 ac  2 k a 2
a
sin−1
1 c
=− a − ac2 +
2 2 a
a a a
uF(u)du uF(u) [k1 ac + k2 (u − ac )]u
√ = √ du = √ du
a 2 − u2 a 2 − u2 a 2 − u2
ac ac ac

a a
(k1 ac − k2 ac )u k2 u2
= √ du + √ du
a 2 − u2 a 2 − u2
ac ac

 k2 u  2 k2 a2 −1 u
= −(k1 ac − k2 ac ) a2 − u2 a
ac − a − u2 a
ac + sin a
ac
2 2 a
 k2 ac  2 π k2 a2 k2 a2 −1 ac
= (k1 ac − k2 ac ) a2 − ac2 + a − ac2 + − sin
2 4 2 a
Substituting these two results into (2.28.1), we obtain

ω2 a2 2 k1 ac  2 k1 a2 −1 ac
− + [− a − ac2 + sin
2 π 2 2 a
 k a 
2 c
+ (k1 ac − k2 ac ) a2 − ac2 + a2 − ac2
2

π k2 a2 k2 a2 −1 ac
+ − sin ]=0 (2.28.2)
4 2 a

2 ac ac ac2
So ω = k2 − (k2 − k1 )[sin−1 +
2
1− ] (2.28.3)
π a a a2

This is the nonlinear frequency-amplitude relationship of the system when F(u)

is shown in Fig. 2.9c.
For case (a), k2 = 0 in (2.28.3), so the nonlinear frequency-amplitude relationship
can be obtained as
2.29 Exercise 2.29 (Nonlinear Equations with a Single Arbitrary Subpartial … 83

2k1 ac ac ac2
ω =
2
[sin−1 + 1− ] (2.28.4)
π a a a2

For case (b), k1 = 0 in (2.28.3), so the nonlinear frequency-amplitude relationship

can be obtained as

2k2 −1 ac ac a2
ω = k2 −
2
[sin + 1 − c2 ] (2.28.5)
π a a a

2.29 Exercise 2.29 (Nonlinear Equations with a Single

Arbitrary Subpartial Nonlinear Term Only)

Solution: Let the solution of the equation be

u = acos(ωt + β)

Substituting it for the given nonlinear differential equation, we obtain

−ω2 acosψ + kan cosn ψ = 0, ψ = ωt + β (2.29.1)

cosn ψ is a periodic function of ψ with a period of 2π , which can be expanded into

a Fourier series

cosn ψ = b1 cosψ + · · ·

where

2π 
π/2
b1 = 1
π
cosψcos n
ψd ψ = 4
π
cosn+1 ψd ψ
√ 0 0
4 π [ 2 (n+2)] [ 1 (n+2)]
1
= π 2 [ 21 (n+3)]
= √2π [ 21 (n+3)]
2

The second equal sign of the above equation holds because n is an odd integer.
Substituting this to (2.29.1), we obtain

2 [ 21 (n + 2)]
−ω2 cosψ + kan−1 √ cosψ + · · · = 0
π [ 21 (n + 3)]

So

2k [ 1 (n + 2)]
ω2 = √ an−1 21 (2.29.2)
π [ 2 (n + 3)]