Classification and Examples

of Differential Equations
and their Applications
Mathematics and Physics for Science and Technology
Series Editor: L.M.B.C. Campos
Director of the Center for Aeronautical
and Space Science and Technology
Lisbon University

Volumes in the series

Topic A – Theory of Functions and Potential Problems
Volume I (Book 1) – Complex Analysis with Applications to Flows
and Fields
L.M.B.C. Campos
Volume II (Book 2) – Elementary Transcendentals with Applications
to Solids and Fluids
L.M.B.C. Campos
Volume III (Book 3) – Generalized Calculus with Applications to Matter
and Forces
L.M.B.C. Campos
Topic B – Boundary and Initial-Value Problems
Volume IV – Ordinary Differential Equations with Applications
to Trajectories and Oscillations
L.M.B.C. Campos
Book 4 – Linear Differential Equations and Oscillators
L.M.B.C. Campos
Book 5 – Non-Linear Differential Equations and Dynamical Systems
L.M.B.C. Campos
Book 6 – Higher-order Differential Equations and Elasticity
L.M.B.C. Campos
Book 7 – Simultaneous Differential Equations and Multi-Dimensional
Vibrations
L.M.B.C. Campos
Book 8 – Singular Differential Equations and Special Functions
L.M.B.C. Campos
Book 9 – Classification and Examples of Differential Equations and
their Applications
L.M.B.C. Campos

For more information about this series, please visit: https://www.crcpress.

com/Mathematics-and-Physics-for-Science-and-Technology/book-series/
CRCMATPHYSCI
 Mathematics and Physics for Science and Technology

Volume IV

Ordinary Differential
Equations with Applications
to Trajectories and Oscillations

Book 9

Classification and Examples

of Differential Equations
and their Applications

By
L.M.B.C. Campos
Director of the Center for Aeronautical
and Space Science and Technology
Lisbon University
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 334 87-2742

© 2020 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Printed on acid-free paper

International Standard Book Number-13: 978-0-367-13724-3 (Hardback)

This book contains information obtained from authentic and highly regarded sources.
Reasonable efforts have been made to publish reliable data and information, but the
author and publisher cannot assume responsibility for the validity of all materials or the
consequences of their use. The authors and publishers have attempted to trace the copyright
holders of all material reproduced in this publication and apologize to copyright holders if
permission to publish in this form has not been obtained. If any copyright material has not
been acknowledged, please write and let us know so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted,
reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means,
now known or hereafter invented, including photocopying, microfilming, and recording, or in
any information storage or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www
.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center,
Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit
organization that provides licenses and registration for a variety of users. For organizations
that have been granted a photocopy license by the CCC, a separate system of payment has been
arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks,

and are used only for identification and explanation without intent to infringe.

Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com

and the CRC Press Web site at

http://www.crcpress.com
to Leonor Campos
Contents

Contents����������������������������������������������������������������������������������������������������������������� vii
List of Classifications and Tables�������������������������������������������������������������������������xi
Preface������������������������������������������������������������������������������������������������������������������� xxi
Acknowledgments��������������������������������������������������������������������������������������������� xxv
About the Author��������������������������������������������������������������������������������������������� xxvii
Physical Quantities������������������������������������������������������������������������������������������� xxix

10. Examples 10.1 to 10.20������������������������������������������������������������������������������������1

EXAMPLE 10.1 One Finite Difference (Two Differential)
Equation(s) with the Same Characteristic Polynomial�������������������� 1
E10.1.1 Linear Differential Equation with Constant
Coefficients������������������������������������������������������������������������������� 1
E10.1.2 Linear Differential Equation with Homogeneous
Coefficients������������������������������������������������������������������������������� 3
E10.1.3 Linear Finite Difference Equation with Constant
Coefficients������������������������������������������������������������������������������� 4
EXAMPLE 10.2 Energies, Dissipation, and Power of
an Oscillation����������������������������������������������������������������������������������������� 5
E10.2.1 Potential, Kinetic, and Total Energies����������������������������������� 5
E10.2.2 Strong Subcritical or Weakly Damped
Oscillations�������������������������������������������������������������������������������7
E10.2.3 Averages over a Period of Energies and Power������������������� 8
E10.2.4 Balance of Forcing and Fast Amplification������������������������ 10
EXAMPLE 10.3 Power-Law Forcing of a Harmonic Oscillator��������������� 11
E10.3.1 Power-Law Monotonic and Oscillating Forcing��������������� 12
E10.3.2 Forcing of an Undamped and Damped Oscillator����������� 13
EXAMPLE 10.4 Five Standards of First-Order Differential
Equations���������������������������������������������������������������������������������������������� 15
EXAMPLE 10.5 Integrating Factor for the Linear First-Order
Differential Equation�������������������������������������������������������������������������� 17
EXAMPLE 10.6 First-Order Differentials in Two Variables��������������������� 18
E10.6.1 Exact First-Order Differentials in Two Variables�������������� 19
E10.6.2 Inexact First-Order Differentials in Two Variables����������� 19
E10.6.3 Integrating Factors for Inexact Differentials���������������������� 20
EXAMPLE 10.7 Paths Near a Stagnation Point of the Second
Degree��������������������������������������������������������������������������������������������������� 21
E10.7.1 Paths Tangent at a Stagnation Point of the
Second Degree������������������������������������������������������������������������ 21

vii
viii Contents

E10.7.2 Paths with Three Asymptotes Crossing at a

Stagnation Point��������������������������������������������������������������������� 23
E10.7.3 Comparison of Velocity Fields with a
Stagnation Point��������������������������������������������������������������������� 24
EXAMPLE 10.8 First-Order Special Differential Equations��������������������� 25
E10.8.1 Cusped Parabola as the Envelope of a Family of
Straight Lines�������������������������������������������������������������������������� 25
E10.8.2 Smooth Parabola as the Envelope of a Family of
Straight Lines�������������������������������������������������������������������������� 26
E10.8.3 First-Order Equation of Degree Three without
Singular Integrals������������������������������������������������������������������ 27
EXAMPLE 10.9 Differential Equations of Order Higher than
the First������������������������������������������������������������������������������������������������� 28
E10.9.1 Linear Differential Equations with Constant or
Variable Coefficients�������������������������������������������������������������� 29
E10.9.2 Non-Linear Second-Order Differential Equations
Omitting the Independent Variable������������������������������������30
E10.9.3 Higher-Order Differential Equations Reducible to
the First Order������������������������������������������������������������������������ 31
E10.9.4 Linear Non-Commutative Differential Operators
with Variable Coefficients����������������������������������������������������� 32
E10.9.5 Non-Linear Exact Differential Equation with
Variable Coefficients�������������������������������������������������������������� 33
E10.9.6 Second-Order Homogeneous Differential
Equation����������������������������������������������������������������������������������34
EXAMPLE 10.10 Bending of a Beam under Traction�������������������������������� 35
E10.10.1 Elastica of a Beam under Traction������������������������������������� 36
E10.10.2 Clamped Beam with a Concentrated Torque������������������37
E10.10.3 Pinned Beam with a Concentrated Force������������������������� 40
E10.10.4 Heavy Clamped-Pinned Beam������������������������������������������ 43
E10.10.5 Cantilever Beam with a Linearly Increasing
Shear Stress����������������������������������������������������������������������������� 45
EXAMPLE 10.11 Linear Bending of a Circular Heavy Plate
with a Circular Hole��������������������������������������������������������������������������� 48
EXAMPLE 10.12 Vibrations of a Membrane under Uniform or
Non-Uniform Tension������������������������������������������������������������������������ 51
E10.12.1 Propagating/Evanescent or Cut-on/Cut-off
Modes�������������������������������������������������������������������������������������� 54
E10.12.2 Turning Point Due to Non-Uniform Tension������������������� 55
EXAMPLE 10.13 Curve as the Tangent to a Vector Field
or as of the Intersection Surfaces������������������������������������������������������ 56
EXAMPLE 10.14 Differentials of First Degree in Three/
Four Variables��������������������������������������������������������������������������������������� 58
E10.14.1 Exact, Inexact, and Non-Integrable Differentials������������ 58
Contents ix

E10.14.2 Immediate and Complete Integrability and

Subsidiary Conditions����������������������������������������������������������� 59
EXAMPLE 10.15 General Boundary Conditions for the
Bending of a Plate������������������������������������������������������������������������������� 60
E10.15.1 Elastic Energy along the Boundary of a Plate�����������������61
E10.15.2 Integration by Parts along a Closed Regular
Boundary��������������������������������������������������������������������������������� 62
E10.15.3 Stress Couple and Turning Moment��������������������������������� 63
E10.15.4 Plate with Arbitrary Closed Regular Boundary�������������� 64
E10.15.5 Isotropic and Pseudo-Isotropic Orthotropic
Plates���������������������������������������������������������������������������������������� 65
E10.15.6 General or Rectangular Clamped Plate���������������������������� 65
E10.15.7 Boundary Conditions for Supported or
Pinned Plates�������������������������������������������������������������������������� 66
E10.15.8 General, Circular, or Rectangular Plate
with Free Boundary��������������������������������������������������������������� 67
E10.15.9 Sets of Boundary Conditions for the
Bending of a Plate������������������������������������������������������������������ 67
EXAMPLE 10.16 Strong Bending of a Pseudo-Isotropic
Orthotropic Plate��������������������������������������������������������������������������������� 68
EXAMPLE 10.17 Non-Linear and Linear Coupled Systems of
Differential Equations������������������������������������������������������������������������� 71
E10.17.1 Simultaneous System of Non-Linear Ordinary
Differential Equations����������������������������������������������������������� 71
E10.17.2 Linear Differential System with a Given
Characteristic Polynomial����������������������������������������������������� 72
EXAMPLE 10.18 Three Coupled Systems with a Matrix of
Characteristic Polynomials���������������������������������������������������������������� 74
E10.18.1 Forced Coupled Linear Differential System
with Constant Coefficients��������������������������������������������������� 75
E10.18.2 Forced Coupled Linear Differential System
with Homogeneous Coefficients����������������������������������������� 77
E10.18.3 Forced Coupled System of Finite Difference
Equations��������������������������������������������������������������������������������� 77
E10.18.4 System of Finite Difference Equations with
Oscillatory Forcing����������������������������������������������������������������78
E10.18.5 Forcing by Power Multiplied by a Cosine����������������������� 78
E10.18.6 Special Case of Constant Forcing�������������������������������������� 80
E10.18.7 General and Complete Solutions��������������������������������������� 81
EXAMPLE 10.19 Asymptotic Stability of a Damped
Non-Linear Oscillator������������������������������������������������������������������������� 82
E10.19.1 Positive-Definite Stability Function���������������������������������� 83
E10.19.2 Negative-Definite Time Derivative of the
Stability Function������������������������������������������������������������������� 83
x Contents

EXAMPLE 10.20 Solutions of the Generalized Circular and

Hyperbolic Differential Equation����������������������������������������������������� 84
E10.20.1 Simple Pole and Regular Integrals������������������������������������ 85
E10.20.2 Double Pole and Elementary Integrals����������������������������� 87
E10.20.3 Triple Pole and Asymptotic Integrals������������������������������� 88
E10.20.4 High-Order Poles and Asymptotic
Analytic Integrals������������������������������������������������������������������� 90
E10.20.5 Circular, Hyperbolic, and Airy Functions
of the Inverse Variable����������������������������������������������������������� 91
Conclusion 10������������������������������������������������������������������������������������������������� 92

Classification 10.1�������������������������������������������������������������������������������������������������94
Classification 10.2����������������������������������������������������������������������������������������������159
References�����������������������������������������������������������������������������������������������������������203
Index���������������������������������������������������������������������������������������������������������������������209
List of Classifications and Tables

Classifications
C 10.1 500 Standards of Single and Simultaneous Systems
of Ordinary Differential Equations����������������������������������������������������� 94
A. Linear Differential Equation with Constant Coefficients�������������������� 95
A. a. Unforced: Standards I–VI��������������������������������������������������������� 95
A. b. Forced—Direct Method: Standards VII–XII��������������������������� 95
A. c. Forced—Method of Inverse Polynomial of Ordinary
Derivatives: Standards XIII–XX������������������������������������������� 96
A. d. Complete Integrals�������������������������������������������������������������������� 97
B. Linear Differential Equation with Homogeneous Coefficients���������� 97
B. a. Unforced: Standards XXI–XXIV����������������������������������������������� 97
B. b. Forced—Direct Method: Standards XXV–XXX���������������������� 97
B. c. Forced—Method of Inverse Polynomial of Homogeneous
Derivatives: Standards XXXI–XXXVII�������������������������������� 98
B. d. Complete Integrals��������������������������������������������������������������������� 99
C. Linear Finite Difference Equation with Constant Coefficients����������� 99
C. a. Unforced: Standards XXXVII–XL��������������������������������������������� 99
C. b. Forced: Standards XLI–XLIV�������������������������������������������������� 100
D. Method of Variation of Parameters: Standard XLV��������������������������� 100
E. Influence or Green’s Function: Standard XLVI����������������������������������� 101
F. Fourier Series-—Discrete Spectrum: Standard XLVII������������������������� 101
G. Fourier Integral—Continuous Spectrum: Standard XLVIII�������������� 101
H. Laplace Transform—Initial Conditions: Standard XLIX������������������� 102
I. First-Order Differential Equations��������������������������������������������������������� 102
I. a. Solvable by Quadratures: Standards L–LIV�������������������������� 102
I. b. Ricatti: Standards LV–LXIII������������������������������������������������������ 103
I. c. Homogeneous Differential Equation: Standards
LXVI–LXV����������������������������������������������������������������������������� 104
J. Differentials����������������������������������������������������������������������������������������������� 104
J. a. First-Order Differential in Two Variables: Standards
LXVI–LXXIII������������������������������������������������������������������������� 104
J. b. First-Order Differential in Three Variables: Standards
LXXIV–LXXVI���������������������������������������������������������������������� 106
J. c. Three-Dimensional Potentials: Standards
LXXVII–LXXXI��������������������������������������������������������������������� 106

xi
xii List of Classifications and Tables

J. d. First-Order Differentials in N ≥ 4 Variables-Standards

LXXXII–LXXXV�������������������������������������������������������������������� 107
J. e. Homogeneous First-Order Differentials in Two
Variables: Standards LXXXVI–LXXXVII��������������������������� 108
J. f. Homogeneous First-Order Differentials in N ≥ 3
Variables: Standards LXXXVIII–LXXXIX������������������������� 108
J. g. Differential of Order M in N Variables:
Standards XC–XCI��������������������������������������������������������������� 108
K. Special Integrals�������������������������������������������������������������������������������������� 109
K. a. Discriminants and Special Points of First-Order
Differential Equations: Standards XCII–XCVI����������������� 109
K. b. Quadratic Discriminants of First-Order Equations:
Standards XCVII–XCVIII���������������������������������������������������� 109
K. c. First-Order Differential Equations That Always Have
Special Integrals: Standards XCIX–C�������������������������������� 110
K. d. First-Order Differential Equations Solvable for the
Slope or a Variable: Standards CI–CIII����������������������������� 110
L. Transformation and Reduction of Order of Differential
Equations�������������������������������������������������������������������������������������������� 111
L. a. Legendre Transformation of Differentials: Standard CIV������� 111
L. b. Dual First-Order Differential Equations: Standards
CV–CVI���������������������������������������������������������������������������������� 111
L. c. Second-Order Differential Equation Reduced
to First-Order: Standards CVII–CXI���������������������������������� 111
L. d. Depression of the Order of Differential Equations
of Order Higher Than Two N > 2: Standards
CXII–CXVI���������������������������������������������������������������������������� 112
L. e. Combinations of Differential Operators:
Standards CXVII–CXVIII���������������������������������������������������� 113
L. f. Higher-Order Homogeneous Differential Equations:
Standards CXIX–CXX���������������������������������������������������������� 113
M. Simultaneous System of Differential Equations:
Standards CXXI–CXXVII����������������������������������������������������������������� 113
N. Simultaneous System of Linear Differential Equations
with Constant Coefficients: Standards CXXVIII–CXXX�������������� 114
N. a. Unforced: Standards CXXXI–CXXXVIII������������������������������� 115
N. b. Forced: Standards CXXXIX–CXLV���������������������������������������� 115
O. Simultaneous System of Linear Differential Equations
with Homogeneous Derivatives: Standard CXLV������������������������ 116
O. a. Unforced: Standards CXLVI–CLV����������������������������������������� 116
O. b. Forced: Standards CLVI–CLXVII������������������������������������������� 117
P. Simultaneous System of Linear Finite Difference Equations
with Constant Coefficients: Standard CLXIII������������������������������� 117
P. a. Unforced: Standards CLXIV–CLXXIII����������������������������������� 118
P. b. Forced: Standards CLXXIV–CLXXVIII���������������������������������� 118
List of Classifications and Tables xiii

Q. Generalized Equation of Mathematical Physics:�������������������������������� 119

Q. a. Reduction to a Helmholtz Equation: Standards
CLXXIX–CLXXXVI�������������������������������������������������������������� 119
Q. b. Hyperspherical (Hypercylindrical) Coordinates:
Standards CLXXXVII–CXCII���������������������������������������������� 120
Q. c. Helmholtz Equation Involving the Laplacian in
Hyperspherical Coordinates: Standards
CXCIII–CXCVIII������������������������������������������������������������������� 120
Q. d. Solution in Hypercylindrical Coordinates:
Standards CXCIX–CCI�������������������������������������������������������� 121
R. Existence, Unicity, Robustness, and Uniformity Theorems:������������� 121
R. a. First-Order Differential Equation with Explicit Slope:
Standards CCII–CCIX��������������������������������������������������������� 121
R. b. Generalized Autonomous System of Differential
Equations: Standards CCX–CCXIII����������������������������������� 122
R. c. Non-Linear Differential Equation of Order N:
Standards CCXIV–CCXVI�������������������������������������������������� 122
R. d. Linear Differential Equation of Order N:
Standards CCXVII–CCXIX������������������������������������������������� 123
S. Stability of Equilibrium of a System����������������������������������������������������� 123
S. a. Classification of Equilibria and Related Concepts:
Standards CCXX–CCXXVIII����������������������������������������������� 123
S. b. Stability of an Equilibrium Specified by a Liapunov
Function: Standards CCXXIX–CCXXXII�������������������������� 124
S. c. Linear Differential Equation with Bounded
Coefficients: Standards CCXXXIII–CCXXXVI����������������� 125
S. d. Stability of an Equilibrium Specified by an
Autonomous Differential System: Standards
CCXXXVII–CCXLV�������������������������������������������������������������� 125
T. Linear Differential Equations with Periodic Coefficients:����������������� 126
T. a. Differential Equation of Arbitrary Order: Standards
CCXLVI–CCXLIX����������������������������������������������������������������� 126
T. b. Second-Order Differential Equation in Invariant Form:
Standard CCL–CCLIV��������������������������������������������������������� 127
U. Linear Differential Equation with Analytic Coefficients������������������� 127
U. a. Classification of Points of a Complex Function
of a Complex Variable: Standards CCLV–CCLXII���������� 127
U. b. Classification of Points of a Linear Differential
Equation: Standards CCLXIII–CCLXVIII������������������������� 128
U. c. Methods of Solution of a Linear Differential
Equation with Analytic Coefficients: Standards
CCLXIX–CCLXXII��������������������������������������������������������������� 128
U. d. Application to The Generalized Hyperbolic
and Circular Functions: Standards
CCLXXIII–CCLXXXIV�������������������������������������������������������� 129
xiv List of Classifications and Tables

U. e. Generalized Hyperbolic (Circular) Differential Equation

with Negative Integer Parameter: Standards
CCLXXXV–CCXC���������������������������������������������������������������� 130
V. Linear Differential Equation with Regular Singularities������������������� 131
V. a. Linear Autonomous System of Differential
Equations with a Regular Singularity–Standards
CCXI–CCXII�������������������������������������������������������������������������� 131
V. b. Linear Differential Equation of Order N with a Regular
Singularity: Standards CCXCIII–CCXCIV����������������������� 131
V. c. Linear Second-Order Differential Equation with a
Regular Singularity: Standards CCXCV–CCCI��������������� 132
V. d. Generalized Bessel Function: Standards
CCCII–CCCVI���������������������������������������������������������������������� 133
V. e. Generalized Neumann Function of Integer Order:
Standards CCCVII–CCCX�������������������������������������������������� 133
V. f. Extension of the Generalized Neumann Function to
Complex Order: Standards CCCXI–CCCXIV������������������ 133
V. g. Generalized, Cylindrical and Spherical Bessel
Differential Equation: Standards CCCXV–CCCXXII������ 134
W. Linear Differential Equation with Irregular Singularity������������������135
W. a. Existence of Regular Integrals Near an Irregular
Singularity: Standards CCCXXIII–CCXXIV��������������������� 135
W. b. Existence of Normal Integrals Near an Irregular
Singularity: Standards CCCXXV–CCXXVIII������������������� 135
W. c. Asymptotic Integrals for Cylindrical and Spherical
Bessel Differential Equations: Standards
CCCXXIX–CCCXXXIII�������������������������������������������������������� 136
W. d. Asymptotic Integrals for the Generalized
Bessel Differential Equation: Standards
CCCXXXIV–CCCXL������������������������������������������������������������ 136
W. e. Bifurcation of the Generalized Bessel Differential
Equation at Infinity: Standards CCCXLI–CCCXLV�������� 137
W. f. General Irregular Integrals of Two Kinds: Standards
CCCXLVI–CCCLIV�������������������������������������������������������������� 137
W. g. Infinite Determinants and Linear Systems: Standards
CCCLV–CCCLVII����������������������������������������������������������������� 138
W. h. Multiple Parametric Resonance and Hill Differential
Equation: Standards CCCLVIII–CCCLXII������������������������ 139
X. Kernal and Path for the Solution by Integral Transforms����������������� 139
X. a. Beta Function or Eulerian Integral of the Second Kind:
Standards CCCLXIII–CCCLXVI���������������������������������������� 139
X. b. Generalized Laplace Transform Along a Path in the
Complex Plane: Standards CCCLXVII–CCCLXXIII������� 140
X. c. Generalized Integral Transform and Adjoint Kernel
Operator: Standards CCCLXXIV–CCLXXIX������������������� 141
List of Classifications and Tables xv

Y. Solution of Linear Second-Order Differential Equations by

Continued Fractions: Standards CCCLXXX–CCCXCV��������������� 141
Z. Higher Transcendental Functions: Special and Extended����������������� 142
Z. a. Generalized Hypergeometric Differential Equation:
Standards CCCXCVI–CDIII����������������������������������������������� 142
Z. b. Gaussian Hypergeometric Differential Equation:
Standards CDIV–CDXIII����������������������������������������������������� 144
Z. c. Legendre, Associated, Ultraspherical and
Hyperspherical Differential Equations: Standards
CDXIV–CDXXVIII��������������������������������������������������������������� 144
Z. d. Jacobi and Chebychev Functions of Two Kinds:
Standards CDXXIX–CDXXXIX������������������������������������������ 146
Z. e. Confluent Hypergeometric, Laguerre, Associated, and
Hermite Functions: Standards CDXL–CDL��������������������� 147
Z. f. Sub- and Superhypergeometric Differential Equations:
Standards CDLI–CDLXI����������������������������������������������������� 148
Z. g. Invariant and Self-Adjoint Linear
Second-Order Differential Equations: Standards
CDLXII–CDLXXV���������������������������������������������������������������� 150
Z. h. First Class of Transformed Generalized Bessel
Differential Equations: Standards
CDLXXVI–CDLXXXVII������������������������������������������������������ 153
Z. i. Second Class of Transformed Generalized Bessel
Differential Equations: Standards CDLXXXVIII–D�������� 155
C 10.2 500 Applications of Differential Equations to Physical and
Engineering Problems������������������������������������������������������������������������� 159
A. Linear, Attractive/Repulsive, Damped/Amplified and
Free/Forced Oscillator��������������������������������������������������������������������� 160
Problems 1–2�������������������������������������������������������������������������������������� 160
A. a. Undamped and Unforced: Problems 3–7����������������������������� 160
A. b. Damped and Unforced: Problems 8–11�������������������������������� 161
A. c. Amplified and Unforced: Problems 12–22���������������������������� 161
A. d. Degenerate Cases: Problems 23–27��������������������������������������� 161
A. e. Undamped, with Sinusoidal Forcing:
Problems 28–30��������������������������������������������������������������������� 161
A. f. Damped, with Sinusoidal Forcing: Problems 31–32������������ 162
A. g. Simultaneous Forcing in Frequency and Damping for
Subcritical Damping: Problems 33–38������������������������������ 162
A. h. Forcing Involving a Power of Time: Problems 39–44��������� 162
A. i. Integrable Forcing Function: Problems 45–50����������������������163
A. j. Forcing by a Discrete Spectrum: Problems 51–54����������������� 163
A. k. Forcing by a Continuous Spectrum: Problems 55–58��������� 163
A. l. Forcing by an Impulse: Problems 59–62�������������������������������� 163
A. m. Forcing by a Continuous Function: Problems 63–67��������� 164
A. n. Forcing by a Gaussian Function: Problems 68–69��������������� 164
xvi List of Classifications and Tables

B. Motion Near a Stagnation Point������������������������������������������������������������ 164

B. a. Paths Near a Stagnation Point of First Degree:
Problems 70–76��������������������������������������������������������������������� 164
B. b. Paths Near a Stagnation Point of Second Degree:
Problems 77–78��������������������������������������������������������������������� 165
B. c. Trajectories Near a Stagnation Point of First Degree:
Problems 79–86��������������������������������������������������������������������� 165
B. d. Two Orthogonal Undamped Oscillators:
Problems 87–89��������������������������������������������������������������������� 166
C. Parametric Resonance���������������������������������������������������������������������������� 166
C. a. Undamped Oscillator with Coefficients Depending
on Time: Problem 90������������������������������������������������������������ 166
C. b. Harmonic Oscillator with Oscillating Support:
Problem 91���������������������������������������������������������������������������� 166
C. c. Stability/Instability/Indifference of Solutions:
Problems 92–93��������������������������������������������������������������������� 167
C. d. Excitation Frequencies Leading to Parametric
Resonance: Problems 94–95������������������������������������������������ 167
C. e. Undamped Parametric Resonance: Problems 96–98����������� 167
C. f. Parametric Resonance with Weak Damping:
Problems 99–101������������������������������������������������������������������� 168
C. g. Suspended Pendulum with Oscillating Support:
Problems 102–104����������������������������������������������������������������� 168
D. Non-Linear, Undamped, Unforced Oscillator������������������������������������ 168
D. a. For a General Potential: Problems 105–106�������������������������� 169
D. b. Biquadratic Potential for a Soft/Hard Spring: Problems
107–113�����������������������������������������������������������������������������������169
D. c. Quartic Potential Generalizing the Biquadratic
Potential: Problems 114–116����������������������������������������������� 169
E. Resonance of a Non-Linear Oscillators: Problems 117–121��������������� 170
F. Electromechanical Dynamo�������������������������������������������������������������������� 170
F. a. Electrical Dynamo: Problems 122–123����������������������������������� 170
F. b. Electromechanical Dynamo: Problems 124–125������������������� 171
F. c. Exact electromechanical Homopolar Dynamo:
Problems 126–127����������������������������������������������������������������� 171
F. d. Self-Excited Dynamo: Problems 128–129������������������������������ 171
G. Circular Pendular Motion���������������������������������������������������������������������� 171
G. a. Large Motion Without Damping: Problems 130–133���������� 171
G. b. Pendular Motion with Quadratic Damping:
Problems 134–136����������������������������������������������������������������� 172
H. Bifurcations, Chaos and Approximations������������������������������������������� 172
H. a. Bifurcations of Dynamical Systems: Problem 137��������������� 172
H. b. Hopf Bifurcation: Problems 138–139������������������������������������172
List of Classifications and Tables xvii

H. c. Mixed Non-Linear Oscillator of van der Pol:

Problems 140–141����������������������������������������������������������������� 172
H. d. Polynomial Potentials and Catastrophes:
Problems 142–143����������������������������������������������������������������� 172
H. e. Chaotic Motions: Problems 144–147������������������������������������� 173
H. f. Limit Cycles: Problems 148–149��������������������������������������������� 173
H. g. Stability of Orbits of Dynamical Systems:
Problems 150–151����������������������������������������������������������������� 173
H. h. Index of a Singularity: Problems 152–156���������������������������� 173
I. Numerical Solution of Ordinary Differential Equations:
Problems 157–160������������������������������������������������������������������������������ 174
J. Optics and Acoustic Rays������������������������������������������������������������������������ 174
J. a. Linear Second-Order Differential Equation:
Problems 161–163����������������������������������������������������������������� 174
J. b. Three Alternative Forms: Problems 164–167������������������������� 175
J. c. Solutions of the Invariant Unforced Linear Second-
Order Differential Equation: Problems 168–170�������������� 175
J. d. Waves in a Reffracting Medium: Problems 171–172������������ 175
J. e. Light and Sound Rays: Problems 173–177����������������������������� 176
K. Buckling of Elastic Beams���������������������������������������������������������������������� 176
K. a. General Equations: Problems 178–183���������������������������������� 176
K. b. Linear Buckling: Problems 184–189��������������������������������������� 176
K. c. Non-Linear Buckling to Lowest Order:
Problems 190–193����������������������������������������������������������������� 176
K. d. Buckling with a Point Translational Spring:
Problems 194–198����������������������������������������������������������������� 177
K. e. Buckling with a Point Rotational Spring:
Problems 199–203����������������������������������������������������������������� 177
K. f. Opposition to, or Facilitation of, Buckling:
Problems 204–208����������������������������������������������������������������� 177
K. g. Bending of Beam Continuously Supported on Springs:
Problems 209–221����������������������������������������������������������������� 178
K. h. Linear Bending of a Beam Under Traction Subject
to Shear Stresses: Problems 222–225��������������������������������� 179
L. Axial Traction/Compression of a Straight Bar:
Problems 226–229�������������������������������������������������������������������������������������� 179
M. Plane Elastic Displacements, Strains, and Stresses:
Problems 230–238������������������������������������������������������������������������������ 180
N. Deflection of an Elastic Membrane Under Anisotropic In-Plane
Stresses: Problems 239–246�������������������������������������������������������������� 180
O. Weak Linear Bending of Elastic Isotropic Plates�������������������������������� 181
O. a. Transverse Displacement with Constant Bending
Stiffness: Problems 247–251������������������������������������������������ 181
xviii List of Classifications and Tables

O. b. Transverse Displacement with Non-Uniform Bending

Stiffness: Problems 252–257������������������������������������������������ 182
O. c. Boundary Conditions for Isotropic Elastic Plates:
Problems 258–265����������������������������������������������������������������� 182
O. d. Weak Linear Bending or Isotropic Circular Plates:
Problems 266–272����������������������������������������������������������������� 182
P. Linear Elastic Stability of Stressed Orthotropic Plates����������������������� 183
P. a. Linear Strong Bending with Anisotropic In-Plane
Stresses: Problems 273–278������������������������������������������������� 183
P. b. Mechanics and Thermodynamics of Deformable Media:
Problems 279–281����������������������������������������������������������������� 183
P. c. Stress-Strain Relation for Elastic Materials:
Problems 282–292����������������������������������������������������������������� 184
P. d. Equation for the Transverse Displacement of a
Pseudo-Isotropic Orthotropic Elastic Plate:
Problems 293–297����������������������������������������������������������������� 184
P. e. Boundary Conditions for a Pseudo-Isotropic
Orthotropic Elastic Plate: Problems 298–306������������������� 185
P. f. Buckling of a Clamped Rectangular Plate with
Normal Stresses: Problems 307–309���������������������������������� 185
P. g. Buckling of a Clamped-Free Plate with Uniaxial
Stresses: Problems 310–312������������������������������������������������� 185
Q. Non-Linear Coupling of Strong Bending
and In-Plane Stresses������������������������������������������������������������������������ 186
Q. a. Balance Equations for the Strong Bending of an
Isotropic Elastic Plate: Problems 313–316������������������������� 186
Q. b. Extension of Balance Equations to a Pseudo-Isotropic
Orthotropic Plate: Problems 317–320�������������������������������� 186
Q. c. Boundary Conditions for the Transverse
Displacement and Stress Function:
Problems 321–327����������������������������������������������������������������� 186
Q. d. Deflection and Bending with an Elliptic Boundary:
Problems 328–329����������������������������������������������������������������� 187
Q. e. Perturbation Expansions of All Orders:
Problems 330–334����������������������������������������������������������������� 187
Q. f. Strong Non-Linear Bending of a Heavy Circular Plate
Under Compression: Problems 335–338��������������������������� 187
R. Waves or Vibrations of Elastic Media��������������������������������������������������� 188
R. a. Non-Linear Waves in Unsteady and Inhomogeneous
Media: Problems 339–344��������������������������������������������������� 188
R. b. Generalized and Analogous Wave Equations:
Problems 345–349����������������������������������������������������������������� 189
R. c. Generalized Classical Bending Wave Equation
in One Dimension with Constant Coefficients:
Problems 350–353����������������������������������������������������������������� 190
List of Classifications and Tables xix

R. d. Non-Dispersive and Dispersive Free Wave

Propagation and Standing Modes:
Problems 354–359����������������������������������������������������������������� 190
R. e. Wave Generation by Out-of-Equilibrium Initial
Conditions and Forcing: Problems 360–365��������������������� 190
S. Waves in Inhomogeneous Media���������������������������������������������������������� 191
S. a. Classification of Wave Equations: Problems 366–369���������� 191
S. b. Quasi-One-Dimensional Waves in Inhomogeneous
Media: Problems 370–375��������������������������������������������������� 191
S. c. Analogies Among Six Types of Waves:
Problems 376–382����������������������������������������������������������������� 193
S. d. Duality Principle for Wave Variables:
Problems 383–388����������������������������������������������������������������� 193
T. Acoustics of Horns����������������������������������������������������������������������������������� 193
T. a. General Properties: Problems 389–393����������������������������������� 193
T. b. Exact Elementary Solutions: Problems 394–402������������������� 194
T. c. Gaussian Horn and Displacement Amplifier:
Problems 403–406����������������������������������������������������������������� 194
T. d. Power-Law Horns and Cylindrical/Spherical Waves:
Problems 407–412����������������������������������������������������������������� 194
T. e. Reflection and Transmission at Interfaces:
Problems 413–420����������������������������������������������������������������� 195
U. Oscillators with Several Degrees-of-Freedom������������������������������������ 195
U. a. Equations of Motion and Energy Balance:
Problems 421–426����������������������������������������������������������������� 195
U. b. Simultaneous Multi-Dimensional Free Oscillations:
Problems 427–433����������������������������������������������������������������� 196
U. c. Multiple Resonance and Forced Motion:
Problems 434–444����������������������������������������������������������������� 196
V. Coupled Two-Dimensional Oscillators������������������������������������������������ 197
V. a. Mechanical and Electrical Analogues:
Problems 445–449����������������������������������������������������������������� 197
V. b. Translational and Rotational Free Oscillations of a
Body: Problems 450–461����������������������������������������������������� 198
V. c. Forced Oscillations, Beats, and Resonance of a Rod:
Problems 462–467����������������������������������������������������������������� 199
W. Principle of the Vibration Absorber: Problems 468–473������������������� 199
X. Radioactive Desintegration Chain: Problems 474–486���������������������� 200
Y. Sequence of Linked Oscillators Problems 487–493����������������������������� 201
Z. Signals Along a Transmission Line: Problems 494–500��������������������� 201
xx List of Classifications and Tables

Tables
T 10.1 Combinations of Boundary Conditions for a Circular Plate
with a Circular Hole������������������������������������������������������������������������������ 49
T 10.2 Constants in the Shape of Neutral Surface of Heavy
Circular Plate with a Circular Hole����������������������������������������������������� 51
T 10.3 Boundary Conditions for the Bending of a Plate������������������������������ 61
T 10.4 General Solution (10.350a, b) of Coupled System of
Finite Difference Equations (10.346a, b)���������������������������������������������� 81
Preface

Volume IV (Ordinary Differential Equations with Applications to Trajectories

and Oscillations) is organized like the preceding three volumes of the series
Mathematics and Physics Applied to Science and Technology: (volume III) Generalized
Calculus with Applications to Matter and Forces; (volume II) Transcendental
Representations with Applications to Solids and Fluids; (volume I) Complex
Analysis with Applications to Flows and Fields. The present book, Classifications
of Differential Equations and Examples of their Applications, is the sixth and last
book of volume IV and the ninth book of the series. This book consists of
chapter 10 of volume IV, containing 20 examples related to the preceding five
books and chapters 1 to 9. This book includes two recollections—the first
details a classification of differential equations into 500 standards and the
second provides a list of 500 applications. The examples and classifications
cover the ensemble of the first five books of volume IV, which correspond
to books 4 to 8 of the series, namely: (i) Book 4—Linear Differential Equations
and Oscillators; (ii) Book 5—Non-Linear Differential Equations and Dynamical
Systems; (iii) Book 6—Higher-Order Differential Equations and Elasticity;
(iv) Book 7—Simultaneous Differential Equations and Multi-Dimensional Vibrations;
(v) Book 8—Singular Differential Equations and Special Functions. The aggregate
of contents is summarized next.

Organization of the Contents

Volume IV consists of ten chapters: (i) the odd-numbered chapters pres-
ent mathematical developments; (ii) the even-numbered chapters contain
physical applications; (iii) the last chapter is a set of 20 detailed examples
of (i) and (ii). The chapters are divided into sections and subsections, for
example, chapter 9, section 9.1, and subsection 9.1.1. Chapter 10 contains
20 examples, Example 10.1 to Example 10.20, some with subsections, such
as E10.1.2. The formulas are numbered by chapters in curved brackets, for
example, (10.2) is equation 2 of chapter 10. When referring to volume I the
symbol I is inserted at the beginning, for example: (i) chapter I.36, sec-
tion I.36.1, subsection I.36.1.2; (ii) equation (I.36.33a). The final part of each
chapter includes: (i) a conclusion referring to the figures as a kind of visual
summary; (ii) the notes, lists, tables, diagrams, and classifications as addi-
tional support. The latter (ii) apply at the end of each chapter, and are num-
bered within the chapter (for example, classification—Classification 10.2,
table—Table 10.3); if there is more than one note, list, table, diagram, or

xxi
xxii Preface

classification, they are numbered sequentially (for example, Table 10.1 to

Table 10.4). The chapter starts with an introductory preview, and related
topics may be mentioned in the notes at the end. The “Series Preface” and
“Mathematical Symbols” from the first book of volume IV are not repeated
in this book and “Bibliography” is not added to this book. The “Physical
Quantities,” “References,” and “Index” in this book cover all six books of
the volume IV to account for the close relationship among them.

Summary of Volume IV
Following the three volumes on complex, transcendental, and generalized
functions, this fourth volume is the first volume on boundary and initial
value problems, and concerns ordinary differential equations with a vari-
ety of applications, including trajectories and oscillations. The differential
equations are classified in 500 standards indicating their properties, meth-
ods of solution, and examples, and include: linear and non-linear ordinary
differential equations and simultaneous systems; in particular, linear equa-
tions with constant coefficients having a characteristic polynomial, and
linear equations with variable coefficients leading to 23 special functions.
The applications consist of 500 problems with physical formulation, math-
ematical solutions, and interpretation of results; the problems include linear
and non-linear oscillators with one or several degrees-of-freedom, with or
without damping, and forcing including ordinary, parametric, multiple, and
non-linear resonance. Other applications include trajectories, orbits, electro-
mechanical systems, and chains with deterministic and chaotic behavior;
also included are deformations and buckling of elastic beams, membranes,
and plates. Included in the contents are:

• Ordinary differential equations of first, second, and higher orders,

and simultaneous systems;
• Linear differential and finite difference equations and simultaneous
systems with characteristic polynomials;
• Linear differential equations with variable coefficients and 23 asso-
ciated special functions, such as generalized Bessel, hypergeometric,
etc.;
• Non-linear differential equations, bifurcations, and dynamical and
chaotic systems;
• Existence, unicity, uniformity, robustness, and stability theorems;
• Applications to trajectories, dynamos, and other electromechanical
systems and chains;
Preface xxiii

• Linear and non-linear oscillators with one or several degrees-of-

freedom, with damping and forcing;
• Linear, non-linear, multiple, and parametric resonance of mechani-
cal and electrical systems;
• Torsional, longitudinal, and bending deformations and buckling of
rods, membranes, and plates.
Acknowledgments

The fourth volume of the series justifies renewing some of the acknowl-
edgments also made in the first three volumes, to those who contributed
more directly to the final form of the volume: Ms. Ana Moura, L. Sousa, and
S. Pernadas for help with the manuscripts; Mr. J. Coelho for all the drawings;
and at last, but not least, to my wife as my companion in preparing this work.

xxv
About the Author

L.M.B.C. Camposwas born on March 28, 1950, in Lisbon,

Portugal. He graduated in 1972 as a mechanical engi-
neer from the Instituto Superior Tecnico (IST) of Lisbon
Technical University. The tutorials as a student (1970)
were followed by a career at the same institution (IST)
through all levels: assistant (1972), assistant with tenure
(1974), assistant professor (1978), associate professor (1982),
chair of Applied Mathematics and Mechanics (1985).
He has served as the coordinator of undergraduate and
postgraduate degrees in Aerospace Engineering since
the creation of the programs in 1991. He is the coordinator of the Scientific
Area of Applied and Aerospace Mechanics in the Department of Mechanical
Engineering. He is also the director and founder of the Center for Aeronautical
and Space Science and Technology.
In 1977, Campos received his doctorate on “waves in fluids” from the
Engineering Department of Cambridge University, England. Afterwards,
he received a Senior Rouse Ball Scholarship to study at Trinity College,
while on leave from IST. In 1984, his first sabbatical was as a Senior Visitor
at the Department of Applied Mathematics and Theoretical Physics of
Cambridge University, England. In 1991, he spent a second sabbatical as an
Alexander von Humboldt scholar at the Max-Planck Institut fur Aeronomic
in Katlenburg-Lindau, Germany. Further sabbaticals abroad were excluded
by major commitments at the home institution. The latter were always com-
patible with extensive professional travel related to participation in scientific
meetings, individual or national representation in international institutions,
and collaborative research projects.
Campos received the von Karman medal from the Advisory Group
for Aerospace Research and Development (AGARD) and Research and
Technology Organization (RTO). Participation in AGARD/RTO included
serving as a vice-chairman of the System Concepts and Integration Panel, and
chairman of the Flight Mechanics Panel and of the Flight Vehicle Integration
Panel. He was also a member of the Flight Test Techniques Working Group.
Here he was involved in the creation of an independent flight test capability,
active in Portugal during the last 30 years, which has been used in national
and international projects, including Eurocontrol and the European Space
Agency. The participation in the European Space Agency (ESA) has afforded
Campos the opportunity to serve on various program boards at the levels of
national representative and Council of Ministers.
His participation in activities sponsored by the European Union (EU)
has included: (i) 27 research projects with industry, research, and academic

xxvii
xxviii About the Author

institutions; (ii) membership of various Committees, including Vice-

Chairman of the Aeronautical Science and Technology Advisory Committee;
(iii) participation on the Space Advisory Panel on the future role of EU in
space. Campos has been a member of the Space Science Committee of the
European Science Foundation, which works with the Space Science Board
of the National Science Foundation of the United States. He has been a
member of the Committee for Peaceful Uses of Outer Space (COPUOS) of
the United Nations. He has served as a consultant and advisor on behalf of
these organizations and other institutions. His participation in professional
societies includes member and vice-chairman of the Portuguese Academy of
Engineering, fellow of the Royal Aeronautical Society, Astronomical Society
and Cambridge Philosophical Society, associate fellow of the American
Institute of Aeronautics and Astronautics, and founding and life member of
the European Astronomical Society.
Campos has published and worked on numerous books and articles.
His publications include 10 books as a single author, one as an editor, and
one as a co-editor. He has published 152 papers (82 as the single author,
including 12 reviews) in 60 journals, and 254 communications to sym-
posia. He has served as reviewer for 40 different journals, in addition to
23 reviews published in Mathematics Reviews. He is or has been member
of the editorial boards of several journals, including Progress in Aerospace
Sciences, International Journal of Aeroacoustics, International Journal of Sound
and Vibration, and Air & Space Europe.
Campos’s areas of research focus on four topics: acoustics, magnetohydro-
dynamics, special functions, and flight dynamics. His work on acoustics has
concerned the generation, propagation, and refraction of sound in flows with
mostly aeronautical applications. His work on magnetohydrodynamics has
concerned magneto-acoustic-gravity-inertial waves in solar-terrestrial and
stellar physics. His developments on special functions have used differintegra-
tion operators, generalizing the ordinary derivative and primitive to complex
order; they have led to the introduction of new special functions. His work
on flight dynamics has concerned aircraft and rockets, including trajectory
optimization, performance, stability, control, and atmospheric disturbances.
The range of topics from mathematics to physics and engineering fits with
the aims and contents of the present series. Campos’s experience in univer-
sity teaching and scientific and industrial research has enhanced his ability to
make the series valuable to students from undergraduate level to research level.
Campos’s professional activities on the technical side are balanced by other
cultural and humanistic interests. Complementary non-technical interests
include classical music (mostly orchestral and choral), plastic arts (painting,
sculpture, architecture), social sciences (psychology and biography), history
(classical, renaissance and overseas expansion) and technology (automotive,
photo, audio). Campos is listed in various biographical publications, includ-
ing Who’s Who in the World since 1986, Who’s Who in Science and Technology
since 1994, and Who’s Who in America since 2011.
Physical Quantities

The location of first appearance is indicated, for example “2.7” means

“section 2.7,” “6.8.4” means “subsection 6.8.4,” “N8.8” means “note 8.8,” and
“E10.13.1” means “example 10.13.1.”

1 Small Arabic Letters

a — acceleration: 2.1.1
— amplitude of oscillation: 2.2.11
— moment arm: 4.7.3
ai — acceleration vector: 6.8.3
an — coefficients of a series: 4.4.10
b — stiffness dispersion parameter of an elastic bar: N6.12
— width of a water channel: N7.12
b — stiffness dispersion parameter for an elastic plate: N6.10
c — speed of light in vacuo: 2.1.9
— phase speed of waves: N5.13, N6.14, N8.1
c1 , c2 — speed of transversal waves in the directions of principal stress
of an anisotropic elastic membrane: N6.7
ce — speed of transversal waves in an elastic string: N6.3
— speed of transversal waves in an isotropic elastic membrane: N6.4
cem — speed of electromagnetic waves: 2.1.11
c — speed of longitudinal waves in an elastic rod: N6.6
cr — isothermal sound speed: 5.5.24
c s — adiabatic sound speed: 5.5.14
ct — speed of torsional waves along an elastic rod: N6.5
cw — speed of water waves along a channel: N7.13

ei — non-unit base vector: N8.8
eijk — three-dimensional permutation symbol: 6.8.4
f — forcing function: 2.1.8

f — force vector per unit area: 6.5.4
fi — force vector per unit volume: 6.8.3
fr — reduced external force: 8.2.1

xxix
xxx Physical Quantities

g — acceleration of gravity: 6.3.14

— determinant of the covariant metric tensor: N8.8
g  — reduced modal force: 8.1.1
g ij — covariant metric tensor: N8.8
g ij — contravariant metric tensor: N8.8
h — friction force: 2.1.1
— amplitude of excitation of parametric resonance: 4.3.2
— thickness of a plate: 6.5.4
— depth of water channel: N7.12
hi — friction force vector: 8.1.1
— scale factors: N8.8
j — restoring force: 2.1.1
— electric current: 2.1.6

j — convective electric current vector: 2.1.9
ji — restoring force vector: 8.1.1
k — resilience of a point translational spring: 2.1.2
— Boltzmann constant: 5.5.16
— transversal wavenumber: N5.14
— curvature: 6.1.11
— wavenumber: N7.25, N8.2
k ′ — differential resilience of a distributed spring: 6.3.1
k — resilience of a point rotational spring: 6.2.6
k1 , k2 — principal curvatures: 6.7.2
kij — matrix of curvatures and cross-curvatures: 6.7
krs — resilience matrix: 8.1.2
 — lengthscale for horns: N7.28
m — mass of a particle or body: 2.1.1
mrs — mass matrix: 8.2.1
n — number of degrees of freedom of a molecule: 5.5.21

n — unit vector normal to a curve: 6.9.7
p — pressure: 5.5.4, 6.5.1
— buckling parameter: 6.1.2
q — electric charge: 2.1.6
— non-linearity parameter for pendular motion: 4.7.13
q — modal coordinates: 8.2.4
s — arc length: 6.1.1
Physical Quantities xxxi

t — time: 2.1.1
— time of reception: 2.1.11
t0 — time of emission: 2.1.11
u — longitudinal displacement of a bar: 6.4.1

u — displacement vector: 6.4.1, 6.8.5
v — velocity: 2.1.1

v — velocity vector: 5.2.1
w — group velocity: N6.16
x — independent variable in ordinary differential equation: 1.1.1
— Cartesian coordinate: 2.1.1

x — position vector of observer: 2.1.11
y — dependent variable in ordinary differential equation: 1.1.1
— Cartesian coordinate: 2.1

y — position vector of source: 2.1.11
z — Cartesian coordinate: 6.1.1
— specific impedance: N7.47

2 Capital Arabic Letters

A — Activity or power: 2.1.4
— anharmonic factor: 4.4.10
— cross-sectional area of a horn: N7.16
— admittance: 8.9.1
— Avogadro number: 5.5.17
An — modal amplitudes: 8.8.5

A — vector magnetic potential: 2.1.10
Aab — elastic compliance matrix: 6.8.10

B — magnetic induction vector: 2.1.9
Bmn — terms in the modal matrix: 8.8.5
C — constant of integration: 1.1.1
— capacity of an electric condenser: 2.1.6
— torsional stiffness of an elastic rod: N6.4

C — bending vector: 6.7.8
Cp — specific heat at constant pressure: 5.5.12
xxxii Physical Quantities

Cv — specific heat at constant volume: 5.5.12

Cab — elastic stiffness matrix: 6.8.7
Cijk — elastic stiffness tensor: 6.8.6
D — drag force: 4.8.13
— bending stiffness of an elastic plate: 6.7.2
D — generalized bending stiffness: 6.8.14
D2 — relative area change: 6.5.5
D3 — relative volume change: 6.5.1

D — electric displacement vector: 2.1.9
E — Young modulus of elasticity: 6.4.1, 6.5.1
E0 , E1 — coefficients in the non-uniform Young modulus: 6.4.4

E — electric field vector: 2.1.9
Eb — elastic energy per unit area of in-plane deformation: 6.9.4
E b — total elastic energy of in-plane deformation: 6.9.5
Ec — elastic energy per unit area for the deflection of an elastic mem-
brane: 6.6.2
Ec — total elastic energy of deflection for a membrane: 6.6.2
E c — total elastic energy of deflection for a plate: 6.9.5
Ed — elastic energy per unit volume for a plate: 6.7.8
E d — total elastic energy for a plate: 6.7.8
E d — elastic energy per unit length along the boundary of a plate: 6.8.16
Êd — total elastic energy along the boundary of a plate: E10.13.1
Ee — electric energy: 2.1.7
Eem — electromagnetic energy: 2.1.7
Em — magnetic energy: 2.1.7
Et — total energy: 2.1.4
Ev — kinetic energy: 2.1.4, 6.8.5
F — force: 3.9.11
— free energy: 5.5.8
— longitudinal force in a bar: 6.4.1

F — force vector: 6.5.4
Fe — electromotive force: 2.1.6
Fm — mechanical force: 2.1.1
F — reduced external force: 8.2.1
G — free enthalpy: 5.5.8
G — modal forces: 8.2.9
Physical Quantities xxxiii

H — friction force on a belt: 2.5.1

— enthalpy: 5.5.8

H — magnetic field vector: 2.1.9
I — moment of inertia: 4.7.3
I — deformation vector: 6.7.9
J — electric current: 4.7.1
— heat flux: 6.4.2

J — electric current vector: 2.1.9
J ± — invariants of acoustic horns: N7.28
K — longitudinal wavenumber: N5.14
— curvature quadratic form: 6.7.3
— reduced wavenumber: N7.29

K — enhanced wavenumber: N7.33
L — induction of a coil or self: 2.1.6, 4.4.7.1
— lift force: 4.8.13
— lengthscale of variation of wave speed: N5.19
LA — lengthscale of variation of the cross-sectional area: N7.17
Lb — lengthscale of variation of the width of a water channel: N7.12
Lc — lengthscale of variation of the torsional stiffness: N7.10
LE — lengthscale of change of the Young modulus: N7.11
LT — lengthscale of change of tension: N7.9
Lε — lengthscale of variation of the dielectric permittivity: N7.15
Lµ — lengthscale of variation of the magnetic permeability: N7.15
M — friction moment on a cam: 2.5.4
— molecular mass: 5.5.17
— moment of the inertia forces: N6.4

M — moment of forces: 6.8.4
M1 , M2 — principal bending moments: 6.7.2
Ma — applied axial moment: N6.4
Mn — normal stress coupled: 6.8.16
— mass of n-th element of a radioactive disintegration
chain: 8.7.1
Mx , My — stress couples: 6.7.4
Mxy — twist couple: 6.7.4
Mrs — modal matrix: 8.2.13
N — number of particles: 5.5.17
xxxiv Physical Quantities


N — unit outer normal to a surface: 5.5.1
— turning moment: 6.7.6
N n — normal turning moment: 6.8.16
N n — augmented normal turning moment: 6.9.5
N x , N y — turning moments: 6.7.6
N rs — undamped modal matrix: 8.2.14
P — reduced pressure perturbation spectrum: N7.23
P2 N — dispersion polynomial: 8.2.2
Pij — dispersion matrix: 8.2.2
Q — heat: 3.9.11
— heat source: 6.4.2
— concentrated torque: E10.10.1
Qr — transformation matrix: 8.2.7
R — electrical resistance: 2.1.6, 4.7.1, 8.9.2
— perfect gas constant: 5.5.17
— radius of curvature: 6.1.1
— reflection coefficient: N7.47
S — entropy: 3.9.11
— dynamo parameter: 4.7.3
— strain: 6.4.1
— surface adsorption coefficient: N7.48
Sa — row of components of the strain tensor: 6.8.7
Sij — strain tensor: 6.5.1, 6.8.8
— scattering matrix: N7.53
T — temperature: 3.9.11
— thrust: 4.8.13
— tangential tension along an elastic string: 6.1.1
— transmission coefficient: N7.51
T — effective axial tension: 6.3.12

T — stress vector: 6.6.2, 6.8.3
Ta — row of components of the stress tensor: 6.8.7
Tij — stress tensor: 6.5.1, 6.8.3
o
Tij — residual stresses: 6.8.6
U — internal energy: 3.9.11
Physical Quantities xxxv

V — specific volume: 5.5.4

— reduced velocity perturbation spectrum: N7.23
W — work: 2.2.2, 3.9.11, 5.5.4
— weight: 4.8.13
Wb — work of deformation: 6.8.5
Wv — work of the inertia force: 6.8.5
Y — inductance: 8.9.1
Z — impedance: 4.7.2, 8.9.1
Z — overall impedance: N7.52
Z0 — impedance of a plane sound wave: N7.47

3 Small Greek Letters

α — phase of an oscillator: 2.2.1
— angle of cylindrical helix: 3.9.5
— coefficient of the cubic term of the quartic potential: 4.5.1
— diffusivity: N8.1
α e — ohmic electrical diffusivity: 2.1.11
β — coefficient of the quartic term in the biquadratic potential: 4.4.4
— amplification/attenuation factor: N7.50
— potential: N8.1
χ — thermal conductivity: 6.4.2
— stiffness ratio parameter: 6.8.16
— damping parameter: 6.8.14
δ ij — identity matrix: 6.5.1, 6.8.4, 8.1.4
ε — dielectric permittivity: 2.1.9
ϕ ± — phase of waves propagating in opposite directions: N6.14
γ — dimensionless non-linearity parameter: 4.4.8
— adiabatic exponent: 5.5.12
λ — damping: 2.1.8
λ  — modal dampings: 8.2.5
λ rs — damping matrix: 8.2.1
µ — kinematic friction coefficient: 2.1.3, N6.10
— magnetic permeability: 2.1.9
xxxvi Physical Quantities

µ rs — kinematic friction matrix: 8.1.3

ν — number of particles per unit volume: 5.5.17
— resilience of distributed translational spring: N6.10
νn — disintegration rate of the mass of the n-th element in a chain: 8.7.1
θ — angle of inclination of a string or a beam: 6.1.11
— temperature: 6.4.2
ρ — mass density per unit volume: 5.8.11
ρ1 — mass density per unit length: N6.3
ρ2 — mass density per unit area: N6.4
σ — Ohmic electrical conductivity: 2.1.9
— Poisson ratio: 6.5.1
— mass density per unit length: N7.8
τ — period of oscillation 2.1.12
— twist: 6.7.13
— torsion: N6.5
τ — decay rate: 2.1.12
ω 0 — natural frequency: 2.1.8
ω 1 — fundamental natural frequency: N6.13
ω n — natural frequency of n-th harmonic: N6.13
ω  — modal frequencies: 8.2.4
ω — oscillation frequency: 2.3.3, 2.4.1
ω n — oscillation frequency of n-th harmonic: N6.17
ω  — oscillation frequencies of modes: 8.2.6
ω — average of applied and natural frequencies: 2.7.5
ω a — applied frequency of sinusoidal forcing: 2.7.1
ω e — excitation frequency of parametric resonance: 4.3.2
ω r — natural frequency for rotary spring: N6.10
ω t — natural frequency for translational spring: N6.10
ω r 1 , ω r 2 — natural frequencies for a vector rotary spring: N6.10
ω 2rs — oscillation matrix: 8.2.1
ψ — non-linearity parameter for a quartic potential: 4.6.1
ζ — transverse deflection of a string, beam, membrane or plate: 6.1.1
Physical Quantities xxxvii

4 Capital Greek Letters

Φ — primal wave variable: N7.19
Φ e — scalar electric potential: 2.1.10
Φ m — mechanical potential energy: 2.1.4, 8.1.2
ϑ — resilience of a rotary spring in one dimension: N6.10
ϑ — resilience vector of a rotary spring in two dimensions: N6.10
Θ — stress function: 6.5.5
Ω — angular velocity of rotation: 4.7.1
Ωij — rotation bivector: 6.8.6
Ψ — dual wave variable: N7.19
Ψ e — dissipation by electrical resistance: 2.1.7
Ψ m — dissipation by mechanical friction: 2.1.4, 8.1.2
10
Examples 10.1 to 10.20

EXAMPLE 10.1 One Finite Difference (Two Differential)

Equation(s) with the Same Characteristic
Polynomial
Solve two linear forced third-order differential equations with constant
(homogeneous) coefficients [E10.1.1 (E10.1.2)] and the same characteristic
polynomial as a linear forced third-order finite difference equation with con-
stant coefficients (E10.1.3).

E10.1.1  Linear Differential Equation with Constant Coefficients

Consider the forced linear third-order differential equation with constant
coefficients:

y ′′′ + y ′′ − 2 y = x 2 + x + 1 + e 3 x + e x + sinh ( 2 x )
(10.1)
+ e −2 x cosh x + e 2 x sin x + e − x cos x + e x sin x

and obtain the complete integral. The linear third-order ordinary differential
equation (10.1) has characteristic polynomial (10.2a–c) of ordinary derivatives
(10.2d):

( )
P3 ( D ) = D3 + D2 − 2 = ( D − 1) D2 + 2 D + 2 = ( D − 1) ( D + 1 − i ) ( D + 1 + i ) ,
(10.2a–c)

d
D≡ : P3′ ( D ) = 3D2 + 2 D = D ( 3D + 2 ) ; (10.2d, e)
dx

the roots (10.3a) of (10.2c) specify the general integral (10.3b) of the unforced
equation:

D = 1, − 1 ± i: y ( x ) = A e x + e − x ( B cos x + C sin x ) , (10.3a, b)

1
2 Differential Equations and Applications

where A, B, C are arbitrary constants of integration. The polynomial forc-

ing term has a particular integral specified by the inverse polynomial of
derivatives:

1 1 1
(
P3 ( D )
x2 + x + 1 = − )
2 1 − D2 ( 1 + D )/2
(
x2 + x + 1 )
(10.4)
1 D2  x2 x
  = − 1 +
2 2
( ) (
+ O D3  x 2 + x + 1 = − )− − 1.
2 2


The remaining particular integrals in (10.1) correspond to the following cases:

1 x2 x
y ( x) − A x −
x
( B cos x + C sin x ) + + +1
2 2

e3x x ex 1  e2x e −2 x  1  e − x e −3 x 
= + +  −  +  + 
P3 ( 3 ) P3′ ( 1) 2  P3 ( 2 ) P3 ( −2 )  2  P3 ( −1) P3 ( −3 ) 

 e(2 + i)x   x e( i − 1)x   e ( 1+ i ) x 

+ Im   + Re   + Im  ,
 P3 ( 2 + i )   P3′ ( i − 1)   P3 ( 1 + i ) 
(10.5)

because; (i) the exponential exp ( 3x ) is non-resonant (1.129a, b); (ii) the
exponential exp(x) is resonant (1.134a, b); (iii) the hyperbolic sine sinh(2x)
is non-resonant (1.147a, b); (iv) the product of exponential exp ( −2x ) and
hyperbolic cosine cosh(x) is not resonant (1.165a, b); (v) the product of the
exponential exp(2x) and circular sine sin(x) is also non-resonant (1.173a, b);
(vi) the product of the exponential exp ( −2x ) by the circular cosine cos x is
resonant (1.174a–c); (vii) the product of the exponential exp(x) by the circu-
lar cosine cos(x) is non-resonant (1.176a, b). The characteristic polynomial
(10.2b) and its first-order derivative (10.2e) evaluate the coefficients (10.6a–i)
in (10.5):

1 1 1 1 1 1 1 1
= , = , = , = − , (10.6a–d)
P3 ( 3 ) 34 P3′ ( 1) 5 P3 ( 2 ) 10 P3 ( −2 ) 6

1 1 1 1 1 1 1 − 5i
=− , =− , = = , (10.6e–g)
P3 ( −1) 2 P3 ( −3 ) 20 P3 ( 2 + i ) 3 ( 1 + 5 i ) 78

1 1 1− 2i 1 1 1+ i
=− =− , = =− , (10.6h, i)
P3′ ( i − 1) 2 (1 + 2i ) 10 P3 ( 1 + i ) 4 ( i − 1) 8
Examples 10.1 to 10.20 3

Substitution of (10.6a–i) in (10.5) leads to the complete integral (10.7) of (10.1):

x2 + x + 2 e3x
y ( x ) = A e x + e − x (B cos x + C sin x) − +
2 34
x e x e 2 x e −2 x e − x e −3 x e 2 x
          +
5
+
20
+
12
−
4
−
40
+
78
( sin x − 5 cos x ) (10.7)
x e− x ex
          −
10
( cos x + 2 sin x ) − ( sin x + cos x ) ,
8

where (A, B, C) are arbitrary constants of integration.

E10.1.2  Linear Differential Equation with Homogeneous Coefficients

Consider the forced linear third-order differential equation with homoge-
neous coefficients:

x 3 y ′′′ + 4 x 2 y ′′ + 2 x y ′ − 2 y = log 2 x + log x + 1 + x 3 + x

cosh ( log x ) cos ( log x )
         + sinh ( 2 log x ) + + x 2 sin ( log x ) +
x 2
x (10.8)
         + x sin ( log x ) ,

and obtain the complete integral. The characteristic polynomial (10.9b–d) of

homogeneous derivatives (10.9a):

d
δ≡x : P3 ( δ ) = δ ( δ − 1)( δ − 2 ) + 4 δ ( δ − 1) + 2δ − 2
dx (10.9a–d)
                                (
= ( δ − 1) δ ( δ + 2 ) + 2  = ( δ − 1) δ + 2 δ + 2 ,
2
)
coincides with (10.9d) ≡ (10.2c). Thus, the substitutions (10.10a, b), which imply
(10.10c, d):

e x ↔ x, x ↔ log x: e ax ↔ x a , x n ↔ log n x , (10.10a–d)

interchange the complete integrals (10.7) [(10.11)]:

B cos ( log x ) + C sin ( log x ) log 2 x + log x + 2 x 3 x log x

y ( x ) = Ax + − +  +
x 2 34 5
x2 1 1 1 x2
            + − − − +  sin ( log x ) − 5 cos ( 2 log x )  (10.11)
20 12 x 2
4 x 40x 3
78 
log x x
            −  cos ( log x ) + 2 sin ( 2 log x )  −  cos ( log x ) + sin ( log x )  ,
10 x  8
4 Differential Equations and Applications

of the forced linear third-order differential equations with constant (10.1)

[homogeneous (10.8)] coefficients.

E10.1.3  Linear Finite Difference Equation with Constant Coefficients

Consider the forced linear third-order finite difference equation with con-
stant coefficients:

y n+ 3 + y n+ 2 − 2 y n = 3n + 2 n− 1 + ( −2 )
n− 1
+
( − )n + ( −3 )n , (10.12)
2 2
and obtain the complete solution. The characteristic polynomial of forward
finite differences (10.13a) of (10.12) is (10.13b) ≡ (10.9d) ≡ (10.2c):

∆y n = y n+ 1 : P3 ( ∆ ) = ∆ 3 + ∆ 2 − 2; y ( x ) ↔ y n , y ( k ) ( x ) ↔ y n+ k , e ax ↔ a n ,
(10.13a–e)
the changes (10.13c–e) imply:

{ }
n
e( ↔ ( a + i b) =
a + ib ) x 1/2
exp  i arc tan ( b/a ) 
n
a2 + b2
(10.13f, g)
2 n/2
                                = a + b 2
exp  i n arc tan ( b/a )  ,

for example:

e − x ± i x ↔ ( −1 ± i ) = ( )
n n
2 e ± i 3π/4 = 2 n/2 e ± i 3π n/4
(10.13h, i)
= 2 n/2  cos ( 3 π n/4 ) ± i sin ( 3 π n/4 )  ,
and also:

( i a ) ± ( − i a ) = an e iπ/2 ± e − iπ/2 ( ) ( )
n n n n
e iax ± e − iax
cos, sin ( ax ) = ↔
{2, 2i} {2, 2i} {2, 2i}
(10.13j–m)
e inπ/2 ± e − inπ/2
= an = a n cos, sin ( nπ/2 ) .
{2, 2i}
Use of the relations (10.13a–c, h, i) transforms: (i) the differential equation
(10.1) with the fourth, sixth, and seventh terms on the right-hand side (r.h.s.)
to the finite difference equation (10.12); (ii) the corresponding terms of the
complete integral (10.7) to solution (10.14):

2 n− 2 ( −2 ) ( − ) − ( −3 ) . n− 2 n n
  πn   π n  3
n
y n = A + 2 n/2  B cos   + C sin    + + − −
  4   4   34 5 3 4 40
(10.14)
Examples 10.1 to 10.20 5

The solution (10.14) [complete integral (10.11)] of the linear finite difference
(differential) equation (10.12) [(10.8)]: (i) have been obtained using the trans-
formations (10.13a–m) [(10.10a–d)] applied to the equivalent linear differ-
ential equation with constant coefficients (10.1) and its complete integral
(10.7); (ii) can be confirmed using directly the methods in the section(s) 1.9
(1.7–1.8).

EXAMPLE 10.2 Energies, Dissipation, and Power

of an Oscillation
For a linear harmonic oscillator, obtain the exact (a) time dependences and
(b) average over a period starting at any time of: (i) the potential, kinetic,
and total energies and the dissipation for free damped subcritical oscilla-
tions, and check the agreement with the particular case of weak damping
(E10.2.1–E10.2.2); (ii) the potential, kinetic, and total energies and the power
of the applied force for the undamped forcing with resonance, and check the
agreement with the particular case of slow amplitude growth in a period
(E10.2.3–E10.2.4).

E10.2.1  Potential, Kinetic, and Total Energies

Omitting the phase shift (10.15a), the potential, kinetic, and total energies of
the harmonic oscillator with subcritical damping (2.106b, c) are given, respec-
tively, by (10.15b–d) using (2.105d):

k k a 2 −2 λt k a 2 −2 λt
α = 0: Φ ( t ) =  x ( t )  = cos 2 ( ωt ) = 1 + cos ( 2ωt )  ,
2
e e
2 2 4
 (10.15a, b)

1 k a 2 −2 λt
Ev ( t ) = m  x ( t )  = ω sin ( ωt ) + λ cos ( ωt ) 
2 2
e
2 2 ω 20

k a 2 −2 λt 2
= e ω sin 2 ( ωt ) + λ 2 cos 2 ( ωt ) + 2 λω cos ( ωt ) sin ( ωt )  (10.15c)
2 ω 20

k a 2 −2 λt  2 λω λ2 − ω2 
= e  1 + 2 sin ( 2ωt ) + cos ( 2ωt )  ,
4  ω 0 ω 2
0 

k a 2 −2 λt  λω λ2 
E ( t ) = Φ ( t ) + Ev ( t ) = e  1 + 2 sin ( 2ωt ) + 2 cos ( 2ωt )  . (10.15d)
2  ω 0 ω 0 
6 Differential Equations and Applications

The exact average values over a period starting at an arbitrary time t and
ending at t + τ1 are given by:

Φ ( t ) , Ev ( t ) , E ( t )
k a2  2 λω λ2 − ω2 2 λω 2λ 2 
          = I0 + I+ , I0 + I− + I+ , 2 I0 + 2 I− + 2 I+  ,
4  ω02
ω02
ω0 ω0 
(10.16a–c)

involving the integrals:

t + 2 π/ω
t+ τ
ω  e −2 λt  ω −2 λt
I0 =
1
τ ∫ t
e −2 λt
dt =
2 π  −2 λ t
=
4π λ
e (
1 − e −4 πλ/ω , (10.17a) )
1 t+ τ
ω  t + 2 π/ω

I± =
τ ∫t
e −2 λt cos, sin ( 2ω t ) dt =
2π
Re, Im 
 ∫
t
e −2 λt + i 2 ω t dt 


=−
ω
4π
Re, Im
λ − iω 
e {
1  −2( λ− iω )t  t + 2 π/ω
t } (10.17b)

ω  λ + iω −2( λ− iω )t
=
4π
Re, Im  2
λ + ω
2
e

1 − e −4 πλ/ω  ;

( )
the latter splits into:

ω 1 − e −4 πλ/ω −2 λt
{I + , I − } = e
4π ω 20 (10.18a, b)
{
                  ×   λ cos ( 2ω t ) − ω sin ( 2ω t ) , λ sin ( 2ω t ) + ω cos ( 2 ω t ) . }
Substituting (10.17a; 10.18a, b) in (10.16a–c) specifies the exact average over a
period (10.19a) of, respectively. (10.19b/c/d) the potential (10.15b), kinetic (10.15c).
and total energies (10.15d) of a subcritically damped harmonic oscillator:

k a 2 −2 λt
( t, t + τ ) ≡ ( t, t + 2π/ω ) : Φ ( t ) , Ev ( t ) , E ( t ) =
16π
e (
1 − e −4 πλ/ω )
 ω λω ω2 ω ωλ
×  + 2 cos ( 2ω t ) − 2 sin ( 2ω t ) , + 2 cos ( 2ω t )
 λ ω 0 ω 0 λ ω0
ω2 ω 2ω λ 
+ sin ( 2ωt ) , 2 + 2 cos ( 2ω t )  .
ω 20 λ ω0 
(10.19a–d)
The total energy decays for strong (E10.2.1) or weak (E10.2.2) damping.
Examples 10.1 to 10.20 7

E10.2.2  Strong Subcritical or Weakly Damped Oscillations

The dissipation (10.20a) is related to the kinetic energy (10.15c) by (10.20b):

2µ
Ψ ( t ) = − µ  x ( t )  = − Ev ( t ) = − 4 λ Ev ( t ) ; (10.20a–c)
2

m

thus, the average of the dissipation over a period (10.19a) is given (10.19c) by
(10.21b):

k a 2 −2 λt
Ψ ( t ) = − 4 λ Ev ( t ) = −
4π
e (
1 − e 4 π λ/ω )
(10.21a, b)
 ω λ2 ω2λ 
                ×   ω + 2 cos ( 2 ω t ) + 2 sin ( 2ω t )  .
 ω0 ω0 

The time derivative of the total energy (10.15d) is balanced by the dissipa-
tion (10.20a) at any time in the absence of external forces, that is, A = 0 in
(2.12); thus, the same relation applies to the time derivative of the average
of the total energy over a period (10.19d) that is given (10.20a–c; 10.21a, b) by
(10.22a):

d  E ( t )  k a 2 −2 λt
dt
=
4π
e (
1 − e 4 πλ/ω )
(10.22a, b)
 ω λ2 ω2 λ 
                      × ω + 2 cos ( 2ω t ) + 2 cos ( 2ω t )  = − ψ ( t ) ,
 ω0 ω0 

minus the dissipation (10.21b), both averaged over a period (10.22b). In the
case of weak damping (10.23a), substitution of (10.23b) gives (10.23c):

4π λ 4π λ
λ 2 << ω 20 ∼ ω 2 : 1 − e −4 πλ/ω ∼ ∼ . (10.23a–c)
ω ω0

Substituting in (10.19b–d) yields, respectively, the mean values over a period

of the potential (10.24a), kinetic (10.24b), and total (10.24c, d) energies:

k a 2 −2 λt 4πλ  ω 0 λ 
Φ (t ) = e  + cos ( 2ω 0 t ) − sin ( 2ω 0 t ) 
16π ω0  λ ω0 
(10.24a)
k a 2 −2 λt  λ  λ2  
= e 1 − sin ( 2ω 0 t ) + O  2   ,
4  ω0  ω0  
8 Differential Equations and Applications

k a 2 −2 λt 4πλ  ω 0 λ 
Ev ( t ) = e + cos ( 2ω 0 t ) + sin ( 2ω 0 t ) 
16π ω 0  λ ω 0 
(10.24b)
k a 2 −2 λt  λ  λ2  
= e 1 + sin ( 2ω 0 t ) + O  2   ,
4  ω0  ω0  

k a 2 −2 λt 4πλ ω 0 k a 2 −2 λt
λ 2 << ω 20 : E (t ) = e 2 = e . (10.24c, d)
16π ω0 λ 2

The further assumption (10.25a) that is met by (10.25b) leads to the equiparti-
tion of potential energy and kinetic energy (10.25c–e):

ω0 k a 2 −2 λt E (t )
sin ( 2ω 0 t ) << : ω 0 >> λ , e = Φ ( t ) = Ev (T ) = . (10.25a–e)
λ 4 2

Thus, for (problem 10) a subcritically damped oscillator (2.106a–c) omitting the
phase (10.15a): (i)(ii) the instantaneous value (average over a period) are given by
(10.15b/c/d) [(10.19b/c/d)] for the potential/kinetic/total energies and by (10.20a–c)
[(10.21a, b)] for the dissipation; (iii) in the case (problem 11) of weak damping (10.24c)
the total energy (10.24d) decays at twice the damping rate; (iv) with the further
assumption (10.25a) there is equipartition (10.25c–e) of potential and kinetic ener-
gies; (v) the assumption (10.25a) is met by (10.25b), which implies negligible damp-
ing, except in the exponential term exp( −2 λ t ), which is much smaller than unity for
2 t > 1/λ but is close to unity for short time t << 2/λ .

E10.2.3  Averages over a Period of Energies and Power

In the case of the resonance of an undamped oscillator, the potential (2.177a),
kinetic (2.177b), and total (2.177c) energies, should be considered, respec-
tively (10.26a/b/c), and also the activity or power (2.180a) of the applied force
(10.26d):

k k f a2 2 m f a2 2
Φ (t ) =  x* ( t )  = t sin 2 ( ω 0 t ) = t sin 2 ( ω 0 t ) , (10.26a)
2
2
2 8ω 0 8

m m f a2
 sin ( ω 0 t ) + ω 0 t cos ( ω 0 t ) 
2
Ev ( t ) =  x * ( t )  =
2
2 
2 8ω0
(10.26b)
m f a2  2
sin ( ω 0 t ) + ω 0 t sin ( 2 ω 0 t ) + ( ω 0 t ) cos 2 ( ω 0 t )  ,
2
          =
8 ω 20  

m 2 2 t 1 
E ( t ) = Φ ( t ) + Ev ( t ) = f a t + sin ( 2 ω 0 t ) + 2 sin 2 ( ω 0 t )  , (10.26c)
8  ω0 ω0 
Examples 10.1 to 10.20 9

m f a2
A ( t ) = f ( t ) x * ( t ) = cos ( ω 0 t )  sin ( ω 0 t ) + ω 0 t cos ( ω 0 t ) 
2 ω0 (10.26d)
m f a2
                             =  sin ( 2 ω 0 t ) + 2 ω 0 t cos 2 ( ω 0 t )  .
4ω 0 
The averages over a period starting at time t and ending at time t + τ 0 are
given by:

m f a2
Φ ( t ) , Ev ( t ) , E ( t ) , A ( t ) =
8 ω 20 (10.27a–d)
1
2{ 1
                    × J − , + J 0 + J + , + J 0 + J + + J − , 4 ω 0 J1 ,
2 }
involving besides (2.116c; 2.117a, b) the integrals:
t +τ 0
{J , J , J } = 1
0 1 ±
τ0 t ∫ {
ω 0 t sin ( 2ω 0 t ) , ω 0 t cos 2 ( ω 0 t ) , ω 20 t 2 cos 2 , sin 2 ( ω 0 t ) dt , }
(10.28a–d)
to be evaluated exactly.
The first integral (10.28a) is evaluated by a single integration by parts:
t +τ 0
1 d
J0 = −
2τ0 ∫t
t  cos ( 2ω 0 t )  dt
dt 
1 ω 0 t + 2 π/ω 0 (10.29)
∫
t +τ 0
     = −  t cos ( 2ω 0 t )  t + cos ( 2ω 0 t ) dt
2τ0 4π t
1 1 t +τ 0 1
     = − cos ( 2ω 0 t ) +  sin ( 2ω 0 t )  t = − cos ( 2ω 0 t ) .
2 8π 2

The second integral (10.28b) is also evaluated by parts:

t +τ 0
1
J1 =
2τ0 ∫t
ω 0 t 1 + cos ( 2ω 0 t )  dt

{ }
t +τ 0
ω 0 2 t +τ0 1 d
= t 
4τ 0   t
+
4τ 0 ∫ t
t  sin ( 2ω 0 t )  dt
dt 
ω0  1 1 t +τ0 (10.30)
∫
t +τ 0
= ( t + τ ) 2
− t 2
+ 
 t sin ( 2 ω t ) 
 − sin ( 2 ω t ) dt
4τ 0   4τ 0
0 0 0
t
4τ 0 t
ω 1 1
= 0 ( 2 t + τ 0 ) + sin ( 2ω 0 t ) +
t +τ 0
 cos ( 2ω 0 t )  t
4 4 8τ 0 ω 0 
ω t π 1
= 0 + + sin ( 2ω 0 t ) ,
2 2 4
so that only the remaining pair is (10.28c, d).
10 Differential Equations and Applications

The last two integrals (10.28c, d) are evaluated by:

t +τ 0
1
J± =
2τ0 ∫t
( ω 0t )2 1 ± cos ( 2ω 0t ) dt

{ }
t +τ 0
ω 20 3 t +τ0 1 d
=
6τ 0
t 
t
±
4τ 0 ω 0 ∫t
( ω 0 t )2  sin ( 2ω 0 t )  dt
dt 

t +τ 0
ω 20  1  1
( ) ∫
t +τ 0
− t 3  ±
3
= t + τ0 ( ω 0 t ) sin ( 2ω 0 t ) 
2
 ω 0 t sin ( 2ω 0 t ) dt
6τ 0   8π  t 2τ0 t

ω 20 1 2 1
6
= (
3 t 2 + 3 t τ 0 + τ 02 ±
8π
)
ω 0 τ 0 ( 2 t + τ 0 ) sin ( 2ω 0 t ) ± cos ( 2ω 0 t ) ;
4
(10.31)

the last term on the r.h.s. of (10.31) is J 0 in (10.28a), which is specified by (10.29),
so that no more integrations by parts are needed:

ω 20 t 2 2π 2  ω 0t π  1
J± = + π ω0 t + ± +  sin ( 2ω 0 t ) ± cos ( 2ω 0 t ) , (10.32a, b)
2 3  2 2 4

to obtain the integrals (10.28c, d).

E10.2.4  Balance of Forcing and Fast Amplification

The integrals (10.32a, b) together with (10.29; 10.30) ≡ (10.33a, b):

1 π ω 0t 1
J 0 = − cos ( 2ω 0 t ) , J1 = + + sin ( 2ω 0 t ) , (10.33a, b)
2 2 2 4

substituted in (10.27a–d) yield (10.34a–d):

m f a2
Φ ( t ) , Ev ( t ) , E ( t ) , , A ( t ) =
16 ω 20

4π 2 1
{
× ω 20 t 2 + 2 π ω 0 t +
3
− ( π + ω 0 t ) sin ( 2ω 0 t ) − cos ( 2ω 0 t ) ,
2

4π 2 1
1 + ω 20 t 2 + 2 π ω 0 t + + ( π + ω 0 t ) sin ( 2ω 0 t ) − cos ( 2ω 0 t ) ,
3 2

8π 2
− cos ( 2ω 0 t ) , 4 π ω 0 + 4ω 20 t + 2ω 0 sin ( 2ω 0 t ) } .
1 + 2ω 20 t 2 + 4π ω 0 t +
3
(10.34a–d)
Examples 10.1 to 10.20 11

The activity of the applied forces (10.34d) equals the rate of change of the
total energy (10.34c):

d d  f a2 m 
 E ( t )  =   2ω 20 t 2 + 4πω 0 t − cos ( 2ω 0 t )  
dt dt  16ω 20   (10.35)
2
fa m
                   =  4 ω 0 t + 4πω 0 + 2ω 0 sin ( 2ω 0 t )  = A ( t ) .
2
16ω 20 

If has been proved that an undamped harmonic oscillator (problem 29) corre-
sponds to: (i) the displacement (2.176b) and velocity (2.176c); (ii) the potential energy
(2.177a) ≡ (10.26a), the kinetic-energy (2.177b) ≡ (10.26b), and the total energy
(2.130) ≡ (10.26c), and the activity (2.180a) ≡ (10.26d) of the applied force; (iii) their
exact average over the period from t to t + τ 0 is given exactly by (10.34a–d); (iv) the
terms growing fastest with time t are:

 fat 
2
E (t ) m f a2 t d
Φ ( t ) = Ev ( t ) = m  = , A ( t ) = =  E ( t )  , (10.36a–c)
 4  2 4 dt

in agreement with (2.178a–c; 2.179b) valid for amplitude varying slowly over a wave
period; (v) within the latter approximation (iv) there is equipartition of the potential
(10.36a) and kinetic (10.36b) energies, whose sum is the total energy (10.36c), grow-
ing like the square of time t due to the work of the applied forces growing linearly
in time.

EXAMPLE 10.3  Power-Law Forcing of a Harmonic Oscillator

Determine the forced displacement for a one-dimensional, linear, undamped
(10.37a) harmonic oscillator with natural frequency ω 0:

λ = 0: x + ω 20 x = f a t n , f a t cos ( ω a t ) , (10.37a–c)

to which an external force is applied whose dependence on time t is specified

by: (i) a power (10.37b); (ii) time multiplied by a sinusoidal function (10.37c)
with frequency ω a. Consider also the forcings:

x + 2 λ x + ω 20 x = f a t , f a t cos ( ω a t ) , (10.38a, b)

in the damped case with damping λ.

12 Differential Equations and Applications

E10.3.1  Power-Law Monotonic and Oscillating Forcing

The undamped (10.37a) harmonic oscillator (10.37b, c) corresponds to the
characteristic polynomial (10.39a):

∂n at
P2 ( D ) = D2 + ω 20 : f0 t n = f0 lim
a→ 0 ∂an
( )
e , (10.39a, b)

and the forcing to (10.37b) ≡ (10.39b). The corresponding response is (problem 39)
obtained using the inverse of the characteristic polynomial:

−1 ∞
 D2  n ( − )m D 2 m t n
∑ω
fa n fa fa
λ = 0:          x* ( t ) = t = 2  1 + 2 
t = 2
P2 ( D ) ω0 ω0  ω0 m= 0
2m
0
(10.40a, b)
≤ n/2
( − )m
∑ω
fa n!
                               = 2 t n− 2 m .
ω0 m= 0
2m
0 ( n − 2 m)!

The forcing (10.37c) leads to the response:

 fa   1 
x* ( t ) = Re  t e iω at  = Re  f a e iω at t
P
 2 ( D )   P2 ( D + iω a ) 
(10.41)
 f a e iω at   f a e iω at 
= Re  t  = Re  2 t .
 ( D + iω a ) + ω 0
2 2 2
 ω0 − ωa + 2 i ωaD + D
2
 

This splits into: (i) the non-resonant (10.42a, b) case (problem 40):

fa   2 iω a D + D2  
−1

λ = 0, ω 0 ≠ ω a : x* ( t ) = Re e iω at  1 + t
2 2
ω0 − ωa  ω 20 − ω 2a  
 

fa   2 i ωaD 2  
= 22
ω − ωa
Re e iω at 1 − ω 2 − ω 2 + O D  t 
  
( )
0  0 a
(10.42a–c)
fa   2 iω a  
= Re e iω at  t − 2 
2
ω − ωa
0
2
  ω 0 − ω 2a  

fa  2 ωa 
=
ω − ω 2a
2 t cos ( ω a t ) + ω 2 − ω 2 sin ( ω a t )  ,
0  0 a 

for which the amplitude of the displacement (10.42c) is a linear func-

tion of time t like the forcing (10.37c) with time O(t) terms in phase; (ii) in
Examples 10.1 to 10.20 13

(problem 41) the resonant case (10.43a, b) does not hold (10.42c) and (10.41)
leads to (10.43c):

 f a e iω at   f e iω at  D  
−1

x* ( t ) = Re 
a
λ = 0: ω 0 ≡ ω a : t  = Re   1 + 2 i ω  t 
 D ( 2 i ω 0 + D )   2 i ω 0 D 0 

 i f a e iω at  D    i f a e iω at  1  
= − Re   1 − 2 i ω  t  = − Re   t − 2 i ω  
 2 ω 0 D 0   2 ω 0 D 0 

 i f a iω at  t 2 t   f a t 2 fa t
= − Re  e  −  = sin ( ω 0 t ) + cos ( ω 0 t ) ,
 2 ω 0  2 2 i ω 0   4 ω 0 4 ω 20
(10.43a–c)
showing that the amplitude of the displacement is a quadratic function
of time t (10.43c) instead of linear for the forcing (10.37c) and the highest
( )
order term O t 2 in (10.43c) is out-of-phase by π/2 relative to the forcing
(10.37c), that is, O ( t ). Setting (10.44b) in (10.42c) leads to (10.44d):
fa t
λ = 0 , ω a = 0 , n = 1: x* ( t ) = , (10.44a–d)
ω 20

which is the same result (10.44d) ≡ (10.40b) for (10.44c). It has been shown that
the undamped (10.37a) harmonic (10.37b) oscillator with forcings (10.37b) [(10.37c)]
has displacements (10.40a, b) [(10.42a, c)/(10.43a, c)], respectively, in the [non-
resonant/resonant cases]. The non-resonant (resonant) cases differ in that the ampli-
tude of the displacement is a function of time t with the same exponent (exponent one
unit higher) than the forcing and the highest powers are in-phase (out-of-phase by π/2).

E10.3.2  Forcing of an Undamped and Damped Oscillator

The undamped (damped) harmonic oscillator (10.37a, b)[(10.38a)] has the
characteristic polynomial (10.39a) [(10.45a)] and the forcing term in (10.37c)
[(10.38a)] is (10.39b) [(10.45b)]:
∂ at
P2 ( D ) = D2 + 2 λ D + ω 20 : f0 t = f0 = lim
a→ 0 ∂a
( )
e . (10.45a, b)

The forcing of the damped harmonic oscillator by a function linear on time t

(10.38a) has response (problem 42):
−1
fa fa fa  2 λ D + D2 
x* ( t ) = t= 2 t= 2 1+  t
P2 ( D ) ω0 + 2 λ D + D 2
ω0  ω 20
(10.46)
fa  2λ 2  fa t 2 fa λ
         = 2
ω0  0
( )
1 − ω 2 D + O D  t = ω 2 − ω 4 .
 0 0
14 Differential Equations and Applications

In the case of λ = 0 of no damping (10.46) simplifies to (10.44d). The forcing

of the damped harmonic oscillator (10.38b) linear on time t with a sinusoidal
factor has response (problem 43):

 fa   f a e iω at 
x* ( t ) = Re  t e iω at  = Re  t
P
 2 ( D )  P
 2 ( D + iω a ) 

 f a e iω at 
= Re  2 t
 ω 0 + 2 λ ( D + iω a ) + ( D + iω a )
2


 f a e iω at 
= Re  2 t
 ω 0 − ω a + 2 iλω a + 2 ( λ + iω a ) D + D
2 2


 f a e iω at  2 ( λ + iω a ) D + D2 
−1

= Re  2 2  1 + 2 2  t
 ω 0 − ω a + 2 iλω a  ω 0 − ω a + 2 iλω a   (10.47)

 2 ( λ + iω a ) D  
 ω 20 − ω 2a − 2 iλω a 
= Re  f a e iω at  1− 2 t 
( )
2
 ω 20 − ω 2a + 4λ 2 ω 2a  ω 0 − ω 2a + 2 iλω a  


= fa t
(ω 2
0 )
− ω 2a cos ( ω a t ) + 2 λω a sin ( ω a t )
(ω )
2 2
0 − ω 2a + 4λ 2 ω 2a

−
(
 ω 20 − ω
2 fa

)
2 2
+ 4λ ω 2 2
2 { 
(
λ  ω 20 + ω 2a )
2
(
− 4 ω 2a ω 2a + λ 2 

)
 a
 a

(
cos ( ω a t ) − ω a  ω 20 − ω 20

)
2
+ 4λ 2 ω 20  sin ( ω a t ) .
 }
In the last term on the r.h.s. of (10.47) was used:

( λ + iω a ) (ω 20 − ω 2a − 2iλω a ) = ( λ + iω a )  ω 20 − ω 2a( ) (
− 4λ 2 ω 2a − 4iλω a ω 20 − ω 2a  )
2 2

 

(
= λ  ω 20 − ω 2a ) (
− 4λ 2 ω 2a + 4ω 2a ω 20 − ω 2a  )
2

 

(
+ i ω a  ω 20 − ω 2a ) (
− 4λ 2 ω 2a − 4λ 2 ω 20 − ω 2a  )
2

 

(
= λ  ω 20 + ω 2a ) (
− 4ω 2a ω 2a + λ 2  )
2

 

+ iω a  ω 20 − ω 2a − 4λ 2 ω 20  . ( )
2

(10.48) 
Examples 10.1 to 10.20 15

In the resonant (10.49a) case (10.47) simplifies to (10.49b):

fa t f a  cos ( ω 0 t ) sin ( ω 0 t ) 
ωa = ω0 : x* ( t ) = sin ( ω 0 t ) +  +  . (10.49a, b)
2 λ ω0 2 λω 0  ω 0 λ 

In (problem 44) the absence (10.42a) of damping (10.47) simplifies to (10.42c).

It has been shown that the damped harmonic oscillator with forcings (10.38a)
[(10.38b)] has displacements (10.46)[(10.47)]. The latter (10.47) simplifies to (10.49b)
[(10.42c)] in the resonant (10.49a) [undamped (10.42a)] case. The undamped (10.43a)
resonant case (10.43b) leads to the displacement (10.43c).

EXAMPLE 10.4 Five Standards of First-Order

Differential Equations
Classify the following five first-order linear differential equations
(10.50a–10.54a):

L:               y ′ = y log x ,            y ( x ) = C e − x x x , (10.50a, b)

y C cos x
LII:             y ′ = − + cos x , y ( x) = + sin x + , (10.51a, b)
x x x

y y2 x
LIII, LXIV: y′ = + ,      y ( x ) = , (10.52a, b)
x x2 C − log x

y 2 − x2
LXIV: y′ = , y ( x ) = x C − 2 log x , (10.53a, b)
xy

−1
1 1 ex  x
eξ 
LV:
x
2
          y ′ = y − y − ; y ( x) = + 2  C +
x x  ∫ dξ , (10.54a, b)
ξ 2 

and use the appropriate method to obtain the respective general integrals
(10.50b–10.54b), where C is an arbitrary constant.
The first differential equation (10.50a) is (subsection 3.2.1), separable (standard L)
and thus, (10.55a) integrable by quadratures leading to (10.55b) ≡ (10.50b):

dy

∫ y = ∫ log x dx, log y = log C + x log x − x. (10.55a, b)
16 Differential Equations and Applications

The second differential equation (10.51a) ≡ (3.28) is (section 3.3) linear

forced (standard LII) with coefficients (10.51a, b), leading (3.20a; 3.31a) to
(10.51c, d):

1  x
dξ  1 (10.56a–c)
P=− ,
x
Q = cos x: X ( x ) = exp −
 ∫ ξ 
 = exp ( − log x ) = ,
x

x x

∫ ∫
Y ( x) = ξ cos ξ dξ = x sin x − sin ξ dξ = x sin x + cos x , (10.56d)

where an integration by parts in (10.56d) was performed; substitution of

(10.56c, d) in (3.31b) leads to the general integral (10.51b).
The third differential equation (10.52a) is both (section 3.4) of a Bernoulli
type (standard LII) and (subsections 3.7.1–3.7.3) homogeneous (standard
LXIV). As a Bernoulli type (3.38) it has coefficients (10.57a, b) leading (3.48a, b)
to (10.57c, d):

1 1  x dξ  1 x
dξ
P = , Q = 2 : X 2 ( x ) = exp −
x x  ξ  x ∫
 = , Y2 ( x ) = −
ξ
= − log x ;
∫
(10.57a–d)

the order is (10.58a) leading on substitution of (10.57c, d) in (3.43b) to (10.58b) ≡

(10.52b):

1 1
n = 2: = ( C − log x ) . (10.58a, b)
y ( x) x

Taken as a homogeneous equation (3.127) then (10.52a) involves the function

(10.59a) leading (3.133) to (10.59b):

y/x
dv C  x
f ( v ) = v + v2 : x = e C exp
∫ v 2
= e exp  −  ,
 y
(10.59a, b)

which can be solved for y leading to (10.52b) ≡ (10.59b) ≡ (10.58b).

The fourth differential equation (10.53a) is (subsections 3.7.1–3.7.3) homo-
geneous (standard LXIV) with (3.127) function (10.60a) leading (3.133) to
(10.60):

1  y /x
  y2 
∫ v dv  = exp  − 2  , (10.60a, b)
f ( v) = v − :          e − C/2 x = exp  −
v    2x 
Examples 10.1 to 10.20 17

which when solved for y gives (10.60b). The fifth differential equation (10.54a)
is of the Riccati type (standard LV), with (3.58) coefficients (10.61a–c) and a
particular integral is (10.61d) leading (3.63a–d) to (10.61e–g):

1 1
Q = − , P = 1, R = − 1 : g1 ( x ) = ,
x x
(10.61a–g)
x
 2 ex 1 x
eξ
               Z ( x ) =  exp
∫  1 − 
ξ
dξ =
x 2
=
X ( x)
, Y ( x) =
∫ ξ2
dξ ;

substitution of (10.61d, e, g) in (3.64b) leads to the general integral (10.54b).

EXAMPLE 10.5 Integrating Factor for the Linear First-Order

Differential Equation
Show that the forced linear first-order differential equation has an integrat-
ing factor that depends only on the independent variable, and use it to obtain
the general integral.
The forced linear first-order differential equation (3.27) corresponds (stan-
dard LII) to the first-order differential in two variables (10.62a, b) with coef-
ficients (10.62c, d), which is inexact because the curl (10.62e, f) is not zero
(10.62g):

0 = dy − ( Py + Q ) dx = A dx + B dy:

∂B ∂ A (10.62a–g)
{ A , B} = {− Py − Q, 1 } , Ω≡ −
∂x ∂y
= − P ≠ 0,

the condition (3.183a) is met (10.63a, b) and thus, there exists (3.183b) an
integrating factor (10.62c, d) depending only on x whose differential is
(10.62e):

Ω  x
 1
−
B
= −P ≡ f ( x) : λ ( x ) = exp −
 ∫ P ( ξ ) dξ  ≡
 X ( x)
, dλ = − P λ dx.

(10.63a–e)

The integrating factor (10.63c) is the same as for the linear unforced first-
order differential equation (3.19a) because it does not depend on Q. The
general integral depends on Q and is obtained multiplying the inexact
18 Differential Equations and Applications

differential (10.62a) by the integrating factor (10.62b) and using its differential
(10.63c) leading to an exact differential (10.64a, b):

0 = λ dy − λ ( Py + Q ) dx = λ dy + y dλ − λ Q dx
(10.64a, b)
 x

= d λ y −
 ∫ λ ( ξ ) Q ( ξ ) dξ  ,


the general integral is (10.65a, b):

x

C = λ ( x) y( x) −
∫ λ ( ξ ) Q ( ξ ) dξ
(10.65a, b)
y ( x) x Q (ξ)
=
X ( x)
−
∫ X (ξ)
dξ ,

which coincides with (3.31a, b) ≡ (10.65a, b), using (10.63e) ≡ (3.20a).

EXAMPLE 10.6  First-Order Differentials in Two Variables

Classify the following eight first-order differentials in two variables
(10.66a–10.73a):

LXVI: (x 2
)
+ y 2 dx + y ( y + 2 x ) dy = 0 , x 3 + y 3 + 3 x y 2 = C, (10.66a, b)

x x2
LXVI : dx − dy = 0,                    x 2 = 2 C y 2 , (10.67a, b)
y2 y3

LXVI: cos x cos y dx − sin x sin y dy = 0,    sin x cos y = C , (10.68a, b)

LXVII: dx + cot x tan y dy = 0,            cos x cos y = C , (10.69a, b)

LXIX: 2 x y dx − x 2 dy = 0,          y = C x 2 , (10.70a, b)

2xy
LXX: dx + ( x − y ) dy = 0,          3 x 2 y − y 3 = C , (10.71a, b)
x+y

LXXI: 3 x y dx + 2 x 2 dx = 0,              x 3 y 2 = C , (10.72a, b)

2x
LXXII: 3 dx − dy = 0 ,         y 2 = C x 3 , (10.73a, b)
y
Examples 10.1 to 10.20 19

into exact (inexact) followed [sections 3.8 (3.9)] by immediate integration

(integration after multiplication by a suitable integrating factor). The roman
numerals at the left identify the standard of the differential equation accord-
ing to the classification 10.1.

E10.6.1  Exact First-Order Differentials in Two Variables

The first three differentials (10.66a–10.68a) have vector of coefficients
(10.74a–10.76a) with zero curl (10.74b–10.76b) and thus, are exact (standard
LXIV) and immediately integrable (10.74c–10.76c):

∂X ∂Y
{X , Y } = {x 2 + y 2 , y 2 + 2 x y } : = 2y = ,
∂y ∂y
(10.74a–c)
 x3 + y 3 
          0 = x 2 dx + y 2 dy + y 2 dx + 2 x y dy = d  + x y2  ,
 3 

x x2  ∂X 2 x ∂Y x dx x 2  x2 
{X , Y } = 
2
, − 3 : =− 3 = , 0 = 2 − 3 dy = d  2  ,
y y  ∂y y ∂x y y  2y 
(10.75a–c)

∂X ∂Y
{X , Y } = {cos x cos y , − sin x sin y} : = − cos x sin y =
∂y ∂x (10.76a–c)
                                 0 = cos x cos y dx − sin x sin y dy = d ( sin x cos y ) ,

leading to the general integrals (10.74c–10.76c) ≡ (10.66b–10.68b).

E10.6.2  Inexact First-Order Differentials in Two Variables

The differentials in two variables (10.69a–10.73a) have a vector of coefficients
(10.77a–10.81a) with non-zero curl (10.77b–10.81b) and hence, are inexact:

{X , Y } = {1, cot x tan y } : Ω = − csc 2 x tan y , λ ( x , y ) = sin x cos y , (10.77a–c)

λ ( X dx + Y dy ) = sin x cos y dy + cos x sin y dy = − d ( cos x cos y ) ; (10.77d)

1
{X , Y } = {2 x y , − x 2 } : Ω = − 4 x, λ ( y ) = : (10.78a–c)
y2

2 x y dx − x 2 dy 2 x x2  x2 
0= 2
= dx − 2 dy = d   ; (10.78d)
y y y  y
20 Differential Equations and Applications

 2xy  2 x2
{X , Y } =  , x − y : Ω = 1− , λ ( x , y ) = x + y: (10.79a–c)
 x + y  ( x + y )2

 2xy   y2 
0 = (x + y) 
dx + ( x − y ) dy  = 2 x y dy + x 2 dy − y 2 dy = d  x 2 y −  ;
x+ y   3
(10.79d)

{X , Y } = {3 x y , 2 x 2 } : Ω = x , λ ( x y ) = x y: (10.80a–c)

( ) (
0  = x y 3 x y dx + 2 x 2 dy = 3 x 2 y 2 dx + 2 x 3 y dy = d x 3 y 2 ; (10.80d) )
 2x  2 x2
{X , Y } = 3, − : Ω=− , λ ( x , y ) = 2 : (10.81a–c)
 y  y y

x2  2x  3 x2 2x3  x3 
0=  3 dx − dy    = dx − dy = d  y 2  ; (10.81d)
y2 y y2 y3

multiplication by the integrating factor (10.77c–10.81c) transforms the inex-

act differentials (10.69a–10.73a) to the exact differentials (10.77d–10.81d) and
specifies the general integrals (10.77d–10.81d) ≡ (10.69b–10.73b). The integrat-
ing factors for inexact differentials are considered next (E10.6.3).

E10.6.3  Integrating Factors for Inexact Differentials

The integrating factor (10.77c) for the (standard LXVII) inexact differential
(10.69a) is found by inspection leading to (10.77d) ≡ (10.69b). The inexact dif-
ferential (10.70a) with coefficients (10.78a, b) and curl (10.78c) meets (standard
LXIX) the condition (3.184a) for (10.82a, b) the existence of an integrating fac-
tor (10.82c, d) ≡ (10.78d) depending only on y:

Ω 2  y
dξ  1
∫
(10.82a, b)
= − = g ( y ): λ ( y ) = exp −2 = .
X y  ξ  y2

The inexact differential (10.71a) with coefficients (10.79a, b) and curl (10.79c)
meets (standard LXX) the condition (3.196a) for the existence (10.83a) of an
integrating factor (10.83b) ≡ (10.79d) depending on x + y:

Ω 1  x+ y
dξ 
∫

= = h ( x + y ): λ ( x + y ) = exp   = x + y. (10.83a, b)
X −Y x+ y  ξ 
Examples 10.1 to 10.20 21

The inexact differential (10.72a) with coefficients (10.80a, b) and curl (10.80c)
meets (standard LXXI) the condition (3.201a) for the existence (10.84a) of an
integrating factor (10.84b) ≡ (10.80d) depending on xy:

Ω 1  xy
dξ 
∫
(10.84a, b)
= = j ( x y ): λ ( x y ) = exp   = x y.
X x −Yy xy  ξ 

The inexact differential (10.73a) with coefficients (10.81a, b) and curl (10.81c)
meets (standard LXXII) the condition (3.206a) for the existence (10.85a) of an
integrating factor (1–0.85b) ≡ (10.81d) depending on y/x:

x2 Ω x y  y  y/x
dξ  x 2
Xx+Yy
= −2 = k :
y x
λ   = exp −2
 x  ∫ =
ξ  y2
. (10.85a, b)

The general integrals of the five inexact differentials in two variables are
obtained using an integrating factor that depends only on: (i) one vari-
able (10.70a; 10.82a, b); (ii–iv) the sum (10.71a; 10.83a, b), the product (10.72a;
10.84a, b), and the ratio (10.73a; 10.85a, b) of two variables; (v) the two variables
(10.70a; 10.77c) in a different way from the preceding (ii–iv).

EXAMPLE 10.7 Paths Near a Stagnation Point of the

Second Degree
Compare the velocity fields specified [problem 77(78)] by (10.86a, b) [(10.89a, b)],
which both have a stagnation point of the second degree at the origin.

E10.7.1  Paths Tangent at a Stagnation Point of the Second Degree

Consider the velocity field:

dx dy
= vx ( x , y ) = x y , = vy ( x , y ) = y 2 − x 2 , (10.86a, b)
dt dt

and determine (problem 77) the paths near the stagnation point.
The paths are specified by the ratio of (10.87a), leading to a first-order ordi-
nary differential equation of degree two, which is not included in the cases
of first degree in section 4.1:

dy y 2 − x 2 y x  y 1
= = − = f  : f ( v) = v − ; (10.87a, b)
dx xy x y  x v
22 Differential Equations and Applications

the first-order ordinary differential equation (10.87a, b) is homogeneous (sec-

tion 3.7) and has (3.133) integral (10.87c, d):
−1
C2 y /x
 1 y2

−
2
+ log x =
∫  − 
v
dv = −
2x2 (10.87c, d)

(
y 2 = x 2 C 2 − 2 log x . )
The integral (10.87d) curves (Figure 10.1). (i) are symmetric relative to
the y-axis (10.88b) and exist in the interval with boundaries (10.88a) and
(10.88 c, d); (ii) pass through the origin (10.88e) and cross (10.88f) the vertical
line x = 1 at y = ± c:

 C2 
0 ≤ x  ( y ) = x ( − y ) ≤ exp    ≡ xmax , y ( 0 ) = 0, y (1) = ± C , (10.88a–f)
 2 

y2 dy
lim = lim ( C − log x ) = ∞ , lim = 0,
x→ 0 x 2 x→ 0 x→± y dx
(10.88g–j)
dy 1
lim = ±C  ,
x→ 1 dx C

y
√2

1 C = √2

1/√2
1 C=1
C=
√2

1
4
√e √e e x

–1/√2

–1

–√2

FIGURE 10.1
An example of singular point of a family of curves or streamlines is the point through which
all streamlines pass with the same tangent direction.
Examples 10.1 to 10.20 23

(iii) the slope is infinite at the origin (10.88g, h) where the tangent is ver-
tical, is zero (10.88i) at the crossing of the diagonal of quadrants and is
(10.88j) at the crossing of the vertical line x = 1. For example the integral
curve: (a) C = 1 crosses the x-axis at the ends of the interval 0 ≤ x ≤ e = 1.649
and has zero slope at x = 1, y = ±1; (b) C 2 crosses the x-axis at the ends
of the interval 0 ≤ x ≤ e = 2.718 and has slope ± 2  1/ 2 at the points
x = 1, y = ± 2 ; (c) C = 1/ 2 crosses the x-axis at the ends of the interval
0 ≤ x ≤ 4 e = 1.284 and has slope ±1/ 2  2 with opposite sign to the case
(b) at x = 1, y = ±1/ 2 .

E10.7.2  Paths with Three Asymptotes Crossing at a Stagnation Point

Consider (problem 78) the velocity field:

dx dy
= vx ( x , y ) = − 2 x y , = vy ( x , y ) = y 2 − x 2 , (10.89a, b)
dt dt

and compare the paths with the preceding case (E10.7.1).

The ratio (10.90a) of (10.89a, b) specifies the paths that again are not included
in (4.9a, b) in section 4.1 and involves (10.90b) one-half of the same homoge-
neous function as in (10.87b):

dy 1 x 2 − y 2 1  x y   y 1 1 
= =  −  = f , f ( v) =  − v  . (10.90a, b)
dx 2 xy 2  y x  x 2v

The integral curves (section 3.7) are specified (3.133) by (10.91a):

1 y/x
v 1  y2  C
log x −
3
log C = 2
∫ 1− 3v
dv = − log  1 − 3 2  ; x 2 − 3 y 2 = .
3  x  x
 (10.91a, b)

The integral curves (10.91b) have asymptotes:

C

x x →∞ x→∞
( x →−0
1
3  )(
C
x  x→ 0
)
= lim x − y 3 x + y 3 , ∞ = lim  x 2 −  = lim y 2 ,
0 = lim

(10.91c–f)

along: (i) the straight lines (10.91c, d) with slopes ±1/ 3 making angles ± π/6
with the x-axis; (ii) along the y-axis (10.91e, f) since y → ±∞ as x → −0. Thus the
asymptotes divide the plane into six angular sectors for ϕ = ± π/6, ± π/2, ± 5π/6
that are the only curves passing through the origin (Figure 10.2).
24 Differential Equations and Applications

π/2

5π/ 6 π/6

7π/ 6 11π / 6

3π / 2

FIGURE 10.2
A different singular point from that in Figure 10.1 is the point through which only six stream-
lines pass, which are the asymptotes of all other streamlines.

E10.7.3  Comparison of Velocity Fields with a Stagnation Point

The velocity fields (10.86a, b) and (10.89a, b) differ (10.86b) ≡ (10.89b) only by
the factor minus two (−2) in (10.86a) ≠ (10.89a), which makes a significant
difference, because: (i) the velocity field (10.86a, b) is neither irrotational nor
incompressible:
   
( )
∇ ∧ v = e3 ∂ v y / ∂ x − ∂ v x / ∂ y = − e3 3 x , ∇.v = ∂ vx / ∂ x + ∂ vy / ∂ y = 3 y ;
 (10.92a, b)
(ii) the velocity field (10.89a, b) is both:

( ) ( )
∂ y 2 − x 2 / ∂ x = − 2 x = ∂ ( −2 x y )/y , ∂( −2 x y )/ ∂ x + ∂ y 2 − x 2 / ∂ y = 0.
 (10.93a, b)

The latter velocity field (10.89a, b) corresponds to the potential flow (section
I.12.19) of complex conjugate velocity:

( )
f ′ = vx − i vy = − 2 x y − i y 2 − x 2 = i ( x + i y ) = i z 2 ;
2
(10.94a, b)
Examples 10.1 to 10.20 25

the stream function Ψ, which is the imaginary part of the complex potential:

C  i z3   i ( x + i y )3  x3
2
= Im f = Im  = Im   = − x y + . (10.95a–d)
3  3  3 3
 

specifies the paths (10.95d) ≡ (10.91b) or streamlines. Thus, (10.88a, b) is the

velocity field of a potential flow in a corner (I.14.75b) of angle θ = π/6 = 60° , so
that the stagnation point at the origin is a saddle point with six asymptotes
(Figure 10.2), which allow a continuous velocity field on both sides of each
asymptote.

EXAMPLE 10.8  First-Order Special Differential Equations

Classify the following four first-order ordinary differential equations
(10.96a–10.99a):
1 1
XCX:   y = y ′x + ,         y ( x ; C ) = C x + ; y ( x ) = 2 x , (10.96a–c)
y′ C

y ′3 C + p 2 /2 − p 3 /3
C:          y = y ′ 2 x + ;   x ( p; C) = , (10.97a–c)
3 (1 − p )−2
y x2
CII:        x = + y ′,                y ( x ; C ) = C ( x − C ) ; y ( x ) = , (10.98a–c)
y′ 4

CIII:    y ′ 3 − y ′ 2 y − 4 x 2 y ′ + 4 x 2 y = 0, (y − x 2
)( )( )
+ C y + x 2 + C y − C e x = 0,
 (10.99a, b)
and obtain the general integrals (10.96b–10.99b) and the special integrals
when they exist (10.96c–10.98c). The roman numerals at left identify the stan-
dard in the classification of differential equations in the recollection 10.1.

E10.8.1  Cusped Parabola as the Envelope of a Family of Straight Lines

The equation (10.96a) is (standard XCIV) a Clairaut-type (5.29a, b) with
(10.100a) and the general integral (5.31c) ≡ (10.100b) ≡ (10.96b):

1 1 ∂y 1 1
h( p) = : y ( x; C) = C x + : = x − 2 = 0, C = , y ( x) = 2 x ,
p C ∂C C x
(10.100a–c)
26 Differential Equations and Applications

y
y = C x + 1/C

y = 2 √−x y = 2 √x

envelope
1/C

x
−b −a a b

FIGURE 10.3
The cusped parabola is the envelope of a family of straight lines whose ordinate at the origin
is the inverse of the slope.

the C-discriminant (10.100c) leads (10.100d) to the special integral (10.100e) ≡

(10.96c). The integral curves are straight lines (10.100b) of slope C and ordinate
1/C at the origin, tangent to a parabola (10.100e) with a vertical cusp at the origin
(Figure 10.3), which is an envelope.
The equation (10.97a) is (standard C) a D’Alembert-type (5.38a–c) with
(10.101a, b):

p3  2  1
g ( p ) = p2 , h( p) =
3
: ϕ ( p ) = exp 

∫ 1 − p  = (1 − p) 2 , (10.101a–d)

leading (5.40a) = (10.101c) to (10.101d); the general integral (5.40b) ≡ (10.102) ≡

(10.97b):

x ( p ; C) =
1
(1 − p ) 2 { ∫
C + p ( 1 − p ) dp = } C + p 2 /2 − p 3 /3
( 1 − p )2
. (10.102)

The parameter p is eliminated between (10.102) and (10.97a) to specify the

integral curves y (x; C).

E10.8.2  Smooth Parabola as the Envelope of a Family of Straight Lines

The Clairaut (10.96a) [D’Alembert (10.97a)] equation (5.45a, b) is solvable for y,
providing an alternate method (subsection 5.4.1) of solution (10.96b, c) [(10.97a, b)].
The equation (10.98a) is (standard CII), solvable for the independent variable
(5.47a, b) with (10.103a), and (5.48a) leads to (10.103b):

y dp
x= + p ≡ g ( y, p ): = 0; p = C, y ( x ; C ) = C ( x − C ) , (10.103a–d)
p dy
Examples 10.1 to 10.20 27

y = x 2 /4

y = Cx − C 2

− C2

FIGURE 10.4
A family of straight lines distinct from that in Figure 10.3 is the case of ordinate at the origin
equal to minus the square of the slope that has a smooth parabola as an envelope.

hence (10.103c) and the general integral is (10.103d) = (10.98b). The C-discriminant
(10.104a) leads (10.104b) to the special integral (10.104c) ≡ (10.98c):

∂y x x2
0= = x − 2C, C= , y ( x) = . (10.104a–c)
∂C 2 4

The integral curves are straight lines (10.103d) of slope C and ordinate −C 2 at the
origin, tangent to the smooth parabola (10.104c), which is their envelope (Figure 10.4).

E10.8.3  First-Order Equation of Degree Three without Singular Integrals

The ordinary differential equation of first-order third-degree (10.99a) can
(standard CIII) be solved (5.62a, b) for y ′, noting that y = y ′ is a root, leading
to the factorization (10.105):

( )
0 = y ′ 3 − y ′ 2 y − 4 x 2 y ′ + 4 x 2 y = ( y ′ − y ) y ′ 2 − 4 x 2 = ( y ′ − y )( y ′ − 2 x )( y ′ + 2 x ) ;
(10.105)
there are three solutions (10.106a–c) of (10.105):

y1− 3 ( x ) = C e x , ± x 2 − C , 0 =  y − y1 ( x )   y − y 2 ( x )   y − y 3 ( x )  , (10.106a–d)

so that the product (10.106a–d) ≡ (10.99b) is the general integral. There is no

special integral.
28 Differential Equations and Applications

EXAMPLE 10.9 Differential Equations of Order

Higher than the First
Classify the following fifteen differential equations of the second order or
higher order (10.107a–10.122a):

1
CXI:         y ′′ = ,            y ( x ) = C2 + C1 x + x log x − x , (10.107a, b)
x

x2
I, CX:       y ′′ = y ′ + x:       y ( x ) = C2 + C1 e x − − x − 1, (10.108a, b)
2

L:             y ′′ = − y ′ tan x y ( x ) = C2 + C1 sin x , (10.109a, b)

1
 3 y ( x ) = C2 +  2 ( C1 + x ) 
3/2
CVIII:      y ′′ = , , (10.110a, b)
y′

3 2
CVIII:      y ′′ =
2
y ,       x ( y ) = C2 +
∫ (C + y )
1
3 −1/2
dy , (10.111a, b)

2y
CVIII:      y ′′ =
3 y′
:         x ( y ) = C2 +
∫ (C + y ) 1
3 −1/3
dy , (10.112a, b)

x ( y ) = C2 +
∫  2 (C sinh y )
−1/2
CVIII:      y ′′ = − cosh y: 1 dy , (10.113a, b)

y′ dy
CVIII:      y ′′ =
y
:        y ( x ) = C2 +
∫ C + log y ,
1
(10.114a, b)

1
     15 y ( x ) = C3 + C2 x + 4 ( C1 + x )
5/2
CXIV:      y ′′′ = : , (10.115a, b)
2 y ′′

CXVI:      y ′′′ = − 2 y ′′ y ′: x ( y ) = C3 +
∫ C
− 1/2
2 + log ( C1 + x )  dy , (10.116a, b)

y ′′ x2
CXIII: y ′′′ = 2
: y ( x ) = C3 + C2 x + − C1−3 ( 1 + C1 x )  log ( 1 + C1 x ) − 1 ,
x 2C1
 (10.117a, b)
N −2
x a+ N −1
CXIII, L:  x y( N ) ( x ) = a y( N − 1) ( x ) : y ( x ) = ∑
n= 0
Cn+ 1 x n + CN
( a + 1)( a + N − 1)
,

 (10.118a, b)
Examples 10.1 to 10.20 29

 1  e x /2 
2

CXVII: y ′′ +  x +
 x
 y ′ + 2 y = 0: y ( x ) = e
− x 2/2
 C2 + C1
 ∫ x
dx  ,

 (10.119a, b)
CXVIII:   x y ′′ y ± y ′y + y ′ 2 x = 0: y + ( x ) = C2 + C1 log x ,
(10.120a–c)
y − ( x ) = C2 x 2 − C1 ,

(
CXIX:      y ′′ y − y ′ 2 + 6 x y 2 = 0:      y ( x ) = C2 exp  x C1 − x 2  , ) (10.121a, b)

CXX:       x 2 y ′′ + x 2 y ′ 2 + x y ′ = 0:    y ( x ) = log ( C2 + C1 log x ) , (10.122a, b)

and obtain the corresponding general integrals (10.107b–10.121b). The roman

numerals at left identify the standard in the classification of differential
equations in the recollection 10.1.

E10.9.1 Linear Differential Equations with Constant

or Variable Coefficients
The linear second-order differential equation (10.107a) involves neither the
dependent variable nor its derivative (standard CXI) and is solvable (5.155a, b)
by two quadratures: (i) the first integration of (10.107a) leads to (10.123a);
(ii) another integration yields the general integral (10.123b) ≡ (10.107b):

y ′ ( x ) = log x + C1 , y ( x ) = x ( log x − 1) + C1 x + C2 . (10.123a, b)

The second-order linear differential equation (10.108a) misses both the inde-
pendent and dependent variables (standard CX) and is also linear with
constant coefficients (1.197a–c) and hence has: (i) characteristic polynomial
(10.124a) with (standard I) distinct roots zero and unity corresponding to the
general integral in the first two terms on the r.h.s. of (10.124b):

1
0 = P2 ( D ) = D2 − D = D ( D − 1) , y ( x ) = C2 + C1 e x + x;
D ( D − 1)
 (10.124a, b)
(ii) the last term on the r.h.s. of (10.124b) is (standard XIII) the particular inte-
gral (10.124c–e):

1 1 x2 x2 x2

D ( D − 1)
x=
D−1 2
= − 1 + D + D2 + O D2 
2
=−
2
( )
− x − 1;

(10.124c–e)
30 Differential Equations and Applications

(iii) sum of (10.124b) and (10.124e) is the complete integral (10.108b). The lin-
ear second-order differential equation with variable coefficients (10.109a) ≡
(10.125a) is a separable first-order differential equation (10.125b) in y ′:

y ′′ d d
= − tan x:                     ( log y ′ ) =  log ( cos x )  , (10.125a–d)
y′ dx dx 

y ′ ( x ) = C1 cos x ,           y ( x ) = C1 sin x + C2 , (10.125c, d)

and two quadratures (10.125c, d) lead to the general integral (10.125d) =

(10.109b).

E10.9.2 Non-Linear Second-Order Differential Equations

Omitting the Independent Variable
The second-order differential equation (10.110a) omitting the independent
variable (standard CVIII) is equivalent to a first-order differential equation
(10.126a) for y ′, which is solved by two quadratures: (i) the first integration of
(10.126a) leads to (10.126b) ≡ (10.126c):

y ′ 2 = 2 ( x + C1 ) , y ′ =  2 ( x + C1 ) 
1/2
y ′dy ′ = dx , ; (10.126a–c)

(ii) a further integration of (10.126c) leads to the general integral (10.110b). The
second-order differential equation (10.111a) is (standard CVIII) also equiva-
lent (5.138a) to a first-order equation (10.127a) for y ′:

3 2
( )
−1/2
y ′dy ′ = y dy , y ′ 2 = y 3 + C1 , dx = y 3 + C1 dy , (10.127a–c)
2

leading to (10.127b, c) and hence, to the general integral (10.111b). The

non-linear second-order differential equation (10.112a) also (standard CVIII)
omits the independent variable and is reducible (5.138a) to a first-order equa-
tion (10.128a), the independent variable:

( )
−1/3
3 y ′ 2 dy ′ = 2 y d y: y ′ 3 = y 2 + C1 , dx = y 2 + C1 dy , (10.128a–c)

leading to (10.128b, c) and hence, to the general integral (10.112b). The differ-
ential equation (10.113a) is (standard CVIII) reducible to first-order and leads
(5.138a) to (10.129a) whose integration yields (10.129b, c):

y ′dy ′ = − cosh y d y: y ′ 2 = 2 ( C1 − sinh y ) , dx =  2 ( C1 − sinh y ) 

−1/2
dy ,
(10.129a–c)
Examples 10.1 to 10.20 31

and hence to the general integral (10.113b). The non-linear differential equa-
tion (10.114a) omits the independent variable (standard CVIII) and leads to a
first-order differential equation (10.130a) of degree two:

 dy ′ 1  dy
0 = y′  − : y ′ = 0: y = C0 ; dy ′ = , (10.130a–d)
 dy y  y

which has: (i) a special integral (10.130b) involving only one constant of inte-
gration (10.130c); (ii) the second factor (10.130d) in (10.130a) leads (10.131a) to
the general integral (10.131b) ≡ (10.114b):

dy
y ′ = C1 + log y , dx = . (10.131a, b)
C1 + log y

The differential equations reducible to the first order may be of the second
(higher) order [E10.9.2 (E10.9.3)].

E10.9.3 Higher-Order Differential Equations Reducible

to the First Order
The third-order non-linear differential equation (10.115a) is (standard CXIV)
of first-order (10.132a) in y ′′ leading to (10.132b, c):

2 y ′′ dy ′′ = dx , ( y ′′ )2 = x + C1 , y ′′ = x + C1 , (10.132a–c)

so that double integration of (10.132c) leads to the general integral (10.115b).

The third-order non-linear differential equation (10.116a) can (standard
CXVI) be factorized (10.133a):

 dy ′′  dy ′′
y′  + 2 y ′′ 2  = 0: y′ = 0 ; y ( x ) = C0 ; = − dy , (10.133a–d)
 dy  2 y ′′ 2

and hence, has: (i) a special integral (10.133b) involving only one constant of
integration (10.133c); (ii) the second factor in (10.133a) leads to (10.133d), which
may be integrated as (10.134a) ≡ (10.134b), implying (10.134c) by a second
integration:

1 dy
y + C1 = : 2 y ′dy ′ = ; y ′ 2 = C2 + log ( C1 + y ) , (10.134a–c)
2 y ′′ y + C1

and yielding the general integral (10.116b) by a third integration. The third-
order differential equation (10.117a) is (standard CXVI) non-linear with
32 Differential Equations and Applications

variable coefficients and: (i) separates (10.135a) to first-order differential

equation for y ′′ leading to (10.135b, c) by a first integration:

dy ′′ dx 1 1 x 1  1 
= : = C1 + , y ′′ = = 1− ; (10.135a–d)
y ′′ 2 x 2 y ′′ x 1 + C1 x C1  1 + C1 x 

(ii) rewriting the rational function (10.135c) in partial fractions (10.135d)

allows (section I.31.8) immediate second integration (10.136):

x
y ′ = C2 + − C1−2 log ( 1 + C1 x ) ; (10.136)
C1

(iii) a third integration leads to the general integral (10.117b). The differen-
tial equation (10.118a) of order N can be solved (standard CXIII) as a linear
unforced differential equation with constant coefficients or as a first-order
separable (standard L) equation (10.137b) for the derivative (10.137a) of order
N − 1, which may be: (i) integrated once (10.137c, d):

z ≡ y( N − 1) ( x ) : x z ′ = a z, log z = a log x + log CN , z = CN x a = y( N − 1) ( x ) ;

(10.137a–e)

(ii) further (N − 1) integrations of (10.137e) lead to the general integral (10.118b).

E10.9.4 Linear Non-Commutative Differential Operators

with Variable Coefficients
The linear second-order differential operator (10.119a) results (standard
CXVII) from (5.187) the composition of two first-order operators:

 d 1  d   d 1 1
0= +
 dx x   dx
+ x y = 

+
 dx x 
( y ′ + x y ) = y ′′ +  x +  y′ + 2 y ;
x
 (10.138a, b)

since the operators have variable coefficients they generally do not commute,
that is, (10.138a, b) is distinct from:

 d   d 1  d  y
0= + x  + y  =  + x  y ′ + 
 dx   dx x   dx  x
(10.139a, b)
 1  1
= y ′′ +  x +  y ′ + y  1 − 2  ,
 x x
Examples 10.1 to 10.20 33

with commutator:

 d 1  d   d   d 1   1
 +   + x −  + x  +   y =  1 + 2  y . (10.140)
 dx x dx   dx   
dx x   x 

The second-order ordinary differential equation (10.119a): (i) splits into two
first-order equations (10.141a, b):

z dz dx C1
z = y ′ + x y: z′ + = 0; + = 0, z= ; (10.141a–d)
x z x x

(ii) the equation (10.141b) ≡ (10.141c) has the integral (10.141d); (iii) substitu-
tion of (10.141d) in (10.141a) leads to a linear differential equation (10.142a) ≡
(10.142b) ≡ (3.27) with coefficients (10.142c; 10.141e):

C1
y′ = −x y + = P y + Q: P ( x ) = − x ,
x
2
(10.142a–f)
 x
  x2  x
e ξ /2
X ( x) = exp  −
 ∫ ξ dξ  = exp  − 
  2
Y ( x ) = C1
∫ ξ
dξ ,

(iv) from (3.20a; 10.142b)[(3.31a; 10.141e)] follow (10.142d, e)[(10.142f)], which

substituted in (3.31b) yield the general integral (10.119b).

E10.9.5 Non-Linear Exact Differential Equation

with Variable Coefficients
The differential equation (10.120a) with + (plus) sign is (standard CXVIII) an
exact differential (92a):

0 = x y ′′y + y ′ 2 x + y ′y = ( x y ′y )′ , ( )
C1 = 2 x y ′y = x y 2 ′ , (10.143a, b)

specifying a first-integral (10.143b); the latter (10.143b) is a first-order differen-

tial equation (10.144a):

( y )′ = Cx ,
2 1
y 2 = C1 log x + C2 , (10.144a, b)
34 Differential Equations and Applications

whose solution is (10.144b) ≡ (10.120b). The differential equation (10.121a) with

– (minus) sign is (standard CXVIII) also an exact differential (10.145a):

x
( )
0 = x y ′′y + y ′ 2 x + y ′y − 2 y ′y = x y ′y − y 2 ′ , C1 = x y ′y − y 2 = y 2 ′ − y 2 ,
2
( )
 (10.145a, b)

leading to: (i) the first integral (10.145b); (ii) the latter is a non-linear first-order
differential equation, which becomes linear (10.145b) ≡ (10.146b) for the vari-
able (10.146a):

z ≡ y2 :
x
2
z ′ − z = C1 ; z′ = P z + Q, { P , Q} ≡ { 2 2 C1
,
x x }
. (10.146a–e)

(iii) the linear first-order differential equation (3.27) ≡ (10.146b) ≡ (10.146c)

with (10.146d, e) leads to (3.20a) ≡ (10.147a, b) and (3.31a) ≡ (10.147c, d):

X x = exp  2 
x x
2 C1 C
( )  ∫ dξ = x 2 ,
ξ 
Y ( x) =
∫ ξ3
dξ = − 21 ; (10.147a–c)
x

(iv) substitution of (10.147b, d) in (3.31b) leads to (10.147e):

z ( x ) = X ( x ) C2 + Y ( x )  = x 2 C2 − C1 =  y ( x )  ,
2
(10.147e, f)

which specifies (10.146a) the general integral (10.147f) ≡ (10.120c).

E10.9.6  Second-Order Homogeneous Differential Equation

The second-order linear differential equation with variable coefficients
(10.121a) is (standard CXIX) homogeneous of the first kind (5.200a–d), as fol-
lows from (10.148a–d):

2
y′ y ′′ y ′′  y ′ 
v≡ , = v′ + v2 : 0= − + 6 x = v ′ + 6 x ; (10.148a–d)
y y y  y 

hence, (10.121a) is reducible to a first-order equation (10.148d) in the depen-

dent variable (10.148a). The solution of (10.148d) is (10.149a):

y′
−3 x 2 = v − C1 = − C1 , − x 3 = log y − C1 x − log C2 , (10.149a, b)
y
Examples 10.1 to 10.20 35

leading to the general integral (10.149b) ≡ (10.121b). The non-linear differential

equation with variable coefficients (10.122a) is (standard CXX) homogeneous
of the second kind (5.208) as follows from (5.211a, 5.212b) ≡ (10.150a, b) leading
to (10.150c, d):

 dw   dw 
y ′ x ≡ w, x 2 y ′′ = w  − 1 : 0 = x 2 y ′′ + x 2 y ′ 2 + y ′x = w  + w ,
 dy   dy 
(10.150a–d)
which is a first-order equation (10.150d) for the dependent variable
(10.150a). The special integral solution w = 0 of (10.150d) is y = const by
(10.150a) and satisfies (10.122a). The general integral solution of (10.150d)
is (10.151a):

dw dy C
= − w: C1 e − y = w = x , 0 = 1 dx − e y dy = C1 log x − e y + C2 ,
dy dx x
(10.151a–c)
leading to (10.151b) and hence (10.150c), to the general integral (10.151d) ≡
(10.122b).

EXAMPLE 10.10  Bending of a Beam under Traction

Consider the linear bending of a uniform beam (6.14f) ≡ (10.152b) under trac-
tion (10.152a) in the presence of a shear stress:

T > 0:                       E I ζ′′′′ − T ζ′′ = f ( x ) , (10.152a, b)

in the following cases: (a) a clamped beam (10.153a–d) with (Figure 10.5a), a
concentrated torque Q in the middle (10.153e); (b) a pinned beam (10.154a–d)
with (Figure 10.5b), a concentrated force F in the middle (10.154e); (c) a
clamped-pinned beam (10.155a–d) under (Figure 10.5c) its own weight
(10.155e); (d) a cantilever beam (10.156a–d) subject (Figure 10.5d) to a shear
stress (10.156e) increasing linearly from the clamped end:

ζ ( 0 ) = 0 = ζ′ ( 0 ) ; ζ ( L ) = 0 = ζ′ ( L ) : f ( x ) = Q δ ′ ( x − L/2 ) , (10.153a–e)

ζ ( 0 ) = 0 = ζ′′ ( 0 ) ; ζ ( L ) = 0 = ζ′′ ( L ) : f ( x ) = F δ ( x − L/2 ) , (10.154a–e)

ζ ( 0 ) = 0 = ζ′ ( 0 ) ; ζ ( L ) = 0 = ζ′′ ( L ) : f ( x ) = ρ g , (10.155a–e)

ζ ( 0 ) = 0 = ζ′ ( 0 ) ; ζ′′ ( L ) = 0 = E I ζ′′′ ( L ) − T ζ′ ( L ) : f ( x ) = q x .(10.156a–e)

36 Differential Equations and Applications

b
F

d
qx

FIGURE 10.5
The linear bending of a beam under traction can be considered for 4 × 4 = 16 combina-
tions of support and loads, of which four are illustrated: (i) for a concentrated torque
(a) [transverse force (b)] or a constant (c) [linearly increasing (d)] shear stress; (ii) for
supports clamped (a) [pinned (b)] at both ends or clamped at one end and pinned (c) [free
(d)] at the other end.

The shear stress (10.154e)[(10.153e)] corresponding to a concentrated trans-

verse force F (torque Q) is specified (III.1.34a, b; III.3.11a–c) [(III.1.45a–c;
III.3.18a–c)] by a Dirac delta function (its derivative) with singularity [subsec-
tion III.1.3 and III.3.1 (III.1.4 and III.3.2)] at the application point.

E10.10.1  Elastica of a Beam under Traction

The problems (10.153a–e; 10.154a–e) may be solved with explicit reference
(chapter III.4) to the generalized functions unit jump, unit impulse, and
derivatives (chapters III.1 and III.3). It is possible to use only ordinary func-
tions in all four problems (10.153 to 155a–e) as shown next. In the absence of
shear stress (10.157a), the differential equation (10.152b) specifying the linear
Examples 10.1 to 10.20 37

deflection of the elastic becomes (10.157c), involving the parameter (10.157b),

which is positive instead of negative for buckling (10.16b):

 d2  d
T  d 
f ( x ) = 0, p 2 =
0 = ζ′′′′ − p 2ζ′′ =  2 
> 0: − p  + p  ζ ( x ).
EI 
 dx dx   dx 
(10.157a–d)

The characteristic polynomial (10.157d) has a double-zero root and a pair

of real symmetric roots ±p, instead of a complex conjugate imaginary pair
of roots ± i p; the general integral (10.158) is obtained from (6.18) replacing p
by ip:

ζ ( x ) = A + B x + C cosh ( p x ) + D sinh ( p x ) , (10.158)

and involves hyperbolic instead of circular functions. The replacement of

periodic functions by monotonic functions implies that there is no buckling
for a beam under traction; bending is possible only under transverse load.
To the general integral (10.158) that specifies the shape of the elastica (10.157c)
without shear stress (10.157a) a particular integral of (10.152b) with loads
must be added to obtain the complete integral specifying the total deflection
in each of (E10.10.2–E10.10.5) the cases (a–d).

E10.10.2  Clamped Beam with a Concentrated Torque

In the case (a) of a beam clamped at both ends (Figure 10.5a) with a concen-
trated moment Q in the middle, the reaction moment at the support is given
by (10.159a) ≡ (III.4.224b):

Q
− = E I ζ1′′( 0 ) ;                        ζ1 ( 0 ) = 0 = ζ1′ ( 0 ) , (10.159a–c)
4

which specifies the curvature, and the coordinate and slope are zero (10.159b, c).
The shape of the beam is skew-symmetric relative to the middle (10.160a)

 L  L
ζ1 ( x ) = − ζ1 ( L − x ) , ζ1   = − ζ1   = 0, (10.160a, b)
 2  2

implying that the displacement is zero (10.160b) there. The conditions

(10.159b, c) applied to (10.158) lead to (10.161a, b), reducing the number of
constants of integration:

{ } {
A + C = 0 = B + p D:             ζ1 ( x ) = A 1 − cosh ( px ) + D sinh ( px ) − px , }
(10.161a–c)
38 Differential Equations and Applications

from four in (10.158) to two in (10.161c). The remaining two constants of inte-
gration are determined substituting (10.161c) in (10.159a) [(10.160b)] leading to
[(10.162a) [(10.162c)]:

Q/4 = EI A p 2 = AT , { } {
D sinh ( pL/2 ) − pL/2 = A cosh ( pL/2 ) − 1 , }
(10.162a–c)

and using (10.157b) in (10.162b). Solving (10.162b, c) for A and D, and substitut-
ing into (10.161c) yields:

L 4T cosh ( pL/2 ) − 1
0≤x≤ : ζ1 ( x ) = 1 − cosh ( px ) +  sinh ( px ) − px  ,
2 Q sinh ( pL/2 ) − pL/2
 (10.163a, b)

the shape (10.163b) of the first half (10.163a) of the elastica of the beam under
traction (10.157b).
If the traction force is weak in the sense (10.164a), then (III.7.6b) ≡ (10.146b)
and (III.7.7b) ≡ (10.146c):

6 6
pL  T L2  p2 x2 p4 x4
( p x ) ≤  2  =  2 E I  << 1:
6
cosh ( px ) − 1 =
2
+
24
(
+ O p6 x6 )
p3 x3 p5 x5
                                                            sinh ( px ) − px =
6
+
120
(
+ O p7 x7 , )
(10.164a–c)

imply (10.164d–f):

cosh ( pL/2 ) − 1
−1
3  p2 x2   p2 x2  
=  1 +
sinh ( pL/2 ) − pL/2 p x  12  
1 +
20 
(
+ O p4 x4 

)


3   1 1  3 px
p x 
1 + p2 x2  −
 12 20 
+ O p4 x4  =
= +
 p x 10
+ O p3 x3 ; ( ) ( )
(10.164d–f)

substitution of (10.164b, c, f) and similar expressions with px replaced by pL/2

leads from (10.163b) to (10.165c) with the approximation (10.165a):

 p 2 x 2 px  6 pL  
pL 
4
4T p2 x2
( px )4 ≤ 
 1 +
12
−   << 1:
2 
+ ;
3  pL 20   Q
ζ1 ( x ) = −
2

(10.165a–c)
Examples 10.1 to 10.20 39

the simplication of (10.165b, c) ≡ (10.166a, b):

2
 4EI  Q x2  2x Tx  L 
T 2 <<  2  : ζ1 ( x ) = − 1 − L + 12EI  x −  , (10.166a, b)
 L  8EI  5  

specifies the shape of the elástica (10.166b) for weak axial traction (10.166a).
The beam equation (10.157c) simplifies to the bar equation (10.167b) in the
absence of axial tension (10.167a):

T = 0: ζ′′′′
0 = 0; ζ0 ( x ) = A + B x + C x 2 + D x 3 , (10.167a–c)

the solution of (10.167b) is a third-degree polynomial (10.167c), which replaces

(10.158) for p = 0 . The four constants of integration in (10.167c) are determined
(10.168a–d) from the boundary conditions (10.159a–c; 10.160b):

Q
0 = ζ01 ( 0 ) = A, 0 = ζ′01 ( 0 ) = B , − = ζ′′01 ( 0 ) = 2 C ,
4EI
(10.168a–d)
 L L
2
 DL 
0 = ζ01   =  C + 2  ;
 2 4

substitution of the non-zero constants (10.168c, d) ≡ (10.169a, b) in (10.167c)

leads to the shape of the elástica (10.169c):

Q 2C Q Q x2  2x 
C=− , D=− = : ζ10 ( x ) = − 1−  , (10.169a–c)
8EI L 4EIL 8EI  L

which coincides with (10.166b) ≡ (III.4.220b) for zero tension (10.167a). Thus,
the (problem 222) linear bending (10.152b) of a uniform beam (6.10a, b) subject
to a traction (10.152a) clamped at both ends (10.153a–d) under a concentrated
torque (10.153e) leads to: (i) the transverse displacement of the first (10.163a, b)
[second(10.170a, b)] half of the elastica, which is skew-symmetric (10.160a):

L 4T
≤ x ≤ L:      ζ1 ( x ) = cosh  p ( L − x )    −  1
2 Q
(10.170a, b)
cosh ( pL/2 ) − 1
                          −  sinh  p ( L − x )  − p ( L − x )  ,
sinh ( pL/2 ) − pL/2
40 Differential Equations and Applications

confirming that the displacement is zero at the clamped ends (10.153a, c) and in the
middle (10.160b); (ii) the slope (10.171b) is symmetric (10.171a):

Qp  cosh ( pL/2 ) − 1 
ζ1′ ( L − x ) = ζ1′ ( x ) = − sinh ( px ) +  cosh ( px ) − 1 ,
4T  sinh ( pL/2 ) − pL/2 
 (10.171a, b)

confirming it vanishes at the clamped end (10.153b, d); (iii) the curvature (10.172c)
is skew-symmetric (10.172b) and specifies the bending moment (10.172a) and
implies that M1 ( 0 ) = −Q/4 in agreement (10.159a) with the bending moment at
the support:

M1 ( x ) = −EI ζ1′′( L − x ) = EIζ1′′( x )

Q  cosh ( pL/2 ) − 1  (10.172a–c)

= − cosh ( px ) + sinh ( px )  ,
4  sinh ( pL/2 ) − pL/2 

(iv) the slope (10.171b) becomes (10.172d) at the mid-position:

 L  Q p 2 − 2 cosh ( pL/2 ) + ( pL/2 ) sinh ( pL/2 )

ζ1′   = . (10.172d)
 2  4T sinh ( pL/2 ) − pL/2

The shape of the elastica simplifies to (10.169c) for a bar in the absence of axial tension
(10.167a); the lowest-order correction for a beam in the case of weak axial traction
(10.166a) is (10.166b).

E10.10.3  Pinned Beam with a Concentrated Force

In the case (b) of a transverse force P concentrated at the middle of the beam,
the traction at the support (10.159a) is replaced by the transverse force (10.173a):

P  L
− 2 ( 0 ) − T ζ′ ( 0 ) ,
= E I ζ′′′ ζ′2   = 0, (10.173a, b)
2  2

and (10.160b) is replaced by a condition of zero slope at the middle (10.173b).

The latter arises because the shape of the beam is symmetric relative to the
middle (10.174a):

ζ2 ( x ) = ζ2 ( L − x ) , ζ′2 ( x ) = − ζ′2 ( L − x ) , ζ′2 ( L/2 ) = − ζ′2 ( L/2 ) = 0,

(10.174a–d)

and hence, the slope (10.174b) is skew-symmetric (10.174c) leading to (10.174d) ≡

(10.173b). Considering the case (Figure 10.5b) of a beam pin-joined at both ends,
Examples 10.1 to 10.20 41

by symmetry (10.175a) it is sufficient to consider in (10.158) only one end

where the boundary conditions (10.154a, b) ≡ (10.175a, b):

ζ ( 0 ) = 0 = ζ′′2 ( 0 ) : A = 0 = C, ζ2 ( x ) = B x + D sinh ( px ) , (10.175a–e)

specify the constants (10.175c, d) and lead to (10.175e). Substitution of (10.175e)

in (10.173a, b) specifies:

P  pL 
− = − T B, 0 = B + p D cosh   , (10.176a, b)
2  2 
the two remaining constants of integration, and hence:

L 2T 1  pL 
0≤x≤ : ζ2 ( x ) = x − sinh ( px ) sech   , (10.177a, b)
2 P p  2 

the shape (10.177a, b) of the elastica of the beam.

In the case of weak traction force (10.178a) the inverse of (10.164b) is (10.178b)
to third order in (10.178c, d):

6 −1
 pL   p2 x2 p4 x4 
( px ) ≤   << 1:   sech ( px ) =  1 +
6
+
 2   2 24 
2
p2 x2 p4 x4  p2 x2  p2 x2 5p4 x4
                                                     = 1 − − +  = 1− + ;
2 24  2  2 24
(10.178a–d)

substitution of (10.164c; 10.178d) in (10.177b) leads to (10.179a):

2T  p2 x2   p 2 L2 5 p 4 L4 
ζ2 ( x ) = 1 −  1 + 1 − +   
Px  6   8 384 
(10.179a)
2 2 2 2 4 4 4 2 2
p x p L 5p L p Lx
=− + − + ,
6 8 384 48

which simplifies to the shape of the elastica (10.179c):

P x  2 3L2 T L2  2 5L2  
2
 4E I 
T 2 <<  2  : ζ02 ( x ) = − x − −  x − 8   , (10.179b, c)
 L  12 EI  4 8EI 
42 Differential Equations and Applications

under weak traction (10.179b). In the absence of axial tension (10.167a) the
beam (10.157c) becomes a bar (10.167b) and the constants in the shape (10.167c)
of the elastica are determined by the boundary conditions (10.175a, b;
10.173a, b) ≡ (10.180a–d):

P
0 = ζ02 ( 0 ) = A, 0 = ζ′′02 ( 0 ) = 2C , − 02 ( 0 ) = 6 D,
= ζ′′′
2EI
(10.180a–d)
 L 3DL2
0 = ζ′02   = B + ;
 2 4

the non-zero constants (10.180c, d) ≡ (10.181a, b) in (10.167c) specify the shape

of the elástica (10.181c) of the bar:

P 3DL2 P L2 P  2 3L2 
D=− , B=− = : ζ20 ( x ) = − x − , (10.181a–c)
12EI 4 16EI 12 EI  4 

which coincides with (10.179c) ≡ (III.4.114b) for zero axial tension (10.167a).
In conclusion, (problem 223) the linear bending (10.152b) of a uniform beam
(6.10a, b) pinned at both ends (10.154a–d) under traction (10.152a) with a concen-
trated force (10.154e) at the middle (Figure 10.5b) leads to: (i) the transverse dis-
placement of the first (10.177a, b) [second (10.182a, b)] half of the elastica, which is
symmetric (10.175a):

L 2T 1  pL 
≤ x ≤ L: ζ2 ( x ) = L − x − sinh  p ( L − x )  sech   , (10.182a, b)
2 P p  2 

confirming that it vanishes at both ends (10.154a, c); (ii) the slope (10.183b):

P   pL  
ζ′2 ( x ) = − ζ′2 ( L − x ) =
2T 1 − cosh ( p x ) sech  2   , (10.183a, b)
 

which is skew-symmetric (10.183a), confirming it vanishes in the middle (10.173b);

(iii) the curvature (10.184c), which is symmetric (10.184b) and specifies the bending
moment (10.184a):

 pL 
M2 ( x ) = − E I ζ′′2 ( x ) = − E I ζ′′2 ( L − x ) = − T sinh ( px ) sech   , (10.184a–c)
 2 
Examples 10.1 to 10.20 43

which vanishes at the pinned ends; (iv) the maximum deflection (slope in modulus) is:

 L  PL  2  pL  
ζ2 max = ζ2   = 1− tanh    , (10.185)
 2  4T  pL  2 

P   pL  
ζ′2 max = ζ′2 ( 0 ) =  1 − sech    = − ζ′2 ( L ) , (10.186a, b)
2T   2 

at the middle (10.185) [at the supports (10.186a, b)]. In the absence of tension (10.167a)
the shape of the elastica of the bar is (10.181a–c) and the lowest-order correction for
a beam under weak traction (10.179b) is (10.179c).

E10.10.4  Heavy Clamped-Pinned Beam

Continuous loads instead of concentrated loads are considered next. The
simplest case is a uniform shear stress (10.155e), such as a beam with mass
density per unit length ρ in a gravity field with uniform acceleration g:

E I ζ′′′′
3 − T ζ ′′
3 = ρ g ; (10.187)

a particular integral of the forced linear differential equation with constant

coefficients (10.187) is given by the method (§3.17) of inverse polynomial of
derivatives:

d ρ g −2 ρ g x2
( )
−1
D≡ : ζ3* ( x ) = E I D4 − T D2 ρg = − D ( 1) = − . (10.188a–d)
dx T 2T

Adding (10.189a) to (10.158) the general integral (10.158) of the unforced

equation (10.157c) specifies the complete integral (10.189b):

ρ g x2
ζ3 ( x ) = ζ ( x ) + ζ3* ( x ) = A + B x − + C cosh ( px ) + D sinh ( px ) .
2T
 (10.189a, b)

In the case (c) of a beam (Figure 10.5c) clamped (10.155a, b) ≡ (10.190a, b) at one
end substitution of (10.189b) yields (10.161a, b), so that only two constants of
integration remain in (10.190c):

ρ g x2
ζ3 ( 0 ) = 0 = ζ3′ ( 0 ) : { } {
ζ3 ( x ) = A 1 − cosh ( px ) + D sinh ( px ) − px − } 2T
.

(10.190a–c)
44 Differential Equations and Applications

The other end is pin-joined (10.155c, d) ≡ (10.191a, b) leading to the linear

inhomogeneous system of equations (10.191c):

 1 − cosh ( pL )
sinh ( pL ) − pL   A   ρ g L2 /2 T 
ζ3 ( L ) = 0 = ζ3′′( L ) : 
 = ;
sinh ( pL )   D   ρ g/p 2 T 
 − cosh ( pL )
 
 
(10.191a–c)

the determinant (10.192b):

p L > 0: { }
∆ ≡ sinh ( pL ) − pL cosh ( pL ) = cosh ( pL ) tanh ( pL ) − pL ≠ 0,
(10.192a–c)

is non-zero (10.192c) for (10.192a), showing that there is no buckling under

traction, in contrast with the case of compression. Since ∆ ≠ 0, the system can
be solved uniquely for A, D:

 A   sinh ( p L ) pL − sinh ( pL )
   ρ g L2 /2 T ∆ 
=    ; (10.193)
 
 D   cosh ( p L ) 1 − cosh ( pL )   ρ g/ p 2 T ∆ 
   

substitution of (10.193) in (10.190c) specifies the shape of the beam under

traction.
In the absence of traction (10.194a), the shape of a bar under its own weight
is (10.187) given by (10.194b):

ρg ρ g x4
T = 0: ζ′′′′
03 = , ζ03 ( 0 ) = + A + B x + C x 2 + D x 3 , (10.194a–c)
EI 24 E I

whose complete integral (10.194c) consists of the sum of: (i) a particular
integral of the unforced equation (10.194b), which is the first term; (ii) the
general integral (10.167c) of the unforced equation (10.167b). The boundary
conditions for clamping (10.190a, b) at one end, and pin-joint (10.191a, b) at
the other end determine the constants of integration in (10.194c), namely
(10.195a–c):

 1 L  C  ρ g L2  1 
A = 0 = B,   =−   , (10.195a–c)
 2 6 L   D  24 E I  12 
Examples 10.1 to 10.20 45

leading to (10.196a, b):

ρ g L2 5ρ g L ρ g x 2  x 2 5 x L L2 
C= , D=− ; ζ03 ( x ) = − + ,
16 E I 48 E I 8 E I  3 6 2
(10.196a–c)

the shape of the bar is given by (10.194c) with (10.195a, b; 10.196a, b) leading
to (10.196c) ≡ (III.4.64).
In conclusion, (problem 224) the linear bending (10.152b) of a heavy (10.155e)
uniform (6.10a, b) beam under traction (10.152a) clamped (10.155a, b) at one end
and pinned (10.155c, d) at the other end leads to: (i) the transverse displacement of
the elastica (10.190c) with (10.193) confirming that the displacement is zero at the
supports (10.190a; 10.191a); (ii) the slope (10.197):

{
ζ3′ ( x ) = p − A sinh ( px ) + D  cosh ( px ) − 1 −} ρgx
T
; (10.197)

(iii) the curvature (10.198b) specifying the bending moment (10.198a):

ρgEI
M3 ( x ) = − E I ζ3′′( x ) = T  A cosh ( px ) − D sinh ( px )  + ; (10.198a, b)
T

(iv) the transverse force (10.199a, b):

F3 ( x ) = E I ζ3′′′ ( x ) − T ζ3′ ( x ) = ρ g x + pT D, f3 ( x ) = F3′ ( x ) = ρ g ,

(10.199a–d)

leading to the shear stress (10.199c), in agreement with (10.199d) ≡ (10.155e). The
bending momenta at the clamped support is (10.200a):

ρgEI
;M3 ( 0 ) = T A +
F3 ( 0 ) = pT D, F3 ( L ) = F3 ( 0 ) + ρ g L ,
T
(10.200a–c)

and the transverse force at the two supports (10.200b, c) differs by the weight of the
beam.

E10.10.5  Cantilever Beam with a Linearly Increasing Shear Stress

The case (d) of a linearly increasing shear stress (10.156e) leads (10.152b) to
(10.201):

E I ζ′′′′
4 − T ζ ′′
4 = q x , (10.201)
46 Differential Equations and Applications

which is a forced linear differential equation with constant coefficients,

whose particular integral can be obtained, as in the case (10.188a–d) of
uniform shear stress:

d q −2 q x3
( )
−1
D≡ : ζ 4* ( x ) = E I D4 − T D2 qx = − D ( x) = − . (10.202a–d)
dx T 6T

Adding (10.202d) to the general integral (10.158) of the unforced case (10.157c)
leads (10.202e) to (10.202f):

q x3
ζ4 ( x ) = ζ ( x ) + ζ 4* ( x ) = A + B x − + C cosh ( px ) + D sinh ( px ) , (10.202e, f)
6T

for the complete integral.

In the case (Figure 10.5d) of a beam clamped at one end (10.156a, b) ≡
(10.203a, b), substitution of (10.202b) yields (10.161a, b) implying (10.203c):

q x3
ζ ( 0 ) = 0 = ζ′ ( 0 ) : { } {
ζ4 ( x ) = A 1 − cosh ( p x ) + D sinh ( p x ) − p x − } 6T
.

(10.203a–c)

The boundary conditions (10.156c) ≡ (10.204a) [(10.156d) ≡ (10.205a)] at the free

end:

q q p 2 L2
ζ4′′′ ( L ) − p 2 ζ4′ ( L ) = 0 :           0 = − + p3D + , (10.204a, b)
T 2T

q L2
ζ4′ ( L ) = 0: A sinh ( pL ) − D  cosh ( pL ) − 1 = − , (10.205a, b)
2Tp

determine the remaining two constants of integration (10.206a) [(10.206b)]:

q q L2 q L2
p3D = − , A = D  tanh ( p L ) − csch ( p L )  − csch ( p L ) ;
T 2E I 2Tp
 (10.206a, b)

substitution of (10.206a, b) into (10.203c), specifies the shape of the beam

under traction.
Examples 10.1 to 10.20 47

In the case of a bar in the absence of traction (10.207a), the equation (10.201)
is replaced by (10.207b):

qx q x5
T = 0: ζ′′′′
04 = , ζ04 ( x ) = + A + B x + C x 2 + D x 3 , (10.207a–c)
EI 120 E I

whose complete integral is (10.207c) the sum of (10.158) with the particular
integral of (10.207b). The constants of integration are determined by the
boundary conditions at the: (i) clamped end (10.156a, b) leading to (10.195a, b);
(ii) at the free end (10.156c, d) leading to (10.208a, b):

q L2 q L3 q L3
04 ( L ) = 6 D +
0 = ζ′′′ , ′′ ( L ) = 2 C + 6 DL +
0 = ζ04 = 2C − ;
2EI 6E I 3E I
(10.208a–c)
substitution of (10.208a) in (10.208b) specifies the remaining constant of inte-
gration (10.208c); substitution of (10.195a, b; 10.208a, c) in (10.207c) yields:

q x2
ζ04 ( x ) =
120 E I
{ }
x 3 − 10 x L2 + 20 L3 , (10.209)

for the shape of the bar.

In conclusion, (problem 225) the linear bending (10.152b) of a uniform (6.10a, b)
cantilever (10.156a–d) beam under compression (10.152a) subject to a shear stress
(10.156e) increasing linearly along its length (10.201) leads to: (i) the transverse dis-
placement of the elastica (10.203c) with (10.206a, b) confirming the zero displacement
at the clamped end (10.156a); (ii) the slope (10.210):

q x2
ζ4′ ( x ) = p  D cosh ( px ) − 1 − A sinh ( px )  − , (10.210)
2T

confirming that it vanishes at the clamped end (10.156b); (iii) the curvature (10.211b),
which specifies the bending moment (10.211a):

q xEI
M4 ( x ) = − E I ζ4′′( x ) = T  A cosh ( p x ) − D sinh ( p x )  + ; (10.211a, b)
T

(iv) the transverse force (10.212a, b):

q E I q x2
+ F4 ( x ) = E I ζ′′′ ( x ) − T ζ′ ( x ) = pT D −
, f 4 ( x ) = F4′ ( x ) = q x ,
T 2
(10.212a–d)
48 Differential Equations and Applications

leading to the shear stress (10.212c) in agreement with (10.212d) ≡ (10.156e).

The displacement (slope) at the free end is given by (10.203c) [(10.210)] with
x = L, and the bending moment (transverse force) at the clamped end by (10.213a)
[(10.213b)]:

qE I
M4 ( 0 ) = T A,              F4 ( 0 ) = pT D − . (10.213a, b)
T

with D given by (10.206a; 10.157b).

Thus, the shape of a beam (bar) under (under no) traction (10.152a)
[(10.167a, b)] has been obtained in the following four cases: (E10.10.2) clamped at
both ends (Figure 10.5a) with a concentrated moment Q at the middle (10.163a, b;
10.170a, b)[(10.166a, b)]; (E10.10.3) pin-joined at both ends (Figure 10.5b) with a
concentrated force P in the middle (10.177a, b; 10.182a, b) [(10.181c)]; (E10.10.4)
under its own weight (Figure 10.5c) with one end clamped and the other pinned
(10.190c; 10.192b, c; 10.193) [(10.196c)]; (E10.10.5) under a linearly increasing load
(Figure 10.5d) with one end clamped and the other free (10.203c; 10.206a, b) [(10.209)].
The first-order correction to the linear approximation was obtained (10.165a–c)
[(10.179b, c)] in the cases (i) [(ii)]. The preceding four cases (E10.10.2–E10.10.5) are
a cross sample of the 16 combinations of four loads (10.153e–10.156e) and four sup-
ports (10.153a–d to 10.156a–d).

EXAMPLE 10.11 Linear Bending of a Circular Heavy

Plate with a Circular Hole
Consider the problem for all combinations of clamped, supported, or free
edges at the inner and outer boundary, with a vertical offset.
The shape of the neutral surface or directrix of a heavy plate with polar
symmetry is specified by (6.502e) ≡ (10.214b):

Eh3 ρ g r4
D= : ζ ( r ) = + C1 + C2 log r + ( C3 − C4 ) r 2 + C4 r 2 log r ,
(
12 1 − σ 2 ) 64D
 (10.214a, b)

where C1− 4 are arbitrary constants to be determined by four boundary condi-

tions. The shape of the neutral surface or directrix (10.214b) implies the slope
(10.215a, b), the curvature (10.216a, b):

dζ ρ g r 3 C2
ζ′ ( r ) ≡ = + + 2 C3 r − C4 r + 2 C4 r log r , (10.215a, b)
dr 16D r
Examples 10.1 to 10.20 49

d 2ζ 3 ρ g r 2 C2
ζ′′ ( r ) ≡ = − 2 + 2 C3 + C4 + 2 C4 log r , (10.216a, b)
dr 2 16 D r

 σ  ρ g r2
Mr ( r ) = − D  ζ′′ + ζ′  = − (3 + σ )
 r  16
(10.217a, b)
 C2 
               + D ( 1 − σ )  2 − C4  − D ( 1 + σ ) C3 + C4 log r  ,
 r 

D ′ ρ g r 4 C4 D
N r ( r ) = −  ( r ζ′ )′  = − − , (10.218a, b)
r  2 r

the stress couple (10.217a, b), and the turning moment (10.218a, b).
The linear bending of a heavy circular plate with a concentric (10.219a) circular
hole (Figure 10.6) leads (Table 10.2) to (problem 272) eight cases (Table 10.1):

b ≤ r ≤ a;
(10.219a–e)
I: ζ ( a ) = 0 = ζ′ ( a ) ,         ζ ( b ) = − d , ζ′ ( b ) = 0 ,

II: ζ ( a ) = 0 = ζ′ ( a ) ,     ζ ( b ) = − d , Mr ( b ) = 0 , (10.220a–d)

III: ζ ( a ) = 0 = ζ′ ( a ) ,        Mr ( b ) = 0 = N r ( b ) , (10.221a–d)

IV: ζ ( a ) = 0 = Mr ( a ) , Mr ( b ) = 0 = N r ( b ) , (10.222a–d)

V: ζ ( a ) = 0 = Mr ( a ) ,      ζ ( b ) = − d , Mr ( b ) = 0, (10.223a–d)

TABLE 10.1
Combinations of Boundary Conditions for a Circular Plate with a Circular Hole
Boundary Outer
Support Clamped Supported Free
Inner Clamped case I, Figure 10.6a VI ≡ reverse case II VII ≡ reverse case III
Supported case II, Figure 10.6b case V, Figure 10.6e VIII ≡ reverse case IV
Free case III, Figure 10.6c case IV, Figure 10.6d statically unstable
Note: There are 3 × 3 = 9 combinations of clamped/pinned/free boundaries for the weak bend-
ing of a circular elastic plate with a circular hole; of these eight are static cases, consisting
of five distinct cases (Table 10.2 and Figures 10.5a–c) and three interchanges.
50 Differential Equations and Applications

e
b
c
a

FIGURE 10.6
The linear bending of a circular elastic plate with a concerntric circular hole (Table 10.1) leads
to five distinct cases (Table 10.2), namely: (a) the case I: clamped at both boundaries; (b) the
case II: clamped at the outer boundary and supported at the inner boundary (and vice-versa
for the case VI); (c) the case III: clamped at the outer boundary and free at the inner boundary
(and vice-versa for the case VII); (d) the case IV: supported at the outer boundary and free at the
inner boundary (and vice-versa for the case VIII): (e) the case V supported at both the outer and
inner boundaries. The case IX free at both boundaries is not static and is omitted.
Examples 10.1 to 10.20 51

TABLE 10.2
Constants in the Shape of Neutral Surface of Heavy Circular Plate with a Circular Hole
Case I II III IV V
Figure 10.6a 10.6b 10.6c 10.6d 10.6e
Outer Clamped Clamped Clamped Supported Supported
Boundary
ζ ( a) = 0 ζ ( a) = 0 ζ ( a) = 0 ζ ( a) = 0 ζ ( a) = 0
ζ′ ( a ) = 0 ζ′ ( a ) = 0 ζ′ ( a ) = 0 Mr ( a ) = 0 Mr ( a ) = 0
Inner Clamped Supported Free Free Supported
Boundary
ζ (b ) = − d ζ (b ) = − d Nr (b ) = 0 Nr (b ) = 0 ζ ( a) = − d
ζ′ ( b ) = 0 Mr ( b ) = 0 Mr ( b ) = 0 Mr ( b ) = 0 Mr ( b ) = 0
Note: For the distinct cases (Figures 10.5a–e) of a circular heavy elastic plate with a concentric
circular hole (Table 10.1) the boundary conditions are indicated.

namely: (I) clamped at both boundaries (Figure 10.6a) with (10.219b–e) a height
difference d; (II) clamped at the outer boundary (10.220a, b) and supported at the
inner boundary (10.220c, d) (Figure 10.6b): (III) clamped at the outer boundary
(10.221a, b) and free at the inner boundary (10.221c, d) (Figure 10.6c); (IV) sup-
ported at the outer boundary (10.222a, b) and free at the inner boundary (10.222c,
d) (Figure 10.6d); (V) supported both at the outer boundary (10.223a, b) and the
inner boundary (10.223c, d) (Figure 10.6e); (VI, VII, VIII) the cases II, III, IV of
unequal boundaries can be reversed, respectively, for (VI) supported outside and
clamped inside, (VII) clamped inside and free outside and (VIII) supported inside
and free outside. The remaining combination IX of 3 × 3 = 9 is free both at
the inner boundary and the outer boundary, which is not statically stable
because there is no support.

EXAMPLE 10.12 Vibrations of a Membrane under Uniform

or Non-Uniform Tension
Consider a homogeneous elastic membrane held at two parallel sides by a
normal uniform tension (Figure 10.7) or non-uniform tension (Figure 10.8).
The tension parallel to the sides is constant and there is no shear stress.
The linear transverse displacement of an elastic membrane is specified by
(10.224b), which balances the stresses (6.369a) against minus the inertia force
(6.754b):

∂2 ζ ∂2 ζ ∂2 ζ
Txy = 0: Txx + Tyy ( x ) = ρ , (10.224a, b)
∂x2 ∂y 2 ∂t 2
52 Differential Equations and Applications

T xy = 0 T yy = const > 0

T xx = const > 0 T xx

0
x

T yy
a

+ +
+
−
−
+
−
+ −
n=2 n=3 n=4

b c d

ζ
e

n<n
cut-off evanescent
x

cut-on
n>n
propagating

FIGURE 10.7
A homogeneous elastic membrane held at two parallel boundaries and (a) like an infi-
nite strip in the orthogonal direction, subject to constant normal tractions and no shear
stress, has transverse modes of oscillation of all orders, for example, (b) two, (c) three, and
(d) four. The modes below (above) a transition value n are (e) cut-on (cut-off) or propagating
(evanescent).
Examples 10.1 to 10.20 53

dT yy / dx > 0

T xy = 0

T xx = const > 0 T xx

0
x

a
T yy

ζ T yy

x
x

+
−
c
+
−

FIGURE 10.8
In Figure 10.6ª, the stress normal to the boundary (a) increases monotonically with dis-
tance along the boundary, each transverse mode has a turning point, such that before
(after) (b) the modes are oscillating (decaying), corresponding to (c) propagating
(evanescent) waves.
54 Differential Equations and Applications

where: (i) the mass density per unit area ρ is constant because the membrane
is homogeneous; (ii) there is no shear stress (10.224a); (iii)(iv) the stress paral-
lel (normal) to the sides is constant (non-uniform or constant). The case of
constant tension is considered first (E10.12.1) leading to propagating (evanes-
cent) that is cut-on (cut-off) modes. The case of non-uniform motion leads to
turning points (E10.12.2), as in the ray theory (notes 5.10–5.19).

E10.12.1  Propagating/Evanescent or Cut-on/Cut-off Modes

The membrane is fixed on two parallel sides (Figure 10.7a) and the boundary
conditions (10.225a, b):

 nπ y 
ζ ( x , y = 0, t ) = 0 = ζ ( x , y = a, t ) : ζn ( x , y , t ) = e iωt sin  X ( x ) , (10.225a–c)
 a 

are met by the y-dependence in the transverse displacement (10.225c), corre-

sponding to transverse modes of order n, of which n = 2, 3, 4 are represented
in Figure 10.7b, c, d. Considering a vibration of frequency ω in (10.225c), the
dependence of the transverse displacement on the longitudinal coordinate is
specified by substitution in (10.224b), leading to (10.226a):

2 1/2
2 2 2 2 π n
d X n /dx + k X n = 0, Txx kn = ρω − Tyy  , (10.226a, b)
n
 a 

where the wavenumbers kn are is specified by (10.226b). The traction parallel

to the supports Txx is always constant. Consider the case in when the traction at
the supports is uniform (Figure 10.7a). A real wavenumber in (10.226b) corresponds
to (10.227a):

ωa ρ
n< ≡ n: X n± ( x ) = C± exp ( ± i kn x ) , (10.227a, b)
π Tyy

to (10.227a) ≡ (10.228b) low-order (Figure 10.7e) propagating or cut-on modes

(10.227b; 10.225c) ≡ (10.228b):

 nπ y 
n < n: ζn± ( x , y , t ) = C± sin  exp  i ( ω t ± kn x )  , (10.228a, b)
 a 

 nπ y 
= C± sin  exp  i kn ( un t ± x )  , (10.228c)
 a 
Examples 10.1 to 10.20 55

propagating in the negative (positive) direction corresponding to the + (–) sign in

(10.228b) ≡ (10.228c) with phase speed (10.229c):

2 −1/2
Txx ω Tyy  π n 
u0 = : un = = u0 1 − , (10.229a–c)
ρ kn ρ  ω a 

where (10.229a) is the speed (10.229b) of the zero-order mode. The modes of order
higher (10.230a) than (10.227a) are (Figure 10.7e) cut-off or evanescent modes
(10.230b; 10.225c) ≡ (10.230c):

 nπy 
n > n: X n± ( x ) = C± exp ( ± kn x ) , ζn± ( x , y , t ) = C± sin  exp ( ± kn x ) cos ( ωt ) ,
 a 
(10.230a–c)

whose amplitude decays as x → + ∞ ( x → − ∞ ) choosing the lower – sign (upper +)

sign in (10.230c).

E10.12.2  Turning Point Due to Non-Uniform Tension

Consider next a fixed mode n and let the transverse traction at the supports be non-
uniform, for example, steadily increasing with x in Figure 10.8a. There is a turning

point x for which the wavenumber (10.226b) vanishes (10.231a):

2  x > x : propagation,
  aω 
kn ( x ) = 0: Tyy ( x ) = ρ  :  (10.231a–d)
 nπ  x < x: evanescence,


and the traction takes the value (10.131b). Before the turning point (10.231c),
there are vibrations like (10.228a–c; 10.229a, b) cut-on or propagating modes
(Figure 10.8b, c); beyond the turning point, the modes decay like cut-off or evanes-
cent modes (10.230a–c). In a strict sense, the wavenumber k only exists (10.226b)
when it is constant, that is for uniform traction (Figure 10.7a–e), corresponding to
sinusoidal vibrations; in the case of non-uniform tension the wavenumber (10.226b)
is not constant:

1/2
ρω 2 Tyy ( x )  nπ 
2
kn ( x ) = −   , (10.232)
Txx Txx  a 
56 Differential Equations and Applications

and the cut-on (cut-off) or propagating (evanescent) wavefields are given by

(10.233a, b) [(10.234a, b)] in the ray approximation (5.274a, b):

 nπy    x
 
ζn ( x , y , t ) = C±  kn ( x ) 
∫ kn ( ξ )  dξ  ,
−1/2
x > x: sin   exp i ω t ±
 a     
 (10.233a, b)

 nπy   x

x < x :    ζn ( x , y , t ) = C±  kn ( x ) 
∫ kn ( ξ ) dξ  ,
−1/2
sin  cos ( ωt ) exp  ±
 a   
 (10.234a, b)

when (7.431a–g) the local wavelength is much smaller than the lengthscale of varia-
tion of the stress:

2
 2π 
 λ n ( x )  2 = 
k ( x )
2 2
(
 ≤ L = Tyy ∂Tyy /dx ) −2
. (10.235)
 n 

If the ray approximation (notes 5.16–5.17, 7.35–7.27) does not hold, then the
notion of wavenumber or wavelength loses its meaning, even locally, and exact
global solutions must be sought to describe accurately vibrations or waves.
The vibrations of a membrane held are similar to quasi-one-dimensional
waves in non-uniform medium (notes 7.28–7.45), so analogous methods
apply.

EXAMPLE 10.13 Curve as the Tangent to a Vector Field

or as of the Intersection Surfaces
Obtain the curves defined as: (i) tangent to the vector field (10.236a):

{X , Y , Z} = {y 2 − z2 , z2 − x 2 , x 2 − y 2 } ;
(10.236a, b)
x 2 ( dx ) − y 2 ( dy ) + ( dz ) − 2 x dx dz = 0.
2 2 2

(ii) specified by the differential of degree two in three variables (10.236b).

From (10.236a), follow the identities (10.237a, b):

dx dy dz
X + Y + Z = 0 = x 2 X + y 2Y + z 2 Z ; = = ; (10.237a–c)
X Y Z
Examples 10.1 to 10.20 57

the curve tangent to the vector field (10.236a) satisfies (7.12a–c) ≡ (10.237c)
leading to (10.238a, b):

dx + dy + dz = 0 = x 2 dx + y 2 dy + z 2 dz: x + y + z = C1 , x 3 + y 3 + z 3 = C2 ,
(10.238a–d)

the corresponding primitives (10.238c, d) show that the two-parameter family of

curves to which the vector field (10.236a) is tangent are the intersection of the cubic
surface (10.238d) with the plane (10.238c).
The differential (10.236b) ≡ (10.239) of second-degree has coefficients
(10.240a–f):

0 = A ( dx ) + B ( dy ) + C ( dz ) + 2 D dx dy + 2 E dx dz + 2 F dy dz ; (10.239)
2 2 2

{ A , B , C , D , E , F } = { x 2 , − y 2 , 1 , 0 , − x , 0 } . (10.240a–f)

which satisfy the condition (3.365b) ≡ (10.240g):

A BC + 2 DE F − A F 2 − BE 2 − C D2 = − x 2 y 2 + x 2 y 2 = 0 ; (10.240g)

this leads to the factorization (3.366) ≡ (10.241a–c):

( )
2
0 = ( C dz + E dx + F dy ) −
2
E 2 − C A dx + F 2 − C B dy
(10.241a–c)
= ( dz − x dx ) − ( y dy ) = ( dz − x dx − y dy )( dz − x dx + y dy ) .
2 2

The solutions (10.242a, b) of (10.241c) show that:

x 2 + y 2 = 2 z + C1 , x 2 − y 2 = 2 z + C2 , (10.242a, b)

the differential (10.236b) of second degree in three variables determines a two-parameter

family of curves specified by the in tersection of circular (10.242a) and hyperbolic
(10.242b) paraboloids with a common axis.
58 Differential Equations and Applications

EXAMPLE 10.14 Differentials of First Degree in

Three/Four Variables
Solve the following first-degree differentials in three variables:

LXXIV: ( )
y 2 dx + 2 x y + z 2 dy + 2 y z dz = 0 ; x y 2 + y z 2 = C ; (10.243a, b)

LXXV: y z dx + 2 x z dy + 3 x y dz = 0 ; x y 2 z 3 = C ; (10.244a, b)

LXXVI: 2 y dx − x dy + z dz = 0 , y = C1 x: C1 x 2 + z 2 = C2 ; (10.245a–c)

and in four variables:

LXXXII: ( ) ( )
y 2 dx + 2 x y + z 2 dy + 2 y z + u2 dz + 2 z ud u = 0:
(10.246a, b)
x y 2 + y z 2 + zu2 = C ;

y z udu + 2 x z udy + 3 x y u + 4 x y z du = 0: x y 2 z 3 u4 = C ; (10.247a, b)

y dx + x dy + z dz − z du = 0, y = C1 x , u = C2 z: 2 C1 x 2 − C2 z 2 = − z 2 .
(10.248a–d)

In the cases where a first-degree differential in three (four) variables is not

exact one (two) subsidiary conditions is (are) used. The Roman numerals at
left indicate the standard in the recollection 10.1.

E10.14.1  Exact, Inexact, and Non-Integrable Differentials

The differential (10.243a) ≡ (10.241a) has vector (10.249b) with zero curl (7.27e) ≡
(10.249c):
   
0 = X . dx , {
X = y 2 , 2 x y + z2 , 2 y z ; } ∇ ∧ X = 0, (10.249a–c)

 
( )
0 = X .dx = y 2 dx + 2 x y dy + z 2 dy + 2 y z dz = d y 2 x + z 2 y , (10.249d)

and hence, is (7.27 a–e) immediately integrable (10.249d) ≡ (10.243a); thus,

the family of surfaces (10.243b) is orthogonal to the vector field (10.249b). The dif-
ferential (10.244a) ≡ (10.243a, b) is (7.28a–d) inexact (10.250c) but integrable
(7.28d) ≡ (10.250d):
     
(
0 = X . dx , X = { y z , 2 x z , 3 x y } ; ∇ ∧ X = { x , − 2 y , z} , X . ∇ ∧ X = 0 ; )
(10.250a–d)
Examples 10.1 to 10.20 59

 
( )
λ = y z 2 : 0 = λ X . dx = y 2 z 3 dx + 2 x y z 3 dy + 3 x y 2 z 2 dz = d x y 2 z 3 , (10.250e, f)

the integrating factor (10.250e) leads to the solution (10.250f) ≡ (10.244b). Thus,
the family of surfaces (10.244b) is orthogonal to the vector field (10.250b). The differ-
ential (10.245a) ≡ (10.251a, b) is (7.29 a–d) inexact (10.251c) and non-integrable
(10.251d):
     
0 = X .dx , X = { 2 y , − x , z} ; ∇ ∧ X = {0, 0 − 3} , X . ∇ ∧ X = −3 z ; ( )
(10.251a–d)

1
y = C1 x: 0 = 2 y dx − x dy + z dz = C1 x dx + z dz =
2
( )
d C1 x 2 + z 2 , (10.251e, f)

the subsidiary condition (10.251e) ≡ (10.245b) leads to the solution (10.251f) ≡

(10.245c). Thus, on the family of planes (10.251e) ≡ (10.245b) lie the family of curves
(10.245c) orthogonal to the vector field (10.251b).

E10.14.2 Immediate and Complete Integrability

and Subsidiary Conditions
The first-degree differential in four variables (10.246a) corresponds (10.252a)
to the vector field (10.252b); the differential is exact because the vector field
satisfies (3.278d). Hence, it is immediately (10.252c) integrable (10.252d) ≡
(10.246b):
  
0 = X .dx: {
X = y 2 , 2 x y + z 2 , 2 y z + u 2 , 2 z u , } (10.252a, b)

( ) ( ) ( ) (
0 = y 2 dx + 2 xy dy + z 2 dy + 2 yz dz + u2 dz + 2 zudu = d y 2 x + z 2 y + u2 z . )
 (10.252c, d)

Thus, the family of hypersurfaces (10.246b) is orthogonal to the four-dimensional

vector field (10.252b). The differential (10.247a) corresponds (10.253a) to the vec-
tor field (10.253b). It is inexact but integrable because the vector field does not
satisfy (3.278d) but does satisfy (3.282). It has the integrating factor (10.253c):
  
0 = X . dx: X = { y z u , 2 x z u , 3 x y u , 4 x y z} , (10.253a, b)

λ = yz 2 u3 : λ ( y z u dx + 2 x z u dy + 3 x y udz + 4 x y z du )

(
           = y 2 z 3 u4 dx + 2 x y z 3 u4 dy + 3 x y 2 z 2 u4 dz + 4 x y 2 z 3 u3 du = d x y 2 z 3 u4 , )
(10.253c–e)
60 Differential Equations and Applications

leading (10.253d) to (10.253e) ≡ (10.247b). Thus, the family of hypersurfaces

(10.247b) is orthogonal to the four-dimensional vector field (10.253b). The
differential (10.248a) corresponds (10.254a) to the vector field (10.254b);
the differential (10.248a) is inexact and non-integrable, since the vector
field does not satisfy (3.282). The subsidiary conditions (10.248b, c) lead to
(10.254c, d):

  
0 = X . dx : X = { y , x , z , − z } , (10.254a, b)

0 = C1 x dx + x d ( C1 x ) + z dz − z d ( C2 z ) = 2C1 x dx + ( 1 − C2 ) z dz

 z2  (10.254c–e)
= d C1 x 2 + ( 1 − C2 )  ,
 2

which coincides with (10.254c) ≡ (10.258d). Thus, the hyperplanes (10.248b, c)

intersect on a plane where lies the family of curves (10.248d) orthogonal to the four-
dimensional vector field (10.254b).

EXAMPLE 10.15 General Boundary Conditions for the

Bending of a Plate
Obtain the general boundary condition (6.609a) for an isotropic (6.453b)
[pseudo-isotropic orthotropic (6.608b)] plate including the explicit stress cou-
ple (6.609b) and turning moment (6.609c).
The method starts with the boundary integral in (6.453b) for an isotropic
plate (E10.15.1) extended to a pseudo-isotropic orthotropic plate using (6.808b).
An integration by parts along the boundary (E10.15.2) leads to the expres-
sion (6.609a) determining (E10.15.3) the stress couple and turning moment,
which is simplified if the plate has a fixed boundary (E10.15.4). Thus, the
general boundary conditions are obtained for the weak bending of a pseudo-
isotropic orthotropic plates (E10.15.5) including the particular case of an iso-
tropic plate. The boundary conditions apply to a plate whose boundary is
an arbitrary closed regular curve (E10.15) including, in particular, circular
and rectangular shapes (subsections 6.7.12–6.7.14). The boundary conditions
are given in the three most important cases of (i) clamped E10.15.6); (ii) sup-
ported or pinned (E10.15.7) and (iii) with a free boundary (E10.15.8) and are
summarized (E10.15.9) in Table 10.3.
Examples 10.1 to 10.20 61

TABLE 10.3
Boundary Conditions for the Bending of a Plate
Arbitrary Regular
Shape of Boundary Curve Circular Rectangular
Figure 6.24a 6.24b 6.24c
Equation (6.467e) (6.467c) (6.467a)
Slope (6.464a) (6.464c) (6.464e)
Stress Couple* (10.261)/(10.280) (10.283b)/(6.471d) (10.283d)/(6.471b)
Turning Moment* (10.262)/(10.281) (6.467d)/(6.467d) (6.467b)/(6.467b)
Clamped* (6.474a, b, d; 10.263b) (6.473a, b, d; 10.284)/ (6.472a–d)/(6.472a–d)
(10.271a–e) (6.474a–d) (6.473a–d)
Figure 6.23a
Supported or Pinned* (6.477a, b, d; 10.285)/ (6.476a, b, d; 10.276)/ (6.475a–d)/(6.475a–d)
(10.273a–e) (6.477a–d) (6.476a–d)
Figure 6.23b, c
Free Boundary* (6.480a, b; 10.278; (6.479a, b, d; 10.276)/ (6.478a, b, d; 10.282)/
(10.277a–e) 10.279)/ (6.479a–d) (6.478a–d)
Figure 6.23d (6.480a–d)
* for a plate made of a pseudo-isotropic orthotropic (isotropic) material before (after) the dash:
* first bending stiffness (6.413c, d) [(6.579a; 6.581a)];
* second bending stiffness (10.255a) [(10.255b)];

* stiffness ratio parameter (10.274a) [(10.275a)]

Note: The boundary conditions for the linear bending of an elastic plate lead to 3 × 3 × 2 = 18
cases combining: (i) three types of boundary: clamped, pinned, and free; (ii) three shapes:
arbitrary closed regular curve, circular, and straight; (iii) two types of elastic material:
isotropic and pseudo-isotropic orthotropic.

E10.15.1  Elastic Energy along the Boundary of a Plate

The boundary condition is (problem 301) the vanishing (10.255c) of the elastic energy
along the boundary on the r.h.s. of (6.453b):

h3
C44 = C44 → D ( 1 − σ ) = const:
12

∫ {D(∇ ζ) ∂ (δζ) − ∂ D(∇ ζ) δζ + C I } ds ,

(10.255a–c)
Eˆ d = 2
n n
2
44 n
∂C

where the first bending stiffness (6.593a; 6.595a) [(6.413c, d)] for a pseudo-
isotropic orthotropic (isotropic) plate need not be constant, and the second
stiffness (10.255b) [(10.255a)] is constant (6.454a). The coefficient of the first
(second) term of (10.255c) is in agreement with (6.609b) [(6.609c)] the turn-
ing moment (6.430a) [the stress couple (6.422a–c; 6.423a–c)] with additional
contributions coming from the third term on the r.h.s. of (10.255c). The latter
62 Differential Equations and Applications


involves the deformation vector I given by (6.448d) whose normal compo-
nent (6.451e) appears in (10.255c). Designating by θ the angle of the normal
with the x-axis (Figure 6.24a) leads to the unit normal (6.462a) [tangent
(6.462c)] vectors and associated derivatives (6.462b) [(6.462d)]. Using (6.462a),
the normal component of the deformation vector (6.448d) is given by:


( ) (
I n = I . n = cos θ  ∂ xy ζ ∂ y ( δζ ) − ∂ yy ζ ∂ x ( δζ )  )
(10.256)
( )
               + sin θ  ∂ xy ζ ∂ x ( δζ ) − ( ∂ xx ζ ) ∂ y ( δζ )  ,

for substitution in the last term of the r.h.s. of (10.255c), leading to an integra-
tion by parts along the boundary (E10.15.2).

E10.15.2  Integration by Parts along a Closed Regular Boundary

Substituting (10.256) in the integral along the boundary in the third term on
the r.h.s. of (10.255c) leads to (10.257):

∫∂c
C44 I n ds =
∫ ∂c
{ ( ) (
C44  2 sin θ cos θ ∂ xy ζ − sin 2 θ ( ∂ xx ζ ) − cos 2 θ ∂ yy ζ  δ ( ∂n ζ ) )
( ) ( )
                     +  cos θ sin θ ∂ yy ζ − ∂ xx ζ + cos 2 θ − sin 2 θ ∂ xy ζ  ∂ s ( δζ ) ds , }
(10.257)

where (6.463b, c) was used to transform the gradient of δζ in Cartesian coor-

{ }
dinates ∂ x , ∂ y to local normal ∂n and tangential ∂ s derivatives. The inte-
grand in the second term in square brackets on the r.h.s. of (10.257) is (10.258):

sin ( 2θ )
X≡
2
( )
∂ yy ζ − ∂ xx ζ + cos ( 2θ ) ∂ xy ζ, (10.258)

leading to an integration by parts along the closed regular boundary (10.259a):

∫ ∂c
C44 X ∂ s ( δζ ) ds +
∫
∂c
(
δζ ∂ s C44 X ds = ) ∫ ∂ (C
∂c
s 44 )
X δζ ds = C44 X δζ  = 0,

(10.259a–d)

where (10.259b) is the difference of the expression in square brackets at the

right-hand side (r.h.s.) and left-hand side (l.h.s) of the same point (10.259c)
and hence, is zero by continuity (10.259d). Substituting (10.259d) in (10.252)
Examples 10.1 to 10.20 63

and then in (10.255c) leads to (6.609a) ≡ (10.260), the elastic energy per unit
length of the boundary:

dEˆ
{ ( ) ( )
E d = d = −  ∂n D ∇ 2ζ  + C44 ∂ s  sin θ cos θ ∂ xx ζ − ∂ yy ζ − cos ( 2θ ) ∂ xy ζ  δζ,
ds }
{ }
                      + D ∇ 2ζ + C44  sin ( 2θ ) ∂ xy ζ − sin 2 θ ∂ xx ζ − cos 2 θ ∂ yy ζ  δ ( ∂n ζ ) ,
(10.260)

where (E10.15.3) minus the coefficient of δζ  ∂n ( δζ )  is the normal compo-

nent of the stress couple (6.609b) [turning moment (6.609c)].

E10.15.3  Stress Couple and Turning Moment

Thus, (problem 302) in the weak linear bending of a thin plate whose bound-
ary is an arbitrary closed regular curve (Figure 6.24a) the stress couple (turning
moment) normal to the boundary (10.260) ≡ (6.609a) is given by (10.261) ≡ (6.609b)
[(10.262) ≡ (6.609c)], which has the dimensions of force (moment) per unit length:

Mn = − D ∇ 2ζ − C44  sin ( 2θ ) ∂ xy ζ − sin 2 θ ∂ xx ζ − cos 2 θ ∂ yy ζ  (10.261)

( ) ( )
N n = ∂n D ∇ 2ζ − C44 ∂ s  sin θ cos θ ∂ xx ζ − ∂ yy ζ − cos ( 2θ ) ∂ xy ζ  . (10.262)

In the case of the nonlinear strong bending of a plate with in-plane stresses, the
augmented turning moment is given by (6.682a–c). If the bending moment D is con-
stant: (i) it can be taken out of the derivative in the first term on the r.h.s. of (10.262);
(ii) from (10.255b) the Poisson ratio is also constant. If (problem 303) the plate is
supported (Figure 6.23b) or pinned (Figure 6.23c) the vanishing of the displacement
on the boundary (10.263a) simplifies the stress couple (10.261) [turning moment
(10.262)] to (10.263b) [(10.263c, d)]:

 dθ  dθ
ζ = 0: Mn = − D  ∂nm ζ + ∂n ζ + C44 ∂n ζ . (10.263a, b)
 ds  ds

Nn dθ
D = const:
D
( )
= ∂n ∇ 2 ζ = ∂nnn ζ +
ds
∂nn ζ . (10.263c, d)

If (problem 304) the plate is clamped (Figure 6.23a) the conditions (10.264a, b)
simplify the stress couple to (10.264c) and leaves the turning moment unchanged
(10.263d) ≡ (10.264d):

Nn dθ
ζ = 0 = ∂n : Mn = − D ∂nn ζ, = ∂nnn ζ + ∂nn ζ . (10.264a–d)
D ds
64 Differential Equations and Applications

The relations (10.261, 10.262; 10.263a–d; 10.264a–d) hold for a plate with any shape
(Figure 6.24a) provided that the boundary is a closed regular curve. The proof of
(10.263a–d) is made next (E10.15.4) before considering the three main types
of boundary conditions (E10.15.6–E.10.15.8) for pseudo-isotropic orthotropic
and isotropic plates (E10.15.5).

E10.15.4  Plate with Arbitrary Closed Regular Boundary

The condition (10.263a) ≡ (10.265a) applies to clamped (Figure 6.23a), sup-
ported (Figure 6.23b), and pinned (Figure 6.23c) plates, and implies that: (i)
the derivatives of all orders along the boundary vanish (10.265b, c); (ii) the
gradient in Cartesian coordinates (6.463b, c) simplifies to (10.265d, e):

ζ = 0: ∂ s ζ = 0 = ∂ ss ζ, {∂ ζ , ∂ ζ} = { cos θ , sin θ } ∂ ζ . (10.265a–e)

x y n

The second-order derivatives of the transverse displacement in Cartesian

coordinates are given (6.463b, c; 10.265d, e) by (10.266a–c):

 dθ 
∂ xy ζ = ( cos θ ∂n − sin θ ∂ s ) sin θ ∂n ζ = cos θ sin θ  ∂nn ζ − ∂n ζ , (10.266a)
 ds 

dθ
∂ yy ζ = ( sin θ ∂n + cos θ ∂ s ) sin θ ∂n ζ = sin 2 θ ∂nn ζ + cos 2 θ ∂n ζ , (10.266b)
ds

dθ
∂ xx ζ = ( cos θ ∂n − sin θ ∂ s ) cos θ ∂n ζ = cos 2 θ ∂nn ζ + sin 2 θ ∂n ζ , (10.266c)
ds

where (10.267a) was used since θ is the direction of the normal, hence, does
not vary along it, implying (10.267b), which is the inverse (10.267c) of the
radius of curvature (6.461c):

dθ 1 dθ
∂n θ = 0,
= ; ζ = 0: ∇ 2ζ = ∂ xx ζ + ∂ yy ζ = ∂ nn ζ +
∂s θ ≡ ∂n ζ ,
ds R ds
(10.267a–f)

from (10.266b, c) follows that the Laplacian (10.267e) along the boundary
(10.267d) is given by (10.267f) in terms of normal derivatives in agreement
with the term in curved brackets in (6.474d). These results apply to pseudo-
isotropic orthotropic plates, which include as the particular case the isotro-
pic plate (E10.15.5).
Examples 10.1 to 10.20 65

E10.15.5  Isotropic and Pseudo-Isotropic Orthotropic Plates

Substituting (10.266a–c) in (10.262) the term in square brackets cancels and
the turning moment simplifies to (10.268a) leading to (10.268b) using the
Lapacian (10.267f):

  dθ 
( )
N n = ∂n D ∇ 2ζ = ∂n  D  ∂nn ζ +
  ds
∂ n ζ   ;

(10.268a, b)

from (10.268b) follows (10.263d) for constant bending stiffness (10.263c).

Substitution of (10.266a–c) in the term in square brackets in (10.261) cancels
the coefficient of ∂nn ζ and leaves only the coefficient of ∂n ζ:

2 sin θ cos θ ∂ xy ζ − sin 2 θ ∂ xx ζ − cos 2 θ ∂ yy ζ

(
= − 2 sin 2 θ cos 2 θ + sin 4 θ + cos 4 θ ) ddsθ ∂ ζ (10.269)
n

dθ dθ
( )
2
= − sin 2 θ + cos 2 θ ∂n ζ = − ∂n ζ .
ds ds

Substitution of (10.269) in (10.261) proves (10.263b) for a pseudo-isotropic

orthotropic plate. Thus, (problem 305) a clamped plate (Figure 6.23b) or a pinned
plate (Figure 6.23c) has a turning moment (10.268a, b), that simplifies to (10.263d) ≡
(10.264d) for constant bending stiffness (10.263c); the turning moment is the same
for isotropic (pseudo-isotropic orthotropic) plates with bending stiffness specified by
(6.413c, d) [(6.593a; 6.595a)]. The same applies for the stress couple (10.264c) of
clamped (Figure 6.23a) plate (10.129a, b). For a supported plate (Figure 6.23b) or
pinned plate (Figure 6.23c) the stress couple (10.263b) for a pseudo-isotropic ortho-
tropic plate simplifies for an isotropic plate (10.255b) to (10.270a):

Mn dθ dθ
− = ∇ 2ζ − ( 1 − σ ) ∂n ζ = ∂nn ζ + σ ∂n ζ , (10.270a, b)
D ds ds

where the Lapacian (10.267f) was used in (10.270b). The boundary conditions are
considered next for the three main cases of (i) clamped plate (E10.15.6), (ii)
supported or pinned plate (E10.15.7), and (iii) free boundary (E10.15.8).

E10.15.6  General or Rectangular Clamped Plate

For the weak linear bending of plate whose boundary is an arbitrary closed
regular curve (10.271a) the clamping (Figure 6.23a) is specified by zero dis-
placement (10.271b) and slope (10.271c), and it implies non-zero stress couple
(10.271d) [turning moment (10.271e)]:

s ( θ ) ∈C1 ( ∂C ) : ζ = 0 = ∂n ζ , Mn ≠ 0 ≠ N n , (10.271a–e)
66 Differential Equations and Applications

given by (10.263b) [(10.263d)] for a pseudo-isotropic orthotropic plate that

simplify to (6.474a–d) for an isotropic plate. For a clamped circular plate, the
boundary conditions in the isotropic case (6.473a–d) hold also for pseudo-
isotropic orthotropic plates replacing (6.473c) by (10.272a, b):

 1  1
(1
Mr = − D  ∂rr ζ + ∂r ζ − C44 ∂r ζ = − D ∂rr ζ − D + C44 ∂r ζ .
 r  r r
) (10.272a, b)

For the straight y = 0 side of a rectangular plate, the boundary conditions

(6.472a, d) for the isotropic case apply also to the pseudo-isotropic ortho-
tropic case.

E10.15.7  Boundary Conditions for Supported or Pinned Plates

The boundary conditions for general shape (10.271a) ≡ (10.273a) for support
plates (Figure 6.23b) or pinning plates (Figure 6.23c) are zero displacement
(10.273b) implying the non-zero turning moment (10.273e) and the non-zero
slope (10.273c) implying the zero stress couple (10.273d):

s ( θ ) ∈C 1 ( ∂C ) : ζ = 0 ≠ ∂ n ζ, Mn = 0 ≠ N n . (10.273a–e)

The non-zero normal component of the turning moment is given by (10.263d)

both for isotropic and pseudo-isotropic orthotropic plates; the vanishing
of the normal component of the stress couple (10.263b) leads to the sec-
ond boundary condition (10.274b) involving the stiffness ratio parameter
(6.606a–c) ≡ (10.274a):

C44 dθ dθ
χ≡ : ∂nn ζ + ∂n ζ = χ ∂n ζ ; (10.274a, b)
D ds ds

in the case of an isotropic plate (6.607a–c) ≡ (10.275a), the boundary condition

(10.274b) simplifies to (10.275b) in agreement with (6.477c) ≡ (10.275b):

dθ
χ = 1 − σ: ∂nn ζ + σ ∂n ζ = 0. (10.275a, b)
ds
In the case of a supported or pinned circular plate, made of an isotropic mate-
rial, the boundary conditions are (6.476a–d), where only (6.476c) is changed
for a pseudo-isotropic orthotropic material to (10.276):

1 χ
∂rr ζ + ∂r ζ = ∂r ζ . (10.276)
r r

For a supported or pinned rectangular plate, the boundary conditions

(6.475a–d) are the same in the isotropic and pseudo-isotropic orthotropic
cases.
Examples 10.1 to 10.20 67

E10.15.8  General, Circular, or Rectangular Plate with Free Boundary

At the free boundary of a plate with arbitrary shape (10.271a) ≡ (10.273a) ≡
(10.277a) neither the displacement (10.277b) [nor the slope (10.277c)] are zero
and thus, the turning moment (10.277e) [and stress (couple (10.277d)] must be
zero:

s ( θ ) ∈C 1 ( ∂C ) : ζ ≠ 0 ≠ ∂n ζ: Mn = 0 = N n . (10.277a–e)

The vanishing of the stress couple (10.261) [turning moment (10.262)] at the
free boundary lead to the first (10.278) [second (10.279)] boundary conditions
for a pseudo-isotropic orthotropic plate involving the parameter (6.606a–c):

−χ −1∇ 2ζ = sin ( 2θ ) ∂ xy ζ − cos 2 θ ∂ xx ζ − sin 2 θ ∂ yy ζ . (10.278)

( ) ( )
χ −1 ∂n ∇ 2ζ = ∂ s  sin θ cos θ ∂ xx ζ − ∂ yy ζ − cos ( 2θ ) ∂ xy ζ  . (10.279)

In the case (10.275a) of an isotropic plate, the general stress couple (10.261)
[turning moment (10.262)] simplifies to (10.280) [(10.281)]:

{ }
Mn = − D ∇ 2ζ − ( 1 − σ )  sin ( 2θ ) ∂ yy ζ − sin 2 θ ∂ yy ζ − cos 2 θ ∂ yy ζ  , (10.280)

( ) ( )
N n = ∂n D ∇ 2ζ − D ( 1 − σ ) ∂ s  sin θ cos θ ∂ xx ζ − ∂ yy ζ − cos ( 2θ ) ∂ yy ζ  . (10.281)

The vanishing of (10.261; 10.262) [(10.280; 10.281)] at the free boundary of a

pseudo-isotropic orthotropic (isotropic) plate leads to the same boundary
conditions (10.278; 10.279) with different stiffness ratio parameter (6.606a–c)
[(6.607a–c)]. At the free boundary of a circular (rectangular) plate, the bound-
ary conditions (6.479a–d) [(6.478a–c)] in the isotropic case also apply in the
pseudo-isotropic orthotropic case, replacing (6.479c) [(6.478c)] by (10.276)
[(10.282)]:

∂ yy ζ + ( 1 − χ ) ∂ xx ζ = 0. (10.282)

The 2 × 3 × 3 = 18 sets of boundary conditions combining different materials

(× 2), shapes (× 3) and support (× 3) are summarized next (E10.15.9).

E10.15.9  Sets of Boundary Conditions for the Bending of a Plate

The 18 sets of boundary conditions (problem 306) for the linear bending of a thin
plate are summarized in Table 10.3 according to three criteria: (i) the boundary is
an arbitrary closed regular curve (Figure 6.24a) and, in particular, [Figure 6.24b(c)]
circular (straight); (ii) the elastic plate is made of a pseudo-isotropic orthotropic
68 Differential Equations and Applications

material, in particular isotropic; (iii) the three main boundary conditions are clamped
(Figure  6.23a), supported (Figure 6.23b), or pinned (Figure 6.23c) or free bound-
ary (Figure 6.23d). The boundary conditions (6.480a–d) involve the displacement,
the slope, and the normal components of the stress couple and the turning moment.
Normal component means radial for a circular plate and along the y-axis for the
straight side along the x-axis of a rectangular plate. For arbitrary/circular/rectangu-
lar plate the boundary is given, respectively, by (6.467e, c, a) and the slope by (6.464a,
c, e). The normal component of the stress couple (turning moment) is given by
(10.261) [(10.262)] for a pseudo-isotropic orthotropic plate and by (10.280) [(10.281)]
for an isotropic plate. The stress couple for a pseudo-isotropic orthotropic (isotropic)
plate simplifies to (10.283b) [(6.471d)] for a circular plate (10.283a) and to (10.283d)
[(6.471b)] for a straight boundary (10.273c):

Mr 1− χ My
r = const:
= ∂rr ζ + − ; y = 0: − = ∂ yy ζ + ( 1 − χ ) ∂ xx ζ .
D r D
(10.283a–d)

The turning moment for a circular (6.467d) [rectangular (6.467b)] plate is the same
for the isotropic and the pseudo-isotropic orthotropic plate.
The general form of the stress general couple (10.261) [turning moment (10.262)]
simplifies: (i) to (10.263b)-[(10.263d)] for supported or pinned boundary (10.263a, c);
(ii) to (10.264c) [(10.264d)] for clamped boundary (10.264a, b). Thus, the boundary con-
ditions for arbitrary/circular/straight boundaries, respectively, of a pseudo-isotropic
orthotropic (isotropic) plate are: (i) for clamped (6.474a, b, d; 10.263b)/(6.473a, b, d;
10.284)/(6.472a–d) [(6.474a–d)/(6.473a–d)/(6.472a–d)]:

 1  1
Mr = − D  ∂rr ζ + ∂r ζ − C44 ∂r ζ ; (10.284)
 r  r

(ii) for supported or pinned boundary (6.477a, b, d; 10.285)/(6.476a, b, d; 10.276)/

(6.475a–d) [(6.477a–d)/(6.476a–d)/(6.475a–d)]:

1− χ
∂rr ζ + ∂r ζ = 0 ; (10.285)
r

(iii) for a free boundary (6.480a, b; 10.278, 10.279)/(6.479a, b, d; 10.276)/(6.478a, b, d;

10.282) [(6.480a–d)/(6.479a–d)/(6.478a–d)].

EXAMPLE 10.16 Strong Bending of a Pseudo-Isotropic

Orthotropic Plate
Generalize the equations of strong bending of a thick elastic plate from an
isotropic material (Foppl, 1907) to a pseudo-isotropic orthotropic material.
Examples 10.1 to 10.20 69

The balance equation (6.656) remains (problem 317) valid (10.286):

f D 2
( ) ( ) ( )(
= ∇ ζ − ∂ yy Θ ( ∂ xx ζ ) − ( ∂ xx Θ ) ∂ yy ζ + 2 ∂ xy Θ ∂ xy ζ , (10.286)
h h
)

replacing the first bending stiffness for an isotropic plate (6.413d) ≡ (10.287a)
by that for a pseudo-isotropic orthotropic plate (6.593a; 6.595a) ≡ (10.287b):

E h3 h3  C13 
← D →  C11 − C13 C  . (10.287a, b)
(
12 1 − σ 2
)
12 23

Concerning the second complementary equation (6.657) of the pair: (i) the
derivation from (6.659a, b) to the first equality in (6.666a–c) is independent
of the type of material; (ii) the elimination of the in-plane displacements is
made (10.288), as in (6.667a), retaining the strains from the first equality in
(6.666a–c) leading to:

1

2
2
( )
∂ yy ( ∂ x ζ ) + ∂ xx ( ∂ x ζ ) − 2 ∂ xy ( ∂ x ζ ) ∂ y ζ  = ∂ yy Sxx + ∂ xx Syy + 2 ∂ xy Sxy ;
2

(10.288)

(iii) the l.h.s. of (10.288) is simplified as in (6.667b):

( ∂ ζ) − ( ∂ ζ)( ∂ ζ) = ∂ ( A
xy
2
xx yy yy 11 ) (
Txx + A12 Tyy + ∂ xx A12 Txx + A22 Tyy )
(10.289)
(
+ 2 ∂ xy A44 Txy ,)

and on the r.h.s. the compliance matrix (6.583) is used for an orthotropic
material, bearing in mind that there are only in-plane stresses (6.312c–e);
(iv) substituting (6.655e) in (10.289) it follows that (problem 318) the comple-
mentary equation relating the transverse displacement to the stress function for the
strong bending of an orthotropic plate (10.290b):

Aab = const: ( ∂ ζ) − ( ∂ ζ)( ∂ ζ)  

xy
2
xx yy
(10.290a, b)
= A22 ∂ xxxx Θ + A11 ∂ yyyy Θ − 2 ( A12 + A44 ) ∂ xxyy Θ ,

assuming that the components of the compliance matrix are constant (10.290a).
The first (10.286) [second (10.290a, b)] equation of the coupled pair is valid
for a pseudo-isotropic (general) orthotropic plate. In the complementary
70 Differential Equations and Applications

equation (10.290b), there are three coefficients (10.291a–c) specified by the

compliance matrix (6.583):

 σ yx 1 
{ A22 , A11 , 2 ( A12 + A44 ) } =  E1 ,
1
Ex
,− 2 +
Ex Gyz 
 . (10.291a–c)
 y

Several coefficients (10.292a–g) of the stiffness matrix (6.586):

{ C11 , C22 , C33 , C44 , C12 , C13 , C23 }

 1 − σ yz σ zy 1 − σ xz σ zx 1 − σ xy σ yx
= , , , 2 Gyz ,
 C0 Ey Ez C0 Ex Ez C0 Ex Ey
σ yx + σ yz σ zx σ zx + σ zy σ yx σ zy + σ zx σ xy  (10.292a–g)
, , ,
C0 Ey Ez C0 Ey Ez C0 Ex Ez 

are needed, namely: (i)(ii) three (one) in the first (10.287b) ≡ (10.293b) [second
(6.595b) ≡ (10.293a)] bending stiffness:

C44 =
h3Gyz
, D=
h3 
1 − σ yz σ zy −
(
Ex σ zx + σ zy σ yx 
2
)
 ; (10.293a, b)
6 12 C0 Ey Ez  Ey σ zy + σ zx σ xy 
 

(iii)(iv) five (six) in the first (6.596a) ≡ (10.294a) [second (6.596b) ≡ (10.294b)]
conditions for a pseudo-isotropic orthotropic material:

( ) (
Ez 1 − σ xy σ yx Ey ( 1 − σ xz σ zx ) − Ex 1 − σ yz σ zy  )
(10.294a)
( ) ( )
2 2
= Ey σ zy + σ zx σ xy  − Ex σ zx + σ zy σ yx  ,

 E 
( ) (
Ez 1 − σ xy σ yx  1 − σ xz σ zx − x σ yx + σ yz σ zx 
Ey
)
  (10.294b)

( ) ( ) (
  = σ xy + σ xz σ zy Ey σ zy + σ zx σ xy − Ex σ yz + σ yz σ zx  . )
In conclusion, (problem 319) the transverse displacement and stress function in
the strong non-linear deflection of an elastic plate, made of a pseudo-isotropic ortho-
tropic material, satisfy the coupled non-linear fourth-order partial differential equa-
tions (10.286; 10.290a, b) where: (i) the first bending stiffness (10.287b) is given
by (10.293b); (ii) the second bending stiffness that appears in the boundary condi-
tions (10.261; 10.262) is given by (10.293a). Four (seven) terms of the compliance
(6.583) [stiffness (6.586)] matrix appear (10.291a–c) [(10.292a–g)] in the balance
Examples 10.1 to 10.20 71

equations (i-ii) and the boundary conditions (ii-iii). The three Young moduli Ei , six
Poisson ratio σ ij, and three shear moduli Gij satisfy (6.584b) [(6.587a–c)] in the com-
pliance (stiffness) matrix for an orthotropic material. In addition, (problem 320) a
pseudo-isotropic orthotropic material satisfies two additional relations (6.596a, b) ≡
(10.294a, b).

EXAMPLE 10.17 Non-Linear and Linear Coupled

Systems of Differential Equations
The simultaneous systems of differential equations are illustrated in the sim-
plest case of two coupled differential equations, either non-linear (E10.17.1) or
linear (E10.17.2).

E10.17.1 Simultaneous System of Non-Linear

Ordinary Differential Equations
Eliminate the system of two simultaneous non-linear second-order ordinary
differential equations:

y ′′z + y z ′ = 0, z ′′ + z y = 0, (10.295a, b)

in two possible ways and conclude that it is of order four.

The system (10.295a, b) is equivalent to the system of four first-order
equations:

dy dz yv
= y′ ≡ u , = z ′ ≡ v , u′ = y ′′ = − , v ′ = z ′′ = − z y . (10.296a–d)
dx dx z

Thus, the order of the system cannot exceed four, but could be less.
Eliminating for y leads to:

 yv  ′
2
y ′y v y 2 v ′ y 2 v z ′  yv
y ′′′y = − y   = − − + = y ′y ′′ + y 3 +  = y ′y ′′ + y 3 + y ′′ 2 ,
 z  z z z2  z 
(10.297a–d)

which is of the third-order. Thus, the system could be of order three or order
four. To find out which is the case, the system is eliminated for z:

z ′z ′′
z ′′′ = − ( z y )′ = − z ′y − y ′z = − y ′z, (10.298a–c)
z
72 Differential Equations and Applications

 z ′z ′′  ′  z ′′  ′
z 2 z ′′′′ = z 2   − z 2 z ′y ′ − z 3 y ′′ = z ′′′ z ′z + z ′′ 2 z − z ′′z ′ 2 + z 2 z ′   + z 2 z ′y ,
 z   z

= 2 z ′′′z ′z + z ′′ 2 z − 2 z ′′z ′ 2 − z ′′z ′z ,

(10.299a–c)
proving that it is of the fourth order.

E10.17.2 Linear Differential System with a Given

Characteristic Polynomial
Consider the coupled system of linear differential equations (10.300a, b):

y ′′ + c z ′′ + α y + β z = 0 = y ′ + z ′ + γ y + δ z , (10.300a, b)

and determine the constant coefficients ( α , β, γ , δ ) so that (10.301a) [(10.302a)]:

c = 0: P3 ( D ) = ( D − 1)( D + 1)( D + 2 ) , (10.301a, b)

P2 ( D ) = λ ( D − µ ) , 
2
c = 1: (10.302a, b)

the characteristic polynomial (10.301b) [(10.302b)]: (i) is of the third (second)

degree; (ii) has three distinct roots ±1, − 2 (a double root µ); and (iii) the coef-
ficient of the leading power is 1( λ ).
The coupled system of linear differential equations with constant
coefficients (10.300a, b) ≡ (10.303):

 D2 + α c D2 + β   y ( x) 
    = 0 , (10.303)
 D + γ               D + δ   z ( x) 
   

has the characteristic polynomial:

( ) (
P3 ( D ) = D2 + α ( D + δ ) − c D2 + β ( D + γ ) )
(10.304a, b)
= ( 1 − c ) D3 + ( δ − c γ ) D2 + ( α − β ) D + α δ − β γ ,

which simplifies to (10.305b) [(10.306b)]:

c = 0:       P3 ( D ) = D3 + δ D2 + ( α − β ) D + α δ − β γ , (10.305a, b)

c = 1:       P2 ( D ) = ( δ − γ ) D2 + ( α − β ) D + α δ − β γ , (10.306a, b)

for (10.301a) ≡ (10.305a) [(10.302a) ≡ (10.306a)].

Examples 10.1 to 10.20 73

In the case (10.301a) ≡ (10.305a) ≡ (10.307a), the characteristic polynomial

(10.301b) ≡ (10.307b) ≡ (10.307c):

c = 0:       ( )
P3 ( D ) = D3 − 1 ( D + 2 ) = D3 + 2 D2 − D − 2 , (10.307a–c)

coincides with (10.305b) if (10.308a–c) hold:

δ = 2:       β − α = 1, β γ − α δ = 2: β = 1 + α, γ = 2, (10.308a–e)

leading to (10.308d, e). Substituting (10.308a, d, e) in (10.300a, b; 10.301a) speci-

fies the differential system (10.309a, b):

y ′′ + α y + ( 1 + α ) z = 0 = y ′ + z ′ + 2 ( y + z ) . (10.309a, b)

The roots (10.310a) of the characteristic polynomial (10.301b) lead to the gen-
eral integral (10.310b):

a1− 3 = ±1, − 2: y ( x ) = C1 e −2 x + C2 cosh x + C3 sinh x , (10.310a, b)

where C1− 3 are arbitrary constants.

In the case (10.302a) ≡ 10.306a) ≡ (10.311a), the characteristic polynomial
(10.302b) ≡ (10.311b):

c = 1: P2 ( D ) = λ D2 − 2 λ µ D + µ 2 λ , (10.311a, b)

coincides with (10.306b) if (10.312a–c) hold:

δ − γ = λ, β − α = 2 λ µ, α δ − β γ = µ 2 λ , (10.312a–c)

the compatibility of (10.312a–c) requires:

α δ − β γ = µ2 λ =
( 2 µ λ )2 = (β − α )2 , (10.313a–d)
4λ 4 (δ − γ )

that the constants ( α , β, γ , δ ) be related by (10.313d) ≡ (10.314a):

β−α
( β − α )2 = 4 ( δ − γ ) ( α δ − β γ ) : λ = δ− γ, 2µ = , (10.314a–c)
δ−γ
74 Differential Equations and Applications

in which case the parameters ( λ , µ ) are given by (10.312a, b) ≡ (10.314b, c). The
differential system (10.300a, b; 10.302a) ≡ (10.315a, b):

y ′′ + z ′′ + α y + β z = 0 = y ′ + z ′ + γ y + δ z, y ( x ) = e µx ( D1 + D2 x ) .
(10.315a–c)
has the general integral (10.315c) where D1− 2 are arbitrary constants.
In the case (10.309a, b) substitution of (10.310b) in (10.309a) ≡ (10.316a) leads
to the second general integral (10.316b):

( 1 + α ) z ( x ) = − y ′′ ( x ) − α y ( x ) = − ( α + 4 ) C1 e −2 x − ( 1 + α ) ( C2 cosh x + C3 sinh x ) .
 (10.316a, b)

In the case (10.315a, b) substitution of (10.315c; 10.317a) in (10.315b) leads to

(10.317b, c):

z ( x ) = e µx ( E1 + E2 x ) : D2 ( µ + γ ) + E2 ( µ + δ ) = 0,
(10.317a–c)
D1 ( µ + γ ) + D2 + E1 ( µ + δ ) + E2 = 0.

Solving (10.317b, c) for E1,2 leads to (10.318a, b):

µ+γ µ+γ γ −δ
E2 = − D2 , E1 = − D1 + D2 , (10.318a, b)
µ+δ µ+δ ( + δ )2
µ

and hence (10.317a) to the general integral (10.319):

µ+δ  γ −δ 
z ( x ) = − e µx  D1 + D2 x − D2 . (10.319)
µ+γ  (µ + γ ) (µ + δ ) 
In conclusion, the coupled system of linear differential equations with constant coef-
ficients (10.300a, b) for (10.301a) [(10.302a)]: (i) simplifies to (10.309a, b) [(10.315a,
b)]; (ii) has the cubic (10.301b) [quadratic (10.302b)] characteristic polynomial if
the four constants satisfy (10.308a, d, e)[(10.314a–c)]; (iii) has the general integral
(10.310b; 10.316b) [(10.315c; 10.319)] where C1,2,3 ( D1,2 ) are arbitrary constants.

EXAMPLE 10.18 Three Coupled Systems with a Matrix

of Characteristic Polynomials
The simultaneous systems with matrix of characteristic polynomials are
illustrated in the simplest case of two coupled forced equations of three types;
(i)(ii) linear differential equations with constant (homogeneous) coefficients
Examples 10.1 to 10.20 75

[E10.18.1 (E10.18.2)]; (iii) linear finite difference equations with constant coef-
ficients (E10.18.3–E10.18.6).

E10.18.1 Forced Coupled Linear Differential System

with Constant Coefficients
Consider the system (10.320a, b):

y ′ + z′ − y = e 2 x + x, y ′′ + z ′ − y + z = e x + x 2 , (10.320a, b)

and obtain the complete integral.

The forced coupled system of linear differential equations with constant
coefficients (10.320a, b) ≡ (10.321a, b):

D − 1  D   y ( x)      
D≡
d
:     = e 2 x  1  + e x  0  +  x  , (10.321a, b)
dx  D − 1 D + 1
2
 z ( x)  0
2
 1   x 
   

has the characteristic polynomial (10.322a–c):

P3 ( D ) =
D − 1            D ( )
= D2 − 1 ( 1 − D ) = − ( D − 1) ( D + 1)
2

D2 − 1       D + 1 = − D3 + D2 + D − 1,
(10.322a–d)

with double (single) root 1(−1) leading to the general integral (10.323a) for y(x),
where (A, B, C) are arbitrary constants:

y ( x) = ( A + B x) ex + C e− x ; z′ ( x ) = y ( x ) − y ′ ( x ) , (10.323a, b)

the general integral (10.323b) for z follows from (10.320a) without forcing and
substitution of (10.323a) leads to (10.323c):

z′ ( x ) = − B e x + 2 C e − x , z ( x) = − B ex − 2 C e− x , (10.323c, d)

and hence by integration to (10.323d).

The forcing by the first term on the r.h.s. of (10.321b), which is a non-
resonant exponential (7.121a, b) uses the inverse matrix of (10.321b) and leads
to the particular integrals (10.324a–d):

 y ( x)  bx 
b + 1 −b  1 
  = lim e   
 z ( x )  b→ 2 P3 ( b )  1 − b 2 b − 1   0 
 
(10.324a–d)
e 2x  b+1  e 2x 
3   −1 
            = lim  =−  =e 
2x
.
P3 ( 2 ) b→ 2  1 − b 2  3  −3   1 
76 Differential Equations and Applications

The forcing by the second term on the r.h.s. of (10.321b), which is a doubly
resonant exponential (7.122a, b) is given by (10.325a–d):

 y ( x)    b + 1 −b   0  
 = lim 1 ∂ 2
2
 bx
e      
 z ( x )  b→1 P3′′( b ) ∂b
  1 − b b − 1
2
    1  

1 ∂2  bx  −b  
= lim e  
b→ 1 2 − 6 b ∂b 2   b − 1  
(10.325a–d)
1   −b   −1  
= − lim e bx  x 2   + 2x 
4 b→1   b − 1   1  

e x  x( x + 2) 
    =  .
4  −2 x 


The forcing by the last term on the r.h.s. of (10.321b) leads to particular inte-
grals (10.326a–d) using the inverse matrix of polynomials of derivatives:

 y ( x)  D + 1         − D   x   1− x 
 = 1     = −
1
 
 z ( x )  P3 ( D )  1 − D D − 1   x 2 
 
2
1 − D − D2 + O D3( )  3 x − x 2 

 1 − x           x 
= − 1 + D + 2 D2 + O D3   ( )
= 2 
 3 x − x 2   x − x + 1 
.

(10.326a–d)

Adding the general integral (10.323a, d) of the unforced system on the l.h.s. of
(10.320a, b) ≡ (10.321a, b) to the particular integrals (10.324d; 10.325d; 10.326d)
corresponding to the three forcing terms on the r.h.s. leads to:

x( x + 2) x
y ( x) = ( A + B x) ex + C e− x − e2x + e + x , (10.327a)
4

x ex
z ( x) = − B ex − 2 C e− x + e2x − + x 2 − x + 1, (10.327b)
2

as the complete integral (10.327a, b) of the coupled forced system of linear differential
equations with constant coefficients (10.320a, b) ≡ (10.321a, b).
Examples 10.1 to 10.20 77

E10.18.2 Forced Coupled Linear Differential System

with Homogeneous Coefficients
Consider the system (10.328a, b):

x y ′ + x z ′ − y = x 2 + log x , x 2 y ′′ + x y ′ + x z ′ − y + z = x + 2 log x , (10.328a, b)

and obtain the general integral.

The forced coupled linear system of differential equations with homoge-
neous coefficients (10.328a, b) ≡ (10.329a, b):

d  δ − 1                       δ      y (x)      
δ≡x :     = x 2  1  + x  0  +  1  log x ,
dx  δ ( δ − 1) + δ − 1 δ + 1   z (x)  0 1   2 
 
 (10.329a, b)

has the same matrix (10.321b) ≡ (10.329b) replacing homogeneous (10.329a) by

ordinary (10.321a) derivatives. Hence, the changes of variables (10.10a, b) lead
from (10.327a, b) to:

C x
y ( x ) = ( A + B log x ) x + − x 2 + log x ( 2 + log x ) + log x , (10.330a)
x 4

2C x
z ( x) = − B x − + x 2 − log x + log 2 x − log x + 1, (10.330b)
x 2

as the complete integral (10.330a, b) of the forced coupled linear system of differential
equations with homogeneous coefficients (10.328a, b) ≡ (10.329a, b).

E10.18.3  Forced Coupled System of Finite Difference Equations

Consider the system (10.331a, b):

y n+ 1 + zn+ 1 − y n = 0, y n+ 2 + zn+ 1 − y n + zn = 0, (10.331a, b)

and obtain the complete solution.

The unforced coupled linear system of finite difference equations with
homogeneous coefficients (10.331a, b) ≡ (10.332a, b):

 ∆ − 1         ∆        yn 
∆ { y n , zn } ≡ { y n + 1 , zn + 1 } :  2    = 0 , (10.332a, b)
 ∆ − 1 ∆ + 1   zn 

coincides with the unforced part of (10.321b) [(10.329b)] replacing ordinary

(10.321a) [homogeneous (10.329a)] derivatives by forward finite differences
78 Differential Equations and Applications

(10.332a). The changes of variable (1.445c, 1.446c, 1.447c) ≡ (10.33a, b) and

(1.448b–d) ≡ (10.333c, d):

e ax ↔ xa ↔ an , xn ↔ log n x ↔ n , (10.333a–d)

lead from the first three (two) terms of (10.327a) ≡ (10.330a) [(10.327b) ≡
(10.330b)] to (10.334a) [(10.334b)]:

yn = A + B n + C ( − ) , zn = − B − 2 C ( − ) ,
n n
(10.334a, b)

as the general solution (10.334a, b) of the coupled unforced system of linear finite
difference equations with constant coefficients (10.331a, b).

E10.18.4  System of Finite Difference Equations with Oscillatory Forcing

Consider the forcing of a coupled linear system of finite difference equations
with constant coefficients by products consisting of (i) power with integral
exponent and (ii) an hyperbolic [and/or (iii) circular] cosine or sine of mul-
tiple angles; in particular, determine a particular solution of the system:

y  + 1 − y  − z = b  cos ( 2 φ ) , y  + 1 + z + 1 + y  = b  cos ( 2 φ ) , (10.335a, b)

for all values of ( b , φ ). The solution of (10.335a, b) includes oscillatory (non-

oscillatory) general (particular) cases [E10.18.5 (E10.18.6)].

E10.18.5  Forcing by Power Multiplied by a Cosine

Consider the coupled linear system of finite difference equations with con-
stant coefficients forced (10.335a, b) ≡ (10.336a, b) product of (i) a power with
integral exponent by (ii) an circular cosine of the double of the same multiple
of an angle:

 ∆ − 1     −1   y    1  

∆ + 1 ∆
   =   b cos ( 2 φ ) = Re b e
 z  1
 i 2 φ
(
. ) (10.336a, b)
     


The matrix of polynomials of finite differences (10.336a) ≡ (10.337a) has the

matrix of co-factors (10.337b):

 ∆ − 1      −1        ∆            1 

Rm , r ( ∆ ) =  , Rm , r ( ∆ ) =   , (10.337a, b)
 ∆ + 1   ∆   −∆ − 1 ∆ − 1 

and the determinant specifying (10.338a–c) the characteristic polynomial:

R2 ( ∆ ) = Det  Rm , r ( ∆ )  = ∆ ( ∆ − 1) + ∆ + 1 = ∆ 2 + 1; (10.338a–c)
Examples 10.1 to 10.20 79

in the solution (7.299b) appear (10.339a, b):

      ∆             1   1   1 + ∆ 
Rr , m ( ∆ ) Gm =    =   . (10.339a, b)
 −∆ − 1   ∆ − 1   1      −2 

Substitution of (10.337b; 10.338c) in (7.299b) leads to the particular solution

(10.340):
 y    ∆   1 + ∆  
  =  limi 2 φ 2    , (10.340)
 z   ∆→b e 1 + ∆  −2  

of (10.335a, b) ≡ (10.336a, b).

The characteristic polynomial (10.338c) ≡ (10.341) appears in (10.340):

( )
R2 b e i 2 φ = b 2 e 4 iφ + 1 = b 2 cos ( 4φ ) + 1 + i b 2 sin ( 4φ ) , (10.341)

in the inverse form (10.342b, c):

1 b 2 cos ( 4φ ) + 1 − i b 2 sin ( 4φ ) b 2 e − i 4φ + 1
b 2 e i 4φ ≠ −1:
= = ,
R2 b e i2φ
( )
b 2 cos ( 4φ ) + 1 + b 4 sin 2 ( 4φ ) 1 + b + 2 b cos ( 4φ )
2 4 2

(10.342a–c)

excluding the special case (10.342a), which is considered separately (E10.18.5).

Substituting (10.342c) in (10.340) leads to (10.343a–c):

 y  
b e i 2 φ ≠ ± i :   = Re 
( )(
 b e i 2 φ  1 + b 2 e − i 4φ ) 1 + b e i 2 φ  
 
 1 + b + 2 b cos ( 4φ )
4 2
 z   − 2  


 e i 2 φ + b e i 2 (  + 1)φ + b 2 e i 2 (  − 2 )φ + b 3 e i 2 ( − 1)φ 
b
= × Re  ,
1 + b 4 + 2 b 2 cos ( 4φ )  −2 e i 2 φ − 2 b 2 e i 2( − 2 )φ 
 
(10.343a–c)

which specifies the particular solution (10.344b, c) of (10.335a, b) ≡ (10.336a, b):

π cos ( 2 φ ) + b 2 cos  2 (  − 2 ) φ 
(b, φ) ≠  1, ± : z = − 2 b  , (10.344a, b)
4 1 + b 4 + 2 b 2 cos ( 4φ )

cos ( 2 φ ) + b cos  2 (  + 1) φ  + b 2 cos  2 (  − 2 ) φ  + b 3 cos  2 (  − 1) φ 

y = b ,
1 + b 4 + 2 b 2 cos ( 4 φ )
(10.344c)

except in the special case (10.344a), which is considered next (E10.18.5).

80 Differential Equations and Applications

E10.18.6  Special Case of Constant Forcing

The case (10.344a), which is excluded from (10.344b, c), corresponds to
(10.343a) ≡ (10.345a) in (10.335a, b) ≡ (10.345b, c):

c ≡ b e i 2 φ = ± i: ( )
y  + 1 − y  − z = Re c  = y  + 1 + z + 1 + y  , (10.345a–c)

which is equivalent to (10.345a–c) ≡ (10.346a–c):

 π
( ) (
Re c  = Re e ± i π /2 : ) y  + 1 − y  − z = cos 
 2 
= y  + 1 + z + 1 + y  .

(10.346a–c)

The system (10.345b, c) ≡ (10.347):

 ∆ − 1           −1   y     1 
    = c   , (10.347a)
 ∆ + 1       ∆   z  1

has for (10.347b) the solutions (10.347c, d), where (10.337b; 10.338c) were used:

 y  c         c              1     1  c    1 + c 

c ≠ ± i:  =   =   . (10.347b–d)
 z  R2 ( c )  − c − 1      c − 1     1  1 + c      −2 
2

The resonant case (10.347e) leads (7.300b) to (10.347f–i):

   y    1 ∂   1 + c   
c ≠ ± i:   = Re  lim   c  −   

c →± i R2′ ( c ) ∂ c
z
       2   

  log c   c  + c  + 1   π   1 ± i  
= Re  lim      = Re c ± i π /2  
 c→± i  2 c   −2   4     −2  

π  cos ( π /2 ) − sin ( π /2 ) 

 
= .
4 −2 cos ( π /2 ) 
 
(10.347e–i)

Thus, the system (10.345b, c) [(10.346b, c)] has particular solution (10.347c, d) ≡
(10.348a–c) [(10.347e, c) ≡ (10.349a–c):

1+ c 2 c
c ≠ ± i: y = c  , z = − ; (10.348a–c)
1 + c2 1 + c2
Examples 10.1 to 10.20 81

π  π  π  π  π

c = ± i: y = cos  − sin  , z = − cos  . (10.349a–c)
4   2   2   2  2 

In conclusion, the particular solution of (10.335a, b) ≡ (10.336a, b) is (10.344b, c)

except (10.344a) ≡ (10.343a) ≡ (10.342a) for the special case (10.346a, c; 10.349b, c).

E10.18.7  General and Complete Solutions

In both cases, the general solution is (10.350a, b):

 π   π 
y  = A cos   + B sin   , (10.350a)
 2  2

 π   π 
z = ( B − A ) cos   − ( A + B ) sin   , (10.350b)
 2  2

whose value is indicated in Table 10.4, where (A,B) are arbitrary constants deter-
mined by initial values. The complete solution adds to the general solution (10.350a,
b), the particular solutions (10.344a–c), simplifying to (10.348a–c) for (10.345a–c)
and (10.349a–c) for (10.349a, b). The general solution (10.350a, b) is obtained as
follows: (i) the characteristic polynomial (10.338c) has roots (10.351a, b) lead-
ing to the solutions (10.351c, d):

 π   π 
∆ = ± i = e ± iπ/2 : (e )
± i π/2 
= e ± i  π/2 = cos   ± i sin   , (10.351a–d)
 2  2

TABLE 10.4
General Solution (10.350a, b) of Coupled System of Finite Difference Equations
(10.346a, b)
= 4n 4n + 1 4n + 2 4n + 3
 π  1 0 –1 0
cos  
 2

 π  0 1 0 –1
sin  
 2
y A B –A –B
z B–A A–B A–B A+B
Note: The general integral (10.350a, b) of the coupled system of linear finite difference equa-
tions with constant coefficients (10.334a, b) leads to four cases, for , a multiple of 4, or
adding 1, 2, or 3.
82 Differential Equations and Applications

of which a real linear combination with arbitrary constants (A, B) was chosen
in the r.h.s. (10.350a); (ii) substituting (10.350a) in the l.h.s. of (10.335a) leads
to (10.351e, f):

  π π   π     π π   π  
z = y  + 1 − y  = A  cos  +  − cos    + B  sin  +  − sin   
  2 2   2     2 2   2 

  π   π     π   π  
= A  − sin   − cos    + B  cos   − sin    ,
  2   2     2   2 
(10.351e–g)

which coincides with (10.351g) ≡ (10.350b).

EXAMPLE 10.19 Asymptotic Stability of a Damped

Non-Linear Oscillator
Prove that the damped oscillator with parameters depending on position
(9.109a) ≡ (9.111b) is asymptotically stable in all cases, including v = 0 when
the total energy (9.112c) as stability function is inconclusive because H = 0 in
(9.113e).
The stability function is chosen (10.352b) as total energy (9.112a) plus an
extra term (10.352c):
x
m 2
f ( x ) = k ( x ) x: H = E + αmv f =
2
v +
∫ 0
f ( ξ ) dξ + α v m f , (10.352a–c)

where (10.352a) is the restoring force, and α is a constant to be chosen at will.

To prove the asymptotic stability, in all cases, it is sufficient to show that
there is a value of α for which: (i) the stability function (10.352c) is positive-
definite (10.353a):

∂ H dx ∂ H dv ∂H µ v + f ∂H
H > 0 > H = + =v − ; (10.353a–d)
∂ x dt ∂ v dt ∂x m ∂v

(ii) its time derivative (10.353c, d) following the differential system is

negative-definite (10.353b). In order to prove the inequalities (10.353a, b), use
will be made of the identity (10.354a, b):

2
 a  a2
β > 0:  + b β > 0 → − 2 ab < + b 2β, (10.354a–c)
 β  β

in the form (10.354c).

Examples 10.1 to 10.20 83

E10.19.1  Positive-Definite Stability Function

The inequality (10.354c) with (10.355a) is applied to the last term of the stabil-
ity function (10.352b) leading to (10.355b, c):

mα 2 m x
m 2
2
v + f 2 = (1 − α ) v2 +
β = 1:
2
H ≥E−
0
f ( ξ ) dξ − (
2
f α. ) ∫
(10.355a–c)

The middle term is compared with the last term on the r.h.s. of (10.355c)
using L’Hôpital rule (I.19.35):

 f ( x ) 
2
2f′ f
lim x = lim = 2 f ′ ( 0 ) = 2 lim  x k ( x )  ′ = 2 k ( 0 ) > 0,
f (10.356a–d)
∫ f ( ξ ) dξ
x→ 0 x→ 0 x→ 0

where k(0) is the resilience of the spring at zero displacement, as for a lin-
ear restoring force. It follows that there exists a constant (10.357a) such that
(10.357b) holds:

mα c  x x
( 1 − α ) v 2 +  1 −
m
 f ( x )  ≤ c
∫ f ( ξ ) dξ ; H ≥
f ( ξ ) dξ ,
∫
2
c > 0: 
0 2 2  0
(10.357a–c)

substitution of (10.357b) in (10.355c) leads to (10.357c), showing that the

Lyapounov or the stability function can be made positive-definite (10.358c)
by choosing α to satisfy (10.358a, b):

 2  x
α < 1, :
 mc 
H > 0; f ( x) > 0 <
∫
0
f ( ξ ) dξ . (10.358a–e)

In (10.357a), as in (10.358d, e), the restoring force is positive; otherwise there

would no stable of equilibrium and the system would not be an oscillator.
Having proved the condition (10.353a) it remains to prove (10.353b).

E10.19.2  Negative-Definite Time Derivative of the Stability Function

Denoting by prime (10.359b)[dot (10.359a)], the spatial (time) derivative
implies (10.359c):

dg dg dg dx
g ≡ , g′ = : g = = g ′ v , (10.359a–d)
dt dx dx dt
84 Differential Equations and Applications

which is used in the time derivative (10.353d) of the stability function

(10.352c), specified by (10.360a–c):

H = m v ( v + α f ) + f v + α m f ′ v 2 = − ( µ v + f )( v + α f ) + v ( f + α m f ′ v )

= − µ v2   + α m f ′ v2 − α f 2 − α µ f v .
(10.360a–c)

The inequality (10.354c) is applied to the last term on the r.h.s. of (10.359c)
leading to (10.360d):

µα  2 f2
H ≤ − µ v 2 + α m v 2 f ′ − α f 2 + v β + . (10.360d)
2  β 

Introducing the upper bound (10.361a) for f ′ ( x ), which must be finite, leads
to (10.361b, c):

 α mB αβ  µ 
B = max f ′ ( x ) :
H ≤ −µ v 2  1 − − − α f 2  1 −   < 0.
 µ 2    2β 
(10.361a–c)

The derivative of the stability function (10.361b) can be made negative-

definite (10.361c) by choosing: (i) a large enough β so that (10.362a), the last
term on the r.h.s. of (10.361b), is negative:

2 1 µ
β> ; α< < , (10.362a–c)
µ mB β 1 + mB
+
µ 2

(ii) a small enough α so that (10.362b), the first term on the r.h.s. of (10.361b),
is also negative; (iii) the value (10.362c) of α is compatible with (10.358a) since
it suffices to comply with the smallest of the two upper bounds. Thus, α
meeting (10.358a; 10.361c) is sufficient to ensure the conditions (10.353a, b) of
asymptotic stability, quod erat demonstrandum (QED).

EXAMPLE 10.20 Solutions of the Generalized Circular

and Hyperbolic Differential Equation
Obtain the solution of the generalized circular (10.363b)[hyperbolic (10.363c)]
differential equations for negative integer values of the parameter (10.363a):

− m ∈|N :                      y ′′ ± x m y = 0, (10.363a–c)

Examples 10.1 to 10.20 85

which is in the cases when the generalized circular and hyperbolic cosine
and sine (subsections 9.4.12–9.4.20) do not apply directly.
The analysis is distinct for: (i)(ii) a simple m = −1 (double m = −2) pole
[E10.20.1(E10.20.2)], for which the origin is a regular singularity leading to
regular integrals (elementary power solutions); (iii)(iv) a pole of third m = −3
(higher m = −4 − 5....) order [E10.20.3(E10.20.4–E10.20.5)], for which the origin
is an irregular singularity and irregular integrals are needed.

E10.20.1  Simple Pole and Regular Integrals

The case of a simple pole (10.364a) leads to the differential equations (10.364b, c):

m = − 1: x y ′′ ± y = 0, y ( x) = ∑x
k=0
k+a
c k ( a ) , (10.364a–d)

which have regular integrals (10.364d) leading to:

∞ ∞

k = n + 1: 0= ∑( k + a)( k + a − 1) xk + a−1 ck ( x) ± ∑ xa+n cn ( a)

k=0 n= 0

∞ (10.365a–c)
= xa ∑n= 0
x n ( n + a )( n + a + 1) cn+ 1 ( a ) ± cn ( a )  .

From (10.365c) follows the recurrence formula for the coefficients (10.366a):

( n + a )( n + a − 1) cn ( a ) =  cn− 1 ( a ) , n = 0: a± = 0 , 1, (10.366a–d)

and for (10.366b) the indices (3.366c, d). The higher index (10.366d) leads
(10.366a) to the coefficients (10.367a–c):

cn− 1 ( 1) c0 ( 1) c0 ( 1)
cn ( 1) =  = () = ()
n n
. (10.367a, b)
( n + 1) n ( n + 1) (
n 2
n − 1) 2
...2 2
( n + 1)! n !

Setting (10.368a) one particular integral of the differential equation (10.364b, c)

is (10.368b):

∞
(  )n x n+ 1
c0 ( 1) = 1:                            y1± ( x ) = ∑
n= 0
n ! ( n + 1)!
, (10.368a, b)

which is (subsection 9.5.9) a regular integral of the first kind.

86 Differential Equations and Applications

Since the indices (10.366c, d) differ by an integer the second particular inte-
gral linearly independent from (10.368b) is a regular integral of the second
kind, second type (9.412) specified by:

∂  
∞

y 0± ( x ) = lim
a→ 0 ∂ a ∑
 a x a + n cn ( a )  , (10.369)
 n= 0 

with the coefficients specified by recurrence (10.370) from (10.366a):

cn ( a ) =
(  )n c 0 ( a ) . (10.370)
( n + a )( n + a − 1)2 ... ( a + 1)2 a

Substituting (10.370) in (10.369) the vanishing factor a → 0 cancels in the numer-

ator and denominator and setting (10.371a) the differentiation in (10.371b):

∂  a   
∞
(  x )n
c0 ( a ) = 1: y 0± ( x ) = lim
a→ 0
x  a +
∂a   ∑  ,
( n + a )( n + a − 1)2 ... ( a + 1)2  
  n= 1

 (10.371a, b)
leads to:

  ∞
(  x )n
y ( x ) = lim  x a 1 + a log x +
±
0 ∑ ( n + a ) ( n + a − 1)2 ... ( a + 1)2
 
a→ 0
n= 1

 1 2 2  
 log x − − ... −  
n+ a n+ a−1 a + 1   
∞
(  x )n  log x − 1 − 2 ... − 2 
= 1+ ∑
n= 1

n ! ( n − 1)!  n n−1


∞
(  x )n log x − ψ 1 + n − ψ n + 2 ψ 1
= 1+ ∑
n= 1

n ! ( n − 1)! 
( ) ( ) ( )

∞
(  x )n ψ 2 + m + ψ 1 + m − 2 ψ 1 ,
= 1 + log x y ±
1 ( x) − ∑
m= 0
 (
m ! ( m + 1)! 
) ( ) ( )
(10.372a–d)

consisting of the sum of three terms: (i) a constant preliminary function

equation equal to unity; (ii) a logarithmic factor multiplying the function of
the first kind (10.368b); (iii) a complementary function involving digamma
functions (9.466a).
Examples 10.1 to 10.20 87

The constant factor −2 ψ ( 1) in the square brackets is a constant multiple of

(10.368b) and can be incorporated in the arbitrary constant c0 in the general
integral (10.373b) of (10.363a, b):

x < ∞:                 y ± ( x ) = C1 y1± ( x ) + C0 y 0± ( x ) . (10.373a, b)

Thus, the general integral (10.373b) of the generalized hyperbolic (circular) differen-
tial equation (10.364b)[(10.364c)] with parameter (10.364a) in (10.363a–c) is (stan-
dard CCLXXXV) a linear combination (10.373b) with arbitrary constants ( C0 , C1 ),
valid in the finite complex x-plane (10.373a) of: (i) a regular integral of the first kind
(10.368b); (ii) a regular integral of the second kind type two (10.372d) consisting
(10.374a) of the sum of: (ii-1) a preliminary function equal to unity; (ii-2) a logarithm
multiplying a function of the first kind:
∞
(  x )n ψ 1 + n + ψ n ,
y 0± ( x ) = log x y1± ( x ) + y 0± ( x ) : y 0± ( x ) = 1 − ∑
n= 1
 (
n ! ( n + 1)! 
) ( )
 (10.374a, b)

(ii-3) a complementary function (9.374b) of order O(1). The case m = − 2 of (10.363b, c)

is the simpler because it leads to exact integrals in finite terms (E10.20.2).

E10.20.2  Double Pole and Elementary Integrals

The generalized hyperbolic (10.36b)[circular (10.363c)] differential equation
with a double pole in the coefficient:

m = − 2:                         x 2 y ′′  y = 0, (10.375a–c)

is linear with homogeneous derivatives (10.375b)[(10.375c)] and has simple

power solutions (10.376a) with exponents (10.376c) that are the roots of the
characteristic polynomial (10.376b):

y ( x) = xa : 0 = a ( a − 1)  1 = P2± ( a ) , 2 a± = 1 ± 1 ± 4 . (10.376a–c)

The upper sign leads to the roots (10.377a) and the general integral (10.377b):

  5   5 
2 a± = 1 ± 5: y ( x ) = x 1/2 C0 cosh  log x  + C1 sinh  log x   ,
  2   2  
 (10.377a, b)

  3   3 
2 a± = 1 ± i 3: y ( x ) = x 1/2 C0 cos  log x  + C1 sin  log x   ,
  2   2  
 (10.378a, b)
88 Differential Equations and Applications

and the lower sign leads to the roots (10.378a) and the general integral (10.378b).
Thus, the generalized hyperbolic (circular) differential equation (10.375b) [(10.375c)]
with a double pole (10.375a) in (10.363b) [(10.363c)] has (standard CCLXXXVI)
general integral (9.377b) [(9.378b)] where (C0 ,C1 ) are arbitrary constants. The cases
of (10.363a–c) remaining are poles of order higher than the second leading to
an irregular singularity at the origin and irregular integrals (E10.20.3).

E10.20.3  Triple Pole and Asymptotic Integrals

The generalized circular (10.363b) [hyperbolic (10.363c)] differential equation
with a pole of order higher than two (10.379a, b) in the coefficient:

p ∈|N 0 ; m = −3 − p: y ′′ ± x −3− p y = 0, (10.379a–d)

has an irregular singularity at the origin (10.379c) [(10.379d)] and hence, the
solution has an essential singularity, which is involving unending descending
powers x − n. This suggests as change of variable the inversion relative to the ori-
gin (9.553a, b) ≡ (10.380a, b) leading (9.555c–e) the differential equation (10.380c):

1
ξ= , w ( ξ ) = y ( x ): 0 = ξ w ′′ + 2 w ′ ± ξ p w. (10.380a–d)
x

In the case of a triple pole (10.381a, b), the differential equation (10.381c, d) has
a regular singular at infinity:

m = −3; p = 0: ξ w ′′ + 2 w ′ ± w = 0, (10.381a–d)

and the solution is specified by regular (10.382a) [irregular (10.382b)] inte-

grals in the inverse (10.380a) [direct x] variable:

∞ ∞

∑n= 0
ξ n+ a en ( a ) = w ( ξ ) = y ( x ) = ∑x
n= 0
− n− a
e n ( a ) , (10.382a, b)

with the index and the coefficients to be determined.

Substitution of (10.382a) in (10.381c, d) leads to (10.383b):

∞ ∞

k = n + 1: 0= ∑( k + a)( k + a + 1) ξ k + a−1 ek ( a) ± ∑ξn+ a en ( a)

k=0 n= 0

∞ (10.383a–c)
= ξa ∑ n= 0
ξ n ( n + a + 1)( n + a + 2 ) en+ 1 ( a ) ± en ( a )  ,
Examples 10.1 to 10.20 89

which is equivalent to (10.383c) with the change of summation variable

(10.383a). The vanishing of the coefficients of all powers in (10.383c) leads to
the recurrence formula (10.384a) for the coefficients:

( n + a ) ( n + a + 1) en ( a ) =  en− 1 ( a ) ; n = 0: a± = 0, −1, (10.384a–d)

setting (10.384b) leads to the índices (10.384c, d). The recurrence formula
(10.384a) may be applied iteratively (10.385b, c) up to the zero order coefficient
to which may be given the value unity (10.385a):

en − 1 ( a ) () n
e1 ( a ) = 1: en ( a ) = 
= .
( n + a )( n + a + 1) ( n + a + 1)( n + a )2 ... ( 2 + a )2 ( 1 + a )
(10.385a–c)

The higher índex (10.384c) together with (10.385a, c) substituted in (10.382b)

lead to:

∞ ∞

∑ n(!( n +) 1)! ,
−n

∑x
x
y ±
0 ( x) = −n
en ( 0 ) = (10.386a, b)
n= 0 n= 0

which is a regular asymptotic integral of the first kind.

Since the indices (10.384c, d) differ by an integer, a linearly independent
solution is (9.412) a regular integral of the second kind, second type:

∂  
∞

w−±1 ( ξ ) = lim
a→−1
( a + 1) ξ a
∂a 

∑ξ
n= 0
n
en ( a ) 

  ∞
(  ξ )m  
= lim
∂
a→ − 1 ∂ a
ξ

a
a + 1 +

∑
m= 1

( m + a + 1)( m + a ) ... ( a + 2 )  
2 2

 ∞
(  ξ )m
= lim ξ a 1 + ( a + 1) log ξ +
a→ − 1

∑ ( m + a + 1)( m + a )2 ... ( a + 2 )2
m= 1
 1 2 2 
 log x − − − ... − 
m+ a+1 m+2 a + 2  
∞
(  ξ )m
= ξ −1 + ξ −1 ∑
m= 1
m ! ( m − 1)!
 1
 log ξ − −
2
m m−1

− ... − 2 

∞
(  ξ )n
=ξ +−1
∑n= 0
 log ξ − ψ ( 2 + n) − ψ ( 1 + n) + 2 ψ ( 1)  ,
n ! ( n + 1)! 
(10.387a–e)
90 Differential Equations and Applications

consisting of a preliminary function, plus the function of the first kind

(10.386b) multiplied by a logarithm plus a complementary function.
Bearing in mind (10.380a, b) it follows that the generalized circular (10.379c)
[hyperbolic (10.379b)] differential equation with a triple pole (10.388a):

m = − 3: x 3 y ′′ ± y = 0, (10.388a–c)

has the general integral (standard CCLXXXVII) valid outside the origin (9.389a),
specified by a linear combination (9.389b) with arbitrary constants ( C0 , C1 ):

x ≠ 0:                       y ( x ) = C0 y 0± ( x ) + C1 y −±1 ( x ) , (10.389a, b)

of asymptotic regular integrals of the first kind (10.386b) and the second kind
(10.387e) ≡ (10.390a):

∞
(  x )− n ψ 2 + n + ψ 1 + n ,
y −1 ( x ) = x − log x y 0 ( x ) − y 2 ( x ) ; y 2 ( x ) =
± ± ± ±
∑n= 0
 (
n ! ( n + 1)! 
) ( )
 (10.390a, b)

the asymptotic regular integral of the second kind (10.390a) is of the second type and
consists of: (i) a preliminary function x; (ii) the integral of the first kind (10.386b)
multiplied by a logarithm; (iii) the complementary function (10.390b). The only
remaining case of solution of the generalized circular (hyperbolic differen-
tial equation (10.363b) [(10.363c)] is the case of a pole of order more than three
that is solvable in terms of the generalized circular (hyperbolic) functions of
the inverse variable (E10.20.4).

E10.20.4  High-Order Poles and Asymptotic Analytic Integrals

The general circular (10.363b) [hyperbolic (10.363c)] differential equation
with a pole of order higher than three (10.391a):

m = − 4, −5,... : y ′′ ± x m y = 0, (10.391a–c)

leads for the inverse variable (10.379b; 10.380a–d) to the differential equation
(10.392a):

0 = ξ w ′′ + 2 w ′ ± ξ − m− 3 w = ( ξ w )′′ ± ξ − m− 4 ξ w , (10.392a, b)

which may be written (10.392b); comparing (10.392b) with (9.313a) [(9.295a)]

it follows that it is a generalized circular (hyperbolic) differential equation
with a non-negative integer parameter (10.393a):

1
− m − 4 = 0, 1, 2 ,... ∈|N 0 :           ξ w (ξ) = y( x), (10.393a, b)
x
Examples 10.1 to 10.20 91

whose solutions are given (10.378b) [(10.377b)] by a linear combination of gen-

eralized circular (10.394) [hyperbolic (10.395)] cosines and sines:

ξ w+ ( ξ ) = B0 cos ( ξ ; − m − 4 ) + B1 sin ( ξ ; − m − 4 ) , (10.394)

ξ w− ( ξ ) = E0 cosh ( ξ ; − m − 4 ) + E1 sinh ( ξ ; − m − 4 ) , (10.395)

with arbitrary constants ( B0 , B1 ) ( E0 , E1 )  . Substituting (10.393b) in (10.394)

[(10.395)] it follows that the general integral of the generalized circular
(10.391b) [hyperbolic (10.391c)] differential equation with a pole of order higher than
three (10.391a) is (standard CCLXXXVIII) a linear combination (10.396)[(10.397)]
with arbitrary constants ( B0 , B1 ) ( E0 , E1 )  of generalized circular (hyperbolic)
cosines and sines of the inverse variable 1/x, with parameter − m − 4 , multiplied by x:

 1  1 
y + ( x ) = x  B0 cos  ; − m − 4 + B1 sin  ; − m − 4  , (10.396)
  x   x 

 1  1 
y − ( x ) = x E0 cosh  ; − m − 4 + E1 sinh  ; − m − 4  . (10.397)
  x   x 

In the case of a pole of order four m = − 4 ( −5 ) these solutions involve (E10.20.5)

the original circular and hyperbolic cosine and sine (Airy functions).

E10.20.5 Circular, Hyperbolic, and Airy Functions

of the Inverse Variable
Setting m = − 4 in (10.396)[(10.397)] it follows that the generalized circular
(10.398a) [hyperbolic (10.399a)] differential equation with a pole of order four has
the (standard CCLXXXIX) general integral specified by (10.398b)[(10.399b)] a lin-
ear combination with arbitrary constants ( B0 , B1 ) ( E0 , E1 )  of circular (hyperbolic)
functions of the inverse variable 1/x multiplied by the variable x:

  1  1 
x 4 y ′′ + y = 0: y ( x ) = x  B0 cos   + B1 sin    , (10.398a, b)
  x   x

  1  1 
x 4 y ′′ − y = 0: y ( x ) = x E0 cosh   + E1 sinh    . (10.399a, b)
  x   x

Setting m = − 5 in (10.396) and using (9.311a–d; 9.312a–d), it follows that the

generalized hyperbolic differential equation with a pole of order five (9.400a) has
(standard CCLXXXIX) general integral specified (10.400b) by a linear combination
92 Differential Equations and Applications

with arbitrary constants ( C0 , C1 ) of the two Airy functions of the inverse variable 1/x
multiplied by the variable x:

  1  1 
x 5 y ′′ + y = 0: y ( x ) = x C0 Ai   + C1 Bi    . (10.400a, b)
  x  x

The general integral of the generalized hyperbolic (9.295a)[circular (9.313a)] differ-

ential equation that is (10.363c)[(10.363b)] with the lower – (upper +) sign has been
obtained for (standard CCXC) all values of the parameter m, namely: (i) for all com-
plex values, excluding negative integers as a linear combination (9.295b) [(9.313b)]
of generalized hyperbolic (9.305; 9.306) ≡ (9.307a–c; 9.308a–c) [circular (9.315;
9.316) ≡ (9.317a–c; 9.318a–c)] cosines and sines including as particular cases the
original functions for zero parameter (9.296a–d) [(9.314a–d)] and also the Airy func-
tions for parameter unity (9.310a–c; 9.311a–d; 9.312a–d); (ii) simple power solutions
with complex exponents (10.377a, b) [(10.378a, b) for double pole (10.375a–c); (iii)
(iv) regular (regular asymptotic) integrals of the first kind (10.368b)[(10.386b)] and
second kind, second type (10.374a, b) [(10.390a, b)] for (10.373a, b) [(10.389a, b)] a
simple (10.364a–c) [triple (10.388a–c)] pole; (v) for poles of order higher than three to
generalized hyperbolic (10.397) [circular (10.396)] functions of the inverse variable,
including the original functions (10.398a, b;10.399a, b) [Airy functions (10.400a, b)]
for poles of order four (five).

Conclusion 10
Two examples of integral curves or streamlines of a flow: (Figure 10.1) all
passing through the origin; (Figure 10.2) not passing through the origin,
except for the six asymptotes. The cusped (smooth) parabola is the envelope
of a one-parameter family of straight lines [Figure 10.3 (10.4)]. A beam, which
is a bar under axial tension, can be subject to 16 possible combinations of
loads and supports, of which four are illustrated (Figures 10.5a–d), using
four types of (i) support: clamped, pinned, clamped-pinned or cantilever;
(ii) loads: concentrated torque or force or uniform or linearly increasing shear
stress. The linear bending of a circular plate with a circular hole allows for
eight combinations (Table 10.1), of which five are distinct (Table 10.2), namely:
(Figure 10.6a/e): clamped/supported at both boundaries; (Figure 10.6b/c)
clamped at the outer boundary and supported/free at the inner boundary;
(Figure 10.6d) supported at the outer boundary and free at the inner bound-
ary. For an elastic membrane without shear stress and under uniform ten-
sion parallel to the supports and with uniform (non-uniform) traction at the
supports [Figure 10.7a (10.8a)], there are modes of all orders n = 1, 2 ,..., for
Examples 10.1 to 10.20 93

example orders 2, 3, 4, respectively, in Figures 10.7b, c, d. The low– (high–)

order modes (Figure 10.7e) are cut-on (cut-off) or propagating (evanescent).
For a given mode or fixed n, a non-uniform traction (Figure 10.8a) causes a
turning point (Figure 10.8b) beyond which propagating modes become eva-
nescent, that is, the cut-on mode becomes the cut-off mode (Figure 10.8c).
The values of the arbitrary constant (Table 10.4) include 16 cases of the general
integral of the linear coupled system of finite difference equations (10.350a,
b). The boundary conditions for the weak bending of an elastic plate depend
(Table 10.3) on the shape, type of support, and material.
Classification 10.1

500 Standards (I to M) of Single and Simultaneous

Systems of Ordinary Differential Equations
Classification 10.1 on the mathematical classes of ordinary differential equa-
tions consists of 500 standards with roman numerals I to M to distinguish
from the arabic numerals 1 to 500 in Classification 10.2 on typical applications
to physics and engineering problems. The 500 standards consist of: (i) 256 dis-
tinct differential equations written explicitly as a look-up table; (ii) 244 related
properties of the solutions that are referred to in the text. Classification 10.1
serves at least two purposes: (a) as a summary of types of differential equa-
tions and methods to solve them, often with one or more examples of their
integrals; (b) in case there is a specific differential equation to be solved, it may
fall into one of the standards, thus indicating a suitable method of solution.
The 500 standards with roman numerals I to M, are divided into 26 sections
with capital letters A to Z, and subsections with a lowercase letter, such as A.a,
A.b, . . . There are six groups. The first group, consisting of the sections A to
H and standards I to XLIX, consists mostly of differential and finite difference
equations having a characteristic polynomial for which the: (i) general inte-
gral can always be determined from the roots of the characteristic polynomial;
(ii) the latter also specifies the complete integral for a variety of forcing func-
tions. The second group, consisting of the sections I to L and standards L to
CXX, includes a variety of linear and non-linear differential equations of the
first and higher orders, whose general and special integrals can be determined
in finite terms. The third group, consisting of the sections M to P and stan-
dards CXXI to CLXXVIII, concerns simultaneous systems of ordinary differ-
ential equations, which can be solved by extensions of the methods that apply
to the first two groups, for example, matrices of characteristic polynomials.
If an ordinary differential equation or a simultaneous system cannot be
solved explicitly by the preceding methods, then a numerical or approxi-
mate solution may be sought. In this case, the theorems of existence and
unicity of solution, ensuring other properties may be useful, form the fifth
group, consisting of the sections R to T and the standards CCII to CCLIV.
The numerical and approximate methods may fail for singular differential
equations, in which case it is critical to know in what form the solution exists.
The linear differential equations with variable coefficients are considered in
the sections U to Z and standards CCII to M, which form the sixth group;
they relate to the special or higher transcendental functions that appear
in many situations. A typical case is the solution of the partial differential
equations of mathematical physics by separation of variables, which appears
as the fourth group, consisting of the section Q and the standards CLXXIX

94
Classification 10.1 95

to CCI. The fifth group relating to the special functions is too extensive to
cover in a single volume; the methods of solution of linear differential equa-
tions with variable coefficients are presented in detail and are illustrated
with generalized Bessel and other special functions. Classification 10.1 with
500 standards on the methods of solution of differential equations is comple-
mented by Classification 10.2, with detailed application to 500 physical and
engineering problems.

A. Linear Differential Equation with Constant

Coefficients − .o.d.e.c. (sections 1.3–1.5):
N

∑
dn y
An = B ( x ) . (10.401)
n= 0
dx n

A. a. Unforced (section 1.3): B ( x ) = 0; roots of the characteristic

polynomial:

0 = PN ( a ) = ∑ A a . (10.402)
n= 0
n
n

*standard I. Single root (subsections 1.3.1–1.3.2);

*standard II. Two distinct real roots (subsection 1.3.3);
*standard III. Pair of complex conjugate roots (subsection 1.3.4);
*standard IV. Multiple root (subsections 1.3.6–1.3.9);
*standard V. Multiple pairs of real roots (subsection 1.3.10);
*standard VI. Multiple pairs of complex conjugate roots (subsection 1.3.10).

A. b.  Forced (section 1.4): Direct Method B ( x ) ≠ 0:

*standard VII. Forcing by an exponential (subsections 1.4.1–1.4.3):

B ( a ) = B e x . (10.403)

*standard VIII. Forcing by hyperbolic cosine or sine (subsection 1.4.4):

B ( x ) = B cosh, sinh ( c x ) ; (10.404)

*standard IX. Forcing by circular cosine or sine (subsection 1.4.5):

B ( x ) = B cos, sin ( a x ) ; (10.405)

96 Differential Equations and Applications

*standard X. Forcing by the product of an exponential and hyperbolic cosine

or sine (subsection 1.4.6):

B ( x ) = B e bx cosh, sinh ( c x ) ; (10.406)

*standard XI. Forcing by the product of an exponential and a circular cosine

or sine (subsection 1.4.7):

B ( x ) = B e bx cos, sin ( a x ) ; (10.407)

*standard XII. Forcing by the product of an exponential by circular and

hyperbolic cosines or sines (subsection 1.4.8):

B ( x ) = B e bx cosh, sinh ( c x ) cos, sin ( a x ) . (10.408)

A. c. Forced (section 1.5) Method of the Inverse

Polynomial of Ordinary Derivatives: B ( x ) ≠ 0
*standard XIII. Forcing by a polynomial (subsections 1.5.1–1.5.7):

B( x) = ∑B
m= 0
m x m ; (10.409)

*standard XIV. Forcing by the product of an exponential by a smooth func-

tion (subsection 1.5.8):

J ( x ) ∈D ∞ (|R ) :                     B ( x ) = e bx J ( x ) ; (10.410a, b)

*standard XV. Forcing by the product of an exponential by a polynomial

(subsections 1.5.9 and 1.5.13):

B ( x ) = e bx ∑B
m= 0
m x m ; (10.411)

*standard XVI. Forcing by the product of a hyperbolic cosine or sine by a

smooth function (subsections 1.5.9 and 1.5.14):

J ( x ) ∈D ∞ (|R ) :   B ( x ) = J ( x ) cosh, sinh ( c x ) ; (10.412a, b)

*standard XVII. Forcing by the product of an exponential by a hyperbolic

cosine or sine and by a smooth function (subsections 1.5.9 and 1.5.16):

J ( x ) ∈D ∞ (|R ) :               B ( x ) = e bx J ( x ) cosh, sinh ( c x ) ; (10.413a, b)

Classification 10.1 97

*standard XVIII. Forcing by the product of a circular cosine or sine by a

smooth function (subsections 1.5.9 and 1.5.15):

J ( x ) ∈D ∞ (|R ) :                  B ( x ) = J ( x ) cos, sin ( a x ) ; (10.414a, b)

*standard XIX. Forcing by the product of an exponential by a circular cosine

or sines by a smooth function (subsections 1.5.9 and 1.5.17):

J ( x ) ∈D ∞ (|R ) :                 B ( x ) = e bx J ( x ) cos, sin ( a x ) , (10.415a, b)

*standard XX. Forcing by the product of an exponential by a circular and a

hyperbolic cosine or sine by a smooth function (subsections 1.5.9 and 1.5.18):

J ( x ) ∈D ∞ (|R ) : B ( x ) = e bx J ( x ) cosh, sinh ( c x ) cos , sin ( a x ) . (10.416a, b)

A. d.  Complete Integrals (subsection 1.5.19; E10.1.1)

B. Linear Ordinary Differential Equation

with Homogeneous Coefficients
.o.d.e.h.c. (section 1.6.1.8)

∑
dn y
An x n = B ( x ) , (10.417)
n= 0
dx n

B. a.  Unforced (section 1.6) B ( x ) = 0; roots of the characteristic

polynomial:

0 = PN ( a ) = ∑A a ( a − 1)...( a − n + 1) , (10.418)
n= 0
n

*standard XXI. Single root (subsections 1.6.2 and 1.6.4);

*standard XXII. Multiple root (subsections 1.6.3–1.6.4);
*standard XXIII. Two distinct real roots (subsection 1.6.6);
*standard XXIV. Pair of complex conjugate roots (subsection 1.6.6).

B. b.  Forced—Direct Method (section 1.7) B ( x ) ≠ 0

*standard XXV. Forcing by a power (subsection 1.7.1):

B ( x ) = x a ; (10.419)
98 Differential Equations and Applications

*standard XXVI. Forcing by a hyperbolic cosine or sine of a logarithm (sub-

section 1.7.2):

B ( x ) = B cosh, sinh ( c log x ) . (10.420)

*standard XXVII. Forcing by a circular cosine or sine of a logarithm (subsec-

tion 1.7.3):

B ( x ) = B cos, sin ( a log x ) ; (10.421)

*standard XXVIII. Forcing by a power multiplied by an hyperbolic cosine or

sine of a logarithm (subsection 1.7.4):

B ( x ) = x b cosh, sinh ( c log x ) . (10.422)

*standard XXIX. Forcing by a power multiplied by a circular cosine or sine of

a logarithm (subsection 1.7.5):

B ( x ) = x b cos, sin ( a log x ) . (10.423)

*standard XXX. Forcing by a power multiplied by hyperbolic and circular

cosines and sines of a logarithm (subsection 1.7.6):

B ( x ) = x b cosh, sinh ( c log x ) cos, sin ( a log x ) . (10.424)

B. c. Forced—Method of the Inverse Characteristic

Polynomial (section 1.8):

B( x) ≠ 0

*standard XXXI. Forcing by a polynomial of logarithms (subsections

1.8.3–1.8.4):

B( x) = ∑B
m= 0
m log m x ;(10.425)

*standard XXXII. Forcing by the product of a power by a smooth function

(subsections 1.8.6 and 1.8.7):

J ( x ) ∈D ∞ (|R ) :                              B ( x ) = x b J ( x ) ; (10.426a, b)

Classification 10.1 99

*standard XXXIII. Forcing by the product of a hyperbolic cosine or sine of a

logarithm by a smooth function (subsections (1.8.6 and 1.8.8):

J ( x ) ∈D ∞ (|R ) :               B ( x ) = cosh, sinh ( c log x ) J ( x ) ; (10.427a, b)

*standard XXXIV. Forcing by the product of a power by the hyperbolic cosine

or sine of a logarithm times a smooth function (subsections 1.8.6 and 1.8.10):

J ( x ) ∈D ∞ (|R ) :             B ( x ) = x b cosh, sinh ( c log x ) J ( x ) ; (10.428a, b)

*standard XXXV. Forcing by the product of a circular cosine or sine of a loga-

rithm by a smooth function (subsections 1.8.6 and 1.8.9):

J ( x ) ∈D ∞ (|R ) :                  B ( x ) = x b cos, sin ( a log x ) J ( x ) ; (10.429a, b)

*standard XXXVI. Forcing by the product of a power by a circular cosine or

sine of a logarithm times a smooth function (subsections 1.8.6 and 1.8.11):

J ( x ) ∈D ∞ (|R ) :                   B ( x ) = x b cos, sin ( a log x ) J ( x ) ; (10.430a, b)

*standard XXXVII. Forcing by the product of a power by circular and hyper-

bolic cosines of sines of a logarithm times a smooth function (subsections
1.8.6 and 1.8.12):

J ( x ) ∈D ∞ (|R ) : B ( x ) = x b cos, sin ( a log x ) cosh, sinh ( c log x ) J ( x ) . (10.431a, b)

B. d.  Complete Integrals: (subsection 1.8.13: E10.1.2)

C. Linear Finite Difference Equation with Constant

Coefficients - .f.d.e.c.c (section 1.9):

∑A y
k=0
n n+ k = Bn . (10.432)

C. a. Unforced (subsections 1.9.3 and 1.9.5–1.9.7); B ( x ) = 0 roots of the

characteristic polynomial (A.2):

*standard XXXVIII. Single root (subsection 1.9.3);

*standard XXXIX. Multiple roots (subsection 1.9.3);
*standard XL. Pair of complex conjugate roots (subsection 1.9.5).
100 Differential Equations and Applications

C. b.  Forced (subsections 1.9.4, 1.9.5, 1.9.8): Bn ≠ 0

*standard XLI. Forcing by a power (subsection 1.9.4, E10.1.3):

Bn = b n ; (10.433)

*standard XLII. Forcing by the product of a power by a circular cosine or sine

(subsection 1.9.5):

Bn = b n cos, sin ( n φ ) . (10.434)

*standard XLIII. Forcing by the product of a power by a hyperbolic cosine or

sine (subsection 1.9.8):

Bn = Bb n cosh, sinh ( n ψ ) ; (10.435)

*standard XLIV. Forced by the product of a power by circular and hyperbolic

cosines and sines (subsection 1.9.8):

Bn = Bb n cos, sin ( nφ ) cosh, sinh ( nψ ) . (10.436)

D.  Method of Variation of Parameters

*standard XLV. (notes 1.2–1.4; section 2.9; subsection 1.3.6; section 3.3) For a
linear differential equation with variable coefficients:

∑
 d   dn y
L     y ( x) ≡ An ( x ) = B ( x ) , (10.437)
  dx   n= 0
dx n

obtains a particular solution of the forced equation (10.437) as a linear combi-

nation (10.438c) of N linearly independent (10.438b) solutions of the unforced
equation (10.438a):

 d  
L     y m ( x ) = 0 , W ( y1 , y 2 ,..., y N ) ≠ 0: y ( x ) =
  dx  
∑
m= 1
Cm ( x ) y m ( x ) ,

(10.438a–c)

with the coefficients not constant, but rather functions of the independent
variable x, to be determined.
Classification 10.1 101

E.  Influence or Green’s Function

*standard XLVI. (notes 1.5–1.8; 2.10–2.11) Is the solution of a linear diferential
equation (10.437) forced by a unit impulse or Dirac function (10.439):

  d
 L    G ( x ; ξ ) = δ ( x − ξ ) , (10.439)
 dx 

and specifies the solution (10.440b) for forcing of (10.437) by a continuous

function (10.440a).

b
B ∈C ( a, b ) : y ( x) =
∫ a
B ( ξ ) G ( x ; ξ ) d ξ . (10.440a, b)

F.  Fourier Series—Discrete Spectrum

*standard XLVII. (notes 1.9–1.10 and 2.4): Forcing of a linear differential equa-
tion with constant coefficients (10.401) by a function of bounded fluctuation
in a finite interval (10.441a) represented by its discrete spectrum (10.441b):

N N
πmx
∑ ∑ B exp  i
dn y
B ∈C ( − L, + L ) :        An = B( x) ≡ m  . (10.441a, b)
n= 0
dx n n= 0
L 

G.  Fourier Integral/Transform—Continuous Spectrum

*standard XLVIII. (notes 1.11–1.14; 2.5–2.9 and 2.12–2.13): Forcing a linear dif-
ferential equation with constant coefficients (10.401) by a function absolutely
integrable (10.442a) and of bounded fluctuation (10.442b) on the real line rep-
resented by its continuous spectrum (10.442c):

∑A
dn y +∞
B ∈L1 ∩ F (−∞ , + ∞):       
n= 0
n
dx n
= B( x) ≡
∫ −∞
B ( k ) e ikx dk. (10.442a–c)
102 Differential Equations and Applications

H.  Laplace Transform—Initial Conditions

*standard XLIX. (notes 1.15–1.25): Forcing of a linear differential equation
with constant coefficients (10.401) by a function absolutely integrable on the
finite positive real line (10.443a) and of slow growth at infinity (10.443b) rep-
resented by its Laplace transform (10.443c):
N

∑
dn y ∞
B ∈L1 (0, L) ∩ V (L , + ∞):
n= 0
An
dx n
= B( x) =
∫
0
B ( s ) e − sx ds. (10.443a–c)

I. First-Order Differential Equations (sections 3.1–3.7;

examples 10.4–10.5):

dy
y′ ≡ :                       F ( x , y ; y ′ ) = 0 (10.444a, b)
dx

I. a.  Solvable by Quadratures (sections 3.2–3.4; examples 10.4–10.5):

*standard L. separable (subsection 3.2.1 and example 10.4 and E10.9.1):

dy X ( x )
X , Y ∈E (|R ) :                         = ; (10.445a, b)
dx Y ( y )

*standard LI. linear unforced (subsection 3.2.2):

P ∈E (|R ) :                y ′ = P ( x ) y ; (10.446a, b)

*standard LII. Linear forced (section 3.3; example 10.4):

P, Q ∈E (|R ) :                y ′ = P ( x ) y + Q ( x ) ; (10.447a, b)

*standard LIII. Bernoulli (1695) non-linear differential equation (section 3.4;

example 10.4):

P, Q ∈E (|R ) :                y ′ = P ( x ) y + Q ( x ) y n . (10.448a, b)

*standard LIV. Particular Ricatti equation (section 3.4):

P, Q ∈E (|R ) :                    y ′ = P ( x ) y + Q ( x ) y 2 . (10.449a, b)

Classification 10.1 103

I. b.  Ricatti (1824) Equation (sections 3.5–3.6; example 10.4):

y ′ = P ( x ) y 2 + P ( x ) y + Q ( x ) . (10.450)

*standard LV. General integral involving two quadratures if one particular

integral is known (subsection 3.5.2 and example 10.4);
*standard LVI, General integral involving one quadrature if two particular
integrals are known (subsection 3.5.3);
*standard LVII. General integral involving no quadratures if three particular
integrals are known (subsections 3.5.4–3.5.5).
*standard LVIII. Transformation to a linear second-order differential equa-
tion (subsections 3.6.1–3.6.2):

u′′ + A   u′ + B  u = 0. (10.451)

*standard LIX. Reducible to a linear second-order differential equation with

constant coefficients and distinct roots a, b of the characteristic polynomial
(subsection 3.6.3):

ab  R′ ( x ) 
y′ = + A + B −  y + R ( x ) y ; (10.452)
2

R ( x)  R ( x) 

*standard LX. As in the standard LIX with a = b double root (subsection

3.6.3):

a2  R′ ( x ) 
y′ = + 2 a −  y + R ( x ) y ; (10.453)
2

R ( x)  R ( x) 

*standard LXI. Reducible to a linear second-order differential equation with

homogeneous coefficients and distinct roots a, b of the characteristic polyno-
mial (subsection 3.6.4):

ab  a + b − 1 R′ ( x ) 
y′ = + −  y + R ( x ) y ; (10.454)
2

x R ( x) 
2
x R ( x) 

*standard LXII. As in the standard LXI with a double a = b root (subsection

3.6.4):

a2  2 a R′ ( x ) 
y′ = + −  y + R ( x ) y . (10.455)
2

x R ( x)  x
2
R ( x) 
104 Differential Equations and Applications

*standard LXIII. Particular Ricatti equation (subsection 3.6.5):

λ
a, b ∈|R:                           y ′ = a x2λ + y + b y 2 . (10.456a–c)
x

I. c.  Homogeneous Differential Equation (section 3.7):

*standard LXIV. Non-linear (subsections 3.7.1–3.7.3; section 4.1; examples 10.4
and 10.7):

 y
y ∈B (|R ) :                                     y ′ = f   ; (10.457a, b)
 x

*standard LXV. Linear (subsections 3.7.1–3.7.2):

y
y′ = a + b . (10.458)
x

J. Differentials (sections 3.8–3.9; notes 3.1–3.24)

J. a. First-Order Differential in Two Variables
(section 3.8; examples 10.5, 10.6):

  
X = {X , Y } : 0 = X ( x , y ) dx + Y ( x , y ) dy = X . dx . (10.459a–c)

*standard LXVI. Exact differential (10.460b) if the integrability condition

(10.460a) is met (subsections 3.8.1–3.8.3; E10.6.1):

∂Y ∂X
Ω ( x, y ) ≡ − = 0: X ( x , y ) dx + Y ( x , y ) dy = dΦ ( x , y ) ; (10.460a, b)
∂x ∂y

*standard LXVII. Inexact differential (10.462a) for which an integrating factor

(10.461b) always exists (subsections 3.8.4–3.8.6; E10.6.2–E10.6.3):

Ω ( x , y ) ≠ 0: X ( x , y ) dx + Y ( x , y ) dy = λ ( x , y ) dΦ ( x , y ) ; (10.461a, b)

*standard LXVIII. Integrating factor (10.462b) depending only (10.462a) on

the independent variable (subsection 3.8.7):

Ω ( x, y )  x

Y ( x, y )
≡ f ( x ): X ( x , y ) dx + Y ( x , y ) dy = exp 
 ∫ f ( ξ ) d ξ  dΦ ( x , y ) ;

 (10.462a, b)
Classification 10.1 105

*standard LXIX. Integrating factor (10.463b) depending only (10.463a) on the

dependent variable (subsection 3.8.7; E10.6.2–E10.6.3):

Ω ( x, y )  y

X ( x, y )
≡ g ( y ): X ( x , y ) dx + Y ( x , y ) dy = exp 
 ∫ g ( η) d η dΦ ( x , y ) ;

 (10.463a, b)

*standard LXX. Integrating factor (10.464b) depending only (10.464a) on the

sum of the variables (subsection 3.8.8; E10.6.2–E10.6.3):

Ω ( x, y )
≡ h ( x + y ): X ( x , y ) dx + Y ( x , y ) dy
X ( x, y ) − Y ( x, y )
 (10.464a, b)
 x+ y

= exp 
 ∫ h ( z ) dz  dΦ ( x , y ) ;


*standard LXXI. Integrating factor (10.465b) depending only (10.465a) on the

product of the variables (subsection 3.8.9; E10.6.2–E10.6.3):

Ω ( x, y )
≡ j ( xy ) : X ( x , y ) dx + Y ( x , y ) dy
x X ( x, y ) − y Y ( x, y )
 (10.465a, b)
 x+ y

= exp 
 ∫ j ( u) du  dΦ ( x , y ) ;


*standard LXXII. Integrating factor (10.466b) a homogeneous function

(10.466a) depending only on the ratio of the two variables (subsection 3.8.10;
E10.6.2–E10.6.3);

x2 Ω ( x, y )  y
≡ k : X ( x , y ) dx + Y ( x , y ) dy
x X ( x, y ) − y Y ( x, y )  x
 (10.466a, b)
 x+ y

= exp 
 ∫ k ( v ) dv  dΦ ( x , y ) ;


*standard LXXIII. Other forms of integrating factor (10.467a) of (10.467b) an

inexact differential (subsection 3.8.11):

λ = x a − 1 y b − 1 :   a y d x + b x d y + x α− a y β− b ( α y d x + β x dy ) =
1
λ
(
d x a y b + xα y β .)
 (10.467a, b)
106 Differential Equations and Applications

J. b.  First-Order Differential in Three Variables (section 3.9):

  
X = {X , Y , Z} : 0 = X ( x , y , z ) dx + Y ( x , y , z ) dy + Z ( x , y , z ) dz = X . dx .
 (10.468a, b)

*standard LXXIV. Exact differential (10.469b) if the vector of coefficients is

(10.469a) irrotational (subsection 3.9.1):
    
Ω≡∇∧X =0 ⇔ X . dx = dΦ ( x ) ; (10.469a, b)

*standard LXXV. Inexact differential with integrating factor (10.470b) if

(10.470a) the vector of coefficients has zero helicity (subsections 3.9.2–3.9.5):
     
H ≡ X. Ω = 0 ⇔ X . dx = λ ( x ) dΦ ( x ) ; (10.470a, b)

*standard LXXVI. Pfaffian (10.471b) if (10.471a) the vector of coefficients has

non-zero helicity (subsections 3.9.5–3.9.7):
   
H≠0 ⇔ X . dx ≠ λ ( x ) dΦ ( x ) . (10.471a, b)

J. c.  Three-Dimensional Potentials (subsections 3.9.8–3.9.11):

*standard LXXVII. A continuously differentiable vector field (10.472a) is the
gradient of a scalar potential (A10.472c) if (10.472b) is irrotational (subsec-
tions 3.9.8–3.9.10):
  
( )
X ∈C 1 |R 2 :                     ∇ ∧ X = 0 ⇔ X = ∇Φ , (10.472a–c)

*standard LXXVIII. A continuously differentiable vector field (10.473a) is

the curl of a vector potential (10.473c) if (10.473b) is solenoidal (subsection
3.9.9):
   
( )
X ∈C 1 |R 3 :                ∇. X = 0 ⇔ X = ∇ ∧ A ; (10.473a–c)

*standard LXXIX. A continuously differentiable vector field (10.474a) is the

sum (10.474b) of an irrotational (10.472b) [solenoidal (10.473b)] part that is the
gradient (10.472c) [curl (10.473c)] of a scalar (vector) potential (subsections
3.9.8–3.9.9):
  
( )
X ∈C 1 |R 3 :                            X = ∇Φ + ∇ ∧ A ; (10.474a, b)
Classification 10.1 107

*standard LXXX. A continuously differentiable vector field (10.475a) is the

sum (10.475b) of the gradient of a scalar potential with the outer product of
the gradient of other two Clebsch potentials (subsections 3.9.8–3.9.9):
 
( )
X ∈C 1 |R 3 :                 X = ∇Φ + ∇ Ξ ∧ ∇Ψ . (10.475a, b)

*standard LXXXI. A continuously differentiable vector field (10.476a) is the

sum (10.476b) of the gradients of two scalar potentials with the second multi-
plied by a third Euler potential (subsections 3.9.8–3.9.9):
 
( )
X ∈C 1 |R 3 :                 X = ∇Φ + Θ∇Ψ . (10.476a, b)

J. d.  First-Order Differential in N ≥ 4 Variables (notes 3.2–3.20):

N

n, m = 1,..., N :               0= ∑ X (x
n= 0
n m ) dxn . (10.477a, b)

*standard LXXXII. Exact differential (10.478b) if (10.478a) the curl bi-vector of

coefficients is zero (notes 3.3 and 3.9):
N

Ωmn =
∂X n ∂X n
−
∂ xm ∂ xm
=0 ⇔ ∑ X (x
n= 1
n m ) dxn = dΦ ( xm ) ; (10.478a, b)

*standard LXXXIII. Inexact differential (10.479c) with integrating factor if

the bi-vector (tri-vector) curl (10.479a) [helicity (10.479b)] is (is not) zero (notes
3.4, 3.5, and 3.10–3.12):
N

Ωmn = 0 ≠ H m , ns = Ωmn X s + Ωns X m + Ωns X m : ∑X n= 1

m ( um ) dxn = λ ( xm ) dΦ ( xm ) ;
(10.479a–c)

*standard LXXXIV. Pfaffian (10.480b) if the tri-vector helicity (10.480a) is non-

zero (notes 3.6–3.10):
N

H mns ≠ 0 ⇔ ∑ X (x
n= 0
n m ) dxn ≠ λ ( xm ) dΦ ( xm ) . (10.480a, b)

*standard LXXXV. Reduction (10.481b) of the number of variables (10.481a) of

a first-order differential (notes 3.11–3.15):
N N −1

0= ∑
n= 0
X n ( xm ) dxn → 0= ∑ Y (y
m= 1
m m ) dyn . (10.481a, b)
108 Differential Equations and Applications

J. e. Homogeneous First-Order Differential in Two Variables: First-

order differential in two variables (10.459c) with homogeneous
coefficients (10.482a, b) of the same degree q (notes 3.17–3.20):

 y  y
X ( x , y ) = x q X  1,  , Y ( x , y ) = x q Y  1,  ; (10.482a, b)
 x  x

*standard LXXXVI. Integrating factor (10.483b) for inexact differential

(10.483a):

∂Y ∂X X ( x , y ) dx + Y ( x , y ) dy
≠ :                  dΦ ( x , y ) = ; (10.483a, b)
∂x ∂y x X ( x, y ) + y Y ( x, y )

*standard LXXXVII. General integral (10.484c) for exact differential (10.484b)

of degree distinct from minus unity (10.484a):

∂X ∂Y
q ≠ −1; ≡ :           x X ( x , y ) + y Y ( x , y ) = C. (10.484a–c)
∂y ∂x

J. f. Homogeneous First-Order Differential in N Variables (note 3.18):

First-order differential in N variables (10.477a, b) whose coefficients
(10.485a) are homogeneous functions (10.485b) of the same degree q:

 x x 
X n ( x1 ,...., xN ) = ( x1 ) X  1, 2 ,..., N  ;
q
m, n = 1,..., N : (10.485a, b)
 x1 x1 

*standard LXXXVIII. Inexact differential (10.486a) with integrating factor

(10.486b):

N
 N 
∂X n ∂X n
≠
∂ xm ∂ xm
: ∑n= 1
X n ( xm ) dxn = 
 n= 1
∑
X n ( xm ) xn  dΦ ( xm ) ;

(10.486a, b)

*standard LXXXIX. General integral (10.487c) for exact differential (10.487b)

of degree distinct from minus unity (10.487a):
N

q ≠ − 1;
∂X n ∂X n
=
∂ xm ∂ xm
:               ∑ X (x
n= 1
n m ) xn = C. (10.487a–c)

J. g.  Differential of Order M in N Variables (notes 3.21–3.22):

N

0= ∑
i1 ,.., iM = 1
X i1 ...i dx i ... dx iM . (10.488)
M
Classification 10.1 109

*standard XC. Second-order differential in two variables (note 3.21):

0 = A ( x , y ) ( dx ) + B ( x , y )( dy ) + D ( x , y ) dx dy . (10.489)
2 2

*standard XCI. Second-order differential in three variables:

0 = A ( dx ) + B ( dy ) + C ( dz ) + 2 D dx dy + 2E dx dz + 2 F dy dz , (10.490a)
2 2 2

which is factorizable (notes 3.23–3.24) if:

A BC + 2 DE F = A F 2 + BE 2 + C D2 . (10.490b)

K. Special Integrals. Integrals not included in the

general integral (sections 5.1–5.4; example 10.8):
K. a.  C, p-discriminants and Special Points of First-Order
Differential Equations (section 5.1):
*standard XCII. Envelope as a special integral (subsection 5.1.1);
*standard XCIII. C–discriminant: envelope, node, and cusp loci (subsec-
tion 5.1.2);
*standard XCIV. p–discriminant: envelope, tac, and cust loci (subsection
5.1.3);
*standard XCV. Classification of special curves (subsection 5.1.4);
*standard XCVI. Extension to special integral of differential equations of
order higher than the first (subsection 5.1.5).

K. b. Quadratic Discriminants of First-Order Differential Equations

(section 5.2):
*standard XCVII. First-order differential equation quadratic in the slope:
2
 dy  dy
0 = L ( x, y )   + M ( x, y ) + N ( x , y ) ; (10.491)
 dx  dx

*standard XCVIII. General integral of first-order differential equation qua-

dratic in the arbitrary constant of integration C:

0 = P ( x , y ) C 2 + Q ( x , y ) C + R ( x , y ) . (10.492)
110 Differential Equations and Applications

K. c. First-Order Differential Equations that Always

Have Special Integrals (section 5.3):
*standard XCIX. Clairaut differential equation (10.493b) involving (10.493a) a
differentiable function (subsections 5.3.1–5.3.2; E10.8.1)

dy  dy 
h ∈D ( R ) :                    y = x + h   ; (10.493a, b)
dx  dx 
*standard C. D’Alembert differential equation (10.494c) involving (10.494a, b)
two differentiable functions (subsections 5.3.3–5.3.4; E10.8.1):

 dy   dy 
g , h ∈D ( R ) :                       y = xg   + h   . (10.494a–c)
 dx   dx 

K. d. First-Order Differential Equations Solvable for

the Slope or a Variable (section 5.4):
*standard CI. First-order differential equation solvable for the dependent
variable (10.495b) and involving (10.495a) a differentiable function (subsec-
tions 5.4.1–5.4.2):

 dy 
( )
h ∈D |R 2 :                            y = h  x,  ;
 dx 
(10.495a, b)

*standard CII. First-order differential equation solvable for the independent

variable (10.496b) and involving (10.496a) a differentiable function (subsec-
tions 5.4.1 and 5.4.3; E10.8.3):

 dy 
( )
g ∈D |R 2 :                          x = g  y ,  .
 dx 
(10.496a, b)

*standard CIII. First-order differential equation of degree M in the slope

(subsection 5.4.4):

M m
0= ∑
m= 0
 dy 
Am   . (10.497)
 dx 
Classification 10.1 111

L. Transformation and Reduction of

Differential Equations (sections 5.5–5.9)
L. a.  Transformation of Differentials (section 5.5.1):
N

dΦ = ∑X
m= 1
n dxn . (10.498)

*standard CIV. Exchange of a variable and a coefficient (10.499b) via the

Legendre transformation (10.499a):

Ψ ( X 1 , x2 ,...., xN ) = Φ ( x1 , x2 ,...., xN ) − x1 , X 1 : dΨ = − x1 d X 1 + ∑X
m= 2
m dxm .

 (10.499a, b)

L. b.  Dual First-Order Differential Equations (subsections 5.5.2–5.5.3):

*standard CV, General case:

 dy   dY dY 
0 = F  x; y,  = F  ;X − Y , X  ; (10.500a, b)
 dx   dX dX 
*standard CVI. Special case:

 dy   dy  dY
0 = Φ x − y − x Ψ   = Φ( Υ ) − Ψ (X ) . (10.501a, b)
 dx   dx  dX

L. c. Second-Order Differential Equation Reduced

to First-Order (section 5.6; E10.9.1–E10.9.2):

dy d2 y
y′ ≡ ,  y ′′ ≡ 2 :                 0 = F ( x ; y , y ′ , y ′′ ) . (10.502a–c)
dx dx

*standard CVII. General case (subsection 5.6.1):

 dy ′ 
0 = F  x ; y , y ′, y ′ . (10.503)
 dy 

*standard CVIII. Omitting the independent variable (subsection 5.6.2; E10.9.2):

 dy ′ 
0 = F ( y , y ′ , y ′′ ) = F  y , y ′ , y ′ . (10.504)
 dy 
112 Differential Equations and Applications

*standard CIX. Omitting both the independent variable and slope (subsec-
tion 5.6.3):

0 = F ( y , y ′′ ) ; (10.505)

*standard CX. Dependent variable missing (subsection 5.6.4):

z ≡ y ′:                           0 = F ( x , y ′ , y ′′ ) = F ( x ; z, z ′ ) ; (10.506a, b)

*standard CXI. Missing both the slope and the dependent variable (subsec-
tion 5.6.4; E10.9.1):

0 = F ( x ; y ′′ ) . (10.507)

L. d. Depression of the Order of Differential Equations

of Order N > 2, Higher than 2 (section 5.7; E10.9.3):

dN y
y( N ) ( x ) ≡
dx N
(
:                         0 = F x ; y ; y ′ , y ′′ ,....y( N ) . ) (10.508a, b)

*standard CXII. Involving only the independent variable (subsection 5.7.1):

( )
0 = F x , y( N ) ; (10.509)

*standard CXIII. Involving only the independent variable and derivatives of

order N and N – 1 (subsection 5.7.2; E10.9.3):

z = y( N − 1) ( x ) :         ( )
0 = F x , y( N − 1) , y( N ) = F ( x ; z, z ′ ) ; (10.510a–c)

*standard CXIV. Involving only the derivatives of orders N – 1 and N – 2

(subsection 5.7.3; E10.9.3):

w = y( N − 1) ( x ) :         ( )
0 = F y( N − 1) , y( N ) = F ( w , w ′ ) . (10.511a–c)

*standard CXV. Involving only derivatives of orders N – 2 and N (subsec-

tion 5.7.4):

u = y( N − 2 ) ( x ) :         ( )
0 = F y( N − 2 ) , y( N ) = F ( u, u′′ ) ; (10.512a–c)

*standard CXVI. Lowest-order derivative appearing is N – P (subsection 5.7.5;

E10.9.3):

z ≡ y( N − P) ( x ) :         ( ) ( )
0 = F x , y( N − P) ,..., y( N ) = F x ; z,..., z( P) . (10.513a–c)
Classification 10.1 113

L. e.  Combinations of Differential Operators (section 5.8; E10.9.4–10.9.6):

*standard CXVII. Factorizable linear differential equation (subsection 5.8.1;
E10.9.4):

  d    d 
 PN    QM    y ( x ) = 0 ; (10.514)
 dx  dx 

*standard CXVIII. Exact differential equation (subsection 5.8.2 and E10.9.5):

(
0 = F x , y , y ′ ,..., y( N ) = ) dp
dx P {(
G x ; y , y ′ ,..., y ( )
N−p
) }. (10.515)
L. f.  Higher-Order Homogeneous Differential Equations (section 5.9):
*standard CXIX. Homogeneous differential equation of the first kind order
N (subsection 5.9.1; E10.9.6):

y   y ′ y ′′ y( N ) 
0 = F  , y ′ , y ′′ ,..., y( N )  = G  x , , ,..., ; (10.516a, b)
x   x y y 

*standard CXX. Homogeneous differential equation of the second kind and

order N (subsections 5.9.2–5.9.3; E10.9.6):

( )
0 = F y , y ′x , y ′′x 2 ,..., y( N ) x N . (10.517)

M. Simultaneous System of Differential

Equations (sections 7.1–7.3):
*standard CXXI. Generalized autonomous system of differential equations
(subsection 7.1.1):

dy m
m = 1,..., M:                         = Ym ( x ; y1 ,...., y M ) ; (10.518a, b)
dx

*standard CXXII. General simultaneous system of M differential equations

(subsection 7.1.2):

m, s = 1,..., M ; N m , s ∈|N : ( )
0 = Fs x ; y m , y m′ , y m′′ ,..., y m( m , s ) ; (10.519a–c)
N
114 Differential Equations and Applications

*standard CXXIII. Implicit autonomous system of differential equations

(subsections 7.2.1–7.2.4):

dxn
m, s = 1,..., N :            = X n ( xm ) ; (10.520a, b)
dt

*standard CXXIV. Family of curves tangent to a vector field (subsections

7.2.1–7.2.4);
*standard CXXV. Hypersurfaces orthogonal to a vector field (subsection
7.2.5);
*standard CXXVI. Order of a simultaneous systems of differential equations
(subsections 7.3.1–7.3.2);
*standard CXXVII. First integral of a system of simultaneous differential
equations (subsection 7.3.4).

N. Simultaneous Linear System of Differential Equations

with Constant Coefficients (sections 7.4–7.5):
*standard CXXVIII. Simultaneous linear system of differential equations
(subsection 7.4.1):

M Nm , r

∑∑
dk y
m, s = 1,..., M ; N m , r ∉|N : Am , r , k ( x ) = Bm ( x ) , (10.521a–c)
m, r =1 k = 0
dx k

*standard CXXIX. Simultaneous linear system of differential equations with

constant coefficients (subsections 7.4.1–7.4.2):

M Nm , r

∑∑
dk y
m = 1,..., M ; N m , r ∉|N : Am , r , k = Bm ( x ) ; (10.522a–c)
m, r =1 k = 0
dx k

*standard CXXX. Simultaneous linear system of differential equations with

homogeneous derivatives (subsection 7.4.1):

M Nm , r

∑∑
dk y
m = 1,..., M ; N m , r ∉|N : Am , r , k x k = Bm ( x ) . (10.523a–c)
m, r =1 k = 0
dx k
Classification 10.1 115

N. a. Unforced (section 7.4): (10.522a–c) with Bm ( x ) = 0 and

characteristic polynomial (subsections 7.4.2–7.4.3):

Nm , r

PN ( a ) = Det { Pm , r ( a )} , Pm , r ( a ) ≡ ∑A
k=0
m, r , k a k . (10.524a, b)

*standard CXXXI. Degenerate and non-degenerate systems (subsections

7.4.3–7.4.4);
*standard CXXXII. Natural integrals for a single root of the characteristic
polynomial (subsection 7.4.5);
*standard CXXXIII. Natural integrals for a multiple root of the characteristic
polynomial (subsection 7.4.6);
*standard CXXXIV. General integral as a linear combination of natural inte-
grals (subsections 7.4.7–7.4.10);
*standard CXXXV. Boundary conditions determining the arbitrary constants
of integration (subsections 7.4.7–7.4.10);
*standard CXXXVI. Compatibility conditions determining the coefficients in
the general integral (subsections 7.4.7–7.4.10);
*standard CXXXVII. Diagonal system for natural integrals corresponding to
single roots of the characteristic polynomial (subsections 7.4.11 and 7.4.13);
*standard CXXXVIII. Two-banded diagonal system for the natural integrals
corresponding to the same multiple root of the characteristic polynomial and
block-banded diagonal system corresponding to distinct multiple roots of
the characteristic polynomial (subsections 7.4.11–7.4.14).

N. b.  Forced (10.522a–c) with Bm ( x ) ≠ 0 (section 7.5):

*standard CXXXIX. Forcing by an exponential in non-resonant, resonant,
and multiply-resonant cases (subsections 7.5.1–7.5.3):

Pm ( x ) = Gm e bx ; (10.525)

*standard CXL. Forcing by the product of an exponential by a hyperbolic

cosine or sine (subsection 7.5.4):

Bm ( x ) = Gm e bx cosh, sinh ( c x ) ; (10.526a, b)

*standard CXLI. Forcing by the product of an exponential by a circular cosine

or sine (subsections 7.5.4–7.5.5):

Bm ( x ) = Gm e bx cos, sin ( g x ) ; (10.527a, b)

116 Differential Equations and Applications

*standard CXLII. Forcing by the product of an exponential by a hyperbolic

and a circular cosine or sine (subsection 7.5.4):

Bm ( x ) = Gm e bx cos, sin h ( cx ) cos, sin ( g x ) . (10.528a–d)

*standard CXLIII. Method of the inverse matrix of polynomials of ordinary

derivatives (subsections 7.5.7–7.5.8);
*standard CXLIV. Complete integral of a forced coupled linear system of dif-
ferential equations with constant coeffficients (subsection 7.5.9).

O. Simultaneous Linear System of Differential Equations

with Homogeneous Derivatives (10.523a–c); (sections 7.6–7.7):
*standard CXLV. Alternative equivalent summation and matrix forms (sub-
section 7.6.2).

O. a.  Unforced System: (10.523a–c) with Bm ( x ) = 0 (section 7.6):

*standard CXLVI. Characteristic polynomial (subsection 7.6.3):

Nm , r

QN ( a ) = Det (Qm , r ) , Qm , r ( a ) ≡ ∑A
k=0
m, r , k a ( a − 1) ... ( a − k + 1) . (10.529a, b)

*standard CXLVII. Natural integrals corresponding to a single root of the

characteristic polynomial (subsection 7.6.4);
*standard CXLVIII. Natural integrals corresponding to multiple roots of the
characteristic polynomial (subsection 7.6.4);
*standard CXLIX. General integral for characteristic polynomial with single
roots only (subsection 7.6.5);
*standard CL. General integral for characteristic polynomial with distinct
multiple roots (subsection 7.6.5);
*standard CLI. Arbitrary constants and boundary conditions (subsections
7.6.5 and 7.6.7);
*standard CLII. Coefficients of the dependent variables and compatibility
conditions (subsections 7.6.5–7.6.6);
*standard CLIII. Diagonal system for natural integrals corresponding to sin-
gle roots of the characteristic polynomial (subsection 7.6.8);
*standard CLIV. Two-banded diagonal system for natural integrals corre-
sponding to multiple roots of the characteristic polynomial (subsection 7.6.8);
*standard CLV. Block-banded diagonal system for several multiple roots of
the characteristic polynomial (subsections 7.6.8–7.6.9).
Classification 10.1 117

O. b.  Forced System (10.523a–c) with Bm ( x ) ≠ 0 (section 7.7):

*standard CLVI. Comparison of simultaneous systems of linear differential
equations with ( α ) constant coefficients; (β ) homogeneous derivatives (sub-
sections 7.7.2–7.7.5);
*standard CLVII. Forcing by a power in non-resonant, resonant, and multi-
resonant cases (subsections 7.7.2–7.7.5):

Bm ( x ) = Gm x b ; (10.530)

*standard CLVIII. Forcing by the product of a power by the hyperbolic cosine

or sine of the multiple of a logarithm (subsection 7.7.6):

Bm ( x ) = Gm x b cosh , sinh ( c log x ) ; (10.531a, b)

*standard CLIX. Forcing by the product of a power by a circular cosine or

sine of the multiple of a logarithm (subsection 7.7.6):

Bm ( x ) = Gm x b cos , sin ( g log x ) ; (10.532a, b)

*standard CLX. Forcing by the product of a power by a hyperbolic and a cir-

cular cosine or sine of a logarithm (subsections 7.7.6–7.7.7):

Bm ( x ) = Gm x b cosh , sinh ( c log x ) cos, sin ( g log x ) ; (10.533a–d)

*standard CLXI. Inverse matrix of polynomials of homogeneous derivatives

(subsections 7.7.8–7.7.9);
*standard CLXII. Complete integral of a forced linear system of differential
equations wih homogeneous derivatives (subsection 7.7.10).

P. Simultaneous System of Linear Finite Difference

Equations with Constant Coefficients (sections 7.8–7.9):

M Nm , r

m = 1,..., M ; N m , r ∈|N : Bm ,  = ∑∑ A
m, =1 k = 0
m, r , k y r , k +  ; (10.534a–c)

*standard CLXIII. The identity (10.534c) ≡ (10.535b) follows using (10.535a)

forward finite diferences (subsections 7.8.1–7.8.2):

M Nm , r

∆ k   yr ,  = yr , + k : Bm ,  = ∑∑ A
m, =1 k = 0
m, r , k ∆ k y r , k +  ; (10.535a, b)
118 Differential Equations and Applications

P. a. Unforced (section 7.8): (10.534c) ≡ (10.535b) with Bm ,  = 0 and

characteristic polynomial:

Nm , r

RN ( ∆ ) = Det {Rm , r ( a )} , Rm , r = ∑A
k=0
m, r , k ∆ k . (10.536a, b)

*standard CLXIV. The roots of the characteristic polynomial specify the nat-
ural sequences (subsection 7.8.3);
*standard CLXV. Natural sequences corresponding to a single root of the
characteristic polynomial (subsection 7.8.4);
*standard CLXVI. Natural sequences corresponding to a multiple root of the
characteristic polynomial (subsection 7.8.4);
*standard CLXVII. General solution for characteristic polynomial with all
roots distinct (subsection 7.8.5);
*standard CLXVIII. General solution for characteristic polynomial with dis-
tinct multiple roots (subsection 7.8.5);
*standard CLXIX. Arbitrary constants determined from starting conditions
(subsections 7.8.5 and 7.8.7);
*standard CLXX. Coefficients of natural sequences determined from com-
patibility conditions (subsections 7.8.5–7.8.6);
*standard CLXXI. Diagonal system of natural sequences corresponding to
single roots of the characteristic polynomial (subsections 7.8.8 and 7.8.10);
*standard CLXXII. Lower triangular system of natural sequences corre-
sponding to multiple roots of the characteristic polynomial (subsection 7.8.8);
*standard CLXXIII. Block-lower triangular system of natural integrals corre-
sponding to distinct multiple roots of the characteristic polynomial (subsec-
tion 7.8.9).

P. b.  Forced (subsection 7.9): Bm ,  ≠ 0 in (10.535b):

*standard CLXXIV. Forcing by powers (subsections 7.9.2–7.9.4):

Bm ,  = Gm b  ; (10.537)

*standard CLXXV. Forcing by the product of a power by a hyperbolic cosine

or sine of the multiple of an angle (subsections 7.9.5–7.9.6):

Bm ,  = Gm b  cosh, sinh ( c  φ ) ; (10.538a, b)

*standard CLXXVI. Forcing by the product of a power by a circular cosine or

sine of a multiple angle (subsection 7.9.5):

Bm ,  = Gm b  cos, sin ( g  φ ) ; (10.539a, b)

Classification 10.1 119

*standard CLXXVII. Forcing by the product of a power by a hyperbolic and a

circular cosine or sine of a multiple angle (subsection 7.9.5):

Bm ,  = Gm b  cos h, sin h ( c  φ ) cos, sin ( g  φ ) ; (10.540a–d)

*standard CLXXVIII. Complete solution of a forced linear system of finite

difference equations with constant coefficients (subsections 7.9.7–7.9.8).

Q. Generalized Isotropic Equation of a Mathematical

Physics (notes 8.1–8.17):

N M

∑n= 0
∂2 F
∂ xn2
+ ∑A
m= 0
m
∂2 F
∂t m
= 0. (10.541)

Q. a.  Reduction to a Helmholtz equation (notes 8.1–8.5):

∇ ≡ 2
∑
n= 1
∂2 F
∂ xn2
: ∇ 2 F + k 2 F = 0. (10.542a, b)

*standard CLXXIX. Reduction of the generalized equation of mathematical

physics (10.541) to a Helmholtz equation (10.542a, b) via (10.543b) a Fourier
(10.442c )≡ (10.543a) transform (notes 8.1–8.2):

M +∞

k ≡ 2
∑A
m= 0
m (−i ω ) m
: F ( xn , t ) =
∫ F ( x , t ) e
n
− iωt
dω . (10.543a, b)
−∞

*standard CLXXX. Solution of the Helmholtz equation (10.542b) in

N-dimensional (10.542a) Cartesian coordinates (note 8.3);
*standard CLXXXI. Cylindrical Bessel differential equation with integer
order (note 8.4; sections 9.5–9.9);
*standard CLXXXII. Solution of the Helmholtz equation (10.542b) in cylindri-
cal coordinates (note 8.4);
*standard CLXXXIII. Solution of the Helmholtz equation (10.542b) in polar
coordinates (note 8.4);
*standard CLXXXIV. Associated Legendre differential equation (notes 8.5
and 9.21);
120 Differential Equations and Applications

*standard CLXXXV. Spherical Bessel differential equation (note 8.5; subsec-

tions 9.5.23–9.5.26 and 9.6.6–9.6.8);
*standard CLXXXVI. Solution of the Helmholtz equation (10.542b) in spheri-
cal coordinates (note 8.5).

Q. b.  Hyperspherical (Hypercylinderical) Coordinates (notes 8.6–8.8):

*standard CLXXXVII (CLXXXVIII). Transformation from N-dimensional
cartesian coordinates to hyperspherical (hypercylindrical) coordinates
(note 8.6);
*standard CLXXXIX (CXC). Inverse coordinate transformation from hyper-
spherical (hypercylindrical) coordinates to N-dimensional cartesian coordi-
nates (note 8.7);
*standard CXCI (CXCII). Base vectors, scale factors, metric tensor, and its
determinant in hyperspherical (hypercylindrical) coordinates (note 8.8).

Q. c. Helmholtz Equation (10.542b) Involving the Laplacian

(10.542a) in Hyperspherical Coordinates:
*standard CXCIII (CXCIV). Helmholtz equation in hyperspherical coordi-
nates in explicit form (nested form to facilitate the separation of variables)
[note 8.9(8.10)];
*standard CXCV. Set of N ordinary differential equations arising from the
solution of the Helmholtz equation (10.542a, b) in hyperspherical coordinates
(standard CXCV) by separation of variables (note 9.11);
*standard CXCVI. Radial dependence (10.544c) of the solution of the Helmholtz
equation (10.542a, b) in hyperspherical coordinates specified by (10.544a) in
terms of Bessel and Neumann functions of order (10.544b; note 8.12):

( )
y ( x ) = J ν ( x ) , Yν ( x ) ; ν2 = a 2 + ( N − 1/ 2 ) , x 2 y ′′ + ( N − 1) x y ′ + x 2 − a 2 y = 0 ;
2

(10.544a–c)

*standard CXCVII. Dependence on the co-latitudes (10.545b) of the solu-

tion of the Helmholtz equation (10.542a, b) in hyperspherical coordinates in
terms of hyperspherical associated Legendre functions (10.545a) of degree n,
order m, and dimension  (note 8.13):

w ( θ ) = Pnm,  ( cos θ ) , Qnm,  ( cos θ ) : w ′′ + ( 1 +  ) cot θ w ′ +  n ( n + 1) − m2 csc 2 θ  w = 0;

 (10.545a, b)

*standard CXCVIII. Solution of the Helmholtz equation (10.542a, b) in hyper-

spherical coordinates in four dimensions, which is next higher dimension
than (standard CLXXXVI) spherical harmonics (note 8.14).
Classification 10.1 121

Q. d.  Solution in Hypercylindrical Coordinates (notes 8.15–8.17):

*standard CXCIX. Solution of the Helmholtz equation (10.542a, b) in four
dimensions in hypercylindrical instead of hyperspherical (standard
CXCVIII) coordinates (note 8.15);
*standard CC. Solution of the generalized equation of mathematical physics
(10.541) in N-dimensional hyperspherical coordinates as a superposition of
hyperspherical harmonics (note 8.16);
*standard CCI. As for standard CC in hypercylindrical coordinates using a
hypercylindrical coordinates using a superposition of hypercylindrical har-
monics (note 8.17).

R. Existence, Unicity, Robustness, and Uniformity Theorems

(section 9.1):
R. a. First-Order Differential Equation with Explicit Slope
(subsections 9.1.1–9.1.14):

dy
= f ( x, y ; λ ) , y ( x0 ) = y 0 + ε: y = y ( x ; y 0 , λ , ε ) . (10.546a–c)
dx

*standard CCII. Transformation to an integral equation and application of

the Picard method of successive approximations (subsections 9.1.2–9.1.3);
*standard CCIII. Lipshitz condition for one function of one variable and
relation with bounded, continuous, and differentiable functions (subsections
9.1.4–9.1.5);
*standard CCIV. Existence theorem for the solution (10.546c) of the differen-
tial equation (10.546a) omitting the parameter λ with the boundary condition
(10.546b) without perturbation ε = 0 (subsections 9.1.6–9.1.8);
*standard CCV. Rectangular region containing the successive approxima-
tions (standard CCII) to the (standard CCV) unique solution (subsection 9.1.9);
*standard CCVI. Unicity theorem for the solution (10.546c) of the differential
equation (10.546a) omitting the parameter λ with the boundary condition
(10.546b) without ε = 0 perturbation (subsection 9.1.10);
*standard CCVII. Robustness condition in which a small pertubation
ε in the boundary condition (10.546b) leads to a small perturbation ε in
the solution (10.546c) of the (10.546a) differential equation (subsections
9.1.11–9.1.13);
*standard CCVIII. Robustness theorem for the solution (10.546c) of the dif-
ferential equation (10.546a) without parameter λ for a small perturbation ε of
(10.546b) the boundary condition (subsections 9.1.12–9.1.13);
122 Differential Equations and Applications

*standard CCIX. Uniformity theorem on the differentiability of the solution

(10.546c) of the differential equation (10.546a) with regards to the parameter λ,
with the unperturbed ε = 0 boundary (10.546b) condition (subsection 9.1.14).

R. b. Generalized Autonomous System of Differential

Equations (subsections 9.1.15–9.1.17):

dy m
= fm ( x , y n ; λ ) , y m ( x0 ) = y m ,0 + ε m y = y ( x ; y 0, m , λ , ε m ) .
m, n = 1,...., M:
dx
(10.547a–d)

*standard CCX. Lipshitz condition (standard CCIII) extended to M functions

of L variables (subsection 9.1.16);
*standard CCXI. Existence and unity of the solution (10.547d) of the
autonomous system of differential equations (10.547a, b) without the
parameter λ, with initial conditions (10.547c) without ε m = 0 perturbations
(subsections 9.1.15–9.1.17);
*standard CCXII. Robustness theorem for the solution (10.547d) of the auton-
omous system of differential equations (10.547a, b) in the absence of the
parameter λ, for small perturbations ε m of (10.547c) the boundary conditions
(subsections 9.1.16–9.1.17);
*standard CCXIII. Uniformity theorem for the solution (10.547d) of the auton-
omous system of differential equations (10.547a, b) with the parameter λ, in
the absence ε m = 0 of perturbations of (10.547c) the boundary conditions (sub-
sections 9.1.16–9.1.17).

R. c. Non-Linear Differential Equation of Order N

(subsections 9.1.18 and 9.1.20):

 dN y 
n = 0,..., N − 1:
dN y
dx N
= F  x ; dx N ; λ  , ( )
y( n) ( x0 ) = y(0n) + ε n , y = y x ; y(0n) , λ , ε n .

(10.548a–d)

*standard CCXIV. Existence and unicity of solution (10.548d) of the N-th order
differential equation (10.548a, b) without parameter λ, with boundary condi-
tions (10.548c) without ε n = 0 perturbations (subsections 9.1.18 and 9.1.20);
*standard CCXV. Robustness theorem for the solution (10.548d) of the non-
linear N-th order differential equation (10.548a, b) without the parameter λ,
and with small pertubations ε n of the (10.548c) boundary condition (subsec-
tions 9.1.18 and 9.1.20);
*standard CCXVI. Uniformity theorem for the solution (10.548d) of the non-
linear N-th order differential equation (10.548a, b) with parameter λ, and
without ε n = 0 perturbation of (10.548c) the boundary conditions (subsections
9.1.18 and 9.1.20).
Classification 10.1 123

R. d.  Linear Differential Equation of Order N (subsections 9.1.19–9.1.20):

∑
dn y
An ( x ; λ ) = B ( x ; λ ) ; m = 0 ;..., N − 1:
n= 0
dx n (10.549a–d)

y( m) ( x0 ) = y(0m) + ε n , ( )
y = y x ; y(0m) , λ , ε m .

*standard CCXVII. Existence and unicity of solution (10.549d) of the lin-

ear Nth order differential equation (10.549a) without parameter λ, and
with unperturbed ε n = 0 boundary (10.549b, c) conditions (subsections
9.1.19–9.1.20);
*standard CCXVIII. Robustness theorem for the solution (10.549d) of the lin-
ear N-th order differential equation (10.549a) without parameter λ, and with
small perturbations ε n of the (10.549b, c) boundary conditions (subsections
9.1.19–9.1.20);
*standard CCXIX. Uniformity theorem for the solution (10.549d) of the lin-
ear N-th order differential equation (10.549a) with parameter λ, without
perturbations ε n = 0 of the (10.549b, c) boundary conditions (subsections
9.1.19–9.1.20).

S.  Stability of Equilibrium of a System (section 9.2)

S. a. Classification of Equilibrium and Related
Concepts (subsections 9.2.1–9.2.3):
*standard CCXX. Seven cases of physical equilibrium consisting of indiffer-
ent, and stable (unstable) in a finite time, or asymptotically or an oscillatory
evolution with time (subsection 9.2.1);
*standard CCXXI. Five cases of mathematical equilibrium consist-
ing of indifferent, stable (unstable) for all time or only asymptotically
(subsection 9.2.1);
*standard CCXXII. Classification of a function into positive (negative) defi-
nite or semi-definite or indefinite (subsection 9.2.2);
*standard CCXXIII. Extended positive (negative) definite function in the case
of dependence on parameter(s) (subsection 9.2.2);
*standard CCXXIV. Five cases of zero force for a conservative system for
which the force is specified by the gradient of a potential (subsection
9.2.2);
*standard CCXXV. Three cases of stability of the equilibrium for (standard
CCXXIV) a conservative system (subsection 9.2.2);
124 Differential Equations and Applications

*standard CCXXVI. Derivative (10.550b) following an (10.550a) autonomous

system (subsection 9.2.3):
 
x = G ( x ) : dH ∂ H dx 
= . = G. ∇H ; (10.550a, b)
dt ∂ x dt

*standard CCXXVII. Derivative (10.551b) following a generalized (10.551a) or

unsteady autonomous system (subsection 9.2.3):
  
dH ∂ H ∂ H dx ∂ H 
x = F ( x , t ) : = + . = + F. ∇ H ; (10.551a, b)
dt ∂t ∂ x dt ∂t

*standard CCXXVIII. Intermediate case between standards CCXXVI and

CCXXVII for an asymptotically (10.552b) autonomous (10.552a) system (sub-
section 9.2.3):
     
x = F ( x , t ) : lim F ( x , t ) = G ( x ) . (10.552a, b)
t→∞

S. b. Stability of an Equilibrium Specified by an Liapunov Function

(subsections 9.2.4–9.2.10):
*standard CCXXIX. First Liapunov theorem on stability, or asymptotic
stability, or instability of an autonomous or generalized autonomous system
(subsections 9.2.4–9.2.7);
*standard CCXXX. Stability of the linear damped oscillator (10.553) with con-
stant mass m and friction coefficient µ ( x ) and the spring resilience k(x) func-
tions of position (subsection 9.2.8):

 d2 d 
m 2 + µ ( x ) + k ( x )  x ( t ) = 0 ; (10.553)
 dt dt 

*standard CCXXXI. Stability of the linear damped oscillator (10.554) with

constant mass m and friction coefficient µ ( t ) and the spring resilience k(t)
functions of time (subsection 9.2.9):

 d2 d 
m 2 + µ ( t ) + k ( t )  x ( t ) = 0 ; (10.554)
 dt dt 

*standard CCXXXII. Particular case of the standard CCXXXI concerning the

instability of the parametric resonance (10.555) of the linear undamped oscil-
lator with natural frequency ω 0, and excitation frequency ω e and amplitude
h (subsection 9.2.10 and section 4.4):

 d2 
 2 + ω 0 1 + 2 h cos ( ω e t )   x ( t ) = 0. (10.555)
2

 dt 
Classification 10.1 125

S. c. Linear Differential Equation with Bounded Coefficients

(subsections 9.2.11–9.2.14):
*standard CCXXXIII. Exponential positive upper (negative lower) asymp-
totic bounds for the solution of a single (simultaneous system of) linear dif-
ferential equation(s) with constant coefficients (subsection 9.2.11);
*standard CLXXXIV. Second Liapunov theorem stating that there exist expo-
nential positive upper (negative lower) asymptotic bounds for the solutions
of an autonomous system of differential equations (10.556a, c) with bounded
(10.556b) coefficients (subsection 9.2.12):

∑
dy n
m, n = 1,...,   M ; Amn ( x ) ≤ B < ∞: = Anm ( x ) y m ( x ) , (10.556a–c)
dx m= 1

*standard CCXXXV. Existence of at least one eigenvalue for the linear auton-
omous system of differential equations (10.556a, c) with bounded (10.556b)
coefficients (subsection 9.2.13);
*standard CCXXXVI. Existence of at least one eigenvalue (standard CCXXXV)
or equivalently existence of exponential positive upper (negative lower)
bounds (standards CCXXXIII–CCXXXIV) for the solution and first N – 1
derivatives of a linear differential equation (10.557c) of order N with bounded
(10.557a, b) coefficients (subsection 9.2.14):

∑
dn y
m = 0,..., N ; Amn ( x ) ≤ B < ∞: Anm ( x ) = 0. (10.557a–c)
m= 1
dx n

S. d. Stability of an Equilibrium Specified by an Autonomous

Differential System (subsections 9.2.15–9.2.21):
*standard CCXXXVII. Solutions of a linear autonomous system (10.556a, c)
with constant coefficients (subsection 9.2.15);
*standard CCXXXVIII. Third Liapunov theorem on the stability of the equi-
librium (10.558c) of a twice-continuously differentiable (10.558b) autonomous
(10.558a, d) system (subsections 9.2.16–9.2.20):

n, m = 1,..., M ; Fn ( xm ) ∈C 2 ( x < ε ) , Fm ( 0 ) = 0 ; (10.558a–c)

dy n     
= Fn ( xm ( t )) , x0 = x ( 0 ) , x = x ( t ; x0 ) , (10.558d–f)
dx

*standard CCXXXIX. Upper bound for the solution (10.558f) of the autono-
mous differential system (10.558a–d) with the initial (10.558e) condition (sub-
sections 9.2.16–9.2.20);
126 Differential Equations and Applications

*standard CCXL. General integral of a linear autonomous differential system

as a linear combination of fundamental solutions (subsection 9.2.17);
*standard CCXLI. Positive-definite quadratic form specifying a stability qua-
dratic form (subsection 9.2.18);
*standard CCXLII. Upper and lower bounds for a quadratic form (subsection
9.2.18);
*standard CCXLIII. Upper bound for the gradient of a quadratic form (sub-
section 9.2.18);
*standard CCXLIV. Third Liapouvov theorem on the upper bound (stan-
dard CCXXXIX) for the solution (standard CCXXXVIII) of the autonomous
(10.558a–f) differential system (subsection 9.2.20);
*standard CCXLV. Stability of an autonomous differential equation (10.559b)
in which the independent variable does not appear explicitly and the func-
tion (10.559a) has continuous second-order derivative (subsection 9.2.21):

dN y  dy d 2 y d N −1 y 
(
F ∈C 2 |R N − 1 : ) dx N
= F y , ,
 dx dx 2 ,....,
dx N − 1 
. (10.559a, b)

T. Linear Differential Equation with

Periodic Coefficients (section 9.3):
M

n = 0,...., N ; An ( t ) = An ( t + τ ) : ∑
m= 0
An ( x )
dn x
dt n
= 0. (10.560a–c)

T. a.  Differential Equation of Arbitrary Order (subsections 9.3.1–9.3.4):

*standard CCXLVI. Floquet theorem on the existence of solutions for a linear
differential equation (10.560a, c) with periodic coefficients (10.560b) in the
case of single or multiple eigenvalues (subsections 9.3.1–9.3.3);
*standard CCXLVII. Four cases of stability of the solution of a linear differ-
ential equation (10.560a, c) with periodic coefficients (10.560b) based on the
eigenvalues and exponents in (standard CCXLVII) the Floquet theorem (sub-
section 9.3.3);
*standard CCXLVIII. Diagonal (banded diagonal) system associated with
single (multiple) eigenvalues (standard CCXLVI) linear differential equation
(10.560a, c) with (10.560b) periodic coefficients (subsection 9.3.4);
*standard CCXLIX. Transformation into a Jordan matrix of the banded
diagonal matrix (standard CCXLVIII) associated with multiple eigenvalues
(standard CCXLVII) of a (standard CCXLVI) linear differential equation
(10.560a, c) with (10.560b) periodic coefficients (subsection 9.3.4).
Classification 10.1 127

T. b. Second-Order Differential Equation in Invariant

Form (10.561b; subsections 9.3.5–9.3.9):

 d2 
I (t ) = I (t + τ ):  2 + I ( t )  x ( t ) = 0. (10.561a, b)
 dt 

*standard CCL. Stable (unstable) and monotonic (oscillatory) cases with

equal or distinct eigenvalues of the linear second differential equation
in invariant form (10.561b) with periodic coefficient (10.561a) (subsections
9.3.5–9.3.6);
*standard CCLI. General integral of a linear second-order differential equa-
tion (10.560c) of order N as a linear combination of N fundamental solutions
(subsection 9.3.7);
*standard CCLII. Eigenvalues of the linear second-order differential equa-
tion (10.561b) in invariant form with periodic coefficients (10.561a) (subsec-
tion 9.3.8);
*standard CCLIII. Fourth Lyapunov theorem on the stability of solutions of
the linear second-order differential equation (10.561b) in invariant form with
periodic coefficients (10.561a) (subsection 9.3.9).

U. Linear Differential Equation with Analytic

Coefficients (section 9.4; example 10.20):
N

∑
dn y
n = 0,...., N : An ∈A (|C ) : An ( x ) = 0. (10.562a–c)
n= 0
dx n

U. a. Classification of Points of Functions of a Complex Variable

(subsections 9.4.1–9.4.4):
*standard CCLIV. Convergence of the Laurent series of a function analytic in
an annulus around an isolated singularity (subsection 9.4.1);
*standard CCLV. Classification of the points of a complex single-valued func-
tion in four classes: regular, simple poles, or multiple poles, and essential
singularities (subsection 9.4.1);
*standard CCLVI. Convergence of the MacLaurin series of a function in a
circle with a regular point at the centre (subsection 9.4.1);
*standard CCCVII. Calculation of the residue of a complex single-valued
function at a simple pole or multiple pole (subsection 9.4.2);
*standard CCLVIII. Distinction between poles and essential singularities of a
single-valued complex function of a complex variable (subsection 9.4.2);
128 Differential Equations and Applications

*standard CCLIX. N branches of the multi-valued complex function N-th

root, and associated Riemann surface, principal branch, branch-point, and
discontinuity across the branch-cut (subsection 9.4.3);
*standard CCLX. As the standard CCLIX for the infinitely many branches of
the many-valued function logarithm of a complex variable (subsection 9.4.4);
*standard CCLXI. Definition and modulus, and argument of the power with
complex base and exponent (subsection 9.4.4);
*standard CCLXII. Classification of the power with complex base and expo-
nent (standard CCLXI) into five cases: ordinary point, zero, pole, multi-
valued, or many-valued branch point (subsection 9.4.4).

U. b. Classification of Points of a Linear Differential Equation

(subsections 9.4.5–9.4.7):
*standard CCLXIII. Three cases of solution of a linear unforced first-order
differential equation; analytic, regular, or irregular integrals (subsections
9.4.5–9.4.6);
*standard CCLXIV. Most general particular solution of a linear differential
equation as a logarithmic irregular integral:
α− 1 +∞

y ( x) = xa ∑
β= 0
gβ logβ x ∑c
n =− ∞
n ,β x n , (10.563)

involving the index a, its multiplicity α, the logarithmic factors gβ and the
coefficients cn,β of the Laurent series (subsection 9.4.6);
*standard CCLXV. Classification of analytic, regular, and irregular integrals
of an unforced linear first-order differential equation in the case of the point
at infinity that is excluded from the standard CCLXII (subsection 9.4.7);
*standard CCLXVI. Upper bound in modulus for the exponential of a com-
plex matrix whose elements are bounded in modulus (subsection 9.4.8);
*standard CCLXVII. – Existence of a solution bounded in modulus for a linear
first-order autonomous system with integrable coefficients (subsection 9.4.8);
*standard CCLXVIII. Analytic, regular, and irregular integrals of a linear
autonomous system of differential equations (subsection 9.4.8).

U. c. Methods of Solution of a Linear Differential Equation with

Analytic Coefficients (subsections 9.4.9–9.4.11):
*standard CCLIX. Existence of solution of the linear differential equation
(10.562c) with analytic coefficients (10.562a, b) as an analytic integral (10.563a)
whose coefficients cn coincide with those of (10.563b), a MacLaurin series
(subsection 9.4.9):
∞ ∞

y ( x) = ∑
n= 0
cn x n = ∑
n= 0
x n ( n)
n!
y ( 0 ) ; (10.563a, b)
Classification 10.1 129

*standard CCLX. Particular case of the standard CCLIX for a linear second-
order differential equation in invariant form (10.564b) with analytic (10.564a)
coefficient (subsection 9.4.9):

∞
 d2 
I ( x) = ∑I
k=0
k xk :  2 − I ( x )  y ( x ) = 0 ;
 dx 
(10.564a, b)

*standard CCLXI. First method of solution of the standard CCLIX, with the
standard CCLX as example, specifying the MacLaurin series (10.563b) as the
analytic integral (subsection 9.4.10);
*standard CCLXII. Second alternative and equivalent method of solution to
the standard CCLXI specifying the recurrence formula for the coefficients cn
of the (10.563a) analytic integral (subsection 9.4.11).

U. d. Application to the Generalized Hyperbolic and

Circular Functions (subsections 9.4.12–9.4.20):
*standard CCLXIII. General integral (10.565b) of the generalized hyperbolic
differential equation (10.565a) as a linear combination with coefficients ( c0 , c1 )
of the generalized hyperbolic cosine and sine with parameter m (subsections
9.4.12–9.4.13):

y ′′ − x m y = 0: y ( x ) = c0 cos h ( x ; m ) + c1 sinh ( x ; m ) ; (10.565a, b)

*standard CCLXIV. Airy differential equation (10.566a) [functions (10.566b)]

as the particular case of parameter unity m = 1 of the (standard CCLXII) gen-
eralized hyperbolic differential equation (10.565a) [cosine and sine (10.565b)]
(subsection 9.4.14):

y ′′ − x y = 0: y ( x ) = c0 Ai ( x ) + c1 Bi ( x ) ; (10.566a, b)

*standard CCLXV. Changing the sign in (10.565a) leads to the generalized

circular differential equation (10.567a) whose general integral (10.567b) is a
linear combination with arbitrary constants ( c0 , c1 ) of the generalized circu-
lar cosine and sine (subsection 9.4.15):

y ′′ + x m y = 0: y ( x ) = c0 cos ( x ; m ) + c1 sin ( x ; m) ; (10.567a, b)

*standard CCLXVI. Extension of the general integral (10.565b) [(10.567b)] of

the generalized hyperbolic (10.565a) [circular (10.567a)] differential equation
as a linear combination of generalized hyperbolic (circular) cosines and sines
from positive integer values [standard CCLXIII (CCLXV)] of the parameter m
to all complex values of m other than negative integers; for negative integer
values of m see the heading U. e., which follows (subsections 9.4.16–9.4.17);
130 Differential Equations and Applications

*standard CCLXVII. Derivative of the generalized hyperbolic (circular) sine

with complex variable (subsection 9.4.18);
*standard CCLXVIII. Derivative of the first and second Airy functions (sub-
section 9.4.18);
*standard CCLXIX. Derivative of the generalized hyperbolic (circular) sine
with complex variable (subsection 9.4.18);
*standard CCLXX. Generalized hyperbolic and circular cosine and sine as
increasing or decreasing functions of real variable and parameter (subsec-
tion 9.4.19);
*standard CCLXXI. Inequalities for the generalized hyperbolic and circular
cosine, and sine of real variable and parameter (subsection 9.4.19);
*standard CCLXXII. Inequalities between hyperbolic and circular cosines
(sines) with real variable and parameter (subsection 9.4.19);
*standard CCLXXXIII. Inequalities for the generalized hyperbolic and circu-
lar secant, cosecant, tangent, and cotangent with real variable and parameter
(subsection 9.4.20);
*standard CCLXXXIV. Inequalities between the generalized hyperbolic and
circular secant (cosecant) with real variable and parameter (subsection 9.4.20).

U. e. Generalized Hyperbolic (Circular) Differential Equation

with Negative Integer Parameter (example 10.20):

m = −1, −2,..., − p ; p ∈|N : y ′′ ± x m y = 0. (10.568a–c)

*standard CCLXXXV. In the case m = −1 of a simple pole in the coefficient

of (10.568c) the general integral integral is a linear combination of regular
integrals of the first and second kinds (E10.20.1);
*standard CCLXXXVI. Elementary power solutions in the case of a double
pole m = −2 in the coefficient of (10.568c) (E10.20.2);
*standard CCLXXXVII. In the case m = −3 of a triple pole in the coefficient of
(10.568c), the general integral is a linear combination of regular asymptotic
integrals of the first and second kinds (E10.20.3);
*standard CCLXXXVIII. In the cases m = −4 − 5,..., of pole of order higher
than 3 the general integral is a linear combination of generalized hyperbolic
(circular) cosines and sines of the asymptotic variable 1/x (E10.20.4);
*standard CCLXXXIX. In the particular case m = − 4 (m = − 5 ) of a quadru-
ple (quintuple) pole of the coefficient of (10.568c) with ±[+] sign, the standard
CCLXXXVIII simplifies to a general integral that is a linear combination of
generalized hyperbolic (circular) cosines and sines [Airy functions of the
first and second kinds] of the asymptotic variable 1/x (E10.20.5);
*standard CCXC. General integral of the generalized hyperbolic (circular)
differential equation (10.568c) with − ( + ) for all values of the parameter, com-
bining the standards CCLXIII–CCLXVI and CCLXXXV–CCLXXXIX (subsec-
tions 9.4.12–9.4.17; E10.20.1–E10.20.5).
Classification 10.1 131

V. A Linear Differential Equation with

Regular Singularities (section 9.5):
V. a. Linear Autonomous System of Differential Equations
with a Regular Singularity (subsections 9.5.1–9.5.4):

∑  a x
dy m 
m, n = 1,..., N : Pm , n ∈A (|C ) : = m, n
+ pm , n ( x )  y n ( x ) .
dx n= 1

(10.569a–d)

*standard CCXCI. Solution of an autonomous linear system of differen-

tial equations (10.569a, b, d) with a regular singularity (10.569c) as a linear
combination of regular integrals of the first kind (10.570b) with índex a and
coefficients cn ( a ) to be determined and radius of convergence (10.570a) (sub-
sections 9.5.1–9.5.2):

x < R: y ( x) = xa ∑x
n= 0
n
cn ( a ) , (10.570a, b)

*standard CCXCII. Example of solution by two equivalent methods of a lin-

ear autonomous system of differential equations with a regular singularity
(subsections 9.5.3–9.5.4).

V. b. Linear Differential Equation of Order N with a

Regular Singularity (subsections 9.5.5–9.5.7).

∑x
dn y
n = 0,..., N ; pn ∈A (|C ) : n
pn ( x ) = 0. (10.571a–c)
n= 0
dx n

*standard CCXCIII. Linear differential equation (10.571c) of order N with a

regular singularity (10.571a, b) whose solutions are all regular integrals (sub-
section 9.5.5);
*standard CCXCIV. General integral near a regular singularity (10.571a, b) of
a linear differential equation (10.571c) of order N as a linear combination of
regular integrals of two kinds (subsections 9.5.6–9.5.7).
132 Differential Equations and Applications

V. c. Linear Second-Order Differential Equation with a

Regular Singularity (subsections 9.5.8–9.5.12):

d2 y dy
p , q ∈A (|C ) : x2 2
+ x p ( x) + q ( x ) y ( x ) = 0. (10.572a–c)
dx dx

*standard CCXCV. General integral integral of (10.572a–c) as a linear combi-

nation of regular integrals of the first kind (10.570a, b) if the índices a±, which
are the roots of the índicial equation (10.573a, b), do not differ by an integer
a+ − a− ∉|N (subsection 9.5.8):

0 = a ( a − 1) + a p ( 0 ) + q ( 0 ) = ( a − a+ )( a − a− ) ; (10.573a, b)

*standard CCXCVI. As the standard CCXCV with coincident índices a+ = a−

so that the general integral is a linear combination of regular integrals of the
first kind and of the second kind, first type (subsection 9.5.9);
*standard CCXCVII. As the standard CCXCV with índices differing by an
integer a+ − a− ∈|N so that the general integral is a linear combination of
regular integral integrals of the first kind and of the second kind, second
type (subsection 9.5.10);
*standard CCXCVIII. Absolute (and uniform) convergence of the regular
integrals of two kinds and two types (standards CCXCV–CCXCVII) in an
annulus around the regular singularity extending up to the nearest singu-
larity (subsection 9.5.11);
*standard CCXCIX. Wronskian (10.574c) of two particular integrals of a lin-
ear second-order differential equation (10.574a, b) in terms of (10.574d) the
coefficient of the first derivative (subsection 9.5.12):

′′ + P ( x ) y1,2
y1,2 ′ ( x ) + Q ( x ) y1,2 ( x ) = 0, W ( y1 , y 2 ) = y1 y 2′ − y 2 y1′ ,
(10.574a–d)
 x

W ( x ) = W ( x0 ) exp 
 ∫ x0
p ( ξ ) dξ  ;


*standard CCC. Two particular integrals (10.574a, b) of a linear second-order

differential equation are linearly independent provided that the Wronskian
(10.574c) ≡ (10.574d) does not vanish at one point (subsection 9.5.13);
*standard CCCI. Wronskian of the generalized Bessel differential equation
(10.575e) with integrals (10.575a–d; subsection 9.5.13):

µ 
y ( x ) = J νµ ( x ) , Yνµ ( x ) , H µ(1,2
,ν ( x ):
) 
( )
x 2 y ′′ + x  1 − x 2  y ′ + x 2 − ν2 y = 0.
 2 
(10.575a–e)
Classification 10.1 133

V. d.  Generalized Bessel Function (subsections 9.5.13–9.5.15):

*standard CCCII. Generalization of the factorial to complex values through the
Euler integral of the first kind as the gamma function, which is meromorphic,
that is, analytic in the whose complex plane except for poles (subsection 9.5.13);
*standard CCCIII. Symmetry formula for the gamma function derived from
the Weierstrass infinite product (subsection 9.5.13);
*standard CCCIV. General integral of the generalized Bessel differential
equation (10.575e) as a linear combination of Bessel functions (10.575a) of
orders ±ν and degree µ if ν is not integer (subsection 9.5.14);
*standard CCCV. Wronskian of the generalized Bessel functions (10.575a)
of orders ± ν and degree µ vanishes when ν is an integer, proving linear
dependence, and failure of the general integral in the standard CCCIV
(subsection 9.5.15);
*standard CCCVI. Linear relation between the generalized Bessel functions
(10.575a) of degree µ and integer orders ν = ± n, in agreement with: (i) the
vanishing of the Wronskian (standard CCCV); (ii) the breakdown of the gen-
eral integral (standard CCCIV) (subsection 9.5.15).

V. e. Generalized Neumann Function of Integer

Order (subsections 9.5.16–9.5.29):
*standard CCCVII. Properties of the digamma function, defined as the loga-
rithmic derivative of the gamma function (standards CCCII–CCCIII), which
is also meromorphic, including residues at the same poles and recurrence
and symmetry formulas (subsection 9.5.16);
*standard CCCVIII. General integral of the generalized Bessel differential
equation (10.575e) with degree µ and non-negative integer order ν = n ∈|N 0
as a linear combination of Bessel (10.575a) and Neumann (10.575b) function
(subsections 9.5.17–9.5.19);
*standard CCCIX. Two alternative expressions for the generalized Neumann
function (10.575b) of degree µ and non-negative integer order ν = n ∈|N 0 involv-
ing preliminary and complementary functions (subsections 9.5.17–9.5.19);
*standard CCCX. Non-zero Wronskian of the generalized Bessel (10.575a)
and Neumann (10.575b) functions of degree µ and non-negative integer
order ν = n ∈|N 0, confirming that: (i) they are linearly independent (standard
CCCIX); (ii) their linear combination specifies (standard CCCVIII) the gen-
eral integral of the generalized Bessel differential equation (10.575e) for non-
negative integer order (subsection 9.5.20).

V. f. Extension of the Generalized Neumann Function to

Complex Order (subsections 9.5.20–9.5.21):
*standard CCCXI. Lemma of the Wronskians states that if two Wronskians
coincide and one function is common the other two functions differ by a
multiple of the first function (subsection 9.5.20);
134 Differential Equations and Applications

*standard CCCXII. Lemma related to the lemma of the Wronskians (stan-

dard CCCXI) states that adding a multiple of the first function to the second
function does not change the Wronskian (subsection 9.5.20);
*standard CCCXIII. Extension of the generalized Neumann function (10.575b)
of degree µ from non-negative integer order ν = n ∈|N 0 to arbitrary complex
order ν =∈|C(subsection 9.5.21);
*standard CCCXIV. General integral of the generalized Bessel differential
equation (10.575e) as a linear combination of the generalized Bessel (10.575a)
and Neumann (10.575b) functions without restrictions neither on the order
(as the standards CCCIV and CCCVIII) nor on the degree (unrestricted in all
cases) (subsection 9.5.21).

V. g. Cylindrical and Spherical Bessel Differential

Equation (subsections 9.5.22–9.5.26):
*standard CCCXV. In the particular case of zero degree µ = 0 in (10.575a–e)
the integrals (10.576a–d) of the original cylindrical Bessel differential equa-
tion (10.576e) (subsection 9.5.22):

( )
y ( x ) = J ν ( x ) , Yν ( x ) , H ν(1,2 ) ( x ) : x 2 y ′′ + x y ′′ + y ′ + x 2 − ν2 y = 0 ; (10.576a–e)

*standard CCCXVI. In the further particular case of zero order ν = 0 the inte-
grals (10.577a–d) of the original Bessel differential equation (10.577e) of order
zero (subsections 9.5.22);

y ( x ) = J 0 ( x ) , Y0 ( x ) , H 0(1,2 ) ( x ) : x y ′′ + y ′ + x y = 0 ; (10.577a–e)

*standard CCCXVII. Comparison of the cylindrical (10.576e) [spheri-

cal (10.578e)] Bessel differential equations and their integrals (10.576a–d)
[(10.578a–d)] including cylindrical (spherical) Bessel (10.576a) [(10.578a)],
Neumann (10.576b) [(10.578b)], and Hankel (10.576c, d) [(10.578c, d)] functions
(subsection 9.5.23):

y ( x ) = jn ( x ) , y n ( x ) , hn(1,2 ) ( x ) : x 2 y ′′ + 2 x y ′ +  x 2 − n ( n + 1)  y = 0, (10.578a–e)

*standard CCCXVIII. General integral of the spherical Bessel differential

equation (10.578e) specified equivalently and alternatively by a linear combi-
nation of: (i) spherical Bessel functions (10.578a) of orders ± n corresponding
to cylindrica Bessel functions (10.576a) of order ν = ± ( n + 1/2 ); (ii) spherical
Bessel (10.578a) and Neumann (10.578b) functions of order n corresponding
to cylindrical Bessel and Neumann functions of order ν = n + 1/2 (subsec-
tion 9.5.23);
Classification 10.1 135

*standard CCCXIX. Power series expansions for the spherical Bessel

(Neumann) functions (10.578a) [(10.578b)] of order n, which have a zero (pole)
of order n ( n + 1) at the origin (subsection 9.5.24);
*standard CCCXX. Six cases of regular integrals of a linear differential equa-
tion of second order (10.572c) with a regular singularity (10.572a, b) consist-
ing of two analytic (four singular) cases of integrals, namely a zero of order n
or finite value (a pole of order n, a power or logarithmic type of branch-point
or their combination) (subsection 9.5.25);
*standard CCCXXI. Spherical Bessel (10.578a) [Neumann (10.578b)] functions
expressed as derivatives involving the circular sines (cosine) (subsection 9.5.26);
*standard CCCXXII. The integrals of the generalized (10.575e), cylindrical
(10.576e), and spherical (10.578e) Bessel differential equation around the regu-
larity at the origin involve ascending power series, whereas around the point
at infinity, which is an irregular singularity, there are irregular descending
asymptotic expansions (next section W) (subsection 9.5.26).

W. Linear Differential Equation with Irregular

Similarities (sections 9.6–9.7)
W. a. Existence of Regular Integrals Near an Irregular
Singularity (subsections 9.6.1–9.6.3):
*standard CCCXXIII. Conditions for the existence of analytic, regular, and
irregular integrals of a linear second-order differential equation with ana-
lytic coefficients (10.579c) in the neighborhood of an isolated singularity
(10.579a, b; subsection 9.6.1):

d2 y dy
P, Q ∈A ( 0 < x − x0 < R ) : 2
+ x P ( x) + Q ( x ) y = 0; (10.579a, b)
dx dx
*standard CCXXIV. Example of the existence of at most one regular integral
near an irregular singularity (subsection 9.6.2).

W. b. Existence of Normal Integral(s) Near an Irregular

Singularity (subsections 9.6.3 –9.6.5):
*standard CCCXXV. Conditions for the existence of two, one, or no normal
integrals (10.580a, b) near an irregular singularity of a linear second-order
differential equation (subsection 9.6.3):

M ∞
 1
Ω  =
 x ∑m= 1
Ωm
xm
:
  1 
y ( x ) = exp Ω    x a
  x
∑x c ( a) ;
m= 0
n
n (10.580a, b)
136 Differential Equations and Applications

*standard CCCXXVI. Analytic, regular, and irregular integrals near the point
at infinity of a linear second-order differential equation (subsection 9.6.4);
*standard CCCXXVII. Comparison of an asymptotic series with an asymp-
totic expansion (subsection 9.6.5);
*standard CCCXXVIII. Properties of asymptotic expansions including (i) the
sum, (ii) the Cauchy rule for the product, (iii) term-by-term differentiation,
and (iv) integration (subsection 9.6.5).

W. c. Asymptotic Integrals for the Cylindrical and Spherical

Bessel Differential Equations (subsections 9.6.6–9.6.9):
*standard CCCXXIX. General integral of the spherical Bessel differential
equation (10.578e) as a linear combination of spherical Hankel functions of
two kinds (subsection 9.6.6);
*standard CCCXXX. Relations among the cylindrical Bessel (10.576a),
Neumann (10.576b) and Hankel (10.576c) functions of two kinds (subsection
9.6.6);
*standard CCCXXXI. Leading asymptotic term for the cylindrical and spher-
ical Hankel functions of two kinds (subsection 9.6.6);
*standard CCCXXXII. Asymptotic expansions for the spherical and cylin-
drical Bessel, Neumann, and Hankel functions of two kinds (subsections
9.6.7–9.6.8);
*standard CCCXXXIII. General integral of the cylindrical Bessel differential
equation (10.576e) as a linear combination of Hankel functions (10.576c, d) of
two kinds (subsection 9.6.8).

W. d. Asymptotic Integrals for the Generalized Bessel

Differential Equation (subsections 9.6.9–9.6.11):
*standard CCCXXXIV. Generalized Hankel function of the first kind (10.575c)
with order ν and non-zero degree µ ≠ 0 as a regular integral of the general-
ized Bessel differential equation (10.575e) near the point at infinity specified
by an asymptotic expansion of descending powers (subsection 9.6.9).
*standard CCCXXXV. Generalized Hankel function of the second kind
(10.575d) with order ν and non-zero degree µ ≠ 0 as a normal integral
(10.580a, b) of the generalized Bessel differential equation (10.575e) near the
point at infinity (subsections 9.6.10–9.6.11);
*standard CCCXXXVI. General integral of the generalized Bessel differential
equation (10.575e) of order ν and non-zero degree µ ≠ 0 as a linear combina-
tion of generalized Hankel functions (10.575c) [(10.575d)] of the first (second)
kind [standard CCCXXXIV (CCCXXXV)] (subsection 9.6.11);
*standard CCCXXXVII. Wronskian of the generalized Hankel functions of
the first (10.575c) and second (10.575d) kinds with order ν and degree µ, show-
ing they are linearly independent for non-zero degree µ ≠ 0, in agreement
with the standard CCCXXXVI (subsection 9.6.11);
Classification 10.1 137

*standard CCCXXXVIII. General integral of the generalized Bessel differen-

tial equation (10.575e) as a linear combination of generalized Neumann func-
tions (10.575b) of order µ and degrees ±ν not integers ν ∉|Z (subsection 9.6.11);
*standard CCCXXXIX. Equivalent and alternative forms of the general inte-
gral of the generalized Bessel differential equation (10.575e) involving the
generalized Bessel (10.575a), Neumann (10.575b), and Hankel functions, the
latter of first (10.575c) and second (10.575d) kind (subsection 9.6.11; table 9.5);
*standard CCCXL. Comparison of the generalized (10.575e) [cylindrical
(10.576e)] Bessel differential equations and their integrals in terms of the
generalized (cylindrical) Bessel (10.575a) [(10.576a)], Neumann (10.575b)
[(10.576b)], and Hankel function of the first (10.575c) [(10.576c)] and second
(10.575d) [(10.576d)] kinds (subsection 9.6.11; table 6.7).

W. e. Bifurcation of the Generalized Bessel Differential

Equation at Infinity (subsections 9.6.12–9.6.15):
*standard CCCXLI. The generalized Bessel (10.575a) and Neumann (10.575b)
[Hankel (10.575c, d)] functions do (do not) tend to the cylindrical Bessel
(10.576a) and Neumann (10.576b) [Hankel (10.576c, d)] functions for zero
degree because in the limit µ → 0 the singularity at the origin (infinity)
remains regular (changes the degree of irregularity) (subsection 9.6.12);
*standard CCCXLII. There is no asymptotic solution of the generalized
Bessel differential equation (10.575e) that tends to the cylindrical Hankel
functions (10.576c, d) as the degree tends to zero µ → 0 (subsections
9.6.12–9.6.13);
*standard CCCXLIII. Alternate proof of the standard CCCXLII by reduction
ad absurdum (subsection 9.6.13);
*standard CCCXLIV. Hopf-type bifurcation of the solution of the general-
ized Bessel differential equation (10.575e) at infinity for zero degree µ = 0
when it reduces to the cylindrical Bessel differential equation (10.576e)
(subsection 9.6.14);
*standard CCCXLV. The computation of the generalized (10.575a–d)/
cylindrical (10.576a–d)/spherical (10.578a–d) Bessel (10.575a/10.576a/10.5
78a), Neumann (10.575b/10.576b/10.578b), and Hankel (10.575c, d/10.576c,
d/10.578c, d) functions of two kinds for small (large) variable is performed
more accurately using a few terms of the power series (asymptotic expansion)
(subsection 9.6.15).

W. f.  General Irregular Integrals of Two Kinds (subsections 9.7.1–9.7.6):

*standard CCCXLVI. Existence of eigenvalues and eigensolutions of a linear
second-order differential equation whose coefficients have an isolated singu-
larity (subsection 9.7.1);
*standard CCCXLVII. Distinct eigenvalues correspond to linearly indepen-
dent eigenintegrals in the standard CCCXLVI (subsection 9.7.1);
138 Differential Equations and Applications

*standard CCCXLVIII. The statement that linearly dependent particular inte-

grals have the same eigenvalues in the standard CCCXLVI is equivalent to
the standard CCCXLVII ( subsection 9.7.1);
*standard CCCXLIX. The existence of at least one irregular integral of the first
kind (10.581) of standard CCCXLVI consisting of a Laurent series (multiplied
by a branch point) with coefficients (índex) to be determined (subsection 9.7.2):

+∞

y ( x ) = ( x − x0 )
a
∑ ( x − x ) c ( a) ; (10.581)
n =−∞
0
n
n

*standard CCCL. If the eigenvalues are distinct (standard CCCXLVII), that

is the indices do not differ by an integer, there are two linearly indepen-
dent irregular integrals of the first kind (standard CCCXLIX) and their
linear combination specifies the general integral (standard CCCXLVI)
(subsection 9.7.2);
*standard CCCLI. The general integral of a linear second-order differential
equation with an isolated singularity (standard CCCXLVI) as a linear combi-
nation of irregular integrals of the first and second kinds when the eigenval-
ues coincide, that is the índices differ by an integer (subsection 9.7.3);
*standard CCCLII. The method alternative to standard CCCLI of obtaining
the irregular integral of the second kind from the irregular integral of the
first kind (subsection 9.7.4);
*standard CCCLIII. The derivation of the regular integrals of the second
kind, first (second) type [standard CCXCVI (CCXCVII)] from the irregular
integral of the second kind (standard CCCLI) by omitting the descending
powers (subsection 9.7.5);
*standard CCCLIV. List of all possible solutions of a linear differential equa-
tion including: (i) analytic integrals near a regular point; (ii) regular integrals
of the first kind, or second kind first or second type near a regular singular-
ity; (iii) irregular integrals of the first and second kind, in particular, normal
integrals, near an irregular singularity (subsection 9.7.6).

W. g.  Infinite Determinants and Linear Systems (subsections 9.7.7–9.7.12):

*standard CCCLV. The irregular integral of a linear differential equation
near an irregular singularity is specified by a Laurent series with a branch-
point with: (i) coefficients satisfying an infinite linear system; (ii) índices the
roots of an infinite determinant (subsection 9.7.8);
*standard CCCLVI. Classification of infinite determinants including (i) diver-
gent, (ii) oscillatory, and (iii–viii) simply, conditionally, absolutely, uniformly,
or totally convergent (subsection 9.7.9);
*standard CCCLVII. If the infinite product (double series) of diagonal (non-
diagonal) elements is absolutely convergent the infinite determinant is con-
vergent (subsection 9.7.10).
Classification 10.1 139

W. h. Multiple Parametric Resonance and Hill Differential

Equation (subsections 9.7.11–9.7.16):

d2 z  
∞

+ b0 +
dθ2  ∑b cos ( 2 θ) u(θ) = 0. (10.582)
=1


*standard CCCLVIII. The parametric resonance (problems 90 to 95) of a har-

monic oscillator excited at a fundamental applied frequency and all its har-
monics leads to a generalization (note 9.33) of the Hill differential equation
(10.582) (subsection 9.7.11);
*standard CCCLIX. The transformation of a linear second-order differential
equation (problem 161) to the invariant form (problem 166) leads to the Hill
differential equation (10.582) if the invariant is an even function of bounded
fluctuation with period π (subsection 9.7.12);
*standard CCCLX. The irregular integral solution of the Hill differential
equation with trignometric (10.582) [power (10.583)] coefficients valid in an
upper-half-plane (circle) in the θ -plane (x-plane) (subsection 9.7.13):

dy  
∞
d2 y
0 = x2
dx 2
+x − b0 +
dx 

∑b ( x
n= 1
n
n
)
+ x − n  y = 0. (10.583)


*standard CCCLXI. Infinite linear system (determinant) specifying the coef-

ficients (índices) of the irregular integral solutions of the Hill differential
equation in trigonometric (10.582) [algebraic (10.583)] form (subsection 9.7.14);
*standard CCCLXII. Evaluation of the Hill infinite determinant whose roots
are the índices that appear in the coefficients of the irregular integral solu-
tions of the Hill differential equation in trigonometric (10.582) or algebraic
(10.583) form (subsections 9.7.15–9.7.16).

X. Kernel and Path of the Solution by

Integral Transforms (section 9.8)
X. a. Beta Function or Eulerian Integral of the
Second Kind (subsections 9.8.1–9.8.4):
*standard CCCLXIII. Evaluation of integrals in terms of the Euler beta func-
tion (subsection 9.8.1);
*standard CCCLXIV. Properties of the beta function, including generaliza-
tion of the permutation from positive integer to complex variables (subsec-
tions 9.8.1–9.8.2);
140 Differential Equations and Applications

*standard CCCLXV. Relation between the gamma (standard CCCII) and beta
(standard CCCLXIII) functions (subsection 9.8.3);
*standard CCCLXVI. Evaluation of integrals in terms of beta (standards
CCCLIII–CCCLV) or gamma (standards CCCII–CCCIII) functions (subsec-
tion 9.8.4).

X. b. Generalized Laplace Transform Along a Path in the

Complex Plane (subsections 9.8.5–9.8.11):
*standard CCCLXVII. Solution of a linear differential equation of order
N whose coefficients are linear functions of the independent variable
(10.584b) by a generalized Laplace transform (10.584a) along a path L in the
complex plane (subsection 9.8.6):

∑( pn + qn x) dxn = 0 ;
dn y
y ( x) =
∫
L
v ( s ) e sx dx:
n= 0
(10.584a, b)

*standard CCCLXVIII. Solution of the linear first-order differential equation

that specifies the transform v(s) in (10.584a) for the standard CCCLXVII (sub-
section 9.8.6);
*standard CCCLXIX. In the case when the coefficients An ( x ) of the linear
differential equation (10.562c) of order N is a polynomial of degree M of the
independent variable [instead of M = 1 in (10.584b)] the Laplace (10.584a)
transform ν ( s ) satisfies a linear differential equation of order M (instead of
M = 1) (subsection 9.8.7);
*standard CCCLXX. Depression of the degree of the cubic coefficients in the
generalized Bessel differential equation (10.575e) to linear in (10.585c) via the
changes of independent (10.585a) and dependent (10.585b) variable (subsec-
tion 9.8.7):

 µ η 2 µν
η = x2 , ( )
y ( x ) = x± ν Ψ ± x2 : η Ψ ′′± +  1 ± ν −
  Ψ ′± +
4  8
Ψ = 0;

(10.585a–c)

*standard CCCLXXI. Choice of paths of integration Γ ± for the generalized

Laplace transform (10.584a) that satisties the generalized Bessel differential
equation (10.575e) (subsections 9.8.8–9.8.9);
*standard CCCLXXII. Integral representation of the generalized Bessel func-
tion (10.575a) as a generalized Laplace (10.584a) parametric integral, but exclud-
ing the original Bessel function (standard CCCLXXIII) (subsection 9.8.9);
*standard CCCLXXIII. Integral representation for the Bessel function
(10.576a) not included in the standard CCCLXXII (subsections 9.8.9–9.8.10).
Classification 10.1 141

X. c. Generalized Integral Transform and Adjoint

Kernel Operator (subsections 9.8.12–9.8.16):
*standard CCCLXXIV. Solution of a linear differential equation (10.586a) has
a general integral transform (10.586b) in which: (i) the kernel k(x,s) satisfies a
partial differential equation involving another linear differential operator M
with s as the independent variable; (ii) the adjoint operator M specifies the
integral transform ν ( s ) ; (iii) the path of integration Γ is such that the bilinear
concomitant takes the same value at both ends (subsection 9.8.12):
N

∑
dn y
n= 0
An ( x )
dx n
= 0: y ( x) =
∫ k ( x, s) v ( s) ds ; (10.586a–b)
Γ

*standard CCCLXXV. Kernel k ( x , s ), transform ν ( s ) and path of integration

Γ in the general integral transform (10.586b) for the solution of the Bessel dif-
ferential equation (10.576e) (subsection 9.8.13);
*standard CCCLXXVI. Integral representations for the Hankel functions of
the first (10.576c) [second (10.576d)] kinds (subsection 9.8.14);
*standard CCCLXXVII. Integral representation for the Neumann (10.576b)
function (subsection 9.8.15);
*standard CCCLXXVIII. Second Integral representation for the Bessel func-
tion (10.576a) distinct from the standard CCCLXXIII (subsection 9.8.16);
*standard CCCLXXIX. Integral representation for the Bessel coefficients, that
is, Bessel functions (10.578a) of integer order, including zero (subsection 9.8.16).

Y. Solution of Linear Second-Order Differential

Equations by Continued Fractions (section 9.9):
*standard CCCLXXX. Logarithmic derivative of the solution of a linear sec-
ond-order differential equation (10.587a) as a continued fraction (10.587b)
(subsections 9.9.1–9.9.2):
1 b1 b2 b1 b2 b y′ d
y = a0 y ′ + b1 y ′′:   ... n = =
a0 + a1 + a2 + a1+ a2 + an + rn y dx
( log y ) ; (10.587a, b)
*standard CCCLXXXI. Example of the rational solution of a linear second-order
differential equation as a terminating continued fraction (subsection 9.9.3);
*standard CCCLXXII. Application of standard CCCLXXX to the confluent
hypergeometric differential equation (10.588c) specifying the representation as
a continued fraction of the logarithmic derivative of the confluent hypergeomet-
ric functions of the first (10.588a) and second (10.588b) kinds (subsection 9.9.4):

y ( x ) = F (α ; γ ; x ) , G (α ; γ ; x ): x y ′′ + ( γ − x ) y ′ − α y = 0; (10.588a–c)
142 Differential Equations and Applications

*standard CCCLXXXIII. Convergence theorem for simple infinite continued

fractions whose coefficients are real and positive (subsections 9.9.5–9.9.6);
*standard CCCLXXXIV. Theorem on the difference of successive continuants
of a continued fraction (subsection 9.9.5);
*standard CCCLXXXV. Recurrence relations for the numerators and denomi-
nators of a continued fraction (subsection 9.9.5);
*standard CCCLXXXVI. The inverse continued fractions exchanging numer-
ators and denominators (subsection 9.9.5);
*standard CCCLXXXVII. Convergence of the continued fraction for the log-
arithmic derivative of the confluent hypergeometric function (10.588a) for
real positive variable x and parameters a, b (subsection 9.9.6);
*standard CCCLXXXVIII. General integral of the confluent hypergeometric
differential equation (10.588c) as a linear combination of confluent hypergeo-
metric functions of the first kind (10.588a) valid in the finite complex x-plane
for γ not zero or a negative integer (subsection 9.9.8);
*standard CCCLXXXIX. Relation between the generalized Bessel function
(10.575a) and the confluent hypergeometric function (10.588a) (subsection
9.9.9);
*standard CCCXC. Continued fraction for the logarithmic derivative of the
generalized Bessel function (10.575a), not valid for the original Bessel func-
tion (10.576a) (subsection 9.9.9);
*standard CCCXCI. Differentiation and recurrence formulas for the cylinder
functions (10.576a–d) (subsection 9.9.10);
*standard CCCXCII. Alternate proof of standard CCCXXI, concerning the
Rayleigh formulas for spherical Bessel (10.578a) and Neumann (10.578b)
functions (subsection 9.9.11);
*standard CCCXCIII. Convergent continued fraction for cylinder functions
(subsection 9.9.12);
*standard CCCXCIV. Convergence theorem for special-limit periodic contin-
ued fractions (subsections 9.9.13–9.9.15);
*standard CCCXCV. Limit of the ratio of two cylinder functions of high order
(subsection 9.9.12).

Z. Higher Transcendental Functions: Special

and Extended (notes 9.7–9.47)
Z. a.  Generalized Hypergeometric Differential Equation (notes 9.9–9.13):

 N M
 


∏
n= 1
 d
 x
dx

− an  − x
 ∏
n= 1
 d
 x
dx
− bm   y ( x ) = 0 ; (10.589)


Classification 10.1 143

*standard CCCXCVI. General integral of the generalized hypergeometric

differential equation (10.589) for N > M as a linear combination of N gener-
alized hypergeometric series convergent in the finite x-plane (note 9.9);
*standard CCCXCVII. Example of the solution of the generalized hypergeo-
metric differential equation (10.589) in the case 3 = N > M = 2 leading to the
differential equation (10.590) (note 9.10):

x 3 y ′′′ + ( 3 − x − α 1 − α 2 − α 3 ) x 2 y ′′
+ 1 − α 1 − α 2 − α 3 + α 1α 2 + α 1α 3 + α 2 α 3 − x ( 1 − β1 − β 2 )  x y ′ (10.590)
− ( α 1α 2 α 3 + β 1 β 2 x ) y = 0 ;

*standard CCCXCVIII. General integral of the generalized hypergeometric

differential equation (10.589) for M > N , which is opposite to the standard
CCCXCVI, as a linear combination of generalized hypergeometric asymp-
totic series converging in the whole complex x-plane, including the point at
infinity, and excluding only the origin (note 9.11);
*standard CCCXCIX. Example of the generalized hypergeometric differ-
ential equation (10.589) in the case 3 = M > N = 2 opposite to the standard
CCCXCVII, leading to the differential equation (10.591) (note 9.12);

x 3 y ′′′ +  ( 3 − β1 − β 2 − β 3 ) x − 1 x 2 y ′′
+  ( 1 − β1 − β 2 − β 3 + β1β 2 + β1β 3 + β 2β 3 ) x y ′ + α 1α 2 − 1  x y ′ (10.591)
− (β1β 2β 3 x − α 1α 2 ) y = 0 ;

*standard CD. Convergence of the ascending generalized hypergeometric

series solutions (standard CCCXCVI) inside the unit disk and on the unit
circle for the generalized hypergeometric differential equation (10.589) in the
case M = N (note 9.13);
*standard CDI. Convergence of the asymptotic generalized hypergeometric
series solutions (standard CCCXCVIII) outside the unit disk and on the unit
circle for the generalized hypergeometric differential equation (10.589) in the
case M = N (note 9.13);
*standard CDII. Existence of little or no overlap between the regions of valid-
ity [standard CD(CDI) of the inner (outer) solutions in ascending (descending)
powers [standard CCCXCV (CCCXVIII)] of the generalized hypergeometric
differential equation (10.589) in the case M = N (note 9.13);
*standard CDIII. Schwartz group of six coordinates transformations that:
(i) interchange between themselves the point x = 0, 1, ∞; (ii) these are the three
regular singularities of the generalized hypergeometric differential equation
(10.589) for M = N; (iii) hence the changes of coordinates (i) relate the gener-
alized hypergeometric functions in distinct overlapping regions of the com
plex x-plane and provide analytic continuation of the solutions (iii) (note 9.13).
144 Differential Equations and Applications

Z. b.  Gaussian Hypergeometric Differential Equation (notes 9.14–9.19):

y ( x ) = F ( α , β ; γ ; x ) , G ( α , β ; γ ; x ) : x ( 1 − x ) y ′′ +  γ − ( α + β + 1) x  y ′ − αβ y = 0,
(10.592a–c)
*standard CDIV. Derivation of the Gaussian hypergeometric differential equa-
tion (10.592c) from the Gaussian hypergeometric series (10.592a) (note 9.14);
*standard CDV. General integral of the Gaussian hypergeometric differential
equation (10.592c) as a linear combination of Gaussian hypergeometric func-
tion of the first kind (10.592a) when γ is not an integer (note 9.15);
*standard CDVI. General integral of the Gaussian hypergeometric differ-
ential equation (10.592c) in the case of γ-positive integer, which is excluded
from the standard CDV, when a Gaussian hypergeometric functions of the
second kind (10.592b) appears (note 9.16);
*standard CDVII. Gaussian hypergeometric function of the second kind
(10.592b) with γ = n, a positive integer, which appears in this case in the solu-
tion of the Gaussian hypergeometric differential equation (10.592a) (note 9.17);
*standard CDVIII. General integral of the Gaussian hypergeometric differen-
tial equation (10.592c) in the case when γ is zero or a negative integer, which
is excluded from the standards CDV and CDVI, and unlike (like) the stan-
dard CDV(CDVI) involves a Gaussian hypergeometric function of the second
kind (10.592b) (note 9.18);
*standard CDIX. Gaussian hypergeometric function of the second kind
(10.592b), which appears in the general integral of the Gaussian hypergeomet-
ric differential equation (10.592c) when γ is zero or a positive integer (note 9.18);
*standard CDX. Particular case of the Gaussian hypergeometric function of
the second kind (10.592b) for γ = 1 (note 9.19);
*standard CDXI. Simplification of the Gaussian hypergeometric function of
the second kind (10.592b) when neither α nor β are negative integers (note 9.19);
*standard CDXII. Simplification of the Gaussian hypergeometric functions
of the first kind (10.592a) or second kind (10.592b) when α or β is a negative
integer (note 9.19);
*standard CDXIII. Summary of the general integrals of the Gaussian hyper-
geometric differential equation (10.592c) involving Gaussian hypergeometric
functions of the first kind (10.592a) or second kind (10.592b) for all cases of
values of the parameters ( α , β, γ ) (note 9.19).

Z. c. Legendre, Associated, Ultraspherical, and Hyperspherical

Differential Equations (notes 9.20–9.25):
*standard CDXIV. Solution of the Legendre differential equation (10.593c)
of degree ν as a linear combination of Legendre functions of the first kind
(10.593a) and second kind (10.593b) (note 9.20):

y ( x ) = Pν ( x ) , Qν ( x ) : (1 − x ) y ′′ − 2 x y ′ + ν ( ν + 1) y = 0 ; (10.593a–c)
2
Classification 10.1 145

*standard CDXV. Relation between the Legendre functions of the first

(10.593a) [second (10.593b)] kind and the Gaussian hypergeometric functions
of the first (10.592a) [second (10.592b)] kind (note 9.20);
*standard CDXVI. Solution of the Legendre differential equation (10.593c)
finite in the unit circle x ≤ 1, which is proportional to the Legendre polyno-
mial (10.593a) with non-negative integer degree (note 9.20);
*standard CDXVII. General integral of the ultraspherical Legendre differ-
ential equation (10.594d) of degree ν and dimension γ as a linear combi-
nation of ultraspherical Legendre functions of the first kind (10.594a) and
second kind (10.594b) for λ = 0, 2, 4,... (note 9.21):

y ( x ) = P ν,λ ( x ) , Q ν,λ ( x ) , R ν,λ ( x ) : (1 − x ) y ′′ − ( 2 + λ ) x y ′ + ν ( ν + 1 + λ ) y = 0 ;

2

(10.594a–d)

*standard CDXVIII. General integral of the standard CDXVII for

λ ≠ 0, ± 2, + 4,... as a linear combination of ultraspherical Legendre functions
of the first kind (10.594a, c) (note 9.21);
*standard CDXIX. General integral of the ultraspherical Legendre differen-
tial equation (10.594d) in the case λ = − 2, − 4,... excluded from the standards
CDXVII and CDXVIII, and the convergence conditions in all three cases for
all values of λ-(note 9.21);
*standard CDXX. Solution of the ultraspherical Legendre differential equa-
tion (10.594d) finite in the unit circle x ≤ 1, which is proportional to the ultra-
spherical Legendre polynomial (10.594a) with non-negative integer degree
(note 9.21);
*standard CDXXI. General integral of the associated Legendre differential
equation (10.595c) of degree ν and order µ as a linear combination of first
(10.595a) and second (10.595b) associated Legendre functions of the first kind
(note 9.22):

y = 0,
y ( x ) = Pνµ ( x ) , Rνµ ( x ) :
 (1 − x ) y ′′ − 2 x y ′ +  ν ( ν + 1) − 1 −µx
2
2

(10.595a–c)

*standard CDXXXII. Spherical harmonics as the solution of the associated

Legendre differential equation (10.595c) finite in the interval x ≤ 1, which
is proportional to the first associated Legendre function of the first kind
(10.595a), with integer degree ν = n larger n > m than the integer order µ = m
(note 9.22);
*standard CDXXIII. Invariance of the associated Legendre differential equa-
tion (10.595c) with regards to the change of sign of the order µ → − µ, except
for the convergence of solutions on the boundary of convergence (note 9.22);
*standard CDXXIV. General integral of the hyperspherical-­associated
Legendre differential equation (10.596c) of degree ν order µ and
146 Differential Equations and Applications

dimension λ as a linear combination of the first (10.596a) and second (10.596b)

hyperspherical-associated Legendre functions (notes 9.23–9.24):

(1 − x ) y ′′ − ( 2 + λ ) x y ′ +  ν ( ν + 1) − 1 −µ x 
2
y ( x ) = Aνµ, λ ( x ) , Bνµ, λ ( x ) : 2
2 y = 0;
 
(10.596a–c)

*standard CDXXV. Particular cases λ = 0 ( λ = 0 = µ ) of (10.596c) leading to the

associated Legendre (10.595c) [Legendre (10.593c)] differential equation and
functions (10.596a, b), which are linear combinations of (10.595a, b) [(10.593a, b)]
(note 9.24);
*standard CDXXVI. Solution of the hyperspherical-associated Legendre dif-
ferential equation (10.596c), which is finite in the unit circle x ≤ 1 and is pro-
portional to the first hyperspherical-associated Legendre function (10.596a)
(note 9.24);
*standard CDXXVII. General integral of the hyperspherical Legendre dif-
ferential equation (10.597c) of degree ν and dimension λ as a linear combi-
nation of first (10.597a) [second (10.597b)] hyperspherical Legendre functions
(note 9.25):

y ( x ) = Pν , λ ( x ) , Rν , λ ( x ) : (1 − x ) y ′′ − ( 2 + λ ) x y ′ + ν ( ν + 1) y = 0 ; (10.597a–c)
2

*standard CDXXVIII. Solution of the hyperspherical Legendre differential

equation (10.597c), which is finite in the circle 0 ≤ 1 − x ≤ 2 for Re ( λ ) > 0 and
is proportional to the first hyperspherical-associated Legendre function
(10.597a) (note 9.25).

Z. d.  Jacobi and Chebychev Functions of Two Kinds (notes 9.26–9.28):

*standard CDXXIX. Relation between the Jacobi function (10.598a) with
degree ν and parameters (a, b) and the Gaussian hypergeometric function of
the first kind (10.592a) (note 9.26);
*standard CDXXX. General integral of the Jacobi differential equation
(10.598b) as a linear combination of (10.598a) Jacobi functions (note 9.26):

y ( x ) = I νa ,b ( x ) : (1 − x ) y ′′ + b − a − ( a + b + 2 ) x  y ′ + ν ( ν + a + b + 1) y = 0.
2

 (10.598a, b)

*standard CDXXXI. Recurrence and differentiation formulas for the

Chebychev functions of the first type (10.599a, b) (note 9.27);
*standard CDXXXII. Relation between the first (10.599a) [second (10.599b)]
Chebychev functions of the first type and the Gaussian hypergeometric
function (133a) of the first kind (note 9.27);
Classification 10.1 147

*standard CDXXXIII. General integral of the Chebychev differential equa-

tion of the first type (10.599c) as a linear combination of the first (10.599a)
[second (10.599b)] Chebychev functions of the first type (note 9.27):

y ( x ) = Tν ( x ) , Sν ( x ) : (1 − x ) y ′′ − x y ′ + ν
2 2
y = 0 ; (10.599a–c)

*standard CDXXXIV. The solution of the Chebychev differential equation

of the first type (10.599a), which is finite in the closed unit disk x > 1 is pro-
portional to the first (10.599a) Chebychev function of the first type (note 9.27);
*standard CDXXXV. Recurrence and differentiational formulas for the
Chebychev functions of the second type (10.600a, b) (note 9.28);
*standard CDXXXVI. Relation between the first (10.600a) [second (10.600b)]
Chebychev function of the second type and the Gaussian hypergeometric
function of the first kind (10.592a) (note 9.28);
*standard CDXXXVII. General integral of the Chebychev differential equa-
tion of the second type (10.600c) as a linear combination of the first (10.600a)
and second (10.600b) Chebychev functions of the second type (note 9.28):

y = ( x ) = U ν ( x ) , Vν ( x ) : (1 − x ) y ′′ − 3 x y ′ + ν ( ν + 2 ) y = 0 ; (10.600a–c)
2

*standard CDXXXVIII. The solution of the Chebychev differential equation

of the second type (10.600c), which is finite in the closed unit disk x ≤ 1 is
proportional to the first (10.600a) Chebychev function of the second type
(note 9.28);
*standard CDXXXIX. The Legendre functions (10.593a) and first Chebychev
functions of the first (10.599a) [second (10.600a)] type with zero or positive
integer degree ν = n are all distinct polynomials of degree n corresponding
to terminating Gaussian hypergeometric series (10.592a) of the first kind
(note 9.28).

Z. e. Confluent Hypergeometric, Laguerre, Associated, and Hermite

Functions (notes 9.29–9.32):
*standard CDXL. Comparison of the Gaussian (10.592c) [confluent (10.588c)]
differential equations and functions of the first kind (10.592a) [(10.588a)]
(note 9.29);
*standard CDXLI. Comparison of the general integrals of the Gaussian
(10.592c) [confluent (10.588c)] hypergeometric differential equations in the
cases involving both functions of the first (10.592a) [(10.588a)] and second
(10.592b) [(10.588b)] kinds (note 9.29);
*standard CDXLII. Relation between the Laguerre function of the first kind
(10.601a) [second (10.601b) kind] and degree ν and the confluent hypergeo-
metric function of the first (10.588a) [second (10.588b)] kind (note 9.30);
148 Differential Equations and Applications

*standard CDXLIII. General integral of the Laguerre differential equation

(10.601c) with degree ν as a linear combination of Laguerre functions of the
first (10.601a) and second (10.601b) kinds (note 9.30):

y ( x ) = Lν ( x ) , N ν ( x ) : x y ′′ + ( 1 − x ) y ′ + ν y = 0. (10.601a–c)

*standard CDXLIV. The solution of the Laguerre differential equation

(10.601c) finite at the origin is proportional to the Laguerre function of the
first kind (10.601a) (note 9.30);
*standard CDXLV. Relation between the associated Laguerre function
(10.602a) of degree ν and order µ and the confluent hypergeometric function
of the first kind (10.588a) (note 9.31);
*standard CDXLVI. General integral of the associated Laguerre differential
equation (10.602b) as a combination of associated Laguerre functions
(10.602a) (note 9.31):

y ( x ) = Lµν ( x ) : x y ′′ + ( 1 + µ − x ) y ′ + ν y = 0. (10.602a, b)

*standard CDXLVII. The solution of the associated Laguerre differen-

tial equation (10.602b) finite at the origin is proportional to one associated
Laguerre function (10.602a) (note 9.31);
*standard CDXLVIII. Relation between the first Hermite function (10.603a) of
degree ν and the confluent hypergeometric function of the first kind (10.588a)
(note 9.32);
*standard CDXLIX. Relation between the second Hermite function (10.603b)
of degree ν and the confluent hypergeometric function of the first kind
(10.588b) (note 9.32);
*standard CDL. General integral of the Hermite differential equation
(10.603c) of degree ν as a linear combination of the first (10.603a) and second
(10.603b) Hermite functions (note 9.32):

y ( x ) = H ν ( x ) , Iν ( x ): y ′′ − 2 x y ′ + 2 ν y = 0. (10.603a–c)

Z. f. Sub- and Superhypergeometric Differential Equations

(notes 9.33–9.38):
*standard CDLI. Subhypergeometric differential equations that have three
regular singularities (one regular singularity and one irregular singular-
ity of degree one) like the Gaussian (confluent) hypergeometric (10.592c)
[(10.588c)] differential equation [standards CCCXCV–CDXXXIX (CDXL-
CDL)] (note 9.33);
*standard CDLII. Relation between the Bessel (10.576a) and Gaussian hyper-
geometric (10.592a) functions (note 9.33);
Classification 10.1 149

*standard CDLIII. Superhypergeometric differential equations that have (a)

either more than three regular singularities or (b) two/one/zero regular sin-
gularities and one irregular singularity of degree at least one/two/three or
(c) or two irregular singularities (note 9.34);
*standard CDLIV. Four alternative equivalent forms of the generalized Hill
differential equation (10.604b) satisfied by (10.604a) the generalized Hill
functions (note 9.34):

w = ( θ ) = H ( b1 ,....., bM ; cos θ ) :
d2 w
dθ2
+ ∑b
m= 1
m cos ( m θ ) = 0. (10.604a, b)

*standard CDLV. Relation between the generalized Hill function (10.604a)

and the extended Gaussian hypergeometric function (10.606a) (note 9.35);
*standard CDLVI. Four equivalent forms of the Mathieu differential equa-
tion (10.605b) whose solutions are Mathieu functions (10.605a) (note 9.35):

y ( x ) = M ( a, q ; x ) : (1 − x ) y ′′ − x y ′ − x y ′ + ( a + 2 q − 4 q x ) y = 0 ;
2 2

 (10.605a, b)

*standard CDLVII. Relation between the Mathieu function (10.605a) and the
extended Gaussian hypergeometric functions (10.606a) (note 9.35);
*standard CDLVIII. General integral of the extended Gaussian hypergeo-
metric differential equation (10.606b) with upper ( α , β ), lower γ, and asymp-
totic ( A1 ,..., AM ) parameters as a linear combination of extended Gaussian
hypergeometric functions of the first kind (10.606a) for γ not an integer
(notes 9.35–9.36):

y ( x ) = F ( α , β ; γ ; A1 ,..., AM ; x ) :

 M
 (10.606a, b)
x ( 1 − x ) y ′′ +  γ − ( α + β + 1) x  y ′ −  α β +

∑m= 1
Am x m  y = 0.


*standard CDLIX. General integral of the extended confluent hypergeomet-

ric differential equation (10.607b) in terms of (10.607a) extended confluent
hypergeometric functions (note 9.37):

 M

y ( x ) = F ( α , β ; γ ; A1 ,..., AM ; x ) : x y ′′ + ( γ − x ) y ′ −  α +

∑A
m= 1
m x m  y = 0;

 (10.607a, b)
150 Differential Equations and Applications

*standard CDLX. Singly- (doubly-) extended second-order linear differential

equations, with coefficient of the dependent variable (its derivative) a poly-
nomial of degree M(L), for example (10.608b) doubly extending the (10.576e)
cylindrical Bessel differential equation (note 9.38);
*standard CDLXI. The general integral of the doubly-extended Bessel dif-
ferential equation (10.608b) is a linear combination of doubly-extended
Bessel functions (10.608a) (note 9.38):

y ( x ) = J ± ν ( A1 ,..., AM ; B1 ,..., BM ; x ) :

 L
  M
 (10.608a, b)
x 2 y ′′ + x  1 +

∑
= 2
B x     y ′ +  x 2 − ν2 +
 
∑
m= 1
Am x m  y = 0.


Z. g. Invariant and Self-Adjoint Linear Second-Order

Differential Equations (notes 9.39–9.40):
*standard CDLXII. Transformation of a linear second-order differential equa-
tion (10.609a) through the transformation factor (10.609b) into a self-adjoint
(10.610a) [invariant (10.610b, c)] differential equation (note 9.39):

 x

y ′′ + P ( x ) y ′ + Q ( x ) y = 0, R ( x ) = exp 
 ∫ p ( ξ ) d ξ  :

(10.609a, b)

 P′ P2 
( R y ′ )′ + RQ y = 0 ; z = y R: z ′′ +  Q −
 2
+
4 
z = 0. (10.610a–c)

*standard CDLXIII. Self-adjoint (10.611a, b) [invariant (10.612a, b)] form of the

confluent hypergeometric differential equation (10.588c) (note 9.39):

y ( x ) = F (α ; γ ; x ): (e −x
)
x γ y ′ ′ − α e − x x γ y = 0, (10.611a, b)

1 α γ γ γ2 
z ( x ) = e − x/2 x γ /2 F ( α ; γ ; x ) : z ′′ −  + − − + z = 0 ; (10.612a, b)
 4 x 2x 2x
2
4 x 2 

*standard CDLXIV. Relation between the confluent hypergeometric function

of the first kind (10.588a) and the Whittaker function (10.613a) (note 9.40);
*standard CDLXV. General integral of the Whittaker differential equa-
tion (10.613b) with parameters (k,m) as a linear combination of Whittaker
functions (10.613a) for 2m not an integer (note 9.40):

 1 k m2 1 
y ( x ) = Wk , m ( x ) : y ′′ −  − + 2 − y = 0 ; (10.613a, b)
4 x x 4 x 2 
Classification 10.1 151

*standard CDLXVI. Relation between the Whittaker (10.613a) and the gener-
alized Bessel (10.575a) function (note 9.41);
*standard CDLXVII. Self-adjoint (10.614a, b) [invariant (10.615a, b)] form of the
generalized Bessel differential equation (10.575e) (note 9.41):

 ′
 µ x2    µ x2 
y ( x ) = J νµ ( x ) :  x exp  −
   y
4  
′ 2 2
 + x − ν exp  −

(
4 
y = 0 , ) (10.614a, b)

 −µ x 2  µ  ν2 1 µ µ x2  
z ( x ) = x exp  Jν ( x ): z ′′ + 1 − 2 + + 1−  z = 0.
 2   x 4x 2
2 8  
 (10.615a, b)

*standard CDLXVIII. Self-adjoint (10.616a, b) [invariant (10.617a, b)] form of

the Gaussian hypergeometric differential equation (10.592c) (note 9.44):

′
y ( x ) = F (α , β; γ ; x ):  x (1 − x )
 γ y ′  − α β x γ ( 1 − x )
α+β+ 1− γ 1+α+β− γ
y = 0;
 (10.616a, b)

(1+α+β− γ )/2
z ( x ) = x γ /2 ( 1 − x ) F (α, β; γ ; x ):

 γ ( 2 − γ ) 1 − ( α + β − γ )2 γ ( 1 + α + β − γ ) − 2 α β  (10.617a, b)
z ′′ +  + +  z = 0.
 4 x
2
4 1 − x2 ( )
2 x (1 − x ) 

*standard CDLXIX. Self-adjoint (10.618a, b) [invariant (10.619a, b)] form of the

hyperspherical-associated Legendre differential equation (10.596c) (note 9.42):

′  µ2 
(
 1 − x2 ) (
y ′  + 1 − x 2 )
1/2 +λ λ− 1/2
y ( x ) = Pνµ, λ ( x ) :  ν ( ν + 1) − 1 − x 2  y = 0 ;
 
 (10.618a, b)

( )
1/4 +λ/2
z ( x ) = 1 − x2 Pνµ, λ ( x ) :

 ν ν + 1 − 1 − λ/2 x 2 1 + λ/2 3 + λ/2 − µ 2  (10.619a, b)

z ′′ + 
( ) +
( )( )  z = 0.
( )
2
 1 − x2 1 − x2 
 

*standard CDLXX. Self-adjoint (10.620a, b) [invariant (10.621a, b)] form of the

ultraspherical Legendre differential equation (10.594d) (note 9.42):

′
(
y ( x ) = P ν , λ ( x ) :  1 − x 2 ) (
y ′  + ν ( ν + 1 + λ ) 1 − x 2 )
1/2 +λ λ− 1/2
y = 0 ; (10.620a, b)
 
152 Differential Equations and Applications

 ν ( ν + 1 + λ ) x 2 ( 1 + λ/2 )( 3 + λ/2 ) 
( )
1/4 +λ/2
z ( x ) = 1 − x2 P ν , λ ( x ) : z ′′ +  +  z = 0.
 1 − x2
 1 − x 2 2
( 
 )
 (10.621a, b)

*standard CDLXXI. Self-adjoint (10.622a, b) [invariant (10.623a, b)] form of the

Jacobi differential equation (10.598b) (note 9.42):

y ( x ) = I νa ,b ( x ) : { (1 − x ) 2 1+ a/2 + b/2
exp ( b − a ) arg tanh x  y ′ }′
( )
a/2 + b/2
+ 1 − x2 exp ( b − a ) arg tanh x  y = 0 ;
 (10.622a, b)

b−a 
( )
1/2 + a/4 + b/4
z ( x ) = 1 − x2 exp  arg tanh x  I νa ,b ( x ) :
 2 

 b − a 2 − 2x b − a 4 + a + b − x2 a + b + 2 a + b + 6
z ′′ 
−
( ) ( )( ) ( )( )
(10.623a, b)

 4 1− x 2 2
( )
a + b + 2 − ν ( ν + a + b + 1) 
+  z = 0.
1 − x2 

*standard CDLXXII. Self-adjoint (10.624a, b) [invariant (10.625a, b)] form of

the first Chebychev differential equation (10.599c) (note 9.43);

y ( x ) = Tν ( x ) : ( ′
)
1 − x 2 y ′ + ν2 1 − x 2 y = 0 ; (10.624a, b)

 1/2 + ν2 5  x  2 
( )
1/4
z ( x ) = 1 − x2 Tν ( x ) : z ′′ +  −    z = 0 ; (10.625a, b)
 1− x
2
4  1 − x2  

*standard CDLXXII. Self-adjoint (10.626a, b) [invariant (10.627a, b)] form of the

second Chebychev differential equation (10.600c) (note 9.43):

′′
(
 1 − x2 ) y ′  + ν ( ν + 2 )( 1 − x ) y = 0 ;
3/2
y ( x ) = U ν ( x ):
3/2
 (10.626a, b)


 3/2 + ν ( ν + 2 ) 21/9 
( )
3/4
z ( x ) = 1 − x2 U ν ( x ): z ′′ +  −  z = 0 ; (10.627a, b)
 1 − x2 1 − x 2  ( )
Classification 10.1 153

*standard CDLXXIV. Self-adjoint (10.628a, b) [invariant (10.629a, b)] form of

the associated Laguerre differential equation (10.602b) (note 9.43):

(e )
′
y ( x ) = Lµν ( x ) : − x 1+µ
x y ′ + ν e − x x 1+µ y = 0 ; (10.628a, b)

 1 + µ + ν 1 − µ2 
z ( x ) = e − x/2 x 1/2 +µ/2 Lµν ( x ) :      z ′′ +  +     z = 0. (10.629a, b)
 x 2 x2 

*standard CDLXXV. Self-adjoint (10.630a, b) [invariant (10.631a, b)] form of the

Hermite differential equation (10.603c) (note 9.43);

( ) ( )
′
y ( x ) = H ν ( x ):  exp − x 2 y ′  + 2 ν exp − x 2 y = 0 ;
  (10.630a, b)

 x2 
z ( x ) = exp  −  H ν ( x ) :
 2
( )
z ′′ + 1 + 2 ν − 4 x 2 z = 0. (10.631a, b)

Z. h. First Class of Transformed Generalized Bessel

Differential Equations (note 9.44):
*standard CDLXXVI. Solution (10.632a) of the first transformed generalized
Bessel differential equation (10.632b) (note 9.44);

( )  1 
y ( x ) = x a J νµ c x b : x 2 y ′′ +  1 − 2 a − µ b c 2 x 2 b  x y ′
 2 
 (10.632a, b)
 2 2 2 2 2 2b  µa 
+ a − b ν + b c x  1 +  y = 0.
  2 b  

*standard CDLXXVII. Particular case a = 0, b = 1, c = i of (10.632a, b) leading

to (10.633a, b):

 µ x2 
y ( x ) = J νµ ( i x ) : x 2 y ′′ +  1 +
 2 
(
x y ′ − x 2 + ν2 y = 0. ) (10.633a, b)

*standard CDLXXVIII. Particular case a = 0, b = 1/2, c = 1 of (10.632a, b) lead-

ing to (10.634a, b):

µx
y ( x ) = J νµ ( x ): 
x 2 y ′′ +  1 −
 4
1 2
4
2
(
 x y ′ + x − ν y = 0. ) (10.634a, b)
154 Differential Equations and Applications

*standard CDLXXIX. Particular case a = 0 , b = 1/2, c = 2 i, ν → 2 ν of (10.632a, b)

leading to (10.635a, b):

(
y ( x ) = J 2µν 2 i x : ) ( )
x 2 y ′′ + ( 1 + µ x ) x y ′ − x + ν2 y = 0. (10.635a, b)

*standard CDLXXX. Particular case a = 0 , c = 1, b = 2 of (10.632a, b) leading to

(10.636a, b):

( )
y ( x ) = J νµ x 2 : ( ) ( )
x 2 y ′′ + 1 − µ x 4 x y ′ + 4 x 4 − ν2 y = 0. (10.636a, b)

*standard CDLXXXI. Particular case a = ν = b , c = 1 of (10.632a, b) leading to

(10.637a, b):

 µ ν 2ν    µ  
( )
y ( x ) = x ν J νµ x ν : x 2 y ′′ +  1 − 2 ν −
 2  

x  x y ′ + ν2 1 − ν2 1 − x 2 ν  1 +    y = 0.
 2  

 (10.637a, b)

*standard CDLXXXII. Particular case a = ν, b = 1 = c of (10.632a, b) leading to

(10.638a, b):

 µ x2   µ ν
y ( x ) = x ν J νµ ( x ) : x y ′′ +  1 − 2 ν −  y ′ + x2  1 +  y = 0. (10.638a, b)
 2   2 

*standard CDLXXXIII. Particular case a = 1/2, b = 1, c = i q , ν = p − 1/2 of

(10.632a, b) leading to (10.639a, b):

µq 3 µ 
( )
y ( x ) = x J pµ− 1/2 i x q : x 2 y ′′ +
2



x y ′ −  p ( p − 1) + q  1 +  x 2  y = 0.
 4 
 (10.639a, b)

*standard CDLXXXIV. Particular case a = 1/2 + p = ν, b = 1, c = i q of (10.632b)

leading to (10.640a, b):

µq 2 µ
y ( x ) = x 1/2 + p J1/2
µ
(
+p i x q : ) 
x y ′′ −  2 p −
 2 

 2 
1 
x  y ′ − q 1 +  p +   x y = 0.
2 
 (10.640a, b)

*standard CDLXXXV. Particular case a = 1/2, b = p , c = i q /p , ν = 1/( 2 p ) of

(10.632b) leading to (10.641a, b):

 q p µ q 2 p−1  µ  2 p−2
y ( x ) = x J1/
µ
( 2 p )  i p x  : y ′′ + 2 p x y′ − q  1 + x y = 0. (10.641a, b)
   4 p 
Classification 10.1 155

*standard CDLXXXVI. Particular case a = 1/2, b = 3/2, c = 2/3, ν = 1/3 of

(10.632a, b) leading to (10.642a, b):

µ  2 3/2  µ 2  µ
y ( x ) = x J1/3  x  : y ′′ − x y ′ +  1 +  x y = 0. (10.642a, b)
3 3  6

*standard CDLXXXVII. Particular case µ = 0 of (10.642a, b) leading to

(10.643a, c), where (10.643c) is an Airy differential equation (10.566a) with
variable − x, and thus, (10.643b) of Airy functions (10.566b) of the same variable:

2 
y ( x ) = x J1/3  x 3/2  = c1 A i ( − x ) + c2 B i ( − x ) : y ′′ + x y = 0. (10.643a–c)
3 

Z. i. Second Class of Transformed Generalized Bessel Differential

Equations (note 9.45):
*standard CDLXXXVIII. Solution (10.644a) of the second transformed gener-
alized Bessel differential equation (10.644c) (note 9.45):

 f ′ g ′′ g ′ µ g ′g 
y ( x ) = f ( x ) J ν ( g ( x )) :
y ′′ −  2 + − − y′
 f g′ g 2  
(10.644a–c)
 f ′  f ′ g ′′ g ′ µ g ′g  f ′′ g′2 
+  2 + − −  − + g ′ 2 − ν2 2  y = 0.
 f  f g′ g 2  f g 

*standard CDLXXXIX. Particular case f = g/g ′ of (10.644a, b) leading to

(10.645a, b):

g ( x) µ µ g ′g
y ( x) = J ν ( g ( x )) : y ′′ + y′
g′ ( x) 2

 g ′′′ 3 g ′′ 2  1 2  g′
2
µ 
+ + g′2 −
4g′ 2
+ 

− ν 
 2
+ ( )
g ′′g − g ′ 2  y = 0.
 2g 4 g 4 
 (10.645a, b)

*standard CDLXC. Particular case f = g a of (10.644a, b) leading to (10.646a, b):

 g ′′ g ′ µ g ′g 
y ( x ) =  g ( x )  J νµ ( g ( x )) : + ( 2 a − 1) −
a
y ′′ +   y′
 g′ g 2 
(10.646a, b)
 g′2
 µa 
(
+  a 2 − ν2 ) g2
+ g′2  1 −
   y = 0.
2 

156 Differential Equations and Applications

*standard CDLXCI. Particular case (10.647a, b):

( )
f ( x ) = x a exp α xβ , g ( x ) = c xb , (10.647a, b)

of (10.644a, b) leading to (10.647c, d):

 µ b c 2 2b 
( ) ( )
y ( x ) = x a exp α xβ J νµ c xb : x 2 y ′′ +  1 − 2 a − 2 α β xβ −
 2
x  x y′


  µ a  2b  µ b c 2 2b  
+  a 2 − b 2 ν2 + b 2 c 2  1 +  x + α β ( 2 a − β ) xβ + α β xβ  α β xβ + x   y = 0,
  2b   2  
 (10.647c, d)

which simplifies to (10.632a, b) for α = 0.

*standard CDXCII. Particular case β = b in (10.647c, d) leading to (10.648a, b):

 µ b c 2 2b 
( ) ( )
y ( x ) = x a exp α x b J νµ c x b : x 2 y ′′ +  1 − 2 a − 2 α b x b −
 2
x  x y′


 µ b c2 
( )
+  a 2 − b 2 ν2 + α b ( 2 a − b ) x b + b 2 c 2 + α 2 x 2 b +
2
( )
a + α b x b x 2 b  y = 0.
 
 (10.648a, b)

*standard CDXCIII. Particular case β = b , α = ± i c of (10.647c, d), which is

equivalent to α = ± i c in (10.648a, b), leading to (10.648c, d):

 µ b c 2 2b 
( )
y ( x ) = x a exp ( ± i c x ) J νµ c x b : x 2 y ′′ +  1 − 2 a  2 i b c x b −
 2
x  x y′


 µ b c2 
+  a 2 − b 2 ν2 ± i b c ( 2 a − b ) x b +
2
( )
a ± i c b x b x 2 b  y = 0.
 
 (10.648c, d)

*standard CDXCIV. Particular case f = x − a , g = c exp ( bx ) of (10.644a, b) lead-

ing to (10.649a, b):

 µ b c 2 2 bx 
( )
y ( x ) = x − a J νµ c e b x : x 2 y ′′ +  2 a +
 2
x e  x y′

(10.649a, b)
   µ a   

+  a ( a − 1) − b 2  ν2 − c 2 e 2 bx  1 + 2
  x  y = 0.
 
  2 b 
 
Classification 10.1 157

*standard CDXCV. Particular case f = x , g = exp ( 1/x ) of (10.644a, b) leading

to (10.650a, b):

µ x 2 2/x  µ x  2/x 2
( )
y ( x ) = x J νµ e 1/ x : x 4 y ′′ −
2
e y ′ +  1 +
  2
 e − ν  y = 0. (10.650a, b)


*standard CDXCVI. Particular case g = x of (10.644a, b) leading to (10.651a, b):

1 f ′ µx
y ( x ) = f ( x ) J νµ ( x ) : y ′′ +  − 2 + y′
 x f 2 
(10.651a, b)
 ν2 f ′  f ′ 1 µ x  f ′′ 
+ 1 − 2 +  2 − − −  y = 0.
 x f  f x 2  f 

( )
a
*standard CDXCVII. Particular case of (10.651a, b) with f = x 2 − 1 leading
to (10.652a, b):

 µ x2 2 
( ) ( ) ( ) ( )
a 2
y ( x ) = x 2 − 1 J νµ ( x ) :  x 2 x 2 − 1 y ′′ + ( 1 − 4 a ) x 2 − 1 + x − 1  ≠ x x2 − 1 y ′
 2 

( )( ) ( )
+  x 2 − ν2 x 2 − 1 + 4 a x 2 − µ a x 4 x 2 − 1  y = 0.
2

 
 (10.652a, b)

*standard CDXCVIII. Particular case f = sec x of (10.651a, b) leading to

(10.653a, b):

 µ x2 
y ( x ) = sec x J νµ ( x ) : x 2 y ′′ +  1 − 2 x tan x + x y′
 2 
(10.653a, b)
 2  µx 2 

−  ν + x tan x  1 +  y = 0.
  2  
158 Differential Equations and Applications

*standard CDXCIX. Particular case f = csc x of (10.651a, b) leading to (10.654a, b):

 µ x2 
y ( x ) = csc x J νµ ( x ) : x 2 y ′′ +  1 − 2 x cot x + x y′
 2 
(10.654a, b)
 2  µ x2  
−  ν + x cot x  1 +  y = 0.
  2  

*standard D. Particular case (10.655a, b) of (10.651a, b) leading to (10.655c):

 x

∫ j ( ξ ) d ξ  = y ( x )  J νµ ( x )  :     
−1
y ( x ) = exp 
 

 µ x2   µ x3 j 
       x 2 y ′′ +  1 − 2 x j +
 2   ( )
x y ′ +  x 2 1 + j 2 − j ′ − x j − ν2 −
2 
y = 0.

(10.655a–c)
Classification 10.2

500 Applications of Differential Equations

to Physical and Engineering Problems
Classification 10.2 consists of the detailed application of differential equa-
tions to 500 physical and engineering problems leading to: (i) 100 distinct
differential equations, all falling into at least one of the 500 standards of the
recollection 10.1; (ii) 500 subcases of solutions depending on the specified
application or combination of parameters. By detailed application is meant
a three-stage process: (i) formulation of the model leading to a differential
equation; (ii) solution of the differential equation with sufficient generality;
(iii) interpretation of the results and assessment of practical consequences.
The 500 problems in Classification 10.2 are divided in 26 sections A to Z, and
form eight groups. The first group, consisting of section A and problems 1 to
69, concerns the linear second-order oscillator with damping or amplifica-
tion and several types of forcing, including ordinary resonance. The second
group, corresponding to section B and problems 70 to 89, concerns the cal-
culation of trajectories and paths, which underlies other groups. The third
group, corresponding to section C and problems 90 to 104, concerns the lin-
ear second-order oscillator with parameters depending on time including
parametric resonance. The fourth group, consisting of sections D to H and
problems 105 to 155, concerns non-linear second-order oscillators, including
non-linear resonance, bifurcations, and chaotic motions. The consideration
of second-order problems concludes with numerical and analytical approxi-
mate methods in the fifth group, consisting of sections I and J and problems
156 to 177.
The deformation and buckling of elastic bodies like bars, beams, and plates
lead to fourth-order differential equations and are considered in the sixth
group, consisting of sections K to Q and problems 178 to 338. The seventh
group, consisting of sections R to T and problems 339 to 420, concerns linear
and non-linear waves in inhomogeneous and unsteady media; the solutions
are obtained for linear, steady, and inhomogeneous media, when the partial
differential equations reduce to ordinary differential equations with variable
coefficients. The eighth group, consisting of the sections U to Z and problems
421 to 500, concerns multidimensional oscillators with several degrees-of-
freedom leading to simultaneous systems of ordinary differential equations
describing combined oscillations and multiple resonance. Both the seventh
and eighth groups concern waves and multidimensional oscillators, which
involve several modes, infinite and finite in number. Four types of resonance
are considered: simple, multiple, parametric, and non-linear, each with or

159
160 Differential Equations and Applications

without damping. The problems often have mechanical electric, acoustic,

and other analogues. Classifications 10.1 and 10.2 provide a summary of the
contents of this book and complement each other as a quick-look guide of
where to find: (a) the solution of a specific differential equation among the
500 standards in Classification 10.1; (b) the indication of some contexts in
which the differential equation arises and the interpretation and application
of the solutions in Classification 10.2.

A. Linear, Attractive/Repulsive, Damped/

Amplified, and Free/Forced Oscillator:

d2 x dx
2
+ 2λ + ω 20 x = f ( t ) . (10.656)
dt dt

*problem 1. Mechanical mass-damper-spring-force system (subsections

2.1.1–2.1.5):

m x + µ x + k x = Fm ( x ) ; (10.657)

*problem 2. Electrical self-resistor-capacitor-battery circuit (subsections

2.1.6–2.1.13):

1
L q + R q + q = Fe ( x ) . (10.658)
c

A. a.  Undamped and Unforced (section 2.2); λ = 0 = f ( t ) :

*problem 3. Harmonic oscillator (subsections 2.2.1–2.2.2; note 1.19); stable
attractor: ω 20 > 0;
*problem 4. Adiabatic invariant (subsection 2.2.3): 2 π dω 0 /dt << ω 20 ;
*problem 5. Small oscillations of suspended pendulum (subsection 2.2.4):
 = − g θ;
θ
*problem 6. Monotonic instability of an inverted pendulum (subsections
2.2.4–2.25):  θ = g θ ;
*problem 7. Unstable repeller (subsection 2.2.5): ω 20 < 0.

A. b.  Damped and Unforced (sections 2.3–2.4): λ > 0 = f ( x ) :

*problem 8. Critical damping (subsections 2.3.1–2.3.2; note 1.29): λ = ω 0;
*problem 9. Supercritical damping (subsections 2.3.3–2.3.4; note 1.21): λ > ω 0;
Classification 10.2 161

*problem 10. Subcritical damping (section 2.4, note 1.22, E10.2.1–E10.2.2);

λ < ω0 ;
*problem 11. Weak damping (subsections 2.4.2–2.4.5; E10.1.4); λ 2 << ω 20.

A. c.  Amplified and Unforced (section 2.5):

*problem 12. Mechanical system on a conveyor belt (subsection 2.5.1):

x + ( λ − ν ) x + ω 20 x = 0. (10.659)

*problem 13. Zero effective damping (subsection 2.5.2): λ = ν;

*problem 14. Critical effective damping (subsection 2.5.2): λ + ν = ω 0;
*problem 15. Supercritical effective damping (subsection 2.5.2): λ + ν > ω 0;
*problem 16. Subcritical effective damping (subsection 2.5.2): λ < ω 0 + ν;
*problem 17. Weak effective damping (subsection 2.5.2): λ ( λ − 2 ν ) << ω 20 − ν2 ;
*problem 18. Critical amplication (subsection 2.5.3): λ = ν − ω 0 > 0;
*problem 19. Supercritical amplification (subsection 2.5.3): λ < ν − ω 0;
*problem 20. Subcritical amplification (subsection 2.5.3): λ > ν − ω 0;
*problem 21. Froude pendulum suspended from a cam with frictional torque
(subsection 2.5.4);
*problem 22. Electric circuit with valve (subsection 2.5.5).

A. d.  Degenerate Cases (section 2.6):

*problem 23. Zero mass and damping (subsection 2.6.1): m = 0 < µ;
*problem 24. Zero mass and amplification (subsection 2.6.1): m = 0 > µ:
*problem 25. Zero mass and zero damping (subsection 2.6.1): m = 0 = µ ;
*problem 26. No spring and no damper (subsection 2.6.1): k = 0 = µ;
*problem 27. Small mass (subsections 2.6.1–2.6.2): m k << µ 2.

A. e.  Undamped, with Sinusoidal Forcing (section 2.7):

λ = 0: x + ω 20   x = f a cos ( ω a t ) . (10.660a, b)

*problem 28. Non-resonant forcing (subsections 2.7.1–2.7.2): ω a ≠ ω 0;

*problem 29. Resonant forcing (subsections 2.7.3–2.7.4, E10.2.3–E10.2.4):
ωa = ω0 ;
*problem 30. Beats (subsections 2.7.5–2.7.6): ( ω 0 − ω a ) << ( ω 0 + ω a ) .
2 2

A. f.  Damped, with Sinusoidal Forcing (subsections 2.8.1–2.8.3):

λ ≠ 0: x + 2 λ x + ω 20 x = f a cos ( ω a t ) . (10.661a, b)

*problem 31. Non-resonant forcing (subsections 2.8.1–2.8.3): ω a ≠ ω 0;

*problem 32. Resonant forcing (subsection 2.8.2): ω a = ω 0.
162 Differential Equations and Applications

A. g. Simultaneous Forcing (10.662b) in Frequency and Damping

(subsections 2.8.4–2.8.6) for subcritical damping (10.662a):

λ < ω0 : x + 2 λ x + ω 20 x = f a exp ( −λ a t ) cos ( ω a t ) . (10.662a, b)

*problem 33. Distinct applied and oscillation (10.662c) frequencies and dis-
tinct applied decay and natural damping (subsection 2.8.4): ω ≠ ω a , λ ≠ λ a ;

1/2
ω = ω 20 − λ 2 . (10.662c)

*problem 34. Monotonically decaying forcing (subsection 2.8.4): ω a = 0;

*problem 35. Coincident applied and oscillation frequencies and distinct
applied decay and natural damping (subsection 2.8.5): ω a = ω , λ a ≠ λ;
*problem 36. Distinct applied and oscillation frequencies and coincident
applied decay and natural damping (subsection 2.8.5): ω a ≠ ω , λ a = λ ;
*problem 37. Coincident applied and frequencies and applied decay equal to
natural damping (subsection 2.8.5): ω a = ω 0 , λ a = λ ;
*problem 38. Resonance only if (i) the applied and oscillation frequencies
coincide and (ii) the applied decay equals the natural damping (subsection
2.8.6): ω a = ω , λ a = λ.

A. h.  Forcing Involving a Power of Time (example 10.3):

x + 2 λ x + ω 20 x = f a t n cos ( ω a t ) . (10.663)

*problem 39. Undamped and non-oscillating (E10.3.1): λ = 0 = ω a;

*problem 40. Linear in time, undamped, oscillating, and non-resonant
(E10.3.1): n = 1, λ = 0 ≠ ω a ≠ ω 0 ;
*problem 41. Problem 40 but resonant (E10.3.1): n = 1, λ = 0 ≠ ω a = ω 0 ;
*problem 42. Linear in time, damped, non-oscillating (E10.13.2): n = 1, λ ≠ 0 = ω a;
*problem 43. Linear in time, damped, oscillating (E10.13.2): n = 1, λ ≠ 0 ≠ ω a ≠ ω 0;
*problem 44. Problem 43 but resonant (E10.13.3): n = 1, λ ≠ 0 ≠ ω a = ω 0 .

A. i.  Integrable Forcing Function (section 2.9): f ∈E (|R ) in (1):

*problem 45. Critical damping (subsections 2.9.1–2.9.3): λ = ω 0;
*problem 46. Undamped (subsection 2.9.3): λ = 0 ≠ ω 0;
*problem 47. Supercritical damping (subsection 2.9.3): λ > ω 0;
*problem 48. Subcritical damping (subsection 2.9.3); λ < ω 0;
*problem 49. Constant forcing (subsection 2.9.4): f ( t ) = f0;
*problem 50. Bounded forcing (subsection 2.9.5): f ∈B (|R ).
Classification 10.2 163

A. j. Forcing by a Discrete Spectrum (10.664b) of a Function

of Bounded Oscillation (10.664a) in a Finite Interval
(10.664a) Represented by a Fourier Series (note 2.4):

+∞
nπt 
f ∈F ( 0, T ) : x + 2 λ x + ω 20 x   = ∑ f cos 
n =−∞
n
T 
 ≡ f ( t ) . (10.664a, b)

*problem 51. Undamped, non-resonant if no harmonic coincides with the

natural frequency: λ = 0 , ωT ≠ n π for all n;
*problem 52. Undamped, resonant if one harmonic coincides with the natu-
ral frequency λ = 0, ω T = m π for one m;
*problem 53. Damped, non-resonant: λ > 0, ωT ≠ n π ;
*problem 54. Damped, resonant: λ > 0 , ω T = m π .

A. k. Forcing by the Continuous Spectrum (10.665b) of a Function of

Bounded Oscillation and Absolutely Integrable on the Real Line
(10.665a) Represented by a Fourier integral (notes 2.5–2.8):
+∞
1
f ∈F ∩ L (−∞ , + ∞) : x + 2 λ x + ω 20 x =
∫ f (ω ) e
−∞
iωt
dω ≡ f ( t ) . (10.665a, b)

*problem 55. Undamped (note 2.5): λ = 0;

*problem 56. Critically damped oscillator (note 2.6): λ = ω 0;
*problem 57. Supercritically damped oscillator (note 2.8): λ > ω 0;
*problem 58. Subcritically damped oscillator (note 2.8): λ < ω 0.

A. l. Influence or Green Function for Forcing by

the Unit Impulse (notes 2.9–2.10):

G + 2 λ G + ω 20 G = δ ( t ) . (10.666)

*problem 59. For the critically damped oscillator (note 2.9): λ = ω 0;

*problem 60. For the subcritically damped oscillator (note 2.9): λ < ω 0;
*problem 61. For the supercritically damped oscillator (note 2.9): λ > ω 0;
*problem 62. For the undamped oscillator (note 2.9): λ = 0.

A. m. Forcing by a Continuous Function (note 2.11):

*problem 63. Using the Green or influence function (note 2.11): f ∈C (|R ) in (1);
*problem 64. Undamped (note 2.11): λ = 0;
*problem 65. Subcritically damped (note 2.11): 0 < λ < ω 0;
*problem 66. Critically damped (note 2.11): λ = ω 0 > 0;
*problem 67. Supercritically damped (note 2.11): λ > ω 0 > 0.
164 Differential Equations and Applications

A. n.  Forcing by a Gaussian Function (notes 2.12–2.13):

*problem 68. Undamped forcing by Gaussian (note 2.12):

 b2 
x + 2 λ x + ω 20 x = f a exp  − 2  (10.667)
 T 

*problem 69. Undamped forcing by Gaussian random noise with mean µ and
variance σ 2 (note 2.13):

 ( t − µ ) 
2
fa
x + 2 λ x + ω 20 x = exp −  . (10.668)
σ 2π  2 σ 2 

B. Motion Near a Stagnation Point

(sections 4.1–4.2; example 10.7):
B. a. Paths Near a Stagnation Point of the First Degree (section 4.1 and
example 10.7):

dy c x + d y
ad ≠ bc: = . (10.669a, b)
dx a x + b y

*problem 70. Contact point of inflexion of paths tangent to the principal line
(subsection 4.1.4) for:

∆ ≡ ( a − d ) + 4 c b = 0 ; (10.670)
2

*problem 71. Saddle point through which pass two principal lines, to which
the paths are asymptotically tangent (subsection 4.1.6): ∆ > 0, b c > a d;
*problem 72. Nodal point through which pass all paths tangent to a princi-
pal line (subsection 4.1.7): ∆ > 0 in (10.670) and:

ad − bc
ϑ=4 ≠ 1, 0. (10.671)
(a + d + ∆ )
2

*problem 73. Focal point through which pass all straight paths (subsec-
tion 4.1.7): ϑ = 1;
*problem 74. Simple point through which passes one of the set of parallel
straight paths (subsection 4.1.7): ϑ = 0;
*problem 75. Common center of elliptical paths (subsection 4.1.8):
∆ < 0, a + d = 0;
*problem 76. Asymptotic point as the limit of spiral paths (subsection 4.1.9):
∆ < 0, a + d ≠ 0.
Classification 10.2 165

B. b.  Paths Near a Stagnation Point of the Second Degree (example 10.7):
*problem 77. Paths tangent to the y-axis at the origin (E10.7.1):

dy y 2 − x 2
= ; (10.672)
dx xy
*problem 78. Paths with three asymptotes crossing at the origin with equal
angles (E10.7.2):

dy x 2 − y 2
2 = . (10.673)
dx xy

B. c.  Trajectories Near a Stagnation Point of First Degree (section 4.2):

dx dy
= a x + b y, = c x + d y . (10.674a, b)
dt dt

*problem 79. Equivalent damped (1) unforced f = 0 oscillator (subsections

4.2.1–4.2.3):

a+d  d2 d 
λ=− , ω 20 = a d − b c:  2 − 2λ + ω 20  x , y ( t ) = 0, (10.675a–c)
2  dt dt 

with (19a) damping λ > 0 (amplification λ < 0) and (19b) natural frequency
ω 20 > 0 (growth rate ω 20 < 0) in (10.675c);
*problem 80. Critical damping corresponding to (problem 70) to a contact
point of inflection (subsection 4.2.4): λ = ω 0;
*problem 81. Supercritical damping corresponding to (problem 72) to a nodal
point (subsection 4.2.5): λ > ω 0 , 0 < r s ≠ 0:

r , s = − λ ± λ 2 − ω 20 ; (10.676a, b)

*problem 82. Supercritical damping corresponding (problem 73) to a focal

point (subsection 4.2.5): λ > ω 0 , r = 0 ≠ s ;
*problem 83. Supercritical damping corresponding (problem 74) to a simple
point (subsection 4.2.5): λ > ω 0 , s = 0 ≠ r ;
*problem 84. Supercritical damping corresponding to a saddle point (sub-
section 4.2.5): λ > ω 0 . r s < 0;
*problem 85. Subcritical damping corresponding (problem 76) to a spiral
fall towards λ > 0 (exist from λ < 0) an asymptotic point (subsection 4.2.6):
λ ≠ 0, λ < ω 0 ;
*problem 86. No damping corresponding (problem 75) to the center of ellip-
tic or circular trajectories (subsection 4.2.6): λ = 0.
166 Differential Equations and Applications

B. d.  Two Orthogonal Undamped Oscillators (subsections 4.2.7–4.2.8):

x ( t ) = a cos ( ω t ) , y ( t ) = b cos ( ω t − ϕ ) . (10.677a, b)

*problem 87. Straight path if in-phase (subsection 4.2.7): ϕ = 0 ;

*problem 88. Elliptic path aligned with the axis when out-of-phase (subsec-
tion 4.2.7): ϕ = π/2;
*problem 89. Elliptic path rotated relative to the axis for intermediate phases
(subsection 4.2.8): 0 < ϕ < π/2.

C.  Parametric Resonance (section 4.3)

C. a. Undamped Oscillator with Coefficients Depending
on Time (subsection 4.3.1):
*problem 90. Oscillator with mass and spring resilience a function of time:

d  dx 
m ( t )  + k ( t ) x ( t ) = 0. (10.678)
dt  dt 

C. b. Parametric Resonance: Harmonic Oscillator with Support

Oscillating in Time with Amplitude h and Excitation Frequency ω e :

d2 x
+ ω 20 1 + 2 h cos ( ω e t )  x ( t ) = 0. (10.679)
dt 2

*problem 91. Exact solution in terms of Mathieu functions (subsections

4.3.2–4.3.11):

2
ωe t  2 ω0 
w (θ) ≡ x (t ) , θ ≡ , a≡ , q ≡ − h a : (10.680a–d)
2  ω e 

d2 w
+  a − 2 q cos ( 2θ )  w ( θ ) = 0. (10.680e)
dθ2 

C. c. Stability/Instability/Indifference of
Solutions (subsections 4.3.3–4.3.6):
*problem 92. For linear unforced second-order differential equation (10.681e)
with (10.681a, b) periodic coefficients (subsections 4.3.3–4.3.5):

n = 0, 1, 2: An ( t ) = An ( t + τ ) , ∑ n= 0
An ( t )
dx x
dt n
= 0; (10.681a–c)
Classification 10.2 167

*problem 93. For undamped harmonic oscillator (10.682b) with (10.682a) peri-
odic “natural frequency” (subsection 4.3.6):

d2 x
ω (t ) = ω (t + τ ): + ω ( t )  x ( t ) = 0.
2
(10.682a, b)
dt 2 

C. d. Excitation Frequencies Leading to Parametric

Resonance (subsections 4.3.10–4.3.11):
*problem 94. Strongest parametric resonance for ω e = 2 ω 0 + 2 ε excitation fre-
quency close to twice the natural frequency (subsection 4.3.10);
*problem 95. Weaker parametric resonance for excitation frequiency (10.683c)
close to a multiple or submultiple (10.683a, b) of the natural frequency (subsec-
tion 4.3.11):

m±1
m, n = 1, 2,...: ω e± =
m
( ω 0 + ε ) , (10.683a–c)
including excitation frequency equal to the natural frequency
( n = 1, m = 2 : ω +
)
e = ω0 .

C. e. Undamped Parametric Resonance (subsections 4.3.12–

4.3.15) by Approximate Solution of (10.679) to the
Lowest Orders O(h n ) in the Excitation Amplitude h:
*problem 96. Parametric resonance by excitation close to the first harmonic
ω e = 2 ω 0 + 2 ε of the natural frequency to the lowest or the first O(h) order in
the excitation amplitude (subsection 4.3.13);
*problem 97. Problem 96 to the next or second order in the excitation ampli-
tude (subection 4.3.14);
*problem 98. Parametric resonance by excitation close to the natural fre-
( )
quency ω e = ω 0 + ε with the second O h2 as the lowest order in the excitation
amplitude (subsection 4.3.15).

C. f. Parametric Resonance with Weak Damping (10.684a, b) for

Several Excitation Frequencies (subsections 4.3.16–4.3.18):

2
 ω h d2 x dx
λ 2 << b 2 ≡  0  : + 2λ + ω 20 1 + 2 h cos ( ω e t )  x ( t ) = 0. (10.684a, b)
 2  dt 2
dt

*problem 99. Strongest parametric resonance for excitation frequency

ω e = 2 ω 0 + 2 ε close to twice the natural frequency (subsections 4.3.16–4.3.17);
168 Differential Equations and Applications

*problem 100. Undamped λ = 0 parametric resonance due to excitation

2 n ω e = ω 0 + ε close to submultiples of the first harmonic of the natural fre-
quency (subsection 4.3.18);
*problem 101. Problem 100 with weak (10.684a) damping (subsection 4.3.18).

C. g. Suspended Pendulum with Support Oscillating Vertically (10.685c)

with Amplitude p Leading to Weakly Damped (10.685a) Small
Amplitude (10.685b) Parametric Resonance (subsection 4.3.19):

g d2θ dθ g  8p 
λ 2 << , θ2 << 1: + 2λ + 1+ cos ( ω e t )  θ ( t ) = 0. (10.685a–c)
L dt 2 dt L  L 

*problem 102. Undamped λ = 0 excitation at ω e = 2 ω 0 at twice the natural

frequency;
*problem 103. Problem 102 with weak (10.685a) damping;
*problem 104. Problem 102 with weak damping and excitation ω e = n ω 0 at a
submultiple of the first harmonic of the natural frequency.

D.  Non-Linear, Undamped, Unforced, Oscillator

(10.686a) hence, constant (10.686b) total energy (sections 4.4–4.5):

2
d2θ dΦ 1  dx 
m = F ( x) = − : m   + Φ ( x ) = E . (10.686a, b)
dt 2 dx 2  dt 

D. a.  For general potential (subsections 4.4.2–4.4.3):

*problem 105. Paths in the phase plane and trajectories in the physical plane
(subsection 4.4.2);
*problem 106. Separatrices between limit cycles and unstable regimes (sub-
section 4.4.3).

D. b. Bi-Quadratic Potential for a Soft/Hard Spring

(subsections 4.4.5–4.4.15):

1 2 β 
Φ( x) = k x  1 + x 2  , (10.687)
2  2 
Classification 10.2 169

*problem 107. Non-linear hard β > 0 spring (subsections 4.4.4–4.4.6);

*problem 108. Non-linear soft β < 0 spring (subsections 4.4.6–4.4.8);
*problem 109. Exact solution for the position as a function of time in terms of
the Jacobian elliptic sine (subsection 4.4.9);
*problem 110. Inverse solution with time as a function of position specified
by a series of circular sines (subsections 4.4.10–4.4.11);
*problem 111. Inverse solution with time as a function of position speci-
fied by a series of Chebychev polynomials of the second kind (subsections
4.4.11–4.4.12);
*problem 112. Effect of non-linearity on the oscillation frequency (subsection
4.4.13);
*problem 113. Effect of non-linearity on the displacement and generation of
harmonics (subsections 4.4.14–4.4.15).

D. c. Quartic Potential (10.688) Generalizing the Bi-Quadratic (10.687)

Potential (section 4.5):

1 2 2 1 
Φ( x) = k x  1 + ε α x + ε 2β x 2  . (10.688)
2  3 2 

*problem 114. Method of parametric expansions for the displacement (10.689a)

and frequency (10.689b) to second order (subsection 4.5.1):

x ( t ) = x1 ( t ) + ε x2 ( t ) + ε 2 x3 ( t ) + ..., ω = ω 0 + ω 1 ε + ω 2 ε 2 + ...; (10.689a, b)

*problem 115. First-order perturbation 0 ( ε ) of displacement (subsection 4.5.2);

( )
*problem 116. Second-order perturbation 0 ε 2 of displacement and oscilla-
tion frequency (subsections 4.5.3–4.5.4).

E.  Resonance of a Non-Linear Oscillator

with (i) linear damping, (ii) non-linear restoring force corresponding to the
quartic potential, and (iii) sinusoidal forcing with amplitude Fa and applied
frequency ω a (section 4.6):

d2 x dx
m
dt 2
+µ
dt
( )
+ k x 1 + ε α x + ε 2 β x 2 = Fa cos (β ω a t ) . (10.690)
170 Differential Equations and Applications

*problem 117. Non-linear resonance with applied frequency close to the nat-
ural frequency ω a = ω 0 + ε leading to the existence of amplitude jumps and
hysterisis loop (subsections 4.6.1–4.6.3);
*problem 118. Problem 117 for subharmonic non-linear resonance with
applied frequency close to half of the natural frequency 2 ω a = ω 0 + ε
leading to smaller amplitude jumps and weaker hysteresis (subsections
4.6.4–4.6.5);
*problem 119. Non-linear resonance with applied frequency ω a = 2 ( ω 0 + ε )
close to the first harmonic of the natural frequency leading to smooth ampli-
tude variation, or amplitude jumps or suppressions of oscillations (subsec-
tions 4.6.6–4.6.7);
*problem 120. Non-linear resonance with applied frequency ω e = 3 ( ω 0 + ε )
close to the second harmonic of the natural frequency leading to no oscilla-
tion, or suppression of oscillations or a threshold amplitude for the existence
of oscillations (subsections 4.6.8–4.6.9);
*problem 121. Non-linear resonance at rational multiples of the natural fre-
quency, that is, submultiples of harmonics of the natural frequencym (sub-
section 4.6.10).

F.  Electromechanical Dynamo (subsections 4.7.1–4.7.9):

F. a.  Electrical Dynamo:

dJ ZΩ
L +RJ = ± J . (10.691)
dt 2π

*problem 122. Counter-rotating electrical dynamo with – (minus) sign in

(10.691) as a brake (subsection 4.7.2);
*problem 123. Co-rotating electrical dynamo with + (plus) sign in (10.691)
with constant, decaying, or growing current (subsection 4.7.2).

F. b. Electromechanical Dynamo (10.692b) with small resistance (10.692a),

compared with the mutual induction (subsections 4.7.3–4.7.4):

2 π R << Z Ω : L J = − R J ± S J 2 . (10.692a, b)

*problem 124. Exponential decay of electric current and angular velocity;

*problem 125. Linear decay if the first term on the r.h.s. of (10.692b) is omitted.
Classification 10.2 171

F. c. Exact Electromechanical Homopolar

Dynamo (subsections 4.7.5–4.7.6):

d2 J  1 d J  dJ R M
− ± b J bJ J + = 0. (10.693)
2
dt  J d t  dt L a L 

*problem 126. Without driving torque M = 0 the electric current and angular
velocity decay to zero (subsection 4.7.5);
*problem 127. Balance of driving and Lorentz force torques M Ω = R J 2 leads
to constant electric current and angular velocity (subsection 4.7.6).

F. d.  Self-Excited Dynamo (subsection 4.7.7):

dv±
=  b v± − a ( v± ) . (10.694)
2
m
dt

*problem 128. For upper sign in (10.694) linear and quadratic damping lead-
ing to decay;
*problem 129. For lower sign in (10.694) linear amplification and quadratic
damping leading to a steady state.

G.  Circular Pendular Motion (subsections 4.7.10–4.7.18)

G. a.  Large Motion without Damping (subsections 4.7.10–4.7.15):

g  + ω 02 sin θ = 0.
ω0 = : θ (10.695a, b)
L

*problem 130. Stopping point at the highest position for E = m g L (subsec-

tions 4.7.11–4.7.12);
*problem 131. Circulatory motion for E > m g L(subsections 4.7.11–4.7.12);
*problem 133. Oscillatory motion with large amplitude for E < m g L
(subsections 4.7.11, 4.7.12).

G. b. Pendular Motion with Quadratic Damping

(subsections 4.7.16–4.7.18):

µ > 0:  + µ θ θ + m g sin θ = 0.
mLθ (10.696a, b)

*problem 134. Decay of oscillations (subsection 4.7.18);

172 Differential Equations and Applications

*problem 135. Start at the highest position (subsection 4.7.18);

*problem 136. Number of circulations before damping causes decay to oscil-
latory motion (subsection 4.7.18).

H. Bifurcations, Chaos, and Approximations

(sections 4.8–4.9; notes 4.1–4.13):
H. a.  Bifurcations of Dynamical Systems (subsections 4.8.1–4.8.9):

x = f ( x ; q ) . (10.697)

*problem 137. Stability and instability at boundaries (subsection 4.8.2).

H. b. Hopf Bifurcation: Gradient Dynamical System

Perturbing a Spiral Flow (subsections 4.8.4–4.8.5):
dr

dt
( )
= r q + µ r 2 . (10.698)

*problem 138. Hopf subcritical bifurcation: µ < 0;

*problem 139. Hopf supercritical bifurcation: µ > 0.

H. c. Mixed Non-Linear Oscillator—van der Pol (subsections 4.8.6–4.8.7):

d2 x dx
+ γ h( x) + ω 20 x = 0. (10.699)
dt 2 dt
*problem 140. Weak γ << 1 damping (subsection 4.8.6);
*problem 141. Strong γ >> 1 damping (subsection 4.8.7).

H. d.  Polynomial Potentials and Catastrophes (subsections 4.8.8–4.8.9):

*problem 142. Fold catastrophe (10.700b) for (10.700a) a cubic potential (subsec-
tion 4.8.8):

x3 d2 x ∂Φ
Φ ( x; a) = − a x: 2
=− = a − x 2 ; (10.700a, b)
3 dt ∂x

*problem 143. Cusp catastrophe (10.701b) for (10.701a) a quartic potential

(subsection 4.8.9):

x 4 a x2 d2 x ∂Φ
Φ ( x; a) = − + − b x: =− = x 2 − a x + b . (10.701a, b)
4 2 dt 2 ∂x
Classification 10.2 173

H. e.  Chaotic Motions (subsections 4.8.10–4.8.14):

*problem 144. Poincaré maps and classification of singularities (subsection
4.8.10);
*problem 145. Passage from a deterministic to a choatic system (subsection
4.8.11);
*problem 146. Transition from a laminar to a turbulent flow (subsection
4.8.12);
*problem 147. Aircraft stall and departure into spin (subsections 4.8.13–4.8.14).

H. f.  Limit Cycles (subsections 4.9.1–4.9.3):

*problem 148. Competing populations of preys and predators (subsections
4.9.1–4.9.2):

dx dy
= ax − b x y , = − c y + d x y . (10.702a, b)
dt dt

*problem 149. Non-existence of a limit cycle (subsection 4.9.3):

∂y ∂v
{u, v} ≡ 
dx dy  1
,  ∈C ( D ) : + ≠ 0. (10.703a, b)
 dt dt  ∂x ∂y

H. g.  Stability of Orbits of Dynamical Systems (subsection 4.9.4):

*problem 150. General dynamical system:

dx dy
u, v ∈C 2 (|R ) : = u ( x, y ) , = v ( x , y ) ; (10.704a–c)
dt dt

*problem 151. Second-order system (10.656) with constant coefficients

(10.705a, b):

dx dv
= v, = −2 λ v − ω 20 x . (10.705a, b)
dt dt

H. h.  Index of a Singularity (subsections 4.9.5–4.9.10):

*problem 152. Index of a domain with several singularities (subsection 4.9.6);
*problem 153. Dynamical system at infinity (subsection 4.9.7);
*problem 154. Index at infinity (subsection 4.9.8);
*problem 155. Classification of singularities by their índices (subsection 4.9.9);
174 Differential Equations and Applications

I.  Numerical Solution of Differential Equations (notes 4.1–4.13):

*problem 156. Discretization of derivatives of any positive integer order
(note 4.8);
*problem 157. Transformation of a differential equation into a finite differ-
ence equation (note 4.9);
*problem 158. Comparison of the solutions of the differential and finite
difference equations (notes 4.10–4.11);
*problem 159. Effect of step size on accuracy (note 4.12);
*problem 160. Comparison of truncation and discretization errors (note 4.13).

J.  Optics and Acoustic Rays (notes 5.1–5.19)

J. a.  Linear Second-Order Differential Equation (notes 5.3–5.5):

d2 y dy
A2 ( x ) 2
+ A1 ( x ) + A0 ( x ) y ( x ) = B ( x ) . (10.706)
dx dx

*problem 161. Non-singular form (note 5.3):

d2 y dy
2
+ P ( x) + Q ( x ) y = f ( x ) . (10.707)
dx dx

*problem 162. Given a particular integral (10.708a) of the unforced equation

(10.706) with B ( x ) = 0 , find another (10.708b) linearly independent (10.708c)
particular integral (notes 5.4–5.5):

y1′′ + P y1′ + Q y1 = 0:       y 2′′ + P y 2′ + Q y 2 = 0 ∧ y 2 ( x ) ≠ C y1 ( x ) ; (10.708a–d)

*problem 162. Idem (10.708a) the general integral (10.709c) of the unforced
(10.709a, b) equation (notes 5.4–5.5):

y1′′ + P y1′ + Q y1 = 0; B ( x ) = 0: y ( x ) = C1 y1 ( x ) + C2 y 2 ( x ) . (10.709a–c)

*problem 163. Idem (10.710a) the complete integral (10.710c) of the forced
(10.710b) equation (notes 5.4–5.5):

y1′′ + P y*′ + Q y* = f , B ( x ) ≠ 0: y ( x ) = y ( x ) + y* ( x ) . (10.710a–c)

Classification 10.2 175

J. b.  Three Alternative Forms (notes 5.6–5.8):

*problem 164. Transformation to an integro-differential equation (note 5.6):

du  x


dx
+ u2 + P u + Q = F ( x ) exp −
 ∫ u ( ξ ) d ξ  ; (10.711)


*problem 165. Transformation to a self-adjoint form (note 5.7):

du  dy 
 R ( x) + Q ( x ) y = T ( x ) ; (10.712)
dx  dx 

*problem 166. Transformation to the invariant form (note 5.8):

d2 v
+ I ( x ) v ( x ) = G ( x ) ; (10.713)
dx 2

*problem 167. Transformation of two distinct differential equations: (i) to the

same invariant form (10.713); (ii) into each other (note 5.9).

J. c. Solutions of the Invariant Unforced Linear Second-Order

Differential Equation (notes 5.10–5.12):

*problem 168. Constant positive, negative or zero invariant (note 5.10);

*problem 169. Variable invariant with change of sign (note 5.10);
*problem 170. Turning point for the Airy differential equation (note 5.11):

d2 Ψ
+ α 2 ( x − xr ) Ψ ( x ) = 0. (10.714)
dx 2

J. d.  Waves in a Refracting Medium (notes 5.13–5.15):

*problem 171. Classical wave equation in a refracting medium with non-
unform propagation speed (note 5.13):

∂2 Φ ∂2 Φ ∂2 Φ 1 ∂2 Φ
+ + = . (10.715)
 c ( x )  ∂t
2
∂x2 ∂y 2 ∂z2 2

*problem 172. Longitudinal wave propagation, evanescence or divergence

(notes 5.14–5.16):

d 2 Ψ  ω 2 
2
+ 2 − k  Ψ ( x ) = 0. (10.716)
  c ( x ) 
2
dx 
176 Differential Equations and Applications

J. e.  Light and Sound Rays (notes 5.16–5.20):

*problem 173. First-order ray approximation (notes 5.16–5.17);
*problem 174. Phase variation along ray paths (note 5.17);
*problem 175. Amplitude variation along ray tubes (note 5.17);
*problem 176. Second-order ray approximation (notes 5.18–5.19);
*problem 177. Validity of the ray approximation (note 5.20).

K. Buckling of Elastic Beams (sections 6.1–6.3; example 10.10)

K. a.  General Equations (subsections 6.1.1–6.1.4):
*problem 178. Curvature of the elastica and tangential tension (subsec-
tion 6.1.1);
*problem 179. Shape of the elastica related to the shear stress and transverse
force (subsection 6.1.2);
*problem 180. Non-linear buckling for a uniform beam (subsection 6.1.2);
*problem 181. Linear buckling, that is for small slope (subsection 6.1.3);
*problem 182. Linear buckling of a uniform beam (subsection 6.1.3);
*problem 183. Lowest-order non-linear correction to problem 182 (subsec-
tion 6.1.4).

K. b.  Linear Buckling (subsections 6.1.5–6.1.9):

d 4ζ d 2ζ
EI − T = 0. (10.717)
dx 4 dx 2
*problem 184. Clamped at both ends (subsection 6.1.5);
*problem 185. Pinned at both ends (subsection 6.1.6);
*problem 186. Clamped at one end, pinned at the other end (subsection 6.1.7);
*problem 187. Non-linear and linear boundary conditions at a free end (sub-
section 6.1.8);
*problem 188. Cantilever beam, which is clamped at one end and free at the
other end (subsection 6.1.9);
*problem 189. Comparison of critical buckling loads for four diferent combi-
nations of supports (subsection 6.1.9).

K. c.  Non-Linear Buckling to the Lowest Order (subsections 6.1.10–6.1.15):

*problem 190. Equation of the elastica (10.718a, c) in the lowest-order, non-
linear (10.718b) approximation (subsection 6.1.10):

 ′ ′
 dζζ′ 2    3 2  ′    ζ′ 2  
ζ′ ≡
 1 − ; ζ′ 3 << 1:
ζ ′′  1 − ζ ′   + p 2
 ζ ′ 1 −  = 0.
 dx 2    2    
 2  

(10.718a–c)
Classification 10.2 177

*problem 191. Cantilever beam (subsections 6.1.10–6.1.12 and 6.1.15);

*problem 192. Beam clamped or pinned at both ends (subsections
6.1.13–6.1.15);
*problem 193. Fundamental buckling mode plus first harmonic (subsec-
tion 6.1.15).

K. d.  Buckling with a Point Translational Spring (subsections 6.2.1–6.2.5):

*problem 194. Boundary condition (10.719b) for a point translational spring at
(10.719a) any position (subsection 6.1.1):

0 ≤ a ≤ L: k ζ ( a ) = E I ζ′′′ ( a ) − T ζ′ ( a ) ; (10.719a, b)

*problem 195. Cantilever beam with a point translational spring at the free
end (subsection 6.2.2);
*problem 196. Clamped beam with a point translational spring in the middle
(subsection 6.2.3);
*problem 197. Pinned beam with a point translational spring in the middle
(subsection 6.2.4);
*problem 198. Clamped-pinned beam with a point translational spring in the
middle (subsection 6.2.5).

K. e.  Buckling with a Point Rotary Spring (subsections 6.2.6–6.2.10):

*problem 199. Boundary conditions (10.720b) for a point rotational spring at
(10.720a) any position (subsection 6.2.6):

0 ≤ a ≤ L: − E I ζ′′ ( a ) = k ζ′ ( a ) ; (10.720a, b)

*problem 200. Cantilever beam with a point rotary spring at the free end
(subsection 6.2.7);
*problem 201. Clamped beam with a point rotary spring in the middle (sub-
section 6.2.8);
*problem 202. Pinned beam with a point rotary spring in the middle (subsec-
tion 6.2.9);
*problem 203. Clamped-pinned beam with a point rotary in the middle (sub-
section 6.2.10).

K. f. Opposition to, or Facilitation of, Buckling

(subsections 6.2.11–6.2.17):
*problem 204. Buckling loads and harmonics of all orders for a cantilever
beam without point spring (subsection 6.2.12);
*problem 205. Fundamental and first two buckling harmonics for problem 204
(subsection 6.2.13);
178 Differential Equations and Applications

*problem 206. Buckling loads of all orders for a cantilever beam with a point
rotary spring at the free end (subsections 6.2.14–6.2.15);
*problem 207. Problem 206 with a point translational spring (subsections
6.2.14 and 6.2.16);
*problem 208. Buckled shape of a cantilever beam with a translational or
rotary spring at the free end (subsections 6.2.14–6.2.17).

K. g.  Bending of a Beam Continuously Supported on Springs (section 6.3):

*problem 209. Force balance equation for the linear deflection of an unforced
uniform beam (subsections 6.3.1–6.3.2):

d 4ζ d 2ζ
EI 4
− T 2 + k ′ζ = 0. (10.721)
dx dx

*problem 210. Pinned beam under traction T > 0 without k ′ = 0, spring sup-
port (subsection 6.3.3);
*problem 211. Pinned bar with attractive springs k ′ > 0 and T = 0, no axial
tension (subsection 6.3.4);
*problem 212. Pinned bar supported on repulsive springs k ′ < 0 without
T = 0, axial tension (subsection 6.3.5);
*problem 213. Pinned beam with balance T 2 = 4 E I k ′ > 0 of axial compression
T < 0 and attractive spring k ′ > 0 (subsection 6.3.6);
*problem 214. Pinned beam with balance T 2 = 4E I k ′ > 0 of axial traction T > 0
and attractive spring k ′ > 0 (subsection 6.3.7);
*problem 215. Pinned beam with predominance T 2 > 4 E I k ′ of axial traction
T > 0 and k ′ > 0 attractive springs (subsection 6.3.8);
*problem 216. Pinned beam with predominance T 2 > 4 E I k ′ of axial compres-
sion T < 0 and k ′ > 0 attractive springs (subsection 6.3.9);
*problem 217. Pinned beam with predominance T 2 > 4 E I k ′ of axial tension
over k ′ < 0 repulsive springs (subsection 6.3.10);
*problem 218. Pinned beam with predominance 0 < T 2 < 4 E I k ′ of axial atrac-
tive springs k ′ > 0 over axial tension (subsection 6.3.11);
*problem 219. Extension of the problems 185 and 210–218 from pinned to can-
tilever, clamped and clamped-pinned beams using an effective axial tension
(subsection 6.3.12);
*problem 220. To all problems 185 and 210–219, apply a combined tension-
spring buckling load that does not affect the shape of the buckled elastica
(subsection 6.3.13);
*problem 221. Extension of problem 219 to (10.722) to bending under own
weight (subsection 6.3.14):

d 4ζ d 4ζ
EI 4
− T 4 + k ′ ζ = ρ g . (10.722)
dx dx
Classification 10.2 179

K. h. Linear Bending (10.723b) of a Beam Under Traction

(10.723a) subject to shear stresses (example 10.10):

d 4ζ d 2ζ
T > 0: EI 4
−T = f ( x ) . (10.723a, b)
dx dx 2

*problem 222. Clamped beam (10.724a–d) with concentrated torque Q at

(10.724e) the middle (E10.10.2):

 L
ζ ( 0 ) = 0 = ζ′ ( 0 ) , ζ ( L ) = 0 = ζ′ ( L ) : f ( x ) = Q δ ′  x −  ; (10.724a–e)
 2

*problem 223. Pinned beam (10.725a–d) with concentrated force F at (10.725e)

the middle (E10.10.3):

 L
ζ ( 0 ) = 0 = ζ′′ ( 0 ) , ζ ( L ) = 0 = ζ′′ ( L ) : f ( x) = F δ′  x −  ; (10.725a–e)
 2

*problem 224. Clamped-pinned beam (10.726a–d) subject to (10.726e) own

weight (E10.10.4):

ζ ( 0 ) = 0 = ζ′ ( 0 ) , ζ ( L ) = 0 = ζ′′ ( L ) : f ( x ) = ρ g ; (10.726a–e)

*problem 225. Cantilever beam (10.727a–d) subject to a shear stress, which is a

linear function (10.727e) of the distance from the clamped support (E10.10.5):

ζ ( 0 ) = 0 = ζ′ ( 0 ) , ζ′′ ( L ) = 0 = E I ζ′′′ ( L ) − T ζ′ ( L ) : f ( x ) = q x . (10.727a–e)

L.  Axial Traction/Compression of a Straight Bar (section 6.4):

− F = ( E u′ )′ = E u′′ + E ′ u′. (10.728a, b)

*problem 226. Analogy with the linear transvese deflection of an elastic

string (subsection 6.4.1);
*problem 227. Analogy with one-dimensional steady heat conduction (sub-
section 6.4.2);
*problem 228. Strain and displacement for an inhomogeneous material (sub-
section 6.4.3);
180 Differential Equations and Applications

*problem 229. Displacement, strain, and stress for an inhomogeneus rod

with Young modulus (10.729b) subject to a (10.729a) constant longitudinal
force (subsection 6.4.4):

x2
F = const: E ( x ) = E0 + ( E1 − E0 ) . (10.729a, b)
L2

M. Plane Elastic Displacements, Strains,

and Stresses (section 6.5):
*problem 230. Relation between the stress and strain tensors (subsection
6.5.1);
*problem 231. In-plane stresses and strains without out-of-plane stresses
(subsection 6.5.2);
*problem 232. Plane elasticity without out-of-plane strains (subsection 6.5.3);
*problem 233. Transformation between in-plane stresses (plane elasticity) in
the [problem 231 (232)] (subsections 6.5.3 and 6.5.8);
*problem 234. Differential equation (10.730) satisfied by the displacement
vector (subsection 6.5.4):
  
0 = E h ( 1 − σ ) ∇ 2 u + ( 1 + σ ) ∇ ( ∇.u)  + 2 ( 1 − σ ) f ; (10.730)

*problem 235. Bi-harmonic differential equation (10.731b) for (10.731a) the

stress function (subsection 6.5.5):

∂4 Θ ∂4 Θ ∂4 Θ
( )
Θ∈D 4 |R 2 : 0=
∂x 4
+
∂y 4
+ 2 2 2 ;
∂x ∂y
(10.731a, b)

*problem 236. Strains and stresses due to the drilling of a circular hole in an
infinite thin elastic plate (subsection 6.5.6);
*problem 237. Stress concentration near a circular hole in an infinite plate
with normal stresses at infinity (subsections 6.5.7–6.5.8);
*problem 238. In-plane deformations, strains, and stresses in a circular disk
subject to a uniform pressure along the periphery (subsection 6.5.9).

N. Deflection of an Elastic Membrane under

Anisotropic In-Plane Stresses (section 6.6):
*problem 239. Elastic strains and energy associated with the deflection of a
membrane subject to anisotropic stresses (subsection 6.6.1);
Classification 10.2 181

*problem 240. Balance equation (10.732b) and boundary condition (10.732a) in

the general non-linear anisotropic case and particular cases (subsection 6.6.2):

∂  ∂ζ 
Tαβ Nβ δζα = 0: 0=
∂ xα  Tαβ ∂ x  + f ( xγ ) ; (10.732a, b)
β

*problem 241. Transformation of the components of a vector or tensor by a

rotation in the plane (subsection 6.6.3);
*problem 242. Green’s function for (10.733) the plane anisotropic Laplace
operator (subsection 6.6.4):

 ∂2 ∂2 ∂2 
T11 + T22 + 2 T12  G ( x1 , x2 ) = δ ( x1 ) δ ( x2 ) ; (10.733)
 ∂ x1 ∂ x1 ∂ x2 ∂ x2 ∂ x1 ∂ x2 

*problem 243. Characteristic lines of zero deflection for an infinite membrane

under anisotropic stresses with no transverse force (subsection 6.6.5);
*problem 244. Linear deflection of a circular membrane under its own weight
(subsection 6.6.6);
*problem 245. Linear deflection of an elliptic membrane under its own weight
(subsection 6.6.6);
*problem 246. Deflection of an elastic membrane under anisotropic stresses
for (10.734) arbitrary loading (subsection 6.6.7):

 ∂2 ∂2 ∂2 
T11 2 + T22 2 + 2 T12  ζ ( x1 , x2 ) = − f ( x1 , x2 ) . (10.734)
 ∂ x1 ∂ x2 ∂ x1 ∂ x2 

O. Weak Linear Bending of Elastic Isotropic

Plates (section 6.7; example 10.11)
O. a. Transverse Displacement with Constant Bending
Stiffness (subsections 6.7.1–6.7.6):
*problem 247. Curvature tensor and principal bending moments (subsec-
tion 6.7.2);
*problem 248. Curvatures and twist in axis rotated relative to the principal
axis (subsection 6.7.3);
*problem 249. Stress and twist couples (subsection 6.7.4);
*problem 250. Elastic energy of weak linear bending (subsection 6.7.5);
182 Differential Equations and Applications

*problem 251. Turning moments and linear weak deflection of an isotropic elas-
tic plane (10.735b) with constant (10.735a) bending stiffness (subsection 6.7.6):

Eh3 f ( x, y ) ∂4 ζ ∂4 ζ ∂4 ζ
D≡ = const: = + + 2 . (10.735a, b)
( )
12 1 − σ 2 D ∂x 4 ∂y 4 ∂x2 ∂y 2

O. b. Transverse Displacement with Non-Uniform

Bending Stiffness (subsections 6.7.7–6.7.11):
*problem 252. Displacement vector and strain, and stress tensor for weak
linear bending (subsection 6.7.7);
*problem 253. Elastic energy for weak linear bending (subsection 6.7.8);
*problem 254. Elastic energy per unit volume or area and total elastic energy
(subsection 6.7.9);
*problem 255. Analogy between the bending and deformation vectors (sub-
section 6.7.9);
*problem 256. Balance equation for the displacement with non-uniform
bending stiffness (subsection 6.7.10);
*problem 257. Comparison of the weak linear bending of elastic bars and
plates (subsection 6.7.11).

O. c. Boundary Conditions for Isotropic Elastic

Plates (subsections 6.7.12–6.7.14):
*problem 258. Clamped, pinned, or supported, and free boundary conditions
(subsection 6.7.12);
*problem 259. Displacement and slope for clamped or pinned boundary con-
ditions (subsection 6.7.12);
*problem 260. Turning moment for free boundary conditions (subsection
6.7.13);
*problem 261. Polar coordinates as a particular case of orthogonal curvilin-
ear coordinates (subsection 6.7.13);
*problem 262. Passage from polar to cartesian coordinates (subsection 6.7.13);
*problem 263. Stress couples for pinned or free boundary conditions (subsec-
tion 6.7.13);
*problem 264. Boundary conditions for rectangular, circular, and arbitrarily
shaped plates (subsection 6.7.14);
*problem 265. Stress couples in polar coordinates (subsection 6.7.14).

O. d. W
 eak Linear Bending of Isotropic Circular
Plates (subsections 6.7.15–6.7.19):
*problem 266. Concentrated force at the center and clamped at the boundary
(subsections 6.7.15–6.7.16);
Classification 10.2 183

*problem 267. Problem 266 pinned or supported at the boundary (subsections

6.7.15–6.7.16);
*problem 268. Bending by a constant transverse force, such as the weight of a
homogeneous plate with constant thickness (subsection 6.7.17);
*problem 269. Heavy uniform circular plate clamped at the boundary (sub-
section 6.7.18);
*problem 270. Problem 269 pinned or supported at the boundary (subsection
6.7.18);
*problem 271. Heavy circular plate hanging from the center with free bound-
ary (subsection 6.7.19);
*problem 272. Circular plate with a concentric circular hole (example 10.11).

P. Linear Elastic Stability of a Stressed

Orthotropic Plate (section 6.8; example 10.15)
P. a. Linear Strong Bending with Anisotropic In-Plane
Stresses (subsections 6.8.1–6.8.2):
*problem 273. Balance equation (10.736) for the transverse displacement (sub-
section 6.8.1):

( )
f = ∇ 2 D ∇ 2ζ − h ∂α (Tαβ ∂β ζ ) . (10.736)

*problem 274. Comparison of elastic stressed plate and beam under tension
(subsection 6.8.1);
*problem 275. Balance equation for the transverse displacement of a uniform
stressed plate and beam (subsection 6.8.1);
*problem 276. Extreme cases of transverse deflection of a membrane and
weak bending of an elastic plate (subsection 6.8.2).

P. b. Mechanics and Thermodynamics of Deformable

Media (subsections 6.8.3–6.8.6):
*problem 277. Momentum equation balancing inertia and volume forces
and stresses (subsection 6.8.3):

dvi
fi + ∂ j Tij = ρ ai = ρ ; (10.737a, b)
dt

*problem 278. Newton law (1686) in the absence of stresses for mass inde-
pendent of time (subsection 6.8.3);
*problem 279. Balance of moments of forces and symmetry of the stress ten-
sor (subsection 6.8.4);
184 Differential Equations and Applications

*problem 280. Work of the volume and surface forces (subsection 6.8.5);
*problem 281. Internal energy, including the work of deformation of the
stresses on the strains (subsection 6.8.6).

P. c.  Stress-Strain Relation for Elastic Materials (subsections 6.8.6–6.8.11):

*problem 282. Residual, elastic, and inelastic stresses (subsection 6.8.6);
*problem 283. Stiffness tensor and matrix, relating the strain and stress
tensors in an anisotropic elastic material (subsection 6.8.7);
*problem 284. Stress-strain relation for an isotropic elastic material (subsec-
tion 6.8.8);
*problem 285. Relations between the two Lamé elastic moduli and the Young
modulus and Poisson ratio (subsection 6.8.8);
*problem 286. Hooke law (1678) for an isotropic elastic material in terms of
various elastic moduli (subsection 6.8.8);
*problem 287. Stiffness matrix for a homoclinic elastic material with x-plane
of symmetry (subsection 6.8.9):
*problem 288. Stiffness matrix for an orthotropic elastic material with three
orthogonal planes of systemetry (subsection 6.8.9);
*problem 289. Compliance matrix as the inverse of the stiffness amtrix (sub-
section 6.8.10);
*problem 290. Compliance matrix for an isotropic elastic material (subsec-
tion 6.8.10);
*problem 291. Stiffness and compliance matrices for an orthotropic elastic
material (subsections 6.8.10–6.8.11);
*problem 292. Comparison of orthotropic and isotropic elastic materials
(subsection 6.8.11).

P. d. Equation for the Transverse Displacement of a Pseudo-Isotropic

Orthotropic Elastic Plate (subsections 6.8.12–6.8.16):
*problem 293. Elastic energy for a homogeneous anisotropic material (sub-
section 6.8.12);
*problem 294. Elastic energy for the weak linear deformation of an ortho-
tropic plate (subsection 6.8.12);
*problem 295. Conditions for a pseudo-isotropic orthotropic elastic material
(subsection 6.8.13);
*problem 296. Generalized bending stiffness (10.738) for (10.735b) a pseudo-
isotropic orthotropic elastic plate (subsection 6.8.14):

D≡
h3 
C −
( C13 )2  = h3 C − ( C23 )2  . (10.738)
 11   22 
12  C33  12  C33 

*problem 297. Comparison of (i) anisotropic, (ii) orthotropic, (iii) pseudo-

isotropic orthotropic, and (iv) isotropic elastic (a) materials and (b) plates
(subsection 6.8.15).
Classification 10.2 185

P. e. Boundary Conditions for a Pseudo-Isotropic

Orthotropic Plate (subsection 6.8.6; example 10.15):
*problem 298. General boundary condition for (i) isotropic and (ii) pseudo-
isotropic orthotropic plates (subsection 6.8.16);
*problem 299. Particular case of constant first and second bending stiffnesses
(subsection 6.8.16);
*problem 300. Relation (10.739) between the elastic energy per unit length of
boundary of a plate and the stress couples and turning moment (subsection
6.8.16):

dEˆ
E d = d = − N x δζ − Mn δ ( ∂n ζ ) ; (10.739)
ds

*problem 301. Vanishing of the total elastic energy along the boundary of a (i)
isotropic or (ii) pseudo-isotropic orthotropic plate (E10.15.1–E10.15.2);
*problem 302. Stress couple and turning moment for a pseudo-isotropic
orthotropic plate of arbitrary shape (E10.15.3);
*problem 303. simplification of the stress couple and turning moment for a
pinned plate of arbitrary shape (E10.15.3–E10.15.4);
*problem 304. Problem 303 for a clamped plate (E10.15.3–E10.15.4);
*problem 305. Comparison of (a) stress couple and (b) turning moment for
(i) clamped or (ii) pinned plate of an (α) isotropic and (β) pseudo-isotropic
orthotropic material (E10.15.5);
*problem 306. Boundary conditions for elastic plates: (i) of arbitrary, circular,
or rectangular shape; (ii) with clamped, pinned, or free boundary; (iii) made of
isotropic or pseudo-isotropic orthotropic elastic material (E10.15.6–E10.15.9).

P. f. Buckling of a Clamped Rectangular Plate with

Normal Stresses (subsection 6.8.17):
*problem 307. Buckled shape of an elastic plate clamped on all four sides and
subject to normal stresses only, that is, no shear stress;
*problem 308. Buckling conditions for two orthogonal compressions or one
compression and a traction;
*problem 309. Comparison of buckling of a clamped rectangular plate and
an elastic beam.

P. g. Buckling of a Clamped-Free Plate with Uniaxial

Stresses (subsections 6.8.18–6.8.20):
*problem 310. Three cases distinguishing traction from two ranges of
compression;
*problem 311. Buckling for weak compression below a threshold;
*problem 312. Buckling for strong compression in a range.
186 Differential Equations and Applications

Q. Non-Linear Coupling of Strong Bending and

In-Plate Stresses (section 6.9; example 10.16)
Q. a. Balance Equations for the Strong Bending of an
Isotropic Elastic Plate (subsections 6.9.1–6.9.3):
*problem 313. Coupled non-linear fourth-order partial differential equations
for the transverse displacement and stress function (subsection 6.9.1);

( ) ( ) ( )
D∇ 2ζ − f − h  ∂ yy Θ ( ∂ xx ζ ) + ( ∂ xx Θ ) ∂ yy ζ − 2 ∂ yy Θ ∂ yy ζ  , (10.740a)

( ) ( )( )
∇ 4 Θ = E  ∂ yy ζ − ∂ xy ζ ∂ yy ζ  ; (10.740b)

2

*problem 314. Degenerate linear case of weak bending (10.741a) of a (10.741b)

stressed plate (subsection 6.9.1):

  ∇ 4ζ = − f , ∇ 4Θ = 0 ; (10.741a, b)

*problem 315. Exact strain tensor due to transverse and in-plane displace-
ments of a plate (subsection 6.9.2);
*problem 316. Using the Hooke law and stress function to eliminate the in-
plane displacements and keep the transverse displacement in (10.740b) (sub-
section 6.9.3).

Q. b. Extension of the Balance Equations to a Pseudo-

Isotropic Orthotropic Plate (example 10.16):
*problem 317. Extension of the balance equation (10.740a) replacing the bend-
ing stiffness for isotropic plates by the generalized bending stiffness (10.738)
pseudo-isotropic orthotropic plates;
*problem 318. Extension of complementary equation orthotropic plate with
constant compliance matrix;
*problem 319. Extension of the complementary equation (10.740b) to an ortho-
tropic plate in terms of generalized Young moduli, Poisson ratios, and shear
moduli;
*problem 320. Two constraints additional to the problem 319 for pseudo-
isotropic orthotropic plates.

Q. c. Boundary Conditions for the Transverse Displacement

and Stress Function (subsections 6.9.4–6.9.5):
*problem 321. Force-stress balance equation in the absence of in-plane inertia
force (subsection 6.9.4);
Classification 10.2 187

*problem 322. Boundary condition for the in-plane stress vector and dis-
placement (subsection 6.9.4);
*problem 323. Principle of virtual work, including elastic energies of bend-
ing, deflection, and deformation (subsection 6.9.5);
*problem 324. Surface integral in problem 323 leading to (10.740a, b) the bal-
ance equations (subsections (subsection 6.9.5);
*problem 325. Boundary integral in problem 323 leading to the boundary
conditions (subsection 6.9.5);
*problem 326. Second boundary condition from problem 325 involving the
augmented turning moment (subsection 6.9.5);
*problem 327. Conditions for clamped, pinned, and free boundaries in the
strong bending of an elastic plate (subsection 6.9.5).

Q. d. Deflection and Bending with an Elliptic Boundary

(subsections 6.9.6–6.9.7):
*problem 328. Transverse deflection with an elliptic boundary satisfying the
balance (10.740a) and complementary (10.740b) equations (subsection 6.9.6);
*problem 329. Incompatibility of the problem 328 with boundary conditions
at low order of approximation (subsection 6.9.7).

Q. e.  Perturbation Expansion to all Orders (subsections 6.9.8–6.9.11):

*problem 330. Perturbation expansion to all orders for the transverse dis-
placement and stress function (subsection 6.9.8);
*problem 331. Balance and complementary equations, strains, and stresses,
and stress couple and turning moment with axial symmetry (subsection 6.9.9);
*problem 332. The problem 330 simplified for the axisymmetric case, which
is dependent only on the distance from the axis (subsection 6.9.10);
*problem 333. Particular integral of the axisymmetric bi-harmonic equation
forced by a power (subsection 6.9.11);
*problem 334. Problem 333 extended to forcing by a polynomial or analytic
function (subsection 6.9.11).

Q. f. Strong Non-Linear Bending of a Heavy Circular Plate

Under Compression (subsections 6.9.12–6.9.17):
*problem 335. Transverse displacement and stress function to lowest non-
linear order of approximation (subsections 6.9.12–6.9.13);
*problem 336. Problem 336 with the arbitrary constants determined for a
clamped circular plate (subsection 6.9.14);
*problem 337. The displacements, strains, stresses, stress couple, and aug-
mented turning moment for the problem 335 (subsections 6.9.14–6.9.17);
*problem 338. General method of solution for the strong non-linear bending
with in-plane stresses, including the particular axisymmetric case (subsec-
tion 6.9.17).
188 Differential Equations and Applications

R.  Waves or Vibrations of Elastic Media (notes 6.1–6.23)

R. a. Non-Linear Waves in Inhomogeneous and
Unsteady Media (notes 6.3–6.9):
*problem 339. Non-linear wave equation (10.742c) for the transverse displace-
ment ζ of an elastic string with mass density per unit length ρ1, tangential
tension T, and shear stress f a , which are all functions (10.742a, b) of position
and time (note 6.3):

{ }
•
∂ζ • ∂ζ  • 2 −1/2 ′
ζ′ ≡ ,ζ≡ :  1  − T ζ′ 1 + ζ′
ρ ζ = f a ( x , t ) ; (10.742a–c)
∂x ∂t

*problem 340. Non-linear wave equation (10.743) for the transverse displace-
ment ζ of an elastic membrane with mass density ρ1 and transverse applied
force f a per unit area and isotropic tangential tension T, which are all func-
tions of position and time (note 6.4):

{ } = f ( x, y, t). (10.743)
•
 •
 ρ2 ζ − ∇. T 1 + ( ∇ζ.∇ζ )
−1/2
a

*problem 341. Linear torsional (10.744a) wave equation (10.744b) for the angle
of rotation φ along a straight elastic rod with moment of inertia I and tor-
sional stiffness of the cross-section C and applied axial torque or moment
Ma , which are all functions of position and time (note 6.5):
•
 I φ•  − C φ′ ′ = M x , t ;
  ( )
•
τ ≡ φ: a( ) (10.744a, b)

*problem 342. Linear wave equation (10.745) for the longitudinal displace-
ment u of a straight elastic rod with mass density per unit length ρ1, Young
modulus E and applied longitudinal force f a , all depending on position and
time (note 6.6).
•
 ρ u•  − E u′ ′ = F x , t ; (10.745)
 1  ( ) a( )

*problem 343. Non-linear wave equation (10.746) for the transverse displace-
ment ζ of an elastic beam with mass density per unit length ρ1, bending
stiffness B, tangential tension T, and applied shear stress f a , which are all
functions of position and time (note 6.7):

{ }′ + {Bζ′ 1 + ζ′ }′′ = f ( x, y, t) ; (10.746)

•
 • 2 −1/2 2 −3/2
 ρ2 ζ − T ζ′ 1 + ζ′ a
Classification 10.2 189

*problem 344. Linear wave equation (10.747) for the transverse displacement
ζ of an elastic stressed plate with mass density ρ2 and transverse force f a per
unit area in-plane stresses Tαβ and bending stiffness D, which are all func-
tions of position and time (note 6.8):
•
 •
2 2
( )
 ρ2 ζ − ∂α (Tαβ ∂β ζ ) + ∇  D ∇ ζ  = f a ( x1 , x2 , t ) . (10.747)

R. b.  Generalized and Analogous Wave Equations (notes 6.9–6.10):

*problem 345. Analogies of wave variables, inertia, restoring effect, wave
speeds, and dispersion stiffness (note 6.9);
*problem 346. Linear general bending wave equation in one-dimension
(10.748) for a wave variable ζ with mass density per unit length ρ1, tangen-
tial tension T, bending stiffness B, friction coefficient µ, translational ν, and
rotary ϑ spring resiliences, and shear stress f a all depending on position and
time (note 6.10):
•
 •
 ρ2 ζ − (T ζ′ ) + ( B ζ′′ ) + µζ+ νζ + ϑ ζ′ = f a ( x , t ) ; (10.748)
•
′ ′′

*problem 347. Linear general bending wave equation in two dimensions

(10.749) for the wave variable ζ with mass density per área ρ2, in-plane stresses
Tαβ, bending stiffness D, friction coefficient µ, translational spring resilience ν
, vector rotary spring resilience ϑ, and transverse force per unit area f a which
are all functions of position and time (note 6.10):
• 
 •
( )
 ρ2 ζ − ∂α (Tαβ ∂β ζ ) + ∇ D ∇ ζ + µζ+ νζ + ϑ. ∇ζ = f a ( x , y , t ) ; (10.749)
•
2 2

*problem 348. Particular case of problem 346 for the classical one-dimensional
bending wave equation (10.750) for the wave variable ζ with damping χ,
translational ω t , and rotational ω r natural frequencies, elastic wave speed,
ce and dispersion stiffness parameter b (note 6.10):
•• •
ζ+ 2 χζ+ ω t2 ζ + ω 2r ζ′ − ce2 ζ′′ + b ζ′′′′ = ρ1−1 f a ( x , t ) ; (10.750)

*problem 349. Particular case of problem 347 for the classic two-dimensional
bi-dimensional bending wave equation (10.751) for the wave variable ζ, as in
the one-dimensional case (10.750) with two wave speeds c1,2 in the principal
stress directions, a vector ω r1,2 natural frequency for the rotary spring and a
dispersion stiffness parameter b (note 6.10):
•• •
ζ+ 2 χζ+ ω t2 ζ + ω 2r 1 ∂1 ζ + ω 2r 2 ∂2 ζ − c12 ∂11 ζ
(10.751)
( ∂11 + ∂22 ) fa ( x, y , t ) .
2 2 2 −1
− c ∂22 ζ + b
2 ζ=ρ 1
190 Differential Equations and Applications

R. c. Generalized Classical Bending Wave Equation in One-

Dimension with Constant Coefficients (notes 6.11–6.13):
*problem 350. Solution by separation of variables of (10.752) the one-­
dimensional classical bending wave equation with constant coefficients and
without forcing (note 6.11):
•• •
ρ1 ζ+ µζ+ νζ + ϑ ζ′ − Tζ′′ + B ζ′′′′ = 0. (10.752)

*problem 351. Fundamental and harmonics of the spatial modes specifying

the wavelengths and wavenumbers (note 6.12);
*problem 352. Dispersion relation for linear transversal waves in an elastic
beam with constant mass density per unit length, tangential tension, bend-
ing stiffness, and translational spring resilience (note 6.13);
*problem 353. Frequencies and periods of the fundamental mode and har-
monics of a uniform elastic string (note 6.14).

R. d. Non-Dispersive and Dispersive Free Wave Propagation

and Standing Modes (notes 6.14–6.17):
*problem 354. Phase and phase speeds of non-dispersive waves with a
permanent waveform, which are solutions of the classical wave equation
(note 6.14);
*problem 355. First-order permanent waveforms propagating in opposite
directions and their superposition with the same amplitude in second-order
standing modes (note 6.15);
*problem 356. Phase speed and group velocity for dispersive waves (note 6.16);
*problem 357. Dispersion relation, phase speed, and group velocity for trans-
verse waves along an elastic beam (note 6.16);
*problem 358. Propagation of a non-dispersive or a dispersive wave packet
(note 6.16);
*problem 359. General free wave field as a solution of the unforced wave
equation specified by a superposition of eigenfunctions, corresponding to
the eigenvalues with arbitrary amplitudes (note 6.17).

R. e. Wave Generation by Out-of-Equilibrium Initial

Conditions and Forcing (notes 6.18–6.23):
*problem 360. Determination of the amplitudes of the eigenfunctions from
the initial conditions, which in the case out-of-equilibrium specify wave gen-
eration (note 6.18);
*problem 361. Wave generation by forcing at an applied frequency with
bounded fluctuation as a function of position (note 6.19);
*problem 362. Damped non-resonant forcing, which is with applied fre-
quency distinct from the fundamental frequency and all its harmonics
(notes 6.20–6.21);
Classification 10.2 191

*problem 363. Damped resonant forcing, which is with applied frequency

equal to the fundamental frequency or equal to any of its harmonics
(note 6.21);
*problem 364. Undamped non-resonant forcing, like problem 363 without
damping (note 6.22);
*problem 365. Undamped resonant forcing, like problem 364 without
damping (note 6.22).

S.  Waves in Inhomogeneous Media (notes 7.1–7.21)

S. a.  Classification of Wave Equations (notes 7.1–7.7):
*problem 366. One-dimensional non-linear waves in an unsteady inhomo-
geneous medium (note 7.3):

 ∂Φ ∂Φ ∂2 Φ ∂2 Φ ∂2 Φ ∂N + M Φ 
0 = F  x , t ; Φ, , , 2 , 2 , 2
,..., N M  ; (10.753)
 ∂ x ∂t ∂ x ∂t ∂ x ∂t ∂ x ∂t 

*problem 367. One-dimensional forced linear waves in an unsteady inhomo-

geneous medium (note 7.3):

N M
∂N + M Φ
∑∑A
n= 0 m= 0
nm ( x, t )
∂ x N ∂t M
= B ( x , t ) ; (10.754)

*problem 368. One-dimensional forced linear waves in a steady homoge-

neous media (note 7.4): as (10.754) with constant coefficients Anm ;
*problem 369. One-dimensional forced linear waves in a steady inhomoge-
neous media (note 7.5): as (10.754) with coefficients Anm depending only on
position x.

S. b. Quasi-One-Dimensional Waves in Inhomogeneous Media

(notes 7.6–7.17):
*problem 370. Wave equation (10.755c) for the transverse displacement ζ in the
free-linear (10.755a) vibrations of an elastic string under tangential tension T
with mass density per unit length (10.755b) independent of time (notes 7.8–7.9):
2
 ∂ζ  ∂σ ∂2 ζ 1 ∂  ∂ζ 
  << 1, = 0: − T ( x , t )  = 0 ; (10.755a–c)
∂x ∂t ∂t 2 σ ( x ) ∂ x  ∂x 
192 Differential Equations and Applications

*problem 371. Wave equation (10.756d) for the torsion τ, which is the rate of
rotation along the axis (10.756a), in the linear-free torsional oscillations of a
straight elastic rod with moment of inertia I and torsional stiffness C of the
cross-section that (10.756b, c) do not depend on time (note 7.10):

∂φ ∂I ∂C ∂2 τ 1 ∂  ∂τ 
τ≡ ; =0= : − C ( x )  = 0 ; (10.756a–d)
∂x ∂t ∂t ∂t 2 I ( x ) ∂ x  ∂x 

*problem 372. Wave equation (10.757c) for the longitudinal displacement u in

the free-linear longitudinal oscillations of an elastic rod with mass density
per unit length σ and Young modulus E that (10.757a, b) do not depend on
time (note 7.11):

∂σ ∂E ∂2 u 1 ∂  ∂u 
=0= : −  E ( x )  = 0 ; (10.757a–c)
∂t ∂t ∂t 2
σ ( x ) ∂x  ∂x 

*problem 373. Wave equation (10.758e) for the transverse displacement ζ of

the free-linear surface waves of an incompressible fluid (10.758a) in a uni-
form gravity field (10.758b) in a channel whose width b and depth h do not
(10.758c, d) depend on time (notes 7.12–7.13):

∂b ∂h ∂2 ζ g ∂  ∂ζ 
ρ, g = const ; =0= : −  h ( x ) b ( x )  = 0 ; (10.758a–e)
∂t ∂t ∂t 2
b ( x ) ∂x  ∂x 

*problem 374. Wave equation (10.759e) for the electric field E of electro-
magnetic waves in a medium with di-electric permittivity ε and mag-
netic permeability µ independent of time (10.759c, d) in the absence of
electric charges and currents (10.759a, b) involving the speed of light in
vacuo c0 (notes 7.14–7.15):

 ∂ε ∂µ ∂2 E c 2 ∂  1 ∂E 
q = 0 = J; =0= : − 0   = 0 ; (10.759a–e)
∂t ∂t ∂t 2
ε ( x ) ∂x  µ ( x ) ∂x 

*problem 375. Wave equation (10.760c) for the acoustic pressure perturbation
p of free-linear (10.760a) sound waves, which is with velocity perturbation
v small compared with the sound speed c s , in a horn, which is a duct with
variable cross-sectional area A, which does not (10.760b) depend on time
(notes 7.16–7.17):

2
 v ∂A ∂2 p c2 ∂  ∂p 
 c  << 1,    ∂t = 0: − s A ( x )  = 0. (10.760a–c)
s ∂t 2
A ( x ) ∂ x  ∂x 
Classification 10.2 193

S. c.  Analogies among Six Types of Waves (notes 7.18–7.19):

*problem 376. Primal (10.761a) and dual (10.761b) wave equations (note 7.18):

∂2 Φ 2 ∂ Φ 1 ∂Φ 
2
−  c ( x )   2 + = 0, (10.761a)
∂t 2
 ∂ t L ( x ) ∂t 

∂2 Φ ∂  2  ∂Ψ Ψ  
−   c ( x )   +   = 0, (10.761b)
 ∂ x L ( x )  
2
∂t ∂ x 

*problem 377. Primary wave variables (note 7.18);

*problem 378. Dual variables (note 7.18);
*problem 379. Inertia effects (note 7.19);
*problem 380. Restoring effects (note 7.20);
*problem 381. Phase speeds (note 7.20);
*problem 382. Lengthscales (note 7.20).

S. d.  Duality Principles for Wave Variables (notes 7.21–7.22):

*problem 383. Duality principle for acoustic horns, problem 375 (note 7.20);
*problem 384. Duality principle for water waves in a channel, problem 373
(note 7.20);
*problem 385. Duality principle for electromagnetic waves in a dielectric,
problem 374 (note 7.20);
*problem 386. Duality principle for the transverse oscillations of elastic
strings, problem 370 (note 7.21);
*problem 387. Duality principle for the torsional oscillations of a rod, prob-
lem 371 (note 7.21);
*problem 388. Duality principle for the longitudinal oscillations of a rod,
problem 372 (note 7.21).

T.  Acoustics of Horns (notes 7.22–7.46)

T. a. General Properties (notes 7.22–7.27):
*problem 389. Horn wave equations for the acoustic pressure and velocity
pertubation spectra (note 7.22);
*problem 390. Invariant form of the horn wave equation for the reduced
acoustic pressure and velocity perturbation spectra (note 7.23);
*problem 391. Polarization relations between the acoustic pressure and veloc-
ity perturbation spectra in original and reduced form (note 7.24);
194 Differential Equations and Applications

*problem 392. Classical wave equation for sound in a duct of constant cross-
section (note 7.25);
*problem 393. Ray approximation for sound propagation in a horn with a
wavelength short compared with the lengthscale of variations in cross-sec-
tion (note 7.26).

T. b.  Exact Elementary Solutions (notes 7.27–7.35):

*problem 394. The exponential horn as the only self-dual duct (note 7.28);
*problem 395. Filtering function as the solution of a filtering equation involv-
ing a cut-off frequency and leading to a reduced wavenumber (note 7.29);
*problem 396. Acoustic pressure and velocity perturbations in an exponen-
tial horn involving the filtering function (note 7.30);
*problem 397. Sound field in a catenoidal horn with a filtering function for
the acoustic pressure perturbation spectrum (note 7.31);
*problem 398. Sound field in an inverse catenoidal horn with a filtering func-
tion for the acoustic velocity pertubation spectrum (note 7.32);
*problem 399. Imaginary change of lengthscale leading from the filtering to
the transparency equation and function (note 7.33);
*problem 400. Sound field in a sinusoidal horn involving a transparency
function for the acoustic pressure perturbation spectrum (note 7.33);
*problem 401. Sound field in an inverse sinusoidal horn involving a transpar-
ency function for the velocity perturbation spectrum (note 7.34);
*problem 402. Five shapes of horn for which the horn wave equation has
exact solutions in terms of elementary functions (note 7.35).

T. c.  Gaussian Horn and Displacement Amplifier (notes 7.36–7.38):

*problem 403. Exact solution of the wave equation in a Gaussian horn in
terms of Hermite functions (note 7.36);
*problem 404. Acoustic pressure and velocity perturbation spectra in a
Gaussian horn in terms of Hermite functions (note 7.36);
*problem 405. Frequencies of oscillation of an elastic rod with Gaussian cross-
section for which the longitudinal displacement is specified by an Hermite
polynomial (notes 7.37–7.38);
*problem 406. Fundamental frequency of Gaussian elastic rod used as a dis-
placement amplifier for which the strain and stress are constant (note 7.38).

T. d.  Power Law Horns and Cylindrical/Spherical Waves (notes 7.39–7.46):

*problem 407. Asymptotic scaling of the sound field in a power-law duct at
large distance when the ray approximation holds (note 7.39);
*problem 408. Acoustic pressure and velocity perturbation spectra in a
power-law horn in terms of Hankel functions (notes 7.40–7.41);
*problem 409. Cylindrical waves in a wedge due to a line-source at the edge
(note 7.42);
Classification 10.2 195

*problem 410. Acoustic pressure and velocity perturbation spectra of cylin-

drical waves in the far-field (notes 7.42–7.43);
*problem 411. Spherical waves in a conical horn due to a point sound source
at the vertex (note 7.44);
*problem 412. Acoustic pressure and velocity perturbation spectra for
spherical waves in a conical horn that coincide as exact and asymptotic
fields (note 7.44).

T. e.  Reflection and Transmission at Interfaces (notes 7.47–7.55):

*problem 413. Impedance of a plane wave (note 7.47);
*problem 414. Locally-reacting boundary condition in terms of the specific
impedance (note 7.47);
*problem 415. Reflection and surface adsorption coefficient for a plane wave
incident normal to a locally reacting wall (note 7.48);
*problem 416. Wavelength and amplification or attenuation of standing
modes trapped between walls with distinct specific impedances (notes
7.49–7.50);
*problem 417. Acoustic boundary conditions at the junction of two ducts with
different cross-sections and containing fluids with distinct mass densities
and sound speeds (note 7.51);
*problem 418. Reflection and transmission coefficients of sound waves at the
junction of two tubes in terms of the overall impedance (note 7.52);
*problem 419. Scattering matrix for sound incident from either side at the
junction between ducts with different cross-sections and containing distinct
fluids (notes 7.53–7.54);
*problem 420. High (low) frequency limits of short (long) waves in the ray
(scattering) approximations to the exact diffraction theory of wave refraction
in inhomogeneous and/or unsteady media (note 7.55).

U. Oscillators with Several Degrees-of-Freedom

(sections 8.1–8.2):
U. a. Equations of Motion and Energy Balance
(section 8.1 and subsection 8.2.1):
*problem 421. Linear restoring and quadratic potential depending on the
position (subsections 8.1.1–8.1.2);
*problem 422. Linear kinetic friction force and quadratic dissipation function
depending on the velocity (subsections 8.1.1 and 8.1.3);
*problem 423. Equation of motion of a multi-dimensional linear oscillator,
with inertia, restoring, friction, and external forces, in the case of coupled or
decoupled degrees-of-freedom (subsections 8.1.1–8.1.4);
196 Differential Equations and Applications

*problem 424. Energy balance equation the work of the external forces minus
the dissipation by friction forces against the total energy, consisting of kinetic
and potential energies (subsections 8.1.5–8.1.6);
*problem 425. Particular case of decoupled oscillators with (i) separate equa-
tions of motion and (ii)(iii) potential (dissipation) functions that involve only
squares of position (velocity) and no cross-products (subsections 8.1.5–8.1.6);
*problem 426. Generalized equation of motion with explicit accelerations xr ,
involving damping λ rs and oscillations ω 2rs matrices, and reduced external
forces fr ( t ) (subsection 8.2.1):
N

r = 1,..., N :
d 2 xr
dt 2
+ ∑  2 λ
s=1
rs
dxs
dt

+ ω 2rs xs  = fr ( t ) .

(10.762a, b)

U. b. Simultaneous Multi-Dimensional Free

Oscillations (subsections 8.2.2–8.2.8):
*problem 427. Modes of a multidimensional linear (10.762a, b) unforced
(10.763a) oscillator as the roots (10.763c) of the determinant (10.763b) of the
dispersion matrix (subsection 8.2.2):
2N

Fr = 0: ( 2
0 = Det δ rs ξ + 2 λ rs ξ + ω 2
rs ) = ∏ ( ξ − ξ ) ; (10.763a–c)
n
n= 1

*problem 428. Modal frequencies for free (10.763a) undamped λ rs = 0 multi-

dimensional oscillations around a position of stable equilibrium (subsec-
tions 8.2.3–8.2.4);
*problem 429. Modal frequencies and dampings for the free (10.763a) damped
linear (10.762a, b) multi-dimensional oscillator (subsection 8.2.5);
*problem 430. Oscillation frequencies appearing in the modal coordinates of
linear free (10.763a) multidimensional (10.762a, b) oscillators (subsection 8.2.6);
*problem 431. The transformation matrix relating the modal to the physical
coordinates gives zero if multiplied by (problem 427) the dispersion matrix
(subsection 8.2.7);
*problem 432. The transformation matrix is specified by the co-factors of the
dispersion matrix evaluated at the modal frequencies and dampings (sub-
section 8.2.7);
*problem 433. Physical coordinates as functions of time for the free (10.763a)
linear multidimensional oscillator (10.762a, b) satisfying the compatibility
and initial conditions (subsection 8.2.8).

U. c.  Multiple Resonance and Forced Motions (subsections 8.2.8–8.2.16):

*problem 434. Transformation from the physical to the modal coordinates:
the inverse of the problem 431 (subsection 8.2.9);
Classification 10.2 197

*problem 435. Transformation between physical and modal forces (subsec-

tion 8.2.9);
*problem 436. Non-diagonal (diagonal) dispersion operators for the linear
multidimensional oscillator with damping and reduced external physical
(nodal) forces (subsection 8.2.10);
*problem 437. Diagonalization (10.764a, b) of the linearized forced multidi-
mensional oscillator (10.762a, b), in terms of modal coordinates q ( t ), reduced
forces g  (  ), frequencies ω , and dampings λ  (subsection 8.2.11):

d 2 q d 2 q
 = 1,..., N : 2
+ 2 λ  2 + ω 2 q = g  ( t ) ; (10.764a, b)
dt dt
*problem 438. Solution of the equations of motion (10.762a, b) for the linear
forced damped multidimensional oscillator via (10.764a, b) modal coordi-
nates (subsections 8.2.11–8.2.12);
*problem 439. Modal matrix for the sinusoidal forcing of the (10.762a, b) lin-
ear damped multidimensional oscillator (subsection 8.2.13);
*problem 440. Sinusoidal forcing of (10.762a, b) the undamped λ rs = 0 linear
multidimensional oscillator (subsection 8.2.14);
*problem 441. Sinusoidal forcing, beats, and resonance of an undamped
linear multidimensional oscillator, in terms of modal coordinates (subsec-
tion 8.2.15);
*problem 442. Sinusoidal forcing of a linear multidimensional damped oscil-
lator, in terms of modal coordinates (subsection 8.2.16);
*problem 443. Diagonalization to a sum of squares of the potential and
kinetic energies and dissipation function of a linear multidimensional
oscillator (10.762a, b), using modal coordinates (subsection 8.2.16);
*problem 444. Three (four) cases of free (forced) oscillations of (10.764a, b) the
linear multidimensional oscillator (subsection 8.2.16).

V.  Coupled Two-Dimensional Oscillators (sections 8.3–8.5)

 x 1   λ 11 λ 12   x 1   ω 12 ω 12 2   x1   f1 ( t ) 
 + 2  + 2  =   . (10.765)
 x2   λ 21 λ 22   x 2   ω 21 ω 22
2
  x2   f2 ( t ) 
  

V. a.  Mechanical and Electrical Analogues (section 8.3):

*problem 445. Modal frequencies for the undamped free oscillations of
two masses connected by three springs between two fixed walls (subsec-
tion 8.3.1);
198 Differential Equations and Applications

*problem 446. Modal frequencies for the undamped free oscillations of two
coupled electrical circuits analogous to the two-dimensional mechanical
oscillator in problem 445 (subsection 8.3.2);
*problem 447. Coupled vertical translational and rotational oscillations of a
two-wheeled vehicle with distinct springs and dampers and choice of the
moment of inertia and position of the center of mass (subsection 8.3.3);
*problem 448. Analogies among the variables, parameters, and forcing of the
coupled electrical (mechanical) circuits in the problem(s) 446 (445 and 447),
which satisfy a similar system of coupled linear differential equations (109)
with constant coefficients (subsection 8.3.4);
*problem 449. Analogous modal matrices, consisting of mass, damping and
resilience matrices, and analogous forcing vectors for problem(s) 446 (445
and 447) of electrical (mechanical) linear two-dimensional oscillators (sub-
section 8.3.5).

V. b. Translational and Rotational Free Oscillations of a Body

(section 8.4):
*problem 450. Modal (natural) frequencies for the coupled (decoupled) free
undamped rotational and translational oscillations of the two-wheeled vehi-
cle in problem 447 (subsection 8.4.1);
*problem 451. Natural frequencies of translation and rotation for the decou-
pled case of two masses at fixed distance, like a diatomic molecule (subsec-
tion 8.4.2);
*problem 452. Extension to the coupled modal frequencies for the diatomic
molecule (problem 451), which is a particular case of problem 447 ≡ 451 (sub-
section 8.4.3.);
*problem 453. Modal frequencies of the coupled oscillations of translation
and rotation of a rod with center of mass in the middle, which is a particular
case of problem 447 ≡ 451 (subsection 8.4.4);
*problem 454. Modal frequencies of the coupled translational and rotational
frequencies of a homogeneous rod, which is a particular case of problem 453
(subsection 8.4.5);
*problem 455. Modal coordinates for the free undamped coupled transla-
tional and rotational oscillations of a homogeneous rod, specifying the
points that oscillate only at one frequency (subsection 8.4.6);
*problem 456. Amplitudes and phases of the oscillations in problem 455,
determined from initial conditions for the translational coordinate and
velocity, and the rotation angle and angular velocity (subsection 8.4.7);
*problem 457. Method of application of initial conditions alternative to prob-
lem 456 (subsection 8.4.8);
*problem 458. Modal frequencies and dampings for the free coupled trans-
lational and rotational oscillations of a homogeneous rod in the case of
subcritical damping (subsection 8.4.9);
Classification 10.2 199

*problem 459. Explicit natural frequencies and dampings for the free decou-
pled translational and rotational oscillations of a homogeneous rod, as a par-
ticular case of problem 458 (subsection 8.4.10);
*problem 460. Four exact relations satisfied by the modal frequencies and
dampings of problem 458, for strong damping and coupling of translational
and rotational motions (subsection 8.4.11);
*problem 461. Weak modal dampings and unchanged modal frequencies for
problem 458 ≡ 460, in the case of weak damping (subsection 8.4.12).

V. c.  Forced Oscillations, Beats, and Resonance of a Rod (section 8.5):

*problem 462. Non-resonant undamped forced coupled translational and
rotational oscillations of a homogeneous rod (subsection 8.5.1);
*problem 463. As in problem 462, the resonant case when the applied fre-
quency of the sinusoidal forcing coincides with one of the modal frequencies
(subsection 8.5.2);
*problem 464. Forced coupled sinusoidal oscillations of an homogeneous rod
in the non-resonant case in terms of modal coordinates (subsection 8.5.3);
*problem 465. As in problem 464, the resonant case of applied frequency of
sinusoidal forcing coincident with one of the modal frequencies (subsections
8.5.3–8.5.4);
*problem 466. As in problem 464, in the case of beats, which is applied fre-
quency close to a modal frequency (subsection 8.5.4);
*problem 467. Sinusoidal forcing with damping of the coupled translational
and rotational oscillations of a inhomogeneous rod, both in the resonant and
non-resonant cases (subsection 8.5.5).

W.  Principle of the Vibration Absorber (section 8.6):

*problem 468. A vibration absorber is a secondary undamped unforced
oscillator coupled to a primary damped forced oscillator (subsection 8.6.1);
*problem 469. The forced vibrations in the primary damped oscillator are
transferred to the secondary damped oscillator if the natural frequency of
the latter equals the frequency applied to the former (subsection 8.6.2);
*problem 470. Proof that the transient oscillations decay in both the pri-
mary and secondary circuits, due to damping in the former, assuming equal
spring resiliences (condenser capacities) in the mechanical (electrical) case
(subsection 8.6.3);
*problem 471. Natural frequencies of the undamped coupled vibration
absorber, which coincide with the modal frequencies in the case of weak
damping (subsection 8.6.4);
200 Differential Equations and Applications

*problem 472. Modal dampings of the vibration absorber in the case of weak
damping (subsection 8.6.4);
*problem 473. Operation of the vibration absorber, including application of
the initial conditions (subsections 8.6.5–8.6.6).

X.  Radioactive Disintegration Chain (section 8.7):

*problem 474. Radioactive disintegration chain in which the total mass is
conserved and ultimately for long time is concentrated in the last element
(subsection 8.7.1);
*problem 475. Decay with time of the mass of the first element of the radioac-
tive disintegration chain (subsection 8.7.2);
*problem 476. Decay with time of the mass of the second element of the radio-
active disintegration chain in the non-resonant case of distinct disintegration
rates (subsection 8.7.2);
*problem 477. As in problem 476, the resonant case of coincidence of the first
two radioactive disintegration rates (subsection 8.7.3);
*problem 478. Decay with time of the mass of the third element in the radio-
active disintegration chain in the rotally non-resonant case when the three
disintegration rates are distinct (subsections 8.7.4–8.7.5);
*problem 479. As in problem 478, the singly resonant case when the first two
radioactive disintegration rates coincide and are distinct from the third (sub-
section 8.7.6);
*problem 480. As in problem 478, the doubly resonant case when all three
disintegration rates coincide (subsection 8.7.7);
*problem 481. Decay with time of the mass of the third element of the radio-
active disintegration chain in the single non-resonant (problem 478) and dou-
bly resonant (problem 480) cases, plus three singly-resonant cases, of which
problem 479 is one (subsection 8.7.7);
*problem 482. Decay of mass with time of the k-th element of the radioactive
disintegration chain in the non-resonant extreme of all disintegration rates
distinct (subsection 8.7.8);
*problem 483. As in problem 482, the opposite maximum resonance with all
disintegration rates equal (subsection 8.7.8);
*problem 484. As in problems 482 and 483, in the intermediate case of mul-
tiple resonances due to the coincidence of some but not all disintegration
rates (subsection 8.7.8);
*problem 485. Accumulation of the total mass of the radioactive disintegra-
tion chain towards the last element as a function of time (subsection 8.7.8);
*problem 486. Determination of all arbitrary constants in the mass as a func-
tion of time of k-th element of the radioactive disintegration chain from the
initial masses of all the preceding elements (subsection 8.7.8).
Classification 10.2 201

Y.  Sequence of Linked Oscillators (section 8.8):

*problem 487. Equations of motion for an N-dimensional forced damped
oscillator, consisting of a sequence of oscillators (subsection 8.8.1);
*problem 488. Modal frequencies of an undamped unforced oscillator with
three degrees-of-freedom associated with a sequence of three masses con-
nected by four springs between two fixed walls (subsection 8.8.2);
*problem 489. Particular cases of problem 488, when the middle mass is much
larger (smaller) than the other two leading to decoupling (strong coupling)
(subsection 8.8.3);
*problem 490. Comparison of a radioactive disintegration chain with N ele-
ments with a sequence of mechanical (electrical) forced oscillators consist-
ing of N masses (selfs) connected by N + 1 springs (capacitors) and dampers
(resistors) with applied mechanical (electromotive) forces (subsection 8.8.4);
*problem 491. Modal frequencies and dampings of the unforced
N-dimensional linear damped oscillator consisting of the sequence of
N oscillators in problem 487 (subsection 8.8.5);
*problem 492. Modal coordinates for the N-dimensional oscillator in prob-
lem 487, with modal dampings and frequencies in problem 491, and amplitudes
and phases determined by a sequence of continued fractions (subsection 8.8.6);
*problem 493. Determination of the modal amplitudes for three (four)-
dimensional oscillators in problem 492 (subsection 8.8.7).

Z.  Signals along a Transmission Line (section 8.9):

*problem 494 (495). Electrical (mechanical) impedance relating the electromo-
tive (mechanical) force to the electric current (velocity) in a circuit (system)
consisting of self (mass), resistor (damper), and capacitor (spring) elements
(subsections 8.9.1–8.9.2);
*problem 496. Resistance (inductance) as the real (imaginary) part of the
impedance, and relation with its inverse or admittance (subsection 8.9.2);
*problem 497. Five cases of operation of an infinite transmission line with
two distinct sets of impedances in series and parallel (subsections 8.9.2–8.9.4);
*problem 498. Two sets of three cases of transmission lines depending on which
of the self/resistance/capacitor lies in parallel or sequence (subsection 8.9.5);
*problem 499. Transmission line without dissipation, that is, without resis-
tances (subsection 8.9.6);
*problem 500. Five regimes of signal passage along a non-dissipative infinite
transmission line, including (a) signal reflection (b) frequency cut-off, and
(c) lossless transmission in a frequency pass band above a cut-off frequency
(subsections 8.9.6–8.9.7).
References

1678 Hooke, R. De Potentia Restitutiva. London, UK.

1686 Newton, I. Principia. Cambridge University Press, reprinted Dover 1934.
1691 Bernoulli, J. Acta Eruditorium 553 (Opera 1, 663).
1691 Bernoulli, J. Differentialrechnung. Manuscript found in 1921.
1695 Bernoulli, J. Acta Erudita 59–67, 537–557.
1696 Leibnitz, G. W. Acta Erudita 145 (Mathematische Werke 5, 329).
1696 L’Hôspital, G. F. A. Annalyse des Infinitement Petits pour l'Intelligence des Courbes.
Paris.
1697 Bernoulli, J. “Problema pure Geometricum Eruditis Propofisum: De
Conoitibus Spheroidibus Quadam.” Acta Evuditorium 113–124.
1724 Ricatti, J. F. “Animadversiones in Equationes Diferentiales Secundi Gradus.”
Acta Erudita Supplementa 8, 66–75.
1729 Euler, L. “Letter to Goldbach.”
1734 Clairaut, A. C. “Solution de Plusieurs Problemes ou il s’agit de Trouver les
Courbes dont la Proprieté Consiste dans une Certaine Relation entre les
Branches Exprimée par Une Equation Donnée.” Histoire de l’Academic de Paris,
196–215.
1743 Euler, L. Miscelanea Berologica 7, 193 (Institutiones Calculis Integralis 2, 375).
1748 D’Alembert, J. R. Histoire de l’Académie de Berlin 4, 275.
1748 D’Alembert, J. R. “Suite des Recherches sur le Calcul Integral. Quatiéme
Partie: Méthodes pour Intégrer Quelques Équations Differentielles.” Histoire
de l’Académie de Berlin 4, 275–291.
1762 D’Alembert, J. R. Miscelanea Turinesia, 381.
1763 D’Alembert, J. R. Histoire de l’Academie de Berlin, 242.
1765 Lagrange, J. L. Miscelanea Turinesia 3, 181 (Oeuvres 1, 473).
1766 Euler, L. “De Motu Vibratorio Tympanorum.” Novi Commentari Academia
Petropolitana 10, 243–260.
1769 Euler, L. Institutiones Calculis Integralis 2, 483.
1771 Van der Monde, A. T. “Memoire sur la Resolutions des Équations.” Histoire de
l’Académie des Sciences de Paris, 365–416 (1774).
1772 Euler, L. Novi Commentary Academia Scientarum Imperialis Petropolitana 16.
1772 Van der Monde, A. T. “Mémoire sur l’Elimination.” Histoire de l’Académie des
Sciences de Paris, 516–532 (1776).
1778 Euler, L. Nova Acta Academia Scientarum Imperialis Petropolitana 7, 58.
1785 Legendre, A. M. “Sur l’Attraction des Spheroides.” Mémoires de Mathématique
et Physique Presentées à l'Académie Royalle des Sciences par Divers Savants, 10.
1807 Poisson, S. D. “Mémoire sur la Théorie du Son.” Journal de l'École Polytechnique
7, 367–390.
1812 Gauss, C. F. “Disquisitiones circam Seriem Infinitam …” Commentationes
Societones Regiae Scientarum Göttensis Recentiores (Werke 1, 185).
1812 Hoene-Wronski, J. Refutation de la Théorie des Fonctions Analytiques de Lagrange.
Paris.
1812 Laplace, P. S. Théorie des Probabilités. Gauthier-Villars, Paris.

203
204 Differential Equations and Applications

1815 Cauchy, A. A. “Mémoire sur les Fonctions qui ne Peuvent Obtenir que Deux
Valeurs.” Journal de l'École Polytechnique 17, 29–107 (Oeuvres, série 2, 1, 91–169,
Gaultier-Villars, 1905, Paris).
1818 Fourier, J. B. Theorie Analytique de la Chaleur. Reprinted Dover, New York, 1953.
1824 Bessel, F. W. “Untersuchung der Theils der Planetarischen Storungen
Welchear aus des Bewegung der Sonne Entsteht.” Berliner Abhandlungen.
1827 Cauchy, A. L. Exercices Mathematiques 2, 159 (Oeuvres 7, 198).
1836 Kummer, E. E. “Ueber die Hypergeometrische Reihe.” Journal für die reine und
angewandte Mathematik 83, 127–172.
1838 Airy, G. B. “On the Intensity of Light Near a Caustic.” Transactions of the
Cambridge Philosophical Society 6, 379–402.
1850 Malmsten, C. J. “Théoremes sur l’Équation Differentielle.” Cambridge and
Dublin Mathematical Journal 5, 180–182.
1853 Riemann, B. “Üeber die Darstellbarkeit einer Function durch eine
Trignometrische Reihe.” Gescimmelte Mathematischan Werke 227–271, reprinted
Dover, New York.
1856 Weiertrass, K. T. W. “Theorie des Abel’schen Functionen.” Journal für
Mathematik 52, 285–379.
1859 Chebychev, P. L. “Sur les Questions de Minima, qui se Ratachent à la
Representation Approximative des Fonctions.” Mémoires de l'Académie
Scientique de St. Petersbourg 7, 199–291 (Oeuvres 1, 271–378).
1859 Jacobi, C. G. J. “Untersuchungen über die Differentialglechung der
Hypergeometrischen Reihe.” Journal für die reine und angewandte Mathematik
56, 149–165.
1863 Maxwell, J. C. Treatise of Electricity and Magnetism. Clarendan Press, Oxford.
1864 Hermite, C. “Sur un Nouveau Development en Série de Fonctions.” Comptes
Rendus de l’Académie des Sciences 58, 93–100, 266, 273 (Oeuvres 2, 293–308,
Gauthiers Villars, 1908).
1864 Lipshitz, R. “De Explicatione per Series Trignometricas Instituenda
Functionum Unices Variablis Arbitrarium, et Praecipue Earum, qua per
Variablis Spatium Finitum Valorum Maximum et Minimum Numerarum
Habent Infinitum Disquisition.” Journal für die reine und angewandte Mathematik
63, 296–308.
1866 Fuchs, J. L. “Zur Theorie der Linearen Differentialgleichungen mit
Verändlichen Koefflizienten.” Journal für die reine und angewandte Mathematik
66, 121–160; 68, 354–385.
1867 Newmann, C. G. Theorie der Bessel’schen Funktionen, Teubner, Leipzig.
1868 Lommel, E. C. J. “Zur Theorie der Bessel’schen Functionen.” Mathematische
Annalen 14, 510–536.
1869 Hankel, H. “Die Cylinderfunktionen Erster and Zweiter Art.” Mathematische
Annalen 1, 467–501.
1873 Frobenius, G. “Ueber die Integration des Linearen Differentialgleichungen
durch Reihen.” Journal für die reine und angewandte Mathematik 76, 214–235.
1873 Mathieu, E. Cours de Physique Mathematique. Gautiers-Villars, Paris.
∞
1879 Laguerre, E. “Sur l’integrale ∫
x
ξ −1   e −ξ   d ξ.” Bulletin de la Societé Mathematique

de France 7, 72–81 (Oeuvres 1, 428–437, reprinted Chelsea, 1971).

1879 Lommel, E. C. J. Studien über der Bessel'shen Functionen. Teubner, Leipzig.
References 205

1880 Pearson, K. “On the Solution of Some Differential Equations by Bessel

Functions.” Messenger 9, 127–131.
1881–6 Poincaré, H. “Mémoire sur les Courbes Définies par Une Equation
Differentielle.” Journal de Mathematiques 7, 375–422; 8, 251–296; 1, 167–244; 2,
151–217.
1883 Floquet, G. Annales Scientifiques de l'École Normale Superieure 12, 47–88.
1883 Floquet, G. “Sur les Équations Differentielles à Coefficients Periodiques.”
Annales Scientifiques de l'École Normale Supérieure 12, 14–88.
1883 Thomé, L. W. Journal für Mathematik 95, 75.
1886 Hill, G. W. “On the Part of the Motion of the Lunar Perigee which is a Function
of the Mean Motions of the Sun and the Moon.” Acta Mathematica 8, 1–36.
1886 Poincaré, H. “Sur les Integrales Irreguliéres des Equations Differentielles.”
Acta Mathematica 8, 295–344.
1886 Poincaré, H. “Sur les Determinants d’Ordre Infini.” Bulletin de la Societé
Mathematique de France 14, 77–90.
1892 Lyapunov, A. “The General Problem of Stability of Motion.” Communications
of the Mathematical Society of Kharkov.
1892 Poincaré, H. Méthodes Nouvelles de la Mécanique Celeste. Gauthier-Villars, Paris.
1892 von Koch, H. “Sur les Determinants Infinis et les Equations Differentielles
Ordinaires.” Acta Mathematica 16, 217–295.
1893 Picard, E. “Sur l’Application de la Méthode des Approximations Sucessives
à l’Étude de Certaines Équations Differentielles.” Journal de Mathématiques
Pures et Appliqués 9, 217–272.
1896 Mellin, H. “Ueber die Fundamentale Wichtigkeit des Satzes ven Cauchy
für die Theorie der Gamma und Hypergeometrischen Funktionen.” Acta
Societarum Scientarum Fennica 21, 1–115.
1904 Rayleigh, J. W. S. “On the Acoustic Shadow of a Sphere.” Philosophical
Transactions of the Royal Society A203, 87–110 (Scientific Papers 5, 149–171).
1904 Whittaker, E. T. “An Expression of Certain Known Functions as Generalized
Hypergeometric Functions.” Bulletin of the American Mathematical Society 10,
125–134.
1907 Foppl, F. Vorlesungen uber Technische Mechanik. Teubner, Leipzig.
1907 Lyapunov, A. “Probléme General de la Stabilité du Mouvement.” Annales de la
Faculté de Sciences de Toulouse 9, 203–469.
1909 Bateman, H. “The Solution of Linear Differential Equations by Means of
Definite Integrals.” Transactions of the Cambridge Philosophical Society 21,
171–196.
1910 von Kármán, “Festigkeit Probleme in Machinenbau.” Enzyklopedie der
Mathematischen Wissenschaften, Springer, Berlin.
1916 Rayleigh, J. W. S. “On the Propagation of Sound in Narrow Tubes of Variable
Cross-section.” Philosophical Magazine 31, 89–96.
1919 Webster, A. G. “Acoustical Impedance and the Theory of Horns and the
Phonograph.” Proceedings of the National Academy of Sciences 5, 275–282.
1922 van der Pol, B. “On Oscillation Hysteresis in a Triode Generator with Two
Degrees-of-Freedom.” Philosophical Magazine 43, 700–719.
1924 Jeffreys, H. “On Certain Approximate Solutions of Linear Differential
Equations of the Second Order.” Proceedings of the London Mathematical Society
23, 428–436.
206 Differential Equations and Applications

1926 Brillouin, L. “La Mécanique Ondulatoire de Schrödinger: Une Méthode

Génerale de Solution por Approximations Successives.” Comptes Redus de
l'Académie des Sciences de Paris 183, 24–26.
1926 Kramers, H. A. “Wellenmechanik und halbzahlige Quantisiering.” Zeitschrift
für Physik 39, 828–840.
1926 Wentzel, G. “Eine Verallgemeinering der Quarterbedindungen für die
Zwecke der Wellenmechanik.” Zeitschrift für Physik 38, 518–529.
1927 Ballantine, S. “On the Propagation of Sound in the General Bessel Horn of
Infinite Length.” Journal of the Franklin Institute 203, 85–101.
1930 Olson, H. F. “A Sound Concentrator for Microphones.” Journal of the Acoustical
Society of America 1, 410–417.
1931 Volterra, V. Leçons sur la Theorie Mathematique de la Lute pour la Vie. Gauthier-
Villars, Paris.
1938 Kontorovich, M. I. and Lebedev, N. N. “A Method for the Solution of Problems
in Diffraction Theory and Related Topics.” Journal of Experimental and
Theoretical Physics 8, 1192–1206 (in Russian).
1943 Hopf, E. “Abzweigung einer periodischen Lösung von einer stationären
Lösung eines Differentialsystems.” Akademie der Wissenschaten Leipzig,
Mathematische – Naturwissenschaften Klasse 95, 3–22.
1946 Salmon, V. “A New Family of Horns.” Journal of the Acoustical Society of America
17, 212–218.
1947 Chebyshev, P. L. Collected works. Moscow. 2, 25–51.
1953 Landau, L. D. and Lifshitz, E. F. Fluid Mechanics. Pergamon Press, Oxford.
1954 Lyapunov, A. Collected works. Moscow.
1962 Bies, D. A. “Tapering Bar of Uniform Stress in Longitudinal Oscillations.”
Journal of the Acoustical Society of America 34, 1567–1572.
1966 Lyapunov, A. Stability of Motion, Academic Press, New York.
1967 Pyle, R. W. “Duality Principle for Horns.” Journal of the Acoustical Society of
America 37, 1178A.
1971 Nagarkar, B. N. and Finch, R. D. “Sinusoidal Horns.” Journal of the Acoustical
Society of America 50, 23–31.
1971 Ruelle, D. and Takens, F. “On the Nature of Turbulence.” Communications in
Mathematical Physics 20, 167–192.
1972 Schwartz, H. A. “Beweis eines für die Theorie der trignometrischen Reihen in
Betracht kommenden Hülfssatzes,” Gesammelte mathematischen Abhandlengen
341–343, reprinted Chelsea, New York.
1984 Campos, L. M. B. C. “On Some General Properties of the Exact Acoustic Fields
in Horns and Baffles.” Journal of Sound and Vibration 95, 177–201.
1985 Campos, L. M. B. C. “On the Fundamental Acoustic Mode in Variable Area
Low Mach Number Nozzles.” Progress in Aerospace Sciences 22, 1–27.
1986 Campos, L. M. B. C. “On Waves in Gases. Part I: Acoustics of Jets, Turbulence,
and Ducts.” Reviews of Modern Physics 58, 117–182.
2000 Campos, L. M. B. C. “On the Singularities and Solutions of the Extended
Hypergeometric Equation.” Integral Transforms and Special Functions 9,
99–120.
2001 Campos, L. M. B. C. “On the Derivation of Asymptotic Expansions for
Special Functions from the Corresponding Differential Equations.” Integral
Transforms and Special Functions 12, 227–236.
References 207

2001 Campos, L. M. B. C. “On the Extended Hypergeometric Equation and

Functions of Arbitrary Degree.” Integral Transforms and Special Functions 11,
233–256.
2001 Campos, L. M. B. C. “On Some Solutions of the Extended Confluent
Hypergeometric Differential Equation.” Journal of Computational and Applied
Mathematics 137, 177–200.
2012 Campos, L. M. B. C. and Cunha, F. S. R. P. “On Hyper-spherical Legendre
Polynomials and Higher Dimensional Multipole Expansions.” Journal of
Inequalities and Special Functions 3, 1–28.
2014 Campos, L. M. B. C. and Marta, A. C. “On the Prevention or Facilitation of the
Buckling of Beams.” International Journal of Mechanical Sciences 79, 95–104.
Index

Acoustic rays, see Optics and acoustic Asymptotic stability of damped non-
rays linear oscillator, 82–84
Acoustics of horns, 193–195 negative-definite stability function,
exact elementary solutions, 194 83–84
Gaussian horn and displacement positive-definite stability function,
amplifier, 194 83
power law horns and cylindrical/ Axial traction, 179–180
spherical waves, 194–195
properties, 193–194 Balance equations for strong bending of
reflection and transmission at isotropic elastic plate, 186
interfaces, 195 Balance of forcing, 10–11
Adjoint kernel operator, 141 Banded diagonal system, 126
Airy differential equation, 175 Bars, 159, 182
Airy functions of inverse variable, 91–92 absence of traction, 47
Amplification, fast, 10–11 elástica of, 42–43
Amplified oscillator, 161 equation, 39
Amplitude, 168 pinned, 178
decays, 55 straight, 179–180
displacement is a quadratic function, Beams
13 bending under traction, see Bending
excitation, 167–168 of beam under traction
excitation frequency and, 166–167 cantilever with linearly increasing
Analogies among six types of waves, shear stress, 45–48
193 clamped with concentrated torque,
Analytic coefficients, 127–130; see also 37–40
Linear differential equations elastica under traction, 36–37
with analytic coefficients heavy clamped-pinned, 43–45
Analytic integrals, 90–91 linear bending under traction, 35–36,
Arbitrary order, 126 36
Associated Laguerre differential pinned with concentrated force,
equation, 147–148 40–43
Associated Legendre differential Beats, 199
equation, 145 Bending of beam under traction, 35–48
Asymptotic integrals cantilever beam with a linearly
analytic and high-order poles, 90–91 increasing shear stress, 45–48
for cylindrical Bessel differential clamped beam with concentrated
equation, 136 torque, 37–40
for generalized Bessel differential elastica of a beam under traction,
equation, 136–137 36–37
for spherical Bessel differential heavy clamped-pinned beam, 43–45
equation, 136 linear bending, 35–36, 36, 179
triple pole and, 88–90 pinned beam with concentrated
Asymptotic point, 165, 166 force, 40–43

209
210 Index

Bernoulli non-linear differential Chebychev differential equation; see also

equation, 102 Differential equations
Beta function, integral transforms, of first type, 147
139–140 of second type, 147
Bifurcations Circular functions of inverse variable,
of dynamical systems, 172 91–92
of generalized Bessel differential Circular heavy plate with circular hole,
equation at infinity, 137 linear bending of, 48–51, 49,
Bi-quadratic potential for soft/hard 50, 51
spring, 169 Circular pendular motion, 171–172
Boundary conditions large motion without damping,
for bending of plate, 67–68, see also 171
General boundary conditions with quadratic damping, 172
for bending of plate Circular plate with free boundary, 67
for isotropic elastic plates, 182 Circulatory motion, 171
for pinned plates, 66 Clairaut differential equation, 110; see
for pseudo-isotropic orthotropic also Differential equations
plates, 185 Clamped beam with concentrated
for supported plates, 66 torque, 37–40; see also Beams
for transverse displacement and Clamped plate, 65–66
stress function, 186–187 Clebsch potentials, 107
Bounded coefficients, linear differential Combinations of differential operators,
equations with, 125 113
Bounded oscillation, 163 Common center, 165
Buckling of clamped-free plate with Complete integrability, 59–60
uniaxial stresses, 185 Complete integrals, 97
Buckling of clamped rectangular plate Complete solutions, 81, 81–82
with normal stresses, 185 Complex variable, points of functions
Buckling of elastic beams, 176–179 of, 127–128
bending of beam continuously Compression of straight bar, 179–180
supported on springs, 178 Confluent hypergeometric function,
general equations, 176 147–148
linear bending of beam under Constant coefficients; see also Linear
traction, 179 differential equations with
linear buckling, 176 constant coefficients
non-linear buckling to lowest order, forced coupled linear differential
176–177 system, 75–76
opposition to or facilitation of linear differential equation, 1–3,
buckling, 177–178 29–30, 95–97
with point rotary spring, 177 linear finite difference equation,
with point translational spring, 4–5
177 simultaneous linear system of
differential equations, 114–116
Cantilever beam with linearly simultaneous system of linear finite
increasing shear stress, difference equations, 117–119
45–48 Contact point, 164, 165
Case of constant forcing, 80–81 Continuous function, 164
Chaotic motions, 173 Continuous spectrum, 101, 163
Characteristic polynomial, 72–82 Coordinates, 120, 121
Index 211

Coupled systems with matrix of of elastic membrane under

characteristic polynomials, anisotropic in-plane stresses,
74–82 180–181
case of constant forcing, 80–81 Deformable media, mechanics and
complete solutions, 81, 81–82 thermodynamics of, 183–184
forced coupled linear differential Degenerate cases, 161
system with constant Diagonal system, 126
coefficients, 75–76 Differential equation(s); see also
forced coupled linear differential Generalized circular and
system with homogeneous hyperbolic differential
coefficients, 77 equation
forced coupled system of finite Airy, 175
difference equations, 77–78 applications to physical and
forcing by power multiplied by engineering problems, 159–201
cosine, 78–79 of arbitrary order, 126
general solutions, 81, 81–82 Clairaut, 110
system of finite difference equations cylindrical Bessel, 134–135
with oscillatory forcing, 78 D’Alembert, 110
Coupled two-dimensional oscillators, depression of order N greater than
197–199 2, 112
beats, 199 doubly-extended Bessel, 150
forced oscillations, 199 dual first-order, 111
mechanical and electrical analogues, factorizable linear, 113
197–198 first-order-differential equations,
resonance of rod, 199 15–17
translational and rotational free first-order special, 25–27, 26, 27
oscillations of body, 198–199 generalized autonomous system of,
Cubic potential, 172 122
Curve as tangent to vector field, higher-order homogeneous, 113
56–57 homogeneous, 104
Cusp catastrophe, 173 linear autonomous system with
Cusped parabola as envelope of family regular singularity, 131
of straight lines, 25–26, 26 linear coupled systems of, 71–74
Cylindrical Bessel differential equation, linear first-order, 17–18
134–135 linear of order N, 123
linear second-order with regular
D’Alembert differential equation, 110 singularity, 132
Damped non-linear oscillator, non-linear coupled systems of, 71–74
asymptotic stability of, 82–84 non-linear of order N, 122
Damped oscillator, 161 numerical solution of, 174
forcing of, 13–15 of order M in N variables, 108–109
sinusoidal forcing with, 162 second-order reduced to first order,
unforced and, 161 111–112
Damping simultaneous linear system with
frequency and, 162 constant coefficients, 114–116
subcritical, 162 simultaneous linear system with
Deflection homogeneous derivatives,
and bending with elliptic boundary, 116–117
187 simultaneous system, 113–114
212 Index

Differential equation(s) (continued) Elastica of beam under traction, 36–37

spherical Bessel linear with regular Elastic displacements, 180
singularities, 134–135 Elastic energy along boundary of plate,
transformation of, 111 61–62
Differential equations of order higher Elastic materials, 184
than first, 28–35 Elastic media, see Bars; Beams;
higher-order reducible to first order, Membranes; Plates, Rods
31–32 Elastic membrane, deflection under
linear non-commutative with anisotropic in-plane stresses,
variable coefficients, 32–33 180–181
linear with constant coefficients, Elastic plate
29–30 isotropic, 182, 186
linear with variable coefficients, linear bending circular with
29–30 concerntric circular hole, 50
non-linear exact with variable Elastic string, 191
coefficients, 33–34 Electrical dynamo, 170
non-linear second-order omitting Electrical self-resistor-capacitor-battery
independent variable, 30–31 circuit, 160
second-order homogeneous Electromagnetic waves, 192
differential equation, 34–35 Electromechanical dynamo, 170–171
Differential operators, combinations of, exact homopolar dynamo, 171
113 self-excited dynamo, 171
Digamma function, 86, 133 with small resistance, 171
Dirac function, 101 Elementary integrals and double pole,
Discrete spectrum, 101, 163 87–88
Discriminants, 109 Elliptic boundary, 187
Dispersive wave, 190 Energies, 5–11, 195–196
Displacement, 192; see also Bars; Beams; averages over period of, 8–10
Membranes; Plates; Rods kinetic, 5–6
elastic, 180 potential, 5–6
Gaussian horn and amplifier, 194 total, 5–6
stress function and boundary Equations of motion, 195–196
conditions for transverse, Euler Beta function, 139
186–187 Euler potential, 107
transverse of pseudo-isotropic Exact differential equations, 113; see also
orthotropic plate, 184 Differential equation(s)
transverse with constant bending first-degree differentials in three/
stiffness, 181–182 four variables, 58–59, 106
transverse with non-uniform non-linear with variable coefficients,
bending stiffness, 182 33–34
Dissipation, 5–11 Exact electromechanical homopolar
Double pole and elementary integrals, dynamo, 171
87–88; see also Integrals Exact elementary solutions, 194
Doubly-extended Bessel differential Exact first-order differentials in two
equation, 150 variables, 19
Dual first-order differential equations, Excitation frequency
111 parametric resonance, 166–167
Duality principles for wave variables, parametric resonance with weak
193 damping for several, 168
Index 213

Existence of normal integral near second order reduced to, 111–112

irregular singularity, 135–136 solvable by quadratures, 102
Existence of regular integrals near solvable for slope/variable, 110
irregular singularity, 135 special integrals in, 110
Existence theorems, 121 in two variables, 18–21, 104–105
Explicit slope, first-order differential First-order special differential
equations with, 121–122 equations, 25–27
Extended confluent hypergeometric cusped parabola as envelope of
differential equation, 149–150 family of straight lines, 25–26,
Extended Gaussian hypergeometric 26
differential equation, 149 first-order equation of degree three
Extension of balance equations to without singular integrals, 27
pseudo-isotropic orthotropic smooth parabola as envelope of
plate, 186 family of straight lines, 26–27,
27
Factorizable linear differential equation, First transformed generalized Bessel
113 differential equation, 153–155;
Fast amplification, 10–11 see also Differential equation(s)
First-degree differential equations in Floquet theorem, 126
three/four variables, 58–60 Focal point, 165, 166
complete integrability, 59–60 Fold catastrophe, 172
exact differentials, 58–59, 106 Force/forcing
immediate integrability, 59–60 balance of, 10–11
inexact differentials, 58–59, 106 case of constant, 80–81
non-integrable differentials, 58–59 by continuous function, 164
subsidiary conditions, 59–60 by continuous spectrum, 163
First-order differential equation(s), coupled linear differential system
15–17 with constant coefficients,
of degree three without singular 75–76
integrals, 27 coupled linear differential system
discriminants and special points of, with homogeneous coefficients,
109 77
dual, 111 coupled system of finite difference
exact differential, 107 equations, 77–78
exact first-order differentials in two of damped oscillator, 13–15
variables, 19 by discrete spectrum, 163
with explicit slope, 121–122 by Gaussian function, 164
having special integrals, 110 Green’s function, 163
homogeneous differential equation, of harmonic oscillator, 11–15
104, 108 by integrable function, 163
inexact differential, 107 involving power of time, 162
inexact first-order differentials in linear differential equations with
two variables, 19–20 constant coefficients, 95–97
integrating factors for inexact linear finite difference equation with
differentials, 20–21 constant coefficients, 100
in N ≥ 4 variables, 107 monotonic, 12–13
Pfaffian, 107 oscillating, 12–13, 199
quadratic discriminants of, 109 pinned beam with concentrated,
Ricatti equation, 102–104 40–43
214 Index

Force/forcing (continued) Generalized circular and hyperbolic

power law, 11–15 differential equation, 84–92,
by power multiplied by cosine, 78–79 129–130; see also Differential
system of finite difference equations equation(s)
with oscillatory, 78 Airy functions of inverse variable,
of undamped oscillator, 13–15 91–92
Fourier integral, 163 circular functions of inverse variable,
Fourier integral/transform, 101 91–92
Fourier series, 101, 163 double pole and elementary
integrals, 87–88
Gamma function, 133 high-order poles and asymptotic
Gaussian function, 164 analytic integrals, 90–91
Gaussian horn and displacement hyperbolic functions of inverse
amplifier, 194 variable, 91–92
Gaussian hypergeometric differential simple pole and regular integrals,
equation, 144, 149 85–87
General boundary conditions for triple pole and asymptotic integrals,
bending of plate, 60–68, 61 88–90
circular plate with free boundary, 67 Generalized Hill differential equation,
elastic energy along boundary of 149
plate, 61–62 Generalized hypergeometric
general clamped plate, 65–66 differential equation, 142–143
general plate with free boundary, 67 Generalized isotropic equation of
integration by parts along closed mathematical physics, 119–121
regular boundary, 62–63 Helmholtz equation involving
for pinned plates, 66 Laplacian in hyperspherical
plate with arbitrary closed regular coordinates, 120, 121
boundary, 64 hypercylinderical coordinates, 120,
pseudo-isotropic orthotropic plates, 121
65 hyperspherical coordinates, 120
rectangular clamped plate, 65–66 reduction to Helmholtz equation,
rectangular plate with free 119–120
boundary, 67 Generalized Laplace transform, 140
sets of boundary conditios for Generalized Neumann function of
bending of plate, 67–68 integer order, 134
stress couple and turning moment, extension to complex order, 134
61, 63–64, 68 General plate with free boundary,
for supported plates, 66 67
General clamped plate, 65–66 General potential for non-linear,
General irregular integrals of two undamped, unforced
kinds, 137–138 oscillator, 169
Generalized autonomous system of General solutions, 81, 81–82
differential equations, 122 Gradient dynamical system, 172
Generalized Bessel differential Green’s function, 101, 163
equations
asymptotic integrals for, 136–137 Harmonic oscillator
bifurcation at infinity, 137 excitation frequency, 166–167
first class of transformed, 153–155 forcing of damped oscillator,
second class of transformed, 155–158 13–15
Index 215

forcing of undamped oscillator, Inexact differential equations

13–15 first-degree in three/four variables,
power-law forcing of, 11–15 58–59, 106
Heavy clamped-pinned beam, 43–45 first-order in two variables, 19–20
Helmholtz equation Infinite determinants and linear
involving Laplacian in systems, 138
hyperspherical coordinates, Influence function, see Green’s function
120, 121 In-plane stresses, 180–181, 183, 186
reduction, 119–120 Integrable forcing function, 163
Hermite differential equation, 147–148; Integral(s)
see also Differential equation(s) analytic, 90–91
Higher-order differential equations; see asymptotic, see Asymptotic integrals
also Differential equation(s) complete, 97
homogeneous, 113 double pole and elementary, 87–88
reducible to first order, 31–32 existence of regular near irregular
High-order poles and asymptotic singularity, 135
analytic integrals, 90–91 first-order differential equations
Hill differential equation, 139, 149 having special, 110
Homogeneous coefficients first-order equation of degree three
forced coupled linear differential without singular, 27
system with, 77 generalized Bessel differential
linear differential equation with, 3–4, equations asymptotic, 136–137
97–99 high-order poles and asymptotic
Homogeneous differential equation, 104 analytic, 90–91
derivatives of simultaneous linear irregular singularity existence of
system, 116–117 regular, 135
of first kind order N, 113 of two kinds, 137–138
higher-order, 113 Integrals transform
of second kind order N, 113 Beta function, 139–140
second-order, 34–35 generalized integral transform and
Homogeneous first-order differential adjoint Kernel operator, 141
equations generalized Laplace transform along
in N variables, 108 path in complex plane, 140
in two variables, 108 Kernel and path of solution by,
Hooke law, 184 139–141
Hopf bifurcation, 172 Integrating factor
Horns, 192; see also Acoustics of horns for inexact differentials, 20–21
Hyperbolic functions of inverse for linear first-order differential
variable, 91–92 equation, 17–18
Hypercylinderical coordinates, 120, 121 Integration by parts along closed
Hypergeometric differential equation, regular boundary, 62–63
142–143 Integro-differential equation, 175; see
Hypergeometric function, 147–148 also Differential equation(s)
Hyperspherical coordinates, 120 Internal energy, 184
Hyperspherical Legendre differential Intersection surfaces, 56–57
equation, 145–146 Invariant differential equation, 150–153;
see also Differential equation(s)
Immediate integrability, 59–60 Invariant form of optics and acoustic
Index of singularity, 173–174 rays, 175
216 Index

Inverse variable with periodic coefficients, 126–127

Airy functions of, 91–92 points of, 128
circular functions of, 91–92 with variable coefficients, 29–30
hyperbolic functions of, 91–92 Linear differential equations with
Irregular singularity analytic coefficients, 127–130
existence of normal integral, 135–136 generalized circular functions,
existence of regular integrals, 135 129–130
Isotropic elastic plate, 182, 186 generalized hyperbolic functions,
129–130
Jacobi differential equation, 146–147; methods of solution, 128–129
see also Differential equation(s) points of functions of complex
Jordan matrix, 126 variable, 127–128
points of linear differential equation,
Kinetic energy, 5–6 128
Linear differential equations with
Laguerre differential equation, 147–148; constant coefficients, 1–3,
see also Differential equation(s) 29–30, 95–97
Laplace transform, 102 complete integrals, 97
Laurent series, 127, 128, 138 forced, 95–97
Legendre differential equation, 144–145; unforced, 95
see also Differential equation(s) Linear differential equations with
associated, 145 irregular similarities, 135–139
hyperspherical, 145–146 asymptotic integrals for cylindrical
ultraspherical, 145 Bessel, 136
Legendre transformation, 111 asymptotic integrals for generalized
Light and sound rays, 176 Bessel, 136–137
Limit cycles, 173 asymptotic integrals for spherical
Linear bending Bessel, 136
of beams under traction, 35–36, 36, bifurcation of generalized Bessel at
179 infinity, 137
circular elastic plate with a existence of normal integral near
concentric circular hole, 50 irregular singularity, 135–136
of circular heavy plate with circular existence of regular integrals near
hole, 48–51, 49, 51 irregular singularity, 135
of elastic isotropic plates, 181–183 general irregular integrals of two
Linear buckling of elastic beams, 176 kinds, 137–138
Linear coupled systems of differential infinite determinants and linear
equations, 71–74 systems, 138
Linear differential equation(s); see multiple parametric resonance and
also Non-linear differential Hill, 139
equations Linear differential equations with
with analytic coefficients, 127–130 regular singularities, 131–135
with bounded coefficients, 125 cylindrical Bessel differential
factorizable, 113 equation, 134–135
with given characteristic polynomial, generalized Bessel function, 133
72–74 generalized Neumann function of
with homogeneous coefficients, 3–4, integer order, 133–134
97–99 linear autonomous system of, 131
of order N, 123 of order N, 131
Index 217

second-order, 132 Mechanics and thermodynamics of

spherical Bessel, 134–135 deformable media, 183–184
Linear elastic stability of stressed Membranes
orthotropic plate, 183–185 deflection of elastic membrane under
boundary conditions for pseudo- anisotropic in-plane stresses,
isotropic, 185 180–181
buckling of clamped-free plate with vibrations of membrane under
uniaxial stresses, 185 uniform/non-uniform tension,
buckling of clamped rectangular 51–56, 52, 53
plate with normal stresses, 185 Mixed non-linear oscillator, 172
linear strong bending with Momentum equation, 183
anisotropic in-plane stresses, Monotonic and oscillating forcing, 12–13
183 Motion(s)
mechanics and thermodynamics of chaotic, 173
deformable media, 183–184 circular pendular, 171–172
stress-strain relation for elastic circulatory, 171
materials, 184 equations and energy balance,
transverse displacement of a pseudo- 195–196
isotropic, 184 multiple resonance and forced,
Linear finite difference equation with 196–197
constant coefficients, 4–5, oscillatory, 171, 172
99–100 Motions near stagnation point, 164–166
forced, 100 of first degree, 164–165
unforced, 99 of second degree, 165
Linear first-order differential equation, trajectories near first degree, 165–166
integrating factor for, 17–18 Multiple parametric resonance, 139
Linear non-commutative differential Multiple resonance and forced motions,
operators with variable 196–197
coefficients, 32–33
Linear second-order differential Negative-definite time derivative of
equations, 174 stability function, 83–84
invariant and self-adjoint, 150–153 Newton law, 183
invariant unforced, 175 Nodal point, 164, 166
optics and acoustic rays, 174 Non-dispersive wave, 190
solution by continued fractions, Non-integrable differentials of first-
141–142 degree in three/four variables,
Linear strong bending with anisotropic 58–59
in-plane stresses, 183 Non-linear, undamped, unforced
Linear systems and infinite oscillator, 168–169
determinants, 138 bi-quadratic potential for soft/hard
Lipshitz condition, 121, 122 spring, 169
for general potential, 169
MacLaurin series, 127–129 quartic potential generalizing the
Mass-damper-spring-force system, bi-quadratic potential, 169
160 Non-linear buckling, 176–177
Mathieu differential equation, 149; see Non-linear coupled systems of
also Differential equation(s) differential equations, 71–74
Mechanical mass-damper-spring-force Non-linear coupling of strong bending
system, 160 and in-plate stresses, 186–187
218 Index

Non-linear differential equations; coupled two-dimensional, 197–199

see also Linear differential damped, 13–15, 161
equation(s) forcing of harmonic, 11–15
Bernoulli, 102 forcing of undamped, 13–15
exact with variable coefficients, harmonic, see Harmonic oscillator
33–34 mixed non-linear, 172
of order N, 122 motion, 171, 172
Non-linear oscillator, 168–169 non-linear, 168–170
mixed, 172 power-law forcing of harmonic, 11–15
resonance of, 170 power of, 5–11
Non-linear second-order differential sequence of linked, 201
equations omitting strong subcritical, 7–8
independent variable, 30–31 system of finite difference equations
Non-linear waves, 188–189 with forcing, 78
Non-uniform tension, see Vibrations of translational and rotational free
membrane under uniform/ oscillations of body, 198–199
non-uniform tension two orthogonal undamped, 166
Numerical solution of differential undamped, 13–15, 168–169
equations, 174; see also weakly damped, 7–8, 168
Differential equation(s) Oscillators with several degrees-of-
freedom, 195–197
One finite difference (two differential) equations of motion and energy
equation(s), 1–5; see also balance, 195–196
Differential equation(s) multiple resonance and forced
linear finite with constant motions, 196–197
coefficients, 4–5 simultaneous multi-dimensional free
linear with constant coefficients, 1–3 oscillations, 196
linear with homogeneous
coefficients, 3–4 Parabola
Optics and acoustic rays, 174–176 cusped, 25–26, 26
integro-differential equation, 175 smooth, 26–27, 27
invariant form, 175 Parametric resonance, 166–168
invariant unforced linear second- excitation frequencies leading to, 167
order differential equation, 175 harmonic oscillator and excitation
light and sound rays, 176 frequency, 166–167
linear second-order differential stability/instability/indifference of
equation, 174 solutions, 167
self-adjoint form, 175 undamped, 167–168
waves in refracting medium, 175 undamped oscillator with
Orthotropic plates, 65 coefficients depending on time,
Oscillating forcing 166
monotonic forcing and, 12–13 with weak damping for several
system of finite difference equations excitation frequencies, 168
with, 78 Pendular motion with quadratic
Oscillation/oscillator damping, 172
amplified, 161 Periodic coefficients, linear differential
asymptotic stability of damped non- equations, 126–127
linear, 82–84 Perturbation expansion, 187
bounded oscillation, 163 Pfaffian, 106, 107
Index 219

Physical and engineering problems, Power of oscillation, 5–11

159–201 averages, 8–10
Picard method, 121 Pseudo-isotropic orthotropic plates, 65
Pinned beam boundary conditions for, 185
with concentrated force, 40–43 extension of balance equations to, 186
heavy clamped, 43–45 strong bending of, 68–71
Plane elastic displacements, 180 transverse diplacement of, 184
Plane elasticity, 180
Plates Quadratic damping of circular pendular
with arbitrary closed regular motion, 172
boundary, 64 Quadratures, 102
boundary conditions for pinned, 66 Quartic potential, 169, 173
boundary conditions for supported, Quasi-one-dimensional waves in
66 inhomogeneous media, 191–192
circular with free boundary, 67
elastic energy along boundary of, Radioactive disintegration chain, 200
61–62 Rectangular clamped plate, 65–66
orthotropic, 65 Rectangular plate with free boundary,
pseudo-isotropic orthotropic, 65 67
rectangular clamped, 65–66 Reflection and transmission at
rectangular with free boundary, 67 interfaces, 195
weak linear bending of elastic Regular integrals and simple pole,
isotropic, 181–183 85–87
weak linear bending of isotropic Resonance; see also Parametric
circular, 182–183 resonance
Point rotary spring, 177 of non-linear oscillator, 170
Points of functions of complex variable, of rod, 199
127–128 Ricatti equation, 17, 102–104
Points of linear differential equation, Riemann surface, 128
128 Robustness condition, 121
Point translational spring, 177 Robustness theorems, 121–123
Poles Rods
asymptotic analytic integrals and elastic, 188, 192, 194
high-order, 90–91 homogeneus, 198, 199
asymptotic integrals and triple, inhomogeneus, 180, 199
88–90 resonance of, 199
elementary integrals and double, Rotary spring, 177
87–88
regular integrals and simple, 85–87 Saddle point, 25, 164, 166
Polynomial potentials and catastrophes, Scalar potential, 106
172–173 Schwartz group, 143
Positive-definite stability function, 83 Second-order differential equations
Potential energy, 5–6 homogeneous, 34–35
Power-law invariant and self-adjoint linear,
forcing of harmonic oscillator, 11–15 150–153
monotonic and oscillating forcing, in invariant form, 127
12–13 linear, 174
Power law horns and cylindrical/ linear differential equation with
spherical waves, 194–195 regular singularities, 132
220 Index

Second-order differential equations Sinusoidal forcing, undamped oscillator

(continued) with, 161
reduced to first order, 111–112 Smooth parabola as envelope of family
solution by continued fractions, of straight lines, 26–27, 27
141–142 Soft/hard spring, 169
Second transformed generalized Bessel Solenoidal, 107
differential equation, 155–158 Sound waves, 192
Self-adjoint differential equation, Special integrals, 109
150–153 Spherical Bessel differential equation,
Self-adjoint form of optics and acoustic 134–135
rays, 175 Spherical harmonics, 145
Self-excited dynamo, 171 Spiral flow, 172
Sequence of linked oscillators, 201 Spring(s)
Sets of boundary conditions for bending bi-quadratic potential for soft/hard,
of plate, 67–68 169
Shear stress, cantilever beam with rotary, 177
linearly increasing, 45–48 translational, 177
Signals along transmission line, 201 Stability function
Simple point, 165, 166 negative-definite time derivative of,
Simple pole and regular integrals, 85–87 83–84
Simultaneous forcing, 162 positive-definite, 83
Simultaneous linear system of Stability of equilibrium of system,
differential equations with 123–126
constant coefficients, 114–116 classification, 123–124
forced, 115–116 specified by autonomous differential
unforced, 115 system, 125–126
Simultaneous linear system of specified by Liapounov function, 124
differential equations with Stability of orbits of dynamical systems,
homogeneous derivatives, 173
116–117 Stagnation point of first degree, 165–166
forced, 117 Stagnation point of second degree,
unforced, 116 21–25
Simultaneous multi-dimensional free comparison of velocity fields with,
oscillations, 196 24–25
Simultaneous system of differential paths near, 165
equations, 113–114 paths tangent at, 21–23, 22
Simultaneous system of linear finite paths with three asymptotes
difference with constant crossing, 23–24, 24
coefficients, 117–119 Stopping point, 171
forced, 118–119 Strains and stresses, 180
unforced, 118 Stress couple, 61, 63–64, 68
Simultaneous system of non-linear Stress-strain relation for elastic
ordinary differential materials, 184
equations, 71–72 Strong bending of pseudo-isotropic
Singularity orthotropic plate, 68–71
index of, 173–174 Strong non-linear bending of
irregular, see Irregular singularity heavy circular plate under
linear differential equations with compression, 187
regular, 131–135 Strong subcritical oscillations, 7–8
Index 221

Subcritical damping, 162 Unforced and amplified oscillator, 161

Subhypergeometric differential Unforced and damped oscillator, 161
equations, 148–150; see also Unforced differential equations, 95
Differential equations linear with constant coefficients,
Subsidiary conditions for first-degree 95
differentials in three/four linear with homogenous coefficients,
variables, 59–60 97
Superhypergeometric differential Unforced oscillator, 168–169
equations, 148–150; see also Unicity theorem, 121
Differential equations Uniformity theorem, 122–123
Surface waves, 192 Unit impulse, see Dirac function
Suspended pendulum with support
oscillating vertically, 168 Van der Pol (mixed non-linear)
System of finite difference equations oscillator, 172
with oscillatory forcing, 78 Variable coefficients
linear differential equations with,
Tension, non-uniform, 55–56 29–30
Three-dimensional potentials, linear non-commutative differential
106–107 operators with, 32–33
Torsion, 192 non-linear exact differential equation
Total energy, 5–6 with, 33–34
Transformation factor, 150 Variation of parameters, 100
Transformation of differential Vector fields, 56–57
equations, 111 with stagnation point, 24–25
Translational and rotational free Vector potential, 106–107
oscillations of body, 198–199 Vibration absorber, 199–200
Translational spring, 177 Vibrations of elastic media, 188–191
Transmission lines, 201 generalized and analogous wave
Transverse displacement of elastic equations, 189
plates generalized classical bending
boundary conditions for, 186–187 wave equation in one-
with constant bending stiffness, dimension with constant
181–182 coefficients, 190
with non-uniform bending stiffness, non-dispersive and dispersive
182 free wave propagation and
of pseudo-isotropic, 184 standing modes, 190
Triple pole and asymptotic integrals, non-linear waves in inhomogeneous
88–90 and unsteady media, 188–189
Turning moment, 61, 63–64, 68 wave generation by out-of-
Two orthogonal undamped oscillators, equilibrium initial conditions
166 and forcing, 190–191
Vibrations of membrane under
Ultraspherical Legendre differential uniform/non-uniform tension,
equation, 145 51–56, 52, 53
Undamped oscillators, 160, 168–169 cut-on/cut-off modes, 54–55
force/forcing of, 13–15 propagating/evanescent modes,
with sinusoidal forcing, 161 54–55
two orthogonal, 166 turning point due to non-uniform
unforced and, 160 tension, 55–56
222 Index

Waves Waves in inhomogeneous media,

analogies, 193 191–193
dispersive, 190 analogies among six types of waves,
of elastic media, 188–191 193
electromagnetic, 192 classification of wave equations,
equations, 193 191
generation by out-of-equilibrium duality principles for wave variables,
initial conditions, 190–191 193
non-dispersive, 190 quasi-one-dimensional, 191–192
non-linear, 188–189 Weak linear bending of elastic isotropic
power law horns and cylindrical/ plates, 181–183
spherical, 194–195 Weak linear bending of isotropic
in refracting medium, 175 circular plates, 182–183
sound, 192 Weakly damped oscillations, 7–8, 168
surface, 192 Whittaker differential equation,
variables, 193 150–151