aor 2H om. eames
Elementary
Differential
Equations
tammernenee eer
SEVENTH EDITION
Earl D. Rainville
Late Professor of Mathematics
University of Michigan
Phillip E. Bedient
Professor of Mathematics -
Franklin and Marshall College
Macmillan Publishing Company
New York
Macmillan
es‘Copyright © 1989, Macmillan Publishing Company, a division of Macmillan, tnc
Printed in the Philippines
All rights reserved. No part of this book may be reproduced or
‘transmitted in any form or by any means, electronic or mechanical,
including photocopying. recording, oF any information storage and
retrieval system, without permission in writing from the publisher
Earlier editions copyright © 1981, 1974, 1969, 1964, 1998, 1952, and 1949 by
‘Macmillan Publishing Company
‘Macmillan Publishing Company
{866 Third Avenue, New York, New York 10022
Collier Macmillan Canada, Inc.
Library of Congress Cataloging in Publication Data
Rainville, Earl David (date)
Elementary differential equations/Ear! D. Rainville, Phillip E
Bedient—Tth ed.
Pp. cm.
Includes index.
ISBN 971-1055-52-X
1. Differential equations. 1, Bedient, Phillip Edward (date)
IL Title,
QA37LR29 1989 811120
S1S3'S—de 19 cP
Priming: 12345678
Year 90123456
Reprinted under Authority of P.D.#285
fas Amended by PD. No. 400 and
P.D.No. 1203 by
EDCA Publishing and Distributing Corporation
488 Quezon Avenuc, Quezon City
Tel. Nos. 71-97-09; 732- 14-83
EXPORT OF THIS BOOK FROM THE PHILIPPINES
PUNISHABLE BY LAW
Preface tothe
Seventh Edition
The publication of a new edition of Professor Rainville’s book is an
acknowledgment of the effectiveness of the simple and direct style that has
U iade earlier editions most popular with students of differential equations
The balance between developing techniques for solving equations and the
theory necessary to support those techniques remains essentially
unchanged, although a number of sections have been revised and some new
‘applications have been added. Sections 56 through $9 are new and present
“A derivation of Kepler's laws from Newton's laws. Section 135 is new and
Gives a presentation of a modified Euler method for solving differential
uations numerically
© The material is arranged to permit great fexibility in the choice of topics
for a semester course, Except for Chapters 1, 2, 5, 16, 17, and 18, and either
yand 7 or 11 and 12, any chapter on ordinary dif
‘omitted without interfering with the study of later chapters. Parts of chap~
omitted in many instances.
7 For a course that aims at reaching power series as rapidly as is consistent
fh some treatment of more elementary methods, a reasonable syllabus
‘should include Chapters 1 and 2; Chapters $, 6, 7, ana 8; parts of Chapters
3 and 15; Chapters 17 and 18; and whatever applications the instructor
Fes to insert.
Chapters 1 through 16 of this book appear separately as A Short Course
"Differential Equations, Seventh Edition, The shorter version is intended
‘courses that do not include discussion of infinite series methods. ‘
‘The author wishes to thank those students and colleagues at
Franklin,
Marshall College whose $0 sey 5
upport of Professor
George Rosenstein are gratefully acknowledged,
er, Pennsylvania1 Definitions, Families of Curves
1. Examples of differential equations
2. Definitions 3
3. The elimination o” arbitrary constants:
4, Families of curves 10, 2
"Equations of Crder One
§. The isoclines of an equation WW
6. An existence theorem 19
7, Separation of variables 21
8. Homogencous functions’ 26 Gama iam
9, Equations with homogeneous coefficients
10. Exact equations
41. The linear equation of order one 37
12, The general olution of linear equation 41
Miscellaneous exercises 440vl Contents Contents
4 eo Topics on Equations of Order One 63
. Integrating factors found by ins
facto spection: 63
19. The determination of integrating factors 67
20. Substitution suggested by the equation 72
21. Bernoulli's equation © 74
22. Coefficients linear in the two variables 77
23. Solutions involving nonelementary integrals. 82
Miscellaneous exercises 85
8 Variation of Parameters
44, Reduction of order 139
45. Variation of parameters 143
46. Solution of y’+y=J(x) 148
Miscellaneous exercises 151
| 9 Inverse Differential Operators
47. The exponential shift 152
48. The operator 1/f(D) 156
49. Evaluation of [1/f(Dyle" 157
$0. Evaluation of (D? + a?)~! sin ax and
(D? +a2)"!cosax 158
5 Linear Differential Equations 87
24. The general linear equation 87
25. An existence and uniqueness theorem 89
26. Linear in: 90
27. The Wronskian a1
28. General solution of a homogeneous equation 94
29. General solution of a nonhomogeneous equation 96
30. Differential operators 97
31. The fundamental laws of ope
32. Some properties of differenti
10 Applications
51. Vibration ofa spring 162
Undamped vibrations 164
j. Resonance 167
}, Damped vibrations 170
The simple pendulum — 174
. Newton's laws and planetary motion 176
Central force and Kepler's second law 178
Kepler's first law 178
Kepler's third law 181
ion 100
operators 102
6 Linear Equations with Constant
Coefficients 105
33. Introduction 105
34, The auxiliary equation; distinct roots 105
35. The auxiliary equation; repeatea roots 108
36. A definition of exp zforimaginaryz — 112
37. The euxiliary equation; imaginary roots 114
38. A note on hyperbolic functions 116
yus-exercises 119
7 Nonhomogencous Equations: Undetermined
Coefficients 121
39. Construction of a nomogeneous equation
from a specified solution 121* Contents
Contents
12 Inverse Transforms 209
70. Definition of an inverse transform 209
71, Partial fractions 213
72. Initial value problems 216
73. Astep function 222
74, Aconvolution theorem 229
75. Special integral equations 234
76. Transform methods and the vibration of springs 239
77. The deflection of beams 243
} Nonlinear Equations
97. Preliminary remarks 324
98. Factoring the left member 325
99. Singular solutions 328
100, The c-discriminant equation 329
101. The p-discriminant equation 331
102. Eliminating the dependent variable 333
103. Clairaut’s equation 335
404, Dependent variable missing 339
105. Independent variable missing 340
106, Thecatenary 343
13 Linear Systems of Equations 250 Miscellaneous exercises 345
78. Introduction 250
79. Elementary elimination calculus 250
80. First-order systems with constant coefficients 254
81, Solution ofa first-order system 256
82. Some matrix algebra 257
83. First-order systems revisited 264
84. Complex eigenvalues 273
85. Repeated eigenvalues 279
86, Nonhomogeneous systems 287
87. Arms races 291 18 Solutions Near Regular Singular Points 364
88. The Laplace transform = 295 112. Regular singular points 364
| 113. The indicial equation 367
114, Form and validity of the solutions near a regular
Power Series Solutions
107 Linear equations and power series 347
108. Convergence of power series 348
109. Ordinary points and singular points 351
110, Validity of the solutions near an ordinary point 352
111, Solutions near an ordinary point 353
4 Electric Circui k 301 singular point 369
; 99, ae ie: 115, Indicial equation with difference of roots
90. Simple a Ks 304 nonintegral 369.
3 aan 116. Differentiation of a product of functions 375
117, Indicial equation with equal roots 377 ¥
: Indicial equation with equal roots, an alternative 383
15 The Existence and Uniqueness of Solutions 314 119. Indicial equation with difference of roots a positive
Dicandeasy somal integer, nonlogarithmic case 387
a sa see 24 Indieiai equation with difference of roots a positive
‘An existence and uniqueness theorem = 314 ti Ss
a A iets sondlton 7 integer, logarithmic case 391
94. A proof of the existence theorem = 318 Solution for large x 397 3
Many-term seucee relations 401.
95. A proof of the uniqueness theorem 321
96. Other existence theorems — 322xii ‘Contents
19 Equations of Hypergeometric Type
124. Equations to be treated in this chapter 410
12S. The factorial function 410
126. The hypergeometric equation 411
127. Laguerre polynomials. 413
410
128. Bessel’s equation with index not an integer 415
129. Bessel's equation with index an integer 415
130. Hermite polynomials 417
131. Legendre polynomials 417
132. The confluent hypergeometric equation 418
20 Numerical Methods
133. General remarks 421
134, Euler's method 421
135. A modification of Euler’s method 425
136. A method of successive approximation 426
421
137. An improvement on the preceding method 427
138. The use of Taylor's theorem 429
139. The Runge-Kutta method — 431
140. Acontinuing method 435
21 «Partial Differential Equations
141. Remarks on partial differential equations — 438
142. Some partial differential equations of applied
mathematics 438
143, Method of separation of variables 441
144. A problem on the conduction of heat in a slab
22 Oxsomsass tes of Pusctions
Ble wa tas ol pctrnonas 453
Orthogonal polynomials 454
Se gage roe 436
‘orthogonal sets 458
438
446
452
Contents
| Fourier Series
151. Orthogonality of a set of sines and cosines
152. Fourier series: an expansion theorem 463
153, Numerical examples of Fourier series 407
154. Fourier sine series. 476
188. Fourier cosine series 479
156. Numerica! Fourier analysis 483 ©
157. Improvement in rapidity of convergence 484
Boundary Value Problems
158. The one-dimensional heat equation 486
159. Experimental verification of the validity
ofthe heat equation 493
160. Surface temperature varying with time 495
161. Heat conduction inasphere 497
162. The simple wave equation 499
163. Laple2e's equation in two dimensions $02
Additional Properties of the Laplace
Transform
164. Power series and inverse transforms 507
© 365. Theerror function S11
166. Bessel functions 518
“167. Differential equations with variable coefficients
"Partial Differential Equations; Transform
ethods
Boundary value problems
The wave equation 527
‘Canonical vari
Diffusion in a slab of finite width
| Diffusion in a quarter-infinite1
Definitions, Families —
of Curves
Examples of Differential Equations
pin equation (3), a derivative may
‘Of differentials. Our aim is to find methods for solving differential
ons, that is, to find the unknown function or functions that satisfy the
ial equation. =
following are examples of differential equations:
oe
ft ena
(+ Pde2 ‘Definitions, Families of Curves Ich
du (Ou, Ow
aie & * a @
ai at
Laat Rt Gin Bm conan, (5)
@v ev
oe tap ©)
‘aw\?
Ce.-. .,
@x a
tegen ®
ty
oe ®
#y ax
ae ae * (19)
Ge a
Baas (i)
‘When an equation involves one or more derivatives with respect to a
‘particular variable, that variable is called an independent variable. A vari-
able is called dependent if a derivative of that variable occurs.
In the equation
ai dit
eta cg Shr et (5)
4 is the dependent variable, ¢ the independent variable, and L, R, C, E, and
@ are called parameters. Tae equation
ov oy,
Ox? * dy?
has one dependent variable V and two independent variables.
Since the equation
0 (6)
(2? + y?) dx — Ixy dy = 0 @)
“may be written
w+y-ry Zo
2 yy eee
G+ NG, — BO
consider either variable to be dependent, the other being the it
one.
ify the
gations given as examples in this section,
Definitions
F order of a differential equation is the order of the Righest-ordered
jive appearing in the equation. For instance,
noe
B+ mz) +9=0
m equation of “order two." It is also referred to as a “second-order
ation.”
ey iyi egbeton
FU IV OD =O ‘
an “nth-order” ordinary differential equation. Under
Bee (ie tancion Fea
pof the ‘other n + I variables x, y,
yi = Le HY
purposes of this book we shall assume that this i
fan equation of the form of equation (2) may
‘one equation of the form of equation (3)
ple, the equationDefinitions, Families of Curves Ich
_ A function $, defined on an interval a < x
0, (2/,/3)x*? is a solution of the equation in Exercise 6,
3. The Elimination of Arbitrary Constants
In practice, differential equations arise in many ways, some of which we
shall encounter later. There is one way of arriving at a differential equation,
however, that is useful in that it gives us a feeling for the kinds of solutions
to be expected. In this section we start with a relation involving arbitrary
constants and, by elimination of those arbitrary constants, come to a differ
‘ential equation consistent with the o-iginal relation. In a sense we start with
the answer and find the prob!=m
‘Methods for the elimination of 2rburary constants vary ith the ws y in
‘which the constants enter the given relation. A mv'ho chat is eificient 0”
‘one problem may be poor for anothe:. Une fact persists throughout
Because each differentiation yields a new relation, the number of derivatives
that need be used is the same as the number of arbitrary constants to be
liminated. We shall in each case determine the differential equation that is
(a) Of order equal to the number of art
tion.
Fy Constents in the given ris
(b) Consistent with that relation.
(€) Free from arbitrary constants.
EXAMPLE (a)
Eliminate* the arbitrary constants c, and c, from the relation
yace™+c,e%. a
‘Since two constants are to be eliminated “tain the two derivatives,
¥ = —deye"™* + Jeze, 2
¥ = Acye"™* + 9e,e", @
‘The elimination of ¢, from equations (2) and (3) yields
of + Dy = Weze™;
+ By diflerentiations and pertinent leplimate mathematical procedures, Elimination 6)
‘ecasare fr instance, snot permitied. es 4
The EI
ination of Arbitrary Constants -
jon of c, from equation (1) and (2) yields
yt
Seze*
y' + dy = 3 + Ih a
yi —y' —6y =0,
TAnother method for obtaining the differential equation in this example
ceeds as follows: We know from a theorem in elementary algebra that
f three equations (1), (2), and (3), considered as equations in the two
ynowns c, and c, , can have solutions only if
4 eae «|
te * Je @
-21 ge%|
flows immediately
this latter method has the advantage of making it easy to see that the
ination of the constants ¢,.¢2...-+ &,{tom a relation of the form
+een"
ye cet eee
iil always lead to a lincar differential equation
fy thy
ay ;
a Ooo Yay ad,
oat Nae
dx
yhich the coefficients do... a ate constants. The study of such —
tial equations will receive much of our attention. ceeDefinitions, Families of Curves Ich
Direct differentiation of the relation yields
Ax — a) + 2yy' = 0,
from which
anxtyy.
‘Therefore, using the original equation, we find that
Ww?'+ P= + w?,
or
vax + doy,
which may be written in the form
(x? = y?) dx + 2xy dy = 0.
Another method will be used in this example as an illustration of a device
that is often helpful. The method is based upon the isolation of an arbitrary
constant.
‘The equation
(oF +P =a
‘may be put in the form
x? + y? — Jax =0,
or
Bey
‘Then differentiation of both members leads to
(2x dx + Jy dy) — (02 + y*) dx
=0,
(x? = y?) dx + Day dy = 0.
as desired. .
It is interesting to speculate here about the significance of x = 0 upon the
argument just used. The student should draw a few of the members of this
family of circles and observe what is peculiar about their behavior at « = 0.
‘The Elimination of Arbitrary Constants:
E (©)
nate B and « from the relation
<= B cos (wt +),
hh w is a parameter (not to be eliminated).
‘we obtain two derivatives of x with respect tof:
& 2 —wB sin (wt + 2
.
exy $x +4=0.
dy tay) th =O.
em -y +4¥)
‘substitution into the original equation leads us to the result
ely? + ey +450. .
examples suggest that in a certain sense the totality of solutions of
stder equation depends on n arbitrary constants, The sense in Which
‘true will be stated in Sections 5 and 12.w Definitions, Families of Curves {Chi
he aaa ea oer os ee oe SF's -0
6 -acectgul ere fees
[Lysates Am yen’ +0 +1
8 y= mx + Asha parameter, mto be eines Ane.
9. y= tax.
WW ywar siete Amy
Me yee tee
2 yds ge
yaa tae
M4 ya qe tee
IS yaxteetee™
Me ye ce te
Tye tte set Ame Sy + 6y = 6x07 10x +?
IR ym eet + ese Am yf —2f +y=0
19. y= Ae + Bre™ Am y'—4y +4y~0
BR y= cet cos 3x + ce sin Ix. Am 7 —4/' + 13y -0.
BL y= c,e cos bx + cp sin br; a and b are parameters.
‘Ams. — 2ay + (0? + b*)y = 0
BR yeent ae Am (e—y — xy +9
Dynex? + ee* Aas ate + 29" +(x? - Dy — x + Ny
Moyer tertaet Ame (e+ Wy # xy y= + 2+?
Boyne sget Ame afl — wy" + (2 — Ny = 22x — y=
Am 7 —5y+6y-0
4. Families of Curves
‘An equation involving » parameter, as well as one or both of the coordi
nates of 4 point in a plane, may represent a family of curves, one curv:
corresponding to each value of the parameter. For instance, the equation
(x - oP + (ye = 207, a)
Families of Curves
FIGURE 1
‘the constant c in equation (1) or in equation (2) is treated as an arbi-
‘constant and eliminated as in the preceding section, the result is called
‘differential equation of the family represented by equation (1). In this
the elimination of c is easily performed by isolating ¢, then differ-
throughout the equation with respect to x. Thus, from
BE
x+y
(x + yh2x dx + 2y dy) — (2 + dx + dy)
+P
rE (x2 + dey — y?) de — (0? = 2xy - yy
differential equation of the family of circles represented byP. n Definitions, Families of Curves 1ch.1
Families of Curves
y
yeti -VDe
Perievoy
\
FIGURE 2
FIGURE 3
a vertical tangent. From
ay = 0 jows that
‘we see that xy =2y30 o
ya(-1+ Six 6) differential equation of the family, We note that (9) is a first-order
ce differential equation. .
ya(-1- JS. 6 E (b)
In Figure 2, the straight lines (5) and (6) appear along with the family (1). differential equation of the family of circles (Figure 4) having tbeir
It is seen that the lines (5) and (6) cut the members of the family of circles in ‘on the y-axis.
‘a member of the family of circles of this example may have its |
precisely those points of vertical tangency.
For a two-parameter family of curves, the differential equation will be of
order two, and such a simple geometric interpretation is not available.
‘anywhere on the y-axis and its radius of any magnitude, we are,
‘with the two-parameter family
+a
EXAMPLE (a) ‘
Find a differential equation satisfied by the tamily of parabolas (Figure 3). hall eliminate both b and r ‘and obtain a second-order
‘having their vertices at the origin and their foci on the y-axis. jon for the family (10).
“An equation of this family of parabolas
ymax, ” x+0-dY =O
so that
ais. y = 2x. ®)
Reeatiy “: i “4“ Definitions, Families of Curves [cnt
y
FIGURE 4
Then
Sw + OPI — Ve + WI _ o,
oF 2
‘50 the desired differential equation is
xf - WP —¥ =0. .
Exercises
In cach exercise, obtain the differential equation of the family of plane curves
described and sketch several representative members of the family.
4. Straight lines through the origin Amy dx -x dy =0.
2 Straight lines through the fixed point (h,k);h and k not to be eliminated.
‘Am (y ~ k) dx ~ (x ~ h) dy = 0.
2 Straight lines with slope and y-intercept equal. Ama y dx —(x + I) dy = 0.
‘Straight lines with slope and x-intercept equal. Ame (y= xy
Straight lines with algebraic sum of the intercepts fixed as k
‘Am (xy — Wy ~ 1) + hy’ = 0.
(6 Straight lines at » fixed distance p from the origi
Ams. (xy — ym pith + 00)
7. Circles with center at the origin. ‘Am. x dx +y dy = 0
‘& Circles with center on the x-axis. Am yy +(yP +1 =0
“9, Circles with fixed radius r and tangent to the x-axis.
. Am (yt MY) + y? + Dry = 0.
Famities of Curves
tangent to the x-axis, Ase. (1+ PP = Dy +14 0FP
with center on the line y = —x, and pasting through the origin.
Ans. (x? — dey ~ yt} de +O? + xy = yd dy =O
lf radius unity. Use the fact that the radius coe oi
‘Am (7)? = 01 +072
Am U1 + OF = YF
olas with vertex on the x-axis, with axis parallel to the y-axis, and with
from focus to vertex fixed asa Am ay? my.
jas with vertex on the y-axis, with axis parallel to the x-axis, and with
tance from focus to vertex fixed as a As
is parallel to the y-axis and with distance from vertex to focts
Am ay = 1.
is parallel to the x-axis and with distance from vertex to focus.
Am 2ay +P =O.
AG a4 (Gyo
ty \dy, dy di \dx, ow
svers to Exercises 17 and 18 are equivalent
ith vertex and focus on the x-axis. Am yy + (YP =O
is parallel to the ax Am yy" HF =.
ntral conics with center atthe origin and vertioes on the coordinate axes.
Ane xy" + xy — yy =O.
‘onfocal central conics
vee
1d b held fixed, Aas (xy! — yoy + x) = (a? = ate
cubies cy’ = x'(e — 0) with a held fixed, Aum. 2x(x — aly = yi3x = 2a
cubics of Exerrise 2¢ with ¢ held fixed and a to be eliminatsd,
‘Ans. 2eytxy'— y= x
quate c#y* = x(x — 0)? witha held xed | Ame ‘aux — aly = Woe — ah
uartics of Exercise 26 with c held fixed and a to be eliminated. ©
cM axy = yP = 2Ixty,
ets 5 ae
ax16 Definitions, Families of Curves (cht
‘31. Circles through the intersections of the circle x* + y* = I and the line y = «
‘Une the “u + ko” form; that ia, the equation
BeyPoitky—ya0
Am (3? — xy ~ y? + 1) dx + (x2 Bay — y?— 1) dy = 0
‘32. Circles theough the fixed points (a, 0) and (—4, 0). Use the method of Exercise
a Am 2xy dx + ()? +a? — x2) dy = 0,
3X The circles r = 2atain 8 ~ cos
‘Ams (con 6 — sin 0) dr + r(c0s 8 + sin 8) d0 ~ 0
34. The cardioids r = o(1 — sin 8). ‘Am. (I sin 6) di +r cos 0:d0— 0
‘3K The cissoids r = a sin 0 tan 6, (See Exercise 29)
‘Ame sin O.cos 6 dr ~ x(t + cos? 6) 40 ~ 0
34 The strophoide r = atsec 4 + tan 8) pm Harsco
‘37, The trisectrices of Maclaurin r = a(4 cos @ ~ sec 8), (See Exercise 30.)
‘Ams. cos 4 cos? @— 1) dr +r sin M4 cos? 8+ 1) do ~ 0
The Isoclines of an Equation
this chapter we study several elementary methods for solving first-order
ial equations. Before we begin that study, we take a brief look at
‘ofthe basic geometrical ideas that are involved.
jider the equation of order one
dy
F=f. wo
‘can think of equation (1) as a machine that assigns to each point (a, 6)
‘domain of f some direction with slope f(a, 6). We can thus speak of
ction field of the differential equation, In @ real sense any solution of
{(1) must have a graph, which at each point has the direction equa-
this basic idea is to draw a short mark at
‘associated with that point. This can* Equations of Order One ICh2
ont
°
ge.
FIGURE 5
EXAMPLE (a)
‘Consider the equation
4
7
‘The isoclines are the straight lines f(x, y) = y=. For each value of « <
‘obtain a line in which, at each point, the direction dictated by the differe
‘ial equation is that number c. For example, at each point along the line
y= 1, equation (2) determines a direction of slope |. In Figure 5 we have
drawn several of these isoclines, indicating the direction associated wh
‘each isocline by short markers. If one start: at any point in the plane an
moves along a curve whose direction is always in the di of the direc
tion marks, then a solution curve is obtained. Several solution curves hove
“ween drawn in Figure 5. .
EXAMPLE (b)
‘Use the method of isoclines (o sketch some of the solution curves for '
«em
Here the isoclines will be the circles x? + y? = c, with ¢ > 0. When c= }
the isocline has radius $; for ¢ = 1, adit. '; for c = 4, radius 2. In Figure ©
‘we have drawn these isoclines, marking each of them with the appropria(e
direction indicator, and finally sketching several curves that represent sol!
tions of equation (3). .
each ‘of the following differential equations, draw several isoctines with appro-
FIGURE 6
tion markers and sketch several solution curves for the equation,
ayy
ab
20 Equations of Order One Ich. 2
the simplest of functions. Before discussing some of the
analytic techniques for finding solutions, we shall state an important
theorem concerning the existence,and uniqueness of solutions, a theorem
which will be discussed in detail in Chapter 15
Consider the equation of order one
Let T denote the rectangular region defined by
Ix-xolSa and = yol Sb,
a region with the point (xo, yo) at its center Suppose that f and (1/¢y ar
continuous functions of x and y in T.
Under the conditions imposed on f(x, y) above, an interval exists about
Xo] — Xo] Sh, and a function y{x) which has the properties.
(a) y = x(x) is a solution of equation (1) on the interval |x ~ xo| Sh
{b) On the interval |x — xo| Sh, (x) satisfies the inequality
Iyeo = yol $b:
(6) Atx= xo. 9 = x0) = Yor
(@) 4x) is unique on the interval |x — xo] = h im the sense that it is the
‘only function that has all of the properties (a), (b), and (c)
The interval {x — xo] Sh may or may not need to be smaller than the
interval |x — xo| S a over which conditions were imposed upon f(x, })
In rough language, the theorem states that if f(x, y) is sulliciently well
behaved near the point (xo, yo). then the differential cyuaiion
aj a
hhas a solution that passes through the point (xo. Yo) and that solution »
unique near (x9. yo).
In Example (a) of Section 5 we can consider (xo. Yo) to be any point 10
the plane, since f(x, y) = y and its partial derivative df/dy = | are contin
lupus in any rectangle. Therefore, our existence theorem assures us tha!
through any point (xp, yo) there is exactly one solution, a situation that we
‘assumed when we sketched the solution curves in Figure 5
“Again in Example (b), Section 5, the function f(x, y) = x7 + y* and its
‘partial derivative dffay = 2y are continuous in any rectangle. It follows that
through any point (xo, yo) in the plane there is exactly one solution curve.
fact that is suggested by the solution curves in Figure 6.
Separation of Variables au
"Separation of Variables
Ibegin our study of the methods for solving first-order, equations By
ng an equation of the form
M dx +N dy
M and N may be functions of both x and y. Some equations of this
s0 simple that they can be put in the form
A(x) dx + Bly) dy = 05 a
Fis, the variables can be separated. Then a solution oan be written at
Bor itis only a matter of finding a function F whose total differential
Heft member of (1). Then F = c, where c is an arbitrary constant, is the
d result
\MPLE (a)
the equation
ye
ax x
fe note that for the function in equation (2), the theorem of the preed-
Tsection applies and assures the existence of a unique continuous solu=
Sthrough any point in the first quadrant. By separating the variables we
write
for x > Oand y>0. @
ee
ce we obtain a family of solutions
In jy =2 In Lx] +6 @
use we are in the first quadrant,
yas? @)
420 & 6
LE (b) Be
quation of Example (a) for x i
ewan of PM ee, Fit iy
ed as before to equation (2). However equation (3) must
typeerxt ty ee i
‘ Sa2 Equations of Order One Ich. 2
Second, if y = 0, we see immediately that since x #0, y = 0 is a solution
of the differential equation (2)
‘As a matter of convenience the solutions given by equation (6) are usually
written
yeex, ”
where c, is taken to be an arbitrary real number, Indeed, this form for the
solutions incorporates the special case y = 0. Thus we say that the family of
curves given in Figure 3, page 13. represents a family of solution curves for
the differential equation (2),
‘We must be cautious, however. The function defined by
a) =<, x>0
=-43, x50,
obtained by piecing together two different parabolic arcs could also be con
sidered a solution of the differential equation, even though this function 1s,
not included in the family of equation (7). The uniqueness statement in the
theorem of Section 6 indicates that, as long as we restrict our attention to «
point (xo. yo) with x #0 and consider a rectangle with center at (ss. 5
containing no points at which x = 0. then in that rectangle there is a un»quc
solution that passes through (Xo, yo) and is continuous in the rectangle. m
EXAMPLE (c)
Solve the equation
(+ dx Ul + Vay
with the “initial condition” that when x = 0, y
If we write this .quation in the form.
dy _ 0+)
dx te
we observe that the right member and its partial derivative with respect (0 )
are continuous near (0, ~1). It follows that a unique solution exists for
equation (8) that passes (urough the point (0, — 1).
equation we get
from which it follows at once that
are tan x + arc tan y = ¢. OO}
Separation of Variables
tangent and is subject to the restriction
— $n 1 In this ext
‘onto the left branch, equation (10), by the initial condition |
x=0.
c tions (10) and (11) can be
Rion betwen acetate aaa
=1), would be:ua Equations of Order One Ich.2
FIGURE 7
prevent the computer from learning of the existence of the other branch of
the hyperbola (11). .
EXAMPLE (d)
Solve the initial value problem
Bay + 1) de ~ y dy = 0, (2)
where x = Oand y = ~2.
Separating the variables in equation (12), we obtain
1
sa) ov
we get a family of solutions given implicitly by
Pr aee(1
Integra
ay-iniy+ tite. (13)
‘Since we seek a member of this family that passes through the point
(0, ~2), we must have
O=-2-Inj-1] +6,
or
en?
‘Thus the solution to the problem is given implicitly by
Pomyoiniy+ i+?
ae
Separation of Variables
should note how the theorem of Section 6 applies to
1 through 6, obtain the particular solution satisfying the initial condi
‘each exercise interpret your answer in the light of the existence
Ams p= ro exp (<2).
met + y?p when x = 2, Am ior fee
Bie wisn s = 27 ~ pe =
ei ohen x= iy a1 ka =
fie = 3x dy: when x= —2.y=1 A y= 729.
idx = 3x dy; when x = 2,y =~. Ans. y= —(x/2P"
condi-
7 through 10, obtain the particular solution satisfying the
2
An. = In oe
An. y= effet = x= 1)
y= 0.
x exp (y — x2), when x
dx + e* dy <0; when x-+ 05, 9+
=P) dr =P? sin 8 d0; when 0 = 0, =a
fax) = gi when x = Xo.0= Y%
11 through 37, obtain the general solution,
Ans. v7? — 05 = 2gtx — xa
mt +2)dy =O
te dy =
ri ee tye ay
Barts +? + OS Wola ae
dx + tan y dy = 0.
snetdy= 0.- arte tert One ena
Bye ers «,
26 yy wet eee:
37. tan? y dy = sin? x de. ee ee stan ye
2B y= cos? x cosy. eae ee eta 1m Be + ain 2x a
adh dat A lee Ans. y= ofeec x + tan x)
30. dx = 1 + 1°) sec? x de Aes, 2x4 sin xm e+ (140)
31 ("+ My =», Tm yh de) — 02
32 a dp + fda + 93 da + ap) = 0. Rie AEB = exp (—30.— 9,
BB (1+ In dx +1 + In dy = 0. ee wine t yiny~<
Moxa — few dy 0. i
Ree enteral srthe circle? + (y — 0) = 0
Moxdx + =x dy = 0.
TAR pe aati tae per hall of the crc x* + (y —
36 a? dx = xa dy, Ams. x =asec®*<
37. yin xn yd + dy = 0, Ans.
xInx+In|Iny|=x +e
8. Homogeneous Functions
Polynomials in which all terms are of the same degree, such as
x3 — xy + 4y3,
ey, 0)
xty + IV.
are called homogeneous polynomials. We wish now to extend the concept of
‘homogeneity so it will apply to functions other than polynomials.
‘If we assign a physical dimension, say length, to each variable x and y
the polynomials in (1), then each polynomial itself also has a physical
dimension, length to some power. This suggests the desired generalization
Mf, when certain variables are thought of as lengths, a function has the
physical dimension length to the kth power, then we shall call that function
homogeneous of degree k in those variables. For example, the function
x
Sx, y= 2y? exp () ve
lengthy’ when x and y are lengths, Therefore, that function
| to be homogeneous of degree 3 in x and y.
Homogeneous Functions
mit the degree k to be any number. The function JE 47
sneous of degree zero in x and y.
ft definition of homogeneity is: The function f(x, 9) is said £0 be
aus of degree k in x and y if, and only if
fax, dy) = 4% Ws
nition is easily extended to functions of more than two vanables:
Bf the function f(x, y) of equation (2), the formal definition of homo-
y Ieads us to consider
° 7
ix + Uy
His = orale
‘see at once that
slax, ay) = BF Mi
ix, y) is homogeneous of degree 3 in x and y, as stated previously.
following theorems prove useful in the next section.
‘4, If M(x, y) and N(x, ») are both homogeneous and af the same
The function Mix, »V/N(x, ) 's homogeneous of degree zero.
fof Theorem I is left to the student.
2, IF f(x. y) ts homogeneous of degree zero in x and y. Sts 15-@
‘of y/x alone.
Let us put y = ox. Then Theorem 2 states that if/(x, 9) is homage: |
Sof degree zero. f(x, ») is a function of v alone, Now A
Fx, y) =Fo% 0x) = XFL, 0) = Fle Oh o
Z
ch the x is now playing the role taken by 2 in the definition (3) By”
¥)) depends on v alone, as stated in Theorem 2: 2
i
sin each exercise whether or not the function is
state the degree of the function.ca Equations of Order One {ch.2
se 6. tan x
rep (2) 8 an
y x
40, xsin = y sin.
9. (xt + yep a
a
a) a
ee
Bt dy
n :
im x= dy
13, (u? + oP
1S»? tan =
acre
+46
oe
1 xtnx—yiny,
1.
18 n~
¥
20, x In x—xIny,
‘Ams. All functions are homogeneous except those of Exercises 2, 5. 6, and 19
9. Equations with Homogeneous Coefficients
‘Suppose that the coefficients M and N in an equation of order one
M(x, y) dx + N(x. y) dy = 0, a
are both homogeneous functions and are of the same degree in x and y. By
Theorems 1 and 2 of Section 8, the ratio M/N is a function of y/x alone
Hence equation (1) may be put in the form
BE
2. a? =o @
‘This suggests the introduction of a new variable » by putting y = vx. Then
(2) becomes
eee tiien ql
in which the variables are separable. We can obtain the solution cof (3) by
‘the method of Section 7, insert y/x for v. and thus arrive at the solution !
(1), We have shown that the substitution y = vx will transform equation a)
“imto an equation in v and x in which the variables are separable,
Peace
Equations with Homogeneous Coefficients, a
above would have been equally successful had we used
to obtain from (1) an equation in y and v. See Example (6).
(x? — xy + y?) dx — xy dy =O. @
the coefficients in (4) are both homogeneous and of degree two in =
Tet us put y = vx. Then (4) becomes
(x2 — x20 + x0) dx — x2x{v dx + x do) = 0
‘which the factor x? should be removed at once. That done, we have 1
(1 — 0 4 0) dx — 0 dx + x do) = 0,
(1-0) dx — xv dv = 0.
of solutions is seen to De
In |x| #o+Injo— t= intel,
xe Nera e é
‘of the original variables, these solutions are given byeI Equations of Order One 1Ch. 2
EXAMPLE (b)
Solve the equation
xy dx + (x? + y3) dy =0. (3)
Again the coefficients in the equation are homogeneous and of degree
two. We could use y = ox, but the relative simplicity of the dx term in (5)
suggests that we put x= vy. Then dx =o dy + y dv, and equation (5) 1s,
replaced by
wytu dy + y dv) + (y+ y*) dy =
wo dy + y de) + (v* + 1) dy =0.
Hence we need to solve
vy dv + (20? + 1) dy =0, o
which leads at once to
In 20? + 1) +4 In |y| = In,
«
YiQu? + I) =e
‘Thus the desired solutions are given by
that is,
V2x? + yy) =e. a
Since the left member of equation (7) cannot be negative, we may, for sym
‘metry’s sake, change the arbitrary constant to cf, writing
PRxe + v?)— eh.
It is worthwhile for the student to attack equation (5) using y = vx. That
‘method leads directly to the equation
(@? + 20) dx + x{0? + 1) dv = 0.
Frequently in equations with homogeneous coefficients, it
| terial whether one uses y = vx oF x = vy. However, it is sometimes:
-guibstitute for the variable whose differential has the simpler coefficient.
Equations with Homogeneous Coefficients
ses 1 through 21, obtain a family of solutions.
xt +?) dx — 2xy dy =0
= 29) dx + (2x + y)dy = 0.
B+ y3) dx — xy dy
ide — (37 + 397) dy ~ 0.
me + Tay + 29%
dx +(x? +?) dy = 0.
wa
An. 2 = 2 + yh
Ame inet +99 Aiea
‘Ane. xt =} A
‘Ans, x4 = 6y? In |y/el
Am. ty +'20) = y+
‘iw ae ae
Ans. xy +2)? = ety — 2a
‘Ams (Go+ 1)? = cf — wh
Ans, 32 + ay? mtx +h
Ans. (x7 + y*P = 6y* In Je/¥|-
Am. (x! + 2y)) mee
‘Ans. yx + y)= cet
Idx + x(x + 0) dv = 0 ‘Ans. x0? = ofx + 20),
tes (y/x) — y] dx + x dy =0. Ans. In.|x/c| = cos (y/a)
Das + sin? (y/x)Ly dx —x dy] = 0. Amn 4x in |t/el = 2y + sin yl
Gay iny ty inx) dx + xin y— In x) dy=0.
‘Ans. (x —3)lnx +yln y= ex
Ey arc tan (y/x)) dx + x arc tan (y/x) dy = 9.
‘Ans. 2y are tan (y/x) = x tn Cet? + Ve.
‘Ans. y In Lylel = =
‘aos. 5? = —2t7 In fest.
‘Ans. arc sin (x/y) = In |y/el
‘Aas. (y ~xXy + 3x) =O.
eA xy — 4y") de — (3?
he + yy de — x dy) +9 dy
den (e+ Jy"
Gxt — xy 39°) de
vipa) dx + Hx, yky dx = x dy) = OF
Fe H(-x, y) is homogeneous in x and 5
23 through 35, find the particular solution indicated:
ays?
Ane. 22x +39) + (8 + In (x +9) = 0.
EEE de ~ x dy = Ocwhen x= Oya t Ame adap
UCP) dx — xy = Ovwben x= J =), A
when x = 1 =n
my) dx +x
‘Gay + V6x2) dx + x2 dy & Of when x= 1 yh f
ss Am x—y= Sy +4
2) dy = 0; when x = y= I
te 4 (x? + dey + 4’
be + 6x2 + 29°) dy = 0; when x= 0 y=
egy + yas — x'0x — dra Oi whee32 Equations of Order One Ic
Bi. 9x — 2y) dx — 16x — y)dy = O; when x= I, y= I.
Am. 3x? -x*y ~ 2)? = 0,
32. ye? + Pax + Sx? — Sy") dy = 0; when x = 2 y= 1
Am 2y? — 2x77? + 3x <0,
BB. (16x + Sy) dx + (3x + y) dy = 0; the curve to pass through the point (1, — 3),
‘Am y+ 3x (y+ 43)In(y + 40,
34 x + 20) dx — x? dv = 0; when x = |. Am 2x? + 2x" — v= 0.
3S Gx? = 2y")y' = 2xy; when x = 0, Am x= 27+ 1
‘36 From Theorems | and 2, page 27, it follows that if F ixhomogeneous of degree
‘ein x and y, F can be written in the form
raed) "
‘Use (A) to prove Euler's theorem that, if F is a homogeneous function of degrce
Kin xand y,
ar oF
Rtg
10. Exact Equations
In Section 7 it was noted that when an equation-can be put in the form
Alx) dx + Bly) dy = 0,
‘a set of solutions can be determined by integration, that is, by finding »
function whose differential is A(x) dx + B(y) dy.
‘That idea can be extended to some equations of the form
M(x, ») dx + N(x, y) dy = 0 0
in which separation of variables may not be possible. Suppose that a func
tion F(x, y) can be found that has for its total differential the expression
M dx +N dy; that is,
dF = M dx +N dy. @)
‘Then certainly
Fix ym °
defines implicitly a set of solutions of (1). For, from (3) it follows th:
dF =O,
or, in view of (2),
M dx + Ndy=0,
Exact Equations | :
‘things, then, are needed: first, to find out under what conditions on
Na function F exists such that its total differential 1s exactly
WN dy; second, if those conditions are satisfied, actually to deter-
function F. Ifa function F exists such that
M dx +N dy
y the total differential of F, we call equation (1) an exact equation:
‘equation
M dx+Ndy=0 a
then by definition F exists such that
dF = M dx +N dy.
calculus,
aF
Sax +54,
+5”
derivatives are continuous. Therefore, if (1) 15 4m
d that these partial
‘equation, then
om
oy
Ip to be exact itis necessay
ly ‘show that, if condition (4) is satisfied,
y) be a function for which
ei
ooM Equations of Order One [Cha
The function ¢ is the result of integrating M dx with respect to x white
holding y constant. Now
te _ aM
dy ax” ay’
hence, if (4) is satisfied, then also
26 oN
Gx dy” Ox” 6)
Let us integrate both sides of this last equation with respect to x, holding y
fixed, In the integration with respect to x, the “arbitrary constant” may be
‘any function of y. Let us call it B'(y), for ease in indicating its integral. Then
integration of (5) with respect to x yields
a
= =N4 BU) )
‘Now a function F can be exhibited, namely,
F = (x, ») — BO),
for which
ote 22 ry) dy
aF 5g ay oY — BU) dy
= M dx +(N + B(y)) dy — BY) dy
=Mdx+N dy.
Hence equation (1) is exact. We have completed a proof of the theorem
stated below
Theorem 3. If M,N, 2M/dy, and ON/2x are continuous functions of x and y
then a necessary and sufficient condition that
M dx+Ndy=0 qa)
be an exact equation is that
am _ aN. “)
dy Ox
Furthermore, the proof contains the germ of a method for obtaining a s),
where Q(x) will be determined from (12). The latter yields
=xy? — 2y + Q(x) = 2x? — xy? — 2y + 3,
Q(x) = 2x? +3.
‘Therefore,
Q(x) = tx* + 3x,
and the desired set of solutions of (11) is defined implicitly by
a dxty! — dey + dxt + 3x = 4c,
A et eam .
Exercises
“Test each of the following equations for exactness and solve the equation. The equa
tions that are ot exact may be solved by methods discussed in the preceding
I
AG tx + pax + x — yNdy = 0. Ame tp dey yin
M6x + y") dx + f2x — By) dy = 0. Am 3x? + xy? ye
Qxy — 323) dx +(x? + yy =0. Am ty + by = 6
(f) ~ Bry + 6x) dx ~*~ 2xy + Nady =O. Am xy? —x*y + Ix? — Dy =o
ay + yhde +(e? — x)dy =O. ed ene = 1"
Be — det Ay — ah dy 0 Am x + dy? = ary te
phere
‘The Linear Equation of Order‘Ove a”
6 by another method,
3y) dx + (2y ~ 32) dy = 0.
8 by another method
B= 3) di + (3u?e? — 3u + 40) dv = 0. Am dite —3u + 2m
© 3x2)?) dx + (008 2y — 2x sin 2y — 2x*y) dy =O
hon} tin 2y + x 008 2y = 2998 =e
‘Ans. 2 arc tan x +In(1 + 9) a
Am 2x + (1 +2P =e
Ans xt + yhm day te
idx 4 (xty + 9) dy = 0.
4 ny?) di + (xy + y + Day) dy
=a) dw + + wie — w) de An (w+ 2) = awe +6
de + (42 — x sec! y) dy = 0 ‘Am. xty—x tan ye
fpieee y — cot x) dx — sin x sin y dy =0. Ame sin xcos y=ln|e sin sh
i tin @ — cos 6) + risin 0 + cos 0) db = 0
‘Asa. 1? + 2nsin @ — con 8) = 6
Say? +6) dx + (x? — 6x"
Am Oya 2ely + Isto yee
IB — 2r cos? @ dr +r cos H2r sin 9 +1 do = 9.
Am rain 0— 7? cost Ome.
hu 2 sin Gy) =e
Am yx? + Ppa
Am xi + y= ey
am. xy? + xy ee
when x = 0,y=
Am. xXx? 3) = 41 —
er 2 4 x41 — xy?) dy = 0; when x= 2p = I
a” me Am xy —y? + Sxy— 3a =
2 x) dx + (x + 2xy — y sin 2x) dy
Y + 2y? sin® x) dx +(x + 2 Oa ee
24 ay) 4 exp (2) dy =
An. xy +) + 2x? + yeap yee
be + xy dy = 2004 9) dy when Ba
meee tn? he Tay = 2 4 (et a) = B92 1S:
Ub y cos (xy)] dx + x cos (xy) dy = 0
dx + (9? + x1) dy =0.
idx + (y? — x7) dy = 0.
ay — 2) dx + xixy + 1) dy =
Si) dx + (x? + By - 3) dy
4+ y—yexp(— x79] dx +e
The Linear Equation of Order One
Qn ied first-order differential equations that were exact, If
0 i i ate
‘of an appropriate factor, which is then called an
d, in Section 7 ws multiplied by an integrating factor
‘variables and thereby obtain an exact equation. —
very little can be said about the theory of int
‘equations, In Chapter 4 we shall prove some
ftasistance in a few isolated situations:J Equations of Order One 1Ch. 2
class of equations, however, where the existence of an inte,
: " grating factor can
be dcraountrated, Thir its caus oflieear equations of order one.
An equation that is linear and of order one in the dependent variable y
‘must by definition (Section 2) be of the form
dy
Als) Fe + Bey = Cl). (1)
By dividing each member of equation (1) by A(x), we obtain
a a 2
2 + Peay =O @
which we choose as the standard form for the linear equation of order onc
For the moment, suppose that there exists for equation (2) a positi«
integrating factor (x) > 0, a function of x alone. Then
ce 2, rey] = x) Cx) @)
must be an exact equation. But (3) is easily put into the form
M dx +Ndy=0
with
M = vPy — 0
and
Nao,
in which v, P, and Q are functions of x alone.
Therefore, ifequation (3) is to be exact, it follows from the requirement
ae ek
éy ~ ox
that » must satisfy the equation
(4)
“From (4), v may be obtained readily, for
Pdxa®,
7
Equation of Order One »
inv [pax
v= exp ({ Pax) ©
if equation (2) has a positive integrating factor independent of ¥,
factor must be as given by equation (5).
ins to be shown that the v given by equation (5) is actually am
dy
aa + Pely= O08), 2
‘multiply (2) by the integrating factor, obtaining
D wio({ rte) Z+ren({rep=cen (fra)
member of (6) is the derivative of the product
yero({? a)
st member of (6) is a function of x only. Hence equation (6) is exact,
is what we wanted to show.
WF course, one integrating factor is sufficient. Hence we may use in the
nt (f P dx) any function whose deriv P.
of the great importance of the ideas just discussed and the fre-
Seecurrence of linear equations of first order, we summarize the steps
in solving such equations:
Put the equation into standard form:
eyes
Z+ Pye d
tain the integrating factor exp (f P dx)
{ultiply both sides of the equ
ting factor.
Wve the resultant exact equation.
n integrating the exact equation that the
ys the product of the dependent“ Equations of Order One tena
EXAMPLE (a)
‘Solve the equation
My — 4x3) dx + x dy = 0.
‘The equation is linear in y. When put in standard form it becomes,
2
Bely-sx when x40 ©
exp (2 In |x) = exp (In x?) = x’.
‘Next apply the integrating factor to (7), thus obtaining the exact equation
4
eZ +ry a8, ®
which may immediately be rewritten as
4
Ee=&. o
By integrating (9) we find that
aty = xt +0. (10)
‘This can be checked. From (10) we get (8) by differentiation. Then ‘he
original differential equation follows from (8) by « simple adjustment
‘Hence (10) defines a set of solutions of the original equation. .
EXAMPLE (b)
Solve the equation
y dx + (3x - xy + Ddy =O.
Since the product y dy occurs
Fe etn Derm the equation isnot linear in y. 1
ydx+(3- yx dy= —2dy
‘and pass to the standard form,
S+Q-)eag meres
JQ -1)a-simi-yren
co)
‘The General Solution of a Linear Equation a
integrating factor for equation (11) is.
exp (3.n |y| — y) = exp In|yl)-€-
= exp (in |yP)-e-?
=ly
Ihat for y > 0, ye~” is an integrating factor for equation (11) and
=yPe"? serves as an integrating factor. In either case we are led
equation
yre-? dx + y'3— ye?x dy = ~2y8e? dye
we get
yreP'dy
xe
= 2y2e? + dye? + de
amily of solutions is defined implicitly by
xy? = 2y? + 4y + 4+ cet +
The General Solution of a Linear Equation
Beginning of this chapter we stated an existence and. walateases
Te: tratorder differential equations. If the differential equation im
happens to be a linear equation, we eam prove &
‘statement.
the linear differential equation
B + Pixdy = Qt) a
that P and Q are continuous functions on the interval <2 -<05
aaa e nn any number in that interval I vy i Sal arbiiracy eas
Mae Geists 2 unique solution y = 248) of differential tuations G9)
‘satisfies the initial condition
this soluti
of this theorem has essentially Deen
‘of equation (1) by the= Equations of Order One [ch
integration gives
pairs:
Since v # 0, we can write
It is a simple matter to show that since » #0 and v is continuous on
a Mey +x Day A dm p= Xe +)
WS Gx? — dey — Ph dx sy eo Am x? me ciy— x)exp(y
Me He! + Phas + Se — Sy) dy = Ov when v= Dy =|
Am. 2y? — 2x*y? + 3x =
‘97. ¥ + ay = b; 2 and b constants Solve by two methods, Am. y= bia + ce ©
‘TR &e — idx — (x + yi dy = 0 Solve by two methods. «Am. x’ — 2xy — =
19, delat = con x con? ‘Am 4injsecx + tan x| = 20+ sin 2 +
aoe Am xsiny+ycosx=c
BA (1+ ay — Aap de + (x — 2) dy =O; when x = 2 y=
Am ixtye + des 2inix—!
ity = 247-9 Am yecty—dexpix*
ER Gy con x + sin* x) dx = sin x dy; when x = bx y= 1, me?
Ams ym 2sin? x sin? $<
UA yids — ay) = 2 dy + 7 de ney inicry
erate asp ecco aa
‘Am Dare tan (y/a) = In [efx + ab(x - «)
Bo-w 2eserty ate eases sca =
es mame ons Ate Am wis dny tyne
ME Bey Be) dx + (2! + Dy) dy = Am ty- ty
nercine 29 through 3 reall that the principal ofthe
aes
ano
“hae seein x + are sn y =
‘Soma ep ada ta temee oo
“The General Solution of 2 Linear Equation s
x TS
— {x or that arc of the ellipse:
any sli nein Figure
Son of Ese condition that when «=O
ame wenn > Grrr that ar of the ligne
Iv te xy + y= 2 that is indicated DY 2 light solid line in Figure 8
toa ee and. 3i have been deleted ihe
that after the answers
ig arcs of the ellipse
erent
pe not solutions of the differential equation:
[iar ax + (Ta ay 8
thi purpose consider the sgn ofthe slope of the carne
‘the equation
Aap ax- /T= Rayne
xercises 29 through 32.
fend solve four problems analogous 10 E
Am tum cel — (e+ Uh
eo = aye x= tye ol — saps
ey t2en% aye me
Bay = Pde — cP oy ant
ae -hoel
Stay ty? - Day = orwnen 1“ Equations of Order One Ich.
BD. Wy? — 3x9) dx + 2? dy =
M0. y= cos x — y sec x; when x = 0, y = 1.
‘Abe, y(1 + tin x) = cos x(x + 2 ~ co
4 Fie that sation oy = 3x +> which passes cvough the point —t.0,
Am. 2x8 = xt +0)
i Am. y= 3x +
en cnn dior Oo poi”
‘Ass. = exp (« + 1) —
mone es:
4. (e? = 1 + 2) dx + (1 = 7) dy = Of when x = By =).
Ame tee tie Hie 4 1 +210 0)
ema ees.
S07 ey Ca ae dew yox- y
LP + yh de ebay + dye Or when xy — tA 2 +
Steamy Am Bax + yim ex's ty)
@ yew ra punk Am yotnztec
eas ee Am. itesy| en
xy mx y+ xy tan, Ame xy con cmc +08 x + x sin
Re eee
0 ee Ans. erty +0)
x sin y + tan x) dx — cos x cos y dy = 0. Ams, x win |c sec ‘
So ey
Am. 2x%(x — 2)y = x? — 5.
felocity of Escape from the Earth
1 probtems involve differential equations of erder one:
if the problem of determining the velocity of a particle projected
Bet direction outward from the earth and acted upon by nly one
Pgravitational attraction of the earth.
Sail assume an initial velocity in & radial direction so thar the
the particle takes place entire
Jy on a line through the center of
to the Newtonian law of gravitation, the acceleration of His
Ie inversely proportional to the square of the distanes fam the
Bike canter of the earth. Let r be that variable Gistanes sams I
Peeine earth, If represents time, ¢ the velocity, Othe! Bartels
a gad k the constant of proportionality i. she Newianlian
ion is ne
Hs negative. When r=.“s Elementary Applications Icha
gravity at the surface of the earth. Thus
from which
‘We wish to express the acceleration in terms of the velocity and the dis
tance. We have a = dv/de and v = dr/dt. Hence 5
do
a-$-5
s0 the differential equation for the velocity is now seen to be
dv _gR?
Fig Reale af a
‘The method of separation of variables applies to equation (1) and leads at
‘once to the set of solutions
2gR?
es
.
Suppose that the particle leaves the earth's surface with the velocity vo
‘Then v= vo when r = R, from which the constant {is easily determined to
be
C= 03 — 20R
Thus, « particle projected in a radial direction outward from the earth's
surface with an initial velocity og will travel with a velocity v given by the
‘equation
0 = WE 5 0 — 29R @
It is of considerable interest to determine whether the
particle will escape
rs velocity is
‘= #9. An examination of the rit member of equation (2) shows
‘that the velocity of the particle will remain positive if, and only if
03 — 29R ZO. eo
Newton's Law of Cooling, -
ity (3) is satisfied, the velocity given by equation (2) will remain
use it cannot vanish, is continuous, and is positive at r=
hand, if (3) is not satisfied, the vj ~ 2gR <0, and there will be
fue of f for which the right member of equation (2) is zero, That
‘would stop, the velocity would change from positive to mens
@ the particle would return to the earth.
Projected from the earth with a velocity 9 such that
Be from the earth. Hence, the minimum such velocity of projections
Jak Oy
Bi the velocity of escape.
Bessie of the earth is approximately R = 3960 miles. The senties:
previty at the surface of the earth is approximately g = 32.16 feet
Biper second (ft/sec), or g — 60X10) > mile/see', For the earth,
ty of escape is easily found to be vy
soltne gravitational pull of other celestial bodies the moos.
Sanus’ and so on, has been neglected in the idealized! proves
Se 1k is not difficult to see that such approximations are j
Be Yatereated in only the critical initial velocity ¥.= Whekies She
Bercy recedes from the earth forcver or becomes for eReOass
Be ioine heavenly body is of no consequence i i: Bears
study we happen to be thinking of the particle 88 80 idealization
Hate rocker then other elements most be considered A resist=
ible. Methods for overcoming
feulties are not suitable topics for discussion here
He aoalized that the formula v, ~ /20R applies ealually well for
Bat escape from the other members of the sales ESSA Jong,
fare given their appropriate values
Newton’s Law of Cooling
Sit has shown that under certain conditions, a good
|to the temperature of an object can be obtained by
‘The temperature of &
the difference in temperatu0 Elementary Apptications Ich. 3
the body itself. We shall assume here that the constant of proportionality is
the same whether the temperature is increasing or decreasing.
‘Suppose, for instance, that a thermometer, which has been at the reading
70°F inside a house, is placed outside where the air temperature is 10°!
‘Three minutes tater it is found that the thermometer reading is 25°F. Ww.
wish to predict the thermometer reading at various later times.
‘Let w (°F) represent the temperature of the thermometer at time ¢ (rn
the time being measured from the instant the thermometer is place:
outside. We are given that when ¢ = 0, u = 70 and when t = 3,u = 25
According to Newton’s law, the time rate of change of temperature, dis
is proportional to the temperature difference (u — 10). Since the thermo
‘meter temperature is decreasing, it is convenient to choose (—K) as the con
stant of proportionality. Thus the wis to be determined from the differential
‘equation
~~ Mu 10, ()
and the conditions that
when t= 0,u= 70 2
and
when t= 3,u= 25. QB)
We need to know the thermometer reading at two different times because
there are two constants to be determined, k in ¢quation (1) and the
“arbitrary” constant that occurs in the solution of differential equation (1)
From equation (1) it follows at once that
w= 10+ Ce™
‘Then condition (2) yields 70 = 10 + C from which C = 60, so we have
u= 10 + 60e 4
‘The value of k will be determined now by using condition (3). Putting
f= 3 and u = 25 into equation (4), we get
25 = 10 + 60e-™,
from which e~** = 4,20 k = $n 4,
‘Thus the temperature is given by the equation
w= 10 + 60 exn /—4¢ In 4), ”
Since In 4 = 1.39, equation (5) may be replaced by
w= 10 + 60 exp (0.461), ©
Simple Chemical Conversion st
ple Chemical Conversion
mount of unconverted substance be known at some specified
fet x = Xo at ¢=0. Then the amount x at any time t >is
by the differential equation
i
di o
dition that x = xo when t = 0, Since the amount x is decreasi
pases, the constant of proportionality in equation (1) is taken to
ion (1) it follows that
Biwhen 1 = 0. Hence C = xq. Thus we have the result
seer Q
fadd another condition, which will enable us to determine k.
Ge known that at the end of half a minute, at ¢ = 30 (sec), two
fofiginal amount xo has already been converted. Let us deter-
jmuch unconv 60 (sec),
Gsthirds of the substance has been converted, one-third remains
d. Hence x 30. Equation (2) now yields the rela-
.d substance remains at €
4x0 when
fk is easily found to be ys In 3, Then with ¢ measured in seconds,
Bt of unconverted substance is given by the equation
Ast In 3). a)
x = Xo exp (
21n 3) = xo"? = Bor
Xp &XP (
Bs of the moon is roughly 1080 miles, The acceleration
oo oe ic 0163p, where g i the acceleration of grav
oe i apomine the velocity of eveape (Or the RANA is
BN2 Elementary Applications res
2 Determine, to two significant figures, the velocity of escape for exch *he ce
tial bodies listed below. The data given are rough and g may be Wuke,. (a
6.1110)" * milejne?. ‘ahd
‘Acceleration of Gravity Radius Answer
a Supece Crlles) _(milesiic)
Venus, 0859 3,800, 63
Mars 038g 2,100 31
Jupiver 26g 43,000 ”
‘Sun 28g 432,000 380
Ganymede O12 1.780 16
3. A thermometer reading 18°F is brough om where the temperatur.
70°F; 1 min later the thermometer reading is 31°F. Determine the temperature
reading as a function of time and, in particular, find the temperature reading §
‘min alter the thermometer is first brought in’o the "sor
Am wm 70— 5: exp (—1 291); whens = 5,u~ 58
4 A thermometer reading 75°F is taken out where the temperature is 20 F The
reading is 30°F 4 min later. Find (a) the thermometer reacling 7 min after 1
thermometer was brought outside and (b) the time taxen for the reading to drop
from 75°F to within a half degree of the air temperature
* Am (a) 23°F; (b) 11.5 min
& At 1:00 pa, a thermometer reading 70°F is taken outside where the air tem
perature is —10°F (ten below zero). At 102 F.m., the reading is 26°F At
1:05 Ps, the thermometer is taken back indoors where the air is at 70°F What
is the thermometer reading at 1:09 p.? Am. 56°F
6 At 9 4m, a thermometer reading 70°F is taken outuvors where the temper
is 15°F Ar 9:05 am, the thermometer reading is 45°F. At 9:10 Aq. the ther
mometer is taken back indoors where the temperature is fixed at 70°F. Find (a)
the reading at 9:20 am. and (b) when the reading, 10 the nearest degree, ll
show ‘ne correct (70°F) indoor temperature. ‘Ams. (a) 58°F; (b) 9:46 40
7. At 2:00 rm, a thermometer reading 80°F is taken outside where the air term
perature is 20°F. At 2:03 ra. the temperature reading yielded by the therm
‘meter is 42°F. Later, the thermometer is brought inside where the air is at 80
‘At 2:10 9. the reading is 71°F. When was the thermometer brought indoors”
im. At 2:05 7
| Suppose that « chemical reaction proceeds according (o the law given in Section
15 above. If half the substance 4 has been converted at the end of 10 sec, find
when nine-tenths of the substance will have been converted Awe. 33 s0c
'%. The conversion of a substance B follows ine law o°ed in Section 15. If only 8
fourth of the substance has been converted at the end of 10 sec, find when
‘nine-tenths of the substance wil, |..ve been converted, ‘Ama. 80 sec
‘I. Foor « substance C, the time rate of conversion is proportional to the square of
the amount x of ure mveried substance Let k be the numerical value of (he
Simple Chemica! Conversion ce
‘of proportionality and let the amount of unconverted substance Be
BE Determine x for ale z0 fon <~ oll + xpht)
tance D, the time rate of conversion is proportional tothe square
Amount x of unconverted substance, Lat & be the numerical value of
tof proportionality. Show thatthe substance ill appear i finite
ibtarmine that time As ta Dale
lances, A and B, nrc being converted into u single compound C tn the
Wt has been shown that, for these substances, the following law of
holds: the time rate of change of the amount x of compound € fs
onal to the product of the «mounts of unconverted substances A and B.
ithe units of measure so chosen that one unit of compound C is formed.
feombination of one unit of A with one unit of B. If at time ¢ = O there
Bits of substance A, b units of substance B, and none of compound ©
‘show that the law of conversion may be expressed by the equation
dx
So ka — xyb — x)
a
Jequation with the given initial condition
abtexp (b- okt =). py og xo tHE
Pe bts ob omer ake +1
lution of Exercise 12, assume that k > 0 and investigate the behavior of
Ame Wb2a,xa:ifb Sax—b
decomposes at a rate proportional (o the quantity of radiam present,
that it is found that in 25 years approximately 1.1% of a certain quan-
rad how long it will take
thaif the original amount of radium to decompose. Ama. 1600 years
itadioactive substance has a half-life of 38 ht. Find how long it takes
M% of the radinactivity to be dissipated Acs. 120 hv.
{population B is known (o have a rate'of growth proportional to &
ween noon and 2 p.m. the population triples, at what time, no con-
ing exerted, should B become 100 times what it was at noon?
‘Ama. About “
ion of an object through a certain medium (air at certain pressures is
the medium furnishes a resisting force proportional to the square
locity of the moving object. Suppose a body falls, due to the action of
pthrough such a medium. Let ¢ represent time, and v represent velocity,
;nward, Let g be the usual constant acceleration of gravity and Tet w:
of the body. Use Newton's law, force equals mass times aoceler-
that the differential equation of the motion is
wa
@
== ho?“ Elementary Applications Icha
I Solve the differential equation of Exercise 17, with the initial condition ip,
‘=r when ¢ = 0. Introduce the constant o* = w/k to simplify the formu.
19, There are mediuims that resist motion through them with a force propoctioy
to the first power of the velocity. For such a medium, state and solve problems
analogous to Exercises 17 and 18, except that for convenience 1 convton,
~ w/k may be introduced to replace the a* of Exercise 18. Show that b huss
dimensions of a velocity
Ae = b+ (0 ~d) exp (
20, Figure 9 shows a weight, w pounds (Ib), sliding down an inclined plane that
makes an angle a with the horizontal. Assume that no force other than gravis
is acting on the weight; that is, there is no friction, no air resistance, and so
At time ¢ = 0, let x = xg and let the initial velocity be vo. Determine x for» 0
Ams x= dat sina + 191 + 5,
FIGURE 9
Along. very smooth board is inchned at an angle of 10° with the horizontal §
‘weight starts from rest 10 ft from the bottom of the board and slides down
tinder the action of gravity alone. Find how ong it will take the wight (© r
the bottom of the board and determine the terminal speed
Ans. 1.9 sec and 10.5 tse
22. Add to the conditions of Exercise 1) a retarding force of magnitude kr. he
ie the velocity. Determine » and x under the assumption that the weight sc
(som rest with x = x9. Use the notation a = ha/w
Ams. v =a 'gsinall —¢ “);x=xo +a *gsinaj—1 te “ +o")
23, A man, standing at 0 in Figure 10, holds a rope of length » 19 which a wen!
sttached, initially at Wy. The man walks to ‘he’. sagging the weigh! 3!
him. When the man is at M, the weight is at W Fiad the differential equation
the path (called the Wactis) of the weight and solve the equation,
gay
ete fat
24. A tank contains 80 gallons (gal) of pure water. A brine solution with 2 Ib/ea!
‘salt enters at 2 gal/min, and the well-stirred mixtur- leaves at the same ‘2
Loglatic Growth and the Price of Commodities s
FIGURE 10
lhe amour of salt in the wank at any time and () the time at which the
fang wi contain 1 b/g ofa
i . ‘Ans. (a). 160{1 — exp (—1/40)); (b) ¢ = 40 In 2 min.
Bank in Exercise 26 determine the limiting vale forthe amount of elt
ink after a long time. How much time must pass before the amount of
Ma youches 80% of this lmiting value?
yy ‘Ans. (a) = 160 Ib; (b) ¢ = 64 min,
am of money P draws inveret compounded continuously, tf ats
Bc dilare in the acsount, determine the time when the
re reac vaiac 2Ps dollars the anal interest cates (a) 2% oF ()
‘km. (a) 30m 2 yeas (e) 25 2 yea
afters 5% interent compounded continuously In savings assent
ri fa) the amount of interest earned in | year on deposit of $100 and
fate if the compounding were done annually,
valent rate if the compounding. ae
istic Growth and the Price of Commodities
re been jevelop models to study the growth:
ict ar een mae 0 se te
he average birthrate per individual is a positive constant anid
gage death rate pet individual is proportional to the population, ~
St) represent the population at time f, then the above: assump
the differential equation sj ;- Elementary Applications Ich
where b and a are positive constants. This equation is commonly called the
Togistic equation and the growth of population determined by it is calieg
logistic growth.
"The variables in the logistic equation may be separated to obtain
dx
o-a~*
or
ate.
Grrta)e-o*
Integrating both sides gives us
ee ea
or
lpaal-ee a
To expedite the study of equation (2), let us further assume that at (= 0
the population is the positive number x. Then equation (2) may be »ri't
ee
Bax" b— ax,
and upon solving for x, we have
bxoe™
Ma, range
It is interesting to not
(3) has a limiting value
iat the population function obtained in equation
bxoe
then ih Shea a
ee Bate Horee
~ im BES
e
, ee
where we have used I"Hospital’s rule to evaluate the limit.
‘We should also note that the logistic equation (1) will dictate a growth or
‘s decline in the population depending upon whether the initial population
is Jess than or greater than b/a.
Logistic Growth and the Price of Commodities ”
‘example of an application in which a first-order differential
is, we consider an economic model of a certain commodity
‘assume that the price P. the supply 5, and the demand D of
ity are functions of time and that the rate of change of the
tional to the difference between the demand and the supply.
@
sitive so that the price will
ithe demand exceeds the supply
Bat models of the commodity marke
‘of the demand and netions
swe assume (hat
Dec-dP an s
will result depending upon
hat are indicated. If, for
a+ bP, o
Ie. and d are positive constants, we obtain a differential equation
aP = k{lc - a) — (d+ OPI ©)
in P. The assumptions (5) reflect the tendency for the demand
fas the price increases and the’ tendency for the supply to
the price increases, both reasonable assumptions for many com
We should also assume that 0 < P-