Sie sind auf Seite 1von 808
® INDIA EDITION DIFFERENTIAL EQUATIONS any RESTRICTED! FOR SALE ONLY IN Shepley L. Ross —__ DIFFERENTIAL —— EQUATIONS ~~ ‘Third Edition i Shepley L. Ross Unicity of New Hampshire WILEY- INDIA | | | | | DIFFERENTIAL EQUATIONS Third Eattion Copyright® 2004 by John Wiley & Sons, Inc. All rights reserved Authorized reprint by Wiley india (P.) Lid, 4435V7, Ansari Road, Daryagan), New Delhi 110 002, All rights reserved, AUTHORIZED REPRINT OF THE EDITION PUBLISHED BY JOHN WILEY & SONS INC., UK. No part ofthis book may be reproduced in any ‘orm without the writen permission ofthe publisher. Limits of Liabitty/Disclaimer of Warranty: The publisher and the author make no representation or warranties with respect to the accuracy or completeness of the ‘contenis of this work and specifically disclaim all warranties, including without livtation warranties of fitness for a particular purpose, No warranty may be ctoated or extended by sales or promotional materials. The advice and strategies contained herein may not be suitable for avery situation. This work is sold win the ‘understening that the publisher is not engaged in rendering legal, ‘accounting, or other professional services. if professional assistance is required, the services of a competent professional person should be sought, Neither the publisher nor the author shall be {iabie for damages arising hereirom, The fact ‘that an organization or Website If referred to in this work a8 a citstion andlor a potential source of further Information does nof man that the author or the ‘publisher endorses the information the organization or Website ray provide or recommendations it may make. Further, readers should be aware that internet ‘Websites listed in this work may have changed or disappeared between when ‘this work was written and when itis read, Wiley also puitishes its books in a variety of electronic formats. Some content that eppears in print may not be avaleble in electronic books. For more Information about Wiley products, visit our website at wnaw.iley.com. Reprint : 2010 Printed at : Rajiv 800k Binding House, Delhi ISBN: 978-81.265-1537-0 == PREFACE———_— ‘This third edition, ike the first two, isan introduetion to the basic methods, theory, and applications of differential equations. A knowledge of elementary calculus is presupposed, ‘Thedetailed style of presentation that characterized the previous editions of the text hhas been retained. Many sections have been taken verbatim from the second edition, while others have been rewritten of cearranged with the sole intention of making them clearer and smoother. As in the earlier editions, the text contains many thoroughly worked out examples. Also, a number of new exercises have been added, and assorted ‘exercise sets rearranged to make them more useful in teaching and learning ‘The book is divided into two main parts. The first part (Chapters 1 through 9) deals: with the material usually found in a one-semester introductory course in ordinary differential equations. This par is also available separately as Introduction o Ordinary Differential Equations, Third Edition ohn Wiley & Sons, New York, 1980). The second part of the present text (Chapters 10 through 14) introduces the reader to certain specialized and more advanced methods and provides an introduction 10 ‘fundamental theory. The table of contents indicates just what topics are treated. The following additions and modifications are specifically noted. 1. Material emphasizing the second-order linear equation has been inserted at appropriate places ia Section 4.1. 2, Newillustrative examples, inciuding an especially detailed introductory one, have been written to clarify the Method of Undetermined Coefficients in Section 43, and a useful table has also been supplied, 3. Matrix multiplication and inversion have been added to the introductory material on linear algebra in Section 7.5. 4. Additional applications now appear in the text in Sections 33 and 72. 5. Section 7.6isa completely new section on the application of matrix algebra fo the solution of linear systems with constant coefficients in the special case of wo ‘equations in two unknown functions. The theory that occupied this section in the iv vestace. second edition now appears in Chapter 11 (see note 9 following). We believe that this change represents a major improvement for both intcoductory and inter- mediate courses. 6. Section 7,7extends the matrix method of Section 7.6to the case of linear systems. with constant coefficients involving m equations in n unknown functions. Several etailed examples illustrate the method for the ease m = 3, 7. Both revised and new material on the Laplace Transform of step functions, translated functions, and periodic functions now appears in Section 9.) 8 The basic existence theory for systems and higher-order equations, formerly located at the beginning of Chapter L1, has now been placed at the end of ‘Chapter 10. This minor change has resulied in better overall organization. 9 Chapter 11, the Theory of Linear Differential Equations, has been changed considerably. Sections 11.1 through 114 present the fundamental theory of linear systems. Much of this material was found in Section 7.6in the second edition, and some additional results are also included here. Sections {1.5 through 11.7 now ‘present the basic theory of the single nth-order equation, making considerable use ‘of the material of the preceding sections. Section 11 8 introduces second-order self-adjoint equations and proceeds through the fundamentals of classical Sturm ‘Theory. We believe that the finear theory is now presented more coherently than in the previous edition 10, An appendix presemts, without proof, he fundamentals of second and third order determinants, ‘The book can be used as a text in several different types of courses. The more or less traditional one-semester introductory course could be based on Chapter ? through. Section 7.4 of Chapter 7if elementary applications are to be included. An alternative fone-semester version omitting applications but including numerical methods and ‘Laplace transforms could be based on Chapters 1,2, 4,6, 7,8, and 9. An introductory ‘course designed to lead quickly to the methods of partial diflerentisl equations could be based on Chapters 1, 2(in part), 4,6, 12, and 14. “The book can also be used as a text in various intermediate courses for juniors and seniors who have already had a one-semesier introduction to the subject. An interme- diate course emphasizing further methods could be based on Chapters 8, 9, 12, 13, and 14. An intermediate course designed as an introduction to fundamental theory could be based on Chapters 10 through 14. We also note that Chapters 13 and 14 can be interchanged advantageously. T am grateful to several anonymous reviewers who made useful comments and suggestions. I thank my cotleagues Wiliam Bonnice and Robert O. Kimball for helpful advies. [also thank my son, Shepley L. Ross, Il, graduate student in mathematics, University of Rochester, Rochester, New York, for his careful reviewing and helpful suggestions, ‘Tam grateful to Solange Abbott for her excellent typing, {am pleased to record my appreciation to Editor Gary Ostedt and the Wiley staff for their constant helpfulness and cooperation. ‘As on several previous occasions, the most thanks goes to my wife who offered encouragement, understanding, patience, and help in many different ways. Thanks, Gin, Shepley L. Ross PART ONE FUNDAMENTAL METHODS AND APPLICATIONS ‘One Differential Equations and Their Solutions LL Clasication of Diferential Equations; The Origin and Application LZ Solutions 1.3 Initia-Value Problems, Boundary-Valve Problems, and Existence of Solutions ‘Two First-Order Equations for Which Exact Solutions Are Obtainable 21 Exact Differential Equations and Integrating Factors 22. Separable Equations and Equations Kedveibie Ths Form ~ 23 Linear Equation and Bemoull Equations 24 Special Imegrating Fators and Transformations Three Applications of First-Order Equations 4.1. Orthogonal and Oblique Trajectories 32 Problemsin Mechanics 433. Rate Problems licit Methods of Solving Higher-Order Linear ial Equations AL Basic Theory of Linear Difeential Equacions 42. The Homogeneous Linear Equation with Constant Coeficiente —— CONTENTS —— 102 102 15 43. "The Method of Undetermined Coefficients 44 Variation of Parameter: 43. The Cauchy-Euler Equation 46 Statements and Proofs of Theorems on the Second-Order Homogeneous Linear Equation ° Five Applications of Second-Order Linear Differential Equations, with Constant Coefficients 5.1 The Differential Equation ofthe Vibrations of @ Mau on a Spring 52. Free, Undamped Motion 53. Free, Damped Motion 54 Forced Motion 53. Resonance Phenomena 5.6 Blectric Cirewie Problems Six Series Solutions of Linear Differential Equations, 6.1 Power Series Solutions About an Ordinary Point 6.2 Solutions About Singular Points; The Method af Feobenivt 63 Besse’ Equation and Bessel Functions Seven Systems of Linear Differential Equations 2.1 Dilfeentiat Operators and an Operator Method 72 Applications 7.3. Basic Theory of Linear Sysems in Normal Form: Two Equations in Two Unknown Fonction: 74 Homogeneous Linear Systems with Constant Cocfisients: Two Equations in Two Unknown Fanctiont 23. Matrices and Vectors 16 The Matri Method for Homogeneous Linear Sytems with Constant Coeficiente: ‘Two Equations in Two Unknown Functions The Matri Method for Homogencous Linear Systems with Constant Coeficints: ‘Equations in e Unknown Functions Eight Approximate Methods of Solving First-Order Equations 8.1. Graphical Methods 82 Power Series Methods 83. The Medhod of Succesive Approximations 84 Numerical Metheds ‘Nine The Laplace Transform 9.1 Definition, Existence, and Basie Properties ofthe Laplace Transform 92. The Invere Transform ant the Convolution 93 Laplace ‘Transform Selition of Linear Differential Equations with Constant Cocticient 94 Laplace Transform Solution of Linear Systems 179 19 132 199 au. a 2 23 232 264 264 28 290 sot a2 346 355 377 37 se 390 394 an su a1 a 453 | cones PART TWO FUNDAMENTAL THEORY AND FURTHER METHODS Ten Existence and Uniqueness Theory 461 | 101 Some Conceps fiom Real Funetion Theory 461 } 102 ‘The Fundamental Bastence and Uniqucness Theorem 473 } 105 Dependence of Solutions on Inti Conditions and an the Funeti 8 104 Extence and Uniguenes Theorems fo Systems and Higher-Order Equations 498 Eleven The Theory of Linear Differential Equations 305 HA Touroduetion 50s 12. Basie Theory ofthe Homogencous Linear Sytem 510 13. Further Theory of the Homogencous Linear System m2 14 The Nonkomogenecus Lineae System 53, 1.3. aie Theory ofthe wh-Order Homogeneous Linea Differasial Equation 48 116 Further Poperis ofthe wh-Order Homogencous Lica Diferetial Equation 558 117. "The mh-Order Nenhomogencous Linear Equation 569 1B. Sturm Theory 373 Twelve Sturm-Liowville Boundary-Value Problems and Fourier Series 588 i 1241, Sturm-Liovle Problems ses j 122. Orthogonal of Characters Factions on ‘ 123 The Bepason oa Fenton na Ser of Ontonormal Function oor { 124 “Tgpeemerc Fourier Sees 8 { Thirteen Nonlinear Dierntial Equations on \ 182. Phase Plane Pats and ii! Poi see | 182 Gaal Pons and Phe Liner Systeme bea 153. Grea Pom and Fao Nonna Sens ost 134 Limit Gees anl Peo Satan esa 133 The Meth o Key a Boga bt Pa Fourteen Partial Differential Equations 15 V4 Some Base Concepts and Examples ns 162. Treated of Sparen of Vrs m 143. Canonical Foo Second-Order Linear quai with Constant Coefiens 148 THA hn Vale bien; Chgracecs : Bt Appendices m Answers a ‘Suggested Reading 807 Index 803 ———PART ONE——— FUNDAMENTAL METHODS AND APPLICATIONS —— CHAPTER ONE—— Differential Equations and Their Solutions ‘The subject of diflerential equations constitutes a large and very important branch of ‘modern mathematics, From the early days of the calculus the subject has been an area of great theoretical esearch and practical applications, and it continues to beso in our day. This much stated, several questions naturally arise. Just what is a diferent equation and what does it signify? Where and how do differential equations originate and of what use are they? Confronted with a differential equation, what does one do il, how does one do it, and what are the results of such activity? These questions indicate three major aspects of the subject: theory, method, and application. The Purpose ofthis chapter is to introduce the reader tothe basic aspects ofthe subject and. atthe same time give a brief survey of the three aspects just mentioned. In the course of, ‘the chapter, we shall find answers to the general questions raised above, answers that ‘will become more and more meaningful as we proceed with the study of diferential ‘equations in the following chapters. 1.1 CLASSIFICATION OF DIFFERENTIAL EQUATIONS; THEIR ORIGIN AND ‘APPLICATION A. Differential Equations and Their Classi DEFINITION Anequation inzolving derivatives of one or more dependent variables with respect to one or ‘mare independent variables is called a differential equation.* in connection with is bac defiaon, we do ror nclade inthe class of dori equations those tsqustons that ate actly tive tdontes For example, wa excude sich expresons a, Lier = set, Law = ober ad 50 forts * os roar 3 4 DIFFERENTIAL EQUATIONS AND THEE SOLUTIONS D Example 1.1 For examples of differential equations we lst the following #4 (2) S+n(Z) on dts dix 1 4 58 1 se msint a a3) Pu Ou Ou ee tg a From the bref lis of diflerential equations in Example 1.1 itis clear that the various variables and derivatives involved in a diflerential equation can occur in a variety of ways, Clearly some kind of classification must be made. To begin with, we classify example 1.2 Equations (1.1) and (1.2) aré ordinary differential equations, In Equation (I.1) the variable xis the single independent variable, and yisa dependent variable. In Equation (1.2) the independent variable is ¢, whereas x is dependent DEFINITION A differential equation inoolving partial derivatives of one or more dependent variables ‘with respect to more than one independent variable is called a partial differential equation. > Example 1.3 Equations (1.3) and (1.4) are partial differential equations. In Equation (1.3) the variables sand tare independent variables and visa dependent variable. In Equation (1.4) there are three independent variables: x, y, and 2; in this equation wis dependent. We further classify differs ial equations, both ordinary and partial, according to the ‘order of the highest derivative appearing in the equation. For this purpose we give the following definition 1.1 CLASSIFICATION OF DHFERENTIAL EQUATIONS; THEI ORIGIN AND ARUCATION 5 DEFINITION ‘The order of the highest ordered derivative involved ina differential equation iscalled che order of the diferental equation D Example 1.4 ‘The ordinary differential equation (1.1) is of the second order, since the highest derivative involved is a second derivative. Equation (1.2) is an ordinary differential equation of the fourth order. The partial diferentia equations 1.3) and{(1.4)areof the first and second orders, respectively. Proceeding with our study of ordinary differential equations, we now introduce the important concept of linearity applied to such equations. This concept will enable us 0 classify these equations still further. DEFINITION A Yinear ordinary differential equation of order n, in the dependent variable y and the independent variable x, is an equation that is in, or can be expressed in, the form anty de ay 7 y aa) + UD act +7 OyasCaD Gy + aly = Bla, where ag isnot identically zero, Observe (1) that the dependent variable yand its various derivatives occur to the first degree only, (2) that no products of y and/or any of its derivatives are present, and (3) that no transcendental functions of y and/or its derivatives occur. D> Example 1.5 ‘The following ordinary differential equations are both linear. In each case y is the dependent variable. Observe that y and its various derivatives occur tothe first degree only and that no products of y and/or any of its des fy 6564 bye BrtsZsoy-o, as ay Py odd oe Gate grt Zax a) DEFINITION ‘nonlinear ordinary dierent equation san rdlnar diferente equation hats not 6 DirremenTiAt EQUATIONS AND THEIR SOLUTIONS. > Example 1.6 ‘The following ordinary diferential equations are all nonlinear: ay a an ey Pres as) a a | Equation (17) i nonlinear because the dependent variable y appears to the second degree in the term 6y", Equation (1.8) owes its nonlinearity to the presence of the term Sidy/dx)?, which involves the third power of the First derivative. Finally, Equation (1.9) is nonlinear because of the term Sy(dy/dx), which involves the product of the dependent variable and its first derivative Linear ordinary differential equations are further classified according to the nature ‘of the coefficients of the dependent variables and their derivatives. For example, Equation ({.5)issaid to be linear with constant coefficients, while Equation (1.6)is linear with variable coeficients B. Origin and Application of Differential Equations Having classified diferential equations in various ways, let us now consider briefly where, and how, such equations actually originate. In this way we shall obtain some indication of the great variety of subjects to which the theory and methods of differential equations may be applied Differential equations occur in connection with numerous problems that are ‘encountered inthe various branches of science and engineering. We indicate a few such problems in the following list, which could easily be extended to fill many pages. rocket, sat oF planet. 1, “The problem of determining the motion of a projecti 2. The problem of determining the charge or current in an electric circuit. 3. The problem of the conduction of heat in & rod or in a slab. 44. The problem of determining the vibrations of a wire or a membrane. 5. The study of the rate of decomposition of a radioactive substance or the rate of growth of a population. 6 The study of the reactions of chemicals. 7. The problem of the determination of curves that have certain geometrical properties. ‘The mathematical formulation of such problems give rise to differential equations. But ust how does this occur? In the situations under consideration ineach of the above problems the objects involved obey certain scientific laws. These laws involve various rates of change of one or more quantities with respect to other quantities. Let use 12 sownons 7 call that such rates of change are expressed mathematically by derivatives. In the ‘mathematical formulation of each of the above situations, the various rates of change are thus expressed by various derivatives and the scientific laws themselves become ‘mathematical equations involving derivatives, tha is, differential equations. In this process of mathematical formulation, certain simplifying assumptions generally have to be made in order that the resulting differential equations be tractable. For example, ifthe actual situation ina certain aspect of the problem is ofa relatively complicated nature, we are often forced to modify this by assuming instead an approximate situation that is of a comparatively simple nature. Indeed, certain relatively unimportant aspects of the problem must often be entirely eliminated. The result of such changes from the actual nature of things means that the resulting diflerential equation is actually that of an idealized situation. Nonetheless, the information obtained from such an equation i of the gfeatest value to the scientist A nnatural question now isthe following: How does one obtain useful information froma differential equation? The answer is essentially that itis possible todo s0, one solves the differential equation to obtain a solution; if this is not possible, one uses the theory of differential equations to obtain information about the solution. To understand the meaning ofthis answer, we must discuss what is meant by a solution of 4a diflerential equation; this is done in the next section. Exercises differential Cassify each of the following differential equations as ordinary or parti tqustiossatethe orderofeachequatoncand deerme wheter heaton oder Smideration ea oc nonin 1 La ety ere. 2 Sg syesinx Boo 4 dye yarn 2 (Eee 0 (t= fn 1.2 SOLUTIONS A. Nature of Solutions We now consider the concept of a solution of the nth-order ordinary differential ‘equation, 8 owreteniat eQuaTiONs AND THUR SOLUTIONS DEFINITION fondo ft =n a dy ty where Fis areal function ofits (n+ 2) arguments x, 1% oo 1. Let f be a reat function defined forall x ina real inereal I and hasing an nth derivative (and hence aso all lower ordered derivatives) forall x€ 1. The function fis called an explicit solution ofthe differential equation(J.10)on 1if i fulfill the following ‘v0 requirements: FL SOM Loe LCT “ is defined for all x € 1, and FL £9, S'.-- SE] =O 8 Jor all x 1. That is, the substitution of f(x) and its various derivations for y and ts corresponding derivatives respectcely, in (I-10) reduces (10) to an identity on 2. relation olx,») = Os caled an implicit solution of 10) this relation defies cat least one eal function f ofthe variable x on an interoal sack hat this function i an explicit solution of (1.10) on this interval 3. Botkexplcit solutions and implicit solutions will usually be called simply solutions Roughly speaking, then, we may say that a solution ofthe differential equation (1-10) isa relation explicit o implicit—between x and y, not containing derivatives, which identically satisfies (1.10) D Example 1.7 “The function f defined for all real x by fle) = 2sin x + 3008 x aan is an explicit solution ofthe differential equation ay Shtys0 ary fora ral x. First note that is defined and has.a second derivative forall real x. Next ‘observe that J") = 208 x —3 sin x, Sta) = ~2sin x— 3.08 x. ‘Upon substituting f(x) ford? y/dx? and f(x) for yin the diferental equation (1-12), it reduces to the identity (~2sin x ~ 3.08 x) + (2 sin x-+ 3.08 x) = 0, 12 sorutions 9 ‘which holds for all real x. Thus the function f defined by (1.11) isan explicit solution of the differential equation (1.12) forall real x. > Example 1.8 ‘The relation x+y? 2550 13) is an implicit solution of the differential equation ay x4 yZn0 a9 ‘on the interval I defined by —5 Example 1.17 Problem. Find a solution f of the diferemtial equation ay Ram (2y such that at x = I this solution f has the value 4, Explanation, First let us be certain that we thoroughly understand this problem. We seck a real function f which fulills the two following requirements: 1. Thefonetion f must satisfy the differential equation (1.21) Thats, the function f must be such that J") = 2x fr ll real x ina ral interval 2. Thefunction f must havethe value 4atx = 1. Thatis the function f must besuch that £0) = 4 Notation. The stated problem may be expressed in the following somewhat abbreviated notation: ay Bory, ax yO) = In this notation we may regard y as representing the desired solution. Thea the differential equation itself obviously represents requirement 1, and the statement (l= 4 stands for requirement 2. More specifically, the notation y(1) = 4 states that the desired solution y must have the value 4 at x = 1; that is, y= 4 at x Solution. We observed in Example 1.9 that the difleential equation (1.21) has a ‘one-parameter family of solutions which we write as yextte, (22) where cisan arbitrary constant, and that each ofthese solutions satisfies requirement L Let us now attempt to determine the constant ¢ so that (1,22) satisfies requirement 2, thats, y = 4.at x= 1, Substituting x = 1, y = 4 into (1.22), we obtain 4 = | + ¢,and hhence ¢ = 3, Now substituting the value c= 3 thus determined back into (1.22), we obtain yan, 16 { } DIFFERENTIAL EQUATIONS AND THER SOLUTIONS which is indeed a solution ofthe differential equation (1.21), which has the value 4 at In other Words, the function f defined by Soy +3, satisfies both of the requirements set forth in the problem. Comment on Requirement 2 and its Notation. Ina problem of this type, require- ment 2 is regarded as a supplementary condition that the solution of the diferential equation must also satisy. The abbreviated notation y(1) = 4, which we used to express this condition, is in some way undesirable, but ithas the advantages of being both customary and convenient. In the application of both first-and higher-order differential equations the problems most frequently encountered are similar to the above introductory problemin that they involve hoth a differential equation and one or more supplementary conditions which the solution of the given differential equation must satisfy. 1C all of the associated ‘supplementary conditions relate to one x value, the problem is called an intial-zalue ‘rable (or one-point boundary-valve problem). If the conditions relate to ro dif- ferent x values, the problem is called a two-point boundary-value problem (or simply a boundary-valve problem). We shall illustrate these concepts with examples and then ‘consider one such type of problem in detail. Concerning notation, we generally employ abbreviated notations for the supplementary conditions that are similar to the ab- breviated notation introduced in Example 1.11 > Example 1.12 0) p= -4 This problem consists in finding a solution of the differential equation Which assumes the value 3at x = | and whose first derivative assumes the value —4 at Both of these conditions relate to one x value, namely, x = I. Thus this is an initial-value problem, We shall see later that this problem has @ unique solution. Example 1.13 13 IMTIALVALUE Prom, soUNDARY.vALUE PROBLEMS 17 In this problem we again seck a solution of the same differential equation, but this time the solution must assume the value 1 at x =O and the value $ at x = n/2. That is, the conditions relate to the two different x values, 0 and 1/2. This is a (two-point) boundary-value problem. This problem also has a unique solution; but the boundary value problem i: Paya, YO=I yln)= 5, has no solution at all! This simple fact may lead one to the correct conclusion that boundary-value problems are not to be taken lightly! We now turn toa mote detailed consideration of the initial-value problem fora frst- order dilferential equation. DEFINITION Consider the first-order differential equation és =f (123) where {is «continuous function of x and y in some domain® D ofthe xy plane: and let (50, 3) be a point of D. The initial-value problem associated with (7.23) is to find ‘solution ofthe differential equation (1.23) defined on some real interval containing Xp, ‘and satisfying the inital condition $600) = Yo! In the customary abbreviated notation, this inital-ualue problem may be written ‘ eo Seah Wx) = Yo. ‘osolvethis problem, wemust inda fonction that not onl satsfos the diferentiat equation (1.23) but that also satsfes the initial condition that it haste value y when x has value x. The gsometrc interpretation of th initial condition, and hence ofthe centr initalvalye problem, i casly understood. The graph of the dsied solution must pas through the pont with coordinates (x, ye). That i, interpreted geometi- cally, the initial-value problem st find an integrl curve ofthe diferntial equation (1.28) that passes theough the point (x) “Themethod of actually finding the desired solution depends upon thenatue of the differential equation ofthe problem, that i, upon the form off}, Cetain special types of iflerential equations have a one-parameter family of solutions whose equation may be found exactly by following definite procedures (ee Chapter 2). the Sifferetial equation ofthe problemi of some such special type, one st obiain the equation of ts one-parameter family of solutions and then applies the initial condition "Nema an open, Connected at For ove unarliar wih ach concept, O may be eed the interior of Some simple closed cure inthe plore. 18 wrenentia. tQuaTions AND THEIR SOLUTIONS {othisequationin an attempt to obtaina“particular” solution @ that satisfies the entire initial-value problem. We shall explain this situation more precisely in the next paragraph. Before doing so, however, we point out that in general one cannot find ‘he equation of a one-parameter family of solutions of the differential equation; approximate methods must then be used (see Chapter 8) Now suppose one can determine the equation als, no) = 0 2) cof a one-parameter family of solutions of the differential equation of the problem. ‘Then, since the initial condition requires that y= yp at x= xo, we let x = Xp and 13 Vo in(1.24) and thereby obtain Bo.¥od = Solving this for cin general we obtain a pasticular value of ¢ which we denote here by x, Wenow replace the arbitrary constant c by the particular constant cy in (1.24, thus ‘obtaining the particular solution 4s, Yea) = 0. ‘The particular explicit solution satistying the two conditions (differential equation and {niial condition) of the problem is then determined from this, if possible. We have already solved one initial-value problem in Example 1.11. We now give another example in order to illustrate the concepts and provedures more thoroughly > Example 1.14 Solve the initial-value problem (125) (126) given that the differential equation(1.25)has a one-parameter family of solutions which -may be written in the form, (127) ‘The condition (1.26) means that we seek the solution of (.25)such that y = 4at x = 3 ‘Thus the pair of values (3,4) must satisy the relation (1.27) Substituting x = 3 and 4 into (1.27), we find vay 9+ 16=c? or c ‘Now substituting this value of c? into (1.27), we have xt yha 2s Solving this for y, we obtain yas VB—H ‘Obviously the positive sign must be chosen to give y the value +4 at x = 3. Thus the function f defined by fl) = JB ~Sex Example 1.15 Con inte den A alte Bovey, wl Let us apply Theorem 11. We first check the hypothesis. Here f(x, y) = x? + y* sn Ty oh of ton and fy ae ots yon of the xy plane. The initial condition y(1) = 3 means that xo = Land yp = 3,and the point (1, 3) certainly lies in some such domain D. Thus all hypotheses are satisfied and the conclusion holds. That is, there is a unique solution @ of the diferential equation dyjdx = x* + y?, defined on some interval |x — 1] Example 1.16 Consider the tio problems: ayy L Rage nr yoy 2 Bete rome Here fx Sewn zin ane ‘These functions are both continuous except for x = 0 (that is, along the y axis). In problem 1.x = 1, yo = 2. The square of side I centered about (1,2) does not contain the y axis, and so both f and df/éy satisty the required hypotheses inthis square. Is interior may thus be taken to be the domain D of Theorem 11; and (1,2) certainly les within it, Thus the conclusion of Theorem 1.1 applies to problem 1 and we know the problem has @ unique solution defined in some sufficiently small interval about xp = J Now let us turn to problem 2. Here xy = 0, yo = 2.At this point neither f nor dff2y are continuous. In other words, the point (0,2) cannot be included in a domain D where ‘the required hypotheses are satisfied. Thus we can no: conclude from Theorem 1.1 that 22 overex, equations ANO THEHE SOLUTIONS problem 2 has a solution, We are not saying that it does not have one. Theorem 1.1 simply gives no information one way or the other. Exercises 1. Show that ya det +20 yo)= 6, YO) =2. Is y = 2% + 4e-* also a solution of this problem? Explain why or why not 2. Given that every solution of uae Rt ya2 may be written in the form y = (x? 4 de~*, for same choice of the arbi constant c solve the following initial-value problems: ay oF ® yee @ Pa yeres, ty Pyare, 0) = 2 wale +3. 4, Given that every solution af ey ay _ oy 8 yao ‘may be written in the form pecs tee, for some choice of the arbitrary constants c, and 3, solve the following initial- value problems: ®y @ Py _ay & -Z- nyo 10) yOr6 4. Every solution ofthe diferential equation @y Gitrno ( Reece Be esc ee OSE aEnREEE 13 INITAL-VALUE PROBLEMS, ROUNDARY-VALUE Poss 23 may be written in the form y = ¢, sinx + ¢; os, for some choice of the arbitrary ‘constants c, and c. Using this information, show that boundary problems (a) and. (©) possess solutions bur that (c) does-n0:. a © Baye yO) = 0, yO) = 1, Wen/2) 41 YR) = “ © Bayeo 30) =0, 0) 1 “ Given that every solution of Py yi dy gt Eh - 3 Sh 4 6% oye ‘may be writtenintheform y = ¢,x + ¢,x7 + ex? forsome choice of the arbitrary constan's ¢,,¢3, and ¢s, solve the initial-value problem consisting of the above dlifferential equation plus the three conditions yQ)=0, y@=2 y"'A=6 . Apply Theorem 1.1 to show that each of the following initial-value problems has 2 unique solution defined on some suficently small inverval |x ~ 1| 0. (8) Carefully graph the solution for which ¢ = 0. Then, using this particular staph, also graph the solutions for which ¢ = 1,¢ = 2, and ¢ = 3. === CHAPTER TWO—— First-Order Equations for Which Exact Solutions Are Obtainable In this chapter we consider certain basic types of first-order equations for which exact solutions may be obtained by definite procedures. The purpose ofthis chapteristo gain ability to recognize these various types and to apply the corresponding methods of solutions. OF the types considered here, the so-called exact equations considered in Section 2.1 arcin asense the most basic, while the separable equations of Section 2.2are {na sense the “easiest.” The most important, from the point of view of applications, are the separable equations of Section 2.2 and the linear equations of Section 23. The remaining types are of various very special forms, and the correponding methods of solution involve various devices. In short, we might describe this chapter as a col- lection of special “methods,” “devices,” “tricks,” oF “recipes. in descending order of kindness! EXACT DIFFERENTIAL EQUATIONS AND INTEGRATING FACTORS, A, Standard Forms of Fitst-Order Differential Equations The first-order differential equations to be studied in this chapter may be expressed in cither the derivative form Pesan ey or he iteremia form Mos hax + NO Nd =O. an 25 26 masronoer EQUATIONS FOR WHICH EXACT SOLUTIONS ARE OBTAINABLE ‘An equation in one of these forms may readily be written in the other form. For example, the equation ay ety & is ofthe form (2.1). It may be written (e+ y)ax+ aay which i of the form (22). The equation (sin + yds + (4 39) y =0, which i of the frm (22), may be written in the form (21) a8, In the form (2.1) itis clear from the notation itself that is regarded as the dependent variable and x as the independent one; but in the form (2.2) we may actually regard cither variable as the dependent one and the other as the independent. However, in this text, in all differential equations of the form (2.2) in x and y, we shall regard y as dependent and x as independent, unless the contrary is specifically tated B. Exact Differential Equations DEFINITION Let F be a function of two real variables such that F has cominuous first partial derivatives in a domain D. The total dillerental dF of the function F is defined by the formula 9 = ED an 4 FLED cs {for ait(x, )¢ D. > ixampe 21 ~ Let F be the function of two teal variables defined by Fis, y)= ay? + Bey for al real (x, 3). Then BFC sg gery, SFE : is 675, SRD e aey 4 20% ‘and the total differential dF is defined by F(x, 9) = (9? + 6x7y) do + Oxy + 2x) dy for all real (x) | | I | } { i | | i | 2.4 WACK DIFFERENTIAL EQUATIONS AND INTEGRATING FACTORS 27 DEFINITION The expression M(x, y) dx + NOs y) dy @3) is called an exact differential ina domain D if there exists a function F of two real vari- ‘ables such that this expression equals the total differential dF(x,y) for all (x,y) €D. That is, expression (2.3) is an exact differential in D if there exists @ function F such that 2PM es.) and FED 9 aces fue yal a cg cated ope aa age ce M(x, y)dx + N(x, ypdy = 0 seated ana trent eution > Example 2.2 The differential equation yi dx + 2xpdy=0 es is an exact diferential equation, since the expression y? dx + 2xy dy is an exact Example 23 We apply the exactness criterion (2.7) to Equations (2.4) and (2.5), introduced in Example 22. For the equation y? dx + 2xydy=0 (24) we have Mey) N(s,9) = 29, OMG 9) _ 9, _ ANU 9) ey a for all(x,»). Thus Equation (24) is exact in every rectangular domain D. On the other hand, for the equa pdx + 2xdy=0, as we have Mex y= NOs 9) = 2x aMix 9) antx,») a jon (2.8) isnot exact in any rectangular domain D, 142 for all(x, 9). Thus Equi 12.1. BRACT DIFERENTIAL EQUATIONS ANO WwTEGaATING FACTORS 3 > example 24 Consider the differential equation Qxsin y+ ye") de + (x? cosy + 3y%e dy = 0. Here M(x y)= 2x sin y + ye N(x y)= x? 60s y+ 39%e, ames 9) ents.) ay a inevey rectangular domain D Ths his diferent equation is exc in vty such = 2x cos y + 3yte* = ‘These examples illustrate the use ofthe test given by (2-7)for determining whether or not anequation of the form M(x, y) dx + N(x, y) dy = Oisexact.Itshould be observed ‘that the equation must be in the standard form M(x, y) dx + N(x, y) dy = Oin order to Use the exactness test (27). Note this carefully: an equation may be encountered inthe nonstandard form M(x, ») dx = N(x, »)dy, and in this form the test (2.7) does not apply. C. The Solution of Exact Differential Equations ‘Now that we havea test with which to determine exactness let us proceed to solve exact differential equations. If the equation M(x, ypde + N(x, »)dy =O is exact in a reétangular domain D, then there exists a fonction F such that FFL») _ sayy oF») i SR Messy and 7 « Nix 9) forall ye. ‘Then the equation may be written BOD gy POD yay oxsingy dF 91= ae ae + y dy=0 orsimply dF (x, y) = 0. ‘The celation F(x, #} = cis obviously a solution of this, wherecis an arbitrary constant ‘We summarize this observation in the following theorem. ‘THEOREM 2.2 ‘Suppose the differential equation M(x, y) dx + NIx, y) dy = 0 satisfies the differentia. Dilley requirements of Theorem 2.1 and s exact wn a rectangular domain D. Then a one- parameter family of solutions of this differential equation is gioen by F(x, ») = ¢, where F isa function such that aF i, Ft FEDS Mex, 9 and : 2D Nix,y) for ail (x ye. ‘and ets an arbitrary constant. Fns1-0806R EQUATIONS FOR WHICH EXACT SOLUTIONS ARE OBTAINABLE Referring to Theorem 2.1, we observe that F(x, y) is given by formula (2.13) However, in solving exact differential equations its neither necessary nar desirable to Use this formule. Instead one obtains F(x, y) either by proceeding asin the proof of Theorem 2.1, Part 2,orby the so-called “method of grouping,” which willbe explained inthe following examples > Example 25 Solve the equation Bx? + dy) dx + Ox? + Iy)dy = 0. ‘Our first éuty is to determine whether or not the equation is exact, Here M(x, y) = 3x? + 4p, NOx 9) = 28? +29, Mix, 9) ant 9) a x for all real (x, yj and so the equation is exact in every rectangular domain D. Thus We must find F such that 66,9) oF») ey = Misys) = 38! + 44y- and = Nu yb = 208 429 From the fist of these, Fo = futon ax + 00)= foe 4+ 4ny) dx + 00) Bey FOOL, 259) _ aya , HOU? eur dy ‘Buc we must have Thus ‘Thus #(y) = ¥? + co, where cp is an arbitrary constant, and so : Fix, ax + Bet ty? +00 ws AND intecRATING Factors 33. 24 Exact oUrimeNALEQUAN Hence a one-parameter family of solution is F(x, y) a4 ety + yb eg me} Combining the constants ¢y and cy we may write this solution as ety ts Pac, wheree = in generality by taking cy = 0 and writing $y) procedure, ois anarbitrary constant. Thestudent will observe that there sno oss y3, We now consider an alternative Method of Grouping. We shall now solve the differential equation of this exam ple by grouping the terms in such a way that its let member appears as the sum of certain exact differentials. We write the diferential equation Gxt + dxyl de + Ox? 4 2y)dy = 0 in the form 3x? dx + (dy dx + 2x? dy) + 2y dv = 0. We now recognize this as dix?) + d(2x?y) + diy?) where ¢ is an arbitrary constant, or dls? + Ixty + y?) = dle. a, From this we have at once xtadty tyme, Clearly this procedure is much quicker, but it requires a good “working knowledge” of differentials and a certain amount of ingenuity to determine just how the terms should be grouped. The standard method may require more "work and take longer, but its perfectly straightforward. [tis recommended for those who like to follow a pattern and for those who have a tendency to jump at conculsions. Just to make certain that we have both procedures well n hand, we shall consider an initial-value problem involving an exact differential equation, > Example 2.6 Solve the initial-value problem (De cos y+ Ixty) dx + (x? — x3 sin y= yhdy = 0, 0) = 2 ‘We fist observe thatthe equation is exact in every rectangular domain D, since aM») iny + 3x2 = ON) Spe = anny 4 act = Rs forall rel (x, 9) 34. emst-onote EQUATIONS FOR WHICH EXACT SOLUTIONS ARE OBTAINABLE Standard Method. We must find F such that Fon 9) a) a My) = 2ec08 y+ 3x4 and ree (ea ee Bo Mag) =~ x7sin yp Then Fe,y)= f Mix, ix +6) = fexcosy +349 2x4 649 = eosy4yt 60), 2FO ys >, dal) ssiny tt OU But also FOI) _ Nyy yy ex =x gin FED Ns yaw yoy and so and hence oun B00, Thus Fes, y) = x eos y +29 2 + oo Hence a one-paraineter family of solutions is Fix, 9) = ¢,, which may be expressed as sPeosy + ty Applying the initial condition y = 2 when x ~ 0, we find ¢ = —2 Thus the solution of ‘he given initial-value problem is sPoos yey Method of Grouping, We group the terms as follows: ede = xtsin yds) + Bx4y dy #8 dy vey