Вы находитесь на странице: 1из 289

Trinity University

Digital Commons @ Trinity


Books and Monographs

2000

Student Solutions Manual for Elementary


Differential Equations and Elementary Differential
Equations with Boundary Value Problems
William F. Trench
Trinity University, wtrench@trinity.edu

Follow this and additional works at: http://digitalcommons.trinity.edu/mono


Part of the Mathematics Commons
Recommended Citation
Trench, William F., "Student Solutions Manual for Elementary Differential Equations and Elementary Differential Equations with
Boundary Value Problems" (2000). Books and Monographs. Book 10.
http://digitalcommons.trinity.edu/mono/10

This Book is brought to you for free and open access by Digital Commons @ Trinity. It has been accepted for inclusion in Books and Monographs by an
authorized administrator of Digital Commons @ Trinity. For more information, please contact jcostanz@trinity.edu.

STUDENT SOLUTIONS MANUAL FOR

ELEMENTARY
DIFFERENTIAL EQUATIONS
AND

ELEMENTARY
DIFFERENTIAL EQUATIONS
WITH BOUNDARY VALUE
PROBLEMS

William F. Trench
Andrew G. Cowles Distinguished Professor Emeritus
Department of Mathematics
Trinity University
San Antonio, Texas, USA
wtrench@trinity.edu
This book has been judged to meet the evaluation criteria set by the Editorial Board of the American Institute of Mathematics in connection with
the Institutes Open Textbook Initiative. It may be copied, modified, redistributed, translated, and built upon subject to the Creative Commons
Attribution-NonCommercial-ShareAlike 3.0 Unported License.

This book was published previously by Brooks/Cole Thomson Learning


Reproduction is permitted for any valid noncommercial educational, mathematical, or scientific purpose.
However, charges for profit beyond reasonable printing costs are prohibited.

TO BEVERLY

Contents
Chapter 1 Introduction

1.2 First Order Equations

Chapter 2 First Order Equations

2.1
2.2
2.3
2.4
2.5
2.6

Linear First Order Equations


Separable Equations
Existence and Uniqueness of Solutions of Nonlinear Equations
Transformation of Nonlinear Equations into Separable Equations
Exact Equations
Integrating Factors

5
8
11
13
17
21

Chapter 3 Numerical Methods

25

3.1 Eulers Method


3.2 The Improved Euler Method and Related Methods

25
29

ii Contents
3.3 The Runge-Kutta Method

34

Chapter 4 Applications of First Order Equations

39

4.1 Growth and Decay


4.2 Cooling and Mixing
4.3 Elementary Mechanics
4.4 Autonomous Second Order Equations
4.5 Applications to Curves
Chapter 5 Linear Second Order Equations

39
40
43
45
46
51

5.1
5.2
5.3
5.4
5.5
5.6
5.7

51
55
58
60
64
75
79

Homogeneous Linear Equations


Constant Coefficient Homogeneous Equations
Nonhomgeneous Linear Equations
The Method of Undetermined Coefficients I
The Method of Undetermined Coefficients II
Reduction of Order
Variation of Parameters

Chapter 6 Applcations of Linear Second Order Equations

85

6.1
6.2
6.3
6.4

85
87
89
90

Spring Problems I
Spring Problems II
The RLC Circuit
Motion Under a Central Force

Chapter 7 Series Solutions of Linear Second Order Equations

108

7.1
7.2
7.3
7.4
7.5
7.6
7.7

91
93
96
102
103
108
118

Review of Power Series


Series Solutions Near an Ordinary Point I
Series Solutions Near an Ordinary Point II
Regular Singular Points; Euler Equations
The Method of Frobenius I
The Method of Frobenius II
The Method of Frobenius III

Chapter 8 Laplace Transforms

125

8.1 Introduction to the Laplace Transform


8.2 The Inverse Laplace Transform
8.3 Solution of Initial Value Problems
8.4 The Unit Step Function
8.5 Constant Coefficient Equations with Piecewise Continuous Forcing
Functions
8.6 Convolution

125
127
134
140
143
152

Contents iii
8.7 Constant Cofficient Equations with Impulses

55

Chapter 9 Linear Higher Order Equations

159

9.1
9.2
9.3
9.4

159
171
175
181

Introduction to Linear Higher Order Equations


Higher Order Constant Coefficient Homogeneous Equations
Undetermined Coefficients for Higher Order Equations
Variation of Parameters for Higher Order Equations

Chapter 10 Linear Systems of Differential Equations

221

10.1
10.2
10.3
10.4
10.5
10.6
10.7

191
192
193
194
201
245
218

Introduction to Systems of Differential Equations


Linear Systems of Differential Equations
Basic Theory of Homogeneous Linear Systems
Constant Coefficient Homogeneous Systems I
Constant Coefficient Homogeneous Systems II
Constant Coefficient Homogeneous Systems II
Variation of Parameters for Nonhomogeneous Linear Systems

Chapter

221

11.1 Eigenvalue Problems for y00 C y D 0


11.2 Fourier Expansions I
11.3 Fourier Expansions II

221
223
229

Chapter 12 Fourier Solutions of Partial Differential Equations

239

12.1
12.2
12.3
12.4

239
247
260
270

The Heat Equation


The Wave Equation
Laplaces Equation in Rectangular Coordinates
Laplaces Equation in Polar Coordinates

Chapter 13 Boundary Value Problems for Second Order Ordinary Differential Equations
13.1 Two-Point Boundary Value Problems
13.2 Sturm-Liouville Problems

273
273
279

CHAPTER 1
Introduction

1.2 BASIC CONCEPTS


1.2.2. (a) If y D ce 2x , then y 0 D 2ce 2x D 2y.
x2
c
2x
c
2x 2 c
x2
c
0
(b) If y D
C , then y 0 D
,
so
xy
C
y
D
C
C D x2.
2
3
x
3
x
3
x
3
x
(c) If
1
2
2
y D C ce x ; then y 0 D 2xce x
2
and


1
2
2
2
2
C ce x D 2xce x C x C 2cxe x D x:
y 0 C 2xy D 2xce x C 2x
2
(d) If
yD

1 C ce
1 ce

x 2 =2
x 2 =2

then
y0

.1

ce

x 2 =2

/. cxe
.1

x 2 =2

/
cxe

.1 C ce
x 2 =2 /2

x 2 =2

/cxe

2cxe x =2
.1 ce x2 =2 /2

D
and
y

D
D

1 C ce
1

ce

.1 C ce

x 2 =2
x 2 =2

x 2 =2 2

/
.1 ce
x
.1 ce 2 =2/2
2

!2

4ce x =2
;
.1 ce x2 =2/2
1

x 2 =2 2

x 2 =2

Chapter 1 Basic Concepts


so

4cx C 4cx
D 0:
.1 ce x2 =2 /2
 3

 3


 3 
x
x
x
C c , then y 0 D x 2 sec2
C c D x 2 1 C tan2
D x 2 .1 C y 2 /.
(e) If y D tan
3
3 x
3
C
c
y D .c1 C c2x/e C sin x C x 2 ; then
(f) If
y 0 D .c1 C 2c2x/e x C cos x C 2x;
2y 0 C x.y 2

y0
2y 0 C y

1/ D

D .c1 C 3c2x/e x sin x C 2;


and
y 00
D c1 e x .1 2 C 1/ C c2 xe x .3 4 C 1/
sin x 2 cos x C sin x C 2 4x C x 2
D
2 cos x C x 2 4x C 2:
2
2
4
and y 00 D c1 e x C 3 , so .1 x/y 00 C xy 0 y D
(g) If y D c1e x C c2x C , then y 0 D c1 e x C c2
2
x
x
x
4.1 x/ 2
2
4.1 x x 2 /
c1 .1 x C x 1/ C c2 .x x/ C
D
x3
x x
x3
c1 sin x C c2 cos x
c
c2 sin x c1 sin x C c2 cos x
1 cos x
0
(h) If y D
C 4x C 8 then y D
C 4 and
1=2
1=2
x
x
2x 3=2

c1 sin x C c2 cos x c1 sin x c2 cos x 3 c1 sin x C c2 cos x
1
2 00
0
2
00
y D
C
, so x y Cxy C x
yD
4
4
x 1=2
x 3=2
x 5=2

3
c1
x 3=2 sin x x 1=2 cos x C x 1=2 sin x C x 1=2 cos x
4


1 1=2
3
1 1=2
3=2
x
sin x C x sin x
x
sin x C c2
x 3=2 cos x C x 1=2 sin x C x 1=2 cos x
2
4


 4
1 1=2
1 1=2
1
1=2
3=2
2
x sin x
x
cos x C x cos x
x
cos x C 4x C x
.4x C 8/ D 4x 3 C 8x 2 C
2
4
4
3x 2.
R
1.2.4. (a) If y 0 D xe x , then y D xe x C e x dx C c D .1 x/e x C c, and y.0/ D 1 ) 1 D 1 C c,
so c D 0 and y D .1 x/e x .
r 
1

2
0
2
(b) If y D x sin x , then y D
cos x C c; y
D 1 ) 1 D 0 C c, so c D 1 and
2
2
1
yD1
cos x 2 .
2
1 d
sin x
(c) Write y 0 D tan x D
D
.cos x/. Integrating this yields y D ln j cos xj C c;
cos x
cos x dx p
p
y.=4/pD 3 ) 3 Dp ln .cos.=4// C c, or 3 D ln 2 C c, so c D 3 ln 2, so y D ln.j cos xj/ C
3 ln 2 D 3 ln. 2j cos xj/.
x5
32
37
(d) If y 00 D x 4 , then y 0 D
C c1 ; y 0 .2/ D 1 )
C c1 D 1 ) c1 D
, so y 0 D
5
5
15
x5
37
x6
37
64
47
. Therefore, y D
.x 2/ C c2; y.2/ D 1 )
C c2 D 1 ) c2 D
, so
5
15
30
15
30
15
6
47 37
x
yD
.x 2/ C
.
15
5
30
Z
R
xe 2x
1
xe 2x
e 2x
xe 2x
e 2x
(e) (A) xe 2x dx D
e 2x dx D
. Therefore, y 0 D
C c1 ;
2
2
2
4
2
4
1
5
5
xe 2x
e 2x
5
xe 2x
y 0 .0/ D 1 )
C c1 D ) c1 D , so y 0 D
C ; Using (A) again, y D
4
4
4
2
4
4
4
e 2x
e 2x
5
xe 2x
e 2x
5
1
29
C x C c2 D
C x C c2 ; y.0/ D 7 )
C c2 D 7 ) c2 D
, so
8
8
4
4
4
4
4
4
2x
2x
xe
e
5
29
yD
C xC .
4 R
4
4
4
R
R
(B) x cos x dx D x sin x
R (f) (A) x sin x dx D x cos x00 C cos x dx D x cos x C sin x and
sin x dx D x sin x C cos x. If y D x sin x, then (A) implies that y 0 D x cos x sin x C c1 ; y 0 .0/ D
3 ) c D 3, so y 0 D x cos x sin x 3. Now (B) implies that y D x sin x Ccos x Ccos x 3x Cc2 D
x sin x C 2 cos x 3x C c2 ; y.0/ D 1 ) 2 C c2 D 1 ) c2 D 1, so y D x sin x C 2 cos x 3x 1.

Section 1.2 Basic Concepts

R
R
(g) If y 000 D x 2 e x , then y 00 D x 2 e x dx D x 2 e x 2 xe x dx D x 2 eRx 2xe x C 2e x C c1 ;
y 00 .0/ D 3 ) 2 C c1RD 3 ) c1 D 1, so (A) y 00 D .x 2 2x C 2/e x C 1. Since .x 2 2x C 2/e x dx D
.x 2 2x C 2/e x
.2x 2/e x dx D .x 2 2x C 2/e x .2x 2/e x C 2e x D .x 2 4x C 6/e x ,
0
x
0
(A) implies that y D .x 2 4x C 6/e
R 2C x C c2 ; yx .0/ D 22 ) 6 C c2xD R 2 ) c2 xD 8, so (B)
0
2
x
y D .x 4x C 6/e C x 8; Since .x 4x C 6/e dx D .x 4x C 6/e
.2x 4/e dx D .x 2
x2
8x Cc3;
4x C6/e x .2x 4/e x C2e x D .x 2 6x C12/e x , (B) implies that y D .x 2 6x C12/e x C
2
2
x
y.0/ D 1 ) 12 C c3 D 1 ) c3 D 11, so y D .x 2 6x C 12/e x C
8x 11.
2
cos 2x
1
7
C c1 ; y 00 .0/ D 3 )
C c1 D 3 ) c1 D ,
(h) If y 000 D 2 C sin 2x, then y 00 D 2x
2
2
2
cos 2x
7
sin 2x
7
00
0
2
0
so y D 2x
C . Then y D x
C x C c2 ; y .0/ D 6 ) c2 D 6, so
2
2
4
2
3
7
x
cos 2x 7 2
1
7
0
2 sin 2x
C x 6. Then y D
C
C x 6xCc3 ; y.0/ D 1 ) Cc3 D 1 ) c3 D ,
y Dx
4
2
3
8
4
8
8
x3
cos 2x
7 2
7
so y D
C
C x
6x C .
3
8
4
8
(i) If y 000 D 2x C 1, then y 00 D x 2 C x C c1; y 00 .2/ D 7 ) 6 C c1 D 7 ) c1 D 1; so y 00 D x 2 C x C 1.
x3 x2
14
26
x3 x2
Then y 0 D
C
C .x 2/ C c2; y 0 .2/ D 4 )
C c2 D 4 ) c2 D
, so y 0 D
C
C
3
2
3
3
3
2
4
3
x
x
1
26
8
5
26
.x 2/
. Then y D
C
C .x 2/2
.x 2/ C c3 ; y.2/ D 1 ) C c3 D 1 ) c3 D
,
3
12
6
2
3
3
3
4
3
x
1
26
5
x
so y D
C
C .x 2/2
.x 2/
.
12
6
2
3
3
1.2.6. (a) If y D x 2 .1 C ln x/, then y.e/ D e 2 .1 C ln e/ D 2e 2; y 0 D 2x.1 C ln x/ C x D 3x C 2x ln x,
so y 0 .e/ D 3e C 2e ln e D 5e; (A) y 00 D 3 C 2 C 2 ln x D 5 C 2 ln x. Now, 3xy 0 4y D 3x.3x C
2x ln x/ 4x 2.1 C ln x/ D 5x 2 C 2x 2 ln x D x 2 y 00 , from (A).
x2
1
1
2
2
5
(b) If y D
C x 1, then y.1/ D C 1 1 D ; y 0 D x C 1, so y 0 .1/ D C 1 D ; (A)
3
3
3
3
3
3

2
x2
2
2
00
2
0
2
y D . Now x
xy C y C 1 D x
x
xC1 C
C x 1 C 1 D x 2 D x 2 y 00 , from (A).
3
3
3
3
(c) If y D .1 C x 2 / 1=2 , then y.0/ D .1 C 02 / 1=2 D 1; y 0 D x.1 C x 2 / 3=2 , so y 0 .0/ D 0; (A)
00
y D .2x 2 1/.1Cx 2 / 5=2 . Now, .x 2 1/y x.x 2 C1/y 0 D .x 2 1/.1Cx 2 / 1=2 x.x 2 C1/. x/.1C
.x 2 1/y x.x 2 C 1/y 0
x 2 / 3=2 D .2x 2 1/.1 C x 2 / 1=2 D y 00 .1 C x 2 /2 from (A), so y 00 D
.
.x 2 C 1/2
x2
1=4
1
x.x 2/
. 1=2/. 3=2/
(d) If y D
, then y.1=2/ D
D ; y0 D
, so y 0 .1=2/ D
D 3;
1 x
1 1=2
2
.1 x/2
.1 1=2/2
2
x2
x
x 2 .x 2/
x2
0
(A) y 00 D
.
Now,
(B)
x
C
y
D
x
C
D
and
(C)
xy
y
D
D
.1 x/3
1 x
1 x
.1 x/2
1 x
2
3
3
0
x
x
x 00
2.x C y/.xy
y/
. From (B) and (C), .x C y/.xy 0 y/ D
D
y , so y 00 D
.
2
3
3
.1 x/
.1 x/
2
x
1.2.8. (a) y D .x c/a is defined and x c D y 1=a on .c; 1/; moreover, y 0 D a.x
a 1
a y 1=a
D ay .a 1/=a .
(b) if a > 1 or a < 0, then y  0 is a solution of (B) on . 1; 1/.

c/a

1.2.10. (a) Since y 0 D c we must show that the right side of (B) reduces to c for all values of x in some

Chapter 1 Basic Concepts

interval. If y D c 2 C cx C 2c C 1,
x 2 C 4x C 4y

D x 2 C 4x C 4c 2 C 4cx C 8c C 4
D x 2 C 4.1 C c/x C 4.c 2 C 2c C 1/
D x 2 C 4.1 C c/ C 2.c C 1/2 D .x C 2c C 2/2 :

x 2 C 4x C 4y D x C 2c C 2 and the right side of (B) reduces to c if x > 2c 2.


x.x C 4/
xC2
(b) If y1 D
, then y10 D
and x 2 C 4x C 4y D 0 for all x. Therefore, y1 satisfies
4
2
(A) on . 1; 1/.
Therefore,

CHAPTER 2
First Order Equations

2.1 LINEAR FIRST ORDER EQUATIONS


y0
2.1.2.
D 3x 2; j ln jyj D x 3 C k; y D ce
y
2.1.4.

y0
D
y

2.1.6.

y0
D
y

yD

3
;
x

ln jyj D 3 ln jxj C k D

1Cx
D
x

1
x

ln jxj3 C k;

1; j ln jyj D

ln jxj

.ln x/2 =2

yD

y0
D
y
y D 3x k .
2.1.10.

y10
D
y1

k
; j ln jyj D
x

3; ln jy1 j D

e 3x
1
C c; y D C ce
3
3
y10
D
y1

3x

ln jxj

c
.
x3

x C k; y D

ln j sin xj C k D

k ln jxj C k1 D ln jx

3x; y1 D e

3x

ce x
; y.1/ D 1 ) c D e;
x

ln jx sin xj C k; y D

j C k1 ; y D cjxj

; y D ue

3x

; u0 e

3x

c
;
x sin x

y.1/ D 3 ) c D 3;

D 1; u0 D e 3x ; u D

2x; ln jy1 j D x 2 ; y1 D e
 2

x
x2
2
uD
C c; y D e x
Cc .
2
2

x2

y10
D
y1

; y D ue

1
1
u
; ln jy1 j D ln jxj; y1 D ; y D ;
x
x
x
3x 2
7 ln jxj 3x
c
u D 7 ln jxj C
C c; y D
C
C .
2
x
2
x
2.1.16.

1
y0
D
cot x; j ln jyj D
y
x

y.=2/ D 2 ) c D  ; y D
.
x sin x

2.1.14.

. y D ce

.x 1/

2.1.8.

2.1.12.

x3

x2

; u0 e

x2

D xe

x2

; u0 D x;

u0
7
7
D 2 C 3; u0 D
C 3x;
x
x
x

Chapter 2 First Order Equations


2

y10
D
y1

1
e x
ue x
2x; ln jy1 j D ln jxj x 2 ; y1 D
; yD
;
x
x
x


x4
x3
c
2
C c; y D e x
C
.
u0 D x 3 ; u D
4
4
x

2.1.18.

u0 e x
D x2e
x

x2

y10
D tan x; ln jy1 j D ln j cos xj; y1 D cos x; y D u cos x; u0 cos x D cos x; u0 D 1;
y1
u D x C c; y D .x C c/ cos x.

.x 2/5
y10
4x 3
5
1

;
D
D
; ln jy1 j D 5 ln jx 2j ln jx 1j D ln
2.1.22.
y1
.x 2/.x 1/
x 2 x 1
x 1
.x 2/5
u.x 2/5
u0 .x 2/5
.x 2/2
1
1
1
; yD
;
D
; u0 D
; uD
C
y1 D
3
x 1
x 1
x 1
x 1
.x 2/
2 .x 2/2
1 .x 2/3
.x 2/5
c; y D
Cc
.
2 .x 1/
.x 1/
2.1.20.

2.1.24.

y10
D
y1

u D xe x

3
; ln jy1 j D 3 ln jxj D ln jxj 3;
x
ex
ex
c
e x C c; y D 2
C 3.
3
x
x
x

y1 D

1
u
; y D 3;
3
x
x

u0
ex
D
; u0 D xe x ;
x3
x2

y10
D
y1

4x
1
; ln jy1 j D 2 ln.1 C x 2 / D ln.1 C x 2 / 2 ; y1 D
; y D
1 C x2
.1 C x 2 /2
0
u
2
2x C c
u
;
D
; u0 D 2; u D 2x C c; y D
; y.0/ D 1 )
2
2
2
2
2
2
.1 C x /
.1 C x /
.1 C x /
.1 C x 2 /2
2x C 1
c D 1; y D
.
.1 C x 2 /2
2.1.26.

2.1.28.

y10
D
y1

2.1.30.

y10
D
y1

u0
D cos x; u0 D
sin x
sin x
sin2 x
1
sin x cos x; u D
C c; y D
C c csc x; y.=2/ D 1 ) c D 21 ; y D .sin x C csc x/.
2
2
2
cot x; ln jy1 j D

; ln jy1 j D

ln j sin xj; y1 D

3 ln jx

1
u
; y D
;
sin x
sin x

1j D ln jx

1j

; y1 D

1
1/3

x 1
.x
u0
1
sin x
1
0
D
C
; u D
C sin x; u D ln jx 1j
.x 1/3
.x 1/4
.x 1/3
x 1
ln jx 1j cos x C c
ln jx 1j cos x
; y.0/ D 1 ) c D 0; y D
.
.x 1/3
.x 1/3

2.1.32.
uD

; ln jy1 j D

3 ln jx

1j D ln jx

1j

; y1 D

u
1/3

.x

cos x C c; y D

y10
2
D
; ln jy1 j D 2 ln jxj D ln.x 2 /; y1 D x 2 ; y D ux 2 ; u0 x 2 D
y1
x
ln jxj C c; y D x 2 .c ln jxj/; y.1/ D 1 ) c D 1; y D x 2 .1 ln x/.

y10
3
D
y1
x
u0
1 C .x
D
.x 1/3
.x

2.1.34.

; y D

x; u0 D

; y D

1
;
x

;
1
.x
.x 1/3
1/ sec2 x
1
ln jx 1j C tan x C c
; u0 D
Csec2 x; u D ln jx 1jCtan xCc; y D
;
1/4
x 1
.x 1/3
ln jx 1j C tan x C 1
y.0/ D 1 ) c D 1; y D
.
.x 1/3
1/3

Section 2.1 Linear First Order Equations

y10
2x
D 2
; ln jy1 j D ln jx 2 1j; y1 D x 2 1; y D u.x 2 1/; u0 .x 2 1/ D x;
y1
x
1


x
1
1
2
2
0
2
; u D ln jx
1j C c; y D .x
1/
ln jx
1j C c ; y.0/ D 4 ) c D 4;
u D 2
x
1 
2
2

1
y D .x 2 1/
ln jx 2 1j 4 .
2
2.1.36.

y0
2
2
2
2
2.1.38. 1 D 2x; ln jy1 j D x 2 ; y1 D e x ; y D ue x ; u0 e x D x 2 ; u0 D x 2 e x ; u D
y
1




Z x
Z x
Z x
2
2
2
2
2
cC
t 2 e t dt; y D e x c C
t 2 e t dt ; y.0/ D 3 ) c D 3; y D e x 3 C
t 2 e t dt .
0

y10
e x tan x
tan x
D 1; ln jy1 j D x; y1 D e x ; y D ue x ; u0 e x D
; u0 D
;
yZ1
x
x


Z x
Z x
x
tan t
tan t
tan t
uDcC
dt ; y D e x c C
dt ; y.1/ D 0 ) c D 0; y D e x
dt.
t
t
t
1
1
1

2.1.40.

y10
D
y1

1
e x
ue x
u0 e x
ex
; ln jy1 j D x
ln jxj; y1 D
; y D
;
D
;
x Z
x
x
x

Z xx
x
x 
e
2
2
2
u0 D e x e x ; u D c C
e t e t dt; y D
cC
e t e t dt ; y.1/ D 2 ) c D 2e;
x
1


Z x1
1
.x 1/
x
t t2
yD
2e
Ce
e e dt .
x
1
2.1.42.

2.1.44. (b) Eqn. (A) is equivalent to


2
D
x

y0

1
x

.B/

y10
2
1
D ; ln jy1 j D 2 ln jxj; y1 D x 2 ; y D ux 2 ; u0 x 2 D
;
y1
x
x
1
1
1
u0 D
; uD
C c, so y D C cx 2 is the general solution of (A) on . 1; 0/ and .0; 1/.
3
2
x
2x
2
(c) From the proof of (b), any solution of (A) must be of the form
8
1

< C c1x 2 ; x  0;
2
yD
.C/

: 1 C c2x 2 ; x < 0;
2

on . 1; 0/ and .0; 1/. Here

for x 0, and any function of the form (C) satisfies (A) for x 0. To complete the proof we must show
that any function of the form (C) is differentiable and satisfies (A) at x D 0. By definition,
y 0 .0/ D lim

x!0

if the limit exists. But

y.x/

y.x/
x
1=2

y.0/
y.x/ 1=2
D lim
x!0
0
x
D

c1 x; x > 0
c2 x; x < 0;

so y 0 .0/ D 0. Since 0y 0 .0/ 2y.0/ D 0  0 2.1=2/ D 1, any function of the form (C) satisfies (A) at
x D 0.
(d) From (b) any solution y of (A) on . 1; 1/ is of the form (C), so y.0/ D 1=2.

Chapter 2 First Order Equations

1=2
and c2 arbitrary is a solution
x02
of the initial value problem on . 1; 1/. Since these functions are all identical on .0; 1/, this does not
contradict Theorem 2.1.1, which implies that (B) (so (A)) has exactly one solution on .0; 1/ such that
y.x0 / D y0 . A similar argument applies if x0 < 0.
(e) If x0 > 0, then every function of the form (C) with c1 D

y0

2.1.46. (a) Let y D c1 y1 C c2 y2 . Then


y 0 C p.x/y

.c1 y1 C c2 y2 /0 C p.x/.c1 y1 C c2 y2 /

c1y10 C c2y20 C c1p.x/y1 C c2 p.x/y2

c1.y10 C p.x/y1 / C c2 .y2 C p.x/y2 / D c1f1 .x/ C c2f2 .x/:

(b) Let f1 D f2 D f and c1 D c2 D 1.


(c) Let f1 D f , f2 D 0, and c1 D c2 D 1.
3x
0 3x
2.1.48. (a) If D tan y, then 0 D .sec2 y/y 0 , so 0 3 D 1; 1 D e 3x ;
 D ue ; u e D 1;
3x
e
1
1
u0 D e 3x ; u D
C c; D C ce 3x D tan y; y D tan 1
C ce 3x .
3
3
3
1
1
u u0
1
2
2
2
(b) If D e y , then 0 D 2yy 0 e y , so 0 C D 2 ; 1 D 2 ; D 2 ; 2 D 2 ; u0 D 1;
x
x
x
x
x
x
 
1=2
c
1
c
1
y2
u D x C c; D C 2 D e ; y D ln
C 2
.
x
x
x
x
0
y
2
y0
2
1
(c) Rewrite the equation as C ln y D 4x. If D ln y, then 0 D , so 0 C D 4x; 1 D 2 ;
y
x
y
x
x

c
c 
u u0
2
0
3
4
2
D 2 ; 2 D 4x; u D 4x ; u D x C c; D x C 2 D ln y; y D exp x C 2 .
x
x
x
x
1
y0
1
3
1
u u0
3
0
0
(d) If D
, then D
,
so

D
;

D
;

D
;
D
;
1
1Cy
.1 C y/2
x
x2
x
x x
x2
3
3 ln jxj C c
1
x
u0 D
; u D 3 ln jxj c; D
D
; y D 1C
.
x
x
1Cy
3 ln jxj C c

2.2 SEPARABLE EQUATIONS


By
2.2.2.

 inspection, y  k (k Dinteger) is a constant solution. Separate variables to find others:
cos y 0
y D sin x; ln.j sin yj/ D cos x C c.
sin y


ln y
.ln y/2
2.2.4. y  0 is a constant solution. Separate variables to find others:
y0 D x2;
D
y
2
x3
C c.
3
1
2.2.6. y  1 and y  1 are constant solutions. For others, separate variables: .y 2 1/ 3=2 yy 0 D 2 ;
x





2
1
1 C cx
x
x
2
1=2
2
1=2
2
.y
1/
D
c D
; .y
1/
D
; .y
1/ D
;
x
x
1 C cx
1 C cx

2

2 !1=2
x
x
2
y D1C
; y D 1C
.
1 C cx
1 C cx

Section 2.2 Separable Equations


2.2.8. By inspection, y  0 is a constant solution. Separate variables to find others:
ln jyj D

y0
D
y

c
1
ln.1 C x 2 / C k; y D p
, which includes the constant solution y  0.
2
1 C x2

2.2.10. .y 1/2 y 0 D 2xC3;

1/3

.y
3

x
;
1 C x2

D x 2 C3xCc; .y 1/3 D 3x 2 C9xCc; y D 1 C 3x 2 C 9x C c/1=3 .



2
y
y0
1
1
0

D x C k; y D ce x2=2 ; y.2/ D
D x;
y D x; ln
2.2.12.
y.y C 1/
y yC1
y C 1
2
yC1
x 2 =2
2
ce
e
e2
2
2
2
; y D .y C 1/ce x =2 ; y.1 ce x =2/ D ce x =2 ; y D
1)cD
;
setting
c
D
2
2
2
1 ce x =2
.x 2 4/=2
e
yields y D
.
2
2 e .x 4/=2




1
1 1
1 1
1 1
1
1
3
2
y0
2.2.14.
D
;
C
y0 D
;
C
y0 D
.y C 1/.y 1/.y 2/
xC1
6yC1 2y 1
3y 2
xC1
yC1 y 1
y 2
6
.y C 1/.y 2/2
c
; ln jy C 1j 3 ln jy 1j C 2 ln jy 2j D 6 ln jx C 1j C k;
D
;
xC1
.y 1/3
.x C 1/6
.y C 1/.y 2/2
256
y.1/ D 0 ) c D 256;
D
.
3
.y 1/
.x C 1/6
!


y0
1
y
jyj
y
2
0
2.2.16.
D 2x;
y D 2x; ln p
D x 2 C k; p
D ce x ;
2
2
2
2
y.1 C y /
y y C1
y C1
y C1
2

ex
1
y
1
2
2
2
y.0/ D 1 ) c D p ; p
D p ; 2y 2 D .y 2 C1/e x ; y 2 .2 e x / D e 2x ; y D p
.
2
2x
2
2
y C1
2e 2 1



y 2
y0
1
1
D x 2 C k; y 2 D ce x2 ;
2.2.18.
D 2x;
y 0 D 2x; ln
.y 1/.y 2/
y 2 y 1
y 1 !
y 1
2
x2
x2
x2
x2
e
e
e
e
4 e x
1 y 2
.
y.0/ D 3 ) c D ;
D
;y 2 D
.y 1/; y 1
D2
; yD
2 y 1
2
2
2
2
2 e x2
The interval of validity is . 1; 1/.





y 2
y0
1
1
1 0
1
1 0
D 2x C k;
2.2.20.
D 1;
y D 1;
y D 2; ln
y.y 2/
2 y 2 y
y 2 y
y
y 2
y 2
D ce 2x ; y.0/ D 1 ) c D 1;
D e 2x ; y 2 D ye 2x ; y.1 C e 2x / D 2;
y
y
2
yD
. The interval of validity is . 1; 1/.
1 C e 2x

2.2.22. y  2 is a constant solution of the differential equation, and it satisfies the initial condition.
Therefore, y  2 is a solution of the initial value problem. The interval of validity is . 1; 1/.
y0
1
D
; tan 1 y D tan
2
1Cy
1 C x2
tan A C tan B
tan.A C B/ D
with A D tan
1 tan A tan B
c D tan k.
2.2.24.

x C k; y D tan.tan

x and B D tan

x C k/. Now use the identity

c to rewrite y as y D

xCc
, where
1 cx

10 Chapter 2 First Order Equations



) c D 0, so (A) cos y D sin x. To obtain
2
y explicity we note that sin x D cos.x C =2/, so (A) can be rewritten as cos y D cos.x C =2/. This
equation holds if an only if one of the following conditions holds for some integer k:
2.2.26. .sin y/y 0 D cos x;

cos y D sin x C c; y. / D

(B) y D x C


C 2k I mbox.C / y D
2


C 2k:
2

Among these choices the only way to satisfy the initial condition is to let k D 1 in (C), so y D
2.2.28. Rewrite the equation as P 0 D

xC

3
:
2

1=/. By inspection, P  0 and P  1= are


P0
1
1
constant solutions. Separate variables to find others:
D a;
P 0 D a;
P
.P
1=/
P
1=
P

P 1=
1
D at C k; (A) P 1= D ce t ; P .1 ce t / D 1=; (B) P D
.
ln

P
P
.1 ce t /
P0 1=
P0
From (A), P .0/ D P0 ) c D
. Substituting this into (B) yields P D
.
P0
P0 C .1 P0/e at
From this limt !1 P .t/ D 1=.
aP .P

2.2.30. If q D rS the equation for I reduces to I 0 D

rI 2 , so

I0
D
I2

r;

1
D
I

rt

1
; so
I0

I0
and limt !1 I.t/ D 0. If q rS , then rewrite the equation for I as I 0 D rI.I /
1 C rI0 t



I
q
I0
1
1 0

D
with D S
. Separating variables yields
D r;
I D r ; ln
r
I.I /
I I
I
I

r t C k; (A)
D ce r t ; I.1 ce r t / D ; (B) I D
. From (A), I.0/ D I0 )
I
1 ce r t
I0
I0
c D
. Substituting this into (B) yields I D
. If q < rS , then > 0 and
I0
I0 C . I0 /e r t
q
limt !1 I.t/ D D S
. If q > rS , then < 0 and limt !1 I.t/ D 0.
r
I D

2.2.34. The given equation is separable if f D ap, where a is a constant. In this case the equation is
y 0 C p.x/y D ap.x/:

.A/

Let P be an antiderivative of p; that is, P 0 D p.


S OLUTION

BY

S EPARATION

OF

VARIABLES . y 0 D

p.x/.y

a/;

y0
y

p.x/; ln jy

aj D

P .x/ C k; y a D ce P.x/ ; y D a C ce P.x/ .


S OLUTION BY VARIATION OF PARAMETERS . y1 D e P.x/ is a solution of the complementary
equation, so solutions of (A) are of the form y D ue P.x/ where u0 e P.x/ D ap.x/. Hence, u0 D
ap.x/e P.x/ ; u D ae P.x/ C c; y D a C ce P.x/ .
2
x5
2
yD
. y1 D x 2 is a solution of y 0
y D 0.
2
x
yCx
x
x5
x3
x
Look for solutions of (A) of the form y D ux 2 . Then u0 x 2 D
D
; u0 D
;
2
.u C 1/x
uC1
uC1


p
p
.1 C u/2
x2
c
.u C 1/u0 D x;
D
C ; u D 1 x 2 C c; y D x 2 1 x 2 C c .
2
2
2
2.2.36. Rewrite the given equation as (A) y 0

Section 2.3 Existence and Uniqueness of Solutions of Nonlinear Equations

11

2.2.38. y1 D e 2x is a solution of y 0 2y D 0. Look for solutions of the nonlinear equation of the


xe 2x
x
.1 u/2
1
form y D ue 2x . Then u0 e 2x D
; u0 D
; .1 u/u0 D x;
D .x 2 c/;
1 u 
1 u
2
2

p
p
u D 1 c x 2 ; y D e 2x 1 c x 2 .
2.3 EXISTENCE AND UNIQUENESS OF SOLUTIONS OF NONLINEAR EQUATIONS

ex C y
1
2y.e x C y/
and fy .x; y/ D 2
are both continuous at all .x; y/
2
2
2
x Cy
x Cy
.x 2 C y 2 /2
.0; 0/. Hence, Theorem 2.3.1 implies that if .x0 ; y0 / .0; 0/, then the initial value problem has a a
unique solution on some open interval containing x0. Theorem 2.3.1 does not apply if .x0 ; y0 / D .0; 0/.
2.3.2. f .x; y/ D

x2 C y2
2y
x2 C y2
and fy .x; y/ D
are both continuous at all .x; y/ such
ln xy
ln xy
x.ln xy/2
that xy > 0 and xy 1. Hence, Theorem 2.3.1 implies that if x0 y0 > 0 and x0 y0 1, then the initial
value problem has unique solution on an open interval containing x0 . Theorem 2.3.1 does not apply if
x0 y0  0 or x0 y0 D 1.
2.3.4. f .x; y/ D

2.3.6. f .x; y/ D 2xy and fy .x; y/ D 2x are both continuous at all .x; y/. Hence, Theorem 2.3.1
implies that if .x0 ; y0 / is arbitrary, then the initial value problem has a unique solution on some open
interval containing x0 .
3
2x C 3y
2x C 3y
and fy .x; y/ D
C4
are both continuous at all .x; y/ such
x 4y
x 4y
.x 4y/2
that x 4y. Hence, Theorem 2.3.1 implies that if x0 4y0 , then the initial value problem has a unique
solution on some open interval containing x0 . Theorem 2.3.1 does not apply if x0 D 4y0 .
2.3.8. f .x; y/ D

4
xy.y 2 1/1=3 is continuous
3
at .x; y/ if and only if y 1. Hence, Theorem 2.3.1 implies that if y0 1, then the initial value
problem has a unique solution on some open interval containing x0, while if y0 D 1, then the initial
value problem has at least one solution (possibly not unique on any open interval containing x0 ).

2.3.10. f .x; y/ D x.y 2

1/2=3 is continuous at all .x; y/, but fy .x; y/ D

1
are both continuous at all .x; y/ such that
2.x C y/1=2
x C y > 0 Hence, Theorem 2.3.1 implies that if x0 C y0 > 0, then the initial value problem has a unique
solution on some open interval containing x0 . Theorem 2.3.1 does not apply if x0 C y0  0.
2.3.12. f .x; y/ D .x C y/1=2 and fy .x; y/ D

2.3.14. To apply Theorem 2.3.1, rewrite the given initial value problem as (A) y 0 D f .x; y/; y.x0 / D y0 ,
where f .x; y/ D p.x/y C q.x/ and fy .x; y/ D p.x/. If p and f are continuous on some open
interval .a; b/ containing x0 , then f and fy are continuous on some open rectangle containing .x0 ; y0 /,
so Theorem 2.3.1 implies that (A) has a unique solution on some open interval containing x0. The
conclusion of Theorem 2.1.2 is more specific: the solution of (A) exists and is unique on .a; b/. For
example, in the extreme case where .a; b/ D . 1; 1/, Theorem 2.3.1 still implies only existence and
uniqueness on some open interval containing x0 , while Theorem 2.1.2 implies that the solution exists and
is unique on . 1; 1/.
2.3.16. First find solutions of (A) y 0 D y 2=5 . Obviously y  0 is a solution. If y 6 0, then we
5
can separate variables on any open interval where y has no zeros: y 2=5 y 0 D 1; y 3=5 D x C c;
3


3
5=3
y D
.x C c/
. (Note that this solution is also defined at x D c, even though y. c/ D 0.
5

12 Chapter 2 First Order Equations


To satisfy the initial condition, let c D 1. Thus, y D

3
.x C 1/5=3
5

is a solution of the initial value

2 3=5
y
are both continuous at
5
all .x; y/ such that y 0, this is the only solution on . 5=3; 1/, by an argument similar to that given in
Example 2.3.7, the function
(
0;
1 < x  35
yD
5=3
3
5
xC1
;
<x <1
5
3

problem on . 1; 1/; moreover, since f .x; y/ D y 2=5 and fy .x; y/ D

5
(To see that y satisfies y 0 D y 2=5 at x D
use an argument similar to that of Discussion 2.3.15-2) For
3
5
every a  , the following function is also a solution:
3
8
5=3
3

;
1 < x < a;

< 5 .x C a/
0;
a  x  53
yD

5=3
: 3
5
;
5x C 1
3 < x < 1:

2.3.18. Obviously, y1  1 is a solution. From Discussion 2.3.18 (taking c D 0 in the two families of
solutions) yields y2 D 1 C jxj3 and y3 D 1 jxj3 . Other solutions are y4 D 1 C x 3, y5 D 1 x 3 ,


1 C x 3 ; x  0;
1 x 3 ; x  0;
y6 D
I y7 D
I
1;
x<0
1;
x<0


1;
x  0;
1;
x  0;
I
y
D
y8 D
9
1 C x3; x < 0
1 x3; x < 0

It is straightforward to verify that all these functions satisfy y 0 D 3x.y 1/1=3 for all x 0. Moreover,
yi .x/ 1
yi0 .0/ D lim
D 0 for 1  i  9, which implies that they also satisfy the equation at x D 0.
x!0
x
2.3.20. Let y be any solution of (A) y 0 D 3x.y 1/1=3 ; y.3/ D 7. By continuity, there is some open
interval I containing x0 D 3 on which y.x/ < 1. From Discussion 2.3.18, y D 1 C .x 2 C c/3=2 on I ;
y.3/ D 7 ) c D 5; (B) y Dp1 .x 2 5/3=2. It now follows that every solution of
p(A) satisfies
y.x/ < 1 and is given by (B) on . 5; 1/; that is, (B) is the unique solution of (A) on . 5; 1/. This
solution can be extended uniquely to .0; 1/ as
p

1;
0p< x  5;
yD
1 .x 2 5/3=2 ;
5<x <1
It can be extended to . 1; 1/ in infinitely many ways. Thus,
p

1;
1 < x  5;
p
yD
1 .x 2 5/3=2 ;
5<x<1
is a solution of the initial value problem on . 1; 1/. Moroever, if  0, then
8
1 < x <p ;
< 1 C .x 2 2 /3=2 ;
yD
1;

p  x  5;
:
1 .x 2 5/3=2 ;
5 < x < 1;

Section 2.4 Transformation of Nonlinear Equations into Separable Equations


and
yD

8
< 1
:

.x 2

2/3=2 ;
1;

.x 2

5/3=2 ;

13

1 < x <p ;

p  x  5;
5 < x < 1;

are also solutions of the initial value problem on . 1; 1/.

2.4 TRANSFORMATION OF NONLINEAR EQUATIONS INTO SEPARABLE EQUATIONS


2
x
y10
2
2
y D
.
Then
D
; ln jy1 j D ln jxj D ln jxj2=7 ; y1 D x 2=7 ;
2.4.2. Rewrite as y 0
7x
7y 6
y1
7x
7
1
1
u7
1
c
6 0
y D ux 2=7 ; u0 x 2=7 D
;
u
u
D
;
D
ln jxj C ; u D .c ln jxj/1=7 ;
6
5=7
7x
7
7
7
7u x
y D x 2=7 .c ln jxj/1=7 .

2x
1
y10
2x
y
D
.
Then
D
; ln jy1 j D ln.1 C x 2 /;
1 C x2
.1 C x 2 /2 y
y1
1 C x2
p
1
u
u0
1
u2
c
0
y1 D
;
y
D
;
D
;
u
u
D
1;
D x C ; u D 2x C c;
2
2
2
2
1pC x
1Cx
1Cx
u.1 C x /
2
2
2x C c
yD
.
1 C x2


y0
1 1
1
2.4.6. 1 D
C 1 ; ln jy1 j D .ln jxj C x/; y1 D x 1=3 e x=3 ; y D ux 1=3 e x=3 ; u0 x 1=3 e x=3 D
y1
3 x
3

1=3
0
1
1
x
u
x
x c
4=3 4x=3 4
; yD
x e
u ; 4 D xe ;
D .x 1/e
; uD
.
u
3u3
3
3.1 x/ C ce x
3.1 x/e x C c1=3
2.4.4. Rewrite as y 0 C

x2
u0
y10
2
2
2
2
2
D x; ln jy1 j D
; y1 D e x =2; y D ue x =2 ; u0 e x =2 D xu3=2 e 3x =4; 3=2 D xe x =4 ;
y1
2
u
1
1
1
2
2
; uD
; y D
. Because
(A)
D 2e x =4 C 2c; u1=2 D
u1=2
c C e x2 =4
.c C e x2 =4 /2
.1 C ce x2 =4 /2
of (A) we must choose c so that y.1/ D 4 and 1 C ce 1=4 < 0. This implies that c D 3e 1=4;


3 .x2 1/=4 2
yD 1
e
.
2
2.4.8.

y10
D 2; ln jy1 j D 2x; y1 D e 2x ; y D ue 2x ; u0 e 2x D 2u1=2e x ; u 1=2 u0 D 2e x ;
y1
2u1=2 D 2e x C 2c; u1=2 D c e x > 0; y.0/ D 1 ) u.0/ D 1 ) c D 2; u D .2 e x /2 ;
y D .2e x 1/2 .
2.4.10.

y0
2
y3
2
1
y D 2 . Then 1 D
; ln jy1 j D 2 ln jxj D ln x 2 ; y1 D 2 ;
x
x
y1
x
x
u u0
u3 u0
1
1
1
1
1
4
y D 2; 2 D 8; 3 D 6;
D
C c; y.1/ D p ) u.1/ D p ) c D
;
2
5
x
x
x
u
x
2u
5x
5
2
2




1=2
1=2
5x 5
5x
uD
;
y
D
.
2.1 C 4x 5/
2.1 C 4x 5 /
Z t
u0
1
1
2.4.14. P D ue at ; u0 e at D au2 e 2at ; 2 D ae at ;
D a
./e a d 
; P D
u
u
P0
0
P0 e at
P0
, which can also be written as P D
. Therefore,
Rt
Rt
1 C aP0 0 ./e a d 
e at C aP0 e at 0 ./e a d 
2.4.12. Rewrite as y 0 C

14 Chapter 2 First Order Equations

limt !1 P .t/ D

8
<

1
0
:
1=aL

if L D 0;
if L D 1;
if 0 < L < 1:

2.4.16. y D ux; u0 x C u D u2 C 2u; (A) u0 x D u.u C 1/. Since u  0 and u  1 are constant
solutions of (A), y  0 and y D x are solutions
The nonconstant solutions
equation.

 of the given
u
u0
1
1
1
1
u
0
D ln jxj C k;
of (A) satisfy D
D ;
u D ; ln
D cx;

u.u C 1/
x
u uC1
x
uC1
uC1
cx
cx 2
; yD
.
u D .u C 1/cx; u.1 cx/ D cx; u D
1 cx
1 cx
1
2.4.18. y D ux; u0 x C u D u C sec; u0 x D sec u; .cos u/u0 D ; sin u D ln jxj C c; u D
x
sin 1 .ln jxj C c/; y D x sin 1 .ln jxj C c/.
x 2 C 2y 2
1
1 C u2
; y D ux; u0 x C u D C 2u; u0 x D
;
xy 
u
u

uu0
1 1
y2
y2
2
2
D
;
ln.1
C
u
/
D
ln
jxjC
k;
ln
1
C
D
ln
x
C
2k;
1
C
D cx 2; x 2 C y 2 D cx 4 ;
1 C u2 p x 2
x2
x2
y D x cx 2 1.
2.4.20. Rewrite the given equation as y 0 D

u0
1
2.4.22. y D ux; u0 x C u D u C u2 ; u0 x D u2 ; 2 D ;
u
x
2
2x
1
2)cD ; uD
; yD
.
2
2 ln jxj C 1
2 ln jxj C 1
2.4.24. Rewrite the given equation as y 0 D
1
uu0
D ;
2
1 C 2u
x

1
D ln jxj C c; y. 1/ D 2 ) u. 1/ D
u

x2 C y2
; y D ux; u0 x C u D
xy

1
u

u; u0 x D

1 C 2u2
;
u

1
ln.1 C 2u2 / D ln jxj C k; x 4.1 C 2u2 / D c; y.1/ D 2 ) u.1/ D 2 ) c D 9;
4

1=2

1=2
9 x4
1 9 x4
1 9 x4
x 4 .1 C 2u2 / D 9; u2 D
;
u
D
;
y
D
.
2x 4
x2
2
x
2
y
y2
C 4 ; y D ux; u0 x C u D 2 C u2 C 4u;
2
x
x



u C 1
u0
1
1
1
1
D
D ;
u0 D ; ln
.u C 1/.u C 2/
x
uC1 uC2
x
u C 2
2 uC1
2
2
) u.1/ D 1 ) c D ;
D x; u C 1 D x.u C 2/;
3 uC2
3
3
3
x.4x 3/
; yD
.
3
2x 3

2.4.26. Rewrite the given equation as y 0 D 2 C


u0 x D u2 C 3u C 2 D .u C 1/.u C 2/;
uC1
ln jxj C k;
D cx; y.1/ D 1
u

 C2
2
4
4x
u 1
x D 1 C x; u D
3
3
2x

1Cu 0
1 C u2 .1 u/u0
1
2.4.28. y D ux; u0 xCu D
; uxD
;
D ; tan
2
1
u
1
u
1
C
u
x


y 1
y2
y 1
tan 1
ln 1 C 2 D ln jxj C c; tan 1
ln.x 2 C y 2 / D c.
x 2
x
x 2

1
ln.1Cu2/ D ln jxjCc;
2

u3 C 2u2 C u C 1
u.u C 1/2 C 1
1
1
D
D uC
; u0 x D
;
.u C 1/2
.u C 1/2
.u C 1/2
.u C 1/2


3
3
1 .u C 1/
y
.u C 1/2 u0 D ;
D ln jxj C c; .u C 1/3 D 3.ln jxj C c/;
C 1 D 3.ln jxj C c/;
x
3
x
.y C x/3 D 3x 3.ln jxj C c/.
2.4.30. y D ux; u0 x C u D

Section 2.4 Transformation of Nonlinear Equations into Separable Equations

15

u
u.u 3/
2.4.32. y D ux; u0 x C u D
; (A) u0 x D
; Since u  0 and u  3 are constant solutions
u 2
2 u
of (A), y  0 and y D 3x are solutions
of the given equation. The nonconstant solutions of (A) satisfy

.2 u/u0
1
1
2 0
3
c
D ;
C
u D
; ln ju 3j C 2 ln juj D 3 ln jxj C k; u2 .u 3/ D 3 ;
u.u 3/
x
u 3
u
x
x
y 2 .y 3x/ D c.
2.4.34. y D ux; u0 x C u D
y
y3
C 3 D ln jxj C c.
x
x

1 C u C 3u3
1
1
D uC
; .1 C 3u2 /u0 D ; u C u3 D ln jxj C c;
2
2
1 C 3u
1 C 3u
x

x 2 xy C y 2
1
1 u
; y D ux; u0 xCu D
1Cu; u0 x D
;
2.4.36. Rewrite the given equation as y 0 D
xy
u
u


0
uu
1
1
1
c
D
; 1C
u0 D
; uC ln ju 1j D ln jxjC k; e u.u 1/ D ; e y=x .y x/ D c.
u 1
x
u 1
x
x
uC1
1
C u; (A) u0 x D
. Since (A) has the constant solution u D 1;
u
u
uu0
1
y D x is a solution of the given equation. The nonconstant solutions of (A) satisfy
D ;
uC1
x


y

1
y
1

0
1
u D ; u ln ju C 1j D ln jxj C c;
ln
1 D ln jxj C c; y x ln jy xj D cy.
uC1
x
x
x
2.4.38. y D ux; u0 x C u D 1 C

dY
d Y dX
dY
dy
D
D
D
, so y D y.x/ satisfies the
dx
dx
dX dx
dX
given equation if and only if Y D Y .X/ satisfies
2.4.40. If x D X

X0 and y D Y

Y0 , then

dY
a.X
D
dX
c.X

X0 / C b.Y
X0 / C d.Y

Y0 / C
;
Y0 / C

which reduces to the nonlinear homogeneous equation


dY
aX C bY
D
dX
cX C d Y
if and only if
aX0 C bY0
cX0 C d Y0

D
D :

.B/

We will now show that if ad bc 0, then it is possible (for any choice of and ) to solve (B).
Multiplying the first equation in (B) by d and the second by b yields
daX0 C dbY0
bcX0 C bd Y0

D d
D b:

Subtracting the second of these equations from the first yields .ad bc/X0 D d b. Since ad bc
d b
0, this implies that X0 D
. Multiplying the first equation in (B) by c and the second by a yields
ad bc
caX0 C cbY0

acX0 C ad Y0

D c

D a:

16 Chapter 2 First Order Equations


Subtracting the first of these equation from the second yields .ad
c a
this implies that Y0 D
.
ad bc

bc/Y0 D c a. Since ad

bc 0

2.4.42. For the given equation, (B) of Exercise 2.4.40 is


2X0 C Y0

X0 C 2Y0

1
4:

Solving this pair of equations yields X0 D 2 and Y0 D 3. The transformed differential equation is

dY
2X C Y
D
:
.A/
dX
X C 2Y
2Cu
2.u 1/.u C 1/
Let Y D uX; u0 X C U D
; (B) u0 X D
. Since u  1 and u  1
1 C 2u
2u C 1
satisfy (B), Y D X and Y D X are solutions of (A). Since X D x C 2 and Y D y 3, it follows
that y D x C 5 and y D x C 1 are solutions of the given equation. The nonconstant solutions
2
1
3
4
.2u C 1/u0
D
;
C
u0 D
; ln ju C 1j C 3 ln ju 1j D
of (B) satisfy
.u 1/.u C 1/
X
uC1
u 1
X
c
4 ln jXj C k; .u C 1/.u 1/3 D 4 ; .Y C X/.Y X/3 D c; Setting X D x C 2 and Y D y 3
X
yields .y C x 1/.y x 5/3 D c.
2.4.44. Rewrite the given equation as y 0 D
u0 x 1=3 D

1
y3 C x
u3 C 1
; y D ux 1=3 ; u0 x 1=3 C
u D
;
2
2=3
3xy
3x
3u2 x 2=3

1
1 u3
2 0
;
u
u
D
;
D .ln jxj C c/; u D .ln jxj C c/1=3 ; y D x 1=3.ln jxj C c/1=3 .
3x 3
3
3x 2=3 u2
1

2.y 2 C x 2 y x 4 /
; y D ux 2 ; u0 x 2 C 2xu D 2x.u2 C
x3
u 1/; (A) u0 x 2 D 2x.u2 1/. Since u  1 and u  1 are constant solutions of (A), y D x 2
u0
and y D x 2 are solutions of the given equation. The nonconstant solutions of (A) satisfy 2
D
u
1




1
1
4
u 1
u 1
2
;
u0 D ; ln
D 4 ln jxj C k;
D cx 4 ; .u 1/ D .u C 1/cx 4 ;
x
u 1 uC1
x
u C 1
uC1
1 C cx 4
x 2 .1 C cx 4 /
u.1 cx 4 / D 1 C cx 4 ; u D
;
y
D
.
1 cx 4
1 cx 4
2.4.46. Rewrite the given equation as y 0 D

u0
D
C1
sec2 x cot x D cot x C tan x; tan 1 u D ln j sin xj ln j cos xj C c D ln j tan xj C c; u D tan.ln j tan xj C
c/; y D tan x tan.ln j tan xj C c/.
p
.y C x/2
1
.u C 1/2
0
2.4.50. Rewrite the given equation as y D
p ; y D ux 1=2 ; u0 x 1=2 C p u D p
;
2x.y C 2 x/
2 x
2 x .u C 2/
1
1
.u C 2/2
1
u0 x 1=2 D p
; .u C 2/u0 D
;
D .ln jxj C c/; .u C 2/2 D ln jxj C c;
2x
2
2
2
x
.u
C
2/
p
p
u D 2 ln jxj C c; y D x 1=2 . 2 ln jxj C c/.
2.4.48. y D u tan x; u0 tan x C u sec2 x D .u2 C u C 1/ sec2 x; u0 tan x D .u2 C 1/ sec2 x;

2.4.52. y1 D

1
2
u
is a solution of y 0 C y D 0. Let y D 2 ; then
x2
x
x

u0
3x 2.u2 =x 4 / C 6x.u=x 2 / C 2
3.u=x/2 C 6.u=x/ C 2
D
D
;
x2
x 2 .2x.u=x 2 / C 3/
x 2 .2.u=x/ C 3/

u2

Section 2.5 Exact Equations

17

3.u=x/2 C 6.u=x/ C 2
. Since (A) is a homogeneous nonlinear equation, we now substitute
2.u=x/ C 3
3v 2 C 6v C 2 0
.v C 1/.v C 2/
.2v C 3/v 0
1
; vxD
;
D ;
u D vx into (A). This yields v 0 x C v D
2v C 3
2v C 3
.v C 1/.v C 2/
x


1
1
1
0
C
v D ; ln j.v C 1/.v C 2/j D ln jxj C k; (B) .v C 1/.v C 2/ D cx. Since
vC1
vC2
x
y.2/ D 2 ) u.2/ D 8 ) v.2/ D 4, (B) implies that p
c D 15. .v C 1/.v C 2/ D 15x; pv 2 C 3v C
3 C 1 C 60x
x. 3 C 1 C 60x/
2 15x D 0. From the quadratic formula, v D
; u D vx D
;
2
2
p
u
3 C 1 C 60x
yD 2 D
.
x
2x
so (A) u0 D

y.ax/
1
yields (B) y10 .x/ D y 0 .ax/  a D y 0 .ax/. Since y 0 .x/ D
a
a
q.y.x/=x/ on some interval I , (C) y 0 .ax/ D q.y.ax/=ax/ on some interval J . Substituting (A) and (B)
into (C) yields y10 .x/ D q.y1 .x/=x/ on J .
2.4.54. Differentiating (A) y1 .x/ D

u0
2.4.56. If y D C 1, then 0 C D x2 ; D ue x ; u0 e x D xu2 e 2x ; 2 D xe
u
1
1
1
e x .x C 1/ c; u D x
; D
;
y
D
1
C
.
e .x C 1/ C c
x C 1 C ce x
x C 1 C ce x

1
u
u0
u2 u0
1
2
2.4.58. If y D C 1, then 0 C D 2 ; 1 D 2 ; D 2 ; 2 D 4 ; 2 D 2 ;
x
x
x
x
x
u
x
1 cx
x
1
1
1
Cc D
; uD
; D
; yD1
.
x
x
1 cx
x.1 cx/
x.1 cx/

1
D
u

1
D
u

2.5 EXACT EQUATIONS


2.5.2. M.x; y/ D 3y cos x C 4xe x C 2x 2e x ; N.x; y/ D 3 sin x C 3; My .x; y/ D 3 cos x D Nx .x; y/,
so the equation is exact. We must find F such that (A) Fx .x; y/ D 3y cos x C 4xe x C 2x 2 e x and (B)
Fy .x; y/ D 3 sin x C 3. Integrating (B) with respect to y yields (C) F .x; y/ D 3y sin x C 3y C .x/.
Differentiating (C) with respect to x yields (D) Fx .x; y/ D 3y cos x C Z 0 .x/. Comparing (D) with
(A) shows that (E) 0 .x/ D 4xe x C 2x 2e x . Integration by parts yields xe x dx D xe x e x and
Z
x 2 e x dx D x 2e x 2xe x C2e x . Substituting from the last two equations into (E) yields .x/ D 2x 2e x .
Substituting this into (C) yields F .x; y/ D 3y sin x C3y C2x 2e x . Therefore, 3y sin x C3y C2x 2e x D c.

2.5.4. M.x; y/ D 2x 2y 2 ; N.x; y/ D 12y 2 4xy; My .x; y/ D 4y D Nx .x; y/, so the equation
is exact. We must find F such that (A) Fx .x; y/ D 2x 2y 2 and (B) Fy .x; y/ D 12y 2 4xy.
Integrating (A) with respect to x yields (C) F .x; y/ D x 2 2xy 2 C .y/. Differentiating (C) with
respect to y yields (D) Fy .x; y/ D 4xy C  0 .y/. Comparing (D) with (B) shows that  0 .y/ D 12y 2 ,
so we take .y/ D 4y 3 . Substituting this into (C) yields F .x; y/ D x 2 2xy 2 C 4y 3 . Therefore,
x 2 2xy 2 C 4y 3 D c.
2.5.6. M.x; y/ D 4x C 7y; N.x; y/ D 3x C 4y; My .x; y/ D 7 3 D Nx .x; y/, so the equation is
not exact.
2.5.8. M.x; y/ D 2x C y; N.x; y/ D 2y C 2x; My .x; y/ D 1 2 D Nx .x; y/, so the equation is not
exact.

18 Chapter 2 First Order Equations


2.5.10. M.x; y/ D 2x 2 C 8xy C y 2 ; N.x; y/ D 2x 2 C
Nx .x; y/, so the equation is not exact.

xy 3
y3
; My .x; y/ D 8x C 2y 4x C
D
3
3

2.5.12. M.x; y/ D y sin xyCxy 2 cos xy; N.x; y/ D x sin xyCxy 2 cos xy; My .x; y/ D 3xy cos xyC
.1 x 2 y 2 / sin xy .xy C y 2 / cos xy C .1 xy 3 / sin xy D Nx .x; y/, so the equation is not exact.
2.5.14. M.x; y/ D e x .x 2 y 2 C 2xy 2 / C 6x; N.x; y/ D 2x 2 ye x C 2; My .x; y/ D 2xye x .x C 2/ D
Nx .x; y/, so the equation is exact. We must find F such that (A) Fx .x; y/ D e x .x 2 y 2 C 2xy 2 / C 6x and
(B) Fy .x; y/ D 2x 2 ye x C2. Integrating (B) with respect to y yields (C) F .x; y/ D x 2 y 2 e x C2yC .x/.
Differentiating (C) with respect to x yields (D) Fx .x; y/ D e x .x 2 y 2 C 2xy 2 / C 0 .x/. Comparing (D)
with (A) shows that 0 .x/ D 6x, so we take .x/ D 3x 2 . Substituting this into (C) yields F .x; y/ D
x 2 y 2 e x C 2y C 3x 2. Therefore, x 2 y 2 e x C 2y C 3x 2 D c.
2.5.16. M.x; y/ D e xy .x 4 y C 4x 3/ C 3y; N.x; y/ D x 5 e xy C 3x; My .x; y/ D x 4e xy .xy C 5/ C 3 D
Nx .x; y/, so the equation is exact. We must find F such that (A) Fx .x; y/ D e xy .x 4 y C 4x 3 / C 3y and
(B) Fy .x; y/ D x 5 e xy C3x. Integrating (B) with respect to y yields (C) F .x; y/ D x 4 e xy C3xy C .x/.
Differentiating (C) with respect to x yields (D) Fx .x; y/ D e xy .x 4 y C 4x 3/ C 3y C 0 .x/. Comparing
(D) with (A) shows that 0 .x/ D 0, so we take .x/ D 0. Substituting this into (C) yields F .x; y/ D
x 4 e xy C 3xy. Therefore, x 4 e xy C 3xy D c.
2.5.18. M.x; y/ D 4x 3y 2 6x 2 y 2x 3; N.x; y/ D 2x 4 y 2x 3 ; My .x; y/ D 8x 3y 6x 2 D
Nx .x; y/, so the equation is exact. We must find F such that (A) Fx .x; y/ D 4x 3 y 2 6x 2 y 2x 3
and (B) Fy .x; y/ D 2x 4 y 2x 3 . Integrating (A) with respect to x yields (C) F .x; y/ D x 4 y 2 2x 3 y
x 2 3x C .y/. Differentiating (C) with respect to y yields (D) Fy .x; y/ D 2x 4 y 2x 3 C  0 .y/.
Comparing (D) with (B) shows that  0 .y/ D 0, so we take .y/ D 0. Substituting this into (C) yields
F .x; y/ D x 4y 2 2x 3y x 2 3x. Therefore, x 4 y 2 2x 3 y x 2 3x D c. Since y.1/ D 3 ) c D 1,
x 4 y 2 2x 3y x 2 3x C 1 D 0 is an implicit solution
of the initial value problem. Solving this for y
p
x C 2x 2 C 3x 1
by means of the quadratic formula yields y D
.
x2
2.5.20. M.x; y/ D .y 3 1/e x ; N.x; y/ D 3y 2 .e x C 1/; My .x; y/ D 3y 2 e x D Nx .x; y/, so the
equation is exact. We must find F such that (A) Fx .x; y/ D .y 3 1/e x and (B) Fy .x; y/ D 3y 2 .e x C 1/.
Integrating (A) with respect to x yields (C) F .x; y/ D .y 3 1/e x C.y/. Differentiating (C) with respect
to y yields (D) Fy .x; y/ D 3y 2 e x C  0 .y/. Comparing (D) with (B) shows that  0 .y/ D 3y 2 , so we take
.y/ D y 3 . Substituting this into (C) yields F .x; y/ D .y 3 1/e x Cy 3 . Therefore, .y 3 1/e x Cy 3 D c.
Since y.0/ D 0 ) c D 1, .y 3 1/e x C y 3 D 1 is an implicit solution of the initial value problem.

 x
e
1 1=3
.
Therefore, y 3 .e x C 1/ D e x 1, so y D
ex C 1
2.5.22. M.x; y/ D .2x 1/.y 1/; N.x; y/ D .x C 2/.x 3/; My .x; y/ D 2x 1 D Nx .x; y/, so the
equation is exact. We must find F such that (A) Fx .x; y/ D .2x 1/.y 1/ and (B) Fy .x; y/ D .x C
2/.x 3/. Integrating (A) with respect to x yields (C) F .x; y/ D .x 2 x/.y 1/ C .y/. Differentiating
(C) with respect to y yields (D) Fy .x; y/ D x 2 x C  0 .y/. Comparing (D) with (B) shows that
 0 .y/ D 6, so we take .y/ D 6y. Substituting this into (C) yields F .x; y/ D .x 2 x/.y 1/ 6y.
Therefore, .x 2 x/.y 1/ 6y D c. Since y.1/ D 1 ) c D 6, .x 2 x/.y 1/ 6y D 6 is an implicit
x2 x C 6
solution of the initial value problem. Therefore, .x 2 x 6/y D x 2 x C 6, so y D
.
.x 3/.x C 2/
2.5.24. M.x; y/ D e x .x 4 y 2 C 4x 3 y 2 C 1/; N.x; y/ D 2x 4ye x C 2y; My .x; y/ D 2x 3 ye x .x C 4/ D
Nx .x; y/, so the equation is exact. We must find F such that (A) Fx .x; y/ D e x .x 4 y 2 C 4x 3y 2 C 1/

Section 2.5 Exact Equations

19

and (B) Fy .x; y/ D 2x 4ye x C 2y. Integrating (B) with respect to y yields (C) F .x; y/ D x 4 y 2 e x C
y 2 C .x/. Differentiating (C) with respect to x yields (D) Fx .x; y/ D e x y 2 .x 4 C 4x 3/ C 0 .x/.
Comparing (D) with (A) shows that 0 .x/ D e x , so we take .x/ D e x . Substituting this into (C) yields
F .x; y/ D .x 4 y 2 C 1/e x C y 2 . Therefore, .x 4 y 2 C 1/e x C y 2 D c.
2.5.28. M.x; y/ D x 2 C y 2 ; N.x; y/ D 2xy; My .x; y/ D 2y D Nx .x; y/, so the equation is exact.
We must find F such that (A) Fx .x; y/ D x 2 C y 2 and (B) Fy .x; y/ D 2xy. Integrating (A) with
x3
respect to x yields (C) F .x; y/ D
C xy 2 C .y/. Differentiating (C) with respect to y yields (D)
3
Fy .x; y/ D 2xy C  0 .y/. Comparing (D) with (B) shows that  0 .y/ D 0, so we take .y/ D 0.
x3
x3
C xy 2 . Therefore,
C xy 2 D c.
Substituting this into (C) yields F .x; y/ D
3
3
2.5.30. (a) Exactness requires that Nx .x; y/ D My .x; y/ D

@ 3 2
.x y C 2xy C 3y 2/ D 2x 3y C 2x C 6y.
@y

x4y
C x 2 C 6xy C g.x/, where g is differentiable.
4
1
@
(b) Exactness requires that Nx .x; y/ D My .x; y/ D
.ln xy C 2y sin x/ D
C 2 sin x. Hence,
@y
y
x
N.x; y/ D
2 cos x C g.x/, where g is differentiable.
y
@
(c) Exactness requires that Nx .x; y/ D My .x; y/ D
.x sin x C y sin y/ D y cos y C sin y. Hence,
@y
N.x; y/ D x.y cos y C sin y/ C g.x/, where g is differentiable.

Hence, N.x; y/ D

2.5.32. The assumptions imply that

@N1
@M2
@N2
@
@M1
D
and
D
. Therefore,
.M1 C M2 / D
@y
@x
@y
@x
@y

@M1 @M2
@N1 @N2
@
C
D
C
D
.N1 C N2/, which implies that .M1 C M2 / dx C .N1 C N2/ dy D 0
@y
@y
@x
@x
@x
is exact on R.
2.5.34. Here M.x; y/ D Ax 2 CBxy CCy 2 and N.x; y/ D Dx 2 CExy CF y 2 . Since My D Bx C2Cy
and Nx D 2Dx C Ey, the equation is exact if and only if B D 2D and E D 2C .
Z y
Z x
2.5.36. Differentiating (A) F .x; y/ D
N.x0 ; s/ dsC
M.t; y/ dt with respect to x yields Fx .x; y/ D
y0

x0

M.x; y/, since the first integral in (A) is independent of x and M.t; y/ is a continuous function of t for
each fixed y. Differentiating
Z x (A) with respect to y and using
Z x the assumption that My D Nx yields
@M
@N
Fy .x; y/ D N.x0 ; y/ C
.t; y/ dt D N.x0 ; y/ C
.t; y/ dt D N.x0 ; y/ C N.x; y/
x0 @y
x0 @x
N.x0 ; y/ D N.x; y/.
2.5.38. y1 D

1
2
u
is a solution of y 0 C y D 0. Let y D 2 ; then
2
x
x
x
u0
D
x2

so u0 D

x2

.x 2

2x.u=x 2/
D
C 2x 2 .u=x 2 / C 1/

2xu
;
x 2 .x 2 C 2u C 1/

2xu
, which can be rewritten as (A) 2xu dx C .x 2 C 2u C 1/ du D 0. Since
C 2u C 1

@
@ 2
.2xu/ D
.x C 2u C 1/ D 2x, (A) is exact. To solve (A) we must find F such that (A) Fx .x; u/ D
@u
@x

20 Chapter 2 First Order Equations


2xu and (B) Fu .x; u/ D x 2 C2uC1. Integrating (A) with respect to x yields (C) F .x; u/ D x 2 uC.u/.
Differentiating (C) with respect to u yields (D) Fu .x; u/ D x 2 C  0 .u/. Comparing (D) with (B) shows
that  0 .u/ D 2u C 1, so we take .u/ D u2 C u. Substituting this into (C) yields F .x; u/ D x 2 u C
u2 C u D u.x 2 C u C 1/. Therefore, u.x 2 C u C 1/ D c. Since y.1/ D 2 ) u.1/ D 2; c D 0.
Therefore, u.x 2 C u C 1/ D 0. Since u  0 does not satisfy u.1/ D 2, it follows that u D x 2 1
1
and y D 1
.
x2


3x C 2u
2
2
2
2
2.5.40. y1 D e x is a solution of y 0 C 2xy D 0. Let y D ue x ; then u0 e x D e x
, so
2x C 3u
3x C 2u
@
u0 D
, which can be rewritten as (A) .3x C2u/ dx C.2x C3u/ du D 0. Since
.3x C2u/ D
2x C 3u
@u
@
.2x C 3u/ D 2, (A) is exact. To solve (A) we must find F such that (A) Fx .x; u/ D 3x C 2u and
@x
3x 2
C 2xu C .u/.
(B) Fu .x; u/ D 2x C 3u. Integrating (A) with respect to x yields (C) F .x; u/ D
2
0
Differentiating (C) with respect to u yields (D) Fu .x; u/ D 2x C  .u/. Comparing (D) with (B) shows
3u2
3x 2
that  0 .u/ D 3u, so we take .u/ D
. Substituting this into (C) yields F .x; u/ D
C 2xu C
2
2
2
2
2
3u
3x
3u
3
. Therefore,
C 2xu C
D c. Since y.0/ D 1 ) u.0/ D 1; c D . Therefore,
2
2
2
2
2
3x 2 C 4xu C 3u2 D 3 is an implicit solution of the initial value problem. Rewriting this as
! C
p3u C 4xu
2
2x C 9 5x
.3x 2 3/ D 0 and solving for u by means of the quadratic formula yields u D
, so
3
!
p
9 5x 2
x 2 2x C
yD e
.
3
2.5.42. Since M dx C N dy D 0 is exact, (A) My D Nx . Since N dx C M dy D 0 is exact, (B)
Mx D Ny . Differentiating (A) with respect to y and (B) with respect to x yields (C) Myy D Nxy and
(D) Mxx D Nyx . Since Nxy D Nyx , adding (C) and (D) yields Mxx C Myy D 0. Differentiating
(A) with respect to x and (B) with respect to y yields (E) Myx D Nxx and (F) Mxy D Nyy . Since
Mxy D Myx , subtracting (F) from (E) yields Nxx C Nyy D 0.
2.5.44. (a) If F .x; y/ D x 2 y 2 , then Fx .x; y/ D 2x, Fy .x; y/ D 2y, Fxx .x; y/ D 2, and
Fyy .x; y/ D 2. Therefore, Fxx C Fyy D 0, and G must satisfy (A) Gx .x; y/ D 2y and (B)
Gy .x; y/ D 2x. Integrating (A) with respect to x yields (C) G.x; y/ D 2xy C .y/. Differentiating (C)
with respect to y yields (D) Gy .x; y/ D 2x C  0 .y/. Comparing (D) with (B) shows that  0 .y/ D 0, so
we take .y/ D c. Substituting this into (C) yields G.x; y/ D 2xy C c.
(b) If F .x; y/ D e x cos y, then Fx .x; y/ D e x cos y, Fy .x; y/ D e x sin y, Fxx .x; y/ D e x cos y,
and Fyy .x; y/ D e x cos y. Therefore, Fxx C Fyy D 0, and G must satisfy (A) Gx .x; y/ D e x sin y
and (B) Gy .x; y/ D e x cos y. Integrating (A) with respect to x yields (C) G.x; y/ D e x sin y C .y/.
Differentiating (C) with respect to y yields (D) Gy .x; y/ D e x cos y C  0 .y/. Comparing (D) with (B)
shows that  0 .y/ D 0, so we take .y/ D c. Substituting this into (C) yields G.x; y/ D e x sin y C c.
(c) If F .x; y/ D x 3 3xy 2 , then Fx .x; y/ D 3x 2 3y 2 , Fy .x; y/ D 6xy, Fxx .x; y/ D 6x,
and Fyy .x; y/ D 6x. Therefore, Fxx C Fyy D 0, and G must satisfy (A) Gx .x; y/ D 6xy and
(B) Gy .x; y/ D 3x 2 3y 2 . Integrating (A) with respect to x yields (C) G.x; y/ D 3x 2y C .y/.
Differentiating (C) with respect to y yields (D) Gy .x; y/ D 3x 2 C  0 .y/. Comparing (D) with (B)
shows that  0 .y/ D 3y 2 , so we take .y/ D y 3 C c. Substituting this into (C) yields G.x; y/ D
3x 2 y y 3 C c.
(d) If F .x; y/ D cos x cosh y, then Fx .x; y/ D sin x cosh y, Fy .x; y/ D cos x sinh y, Fxx .x; y/ D

Section 2.6 Exact Equations

21

cos x cosh y, and Fyy .x; y/ D cos x cosh y. Therefore, Fxx C Fyy D 0, and G must satisfy (A)
Gx .x; y/ D cos x sinh y and (B) Gy .x; y/ D sin x cosh y. Integrating (A) with respect to x yields
(C) G.x; y/ D sin x sinh y C .y/. Differentiating (C) with respect to y yields (D) Gy .x; y/ D
sin x cosh y C 0 .y/. Comparing (D) with (B) shows that  0 .y/ D 0, so we take .y/ D c. Substituting
this into (C) yields G.x; y/ D sin x sinh y C c.
(e) If F .x; y/ D sin x cosh y, then Fx .x; y/ D cos x cosh y, Fy .x; y/ D sin x sinh y, Fxx .x; y/ D
sin x cosh y, and Fyy .x; y/ D sin x cosh y. Therefore, Fxx C Fyy D 0, and G must satisfy (A)
Gx .x; y/ D sin x sinh y and (B) Gy .x; y/ D cos x cosh y. Integrating (A) with respect to x yields
(C) G.x; y/ D cos x sinh y C .y/. Differentiating (C) with respect to y yields (D) Gy .x; y/ D
cos x cosh y C  0 .y/. Comparing (D) with (B) shows that  0 .y/ D 0, so we take .y/ D c. Substituting this into (C) yields G.x; y/ D cos x sinh y C c.
2.6 INTEGRATING FACTORS
1
2.6.2. (a) and (b). To show that .x; y/ D
is an integrating factor for (A) and that (B) is exact,
.x y/2




xy
y2
@
xy
x2
@
D
and
D
. By Theorem 2.5.1
it suffices to observe that
2
@x x y
.x y/
@y x y
.x y/2
this also shows that (C) is an implicit solution of (B). Since .x; y/ is never zero, any solution of (B) is
a solution of (A).
(c) If we interpret (A) as y 2 C x 2 y 0 D 0, then substituting y D x yields x 2 C x 2  1 D 0.
(N OTE: In Exercises 2.6.32.6.23, the given equation is multiplied by an integrating factor to produce
an exact equation, and an implicit solution is found for the latter. For a complete analysis of the relationship between the sets of solutions of the two equations it is necessary to check for additional solutions of
the given equation along which" the integrating factor is undefined, or for solutions of the exact equation
along which" the integrating factor vanishes. In the interests of brevity we omit these tedious details
except in cases where there actually is a difference between the sets of solutions of the two equations.)
2.6.4. M.x; y/ D 3x 2 y; N.x; y/ D 2x 3 ; My .x; y/ Nx .x; y/ D 3x 2 6x 2 D 3x 2 ; p.x/ D
My .x; y/ Nx .x; y/
3x 2
3 R
3
D
D
;
p.x/ dx D
ln jxj; .x/ D P .x/ D x 3=2 ; thereN.x; y/
2x 3
2x
2
fore 3x 1=2y dx C 2x 3=2 dy D 0 is exact. We must find F such that (A) Fx .x; y/ D 3x 1=2y and (B)
Fy .x; y/ D 2x 3=2. Integrating (A) with respect to x yields (C) F .x; y/ D 2x 3=2 y C .y/. Differentiating (C) with respect to y yields (D) Fy .x; y/ D 2x 3=2 C  0 .y/. Comparing (D) with (B) shows that
 0 .y/ D 0, so we take .y/ D 0. Substituting this into (C) yields F .x; y/ D 2x 3=2y, so x 3=2y D c.
2.6.6. M.x; y/ D 5xy C 2y C 5; N.x; y/ D 2x; My .x; y/ Nx .x; y/ D .5x C 2/ 2 D 5x; p.x/ D
My .x; y/ Nx .x; y/
5x
5 R
5x
D
D ; p.x/ dx D
; .x/ D P .x/ D e 5x=2 ; therefore e 5x=2 .5xy C
N.x; y/
2x
2
2
2y C 5/ dx C 2xe 5x=2 dy D 0 is exact. We must find F such that (A) Fx .x; y/ D e 5x=2.5xy C 2y C 5/
and (B) Fy .x; y/ D 2xe 5x=2. Integrating (B) with respect to y yields (C) F .x; y/ D 2xye 5x=2 C .x/.
Differentiating (C) with respect to x yields (D) Fx .x; y/ D 5xye 5x=2 C 2ye 5x=2 C 0 .x/. Comparing
(D) with (A) shows that 0 .x/ D 5e 5x=2, so we take .x/ D 2e 5x=2. Substituting this into (C) yields
F .x; y/ D 2e 5x=2.xy C 1/, so e 5x=2 .xy C 1/ D c.
2.6.8. M.x; y/ D 27xy 2 C 8y 3 ; N.x; y/ D 18x 2y C 12xy 2 ; My .x; y/ Nx .x; y/ D .54xy C
My .x; y/ Nx .x; y/
18xy C 12y 2
1
24y 2 / .36xy C 12y 2 / D 18xy C 12y 2 ; p.x/ D
D
D ;
2 y C 12y 2 x
N.x;
y/
18x
x
R
p.x/ dx D ln jxj; .x/ D P .x/ D x; therefore .27x 2 y 2 C 8xy 3 / dx C .18x 3y C 12x 2y 2 / dy D 0 is
exact. We must find F such that (A) Fx .x; y/ D 27x 2y 2 C 8xy 3 and (B) Fy .x; y/ D 18x 3y C 12x 2y 2 .
Integrating (A) with respect to x yields (C) F .x; y/ D 9x 3y 2 C 4x 2 y 3 C .y/. Differentiating (C)

22 Chapter 2 Integrating Factors


with respect to y yields (D) Fy .x; y/ D 18x 3 y C 12x 2y 2 C  0 .y/. Comparing (D) with (B) shows
that  0 .y/ D 0, so we take .y/ D 0. Substituting this into (C) yields F .x; y/ D 9x 3 y 2 C 4x 2y 3 , so
x 2 y 2 .9x C 4y/ D c.


1
2
2
2.6.10. M.x; y/ D y ; N.x; y/ D xy C 3xy C
; My .x; y/ Nx .x; y/ D 2y .y 2 C 3y/ D
y
Nx .x; y/ My .x; y/
y.y C 1/
1 R
y.y C 1/; q.y/ D
D
D 1 C ; q.y/ dy D y ln jyj; .y/ D
M.x; y/
y2
y
Q.y/ D ye y ; therefore y 3 e y dx C e y .xy 3 C 3xy 2 C 1/ dy D 0 is exact. We must find F such that
(A) Fx .x; y/ D y 3 e y and (B) Fy .x; y/ D e y .xy 3 C 3xy 2 C 1/. Integrating (A) with respect to x
yields (C) F .x; y/ D xy 3 e y C .y/. Differentiating (C) with respect to y yields (D) Fy .x; y/ D
xy 3 e y C 3xy 2 e y C  0 .y/. Comparing (D) with (B) shows that  0 .y/ D e y , so we take .y/ D e y .
Substituting this into (C) yields F .x; y/ D xy 3 e y C e y , so e y .xy 3 C 1/ D c.
2.6.12. M.x; y/ D x 2 y C 4xy C 2y; N.x; y/ D x 2 C x; My .x; y/ Nx .x; y/ D .x 2 C 4x C 2/ .2x C
.x C 1/2
1 R
My .x; y/ Nx .x; y/
1/ D x 2 C 2x C 1 D .x C 1/2 ; p.x/ D
D
D 1 C ; p.x/ dx D
N.x; y/
x.x C 1/
x
x C ln jxj; .x/ D P .x/ D xe x ; therefore e x .x 3 y C 4x 2 y C 2xy/ dx C e x .x 3 C x 2 / dy D 0 is exact.
We must find F such that (A) Fx .x; y/ D e x .x 3 y C 4x 2y C 2xy/ and (B) Fy .x; y/ D e x .x 3 C x 2 /.
Integrating (B) with respect to y yields (C) F .x; y/ D y.x 3 C x 2 /e x C .x/. Differentiating (C) with
respect to x yields (D) Fx .x; y/ D e x .x 3 y C 4x 2y C 2xy/ C 0 .x/. Comparing (D) with (A) shows
that 0 .x/ D 0, so we take .x/ D 0. Substituting this into (C) yields F .x; y/ D y.x 3 C x 2 /e x D
x 2 y.x C 1/e x , so x 2y.x C 1/e x D c.
2.6.14. M.x; y/ D cos x cos y; N.x; y/ D sin x cos y

sin x sin y C y; My .x; y/ Nx .x; y/ D


cos x cos y
Nx .x; y/ My .x; y/
cos x sin y .cos x cos y cos x sin y/ D cos x cos y; q.y/ D
D
D
M.x; y/
cos x cos y
R
1; q.y/ dy D 1; .y/ D Q.y/ D e y ; therefore e y cos x cos y dx C e y .sin x cos y sin x sin y C
y/ dy D 0 is exact. We must find F such that (A) Fx .x; y/ D e y cos x cos y and (B) Fy .x; y/ D
e y .sin x cos y sin x sin y C y/. Integrating (A) with respect to x yields (C) F .x; y/ D e y sin x cos y C
.y/. Differentiating (C) with respect to y yields (D) Fy .x; y/ D e y .sin x cos y sin x sin y/ C  0 .y/.
Comparing (D) with (B) shows that  0 .y/ D ye y , so we take .y/ D e y .y 1/. Substituting this into
(C) yields F .x; y/ D e y .sin x cos y C y 1/, so e y .sin x cos y C y 1/ D c.
2.6.16. M.x; y/ D y sin y; N.x; y/ D x.sin y y cos y/; My .x; y/ Nx .x; y/ D .y cos y C sin y/
Nx .x; y/ My .x; y/
2 cos y R
.sin y y cos y/ D 2y cos y; q.y/ D
D
; q.y/ dy D 2 ln j sin yj;
sin y

N.x; y/ 
y
1
y cos y
1
.y/ D Q.y/ D
; therefore
dx C x
dy D 0 is exact. We must
2
2
sin y
sin y
sin y
sin
 y

y
1
y cos y
find F such that (A) Fx .x; y/ D
and (B) Fy .x; y/ D x
. Integrating (A)
sin y
sin y
sin2 y
xy
with respect to x yields (C) F .x; y/ D
C .y/. Differentiating (C) with respect to y yields (D)
sin y


1
y cos y
Fy .x; y/ D x
C  0 .y/. Comparing (D) with (B) shows that  0 .y/ D 0, so we take
sin y
sin2 y
xy
xy
.y/ D 0. Substituting this into (C) yields F .x; y/ D
, so
D c. In addition, the given
sin y
sin y
equation has the constant solutions y D k , where k is an integer.
2.6.18. M.x; y/ D y C xy; N.x; y/ D x C xy; My .x; y/ Nx .x; y/ D . C x/ . C y/;
and p.x/N.x; y/ q.y/M.x; y/ D p.x/x. C y/ q.y/y. C x/. so exactness requires that

Section 2.6 Exact Equations

23

. C x/

. C y/ D p.x/x. C y/ q.y/y. C x/, which holds if p.x/x D 1 and q.y/y D


R
1
1 R
1
; q.y/ D
;
p.x/ dx D ln jxj;
q.y/ dy D ln jyj; P .x/ D ;
1. Thus p.x/ D
x
y
x




1
1

Q.y/ D ; .x; y/ D
. Therefore,
C dx C
C dy D 0 is exact. We must find F
y
xy
x
y

such that (A) Fx .x; y/ D C and (B) Fy .x; y/ D


C . Integrating (A) with respect to x yields
x
y
(C) F .x; y/ D ln jxj C x C .y/. Differentiating (C) with respect to y yields (D) Fy .x; y/ D  0 .y/.

Comparing (D) with (B) shows that  0 .y/ D C , so we take .y/ D ln jyj C y. Substituting this
y
into (C) yields F .x; y/ D ln jxj C x C ln jyj C y, so jxj jyj e x e y D c. The given equation also
has the solutions x  0 and y  0.
2.6.20. M.x; y/ D 2y; N.x; y/ D 3.x 2 C x 2 y 3 /; My .x; y/ Nx .x; y/ D 2 .6x C 6xy 3 /; and
p.x/N.x; y/ q.y/M.x; y/ D 3p.x/.x 2 C x 2y 3 / 2q.y/y. so exactness requires that (A) 2 6x
6xy 3 D 3p.x/x.x C xy 3 / 2q.y/y. To obtain similar terms on the two sides of (A) we let p.x/x D a
and q.y/y D b where a and b are constants such that 2 6x 6xy 3 D 3a.x C xy 3 / 2b, which holds if
R
2
1 R
a D 2 and b D 1. Thus, p.x/ D
; q.y/ D
; p.x/ dx D 2 ln jxj; q.y/ dy D ln jyj;
x
y


1
1
1
2
1
2
P .x/ D 2 ; Q.y/ D ; .x; y/ D 2 . Therefore, 2 dx C 3
C y dy D 0 is exact. We must
x
y
x y
x 
y
1
2
find F such that (B) Fx .x; y/ D 2 and (C) Fy .x; y/ D 3
C y 2 . Integrating (B) with respect to x
x
y
2
yields (D) F .x; y/ D
C .y/. Differentiating (D) with respect to y yields (E) Fy .x; y/ D  0 .y/.
x


1
2
0
Comparing (E) with (C) shows that  .y/ D 3
C y , so we take .y/ D y 3 C 3 ln jyj. Substituting
y
2
2
this into (D) yields F .x; y/ D
C y 3 C 3 ln jyj, so
C y 3 C 3 ln jyj D c. The given equation also
x
x
has the solutions x  0 and y  0.
2.6.22. M.x; y/ D x 4 y 4 ; N.x; y/ D x 5y 3 ; My .x; y/ Nx .x; y/ D 4x 4 y 3 5x 4 y 3 D x 4 y 3 ;
and p.x/N.x; y/ q.y/M.x; y/ D p.x/x 5 y 3 q.y/x 4 y 4 . so exactness requires that x 4 y 3 D
p.x/x 5 y 3 q.y/x 4 y 4 , which is equivalent to p.x/x q.y/y D 1. This holds if p.x/x D a and
aC1 R
a
q.y/y D a C 1 where a is an arbitrary real number. Thus, p.x/ D ; q.y/ D
; p.x/ dx D
x
y
R
a ln jxj; q.y/ dy D .a C 1/ lnjyj; P .x/ D jxja ; Q.y/ D jyjaC1 ; .x; y/ D jx a jjyjaC1 . Therefore,
jxja jyjaC1 x 4 y 4 dx C x 5 y 3 dy D 0 is exact for any choice of a. For simplicity we let a D 4, so (A)
is equivalent to y dx C x dy D 0. We must find F such that (B) Fx .x; y/ D y and (C) Fy .x; y/ D x.
Integrating (B) with respect to x yields (D) F .x; y/ D xy C .y/. Differentiating (D) with respect to y
yields (E) Fy .x; y/ D x C  0 .y/. Comparing (E) with (C) shows that  0 .y/ D 0, so we take .y/ D 0.
Substituting this into (D) yields F .x; y/ D xy, so xy D c.
2.6.24. M.x; y/ D x 4y 3 C y; N.x; y/ D x 5 y 2 x; My .x; y/ Nx .x; y/ D .3x 4 y 2 C 1/ .5x 4 y 2
My .x; y/ Nx .x; y/
2x 4y 2 2
2 R
1/ D 2x 4 y 2 C 2; p.x/ D
D
D
; p.x/ dx D 2 ln jxj;
5
2
N.x; y/
x y
x
x


1
y 
1
dy D 0 is exact. We must find
.x/ D P .x/ D 2 ; therefore x 2 y 3 C 2 dx C x 3 y 2
x
x
x


y 
1
2 3
3 2
F such that (A) Fx .x; y/ D x y C 2 and (B) Fy .x; y/ D x y
. Integrating (A) with
x
x

24 Chapter 2 Integrating Factors


x 3y 3
y
respect to x yields (C) F .x; y/ D
C .y/. Differentiating (C) with respect to y yields (D)
3
x
1
Fy .x; y/ D x 3 y 2
C  0 .y/. Comparing (D) with (B) shows that  0 .y/ D 0, so we take .y/ D 0.
x
x3y3 y
x3y3 y
, so
D c.
Substituting this into (C) yields F .x; y/ D
3
x
3
x
2.6.26. M.x; y/ D 12xy C 6y 3 ; N.x; y/ D 9x 2 C 10xy 2 ; My .x; y/ Nx .x; y/ D .12x C 18y 2 /
.18x C10y 2 / D 6x C8y 2 ; and p.x/N.x; y/ q.y/M.x; y/ D p.x/x.9x C10y 2 / q.y/y.12x C6y 2 /,
so exactness requires that (A) 6x C 8y 2 D p.x/x.9x C 10y 2 / q.y/y.12x C 6y 2 /. To obtain similar
terms on the two sides of (A) we let p.x/x D a and q.y/y D b where a and b are constants such that
6x C 8y 2 D a.9x C 10y 2 / b.12x C 6y 2 /, which holds if 9a 12b D 6, 10a 6b D 8; that is,
R
2
2 R
a D b D 2. Thus p.x/ D ; q.y/ D ; p.x/ dx D 2 ln jxj; q.y/ dy D 2 ln jyj; P .x/ D x 2 ;
x
y
Q.y/ D y 2 ; .x; y/ D x 2 y 2 . Therefore, .12x 3 y 3 C 6x 2y 5 / dx C .9x 4 y 2 C 10x 3 y 4 / dy D 0 is exact.
We must find F such that (B) Fx .x; y/ D 12x 3 y 3 C 6x 2 y 5 and (C) Fy .x; y/ D 9x 4y 2 C 10x 3y 4 .
Integrating (B) with respect to x yields (D) F .x; y/ D 3x 4y 3 C 2x 3 y 5 C .y/. Differentiating (D)
with respect to y yields (E) Fy .x; y/ D 9x 4 y 2 C 10x 3y 4 C  0 .y/. Comparing (E) with (C) shows that
 0 .y/ D 0, so we take .y/ D 0. Substituting this into (D) yields F .x; y/ D 3x 4y 3 C 2x 3y 5 , so
x 3 y 3 .3x C 2y 2 / D c.


2.6.28. M.x; y/ D ax m yCby nC1 ; N.x; y/ D cx mC1 Cdxy n ; My .x; y/ Nx .x; y/ D ax mC1 C .n C 1/by n
.m C 1/cx m C dy n ; p.x/N.x; y/ q.y/M.x; y/ D xp.x/.cx m C dy n /  yp.y/.ax m C by n /. Let
(A) xp.x/ D and (B) yp.y/ D , where and are to be chosen so that ax mC1 C .n C 1/by n
.m C 1/cx m C dy n D .cx m C dy n / .ax m C by n /, which will hold if
c
d
Since ad

a
b

D
D

bc 0 it can be verified that D

(B), p.x/ D

a .m C 1/c
d C .n C 1/b
aB
ad

Ddf
Ddf

bA
cB
and D
bc
ad

A
B:

.C/
dA
satisfy (C). From (A) and
bc

and q.y/ D , so .x; y/ D x y is an integrating factor for the given equation.


x
y

My .x; y/ Nx .x; y/
D p.x/ and
N.x; y/
R
Theorem 2.6.1 implies that .x/ e p.x/R dx is an integrating factor for (C).
(b) Multiplying (A) through  D e p.x/ dx yields (D) .x/y 0 C 0 .x/yZ D .x/f .x/, which is
2.6.30. (a) Since M.x; y/ D p.x/y

f .x/ and N.x; y/ D 1,

equivalent to ..x/y/0 D .x/f .x/. Integrating this yields .x/y D c C .x/f .x/ dx, so y D


Z
1
1
c C .x/f .x/ dx , which is equivalent to (B) since y1 D
is a nontrivial solution of y 0 C
.x/

p.x/y D 0.

CHAPTER 3
Numerical Methods

3.1 EULERS METHOD


3.1.2. y1 D 1:200000000; y2 D 1:440415946; y3 D 1:729880994
3.1.4. y1 D 2:962500000; y2 D 2:922635828; y3 D 2:880205639

3.1.6.

x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

h D 0:1
2.000000000
2.100000000
2.514277288
3.317872752
4.646592772
6.719737638
9.876155616
14.629532397
21.751925418
32.399118931
48.298147362

h D 0:05
2.000000000
2.169990965
2.649377900
3.527672599
4.955798226
7.171467977
10.538384528
15.605686107
23.197328550
34.545932627
51.492825643

3.1.8.

x
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.45
1.50

h D 0:05
2.000000000
2.250000000
2.536734694
2.867950854
3.253613825
3.706750613
4.244700641
4.891020001
5.678467290
6.653845988
7.886170437

h D 0:025
2.000000000
2.259280190
2.559724746
2.910936426
3.325627715
3.820981064
4.420781829
5.158883503
6.085075790
7.275522641
8.852463793

h D 0:025
2.000000000
2.202114518
2.713011720
3.628465025
5.106379369
7.393322991
10.865186799
16.088630652
23.913328531
35.610005377
53.076673685
h D 0:0125
2.000000000
2.264490570
2.572794280
2.935723355
3.367843117
3.889251900
4.528471927
5.327348558
6.349785943
7.698316221
9.548039907

25

Exact
2.000000000
2.232642918
2.774352565
3.726686582
5.254226636
7.612186259
11.188475269
16.567103199
24.623248150
36.665439956
54.647937102
Exact
2.000000000
2.270158103
2.587150838
2.963263785
3.415384615
3.967391304
4.654198473
5.528980892
6.676923077
8.243593315
10.500000000

26 Chapter 3 Numerical Methods

3.1.10.

3.1.12.

3.1.14.

x
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0

x
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0

h D 0:1
h D 0:05
h D 0:025
1.000000000 1.000000000 1.000000000
0.920000000 0.921898275 0.922822717
0.847469326 0.851018464 0.852746371
0.781779403 0.786770087 0.789197876
0.722453556 0.728682209 0.731709712
0.669037867 0.676299618 0.679827306
0.621054176 0.629148585 0.633080163
0.578000416 0.586740390 0.590986601
0.539370187 0.548588902 0.553070392
0.504674296 0.514228603 0.518877246
0.473456737 0.483227470 0.487986391
Approximate Solutions

x
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0

h D 0:1
0.000000000
-0.100000000
-0.199000000
-0.294996246
-0.386095345
-0.470695388
-0.547627491
-0.616227665
-0.676329533
-0.728190908
-0.772381768

h D 0:05
0.000000000
-0.099875000
-0.198243434
-0.293129862
-0.382748403
-0.465664569
-0.540901018
-0.607969574
-0.666833345
-0.717819639
-0.761510960

h D 0:025
0.000000000
-0.099780455
-0.197800853
-0.292110713
-0.380986158
-0.463078857
-0.537503081
-0.603849795
-0.662136956
-0.712718751
-0.756179726

h D 0:1
0.0000
-0.0384
-0.0745
-0.1092
-0.1428
-0.1752
-0.2062
-0.2356
-0.2631
-0.2889
-0.3129

h D 0:05 h D 0:025
0.0000
0.0000
-0.0189
-0.0094
-0.0368
-0.0183
-0.0540
-0.0268
-0.0707
-0.0351
-0.0868
-0.0432
-0.1023
-0.0509
-0.1170
-0.0583
-0.1310
-0.0653
-0.1441
-0.0719
-0.1563
-0.0781
Residuals

Exact"
0.000000000
-0.099664000
-0.197315517
-0.291036003
-0.379168221
-0.460450590
-0.534085626
-0.599737720
-0.657473792
-0.707670533
-0.750912371

Eulers method
h D 0:1
h D 0:05
h D 0:025
Exact"
2.000000000
2.000000000
2.000000000
2.000000000
2.420000000
2.440610764
2.451962006
2.464119569
2.922484288
2.972198224
2.999753046
3.029403212
3.524104434
3.614025082
3.664184099
3.718409925
4.244823572
4.389380160
4.470531822
4.558673929
5.108581185
5.326426396
5.449503467
5.583808754
6.144090526
6.459226591
6.638409411
6.834855438
7.385795229
7.828984275
8.082588076
8.361928926
8.875017001
9.485544888
9.837137672 10.226228709
10.661332618 11.489211987 11.969020902 12.502494409
12.804226135 13.912944662 14.559623055 15.282004826
Euler semilinear method
h D 0:1
h D 0:05
h D 0:025
Exact"
2.000000000
2.000000000
2.000000000
2.000000000
2.467233571
2.465641081
2.464871435
2.464119569
3.036062650
3.032657307
3.031011316
3.029403212
3.729169725
3.723668026
3.721008466
3.718409925
4.574236356
4.566279470
4.562432696
4.558673929
5.605052990
5.594191643
5.588940276
5.583808754
6.862874116
6.848549921
6.841623814
6.834855438
8.398073101
8.379595572
8.370660695
8.361928926
10.272163096 10.248681420 10.237326199 10.226228709
12.560265110 12.530733531 12.516452106 12.502494409
15.354122287 15.317257705 15.299429421 15.282004826

x
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0

Section 3.1 Eulers Method

3.1.16.

x
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0

3.1.18.

x
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0

Eulers method
h D 0:2
h D 0:1
h D 0:05
Exact"
2.000000000 2.000000000 2.000000000 2.000000000
1.768294197 1.786514499 1.794412375 1.801636774
1.603028371 1.628427487 1.639678822 1.650102616
1.474580412 1.502563111 1.515157063 1.526935885
1.368349549 1.396853671 1.409839229 1.422074283
1.276424761 1.304504818 1.317421794 1.329664953
1.194247156 1.221490111 1.234122458 1.246155344
1.119088175 1.145348276 1.157607418 1.169334346
1.049284410 1.074553688 1.086419453 1.097812069
0.983821745 1.008162993 1.019652023 1.030719114
0.922094379 0.945604800 0.956752868 0.967523153
Euler semilinear method
h D 0:2
h D 0:1
h D 0:05
Exact"
2.000000000 2.000000000 2.000000000 2.000000000
1.806911831 1.804304958 1.802978526 1.801636774
1.659738603 1.654968381 1.652547436 1.650102616
1.540257861 1.533652916 1.530308405 1.526935885
1.438532932 1.430361800 1.426232584 1.422074283
1.348782285 1.339279577 1.334486249 1.329664953
1.267497415 1.256876924 1.251528766 1.246155344
1.192497494 1.180958765 1.175157264 1.169334346
1.122416379 1.110147777 1.103988310 1.097812069
1.056405906 1.043585743 1.037158237 1.030719114
0.993954754 0.980751307 0.974140320 0.967523153

x
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0

Eulers method
h D 0:2
h D 0:1
h D 0:05
Exact"
1.000000000 1.000000000 1.000000000 1.000000000
1.200000000 1.186557290 1.179206574 1.171515153
1.333543409 1.298441890 1.280865289 1.263370891
1.371340142 1.319698328 1.295082088 1.271251278
1.326367357 1.270160237 1.243958980 1.218901287
1.233056306 1.181845667 1.158064902 1.135362070
1.122359136 1.080477477 1.060871608 1.042062625
1.013100262 0.981124989 0.965917496 0.951192532
0.914000211 0.890759107 0.879460404 0.868381328
0.827848558 0.811673612 0.803582000 0.795518627
0.754572560 0.743869878 0.738303914 0.732638628
Euler semilinear method
h D 0:2
h D 0:1
h D 0:05
Exact"
1.000000000 1.000000000 1.000000000 1.000000000
1.153846154 1.162906599 1.167266650 1.171515153
1.236969953 1.250608357 1.257097924 1.263370891
1.244188456 1.258241892 1.264875987 1.271251278
1.195155456 1.207524076 1.213335781 1.218901287
1.115731189 1.125966437 1.130768614 1.135362070
1.025938754 1.034336918 1.038283392 1.042062625
0.937645707 0.944681597 0.948002346 0.951192532
0.856581823 0.862684171 0.865583126 0.868381328
0.784832910 0.790331183 0.792963532 0.795518627
0.722610454 0.727742966 0.730220211 0.732638628

x
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0

27

28 Chapter 3 Numerical Methods

3.1.20.

x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

3.1.22.

x
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0

x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

h D 0:1
1.000000000
0.700000000
0.498330000
0.356272689
0.254555443
0.181440541
0.128953069
0.091393543
0.064613612
0.045585102
0.032105117

h D 0:1
1.000000000
0.740818221
0.555889275
0.418936461
0.315890439
0.237908421
0.178842206
0.134165506
0.100450939
0.075073968
0.056020154

x
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0

Euler semilinear method


h D 0:05
h D 0:025
1.000000000 1.000000000
0.743784320 0.745143557
0.558989106 0.560410719
0.421482025 0.422642541
0.317804400 0.318668549
0.239287095 0.239902094
0.179812811 0.180239888
0.134840668 0.135133367
0.100918118 0.101117514
0.075396974 0.075532643
0.056243980 0.056336491

h D 0:1
1.000000000
1.000000000
1.020500000
1.053489840
1.093521685
1.137137554
1.182269005
1.227745005
1.272940309
1.317545833
1.361427907

h D 0:1
1.000000000
0.982476904
0.988105346
1.009495813
1.041012955
1.078631301
1.119632590
1.162270287
1.205472927
1.248613584
1.291345518

Eulers method
h D 0:05
h D 0:025
1.000000000 1.000000000
0.725841563 0.736671690
0.532982493 0.547988831
0.392592562 0.408724303
0.289040639 0.304708942
0.212387189 0.226758594
0.155687255 0.168375130
0.113851516 0.124744976
0.083076641 0.092230966
0.060505907 0.068068776
0.043997045 0.050159310

Exact"
1.000000000
0.746418339
0.561742917
0.423724207
0.319467408
0.240464879
0.180626161
0.135394692
0.101293057
0.075650324
0.056415515

Eulers method
h D 0:05
h D 0:025
1.000000000 1.000000000
1.005062500 1.007100815
1.026752091 1.029367367
1.059067423 1.061510137
1.097780573 1.099748225
1.140059654 1.141496651
1.184090031 1.185056276
1.228755801 1.229350441
1.273399187 1.273721920
1.317651554 1.317786528
1.361320824 1.361332589

Euler semilinear method


h D 0:05
h D 0:025
1.000000000 1.000000000
0.996114142 1.002608435
1.010577663 1.021306044
1.037358814 1.050731634
1.071994816 1.086964414
1.111346365 1.127262285
1.153300133 1.169781376
1.196488725 1.213325613
1.240060456 1.257146091
1.283506001 1.300791772
1.326535737 1.344004102

Exact"
1.000000000
0.746418339
0.561742917
0.423724207
0.319467408
0.240464879
0.180626161
0.135394692
0.101293057
0.075650324
0.056415515

Exact"
1.000000000
1.008899988
1.031723469
1.063764243
1.101614730
1.142903776
1.186038851
1.229985178
1.274092525
1.317967533
1.361383810

Exact"
1.000000000
1.008899988
1.031723469
1.063764243
1.101614730
1.142903776
1.186038851
1.229985178
1.274092525
1.317967533
1.361383810

Section 3.2 The Improved Euler Method and Related Methods


3.2 THE IMPROVED EULER METHOD AND RELATED METHODS
3.2.2. y1 D 1:220207973; y2 D 1:489578775 y3 D 1:819337186
3.2.4. y1 D 2:961317914; y2 D 2:920132727; y3 D 2:876213748.

3.2.6.

x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

3.2.8.

x
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.45
1.50

3.2.10.

x
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0

h D 0:1
2.000000000
2.257138644
2.826004666
3.812671926
5.387430580
7.813298361
11.489337756
17.015861211
25.292140630
37.662496723
56.134480009
h D 0:05
2.000000000
2.268367347
2.582607299
2.954510022
3.400161788
3.942097142
4.612879780
5.461348619
6.564150753
8.048579617
10.141969585

h D 0:05
2.000000000
2.238455342
2.786634110
3.747167263
5.285996803
7.660199197
11.260349005
16.674352914
24.783149862
36.903828191
55.003390448
h D 0:025
2.000000000
2.269670336
2.585911295
2.960870733
3.411212150
3.960434900
4.642784826
5.510188575
6.645334756
8.188335998
10.396770409

h D 0:025
2.000000000
2.234055168
2.777340360
3.731674025
5.261969043
7.623893064
11.206005869
16.593267820
24.662262731
36.723608928
54.734674836

Exact
2.000000000
2.232642918
2.774352565
3.726686582
5.254226636
7.612186259
11.188475269
16.567103199
24.623248150
36.665439956
54.647937102

h D 0:0125
2.000000000
2.270030868
2.586827341
2.962638822
3.414293964
3.965570792
4.651206769
5.524044591
6.668600859
8.228972215
10.472502111

h D 0:1
h D 0:05
h D 0:025
1.000000000 1.000000000 1.000000000
0.923734663 0.923730743 0.923730591
0.854475600 0.854449616 0.854444697
0.791650344 0.791596016 0.791584634
0.734785779 0.734703826 0.734686010
0.683424095 0.683318666 0.683295308
0.637097057 0.636973423 0.636945710
0.595330359 0.595193634 0.595162740
0.557658422 0.557513000 0.557479947
0.523638939 0.523488343 0.523453958
0.492862999 0.492709931 0.492674855
Approximate Solutions

Exact
2.000000000
2.270158103
2.587150838
2.963263785
3.415384615
3.967391304
4.654198473
5.528980892
6.676923077
8.243593315
10.500000000

h D 0:1
0.00000
0.00004
0.00035
0.00078
0.00125
0.00171
0.00213
0.00250
0.00283
0.00311
0.00335

h D 0:05 h D 0:025
0.000000
0.000000
0.000001
-0.000001
0.000068
0.000015
0.000167
0.000039
0.000277
0.000065
0.000384
0.000091
0.000483
0.000115
0.000572
0.000137
0.000650
0.000156
0.000718
0.000173
0.000777
0.000187
Residuals

29

30 Chapter 3 Numerical Methods

3.2.12.

3.2.14.

x
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0

3.2.16.

x
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0

h D 0:1
0.000000000
-0.099500000
-0.196990313
-0.290552949
-0.378532718
-0.459672297
-0.533180153
-0.598726853
-0.656384109
-0.706530934
-0.749751364

h D 0:05
0.000000000
-0.099623114
-0.197235180
-0.290917718
-0.379013852
-0.460262848
-0.533868468
-0.599496413
-0.657214624
-0.707400266
-0.750637632

h D 0:025
0.000000000
-0.099653809
-0.197295585
-0.291006784
-0.379130237
-0.460404546
-0.534032512
-0.599678824
-0.657410640
-0.707604759
-0.750845571

Exact"
0.000000000
-0.099664000
-0.197315517
-0.291036003
-0.379168221
-0.460450590
-0.534085626
-0.599737720
-0.657473792
-0.707670533
-0.750912371

Improved Euler method


h D 0:1
h D 0:05
h D 0:025
Exact"
2.000000000
2.000000000
2.000000000
2.000000000
2.461242144
2.463344439
2.463918368
2.464119569
3.022367633
3.027507237
3.028911026
3.029403212
3.705511610
3.714932709
3.717507170
3.718409925
4.537659565
4.553006531
4.557202414
4.558673929
5.551716960
5.575150456
5.581560437
5.583808754
6.787813853
6.822158665
6.831558101
6.834855438
8.294896222
8.343829180
8.357227947
8.361928926
10.132667135 10.200955596 10.219663917 10.226228709
12.373954732 12.467758807 12.493470722 12.502494409
15.107600968 15.234856000 15.269755072 15.282004826
Improved Euler semilinear method
h D 0:1
h D 0:05
h D 0:025
Exact"
2.000000000
2.000000000
2.000000000
2.000000000
2.464261688
2.464155139
2.464128464
2.464119569
3.029706047
3.029479005
3.029422165
3.029403212
3.718897663
3.718531995
3.718440451
3.718409925
4.559377397
4.558849990
4.558717956
4.558673929
5.584766724
5.584048510
5.583868709
5.583808754
6.836116246
6.835170986
6.834934347
6.834855438
8.363552464
8.362335253
8.362030535
8.361928926
10.228288880 10.226744312 10.226357645 10.226228709
12.505082132 12.503142042 12.502656361 12.502494409
15.285231726 15.282812424 15.282206780 15.282004826

x
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0

x
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0

h D 0:2
2.000000000
1.801514185
1.649911580
1.526711768
1.421841570
1.329441172
1.245953205
1.169162994
1.097677870
1.030626179
0.967473721

Improved Euler method


h D 0:1
h D 0:05
2.000000000 2.000000000
1.801606135 1.801629115
1.650054870 1.650090680
1.526879870 1.526921882
1.422016119 1.422059743
1.329609020 1.329650971
1.246104819 1.246142713
1.169291515 1.169323639
1.097778523 1.097803683
1.030695880 1.030713305
0.967510790 0.967520062

Exact"
2.000000000
1.801636774
1.650102616
1.526935885
1.422074283
1.329664953
1.246155344
1.169334346
1.097812069
1.030719114
0.967523153

Section 3.2 The Improved Euler Method and Related Methods

x
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0

3.2.18.

x
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0

3.2.20.

Improved Euler semilinear method


h D 0:2
h D 0:1
h D 0:05
2.000000000 2.000000000 2.000000000
1.801514185 1.801606135 1.801629115
1.649911580 1.650054870 1.650090680
1.526711768 1.526879870 1.526921882
1.421841570 1.422016119 1.422059743
1.329441172 1.329609020 1.329650971
1.245953205 1.246104819 1.246142713
1.169162994 1.169291515 1.169323639
1.097677870 1.097778523 1.097803683
1.030626179 1.030695880 1.030713305
0.967473721 0.967510790 0.967520062

Exact"
2.000000000
1.801636774
1.650102616
1.526935885
1.422074283
1.329664953
1.246155344
1.169334346
1.097812069
1.030719114
0.967523153

Improved Euler method


h D 0:2
h D 0:1
h D 0:05
Exact"
1.000000000 1.000000000 1.000000000 1.000000000
1.166771705 1.170394902 1.171244037 1.171515153
1.255835116 1.261642355 1.262958788 1.263370891
1.263517157 1.269528214 1.270846761 1.271251278
1.212551997 1.217531648 1.218585457 1.218901287
1.130812573 1.134420589 1.135150284 1.135362070
1.039104333 1.041487727 1.041938536 1.042062625
0.949440052 0.950888923 0.951132561 0.951192532
0.867475787 0.868263999 0.868364849 0.868381328
0.795183973 0.795523696 0.795530315 0.795518627
0.732679223 0.732721613 0.732667905 0.732638628
Improved Euler semilinear method
h D 0:2
h D 0:1
h D 0:05
Exact"
1.000000000 1.000000000 1.000000000 1.000000000
1.170617859 1.171292452 1.171459576 1.171515153
1.261629934 1.262938347 1.263262919 1.263370891
1.269173253 1.270734290 1.271122186 1.271251278
1.216926014 1.218409355 1.218778420 1.218901287
1.133688235 1.134944960 1.135257876 1.135362070
1.040721691 1.041728386 1.041979126 1.042062625
0.950145706 0.950931597 0.951127345 0.951192532
0.867573431 0.868179975 0.868331028 0.868381328
0.794899034 0.795364245 0.795480063 0.795518627
0.732166678 0.732521078 0.732609267 0.732638628

x
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0

x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

h D 0:1
1.000000000
0.749165000
0.565942699
0.428618351
0.324556426
0.245417735
0.185235654
0.139546094
0.104938506
0.078787731
0.059071894

Improved Euler method


h D 0:05
h D 0:025
1.000000000 1.000000000
0.747022742 0.746561141
0.562667885 0.561961242
0.424803657 0.423978964
0.320590918 0.319732571
0.241558658 0.240723019
0.181643813 0.180866303
0.136310496 0.135610749
0.102096319 0.101482503
0.076340645 0.075813072
0.056999028 0.056553023

Exact"
1.000000000
0.746418339
0.561742917
0.423724207
0.319467408
0.240464879
0.180626161
0.135394692
0.101293057
0.075650324
0.056415515

31

32 Chapter 3 Numerical Methods

x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

3.2.22.

x
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0

3.2.24.

Improved Euler semilinear method


h D 0:1
h D 0:05
h D 0:025
1.000000000 1.000000000 1.000000000
0.745595127 0.746215164 0.746368056
0.560827568 0.561515647 0.561686492
0.422922083 0.423524585 0.423674586
0.318820339 0.319306259 0.319427337
0.239962317 0.240339716 0.240433757
0.180243441 0.180530866 0.180602470
0.135106416 0.135322934 0.135376855
0.101077312 0.101239368 0.101279714
0.075489492 0.075610310 0.075640381
0.056295914 0.056385765 0.056408124

x
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0

h D 0:1
1.000000000
1.010250000
1.033547273
1.065562151
1.103145347
1.144085693
1.186878796
1.230530804
1.274404357
1.318104153
1.361395309

Improved Euler method


h D 0:05
h D 0:025
1.000000000 1.000000000
1.009185754 1.008965733
1.032105322 1.031811002
1.064135919 1.063849094
1.101926450 1.101685553
1.143140125 1.142957158
1.186202854 1.186075600
1.230088035 1.230007943
1.274147657 1.274104430
1.317987551 1.317971490
1.361379259 1.361382239

Improved Euler semilinear method


h D 0:1
h D 0:05
h D 0:025
1.000000000 1.000000000 1.000000000
1.012802674 1.009822081 1.009124116
1.038431870 1.033307426 1.032108359
1.072484834 1.065821457 1.064263950
1.111794329 1.104013534 1.102197168
1.154168041 1.145554968 1.143547198
1.198140189 1.188883373 1.186728849
1.242762459 1.232984559 1.230712361
1.287441845 1.277221941 1.274850828
1.331821976 1.321210992 1.318753047
1.375699933 1.364730937 1.362193997
x
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0

h D 0:1
1.000000000
1.151019287
1.238798618
1.289296258
1.317686801
1.333073855
1.341027170
1.345001345
1.347155352
1.348839325
1.350890736

Exact"
1.000000000
0.746418339
0.561742917
0.423724207
0.319467408
0.240464879
0.180626161
0.135394692
0.101293057
0.075650324
0.056415515

h D 0:05
1.000000000
1.153270661
1.241884421
1.292573128
1.320866599
1.336036248
1.343732006
1.347446389
1.349355473
1.350816158
1.352667599

Exact"
1.000000000
1.008899988
1.031723469
1.063764243
1.101614730
1.142903776
1.186038851
1.229985178
1.274092525
1.317967533
1.361383810

Exact"
1.000000000
1.008899988
1.031723469
1.063764243
1.101614730
1.142903776
1.186038851
1.229985178
1.274092525
1.317967533
1.361383810

h D 0:025
1.000000000
1.153777957
1.242580821
1.293313355
1.321585242
1.336705820
1.344343232
1.347998652
1.349852082
1.351261995
1.353067951

Exact
1.000000000
1.153937085
1.242799540
1.293546032
1.321811247
1.336916440
1.344535503
1.348172348
1.350008229
1.351402121
1.353193719

Section 3.2 The Improved Euler Method and Related Methods

3.2.26.

x
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.45
1.50

h D 0:05
2.000000000
2.268496358
2.582897367
2.954995034
3.400872342
3.943047906
4.614039436
5.462568051
6.564985580
8.047824947
10.136329642

3.2.28.

x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

h D 0:1
1.000000000
0.984142840
0.965066124
0.942648578
0.916705578
0.886970525
0.853066054
0.814458249
0.770380571
0.719699643
0.660658411

3.2.30.

x
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0

h D 0:1
0.000000000
-0.099666667
-0.197322275
-0.291033227
-0.379131069
-0.460350276
-0.533897316
-0.599446325
-0.657076288
-0.707175010
-0.750335016

h D 0:025
2.000000000
2.269703943
2.585985695
2.960992388
3.411384294
3.960651794
4.643018510
5.510357362
6.645224236
8.187384679
10.393419681
h D 0:05
1.000000000
0.984133302
0.965044455
0.942611457
0.916648569
0.886887464
0.852948011
0.814291679
0.770143777
0.719355385
0.660136630
h D 0:05
0.000000000
-0.099665005
-0.197317894
-0.291036361
-0.379160444
-0.460427667
-0.534041581
-0.599668984
-0.657379719
-0.707553135
-0.750775571

h D 0:0125
2.000000000
2.270043628
2.586855275
2.962683751
3.414355862
3.965644965
4.651277424
5.524069547
6.668472955
8.228413044
10.470731411

h D 0:025
1.000000000
0.984130961
0.965039117
0.942602279
0.916634423
0.886866778
0.852918497
0.814249848
0.770083998
0.719267905
0.660002840
h D 0:025
0.000000000
-0.099664307
-0.197316222
-0.291036258
-0.379166504
-0.460445166
-0.534075026
-0.599721072
-0.657450947
-0.707641993
-0.750879100

33

Exact
2.000000000
2.270158103
2.587150838
2.963263785
3.415384615
3.967391304
4.654198473
5.528980892
6.676923077
8.243593315
10.500000000

Exact"
1.000000000
0.984130189
0.965037353
0.942599241
0.916629732
0.886859904
0.852908668
0.814235883
0.770063987
0.719238519
0.659957689
Exact"
0.000000000
-0.099674132
-0.197355914
-0.291123993
-0.379315647
-0.460662347
-0.534359685
-0.600066382
-0.657845646
-0.708072516
-0.751331499

3.2.32. (a) Let xi D a C ih, i D 0; 1; : : : ; n. If y is the solution of the initial value problem
Rb
y 0 D f .x/, y.a/ D 0, then y.b/ D a f .x/ dx. The improved Euler method yields yi C1 D yi C
:5h .f .a C ih/ C f .a C .i C 1/h//, i D 0; 1; : : : ; n 1, where y0 D a and yn is an approximation to
Z b
f .x/ dx. But
a

yn D

n
X1
i D0

.yi C1

yi / D :5h .f .a/ C f .b// C h

n 1
X
i D1

f .a C ih/:

(c) The local truncation error is a multiple of y 000 .xQ i / D f 00 .xQ i /, where xi < xQ i < xi C1 . Therefore,
the quadrature formula is exact if f is a polynomial of degree < 2.
Z b
(d) Let E.f / D
f .x/ dx yn . Note that E is linear. If f is a polynomial of degree 2, then
a

34 Chapter 3 Numerical Methods


f .x/ D f0 .x/ C K.x
a/2 /

E..x

a/h2

K.b

E.f / D

a/2 where deg.f0 /  1. Since E.f0 / D 0 from (c) and


D

.b

a/3

3
 3
n
D h3
3

a/2 h
2

.b
n2
2

n.n

h3

n
X1

i2

i D1

1/.2n
6

1/

nh3
D
6

.b

a/h2
;
6

; therefore the error is proportional to h2 .

3.3 THE RUNGEKUTTA METHOD


3.3.2. y1 D 1:221551366; y2 D 1:492920208
3.3.4. y1 D 2:961316248; y2 D 2:920128958.

3.3.6.

x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

3.3.8.

x
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.45
1.50

3.3.10.

x
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0

h D 0:1
2.000000000
2.232752507
2.774582759
3.727068686
5.254817388
7.613077020
11.189806778
16.569088310
24.626206255
36.669848687
54.654509699
h D 0:05
2.000000000
2.270153785
2.587139846
2.963242415
3.415346864
3.967327077
4.654089950
5.528794615
6.676590929
8.242960669
10.498658198

h D 0:05
2.000000000
2.232649573
2.774366625
3.726710028
5.254263005
7.612241222
11.188557546
16.567225975
24.623431201
36.665712858
54.648344019
h D 0:025
2.000000000
2.270157806
2.587150083
2.963262317
3.415382020
3.967386886
4.654191000
5.528968045
6.676900116
8.243549415
10.499906266

h D 0:025
2.000000000
2.232643327
2.774353431
3.726688030
5.254228886
7.612189662
11.188480365
16.567110808
24.623259496
36.665456874
54.647962328

Exact
2.000000000
2.232642918
2.774352565
3.726686582
5.254226636
7.612186259
11.188475269
16.567103199
24.623248150
36.665439956
54.647937102

h D 0:0125
2.000000000
2.270158083
2.587150789
2.963263689
3.415384445
3.967391015
4.654197983
5.528980049
6.676921569
8.243590428
10.499993820

h D 0:1
h D 0:05
h D 0:025
1.000000000 1.000000000 1.000000000
0.923730622 0.923730677 0.923730681
0.854443253 0.854443324 0.854443328
0.791581155 0.791581218 0.791581222
0.734680497 0.734680538 0.734680541
0.683288034 0.683288051 0.683288052
0.636937046 0.636937040 0.636937040
0.595153053 0.595153029 0.595153028
0.557469558 0.557469522 0.557469520
0.523443129 0.523443084 0.523443081
0.492663789 0.492663738 0.492663736
Approximate Solutions

Exact
2.000000000
2.270158103
2.587150838
2.963263785
3.415384615
3.967391304
4.654198473
5.528980892
6.676923077
8.243593315
10.500000000

h D 0:1
0.000000000
-0.000000608
-0.000000819
-0.000000753
-0.000000523
-0.000000224
0.000000079
0.000000351
0.000000578
0.000000760
0.000000902

h D 0:05
0.0000000000
-0.0000000389
-0.0000000529
-0.0000000495
-0.0000000359
-0.0000000178
0.0000000006
0.0000000171
0.0000000309
0.0000000421
0.0000000508
Residuals

h D 0:025
0.00000000000
-0.00000000245
-0.00000000335
-0.00000000316
-0.00000000233
-0.00000000122
-0.00000000009
0.00000000093
0.00000000179
0.00000000248
0.00000000302

Section 3.3 The RungeKutta Method

3.3.12.

3.3.14.

x
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0

3.3.16.

x
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0

h D 0:1
0.000000000
-0.099663901
-0.197315322
-0.291035700
-0.379167790
-0.460450005
-0.534084875
-0.599736802
-0.657472724
-0.707669346
-0.750911103

h D 0:05
0.000000000
-0.099663994
-0.197315504
-0.291035983
-0.379168194
-0.460450552
-0.534085579
-0.599737663
-0.657473726
-0.707670460
-0.750912294

h D 0:025
0.000000000
-0.099664000
-0.197315516
-0.291036001
-0.379168220
-0.460450587
-0.534085623
-0.599737717
-0.657473788
-0.707670529
-0.750912367

Exact"
0.000000000
-0.099664000
-0.197315517
-0.291036003
-0.379168221
-0.460450590
-0.534085626
-0.599737720
-0.657473792
-0.707670533
-0.750912371

RungeKutta method
h D 0:1
h D 0:05
h D 0:025
Exact"
2.000000000
2.000000000
2.000000000
2.000000000
2.464113907
2.464119185
2.464119544
2.464119569
3.029389360
3.029402271
3.029403150
3.029403212
3.718384519
3.718408199
3.718409812
3.718409925
4.558632516
4.558671116
4.558673746
4.558673929
5.583745479
5.583804456
5.583808474
5.583808754
6.834762639
6.834849135
6.834855028
6.834855438
8.361796619
8.361919939
8.361928340
8.361928926
10.226043942 10.226216159 10.226227891 10.226228709
12.502240429 12.502477158 12.502493285 12.502494409
15.281660036 15.281981407 15.282003300 15.282004826
RungeKutta semilinear method
h D 0:1
h D 0:05
h D 0:025
Exact"
2.000000000
2.000000000
2.000000000
2.000000000
2.464119623
2.464119573
2.464119570
2.464119569
3.029403325
3.029403219
3.029403212
3.029403212
3.718410105
3.718409936
3.718409925
3.718409925
4.558674188
4.558673945
4.558673930
4.558673929
5.583809105
5.583808776
5.583808755
5.583808754
6.834855899
6.834855467
6.834855440
6.834855438
8.361929516
8.361928963
8.361928928
8.361928926
10.226229456 10.226228756 10.226228712 10.226228709
12.502495345 12.502494468 12.502494413 12.502494409
15.282005990 15.282004899 15.282004831 15.282004826

x
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0

x
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0

h D 0:2
2.000000000
1.801636785
1.650102633
1.526935904
1.422074302
1.329664970
1.246155357
1.169334355
1.097812074
1.030719113
0.967523147

RungeKutta method
h D 0:1
h D 0:05
2.000000000 2.000000000
1.801636775 1.801636774
1.650102617 1.650102616
1.526935886 1.526935885
1.422074284 1.422074283
1.329664954 1.329664953
1.246155345 1.246155344
1.169334347 1.169334346
1.097812070 1.097812069
1.030719114 1.030719114
0.967523152 0.967523153

Exact"
2.000000000
1.801636774
1.650102616
1.526935885
1.422074283
1.329664953
1.246155344
1.169334346
1.097812069
1.030719114
0.967523153

35

36 Chapter 3 Numerical Methods

x
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0

3.3.18.

x
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0

3.3.20.

RungeKutta semilinear method


h D 0:2
h D 0:1
h D 0:05
2.000000000 2.000000000 2.000000000
1.801636785 1.801636775 1.801636774
1.650102633 1.650102617 1.650102616
1.526935904 1.526935886 1.526935885
1.422074302 1.422074284 1.422074283
1.329664970 1.329664954 1.329664953
1.246155357 1.246155345 1.246155344
1.169334355 1.169334347 1.169334346
1.097812074 1.097812070 1.097812069
1.030719113 1.030719114 1.030719114
0.967523147 0.967523152 0.967523153

Exact"
2.000000000
1.801636774
1.650102616
1.526935885
1.422074283
1.329664953
1.246155344
1.169334346
1.097812069
1.030719114
0.967523153

RungeKutta method
h D 0:2
h D 0:1
h D 0:05
Exact"
1.000000000 1.000000000 1.000000000 1.000000000
1.171515610 1.171515156 1.171515152 1.171515153
1.263365845 1.263370556 1.263370869 1.263370891
1.271238957 1.271250529 1.271251232 1.271251278
1.218885528 1.218900353 1.218901230 1.218901287
1.135346772 1.135361174 1.135362016 1.135362070
1.042049558 1.042061864 1.042062579 1.042062625
0.951181964 0.951191920 0.951192495 0.951192532
0.868372923 0.868380842 0.868381298 0.868381328
0.795511927 0.795518241 0.795518603 0.795518627
0.732633229 0.732638318 0.732638609 0.732638628
RungeKutta semilinear method
h D 0:2
h D 0:1
h D 0:05
Exact"
1.000000000 1.000000000 1.000000000 1.000000000
1.171517316 1.171515284 1.171515161 1.171515153
1.263374485 1.263371110 1.263370904 1.263370891
1.271254636 1.271251485 1.271251291 1.271251278
1.218903802 1.218901442 1.218901297 1.218901287
1.135363869 1.135362181 1.135362077 1.135362070
1.042063952 1.042062706 1.042062630 1.042062625
0.951193560 0.951192595 0.951192536 0.951192532
0.868382157 0.868381378 0.868381331 0.868381328
0.795519315 0.795518669 0.795518629 0.795518627
0.732639212 0.732638663 0.732638630 0.732638628

x
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0

x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

h D 0:1
1.000000000
0.746430962
0.561761987
0.423746057
0.319489811
0.240486460
0.180646105
0.135412569
0.101308709
0.075663769
0.056426886

RungeKutta method
h D 0:05
h D 0:025
1.000000000 1.000000000
0.746418992 0.746418376
0.561743921 0.561742975
0.423725371 0.423724274
0.319468612 0.319467478
0.240466046 0.240464947
0.180627244 0.180626225
0.135395665 0.135394749
0.101293911 0.101293107
0.075651059 0.075650367
0.056416137 0.056415552

Exact"
1.000000000
0.746418339
0.561742917
0.423724207
0.319467408
0.240464879
0.180626161
0.135394692
0.101293057
0.075650324
0.056415515

Section 3.3 The RungeKutta Method

x
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

3.3.22.

x
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0

3.3.24.

RungeKutta semilinear method


h D 0:1
h D 0:05
h D 0:025
1.000000000 1.000000000 1.000000000
0.746416306 0.746418217 0.746418332
0.561740647 0.561742780 0.561742908
0.423722193 0.423724084 0.423724199
0.319465760 0.319467308 0.319467402
0.240463579 0.240464800 0.240464874
0.180625156 0.180626100 0.180626158
0.135393924 0.135394645 0.135394689
0.101292474 0.101293021 0.101293055
0.075649884 0.075650297 0.075650322
0.056415185 0.056415495 0.056415514

x
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3.0

h D 0:1
1.000000000
1.008912398
1.031740789
1.063781819
1.101630085
1.142915917
1.186047678
1.229991054
1.274095992
1.317969153
1.361384082

RungeKutta method
h D 0:05
h D 0:025
1.000000000 1.000000000
1.008900636 1.008900025
1.031724368 1.031723520
1.063765150 1.063764295
1.101615517 1.101614774
1.142904393 1.142903811
1.186039295 1.186038876
1.229985469 1.229985194
1.274092692 1.274092535
1.317967605 1.317967537
1.361383812 1.361383809

RungeKutta semilinear method


h D 0:1
h D 0:05
h D 0:025
1.000000000 1.000000000 1.000000000
1.008913934 1.008900843 1.008900041
1.031748526 1.031725001 1.031723564
1.063798300 1.063766321 1.063764371
1.101656264 1.101617259 1.101614886
1.142951721 1.142906691 1.142903955
1.186092475 1.186042105 1.186039051
1.230043983 1.229988742 1.229985397
1.274156172 1.274096377 1.274092762
1.318035787 1.317971658 1.317967787
1.361456502 1.361388196 1.361384079
x
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1.70
1.80
1.90
2.00

h D :1
0.142854841
0.053340745
-0.046154629
-0.153363206
-0.266397049
-0.383721107
-0.504109696
-0.626598326
-0.750437351
-0.875050587
-1.000000000

Exact"
1.000000000
0.746418339
0.561742917
0.423724207
0.319467408
0.240464879
0.180626161
0.135394692
0.101293057
0.075650324
0.056415515

h D :05
0.142857001
0.053341989
-0.046153895
-0.153362764
-0.266396779
-0.383720941
-0.504109596
-0.626598268
-0.750437320
-0.875050574
-1.000000000

Exact"
1.000000000
1.008899988
1.031723469
1.063764243
1.101614730
1.142903776
1.186038851
1.229985178
1.274092525
1.317967533
1.361383810

Exact"
1.000000000
1.008899988
1.031723469
1.063764243
1.101614730
1.142903776
1.186038851
1.229985178
1.274092525
1.317967533
1.361383810

h D :025
0.142857134
0.053342066
-0.046153849
-0.153362736
-0.266396762
-0.383720931
-0.504109589
-0.626598264
-0.750437318
-0.875050573
-1.000000000

Exact
0.142857143
0.053342071
-0.046153846
-0.153362734
-0.266396761
-0.383720930
-0.504109589
-0.626598264
-0.750437318
-0.875050573
-1.000000000

37

38 Chapter 3 Numerical Methods


x
0.50
0.60
0.70
0.80
0.90
1.00
1.10
1.20
1.30
1.40
1.50

3.3.26.

h D :1
-8.954103230
-5.059648314
-2.516755942
-0.752508672
0.530528482
1.500000000
2.256519743
2.863543039
3.362731379
3.782361948
4.142171279

h D :05
-8.954063245
-5.059633293
-2.516749850
-0.752506238
0.530529270
1.500000000
2.256519352
2.863542454
3.362730700
3.782361231
4.142170553

h D :025
-8.954060698
-5.059632341
-2.516749465
-0.752506084
0.530529319
1.500000000
2.256519328
2.863542417
3.362730658
3.782361186
4.142170508

Exact
-8.954060528
-5.059632277
-2.516749439
-0.752506074
0.530529323
1.500000000
2.256519326
2.863542415
3.362730655
3.782361183
4.142170505

3.3.28. (a) Let xi D a C ih, i D 0; 1; : : : ; n. If y is the solution of the initial value problem
Rb
y 0 D f .x/, y.a/ D 0, then y.b/ D a f .x/ dx. The Runge-Kutta method yields yi C1 D yi C
h
.f .a C ih/ C 4f .a C .2i C 1/h=2/ C f .a C .i C 1/h//, i D 0; 1; : : : ; n 1, where y0 D a and yn
6
Z
b

f .x/ dx. But

is an approximation to

yn D

n 1
X
i D0

.yi C1

yi / D

n 1

i D1

i D1

hX
2h X
h
.f .a/ C f .b// C
f .a C ih/ C
f .a C .2i
6
3
3

1/h=2/ :

(c) The local truncation error is a multiple of y .5/ .xQ i / D f .4/ .xQ i /, where xi < xQ i < xi C1 . Therefore,
the quadrature formula is exact if f is a polynomial of degree < 4.
Z b
(d) Let E.f / D
f .x/ dx yn . Note that E is linear. If f is a polynomial of degree 4, then
a

f .x/ D f0 .x/ C K.x


E..x

E.f / D

.b

a/4 /

a/4 where deg.f0 /  3 and K is constant. Since E.f0 / D 0 from (c) and

n 1
n
a/4 h h5 X 4 2h5 X
i
.i 1=2/4
5
6
3
3
i D1
i D1
 5
  5

 5
4
4
3
n
n
n
n
n
2n
n3
7n
5 n
D h
C
C
5
6
15
6
9
90
15
9
360
5
4
nh
.b a/h
D
D
;
120
120

.b

a/5

.b

a/h4
; thus, the error is proportional to h4 .
120

CHAPTER 4
Applications of First Order Equations

4.1 GROWTH AND DECAY


ln 2
4.1.2. k D ln 2 and  D 2 ) k D
; Q.t/ D Q0 e
2
T
ln
2
2
ln
10
Q0 e T ln 2=2 ; ln 10 D
;T D
days.
2
ln 2

t ln 2=2

; if Q.T / D

4.1.4. Let t1 be the elapsed


 time since the tree died. Since p.t/ D e
p1
t1
ln.p0 =p1 /
.t1 ln 2/=
p0 e
, so ln
D
ln 2 and t1 D 
.
p0

ln 2
4.1.6. Q D Q0 e k t ; Q1 D Q0 e


1
Q1
kD
ln
.
t2 t1
Q2

k t1

; Q2 D Q0 e

k t2

Q2
D e
Q1

.t ln 2/

k.t2 t1 /

Q0
Q0
, then
D
10
10

, it follows that p1 D

; ln

Q1
Q2

D k.t2

t1 /;

4.1.8. Q0 D :06Q; Q.0/ D Q0 ; Q D Q0 e :06t . We must find  such that Q./ D 2Q0 ; that is,
ln 2
50 ln 2
Q0 e :06 D 2Q0 , so :06 D ln 2 and  D
D
yr.
:06
3
4.1.10. (a) If T is the time to triple the value, then Q.T / D Q0 e :05T D 3Q0 , so e :05T D 3. Therefore,
:05T D ln 3 and T D 20 ln 3.
(b) If Q.10/ D 100000, then Q0 e :5 D 100000, so Q0 D 100000e :5
Q2
Q0
1
1
t
1 1
; Q.0/ D 50;
D
;
D
C c; Q.0/ D 50 ) c D
;
D
2
Q2
2 Q
2
50 Q
t
1
1 C 25t
50
1
C
D
; QD
. Now Q.T / D 25 ) 1 C 25T D 2 ) 25T D 1 ) T D
years.
2 50
50
1 C 25t
25

4.1.12. Q0 D

ln 2
3Q0
4.1.14. Since  D 1500, k D
; hence Q D Q0 e .t ln 2/=1500 . If Q.t1 / D
, then e .t1 ln 2/=1500 D
1500
4

 
 
ln 43
3
ln 2
3
4
4
; t1
D ln
D ln
; t1 D 1500
. Finally, Q.2000/ D Q0 e 3 ln 2 D 2 4=3 Q0 .
4
1500
4
3
ln 2

39

40 Chapter 4 Applications of First Order Equations


S
S
; S.0/ D 20. Rewrite the differential equation in (A) as (B) S 0 C
D 1. Since
10
10
S1 D e t =10 is a solution of the complementary equation, the solutions of (B) are given by S D ue t =10,
where u0 e t =10 D 1. Therefore, u0 D e t =10; u D 10e t =10 C c; S D 10 C ce t =10. Now S.0/ D 20 )
c D 10, so S D 10 C 10e t =10 and limt !1 S.t/ D 10 g.
4.1.16. (A) S 0 D 1

V
4.1.18. (A) V 0 D 750 C
; V .0/ D 25000. Rewrite the differential equation in (A) as (B) V 0
20
V
D 750. Since V1 D e t =20 is a solution of the complementary equation, the solutions of (B) are
20
given by V D ue t =20, where u0 e t =20 D 750. Therefore, u0 D 750e t =20; u D 15000e t =20 C c;
V D 15000 C ce t =20; V .0/ D 25000 ) c D 10000. Therefore, V D 15000 C 10000e t =20.




p p2
1
p0
1 1
1
1 0
1
1
1 0
0
D
p.p 4/;
D
;
p D
;
p D
4.1.20. p D
8
p.p 4/
8 4 p 4 p
8 p 4 p

2 8
1 p 4
t
p 4
24 p 4
24 t =2
;
D
C k;
D ce t =2; p.0/ D 100 ) c D
;
D
e
; p 4D
2
p 
2
25
p
25
 p
24
24 t =2
4
100
pe t =2; p 1
e
D 4; p D
D
.
24
t
=2
25
25
24 24e t =2
1 25 e
4.1.22. (a)P 0 D rP

12M .

12M
12M r t
(b) P D ue ; u e D 12M ; u0 D 12Me r t ; u D
e
C c; P D
C ce r t ; P .0/ D
r
r
12M
12M
P0 ) c D P0
;P D
.1 e r t / C P0 e r t .
r
r
rP0
(c) Since P .N / D 0, the answer to (b) implies that M D
12.1 e rN /
rt

0 rt

4.1.24. The researchers salary is the solution of the initial value problem S 0 D aS; S.0/ D S0 .
Therefore, S D S0 e at . If P D P .t/ is the value of the trust fund, then P 0 D S0 e at C rP , or
P 0 rP D S0 e at . Therefore, (A) P D ue r t , where u0 e r t D S0 e at , so (B) u0 D S0 e .a r /t . If
S0 at
S0 .a r /t
a r , then (B) implies that u D
e
C c, so (A) implies that P D
e C ce r t . Now
r a
r
a


S0 at
S0
S0
; therefore P D
e C P0
e r t . We must choose
P .0/ D P0 ) c D P0
r a
r
a
r
a


S0 aT
S0
P0 so that P .T / D 0; that is, P D
e C P0
e rT D 0. Solving this for P0 yields
r a
r a
S0 .1 e .a r /T /
P0 D
. If a D r , then (B) becomes u0 D S0 , so u D S0 t C c and (A) implies that
r a
P D . S0 t C c/e r t . Now P .0/ D P0 ) c D P0 ; therefore P D . S0 t C P0 /e r t . To make P .T / D 0
we must take P0 D S0 T .
4.1.26. Q0 D

at
1 C btQ2

kQ; limt !1 Q.t/ D .a=bk/1=3 .

4.2 COOLING AND MIXING


4.2.2. Since T0 D 100 and TM D 10, T D 10 C 110e k t . Now T .1/ D 80 ) 80 D 10 C 110e
9
11
11
so e k D
and k D ln . Therefore, T D 10 C 110e t ln 9 .
11
9

4.2.4. Let T be the thermometer reading. Since T0 D 212 and TM D 70, T D 70 C 142e k t . Now
55
1 142
T .2/ D 125 ) 125 D 70 C 142e 2k , so e 2k D
and k D ln
. Therefore, (A) T D
142
2
55

Section 4.2 Cooling and Mixing


70 C 142e

t
2

ln

142
55

(a) T .2/ D 70 C 142e

2 ln

142
55

55
D 70 C 142
142

2

41

 91:30F.

1

142

142
(b) Let  be the time when T ./ D 72, so 72 D 70 C 142e 2 ln 55 , or e 2 ln 55 D
. Therefore,
71
ln 71
 D 2 142  8:99 min.
ln 55
(c) Since (A) implies that T > 70 for all t > 0, the thermometer will never read 69 F.
4.2.6. Since TM D 20, T D 20 C .T0 20/e k t . Now T0 5 D 20 C .T0 20/e 4k and T0 7 D




T0 25
T0 27
T0 27
T0 25 2
D e 4k and
D e 8k , so
D
,
20 C .T0 20/e 8k . Therefore,
T0 20
T0 20
T0 20
T0 20
which implies that .T0 20/.T0 27/ D .T0 25/2 , or T02 47T0 C 540 D T02 50T0 C 625; hence
3T0 D 85 and T0 D .85=3/ C .
3
3
Q; Q.0/ D 0. Rewrite the differential equation as (A) Q0 C
Q D 3. Since
40
40
Q1 D e 3t =40 is a solution of the complementary equation, the solutions of (A) are given by Q D
ue 3t =40 where u0 e 3t =40 D 3. Therefore,u0 D 3e 3t =40, u D 40e 3t =40 C c, and Q D 40 C ce 3t =40.
Now Q.0/ D 0 ) c D 40, so Q D 40.1 e 3t =40/.
4.2.8. Q0 D 3

3
Q
Q
3
; Q.0/ D 10. Rewrite the differential equation as (A) Q0 C
D . Since
2
20
20
2
Q1 D e t =20 is a solution of the complementary equation, the solutions of (A) are given by Q D ue t =20
3
3
where u0 e t =20 D . Therefore,u0 D e t =20, u D 30e t =20 C c, and Q D 30 C ce t =20. Now
2
2
Q
Q.0/ D 10 ) c D 20, so Q D 30 20e t =20 and K D
D :3 :2e t =20.
100

4.2.10. Q0 D

Q
Q
, or (A) Q0 C
D 10. Since Q1 D e t =5 is a solution of the complementary
5
5
equation, the solutions of (A) are given by Q D ue t =5 where u0 e t =5 D 10. Therefore,u0 D 10e t =5,
u D 50e t =10 C c, and Q D 50 C ce t =5. Since limt !1 Q.t/ D 50, the mininum capacity is 50 gallons.

4.2.12. Q0 D 10

4.2.14. Since there are 2t C 600 gallons of mixture in the tank at time t and mixture is being drained
2
at 4 gallons/min, Q0 D 3
Q; Q.0/ D 40. Rewrite the differential equation as (A) Q0 C
t C 300
2
1
Q D 3. Since Q1 D
is a solution of the complementary equation, the solutions
t C 300
.t C 300/2
u
u0
where
D 3. Therefore, u0 D 3.t C 300/2 , u D
of (A) are given by Q D
.t C 300/2
.t C 300/2
c
.t C 300/3 C c, and Q D t C 300 C
. Now Q.0/ D 40 ) c D 234  105, so Q D
.t C 300/2
234  105
t C 300
; 0  t  300.
.t C 300/2
4.2.16. (a) S 0 D km .S Tm /; S.0/ D 0, so (A) S D Tm C .S0 Tm /e km t . T 0 D k.T S / D
k T Tm .S0 Tm /e km t , from (A). Therefore,T 0 CkT D kTm Ck.S0 Tm /e km t ; T D ue k t ;
k
(B) u0 D kTm e k t C k.S0 Tm /e .k km /t ; u D Tm e k t C
.S0 Tm /e .k km /t C c; T .0/ D T0 )
k km
k
k
k
c D T0 Tm
.S0 Tm /; u D Tm e k t C
.S0 Tm /e .k km /t CT0 Tm
.S0 Tm /;
k km
k km
k km

42 Chapter 4 Applications of First Order Equations



k.S0 Tm /  kmt
e
e kt .
.k km /
(b) If k D km (B) becomes (B) u0 D kTm e k t C k.S0 Tm /; u D Tm e k t C k.S0 Tm /t C c; T .0/ D
T0 ) c D T0 Tm ; u D Tm e k t Ck.S0 Tm /t C.T0 Tm /; T D Tm Ck.S0 Tm /te k t C.T0 Tm /e k t .
(c) limt !1 T .t/ D limt !1 S.t/ D Tm in either case.


V0
1
1
0
2
D b;
V 0 D a;
4.2.18. V D aV
bV D bV .V
b=a/;
V .V a=b/
V a=b V

V a=b
1
D at C k; (A) V a=b D ce at ; (B) V D a
ln
. Since V .0/ D V0 , (A)

V
V
b 1 ce at
V0 a=b
a
V0
)c D
. Substituting this into (B) yields V D
so limt !1 V .t/ D
V0
b V0 .V0 a=b/ e at
a=b
T D Tm C .T0

Tm /e

kt

r
QnC1 D r cn .t/; n D
W
r t =w
0
r t =W
0; 1; : : : , where c0.t/  c. Therefore,QnC1 D unC1 e
; (A) unC1 D r e
cn .t/. In particular,
r t =W
r t =W
r t =W
with n D 0, u1 D cW .e 0
1/, so Q1 D cW .1 1
e
/ and c1 D c.1 e
/. We will shown
 j
n 1
X
1 rt
A. This is true for n D 1; if it is true for a given n,
by induction that cn D c @1 e r t =W
j W
0
4.2.20. If Qn .t/ is the number of pounds of salt in Tn at time t, then QnC1
C

j D0

then, from (A),

u0nC1

D cre

r t =W

so (since QnC1 .0/ D 0),

@1

r t =W

1
 
n 1
X
1 rt j
A D c r e r t =W
j W

j D0

unC1 D cW .e r t =W

1/

n
X1

j D0

 
n 1
X
1 rt j
cr
;
j W
j D0

1
r j C1 j C1
t
:
.j C 1/ W j

Therefore,
cnC1

1
D
unC1 e
W

r t =W

D c @1

r t =W

which completes the induction. From this, limt !1 cn .t/ D c.

1
 
n
X
1 rt j A
;
j W

j D0

4.2.22. Since the incoming solution contains 1/2 lb of salt per gallon and there are always 600 gallons in the tank, we
 conclude
 that limt !1 Q.t/ D 300. To verify this rigorously, note
Z t intuitively
1
that Q1 .t/ D exp
a./ d  is a solution of the complementary equation, (A) Q1 .0/ D 1,
150 0
0
0
and (B) limt !1 Q1 .t/
Z t D 0 (since limt !1 a.t/ D 1). Therefore,Q D QZ1 u;t Q1 u D 2; u D
2
d
d
; u D Q0 C 2
(see (A)), and Q.t/ D Q0 Q1 .t/ C 2Q1 .t/
. From (B),
Q1
1 ./
0 QZ
0 Q1 ./
t
d
, a 0  1 indeterminate form. By LHospitals rule, lim Q.t/ D
lim Q.t/ D 2 lim Q1 .t/
t !1
t !1
t !1
Q
1 ./
0


Q10 .t/
1
Q1 .t/
2 lim
D 2 lim 0
D 300.
t !1 Q1 .t/
t !1 Q .t/
Q12 .t/
1
vspace*10pt

Section 4.3 Elementary Mechanics

43

4.3 ELEMENTARY MECHANICS


192
k
D 6 sl, so 6v 0 D 192 kv, or (A) v 0 C v D 32.
32
6
Since v1 D e k t =6 is a solution of the complementary equation, the solutions of (A) are v D ue k t =6
192 k t =6
192
where u0 e k t =6 D 32. Therefore,u0 D 32e k t =6; u D
e
C c; v D
C ce k t =6. Now
k
k
192
192
192
v.0/ D 0 ) c D
. Therefore,v D
.1 e k t =6/ and limt !1 v.t/ D
D 16 ft/s, so
k
k
k
2t
k D 12 lb-s/ft and v D 16.1 e /.
4.3.2. The firefighters mass is m D

64000
D 2000, so 2000v 0 D 50000 2000v, or (A) v 0 Cv D 25. Since v1 D e t is a solution
32
of the complementary equation, the solutions of (A) are v D ue t where u0 e t D 25. Therefore,u0 D
25e t ; u D 25e t C c; v D 25 C ce t . Now v.0/ D 0 ) c D 25. Therefore,v D 25.1 e t / and
limt !1 v.t/ D 25 ft/s.
4.3.3. m D

4.3.4. 20v 0 D 10

1
1
1
v, or (A) v 0 C v D . Since v1 D e
2
20
2

t =40

is a solution of the complementary

1
e t =40
. Therefore,u0 D
; u D
2
2
t =40
t =40
t =40
20e
C c; v D 20 C ce
. Now v.0/ D 7 ) c D 27. Therefore,v D 20 27e
.
equation, the solutions of (A) are v D ue

t =40

where u0 e

t =40

3200
D 100 sl. The component of the gravitational force in the direction of motion is
32
v0
3200 cos.=3/ D 1600 lb. Therefore, 100v 0 D 1600Cv 2. Separating variables yields
D
.v 40/.v C 40/



v 40
1
1
4
1
D 4t Ck and v 40 D ce 4t =5. Now v.0/ D
, or
D . Therefore, ln
100
v 40 v C 40
5
v C 40
5
v C 40
4t =5
40.13 C 3e 4t =5/
13
v 40
13e
40.3 C 13e 4t =5/
64 ) c D
,
or
v
D
.
; therefore
D
, so v D
3
v C 40
3
3 13e 4t =5
13 3e 4t =5

mg  k t =m
mg
4.3.8. From Example 4.3.1, (A) v D
C v0 C
e
. Integrating this yields (B) y D
k
k


m
mg
m
mg 
mgt
v0 C
e k t =m C c. Now y.0/ D y0 ) c D y0 C
v0 C
. Substituting this
k
k
k
k
k
into (B) yields
mgt
m
mg  k t =m
m
mg 
y D
v0 C
e
C y0 C
v0 C
k
k
k
k
k
mg 
mg  k t =m
m
v0 gt C
v0 C
e
D y0 C
k
k
k
m
D y0 C .v0 v gt/
k
where the last equality follows from (A).
4.3.6. m D

256
1
D 8 sl. Since the resisting force is 1 lb when jvj D 4 ft/s, k D
. Therefore,
32
16
0

1 2
1 2
v
1
8v 0 D 256 C
v D
v
.64/2 . Separating variables yields
D
, or
16
16
.v
64/.v
C
64/
128



v 64
1
1
D t C k and v 64 D ce t . Now v.0/ D 0 ) c D
v 0 D 1. Therefore,ln
v 64 v C 64
v C 64
v C 64
v 64
64.1
et /
64.1 e t /
t
1; therefore
D e , so v D
, or v D
. Therefore, limt !1 v.t/ D
v C 64
1 C et
1Ce t
64.
4.3.10. m D

44 Chapter 4 Applications of First Order Equations


s

k
v0
kv 2 D mg.1 C 2 v 2/, where D
. Therefore,(A)
D g.
mg
1 C 2v2
Z
Z
dv
1
du
1
1
D
D
tan 1 u D
tan 1 . v/. ThereWith the substitution u D v,
2
2
2
1C v

1Cu


1
1
1
fore, tan 1 . v/ D gt C c. Now v.0/ D v0 ) c D tan 1 . v0 /, so tan 1 . v/ D gt C



s
!
r
1
m
k
1
1
1
1
tan . v0 /. Since v.T / D 0, it follows that T D
tan v0 D
tan
v0
.

g
kg
mg
(b) Replacing qt by t
T and setting v0 D 0in the answer to the previous exercise yields v D
r
gk
mg 1 e 2 m .t T /
q
.
k
.t T /
2 gk
m
1Ce
4.3.12. (a) mv 0 D

mg

4.3.14. (a) mv 0 D mg C f .jvj/; since s D jvj D v, (A) ms 0 D mg f .s/.


(b) Since f is increasing and limt !1 f .s/  mg, mg f .s/ > 0 for all s. This and (A) imply that s
is an increasing function of t, so either (B) limt !1 s.t/ D 1 or (C ) limt !1 s.t/ D s < 1. However,
(A) and (C) imply that s 0 .t/ > K D g f .s/=m for all t > 0. Consequently, s.t/ > s0 C Kt for all
t > 0, which contradicts (C) because K > 0.
(c) There is a unique positive number sT such that f .sT / D mg, and s  sT is a constant solution of
(A). Now suppose that s.0/ < sT . Then Theorem 2.3.1 implies that (D) s.t/ < sT for all t > 0, so (A)
implies that s is strictly increasing. This and (D) imply that limt !1 s.t/ D s  sT . If s < sT then (A)
implies that s 0 .t/ > K D g f .s/=m. Consequently, s.t/ > s.0/ C Kt, which contradicts (D) because
K > 0. Therefore, s.0/ < sT ) limt !1 s.t/ D sT . A similar proof with inequlities reversed shows
that s.0/ > sT ) limt !1 s.t/ D sT .
p
4.3.16. (a) (A) mv 0 D mg C k jvj; since the magnitude of the resistance is 64 lb when v Dp16 ft/s,
4k D 64, so kpD 16 lb s1=2 =ft1=2 . Since m D 2 and g D 32, (A) becomes 2v 0 D 64 C 16 jvj, or
v 0 D 32 C 8 jvj.
p
(b) From Exercise 4.3.14(c), vT is the negative number such that 32 C 8 jvT j D 0; thus, vT D 16
ft/s.
p
4.3.18. With h D 0, ve D 2gR, where R is the radius of the moon and g is
rthe acceleration due to
5:31
2  5:31  1080
gravity at the moons surface. With length in miles, g D
mi/s2 , so ve D
 1:47
5280
5280
miles/s.
4.3.20. Suppose that there is a number ym such that y.t/  ym for all t  0 and let D

gR2
.
.ym C R/2

d 2y
 for all t  0. Integrating this inequality from t D 0 to t D T > 0 yields v.T / v0 
dt 2
v0
T , or v.T /  v0 T , so v.T / < 0 for T >
. This implies that the vehicle must eventually fall

back to Earth, which contradicts the assumption that it continues to climb forever.

Then

4.4 AUTONOMOUS SECOND ORDER EQUATIONS


4.4.1. y D 0 is a stable equilibrium. The phase plane equivalent is v
v2 C

y4
D c.
4

dv
C y 3 D 0, so the trajectories are
dy

Section 4.4 Autonomous Second Order Equations


4.4.2. y D 0 is an unstable equilibrium. The phase plane equivalent is v
are v 2 C

dv
C y 2 D 0, so the trajectories
dy

2y 3
D c.
3

4.4.4. y D 0 is a stable equilibrium. The phase plane equivalent is v


are v 2

45

dv
C ye
dy

D 0, so the trajectories

.y C 1/ D c.

4.4.6. p.y/ D y 3

4y D .y C 2/y.y
y.y

2/, so the equilibria are 2; 0; 2. Since

2/.y C 2/

<0
>0

if y < 2 or 0 < y < 2;


if 2 < y < 0 or y > 2;

dv
C y 3 4y D 0, so the trajectories
dy
are 2v 2 C y 4 8y 2 D c. Setting .y; v/ D .0; 0/ yields c D 0, so the equation of the separatrix is
2v 2 y 4 C 8y 2 D 0.
0 is unstable and 2; 2 are stable. The phase plane equivalent is v

4.4.8. p.y/ D y.y

2/.y

y.y

1/.y C 2/, so the equilibria are 2; 0; 1; 2. Since

2/.y

1/.y C 2/

>0
<0

if y < 2 or 0 < y < 1 or y > 2;


if 2 < y < 0 or 1 < y < 2;

dv
Cy.y 2/.y 1/.y C2/ D 0, so
dy
the trajectories are 30v 2 Cy 2 .12y 3 15y 2 80yC120/ D c. Setting .y; v/ D . 2; 0/ and .y; v/ D .1; 0/
yields c D 496 and c D 37 respectively, so the equations of the separatrices are 30v 2 Cy 2 .12y 3 15y 2
80y C 120/ D 496 and 30v 2 C y 2 .12y 3 15y 2 80y C 120/ D 37.
0; 2 are stable and 2; 1 are unstable. The phase plane equivalent is v

4.4.10. p.y/ D y 3 ay. If a  0, then p.0/ D 0, p.y/ > 0 if y > 0, and p.y/ < 0 if y < 0, so 0 is
stable. If a > 0, then
p
p
p
p
a/.y C a/ > 0 if
a < y < 0 or y > a;
y 3 ay D y.y
p
p
< 0 if y <
a or 0 < y < a;
p
p
so
a and a are stable and 0 is unstable. We say that a D 0 is a critical value because it separates
the two cases.
4.4.12. p.y/ D y ay 3 . If a  0, then p.0/ D 0, p.y/ > 0 if y > 0, and p.y/ < 0 if y < 0, so 0 is
stable. If a > 0, then
p
p
p
p
y ay 3 D ay.y 1= a/.y C 1= a/ > 0 if y < 1= a < y < 0 or 0 < y < 1= a
p
p
< 0 if 1= a < y < 0 or y > 1= a;
p
p
so
a and a are unstable and 0 is stable. We say that a D 0 is a critical value because it separates
the two cases.
4.4.24. (a) Since v 0 D p.y/  k and v.0/ D 0, v  kt and therefore y  y0 C kt 2 =2 for 0  t < T .
(b) Let 0 <  < . Suppose that y is the solution of the initial value problem (A) y 00 C p.y/ D
0; y.0/ D y0 ; y 0 .0/ D 0, where y < y0 < y C . Now let Y D y y and P .Y / D p.Y C
y/. Then P .0/ D 0 and P .Y / < 0 if 0 < Y   . Morover, Y is the solution of Y 00 C p.Y / D

46 Chapter 4 Applications of First Order Equations


0; Y .0/ D Y0 ; Y 00 .0/ D 0, where Y0 D y0 y, so 0 < Y0 < . From (a), Y .t/   for some t > 0.
Therefore,y.t/ > y C  for some t > 0, so y is an unstable equilibrium of y 00 C p.y/ D 0.
4.5 APPLICATIONS TO CURVES

4.5.2. Differentiating (A) e xy D cy yields (B) .xy 0 C y/e xy D cy 0 . From (A), c D


this into (B) and cancelling e xy yields xy 0 C y D
4.5.4. Solving y D x 1=2 Ccx for c yields c D
or xy 0

yD

x 1=2
.
2

4.5.6. Rewriting y D x 3 C

y
x
x

1=2

y2
.
.xy 1/

, and differentiating yields 0 D

y0
x

y x 3=2
C
,
x2
2

c
as xy D x 4 C c and differentiating yields xy 0 C y D 4x 3.
x

4.5.8. Rewriting y D e x C c.1 C x 2 / as


2xy
ex
D
.1 C x 2 /2
1 C x2

y0
, so y 0 D
y

e xy
. Substituting
y

ex
y0
y
D
C c and differentiating yields
2
2
1Cx
1Cx
1 C x2

2xe x
, so .1 C x 2 /y 0
.1 C x 2/2

2xy D .1

x/2 e x .

4.5.10. If (A) y D f C cg, then (B) y 0 D f 0 C cg0 . Mutiplying (A) by g0 and (B) by g yields (C)
yg0 D fg0 Ccgg0 and (D) y 0 g D f 0 gCcg0 g, and subtracting (C) from (D) yields y 0 g yg0 D f 0 g fg0 .
4.5.12. Let .x0 ; y0/ be the center and r be the radius of a circle in the family. Since . 1; 0/ and .1; 0/ are
on the circle, .x0 C 1/2 C y02 D .x0 1/2 C y02 , which implies that x0 D 0. Therefore,the equation of
the circle is (A) x 2 C .y y0 /2 D r 2. Since .1; 0/ is on the circle, r 2 D 1 C y02 . Substituting this into
x2 C y2 1
(A) shows that the equation of the circle is x 2 C y 2 2yy0 D 1, so 2y0 D
. Differentiating
y
0
0 2
2
0 2
2
y.2x C 2yy / y .x C y
1/ D 0, so y .y
x C 1/ C 2xy D 0.
4.5.14. From Example 4.5.6 the equation of the line tangent to the parabola at .x0 ; x02/ is (A) y D
x02 C 2x0 x.
(a) From (A), .x; y/ D .5; 9/ is on the tangent line through .x0 ; x02/ if and only if 9 D x02 C 10x0,
or x02 10x0 C 9 D .x0 1/.x0 9/ D 0. Letting x0 D 1 in (A) yields the line y D 1 C 2x, tangent
to the parabola at .x0 ; x02 / D .1; 1/. Letting x0 D 9 in (A) yields the line y D 81 C 18x, tangent to the
parabola at .x0 ; x02 / D .9; 81/.
(b) From (A), .x; y/ D .6; 11/ is on the tangent line through .x0 ; x02/ if and only if 11 D x02 C 12x0,
or x02 12x0 C 11 D .x0 1/.x0 11/ D 0. Letting x0 D 1 in (A) yields the line y D 1 C 2x, tangent
to the parabola at .x0 ; x02 / D .1; 1/. Letting x0 D 11 in (A) yields the line y D 121 C 22x, tangent to
the parabola at .x0 ; x02 / D .11; 21/.
(c) From (A), .x; y/ D . 6; 20/ is on the tangent line through .x0 ; x02 / if and only if 20 D x02 12x0,
or x02 C 12x0 C 20 D .x0 C 2/.x0 C 10/ D 0. Letting x0 D 2 in (A) yields the line y D 4 4x,
tangent to the parabola at .x0 ; x02/ D . 2; 4/. Letting x0 D 10 in (A) yields the line y D 100 20x,
tangent to the parabola at .x0 ; x02/ D . 10; 100/.
(d) From (A), .x; y/ D . 3; 5/ is on the tangent line through .x0 ; x02/ if and only if 5 D x02 6x0,
or x02 C 6x0 C 5 D .x0 C 1/.x0 C 5/ D 0. Letting x0 D 1 in (A) yields the line y D 1 2x, tangent
to the parabola at .x0 ; x02 / D . 1; 1/. Letting x0 D 5 in (A) yields the line y D 25 10x, tangent to
the parabola at .x0 ; x02 / D . 5; 25/.

Section 4.5 Applications to Curves

47

4.5.15. (a) If .x0 ; y0 / is any point on the circle such that x0 1 (and therefore y0 0), then
x0
. Therefore,the equation of the tangent line is
differentiating (A) yields 2x0 C 2y0 y00 D 0, so y00 D
y0
x0
y D y0
.x x0 /. Since x02 C y02 D 1, this is equivalent to (B).
y0
1
1
x0
on the tangent line, we can rewrite (B) as y xy 0 D
. Hence (F)
D
(b) Since y 0 D
y0
y0
.y xy 0 /2
.y xy 0 /2 1
x02
0 2
y02 and (G) x02 D 1 y02 D
.
Since
.y
/
D
, (F) and (G) imply that .y 0 /2 D
.y xy 0 /2
y02
.y xy 0 /2 1, which implies (C).
(c) Using the quadratic formula to solve (C) for y 0 yields
p
xy x 2 C y 2 1
0
y D
.H/
x2 1


1 x0 x
1 x0 x 2
if .x; y/ is on a tangent line with slope y 0 . If y D
, then x 2 C y 2 1 D x 2 C
1D
y0
y0
2

x0
x0
x x0
(since x02 C y02 D 1). Since y 0 D
, this implies that (H) is equivalent to
D
y0 
y0
y0

x x0
1
x.1 x0 x/ x x0
D

, which holds if and only if we choose the " so that

2
x 1
y0
y0
 y0
x x0
x x0
. Therefore,we must choose D if
> 0, so (H) reduces to (D), or D C if
y0
y0
x x0
< 0, so (H) reduces to (E).
y0
x
(d) Differentiating (A) yields 2x C 2yy 0 D 0, so y 0 D
on either semicircle. Since (D) and (E)
y
xy
x
both reduce to y 0 D
D
(since x 2 C y 2 D 1) on both semicircles, the conclusion follows.
1 x2
y
p
25 49
0
(e) From (D) and (E) the slopes of tangent lines from (5,5) tangent to the circle are y D
D
24
3
1 C 3x=5
4
1 4x=5
3 4
; . Therefore, tangent lines are y D 5 C .x 5/ D
and y D 5 C .x 5/ D
,
4 3
4
4=5
3
3=5
which intersect the circle at . 3=5; 4=5/ .4=5; 3=5/, respectively. (See (B)).
4.5.16. (a) If .x0 ; y0 / is any point on the parabola such that x0 > 0 (and therefore y0 0), then
1
differentiating (A) yields 1 D 2y0 y00 , so y00 D
. Therefore,the equation of the tangent line is y D
2y0
1
y0 C
.x x0 /. Since x0 D y02 , this is equivalent to (B).
2y0
1
y0
(b) Since y 0 D
on the tangent line, we can rewrite (B) as
D y xy 0 . Substituting this into (B)
2y0
2
x
yields y D .y xy 0 / C
, which implies (C).
4.y xy 0 /
(c) Using the quadratic formula to solve (C) for y 0 yields
p
y y2 x
0
y D
.F/
2x


y0
x
1
x 2
0
2
if .x; y/ is on a tangent line with slope y . If y D
C
, then y
x D
y0
so (F)
2
2y0
4
y0

48 Chapter 4 Applications of First Order Equations

x
x
y0
1
y0
y0
is equivalent to
D
which holds if and only if we choose the " so that
2y
4x

0


x
x
D
y0
. Therefore,we must choose D C if x > y02 D x0, so (F) reduces to (D),
y0

y0
y0
or D if x < y02 D x0, so (F) reduces to (E).
1
(d) Differentiating (A) yields 1 D 2yy 0 , so y 0 D
on either half of the parabola. Since (D) and (E)
2y
both reduce to this if x D y 2 , the conclusion follows.
y0 C

4.5.18. The equation of the line tangent to the curve at .x0 ; y.x0 // is y D y.x0 / C y 0 .x0 /.x x0 /.
y 0 .x0 /x0
2y
y0
2
D 0. Since x0 is arbitrary, it follows that y 0 D
, so
D ,
Since y.x0 =2/ D 0, y.x0 /
2
x
y
x
ln jyj D 2 ln jxj C k, and y D cx 2 . Since .1; 2/ is on the curve, c D 2. Therefore,y D 2x 2 .
4.5.20. The equation of the line tangent to the curve at .x0 ; y.x0 // is y D y.x0 / C y 0 .x0 /.x x0 /.
Since .x1 ; y1 / is on the line, y.x0 / C y 0 .x0 /.x1 x0 / D y1 . Since x0 is arbitrary, it follows that
1
y0
y C y 0 .x1 x/ D y1 , so
D
, ln jy y1 j D ln jx x1 j C k, and y y1 D c.x x1/.
y y1
x x1
4.5.22. The equation of the line tangent to the curve at .x0 ; y.x0 // is y D y.x0 / C y 0 .x0 /.x
y.0/ D x0 , x0 D y.x0 / y 0 .x0 /x0 . Since x0 is arbitrary, it follows that x D y xy 0 , so (A) y 0
1
The solutions of (A) are of the form y D ux, where u0 x D 1, so u0 D
. Therefore,u D
x
and y D x ln jxj C cx.

x0 /. Since
y
D 1.
x
ln jxj C c

x x0
. Since y.0/ D
y 0 .x0 /
x0
y2
x2
c
2y.x0 /, y.x0 / C 0
D 2y.x0 /. Since x0 is arbitrary, it follows that y 0 y D x, so (A)
D
C
y .x0 /
2
2
2
p
and y 2 D x 2 C c. Now y.2/ D 1 $ c D 3. Therefore,y D x 2 3.
4.5.24. The equation of the line normal to the curve at .x0 ; y0 / is y D y.x0 /

4.5.26. Differentiating the given equation yields 2x C 4y C 4xy 0 C 2yy 0 D 0, so y 0 D

x C 2y
2x C y

2x C y
is a differential equation for the
x C 2y
2Cu
2.u2 1/
orthogonal trajectories. Substituting y D ux in (A) yields u0 x C u D
, so u0 x D
and
1 C 2u
1 C 2u


1 C 2u
2
3
1
4
u0 D
, or
C
u0 D
. Therefore, 3 ln ju 1jCln juC1j D 4 ln jxjC
.u 1/.u C 1/
x
u 1
uC1
x
k
y
K, so .u 1/3 .u C 1/ D 4 . Substituting u D yields the orthogonal trajectories .y x/3 .y C x/ D k.
x
x

is a differential equation for the given family, and (A) y 0 D

y.1 C 2x 2 /
2
2
4.5.28. Differentiating yields ye x .1C2x 2 /Cxe x y 0 D 0, so y 0 D
is a differential equation
x
x
for the given family. Therefore,(A) y 0 D
is a differential equation for the orthogonal
y.1 C 2x 2 /
x
y2
1
k
trajectories. From (A), yy 0 D
, so
D
ln.1 C 2x 2/ C , and the orthogonal trajectories
2
1 C 2x
2
4
2
1
2
2
are given by y D
ln.1 C 2x / C k.
2

Section 4.5 Applications to Curves


4.5.30. Differentiating (A) y D 1 C cx 2 yields (B) y 0 D 2cx. From (C), c D

1
x2

49

. Substituting this

2.y 1/
into (B) yields the differential equation y 0 D
for the given family of parabolas. Therefore,y 0 D
x
x
is a differential equation for the orthogonal trajectories. Separating variables yields 2.y
2.y 1/
x2
9
x2 9
C k. Now y. 1/ D 3 $ k D , so .y 1/2 D
C . Therefore,(D)
1/y 0 D x, so .y 1/2 D
2
2
2
2
r
9 x2
. This curve interesects the parabola (A) if and only if the equation (C) cx 2 D
y D 1C
2
r
9 x2
has a solution x 2 in .0; 9/. Therefore,c > 0 is a necessaary condition for intersection. We will
2
show that it is also sufficient. Squaring both sides of (C) and simplifying
yields 2c 2 x 4 Cx 2 9 D 0. Using
p
1 C 1 C 72c 2
the quadratic formula to solve this for x 2 yields x 2 D
. The condition x 2 < 9 holds if
4c 2
p
and only if 1C 1 C 72c 2 < 36c 4, which is equivalent to 1C72c 2 < .1C36c 2/2 D 1C72c 2 C1296c 4,
which holds for all c > 0.
4.5.32. The angles  and 1 from the x-axis to the tangents to C and C1 satisfy tan  D f .x0 ; y0 / and
f .x0 ; y0 / C tan
tan  C tan
tan 1 D
D
D tan. C /. Therefore, assuming  and 1 are both in
1 f .x0 ; y0 / tan
1 tan  tan
0; 2 /, 1 D  C .
4.5.34. Circles centered at the origin are given by x 2 C y 2 D r 2 . Differentiating yields 2x C 2yy 0 D 0,
.x=y/ C tan
x
so y 0 D
is a differential equation for the given family, and y 0 D
is a differential
y
1 C .x=y/ tan
1=u C tan
1 C u tan
equation for the desired family. Substituting y D ux yields u0 x C u D
D
.
1 C .1=u/ tan
u C tan
1 C u2 u C tan 0
1
1
Therefore,u0 x D
,
u D
and ln.1 C u2 / C tan tan 1 u D ln jxj C k.
u C tan 1 C u2
x
2
y
1
2
2
1 y
Substituting u D yields ln.x C y / C .tan / tan
D k.
x
2
x

CHAPTER 5
Linear Second Order Equations

5.1 HOMOGENEOUS LINEAR EQUATIONS


5.1.2. (a) If y1 D e x cos x, then y10 D e x .cos x sin x/ and y100 D e x .cos x sin x sin x cos x/ D
2e x sin x, so y100 2y10 C 2y1 D e x . 2 sin x 2 cos x C 2 sin x C 2 cos x/ D 0. If y2 D e x sin x, then
0
y2 D e x .sin x C cos x/ and y200 D e x .sin x C cos x C cos x sin x/ D 2e x cos x, so y200 2y20 C 2y2 D
e x .2 cos x 2 sin x 2 cos x C 2 sin x/ D 0.
(b) If (B) y D e x .c1 cos x C c2 sin x/, then
y 0 D e x .c1.cos x

sin x/ C c2 .sin x C cos x//

.C/

and
y 00

D
D

c1e x .cos x sin x sin x cos x/


Cc2e x .sin x C cos x C cos x sin x/
2e x . c1 sin x C c2 cos x/;

so
y 00

2y 0 C 2y

D c1 e x . 2 sin x

2 cos x C 2 sin x C 2 cos x/

Cc2 e x .2 cos x

2 cos x C 2 sin x/ D 0:

2 sin x

(c) We must choose c1 and c2 in (B) so that y.0/ D 3 and y 0 .0/ D 2. Setting x D 0 in (B) and (C)
shows that c1 D 3 and c1 C c2 D 2, so c2 D 5. Therefore, y D e x .3 cos x 5 sin x/.
(d) We must choose c1 and c2 in (B) so that y.0/ D k0 and y 0 .0/ D k1 . Setting x D 0 in (B) and (C)
shows that c1 D k0 and c1 C c2 D k1 , so c2 D k1 k0 . Therefore, y D e x .k0 cos x C .k1 k0 / sin x/.
5.1.4. (a) If y1 D

1
x
.x 2

, then y10 D

1
.x

1/2

1/y100 C 4xy10 C 2y1

and y100 D

2
.x

1/3

, so

2.x 2 1/
.x 1/3

2.x C 1/ 4x C 2.x
.x 1/2

4x
2
C
2
.x 1/
x 1
1/

D 0:

1/y200 C 4xy20 C 2y2 D 0. The general solution on each of the


c1
c2
intervals . 1; 1/, . 1; 1/, and .1; 1/ is (B) y D
C
.
x 1
xC1

Similar manipulations show that .x 2

51

52 Chapter 5 Linear Second Order Equations


c1
c2
. We must choose c1 and c2 in (B) so
.x 1/2
.x C 1/2
0
that y.0/ D 5 and y .0/ D 1. Setting x D 0 in (B) and (C) shows that c1 C c2 D 5; c1 c2 D 1.
2
3
Therefore,c1 D 2 and c2 D 3, so y D
on . 1; 1/.
x 1 xC1
(d) The Wronskian of fy1 ; y2 g is
(b) Differentiating (B) yields (C) y 0 D

2
x 1
xC1
;
D 2
1
1

.x
1/2

.x 1/2
.x C 1/2
Z x
Z x
4x
4t
, so
p.t/ dt D
dt D ln.x 2
so W .0/ D 2. Since p.x/ D 2
2
x
1
t
1
0
0
2
2
2
implies that W .x/ D W .0/e ln.x 1/ D 2
,
consistent
with (D).
.x
1/2

W .x/ D

Rx

Rx

.D/

1/2 , Abels formula

5.1.6. From Abels formula, W .x/ D W . /e 3  .t C1/ dt D 0  e 3  .t C1/ dt D 0.


Z x
Z x
dt
1
5.1.8. p.x/ D ; therefore
p.t/ dt D
D ln x, so Abels formula yields W .x/ D W .1/e
x
t
1
1
1
.
x
5.1.10. p.x/ D
uD

K
e
4

4x

2; P .x/ D

2x; y2 D uy1 D ue 3x ; u0 D

. Choose K D 4; then y2 D e

5.1.12. p.x/ D

2a; P .x/ D

4x 3x

De

Ke P.x/
Ke 2x
D
D Ke
2
e 6x
y1 .x/

ln x

4x

2ax; y2 D uy1 D ue ax ; u0 D

Choose K D 1; then y2 D xe ax .

Ke P.x/
Ke 2ax
D
D K; u D Kx.
e 2ax
y12 .x/

Ke P.x/
K
1
Kx
; P .x/ D ln x; y2 D uy1 D ux; u0 D
D 2 D ; u D K ln x.
x
x
x
y12 .x/
Choose K D 1; then y2 D x ln x.
5.1.14. p.x/ D

5.1.16. p.x/ D
uD

1
; P .x/ D
x

ln jxj; y2 D uy1 D ux 1=2e 2x ; u0 D

Ke 4x
. Choose K D 4; then y2 D e
4

4x

.x 1=2e 2x / D x 1=2e

2x

Ke P.x/
Kx
D
De
2
xe 4x
y1 .x/

4x

2
Ke P.x/
Kx 2
; P .x/ D 2 ln jxj; y2 D uy1 D ux cos x; u0 D
D
D
x
x 2 cos2 x
y12 .x/
K sec2 x; u D K tan x. Choose K D 1; then y2 D tan x.x cos x/ D x sin x.

5.1.18. p.x/ D

3x C 2
3
D 1
; P .x/ D x
3x 1
3x 1
x
P.x/
Ke
K.3x 1/e
u0 D
D
D K.3x 1/e 3x ; u D
2
e 4x
y1 .x/
y2 D xe 3x e 2x D xe x .
5.1.20. p.x/ D

1j; y2 D uy1 D ue 2x ;

ln j3x
Kxe

3x

. Choose K D

1; then

Section 5.1 Homogeneous Linear Equations

53

2.2x 2 1/
2
2
D 2
C ; P .x/ D 2x ln j2x C 1j C 2 ln jxj; y2 D
x.2x C 1/
2x C 1
x
u
Ke P.x/
K.2x C 1/e 2x 2
D
x D K.2x C 1/e 2x ; u D Kxe 2x . Choose K D 1; then
uy1 D ; u0 D
x
x2
y12 .x/
xe 2x
y2 D
D e 2x .
x
5.1.22. p.x/ D

5.1.24. Suppose that y  0 on .a; b/. Then y 0  0 and y 00  0 on .a; b/, so y is a solution of (A)
y 00 C p.x/y 0 C q.x/y D 0; y.x0 / D 0; y 0 .x0 / D 0 on .a; b/. Since Theorem 5.1.1 implies that (A) has
only one solution on .a; b/, the conclusion follows.
5.1.26. If f1 ; 2 g is a fundamental set of solutions of (A) on .a; b/, then every solution y of (A) on .a; b/
is a linear combination of f1 ; 2 g; that is, y D c1 1 C c2 2 D c1.y1 C y2 / C c2 . y1 C y2 / D
.c1 C c2 /y1 C .c1 C c2 /y2 , which shows that every solution of (A) on .a; b/ can be written as a
linear combination of fy1 ; y2 g. Therefore,fy1 ; y2 g is a fundamental set of solutions of (A) on .a; b/.
5.1.28. The Wronskian of fy1 ; y2 g is

y1 y2

W D
y0 y0
1
2
00


y1

D 0
y
1
0

ky1
D k.y1 y10
ky 0

y10 y1 / D 0:

nor y2 can be a solution of y C p.x/y C q.x/y D 0 on .a; b/.



5.1.30. W .x0 / D y1 .x0 /y20 .x0 / y10 .x0 /y2 .x0 / D 0 if either y1 .x0 / D y2 .x0 / D 0 or y10 .x0 / D
y20 .x0 / D 0, and Theorem 5.1.6 implies that fy1 ; y2 g is linearly dependent on .a; b/.
5.1.32. Let x0 be an arbitrary point in .a; b/. By the motivating argument preceding Theorem 5.1.4, (B)
W .x0 / D y1 .x0 /y20 .x0 / y10 .x0 /y2 .x0 / 0. Now let y be the solution of y 00 C p.x/y 0 C q.x/y D
0; y.x0 / D y1 .x0 /; y 0 .x0 / D y10 .x0 /. By assumption, y is a linear combination of fy1 ; y2g on .a; b/;
that is, y D c1 y1 C c2 y2 , where
c1 y1 .x0 / C c2 y2 .x0 /

c1 y10 .x0 / C c2 y20 .x0 /

D y1 .x0 /

D y10 .x0 /:

Solving this system by Cramers rule yields

1
1 y1 .x0 / y2 .x0 /
c1 D
0
D 1 and c2 D
0

W .x0 / y1 .x0 / y2 .x0 /


W .x0 /

y1 .x0 /

0
y .x0 /
1

y1 .x0 /
D 0:
y1 .x0 /

Therefore,y D y1 , which shows that y1 is a solution of (A). A similar argument shows that y2 is a
solution of (A).

5.1.34. Expanding the determinant by cofactors of its


exercise can be written as
0

0
y y1 y2 y 0 y1 y2

W y100 y200 W y100 y200


which is of the form (A) with

pD

1
W

y1

00
y
1

y2

y 00
2

and

first column shows that the first equation in the

y 00 y1

W y0
1
1
qD
W

0
y
1
00
y
1

y2
D 0;
y20

y20
:
y 00
2

54 Chapter 5 Linear Second Order Equations


5.1.36. Theorem 5.1.6 implies that that there are constants c1 and c2 such that (B) y D c1y1 C c2 y2 on
.a; b/. To see that c1 and c2 are unique, assume that (B) holds, and let x0 be a point in .a; b/. Then (C)
y 0 D c1 y10 C c2y20 . Setting x D x0 in (B) and (C) yields
c1 y1 .x0 / C c2 y2 .x0 /
c1 y10 .x0 / C c2 y20 .x0 /

D y.x0 /
D y 0 .x0 /:

Since Theorem 5.1.6 implies that y1 .x0 /y20 .x0 / y10 .x0 /y2 .x0 / 0, the argument preceding Theorem 5.1.4 implies that c1 and c2 are given uniquely by
c1 D

y20 .x0 /y.x0 /


y1 .x0 /y20 .x0 /

y2 .x0 /y 0 .x0 /
y1 .x0 /y 0 .x0 /
c
D
2
y10 .x0 /y2 .x0 /
y1 .x0 /y20 .x0 /

y10 .x0 /y.x0 /


:
y10 .x0 /y2 .x0 /

5.1.38. The general solution of y 00 D 0 is y D c1 C c2 x, so y 0 D c2 . Imposing the stated initial


conditions on y1 D c1 C c2 x yields c1 C c2 x0 D 1 and c2 D 0; therefore c1 D 1, so y1 D 1. Imposing
the stated initial conditions on y2 D c1 C c2x yields c1 C c2x0 D 0 and c2 D 1; therefore c1 D x0 , so
y2 D x x0. The solution of the general initial value problem is y D k0 C k1 .x x0 /.
5.1.40. Let y1 D a1 cos !x C a2 sin !x and y2 D b1 cos !x C b2 sin !x. Then
a1 cos !x0 C a2 sin !x0

D 1

!. a1 sin !x0 C a2 cos !x0 /

D 0

b1 cos !x0 C b2 sin !x0


!. b1 sin !x0 C b2 cos !x0 /

D 0
D 1:

and

sin !x0
cos !x0
, and b2 D
.
!
!
1
Therefore, y1 D cos !x0 cos !x C sin !x0 sin !x D cos !.x x0 / and y2 D . sin !x0 cos !x C
!
1
cos !x0 sin !x/ D sin !.x x0 /. The solution of the general initial value problem is y D k0 cos !.x
!
k1
x0 / C
sin !.x x0 /.
!

Solving these systems yields a1 D cos !x0 , a2 D sin !x0 , b1 D

5.1.42. (a) If y1 D x 2 , then y10 D 2x and y100 D 2, so x 2y100 4xy10 C 6y1 D x 2 .2/ 4x.2x/ C 6x 2 D 0
for x in . 1; 1/. If y2 D x 3 , then y20 D 3x 2 and y200 D 6x, so x 2 y200 4xy20 C 6y2 D x 2 .6x/
4x.3x 2 / C 6x 3 D 0 for x in . 1; 1/. If x 0, then y2 .x/=y1 .x/ D x, which is nonconstant on
. 1; 0/ and .0; 1/, so Theorem 5.1.6 implies that fy1 ; y2 g is a fundamental set of solutions of (A) on
each of these intervals.
5.1.6 and (a) imply that y satisfies (A) on . 1; 0/ and on .0; 1/ if and only if y D
 (b) Theorem
a1 x 2 C a2 x 3; x > 0;
Since y.0/ D 0 we can complete the proof that y is a solution of (A) on
b1 x 2 C b2 x 3; x < 0:
. 1; 1/ by showing that y 0 .0/ and y 00 .0/ both exist if and only if a1 D b1 . Since

y.x/ y.0/
a1 x C a2 x 2 ; if x > 0;
D
b1 x C b2 x 2 ; if x < 0;
x 0

y.x/ y.0/
2a1 x C 3a2 x 2 ; x  0;
0
0
it follows that y .0/ D lim
D 0. Therefore, y D
Since
2b1x C 3b2 x 2 ; x < 0:
x!0
x 0

0
0
0
0
y .x/ y .0/
y .x/ y .0/
2a1 C 3a2 x; if x > 0;
D
it follows that y 00 .0/ D lim
exists if and
2b
C
3b
x;
if
x
<
0;
x!0
x 0
x 0
1
2

Section 5.2 Constant Coefficient Homogeneous Equations

55

only if a1 D b1 . By renaming a1 D b1 D c1, a2 D c2, and b2 D c3 we see that y is a solution of (A) on


c1 x 2 C c2x 3 ; x  0;
. 1; 1/ if and only if y D
c1 x 2 C c3x 3 ; x < 0:
(c) We have shown that y.0/ D y 0 .0/ D 0 for any choice of c1 and c2 in (C). Therefore,the given
initial value problem has a solution if and only if k0 D k1 D 0, in which case every function of the form
(C) is a solution.
(d) If x0 > 0, then c1 and c2 in (C) are uniquely determined by k0 and k1 , but c3 can be chosen
arbitrarily. Therefore,(B) has a unique solution on .0; 1/, but infinitely many solutions on . 1; 1/. If
x0 < 0, then c1 and c3 in (C) are uniquely determined by k0 and k1 , but c2 can be chosen arbitrarily.
Therefore,(B) has a unique solution on . 1; 0/, but infinitely many solutions on . 1; 1/.
5.1.44. (a) If y1 D x 3 , then y10 D 3x 2 and y100 D 6x, so x 2 y100 6xy10 C 12y1 D x 2.6x/ 6x.3x 2/ C
12x 3 D 0 for x in . 1; 1/. If y2 D x 4 , then y20 D 4x 3 and y200 D 12x 2 , so x 2 y200 6xy20 C 12y2 D
x 2 .12x 2 / 6x.4x 3 / C 12x 4 D 0 for x in . 1; 1/. If x 0, then y2 .x/=y1 .x/ D x, which is
nonconstant on . 1; 0/ and .0; 1/, so Theorem 5.1.6 implies that fy1 ; y2 g is a fundamental set of
solutions of (A) on each of these intervals.
(b) Theorem 5.1.2 and (a) imply that y satisfies (A) on . 1; 0/ and on .0; 1/ if and only if (C)
a1 x 3 C a2 x 4 ; x > 0;
y D
Since y.0/ D 0 we can complete the proof that y is a solution of
b1 x 3 C b2 x 4 ; x < 0:
(A) on . 1; 1/ by showing
that y 0 .0/ and y 00 .0/ both exist for any choice of a1 , a2 , b1 , and b2 .

2
y.x/ y.0/
y.x/ y.0/
a1 x C a2 x 3 ; if x > 0;
Since
D
it follows that y 0 .0/ D lim
D 0.
b1x 2 C b2 x 3; if x < 0;
x!0
x 0 
x 0

y 0 .x/ y 0 .0/
3a1 x 2 C 4a2 x 3 ; x  0;
3a1 x C 4a2 x 2 ; if x > 0;
Therefore, y 0 D
Since
D
2
3
3b1x C 4b2 x ; x < 0:
3b1x C 4b2 x 2 ; if x < 0;
x 0
0
0
y .x/ y .0/
it follows that y 00 .0/ D lim
D 0. Therefore,(B) is a solution of (A) on . 1; 1/.
x!0
x 0
0
(c) We have shown that y.0/ D y .0/ D 0 for any choice of a1 , a2 , b1 , and b2 in (B). Therefore,the
given initial value problem has a solution if and only if k0 D k1 D 0, in which case every function of the
form (B) is a solution.
(d) If x0 > 0, then a1 and a2 in (B) are uniquely determined by k0 and k1 , but b1 and b2 can be chosen
arbitrarily. Therefore,(C) has a unique solution on .0; 1/, but infinitely many solutions on . 1; 1/.
If x0 < 0, then b1 and b2 in (B) are uniquely determined by k0 and k1 , but a1 and a2 can be chosen
arbitrarily. Therefore,(C) has a unique solution on . 1; 0/, but infinitely many solutions on . 1; 1/.
5.2 CONSTANT COEFFICIENT HOMOGENEOUS EQUATIONS
5.2.2. p.r / D r 2

4r C 5 D .r

2/2 C 1; y D e 2x .c1 cos x C c2 sin x/.

5.2.4. p.r / D r 2

4r C 4 D .r

2/2 ; y D e 2x .c1 C c2x/.

5.2.6. p.r / D r 2 C 6r C 10 D .r C 3/2 C 1; y D e


5.2.8. p.r / D r 2 C r D r .r C 1/; y D c1 C c2 e

3x

5.2.10. p.r / D r 2 C 6r C 13y D .r C 3/2 C 4; y D e


5.2.12. p.r / D 10r 2

3r

1 D .2r

.c1 cos x C c2 sin x/.


3x

1/.5r C 1/ D 10.r

.c1 cos 2x C c2 sin 2x/.


1=2/.r C 1=5/; y D c1e

x=5

C c2 e x=2 .

5.2.14. p.r / D 6r 2 r 1 D .2r 1/.3r C 1/ D 6.r 1=2/.r C 1=3/; y D c1e x=3 C c2 e x=2;
c1 x=3 c2 x=2
c1 c2
y0 D
e
C e ; y.0/ D 10 ) c1 C c2 D 10; y 0 .0/ D 0 )
C
D 0; c1 D 6; c2 D 4;
3
2
3
2
x=2
x=3
y D 4e
C 6e
.

56 Chapter 5 Linear Second Order Equations


5.2.16. p.r / D 4r 2 4r 3 D .2r 3/.2r C 1/ D 4.r 3=2/.r C 1=2/; y D c1 e x=2 C c2e 3x=2 ;
c1 x=2
3c2 3x=2
13
13 0
23
c1
3c2
23
y0 D
e
C
e
; y.0/ D
) c1 C c2 D
; y .0/ D
)
C
D
;
2
2
12
12
24
2
2
24
1
3
e x=2
3e 3x=2
c1 D ; c2 D ; y D
C
.
3
4
3
4
5.2.18. p.r / D r 2 C 7r C 12 D .r C 3/.r C 4/; y D c1e 4x C c2 e 3x ; y 0 D 4c1e 4x 3c2 e
y.0/ D 1 ) c1 C c2 D 1; y 0 .0/ D 0 ) 4c1 3c2 D 0; c1 D 3, c2 D 4; y D 3e 4x 4e

3x
3x

;
.

e x=6
1/2 D 36.r 1=6/2 ; y D e x=6 .c1 C c2 x/; y 0 D
.c1 C
6
5
c
5
1
c2 x/ C c2 e x=6 ; y.0/ D 3 ) c1 D 3; y 0 .0/ D )
C c2 D ) c2 D 2; y D e x=6 .3 C 2x/.
2
6
2

5.2.20. p.r / D 36r 2

12r C 1 D .6r

5.2.22. (a) From (A), ay 00 .x/ C by 0 .x/ C cy.x/ D 0 for all x. Replacing x by x x0 yields (C)
ay 00 .x x0 / C by 0 .x x0 / C cy.x x0 / D 0. If .x/ D y.x x0 /, then the chain rule implies that
0 .x/ D y 0 .x x0 / and 00 .x/ D y 00 .x x0 /, so (C) is equivalent to a00 C b0 C c D 0.
(b) If fy1 ; y2 g is a fundamental set of solutions of (A) then Theorem 5.1.6 implies that y2 =y1 is
2 .x/
y2 .x x0 /
nonconstant. Therefore,
D
is also nonconstant, so Theorem 5.1.6 implies that f1 ; 2 g
1 .x/
y1 .x x0 /
is a fundamental set of solutions of (A).
(c) Let p.r / D ar 2 C br C c be the characteristic polynomial of (A). Then:
 If p.r / D 0 has distinct real roots r1 and r2 , then the general solution of (A) is
y D c1 e r1.x

x0 /

C c2 e r2.x

x0 /

 If p.r / D 0 has a repeated root r1 , then the general solution of (A) is


y D e r1.x

x0 /

.c1 C c2 .x

x0 //:

 If p.r / D 0 has complex conjugate roots r1 D  C i ! and r2 D 


general solution of (A) is
y D e .x
5.2.24. p.r / D r 2

7 D .r

6r

.c1 cos !.x

x0 / C c2 sin !.x

x0 //:

7/.r C 1/;
y
y0

1
y.2/ D
) c1 C c2 D
3
1 .x 2/ 2 7.x 2/
e
e
.
3
3

x0 /

i ! .where ! > 0/, then the

D c1 e .x 2/ C c2 e 7.x 2/ I
D
c1 e .x 2/ C 7c2e 7.x 2/ I

1
; y 0 .2/ D
3

5 )

c1 C 7c2 D

5; c1 D

5.2.26. p.r / D 9r 2 C 6r C 1 D .3r C 1/2 D 9.r C 1=3/2 ;


y
y0

D e
D

.x 2/=3

1
e
3

.c1 C c2 .x

.x 2/=3

2// I

.c1 C c2 .x

2// C c2 e

.x 2/=3

1
; c2 D
3

2
; y D
3

Section 5.2 Constant Coefficient Homogeneous Equations


y.2/ D 2 ) c1 D 2; y 0 .2/ D
5.2.28. p.r / D r 2 C 3;
y
y0

14
)
3

c1
C c2 D
3

14
) c2 D
3

4; y D e

p 
p 


D c1cos 3 x
C c2 sin 3 x
I
3
3 

p 
p
p
p

3c1sin 3 x
C 3c2 cos 3 x
D
3

y D 2 cos

p 
3 x

p 
1
p sin 3 x
3


3

.2

4.x

2//.


I
3

1
p ;
3

y.=3/ D 2 ) c1 D 2; y 0 .=3/ D 1 ) c2 D

.x 2/=3

57


:
3

5.2.30. y is a solution of ay 00 C by 0 C cy D 0 if and only if


y
y0

D
D

c1 e r1.x
r1c1 e r1.x

x0 /

C e r2.x
x0 /
C r2 e r2 .x

x0 /

x0 /:

Now y1 .x0 / D k0 and y10 .x0 / D k1 ) c1 C c2 D k0 ; r1c1 C r2 c2 D k1 . Therefore,c1 D


and c2 D

k1 r1 k0
. Substituting c1 and c2 into the above equations for y and y yields
r2 r1
y

D
D

r2 k0 k1 r1.x x0 / k1 r1 k0 r2 .x x0 /
e
C
e
r2 r1
r2 r1

k0  r1.x x0 /
k1  r2.x
r2 e
r1 e r2.x x0 / C
e
r2 r1
r2 r1

x0 /

e r1.x

x0 /

r2 k0 k1
r2 r1

5.2.32. y is a solution of ay 00 C by 0 C cy D 0 if and only if


y D e .x

x0 /

.c1 cos !.x

x0 / C c2 sin !.x

x0 //

(A)

and
y0

e .x x0 / .c1 cos !.x x0 / C c2 sin !.x x0 //


C!e .x x0 / . c1 sin !.x x0/ C c2 cos !.x x0 // :

k1 k0
Now y1 .x0 / D k0 ) c1 D k0 and y10 .x0 / D k1 ) c1 C !c2 D k1 , so c2 D
. Substituting
!
c1 and c2 into (A) yields




k1 k0
y D e .x x0 / k0 cos !.x x0 / C
sin !.x x0 / :
!
5.2.34. (b)
e i1 e i2

D .cos 1 C i sin 1 /.cos 2 C i sin 2 /


D .cos 1 cos 2 sin 1 sin 2 / C i.sin 1 cos 2 C cos 1 sin 2 /
D cos.1 C 2 / C i sin.1 C 2 / D e i.1 C2 / :

58 Chapter 5 Linear Second Order Equations


(c)
e 1C2

D e .1 Ci1 /C.2 Ci2 / D e .1 C2 /Ci.1 C2 /

D
D
D
D

e .1 C2 / e i.1C2 / (from (F) with D 1 C 2 and D 1 C 2 )


e 1 e 2 e i.1 C2 / (property of the realvalued exponential function)
e 1 e 2 e i1 e i2 (from (b))
e 1 e i1 e 2 e i2 D e 1Ci1 e 2 Ci2 D e 1 e 2 :

(d) The real and imaginary parts of 1 D e .Ci !/x are u1 D e x cos !x and v1 D e x sin !x, which
are both solutions of ay 00 C by 0 C cy D 0, by Theorem 5.2.1(c). Similarly, the real and imaginary parts
of 2 D e . i !/x are u2 D e x cos. !x/ D e x cos !x and v1 D e x sin. !x/ D e x sin !x, which
are both solutions of ay 00 C by 0 C cy D 0, by Theorem 5.2.1,(c).
5.3 NONHOMOGENEOUS LINEAR EQUATIONS
5.3.2. The characteristic polynomial of the complementary equation is p.r / D r 2 4r C5 D .r 2/2 C1,
so fe 2x cos x; e 2x sin xg is a fundamental set of solutions for the complementary equation. Let yp D
A C Bx; then yp00 4yp0 C 5yp D 4B C 5.A C Bx/ D 1 C 5x. Therefore,5B D 5; 4B C 5A D 1, so
B D 1, A D 1. Therefore,yp D 1 C x is a particular solution and y D 1 C x C e 2x .c1 cos x C c2 sin x/
is the general solution.
5.3.4. The characteristic polynomial of the complementary equation is p.r / D r 2 4r C 4 D .r 2/2 , so
fe 2x ; xe 2x g is a fundamental set of solutions for the complementary equation. Let yp D A C Bx C Cx 2 ;
then yp00 4yp0 C 4yp D 2C 4.B C 2Cx/ C 4.A C Bx C Cx 2 / D .2C 4B C 4A/ C . 8C C 4B/x C
4Cx 2 D 2 C 8x 4x 2. Therefore,4C D 4; 8C C 4B D 8; 2C 4B C 4A D 2, so C D 1,
B D 0, and A D 1. Therefore,yp D 1 x 2 is a particular solution and y D 1 x 2 C e 2x .c1 C c2 x/ is
the general solution.
5.3.6. The characteristic polynomial of the complementary equation is p.r / D r 2 C 6r C 10 D .r C
3/2 C 1, so fe 3x cos x; e 3x sin xg is a fundamental set of solutions for the complementary equation.
Let yp D A C Bx; then yp00 C 6yp0 C 10yp D 6B C 10.A C Bx/ D 22 C 20x. Therefore,10B D
20; 6B C 10A D 22, so B D 2, A D 1. Therefore,yp D 1 C 2x is a particular solution and (A)
y D 1 C 2x C e 3x.c1 cos x C c2 sin x/ is the general solution. Now y.0/ D 2 ) 2 D 1 C c1 ) c1 D 1.
Differentiating (A) yields y 0 D 2 3e 3x .c1 cos x C c2 sin x/ C e 3x . c1 sin x C c2 cos x/, so y 0 .0/ D
2 ) 2 D 2 3c1 C c2 ) c2 D 1. y D 1 C 2x C e 3x .cos x sin x/ is the solution of the initial
value problem.
  
   
A
2
1
8
3A
6
2
2 00
0
5.3.8. If yp D , then x yp C 7xyp C 8yp D A x
C 7x
C
D
D
if
3
2
x
x
x
x
x
x
2
A D 2. Therefore,yp D is a particular solution.
x
5.3.10. If yp D Ax 3 , then x 2 yp00
Therefore,yp D


1
x.3x 2 / C x 3 D 4Ax 3 D 2x 3 if A D .
2

xyp0 C yp D A x 2 .6x/

x3
is a particular solution.
2

5.3.12. If yp D Ax

1=3

, then

x 2 yp00

xyp0

C yp D A x

2x
9

5=3

10A 1=3
x
D 10x 1=3 if A D 9. Therefore,yp D 9x 1=3 is a particular solution.
9

Cx

2=3

Cx

1=3

Section 5.3 Nonhomogeneous Linear Equations

59

  
 

A
3
3
2 12
2 00
0
,
then
x
y
C
3xy
3y
D
A
x
C
3x
C
D 0. Therefore,yp
p
p
p
x3
x5
x4
x3
is not a solution of the given equation for any choice of A.

5.3.14. If yp D

5.3.16. The characteristic polynomial of the complementary equation is p.r / D r 2 C5r 6 D .r C6/.r
1/, so fe 6x ; e x g is a fundamental set of solutions for the complementary equation. Let yp D Ae 3x ; then
1
e 3x
yp00 C 5yp0 6yp D p.3/Ae 3x D 18Ae 3x D 6e 3x if A D . Therefore,yp D
is a particular solution
3
3
e 3x
C c1 e 6x C c2 e x is the general solution.
and y D
3
5.3.18. The characteristic polynomial of the complementary equation is p.r / D r 2 C8r C7 D .r C1/.r C
7/, so fe 7x ; e x g is a fundamental set of solutions for the complementary equation. Let yp D Ae 2x ;
then yp00 C 8yp0 C 7yp D p. 2/Ae 2x D 5Ae 2x D 10e 2x if A D 2. Therefore,yp D 2e 2x is
a particular solution and (A) y D 2e 2x C c1 e 7x C c2 e x is the general solution. Differentiating (A)
yields y 0 D 4e 2x 7c1 e 7x c2 e x . Now y.0/ D 2 ) 2 D 2 C c1 C c2 and y 0 .0/ D 10 )
10 D 4 7c1 c2 . Therefore,c1 D 1 and c2 D 1, so y D 2e 2x e 7x C e x is the solution of the
initial value problem.
5.3.20. The characteristic polynomial of the complementary equation is p.r / D r 2 C 2r C 10 D .r C
1/2 C 9, so fe x cos 3x; e x sin 3xg is a fundamental set of solutions for the complementary equation.
4
45 x=2
If yp D Ae x=2 , then yp00 C 2yp0 C 10yp D p.1=2/Ae x=2 D
Ae
D e x=2 if A D
. Therefore,
4
45
4 x=2
4 x=2
yp D
e
is a particular solution and y D
e Ce x .c1 cos 3x Cc2 sin 3x/ is the general solution.
45
45
5.3.22. The characteristic polynomial of the complementary equation is p.r / D r 2 7r C 12 D .r
4/.r 3/. If yp D Ae 4x , then yp00 7yp0 C 12yp D p.4/Ae 4x D 0  e 4x D 0, so yp00 7yp0 C 12yp 5e 4x
for any choice of A.
5.3.24. The characteristic polynomial of the complementary equation is p.r / D r 2 8r C 16 D .r 4/2 ,
so fe 4x ; xe 4x g is a fundamental set of solutions for the complementary equation. If yp D A cos x C
B sin x, then yp00 8yp0 C16yp D .A cos x CB sin x/ 8. A sin x CB cos x/C16.A cos x CB sin x/ D
.15A 8B/ cos x C .8A C 15B/ sin x, so 15A 8B D 23; 8A C 15B D 7, which implies that A D 1
and B D 1. Hence yp D cos x sin x and y D cos x sin x C e 4x .c1 C c2 x/ is the general solution.
2
2
5.3.26. Thepcharacteristic
p polynomial of the complementary equation is p.r / D r 2r C3 D .r 1/ C2,
x
x
so fe cos 2x; e sin 2xg is a fundamental set of solutions for the complementary equation. If yp D
A cos 3x C B sin 3x, then yp00 2yp0 C 3yp D 9.A cos 3x C B sin 3x/ 6. A sin 3x C B cos 3x/ C
3.A cos 3x C B sin 3x/ D .6A C 6B/ cos 3x C .6A 6B/ sin 3x, so 6A 6B D 6; 6A 6B D 6,
which implies
p that A Dp1 and B D 0. Hence yp D cos 3x is a particular solution and y D cos 3x C
e x .c1 cos 2x C c2 sin 2x/ is the general solution.

5.3.28. The characteristic polynomial of the complementary equation is p.r / D r 2 C 7r C 12 D .r C


3/.r C 4/, so fe 4x ; e 3x g is a fundamental set of solutions for the complementary equation. If yp D
A cos 2x C B sin 2x, then yp00 C 7yp0 C 12yp D 4.A cos 2x C B sin 2x/ C 14. A sin 2x C B cos 2x/ C
12.A cos x CB sin x/ D .8AC14B/ cos 2x C.8B 14a/ sin 2x, so 8AC14B D 2; 14AC8B D 36,
which implies that A D 2 and B D 1. Hence yp D 2 cos 2x C sin 2x is a particular solution and
(A) y D 2 cos 2x C sin 2x C c1 e 4x C c2e 3x is the general solution. Differentiating (A) yields
y 0 D 2 sin 2x C 2 cos 2x 4c1e 4x 3c2e 3x . Now y.0/ D 3 ) 3 D 2 C c1 C c2 and y 0 .0/ D
3 ) 3 D 2 4c1 3c2 . Therefore, c1 D 2 and c2 D 3, so y D 2 cos 2x C sin 2x C 2e 4x 3e 3x
is the solution of the initial value problem.

60 Chapter 5 Linear Second Order Equations I


5.3.30. fcos !0 x; sin !0 xg is a fundamental set of solutions of the complementary equation. If yp D
A cos !x C B sin !x, then yp00 C !02 yp D ! 2 .A cos !x C B sin !x/ C !02 .A cos !x C B sin !x/ D
M
N
.!02 ! 2 /.A cos !x C B sin !x/ D M cos !x C N sin !x if A D 2
and B D 2
.
2
!0 !
!0 ! 2
Therefore,
1
.M cos !x C N sin !x/
yp D 2
!0 ! 2
is a particular solution of the given equation and
yD

1
!02

!2

.M cos !x C N sin !x/ C c1 cos !0 x C c2 sin !0 x

is the general solution.


5.3.32. If yp D A cos !x C B sin !x, then ayp00 C byp0 C cyp D a! 2 .A cos !x C B sin !x/ C




b!. A sin !xCB cos !x/Cc.A cos !xCB sin !x/ D .c a! 2 /A C b!B cos !xC b!A C .c a! 2 /B sin !x.
Therefore, yp is a solution of (A) if and only if the set of equations (B) .c
a! 2 /A C b!B D
2
2 2
2
M; b!A C .c a! /B D N has a solution. If .c a! / C .b!/ 0, then (B) has the so.c a! 2 /M b!N
.c a! 2 /N C b!M
lution A D
,
B
D
, and yp D A cos !x C B sin !x is a
.c a! 2 /2 C .b!/2
.c a! 2 /2 C .b!/2
solution of (A). If .c a! 2 /2 C .b!/2 D 0 (which is true if and only if the left side of (A) is of the form
a.y 00 C ! 2 y/, then the coefficients of A and B in (B) are all zero, so (B) does not have a solution, so (A)
does not have a solution of the form yp D A cos !x C B sin !x.
5.3.34. From Exercises 5.3.2 and 5.3.17, yp1 D 1 C x and yp2 D e 2x are particular solutions of
y 00 4y 0 C 5y D 1 C 5x and y 00 4y 0 C 5y D e 2x respectively, and fe 2x cos x; e 2x sin xg is a fundamental
set of solutions of the complementary equation. Therefore,yp D yp1 C yp2 D 1 C x C e 2x is a particular
solution of the given equation, and y D 1 C x C e 2x .1 C c1 cos x C c2 sin x/ is the general solution.
5.3.36. From Exercises 5.3.4 and 5.3.19, yp1 D 1 x 2 and yp2 D e x are particular solutions of
y 00 4y 0 C 4y D 2 C 8x 4x 2 and y 00 4y 0 C 4y D e x respectively, and fe 2x ; xe 2x g is a fundamental
set of solutions of the complementary equation. Therefore,yp D yp1 C yp2 D 1 x 2 C e x is a particular
solution of the given equation, and y D 1 x 2 C e x C e 2x .c1 C c2x/ is the general solution.
5.3.38. From Exercises 5.3.6 and 5.3.21, yp1 D 1 C 2x and yp2 D e 3x are particular solutions of
y 00 C 6y 0 C 10y D 22 C 20x and y 00 C 6y 0 C 10y D e 3x respectively, and fe 3x cos x; e 3x sin xg is a
fundamental set of solutions of the complementary equation. Therefore,yp D yp1 Cyp2 D 1C2x Ce 3x
is a particular solution of the given equation, and y D 1 C 2x C e 3x .1 C c1 cos x C c2 sin x/ is the
general solution.
5.3.40. Letting c1 D c2 D 0 shows that (A) yp00 C p.x/yp0 C q.x/yp D f . Letting c1 D 1 and c2 D 0
shows that (B) .y1 C yp /00 C p.x/.y1 C yp /0 C q.x/.y1 C yp / D f . Now subtract (A) from (B) to see
that y100 C p.x/y10 C q.x/y1 D 0. Letting c1 D 0 and c2 D 1 shows that (C) .y2 C yp /00 C p.x/.y2 C
yp /0 C q.x/.y2 C yp / D f . Now subtract (A) from (C) to see that y200 C p.x/y20 C q.x/y2 D 0.
5.4 THE METHOD OF UNDETERMINED COEFFICIENTS I
5.4.2. If y D ue 3x , then y 00 6y 0 C5y D e 3x .u00 6u0 C 9u/ 6.u0 3u/ C 5u D e 3x .35 8x/,
so u00 12u0 C32u D 35 8x and up D ACBx, where 12B C32.ACBx/ D 35 8x. Therefore,32B D

1
x
x
8, 32A 12B D 35, so B D
, A D 1, and up D 1
. Therefore, yp D e 3x 1
.
4
4
4

Section 5.4 The Method of Undetermined Coefficients I

61

5.4.4. If y D ue 2x , then y 00 C 2y 0 C y D e 2x .u00 C 4u0 C 4u/ C 2.u0 C 2u/ C u D e 2x . 7 15x C


9x 2 / so u00 C 6u0 C 9u D 7 15x C 9x 2 and up D A C Bx C Cx 2 , where 2C C 6.B C 2Cx/ C
9.A C Bx C Cx 2 / D 7 15x C 9x 2. Therefore,9C D 9, 9B C 12C D 15, 9A C 6B C 2C D 7,
so C D 1, B D 3, A D 1, and up D 1 3x C x 2 . Therefore,yp D e 2x .1 3x C x 2 /.
5.4.6. If y D ue x , then y 00 y 0 2y D e x .u00 C 2u0 C u/ .u0 C u/ 2u D e x .9 C 2x 4x 2 / so
u00 C u0 2u D 9 C 2x 4x 2, and up D A C Bx C Cx 2, where 2C C .B C 2Cx/ 2.A C Bx C Cx 2/ D
9 C 2x 4x 2 . Therefore, 2C D 4, 2B C 2C D 2, 2A C B C 2C D 9, so C D 2, B D 1, A D 2,
and up D 2 C x C 2x 2 . Therefore, yp D e x . 2 C x C 2x 2/.
5.4.8. If y D ue x , then y 00 3y 0 C 2y D e x .u00 C 2u0 C u/ 3.u0 C u/ C 2u D e x .3 4x/, so
u00 u0 D 3 4x and up D Ax C Bx 2 , where 2B .A C 2Bx/ D 3 4x. Therefore, 2B D 4,
A C 2B D 3, so B D 2, A D 1, and up D x.1 C 2x/. Therefore,yp D xe x .1 C 2x/.
5.4.10. If y D ue 2x , then 2y 00 3y 0 2y D e 2x 2.u00 C 4u0 C 4u/ 3.u0 C 2u/ 2u D e 2x . 6 C
10x/, so 2u00 C 5u0 D 6 C 10x and up D Ax C Bx 2 , where 2.2B/ C 5.A C 2Bx/ D 6 C 10x.
Therefore,10B D 10, 5A C 4B D 6, so B D 1, A D 2, and up D x. 2 C x/. Therefore,yp D
xe 2x . 2 C x/.
5.4.12. If y D ue x , then y 00

2y 0 C y D e x .u00 C 2u0 C u/

2.u0 C u/ C u D e x .1  6x/, 
so
1
u00 D 1 6x Integrating twice and taking the constants of integration to be zero yields up D x 2
x .
2


1
Therefore, yp D x 2 e x
x .
2
 



2u0
u
u
5.4.14. If y D ue x=3 , then 9y 00 C 6y 0 C y D e x=3 9 u00
C
C 6 u0
Cu D
3
9
3
1
x=3
2
2
00
2
00
e
.2 4x C 4x /, so 9u D 2 4x C 4x , or u D .2 4x C 4x /. Integrating twice
9
x2
and taking the constants of integration to be zero yields up D
.3 2x C x 2 /. Therefore, yp D
27
x 2 e x=3
.3 2x C x 2 /.
27
5.4.16. If y D ue x , then y 00 6y 0 C 8y D e x .u00 C 2u0 C u/ 6.u0 C u/ C 8u D e x .11 6x/, so
u00 4u0 C 3u D 11 6x and up D A C Bx, where 4B C 3.A C Bx/ D 11 6x. Therefore,3B D 6,
3A 4B D 11, so B D 2, A D 1 and up D 1 2x. Therefore,yp D e x .1 2x/. The characteristic
polynomial of the complementary equation is p.r / D r 2 6r C 8 D .r 2/.r 4/, so fe 2x ; e 4x g is a
fundamental set of solutions of the complementary equation. Therefore, y D e x .1 2x/C c1 e 2x C c2 e 4x
is the general solution of the nonhomogeneous equation.
5.4.18. If y D ue x , then y 00 C 2y 0 3y D e x .u00 C 2u0 C u/ C 2.u0 C u/ 3u D 16xe x , so
u00 C 4u0 D 16x and up D Ax C Bx 2 , where 2B C 4.A C 2Bx/ D 16x. Therefore,8B D 16,
4AC 2B D 0, so B D 2, A D 1, and up D x.1 2x/. Therefore,yp D xe x .1 2x/. The characteristic
polynomial of the complementary equation is p.r / D r 2 C 2r 3 D .r C 3/.r 1/, so fe x ; e 3x g is a
fundamental set of solutions of the complementary equation. Therefore, y D xe x .1 2x/Cc1e x Cc2e 3x
is the general solution of the nonhomogeneous equation.
5.4.20. If y D ue 2x , then y 00 4y 0 5y D e 2x .u00 C 4u0 C 4u/ 4.u0 C 2u/ 5u D 9e 2x .1 C x/, so
u00 9u D 9 C 9x and up D A C Bx, where 9.A C Bx/ D 9 C 9x. Therefore, 9B D 9, 9A D 9,
so B D 1, A D 1, and up D 1 x. Therefore,yp D e 2x .1 C x/. The characteristic polynomial
of the complementary equation is p.r / D r 2 4r 5 D .r 5/.r C 1/, so fe x ; e 5x g is a fundamental

62 Chapter 5 Linear Second Order Equations I


set of solutions of the complementary equation. Therefore,(A) y D e 2x .1 C x/ C c1 e x C c2 e 5x is
the general solution of the nonhomogeneous equation. Differentiating (A) yields y 0 D 2e 2x .1 C x/
e 2x c1 e x C 5c2e 5x . Now y.0/ D 0; y 0 .0/ D 10 ) 0 D 1 C c1 C c2 ; 10 D 3 c1 C 5c2, so
c1 D 2, c2 D 1. Therefore,y D e 2x .1 C x/ C 2e x e 5x is the solution of the initial value problem.
5.4.22. If y D ue x , then y 00 C 4y 0 C 3y D e x .u00 2u0 C u/ C 4.u0 u/ C 3u D e x .2 C 8x/, so
u00 C 2u0 D 2 8x and up D Ax C Bx 2 , where 2B C 2.A C 2Bx/ D 2 8x. Therefore,4B D 8,
2A C 2B D 2, so B D 2, A D 1, and up D x.1 2x/. Therefore,yp D xe x .1 2x/. The
characteristic polynomial of the complementary equation is p.r / D r 2 C 4r C 3 D .r C 3/.r C 1/,
so fe x ; e 3x g is a fundamental set of solutions of the complementary equation. Therefore,(A) y D
xe x .1 2x/ C c1e x C c2 e 3x is the general solution of the nonhomogeneous equation. Differentiating
(A) yields y 0 D xe x .1 2x/ C e x .1 4x/ c1 e x 3c2e 3x . Now y.0/ D 1; y 0 .0/ D 2 ) 1 D
c1 C c2 ; 2 D 1 c1 3c2 , so c1 D 2, c2 D 1. Therefore,y D e x .2 C x 2x 2 / e 3x is the solution
of the initial value problem.
5.4.24. We must find particular solutions yp1 and yp2 of (A) y 00 C y 0 C y D xe x and (B) y 00 C
y 0 C y D e x .1 C 2x/, respectively. To find a particular solution of (A) we write y D ue x . Then
y 00 C y 0 C y D e x .u00 C 2u0 C u/ C .u0 C u/ C u D xe x so u00 C 3u0 C 3u D x and up D A C Bx,
1
1
where 3B C 3.A C Bx/ D x. Therefore,3B D 1, 3A C 3B D 0, so B D , A D
, and up D
3
3
x
e
1
.1 x/, so yp1 D
.1 x/. To find a particular solution of (B) we write y D ue x . Then
3
3 0
00
0
x
00
y C y C y D e .u
2u C u/ C .u0 u/ C u D e x .1 C 2x/, so u00 u0 C u D 1 C 2x and
up D A C Bx, where B C .A C Bx/ D 1 C 2x. Therefore, B D 2, A B D 1, so A D 3, and
ex
up D 2 C 3x, so yp2 D e x .3 C 2x/. Now yp D yp1 C yp2 D
.1 x/ C e x .3 C 2x/.
3
5.4.26. We must find particular solutions yp1 and yp2 of (A) y 00 8y 0 C 16y D 6xe 4x and (B) y 00
8y 0 C 16y D 2 C 16x C 16x 2, respectively. To find a particular solution of (A) we write y D ue 4x . Then
y 00 8y 0 C 16y D e x .u00 C 8u0 C 16u/ 8.u0 C 4u/ C 16u D 6xe 4x , so u00 D 6x, up D x 3. and
yp1 D x 3 e 4x . To find a particular solution of (B) we write yp D ACBxCCx 2 . Then yp00 8yp0 C16yp D
2C 8.B C2Cx/C16.ACBx CCx 2 / D .16A 8B C2C /C.16B 16C /x C16Cx 2 D 2C16x C16x 2
if 16C D 16, 16B 16C D 16, 16A 8B C 2C D 2. Therefore,C D 1, B D 2, A D 1, and
yp2 D 1 C 2x C x 2 . Now yp D yp1 C yp2 D x 3 e 4x C 1 C 2x C x 2 .
5.4.28. We must find particular solutions yp1 and yp2 of (A) y 00 2y 0 C 2y D e x .1 C x/ and (B)
y 00 2y 0 C 2y D e x .2 8x C 5x 2/, respectively. To find a particular solution of (A) we write y D ue x .
Then y 00 2y 0 C 2y D e x .u00 C 2u0 C u/ 2.u0 C u/ C 2u D e x .1 C x/, so u00 C u D 1 C x and
up D 1 C x, so yp1 D e x .1 C x/. To find a particular solution of (B) we write y D ue x . Then
y 00 2y 0 C 2y D e x .u00 2u0 C u/ 2.u0 u/ C 2u D e x .2 8x C 5x 2 /, so u00 4u0 C 5u D
2 8x C 5x 2 and up D A C Bx C Cx 2 , where 2C 4.B C 2Cx/ C 5.A C Bx C Cx 2 / D 2 8x C 5x 2.
Therefore,5C D 5, 5B 8C D 8, 5A 4B C 2C D 2, so C D 1, B D 0, A D 0, and up D x 2 .
Therefore,yp2 D x 2 e x . Now yp D yp1 C yp2 D e x .1 C x/ C x 2 e x .


5.4.30.
(a) If y D ue x , then ay 00 C by 0 C cy
D e x a.u00 C 2u0 C 2 u/ C b.u0 C u/ C cu D


e x au00 C .2a C b//u0 C .a 2 C b C c/u D e x .au00 Cp 0 ./u0 Cp./u/. Therefore,ay 00 Cby 0 C
cy D e x G.x/ if and only if au00 C p 0 ./u0 C p./u D G.x/.
(b) Substituting up D A C Bx C Cx 2 C Dx 3 into (B) yields
a.2C C 6Dx/ C p 0 ./.B C 2Cx C 3Dx 2 / C p./.A C Bx C Cx 2 C Dx 3 /
D p./A C p 0 ./B C 2aC C p./B C 2p 0 ./C C 6aDx
Cp./C C 3p 0./Dx 2 C p./Dx 3 D g0 C g1 x C g2 x 2 C g3 x 3

Section 5.4 The Method of Undetermined Coefficients I


if

D g3
D g2
D g1
D g0 :

p./D
p./C C 3p 0 ./D
p./B C 2p 0 ./C C 6aD
p./A C p 0 ./B C 2aC

63

.C/

Since e x is not a solution of the complementary equation, p./ 0. Therefore,the triangular system
(C) can be solved successively for D, C , B and A.
(c) Since e x is a solution of the complementary equation while xe x is not, p./ D 0 and p 0 ./ 0.
Therefore, (B) reduces to (D) au00 C p 0 ./u D G.x/. Substituting up D Ax C Bx 2 C Cx 3 C Dx 4 into
(D) yields
a.2B C 6Cx C 12Dx 2/ C p 0 ./.A C 2Bx C 3Cx 2 C 4Dx 3 /
D .p 0 ./A C 2aB/ C .2p 0 ./B C 6aC /x C .3p 0 ./C C 12aD/x 2
C4p 0 ./Dx 3 D g0 C g1 x C g2 x 2 C g3 x 3
if

4p 0 ./D
3p ./C C 12aD
2p 0 ./B C 6aC
p 0 ./A C 2aB
0

D g3
D g2
D g1
D g0 :

Since p 0 ./ 0 this triangular system can be solved successively for D, C , B and A.


(d) Since e x and xe x are solutions of the complementary equation, p./ D 0 and p 0 ./ D 0.
G.x/
Therefore, (B) reduces to (D) au00 D G.x/, so u00 D
. Integrating this twice and taking the
a
g
g1
g2
g3 3 
0
constants of integration yields the particular solution up D x 2
C x C x2 C
x .
2
6
12
20

5.4.32. If yp D Axe 4x , then yp00


if A D 1, so yp D 5xe 4x .

7yp0 C 12yp D .8 C 16x/

7.1 C 4x/ C 12xAe 4x D Ae 4x D 5e 4x

5.4.34. If yp D e 3x .A C Bx C Cx 2 /, then
yp00

3yp0 C 2yp

e 3x .9A C 6B C 2C / C .9B C 12C /x C 9Cx 2

D
D

e 3x .2A C 3B C 2C / C .2B C 6C /x C 2Cx 2


e 3x . 1 C 2x C x 2 /

3e 3x .3A C B/ C .3B C 2C /x C 3Cx 2


C2e 3x .A C Bx C Cx 2 /

if 2C D 1; 2B C 6C D 2; 2A C 3B C 2C D
yp D

e 3x
.1 C 2x
4

5.4.36. If yp D e

x=2

1. Therefore,C D

1
,B D
2

1
,A D
2

2x 2 /.
.Ax 2 C Bx 3 C Cx 4 /, then

4yp00 C 4yp0 C yp

D e

x=2

8A

.8A

24B/x C .A

12B C 48C /x 2

Ce x=2 .B 16C /x 3 C Cx 4
Ce x=28Ax .2A 12B/x 2 .2B 16C /x 3 2Cx 4
Ce x=2 .Ax 2 C Bx 3 C Cx 4 /
D e x=2 .8A C 24Bx C 48Cx 2 / D e x=2. 8 C 48x C 144x 2/

1
, and
4

64 Chapter 5 Linear Second Order Equations


if 48C D 144, 24B D 48, and 8A D 8. Therefore,C D 3, B D 2, A D 1, and yp D x 2 e x=2. 1 C
2x C 3x 2 /.
R
5.4.38. If y D e x P .x/ dx, then y 0 D e x P .x/. Let y D ue x ; then .u0 Cu/e x D e x P .x/, which
implies (A). We must show that it is possible to choose A0 ; : : : ; Ak so that (B) .A0 C A1 x    C Ak x k /0 C
.A0 C A1 x    C Ak x k / D p0 C p1 x C    C pk x k . By equating the coefficients of x k ; x k 1; : : : ; 1
(in that order) on the two sides of (B), we see that (B) holds if and only if Ak D pk and .k j C
1/Ak j C1 C Ak D pk j ; 1  j  k.
R
5.4.40. If y D x k e x dx, then y 0 D x k e x . Let y D ue x ; then .u0 C u/e x D x k e x , so
u0 C u D x k . This equation has a particular solution up D A0 C A1 x    C Ak x k , where (A) .A0 C
A1 x    C Ak x k /0 C .A0 C A1 x    C Ak x k / D x k . By equating the coefficients of x k ; x k 1; : : : ; 1
on the two sides of (A), we see that (A) holds if and only if Ak D 1 and .k j C 1/Ak j C1 C
1
k
k.k 1/
Ak j D 0; 1  j  k. Therefore, Ak D , Ak 1 D
, Ak 2 D
, and, in general,
2

3
j
. 1/ k
k.k 1/    .k j C 1/
Ak j D . 1/j
D j C1
; 1  j  k. By introducing the index
j
C1

.k j /
k
. 1/k k X . x/r
. 1/k r k
r D k j we can rewrite this as Ar D k r C1 ; 0  r  k. Therefore, up D
r

r
kC1
k

and y D

. 1/ ke
kC1

r D0

k
X
. x/r
C c.
r
r D0

5.5 THE METHOD OF UNDETERMINED COEFFICIENTS II


5.5.2. Let
yp
yp0
yp00

D .A0 C A1 x/ cos x C .B0 C B1x/ sin xI then

D .A1 C B0 C B1x/ cos x C .B1 A0 A1 x/ sin x


D .2B1 A0 A1 x/ cos x .2A1 C B0 C B1 x/ sin x; so
yp00 C 3yp0 C yp

D .3A1 C 3B0 C 2B1 C 3B1 x/ cos x


C.3B1 3A0 2A1 3A1 x/ sin x

D .2

6x/ cos x

9 sin x

if 3B1 D 6, 3A1 D 0, 3B0 C 3A1 C 2B1 D 2, 3A0 C 3B1 C 2A1 D


B1 D 2, A0 D 1, B0 D 2, and yp D cos x C .2 2x/ sin x.

9. Therefore,A1 D 0,

5.5.4. Let y D ue 2x . Then


y 00 C 3y 0

2y

if u00 C 7u0 C 8u D 5 cos 2x


u00p C 7u0p C 8up


D e 2x .u00 C 4u0 C 4u/ C 3.u0 C 2u/
D e 2x .u00 C 7u0 C 8u/ D


2u

e 2x .5 cos 2x C 9 sin 2x/

9 sin 2x. Now let up D A cos 2x C B sin 2x. Then


D

4.A cos 2x C B sin 2x/ C 14. A sin 2x C B cos 2x/


C8.A cos 2x C B sin 2x/
D .4A C 14B/ cos 2x .14A 4B/ sin 2x
D
5 cos 2x 9 sin 2x

Section 5.5 The Method of Undetermined Coefficients II


1
,B D
2

if 4A C 14B D 5, 14A C 4B D 9. Therefore,A D


1
up D .cos 2x
2
2x

5.5.6. Let y D ue

u0

2y

D e

D e
D e

2x
2x
2x

4u D .4 C 20x/ cos 3x C .26

e 2x
.cos 2x
2

sin 2x/:

 00
.u

4u0 C 4u/ C 3.u0

.u00 u0 4u/
.4 C 20x/ cos 3x C .26

2u/


2u

32x/ sin 3x

32x/ sin 3x. Let

D .A0 C A1 x/ cos 3x C .B0 C B1 x/ sin 3xI then


D .A1 C 3B0 C 3B1x/ cos 3x C .B1 3A0 3A1x/ sin 3x

up
u0p
u00p

u00p

1
,
2

. Then

y 00 C 3y 0

if u00

sin 2x/; and yp D

65

D .6B1

u0p

4up

13A1
3A1

3B1
13B1

9A0

9A1 x/ cos 3x

.2A1 C 9B0 C 9B1 x/ sin 3x; so

13A0 C A1 C 3B0 6B1 C .13A1 C 3B1 /x cos 3x


13B0 C B1 3A0 C 6A1 C .13B1 3A1 /x sin 3x
D .4 C 20x/ cos 3x C .26 32x/ sin 3x if
D
D

20
and
32

13A0
3A0

3B0
13B0

A1 C 6B1
6A1
B1

D
4
D 26:

From the first two equations, A1 D 2, B1 D 2. Substituting these in the last two equations yields
13A0 3B0 D 10, 3A0 13B0 D 16. Solving this pair yields A0 D 1, B0 D 1. Therefore,
up D .1

2x/.cos 3x

sin 3x/ and yp D e

2x

.1

2x/.cos 3x

sin 3x/:

5.5.8. Let
yp
yp0
yp00

D .A0 x C A1 x 2 / cos x C .B0 x C B1 x 2 / sin xI then





D A0 C .2A1 C B0 /x C B1 x 2 cos x C B0 C .2B1


D 2A1 C 2B0 .A0 4B1/x A1 x 2 cos x


C 2B1 2A0 .B0 C 4A1 /x B1x 2 sin x; so

yp00 C yp

D .2A1 C 2B0 C 4B1 x/ cos x C .2B1

D . 4 C 8x/ cos x C .8

A0 /x

2A0


B1 x 2 sin x

4A1 x/ sin x

4x/ sin x

if 4B1 D 8, 4A1 D 4, 2B0 C 2A1 D 4, 2A0 C 2B1 D 8. Therefore, A1 D 1, B1 D 2, A0 D 2,


B0 D 3, and yp D x .2 x/ cos x C .3 2x/ sin x.
5.5.10. Let y D ue

. Then

y 00 C 2y 0 C 2y

D e
D e

.u00 2u0 C u/ C 2.u0


x 00
.u C u/ D e x .8 cos x



u/ C 2u
6 sin x/

66 Chapter 5 Linear Second Order Equations


if u00 C u D 8 cos x

6 sin x. Now let


up
u0p
u00p

D Ax cos x C Bx sin xI then


D .A C Bx/ cos x C .B Ax/ sin x

D .2B

Ax/ cos x

u00p C up D 2B cos x

.2A C Bx/ sin x; so

2A sin x D 8 cos x

6 sin x

if 2B D 8, 2A D 6. Therefore,A D 3, B D 4, up D x.3 cos x C 4 sin x/, and yp D xe


4 sin x/.

.3 cos x C

5.5.12. Let
yp
yp0

D
D

yp00

yp00 C 2yp0 C yp

.A0 C A1 x C A2 x 2 / cos x C .B0 C B1 x C B2 x 2 / sin xI then




A1 C B0 C .2A2 C B1/x C B2 x 2 cos x


C B1 A0 C .2B2 A1 /x A2 x 2 sin x;


A0 C 2A2 C 2B1 .A1 4B2/x A2 x 2 cos x


C B0 C 2B2 2A1 .B1 C 4A2 /x B2 x 2 sin x; so



D 2 A1 C A2 C B0 C B1 C .2A2 C B1 C 2B2/x C B2 x 2 cos x


C2 B1 C B2 A0 A1 C .2B2 A1 2A2 /x A2 x 2 sin x

D 8x 2 cos x
(i)

2B2
2A2

4x sin x if

D 8
; (ii)
D 0

2B1 C 4A2 C 4B2


2A1 4A2 C 4B2

2B0 C 2A1 C 2B1 C 2A2


2A0 2A1 C 2B1 C 2B2

(iii)

D
D

D
D

0
;
4

0
:
0

From (i), A2 D 0, B2 D 4. Substituting these into (ii) and solving for A1 and B1 yields A1 D 10,
B1 D 8. Substituting the known coefficients into (iii) and solving for A0 and B0 yields A0 D 14,
B0 D 2. Therefore,yp D .14 10x/ cos x .2 C 8x 4x 2/ sin x.
5.5.14. Let
yp
yp0
yp00

D .A0 C A1 x C A2 x 2 / cos 2x C .B0 C B1 x C B2 x 2 / sin 2xI then




D A1 C 2B0 C .2A2 C 2B1/x C 2B2x 2 cos 2x


C B1 2A0 C .2B2 2A1 /x 2A2 x 2 sin 2x


D
4A0 C 2A2 C 4B1 .4A1 8B2 /x 4A2x 2 cos 2x


C 4B0 2B2 4A1 .4B1 C 8A2 /x 4B2 x 2 sin 2x; so

yp00 C 3yp0 C 2yp

D 2A0 C 3A1 C 4A2 C 6B0 C 4B1


.2A1 6A2 6B1 8B2 /x
C 2B0 C 3B1 C 4B2 6A0
.2B1

D .1

6B2 C 6A1 C 8A2 /x

4x 2 / cos 2x

.2A2
4A1


6B2 /x 2 cos 2x


.2B2 C 6A2 /x 2 sin 2x

.1 C 7x C 2x 2 / sin 2x if

Section 5.5 The Method of Undetermined Coefficients II


2A2 C 6B2
6A2 2B2

(i)

D
D

4
; (ii)
2

2A1 C 6B1 C 6A2 C 8B2


6A1 2B1 8A2 C 6B2

2A0 C 6B0 C 3A1 C 4B1 C 2A2


6A0 2B0 4A0 C 3B1 C 2B2

(iii)

D
D

D
D

67

1
;
7

1
:
1

1
1
. Substituting these into (ii) and solving for A1 and B1 yields A1 D 0,
From (i), A2 D , B2 D
2
2
B1 D 0. Substituting the known coefficients into (iii) and solving for A0 and B0 yields A0 D 0, B0 D 0.
x2
.cos 2x sin 2x/.
Therefore, yp D
2
5.5.16. Let y D ue x . Then
y 00

if u00 D .3 C 4x
up
u0p
u00p

2y 0 C y



D e x .u00 C 2u0 C u/ 2.u0 C u/ C u D e x u00


D
e x .3 C 4x x 2 / cos x C .3 4x x 2 / sin x

x 2 / cos x

.3

x 2 / sin x. Now let

4x

D .A0 C A1 x C A2 x 2 / cos x C .B0 C B1 x C B2x 2 / sin xI then




D A1 C B0 C .2A2 C B1 /x C B2 x 2 cos x


C B1 A0 C .2B2 A1 /x A2 x 2 sin x;


D
A0 C 2A2 C 2B1 .A1 4B2 /x A2 x 2 cos x


C B0 C 2B2 2A1 .B1 C 4A2 /x B2x 2 sin x

.3 C 4x

A2
B2

(i)

x 2/ cos x
D
D

(iii)

1
; (ii)
1

.3

4x

x 2 / sin x if

A1 C 4B2
B1 4A2

A0 C 2B1 C 2A2
B0 2A1 C 2B2

D
D

D
D

4
;
4

3
:
3

From (i), A2 D 1, B2 D 1. Substituting these into (ii) and solving for A1 and B1 yields A1 D 0,
B1 D 0. Substituting the known coefficients into (iii) and solving for A0 and B0 yields A0 D 1, B0 D 1.
Therefore,up D .1 x 2 /.cos x C sin x/ and yp D e x .1 x 2 /.cos x C sin x/.
5.5.18. Let y D ue

. Then

y 00 C 2y 0 C y
if u00 D .5

2x/ cos x
up
u0p
u00p

D e
D e


.u00 2u0 C u/ C 2.u0 u/ C u
x 00
u D e x .5 2x/ cos x .3 C 3x/ sin x


.3 C 3x/ sin x. Let


D

D
D
D

.A0 C A1 x/ cos x C .B0 C B1 x/ sin xI then

.A1 C B0 C B1 x/ cos x C .B1 A0 A1 x/ sin x


.2B1 A0 A1 x/ cos x .2A1 C B0 C B1 x/ sin x
.5 2x/ cos x .3 C 3x/ sin x

if A1 D 2, B1 D 3, A0 C 2B1 D 5, B0 2A1 =-3. Therefore, A1 D 2, B1 D 3, A0 D 1,


B0 D 1, up D e x .1 C 2x/ cos x .1 3x/ sin x, and yp D e x .1 C 2x/ cos x .1 3x/ sin x.

68 Chapter 5 Linear Second Order Equations


5.5.20. Let
D .A0 x C A1 x 2 C A2 x 3/ cos x C .B0 x C B1 x 2 C B2 x 3 / sin xI then


D A0 C .2A1 C B0 /x C .3A2 C B1/x 2 C B2 x 3 cos x


C B0 C .2B1 A0 /x C .3B2 A1 /x 2 A2 x 3 sin x


D 2A1 C 2B0 .A0 6A2 4B1/x .A1 6B2/x 2 A2 x 3 cos x


C 2B1 2A0 .B0 C 6B2 C 4A1 /x .B1 C 6A2 /x 2 B2 x 3 sin x; so

yp
yp0
yp00

yp00 C yp



2A1 C 2B0 C .6A2 C 4B1 /x C 6B2 x 2 cos x


C 2B1 2A0 C .6B2 4A1 /x 6A2 x 2 sin x

.2 C 2x/ cos x C .4 C 6x 2 / sin x if

D
6B2
6A2

(i)

D
D

0
; (ii)
6

4B1 C 6A2
4A1 C 6B2

D 2
; (iii)
D 0

2B0 C 2A1
2A0 C 2B1

D 2
:
D 4

From (i), A2 D 1, B2 D 0. Substituting these into (ii) and solving for A1 and B1 yields A1 D 0,
B1 D 2. Substituting the known coefficients into (iii) and solving for A0 and B0 yields A0 D 0, B0 D 2.
Therefore,yp D x 3 cos x C .x C 2x 2 / sin x.
5.5.22. Let y D ue x . Then
y 00

if u00

7y 0 C 6y



D e x .u00 C 2u0 C u/ 7.u0 C u/ C 6u
D e x .u00 5u0 / D e x .17 cos x 7 sin x/

5u0 D 17 cos x C 7 sin x. Now let up D A cos x C B sin x. Then


u00p

5u0p

.A cos x C B sin x/

D . A

5B/ cos x

.B

5. A sin x C B cos x/

5A/ sin x D 17 cos x C 7 sin x

if A 5B D 17, 5A B D 7. Therefore,A D 2, B D 3, up D 2 cos x C 3 sin x, and yp D


e x .2 cos x C 3 sin x/. The characteristic polynomial of the complementary equation is p.r / D r 2
7r C 6 D .r 1/.r 6/, so fe x ; e 6x g is a fundamental set of solutions of the complementary equation.
Therefore, (A) y D e x .2 cos x C 3 sin x/ C c1 e x C c2 e 6x is the general solution of the nonhomogeneous
equation. Differentiating (A) yields y 0 D e x .2 cos x C 3 sin x/C e x . 2 sin x C 3 cos x/C c1 e x C 6c2e 6x ,
so y.0/ D 4; y 0 .0/ D 2 ) 4 D 2 C c1 C c2; 2 D 2 C 3 C c1 C 6c2 ) c1 C c2 D 2; c1 C 6c2 D 3,
so c1 D 3, c2 D 1, and y D e x .2 cos x C 3 sin x/ C 3e x e 6x .
5.5.24. Let y D ue x . Then
y 00 C 6y 0 C 10y



D e x .u00 C 2u0 C u/ C 6.u0 C u/ C 10u
D e x .u00 C 8u0 C 17u/ D 40e x sin x

if u00 C 8u0 C 17u D 40 sin x. Let up D A cos x C B sin x. Then


u00p C 6u0p C 17up

.A cos x C B sin x/ C 8. A sin x C B cos x/

C17.A cos x C B sin x/


D .16A C 8B/ cos x .8A

16B/ sin x D 40 sin x

Section 5.5 The Method of Undetermined Coefficients II

69

if 16A C 8B D 0, 8A C 16B D 40. Therefore,A D 1, B D 2, and yp D e x .cos x 2 sin x/.


The characteristic polynomial of the complementary equation is p.r / D r 2 C 6r C 10 D .r C 3/2 C 1,
so fe 3x cos x; e 3x sin xg is a fundamental set of solutions of the complementary equation, and (A)
y D e x .cos x 2 sin x/ C e 3x .c1 cos x C c2 sin x/ is the general solution of the nonhomogeneous
equation. Therefore,y.0/ D 2 ) 2 D 1 C c1 , so c1 D 1. Differentiating (A) yields y 0 D e x .cos x
2 sin x/ e x .sin x C2 cos x/ 3e 3x .c1 cos x Cc2 sin x/Ce 3x . c1 sin x Cc2 cos x/. Therefore,y 0 .0/ D
3 ) 3 D 1 2 3c1 C c2, so c2 D 1, and y D e x .cos x 2 sin x/ C e 3x .cos x C sin x/.
5.5.26. Let y D ue 3x . Then
y 00

3y 0 C 2y


D e 3x .u00 C 6u0 C 9u/

D e 3x .u00 C 3u0 C 2u/ D e 3x 21 cos x

if u00 C 3u0 C 2u D 21 cos x


up
u0p
u00p

D
D
D


3.u0 C 3u/ C 2u

.11 C 10x/ sin x

.11 C 10x/ sin x. Now let


.A0 C A1 x/ cos x C .B0 C B1 x/ sin xI then
.A1 C B0 C B1 x/ cos x C .B1 A0 A1 x/ sin x
.2B1 A0 A1 x/ cos x .2A1 C B0 C B1 x/ sin x; so

u00 C 3u0 C 2u D A0 C 3A1 C 3B0 C 2B1 C .A1 C 3B1/x cos x


C B0 C 3B1 3A0 2A1 C .B1 3A1 /x sin x
D 21 cos x .11 C 10x/ sin x if
A1 C 3B1
3A1 C B1

D
D

0
and
10

A0 C 3B0 C 3A1 C 2B1


3A0 C B0 2A1 C 3B1

D
D

21
:
11

From the first two equations A1 D 3, B1 D 1. Substituting these in last two equations yields and
solving for A0 and B0 yields A0 D 2, B0 D 4. Therefore, up D .2 C 3x/ cos x C .4 x/ sin x and
yp D e 3x .2 C 3x/ cos x C .4 x/ sin x. The characteristic polynomial of the complementary equation
is p.r / D r 2 3r C2 D .r 1/.r 2/, so fe x ; e 2x g is a fundamental set of solutions of the complementary
equation, and (A) y D e 3x .2 C 3x/ cos x C .4 x/ sin x C c1 e x C c2 e 2x is the general solution of the
nonhomogeneous equation. Differentiating (A) yields
y0

D 3e 3x .2 C 3x/ cos x C .4 x/ sin x


Ce 3x .7 x/ cos x .3 C 3x/ sin x C c1 e x C 2c2e 2x :

Therefore,y.0/ D 0; y 0 .0/ D 6 ) 0 D 2 C c1 C c2 ; 6 D 6 C 7 C c1 C 2c2, so c1 C c2 D 2; c1 C 2c2 D


7. Therefore, c1 D 3, c2 D 5, and y D e 3x .2 C 3x/ cos x C .4 x/ sin x C 3e x 5e 2x .
5.5.28. We must find particular solutions yp1 , yp2 , and yp3 of (A) y 00 C y D 4 cos x 2 sin x and (B)
y 00 C y D xe x , and (C) y 00 C y D e x , respectively. To find a particular solution of (A) we write
yp1
yp0 1
yp001

D
D
D

Ax cos x C Bx sin xI then


.A C Bx/ cos x C .B Ax/ sin x
.2B Ax/ cos x .2A C Bx/ sin x; so

yp001 C yp1 D 2B cos x 2A sin x D 4 cos x 2 sin x if 2B D 4, 2A D 2. Therefore, A D 1, B D 2,


and yp1 D x.cos x C 2 sin x/. To find a particular solution of (B) we write y D ue x . Then


y 00 C y D e x .u00 C 2u0 C u/ C u
D e x .u00 C 2u0 C 2u/ D xe x

70 Chapter 5 Linear Second Order Equations


if u00 C 2u0 C 2u D x. Now up D A C Bx, where 2B C 2.A C Bx/ D x. Therefore, 2B D 1,
1
1
1
ex
2A C 2B D 0, so B D , A D
, up D
.1 x/, and yp2 D
.1 x/. To find a particular
2
2
2
2
1
solution of (C) we write yp3 D Ae x . Then yp003 C yp3 D 2Ae x D e x if 2A D 1, so A D and
2
e x
ex
e x
Now yp D yp1 C yp2 C yp3 D x.cos x C 2 sin x/
.1 x/ C
.
yp3 D
2
2
2
5.5.30. We must find particular solutions yp1 , yp2 and yp3 of (A) y 00 2y 0 C 2y D 4xe x cos x, (B)
y 00 2y 0 C 2y D xe x , and (C) y 00 2y 0 C 2y D 1 C x 2 , respectively. To find a particular solution of
(A) we write y D ue x . Then y 00 2y 0 C 2y D e x .u00 C 2u0 C u/ 2.u0 C u/ C 2u D e x .u00 C u/ D
4xe x cos x if u00 C u D 4x cos x. Now let
up
u0p
u00p

D .A0 x C A1 x 2 / cos x C .B0 x C B1 x 2 / sin xI then





D A0 C .2A1 C B0 /x C B1 x 2 cos x C B0 C .2B1


D 2A1 C 2B0 .A0 4B1 /x A1 x 2 cos x


C 2B1 2A0 .B0 C 4A1 /x B1 x 2 sin x; so

u00p C up

D .2A1 C 2B0 C 4B1 x/ cos x C .2B1

A0 /x

2A0


B1x 2 sin x

4A1 x/ sin x

D 4x cos x

if 4B1 D 4, 4A1 D 0, 2B0 C 2A1 D 0, 2A0 C 2B1 D 0. Therefore, A1 D 0, B1 D 1, A0 D 1,


B0 D 0, up D x.cos x C x sin x/, and yp1 D xe x .cos x C x sin x/. To find a particular solution of (B)
we write y D ue x . Then


y 00 2y 0 C 2y D e x .u00 2u0 C u/ 2.u0 u/ C 2u
D e x .u00 4u0 C 5u/ D xe x

if u00

4u0 C 5u D x. Now up D A C Bx where 4B C 5.A C Bx/ D x. Therefore, 5B D 1,


4
1
e x
1
5A 4B D 0, B D , A D
, up D
.4 C 5x/, and yp2 D
.4 C 5x/. To find a particular
5
25
25
25
2
solution of (C) we write yp3 D A C Bx C Cx . Then
yp003

2yp0 3 C 2yp3

D 2C

D .2A

2.B C 2Cx/ C 2.A C Bx C Cx 2 /


2B C 2C / C .2B

4C /x C 2Cx 2 D 1 C x 2

1
x2
, B D 1, A D 1, and yp3 D 1 C x C .
2
2
x
2
e
x
.4 C 5x/ C 1 C x C
.
D xe x .cos x C x sin x/ C
25
2

if 2A 2B C 2C D 1, 2B 4C D 0, 2C D 1. Therefore,C D
Now yp D yp1 C yp2 C yp3

5.5.32. We must find particular solutions yp1 and yp2 of (A) y 00 4y 0 C4y D 6e 2x and (B) y 00 4y 0 C4y D
25 sin x, respectively. To find a particular solution of (A), let y D ue 2x . Then


y 00 4y 0 C 4y D e 2x .u00 C 4u0 C 4u/ 4.u0 C 2u/ C 4u
D e 2x u00 D 6e 2x

if u00 D 6. Integrating twice and taking the constants of integration to be zero yields up D 3x 2 , so
yp1 D 3x 2e 2x . To find a particular solution of (B), let yp2 D A cos x C B sin x. Then
yp002

4yp0 2 C 4yp2

.A cos x C B sin x/ 4. A sin x C B cos x/


C4.A cos x C B sin x/
D .3A 4B/ cos x C .4A C 3B/ sin x D 25 sin x

Section 5.5 The Method of Undetermined Coefficients II

71

if 3A 4B D 0, 4A C 3B D 25. Therefore,A D 4, B D 3, and yp2 D 4 cos x C 3 sin x. Now


yp D yp1 C yp2 D 3x 2 e 2x C 4 cos x C 3 sin x. The characteristic polynomial of the complementary
equation is p.r / D r 2 4r C 4 D .r 2/2 , so fe 2x ; xe 2x g is a fundamental set of solutions of the
complementary equation. Therefore,(C) y D 3x 2 e 2x C 4 cos x C 3 sin x C e 2x .c1 C c2 x/ is the general
solution of the nonhomogeneous equation. Now y.0/ D 5 ) 5 D 4 C c1 , so c1 D 1. Differentiating (C)
yields y 0 D 6e 2x .x C x 2/ 4 sin x C 3 cos x C 2e 2x .c1 C c2x/ C c2 e 2x , so y 0 .0/ D 3 ) 3 D 3 C 2 C c2.
Therefore,c2 D 2, and y D .1 2x C 3x 2/e 2x C 4 cos x C 3 sin x.
5.5.34. We must find particular solutions yp1 and yp2 of (A) y 00 C 4y 0 C 4y D 2 cos 2x C 3 sin 2x
and (B) y 00 C 4y 0 C 4y D e x , respectively. To find a particular solution of (A) we write yp1 D
A cos 2x C B sin 2x. Then
yp001 C 4yp0 1 C 4yp1

4.A cos 2x C B sin 2x/ C 8. A sin 2x C B cos 2x/


C4.A cos 2x C B sin 2x/ D 8A sin 2x C 8B cos 2x

D 2 cos 2x C 3 sin 2x

1
3
1
3
, B D , and yp1 D
cos 2x C sin 2x. To find a
8 x
4
8
4
particular solution of (B) we write yp2 D Ae . Then yp002 C 4yp0 2 C 4yp2 D A.1 4 C 4/e x D
3
1
Ae x D e x if A D 1. Therefore,yp2 D e x . Now yp D yp1 C yp2 D
cos 2x C sin 2x C e x .
8
4
The characteristic polynomial of the complementary equation is p.r / D r 2 C 4r C 4 D .r 2/2 ,
so fe 2x ; xe 2x g is a fundamental set of solutions of the complementary equation. Therefore,(C) y D
3
1
cos 2x C sin 2x C e x C e 2x .c1 C c2 x/ is the general solution of the nonhomogeneous equation.
8
4
13
3
3
Now y.0/ D 1 ) 1 D
C 1 C c1 , so c1 D
. Differentiating (C) yields y 0 D sin 2x C
8
8
4
1
1
3
x
2x
2x
0
cos 2x e
2e .c1 C c2x/C c2 e
, so y .0/ D 2 ) 2 D
1 2c1 C c2. Therefore,c2 D
,
2
2
4
1
13 2x 3
3
and y D
cos 2x C sin 2x C e x
e
xe 2x .
8
4
8
4

if 8B D 2,

8A D 3. Therefore,A D

5.5.36. (a), (b), and (c) require only routine manipulations. (d) The coefficients of sin !x in yp0 , yp00 ,
ayp00 Cbyp0 Ccyp , and yp00 C! 2 yp can be obtained by replacing A by B and B by A in the corresponding
coefficients of cos !x.
5.5.38. Let y D ue x . Then
ay 00 C by 0 C cy



D e x a.u00 C 2u0 C 2 u/ C b.u0 C u/ C cu


D e x au00 C .2a C b/u0 C .a2 C b C c/u


D e x au00 C p 0 ./u0 C p./u
D e x .P .x/ cos !x C Q.x/ sin !x/ if

(A) au00 C p 0 ./u0 C p./uP .x/ cos !x C Q.x/ sin !x, where p.r / D ar C br C c is that characteristic
polynomial of the complementary equation (B) ay 00 C by 0 C cy D 0. If e x cos !x and e x sin !x are
not solutions of (B), then cos !x and sin !x are not solutions of the complementary equation for (A).
Then Theorem 5.5.1 implies that (A) has a particular solution
up D .A0 C A1 x C    C Ak x k / cos !x C .B0 C B1 x C    C Bk x k / sin !x;
and yp D up e x is a particular solution of the stated form for the given equation. If e x cos !x and
e x sin !x are solutions of (B), then cos !x and sin !x are solutions of the complementary equation for

72 Chapter 5 Linear Second Order Equations


(A). Then Theorem 5.5.1 implies that (A) has a particular solution
up D .A0 x C A1 x 2 C    C Ak x kC1 / cos !x C .B0 x C B1 x 2 C    C Bk x kC1 / sin !x;
and yp D up e x is a particular solution of the stated form for the given equation.
R
5.5.40. (a) Let y D x 2 cos x dx; then y 0 D x 2 cos x Now let

(i)

.A0 C A1 x C A2 x 2 / cos x C .B0 C B1 x C B2 x 2 / sin xI then




A1 C B0 C .2A2 C B1/x C B2 x 2 cos x


C B1 A0 C .2B2 A1 /x A2 x 2 sin x D x 2 cos x if

D
D

yp
yp0

D 1
; (ii)
D 0

B2
A2

B1 C 2A2
A1 C 2B2

D 0
; (iii)
D 0

B0 C A1
A0 C B1

D
D

0
:
0

Solving these equations yields A2 D 0, B2 D 1, A1 D 2, B1 D 0, A0 D 0, B0 D 2. Therefore,yp D


2x cos x .2 Rx 2 / sin x and y D 2x cos x .2 x 2 / sin x C c.
(b) Let y D x 2 e x cos x dx D ue x ; then y 0 D .u0 C u/e x D x 2 e x cos x if u0 C u D x 2 cos x. Now
let
D .A0 C A1 x C A2 x 2 / cos x C .B0 C B1 x C B2x 2 / sin xI then


D A1 C B0 C .2A2 C B1 /x C B2 x 2 cos x


C B1 A0 C .2B2 A1 /x A2 x 2 sin x; so

up
u0p

u00p C up



A0 C A1 C B0 C .A1 C 2A2 C B1 /x C .A2 C B2 /x 2 cos x


C B0 C B1 A0 C .B1 C 2B2 A1 /x C .B2 A2 /x 2 sin x

D x 2 cos x if
(i)

A2 C B2
A2 C B2

D 1
; (ii)
D 0

(iii)

A1 C B1 C 2A2
A1 C B1 C 2B2

A0 C B0 C A1
A0 C B0 C B1

D 0
;
D 0

D 0
:
D 0

1
1
, B2 D . Substituting these into (ii) and solving for A1 and B1 yields A1 D 0,
2
2
1
1
B1 D 1. Substituting these into (iii) and solving for A0 and B0 yields A0 D
, B0 D . Therefore,
2
2


1
ex 
up D
.1 x 2 / cos x .1 x/2 sin x and y D
.1 x 2 / cos x .1 x/2 sin x .
2
2
R
(c) Let y D xe x sin 2x dx D ue x ; then y 0 D .u0 u/e x D xe x sin 2x if u0 u D x sin 2x.
Now let

From (i), A2 D

up

D .A0 C A1 x/ cos 2x C .B0 C B1x/ sin 2xI then

u0p

D .A1 C 2B0/ C 2B1x cos 2x C .B1

u00p

up

2A0 /

2A1 x sin 2x; so

D A0 C A1 C 2B0 .A1 2B1/x cos 2x


C B0 C B1 2A0 .B1 C 2A1 /x sin 2x D x sin 2x if

Section 5.5 The Method of Undetermined Coefficients II


A1 C 2B1
2A1
B1

(i)

2
From (i), A1 D
, B1 D
5
3
. Therefore,
B0 D
25

D 0
; (ii)
D 1

A0 C 2B0 C A1
2A0
B0 C B1

D 0
:
D 0

1
. Substituting these into (ii) and solving for A0 and B0 yields A0 D
5

1
.4 C 10x/ cos 2x
25

up D

.3

73

4
,
25

5x/ sin 2x C c and

e x
.4 C 10x/ cos 2x .3 5x/ sin 2x C c:
25
R
(d) Let y D x 2 e x sin x dx D ue x ; then y 0 D .u0 u/e x D x 2 e x sin x if u0
Now let
yp D

up
u0p

u00p

up

u D x 2 sin x.

D .A0 C A1 x C A2 x 2 / cos x C .B0 C B1 x C B2x 2 / sin xI then




D A1 C B0 C .2A2 C B1 /x C B2 x 2 cos x


C B1 A0 C .2B2 A1 /x A2 x 2 sin x; so
D


A0 C A1 C B0 .A1 2A2 B1/x .A2 B2 /x 2 cos x


C B0 C B1 A0 .B1 2B2 C A1 /x .B2 C A2 /x 2 sin x

D x 2 sin x if
(i)

A2 C B2
A2 B2

D 0
; (ii)
D 1

(iii)

A1 C B1 C 2A2
A1 B1 C 2B2

A0 C B0 C A1
A0 B0 C B1

D 0
;
D 0

D 0
:
D 0

1
1
, B2 D
. Substituting these into (ii) and solving for A1 and B1 yields A1 D 1,
2
2
1
1
B1 D 0. Substituting these into (iii) and solving for A0 and B0 yields A0 D
, B0 D . Therefore,
2
2

From (i), A2 D

up D
yD

(e) Let y D
up
u0p

e
2
e


.1 C x/2 cos x


.1 C x/2 cos x

.1
.1


x 2 / sin x and


x 2 / sin x C c:

2
x 3 e x sin x dx D ue x ; then y 0 D .u0 C u/e x D x 3 e x sin x if u0 C u D x 3 sin x. Now let

D .A0 C A1 x C A2 x 2 C A3 x 3 / cos x C .B0 C B1 x C B2 x 2 C B3x 3 / sin xI then




D A1 C B0 C .2A2 C B1 /x C .3A3 C B2 /x 2 C B3 x 3 cos x


C B1 A0 C .2B2 A1 /x C .3B3 A2 /x 2 A3 x 3 sin x; so
u00p C up

D A0 C A1 C B0 C .A1 C 2A2 C B1 /x


C.A2 C 3A3 C B2 /x 2 C .A3 C B3 /x 3 cos x
C B0 C B1 A0 C .B1 C 2B2 A1 /x

C.B2 C 3B3 A2 /x 2 C .B3 A3 /x 3 sin x D x 3 sin x if

74 Chapter 5 Linear Second Order Equations


A3 C B3
A3 C B3

(i)

D 0
; (ii)
D 1

A1 C B1 C 2A2
A1 C B1 C 2B2

(iii)

A2 C B2 C 3A3
A2 C B2 C 3B3

D 0
; (iv)
D 0

D 0
;
D 0

A0 C B0 C A1
A0 C B0 C B1

D 0
:
D 0

1
1
3
, B3 D . Substituting these into (ii) and solving for A2 and B2 yields A2 D ,
2
2
2
3
3
, B1 D
. Substituting
B2 D 0. Substituting these into (iii) and solving for A1 and B1 yields A1 D
2
2
3
these into (iv) and solving for A0 and B0 yields A0 D 0, B0 D . Therefore,
2
From (i), A3 D

1
x.3
2

up D

3x C x 2 / cos x

.3


3x C x 3 / sin x and


ex 
x.3 3x C x 2 / cos x .3 3x C x 3 / sin x C c:
2
R
(f) Let y D e x x cos x .1 C 3x/ sin x dx D ue x ; then y 0 D .u0 Cu/e x D e x x cos x
if u0 C u D x cos x .1 C 3x/ sin x. Now let
yD

D
D

up
u0p

.A0 C A1 x/ cos x C .B0 C B1 x/ sin xI then


A1 C B0 C B1 x cos x C B1 A0 A1 x sin x; so

u00p C up

(i)

.1 C 3x/ sin x

D A0 C A1 C B0 C .A1 C B1 /x cos x

C B0 C B1 A0 C .B1 A1 /x sin x
D x cos x .1 C 3x/ sin x if

A1 C B1
A1 C B1

D
D

1
; (ii)
3

A0 C B0 C A1
A0 C B0 C B1

D
D

0
:
1

From (i), A1 D 2, B1 D 1. Substituting these into (ii) and solving for A0 and B0 yields A0 D 1, B0 D
1. Therefore, up D .1 2x/ cos x C .1 C x/ sin x and y D e x .1 2x/ cos x C .1 C x/ sin xC
c.
R


(g) Let y D e x .1 C x 2/ cos x C .1 x 2 / sin x dx D ue x ; then


y 0 D .u0 u/e x D e x .1 C x 2 / cos x C .1 x 2 / sin x

if u0

u D .1 C x 2 / cos x C .1
up
u0p

u00p

up

x 2 / sin x. Now let

D .A0 C A1 x C A2 x 2 / cos x C .B0 C B1 x C B2x 2 / sin xI then




D A1 C B0 C .2A2 C B1 /x C B2 x 2 cos x


C B1 A0 C .2B2 A1 /x A2 x 2 sin x; so
D


A0 C A1 C B0 .A1 2A2 B1/x .A2 B2 /x 2 cos x


C B0 C B1 A0 .B1 2B2 C A1 /x .B2 C A2 /x 2 sin x

D .1 C x 2 / cos x C .1
(i)

A2 C B2
A2 B2

D
D

x 2 / sin x if

1
; (ii)
1

A1 C B1 C 2A2
A1 B1 C 2B2

D 0
;
D 0

Section 5.6 Reduction of Order


(iii)

A0 C B0 C A1
A0 B0 C B1

75

D 1
:
D 1

From (i), A2 D 0, B2 D 1. Substituting these into (ii) and solving for A1 and B1 yields A1 D 1,
B1 D 1. Substituting these into (iii) and solving for A0 and B0 yields A0 D 0, B0 D 0. Therefore,up D
x cos x C x.1 C x/ sin x and y D e x x cos x C x.1 C x/ sin x C c.
5.6 REDUCTION OF ORDER
(N OTE: The term uy100 is indicated by    " in some of the following solutions, where y100 is complicated.
Since this term always drops out of the differential equation for u, it is not necessary to include it.)
5.6.2. If y D ux, then y 0 D u0 x C u and y 00 D u00Zx C 2u0 , so x 2 y 00 C xy 0 y D x 3 u00 C 3x 2u0 D
3
4
3
1
4
if u0 D , where (A) 0 C D 5 . Since
dx D 3 ln jxj, 1 D 3 is a solution of the
x2
x
x
x
x
v
complementary equation for (A). Therefore,the solutions of (A) are of the form (B) D 3 , where
x
v0
4
4
4
4
C1
4
C1
0
0
D 5 , so v D 2 . Hence, v D
C C1 ; u D D
C 3 (see (B)); u D
C C2 ;
x3
x
x
x
x4
x
3x 3 2x 2
C1
4
c2
4
y D ux D
C C2 x, or y D
C c1 x C . As a byproduct, fx; 1=xg is a fundamental set
3x 2 2x
3x 2
x
of solutions of the complementary equation.
5.6.4. If y D ue 2x , then y 0 D .u0 C 2u/e 2x and y 00 D .u00 C 4u0 C 4u/e 2x , so y 00 3y 0 C 2y D
1
e 2x
0
0
.u00 C u0 /e 2x D
if
u
D
,
where
(A)

D
. Since 1 D e x is a solution of
1Ce x
1Ce x
the complementary equation for (A), the solutions of (A) are of the form (B) D ve x , where v 0 e x D
e 2x
e x
, so v 0 D
. Hence, v D ln.1 C e x / C C1 ; u0 D D e x ln.1 C e x / C C1 e x
x
1Ce
1Ce x
(see (B)); u D .1 C e x / ln.1 C e x / 1 e x C1 e x C C2 ; y D ue 2x D .e 2x C e x / ln.1 C e x /
.C1 C 1/e x C .C2 1/e 2x , or y D .e 2x C e x / ln.1 C e x / C c1 e 2x C c2e x . As a byproduct, fe 2x ; e x g
is a fundamental set of solutions of the complementary equation.
!
!
x 1=2 x
x 1=2 x
1=2
1=2
1=2 x
0
0 1=2 x
00
00 1=2 x
0
5.6.6. If y D ux e , then y D u x e Cu x
C
e and y D u x e C2u x
C
e C
2
2
   so 4x 2 y 00 C .4x

8x 2 /y 0 C .4x 2 4x 1/yZD e x .4x 5=2 u00 C 8x 3=2 u0 / D 4x 1=2e x .1 C 4x/ if


2
1 C 4x
2
2
u0 D , where (A) 0 C D
. Since
dx D 2 ln jxj, 1 D 2 is a solution of the
x
x2
x
x
v
complementary equation for (A). Therefore,the solutions of (A) are of the form (B) D 2 , where
x
v0
1 C 4x
1
C1
0
2
0
D
, so v D 1 C 4x. Hence, v D x C 2x C C1 ; u D D
C 2 C 2 (see
x2
x2
x
x
C1
1=2 x
x
3=2
1=2
1=2
(B)); u D ln x C 2x
C C2 ; y D ux e D e .2x
C x ln x C1 x
C C2 x 1=2/, or
x
y D e x .2x 3=2 C x 1=2 ln x C c1x 1=2 C c2x 1=2 /. As a byproduct, fx 1=2 e x ; x 1=2e x g is a fundamental
set of solutions of the complementary equation.
2

5.6.8. If y D ue x , then y 0 D u0 e x 2xue x and y 00 D u00 e x 4xu0 e x C    , so y 00 C 4xy 0 C


2
2
.4x 2 C 2/y D u00 e x D 8e x.xC2/ D 8e x e 2x if u00 D 8e 2x . Therefore,u0 D 4e 2x C C1 ;
2
2
2
u D 2e 2x C C1 x C C2 , and y D ue x D e x .2e 2x C C1 x C C2 /, or y D e x .2e 2x C c1 C c2x/.
2
2
As a byproduct, fe x ; xe x g is a fundamental set of solutions of the complementary equation.

76 Chapter 5 Linear Second Order Equations


5.6.10. If y D uxe

, then y 0 D u0 xe

1/ and y 00 D u00 xe

2u0 e

1/ C    ,
3x
e
so x 2 y 00 C 2x.x 1/y 0 C .x 2 2x C 2/y D x 3u00 D x 3 e 2x if u00 D e 3x . Therefore,u0 D
C C1 ;
3
3x
2x
2x
e
xe
xe
C C1 x C C2 , and y D uxe x D
C xe x .C1 x C C2 /, or y D
C xe x .c1 C c2 x/.
uD
9
9
9
As a byproduct, fxe x ; x 2 e x g is a fundamental set of solutions of the complementary equation.
ue

.x

.x

5.6.12. If y D ue x , then y 0 D .u0 C u/e x and y 00 D .u00 C 2u0 C u/e x , so .1

2x/y 00 C 2y 0 C
4 4x
.2x 3/y D e x .1 2x/u00 C .4 4x/u0 D .1 4x C 4x 2/e x if u0 D , where (A) 0 C
D
1 2x

Z
Z 
4 4x
2
1 2x. Since
dx D
2C
dx D 2x ln j1 2xj, 1 D .1 2x/e 2x is
1 2x
1 2x
a solution of the complementary equation for (A). Therefore,the solutions of (A) are of the form (B)
e 2x
D v.1 2x/e 2x , where v 0 .1 2x/e 2x D .1 2x/, so v 0 D e 2x . Hence, v D
C C1 ;
2


1
.2x 1/2
u0 D D
C C1 e 2x .1 2x/ (see (B)); u D
C C1 xe 2x C C2 ; y D ue x D
2
8
.2x 1/2 e x
.2x 1/2 e x
C C1 xe x C C2 e x , or y D
C c1 e x C c2 xe x . As a byproduct, fe x ; xe x g
8
8
is a fundamental set of solutions of the complementary equation.
, so 2xy 00 C .4x C
1
3
1/y 0 C .2x C 1/y D e x .2xu00 C u0 / D 3x 1=2 e x if u0 D , where (A) 0 C
D x 1=2 . Since
2x
2
Z
1
1
dx D ln jxj, 1 D x 1=2 is a solution of the complementary equation for (A). Therefore,the
2x
2
3
3
solutions of (A) are of the form (B) D vx 1=2, where v 0 x 1=2 D x 1=2, so v 0 D . Hence,
2
2
3x
3 1=2
1=2
0
3=2
1=2
v D
C C1 ; u D D x
C C1 x
(see (B)); u D x
C 2C1 x
C C2 ; y D ue x D
2
2
e x .x 3=2 C 2C1 x 1=2 C C2 /, or y D e x .x 3=2 C c1 C c2x 1=2 / As a byproduct, is a fe x ; x 1=2e x g
fundamental set of solutions of the complementary equation.
5.6.14. If y D ue

, then y 0 D .u0

u/e

and y 00 D .u00

2u0 C u/e

u
u0
00
00 1=2
and
y
D
u
x
C
C    so 4x 2 y 00
2x 1=2
x 1=2
4x.x C 1/y 0 C .2x C 3/y D 4x 5=2.u00 u0 / D 4x 5=2e 2x if u0 D , where (A) 0 D e 2x . Since
1 D e x is a solution of the complementary equation for (A), the solutions of (A) are of the form (B)
x
0
2x
D ve x , where v 0 e x D e 2x , so v 0 D e x . Hence,
C1 e x (see (B));
 2x v D e C C1; u D D e  C

2x
2x
e
e
e
uD
C C1 e x C C2 ; y D ux 1=2 D x 1=2
C C1 e x C C2 , or y D x 1=2
C c1 C c2 e x .
2
2
2
As a byproduct, fx 1=2; x 1=2e x g is a fundamental set of solutions of the complementary equation.
5.6.16. If y D ux 1=2, then y 0 D u0 x 1=2 C

5.6.18. If y D ue x , then y 0 D .u0 C u/e x and y 00 D .u00 C 2u0 C u/e x , so xy 00 C .2 2x/y 0 C


u00
2
C1
C1
.x 2/y D e x .xu00 C 2u0 / D 0 if 0 D
; ln ju0 j D 2 ln jxj C k; u0 D 2 ; u D
C C2 .
u
x
x
x


C1
Therefore,y D ue x D e x
C C2 is the general solution, and fe x ; e x =xg is a fundamental set of
x
solutions.
u
2u0
and y 00 D u00 ln jxj C
   , so x 2 .ln jxj/2 y 00
x
x
.2x ln jxj/y 0 C .2 C ln jxj/y D x 2 .ln jxj/3 u00 D 0 if u00 D 0; u0 D C1 ; u D C1 x C C2 . Therefore,y D
u ln jxj D .C1 x C C2 / ln jxj is the general solution, and fln jxj; x ln jxjg is a fundamental set of solutions.
5.6.20. If y D u ln jxj, then y 0 D u0 ln jxj C

Section 5.6 Reduction of Order

77

5.6.22. If y D ue x , then y 0 D u0 e x Cue x and y 00 D u00 e x C2u0 e x Cue x , so xy 00 .2xC2/y 0 C.xC2/y D


u00
2
C1 x 3
e x .xu00 2u0 / D 0 if 0 D ; ln ju0 j D 2 ln jxj C k; u0 D C1 x 2 ; u D
C C2 . Therefore,
u
x
3

3
C1 x
C C2 e x is the general solution, and fe x ; x 3 e x g is a fundamental set of solutions.
y D ue x D
3
5.6.24. If y D ux sin x, then y 0 D u0 x sin x C u.x cos x C sin x/ and y 00 D u00 x sin x C 2u0 .x cos x C
u00
2 cos x
sin x/C   , so x 2 y 00 2xy 0 C.x 2 C2/y D .x 3 sin x/u00 C2.x 3 cos x/u0 D 0 if 0 D
; ln ju0 j D
u
sin x
C1
2 ln j sin xjCk; u0 D
; u D C1 cot x CC2. Therefore, y D ux sin x D x. C1 cos x CC2 sin x/
sin2 x
is the general solution, and fx sin x; x cos xg is a fundamental set of solutions.
u
u0
00
00 1=2
and
y
D
u
x
C
C    so 4x 2 .sin x/y 00
2x 1=2
x 1=2
u00
cos x
4x.x cos x C sin x/y 0 C .2x cos x C 3 sin x/y D 4x 5=2.u00 sin x u0 cos x/ D 0 if 0 D
; ln ju0 j D
u
sin x
ln j sin xj C k; u0 D C1 sin x; u D C1 cos x C C2 . Therefore,y D ux 1=2 D . C1 cos x C C2 /x 1=2 is
the general solution, and fx 1=2; x 1=2 cos xg is a fundamental set of solutions.
5.6.26. If y D ux 1=2, then y 0 D u0 x 1=2 C

u
u00 2u0
and y 00 D
C    , so .2x C 1/xy 00 2.2x 2 1/y 0 4.x C
2
x
x
x2
u00
4x C 4
2
1/y D .2x C 1/u00 .4x C 4/u0 D 0 if 0 D
D 2C
; ln ju0 j D 2x C ln j2x C 1j C k;
u
2x C 1
2x C 1
u
C2
u0 D C1 .2x C 1/e 2x ; u D C1 xe 2x C C2 . Therefore,y D D C1 e 2x C
is the general solution, and
x
x
2x
f1=x; e g is a fundamental set of solutions.
5.6.28. If y D

u0
u
, then y 0 D
x
x

5.6.30. If y D ue 2x , then y 0 D .u0 C 2u/e 2x and y 00 D .u00 C 4u0 C 4u/e 2x , so xy 00 .4x C 1/y 0 C
1
C1 x 2
u00
C C2 .
.4x C 2/y D e 2x .xu00 u0 / D 0 if 0 D ; ln ju0 j D ln jxj C k; u0 D C1 x; u D
u  x
2

2
C1 x
Therefore, y D ue 2x D e 2x
C C2 is the general solution, and fe 2x ; x 2 e 2x g is a fundamental
2
set of solutions.
5.6.32. If y D ue 2x , then y 0 D .u0 C 2u/e 2x and y 00 D .u00 C 4u0 C 4u/e 2x , so .3x 1/y 00
u00
9x 6
3
.3x C 2/y 0 .6x 8/y D e 2x .3x 1/u00 C .9x 6/u0 D 0 if 0 D
D 3C
.
u
3x 1
3x 1
0
0
3x
3x
Therefore,ln ju j D 3x C ln j3x 1j C k, so u D C1 .3x 1/e
, u D C1 xe
C C2 . Therefore,the
general solution is y D ue 2x D C1 xe x C C2 e 2x , or (A) y D c1 e 2x C c2 xe x . Now y.0/ D 2 )
c1 D 2. Differentiating (A) yields y 0 D 2c1e 2x C c2 .e x xe x /. Now y 0 .0/ D 3 ) 3 D 2c1 C c2, so
c2 D 1 and y D 2e 2x xe x .
5.6.34. If y D ux, then y 0 D u0 x Cu andZy 00 D u00 x C2u0 , so x 2 y 00 C2xy 0 2y D x 3 u00 C4x 2u0 D x 2 if
4
1
4
1
u0 D , where (A) 0 C D . Since
dx D 4 ln jxj, 1 D 4 is a solution of the complementary
x
x
x
x
v
v0
1
equation for (A). Therefore,the solutions of (A) are of the form (B) D 4 , where 4 D , so v 0 D x 3 .
x
x
x
x4
1 C1
x C1
Hence, v D
CC1; u0 D D C 4 (see (B)); u D
CC2 . Therefore,the general solution is
4
4 x
4 3x 3
2
2
x
C1
x
c2
x
c2
y D ux D
C C2 x, or (C) y D
C c1 x C 2 . Differentiating (C) yields y 0 D C c1 2 3 :
4 3x 2
4
x
2
x

78 Chapter 5 Linear Second Order Equations


Now y.1/ D

5 0
3
; y .1/ D ) c1 C c2 D 1; c1
4
2

2c2 D 1, so c1 D 1, c2 D 0 and y D

x2
C x.
4

5.6.36. If y D uy1 , then y 0 D u0 y1 C uy10 and y 00 D u00 y1 C 2u0 y10 C uy100 , so y 00 C p1 .x/y 0 C p2 .x/y D
u00
y0
y1 u00 C .2y10 C p1 y1 /u0 D 0 if u is any function such that (B) 0 D 2 1 p1 . If ln ju0 .x/j D
u
y1
 Z x

Z x
1
0
2 ln jy1 .x/j
p1 .t/ dt, then u satisfies (B); therefore, if (C) u .x/ D 2
exp
p1 .s/ ds ,
y1 .x/
x0
x0
 Z t

Z x
1
exp
p1 .s/ds satisfies (C), y2 D uy1 is a solution
then u satisfies (B). Since u.x/ D
2
x0 y1 .t/
x0
y2
D u is nonconstant, Theorem 5.1.6 implies that fy1 ; y2 g is a fundamental set
of (A) on .a; b/. Since
y1
of solutions of (A) on .a; b/.
5.6.38. (a) The associated linear equation is (A) 00 C k 2 D 0, with characteristic polynomial p.r / D
r 2 C k 2 . The general solution of (A) is D c1 cos kx C c2 sin kx. Since 0 D kc1 sin kx C kc2 cos kx,
kc1 sin kx C kc2 cos kx
0
yD
D
.

c1 cos kx C c2 sin kx
(b) The associated linear equation is (A) 00 30 C 2 D 0, with characteristic polynomial p.r / D
2
r
3r C 2 D .r 1/.r 2/. The general solution of (A) is D c1 e x C c2 e 2x . Since 0 D c1 e x C c2e 2x ,
0
c1 C 2c2 e x
yD
D
.

c1 C c2 e x
(c) The associated linear equation is (A) 00 C50 6 D 0, with characteristic polynomial p.r / D r 2 C
5r 6 D .r C 6/.r 1/. The general solution of (A) is D c1 e 6x C c2e x . Since 0 D 6c1 e 6x C c2e x ,
0
6c1 C c2 e 7x
yD
D
.

c1 C c2 e 7x
(d) The associated linear equation is (A) 00 C 80 C 7 D 0, with characteristic polynomial p.r / D
2
r C 8r C 7 D .r C 7/.r C 1/. The general solution of (A) is D c1 e 7x C c2e x . Since 0 D
7c1 C c2 e 6x
0
7c1 e 7x 2c2e x , y D
D
.

c1 C c2 e 6x
(e) The associated linear equation is (A) 00 C 140 C 50 D 0, with characteristic polynomial p.r / D
r 2 C 14r C 50 D .r C 7/2 C 1. The general solution of (A) is D e 7x .c1 cos x C c2 sin x/. Since
0 D 7e 7x .c1 cos x C c2 sin x/ C e 7x . c1 sin x C c2 cos x/ D .7c1 c2 / cos x .c1 C 7c2 / sin x,
0
.7c1 c2 / cos x C .c1 C 7c2/ sin x
yD
D
.

c1 cos x C c2 sin x
1
1
(f) The given equation is equivalent to (A) y 0 C y 2
y
D 0. The associated linear equation is
6
6



1
1
1
1
1
1
(B) 00
0
D 0, with characteristic polynomial p.r / D r 2
r
D rC
r
.
6
6
6
6
3
2
c1 x=3
c2
0
The general solution of (B) is D c1 e x=3 C c2 e x=2. Since 0 D
e
C e x=2, y D
D
3
2

2c1 C 3c2e 5x=6


.
6.c1 C c2 e 5x=6 /
1
1
(g) The given equation is equivalent to (A) y 0 C y 2
yC
D 0. The associated linear equation is
3
36


1
1
1 2
00 1 0
2 1
(B)
C D 0, with characteristic polynomial p.r / D r
rC
D r
. The general
3
36
3
36
6
e x=6
e x=6
solution of (B) is D e x=6 .c1 C c2x/. Since 0 D
.c1 C c2 x/ C c2 e x=6 D
.c1 C c2.x C 6//,
6
6

Section 5.7 Variation of Parameters


yD

79

0
c1 C c2 .x C 6/
D
.

6.c1 C c2 x/



0
00
r 0 .x/
5.6.40. (a) Suppose that is a solution of (B) and let y D
. Then (D) C p.x/
yCq.x/ D
r
r
r .x/


2
1 0
r 0 0
00
r0
00
r0
00
D
ry 2
y, so
D y 0 C ry 2 C y. Therefore, (D)
0 and y 0 D
2
r
r
r
r
r
r
r
implies that y satisfies (A). Now suppose that y is a solution of (A) and let be any function such that
r0
r0
0 D ry. Then 00 D r 0 y C ry 0 C ry0 D 0 C .y 0 C ry 2 /r D 0 .p.x/y C q.x//r , so
r
r
r0 0
00

C p.x/ry C q.x/r D 0, which implies that satisfies (B), since ry D 0 .


r
(b) If f1 ; 2 g is a fundamental set of solutions of (B) on .a; b/, then D c1 1 C c2 2 is the general
solution of (B) on .a; b/. This and (a) imply that (C) is the general solution of (A) on .a; b/.
5.7 VARIATION OF PARAMETERS
5.7.2. (A) yp D u1 cos 2x C u2 sin 2x;
u01 cos 2x C u02 sin 2x
2u01 sin 2x C 2u02 cos 2x

D 0
(B)
D sin 2x sec2 x:

(C)

Multiplying (B) by 2 sin 2x and (C) by cos 2x and adding the resulting equations yields 2u02 D tan 2x,
tan2 2x
1 sec2 2x
tan 2x
so u02 D
. Then (B) implies that u01 D u02 tan.2x/ D
D
. Therefore,u1 D
2
2
2
tan 2x
ln j cos 2xj
sin 2x ln j cos 2xj
x cos 2x
sin 2x
x
and u2 D
. Now (A) yields yp D
C
.
2
4
4
4
2
4
sin 2x ln j cos 2xj x cos 2x
C
.
Since sin 2x satisfies the complementary equation we redefine yp D
4
2
5.7.4. (A) yp D u1 e x cos x C u2 e x sin x;
u01 .e x

cos x

u01 e x cos x C u02 e x sin x


e x sin x/ C u02 .e x sin x C e x cos x/

D 0
(B)
x
D 3e sec x:

(C)

Subtracting (B) from (C) and cancelling e x from the resulting equations yields
u01 cos x C u02 sin x
u01 sin x C u02 cos x

D 0
(D)
D 3 sec x:

(E)

Multiplying (D) by sin x and (E) by cos x and adding the results yields u02 D 3. From (D), u01 D
u02 tan x D 3 tan x. Therefore u1 D 3 ln j cos xj, u2 D 3x. Now (A) yields yp D 3e x .cos x ln j cos xjC
x sin x/.
5.7.6. (A) yp D u1 e x C u2 e

;
u01 e x C u02 e

D 0

u01 e x

Adding (B) to (C) yields 2u01 e x D


1

2
e

2x

u02 e

(B)

4e x
:
1 C e 2x

(C)

4e x
2e 2x
0
,
so
u
D
. From (B), u02 D
1
1 C e 2x
1 e 2x

2e 2x
. Using the substitution v D e
1 e 2x

2x

e 2x u01 D

we integrate u01 to obtain u1 D ln.1

2x

/.

80 Chapter 5 Linear Second Order Equations


Using the substitution v D e 2x we integrate u02 to obtain u1 D ln.1
e x ln.1 e 2x / e x ln.e 2x 1/.
5.7.8. (A) yp D u1 e x C u2

ex
;
x
ex
u01 e x C u02
D
x
 x

e
ex
D
u01 e x C u02
x
x2

u02 e x
e 2x
D
, so u02 D
x2
x
e 2x
.
u1 D e x , u2 D xe x C e x . Now (A) yields yp D
x

Subtracting (B) from (C) yields

5.7.10. (A) yp D u1 e

x2

e 2x /. Now (A) yields yp D

C u2 xe

2xu01 e

x2

x2

(B)

e 2x
:
x

(C)

xe x . From (B), u01 D

u02
D e x . Therefore
x

u01 e

x2

C u02 .e

x2

C u02 xe
2x 2 e

x2
x2

D 0
D 4e

(B)
x.xC2/

(C)

Multiplying (B) by 2x and adding the result to (C) yields u02 e x D 4e x.xC2/ , so u02 D 4e 2x . From
(B), u01 D u02 x D 4xe 2x . Therefore u1 D .2x C 1/e 2x , u2 D 2e 2x . Now (A) yields yp D
e x.xC2/ .
5.7.12. (A) yp D u1 x C u2 x 3;
u01 x C u02 x 3

u01 C 3u02 x 2

(B)
4

2x sin x
D 2x 2 sin x
x2

(C):

1
and subtracting the result from (C) yields 2x 2 u02 D 2x 2 sin x, so u02 D sin x. From
x
(B), u01 D u02 x 2 D x 2 sin x. Therefore u1 D .x 2 2/ cos x 2x sin x, u2 D cos x. Now (A) yields
yp D 2x 2 sin x 2x cos x.
p
p
5.7.14. (A) yp D u1 cos x C u2 sin x;
p
p
u01 cos x C u02 sin x D 0
(B)
p
p
p
x
cos
x
sin
x
sin
u01 p C u02 p
D
(C):
4x
2 x
2 x
p
p
sin x
u0
Multiplying (B) by p
and (C) by cos x and adding the resulting equations yields p2 D
2 x
2 x
p
p
p
p
2p
p
sin x cos x
sin
x
cos
x
sin
x
, so u02 D
p
. From (B), u01 D u02 tan x D
p . Therefore,
4x p
2 x
2p x p
p
p
p
2p
sin x cos x
x
sin
x
sin x
x cos x
u1 D
, u2 D
. Now (A) yields yp D
. Since
2
2
2
2
p
p2
p
x cos x
sin x satisfies the complementary equation we redefine yp D
.
2

Multiplying (B) by

Section 5.7 Variation of Parameters

81

5.7.16. (A) yp D u1 x a C u2 x a ln x;
u01 x a C u02 x a ln x
au01 x a

C u02 .ax a

ln x C x a

D 0

x
D xa
x2

(B)
aC1

(C):

a
and subtracting the result from (C) yields u02 x a 1 D x a 1, so u02 D 1. From (B),
x
ln x. Therefore, u1 D x ln x, u2 D x. Now (A) yields yp D x aC1.

Multiplying (B) by
u01 D u02 ln x D
2

5.7.18. yp D u1 e x C u2 e

x2

;
2

u01 e x C u02 e
2u01 xe

x2

2u02 xe

x2

x2

0
(B)
8x 5
D 8x 4 :
x

(B)

Multiplying (B) by 2x and adding the result to (C) yields 4u01xe x D 8x 4 , so u01 D 2x 3 e x . From (B),
2
2
2
2
u02 D u01 e 2x D 2x 3e x . Therefore u1 D e x .x 2 C 1/, u2 D e x .x 2 1/. Now (A) yields
2
yp D 2x .
p
p
5.7.20. (A) yp D u1 xe 2x C u2 xe 2x ;
p
p
(B)
u01 xe 2x C u02 xe 2x D 0




5=2
p
p
p
1
1
8x
u01 e 2x 2 x C p
u02 e 2x 2 x
p
D
(C):
D2 x
4x 2
2 x
2 x
1
Multiplying (B) by
, subtracting the result from (C), and cancelling common factors from the resulting
2x
equations yields
u01 e 2x C u02 e
u01 e 2x u02 e
Adding (D) to (E) yields 2u01 e 2x D 1, so u01 D
u1 D

2x

, u2 D

2x

e 2x
. Now (A) yields yp D
4

5.7.22. (A)yp D u1 xe x C u2 xe

u01 .x C 1/e x

u02 .x

1/e

2x

. From (D), u02 D

2p
x
.
2

u01 e 4x D

e 2x
. Therefore,
2

0
(B)
4
3x
D 3x 2
x2

(C):

1
, subtracting the resulting equation from (C), and cancelling common factors yields
x
u01 e x C u02 e
u01 e x u02 e

Adding (D) to (E) yields 2u01e x D 3x, so u01 D


u1 D

(D)
(E)

;
u01 xe x C u02 xe

Multiplying (B) by

D 0
D 1:

2x

3e x .x C 1/
, u2 D
2

3e x .x
2

1/

x
x

3xe
2

D 0
D 3x:

(D)
(E)

. From (D), u02 D

. Now (A) yields yp D 3x 2 .

u01 e 2x D

3xe x
. Therefore
2

82 Chapter 5 Linear Second Order Equations


5.7.24. (A) yp D

u1
C u2 x 3 ;
x
u01
C u02 x 3
x
u01
C 3u02 x 2
x2

D 0
D

(B)

x 3=2
Dx
x2

1=2

(C)

1
x 5=2
and adding the result to (C) yields 4u02x 2 D x 1=2 , so u02 D
. From (B),
x
4
x 3=2
x 5=2
x 3=2
4x 3=2
. Therefore u1 D
, u2 D
. Now (A) yields yp D
.
4
10
6
15

Multiplying (B) by
u01 D u02 x 4 D

5.7.26. (A) yp D u1 x 2 e x C u2 x 3 e x ;
u01 x 2 e x C u02 x 3 e x

u01 .x 2 e x C 2xe x / C u02 .x 3 e x C 3x 2e x / D

0
(B)
2xe x
2e x
D
:
x2
x

(C)

Subtracting (B) from (C) and cancelling common factors in the resulting equations yields
u01 C u02 x
2u01 x C 3u02 x 2

D 0
2
D
:
x

(D)
(E)

2
2
Multiplying (D) by 2x and subtracting the result from (E) yields x 2u02 D , so u02 D 3 . From (D),
x
x
2
1
2
u01 D u02 x D
. Therefore u1 D , u2 D
. Now (A) yields yp D xe x .
2
x
x
x2
5.7.28. (A) yp D u1 x C u2 e x ;
u01 x C u02 e x
u01 C u02 e x

D 0
(B)
2.x 1/2 e x
D 2.x
D
x 1

1/e x :

(C)

Subtracting (B) from (C) yields u01 .1 x/ D 2.x 1/e x , so u01 D 2e x . From (B), u02 D
Therefore, u1 D 2e x , u2 D x 2 . Now (A) yields yp D xe x .x 2/.
5.7.30. (A) yp D u1 e 2x C u2 xe

D 2x.

u01 e 2x C u02 xe
2u01 e 2x C u02 .e

u01 xe

xe

x
x

D 0
(B)
.3x 1/2 e 2x
D
D .3x
3x 1

1/e 2x :

(C)

Multiplying (B) by 2 and subtracting the result from (C) yields u02 .1 3x/e x D .3x 1/e 2x , so
x2
e 3x
u02 D e 3x . From (B), u01 D u02 xe 3x D x. Therefore u1 D
, u2 D
. Now (A) yields
2
3
2x
2x
xe .3x 2/
xe .3x 2/
yp D
. The general solution of the given equation is y D
Cc1 e 2x Cc2xe x .
3
3
e 2x .3x 2 C x 1/
Differentiating this yields y 0 D
C 2c1e 2x C c2.1 x/e x . Now y.0/ D 1; y 0 .0/ D
3
1
1
e 2x .3x 2 2x C 6/
xe x
2 ) c1 D 1; 2 D
C 2c1 C c2, so c2 D , and y D
C
.
3
3
6
3

Section 5.7 Variation of Parameters


5.7.32. (A) yp D u1 .x

u01 .x

1/e x C u2 .x

1/e x C u02 .x

83

1/;
1/

u01 xe x C u02

D 0
(B)
.x 1/3 e x
D
D .x
.x 1/2

1/e x :

(C)

From (B), u01 D u02 e x . Substituting this into (C) yields u02 .x 1/ D .x 1/e x , so u02 D e x , u01 D 1.
Therefore u1 D x, u2 D e x . Now (A) yields yp D e x .x 1/2 . The general solution of the given equation
is y D .x 1/2 e x C c1.x 1/e x C c2.x 1/. Differentiating this yields y 0 D .x 2 1/e x C c1 xe x C c2.
Now y.0/ D 4; y 0 .0/ D 6 ) 4 D 1 c1 c2 ; 6 D 1 C c2 , so c1 D 2; c2 D 5 and
y D .x 2 1/e x 5.x 1/.
u2
5.7.34. (A) yp D u1 x C 2 ;
x
u0
u01 x C 22 D 0
(B)
x
2x 2
2u02
u01
D
D 2:
(C)
x3
x2
2
2
Multiplying (B) by and adding the result to (C) yields 3u01 D 2, so u01 D
. From (B), u02 D
x
3
2x 3
2x
x4
x2
u01 x 3 D
. Therefore u1 D
, u2 D
. Now (A) yields yp D
. The general solution
3
3
6
2
2
c2
2c2
x
of the given equation is y D
C c1 x C 2 . Differentiating this yields y 0 D x C c1
. Now
2
x
x3
1
1
y.1/ D 1; y 0 .1/ D 1 ) 1 D
C c1 C c2 ; 1 D 1 C c1 2c2, so c1 D 1; c2 D , and
2
2
x2
1
yD
C x C 2.
2
2x
5.7.36. Since y D yp

a1 y1

a2 y2 ,

00

P0 .x/y C P1 .x/y 0 C P2 .x/y

D P0 .x/.yp

a1 y1

a2 y2 /00

CP1 .x/.yp a1 y1 a2 y2 /0
CP2 .x/.yp a1 y1 a2 y2 /
D .P0 .x/yp00 C P1 .x/yp0 C P2 .x/yp /


a1 P0 .x/y100 C P1 .x/y10 C P2 .x/y1


a2 P0 .x/y200 C P1 .x/y20 C P2 .x/y2
D F .x/ a1  0 a2  0 D F .x/I

hence y is a particular solution of (A).


is a solution of (A) on .a; 1/ if u01 e x C u02 e x D 0 and u01 e x
e xf 0
ex f
u02 e x D f .x/. Solving these two equations yields u01 D
, u2 D
. The functions u1 .x/ D
2
2
Z x
Z x
1
1
e t f .t/ dt and u2 .x/ D
e t f .t/ dt satisfy these conditions. Therefore,
2 0
2 0
Z
Z
ex x t
e x x t
yp .x/ D
e f .t/ dt
e f .t/ dt
2 0
2 0
Z
Z x


1 x
.x t/
D
f .t/ e .x t / e
dt D
f .t/ sinh.x t/ dt:
2 0
0

5.7.38.

(a) yp D u1 e x C u2 e

84 Chapter 5 Linear Second Order Equations


is a particular solution of y 00
yp0 .x/

D
D
D

y D f .x/. Differentiating yp yields


Z
Z
ex x t
ex
e x x t
e x x
e f .t/ dt C e x C
e f .t/ dt
e
2 0
2
2 0
2
Z
Z
e x x t
ex x t
e f .t/ dt C
e f .t/ dt
2 0
2 0
Z x
Z x


1
.x t /
.x t /
f .t/ e
Ce
dt D
f .t/ cosh.x t/ dt:
2 0
0

Since yp .x0 / D yp0 .x0 / D 0, the solution of the initial value problem is
y

D yp C k0 cosh x C k1 sinh x
Z x
D k0 cosh x C k1 sinh x C
sinh.x

t/f .t/ dt:

The derivative of the solution is


y0

D yp0 C k0 sinh x C k1 cosh x


Z x
D k0 sinh x C k1 cosh x C
cosh.x
0

t/f .t/ dt:

CHAPTER 6
Applications of Linear Second Order
Equations

6.1 SPRING PROBLEMS I


g
32
k
D
D
D 320 the equation of motion is (A) y 00 C 320y D 0. The general
m
l
:1 p
p
p
p
p
solution of (A) is y D c1 cos 8 5t C c2 sin 8 5t, so y 0 D 8 5. c1 sin 8 5t C c2 cos 8 5t/. Now
p
p
1
1
1
1
1
y.0/ D
) c1 D
and y 0 .0/ D 2 ) c2 D
p , so y D
cos 8 5t
p sin 8 5t ft.
4
4
4
4 5
4 5
6.1.2. Since

g
32
k
D
D
D 64 the equation of motion is (A) y 00 C 64y D 0. The general solution
m
l
:5
1
1
of (A) is y D c1 cos 8t C c2 sin 8t, so y 0 D 8. c1 sin 8t C c2 cos 8t/. Now y.0/ D ) c1 D and
4
4
p
1
1
1
17
1
0
y .0/ D
) c2 D
, so y D cos 8t
sin 8t ft; R D
ft; !0 D 8 rad/s; T D =4 s;
2
16
4
16
16

  :245 rad  14:04 .


6.1.4. Since

.9:8/10
mg
D
D 140, the equation of motion of the 2 kg mass is (A) y 00 C 70y D
l
:7
p
p
p
p
0. The general solution of (A) is y D c1 cos 70t C c2 sin 70t, so y 0 D 70. c1 sin 70t C
p
p
p
1
2
1
1
2
c2 cos 70t/. Now y.0/ D
) c1
D and y 0 .0/ D 2 ) c2 D p , so y D
cos 70 t C p sin 70 t
4
4
4
70
70
r
p
p
1 67
m; R D
m; !0 D 70 rad/s; T D 2= 70 s;   2:38 rad  136:28.
4 35
6.1.6. Since k D

k
g
32
D
D
D 64 the equation of motion is (A) y 00 C 64y D 0. The general solution
m
l
1=2
1
1
of (A) is y D c1 cos 8t C c2 sin 8t, so y 0 D 8. c1 sin 8t C c2 cos 8t/. Now y.0/ D ) c1 D and
2
2
3
1
3
0
y .0/ D 3 ) c2 D
, so y D cos 8t
sin 8t ft.
8
2
8
6.1.8. Since

64
D 2, so the equation of motion is 2y 00 C 8y D 2 sin t, or (A) y 00 C 4y D sin t. Let
32
yp D A cos t C B sin t; then yp00 D A cos t B sin t, so yp00 C 4yp D 3A cos t C 3B sin t D sin t if
1
1
3A D 0, 3B D 1. Therefore,A D 0, B D , and yp D sin t. The general solution of
3
3
6.1.10. m D

85

86 Chapter 6 Applications of Linear Second Order Equations


1
1
1
(A) is (B) y D sin t C c1 cos 2t C c2 sin 2t, so y.0/ D
) c1 D . Differentiating (B) yields
3
2
2
1
1
5
0
0
y D cos t 2c1 sin 2t C 2c2 cos 2t, so y .0/ D 2 ) 2 D
C 2c2, so c2 D . Therefore,
3
3
6
1
1
5
y D sin t C cos 2t C sin 2t ft.
3
2
6
4
1
mg
1
1
D and k D
D 4, so the equation of motion is y 00 C 4y D sin 8t, or (A)
6.1.12. m D
32
8
l
8
4
y 00 C 32y D 2 sin 8t. Let yp D A cos 8t C B sin 8t; then yp00 D 64A cos t 64B sin 8t, so yp00 C 32yp D
1
32A cos 8t 32B sin 8t D 2 sin 8t if 32A D 0, 32B D 2. Therefore,A D 0, B D
, and
16
p
p
1
1
yp D
sin 8t. The general solution of (A) is (B) y D
sin 8t C c1 cos 4 2t C c2 sin 4 2t, so
16
16
p
p
p
1
1
1
y.0/ D ) c1 D . Differentiating (B) yields y 0 D
cos 8t C4 2. c1 sin 4 2t Cc2 cos 4 2t/, so
3
3
2
p
p
1 p
1
1
1
1
0
y .0/ D 1 ) 1 D
C4 2c2 , so c2 D
p . Therefore,y D
sin 8t C cos 4 2t
p sin 4 2t ft.
2
16
3
8 2
8 2
r
2
m
6.1.14. Since T D
D 2
the period is proportional to the square root of the mass. Therefore,
!0
k
p
doubling
p the mass mutiplies the period by 2; hence the period of the system with the 20 gm mass is
T D 4 2 s.

3
mg
6
3 00
6
D
and k D
D
D 18 so the equation of motion is
y C 18y D
32
16
l
1=3
16
64
4 sin !t 6 cos !t, or (A) y 00 C 96y D
sin !t 32 cos !t. The displacement will be unbounded if
3
p
p
p
p
64
! D 96 D 4 6, in which case (A) becomes (B) y 00 C 96y D
sin 4 6t 32 cos 4 6t. Let
3
p
p
yp D At cos 4 6t C Bt sin 4 6tI then
p
p
p
p
yp0 D .A C 4 6Bt/ cos 4 6t C .B 4 6At/ sin 4 6t
p
p
p
p
yp00 D .8 6B 96At/ cos 4 6t .8 6A C 96Bt/ sin 4 6t; so
p
p
p
p
p
p
64
yp00 C 96yp D 8 6B cos 4 6t 8 6A sin 4 6t D
sin 4 6t 32 cos 4 6t
3


p
p
p
p
8
4
t
8
64
if 8 6B D 32, 8 6A D
. Therefore, A D
cos 4 6t C 4 sin 4 6t .
p , B D p , and yp D p
3
3 6
6
6 3
The general solution of (B) is


p
p
p
p
t
8
yD p
cos 4 6t C 4 sin 4 6t C c1 cos 4 6t C c2 sin 4 6t;
.C/
6 3
6.1.16. m D

so y.0/ D 0 ) c1 D 0. Differentiating (C) yields






p
p
p
p
8
4
8
y0 D
p cos 4 6t C p sin 4 6t
4t
sin 4 6t C 4 cos 4 6t
3
3 6
6
p
p
p
C4 6. c1 sin 4 6t C c2 cos 6t/;
p
8
1
so y 0 .0/ D 0 ) 0 D
p C 4 6c2, and c2 D . Therefore,
9
3 6

p  1
p
p
t
8
yD p
cos 4 6t C 4 sin 4 6t C sin 4 6t ft:
9
6 3

Section 6.2 Spring Problems II

87

6.1.18. The equation of motion is (A) y 00 C !02 y D 0. The general solution of (A) is y D c1 cos !0 t C
c2 sin !0 t. Now y.0/ D y0 ) c1 D y0 . Since y 0 D !0 . c1 sin !0 t C c2 cos !0 t/, y 0 .0/ D v0 ) c2 D
v0
v0
. Therefore, y D y0 cos !0 t C
sin !0 t;
!0
!0
1 p
v0
y0 !0
RD
I sin  D p
:
.!0 y0 /2 C .v0 /2 I cos  D p
!0
.!0 y0 /2 C .v0 /2
.!0 y0 /2 C .v0 /2

Discussion 6.1.1 In Exercises 19, 20, and 21 we use the r


fact that in a springmass system with mass m
m
and spring constant k the period of the motion is T D 2
. Therefore, if we have two systems with
k
s
T2
m2 k1
masses m1 and m2 and spring constants k1 and k2 , then the periods are related by
D
. We
T1
m1 k2
will use this formula in the solutions of these exercises.
s
p
p
T2
2m1
6.1.20. Let m2 D 2m1 . Since k1 D k2 ,
D
D 2, so T2 D 2T1 .
T1
m1
s
k1
6.1.21. Suppose that T2 D 3T1 . Since m1 D m2 ,
D 3, k1 D 9k2 .
k2
6.2 SPRING PROBLEMS II
16
1
mg
D
D 5 the equation of motion is y 00 Cy 0 C5y D 0, or (A) y 00 C2y 0 C10y D 0.
l
3:2
2
The characteristic polynomial of (A) is p.r / D r 2 C2r C10 D .r C1/2 C9. Therefore,the general solution
of (A) is y D e t .c1 cos 3t C c2 sin 3t/, so y 0 D y C 3e t . c1 sin 3t C c2 cos 
3t/. Now y.0/ D 
3
1
1
t
0
and y .0/ D 2 ) c1 D 3 and 2 D 3 C 3c2, or c2 D
. Therefore,y D e
3 cos 3t C sin 3t
3
3
p
82 t
ft. The timevarying amplitude is
e ft.
3
96
mg
6.2.4. Since k D
D
D 30 the equation of motion is 3y 00 C 18y 0 C 30y D 0, or (A) y 00 C 6y 0 C
l
3:2
10y D 0. The characteristic polynomial of (A) is p.r / D r 2 C 6r C 10 D .r C 3/2 C 1. Therefore,the
general solution of (A) is y D e 3t .c1 cos t C c2 sin t/, so y 0 D 3y C e 3t . c1 sin t C c2 cos t/.
5
5
15
63
Now y.0/ D
and y 0 .0/ D 12 ) c1 D
and 12 D
C c2 , or c2 D
. Therefore,
4
4
4
4
3t
e
yD
.5 cos t C 63 sin t/ ft.
4
mg
8
1
3
6.2.6. Since k D
D
D 25 the equation of motion is y 00 C y 0 C 25y D 0, or (A) y 00 C 6y 0 C
l
:32
4
2
2
2
100y D 0. The characteristic polynomial of
(A)
is
p.r
/
D
r
C
6r
C
100
D .r C
p
p
p 3/ C3t91. Therefore,the
p
3t
0
general solution of (A) is y D e .c1 cos 91t Cc2 sin 91t/, so y D 3yC 91e . c1 sin 91t C
p
p
1
1
3
11
c2 cos 91t/. Now y.0/ D
and y 0 .0/ D 4 ) c1 D
and 4 D
C 91c2, or c2 D p .
2
2
2
2 91


p
p
1 3t
11
Therefore,y D e
cos 91t C p sin 91t ft.
2
91
6.2.2. Since k D

mg
20  980
D
D 3920 the equation of motion is 20y 00 C 400y 0 C 3920y D 0,
l
5
or (A) y 00 C 20y 0 C 196y D 0. The characteristic polynomial of (A) is p.r / D r 2 C 20r C 196 D

6.2.8. Since k D

88 Chapter 6 Applications of Linear Second Order Equations


p
p
.r C 10/2 C 96.pTherefore,the general
solution of
(A) is y D e 10t .c1 cos 4 6t C c2 sin 4 6t/, so
p
p
y 0 D 10y C 4 6e 10t . c1 sin 4 6t C c2 cos 4 6t/. 
Now y.0/ D 9 and y 0 .0/ D 0 ) c1 D 9 and
p
p
p
45
45
0 D 90 C 4 6c2 , or c2 D p . Therefore, y D e 10t 9 cos 4 6t C p sin 4 6t cm.
2 6
2 6
mg
32
D
D 32 the equation of motion is (A) y 00 C 3y 0 C 32y D 0. The characterisl
1


3 2 119
tic polynomial of (A) is p.r / D r 2 C3r C32 D r C
C
. Therefore,the general solution of (A) is
2
4
!
!
p
p
p
p
p
3
119
119
119 3t =2
119
119
3t =2
0
t C c2 sin
t , so y D
yC
e
c1 sin
t C c2 cos
t .
yDe
c1 cos
2
2
2
2
2
2
p
1
119
1
3
9
Now y.0/ D and y 0 .0/ D 3 ) c1 D and 3 D
C
c2 , or c2 D
p
. Therefore,
2
2 !
4
2
2
119
p
p
3
1
119
9
119
y D e 2t
cos
t
p
sin
t ft.
2
2
2
2 119
6.2.10. Since k D

2
25
1 00
1
25
mg
D
D
the equation of motion is
y C y0 C
y D 0, or (A)
l
:32
4
16
8
4
2
y 00 C 2y 0 C 100y D 0. The characteristic polynomial of (A) is p.r
C 100 D .r C
p/ D r C 2r p
1/2 C 99. Therefore,the general solution of (A) is y D e t .c1 cos 3 11t C c2 sin 3 11t/, so y 0 D
p
p
p
1
1
y C 3 11e t . c1 sin 3 11t C c2 cos 3 11t/. Now y.0/ D
and y 0 .0/ D 5 ) c1 D
and
3
3


p
p
p
1
14
1
14
5 D C 3 11c2 , or c2 D p . Therefore,y D e t
cos 3 11t C p sin 3 11t ft.
3
3
9 11
9 11
6.2.12. Since k D

mg
D 32 the equation of motion is (A) y 00 C 12y 0 C 32y D 0. The characteristic
l
polynomial of (A) is p.r / D r 2 C 12r C 32 D .r C 8/.r C 4/. Therefore, the general solution of (A) is
2
2
and y 0 .0/ D 0 ) c1 C c2 D ;y D c1e 8t C c2 e 4t , so y 0 D 8c1e 8t 4c2e 4t . Now y.0/ D
3
3
2
4
2
8t
4t
8c1 4c2 D 0, so c1 D
, c2 D , and y D
.e
2e /.
3
3
3
6.2.14. Since k D

100  980
mg
D
D 100 the equation of motion is 100y 00 C 600y 0 C 1000y D 0, or
l
98
(A) y 00 C 6y 0 C 10y D 0. The characteristic polynomial of (A) is p.r / D r 2 C 6r C 10 D .r C 3/2 C 1.
Therefore, the general solution of (A) is y D e 3t .c1 cos t C c2 sin t/, so y 0 D 3y C e t . c1 sin t C
c2 cos t/. Now y.0/ D 10 and y 0 .0/ D 100 ) c1 D 10 and 100 D 30 C c2 , or c2 D 70.
Therefore, y D e 3t .10 cos t 70 sin t/ cm.
6.2.16. Since k D

6.2.18. The equation of motion is (A) 2y 00 C 4y 0 C 20y D 3 cos 4t 5 sin 4t. The steady state component
of the solution of (A) is of the form yp D A cos 4t CB sin 4t; therefore yp0 D 4A sin 4t C4B cos 4t and
yp00 D 16A cos 4t 16B sin 4t, so 2yp00 C 4yp0 C 20yp D . 12A C 16B/ cos 4t .16A C 12B/ sin 4t D
11
27
3 cos 4t 5 sin 4t if 12A C 16B D 3, 16A 12B D 5; therefore A D
,B D
, and
100
100
11
27
yp D
cos 4t C
sin 4t cm.
100
100
mg
9:8
6.2.20. Since k D
D
D 20 the equation of motion is (A) y 00 C 4y 0 C 20y D 8 sin 2t 6 cos 2t.
l
:49
The steady state component of the solution of (A) is of the form yp D A cos 2t C B sin 2t; therefore

Section 6.3 The RLC Circuit

89

yp0 D 2A sin 2t C 2B cos 2t and yp00 D 4A cos 2t 4B sin 2t, so yp00 C 4yp0 C 20yp D .16A C
8B/ cos 2t .8A 16B/ sin 2t D 8 sin 2t 6 cos 2t if 16A C 8B D 6, 8A C 16B D 8; therefore
1
1
1
1
, B D , and y D
cos 2t C sin 2t m.
AD
2
4
2
4
6.2.22. If e r1t .c1 C c2 t/ D 0, then (A) c1 C c2 t D 0. If c2 D 0, then c1 0 (by assumption), so (A) is
impossible. If c1 0, then the left side of (A) is strictly monotonic and therefore cannot have the same
value for two distinct values of t.
c
y C !1 e ct =2m. c1 sin !1 t C
6.2.24. If y D e ct =2m.c1 cos !1 t C c2 sin !1 t/, then y 0 D
2m
cy0
1 
cy0 
c2 cos !1 t/, so y.0/ D y0 and y 0 .0/ D v0 ) c1 D y0 and v0 D
Cc2 !1 , so c2 D
v0 C
t
2m
!1
2m




1
cy0
and y D e ct =2m y0 cos !1 t C
v0 C
sin !1 t .
!1
2m

6.2.26. If y D e r1t .c1 C c2t/, then y 0 D r1 y C c2 e r1t , so y.0/ D y0 and y 0 .0/ D v0 ) c1 D y0 and
v0 D r1 y0 C c2, so c2 D v0 r1 y0 . Therefore, y D e r1t .y0 C .v0 r1 y0 /t/.

6.3 THE RLC CIRCUIT


1 00
6.3.2.
Q C 2Q0 C 100Q D 0; Q00 C 40Q0 C 2000Q D 0; r 2 C 40r C 2000 D .r C 20/2 C
20
1600 D 0; r D 20 40i ; Q D e 20t .2 cos 40t C c2 sin 40t/ (since Q0 D 2); I D Q0 D
21
e 20t ..40c2 40/ cos 40t .20c2 C 80/ sin 40t/; I0 D 2 ) 40c2 40 D 2 ) c2 D
, so
20
20t
20c2 C 80 D 101; I D e
.2 cos 40t 101 sin 40t/.
1 00
Q C 6Q0 C 250Q D 0; Q00 C 60Q0 C 2500Q D 0; r 2 C 60r C 2500 D .r C 30/2 C
10
1600 D 0; r D 30 40i ; Q D e 30t .3 cos 40t C c2 sin 40t/ (since Q0 D 3); I D Q0 D
e 30t ..40c2 90/ cos 40t .30c2 C 120/ sin 40t/; I0 D 10 ) 40c2 90 D 10 ) c2 D 2, so
30c2 120 D 180; I D 10e 30t .cos 40t C 18 sin 40t/.
6.3.4.

6.3.6. Qp D A cos 10t C B sin 10t; Qp0 D 10B cos 10t 10A sin 10t; Qp00 D 100A cos 10t
1 00
Q C 3Qp0 C 100Q9 D .90A C 30B/ cos 10t .30A 90B/ sin 10t D 5 cos 10t
100B sin 10t;
10 p
5 sin 10t, so 90A C 30B D 5, 30A C 90B D 5. Therefore, A D 1=15, B D 1=30, Qp D
1
cos 10t sin 10t
, and Ip D
.cos 10t C 2 sin 10t/.
15
30
3
6.3.8. Qp D A cos 50t C B sin 50t; Qp0 D 50B cos 50t 50A sin 50t; Qp00 D 2500A cos 50t
1
2500B sin 50t; Qp00 C 2Qp0 C 100Qp . 150AC 100B/ cos 50t .100AC 150B/ sin 50t D 3 cos 50t
10
6 sin 50t, so 150A C 100B D 3, 100A C 1500B D 6. Therefore,A D 3=650, B D 12=325,
3
3
Qp D
.cos 50t C 8 sin 50t/, and Ip D
.8 cos 50t sin 50t/.
650
13
6.3.10. Qp D A cos 30t C B sin 30t; Qp0 D 30B cos 30t 30A sin 30t; Qp00 D 900A cos 30t
1
900B sin 30t; Qp00 C 4Qp0 C 125Qp D .80A C 120B/ cos 30t .120A 80B/ sin 30t D 15 cos 30t
20
3
30 sin 30t, so 80AC120B D 15, 120AC80B D 30, A D 3=13, B D 3=104, Qp D
.8 cos 30t
104
45
sin 30t/, and Ip D
.cos 30t C 8 sin 30t/.
52

90 Chapter 6 Applications of Linear Second Order Equations


6.3.12. Let  D .!/ be the amplitude of Ip . From the solution of Exercise 6.3.11, Qp D A cos !t C
.1=C L! 2 /U R!V
R!U C .1=C L! 2 /V
, B D
, and D .1=C
B cos !t, where A D

2 2
2 2
0
2
L! / C R ! . Since Ip D Qp D !. A sin !t C B cos !t/, it follows that  .!/ D ! 2 .A2 C B 2 / D
U2 CV2

, with .!/ D 2 D .1=C! L!/2 C R2 , which attains it mininmum value R2 when


.!/
!
p
U2 CV2
1
! D !0 D p
. The maximum amplitude of Ip is .!/ D
.
R
LC
6.4 MOTION UNDER A CENTRAL FORCE
6.4.2. Let h D r0200 ; then  D

h2

. Since r D
k
1 C e cos.

/

, it follows that (A) e cos.


0

/ D


r
1. Differentiating this with respect to t yields e sin. / 0 D
, so (B) e sin. / D
r 0
r2
r
, since r 2  0  h. Squaring and adding (A) and (B) and setting t D 0 in the result yields e D
"h
2  0 2 #1=2

r0
1 C
. If e D 0, then 0 is undefined, but also irrelevant; if e 0, then set t D 0
r0
h


r 0
1 
in (A) and (B) to see that  D 0 , where   <  , cos D
1 and sin D 0
e r0
eh
1
1
1
d 2u
6
f
.1=u/.
Let
u
D
D
;
then
D
D 6cu2 .
2
2
2
2
mh u
r
c
d 
c 4 
6c
1
1
6cu2 C u D
f .1=u/, so f .1=u/ D mh2 .6cu4 C u3 / and f .r / D mh2
C 3 .
mh2 u2
r4
r


mk
d 2u
k
6.4.6. (a) With f .r / D
,
Eqn.
6.4.11
becomes
(A)
C
1
u D 0. The initial conditions
r3
d 2
h2
1
du
r00
imply that u.0 / D
and
.0 / D
(see Eqn. (6.4.)
r
d
h
0 1=2

k
d 2u
(b) Let D 1
. (i) If h2 < k, then (A) becomes
2 u D 0, and the solution

2
h
d 2
r00
1
of the initial value problem for u is u D
cosh .
0 /
sinh .
0 /; therefore r D
r0
h

 1
r0 r00
d 2u
r0 cosh . 0 /
sinh . 0 /
. (ii) If h2 D k, then (A) becomes
D 0, and the solu h
d 2

 1
1
r00
r0r00
tion of the initial value problem for u is u D
. 0 /; therefore r D r0 1
. 0 /
.
r0
h
h
d 2u
(iii) If h2 > k, then (A) becomes
C 2 u D 0, and the solution of the initial value problem for u is
d 2

 1
1
r00
r0r00
uD
cos . 0 /
sin . 0 /; therefore r D r0 cos . 0 /
sin . 0 /
.
r0
h
h
6.4.4. Recall that (A)

d 2u 2
 D
d

CHAPTER 7
Series Solutions of Linear Second
Equations

7.1 REVIEW OF POWER SERIES


P
m
7.1.2. From Theorem 7.1.3, 1
jj > 1=L. Therefore,
mD0 bm converges
p if jj < 1=L and diverges if p
P1
2
x0 / converges if jx x0 j < 1= L and diverges if jx x0 j > 1= L.
mD0 bm .x
P
m
7.1.4. From Theorem 7.1.3, 1
if jj < 1=L and diverges if jjp> 1=L. Therefore,
mD0 bm converges
p
P1
k
k
km
b
.x
x
/
converges
if
jx
x
j
<
1=
L
and diverges if jx x0 j > 1= L.
m
0
0
mD0
2

00

2 0

7.1.12. .1 C 3x /y C 3x y
2

1
X

nD0

an x n D
2a0 C

2a2

1
X

2y D

1
X

.n C 2/.n C 1/anC2 x n C 3

nD0

1
X

2
3

nD0
1
X

nD0

an x C 4

nan x n C 3

nD0
1
X

nD0

n.n


nD1

1
X

nD0
1
X

nD0

an x n D

1
X

nD0

1
X

1
X

C3

1
X

2/an C 3.n

1/

n.n 1/an x n

nD2

.n C 2/.n C 1/anC2 x n C 2

nD0

C2

1
X

1/an x C 3

n.n

nD2

1/nan x n C 3

n.n

nD1

an x n C 4

nD1
1
X

1
X

.n C 2/.n C 1/anC2 C .3n.n

1
X

.n

1/an

1
X

nD2

n.n

nD0

1/an

an x n D

1/an x n C 2

1
X

nD0

1
X

1
X

nD2

1 x

1
X

1
X

nD0

nan x n C 3

n.n

nD2

.n C 2/.n C 1/anC2 C 2.n C 1/anC1 C .n2


91

1
X

an x n

nD0


2n C 3/an x n .

1
X

an x n D

1/an x n C 2

.n C 2/.n C 1/anC2 x n

.n C 1/anC1 x n

nD0

1/an x n

n.n

nan x nC1

nD0

1
X

an x
.n C 2/.n C 1/anC2 C 2.n C 1/anC1 C .2n2

x/y 0 C 3y D

n
1x

n.n 1/an x n C 2

nD0

1
X

nD1

nD1

7.1.14. .1 C x 2 /y 00 C .2
1
X

1/an x

nD2

7.1.13. .1 C 2x 2 /y 00 C .2 3x/y 0 C 4y D
1
X

n.n

n 2

1
X

nD0

1
X

nan x n

nD1

.n C 1/anC1 x n


5n C 4/an x n .

1/an x n C 2

1
X

nD1

nan x n

92 Chapter 7 Series Solutions of Linear Second Equations


1
X

7.1.16. Let t D xC1; then xy 00 C.4C2x/y 0 C.2Cx/y D . 1Ct/y 00 C.2C2t/y 0 C.1Ct/y D


1/an t n

1
X

1/an t n

n.n

nD2

2/.nC 1/anC2 t n C

1
X

C2

1
X

nan t n

nD1

.nC 1/nanC1 t n C 2

nD0
1
X

. 2a2 C2a1 Ca0 /C

nD1

1
X

nD0

C2

1
X

nD1

nan t n C

.nC 1/anC1 t n C 2

1
X

nD0
1
X

nD0

an t n C

nan t n C

1
X

nD0
1
X

nD0

an t nC1 D
an t n C

1
X

an

1t

n
1 .xC2/ .

1/x r

C .r

.n C

nD1

.n C 2/.n C 1/anC2 C .n C 1/.n C 2/anC1 C .2n C 1/an C an

1
1
X
X
nan x n 1 C r x r 1
an x n D
.n C r /x nCr 1
nD0 "
nD0 #
nD0
1
1
X
X
d
d
y 00 D
y 0 .x/ D
xr 1
.n C r /an x n D x r 1
.n C r /nan x n
dx
dx
nD0
nD0
1
X
r /an x n D
.n C r /.n C r 1/an x nCr 2 .

nD0

1
X

7.1.20. y 0 .x/ D x r

n.n

nD2
1
X

1
X

nD0

.n C

nD0

7.1.22. x 2 .1 C x/y 00 C x.1 C 2x/y 0 .4 C 6x/y D .x 2 y 00 C xy 0


1
1
X
X
.n C r /.n C r 1/ C .n C r / 4an x nCr C
.n C r /.n C r

nD0
1
X

nD0

.n C r

2/.n C r C 2/an x

r C 2/an x nCr C
bn D .n C r

1
X

nD1

nCr

1
X

nD0

.n C r C 2/.n C r

nD0

nCr
D xr
1x

2/.n C r C 2/an C .n C r C 2/.n C r

1
X

nD0

3/an

1,

1
X

6y/ D

6an x nCr C1 D

1/ C 2.n C r /

2/an x nCr C1 D

.n C r C 3/.n C r

3/an

4y/ C x.x 2 y 00 C 2xy 0

.n C r

nD0

bn x n with b0 D .r

2/.n C

2/.r C 2/a0 and

n  1.

7.1.24. x 2 .1 C 3x/y 00 C x.2 C 12x C x 2 /y 0 C 2x.3 C x/y D .x 2 y 00 C 2xy 0 / C x.3x 2 y 00 C 12xy 0 C


1
1
X
X
6y/ C x 2 .xy 0 C 2y/ D
.n C r /.n C r 1/ C 2.n C r /anx nCr C
3.n C r /.n C r 1/ C 12.n C
r / C 6an x

nCr C1

r C 2/an x nCr C1 C
1/an

1x

nCr

1
X

nD2

1
X

nD0
1
X

nD0

nD0

.n C r / C 2an x

nCr C2

.n C r C 2/an x nCr C2 D

.n C r /an

2x

nCr

D xr

1
X

nD0

1
X

nD0
1
X

nD0

.n C r /.n C r C 1/an x nCr C 3

.n C r /.n C r C 1/an x nCr C 3

nD0

1
X

.n C r C 1/.n C

nD0
1
X

.n C r /.n C r C

nD1

bn x n with b0 D r .r C 1/a0 , b1 D .r C 1/.r C 2/a1 C

3.r C 1/.r C 2/a0 , bn D .n C r /.n C r C 1/an C 3.n C r /.n C r C 1/an

C .n C r /an

2,

n  2.

7.1.26. x 2 .2 C x 2 /y 00 C 2x.5 C x 2/y 0 C 2.3 x 2 /y D .2x 2 y 00 C 10xy 0 C 6y/C x 2 .x 2 y 00 C 2xy 0 2y/ D


1
1
X
X
2.n C r /.n C r 1/ C 10.n C r / C 6an x nCr C
.n C r /.n C r 1/ C 2.n C r / 2an x nCr C2 D

nD0
1
X

nD0

.n C r C 1/.n C r C 3/an x nCr C

1
X

nD0

nD0

.n C r

1/.n C r C 2/an x nCr C2 D 2

1
X

nD0

.n C r C 1/.n C

Section 7.2 Series Solutions Near an Ordinary Point I


r C 3/an x nCr C

1
X

nD2

.n C r

3/.n C r /an

2x

nCr

1
X

D xr

nD0

93

bn x n with b0 D 2.r C 1/.r C 3/a0 ,

b1 D 2.r C 2/.r C 4/a1 bn D 2.n C r C 1/.n C r C 3/an C .n C r

3/.n C r /an

2,

n  2.

7.2 SERIES SOLUTIONS NEAR AN ORDINARY POINT I


7.2.2. p.n/ D n.n
a2m D

1/ C 2n

. 1/m
a0 ; a2mC3 D
2m 1

2 D .n C 2/.n

1/; anC2 D

n 1
a ;
nC1 n
1
X

a2mC2 D

2m 1
a2m , so
2m C 1

m
x 2m
a2mC1 D 0 if m  0; y D a0
. 1/mC1
C a1 x.
mC2
2m 1
mD0

7.2.4. p.n/ D n.n 1/ 8n 12 D .n C 3/.n C 4/; anC2 D .nC3/.nC4/


a ; a2mC2 D
.nC2/.nC1/ n
.m C 2/.2m C 3/
.m C 2/.2m C 5/
a2m , so a2m D .mC1/.2mC1/a0 ; a2mC3 D
a2mC1 so a2mC1 D
.m C 1/.2m C 1/
.m C 1/.2m C 3/
1
1
X
.m C 1/.2m C 3/
a1 X
a1 ; y D a0
.m C 1/.2m C 1/x 2m C
.m C 1/.2m C 3/x 2mC1 .
3
3
mD0
mD0
.4m C 1/2
1
.2n C 1/2
.2nC1/2
an ; a2mC2 D
7.2.6. p.n/ D n.n 1/C2nC D
; anC2 D 4.nC2/.nC1/
a2m ,
4
8.m C 1/.2m C 1/
3 4
2
m
Y1 .4j C 1/2
.4m C 3/2
m4
5 1 a0 ; a2mC3 D
so a2m D . 1/
a2mC1 so a2mC1 D
2j C 1
8m m
8.2m C 3/.m C 1/
j D0
3
2
3
2
3
2
1
m
1
m
m
X
Y1 .4j C 1/2 x 2m
X
Y1 .4j C 3/2 x 2mC1
Y1 .4j C 3/2
1
5
5
5
. 1/m 4
a1 ; y D a0
. 1/m 4
C a1
. 1/m 4
m m
2j C 3
8m m
2j
C
1
8
2j
C
3
8m m
mD0
mD0
j D0

j D0

7.2.8. p.n/ D n.n 1/ 10nC28 D .n 7/.n 4/; anC2 D

j D0

5
35
a2 D
a0 , a2m D 0 if m  3; a2mC3 D
6
3
1
3
1
1
a3 D 3a1 , a5 D
a3 D a1 , a7 D
a5 D
a1 ;
5
35
5
 21


3 5
1 7
35 4
3
2
y D a0 1 14x C x C a1 x 3x C x C x .
3
5
35

so a2 D

2.2m 7/.m 2/
a2m ,
2.m C 1/.2m C 1/
.m 3/.2m 3/
a2mC1 , so
.2m C 3/.m C 1/

.n 7/.n 4/
.nC2/.nC1/ an ; a2mC2

14a0 , a4 D

2n C 3
4m C 3
7.2.10. p.n/ D 2n C 3; anC2 D
an ; a2mC2 D
a2m , so a2m D
.n
C
2/.n
C
1/
2.m
C
1/.2m
3
2 C 1/
3
2
m
m
1
Y
Y1 4j C 3 . 1/m
4m C 5
4j C 5 . 1/m
5
5
4
a0 ; a2mC3 D
a2mC1 so a2mC1 D 4
a1 ;
m
2j C 1 2 m
2.2m C 3/.m C 1/
2j C 3 2m m
j D0
j D0
2
3
2
3
1
m
1
m
X
Y1 4j C 3 x 2m
X
Y1 4j C 5 x 2mC1
m4
m
5
5
y D a0
. 1/
C a1
. 1/ 4
2j C 1 2m m
2j C 3 2m m
mD0

7.2.12. p.n/ D 2n.n

a1 D y 0 .0/ D 1.

7.2.13. p.n/ D 8n.n

y 0 .0/ D 1.

mD0

j D0

1/

9n

6 D .n

1/ C 2 D 2.2n

j D0

6/.2n C 1/; anC2 D


1/2 ; anC2 D

.n 6/.2n C 1/
an ; a0 D y.0/ D 1;
.n C 2/.n C 1/

2.2n 1/2
an ; a0 D y.0/ D 2; a1 D
.n C 2/.n C 1/

94 Chapter 7 Series Solutions of Linear Second Order Equations


1
1
1
an ; a2mC2 D
a2m , so a2m D
a0 ;
.n C 2/.n C 1/
.2m C 2/.2m C 1/
.2m/
1
1
X
X
1
1
.x 3/2m
.x 3/2mC1
D
a2mC1 , so a2mC1 D
a1 ; y D a0
C a1
.
.2m C 3/.2m C 1/
.2m C 1/
.2m/
.2m C 1/
mD0
mD0

7.2.16. p.n/ D
a2mC3

1; anC2 D

7.2.18. Let t D x

2t 2/y 00

10ty 0

2n.n 2 1/ 10n 36 D
m 1
2.n C 3/
2m C 3
1 4Y
2.n C 1/.n C 3/; anC2 D
an ; a2mC2 D
a2m , so a2m D
.2j C 3/5 a0 ;
nC2
mC1
m
j D0
2
3
1
m
1
m
X
Y
4.m C 2/
4 .m C 1/
.x 1/2m
4
a2mC1 , so a2mC1 D Qm 1
a1 ; y D a0
.2j C 3/5
a2mC3 D
2m C 3
m
j D0 .2j C 3/
mD0
1; then .1

6y D 0; p.n/ D

j D0

1
X

4m .m C 1/
C a1
.x
Qm 1
j D0 .2j C 3/
mD0

1/2mC1 .



3t 2
7.2.20. Let t D x C 1; then 1 C
y 00
3
1/2 ; anC2 D

3.n C 1/
an ; a2mC2 D
2.n C 2/

9t 0 3
3
9
3
3
y C y D 0; p.n/ D n.n 1/ C n C D .n C
2
2
2
2
2
2
3 2
m
1
Y
3.2m C 1/
3m
a2m , so a2m D . 1/m 4
.2j C 1/5 m a0 ;
4.m C 1/
4 m
j D0

3.m C 1/
3m m
a2mC3 D
a2mC1 , so a2mC1 D . 1/m Qm 1
a1 ;
2m C 3
j D0 .2j C 3/
2
3
1
m
1
X
Y1
X
3m
3m m
y D a0
. 1/m 4
.2j C 1/5 m .x C 1/2m C a1
. 1/m Qm 1
.x C 1/2mC1 .
4 m
.2j
C
3/
j D0
mD0

mD0

j D0

7.2.22. p.n/ D n C 3; anC2 D

nC3
an ; a0 D y.3/ D 2; a1 D y 0 .3/ D 3.
.n C 2/.n C 1/

7.2.24. Let t D x 3; .1 C 4t 2 /y 00 C y D 0; p.n/ D .4n.n 1/ C 1 D 2n 1/2 ; anC2 D


.2n 1/2
an ; a0 D y.3/ D 4; a1 D y 0 .3/ D 6.
.n C 2/.n C 1/


20 0
2
20
2t 2
7.2.26. Let t D x C 1; 1 C
y 00
ty C 20y D 0; p.n/ D n.n 1/
n C 20 D
3
3
3
3
2.n 6/.n 5/
2.n 6/.n 5/
; anC2 D
an ; a0 D y. 1/ D 3; a1 D y 0 . 1/ D 3.
3
3.n C 2/.n C 1/
p.n/
p.2m/
7.2.28. From Theorem 7.2.2, anC2 D
an ; a2mC2 D
a2m , so
.n C 2/.n C 1/
.2m C22/.2m C 1/
2
3
3
m
m
Y1
Y1
. 1/m
p.2m C 1/
. 1/m
a2m D 4
p.2j /5
a0 ; a2mC3 D
a2m , so a2mC1 D 4
p.2j C 1/5
a1 .
.2m/
.2m C 3/.2m C 2/
.2m C 1/
j D0

j D0

7.2.30. (a) Here p.n/ D n.n 1/ C 2bn . C 2b 1/ D .n /.n C C 2b


Exercise 7.2.28 implies that y1 and y2 have the stated forms. If D 2k, then
2
3
1
m
X
Y1
x 2m
4
y1 D
.2j 2k/.2j C 2k C 2b 1/5
.2m/
mD0

j D0

1/, so

.C/:

Section 7.2 Series Solutions Near an Ordinary Point I

95

If D 2k C 1, then
y2 D

1
X

mD0

m
Y1

.2j

j D0

2k/.2j C 2k C 2b/5

x 2mC1
:
.2m C 1/

.D/

Q
Since 2b is not a negative integer and jmD01 .2j 2k/ D 0 if m > k, y1 in (C) and y2 in (D) have the
stated properties. This implies the conclusions regarding Pn .
(b) Multiplying (A) through by .1 x 2 /b 1 yields
x 2 /b Pn0 0 D n.n C 2b

1/.1

x 2 /b

x 2 /b Pm0 0 D m.m C 2b

1/.1

x 2 /b

.1

Pn :

.E/

(c) Therefore,
.1

Pm :

.F/

Subtract Pn times (F) from Pm times (E) to obtain (B).


(d) Integrating the left side of (B) by parts over 1; 1 yields zero, which implies the conclusion.
7.2.32. (a) Let Ly D .1 C x 3/y 00 C x 2 y 0 C xy. If y D
1
X

p.n/an x

nC1

nD0

D 2a2 C

1
X

nD0

1
X

nD0

1
X

an x n , then Ly D

n.n

1/an x n

nD2

.n C 3/.n C 2/anC3 C p.n/an x nC1 D 0 if and only if a2 D 0 and

p.n/
an for n  0.
.n C 3/.n C 2/

anC3 D

7.2.34. p.r / D
. 1/m 2m

2r .r

m
Y1

m
Y1 . 2/.3j C 2/2
p.3j /
D
D
3j C 2
3j C 2

10r

m
Y1

m
Y1 . 2/.3j C 3/2
. 1/m 2m .m/2
p.3j C 1/
D
D Qm 1
. Substituting (A)
3j C 4
3j C 4
j D0 .3j C 4/

m
Y1
j D0

2.r C 2/2 ; (A)

8 D

1/

.3j C 2/; (B)

j D0

j D0

j D0

j D0

and (B) into the result 2


of Exercise 7.2.32(c)
yields
3
1  m m
1
1
3m
X
Y
X
2
x
6m m
4
y D a0
.3j C 2/5
C a1
x 3mC1 .
Qm 1
3
m
.3j
C
4/
j D0
mD0
mD0
j D0

7.2.36. p.r / D 2r .r
(B)

m
Y1
j D0

1/ C 6r C 24 D 2.r

m
Y1 . 6/.j C 1/.3j
p.3j C 1/
D
3j C 4
3j C 4
j D0

6/.r C 2/; (A)


5/

m
Y1

m
Y1
p.3j /
D
. 6/.j
3j C 2

j D0
m
Y1
m m

D . 1/ 6 m

(B) into the result of Exercise 7.2.32(c) yields


2
3
1
m
X
Y1 3j 5
5 x 3mC1 .
y D a0 .1 4x 3 C 4x 6/ C a1
2m 4
3j
C
4
mD0

j D0

2/.

j D0

3j 5
. Substituting (A) and
3j C 4

j D0

7.2.38. (a) Let Ly D .1 C x kC2/y 00 C x kC1 y 0 C x k y. If y D


1/an x n

1
X

nD0

p.n/an x nCk D

1
X

nD k

.n C k C 2/.n C k

1
X

nD0

an x n , then Ly D

1/anCkC2 x nCk C

1
X

nD0

1
X

n.n

nD2

.n C k C 2/.n C

96 Chapter 7 Series Solutions of Linear Second Order Equations


k C 1/anCkC2 C p.n/an x nCk D 0 if and only if ak D 0 for 2  n  k C 1 and (A) anCkC1 D
p.n/
an for n  0.
.n C k C 2/.n C k C 1/
(b) If an D 0 the anC.kC2/m D 0 for all m  0, from (A).
m
Y1

m
Y1 p.4j C 1/
p.4j /
1
1
D Qm 1
; (B)
D Qm 1
.
7.2.40. k D 2 and p.r / D 1; (A)
4j C 3
.4j
C
5/
j D0 .4j C 3/
j D0 4j C 5
j D0
j D0
Substituting (A) and (B) into the result of Exercise t.2.38(c) yields
1
1
X
X
x 4m
x 4mC1
y D a0
. 1/m
C
a
. 1/m
.
Q
Q
1
m
1
4m m j D0 .4j C 3/
4m m jmD01 .4j C 5/
mD0
mD0

7.2.42. k D 6 and p.r / D r .r 1/ 16r C72 D .r 9/.r 8/; (A)


(B)

m
Y1
j D0

m
Y1 8.j 1/.8j
p.8j C 1/
D
.8j C 9/
8j C 9

7/

m
Y1
j D0

m
Y1 8.j 1/.8j
p.8j /
D
8j C 7
8j C 7

9/

j D0

j D0

Substituting
 (A) and
 (B) into
 the result
 of Exercise 7.2.38(c) yields
9 8
7 9
y D a0 1
x C a1 x
x .
7
9
7.2.44. k D 4 and p.r / D r C 6; (A)
(B)

m
Y1
j D0

p.6j C 1/
D 1;
.6j C 7/

m
Y1

m
Y1 6.j C 1/
6m m
p.6j /
D
D Qm 1
;
6j C 5
6j C 5
j D0 .6j C 5/

j D0

j D0

Substituting (A) and (B) into the result of Exercise 7.2.38(c) yields
1
1
X
X
x 6m
x 6mC1
m
y D a0
. 1/ Qm 1
C a1
. 1/m m .
6 m
j D0 .6j C 5/
mD0
mD0

7.3 SERIES SOLUTIONS NEAR AN ORDINARY POINT II


1
1
X
X
n
2 00
0
7.3.2. If y D
an x , then .1 C x C 2x /y C .2 C 8x/y C 4y D
n.n
1/an x n

nD0
1
X

C2

1/an x

n 2

n.n

nD2

1/an x n C 2

1
X

nan x n

nD1

C8

1
X

nD1

nan x n C 4

1
X

nD0

an x n D

1
X

nD0

1/.anC2 C anC1 C 2an /x n D 0 if anC2 D anC1 2an , an  0. Starting with a0 D


yields y D 1 C 2x 4x 3 C 4x 4 C 4x 5 12x 6 C 4x 7 C   .
7.3.4. If y D
1/an x n

1
X

an x n , then .1 C x C 3x 2/y 00 C .2 C 15x/y 0 C 12y D

nD0
1
X

C3

nD2

nD2

n.n

1/an x n C 2

1
X

nan x n

nD1

C 15

1
X

nD1

1
X

x2

nan x n C 12

7 3 15 4 45 5
x C x C x
2
2
8

1/an x n

nD2

1/anC2 C .n C 1/.n C 2/anC1 C 3.n C 2/2an x n D 0 if anC2 D anC1


with a0 D 0 and a1 D 1 yields y D x

n.n

1
X

nD0

an x n D

1
X

n.n

nD2

.n C 2/.n C

1 and a1 D 2

1
X

nD0

1
X

n.n

nD2

.n C 2/.n C

3.n C 2/
an , an  0. Starting
nC1
261 6 207 7
x C
x C   .
8
16

Section 7.3 Series Solutions Near an Ordinary Point II

7.3.6. If y D
3

1
X

1
X

nD0

an x n , then .3 C 3x C x 2 /y 00 C .6 C 4x/y 0 C 2y D 3

1/an x n

n.n

nD2

1
X

1/an x n C 6

n.n

nD2

1
X

nan x n

C4

nD1

1
X

nD1

1
X

n.n

nD2
1
X

nan x n C 2

nD0

97

1/an x n

an x n D

1
X

nD0

.n C

2/.n C 1/3anC2 C 3anC1 C an x n D 0 if anC2 D anC1 an =3, an  0. Starting with a0 D 7 and


16 2 13 3 23 4 10 5
7 6 1 7
a1 D 3 yields y D 7 C 3x
x C x
x C x
x
x C   .
3
3
9
9
27
9
7.3.8. The equation is equivalent to .1 C t C 2t 2 /y 00 C .2 C 6t/y 0 C 2y D 0 with t D x 1. If
1
1
1
X
X
X
yD
an t n , then .1 C t C 2t 2 /y 00 C .2 C 6t/y 0 C 2y D
n.n 1/an t n 2 C
n.n 1/an t n 1 C
nD0

1
X

1/an t C 2

n.n

nD2

1
X

nan t

n 1

nD1

C6

1
X

nD1

nD2

nan t C 2

2/anC1 C2.nC1/an t n D 0 if anC2 D anC1


yields y D 1

4
1/ C .x
3

.x

4
.x
3

1/3

1
X

nD0

an t D

1
X

nD0

nD2

.n C 1/.n C 2/anC2 C .n C

2.n C 1/
an , an  0. Starting with a0 D 1 and a1 D 1
nC2
4
136
104
1/4
.x 1/5 C
.x 1/6
.x 1/7 C   .
5
45
63

7.3.10. The equation is equivalent to .1 C t C t 2 /y 00 C .3 C 4t/y 0 C 2y D 0 with t D x 1. If


1
1
1
X
X
X
yD
an t n , then .1 C t C t 2 /y 00 C .3 C 4t/y 0 C 2y D
n.n 1/an t n 2 C
n.n 1/an t n 1 C
nD0

1
X

n.n 1/an t n C3

.x

nD2

1
X

nan t n

nD1

C4

nD1

nan t n C2

1
X

nD0

nD2

an t n D

1
X

nD0

nD2

.nC1/.nC2/anC2 C.nC3/anC1 C

nC3
anC1 an , an  0. Starting with a0 D 2 and a1 D 1 yields y D
nC2
5
19
7
59
1091
1/2 C .x 1/3
.x 1/4 C .x 1/5 C .x 1/6
.x 1/7 C   
3
12
30
45
630

.n C 2/an t n D 0 if anC2 D
1
.x
2

1/

1
X

7.3.12. The equation is equivalent to .1 C 2t C t 2 /y 00 C .1 C 7t/y 0 C 8y D 0 with t D x 1. If y D


1
1
1
1
X
X
X
X
an t n , then .1C2t Ct 2 /y 00 C.1C7t/y 0 C8y D
n.n 1/an t n 2 C2
n.n 1/an t n 1 C
n.n

nD0

1/an t n C

1
X

nan t n

nD1

C7

1
X

nD1

nan t n C 8

7.3.16. If y D
2

1
X

nD1

nan x

2.x
1
X

nD1

0 if anC2 D anC1

3.x

an x n , then .1 x/y 00

nD0
1
X
n 1

1/

nan x n C

1
X

nD0

nD2

an t n D

1
X

nD0

nD2

nD2

.n C 2/.n C 1/anC2 C .n C 1/.2n C 1/anC1 C

2n C 1
nC4
anC1
an , an  0. Starting with a0 D 1 and a1 D 2
nC2
nC1
42
604
1/2 C 8.x 1/3 4.x 1/4
.x 1/5 C 19.x 1/6
.x 1/7 C   
5
35

.nC2/.nC4/an t n D 0 if anC2 D
yields y D 1

1
X

nD0

.2 x/y 0 C y D

an x n D

an
,an  0.
nC2

1
X

nD0

1
X

nD2

n.n 1/an x n

1
X

n.n 1/an x n

nD2

.nC2/.nC1/anC2 .nC2/.nC1/anC1 C.nC1/an x n D

98 Chapter 7 Series Solutions of Linear Second Order Equations

7.3.18. If y D
1
X

nan x n

nD1

anC2 D

1
X

an x n , then .1 C x 2 /y 00 C y 0 C 2y D

nD0
1
X

C2

nD0

an x n D

1
X

nD0

1
X

1/an x n

n.n

nD2

1
X

1/an x n C

n.n

nD2

.n C 2/.n C 1/anC2 C .n C 1/anC1 C .n2

n C 2/an x n D 0 if

n2 n C 2
an .
.n C 2/.n C 1/

1
anC1
nC2

7.3.20. The equation is equivalent to .3 C 2t/y 00 C .1 C 2t/y 0 .1 2t/y D 0 with t D x 1. If


1
1
1
X
X
X
y D
an t n , then .3 C 2t/y 00 C .1 C 2t/y 0 .1 2t/y D 3
n.n 1/an t n 2 C 2
n.n
nD0

1/an t

n 1

1
X

nan t

n 1

nD1

C2

1
X

nan t

nD1

1
X

nD0

an t C 2

1
X

nD2

an t

nC1

nD0

nD2

D .6a2 C a1

1
X

a0 / C

nD1

3.n C

a1 a0
2/.n C 1/anC2 C .n C 1/.2n C 1/anC1 C .2n 1/an C 2an 1 t n D 0 if a2 D
and anC2 D
6
2n 1
2
2n C 1
anC1
an
an 1 , n  1. Starting with a0 D 1 and
3.n C 2/
3.n C 2/.n C 1/
3.n C 2/.n C 1/
1
1
5
73
a1 D 2 yields y D 1 2.x 1/ C .x 1/2
.x 1/3 C .x 1/4
.x 1/5 C   .
2
6
36
1080
7.3.22. The equation is equivalent to .1Ct/y 00 C.2 2t/y 0 C.3Ct/y D 0 with t D xC3. If y D
then .1 C t/y 00 C .2
2

1
X

nD1

nan t C 3

1/anC1

.2n

1
X

nD0

2t/y 0 C .3 C t/y D
n

an t C

3/an C an

1
X

an t

nC1

nD0
1 t

1
X

n.n

1/an t n

nD2

D .2a2 C 2a1 C 3a0 / C

D 0 if a2 D

1
X

nD2
1
X

nD1

n.n

1/an t n

2.x C 3/

C2

2a1 C 3a0
and anC2 D
2

anC1 C

2 yields
11
67
.x C 3/ C .x C 3/3
.x C 3/4 C .x C 3/5 C   .
12
60

1
X

n.n 1/an t n

nD2

an t nC1 D .2a2 C 3a1 C a0 / C

if a2 D

3a1 C a0
and anC2 D
2

a1 D 3 yields y D 2

nan t n

nD1

.1 C 2t/y 00 C 3y 0 C .1 t/y D

nD0

nD0

.2n 3/an an 1
,
.n C 2/.n C 1/

7.3.24. The equation is equivalent to .1C2t/y 00 C3y 0 C.1 t/y D 0 with t D xC1. If y D

1
X

an t n ,

.n C 2/.n C 1/anC2 C .n C 2/.n C

n  1. Starting with a0 D 2 and a1 D


yD2

1
X

1
X

1
X

nD1

C2

1
X

nD2

n.n 1/an t n

C3

1
X

nan t n

nD1

.n C 2/.n C 1/anC2 C .2n C 3/.n C 1/anC1 C an

1
X

an t n , then

nD0
1

1
X

an t n

nD0

an

1 t

D0

2n C 3
an an 1
anC1
, n  1. Starting with a0 D 2 and
nC2
.n C 2/.n C 1/
7
197
287
3.x C 1/ C .x C 1/2 5.x C 1/3 C
.x C 1/4
.x C 1/5 C   .
2
24
20

7.3.26. The equation is equivalent to .6

2t/y 00 C .3 C t/y D 0 with t D x

2. If y D

1
X

nD0

an t n , then

Section 7.3 Series Solutions Near an Ordinary Point II


2t/y 00 C .3 C t/y D 6

.6

1
X

.12a2 C 3a0 / C

nD1

1
X

1/an t n

n.n

nD2

1
X

1/an t n

n.n

C3

nD2

6.n C 2/.n C 1/anC2

1
X

an t n C

nD0

2.n C 1/nanC1 C 3an C an

1 t

1
X

nD0

an t nC1 D

D 0 if a2 D

n
3an C an 1
anC1
, n  1. Starting with a0 D 2 and a1 D
3.n C 2/
6.n C 2/.n C 1/
1
2
49
23
4.x 2/
.x 2/2 C .x 2/3 C
.x 2/4 C
.x 2/5 C   .
2
9
432
1080

and anC2 D
yD2

7.3.28. The equation is equivalent to .2 C 4t/y 00


.2 C 4t/y 00

a0 / C

.4a2

2t/y D 2

.1
1
X

nD1

1
X

n.n

1/an t n

nD2

2t/y D 0 with t D x C 4. If y D

.1
C4

1
X

1/an t n

n.n

1
X

nD2

nD0

2.n C 2/.n C 1/anC2 C 4.n C 1/nanC1

an C 2an

1 t

an t n C 2

1
X

nD0
1
X

nD0

99

a0
4

4 yields

an t n , then
an t nC1 D

D 0 if a2 D

a0
and
4

an 2an 1
2n
anC2 D
anC1 C
, n  1. Starting with a0 D 1 and a1 D 2 yields y D
nC2
2.n C 2/.n C 1/
1
1
65
67
1 C 2.x C 1/
.x C 1/2 C .x C 1/3
.x C 1/4 C .x C 1/5 C   .
4
2
96
80
N=5; b=zeros(N,1); b(1)=-1;b(2)=2; b(3)=b(1)/4; for n=1:N-2 b(n+3)=-2*n*b(n+2)/(n+2)+(b(n+1)2*b(n))/(2*(n+2)*(n+1)); end
7.3.29. Let Ly D .1 C x C x 2 /y 00 C . C x/y 0 C y. If y D
1/an x n
1
X

1
X

n.n 1/an x n

nD2

.n C 2/.n C 1/anC2 x n C

nD0
1
X

nD0

nan x n C 

n.n

1
X

nD0

an x n D

1
X

nD0

1
X

nD0

1
X

nD2

n.n 1/an x n C

.n C 1/nanC1 x n C

1
X

nD0

nD0

nan x n

nD1

1
X

1
X

n.n

an x n , then Ly D

1
X

nD1

nan x n C

1/an x n C

1
X

nD0

1
X

n.n

nD2
1
X

nD0

an x n D

.n C 1/anC1 x n C

bn x n , where bn D .n C 1/.n C 2/anC2 C .n C 1/. n C /anC1 C

1/ C n C an , which implies the conclusion.

7.3.30. (a) Let D 2, D 4, and  D 2 in Exercise 7.3.29 to obtain (B).


(b) If an D c1 r1n C c2 r2n , then anC2 C anC1 C an D c1 r1n .r12 C r C / C c2 r2n .r22 C r2 C
/ D c1 r1n P0 .r1 / C c2 r2n PP
0 .r2 / D 0, so fan g satisfies (B). Since 1=r1 and 1=r2 are the zeros of P0 ,
n
n n
Theorem 7.2.1 implies that 1
nD0 .c1 r1 C c2 r2 /x is a solution of (A) on . ; /.
1
X
1
(c) If jxj < , then jr1 xj <  and jr2 xj < 1, so
rin x n D
D yi , i D 1; 2. Therefore, (b)
1
ri x
nD0
implies that fy1 ; y2g is a fundamental set of solutions of (A) on . ; /.
(d) (A) can written as P0 y 00 C 2P00 y 0 C P000 y D .P0 y/00 D 0. Therefore,P0 y D a C bx where a and
b are arbitrary constants, and a partial fraction expansion shows that the general solution of (A) on any
a C bx
c1
c2
interval not containing 1=r1 or 1=r2 is y D
D
C
D c1 y1 C c2 y2 .
P0 .x/
1 r1 x
1 r2 x
(e) If an D c1 r1n C c2r2n , then anC2 C anC1 C an D c1r1n .r12 C r C / C c2r1n .n C 2/r12 C .n C
1/r2 C n/ D .c1 C nc2 /r1n P0 .r1 / C c2 r1n P00 .r1 / D 0, so fan g satisfies (B). Since 1=r1 is the only zero

100 Chapter 7 Series Solutions of Linear Second Order Equations


of P0 , Theorem 7.2.1 implies that
(f) If jxj < , then jr1 xj < ,
1
X

P1

n n
nD0 .c1 C c2 n/r1 /x is
1
X
1
so
r1n x n D
1
r1 x
nD0

a solution of (A) on . ; /.


D y1 . Differentiating this and multiplying

r1 x
D r1 y2 . Therefore, (e) implies that fy1 ; y2g is a
.1 r1 x/2
nD0
fundamental set of solutions of (A) on . ; /.
(g) The argument is the same as in (c), but now the partial fraction expansion can be written as y D
a C bx
c1
c2 x
D
C
D c1y1 C c2y2 .
P0 .x/
1 r1 x
.1 r2 x/2
the result by x shows that

7.3.32. If y D
3

1
X

nD0

2an

an x C 2

2 x

1
X

nD0

1
X

nr1n x n D

00

an x , then y C 2xy C .3 C 2x /y D

an x

nD0

D 0 if a2 D

nC2

D .2a2 C 3a0 / C .6a3 C 5a1 /x C

3a0 =2, a3 D

with a0 D 1 and a1 D 2 yields y D 1


7.3.34. If y D
1
X

1
X

nD0

3a0 / C .6a3 C 2a1 /x C


a1 =3, and anC2 D

a1 D 2 yields y D 6

2
2x C 9x 2 C x 3
3

1
X

nD0

4an

an x n C 4

2 x

1
X

an x n , then y 00

nD0

1
X

nD0

an x n

2 x

nD0

1
X

nD0

nD0

an

1
X

1
X

1
X

nD1

nan x n C

.n C 2/.n C 1/anC2 C .2n C 3/an C

n.n 1/an x n

nD2

C5

1
X

, n  2. Starting

nan x n 3

nD1

1
X

nD0

3/an C an

an x n C

2 x

D0

.5n 3/an C an 2
, n  2. Starting with a0 D 6 and
.n C 2/.n C 1/
23 4
3 5
x
x C   .
4
10

a1 /x C

2x 4

x3

1/an x n

n.n

nD2

1
X

nD2

x 2 /y D 3

nD1

.n C 2/.n C 1/anC2

1
X

nD2

1
X

1
X

n.n

.3n

nan x n C
2/an C

, n  2. Starting with a0 D 3

1/an x n

C2

1
X

nD1

nan x n C

3.n C 2/.n C 1/anC2 C .2n C 4/an

nD2

.2n C 4/an an 2
, n  2. Starting with
3.n C 2/.n C 1/
19 4 13 5
x C x C .
54
60

a1 =3, and anC2 D

4
a0 D 2 and a1 D 3 yields y D 2 C 3x C x 2
3

1
X

.3n 2/an 4an


.n C 2/.n C 1/
17 5
x C   .
20

an x nC2 D .6a2 C 4a0 / C .18a3 C 6a1 /x C


2a0 =3, a3 D

nD2

3xy 0 C .2 C 4x 2/y D

3x 2 C x 3

C2

nD2

1
X

nD2

an x n , then 3y 00 C 2xy 0 C .4

D 0 if a2 D

1/an x

.n C 2/.n C 1/anC2 C .5n

D 0 if a2 D a0 , a3 D a1 =6, and anC2 D

7.3.38. If y D
1
X

1
X

an x nC2 D .2a2 C 2a0 / C .6a3

and a1 D 6 yields y D 3 C 6x

n.n

n 2

.2n C 3/an C 2an


.n C 2/.n C 1/
3 2 5 3 17 4 11 5
x C x C x
x C   .
2
3
24
20

2x

if a2 D 3a0 =2, a3 D

7.3.36. If y D

1
X

5a1 =6, and anC2 D

an x n , then y 00 C5xy 0 .3 x 2 /y D

an x nC2 D .2a2

nD0

Section 7.3 Series Solutions Near an Ordinary Point II

7.3.40. If y D
1/an x n

1
X

nD0

1
X

nD1

an x n , then .1 C x/y 00 C x 2 y 0 C .1 C 2x/y D

nan x nC1 C

1
X

nD0

an x n C 2

1
X

1
X

nD2
1
X

an x nC1 D .2a2 C a0 / C

nD0

n.n

nD1

1/an x n

1
X

101

n.n

nD2

.n C 2/.n C 1/anC2 C .n C

.n C 1/nanC1 C an C .n C 1/an
.n C 2/.n C 1/
1 3 3 4
31 5
2
n  1. Starting with a0 D 2 and a1 D 3 yields y D 2 C 3x C x
x
x C
x C   .
6
4
120

1/nanC1 C an C .nC 1/an

7.3.42. If y D
2

1
X

nan x

nD1

1
X

D 0 if a2 D

nD1

nan x nC1 C

1
X

nD1

a0 =2 and anC2 D

an x n , then .1Cx 2 /y 00 C.2Cx 2 /y 0 Cxy D

nD0
1
X
n 1

1 x

1
X

n.n 1/an x n

nD2
1
X

an x nC1 D .2a2 C2a1 /C

nD1

1
X

nD2

n.n 1/an x n C

.nC2/.nC1/anC2 C2.nC1/anC1 C

2.n C 1/anC1 C n.n 1/an C nan 1


, n  1.
.n C 2/.n C 1/
23
23 4 11 5
Starting with a0 D 3 and a1 D 5 yields y D 3 C 5x 5x 2 C x 3
x C x C   .
6
12
30

n.n 1/an C nan

1 x

D 0 if a2 D a1 and anC2 D

7.3.44. The equation is equivalent to y 00 C .1 C 3t 2/y 0 C .1 C 2t/y D 0 with t D x


then y 00 C .1 C 3t 2 /y 0 C .1 C 2t/y D
2

1
X

nD0

an t nC1 D .2a2 C a1 C a0 / C

if a2 D .a1 C a0 /=2 and anC2 D


and a1 D 3 yields

1
3.x C 2/ C .x C 2/2
2

yD2

1
X

n.n 1/an t n

nD2

1
X

nD1

1
X

nan t n

nD1

C3

1
X

nD1

2. If y D

nan t nC1 C

.n C 2/.n C 1/anC2 C .n C 1/anC1 C an C .3n

.n C 1/anC1 C an C .3n
.n C 2/.n C 1/

1/an

1
31
.x C 2/3 C .x C 2/4
3
24

53
.x C 2/5 C   .
120

nD0

1/an

t 2 /y 00

then .1
8

1
X

nan t n

nD1

9/an

1
X

nD1

.n

2/an

8t C t 2 /y 0 C ty D

.7

nan t nC1 C
1 t

1
X

nD0

n.n

1/an t n

nD2

an t nC1 D .2a2 7a1 /C

D 0 if a2 D 7a1 =2 and anC2 D

n  1. Starting with a0 D 2 and a1 D 1 yields


7
43
y D 2 .x C 2/
.x C 2/2
.x C 2/3
2
6

1
X

1
X

n.n

an t n ,

nD0

1
X

an t n C

1 t

D0

, n  1. Starting with a0 D 2

7.3.46. The equation is equivalent to .1 t 2 /y 00 .7 8t Ct 2 /y 0 Cty D 0 with t D xC2. If y D


1
X

1
X

1/an t n

nD2

1
X

1
X

an t n ,

nD0

nan t n

nD1

.nC2/.nC1/anC2 7.nC1/anC1 n.n

nD1

7.n C 1/anC1 C n.n 9/an C .n


.n C 2/.n C 1/

203
.x C 2/4
24

2/an

167
.x C 2/5 C   .
30

7.3.48. The equation is equivalent to .1 C 3t C 2t 2/y 00 .3 C t t 2 /y 0 .3 C t/y D 0 with t D x 1. If


1
1
1
X
X
X
yD
an t n , then .1 C 3t C 2t 2 /y 00 .3 C t t 2 /y 0 .3 C t/y D
n.n 1/an t n 2 C 3
n.n
nD0

nD2

nD2

102 Chapter 7 Series Solutions of Linear Second Order Equations


1/an t n

C2

1
X

n.n 1/an t n 3

nD2

.2a2 3a1 3a0 /C

1
X

nD1

1
X

nan t n

nD1

1
X

nD1

nan t n C

1
X

nan t nC1 3

nD1

1
X

an t n

nD0

1
X

nD0

an t nC1 D

.nC2/.nC1/anC2 C3.n2 1/anC1 C.2n2 3n 3/.nC1/an C.n 2/an

1 t

D0

1/anC1 C .2n2 3n 3/.n C 1/an C .n 2/an 1


,n
.n C 2/.n C 1/
3
1
1
1. Starting with a0 D 1 and a1 D 0 yields y D 1 C .x 1/2 C .x 1/3
.x 1/5 C   .
2
6
8
7.4 REGULAR SINGULAR POINTS; EULER EQUATIONS

if a2 D 3.a1 C a0 /=2 and anC2 D

3.n2

7r C 7 D .r

1/; y D c1 x C c2x 7 .

7.4.2. p.r / D r .r

1/

7.4.4. p.r / D r .r

1/ C 5r C 4 D .r C 2/2 ;y D x

7.4.6. p.r / D r .r

1/

7/.r

3r C 13 D .r

7.4.8. p.r / D 12r .r

1/

5r C 6 D .3r

7.4.10. p.r / D 3r .r

1/

r C 1 D .r

.c1 C c2 ln x/

2/2 C 9; y D x 2 c1 cos.3 ln x/ C c2 sin.3 ln x/.


2/.4r
1/.3r

3/; y D c1x 2=3 C c2 x 3=4 .


1/; y D c1x C c2 x 1=3 .

7.4.12. p.r / D r .r

1/ C 3r C 5 D .r C 1/2 C 4; y D

7.4.14. p.r / D r .r

1/

r C 10 D .r

1
c1 cos.2 ln x/ C c2 sin.2 ln x
x

1/2 C 9; y D x c1 cos.3 ln x/ C c2 sin.3 ln x/.

c1
C c2x 1=2 .
x





1
1
1
2
7.4.18. p.r / D 2r .r 1/C 10r C 9 D 2.r C 2/ C 1; y D 2 c1 cos p ln x C c2 sin p ln x .
x
2
2
7.4.16. p.r / D 2r .r

1/ C 3r

1 D .r C 1/.2r

1/; y D

7.4.20. If p.r / D ar .r 1/ C br C c D a.r r1 /2 , then (A) p.r1 / D p 0 .r1 / D 0. If y D ux r1 , then


y 0 D u0 x r1 C r1 ux r1 1 and y 00 D u00 x r1 C 2r1u0 x1r1 1 C r1 .r1 1/x1r1 2 , so
ax 2 y 00 C bxy 0 C cy

D
D

ax r1C2 u00 C .2ar1 C b/x r1 C1 u0 C .ar1 .r1 1/ C br1 C c/ x r1 u


r C1
ax r1C2 u00 C p 0 .r1 /x11 u0 C p.r /x r1 u D ax r1 C2 u00 ;

from (A). Therefore,u00 D 0, so u D c1 C c2x and y D x r1 .c1 C c2 x/.


7.4.22. (a) If t D x 1 and Y .t/ D y.t C 1/ D y.x/, then .1 x 2 /y 00 2xy 0 C . C 1/y D
d 2Y
dY
t.2 C t/ 2
2.1 C t/
C . C 1/Y D 0, so y satisfies Legendres equation if and only if Y
dt
dt
2
d Y
dY
d 2Y
satisfies (A) t.2 C t/ 2 C 2.1 C t/
. C 1/Y D 0. Since (A) can be rewritten as t 2 .2 C t/ 2 C
dt
dt
dt
dY
2t.1 C t/
. C 1/tY D 0, (A) has a regular singular point at t D 00 .
dt
d2Y
(b) If t D x C 1 and Y .t/ D y.t 1/ D y.x/, then .1 x 2 /y 00 2xy 0 C . C 1/y D t.2 t/ 2 C
dt
dY
d2Y
2.1 t/
C . C 1/Y , so y satisfies Legendres equation if and only if Y satisfies (B) t.2 t/ 2 C
dt
dt

Section 7.5 The Method of Frobenius I

103

dY
d 2Y
dY
C . C 1/Y , Since (B) can be rewritten as (B) t 2 .2 t/ 2 C 2t.1 t/
C . C 1/tY ,
2.1 t/
dt
dt
dt
(B) has a regular singular point at t D 00 .
7.5 The Method of Frobenius I
7.5.2. p0 .r / D r .3r 1/; p1 .r / D 2.r C 1/; p2 .r / D 4.r C 2/.
2
2an 1 .r / 4an 2 .r /
a1 .r / D
; an .r / D
, n  1.
3r C 2
3n C 3r 1
2an 1 .1=3/ 4an 2 .1=3/
, n 1;
r1 D 1=3; a1 .1=3/ D 2=3; an .1=3/ D
3n


2
8
40
y1 D x 1=3 1
x C x2
x3 C    .
3
9
81
2an 1 .0/ 4an 2 .0/
, n  1;
r2 D 0; a1 .0/ D 1; an .0/ D
3n 1
6
4 3
y2 D 1 x C x 2
x C   .
5
5
7.5.4. p0 .r / D .r C 1/.4r 1/; p1 .r / D 2.r C 2/; p2 .r / D 4r C 7.
2
2
1
a1 .r / D
; an .r / D
an 1 .r /
an 2 .r /, n  1.
4r C 3
4n C 4r 1
nCr C1
1
4
r1 D 1=4; a1 .1=4/ D 1=2; an .1=4/ D
an 1 .1=4/
an 2 .1=4/, n 1;
2n
4n
C5


19 2
1571 3
1
y1 D x 1=4 1
x
x C
x C .
2
104
10608
2
1
r2 D 1; a1 . 1/ D 2; an . 1/ D
an 1 . 1/
an 2 . 1/, n  1;
4n 5
n

11 2 1 3
y2 D x 1 1 C 2x
x
x C .
6
7
7.5.6. p0 .r / D r .5r 1/; p1 .r / D .r C 1/2 ; p2 .r / D 2.r C 2/.5r C 9/.
r C1
nCr
a1 .r / D
; an .r / D
an 1 .r / 2an 2 .r /, n  1.
5r C 4
5n C 5r 1
5n C 1
an 1 .1=5/ 2an 2 .1=5/, n 1;
r1 D 1=5; a1 .1=5/ D 6=25; an .1=5/ D
25n 

6
1217
41972
y1 D x 1=5 1
x
x2 C
x3 C    .
25
625
46875
n
r2 D 0; a1 .0/ D 1=4; an .0/ D
an 1 .0/ 2an 2 .0/, n  1;
5n 1
1 2 35 3 11 4
y2 D x
x
x C x C   .
4
18
12
7.5.8. p0 .r / D .3r 1/.6r C 1/; p1 .r / D .3r C 2/.6r C 1/; p2 .r / D 3r C 5.
6r C 1
6n C 6r 5
1
; an .r / D
an 1 .r /
an 2 .r /, n  1.
a1 .r / D
6r C 7
6n C 6r C 1
6n C 6r C 1
2n 1
1
r1 D 1=3; a1 .1=3/ D 1=3; an .1=3/ D
an 1 .1=3/
an 2 .1=3/, n 1;
2n
6n C 3

C1
1
2
5 3
y1 D x 1=3 1
x C x2
x C .
3
15
63
n 1
1
r2 D 1=6; a1 . 1=6/ D 0; an . 1=6/ D
an 1 . 1=6/
an 2 . 1=6/, n  1;
n
6n


1 2
1
y2 D x 1=6 1
x C x3 C    .
12
18

104 Chapter 7 Series Solutions of Linear Second Order Equations


7.5.10. p0 .r / D .2r C 1/.5r 1/; p1 .r / D .2r 1/.5r C 4/; p2 .r / D 2.2r C 5/.5r 1/.
2r 1
2n C 2r 3
10n C 10r 22
a1 .r / D
; an .r / D
an 1 .r /
an 2 .r /, n  1.
2r C 3
2n C 2r C 1
5n C 5r 1
10n 13
2n 4
r1 D 1=5; a1 .1=5/ D 3=17; an .1=5/ D
an 1 .1=5/
an 2 .1=5/, n 1;
10n C7
n

7 2
547 3
3
x
x C .
y1 D x 1=5 1 C x
17
153
5661
n 2
20n 54
r2 D 1=2; a1 . 1=2/ D 1; an . 1=2/ D
an 1 . 1=2/
an 2 . 1=2/, n  1;
10n 7

n
14
556 3
y2 D x 1=2 1 C x C x 2
x C .
13
897
1
an 1 .r /.
nCr C1
1
X
2
. 2/n
Qn
an 1 .1=2/; y1 D x 1=2
r1 D 1=2; an .1=2/ D
x n.
2n C 3
.2j
C
3/
j
D1
nD0
1
n
X
1
.
1/
r2 D 1; an . 1/ D
an 1 . 1/; y2 D x 1
xn.
n
n
nD0

7.5.14. p0 .r / D .r C 1/.2r

1/; p1 .r / D 2r C 1; an .r / D

1
an
2n C 2r 1
1
n
X . 1/
1
r1 D 1=2; an .1=2/ D
an 1 .1=2/; y1 D x 1=2
x n.
2n
2n n
nD0
1
1 X
1
. 1/n
Qn
r2 D 2; an . 2/ D
an 1 . 2/; y2 D 2
xn.
2n 5
x nD0 j D1 .2j 5/

7.5.16. p0 .r / D .r C 2/.2r

1/; p1 .r / D r C 3; an .r / D

1 .r /.

2
an
.n C r 1/.2n C 2r 1/
1
X
2
2n
Qn
xn.
r1 D 1; an .1/ D
/an 1 .1/; y1 D x
n.2n C 1
n
.2j
C
1/
j D1
nD0
1
X
2
2n
Qn
r2 D 1=2; an .1=2/ D
x n.
/an 1 .1=2/; y2 D x 1=2
n.2n 1
n
.2j
1/
j D1
nD0

7.5.18. p0 .r / D .r

1/.2r

1/; p1 .r / D 2; an .r / D

1 .r /.

nCr 4
3; an .r / D
an 1 .r /.
.n Cr 1/.3n C 3r C 1/

n 3
2
1
an 1 .1/; y1 D x 1 C x C x 2 .
r1 D; an .1/ D
n.3n C 4/
7
70
0
1
1
n
n
X
Y
3n 13
.
1/
3j
13
@
A x n.
r2 D 1=3; an . 1=3/ D
an 1 . 1=3/; y2 D x 1=3
n n
3n.3n 4/
3
3j
4
nD0

7.5.20. p0 .r / D .r

1/.3r C 1/; p1 .r / D r

j D1

nCr C1
an 1 .r /.
4n C 4r 1
1
X
nC2
. 1/n .n C 2/ n
Q
r1 D 1; an .1/ D
an 1 .1/; y1 D x
x .
4n C 3
2 njD1 .4j C 3/
nD0
1
n
X
4n C 5
. 1/n Y
r2 D 1=4; an .1=4/ D
an 1 .1=4/; y2 D x 1=4
.4j C 5/x n
n n
16n
16
nD0

7.5.22. p0 .r / D .r

1/.4r

1/; p1 .r / D r .r C 2/; an .r / D

j D1

Section 7.5 The Method of Frobenius I

105

2n C 2r C 1
an 1 .r /.
3n0C 3r 1 1
 
n
1
X
6n C 5
. 1/n 2 n Y
@
r1 D 1=3; an .1=3/ D 2
an 1 .1=3/; y1 D x 1=3
.6j C 5/A x n ;
9n
n
9
nD0
j D1
0
1
1
n
X
Y
2n 1
2j
1
A xn
an 1 . 1/; y2 D x 1
. 1/n 2n @
r2 D 1; an . 1/ D 2
3n 4
3j 4

7.5.24. p0 .r / D .r C 1/.3r

1/; p1 .r / D 2.r C 2/.2r C 3/; an .r / D 2

nD0

7.5.28. p0 .r / D .2r

j D1

.n C r /2
an 1 .r /.
.2n C 2r 1/.4n C 4r 1/


4n2 C 4n C 1
9
5 2
245 3
an 1 .1=2/; y1 D x 1=2 1
xC
x
x C .
8n.4n C 1/
128
39936
 40

16n2 C 8n C 1
25
675
38025 3
1=4
2
an 1 .1=4/; y2 D x
1
xC
x
x C .
32n.4n 1/
96
14336
5046272

1/.4r

r1 D 1=2; an .1=2/ D
r2 D 1=4; an .1=4/ D

1/; p1 .r / D .r C 1/2 ; an .r / D

.3n C 3r 2/
1/.2r C 1/; p1 .r / D .2r C 1/.3r C 1/; an .r / D
an .r /.
.2n C 2r C 1/


6n 1
5
55
935 3
r1 D 1=2; an .1=2/ D
an 1 .1=2/; y1 D x 1=2 1
x C x2
x C .
4.n C 1/
96
1536
8

1
5 2
55 3
6n 7
1=2
r2 D 1=2; an . 1=2/ D
an 1 . 1=2/; y2 D x
1C x
x
x C .
4n
4
32
384

7.5.30. p0 .r / D .2r

.n C r /.n C r C 1/
an .r /.
.2n C2r C 1/.3n C 3r C 1/

10
200 2
17600 3
.3n 1/.3n C 2/
1=3
an 1 . 1=3/; y1 D x
1
xC
x
x C .
9n.6n C 1/
63
7371
3781323


.2n 1/.2n C 1/
3
9 2
105 3
1=2
an 1 . 1=2/; y2 D x
1
xC
x
x C .
4n.6n 1/
20
352
23936

7.5.32. p0 .r / D .2r C 1/.3r C 1/; p1 .r / D .r C 1/.r C 2/; an .r / D


r1 D

1=3; an . 1=3/ D

r2 D

1=2; an . 1=2/ D

7.5.34. p0 .r / D .2r

1/.4r

r1 D 1=2; a2m .1=2/ D

8m 3
a2m
8m C 1

r2 D 1=4; a2m .1=4/ D

2m 1
a2m
2m

7.5.36. p0 .r / D r .3r

1/; p2 .r / D .r

r1 D 1=3; a2m .1=3/ D


r2 D 0; a2m .0/ D
7.5.38. p0 .r / D .2r

8m C 4r 5
.2r C 3/.4r C 3/; a2m .r / D
a2m
8m
0
1 C 4r 1
1
m
X
Y
8j 3 A 2m
1=2
@
.1=2/;
y
D
x
x .
2
1
8j
C1
mD0 j D1
0
1
1
m
X
Y
1
1=4
@ .2j 1/A x 2m
2 .1=4/; y2 D x
m m
2
mD0

1/; p2 .r / D

6m 17
a2m
18m

2m
6m
1/.3r

6
a2m
1

2 .r /.

j D1

2m C r 6
a2m
6m C 3r0 1
1
m
X
. 1/m Y
1=3
@
.6j
2 .1=3/; y1 D x
18m m
mD0

4/.r C 2/; a2m .r / D

4 2
8 4
x
2 .0/; y2 D 1 C x C
5
55

1/; p2 .r / D .r C 1/.3r C 5/; a2m .r / D

j D1

2 .r /.

17/A x 2m .

2m C r 1
a2m
4m C 2r 1

2 .r /.

106 Chapter 7 Series Solutions of Linear Second Order Equations


0
1
m
Y
1
1=2
@ .4j 1/A x 2m .
2 .1=2/; y1 D x
m m
8
mD0
0 j D1
1
1
m
X
Y
6m 2
3j
1
A x 2m .
a2m 2 .1=3/; y2 D x 1=3
2m @
r2 D 1=3; a2m .1=3/ D
12m 1
12j
1
mD0
1
X

4m 1
r1 D 1=2; a2m .1=2/ D
a2m
8m

j D1

7.5.40. p0 .r / D .2r

4m 1
a2m
4.2m C 1/

r1 D 1=2; a2m .1=2/ D


r2 D

4m 3
a2m
8m

1=2; a2m . 1=2/ D

7.5.42. p0 .r / D .r C 1/.3r
r1 D 1=3; a2m .1=3/ D
r2 D

2m C r 1
a2m 2 .r /.
4m
0 C 2r C 1 1
1
m
X
. 1/m @ Y 4j 1 A 2m
1=2
x .
2 .1=2/; y1 D x
4m
2j C 1
mD0
j D1
0
1
1
m
m
X
Y
. 1/
1=2
@ .4j 3/A x 2m
2 . 1=2/; y2 D x
m m
8
mD0

1/.2r C 1/; p1 .r / D .r C 1/.2r C 3/; a2m .r / D

1; a2m . 1/ D

2m C r 3
a2m 2 .r /.
02m C r C 11
1
m
X
Y
3j 4 A 2m
1=3
m@
.1=3/;
y
D
x
.
1/
x .
2
1
3j
C2
mD0

1/; p1 .r / D .r
3m 4
a2m
3m C 2

a2m

2.

j D1

1/.3r C 5/; a2m .r / D

j D1

1/; y2 D x

.1 C x 2 /

.2m C r 2/2
a2m 2 .r /.
.2m C r C 1/.4m 0
C 2r 1/
1
1
m
m
2
X
Y
.4m 3/2
.
1/
.4j
3/
@
A x 2m .
r1 D 1=2; a2m .1=2/ D
a2m 2 .1=2/; y1 D x 1=2
m m
8m.4m C 3/
8
4j
C
3
mD0
0 j D1
1
1
m
2
m
2
X
Y
.2m 3/
.
1/
.2j
3/
@
A x 2m .
r2 D 1; a2m . 1/ D
a2m 2 . 1/; y2 D x 1
m m
2m.4m 3/
2
4j
3
mD0

7.5.44. p0 .r / D .r C 1/.2r

1/; p1 .r / D r 2; a2m .r / D

j D1

1
a2m 2 .r /.
6m C 3r C 1
1
X
1
. 1/m
Q
r1 D 1=3; a2m .1=3/ D
a2m 2 .1=3/; y1 D x 1=3
x 2m .
m
m
2.3m C 1/
2
.3j
C
1/
j D1
mD0
1
m
X
1
.
1/
r2 D 1=3; a2m . 1=3/ D
a2m 2 . 1=3/; y2 D x 1=3
x 2m
m m
6m
6
mD0

7.5.46. p0 .r / D .3r

1/.3r C 1/; p1 .r / D 3r C 5; a2m .r / D

2m C r C 1
1/; p2 .r / D .r C 3/2 ; a2m .r / D
a2m 2 .r /.
2.8m C 4r 1/


8m C 5
13 2
273 4
2639 6
r1 D 1=4; a2m .1=4/ D
a2m 2 .1=4/; y1 D x 1=4 1
x C
x
x C .
64m
64
8192
524288

m
1 2
2 4
2 6
1
r2 D 1; a2m . 1/ D
a2m 2 . 1/; y2 D x
1
x C x
x C .
8m 5
3
33
209

7.5.48. p0 .r / D 2.r C 1/.4r

7.5.50. p0 .r / D .2r

1/.2r C 1/; p2 .r / D .2r C 5/2 ; a2m .r / D

4m C 2r C 1
a2m
4m C 2r 1

2 .r /.

Section 7.5 The Method of Frobenius I

107



2m C 1
3 2 15 4 35 6
a2m 2 .1=2/; y1 D x 1=2 1
x C x
x C .
2m
2
8
16

2m
8
16 6
1=2
2
4
r2 D 1=2; a2m . 1=2/ D
a2m 2 . 1=2/; y2 D x
1 2x C x
x C .
2m 1
3
5
P
P1
n
n
7.5.52. (a) Multiplying (A) c1 y1 C c2y2  0 by x r2 yields c1x r1 r2 1
nD0 an x C c2
nD0 bn x D 0,
0 < x < . Letting x ! 0C shows that c2 D 0, since b0 D 1. Now (A) reduces to c1y1  0, so c1 D 0.
Therefore,y1 and y2 P
are linearly independent on .0;
/.
P1
1
n
n
(b) Since y1 D
a
.r
/x
and
y
D
2
nD0 n 1
nD0 an .r2 /x are linearly independent solutions of
Ly D 0 .0; /, fy1 ; y2 g is a fundamental set of solutions of Ly D 0 on .0; /, by Theorem 5.1.6.
r1 D 1=2; a2m .1=2/ D

7.5.54. (a) If x > 0, then jxjr x n D x nCr , so the assertions are obvious. If x < 0, then jxjr D . x/r , so
r . x/r
r jxjr
d
r jxjr n
d
jxjr D r . x/r 1 D
D
. Therefore,(A)
.jxjr x n / D
x C jxjr .nx n 1 / D
dx
x
x
dx
x
d2
d
.n C r /jxjr x n 1 and
.jxjr x n / D .n C r / .jxjr x n 1 / D .n C r /.n C r 1/jxjr x n 2 , from (A)
dx 2
dx
with n replaced by n 1.
p2 .n C r 2/
an 2 .r /,
p0 .n C r /
r  0, which implies that a2mC1 .r / D 0 for m D 1; 2; 3; : : : . Therefore,Eqn. (7.5.12) actually reduces
p2 .2m C r 2/
to a0 .r / D 1, a2m .r / D
, which holds because of condition (A).
p0 .2m C r /
(b) Similar to the proof of Exercise 7.5.55(a).
(c) p0 .2m C r1/ D 2m0.2m C r1 r2 /, which is nonzero if m > 0, since r1 r2  0. Therefore,
the assumptions of Theorem 7.5.2 hold with r D r1 , and Ly1 D p0 .r1 /x r1 D 0. If r1 r2 is not an
even integer, then p0 .2m C r2/ D 2m0 .2m r1 C r2/ 0, m D 1; 2;    . Hence, the assumptions
of Theorem 7.5.2 hold with r D r2 and Ly2 D p0 .r2 /x r2 D 0. From Exercise 7.5.52, fy1 ; y2 g is a
fundamental set of solutions.
(d) Similar to the proof of Exercise 7.5.55(c).
7.5.56. (a) Here p1  0, so Eqn. (7.5.12) reduces to a0 .r / D 1, a1 .r / D 0, an .r / D

7.5.58. (a) From Exercise 7.5.57, bn D 0 for n  1.


7.5.60. (a) .0 C1 x C2 x 2/
1
X

1
X

nD0

an x n D 0 a0 C.0 a1 C1a0 /x C

1
X

nD2

.0 an C1 an

1 C2 an 2 /x

0 a0
.
2

0
1 x C 2 x
nD0
p1 .r 1/
1
p2 .r 2/
2
D
and
D
, then Eqn. (7.5.12) is equivalent to a0 .r / D 1, 0a1 .r /C
(b) If
p0 .r /
0
p0 .r /
0
1 a0 .r / D 0, 0an .r / C 1 an 1 .r / C 2 an 2 .r / D 0, n  2. Therefore,Theorem 7.5.2 implies the
conclusion.

1, so

an x n D

7.5.62. p0 .r / D .2r

1/.3r

1/; p1 .r / D 0; p2 .r / D 2.2r C 3/.3r C 5/;

p1 .r 1/
1
D0D
;
p0 .r /
0

p2 .r 2/
2
x 1=3
x 1=2
D2D
; y1 D
; y2 D
.
2
p0 .r /
0
1 C 2x
1 C 2x 2
7.5.64. p0 .r / D 5.3r

1/.3r C 1/; p1 .r / D .3r C 2/.3r C 4/; p2 .r / D 0;

p2 .r 2/
2
x 1=3
x 1=3
D0D
; y1 D
; y2 D
.
p0 .r /
0
5Cx
5Cx

p1 .r 1/
1
1
D
D
;
p0 .r /
5
0

108 Chapter 7 Series Solutions of Linear Second Order Equations


7.5.66. p0 .r / D .2r
3D

3/.2r

1/; p1 .r / D 3.2r

1/.2r C 1/; p2 .r / D .2r C 1/.2r C 3/;

p1 .r 1/
D
p0 .r /

2
x 1=2
x 3=2
1 p2 .r 2/
;
D1D
; y1 D
;
y
D
.
2
0
p0 .r /
0
1 C 3x C x 2
1 C 3x C x 2

7.5.68. p0 .r / D 3.r

1/.4r

1/; p1 .r / D 2r .4r C 3/; p2 .r / D .r C 1/.4r C 7/;

p1 .r 1/
2
1
D D
;
p0 .r /
3
0

p2 .r 2/
1
2
x
x 1=4
D D
; y1 D
; y2 D
.
2
p0 .r /
3
0
3 C 2x C x
3 C 2x C x 2
7.6 THE METHOD OF FROBENIUS II
7.6.2. p0 .r / D .r C 1/2 ; p1 .r / D .r C 2/.r C 3/; p2 .r / D .r C 3/.2r 1/;
r C3
nCr C2
2n C 2r 5
a1 .r / D
; an .r / D
an 1 .r /
an 2 .r /, n  2.
r C2
nCr C1
nCr C1
1
nCr C2 0
2n C 2r 5 0
1
a10 .r / D
; a0 .r / D
a .r /
a .r / C
an 1 .r /
.r C 2/2 n
nCr C1 n 1
nCr C1 n 2
.n C r C 1/2
7
an 2 .r /, n  2.
.n C r C 1/2
nC1
2n 7
r1 D 1; a1 . 1/ D 2; an . 1/ D
an 1 . 1/
an 2 . 1/, n  2;
n
n

9
20 3
y1 D x 1 1 2x C x 2
x C ;
2
3
nC1
2n 7
1
7
a10 . 1/ D 1; an0 . 1/ D
an 1 . 1/
an 2 . 1/ C 2 an 1 . 1/
an 2 . 1/, n  2;
n
n
n
n2
15
133 2
y2 D y1 ln x C 1
xC
x C   .
4
18
7.6.4. p0 .r / D .2r 1/2 ; p1 .r / D .2r C 1/.2r C 3/; p2 .r / D .2r C 1/.2r C 3/;
.2n C 2r C 1/an 1 .r / .2n C 2r 3/an 2 .r /
2r C 3
a1 .r / D
; an .r / D
, n  2.
2r C 1
2n C 2r 1
0
0
4
.2n C 2r C 1/an 1 .r / .2n C 2r 3/an 2 .r / 4.an 1 .r / an 2 .r //
a10 .r / D
; a0 .r / D
C
;
.2r C 1/2 n
2n C 2r 1
.2n C 2r 1/2
n  2.
.n C 1/an 1 .1=2/ C .n 1/an 2 .1=2/
; n  2;
r1 D 1=2; a1 .1=2/ D 2; an .1=2/ D
n


5
y1 D x 1=2 1 2x C x 2 2x 3 C    ;
2
.n C 1/an0 1 .1=2/ C .n 1/an0 2 .1=2/
an 1 .1=2/ an 2 .1=2/
a10 .1=2/ D 1; an0 .1=2/ D
C
,n
n
n2
2;


9
17 2
3=2
y2 D y1 ln x C x
1
x C x C .
4
6
7.6.6. p0 .r / D .3r C 1/2 ; p1 .r / D 3.3r C 4/; p2 .r / D 2.3r C 7/;
3
3an 1 .r / C 2an 2 .r /
a1 .r / D
; an .r / D
; n  2;
3r C 4
3n C 3r C 1
9
3an0 1 .r / C 2an0 2 .r /
9an 1 .r / 6an 2 .r /
a10 .r / D
; an0 .r / D
C
; n  2.
2
.3r C 4/
3n C 3r C 1
.3n C 3r C 1/2
3an 1 . 1=3/ C 2an 2 . 1=3/
r1 D 1=3; a1 . 1=3/ D 1; an . 1=3/ D
, n  2;
3n


5
1 3
y1 D x 1=3 1 x C x 2
x C ;
6
2

Section 7.6 The Method of Frobenius II


3an0 1 .r / C 2an0 2 .r /
3an
a10 . 1=3/ D 1; an0 . 1=3/ D
C
3n


11
25
x C x2 C    .
y2 D y1 ln x C x 2=3 1
12
36

1 .r /

2an
3n2

2 .r /

109

; n  2;

7.6.8. p0 .r / D .r C 2/2 ; p1 .r / D 2.r C 3/2 ; p2 .r / D 3.r C 4/;


3an 2 .r /
a1 .r / D 2; an .r / D 2an 1 .r /
; n  2;
nCr C2
0
3an 2 .r /
3an 2 .r /
C
, n  2.
a10 .r / D 0; an0 .r / D 2an0 1 .r /
nCr C2
.n C r C 2/2
3an 2 . 2/
r1 D 2; a1 . 2/ D 2; an . 2/ D 2an 1 . 2/
, n  2;
n


5
y1 D x 2 1 2x C x 2 3x 3 C    ;
2
3an 2 . 2/
3an 2 . 2/
a10 . 2/ D 0; an0 . 2/ D 2an 1 . 2/
C
; n  2;
n
n2
3 13
y2 D y1 ln x C
x C   .
4
6
7.6.10. p0 .r / D .4r C 1/2 ; p1 .r / D 4r C 5; p2 .r / D 2.4r C 9/;
1
an 1 .r / C 2an 2 .r /
a1 .r / D
; an .r / D
; n  2;
4r C 5
4n C 4r C 1
0
0
an 1 .r / C 2an 2 .r /
4an 1 .r / C 8an 2 .r /
4
a10 .r / D
; a0 .r / D
C
; n  2.
.4r C 5/2 n
4n C 4r C 1
.4n C 4r C 1/2
an 1 . 1=4/ C 2an 2 . 1=4/
r1 D 1=4; a1 . 1=4/ D 1=4; an . 1=4/ D
; n  2;
4n


1
7 2
23 3
y1 D x 1=4 1
x
x C
x C ;
4
32
384
an0 1 . 1=4/ C 2an0 2 . 1=4/ an 1 . 1=4/ C 2an 2 . 1=4/
a10 . 1=4/ D 1=4; an0 . 1=4/ D
C
;n
4n
4n2
2;


5
157 2
1
C x
x C .
y2 D y1 ln x C x 3=4
4
64
2304
7.6.12. p0 .r / D .2r

1/2 ; p1 .r / D 4;

4
an 1 .r /;
.2n C 2r 1/2
n
. 4/
an .r / D Qn
.
1/
j D1 .2j C 2r

an .r / D

By logarithmic differentiation, an0 .r / D an .r /


. 1/n
;
r1 D 1=2; an .1=2/ D
2
0 .n/
1
n
X
1
A;
an0 .1=2/ D an .1=2/ @ 2
j
1
X
. 1/n n
y1 D x 1=2
x ;
.n/2
nD0

j D1

n
X

j D1

2
2j C 2r

110 Chapter 7 Series Solutions of Linear Second Order Equations

y2 D y1 ln x

0
1
1
n
n
X
X
.
1/
1
@
A xn;
2x 1=2
2
.n/
j
nD1
j D1

2/2 ; p1 .r / D r 2 ; an .r / D

7.6.14. p0 .r / D .r

.n C r
.n C r

1/2
an
2/2

1 .r /;

an .r / D . 1/n

.n C r 1/2
;
.r 1/2

2n.r C n 1/
; r1 D 2; an .2/ D . 1/n .n C 1/2 ; an0 .2/ D . 1/nC1 2n.n C 1/;
.r 1/3
1
1
X
X
y1 D x 2
. 1/n .n C 1/2 x n ; y2 D y1 ln x 2x 2
. 1/n n.n C 1/x n .
an0 .r / D . 1/nC1
nD0

nD1

7.6.16. p0 .r / D .5r 1/2 ; p1 .r / D r C 1;


.n C r /
an 1 .r /;
an .r / D
.5n C 5r 1/2
n
Y
.j C r /
an .r / D . 1/n
;
.5j C 5r 1/2
j D1

By logarithmic differentiation,
n
X
.5j C 5r C 1/
an0 .r / D an .r /
;
.j C r /.5j C 5r 1/
j D1

n
Y
.5j C 1/
r1 D 1=5; an .1=5/ D . 1/
;
125n .n/2
n

j D1

n
X
5j C 2
;
an0 .1=5/ D an .1=5/
j.5j C 1/
j D1
Q
1
X
. 1/n njD1 .5j C 1/ n
1=5
y1 D x
x ;
125n .n/2
nD0
0
1
Qn
1
n
n
X
X
.
1/
.5j
C
1/
5j
C
2
j
D1
@
A x n.
y2 D y1 ln x x 1=5
125n.n/2
j.5j C 1/
nD1

7.6.18. p0 .r / D .3r 1/2 ; p1 .r / D .2r


.2n C 2r 3/2
an .r / D
an 1 .r /;
.3n C 3r 1/2
n
Y
.2j C 2r 3/2
an .r / D . 1/n
;
.3j C 3r 1/2
j D1

j D1

1/2 ;

By logarithmic differentiation,
n
X
1
an0 .r / D 14an .r /
;
.2j C 2r 3/.3j C 3r 1/
j D1
Q
. 1/n njD1 .6j 7/2
r1 D 1=3; an .1=3/ D
;
81n .n/2
n
X
1
an0 .1=3/ D 14an .1=3/
/;
j.6j 7/
j D1
Q
1
X
. 1/n njD1 .6j 7/2 n
1=3
y1 D x
x ;
81n .n/2
nD0

Section 7.6 The Method of Frobenius II

y2 D y1 ln x C 14x

1
X
. 1/n
1=3
nD1

Qn

j D1 .6j
81n .n/2

7/2

0
@

n
X

j D1

1
1
/A x n .
j.6j 7/

7.6.20. p0 .r / D .r C 1/2 ; p1 .r / D 2.r C 2/.2r C 3/;


n
Y
2.2n C 2r C 1/
2j C 2r C 1
an .r / D
an 1 .r /, n  1; an .r / D 2n
;
nCr C1
j Cr C1
j D1

By logarithmic differentiation,
n
X
1
;
an0 .r / D an .r /
.j C r C 1/.2j C 2r C 1/
j D1
Q
2n njD1 .2j 1/
;
r1 D 1; an . 1/ D
n
n
X
1
an0 . 1/ D an . 1/
;
j.2j 1/
j D1
Qn
1
n
1/ n
1X2
j D1 .2j
y1 D
x ;
x nD0
n
0
1
Qn
1
n
1 X 2n j D1 .2j 1/ X
1
@
A x n.
y2 D y1 ln x C
x nD1
n
j.2j 1/
j D1

7.6.22. p0 .r / D 2.r
2n C 2r
an .r / D
2.n C r
n
. 1/n Y
an .r / D
2n

j D1

2/2 ; p1 .r / D .r
1
an 1 .r /;
2/
2j C 2r 1
;
j Cr 2

1/.2r C 1/;

By logarithmic differentiation,
n
X
1
an0 .r / D 3an .r /
;
.j C r 2/.2j C 2r 1/
j D1
Q
. 1/n njD1 .2j C 3/
r1 D 2; an .2/ D
;
2n n
n
X
1
an0 .2/ D 3an .2/
;
j.2j C 3/
j D1
Q
1
X
. 1/n njD1 .2j C 3/ n
y1 D x 2
x ;
2n n
nD0
0
1
Qn
1
n
n
X
X
.
1/
.2j
C
3/
1
j D1
@
A x n.
y2 D y1 ln x 3x 2
n n
2
j.2j
C
3/
nD0
j D1

7.6.24. p0 .r / D .r 3/2 ; p1 .r / D 2.r 1/.r C 2/;


2.n C r 2/.n C r C 1/
an .r / D
an 1 .r /;
.n C r 3/2
2.n C r 2/.n C r C 1/ 0
2.5n C 5r 7/
an0 .r / D
an 1 .r /
an
2
.n C r 3/
.n C r 3/3
2.n C 1/.n C 4/
r1 D 3; an .3/ D
an 1 .3/;
n2

1 .r /;

111

112 Chapter 7 Series Solutions of Linear Second Order Equations


y1 D x 3 .1 C 20x C 180x 2 C 1120x 3 C   ;
2.n C 1/.n C 4/ 0
2.5n C 8/
an0 .3/ D
an 1 .3/
an
n2
n3

6968
y2 D y1 ln x x 4 26 C 324x C
x2 C   
3

1 .3/;

7.6.26. p0 .r / D r 2; p1 .r / D r 2 C r C 1;
.n2 C n.2r 1/ C r 2 r C 1/
an 1 .r /;
an .r / D
.n C r /2
2
2
.n C n.2r 1/ C r
r C 1/ 0
.n C r 2/
an0 .r / D
an 1 .r /
an
2
.n C r /
.n C r /3
.n2 n C 1/
r1 D 0; an .0/ D
an 1 .0/;
n2
3
7 3
x C   ;
y1 D 1 x C x 2
4
12
2
.n 2/
.n
n C 1/ 0
an0 .0/ D
an 1 .0/
an 1 .0/;
2
n
n3 
3
5
y2 D y1 ln x C x 1
x C x2 C    .
4
9
7.6.28. p0 .r / D .r

1/2 ; p2 .r / D r C 1;

. 1/m
1
a2m 2 .r /, n  1; a2m .r / D Qm
2m C r 1
j D1 .2j C r
By logarithmic differentiation,
m
X
0
.r / D a2m .r /
a2m
;

a2m .r / D

j D1

. 1/m
;
2m m
m
X1
1
a2m .1/
;
2
j

r1 D 1; a2m .1/ D
0
a2m
.1/ D

j D1

1
X
. 1/m 2m
y1 D x
x ;
2m m
mD0

y2 D y1 ln x

0
1
1
m
x X . 1/m @ X 1 A 2m
x .
2
2m m
j

7.6.30. p0 .r / D .2r

mD1

j D1

1/2 ; p2 .r / D 2r C 3;

1
a2m 2 .r /;
4m C 2r 1
m
. 1/
a2m .r / D Qm
;
1/
j D1 .4j C 2r
By logarithmic differentiation,
m
X
1
0
a2m
.r / D 2a2m .r /
;
4j C 2r 1

a2m .r / D

1 .r /;

j D1

r1 D 1=2; a2m .1=2/ D

. 1/m
;
4m m

1/

Section 7.6 The Method of Frobenius II


m

0
a2m
.1=2/ D

y1 D x 1=2

X1
1
a2m .1=2/
;
2
j
j D1

1
X
. 1/m 2m
x ;
4m m

mD0

y2 D y1 ln x

0
1
1
m
x 1=2 X . 1/m @ X 1 A 2m
x .
2 mD1 4m m
j
j D1

7.6.32. p0 .r / D .2r
a2m .r / D . 1/m

1/2 ; p2 .r / D .r C 1/.2r C 3/; a2m .r / D

m
Y

2j C r 1
;
4j C 2r 1

j D1

By logarithmic differentiation,
m
X
0
a2m
.r / D a2m .r /
.2j C r
j D1

r1 D 1=2; a2m .1=2/ D


0
a2m
.1=2/ D a2m .1=2/

y1 D x 1=2

1
X
. 1/m

mD0

y2 D y1 ln x C

1=2

. 1/m

m
X

j D1
Q
m

1
1/.4j C 2r

Qm

j D1 .4j
8m m

1
2j.4j

j D1 .4j

1/
1/

8m m

1
X
. 1/m

mD1

Qm

1/

1/

x 2m ;

j D1 .4j

1/

8m m

7.6.34. p0 .r / D .4r C 1/2 ; p2 .r / D .r


2m C r 3
a2m .r / D
a2m 2 .r /;
8m C 4r C 1
m
Y
2j C r 3
a2m .r / D . 1/m
;
8j C 4r C 1

2m C r 1
a2m
4m C 2r 1

0
@

m
X

j D1

1/.4r C 9/;

1
A x 2m .
j.4j 1/

j D1

By logarithmic differentiation,
m
X
0
a2m .r / D a2m .r /
.2j C r

13
;
3/.8j C 4r C 1/
j D1
Q
. 1/m m
13/
j D1 .8j
r1 D 1=4; a2m . 1=4/ D
;
.32/m m
m
X
13
0
a2m
. 1=4/ D a2m . 1=4/
;
2j.8j 13/
j D1
Q
1
X
. 1/m m
13/ 2m
j D1 .8j
1=4
y1 D x
x ;
.32/m m
mD0
0
1
Qm
1
m
13 1=4 X . 1/m j D1 .8j 13/ X
1
@
A x 2m .
y2 D y1 ln x C x
m m
2
.32/
j.8j
13/
mD1
j D1

7.6.36. p0 .r / D .2r

1/2 ; p2 .r / D 16r .r C 1/;

2 .r /;

113

114 Chapter 7 Series Solutions of Linear Second Order Equations


16.2m C r 2/.2m C r 1/
a2m 2 .r /;
.4m C 2r 1/2
m
Y
.2j C r 2/.2j C r 1/
;
a2m .r / D . 16/m
.4j C 2r 1/2
a2m .r / D

j D1

By logarithmic differentiation,
m
X
0
a2m
.r / D a2m .r /
.2j C r
j D1

r1 D 1=2; a2m .1=2/ D


0
a2m
.1=2/ D a2m .1=2/

y1 D x 1=2

1
X
.

mD0

. 1/m

m
X

8j C 4r 5
2/.2j C r 1/.4j C 2r

Qm

j.4j

3/.4j
j D1 .4j
m
2
4 .m/
8j 3
3/.4j

j D1
Q
1/m m
3/.4j
j D1 .4j
m
2
4 .m/

y2 D y1 ln x C x 1=2

1
X
. 1/m

mD1

1/
1/

Qm

j D1 .4j

1/

;
x 2m ;

3/.4j

4m .m/2

7.6.38. p0 .r / D .r C 1/2 ; p2 .r / D .r C 3/.2r


4m C 2r 5
a2m .r / D
a2m 2 .r /;
2m C r C 1
m
Y 4j C 2r 5
a2m .r / D . 1/m
;
2j C r C 1

1/

1/;

1/

0
@

m
X

j D1

j.4j

j D1

8j 3
3/.4j

By logarithmic differentiation,
m
X
7
0
a2m
.r / D a2m .r /
;
.2j C r C 1/.4j C 2r 5/
j D1
Q
. 1/m m
7/
j D1 .4j
r1 D 1; a2m . 1/ D
;
m
2 m
m
X
7
0
a2m
. 1/ D a2m . 1/
;
2j.4j 7/
j D1
Qm
1
1 X . 1/m j D1 .4j 7/ 2m
y1 D
x ;
x mD0
2m m
0
1
Qm
1
m
7 X . 1/m j D1 .4j 7/ X
1
@
A x 2m .
y2 D y1 ln x C
2x mD1
2m m
j.4j 7/
j D1

1/2 ; p2 .r / D r C 1;
1
a2m .r / D
a2m 2 .r /;
2m C r 1
1
1
0
a2m
.r / D
a0
.r / C
2m C r 1 2m 2
.2m C r
1
r1 D 1; a2m .1/ D
a2m 2 .1/;
2m


1 2 1 4
1 6
y1 D x 1
x C x
x C ;
2
8
48

7.6.40. p0 .r / D .r

1/2

a2m

2 .r /;

1/

A x 2m .

Section 7.6 The Method of Frobenius II

115

1
1 0
a2m 2 .1/ C
a2m 2 .1/, m  1;
2m 
4m2

3 2
11 4
3 1
y2 D y1 ln x C x
x C
x C .
4 32
576
0
a2m
.1/ D

7.6.42. p0 .r / D 2.r C 3/2 ; p2 .r / D r 2 2r C 2;


4m2 C 4m.r 3/ C r 2 6r C 10
a2m .r / D
a2m 2 .r /;
2.2m C r C 3/2
4m2 C 4m.r 3/ C r 2 6r C 10 0
12m C 6r 19
0
a2m 2 .r /
a2m
a2m
.r / D
2.2m C r C 3/2
.2m C r C 3/3
2
4m
24m C 37
r1 D 3; a2m . 3/ D
a2m 2 . 3/;
8m2


17 2
85 4
85 6
3
x C
x
x C ;
y1 D x
1
8
256
18432
2
4m
24m C 37 0
37 12m
0
a2m
. 3/ D
a2m 2 . 3/ C
a2m 2 . 3/, m  1;
2
8m3 
8m
25 471 2
1583 4
y2 D y1 ln x C x 1
x C
x C .
8
512
110592

2 .r /;

7.6.44. p0 .r / D .r C 1/2 ; p1 .r / D 2.2 r /.r C 1/; r1 D 1.


n
Y
2.n C r /.n C r 3/
.j C r /.j C r 3/
n
an .r / D
a
.r
/;
a
.r
/
D
2
, n  0. Therefore,an . 1/ D
n 1
n
.n C r C 1/2
.j C r C 1/2
j D1

0 if n  1 and y1 D 1=x. If n  4, then an .r / D .r C1/2 bn .r /, where bn0 . 1/ exists; therefore an0 . 1/ D


4.r 2/.r 1/
2.r 2/
0 if n  4. For r D 1; 2; 3, an .r / D .r C 1/cn .r /, where c1 .r / D
, c2 .r / D
,
2
.r C 2/
.r C 2/.r C 3/2
8r .r 2/.r 1/
c3 .r / D
. Hence, a10 . 1/ D c1. 1/ D 6, a20 . 1/ D c2. 1/ D 6, a30 . 1/ D
.r C 2/.r C 3/.r C 4/2
8 2
c3 . 1/ D 8=3, and y2 D y1 ln x 6 C 6x
x .
3
7.6.46. p0 .r / D .r C 1/2 ; p1 .r / D .r 1/.r C 2/; r1 D 1.
m
Y
nCr 2
j Cr 2
an .r / D
an 1 .r /; an .r / D
, n  0. Therefore,a1 . 1/ D
nCr C1
j Cr C1
and an . 1/ D 0 if n  3, so y1 D

j D1
2

.x

1/

2, a2 . 1/ D 1,

r 1 0
3
r .r 1/
6.r 2 C 2r 1/
0
0
, a1 .r / D
,
a
.
1/
D
3;
a
.r
/
D
,
a
.r
/
D
,
2
r C2
.r C 2/2 1
.r C 2/.r C 3/ 2
.r C 2/2 .r C 3/2
r .r 1/
a20 . 1/ D 3; if n  3 an .r / D .r C 1/cn .r / where cn .r / D
, so
.n C r /.n C r 1/.n C r C 1/
1
X
2
1
an0 . 1/ D cn . 1/ D
and y2 D y1 ln x C 3 3x C 2
x n.
2
n.n 2/.n 1/
n.n
1/
nD2
a1 .r / D

7.6.48. p0 .r / D .r 2/2 ; p1 .r / D .r 5/.r 1/; r1 D 2.


nCr 6
an .r / D
an 1 .r /;
nCr 2
m
Y j Cr 6
an .r / D
, n  0. Therefore,a1 .2/ D 3, a2 .2/ D 3, a3 .2/ D
j Cr 2
j D1

n  4, so y1 D x 2 .1

x/3 .

1, and an .2/ D 0 if

116 Chapter 7 Series Solutions of Linear Second Order Equations


r 5 0
4
, a1 .r / D
, a0 .2/ D 4;
r 1
.r 1/2 1
.r 5/.r 4/ 0
4.2r 2 10r C 5/ 0
a2 .r / D
, a2 .r / D
, a2 .2/ D 7;
r .r 1/
r 2 .r 1/2
4
.r 5/.r 4/.r 3/ 0
12.r
8r 3 C 16r 2 5/ 0
, a3 .r / D
, a3 .2/ D 11=3; if n  4, then
a3 .r / D
r .r 1/.r C 1/
r 2.r 1/2 .r C 1/2
.r 5/.r 4/.r 3/
an .r / D .r 2/cn .r / where cn .r / D
, so an0 .2/ D
.n C r 5/.n C r 4/.n C r 3/.n C r 2/
6
cn .2/ D
and
n.n 2/.n2 1/
!
1
X
11 2
1
3
n
y2 D y1 ln x C x 4 7x C x
6
x .
3
n.n 2/.n2 1/
a1 .r / D

nD3

7.6.50. p0 .r / D .3r 1/2 ; p2 .r / D 7 3r ; r1 D 1=3.


6m C 3r 13
a2m 2 .r /;
a2m .r / D
.6m C 3r 1/2
m
Y
6j C 3r 13
a2m .r / D
, m  0. Therefore,a2 .1=3/ D 1=6 and a2m .1=3/ D 0 if m  2, so
.6j C 3r 1/2
j D1


1 2
y1 D x 1=3 1
x .
6
3r 7
3.19 3r / 0
a2 .r / D
; a0 .r / D
; a .1=3/ D 1=4. If m  2, then a2m .r / D .r 1=3/c2m .r /
.3r C 5/2 2
.3r C 5/3 2
3.3r 7/
0
Q
where c2m .r / D
, so a2m
.1=3/ D c2m .1=3/ D
.6m C 3r 7/.6m C 3r 1/ m
1/
j D1 .6j C 3r
1
1
, and
m
1
12 6
.m 1/m m
!
1
1 X
1
7=3 1
2m
y2 D y1 ln x C x
x
.
4 12
6m m.m C 1/.m C 1/
mD1

7.6.52. p0 .r / D .2r C 1/2 ; p2 .r / D 7 2r ; r1 D 1=2.


4m C 2r 11
a2m .r / D
a2m 2 .r /;
.4m C 2r C 1/2
m
Y
4j C 2r 11
a2m .r / D
, m  0. Therefore,a2 . 1=2/ D 1=2, a4 . 1=2/ D 1=32, and
.4j C 2r C 1/2
j D1


1 2
1
a2m . 1=2/ D 0 if m  3, so y1 D x 1=2 1
x C x4 .
2
32
2r 7
2.19 2r / 0
0
a2 .r / D
, a .r / D
, a . 1=2/ D 5=8,
.2r C 5/2 2
.2r C 5/3 2
.2r 7/.2r 3/
4.8r 3 60r 2 146r C 519/ 0
0
a4 .r / D
,
a
.r
/
D
, a4 . 1=2/ D 9=128; if
4
.2r C 5/2 .2r C 9/2
.2r C 5/3 .2r C 9/3
m  3, then a2m .r / D .r C 1=2/c2m .r / where
2.2r 7/.2r 3/
0
Q
c2m .r / D
, so a2m
. 1=2/ D c2m . 1=2/ D
.4m C 2r 7/.4m C 2r 3/.4m C 2r C 1/ m
.4j
C
2r
C
1/
j D1
1
, and
m
4 .m 2/.m 1/m m

Section 7.6 The Method of Frobenius II

y2 D y1 ln x C x

3=2

5
8

1
9 2 X
x C
128
4mC1 .m
mD2

n
. 1/n Y
p1 .j C r1
0n .n/2
j D1

j D1

n
. 1/n Y p1 .j C r 1/
. Therefore, an .r1 / D
0n
.j C r r1 /2
j D1

1/. Theorem 7.6.2 implies Ly1 D 0.

(b) From (A), ln jan .r /j D


n  0
X
p1 .j C r
an .r /
p1 .j C r

!
1
2m
x
.
1/m.m C 1/.m C 1/

r1 /2 , then (A) an .r / D

7.6.54. (a) If p0 .r / D 0 .r

117

1/
1/

n ln j0 j C

2
j Cr

r1

n
X

j D1

and

.ln jp1 .j C r

an0 .r1 /

1/j

2 ln jj C r

n  0
X
p1 .j C r1
D an .r1 /
p1 .j C r1
j D1

r1 j/, so an0 .r / D
1/
1/


2
. Theoj

rem 7.6.2 implies that Ly2 D 0.


(c) Since p1 .r / D 1 , y1 and y2 reduce to the stated forms. If 1 D 0, then y1 D x r1 and y2 D
r1
x ln x, which are solutions of the Euler equation 0 x 2 y 00 C 0 xy 0 C 0 y.
Q
7.6.54. (a) Ly1 D p0 .r1 /x r1 D 0. Now use the fact that p0 .j C r1 / D 0 j 2 , so njD1 p0 .j C r1 / D
0n .n/2 .
1
X
(b) From Theorem 7.6.2, y2 D y1 ln x C x r1
an0 .r1 /x n is a second solution of Ly D 0. Since
nD1

n
n
X
. 1/n Y p1 .j C r 1/
an .r / D
,
(A)
ln
ja
.r
/j
D
n
ln
j
j
C
ln jp1 .j C r
n
0
n
0
.j C r r1 /2
j D1

j D1

1/j

n
X

j D1

ln jj C

r r1 j, provided that p1 .j Cr 1/ and j Cr r1 are nonzero for all positive integers j . Differentiating (A)
n
n
X
X
p10 .j C r1 1/
1
a0 .r1 /
and then setting r D r1 yields n
D
2
, which implies the conclusion.
an .r1 /
p2 .j C r1 1/
j
j D1
j D1
 
n
X
1
. 1/n 1 n
0
and Jn D 2
. If
(c) In this case p1 .r / D 1 and p1 .r / D 0, so an .r1 / D
.n/2 0
j
j D1

1 D 0, then y1 D x r1 and y2 D x r1 ln x, while the differential equation is an Euler equation with


indicial polynomial 0 .r r12 /. See Theorem 7.4.3.
7.6.56. p0 .r / D r 2; p1 .r / D 1; r1 D 0. a2m .r / D

a2m 1 .r /
. 1/m
Q
,
m

1;
a
.r
/
D
,
2m
m
2
.2m C r /2
j D1 .2j C r /

1
X
. 1/m
. 1/m 2m
,
so
y
D
x .
1
m
2
4 .m/
4m .m/2
mD0
m
X
1
0
0
By logarithmic differentiation, a2m
.r / D 2a2m .r /
, so a2m
.0/ D
2m C r
j D1
0
1
1
m
X
X
. 1/m
1
@
A x 2m .
y2 D y1 ln x
m .m/2
4
j
mD1

m  0. Therefore,a2m .0/ D

a2m .0/

m
X
1
and
j

j D1

j D1

7.6.58. p0 .r / D .2r 1/2 ; p1 .r / D .2r C1/2; p2 .r / D 0;


y1 D

x 1=2
x 1=2 ln x
; y2 D
.
1Cx
1Cx

p1 .r 1/
1 p2 .r 2/
2
D1D
;
D0D
;
p0 .r /
0
p0 .r /
0

118 Chapter 7 Series Solutions of Linear Second Order Equations


1/2 ; p1 .r / D 0; p2 .r / D

7.6.60. p0 .r / D 2.r
1=2 D

2
x
x ln x
; y1 D
; y2 D
.
2
0
2 x
2 x2
1/2 ; p1 .r / D 3r 2; p2 .r / D 0;

7.6.62. p0 .r / D 4.r
y1 D

.r C 1/2 ;

x
x ln x
; y2 D
.
4 C 3x
4 C 3x

p1 .r 1/
1 p2 .r 2/
2
D 3=4 D
;
D0D
;
p0 .r /
0
p0 .r /
0

1/2 ; p1 .r / D 2r 2; p2 .r / D .r C 1/2;

7.6.64. p0 .r / D .r

p1 .r 1/
1 p2 .r 2/
D 0 D
;
D
p0 .r /
0
p0 .r /

p1 .r 1/
1 p2 .r 2/
D 2D
;
D1D
p0 .r /
0
p0 .r /

2
x
x ln x
; y1 D
; y2 D
.
2
0
.1 x/
.1 x/2
7.6.66. See the proofs of Theorems 7.6.1 and 7.6.2.
7.7 THE METHOD OF FROBENIUS III
7.7.2. p0 .r / D r .r

1/; p1 .r / D 1; r1 D 1; r2 D 0; k D r1
1
an .r / D
an 1 .r /;
.n C r /.n C r 1/
. 1/n
an .r / D Qn
;
1/
j D1 .j C r /.j C r
. 1/n
an .1/ D
;
n.n C 1/
1
X . 1/n
y1 D x
x n;
n.n
C
1/
nD0
D 1; C D p1 .0/a0 .0/ D 1.
By logarithmic differentiation,
n
X
2n C 2r 1
an0 .r / D an .r /
;
.n C r /.n C r 1/
an0 .1/ D an .1/
y2 D 1

j D1
n
X

j D1

y1 ln x C x

r2 D 1;

2j C 1
;
j.j C 1/
1
X

nD1

0
1
n
. 1/n @ X 2j C 1 A n
x .
n.n C 1/
j.j C 1/
j D1

7.7.4. p0 .r / D r .r 1/; p1 .r / D r C1; r1 D 1; r2 D 0; k D r1 r2 D 1; an .r / D


. 1/n
j D1 .j C r

Qn

1/

; an .1/ D

1
X
. 1/n
. 1/n n
; y1 D x
x D xe
n
n
nD0

By logarithmic differentiation, an0 .r / D an .r /

n
X

j D1

1
j Cr

an 1 .r /
; an .r / D
nCr 1

; D 1; C D

p1 .0/a0 .0/ D

n
X
1
; an0 .1/ D an .1/
; y2 D 1
1
j
j D1

1.

0
1
1
n
X
. 1/n X 1
@
A
y1 ln x C x
n
j
nD1

7.7.6. p0 .r / D .r 1/.r C 2/; p1 .r / D r C 3; r1 D 1; r2 D 2; k D r1 r2 D 3. an .r / D


1
X
1
. 1/n
. 1/n
. 1/n n
an 1 .r /; an .r / D Qn
; an .1/ D
; y1 D x
x D xe x ;
nCr 1
n
n
.j
C
r
1/
j D1
nD0

j D1

Section 7.7 The Method of Frobenius III




1
1
p1 .0/
a2 . 2/ D 1=2.
1 C x C x2 ; C D
2
2
3

n
n
X
X
1
1
an .r /
; an0 .1/ D an .1/
; y2 D x 2 1 C
j Cr 1
j

D x

j D1

j D1

119

By logarithmic differentiation, an0 .r / D


0
0
1 1

1
n
X
1
1 2
1
. 1/n X 1
@y1 ln x x
@
A x n A;
xC x
2
2
2
n
j
nD1
j D1

7.7.8. p0 .r / D .r C 2/.r C 7/; p1 .r / D 1; r1 D 2; r2 D 7; k D r1 r2 D 5; an .r / D


n
Y
. 1/n
an 1 .r /
. 1/n
; an .r / D Qn
; an . 2/ D 120
;
.n C r C 2/.n C r C 7/
n.n C 5/
j D1 .j C r C 2/.j C r C 7/
j D1

1
120 X . 1/n
x n;
x2
n.n C 5/
nD0


1
1
1 3
1 4
p1 . 3/
7
x C
x ;C D
a4 .
Dx
1 C x C x2 C
4
24
144
576
5
n
X
2j C 2r C 9
mic differentiation, an0 .r / D an .r /
; a0 .
.j C r C 2/.j C r C 7/ n
j D1
0


1
1
1
1
1
@y1 ln x
y2 D x 7 1 C x C x 2 C
x3 C
x4
4
24
144
576
2880

y1 D

7/ D

1=2880. By logarith-

n
X
2j C 5
;
j.j C 5/
j D1
0
1 1
1
n
120 X . 1/n @ X 2j C 5 A n A
x .
x2
n.n C 5/
j.j C 5/

2/ D an . 2/

nD1

j D1

7.7.10. p0 .r / D r .r 4/; p1 .r / D .r 6/.r 5/; r1 D 4; r2 D 0; k D r1 r2 D 4; an .r / D


n
Y
.j C r 7/.j C r 6/
.n C r 7/.n C r 6/
an 1 .r /; an .r / D . 1/n
. Setting r D 4 yields
.n C r /.n C r 4/
.j C r /.j C r 4/
j D1


2
p1 .3/
.r 6/.r 5/
4
y1 D x 1
x . D 1C10xC50x 2C200x 3; C D
a3 .0/ D 300. a1 .r / D
;
5
4
.r 3/.r C 1/
2
2
3.3r
22r C 31/ 0
.r 6/.r 5/
a10 .r / D
; a1 .4/ D 27=25. a2 .r / D .r 4/c2 .r /, with c2 .r / D
,
2
2
.r 3/ .r C 1/
.r 3/.r 2/.r C 1/.r C 2/
2
0
0
so a20 .4/ D c2 .4/ D 1=30. If n  3, then
 an .r / D .r 4/ bn .r /where bn .4/ exists, so an .4/ D 0 and
1 6
27
y2 D 1 C 10x C 50x 2 C 200x 3 300 y1 ln x C x 5
x .
25
30
7.7.12. p0 .r / D .r 2/.r C 2/; p1 .r / D 2r 1; r1 D 2; r2 D 2; k D r1 r2 D 4; an .r / D
n
Q

Y
2j C 2r 1
2n C 2r 1
n
2j C3
1
an 1 .r /; an .r / D
; an .2/ D n
j D1 j C4 ;
.j C r 2/.j C r C 2/
.n C r 2/.n C r C 2/
j D1
0
1


1
n
X
1 Y 2j C 3
1 2
1 3
p1 .1/
2
n
2
@
A
y1 D x
x ; D x
1CxC x
x ;C D
a3 . 2/ D 1=16.
n
j C4
4
12
4
nD0
j D1

By logarithmic differentiation,
n
n
X
X
j 2 C j.2r 1/ C r 2 r C 4
.j 2 C 3j C 6/
an0 .r / D 2an .r /
; an0 .2/ D 2an .2/
;
.j C r 2/.j C r C 2/.2j C 2r 1/
j.j C 4/.2j C 3/
j D1
j D1
0
10
1


1
n
n
2 X
2
Y
X
1
1
1
x
1
2j
C
3
.j
C
3j
C
6/
@
A@
A x n.
y2 D x 2 1 C x C x 2
x3
y1 ln x C
4
12
16
8 nD1 n
j C4
j.j C 4/.2j C 3/
j D1

j D1

7.7.14. p0 .r / D .r C 1/.r C 7/; p1 .r / D .r C 5/.2r C 1/; r1 D 1; r2 D 7; k D r1


n
Y
.n C r C 4/.2n C 2r 1/
.j C r C 4/.2j C 2r 1/
r2 D 6; an .r / D
an 1 .r /; an .r / D . 1/n
;
.n C r C 1/.n C r C 7/
.j C r C 1/.j C r C 7/
j D1

120 Chapter 7 Series Solutions of Linear Second Order Equations


0
n
. 1/n Y .j C 3/.2j
@
an . 1/ D
n
j C6
C D

0
1
n
1 X . 1/n Y .j C 3/.2j
A; y1 D
@
x nD0 n
j C6
j D1
j D1


p1 . 2/
26
143 2
a5 . 7/ D 0; y2 D x 7 1 C x C
x .
6
5
20
3/

3/

A x n; D x



26
143 2
1C xC
x ;
5
20

7.7.16. p0 .r / D .3r
.n C r
an .r / D
.3n C 3r

10/.3r C 2/; p1 .r / D r .3r 4/; r1 D 10=3; r2 D 2=3; k D r1 r2 D 4;


n
Y
.j C r 1/.3j C 3r 7/
1/.3n C 3r 7/
an 1 .r /; an .r / D . 1/n
;
10/.3n C 3r C 2/
.3j C 3r 10/.3j C 3r C 2/
j D1
0
1
0
1
n
1
n
n
X
Y
. 1/n .n C 1/ Y 3j C 7
.
1/
.n
C
1/
3j
C
7
@
A; y1 D x 10=3
@
A x n; D
an .10=3/ D
n
9n
j C4
9
j
C
4
nD0
j D1

 j D1


4
1
p
.7=3/
4
1 2
1
2=3
2
2=3
x
1C x
x ;C D
a3 . 2=3/ D 0; y2 D x
1C x
x .
27
243
36
27
243
7.7.18. p0 .r / D .r
3/.r C 2/; p1 .r / D .r C 1/2 ; r1 D 3; r2 D 2; k D r1 r2 D 5;
n
Y
.n C r /2
.j C r /2
an 1 .r /; an .r / D . 1/n
; an .3/ D
an .r / D
.n C r 3/.n C r C 2/
.j C r 3/.j C r C 2/
j D1
0
1


n
1
Q

2
n
Y
X
n
1
p1 .2/
2
.
1/
.j
C
3/
n
. 1/
.j C3/
3
n
2
@
A
; y1 D x
x ; D x
1C x ;C D
a4 . 2/ D
j D1 j C5
n
n
j C5
4
5
nD0
j D1


1
2
0; y2 D x
1C x .
4
7.7.20. p0 .r / D .r 6/.r 1/; p1 .r / D .r  8/.r 4/; r1 D
D
 6; r2 D 1; k D r1 r2 D 5; an .r / 
.n C r 9/.n C r 5/
2
1 2
21
21 2 35 3
6
an 1 .r /; y1 D x 1 C x C x ; D x 1 C x C x C x ;
.n C r 6/.n C r 1/
3
7 
4
2
4

21 2 35 3
p1 .5/
21
C D
a5 .1/ D 0; y2 D x 1 C x C x C x .
6
4
2
4
7.7.22. p0 .r / D r .r 10/; p1 .r / D 2.r 6/.r C 1/; r1 D 10; r2 D 0; k D r1 r2 D 10; an .r / D
.n C r 9/.n C r 8/.n C r 7/
. 1/n 2n .n C 1/.n C 2/.n C 3/
2.n C r 7/
an 1 .r /; an .r / D . 2/n
; an .10/ D
;
n C r 10
.r 9/.r 8/.r 7/
6


1
x 10 X
4
5
40 3 40 4 32 5 16 6
y1 D
. 1/n 2n .n C 1/.n C 2/.n C 3/x n ; D 1
x C x2
x C x
x C x ;
6 nD0
3
3
21
21
21
21


p1 .9/
4
5
40 3 40 4 32 5 16 6
C D
a9 .0/ D 0; y2 D 1
x C x2
x C x
x C x .
10
3
3
21
21
21
21
Pk 1
r2
2m
Note: in the solutions to Exercises 7.7.237.7.40, D x
.
mD0 a2m .r2 /x

7.7.24. p0 .r / D .r 6/.r 2/; p2 .r / D r ; r1 D 6; r2 D 2; k D .r1 r2/=2 D 2; a2m .r / D


1
X
a2m 2 .r /
. 1/m
. 1/m
. 1/m 2m
2
; a2m .r / D Qm
x D x 6 e x =2 ;
; a2m .6/ D m ; y1 D x 6
m
2m C r 6
2 m
2 m
6/
j D1 .2j C r
mD0


m
X
1 2
p2 .4/
1
2
0
D x 1C x ;C D
a2 .2/ D 1=2. By logarithmic differentiation, a2m .r / D a2m .r /
2
4
2j C r
j D1
0
1


m
1
m
6 X
m
X
X
1
1
1
x
. 1/
1
0
@
A x 2m .
a2m
.6/ D a2m .6/
; y2 D x 2 1 C x 2
y1 ln x C
2j
2
2
4 mD1 2m m
j
j D1

j D1

Section 7.7 The Method of Frobenius III

121

7.7.26. p0 .r / D .r 1/.r C 1/; p2 .r / D 2r C 10; r1 D 1; r2 D 1; k D .r1 r2 /=2 D 1;


m
Y
2.2m C r C 3/
2j C r C 3
a2m .r / D
a2m 2 .r /; a2m .r / D . 2/m
;
.2m C r 1/.2m C r C 1/
.2j C r 1/.2j C r C 1/
1
. 1/ .m C 2/
x X . 1/m .m C 2/ 2m
; y1 D
x ;Dx
a2m .1/ D
2 m
2 mD0
m
By logarithmic differentiation,
m
X
.4j 2 C 4j.r C 3/ C r 2 C 6r C 1/
0
;
a2m
.r / D a2m .r /
.2j C r 1/.2j C r C 1/.2j C r C 3/

j D1

0
a2m
.1/ D a2m .1/

y2 D x

j D1
m
X

j D1

;C D

p2 . 1/
a0 . 1/ D
2

4.

j 2 C 4j C 2
;
2j.j C 1/.j C 2/

0
1
m
1
X
. 1/m .m C 2/ @ X j 2 C 4j C 2 A 2m
4y1 ln x C x
x .
m
j.j C 1/.j C 2/
mD1

j D1

7.7.28. p0 .r / D .2r C 1/.2r C 5/; p2 .r / D 2r C 3; r1 D 1=2; r2 D 5=2; k D .r1 r2 /=2 D 1;


m
Y
.4m C 2r 1/
.4j C 2r 1/
a2m 2 .r /; a2m .r / D . 1/m
;
a2m .r / D
.4m C 2r C 1/.4m C 2r C 5/
.4j C 2r C 1/.4j C 2r C 5/
j D1
Q
Q
1
X
. 1/m m
1/
. 1/m m
1/ 2m
j D1 .2j
j D1 .2j
1=2
a2m . 1=2/ D
; y1 D x
x ; D x 5=2 ; C D
m
m
8 m.m C 1/
8
m.m
C
1/
mD0
p2 . 5=2/
a0 . 5=2/ D 1=4. By logarithmic differentiation,
8
m
m
X
X
.16j 2 C 8j.2r 1/ C 4r 2 4r 11/
2j 2 2j 1
0
0
a2m
.r / D 2a2m .r /
; a2m
. 1=2/ D a2m . 1=2/
;
.4j C 2r 1/.4j C 2r C 1/.4j C 2r C 5/
2j.j C 1/.2j 1/
j D1
j D1
0
1
Qm
1
m
m
2
X
X
.
1/
.2j
1/
1
2j
1
2j
j D1
@
A x 2m .
y2 D x 5=2 C y1 ln x x 1=2
mC1 m.m C 1/
4
8
j.j
C
1/.2j
1/
mD1
j D1

7.7.30. p0 .r / D .r 2/.r C 2/; p2 .r / D 2.r C 4/; r1 D 2; r2 D 2; k D .r1 r2 /=2 D 2; a2m .r / D


1
X
2
2m
1 2m
1
2
a2m 2 .r /; a2m .r / D Qm
; a2m .2/ D
; y1 D x 2
x D x2ex ;
2m C r 2
m
m
2/
j D1 .2j C r
mD0
m
X
p2 .0/
1
0
D x 2 .1 x 2 /; C D
a2 . 2/ D 2. By logarithmic differentiation, a2m
.r / D a2m .r /
4
2j C r
j D1
0
1
m
1
m
X
X
1
1 X1
0
@
A x 2m .
a2m
.2/ D a2m .2/
; y2 D x 2 .1 x 2 / 2y1 ln x C x 2
2j
m
j
mD1
j D1

j D1

7.7.32. p0 .r / D .3r 13/.3r 1/; p2 .r / D 2.5 3r /; r1 D 13=3; r2 D 1=3; k D .r1 r2 /=2 D 2;


m
Y
2.6m C 3r 11/
.6j C 3r 11/
a2m .r / D
a2m 2 .r /; a2m .r / D 2m
;
.6m C 3r 13/.6m C 3r 1/
.6j C 3r 13/.6j C 3r 1/
j D1
Qm


1 Qm
X
2 2
j D1 .3j C 1/
j D1 .3j C 1/ 2m
13=3
1=3
a2m .13=3/ D m
; y1 D 2x
x ; D x
1C x ; C D
9 m.m C 2/
9m m.m C 2/
9
mD0
m
X
p2 .7=3/
.12j 2 C 4j.3r 11/ C 3r 2 22r C 47/
0
a2 .1=3/ D 2=81. By logarithmic differentiation, a2m
.r / D 9a2m .r /
;
36
.6j C 3r 13/.6j C 3r 11/.6j C 3r 1/
j D1

122 Chapter 7 Series Solutions of Linear Second Order Equations


m
X

3j 2 C 2j C 2
;
2j.j C 2/.3j C 1/
j D1
0
0
1
1


1 Qm
m
2
X
X
.3j
C
1/
2
2
3j
C
2j
C
2
j D1
@y1 ln x x 13=3
@
A x 2m A.
y2 D x 1=3 1 C x 2 C
m m.m C 2/
9
81
9
j.j
C
2/.3j
C
1/
mD0

0
a2m
.13=3/ D a2m .13=3/

j D1

7.7.34. p0 .r / D .r 2/.r C 2/; p2 .r / D 3.r 4/; r1D 2; r2 D 2; k D .r1 r2 /=2


 D 2;
3.2m C r 6/
1 2
9 2
2
2
a2m .r / D
a2m 2 .r /; y1 D x 1
x ; D x
1C x ; C D
.2m C r 2/.2m C r C 2/
2
2
p2 .0/
3.r 4/ 0
3.r 2 8r 16/ 0
a2 . 2/ D 27=2; a2 .r / D
, a .r / D
, a2 .2/ D 7=12. If m  2,
4
r .r C 4/ 2
r 2 .r C 4/2
m
3 .r 4/
Q
, so
then a2m .r / D .r 2/c2m .r / where c2m .r / D
.2m C r 4/.2m C r 2/ m
j D1 .2j C r C 2/



1
3 m
X
9
27
7
2
0
a2m
.2/ D c2m .2/ D
; y2 D x 2 1 C x 2
y1 ln x C x 4 x 2
m.m 1/.m C 2/
2
2
12
m.m
mD2


3 m
2

1/.m C 2/

2m

7.7.36. p0 .r / D .2r 5/.2r C 7/; p2 .r / D .2r 1/2 ; r1 D 5=2; r2 D 7=2; k D .r1 r2 /=2 D 3;
.2r 1/.2r C 3/.2r C 7/
4m C 2r 5
a2m .r / D
a2m 2 .r /; a2m .r / D
; a2m .5=2/ D
4m C 2r C 7
.4m C 2r 1/.4m C 2r C 3/.4m C 2r C 7/
1
X
. 1/m
. 1/m
; y1 D x 5=2
x 2m ; D x 7=2 .1 C x 2 /2 C D
.m C 1/.m C 2/.m C 3/
.m
C
1/.m
C
2/.m
C
3/
mD0
p2 .1=2/
7=2
a4 . 7=2/ D 0; y2 D x
.1 C x 2 /2 .
24
7.7.38. p0 .r / D .r 3/.r C 7/; p2 .r / D r .r C 1/; r1 D 3; r2 D 7; k D .r1 r2 /=2 D 5; a2m .r / D
m
Y
.2j C r 2/.2j C r 1/
.2m C r 2/.2m C r 1/
a2m 2 .r /; a2m .r / D . 1/m
; a2m .3/ D
.2m C r 3/.2m C r C 7/
.2j C r 3/.2j C r C 7/
j D1
0
1
0
1


m
1
m
Y
X
Y
mC1
2j C 1
mC1
2j C 1
21
35
35
A; y1 D x 3
A x 2m ; D x 7 1 C x 2 C x 4 C x 6
. 1/m m @
. 1/m m @
2
j C5
2
j C5
8
16
64
mD0
j D1
j D1


p2 .1/
21
35
35
C D
a8 . 7/ D 0; y2 D x 7 1 C x 2 C x 4 C x 6 .
10
8
16
64
7.7.40. p0 .r / D .2r

3/.2r C 5/; p2 .r / D .2r 1/.2r C 1/; r1 D 3=2; r2 D 5=2; k D


m
Y
4m C 2r 5
4j C 2r 5
.r1 r2 /=2 D 2; a2m .r / D
a2m 2 .r /; a2m .r / D . 1/m
; a2m .3=2/ D
4m C 2r C 5
4j C 2r C 5
j D1
Q
Q


1
X
. 1/m m
1/
. 1/m m
1/ 2m
3 2
p2 . 1=2/
j D1 .2j
j D1 .2j
3=2
5=2
; y1 D x
x ; D x
1C x C D
a2 . 5=2/ D
m
1
m
1
2
.m C 2/
2
.m C 2/
2
16
mD0


3
0; y2 D x 5=2 1 C x 2 .
2
7.7.42. p0 .r / D r 2  2 ; p2 .r / D 1; r1 D ; r2 D ; k D .r1 r2 /=2 D ; a2m .r / D
a2m 2 .r /
. 1/m
. 1/m
; a2m .r / D Qm
; a2m ./ D m Qm
;
.2m C r C /.2m C r /
/
4 m j D1 .j C /
j D1 .2j C r C /.2j C r
1
X
. 1/m
. 1/m

a2m . / D m Qm
, j D 0; : : : ;  1; y1 D x
Q
x 2m ; D
m
m m
4 m j D1 .j /
4
.j
C
/
j
D1
mD0

Section 7.7 The Method of Frobenius III


 1
X

. 1/m
Qm
j D1 .j

123

p2 . 2/
a2 2 . /
2
a2 2 . / D
D
. By
 .
2
2
4
1/
mD0
m
m
X
X
2j C 
2j C 
0
0
logarithmic differentiation, a2m
.r / D 2a2m .r /
; a2m
./ D a2m ./
;
.2j C r C /.2j C r /
2j.j C /
j D1
j D1
0
0
1
1
m
 1
1
m
 X
m
X
X
2
2j
C

.
1/
x
.
1/
@
@y1 ln x
A x 2m A.
Qm
x 2m
Qm
y2 D x 
 .
m m
m m
4
1/
2
j.j
C
/
4
.j
/
4
.j
C
/
j D1
j D1
mD0
mD1
x

4m m

/

x 2m ; C D

j D1

7.7.44. Since an .r2 / D

p1 .n C r2 1/
an
p0 .n C r2 /

p1 .r1 1/
ak
k0

But C D
and only if

k
Y

j D1

p1 .r2 C j

1 .r2 /

1 .r2 /, 1  n  k 1, ak

p1 .r2 C k
k0

1/

ak

1 .r2 /

k
1 .r2 / D . 1/

D . /k

1/ D 0.

Qk

kY
1

j D1

p1 .r2 C j 1/
.
p0 .r2 C j /

p1 .r2 C j 1/
D 0 if
Qk 1
k0 j D1 p0 .r2 C j /
j D1


1 n
Qn
7.7.46. Since p1 .r / D 1 , an .r / D
an 1 .r / and (A) an .r / D . 1/
0 .n C r
r2 /
0
j D1 .j C r




. 1/n 1 n
1
1
. 1/n 1 n
Therefore, an .r1 / D
Qn
for n  0 (so Ly1 D 0) and an .r2 / D
Qn
n
0
n
0
k/
j D1 .j C k/
j D1 .j
!
k 1
1
X
X
for n D 0; : : : ; k 1. Ly2 D 0 if y2 D x r2
an .r2 /x n C C y1 ln x C x r1
an0 .r1 /x n if C D
1
r1 /.n C r

nD0

nD1

 
 k
1
1 . 1/k 1 1 k 1 . 1/k 1
1
1
ak 1 .r2 / D
D
. From (A), ln jan .r /j D
k0
k0 .k 1/ 0
.k 1/
k.k 1/ 0
n

n 
X
1 X
1
1
n ln
..ln jj C r r1 j C ln jj C r r2 j/, so an0 .r / D an .r /
C
0
j C r r1
j C r r2
j D1

j D1

n
X
2j C k
and an0 .r1 / D an .r1 /
.
j.j C k/
j D1



@y
7.7.48. (a) From Exercise 7.6.66(a) of Section 7.6, L
.x; r / D p00 .r /x r Cx r p0 .r / ln x. Setting r D
@r
!
1
X
r1 yields L y1 ln x C x r1
an0 .r1 / D p00 .r1 /x r1 . Since p00 .r / D 0.2r r1 r2 /, p00 .r1 / D k0 .
nD1
!
1
1
X
X
r2
n
(b) From Exercise 7.5.57 of Section 7.5, L x
an .r2 /x
D x r2
bn x n , where b0 D p0 .r2 / D
nD0

0 and bn D
while bk D

n
X

j D0
k
X

j D0

pj .n C r2

pk .k C r2

j /an
j /ak

j .r2 /

nD0

if n  1. From the definition of fan .r2 /g, bn D 0 if n k,

j .r2 / D

k
X

j D1

pj .r1

j /ak

j .r2 /.

(d) Let faQ n .r2 /g be the coefficients that would obtained if aQ k .r2 / D 0. Then an .r2 / D aQ n .r2 / if n D

1
r1 /.j C

124 Chapter 7 Series Solutions of Linear Second Order Equations

0; : : : ; k

1, and (A) an .r2 /

aQn .r2 / D

n
Xk
1
pj .n C r2
p0 .n C r2 /
j D0

j /.an

j .r2 /

aQ n

j r2 /

if n > k.

Now let cm D akCm .r2 / aQ kCm .r2 /. Setting n D m C k in (A) and recalling the k C r2 D r1 yields (B)
m
X
1
cm D
pj .m C r1 j /cm j . Since ck D ak .r2 /, (B) implies that cm D ak .r2 /am .r1 /
p0 .m C r1 /
j D0

for all m  0, which implies the conclusion.

CHAPTER 8
Laplace Transforms

8.1 INTRODUCTION TO THE LAPLACE TRANSFORM





1
1
1
s2 C 2
1 t
t
8.1.2. (a) cosh t sin t D
e sin t C e sin t $
C
D
.
2
2
.s C 1/2 C 1
.s 1/2 C 1.s C 1/2 C 1

 2 .s 1/ C 1
1 cos 2t
1 1
s
2
(b) sin2 t D
$
D
.
2
2
2 
s .s C 4/
s.s C 4/
2
s
s2 C 8
1 1
C 2
D
.
(c) cos2 2t D
2 s
s C 16
s.s 2 C 16/


.e t C e t /2 /
.e 2t C 2 C e 2t /
1
1
2
1
s2 2
2
(d) cosh t D
D
$
C C
D
.
4
4 s 2
s
sC2
s.s 2 4/
 4
2t
2t
1
1
1
te
te
4s
$
D 2
.
(e) t sinh 2t D
2
2
2
2 .s 2/
.s C 2/
.s
4/2
sin 2t
1
(f) sin t cos t D
$ 2
.
2
s C4
1 sC1
(g) sin.t C =4/ D sin t cos.=4/ C cos t cos.=4/ $ p 2
.
2 s C1
s
s
5s
(h) cos 2t cos 3t $ 2
D 2
.
2
s C4 s C9
.s C 4/.s 2 C 9/
2
s
s 3 C 2s 2 C 4s C 32
(i) sin 2t C cos 4t $ 2
C 2
D
.
s C4
s C 16
.s 2 C 4/.s 2 C 16/
Z 1
Z 1
Z 1
st
st
0
8.1.6. If F .s/ D
e f .t/ dt, then F .s/ D
. te /f .t/ dt D
e st .tf .t// dt. Apply0

ing this argument repeatedly yields the assertion.

8.1.8. Let f .t/ D 1 and F .s/ D 1=s. From Exercise 8.1.6, t n $ . 1/n F .n/ .s/ D n=s nC1 .
st
.s s0 /t
8.1.10. If jf .t/j  Me s0 t for t  t0 , then
for t  t0 . Let g.t/ D e st f .t/,
R 1 jf .t/e j  Me
.s s0 /t
w.t/ D Me
, and  D t0 . Since t0 w.t/ dt converges if s > s0, F .s/ is defined for s > s0 .

8.1.12.
Z
1 T
e
s 0

0
st

st

Z

f ./ d 
0

dt D

st

T
Z

f ./ d  C
s
0

e
0

st

f .t/ dt D

sT

f ./ d C
0

f .t/ dt. Since f is of exponential order s0 , the second integral on the right converges to
125

126 Chapter 8 Laplace Transforms


Z T
1
sT
L.f / as T ! 1 (Exercise 8.1.10). Now it suffices to show that (A) lim e
f ./ d  D 0
T !1
s
0
s0 t
that jf .t/j  Me
if t  t0 and jf .t/j  K if 0  t  t0 , and let T > t0 . Then
ifZ s > s0 . Suppose

Z t0
Z T
Z T
T

Me s0 T

f ./ d  
f ./ d  C
f ./ d  < Kt0 C M
e s0  d  < Kt0 C
, which

t0

s0
t0
0
proves (A).
Z T
Z T
8.1.14. (a) If T > 0, then
e st f .t/ dt D
e .s s0 /t .e s0 t f .t// dt. Use integration by parts with
0
0
Z T
.s s0 /t
s0 t
.s s0 /t
uDe
, dv D e
f .t/ dt, du D .s s0 /e
, and v D g t obtain
e st f .t/ dt D
0
T
Z T
Z T

e .s s0 /t g.t/ C .s s0 /
e .s s0 /t g.t/ dt. Since g.0/ D 0 this reduces to
e st f .t/ dt D
0
0
Z T0
.s s0 /T
.s s0 /t
e
g.T / C .s s0 /
e
g.t/ dt. Since jg.t/j  M for all t  0, we can let t ! 1 to
0
Z 1
Z 1
conclude that
e st f .t/ dt D .s s0 /
e .s s0 /t g.t/ dt if s > s0 .
0

(b) If F .s0 / exists, then g.t/ is bounded on 0; 1/. Now apply (a).
Z t

1 d
j sin.e t / sin.1/j
2
(c) Since f .t/ D
sin.e t /,
f ./ d  D
 1 for all t  0. Now apply (a)
2 dt
2
0
with s0 D 0.

Z 1
Z
1 1 x
. C 1/
x e x 1
8.1.16. (a) ./ D
x 1 e x dx D
C
x e dx D
.

0
0
0
Z 1
(b) Use induction. .1/ D
e x dx D 1. If (A) .n C 1/ D n, then .n C 2/ D .n C 1/.n C 1/
0

(from (a)) D .n CZ1/n (from (A)) D .n C 1/.


1

(c) . C 1/ D

xe

. C 1/
.

dt. Let x D st. Then . C 1/ D

1
0

.st/ e

st

s dt, so

1
0

st

t dt D

  s

e s .s C 1/
e .s 1/
e 2s
1
C
C
D
s2
s2
s2
s2
0
0
1
2e s
e 2s
1
.1 e s /2
.1 e s /2
1 e s
1
s
C
C
D
.
Therefore,F
.s/
D
D
D 2 tanh .
s2
s2
s2
s2
s 2 .1 e 2s /
s 2.1 C e s /
s
2
Z 1
Z 1=2
Z 1
1
e s=2
e s
e s=2
2e s=2
(b)
e st f .t/ dt D
e st dt
e st dt D
C
D
C
s
s
s
s
s
0
0
1=2
e s
1
.1 e s=2 /2
.1 e s=2 /2
1 e s=2
1
s
C D
. Therefore,F .s/ D
D
D tanh .
s
s=2
s Zs
s
s.1 e /
s
4
s.1 C e
/
Z 

1 C e s
1 C e s
1
s
st
st
(c)
e f .t/ dt D
e sin t dt D 2
. Therefore,F .s/ D 2
coth
.
s / s 2 C 1
.s
C
1/
.s
C
1/.1
e
2
0Z
Z0 2

1 C e s
1 C e s
.
Therefore,F
.s/
D
D
(d)
e st f .t/ dt D
e st sin t dt D
.s 2 C 1/
.s 2 C 1/1 C e 2s
0
0
1
.
2
.s C 1/.1 e s /

8.1.18. (a)

st

f .t/ dt D

st

t dtC

st

.2 t/ dt D

Section 8.2 The Inverse Laplace Transform

127

8.2 THE INVERSE LAPLACE TRANSFORM




2s C 3
2.s 7/ C 17
2
17
2
17 3
17 3
7t
2
8.2.2. (a)
D
D
C
D
C
$
e
t
C
t
.
.s 7/4
.s 7/4
.s 7/3 .s 7/4
.s 7/3 6 .s 7/4
6
s2 1
.s 2/ C 22 1
.s 2/2 C 4.s 2/ C 3
1
4
3
D
D
D
C
C
D
(b)
6
6
6
4
5
.s 2/
.s 2/
.s 2/
.s 2/6
 .s 2/
 .s 2/
1 3
1 4
1
5
1 3 1 4
1
C
C
$
t C t C t 5 e 2t .
4
5
6
6 .s 2/
6 .s 2/
40 .s 2/
6
6
40 

sC5
.s C 3/
2
3
2
3t
(c) 2
D
C
$
e
cos
3t
C
sin
3t
.
s C 6s C 18
.s C 3/2 C 9 3 .s C 3/2 C 9
3
2s C 1
s
1 3
1
(d) 2
D2 2
C
$ 2 cos 3t C sin 3t.
s C9
s C9
3 s2 C 9
3
s
.s C 1/ 1
1
1
D
D
$ .1 t/e t .
(e) 2
s C 2s C 1
.s C 1/2
s C 1 .s C 1/2
sC1
s
1 3
1
(f) 2
D 2
C
$ cosh 3t C sinh 3t.
2
s
9
s
9 3s
9
3
(g) Expand the numerator in powers of s C 1: s 3 C 2s 2 s 3 D .s C 1/ 13 C 2.s C 1/ 12
s 3 C 2s 2 s 3
1
1
.s C 1/ 1 3 D .s C 1/3 .s C 1/2 2.s C 1/ 1; therefore
D
4
.s
C
1/
s
C
1
.s
C
1/2


1
6
2
1
$ 1 t t2
t3 e t.
.s C 1/3 6 .s C 1/4
6


.s 1/
5
2
5
2s C 3
t
(h)
D
2
C
$
e
2
cos
2t
C
sin
2t
.
.s 1/2 C 4
.s 1/2 C 4
2 .s 1/2 C 4
2
1
s
(i)
$ 1 cos t.
2
s
s C1
3s C 4
3s
4
3s C 4
3s C 4
(j) 2
D 2
C 2
$ 3 cosh t C 4 sinh t. Alternatively, 2
D
D
s
1
s
1
s
1
s
1
.s
1/.s C 1/


t
t
7
1
7e
e
1
$
.
2 s 1 sC1
2
3
4s C 1
1
s
1 3
1
(k)
C 2
D3
C4 2
C
$ 3e t C 4 cos 3t C sin 3t.
2
s 1
s C9
s 1
s C9
3s C9
3
3
2s C 6
1
s
2
2t
(l)
D3
2 2
3 2
$ 3te
2 cos 2t 3 sin 2t.
.s C 2/2 s 2 C 4
.s C 2/2
s C4
s C4
8.2.4. (a)
.s 2
where

A
B
Cs C D
2 C 3s
D
C
C 2
;
C 1/.s C 2/.s C 1/
sC2
s C1
s C1

A.s 2 C 1/.s C 1/ C B.s 2 C 1/.s C 2/ C .C s C D/.s C 2/.s C 1/ D 2 C 3s:


5A
2B
A C 2B C 2D
ACB CC

Solving this system yields A D


.s 2

4
,B D
5

2 C 3s
C 1/.s C 2/.s C 1/

D
D
D
D

4
1
2
0

1
,C D
2
D
$

.set s D 2/I
.set s D 1/I
.set s D 0/I
.equate coefficients of s 3 /:
3
11
,D D
. Therefore,
10
10

4 1
1 1
1 3s 11
5 s C 2 2 s C 1 10 s 2 C 1
4 2t 1 t
3
11
e
e
cos t C
sin t:
5
2
10
10

128 Chapter 8 Laplace Transforms


(b)
3s 2 C 2s C 1
As C B
C.s C 1/ C D
D 2
C
;
2
2
.s C 1/.s C 2s C 2/
s C1
.s C 1/2 C 1

where

.As C B/..s C 1/2 C 1/ C .C.s C 1/ C D/.s 2 C 1/ D 3s 2 C 2s C 1:


2B C C C D
A C B C 2D
2B C C C D
ACC

D
D
D
D

1
2
1
0

.set s D 0/I
.set s D 1/I
.set s D 0/I
.equate coefficients of s 3/:

Solving this system yields A D 6=5, B D 2=5, C D 6=5, D D 7=5. Therefore,




3s 2 C 2s C 1
1 6s C 2 6.s C 1/ 7
D
.s 2 C 1/.s 2 C 2s C 2/
5 s 2 C 1 .s C 1/2 C 1
2
6 t
7
6
cos t C sin t
e cos t C e
$
5
5
5
5

sin t:

(c) s 2 C 2s C 5 D .s C 1/2 C 4;
.s

3s C 2
A
B.s C 1/ C C
D
C
;
2
2/..s C 1/ C 4/
s 2
.s C 1/2 C 4

where

A .s C 1/2 / C 4 C .B.s C 1/ C C / .s
13A D
4A 3C D
ACB D

Solving this system yields A D

8
,B D
13

3s C 2
2/..s C 1/2 C 4/

.s

8
1
0

2/ D 3s C 2:

.set s D 2/I
.set s D 1/I
.equate coefficients of s 2 /:

8
15
,C D
. Therefore,
13
13


1
8
8.s 1/ 15
D
13 s 2
.s C 1/2 C 4
8 2t
8 t
15
$
e
e cos 2t C e
13
13
26

sin 2t:

(d)
.s
where

3s 2 C 2s C 1
A
B
C
D
D
C
C
C
;
2
2
1/ .s C 2/.s C 3/
s 1
.s 1/
sC2 sC3

1/ C B/.s C 2/.s C 3/ C .C.s C 3/ C D.s C 2//.s

.A.s

12B
9C
16D
ACC CD

D
D
D
D

3s 2 C 2s C 1
1/2 .s C 2/.s C 3/

6
9
22
0

1/2 D 3s 2 C 2s C 1:

.set s D 1/I
.set s D 2/I
.set s D 3/I
.equate coefficients of s 3 /:

Solving this system yields A D 3=8, B D 1=2, C D 1, D D 11=8. Therefore,


.s

3 1
1
1
1
C
C
2
8s 1
2 .s 1/
sC2
3 t 1 t
11
e C te C e 2t
e 3t :
8
2
8

11 1
8 s C3

Section 8.2 The Inverse Laplace Transform

129

(e)
2s 2 C s C 3
A
B
C
D
D
C
C
C
;
2
2
2
.s 1/ .s C 2/
s 1
.s 1/
sC2
.s C 2/2

where

1/ C B/.s C 2/2 C .C.s C 2/ C D/.s

.A.s

9B
9D
4A C 4B C 2C C D
ACC

D
D
D
D

1/2 D 2s 2 C s C 3:

.set s D 1/I
.set s D 2/I
.set s D 0/I
.equate coefficients of s 3 /:

6
9
3
0

Solving this system yields A D 1=9, B D 2=3, C D 1=9, D D 1. Therefore,


2s 2 C s C 3
.s 1/2 .s C 2/2

D
$

(f)

1 1
2
1
1 1
1
C
C
9s 1
3 .s 1/2 9 s C 2
.s C 2/2
1 t 2 t 1 2t
e C te
e
C te 2t :
9
3
9

3s C 2
A
B
Cs C D
D
C
C 2
;
.s 2 C 1/.s 1/2
s 1
.s 1/2
s C1

where
A.s

1/.s 2 C 1/ C B.s 2 C 1/ C .C s C D/.s

Setting s D 1 yields 2B D 5, so B D
A.s

1/2 D 3s C 2:

.A/

5
. Substituting this into (A) shows that
2

1/.s 2 C 1/ C .C s C D/.s

1/2

D 3s C 2
5s 2

5 2
.s C 1/
2
6s C 1
.s
D
2

Therefore,
A.s 2 C 1/ C .C s C D/.s
2A D
A D D
ACC D

5s
2

1/

2 .set s D 1/I
1=2 .set s D 0/I
0 .equate coefficients of s 2 /:
3
. Therefore,
2

Solving this system yields A D 1, C D 1, D D


3s C 2
.s 2 C 1/.s 1/2

1/ D

1/.5s
2

D
$

5
1
s 3=2
C 2
2 .s 1/2
s C1
5
3
e t C te t C cos t
sin t:
2
2
s

8.2.6. (a)
.s 2
where
.A.s

A.s 1/ C B
C.s C 1/ C D
17s 15
D
C
2s C 5/.s 2 C 2s C 10/
.s 1/2 C 4
.s C 1/2 C 9

1/ C B/..s C 1/2 C 9/ C .C.s C 1/ C D/..s

1/2 C 4/ D 17s

15:

130 Chapter 8 Laplace Transforms


13B C 8C C 4D
18A C 9B C 8D
10A C 10B C 5C C 5D
ACC

D
D
D
D

2
32
15
0

Solving this system yields A D 1, B D 2, C D 1, D D


.s 2

17s 15
2s C 5/.s 2 C 2s C 10/

D
$

.set s D 1/I
.set s D 1/I
.set s D 0/I
.equate coefficients of s 3/:
4. Therefore,

.s 1/ C 2
.s 1/2 C 4

.s C 1/ C 4
.s C 1/2 C 9


4
e t .cos 2t C sin 2t/ e t cos 3t C sin 3t :
3

(b)
.s 2
where
.A.s

8s C 56
A.s 3/ C B
C.s C 1/ C D
D
C
2
2
6s C 13/.s C 2s C 5/
.s 3/ C 4
.s C 1/2 C 4

3/ C B/..s C 1/2 C 4/ C .C.s C 1/ C D/..s

20B C 16C C 4D
16A C 4B C 20D
15A C 5B C 13C C 13D
ACC

3/2 C 4/ D 8s C 56:

D 80 .set s D 3/I
D 48 .set s D 1/I
D 56 .set s D 0/I
D
0 .equate coefficients of s 3 /:

Solving this system yields A D 1, B D 3, C D 1, D D 1. Therefore,


.s 2

8s C 56
6s C 13/.s 2 C 2s C 5/

D
$

.s 3/ C 3
.s C 1/ C 1
C
2
.s 3/ C 4
.s C 1/2 C 4


3
e 3t
cos 2t C sin 2t C e
2



1
cos 2t C sin 2t :
2

(c)
s C9
.s 2 C 4s C 5/.s 2

where

.A.s C 2/ C B/..s

4s C 13/

A.s C 2/ C B
C.s 2/ C D
C
.s C 2/2 C 1
.s 2/2 C 9

2/2 C 9/ C .C.s

25B 4C C D
36A C 9B C 17D
26A C 13B 10C C 5D
ACC

2/ C D/..s C 2/2 C 1/ D s C 9:

D
7 .set s D 2/I
D 11 .set s D 2/I
D
9 .set s D 0/I
D
0 .equate coefficients of s 3 /:

Solving this system yields A D 1=8, B D 1=4, C D 1=8, D D 1=4.



sC9
1 .s C 2/ C 2
.s
D
D
2
2
2
.s C 4s C 5/.s
4s C 13/
8 .s C 2/ C 1 .s


1
1
2t
$ e
cos t C sin t
8
4

Therefore,

2/ 2
2/2 C 3

1
2t
e
cos 3t
8


1
sin 3t :
12

(d)
.s 2
where
.A.s

3s 2
A.s 2/ C B
C.s 3/ C D
D
C
4s C 5/.s 2 6s C 13/
.s 2/2 C 1
.s 3/2 C 4

2/ C B/..s

3/2 C 4/ C .C.s

3/ C D/..s

2/2 C 1/ D 3s

2:

Section 8.2 The Inverse Laplace Transform


5B C C D
4A C 4B C 2D
26A C 13B 15C C 5D
ACC

D
D
D
D

.set s D 2/I
.set s D 3/I
.set s D 0/I
.equate coefficients of s 3 /:

4
7
2
0

Solving this system yields A D 1, B D 1=2, C D 1, D D 1=2. Therefore,




3s 2
1 2.s 2/ C 1 2.s 3/ 1
D D
.s 2 4s C 5/.s 2 6s C 13/
2 .s 2/2 C 1 .s 3/2 C 4




1
1
$ e 2t cos t C sin t
e 3t cos 2t
sin 2t :
2
4
(e)
.s 2
where

3s 1
A.s 1/ C B
C.s C 1/ C D
D
C
2
2
2s C 2/.s C 2s C 5/
.s 1/ C 1
.s C 1/2 C 4

1/ C B/..s C 1/2 C 4/ C .C.s C 1/ C D/..s

.A.s

8B C 2C C D
8A C 4B C 5D
5A C 5B C 2C C 2D
A C 5B C C

D
D
D
D

1/2 C 1/ D 3s

1:

.set s D 1/I
.set s D 1/I
.set s D 0/I
.equate coefficients of s 3/:

2
4
1
0

Solving this system yields A D 1=5, B D 2=5, C D 1=5, D D 4=5. Therefore,




1 .s 1/ C 2
.s C 1/ C 4
3s 1
D
:
.s 2 2s C 2/.s 2 C 2s C 5/
5 .s 1/2 C 1 .s C 1/2 C 4




2
1
2
1
$ et
cos t C sin t
e t
cos 2t C sin 2t :
5
5
5
5
(f)
.4s 2
where
.A.s

20s C 40
A.s 1=2/ C B
C.s C 1=2/ C D
D
C
2
2
4s C 5/.4s C 4s C 5/
.s 1=2/ C 1
.s C 1=2/2 C 1

1=2/ C B/..s C 1=2/2 C 1/ C .C.s C 1=2/ C D/..s


2B C C C D
A C B C 2D
5A C 10B C 5C C 10D
ACC

D
D
D
D

25=8
15=8
20
0

1=2/2 C 1/ D

5s C 10
:
4

.set s D 1=2/I
.set s D 1=2/I
.set s D 0/I
.equate coefficients of s 3 /:

Solving this system yields A D 1, B D 9=8, C D 1, D D 1=8. Therefore,




20s C 40
1
8.s 1=2/ C 9
8.s C 1=2/
D
C
.4s 2 4s C 5/.4s 2 C 4s C 5/
8 .s 1=2/2 C 1
.s C 1=2/2 C 1




9
1
$ e t =2
cos t C sin t C e t =2 cos t
sin t :
8
8
8.2.8. (a)
.s 2

2s C 1
C 1/.s 1/.s

3/

A
s

B
s

Cs C D
s2 C 1

131

132 Chapter 8 Laplace Transforms


where
1//.s 2 C 1/ C .C s C D/.s

3/ C B.s

.A.s

D
D
D
D

4A
20B
3A B C 3D
ACB CC

1/.s

3/ D 2s C 1:

.set s D 1/I
.set s D 3/I
.set s D 0/I
.equate coefficients of s 3/:

3
7
1
0

Solving this system yields A D 3=4, B D 7=20, C D 2=5, D D 3=10. Therefore,


.s 2

2s C 1
C 1/.s 1/.s

3/

3 1
7 1
2 s
3
1
C
C
2
2
4 s 1 20 s 3
5 s C 1 10 s C 1
3 t
7 3t
2
3
e C e C cos t
sin t:
4
20
5
10

D
$

(b)

where

sC2
.s 2 C 2s C 2/.s 2
.A.s C 1/ C B.s

1/

A
s

B
C.s C 1/ C D
C
sC1
.s C 1/2 C 1

1//..s C 1/2 C 1/ C .C.s C 1/ C D/.s 2

10A
2B
2B C D
ACB CC

2A

D
D
D
D

1/ D s C 2:

.set s D 1/I
.set s D 1/I
.set s D 0/I
.equate coefficients of s 3 /:

3
1
2
0

Solving this system yields A D 3=10, B D 1=2, C D 1=5, D D 3=5. Therefore,


sC2
.s 2 C 2s C 2/.s 2

1/

D
$

3 1
1 1
1
sC1
C
10 s 1 2 s C 1
5 .s C 1/2 C 1
1
3 t 1 t
e
e C e t cos te t sin t:
10
2
5

3
1
5 .s C 1/2 C 1

(c)
.s 2
where

2s 1
2s C 2/.s C 1/.s

.A.s C 1/ C B.s

2A

2//..s

2/

A
s

1/2 C 1/ C .C.s

6A
15B
4B C 2C 2D
ACB CC

D
D
D
D

3
3
1
0

B
C.s 1/ C D
C
sC1
.s 1/2 C 1

1/ C D/.s

2/.s C 1/ D 2s

1:

.set s D 2/I
.set s D 1/I
.set s D 0/I
.equate coefficients of s 3 /:

Solving this system yields A D 1=2, B D 1=5, C D 7=10, D D 1=10. Therefore,


.s 2

2s 1
2s C 2/.s C 1/.s

2/

D
$

1 1
1 1
7
s 1
C
2s 2
5 s C 1 10 .s 1/2 C 1
1 2t
1
7 t
1 t
e C e t
e cos t
e sin t:
2
5
10
10

(d)
.s 2

s 6
A
B
Cs C D
D
C
C 2
2
1/.s C 4/
s 1
sC1
s C4

1
10 .s

1
1/2 C 1

Section 8.2 The Inverse Laplace Transform


where
1//.s 2 C 4/ C .C s C D/.s 2

.A.s C 1/ C B.s

D
D
D
D

10A
10B
4A 4B D
ACB CC

5
7
6
0

1/ D s

6:

.set s D 1/I
.set s D 1/I
.set s D 0/I
.equate coefficients of s 3 /:

Solving this system yields A D 1=2, B D 7=10, C D 1=5, D D 6=5. Therefore,


.s 2

s 6
1/.s 2 C 4/

1 1
7 1
1 s
3
1
C
C C 2
2s 1
10 s C 1 5 s 2 C 4
5
s C4
1 t
7
1
3
e C e t
cos 2t C sin 2t:
2
10
5
5

(e)
s.s
where
.A.s

A
B
C.s 1/ C D
2s 3
D C
C
2
2/.s
2s C 5/
s
s 2
.s 1/2 C 4
1/2 C 4/ C .C.s

2/ C Bs/..s

10A
10B
4A C 4B D
ACB CC

D
D
D
D

3
1
1
0

1/ C D/s.s

2/ D 2s

3:

.set s D 0/I
.set s D 2/I
.set s D 1/I
.equate coefficients of s 3 /:

Solving this system yields A D 3=10, B D 1=10, C D 2=5, D D 1=5. Therefore,


2s 3
2/.s 2 2s C 5/

s.s

3
1 1
2
s 1
1
C
C
2
10s
10 s 2 5 .s 1/ C 4
5 .s
3
1 2t 2 t
1 t
C e
e cos 2t C e sin 2t:
10 10
5
10

1
1/2 C 4

(f)
.s 2
where
.A.s

5s 15
4s C 13/.s 2/.s

2/ C B.s

26A

10A
9B
13B 4C C 2D
ACB CC

5s 15
4s C 13/.s 2/.s

1/

A
s

2/2 C 9/ C .C.s

1//..s

Solving this system yields A D 1, B D


.s 2

1/

D
D
D
D

5=9, C D

et

1
s

1
5 2t
e
9

10
5
15
0

B
s

C.s 2/ C D
.s 2/2 C 9

2/ C D/.s

1/.s

2/ D 5s

15:

.set s D 1/I
.set s D 2/I
.set s D 0/I
.equate coefficients of s 3/:

4=9, D D 1. Therefore,
5 1
4
9 s 2 9 .s
4 2t
e cos 3t C
9

2
C
C9
.s

2/2

1 2t
e sin 3t:
3

1
2/2 C 9

133

134 Chapter 8 Laplace Transforms


8.2.10. (a) Let i D 1. (The proof for i D 2; : : : ; n) is similar. Multiplying the given equation through by
s s1 yields


P .s/
A2
An
D A1 C .s s1/
CC
;
.s s2/    .s sn /
s s2
s sn
and setting s D s1 yields A1 D

.s1

P .s1 /
s2 /    .s2

sn /

.
P .s/
D
s1 /Q1 .s/
P .s/
s1 shows that
D AC
Q1 .s/

(b) From calculus we know that F has a partial fraction expansion of the form

.s

A
C G.s/ where G is continuous at s1 . Multiplying through by s
s1
P .s1 /
.s s1 /G.s/. Now set s D s1 to obtain A D
.
Q.s1 /
(c) The result in (b) is generalization of the result in (a), since it shows that if s1 is a simple zero of the
denominator of the rational function, then Heavisides method can be used to determine the coefficient of
1=.s s1/ in the partial fraction expansion even if some of the other zeros of the denominator are repeated
or complex.
s

8.3 SOLUTION OF INITIAL VALUE PROBLEMS


8.3.2.
.s 2
Since .s 2

6/ D .s

2 C s.s 1/
D
s.s 3/.s C 2/

1
8
4
C e 3t C e
3
15
5

2t

.s 2
4 D .s

2 C s.s
s

1/

1
8 1
4 1
C
C
3s
15 s 3
5sC2

4/Y .s/ D

2
s

C . 1 C s/ D

2 C .s

1/.s
3

3/

2/.s C 2/,
Y .s/ D

and y D

1D

8.3.4.

Since s 2

2
Cs
s

3/.s C 2/,
Y .s/ D

and y D

6/Y .s/ D

1 2t 17
e C e
4
20

2t

2 C .s 1/.s 3/
D
.s 2/.s C 2/.s 3/

1 1
17 1
2 1
C
C
4s 2
20 s C 2
5s 3

2
C e 3t .
5

8.3.6.
.s 2 C 3s C 2/Y .s/ D

6
s

C . 1 C s/ C 3 D

6 C .s

1/.s C 2/
:
s 1

Since s C 3s C 2 D .s C 2/.s C 1/,


Y .s/ D
and y D e t C 2e

2t

2e t .

6 C .s 1/.s C 2/
1
2
D
C
.s 1/.s C 2/.s C 1/
s 1
sC2

2
sC1

Section 8.3 Solution of Initial Value Problems


8.3.8.
.s 2
Since s 2

3s C 2/Y .s/ D

3s C 2 D .s

1/.s

Y .s/ D
and y D 4e t

2
s

2 C .s 3/.s 4/
4
D
.s 1/.s 2/.s 3/
s 1

3s C 2/Y .s/ D

3s C 2 D .s

1/.s

Y .s/ D
5 t
e
2

1
s

C. 4

2/Y .s/ D

2/ D .s C 2/.s

2t

4
C .3 C 2s/ C 2 D
s

4 C s.5 C 2s/
2
D
s.s C 2/.s 1/
s

.s

1
s

3/.s C 1/
:
3

4
s

1 1
2s 3

.s 2
6 D .s

1
1
C
;
sC2
s 1

6/Y .s/ D

2
Cs
s

1D

2 C s.s
s

1/

3/.s C 2/,
2 C s.s 1/
D
s.s 3/.s C 2/

Y .s/ D
8
4
1
C e 3t C e
3
15
5

2t

1
8 1
4 1
C
C
3s
15 s 3
5sC2

8.3.16.
.s 2
1 D .s

4 C s.5 C 2s/
:
s

C et .

8.3.14.

Since s 2

4
s

1/,

Y .s/ D

and y D

4/

1
4e 2t C e 3t .
2

Since .s 2 C s

s/ C 3 D

5 1
1 .s 3/.s C 1/
D
.s 1/.s 2/.s 3/
2s 1

.s 2 C s

Since s 2

3/.s
s 3

2/,

8.3.12.

and y D 2

2 C .s

4e 2t C e 3t .
.s 2

and y D

3D

2/,

8.3.10.

Since s 2

C . 1 C s/

1/Y .s/ D

1
1 C s2
Cs D
:
s
s

1/.s C 1/,
Y .s/ D

and y D 1 C e t C e t .

1 C s2
D
s.s 1/.s C 1/

1
1
1
C
C
s
s 1
sC1

135

136 Chapter 8 Laplace Transforms


8.3.18.
.s 2 C s/Y .s/ D

2
s

C .4

.s 3/2
:
s 3

1D

s/

Since s C s D s.s C 1/,


Y .s/ D
and y D

7
3

7
e
2

2 .s 3/2
7
D
s.s C 1/.s 3/
3s

7 1
1 1
C
2sC1
6s 3

1
C e 3t .
6

8.3.20.
.s 2 C 1/Y .s/ D

1
C 2;
s2

so Y .s/ D

Substituting x D s 2 into
1
1
D
.x C 1/x
xC1
so Y .s/ D

1
x

.s 2 C 5s C 6/Y .s/ D
Since s 2 C 5s C 6 D .s C 2/.s C 3/,
Y .s/ D
t

C 4e

2t

.s 2

4e

3t

s2

1
;
C1

2s

3 D .s

3/Y .s/ D

2s

2
2 C .s C 1/.s C 8/
C .3 C s/ C 5 D
:
sC1
sC1

3/.s C 1/,

10s
C .7 C 2s/
C1

s2

4D

10s
C .2s C 3/:
C1

s2

10s
2s C 3
C
:
3/.s C 1/.s 2 C 1/
.s 3/.s C 1/

.s

2s C 3
9 1
D
.s 3/.s C 1/
4s 3
.s

4
sC3

Y .s/ D

where

1
1
D 2
2
C 1/s
s

1
4
2 C .s C 1/.s C 8/
D
C
.s C 1/.s C 2/.s C 3/
sC1
sC2

8.3.24.

Since s 2

.s 2

1
2
C 2
:
2
C 1/s
s C1

1
1
C 2
and y D t C sin t.
s2
s C1

8.3.22.

and y D e

yields

.s 2

1 1
9
$ e 3t
4sC1
4

1 t
e :
4

10s
A
B
Cs C D
D
C
C 2
3/.s C 1/.s 2 C 1/
s 3
sC1
s C1

.A.s C 1/ C B.s

3//.s 2 C 1/ C .C s C D/.s

40A
8B
A 3B 3D
ACB CC

D
D
D
D

30
10
0
0

3/.s C 1/ D 10s:

.set s D 3/I
.set s D 1/I
.set s D 0/I
.equate coefficients of s 3 /:

.A/
.B/

Section 8.3 Solution of Initial Value Problems


Solving this system yields A D 3=4, B D 5=4, C D 2, D D
.s

10s
3/.s C 1/.s 2 C 1/

D
$

From this, (A), and (B), y D


8.3.26.

137

1. Therefore,

3 1
5 1
2s C 1
C
4s 3
4 s C 1 s2 C 1
3 3t 5 t
e C e
2 cos t sin t:
4
4

2 cos t C 3e 3t C e t .

sin t

16
9s
C 2
C s; so
C4
s C1
16
9s
s
C 2
C 2
:
Y .s/ D 2
.s C 4/2
.s C 4/.s 2 C 1/
s C4
.s 2 C 4/Y .s/ D

s2

From the table of Laplace transforms,


t cos 2t

$
D

s2 4
s2 C 4
D
.s 2 C 4/2
.s 2 C 4/2
1
8
:
2
2
s C 4 .s C 4/2

.s 2

8
C 4/2

Therefore,
8
1
D 2
.s 2 C 4/2
s C4

Substituting x D s 2 into

and multiplying by s yields

L.t cos 2t/; so

16
$ sin 2t
.s 2 C 4/2

3
9
D
.x C 4/.x C 1/
xC1

9s
3s
D 2
.s 2 C 4/.s 2 C 1/
s C1
Finally,

2t cos 2t:

3
xC4

3s
$ 3 cos t
s2 C 4

3 cos 2t:

s
$ cos 2t:
C4
.2t C 2/ cos 2t C sin 2t C 3 cos t.
s2

Adding (A), (B), and (C) yields y D


28.

.s 2 C 2s C 2/Y .s/ D

.A/

.B/

.C/

2
C . 7 C 2s/ C 4:
s2

Since .s 2 C 2s C 2/ D .s C 1/2 C 1,
Y .s/ D

2
2s 3
C
:
s 2 ..s C 1/2 C 1/
.s C 1/2 C 1

2s 3
2.s C 1/ 5
D
$ e t .2 cos t
2
.s C 1/ C 1
.s C 1/2 C 1

5 sin t/:

A
B
C.s C 1/ C D
2
D C 2C
;
s 2..s C 1/2 C 1/
s
s
.s C 1/2 C 1

.A/

.B/

138 Chapter 8 Laplace Transforms



where .As C B/ .s C 1/2 C 1 C s 2 .C.s C 1/ C D/ D 2.
D
D
D
D

2B
ACB CD
ACC
2A C B C C C D

.set s D 0/I
.set s D 1/I
.equate coefficients of s 3 /I
.equate coefficients of s 2 /:

2
2
0
0

Solving this system yields A D 1, B D 1, C D 1, D D 0. Therefore,


s 2..s

1
1
.s C 1/
C 2C
$
s
s
.s C 1/2 C 1

2
D
C 1/2 C 1/

From this, (A), and (B), y D


8.3.30. .s 2 C 4s C 5/Y .s/ D

1 C t C e t .cos t

1Ct Ce

cos t:

5 sin t/.

.s C 1/ C 3
C 4. Since .s 2 C 4s C 5/ D .s C 2/2 C 1,
.s C 1/2 C 1

Y .s/ D

..s C

1/2

4
sC4
C
:
2
C 1/..s C 2/ C 1/
.s C 2/2 C 1

4
$ 4e
.s C 2/2 C 1

2t

.A/

sin t:

.B/

sC4
A.s C 1/ C B
C.s C 2/ C D
D
C
;
2
2
..s C
C 1/..s C 2/ C 1/
.s C 1/ C 1
.s C 2/2 C 1


where .A.s C 1/ C B/ .s C 2/2 C 1 C .C.s C 2/ C D/ .s C 1/2 C 1 D 4 C s.
1/2

5A C 5B C 4C C 2D
2B C C C D
A C B C 2D
ACC

D
D
D
D

.set s D 0/I
.set s D 1/I
.set s D 2/I
.equate coefficients of s 3/:

4
3
2
0

Solving this system yields A D 1, B D 1, C D 1, D D 0. Therefore,


..s C

1/2

s C4
.s C 1/ C 1
s C2
D
C
; $ e t . cos t C sin t/ C e
2
2
C 1/..s C 2/ C 1/
.s C 1/ C 1
.s C 2/2 C 1

From this, (A), and (B), y D e t . cos t C sin t/ C e


8.3.32.
.2s 2
Since 2s

3s

3s

2 D .s

Y .s/ D
and y D

1 2t
e
5

2/Y .s/ D

.cos t C 4 sin t/.

C 2. 2 C s/

3D

4 C .2s 7/.s
s 1

1/

2/.2s C 1/,

4 C .2s
2.s 2/.s

4 t
32
e C e
3
15

2t

t =2

7/.s 1/
1 1
D
1/.s C 1=2/
5s 2

4 1
32
1
C
3s 1
15 s C 1=2

8.3.34.
.2s 2 C 2s C 1/Y .s/ D

2
2
C 2. 1 C s/ C 2 D 2 C 2s:
s2
s

2t

cos t:

Section 8.3 Solution of Initial Value Problems

139

Since 2s 2 C 2s C 1 D 2..s C 1=2/2 C 1=4/,


Y .s/ D

s 2..s

1
s
C
:
2
C 1=2/ C 1=4/
..s C 1=2/2 C 1=4/

s
$e
..s C 1=2/2 C 1=4/

where

t =2

.cos.t=2/

sin.t=2//:

.A/
.B/

A
B
C.s C 1=2/ C D
1
D C 2C
s 2 ..s C 1=2/2 C 1=4/
s
s
..s C 1=2/2 C 1=4/
.As C B/..s C 1=2/2 C 1=4/ C .C.s C 1=2/ C D/s 2 D 1:
B
A C 2B C 2D
5A C 10B C 2C C 2D
ACC

D
D
D
D

2
8
8
0

.set s D 0/I
.set s D 1=2/I
.set s D 1=2/I
.equate coefficients of s 3 /:

Solving this system yields A D 4, B D 2, C D 4, D D 0. Therefore,


1
s 2 ..s C 1=2/2 C 1=4/
This, (A), and (B) imply that y D e

t =2

4
2
4.s C 1=2/
C 2C
s
s
.s C 1=2/2 C 1=4

D
$

4 C 2t C 4e

t =2

sin.t=2// C 2t

.5 cos.t=2/

cos.t=2/:
4.

8.3.36.
.4s 2 C 4s C 1/Y .s/ D

3Cs
3Cs
C 4. 1 C 2s/ C 8 D 2
C 4. 1 C 2s/ C 8s C 4:
2
s C1
s C1

Since 4s 2 C 4s C 1 D 4.s C 1=2/2 ,


Y .s/ D

where

3Cs
2
C
:
4.s C 1=2/2.s 2 C 1/
s C 1=2

.A/

A
B
Cs C D
3Cs
D
C
C 2
2
2
2
4.s C 1=2/ .s C 1/
s C 1=2 .s C 1=2/
s C1
.A.s C 1=2/ C B/.s 2 C 1/ C .C s C D/.s C 1=2/2 D
10B
2A C 4B C D
12A C 8B C 9C C 9D
ACC

D
D
D
D

5
3
4
0

3Cs
:
4

.set s D 1=2/I
.set s D 0/I
.set s D 1/I
.equate coefficients of s 3/:

Solving this system yields A D 3=5, B D 1=2, C D 3=5, D D 1=5. Therefore,


3Cs
4.s C 1=2/2.s 2 C 1/

D
$

3
1
1
1
1 3s C 1
C
:
2
5 s C 1=2 2 .s C 1=2/
5 s2 C 1
3 t =2 1
1
e
C te t =2
.3 cos t C sin t/:
5
2
5

140 Chapter 8 Laplace Transforms


Since

2
$ 2e
s C 1=2

t =2

t =2

, this and (A) imply that y D

10

1
.3 cos t C sin t/.
5

.5t C 26/

8.3.38. Transforming the initial value problem


ay 00 C by 0 C cy D 0; y.0/ D 1; y 0 .0/ D 0
as C b
. Therefore, y1 D L
as 2 C bs C c
satisfies the initial conditions y1 .0/ D 1, y10 .0/ D 0.
Transforming the initial value problem

yields .as 2 C bs C c/Y .s/ D as C b, so Y .s/ D

as C b
as 2 C bs C c

a
2
as C bs C c

ay 00 C by 0 C cy D 0; y.0/ D 0; y 0 .0/ D 1
a
. Therefore, y2 D L
2
as C bs C c
satisfies the initial conditions y1 .0/ D 0, y10 .0/ D 1.

yields .as 2 C bs C c/Y .s/ D a, so Y .s/ D


8.4 THE UNIT STEP FUNCTION
8.4.2.
Z

L.f / D

st

f .t/ dt D

1
st

e
0

t dt C

This and (B) imply that L.f / D .1


Alternatively, f .t/ D t

u.t

8.4.4.
L.f / D

Letting t D x C 1 in the last integral yields


Z 1
e st .t 1/ dt D

1
0

/L.t/ D

f .t/ dt D

s.xC1/

st

dt in (A) to obtain
st

.t

1/ dt:

dt C

s2

st

.t C 2/ dt:

R1
To relate the first term to a Laplace transform we add and subtract 1 e st dt in (A) to obtain
Z 1
Z 1
Z 1
L.f / D
e st dt C
e st .t C 1/ dt D L.t/ C
e st .t C 1/ dt:
0

Letting t D x C 1 in the last integral yields


Z 1
Z
e st .t C 1/ dt D

This and (B) imply that L.f / D L.1/ C e

s.xC1/

.B/

.A/

x dx D e s L.t/:

e s /L.t/ D

dt:

e
s2

1/ $ .1

1/.t

st

st

1
R 1 st
t
1 e

To relate the first term to a Laplace transform we add and subtract


Z 1
Z 1
L.f / D
e st t dt C
e st .1 t/ dt D L.t/
0

.x C 2/ dx D e

1
L.t C 2/ D C e
s

L.t C 2/:


1
2
C
.
s2
s

.A/

.B/

Section 8.4 The Unit Step Function

141

Alternatively,
1
1/.t C 1/ $ L.1/ C e L.t C 2/ D C e
s
s

f .t/ D 1 C u.t
8.4.6.

L.f / D

st

f .t/ dt D


1
2
C
.
s2
s

st 2

t D L.t 2 /

t 2 dt:

.A/

Letting t D x C 1 in the last integral yields


Z 1
Z 1
st 2
e t dt D
e s.xC1/ .t 2 C 2t C 1/ dx D e s L.t 2 C 2t C 1/:
1

This and (A) imply that


L.f / D L.t 2 / C e s L.t 2 C 2t C 1/ D

2
s3


2
2
1
C
C
:
s3
s2
s

Alternatively,
2

f .t/ D t .1

u.t

1// $ L.t / C e

8.4.8. f .t/ D t 2 C 2 C u.t


L u.t

t2

1/.t


2/

2
2
it follows that F .s/ D 3 C
s
s
8.4.10. f .t/ D e
e

.sC2/

L.e

2t

t2

1/.t

2
L.t C 2t C 1/ D 3
s
2

1/.e

2
2
1
C 2C
s3
s
s

2
2
C and
3
s
s

e s L .t C 1/ .t C 1/2 2


1
2
2
e s L.t 2 C t C 2/ D e s 3 C 2 C
;
s
s
s

1
2
C 2C
.
s
s

D
2
s3

e t / $ L.e t / C e s L.e
1 e .sC1/
e .sC2/
e .sC1/ L.e t / D
C
.
sC1
sC2
C u.t

2/. Since t 2 C 2 $
D

2t

8.4.12.f .t/ D u.t


1/ u.t 2/t $ e s L.t C 1/

1
1
1
2
De s 2C
e 2s
C
.
s
s
s2
s

2s

2.t C1/

L.e

t 1

/ D L.e t / C

L.t C 2/

8.4.14.
f .t/

1
s2

2u.t
2e s
s2

1
1/ C u.t 2/.t C 4/ $ 2
s


1
6
C e 2s 2 C
:
s
s

1/.t

8.4.16. f.t/ D 2 2u.t 1/t C u.t 3/.5t


2
2
2
5
13
e s 2C
C e 3s 2 C
.
s
s
s
s
s

2/ $ L.2/

2e s L.t/ C e

2s

2e s L.t C 1/ C e

L.t C 6/

3s

L.5t C 13/ D

142 Chapter 8 Laplace Transforms


8.4.18. f .t/ D .t C 1/2 C u.t


1/ .t C 2/2 .t C 1/2 D t 2 C 2t C 1 C u.t


2
2
1
2
5
2
s
s
C
.
L.t C 2t C 1/ C e L.2t C 5/ D 3 C 2 C C e
s
s
s
s2
s
1
1
1
8.4.20.
D
$1 e
s.s C 1/
s sC1

)e


1
$ u.t 1/ 1
s.s C 1/

.t 1/

1/.2t C 3/ $

0;
1

.t 1/

0  t < 1;
; t  1:

8.4.22.

3
1
t )e
$ u.t 1/ .3 .t 1// D u.t 1/.4 t/I
s s2


1
1
1
3s 1
C 2 $1Ct )e
C 2 $ u.t 3/ .1 C .t 3// D u.t 3/.t 2/I
s
s
s
s
3
s

1
$3
s2

therefore

h.t/ D 2 C t C u.t

8.4.24.

t/ C u.t

1/.4

3/.t

1 2s
5 2.s C 2/
D
$e
s 2 C 4s C 5
.s C 2/2 C 1

therefore,

h.t/

D u.t
D u.t
(

 /e
 /e

2.t /
2.t /

2.t /

2t

.5 sin t

2 cos t/I

.5 sin.t
.2 cos t

0;

8
2 C t; 0  t < 1;

<
6;
1  t < 3;
2/ D

:
t C 4; t  3:

.2 cos t

 / 2 cos.t  //
5 sin t/
0  t < ;
:
5 sin t/; t  :

s C1
3.s 3/
4
1
3.s 3/
. Since
D
and
.s C 1/.s 2/ .s 1/.s 2/
.s C 1/.s 2/
sC1 s 2
sC1
3
2
4
4
2
D
, F .s/ D
C
$ 4e t 4e 2t C2e t . Therefore,e 2s F .s/ $
.s 1/.s 2/
s 2 s 1
s C(1 s 2 s 1


0;
0  t < 2;
u.t 2/ 4e .t 2/ 4e 2.t 2/ C 2e .t 2/ D
.t 2/
2.t 2/
.t 2/
4e
4e
C 2e
; t  2:
8.4.26. Denote F .s/ D

8.4.28.
3
s

1
$3
s3

t2
)e
2

2s

3
s

1
s3

$ u.t


2/ 3

1
e 4s
$
t
)
$ u.t
s2
s2

2/2
2

4/.t

4/I

.t

D u.t

 2

t
2/
C 2t C 1 I
2

Section 8.5 Constant Coeefficient Equations with Piecewise Continuous Forcing Functions

143

therefore
h.t/


t2
C 2t C 1 C u.t
2
0t <2

D 1 t C u.t 2/
8
1 t 2;

< 3t 2
C 2t C 2; 2  t < 4;
D
2

3t

:
C 3t 2; t  4:
2

4/.t

4/

8.4.30. Let T be an arbitrary positive number. Since limm!1 tm D 1, only finitely many members of
ftm g are in 0; T . Since fm is continuous on tm ; 1/ for each m, f is piecewise continuous on 0; T .
If tM  t < tM C1 , then u.t tm / D 1 if m  M , while u.t tm / D 0 if m > M . Therefore,
M
X
.fm .t/ fm 1 .t// D fM .t/
f .t/ D f0 .t/ C
mD1

8.4.32. Since

1
X

Km

converges if  > 0,

mD0

1
X

tm

converges if  > 0, by the comparison test.

mD0

Therefore,(C) of Exercise 8.3.31 holds if s > s0 C  if  is any positive number. This implies that it holds
if s > s0 .
8.4.34. Let tm D m and fm .t/ D . 1/m , m D 0; 1; 2; : : : . Then fm .t/
fm 1 .t/ D . 1/m 2, so
!
1
1
X
X
1
f .t/ D 1 C 2
. 1/m u.t m/ and F .s/ D
1C2
. 1/m e ms Substituting x D e s in the
s
mD1
mD1


1
X
1
2e s
1 1 e s
x
m m
identity
. 1/ x D
(jxj < 1) yields F .s/ D
1
D
.
1Cx
s
1Ce s
s 1Ce s
mD1

8.4.36. Let tm D m and fm .t/ D . 1/m m, m D 0; 1; 2; : : : . Then fm .t/ fm 1 .t/ D . 1/m .2m 1/,
1
1
X
1 X
so f .t/ D
. 1/m .2m 1/u.t m/and F .s/ D
. 1/m .2m 1/e ms . Substituting x D e s
s
mD1
mD1
1
1
X
X
x
x
in the identities
. 1/m x m D
and
. 1/m mx m D
(jxj < 1) yields F .s/ D
1Cx
.1 C x/2
mD1
mD1


1
e s
2e s
1 .1 e s /
D
.
s
s
2
s 1Ce
.1 C e /
s .1 C e s /2
8.5 CONSTANT COEEFFICIENT EQUATIONS WITH PIECEWISE CONTINUOUS FORCING
FUNCTIONS

8.5.2. y 00 C y D 3 C u.t
L .u.t

4/.2t
4/.2t

8/; y.0/ D 1; y 0 .0/ D 0. Since


8// D e

4s

L .2.t C 4/

.s 2 C 1/Y .s/ D

8/ D e

4s

3
2e 4s
C
C s:
s
s2

L.2t/ D

2e 4s
;
s2

144 Chapter 8 Laplace Transforms


3
2e 4s
s
C
C 2
s.s 2 C 1/
s 2.s 2 C 1/
s C1




1
s
1
1
s
4s
D 3
C
2e
C 2
s s2 C 1
s2 s2 C 1
s C1


3
2s
1
1
4s
C
2e
:
D
s
s2 C 1
s2 s2 C 1

Y .s/

Since
1
s2
yD3

1
$t
s2 C 1

2 cos t C 2u.t

8.5.4. y 00

4s

sin t ) e

4/ .t

y D e 2t C u.t

sin.t

1
s2

1
s2 C 1

$ u.t

4/ .t

sin.t

4// ;

4//.

e 2t /; y.0/ D 3; y 0 .0/ D 1. Since




e4
1
L.u.t 2/.1 e 2t // D e 2s L.1 e 2.t C2// D e 2s
;
s s 2


1
e4
1
.s 2 1/Y .s/ D
C e 2s
C . 1 C 3s/:
s 2
s
s 2
2/.1

Therefore,
Y .s/

.s

1
3s 1
C
.s 1/.s
.s 1/.s C 1/
 C 1/.s 2/
e4
1
2s
Ce
s.s 1/.s C 1/ .s 1/.s C 1/.s

1
1/.s C 1/.s

e 2s e 4
1/.s C 1/.s

.s

2/

1 1
1 1
1 1
C
C
2s 1
6sC1
3s 2
1 t 1 t 1 2t
$
e C e C e I
2
6
3


1 t C2 1 .t 6/ 1 2t
$ u.t 2/
e
C e
C e
I
2/
2
6
3
2/

1
D
1/.s C 1/

1
1 1
1 1
1
1
C
C
$ 1 C et C e t I
s.s
s
2s 1
2sC1
2
2


2s
1
1
e
$ u.t 2/
1 C e t 2 C e .t 2/ I
s.s 1/.s C 1/
2
2
.s
Therefore,
1
13
y D et C e
2
6

3s 1
1
2
D
C
$ e t C 2e t :
1/.s C 1/
s 1
sC1

1
C e 2t C u.t
3

2/

1
1 C et
2

1
C e
2

8.5.6. Note that j sin tj D sin t if 0  t <  , while j sin tj D


value problem as
y 00 C 4y D sin t

2u.t

 / sin t C u.t

.t 2/

1
C e t C2
2

1
e
6

.t 6/


1 2t
e
:
3

sin t if   t < 2 . Rewrite the initial

2 / sin t; y.0/ D 3; y 0 .0/ D 1:

Section 8.5 Constant Coeefficient Equations with Piecewise Continuous Forcing Functions

145

Since
 / sin t/ D e

L .u.t

s

L.sin.t C  // D e

s

L.sin t/

and
2 / sin t/ D e

L .u.t
.s 2 C 4/Y .s/ D

s

1 C 2e
Ce
2
.s C 1/

2s

e s
.s 2 C 1/.s 2 C 4/

L.sin.t C 2 // D e

C1


1
D
.s 2 C 1/.s 2 C 4/

therefore

2s

3s; so Y .s/ D

1
s2 C 1

1
s2 C 4

2s

L.sin t/;

1 C 2e s C e 2s
1 3s
C 2
:
2
2
.s C 1/.s C 4/
s C4
1
sin t
3

1
sin 2tI
6

1
1
u.t  /
sin.t  /
sin 2.t
3
6


1
1
u.t  /
sin t C sin 2t
3
6

$
D


/

and
e 2s
.s 2 C 1/.s 2 C 4/

u.t

u.t

1
2 /
sin.t
3

1
2 /
sin t
3

1
2 /
sin 2.t
6

1
sin 2t I
6


2 /

therefore
1
y D sin 2t
3

1
3 cos 2t C sin t
3

2u.t

8.5.8. y 00 C 9y D cos t C u.t

3=2/.sin t

L .u.t

cos t//

3=2/.sin t


1
1
/
sin t C sin 2t C u.t
3
6

1
2 /
sin t
3


1
sin 2t :
6

cos t/; y.0/ D 0; y 0 .0/ D 0. Since


D e

3s=2

L .sin.t C 3=2/

3s=2

L.cos t C sin t/;

cos.t C 3=2//

s C1
1
sC1
; so Y .s/ D 2
e 3s=2 2
:
2
2
s C1
.s C 1/.s C 9/
.s C 1/.s 2 C 9/




1
1
1
1
1
1
D
$
sin t
sin 3t
and
.s 2 C 1/.s 2 C 9/
8 s2 C 1 s2 C 9
8
3


s
1
s
s
1
D
$ .cos t cos 3t/ :
.s 2 C 1/.s 2 C 9/
8 s2 C 1 s2 C 9
8

.s 2 C 9/Y .s/ D

s2

1
C1

3s=2

sC1
.s 2 C 1/.s 2 C 9/

D
$



sC1
sC1 sC1
sC1
D
.s 2 C 1/.s 2 C 9/
8
s2 C 1 s2 C 9


1
1
cos t C sin t cos 3t
sin 3t ; so
8
3

146 Chapter 8 Laplace Transforms


3s=2

sC1
.s 2 C 1/.s 2 C 9/

u.t

3=2/
.cos.t
8

3=2/ C sin.t

3=2/

=2/

1
sin 3.t
3=2/
3


u.t 3=2/
1
sin t cos t C sin 3t
cos 3t :
8
3


1
3=2/ sin t cos t C sin 3t
cos 3t .
3
cos 3.t

D
Therefore, y D

1
.cos t
8

8.5.10. y 00 C y D t

cos 3t/

1
u.t
8

 /t; y.0/ D 0; y 0 .0/ D 0. Since

2u.t

1

L .u.t  /t/ D e
L .t C  / D e
C
s2
s


1

1
.s 2 C 1/Y .s/ D 2 2e s 2 C
I
s
s
s
s

Y .s/

s


1

C
s 2 .s 2 C 1/
s.s 2 C 1/


1
1
s
2e
2 e
s2 s2 C 1

1
D
2e
2
2
s .s C 1/


1
1
D
s2 s2 C 1

s

s

1
s


s
:
s2 C 1

Since
1
s2

1
s2 C 1

u.t

s

sin t ) e
 / .t


1
1
s2 s2 C 1
sin.t  // D u.t

 C sin t/

 /.t

and
1
s
yDt

sin t

2u.t

8.5.12. y 00 C y D t

s
2
s C1

u.t

cos t ) e

s

 / .1

cos.t

1
s


s
s2 C 1
 // D u.t  /.1 C cos t/;

 /.t C sin t C  cos t/.


2 /t; y.0/ D 1; y 0 .0/ D 2I

3u.t

2 /t/ D e

L.u.t

.s 2 C 1/Y .s/ D
Y .s/ D

e 2s
$ u.t
s 2.s 2 C 1/

2s

L.t C 2 / D e
3e
s2

1 3e 2s
s 2 .s 2 C 1/

1
1
D 2
s 2 .s 2 C 1/
s
2 /..t

2

2s

2
1
C
2
s
s

2s

6 e 2s
C 2 C sI
s
6 e 2s
2Cs
C 2
I
2
s.s C 1/
s C1

1
$t
s2 C 1

sin.t

sin tI

2 // D u.t

2 /.t

2

sin t/I

Section 8.5 Constant Coeefficient Equations with Piecewise Continuous Forcing Functions
1
1
D
C 1/
s

s.s 2
e 2s
$ u.t
s.s 2 C 1/

y D t C sin t C cos t

2 /.1

s2

s
$1
C1

cos tI

2 // D u.t

cos.t

2 /.1

cos t/I

2Cs
$ 2 sin t C cos tI
s2 C 1
2 /.3t 3 sin t 6 cos t/.

u.t

8.5.14. y 00

4y 0 C 3y D 1 C 2u.t 1/; y.0/ D 0; y 0 .0/ D 0;


1 C 2e s
1 C 2e s
.s 2 4s C 3/Y .s/ D
; Y .s/ D
;
s
s.s 1/.s 3/
1
1
1 1
1 1
1
1
1 t
D
C
$ C e 3t
e ;
s.s 1/.s 3/
3s
6 s 3 2 s 1
3
6
2
s
e
1
1
1 t 1
$ u.t 1/
C e 3.t 1/
e
;
s.s 1/.s 3/
3 
6
2

1 1 3t
1
2
1
yD
e C e t C u.t 1/
C e 3.t 1/ e t 1 .
3 6
2
3
3

8.5.16. y 00 C 2y 0 C y D 4e t

4u.t

L 4u.t

1/e t ; y.0/ D 0; y 0 .0/ D 0. Since





4e
1/e t D e s L 4e .t C1/ D
s

.s 2 C 2s C 1/Y .s/ D
Y .s/ D

.s

where

4
1/.s C 1/2

A.s C 1/2 C B.s


A
C
ACB

1/.s C 1/ C C.s

D et

2te

eu.t

D et

2te

u.t

4y 0 C 4y D e 2t

2u.t
L u.t

1/ D 4:

1, C D 2. Therefore,

1
1
D
1/.s C 1/2
s 1

.s

4e sC1
; so
1
s 1
4e sC1
:
.s 1/.s C 1/2

D 1
.set s D 1/I
D
2 .set s D 1/I
D 0
.equate coefficients of s 2 /:

Solving this system yields A D 1, B D

8.5.18. y 00

A
B
C
1
D
C
C
;
2
1/.s C 1/
s 1
sC1
.s C 1/2

.s

sC1

1
sC1


1/ e t 1

1/ e t e

2
.s C 1/2
e

.t 1/

.t 2/

2.t

and

2.t
1/e

2/e 2t ; y.0/ D 0; y 0 .0/ D 1. Since



2/e 2t D e

2s


e 2sC4
L e 2t C4 D
;
s 2

1/e

.t 1/

.t 2/

147

148 Chapter 8 Laplace Transforms


.s 2

4s C 4/Y .s/ D
Y .s/ D

1
2/3

.s

therefore y D

t 2 e 2t
2

t 2 e 2t
e
)
2
.s
te 2t

u.t

8.5.20. y 00 C 2y 0 C 2y D 1 C u.t

1
s

2sC4

2/3

2e
.s

2/3

.s
$

2e
s

e4
u.t
2

2sC4

2sC4

2/3

2/e 2.t

1;

2/2

.s
2/

.t

so
:

2/2 D u.t

2/

.t

2/2 e 2t
I
2

2/2 e 2t .

2/.t

3 /.t C 1/; y.0/ D 2; y 0 .0/ D 1;




1
2 1
2s
2s
C
I
L.u.t 2 /.t 1// D e
L..t C 2 1// D e
s2
s


3 C 1
1
C
I
L.u.t 3 /.t C 1// D e 3s L..t C 3 C 1// D e 3s
s2
s




1
1
2 1
1
3 C 1
2
2s
3s
.s C 2s C 2/Y .s/ D C e
C
e
C
C . 1 C 2s/ C 4:
s
s2
s
s2
s

Let G.s/ D

2 /.t

1/

u.t

1
1
, H.s/ D
; then
s.s 2 C 2s C 2/
s.s 2 C 2s C 2/
Y .s/ D Y1 .s/ C e

where

2s

Y2 .s/

3s

Y3 .s/;

(A)

2s C 3
;
C 2s C 2
Y2.s/ D H.s/ C .2 1/G.s/;
Y1 .s/ D G.s/ C

Let yi .t/ D L

(B)

s2

(C)

Y3 .s/ D H.s/ C .3 C 1/G.s/:

(D)

.Yi .s//, .i D 1; 2; 3/. From (A),


y.t/ D y1 .t/ C u.t

Find L

2 /y2.t

2 /

u.t

3 /y3 .t

3 /:

.G.s//:
G.s/ D

A
B.s C 1/ C C
C
s
.s C 1/2 C 1

where A..s C 1/2 C 1/ C .B.s C 1/ C C /s D 1. Setting s D 0 yields A D 1=2; setting s D


A C D 1, so C D 1=2; since A C B D 0 (coefficient of x 2 ), B D 1=2. Therefore,


1 1
.s C 1/ C 1
1 1 t
G.s/ D
$
e .cos t C sin t/:
2 s
.s C 1/2 C 1/
2 2
Find L

(E)

.H.s//:
H.s/ D

A
B
C.s C 1/ C D
C
C
s
sC2
.s C 1/2 C 1

1 yields

(F)

Section 8.5 Constant Coeefficient Equations with Piecewise Continuous Forcing Functions

149

where .As C B/..s C 1/2 C 1/ C .C.s C 1/ C D/s 2 D 1.


2B
ACB CD
5A C 5B C 2C C D
ACB D 0

D
D
D
D

1
1
1
0

.set s D 0/I
.set s D 1/I
.set s D 1/I
.equate coefficients of s 3 /:

Solving this system yields A D 1=2, B D 1=2, C D 1=2, D D 0; therefore




1 1
1
sC1
1
H.s/ D
$
.1 t e
2
2
2 s s
.s C 1/ C 1/
2
Since

cos t/:

(G)

2.s C 1/ C 1
2s C 3
D
$ e t .2 cos t C sin t/;
s 2 C 2s C 2
.s C 1/2 C 1

(B) and (F)) imply that

y1 .t/ D

1 t
1
e .3 cos t C sin t/ C :
2
2

(H)

From (C), (F)), and (G),


y2 .t/ D 
so
y2 .t

1C

e

2 / D

t
C .
2

.t 2/


.

1/e

2

cos t

1/ cos t C

1
2

2

1
2

sin t;

sin t C 1


t
:
2

(I)

From (D), (F)), and (G),


y3 .t/ D

1
2

so
y3 .t

3 / D


e t .3 cos t C .3 C 1/ sin t C t C 3 / ;
1
e
2

.t 3/


.3 cos t C .3 C 1/ sin t C t/ :

Now (E), (20), (I), and (J)


1
1
y D e t .3 cos t C sin t/ C imply that
2
2



2 1
.t 2/
u.t 2 / e
. 1/ cos t C
sin t C 1
2


1
.t 3/
u.t 3 / e
.3 cos t C .3 C 1/ sin t/ C t .
2
8.5.22. (a) f .t/ D

1
X

nD0

u.t

1
1X
n /; F .s/ D
e
s nD0

ns

t
2

(J)

1
X
1
; Y .s/ D
e
s.s 2 C 1/ nD0

ns

1
D
s.s 2 C 1/

s
e ns
$
1
cos
t;
$ u.t n /.1 cos.t n // D u.t n /.1 . 1/n cos t/;
s2 C 1
s.s 2 C 1/
1
m
X
X
y.t/ D
u.t n /.1 . 1/n cos t/. If m  t < .m C 1/ , y.t/ D
.1 . 1/n cos t/. Therefore,
nD0
nD0
(
2m C 1 cos t; 2m  t < .2m C 1/ .m D 0; 1; : : : /
y.t/ D
.
2m;
.2m 1/  t < 2m .m D 1; 2; : : : /
1
s

150 Chapter 8 Laplace Transforms


1
X

(b) f .t/ D

2n /t; F .s/ D

u.t

nD0

1
X

1
X

2ns

nD0

L.t C 2n s/ D

1
2n
1
C
D 2
e 2ns Yn .s/, where Yn .s/ D 2 2
2
s .s C 1/ s.s C 1/
s
nD0
t

sin t C 2n

2n /yn .t/ D u.t

2n /.t

2n cos t/; therefore y.t/ D

sin t

m
X

.t

1
X


1
2n
C
; Y .s/ D
s2
s

1
2n
C
C1
s

2n
$ yn .t/ D
s2 C 1

1
X

2ns

u.t

2n /.t

Yn .s/ $ u.t

sin t

2n cos t/.

nD0

2n cos t/ D .m C 1/.t

sin t

nD0

2ns

2n / D sin t, e

If 2m  t < 2.m C 1/ , then


y.t/ D

nD0

s2

2n / D cos t and sin.t

2n cos t. Since cos.t

1
X

sin t

m cos t/:

!
1
1
X
X
1
1
n
ns
(c) f .t/ D 1C2
. 1/ u.t n /; F .s/ D
1C2
. 1/ e
; Y .s/ D
1
C
2
. 1/n e
2 C 1/
s
s.s
nD1
nD1
nD1
1
1
s
e ns
D
$ 1 cos t;
$ u.t n /.1 cos.t n // D u.t n /.1
s.s 2 C 1/
s
s2 C 1
s.s 2 C 1/
1
X
. 1/n cos t/; y.t/ D 1 cos t C 2
. 1/n u.t n /.1 . 1/n cos t/. If m  t < .m C 1/ ,
n

ns

nD1

y.t/ D 1
(d) f .t/ D

1
X

u.t

nD0

1 1
1
1
$ .e t Ce
2sC1 s
2
If m  t < .m C 1/,
y.t/

D
D

cos t C 2

m
X

. 1/n .1

nD1

1
1X
n/; F .s/ D
e
s nD0
t

2/;

e
s.s 2

m
1 X t
e
2

nD0
mC1

ns

1/

.s 2

1
X

nD0

ns

; Y .s/ D

u.t

n/
2

.t n/

Ce

1 e
.e t
2.1 e/

(e) f .t/ D .sin tC2 cos t/

. 1/n cos t/ D . 1/m


1
s.s 2

et C e


1
2 D .e t
2

C e t/

u.t 2n /; F .s/ D

.2m C 1/ cos t:

1
X

1
1 1
D
C
s.s 2 1/
2s 1
1


1X
2 ; y.t/ D
u.t n/ e t n C e
2

1/ nD0
t

ns

nD0

C e t/

m
X

en

1
1 C 2s X
e
s 2 C 1 nD0

2ns

; Y .s/ D

1
X
1 C 2s
e
.s 2 C 1/.s 2 C 2s C 2/ nD0

.As C B/..s C 1/2 C 1/ C .C.s C 1/ C D/.s 2 C 1/ D 1 C 2s:


1
1
3
0

1:

where
D
D
D
D


2 .

nD0

1 C 2s
As C B
C.s C 1/ C D
D 2
C
2
C 1/.s C 2s C 2/
s C1
.s C 1/2 C 1

2B C C C D
A C B C 2D
5A C 5B C 4C C 2D
ACC

.t n/

.set s D 0/I
.set s D 1/I
.set s D 1/I
.equate coefficients of s 3/:

2ns

Section 8.5 Constant Coeefficient Equations with Piecewise Continuous Forcing Functions

151

Solving this system yields A D 0, B D 1, C D 0, D D 1. Therefore,


.s 2

Since sin.t

1 C 2s
C 1/.s 2 C 2s C 2/

s2

2n / D sin t,
e

2ns

1 C 2s
$ u.t
.s 2 C 1/.s 2 C 2s C 2/

so
y.t/ D sin t

1
X

y.t/ D sin t
1
X

m 
X

nD0

n/; F .s/ D

u.t

nD0

s.s

.t 2n/

1
1/.s

2/

1
1X
e
s nD0

e ns
s.s 1/.s
y.t/ D
If m  t < m C 1,
y.t/

D
D

m 
X
1

2e t

nD0

mC1
2

et

1
2s
2/

ns

; Y .s/ D

1
u.t
2

1
1X
u.t
2 nD0

C e 2.t

m1

e
1

n/

mC1


n/ 1

2/

1 1
1
$
1
2s 2
2


n/ 1
2e t

mC1
2

1
C e 2.t
2

2e t
n

m/ 1

e
1

sin t;

sin t:


2e t C e 2t I

C e 2.t

C e 2.t

et

e 2.mC1/
1 e 2

1
1/.s

s.s

.t 2n/

.t 2n/

D mC1


2n / 1


2n / 1

u.t

nD0

If 2m  t < 2.m C 1/ ,

(f) f .t/ D

1
1
C 1 .s C 1/2 C 1

1 e t sin t:

n/

m
X

n/

1
e n C e 2.t
2
nD0

m/

e 2n

nD0

2mC2

e2

m
X

by 0 .t/ cy.t/
on .; t0/ and .t0 ; /, y 00 .t0 C/ D
a
f .t0 C/ by 0 .t0 / cy.t0 /
f .t0 / by 0 .t0 / cy.t0 /
, and y 00 .t0 / D
. This implies the conclusion.
a00
a
(b) Since y has a junp discontinouity at t0 , applying Exercise 8.4.23(c) to y 0 shows that y 0 is not
differentiable at t0 . Therefore,y cannot satisfy (A) on .; / if f has a jump discontinuity at some t0 in
.; /.
8.5.24. (a) The assumptions imply that y 00 .t/ D

f .t/

8.5.26. If 0  t < t0 , then y.t/ D 0 .t/. Therefore, y.0/ D 0 .0/ D k0 and y 0 .0/ D 0 .0/ D k1 , and
ay 00 C by 0 C cy D a000 C b00 C c0 D f0 .t/ D f .t/;

0 < t < t0 :

152 Chapter 8 Laplace Transforms


Now P
suppose that 1  m  n. For convenience, define tnC1 D 1. If tm  t < tmC1 , then
y.t/ D m
kD0 m .t/, so
ay 00 C by C cy D

m
X

.a00k C b0k C ck / D f0 C

kD0

m
X

.fk

fk

kD1

1/

D fm D f;

tm < t < tmC1 :

Thus, y satisfies ay 00 C by 0 C cy D f on any open interval that does not contain any of the points t1 ,
t2 ,. . . , tn .
Since .tm / D 0 .tm / for m D 1; 2; : : : , y and y 0 are continuous on 0; 1/. Since y 00 .t/ D .by 0 .t/ C
cy.t//=a if t tm (m D 1; 2; : : : ), y 00 has limits from the left at t1 ; : : : ; tn .
8.6 CONVOLUTION
a
s
as
and cos bt $ 2
, so H.s/ D 2
.
2
2
2
Ca
s Cb
.s C a /.s 2 C b 2 /
1
a
a
(b) e t $
and sin at $ 2
, so H.s/ D
.
2
s 1
s Ca
.s 1/.s 2 C a2 /
a
1
as
(c) sinh at $ 2
and cosh at $ 2
, so H.s/ D 2
.
2
2
s
a
s
a
.s
a2 /2
2!s
s 2 !2
2!s.s 2 ! 2 /
(d) t sin !t $ 2
and
t
cos
!t
$
,
so
H.s/
D
.
.s C ! 2 /2
.s 2 C ! 2 /2
.s 2 C ! 2 /4
Z t
Z t


(e) e t
sin ! cos !.t / d  D
.e  sin !/ e .t  / cos !.t / d ; e t sin !t $

8.6.2. (a) sin at $

s2

s 1
.s 1/!
and e cos !t $
, so H.s/ D
.
2
2
.s 1/ C !
..s 1/2 C ! 2 /2
Z t
Z t
(f) e t
 2 .t /e  d  D
 2 e 2 .t /e .t  / d ; t 2 e 2t $

!
1/2 C ! 2

.s

H.s/ D
(g) e

.sZ

2
2/3 .s

e
0

1/2

and te t $

2/3

 cos !.t

/ d  D

e

2

.t  /

2t

cos !.t / d ; te


/ d ; e 3t $

1
s

1
.
3/ ..s 1/2 1/
1
2
2
(i) te 2t $
and sin 2t $ 2
, so H.s/ D
.
2
2
.s 2/
s C4
.s 2/ .s 2 C 4/
6
1
6
(j) t 3 $ 4 and e t $
, so H.s/ D 4
.
s
s 1
s .s 1/
3
3  6
6
(k) t 6 $ 7 and e t sin 3t $
, so H.s/ D 7
.
2
s
.s C 1/ C 9
s .s C 1/2 C 9
2
6
12
(l) t 2 $ 3 and t 3 $ 4 , so H.s/ D 7 .
s
s
s
7
2
2  7
7
t
(m) t $ 8 and e sin 2t $
, so H.s/ D 8
.
s
.s C 1/2 C 4
s .s C 1/2 C 4
24
2
48
(n) t 4 $ 5 and sin 2t $ 2
, so H.s/ D 5 2
.
s
s C4
s .s C 4/

so H.s/ D

.s

1
.s

1/2

, so

sC1
sC1
, so H.s/ D
.
2
2
2
.s C 1/ ZC !
.s C 2/
.s C 1/2 C ! 2
Z

t
t
(h) e t
e 2 sinh.t / d  D
e 3 e .t  / sinh.t
0

2
.s

1
and e
.s C 2/2

and e t sinh t $

.s

cos !t $

1
1/2

Section 8.6 Convolution

153



1
Y .s/
1
1
s2 C 1
1
1
8.6.4. (a) Y .s/ D 2
;
Y
.s/
1
C
D
;
Y
.s/
D 2 ; Y .s/ D 2
, so y D sin t.
2
2
2
s
s2
s
s
s
s
s
C1


2
1
2sY .s/
2s
1
.s C 1/
1
(b) Y .s/ D 2
; Y .s/ 1 C 2
D 2
; Y .s/ 2
D 2
; Y .s/ D
2
s C1
s C1
s C1
s C1
s C1
s C1
1
, so y D te t .
.s C 1/2


2sY .s/
2s
1
.s 1/2
1
.s 2 C 1/
A
1
; Y .s/ 1
D ; Y .s/ 2
D ; Y .s/ D
D
C
(c) Y .s/ D C 2
2
s
s C1
s C1
s
s C1
s
s.s 1/2
s
B
C
C
, where A.s 1/2 C Bs.s 1/ C C s D s 2 C 1. Setting s D 0 and s D 1 shows that A D 1
s 2 .s 1/2
1
1
and C D 2; equating coefficients of s 2 yields A C B D 1, so B D 0. Therefore, Y .s/ D C
,
s
.s 1/2
so y D 1 C 2te t .




1
Y .s/
1
1
s
1
sC1
1
1
(d) Y .s/ D 2 C
; Y .s/ 1
D 2 ; Y .s/
D 2 ; Y .s/ D
D 2 C 3,
s
sC1
s C1
s
sC1
s
s3
s
s
t2
so y D t C .
2


sY .s/
s
1
s3
4s 2 C 1
1
(e) sY .s/ 4 D 2 C 2
; Y .s/ s
D
4
C
;
Y
.s/
D
; Y .s/ D
s
s C1
s2 C 1
s2
s2 C 1
s2
2
2
4
2
.4s C 1/.s C 1/
4s C 5s C 1
4
5
1
5
1
D
D C 3 C 5 , so y D 4 C t 2 C t 4 .
5
5
s
s
s  s
2
24
 s
Y .s/
1
s 1
s2
s 1
s 1
s 1
(f) Y .s/ D 2
C 2
; Y .s/ 1
D 2
; Y .s/ 2
D 2
; Y .s/ D 2 D
2
s C1 s C1
s C1
s C1
s C1
s C1
s
1
1
, so y D 1 t.
s
s2
Z t
Z 0
Z t
8.6.6. Substituting x D t  yields
f .t /g./ d  D
f .x/g.t x/. dx/ D
f .x/g.t
0
t
0
Z t
x/ dx D
f ./g.t / d .
0

k0 .as C b/ C k1 a
F .s/
C
. Since
p.s/
p.s/
a.r1 C r2 /, (A) can be rewritten as

8.6.8. p.s/Y .s/ D F .s/ C a.k1 C k0 s/ C bk0 , so (A) Y .s/ D


p.s/ D a.s

r1 /.s

r2 / and therefore b D

k0 .s r1 r2 /
k1
C
:
.s r1 /.s r2 /
.s r1 /.s r2 /


1
1
1
e r2 t e r1 t
1
D
$
;
.s r1 /.s r2 /
r2 r1 s r2 s r1
r2 r1
so the convolution theorem implies that
Z
F .s/
1 t e r2  e r1 
$
f .t / d :
a.s r1 /.s r2 /
a 0
r2 r1
Y .s/ D

s
.s

a.s

F .s/
r1/.s

r2 /

r1 r2
r2
1
D
r1 /.s r2 /
r2 r1 s r1

r1
r2

1
r1 s

r2

r2 e r 1 t
r2

r1 e r 2 t
:
r1

Therefore,
y.t/ D k0

r2 e r 1 t
r2

r1 e r 2 t
e r2 t
C k1
r1
r2

e r1 t
1
C a
r1
a

t
0

e r2 
r2

e r1 
f .t
r1

/ d :

154 Chapter 8 Laplace Transforms


F .s/
k0 .as C b/ C k1 a
C
. Since
p.s/
p.s/
/2 C ! 2 and therefore b D 2a, (A) can be rewritten as

8.6.10. p.s/Y .s/ D F .s/ C a.k1 C k0 s/ C bk0 , so (A) Y .s/ D


p.s/ D a.s

Y .s/ D
.s

a.s

k0 .s 2/
F .s/
C
C
/2 C ! 2
.s /2 C ! 2
.s

k1
:
/2 C ! 2

1
1
$ e t sin !t, so the convolution theorem implies that
2
2
/ C !
!
Z t
F .s/
1
$
e t f .t / sin ! d :
a.s /2 C ! 2
a! 0


s 2
.s / 

t
D
$e
cos !t
sin !t :
.s /2 C ! 2
.s /2 C ! 2
!

Therefore,

 
y.t/ D e t k0 cos !t


sin !t
!


Z t
k1
1
sin !t C
e t f .t
!
a! 0

/ sin ! d :

8.6.12. (a)
ay 00 C by 0 C cy D f0 .t/ C u.t
p.s/Y .s/ D F0 .s/ C L.u.t
Y .s/ D

f0 .t//; y.0/ D 0; y 0 .0/ D 0I

t1 /.f1 .t/
t1 /.f1 .t/

f0 .t/// D F0 .s/ C e

st1

L.g/I

st1

F0 .s/ C e
G.s/
:
p.s/

.B/

1
$ w.t/, the convolution theorem implies that
p.s/
Z t
Z t
F0 .s/
G.s/
$
w.t /f0 ./ d  and
$
w.t /g./ d :
p.s/
p.s/
0
0
Z t
e st1 G.s/
Now Theorem 8.4.2 implies that
$ u.t t1 /
w.t t1 /g./ d , and (B) implies that
p.s/
0
Z t
Z t t1
y.t/ D
w.t /f0 ./ d  C u.t t1 /
w.t t1 /g./ d :
(b) Since F0 .s/ $ f0 .t/, G.s/ $ g.t/, and

Rt

Rt

(c) Let 0 .t/ D 0 w.t /f0 ./ d  and 1 .t/ D 0 w.t /g./ d . Then y.t/ D 0 .t/ C u.t
t1 /1 .t t1 /. Using Leibnizs rule as in the solution of Exercise 8.6.11(b) shows that
Z t
Z t
0
0
0
0 .t/ D
w .t /f0 ./ d ; 1 .t/ D
w 0 .t /g./ d ; t > 0;
0

000 .t/ D

f0 .t/
C
a

if t > 0, and that

w 00.t
0

a000 C b00 C c0 D f0 .t/

/f0 ./ d ; 001 .t/ D

and

g.t/
C
a

w 00 .t

/g./ d ;

t > 0;

a001 C b01 C c1 D f1 .t C t1 /

f0 .t C t1 /;

t > 0:

Section 8.7 Constant Coefficient Equations with Impulses

155

This implies the stated conclusion for y 0 and y 00 on .0; t/ and .t; 1/, and that ay 00 C by 0 C cy D f .t/
on these intervals.
(d) Since the functions 0 .t/ and h.t/ D u.t t1 /1 .t t1 / are both continuous on 0; 1/ and h.t/ D 0
if 0  t  t1 , y is continuous on 0; 1/. From (c), y 0 is continuous on 0; t1/ and .t1 ; 1/, so we need
only show that y 0 is continuous at t1 . For this it suffices to
that h0 .t1 / D 0. Since h.t1 / D 0 if
Z tshow
t1
h.t/ h.t1 /
D 0. If t > t1 , then h.t/ D
w.t t1 /g./ d . Since h.t1 / D 0,
t  t1 , (B) lim
t !t1
t t1
0
Z t t1

h.t/ h.t1 /


jw.t t1 /g./j d :
.B/

t t
1
0
Since g is continuous from the right at 0, we can choose constants T > 0 and M > 0 so that jg./j < M
if 0    T . Then (B) implies that

Z t t1
h.t/ h.t1 /

M
jw.t t1 /j d ; t1 < t < t1 C T:
.C/
t t

1
0

Now suppose  > 0. Since w.0/ D 0, we can choose T1 such that 0 < T1 < T and jw.x/j < =M if
0  x < T1 . If t1 < t < t1 C T1 and 0    t t1 , then 0  t t1  < T1 , so (C) implies that

h.t/ h.t1 /

< ; t1 < t < t1 C T:


t t

1
Therefore, lim

t !t1 C

h.t/
t

h.t1 /
D 0. This and (B) imply that h0 .t1 / D 0.
t1

8.7 CONSTANT COEFFICIENT EQUATIONS WITH IMPULSES


10
10 C .s C 1/.7s 2/
2
5
C . 9 C 7s/ C 7; YO .s/ D
D
C
;
sC1
.s
1/.s
C
2/.s
C
1/
s
C
2
s
C
1




1
1
1
1
1
1
e t e 2t
yO D 2e 2t C 5e t ;
D
D
;w DL 1
D
;
p.s/
.s C 2/.s 1/
3 s 1 sC2
p.s/
3


5
y D 2e 2t C 5e t C u.t 1/ e .t 1/ e 2.t 1/ .
3
8.7.2. .s 2 C s

2/YO .s/ D

8.7.4. .s 2 C 1/YO .s/ D

3
s2 C 9

1 C s;





3
s 1
3
1
1
s 1
1 8s 5
3
C
D
C
D
;
.s 2 C 1/.s 2 C 9/
s2 C 1
8 s2 C 1 s2 C 
9
s2 C 1
8 s2 C 1 s2 C 9
1
1
1
1
1
yO D .8 cos t 5 sin t sin 3t/;
D 2
;w DL 1
D sin t; y D .8 cos t 5 sin t
8
p.s/
s C1
p.s/
8
sin 3t/ 2u.t =2/ cos t.
YO .s/ D

8.7.6. .s 2

1/YO .s/ D

8
C1
s

1
1
1 1
D
D
p.s/
.s 1/.s C 1/
2s 1
3e t 8 C 2u.t 2/ sinh.t 2/;
8.7.8. .s 2 C 4/YO .s/ D

8
s

8 C s.1 s/
4
3
8
D
C
; yO D 4e t C 3e t 8;
s.s 1/.s C 1/ s 1 s C 1 s
1
1
et C e t
;w D L 1
D
D sinh t; y D 4e t C
sC1
p.s/
2

s; YO .s/ D

C 8s; (A) YO .s/ D

Bs C C
where A.s 2 C 4/ C .Bs C C /.s
s2 C 4

.s

8
8s
C
;
2/.s 2 C 4/ s 2 C 4 .s

8
A
D
C
2/.s 2 C 4/
s 2

2/ D 8. Setting s D 2 yields A D 1; setting s D 0

156 Chapter 8 Laplace Transforms


yields 4A 2C
8
.s 2/.s 2 C
 4/
1
L 1
D
p.s/

D 8, so C D 2; A C B D 0 (coefficient of x 2 ), so B D A D 1; therefore
1
sC2
1
1
D
, so (A) implies that yO D e 2t C7 cos 2t sin 2t;
D 2
;w D
s 2 s2 C 4
p.s/
s C4
1
1
sin 2t. Since sin.2t  / D sin 2t, y D e 2t C 7 cos 2t sin 2t
u.t =2/ sin 2t.
2
2
1

1 s.s 1/
A
B
C
D
C
C
2
s 1
.s 1/.s C 1/
s 1 s C 1 .s C 1/2
where A.s C 1/2 C .B.s C 1/ C C /.s 1/ D 1 s.s 1/. Setting s D 1 yields A D 1=4; setting s D 1
O
yields C D 1=2; since A C B D 1 (coefficient of s 2), B D 1 A D 5=4. Therefore,

 Y .s/ D
1 1
5 1
1
1
1 t 1 t
1
1
1
C
; yO D e C e .2t 5/;
D
;w D L 1
D te t ;
4 s 1 4 s C 1 2 .s C 1/2
4
4
p.s/
.s C 1/2
p.s/
1
1
y D e t C e t .2t 5/ C 2u.t 2/.t 2/e .t 2/.
4
4
8.7.10. .s 2 C 2s C 1/YO .s/ D

C .2 s/ 2; YO .s/ D

1
.s C 1/ C 1
8.7.12. .s 2 C 2s C 2/YO .s/ D .2 s/ 2; Y .s/ D
; yO D e t .sin t cos t/;
D
2C1
.s
C
1/
p.s/


1
1
;w D L 1
D e t sin t. Since sin.t  / D sin t and sin.t 2 / D sin t,
.s C 1/2 C 1
p.s/
y D e t .sin t cos t/ e .t / u.t  / sin t 3u.t 2 /e .t 2/ sin t.
1 C s.7 2s/
7 1
6
1
1
1
C2.2 s/C3; YO .s/ D
D
;
s
2s.s C1=2/.s 2/
10s 2 5 s C1=2 2s
1
1
1
1
1
7 2t 6 t =2 1 1
yO D
e
e
;
D
D
;wDL 1
D
10
5
2 p.s/
2.s C 1=2/.s 2/
5 s 2 s C 1=2
p.s/


1 2t
7 2t 6 t =2 1
1
.e
e t =2/; y D
e
e
C u.t 2/ e 2.t 2/ e .t 2/=2 ;
5
10
5
2
5


s
1
1
s
s
s
2
O
O
8.7.16. .s C 1/Y .s/ D 2
1; Y .s/ D 2
D
s C4
.s C 1/.s 2 C 4/ s 2 C1
3 s2 C 1 s2 C 4

1
1
1
1
1
; yO D .cos t cos 2t 3 sin t/;
D 2
;w DL 1
D sin t. Since sin.t
s2 C 1
3
p.s/
s C1
p.s/
=2/ D cos t and sin.t  / D sin t,
8.7.14. .2s 2 3s 2/YO .s/ D

yD

1
.cos t
3

8.7.18. .s 2 C 2s C 1/YO .s/ D

cos 2t
1
s

.s

3 sin t/

=2/ cos t C 3u.t

2u.t

1; (A) YO .s/ D

1
1/.s C 1/2

.s

 / sin t:

1
;
.s C 1/2

1
A
B
C
D
C
C
2
1/.s C 1/
s 1 sC1
.s C 1/2

where A.s C 1/2 C .B.s C 1/ C C /.s 1/ D 1. Setting s D 1 yields A D 1=4; setting s D


C D 1=2; since A C B D 0 (coefficient of s 2), B D A D 1=4. Therefore,
.s

1
1 1
D
1/.s C 1/2
4s 1

1 1
4sC1

1
1
:
2 .s C 1/2

This and (A) imply that


1 1
YO .s/ D
4s 1

1 1
4sC1

3
1
I
2 .s C 1/2

1 yields

Section 8.7 Constant Coefficient Equations with Impulses




 1
1 t
1
1
1
e
e t .1 C 6t/ ;
D
;
w
D
L
D te t ;
4
p.s/
.s C 1/2
p.s/

1 t
yD
e
e t .1 C 6t/
u.t 1/.t 1/e .t 1/ C 2u.t 2/.t 2/e
4

157

yO D

.t 2/

8.7.20. y 00 C 4y D 1 2u.t =2/ C .t  / 3.t 3=2/; y.0/ D 1; y 0 .0/ D 1. .s 2 C 4/YO .s/ D




1 2e s=2
1 2e s=2
s 1
1
1 1
s
O
C s 1; Y .s/ D
C 2
. Since
D
, YO .s/ D
2
s
s.s
s.s 2 C 4/
4 s s2 C 4
 C 4/
s C 4
1 3 s
1
1
s
3
1
1 1
C 2
2e s=2
. yO D cos 2t
sin 2t C C u.t =2/.1Ccos 2t/.
2
2
4s 4 s C
s s C4
4
2
4 4
 4 s C 4
1
1
1
wDL
D sin 2t. Since sin 2.t  / D sin 2t and sin 2.t 3=2/ D sin 2t,
p.s/
2
3
1
1 1
1
3
y D cos 2t
sin 2t C C u.t =2/.1 C cos 2t/ C u.t  / sin 2t C u.t 3=2/ sin 2t.
4
2
4 4
2
2
8.7.26. w.t/ D e

st0

sin t; fh .t/ D

u.t

t0 /

u.t
h

s.t0 Ch/

t0

h/

; .s 2 C 2s C 2/Yh.s/ D

1e
h

st0

e
s

s.t0 Ch/


1e
e
1
.s C 1/ C 1
1
1
;
D
C
$
1 e t .cos t C sin t/ ;
2
2
2
h s.s C 2s C 2/ s.s C 2s C 2/
2 ..s C 1/ C 1/ 2s
2
8
0;
0  t < t0 ;

h
i
<
1
1 e .t t0 / .cos.t t0 / C sin.t t0 // ;
t0  t < t0 C h;
yh .t/ D
2h

i
.t t0 / h

: e
e h .cos.t t0 h/ C sin.t t0 h// cos.t t0 / sin.t t0 / ; t  t0 C h:
2h
Yh .s/ D

8.7.28. w.t/ D e
t

2t

1/2
2

t0 /

u.t
h

<

t0

h/

1e
h
1
1
e 2t
C
$
sC1
2s
2

; .s 2 C3sC2/Yh.s/ D

1
k/ D
2

.e

: .e

1
X

u.t

.t t0 /

e /

kD0

2.e

k/ e t

kD0
m
mC1
X
t
ke

1/2

1 t
.e
1/

e
s
e

s.t0 Ch/

1
D
2

2h
.e
2h

1/

t0  t < t0 C h;

.t t0 h/

1
s2

1/2

; t  t0 C h:

1 t
e
2

1
X

; y D yO C
u.t

1
m

1 X t
k
. If m  t < m C 1, then y D
e
2

kD0

kD0

(b) .s C 1/YO .s/ D 1, so yO D w D L


2

st0

0  t < t0 ;

0;
.t t0 /

1/YO .s/ D 1, so yO D w D L

8.7.30. (a) .s 2

1 t
.e
2

u.t

yh .t/ D

k/w.t

; fh .t/ D

1 e st0 e s.t0 Ch/


1
1
;
D
h s.s C 1/.s C 2/ s.s C 1/.s C 2/
2.s C 2/

Yh .s/ D
.e

t k

e t /.
1

1
2
s C1

D sin t; y D yO C

1
X

kD0

u.t

2k /w.t

2k / D

158 Chapter 8 Laplace Transforms

sin t

1
X

2k /. If 2m  t < 2.m C 1/ , then y D .m C 1/ sin t.

u.t

kD0

(c) .s
k/w.t
e 2.t

m/

3s C 2/YO .s/ D 1, so yO D w D L

k/ D
m
X

1
X

kD0

e 2k

et

kD0

(d) w D L
yD

0;
sin t;


k/ e 2.t

u.t
m

m
X

kD0
1

1
s2 C 1

k/

e k D e 2.t

et
m/ e

D sin t; y D

2m  t < .2m C 1/;

.s

1
1/.s

2/

D .e 2t

. If m  t < m C 1, then y D

2mC2

e2
1
X

kD1

.2m C 1/  t < .2m C 2/;

1
1
u.t

e .t

m/ e

k /w.t

mC1

1
1

m 
X

e 2.t

k/

kD0

1
X

u.t

kD0

et

k / D sin t

.m D 0; 1; : : : /.

e t /; y D yO C

1
X

. 1/k u.t

kD1

k /, so

CHAPTER 9
Linear Higher Order Equations
9.1 INTRODUCTION TO LINEAR HIGHER ORDER EQUATIONS
c3
2c3
c3
9.1.2. From Example 9.1.1, y D c1x 2 Cc2x 3 C y 0 D 2c1x C3c2x 2
, and y 00 D 2c1 C6c2x C 3 ,
x
x2
x
where
c1 c2 c3 D
4
2c1 C 3c2 c3 D
14
2c1 6c2 2c3 D
20;
so c1 D 2, c2 D 3, c3 D 1, and y D 2x 2

3x 3 C

1
.
x

9.1.4. The general solution of y .n/ D 0 can be written as y.x/ D


n
X1

mDj

m.m 1/    .m j C1/cm .x x0 /m

n
X1

cm.x

mD0

, y .j / .x0 / D j cj . Therefore, yi D

x0 /m . Since y .j / .x/ D
.x x0 /i 1
; 1  i  n.
.i 1/

9.1.6. We omit the verification that the given functions


solutions of the given equations.
are

ex e x

1 1
xe x
0
x

1 1 D
e x e x .1 x/ ; W .0/ D 1
(a) The equation is normal on . 1; 1/. W .x/ D e
e x e x e x .x 2/
1 1
2
4. Apply Theorem 9.1.4.
(b) The equation
is normal on . 1; 1/.
x

e x cos 2x
e x sin 2x
x

x
x

e .cos 2x 2 sin 2x/


e .2 cos 2x C sin 2x/ ;
W .x/ D e
x
e x .3cos 2x C 4 sin 2x/ e x .4 cos 2x 3 sin 2x/
e
1 1 0

W .0/ D 1 1 2 8. Apply Theorem 9.1.4.


1
3 4
(c) The equation
on
x is normal
. 1; 0/ and .0; 1/.
x
e

e
x
x

e
1 D 2x. Apply Theorem 9.1.4.
W .x/ D e
ex e x 0
(d) The equation
is normal on . 1; 0/ and .0; 1/.

e x =x
e x =x
1

W .x/ D
e x .1=x 1=x 2 /
e x .x C 1/=x 2
0 D 2=x 2. Apply Theorem 9.1.4.
e x .1=x 2=x 2 C 2=x 3/ e x .x 2 C 2x C 2/=x 3 0
159

160 Chapter 9 Linear Higher Order Equations

x x2 ex
0 0 1

(e) The equation is normal on . 1; 1/. W .x/ D 1 2x e x ; W .0/ D 1 0 1 D 2; Apply


0 2 ex
0 2 1
Theorem 9.1.4.
(f) The equation
is normal on . 1; 1=2/
and .1=2; 1/.

2x
x ex e x

1 ex
e x 2e 2x
W .x/ D
D e 2x .12x 6/. Apply Theorem 9.1.4.
x
e x 4e .2x/
0 e
0 ex
e x
8e 2x
(g) The equation
is normal on . 1; 0/

and .0; 1/.


1 x 2 e 2x
e 2x

0 2x 2e 2x
2e 2x
W .x/ D
2x
2x D 128x. Apply Theorem 9.1.4.
0
2
4e
4e

0 0 8e 2x
8e 2x

x
 Z x
 Z x

9.1.8. From Abels formula, (A) W .x/ D W .=2/ exp


tan t dt ;
tan t dt D ln cos t
D
=4
=4
=4
p
p
ln. 2 cos x/; therefore (A) implies that W .x/ D 2K cos x.

1 ex
1 1
e x
1

9.1.10. (a) W .x/ D 0 e x


e x D .e x /.e x / 0 1
1 D 2.
x
0 ex

e
0 1
1
x

e x sin x
e x cos x
x

x
x

e .cos x C sin x/ e .cos x sin x/ D


(b) W .x/ D e
ex

2e x cos x
2e x sin x

sin x
cos x

3x
D e 1 cos x C sin x cos x sin x
1
2 cos x
2 sin x

sin x
cos x

3x

cos x
sin x
D e 0

1 2 cos x sin x
2 sin x cos x

1
sin x
cos x

cos x
sin x D e 3x
D e 3x 0
0
sin x
cos x

2 x C 1 x2 C 2

1
2x D 4.
(c) W .x/ D 0
0

0
2


x
ln jxj
1=x 2
x ln jxj
1=x 1

1=x 2
1=x 2 D 1 ln jxj C 1
(d) W .x/ D 1 ln jxj C 1
0
1
2=x 2
1=x
2=x 3 0

1
1 ln jxj
ln jxj
1
1

1
1
1 ln jxj C 1
1 D 2 0
1
2
D
x 2 0
x
0
1
2
1
2

1 ln jxj
1

1
0
1
2 D 4=x 2:
D
x 2 0
0
4

Section 9.1 Introduction to Linear Higher Order Equations


2

6
6
6
6
(e) W .x/ D 6
6
6
4

1 x
0 1
0 0
0 0
:: ::
: :
0 0

x
e

(f) W .x/ D e x
ex
(g) W .x/

x 2 =2 x 3 =3
x
x 2=2
1
x
0
1
::
::
:
:
0 0
e x x 1
e x 1 D 1
e x 0 1

D
D
D
D

x x2

(h) W .x/ D 1 2x
0 2
x 2
D e .x
2x C 2/.

(i) W .x/

e x =x
x
e =x e x =x 2
x
e =x 2e x =x 2 C 2e x =x 3

D 2x.

e x =x
1
e x =x e x =x 2
0
x
x
2
x
3
e =x C 2e =x C 2e =x
0

1=x
1=x
1
2
2
1=x 1=x
1=x 1=x
0
2
3
2
3
1=x 2=x C 2=x
1=x C 2=x C 2=x
0

1=x 1=x 2
1=x 1=x 2

2
3
2
3
1=x 2=x C 2=x
1=x C 2=x C 2=x

1=x 1=x 2
2=x
D 2=x 2:
1=x 2=x 2 C 2=x 3 4=x 2

x x2 1
 
  2

e x

2x 1
x
1
e x D e x 1 2x 1 D e x x
2 1
2 1
0 2 1
ex

3
x n =n
x n 1 =.n 1/ 7
7
x n 2 =.n 2/ 7
7
D 1.
x n 3 =.n 3/ 7
7
7
::
5
:

1

1 x 1
1
x
1 1 D 0
2 1 x
1 0 0
0
x




::
:

x x3
1 3x 2
0 6x
0
6

2x 3

6x

x 4 0
0

1=x
1=x 2
2=x 3
6=x 4
2=x
2=x 3
6=x 4
2=x 4
8=x 4
0


0
2x 3
2=x

1 3x 2
1=x 2
D
0
6x
2=x 3

0
6
6=x 4

2 2=x 4
3=x 2

4
4
D x 6 2=x 4
6=x

6
24=x 5
6=x 4

3=x 5
15=x 5 D 240=x 5:
15=x 5
1=x 2
2=x 3
6=x 4
24=x 5

3=x 2
2=x 3
6=x 4
24=x 5

3=x 5
6=x 5
24=x 5

161

162 Chapter 9 Linear Higher Order Equations

1
ex
e x x e 2x

x
x
2x

e
e
1 2e
2x 1
x
x
2x D e
1
e
e
0 4e

x
x
2x
1
e
e
0 8e

1 x 1 C x 0 1 2x

1
1
2
2x
D
D e

1
0
4
1

1
0
8

2 1 C x 1 2x

2x
2x

D
e 2
1
4
D e
0

1
8

(j) W .x/

(k)

W .x/

e 2x
2e 2x
4e 2x
8e 2x
2
4
8

e
2e
4e
8e

2x

1
2x
0
2x
0
2x
0

2 2x 0
4 2 D 8
8 0 0

1 x
1 1
1 0
1 0

2x
e

1
2
4
8

1
1

2 1Cx
0
x
0
1

x2
2x
2
0

2
4
8

1Cx
1
1

1
2
4
8

2x
2 D
0

1 2x
3 C 2x
8
1
2
4
8

2x

D 6e 2x .2x

1/:

1 x 2
0 2x
0 2
0 0

128x:

9.1.12. Let y be an arbitrary solution of Ly D 0 on .a; b/. Since f1 ; : : : ; n g is a fundamental set of


n
X
solutions of Ly D 0 on .a; b/, there are constants c1, c2 ,. . . , cn such that y D
ci yi . Therefore,
y D

n
X

ci

i D1

n
X

j D1

aij yj D

n
X

j D1

Cj yj , with Cj D

solutions of Ly D 0 on .a; b/.

n
X
i D1

i D1

aij ci . Hence fy1 ; : : : ; yn g is a fundamental set of

9.1.14. Let y be a given solution of Ly D 0 and D


r D 0; : : : ; n

n
X

j D1

y .j

1/

.x0 /yj Then .r / .x0 / D y .r /.x0 /,

1. Since the solution of every initial value problem is unique (Theorem 9.1.1), D y.

9.1.16. If fy1 ; y2; : : : ; yn g is linearly dependent on .a; b/ there are constants c1 ; : : : ; cn , not all zeros,
such that c1y1 C c2 y2 C    C cn yn . Let k be the smallest integer such that ck 0. If k D 1, then
1
1
y1 D .c2 y2 C  Ccn yn /; if 1 < k < n, then yk D 0y1 C  C0yk 1 C .ckC1 ykC1 C  Ccn yn /;
c1
ck
if k D n, then yn D 0, so yn D 0  y1 C 0  y2 C    C 0  yn .
P
9.1.18. Since F D f1i1 f2i2 ; : : : ; fnin ,
X
X
X
F0 D
f1i0 1 f2i2 ; : : : ; fnin C
f1i1 f2i0 2 ; : : : ; fnin C    C
f1i1 f2i2 ; : : : ; fni0 n
D

F1 C F2 C    C Fn :

Section 9.1 Introduction to Linear Higher Order Equations


9.1.20. Since yj.n/ D

W0 D

P
n

n
X

.Pk =P0 /yj.n

k/

163

, Exercise 9.1.19 implies that

kD1

y1

y2

y10

y20

::
:

::
:

y1.n

2/

y2.n

.n k/
kD1 .Pk =P0 /y1

so Exercise 9.1.17 implies that

W0 D



yn0


:

::
:



yn.n

::
2/

.n k/
kD1 .Pk =P0 /y2

Pn

y1

y10

n
X
Pk
::

:
P0
.n 2/
kD1
y
1

y .n k/
1

yn

y2
y20
::
:


::

.n k/
kD1 .Pk =P0 /yn

Pn





yn
yn0
::
:

.n 2/



yn

.n k/



yn

y2
y2

2/

.n 2/

.n k/

However, the determinants on the right each have two identical rows if k D
: : : ; n. Therefore,W 0 D
Z 2;
x
0
P1 W
W
P1
W .x/
P1 .t/
. Separating variables yields
D
; hence ln
D
dt, which implies
P0
W
P0
W .x0 /
x0 P0 .t/
Abels formula.
9.1.22. See the proof of Theorem 5.3.3.
9.1.24. (a)

x x2 1 x2 C 1 2
x 2 1 x 2 C 1
x

1 x 2 C 1

P0 .x/ D
2x
2x D 1
0
0
D
D
2
2

0
2
2
2
2

x x2 1 x2 C 1
x x2 1 x2 C 1
1

D 0; P3 .x/ D 0
P1 .x/ D 1
2x
2x D 0; P2 .x/ D 0
2
2

0
0
0
0
0
0. Therefore, 4y 000 D 0, which is equivalent to y 000 D 0.
(b)

1
0
e
0 x
1 x
e x x

e x 1 D 1
1 1 D 0
2 1 D 2xI
P0 D e
1
ex
1
1 0
1 0
e x 0

x

e
1 x 0
2 x
e x x 1
x
0 1 D 2I
P1 D e
e x 1 D 1
1 1 D 0
x

ex

1
1 0
e
0
1
1 0

0
e
1
1 x
0 x
e x x
x

1 0 D 1
1 0 D 2xI
e x 0 D 1
P2 D e
0
ex
1
1 0
2 0
e x 0

4I

2x
2
0

2x
2
0

164 Chapter 9 Linear Higher Order Equations

Therefore,
(c)

x
e

P3 D e x
ex

e
e
e

x
x
x

2xy 000 C 2y 00 C 2xy 0

x
e

P0 .x/ D e x
ex

P1 .x/ D

x
e

P2 .x/ D e x
ex

Therefore, .3
(d)

xe
e x .1
e x .x
ex
ex
ex

x/
2/

xe
e .1
e x .3

xe
e .x
e x .3
x


1

D 1

1


1 1 0
1 0 D 1
1 0 1

0
2
1

2y D 0, which is equivalent to xy 000

2/
x/

P3 .x/

2x/y 000 C 2y 00 C .2x

P0 .x/

1
0
0

x
1

1 1 x 0 D

1 x 2 0

1 1
x
1
x/ 0 D 1 1 x 0 D
x/ 0 1 3 x 0

1
x
1
1

0 D 1 x 2 0 D
1 3 x 0
0
x

x
e

x e x .1 x/ 0

D e
e .x 2/ 0 D 0:
e x e x .3 x/ 0
1
0
0

1
0
0
y 00

D 2:

1 1
1 x

1 1
1 3
1 x
1 3

xy 0 C y D 0.

x
D3
2

2xI

x
D 2I
x

2
D 2x
x

5I

5/y 0 D 0.

x x2 2
x x2 1
x x2 ex

1 2x e x D e x 1 2x 1 D e x 1 2x 2
D

0
0 2 1
0 2 ex
2

2
x x
2
D e x .x 2 2x C 2/I
D
e x
1 2x 2

x x2 ex

x x2
D x2ex I
P1 .x/ D 1 2x e x D e x
1 2x
0 0 ex

x x2 ex

x2
x x
x
D 2xe x I

P2 .x/ D 0 2 e D e

0
2
0 0 ex

1 2x e x

x 1 2x
x

D 2e x :
P3 .x/ D 0 2 e D e
0 2
0 0 ex

0
0
1

Therefore, e x .x 2 2x C 2/y 000 C x 2 e x y 00 2xe x y 0 C 2e x y D 0; which is equivalent to .x 2 2x C


2/y 000 x 2 y 00 C 2xy 0 2y D 0.
(e)

x x2
1 x
1 x
1=x
1=x 2
1=x 2

1 2x
1=x 2 D x 1 2x
1=x 2 D x 0 x
2=x 2
P0 .x/ D

0 2
0 2
0 2
2=x 3
2=x 3
2=x 3

x
6
2=x 2

D
x
D
I
2
2=x 3
x

Section 9.1 Introduction to Linear Higher Order Equations

6 000
y
x

Therefore,
(f)
P0 .x/

P1 .x/

P1 .x/ D 1
0

P2 .x/ D

P3 .x/ D 0
0

6 00 12 0
y C 3y
x2
x

D 6 x

4
x 1

x x2
1=x
6 x
3
0 2
2=x D 4
0
x
0 0
6=x 4

2x
1=x 2
6 1
2
2=x 3 D
x4 0
0
6=x 4
1=x
1=x 2
6=x 4

x 2
D
2x

6
I
x2

2x
D
2

12
:
x4

12
x 2
D 3I
2
x

12
y D 0, which is equivalent to x 3y 000 C x 2 y 00
x4

2xy 0 C 2y D 0.

x C 1 e x e 3x
xC1 1 1
xC1 0

4x
x
3x
1
D e 4x 1
e
3e
1
1
3
0
D
D
e

0
0
0
e x 9e 3x
1 9
1

xC1
8
D e 4x
D 2e 4x .1 3x/I
1
6

x C 1 ex
x C1 1 1
e 3x

4x
x
3x

1
e
3e
1 3
D
De 1

0
0
e x 27e 3x
1 27

xC1
26
D 2e 4x .12x 1/I
D
e 4x
1
24
P2 .x/

D
D

Therefore,

x2
2x
0

1 ex

P3 .x/ D 0 e x
0 ex
2e 4x .1

xC1 0

D e 4x 1
0

0
1

x C 1 ex
xC1
e 3x

4x
x
3x
0
e
9e
D e 0

0
0
e x 27e 3x

xC1 1 1

4x
1 9 D 18e 4x .x C 1/I
e 0
0
0 18

1 1 3
1 1
3e 3x

9e 3x D e 4x 0 1 9 D e 4x 0 1
0 1 27
0 0
27e 3x

3x/y 000 C 2e 4x .12x

1/y 00

1 1
1 9
1 27

3
9
18

1/y 000

.12x

1/y 00 C 9.x C 1/y 0

9y D 0:

8
6
9

26
24
27

D 18e 4x :

18e 4x .x C 1/y 0 C 18e 4x y D 0;

which is equivalent to
.3x

165

166 Chapter 9 Linear Higher Order Equations


(g)

P0 .x/

P1 .x/

1 x2
x x3
1=x
1=x 2
1=x 2
1=x 3

2
2
3
2
2

1 3x
1=x
2=x
1 3x
1=x
2=x 3
D x
3
4
3

0 6x
2=x
6=x
2=x
6=x 4

0 6x
4
5
4

0
6
6=x
24=x
0
6
6=x
24=x 5

1 x2

1=x 2
1=x 3

2x 2
2=x 2
3=x 3
2
3
0 2x 2

2=x
3=x
D x
D x 6x
2=x 3
6=x 4
2=x 3
6=x 4
4
0 6x

6
6=x
24=x 5
0
6
6=x 4
24=x 5

2x
2x
2=x 3
3=x 4
2=x 3
3=x 4

D x 2 6x
2=x 3
6=x 4 D x 2 0
8=x 3 15=x 4 D
6
0
6=x 4
24=x 5
0
15=x 5

P2 .x/

P3 .x/

1 x2
x x3
1=x
1=x 2
1=x 2
1=x 3

1 3x 2
1 3x 2
1=x 2
2=x 3
1=x 2
2=x 3

3
4 D x
3
0 6x
2=x
6=x
2=x
6=x 4

0 6x
5
6
5

0
0
24=x
120=x
0
0
24=x
120=x 6

1 x2

1=x 2
1=x 3

2x
2=x 3
3=x 4
2
3
0 2x 2

2=x
3=x
6=x 4
x
D x 2 6x 2=x 3
2=x 3
6=x 4
5
0 6x

0 24=x
120=x 6
0
0
24=x 5 120=x 6

2x
2=x 3
3=x 4
3

1200
15=x 4
2
3 8=x
3
4
8=x
15=x D 2x
x 0
I
5
6 D
24=x
120=x
x6
0 24=x 5 120=x 6

1 x2
x x3
1=x
1=x 2
1=x 2
1=x 3

2
2
3
2
2

1 3x
1=x
2=x
1=x
2=x 3
1 3x
4
5 Dx
4
0
6
6=x
24=x
6
6=x
24=x 5
0
5
6
5

0
0
24=x
120=x
0
0
24=x
120=x 6

1 x2

1=x 2
1=x 3

2
2=x 4
3=x 5
2
3
0 2x 2

2=x
3=x
3
D x
6=x 4
24=x 5
4
5 Dx 6
0
6
6=x
24=x
5

0 24=x
120=x 6
0
0
24=x 5 120=x 6

2=x 4
3=x 5

720
0
15=x 5
3
3
5
D x 0
0
15=x D 2x
I
5
6 D
24=x
120=x
x7
0 24=x 5 120=x 6

240
x5

x x3
1=x
1=x 2
6 2=x 4
6=x 5
3
4

0 6x 2=x
6=x
D x2 6
6=x 4
24=x 5

0 6
6=x 4
24=x 5
5
0 24=x
120=x 6
0 0 24=x 5 120=x 6

6 2=x 4

6=x 5

8=x 4
1440
30=x 5
2
2
4
5
8=x
30=x D 6x
x 0
I
5
6 D
24=x
120=x
x6
0 24=x 5 120=x 6

Section 9.1 Introduction to Linear Higher Order Equations

P4 .x/

Therefore,

1 3x 2
1=x 2
2=x 3
6 2=x 4

6=x 5
3
4

0 6x
2=x
6=x
4
5

D
x
6
6=x
24=x
4
5

0
6
6=x
24=x
5
6

0
0
24=x
120=x
5
6
0
0
24=x
120=x

6 2=x 4

6=x 5

8=x 4
1440
30=x 5
4
5

8=x
30=x
:
D x 0
D 6x
5
6 D

24=x
120=x
x9
0
0
24=x 5 120=x 6

240 .4/
y
x5

1200 000 720 00 1440 0


y C 7y C 8 y
x6
x
x

1440
y D 0;
x9

which is equivalent to x 4 y .4/ C 5x 3y 000 3x 2y 00 6xy 0 C 6y D 0.


(h)

x
1
x ln jxj
1=x
x 2
ln jxj
1=x 2 x

2
1 ln jxj C 1
1 ln jxj C 1
1=x
2x
1=x 2 2x
P0 .x/ D
D x
3

1=x
2=x
2
1=x
2=x 3 2
0
0
2
4
2
0

1=x
6=x
0
0
1=x
6=x 4 0

1 ln jxj

1=x 2 x

1
2=x 2 x
2
0

1
2=x
x
D x 1=x
2=x 3 2
D x

1=x
2=x 3 2
2
0

1=x
6=x 4 0
0
1=x 2
6=x 4 0

1
2=x 2 x
2=x 2
x

4=x 3
3
3

2=x
2 Dx 0
4=x
1 D x
D x 1=x
8=x 4
1=x 2
0
6=x 4 0
8=x 4 1=x
P1 .x/

x
1
0
0

0
x
0
0

x 0
0

x ln jxj
ln jxj C 1
1=x
2=x 3
ln jxj
1
1=x
2=x 3
2=x 2
4=x 3
28=x 5

1=x
1=x 2
2=x 3
24=x 5

1=x 2
2=x 2
2=x 3
24=x 5

x2
2x
2
0

x
x
2
0

D
1

2=x 2

12
1
D 3I

1=x
x

1
ln jxj
1=x 2

1=x 2
D x 1 ln jxj C 1

0
1=x
2=x 3

0
2=x
24=x 5

1
2=x 2 x

x 1=x
2=x 3 2
2=x 3 24=x 5 0

4=x 3
x
28=x 5

36
1

2 D 4I
2=x
x

x
2x
2
0

167

168 Chapter 9 Linear Higher Order Equations

P2 .x/

1
x
x ln jxj
1=x
x 2
ln jxj

1 ln jxj C 1
1 ln jxj C 1
1=x
2x
D x
0
1=x 2
6=x 4 0
1=x 2
0
3
5

0
2=x
24=x
0
0
2=x 3

1 ln jxj

1=x 2 x

1
2=x 2
2
0

1
2=x
x
2
D x 1=x
6=x 4
D x
2
4

0
1=x
6=x
0
3

2=x
24=x 5
0 2=x 3 24=x 5 0

2=x 2
x

8=x 4
1=x
4

8=x
1=x D x
D x 0
D
28=x 5
2=x 2
0 28=x 5
2=x 2

P3 .x/

x x ln jxj
1=x
x 2
0
1=x
2=x 3
2
D
2
4
0
1=x
6=x
0
0 2=x 3 24=x 5 0

1=x

2=x 3
2

x 0
4=x
2=x D
0
20=x 5
4=x 2

P4 .x/

D
D

Therefore,

4=x 4

20=x 5

which is equivalent to x 4 y .4/ C 3x 2y 000


(i)
x
e
x
e
P0 .x/ D x
ex
e

2x 0
D e
2
0

D 2e 2x

2x
D
2e

12 00 24 0
y C 6y
x5
x

x 2y 00 C 2xy 0
e
e
e
e

x
x
x
x

x
1
0
0

1 x
1 1
1 0
1 0
1
0
1

1
x
0

2=x 3
6=x 4
24=x 5

2
0
0

24
y D 0;
x7

2y D 0.

1
e 2x

2e 2x
2x 1
De
4e 2x
1
1
8e 2x

2
1
1

2
0
1
D e 2x
4
0
0
0
8
1

1
2

3 D 2e 2x 0
0
8

x 3
D 6e 2x .2x
1 6

1/I

1 x
1 1
1 0
1 0
x
1
x
0
1
x
1

2
0
0

24
2=x
D 6I
4=x 2
x

12 .4/ 36 000
y C 4y
x3
x

x
2x
0
0

12
I
x5

2=x 3
6=x 4
24=x 5

1=x

x 1=x 2
2=x 3

1 ln jxj C 1
1=x 2 2x
1=x
0
1=x
2=x 3 2
D 1=x 2
2
4

0
1=x
6=x
0
2=x 3
0
2=x 3
24=x 5 0

1=x 2
24
6=x 4

2
D
:
2=x 3 24=x 5
x7

1=x 2
1=x 2
6=x 4
24=x 5

x
0
0

2
3
6

1
2
4
8

1
2
3
8

Section 9.1 Introduction to Linear Higher Order Equations

P1 .x/

ex
ex
ex
ex

D e 2x

P2 .x/

P3 .x/

P4 .x/

e
e
e
e

0
0
2
0

D
Therefore,
6e 2x .2x

e
e
e
e

x
1
0
0

x
x
x

0
2
2
0

2x
2e 0
0

ex
ex
ex
ex

2x
e

ex
ex
D
ex
ex

2
D
e 2x
0
0

2x
D 2e 0
0

1 x
1 1
1 0
1 0

x
x
x
x

x
1
x
x
0
0
0

1 x
1 0
1 0
1 0
ex
ex
ex
ex

2x
e

x
x
x
x

x
1
0
0

1
1
1
1

x
1
0
0

x
x
x

1
3
15

1
e 2x
1 x 1

2x
1
2e
1
1 2
2x
D e
1
4e 2x
1
0 4

1
16e 2x
1 0 16

1
1 x 1

2
D 2e 2x 1 0 4

4
1 0 16
16

3
2x
D 2e 2x x

x 15 D 24xe I

e 2x

2x

2e
2x

2x D e

8e

2x

16e

0
1

2
2x 2
D
e

0
8

0
16

2x

6 D 2e
15

e 2x

4e 2x
2x
D e
8e 2x

16e 2x

1
1

4
2x
D 2e 0

8
0
16

e x
e x
e x
e x

0
2
0
2

2e 2x

1/y .4/

e
e
e
e

1
1
1
1
1
1
0

1
2
8
16

x 1
1

1 2
2x
D
2e

0
1 6
1
0 16

6
D 6e 2x .5 2x/I
15

1
1
1
1

1 x
1 1
1 0
1 0

1
1
0
1
1
x
1
1
1
1

1 x
1 0
1 0
1 0

x
x
x

1 2e 2x

0 4e 2x
2x
De
0 8e 2x

0 16e 2x

0
1 2

0 4
2x 2
De
0 8
0
0
0 16

1 2

0 8 D 2e 2x

0 12

24xe 2x y 000 C 6e 2x .5

169

1
4
8
16

D 2e 2x

1
3
15
1
1
1
1

1
1
1
1

1
1
1
0
1
0
0

1
0
0
0
1
1
0

x
x

x
1
0

3
D 24xe 2x I
15

1 2
0 4
0 8
0 16

2
4
8
12

2
6 D 24e 2x :
12

2x/y 00 C 24xe 2x y 0

1
6
16

24e 2x y D 0;

170 Chapter 9 Linear Higher Order Equations


1/y .4/

which is equivalent to .2x

4xy 000 C .5

2x/y 00 C 4xy 0

4y D 0.

9.1.24. (j)

P0 .x/

P1 .x/

P2 .x/

P3 .x/

e 2x
2e 2x
4e 2x
8e 2x
2
4
8
e 2x
2e 2x
4e 2x
16e 2x
2
4
16

2x

1
2x
0
2x
0
2x
0

2 2x 2
4 2 D 0
8 0 0

x2
2x
2
0

e
2e
4e
16e

2x

2
8
0


1 x 2
2x
0 2x
D
2x
0 0
2x
0 0

2 2x
8
8

8 0 D 2x
16
16
16 0

e 2x
4e 2x
8e 2x
16e 2x

e
2e
8e
16e

2x

e 2x
1 x2
2x
4e
0 2
8e 2x 0 0
16e 2x 0 0

4
4 2
8

8
8 0 D 2
16
16 16 0

2e 2x
2e 2x

2x
4e
4e 2x
P4 .x/ D
2x
8e 2x
8e 2x
16e
16e 2x

1 1 x 2
2 0 2x
4 0 2
8 0 0

2x
2 4x D 128x
8x

1
2
4
8

1
1 x 2

2x
2
0 2x
D
2x
0 2
4
2x
16
0 0

2
2 2x
2
2x

4 2 D 0
8 2 4x
0
16 0
0
8

e 2x
2e 2x
8e 2x
16e 2x
2
8
16

e
2e
4e
8e

1 1 x2
2 0 2x
4 0 2
16 0 0

D 128I

x2
2
0
0

x2
2x
0
0

1
1 1
2
2 0
8
8 0
16 16 0

D 512xI

1
4
8
16

1
4
8
16

1
0
0
0

8
D 512I
16

0 2x
0 2
D 0:
0 0
0 0

Therefore, 128xy .4/ C128y 000 C512xy 00 512y D 0, which is equivalent to xy .4/ y 000 4xy 00 C4y 0 D 0.
9.2 HIGHER ORDER CONSTANT COEFFICIENT HOMOGENEOUS EQUATIONS
9.2.2. p.r / D r 4 C 8r 2
9.2.4. p.r / D 2r 3 C 3r 2

9 D .r
2r

1/.r C 1/.r 2 C 9/; y D c1e x C c2 e


3 D .r

C c3 cos 3x C c4 sin 3x.

1/.r C 1/.2r C 3/; y D c1 e x C c2 e

C c3 e

3x=2

Section 9.2 Higher Order Constant Coefficient Homogeneous Equations


9.2.6. p.r / D 4r 3

8r 2 C 5r

1 D .r

171

1/2 ; y D c1 e x C e x=2 .c2 C c3 x/.

1/.2r

9.2.8. p.r / D r 4 C r 2 D r 2.r 2 C 1/; y D c1 C c2 x C c3 cos x C c4 sin x.


p
p
9.2.10. p.r / D r 4 C 12r 2 C 36 D .r 2 C 6/2 ; y D .c1 C c2x/ cos 6x C .c3 C c4 x/ sin 6x.
9.2.12. p.r / D 6r 4 C 5r 3 C 7r 2 C 5r C 1 D .2r C 1/.3r C 1/.r 2 C 1/; y D c1 e
c3 cos x C c4 sin x.
9.2.14. p.r / D r 4 4r 3 C 7r 2 6r C 2 D .r
9.2.16. p.r / D r 3 C 3r 2
y
y0
y 00

3 D .r

D c1 e x C c2 e
D c1 e x c2 e
D c1 e x C c2 e

y
y0
y 00

2r

4 D .r

x
x
x

C c3e 3x
3c3 e 3x I
C 9c3e 3x
x

3x

5e

c1 C c2 C c3
c1 c2 3c3
c1 C c2 C 9c3

0
14 I
40

2/.r 2 C 2r C 2/;
D
D
D

6
3 I
22

e 2x .c1 C c2 x C c3x 2 /
c1
e 2x .2c1 C c2 C .2c2 C 2c3/x C 2c3 x 2/
I
2c1 C c2
2e 2x .2c1 C 2c2 C c3 C 2.c2 C 2c3/x C 2c3x 2 / 4c1 C 4c2 C 2c3

D
D
D

1
1
4

6r 2 C 12r

cos x

8 D .r

c1 D 1, c2 D 3, c3 D 2; y D e 2x .1
9.2.22. p.r / D 8r 3
y
y0
y 00

4r 2

3e

3x C 2x 2 /.
1/2 ;

2r C 1 D .2r C 1/.2r

D e x=2.c1 C c2x/ C c3e x=2


D 12 e x=2.c1 C 2c2 C c2 x/ 21 c3 e
D 14 e x=2.c1 C 4c2 C c2 x/ C 14 c3 e
2x/ C 3e

9.2.24. p.r / D r 4

8 D .r

D
D
D
D

sin x C 4e 2x .

2/3 ;

c1 D 1, c2 D 2, c3 D 3; y D e x=2.1

y
y0
y 00
y 000

D
D
D

c1 C c3
c1 C c2 C 2c3
2c2 C 4c3

9.2.20. p.r / D r 3
D
D
D

x=3

6r 3 C 7r 2 C 6r

1/2 .r 2 2r C 2/; y D e x .c1 C c2x C c3 cos x C c4 sin x/.

D e x .c1 cos x C c2 sin x/ C c3e 2x


D
e x ..c1 c2 / cos x C .c1 C c2 / sin x/ C 2c3e 2x I
D e x .2c1 sin x 2c2 cos x/ C 4c3e 2x

c1 D 2, c2 D 3, c3 D 4; y D 2e

y
y0
y 00

C c2 e

1/.r C 1/.r C 3/;

c1 D 2, c2 D 3, c3 D 5; y D 2e x C 3e
9.2.18. p.r / D r 3

x=2

x=2

1/.r

x=2
x=2

c1 C c3
1
1
c
C
c
c
2
2 1
2 3
1
1
c
C
c
C
c
2
4 1
4 3

4
3 I
1

.
2/.r

4/.r C 1/;

c1e x C c2 e 2x C c3 e 4x C c4e x
c1 C c2 C c3 C c4
c1e x C 2c2 e 2x C 4c3e 4x c4 e x
c1 C 2c2 C 4c3 c4
I
c1e x C 4c2 e 2x C 16c3e 4x C c4e x
c1 C 4c2 C 16c3 C c4
c1e x C 8c2 e 2x C 64c3e 4x c4 e x
c1 C 8c2 C 64c3 c4

c1 D 4, c2 D 1, c3 D 1, c4 D 2; y D 4e x C e 2x

D
D
D

e 4x C 2e

D
D
D
D

2
8
I
14
62

172 Chapter 9 Linear Higher Order Equations


9.2.26. p.r / D r 4 C 2r 3
y
y0
y 00
y 000

D
D
D
D

2r 2

8 D .r

8r

2/.r C 2/.r 2 C 2r C 2/;

c1 e 2x C c2e 2x C e x .c3 cos x C c4 sin x/


2c1e 2x 2c2e 2x e x ..c3 c4/ cos x C .c3 C c4 / sin x/
I
4c1e 2x C 4c2e 2x C e x .2c3 sin x 2c4 cos x/
8c1e 2x 8c2e 2x C e x ..2c3 C 2c4/ cos x C 2.c4 c3 / sin x/
2c1
8c1

c1 C c2 C c3
2c2 c3 C c4
4c1 C 4c2 2c4
8c2 C 2c3 C 2c4

D
D
D
D

5
2
I
6
8

c1 D 1, c2 D 1, c3 D 3, c4 D 1; y D e 2x C e 2x C e x .3 cos x C sin x/.


x

e
1 0 1
1 0
xe x
e 2x
x

x
2x

e .x C 1/ 2e I W .0/ D 1 1 2 D 1 1
9.2.28. (a) W .x/ D e
e x e x .x C 2/ 4e 2x
1 2 4
1 2
1:


cos 2x
1 0 1 1
sin 2x
e 3x


(b) W .x/ D 2 sin 2x 2 cos 2x 3e 3x I W .0/ D 0 2 3 D 0
4 cos 2x
4 0 9 4
4 sin 2x 9e 3x
26:

1
e x cos x
e x sin x
e x

(c) W .x/ D e x .cos x C sin x/ e x .cos x sin x/ e x I W .0/ D 1

0
2e x sin x
2e x cos x
ex

1 2

2 1 D 5:

1 x x2 ex
1 0 0 1

0 1 2x e x
0 1 0 1

(d) W .x/ D
I
W
.0/
D
x
0 0 2 1 D 1:
0 0 2 ex

0 0 0 1
0x 0 0x e

e
e
cos x
sin x
x
x
e
e
sin x
cos x
(e) W .x/ D x
I
x
e
cos x
sin x
ex
x
e

e
sin x
cos x

1
1
0 1
1
1
0

1
1

1
0
1 1
1
0
1
1
W .0/ D
D
D
2

1
1
0 1
1
1
0
1
1
1

1
0
1
0
0
0
2

1
1

1
1

1
0 D 4
D
2 1
D 8:
1
1
0
0
2

cos x
sin x
e x cos x
e x sin x

x
x

sin x
cos x
e .cos x sin x/
e .cos x C sin x/
(f) W .x/ D
x
cos
x
sin
x
2e
sin
x
2e x cos x

x
x

sin x
cos x
e .2 cos x C 2 sin x/ e .2 cos x 2 sin x/

D 1 1 D

2 3

0 0
2 3
2 3 D
D
0 13
0 13

0 1 1
0 1
1 1 D 0
1 2 D
2 1 0
2 1
0
1
3

1
1
1

1
0
1

Section 9.2 Higher Order Constant Coefficient Homogeneous Equations

W .0/


1
0
1 0

0
1
1 1

D
1
0
0 2

0
1
2 2


1
1 1 1 1

0
1 2 D 0 1

0
1 3 0 0

1
0
0
0

0
1
0
1

1
2
5

D 5:

1
1
1
2

0
1
2
2

1
0
1

173

1 1
1 2
2 2

9.2.40. (a) Since y D Q1 .D/P1 .D/y C Q2 .D/P2 .D/y and P1 .D/y D P2 .D/y D 0, it follows that
y D 0.
(b) Suppose that (A) a1 u1 C    C ar ur C b1v1 C    C bs vs D 0, where a1 ; : : : ; ar and b1 ; : : : ; bs are
constants. Denote u D a1 u1 C    C ar ur and v D b1 v1 C    C bs vs . Then (B) P1 .D/u D 0 and (C)
P2 .D/v D 0. Since uCv D 0, P2 .D/.uCv/ D 0. Therefore,0 D P2 .D/.uCv/ D P2 .D/uCP2 .D/v.
Now (C) implies that P2 .D/u D 0. This, (B), and (a) imply that u D a1 u1 C    C ar ur D 0, so
a1 D    D ar D 0, since u1 ; : : : ; ur are linearly independent. Now (A) reduces to b1 v1 C  Cbs vs D 0,
so b1 D c    D bs D 0, since v1 ; : : : ; vs are linearly independent. Therefore,u1 ; : : : ; ur ; v1; : : : ; vr are
linearly independent.
(c) It suffices to show that fy1 ; y2 ; : : : ; yn g is linearly independent. Suppose that c1 y1 C  Ccnyn D 0.
We may assume that y1 ; : : : ; yr are linearly independent solutions of p1 .D/y D 0 and yr C1 ; : : : ; yn are
solutions of P2 .D/ D p2 .D/    pk .D/y D 0. Since p1 .r / and P2 .r / have no common factors, (b)
implies that (A) c1 y1 C    C cr yr D 0 and (B) cr C1 yr C1 C    C cn yn D 0. Now (A) implies that
c1 D    D cr D 0, since y1 ; : : : ; yr are linearly independent. If k D 2, then yr C1 ; : : : ; yn are linearly
independent, so cr C1 D    D cn D 0, and the proof is complete. If k > 2 repeat this argument, starting
from (B), with p1 replaced by p2 , and P2 replaced by P3 D p3    pn .
9.2.42. (a)
.cos A C i sin A/.cos B C i sin B/

D .cos A cos B

sin A sin B/

C.cos A sin B C sin A cos B/


D cos.A C B/ C i sin.A C B/:

(b) Obvious for n D 0. If n D 1 write


1
cos  C i sin 

D
D

1
cos  i sin 
cos  C i sin  cos  i sin 
cos  i sin 
D cos  i sin  D cos. / C i sin. /:
cos2  C sin2 

(d) If n is a negative integer, then (B) .cos  C i sin /n D

1
. From the hint, (C)
.cos  C i sin /jnj

1
D .cos  i sin /jnj D .cos. / C i sin. //jnj . Replacing  by  and n by
.cos  C i sin /jnj
jnj in (A) shows that (D) .cos. / C i sin. //jnj D cos. jnj/ C i sin. jnj/. Since jnj D n, (E)
cos. jnj/ C i sin. jnj/ D cos n C i sin n. Now (B), (C), (D), and (E) imply (A).
(e) From (A), nk D cos 2k C i sin 2k D 1 and kn D cos.2k C 1/ C i sin.2k C 1/ D cos.2k C
1/ D cos  D 1.
(f) From (e),  1=n 0 ; : : : ;  1=nn 1 are all zeros of n . Since they are distinct numbers, n  has
the stated factoriztion.

174 Chapter 9 Linear Higher Order Equations


From (e),  1=n0 ; : : : ;  1=nn
the stated factoriztion.

are all zeros of n C . Since they are distinct numbers, n C  has

2k
2k
1 D .r 0 /.r 1 /.r 2 / where k D cos
C i sin
, k D 0; 1; 2.
3
!
p
p
3

1
3
1
3
1 2 3
Ci
, and 2 D
i
. Therefore, p.r / D .r 1/
rC
C
,
Hence, 0 D 1, 1 D
2
2
2
2
2
4
(
p !
p !)
3
3
so e x ; e x=2 cos
x ; e x=2 sin
x
is a fundamental set of solutions.
2
2
.2k C 1/
.2k C 1/
(b) p.r / D r 3 C1 D .r 0 /.r 1 /.r 2 / where k D cos
Ci sin
, k D 0; 1; 2.
3
!
p
p
3
2
1
3
1
3
1
3
, 1 D 1 2 D
i
. Therefore, p.r / D .r C 1/
r
C
, so
Hence, 0 D C i
2
2
2
2
2
4
(
p !
p !)
3
3
e x ; e x=2 cos
x ; e x=2 sin
x
is a fundamental set of solutions.
2
2
p
p
p
p
.2k C 1/
(c) p.r / D r 4 C64 D .r 2 20 /.r 2 21 /.r 2 22 /.r 2 23 /, where k D cos
C
4
1Ci
1Ci
1 i
1 i
.2k C 1/
i sin
, k D 0; 1; 2; 3. Therefore, 0 D p , 1 D p , 2 D p , and 3 D p ,
4
2
2
2
2
so p.r / D ..r 2/2 C 4/..r C 2/2 C 4/ and fe 2x cos 2x; e 2x sin 2x; e 2x cos 2x; e 2x sin 2xg is a
fundamental set of solutions.
2k
2k
(d) p.r / D r 6 1 D .r 0 /.r 1 /.r 2 /.r 3 /.r 4 /.r 5 / where k D cos
Ci sin
,
6
6
p
p
p
1
3
1
3
1
3
k D 0; 1; 2; 3; 4; 5. Therefore, 0 D 1, 1 D C i
, 2 D
Ci
, 3 D 1, 4 D
i
,
2
2
2
2
2
2
!
!
p

2

2
1
3
1
3
1
3
and 5 D
i
, so p.r / D .r 1/.r C 1/
r
C
rC
C
and
2
2
2
4
2
4
(
!
!
!
p
p
p
p !)
3
3
3
3
x
x
x=2
x=2
x=2
x=2
e ; e ; e cos
x ; e sin
x ;e
cos
x ;e
sin
x
is a funda2
2
2
2
mental set of solutions.
(e) p.r / D r 6 C 64 D .r
20 /.r
21 /.r
22/.r
23/.rp 24/.r
25 / where
p k D
.2k C 1/
.2k C 1/
3 i
3 i
cos
Ci sin
, k D 0; 1; 2; 3; 4; 5. Therefore, 0 D
C , 1 D i , 2 D
C ,
6
6
2
2
2
2
p
p
p
p
3 i
3 i
3 D
, 4 D i , and 5 D
, so p.r / D .r 2 C 4/..r
3/2 C 1/..r C 3/2 C 1/ and
2
2 p
2
2p
p
p
fcos 2x; sin 2x; e 3x cos x; e 3x sin x; e 3x cos x; e 3x sin xg is a fundamental set of solutions.
(f) p.r / D .r 1/6 1 D .r 1 0 /.r 1 1 /.r 1 2 /.r 1 3 /.r 1p 4 /.r 1 5 / where
p k D
2k
2k
1
3
1
3
cos
C i sin
, k D 0; 1; 2; 3; 4; 5. Therefore, 0 D 1, 1 D C i
, 2 D
Ci
, 3 D
6
6
2
! 2
!
p
p
2
22

2
1
3
1
3
3
3
1
3
1, 4 D
i
, and 5 D
i
, so p.r / D r .r 2/
r
C
r
C
2
2
2
2
2
4
2
4
(
p !
p !
p !
p !)
3
3
3
3
and 1; e 2x ; e 3x=2 cos
x ; e 3x=2 sin
x ; e x=2 cos
x ; e x=2 sin
x
is a funda2
2
2
2
mental set of solutions.
r6 1
(g) p.r / D r 5 C r 4 C r 3 C r 2 C r C 1 D
. Therefore, from the solution of (d) p.r / D
r 1
9.2.43. (a) p.r / D r 3

Section 9.3 Undetermined Coefficients for Higher Order Equations


! 
!

3
1 2 3
C
rC
C
and
4
2
4
p !
p !
p !
3
3
3
x
x=2
x=2
x=2
e ;e
cos
x ;e
sin
x ;e
cos
x ;e
2
2
2
tal set of solutions.

.r C 1/
r
(

1
2

2

x=2

sin

3
x
2

!)

175

is a fundamen-

9.3 UNDETERMINED COEFFICIENTS FOR HIGHER ORDER EQUATIONS


9.3.2. If y D u 3x , then y 000 2y 00 5y 0 C 6y D e 3x .u000 11u00 C 34u0 24u/ 2.u00 6u0 C
9u/ 5.u0 3u/ C 6u D e 3x .u000 11u00 C 34u0 24u/. Let up D A C Bx C Cx 2 , where
. 24A C 34B 22C / C . 24B C 68C /x 24Cx 2 D 32 23x C 6x 2 . Then C D 1=4, B D 1=4,
e 3x
.3 x C x 2/.
A D 3=4 and yp D
4
9.3.4. If y D ue 2x , then y 000 C 3y 00 y 0 3y D e 2x .u000 6u00 C 12u0 8u/ C 3.u00 4u0 C 4u/
.u0 2u/ 3u D e 2x .u000 3u00 u0 C 3u/. Let up D A C Bx C Cx 2 , where .3A B 6C / C
.3B 2C /x C 3Cx 2 D 2 17x C 3x 2 . Then C D 1, B D 5, A D 1, and yp D e 2x .1 5x C x 2 /.
9.3.6. If y D ue x , then y 000 Cy 00 2y D e x .u000 C3u00 C3u0 Cu/C.u00 C2u0 Cu/ 2u D e x .u000 C4u00 C
5u0 /. Let up D x.ACBx CCx 2/, where .5AC8B C6C /C.10B C24C /x C15Cx 2 D 14C34x C15x 2.
Then C D 1, B D 1, A D 0, and yp D x 2 e x .1 C x/.
9.3.8. If y D ue x , then y 000 y 00 y 0 C y D e x .u000 C 3u00 C 3u0 C u/ .u00 C 2u0 C u/ .u0 C u/ C u D
e x .u000 C 2u00 /. Let up D x 2.A C Bx/ where .4A C 6B/ C 12Bx D 7 C 6x. Then B D 1=2, A D 1,
x2ex
and yp D
.2 C x/.
2
9.3.10. If y D ue 3x , then y 000 5y 00 C3y 0 C9y D e 3x .u000 C9u00 C27u0 C27u/ 5.u00 C6u0 C9u/C3.u0 C
3u/C9u D e 3x .u000 C4u00 /. Let up D x 2 .ACBxCCx 2/, where .8AC6B/C.24B C24C /xC48Cx 2 D
22 48x 2 . Then C D 1, B D 1, A D 2, and yp D x 2 e 3x .2 C x x 2 /.
9.3.12. If y D ue x=2 , then 8y 000

12y 00 C 6y 0

y D e x=28.u000 C 3u00=2 C 3u0=4 C u=8/ 12.u00 C u0 C


1 C 4x
u=4/ C 6.u0 C u=2/ u D 8e x=2u000 , so u000 D
. Integrating three times and taking the constants
8
x3
x 3 e x=2
of integration to be zero yields up D
.1 C x/. Therefore, yp D
.1 C x/.
48
48
9.3.14. If y D ue 2x , then y .4/ C3y 000 Cy 00 3y 0 2y D e 2x .u.4/ C8u000 C24u00 C32u0 C16u/C3.u000 C
6u00 C 12u0 C 8u/ C .u00 C 4u0 C 4u/ 3.u0 C 2u/ 2u D e 2x .u.4/ C 11u000 C 43u00 C 69u0 C 36u/. Let
up D A C Bx where .36A C 69B/ C 36Bx D 33 36x. Then B D 1, A D 1, and yp D e 2x .1 x/.
9.3.16. If y D ue x , then 4y .4/ 11y 00 9y 0 2y D e x 4.u.4/ C 4u000 C 6u00 C 4u0 C u/ 11.u00 C
2u0 C u/ 9.u0 C u/ 2u D e x .4u.4/ C 16u000 C 13u00 15u0 18u/. Let up D A C Bx where
ex
.18A C 15B/ 18Bx D 1 C 6x. Then B D 1=3, A D 1=3, and yp D
.1 x/.
3
9.3.18. If y D ue x , then y .4/ 4y 000 C6y 00 4y 0 C2y D e x .u.4/ C4u000 C6u00 C4u0 Cu/ 4.u000 C3u00 C
3u0 C u/ C 6.u00 C 2u0 C u/ 4.u0 C u/ C 2u D e x .u.4/ C u/. Let up D A C Bx C Cx 2 C Dx 3 C Ex 4
where .A C 24E/ C Bx C Cx 2 C Dx 3 C Ex 4 D 24 C x C x 4. Then E D 1 D D 0, C D 0 B D 1,
A D 0, and yp D xe x .1 C x 3 /.
9.3.20. If y D ue 2x , then y .4/ C y 000 2y 00
.u000 C 6u00 C 12u0 C 8u/ 2.u00 C 4u0 C 4u/

6y 0 4y D e 2x .u.4/ C 8u000 C 24u00 C 32u0 C 16u/ C


6.u0 C 2u/ 4u D e 2x .u.4/ C 9u000 C 28u00 C 30u0 /. Let

176 Chapter 9 Linear Higher Order Equations


up D x.AC Bx C Cx 2 / where .30AC 56B C 54C /C .60B C 168C /x C 90Cx 2 D .4 C 28x C 15x 2 /.
xe 2x
.1 x 2 /.
Then C D 1=6, B D 0, A D 1=6, and yp D
6
9.3.22. If y D ue x , then y .4/ 5y 00 C 4y D e x .u.4/ C 4u000 C 6u00 C 4u0 C u/ 5.u00 C 2u0 C u/ C 4u D
e x .u.4/ C 4u000 C u00 6u0 /. Let up D x.A C Bx C Cx 2 / where . 6A C 2B C 24C / C . 12B C
xe x
.1 C x 2 /.
6C /x 18Cx 2 D 3 C x 3x 2 , so C D 1=6, B D 0, A D 1=6. Then yp D
6
9.3.24. If y D ue 2x , then y .4/ 3y 000 C 4y 0 D e 2x .u.4/ C 8u000 C 24u00 C 32u0 C 16u/ 3.u000 C
6u00 C 12u0 C 8u/ C 4.u0 C 2u/ D e 2x .u.4/ C 5u000 C 6u00 /. Let up D x 2 .A C Bx C Cx 2 / where
.12A C 30B C 24C / C .36B C 120C /x C 72Cx 3 D 15 C 26x C 12x 2. Then C D 1=6, B D 1=6,
x 2 e 2x
.3 C x C x 2 /.
A D 1=2, and yp D
6
9.3.26. If y D ue x , then 2y .4/ 5y 000 C 3y 00 C y 0 y D e x 2.u.4/ C 4u000 C 6u00 C 4u0 C u/ 5.u000 C
3u00 C 3u0 C u/ C 3.u00 C 2u0 C u/ C .u0 C u/ u D e x .2u.4/ C 3u000 /. Let up D x 3 .A C Bx/, where
x3ex
.18A C 48B/ C 72Bx D 11 C 12x. Then B D 1=6, A D 1=6, and yp D
.1 C x/.
6
9.3.28. If y D ue 2x , then y .4/ 7y 000 C 18y 00 20y 0 C 8y D e 2x .u.4/ C 8u000 C 24u00 C 32u0 C
16u/ 7.u000 C 6u00 C 12u0 C 8u/ C 18.u00 C 4u0 C 4u/ 20.u0 C 2u/ C 8u D e 2x .u.4/ C u000 /. Let
up D x 3 .A C Bx C Cx 2 / where .6A C 24B/ C .24B C 120C /x C 60Cx 2 D 3 8x 5x 2 . Then so
x 3 e 2x
C D 1=12 B D 1=12, A D 1=6, and yp D
.2 C x x 2 /.
12
9.3.30. If y D ue x , then y 000 C y 00 4y 0 4y D e x .u000 3u00 C 3u0 u/ C .u00 2u0 C u/
u/ 4u D e x .u000 2u00 3u0 /. Let up D .A0 C A1 x/ cos 2x C .B0 C B1 x/ sin 2x, where
8A1 14B1
14A1 C 8B1
8A0 14B0 15A1 8B1
14A0 C 8B0 C 8A1 15B1
Then A1 D 1, B1 D 1, A0 D 1, B0 D 1, and yp D e

.1

D
D
D
D

4.u0

22
6
1
1:

x/ cos 2x C .1 C x/ sin 2x.

9.3.32. If y D ue x , then y 000 2y 00 Cy 0 2y D e x .u000 C3u00 C3u0 Cu/ 2.u00 C2u0 Cu/C.u0 Cu/ 2u D
e x .u000 C u00 2u/. Let up D .A0 C A1 x C A2 x 2 / cos 2x C .B0 C B1 x C B2 x 2 / sin 2x where

6A0 8B0
8A0 6B0

6A2 8B2
8A2 6B2
6A1 8B1 24A2 C 8B2
8A1 6B1 8A2 24B2
12A1 C 4B1 C 2A2 C 12B2
4A1 12B1 12A2 C 2B2

D
D
D
D
D
D

4
3
5
5
9
6:

Then A2 D 0, B2 D 1=2; A1 D 1=2, B1 D 1=2; A0 D 1=1, B0 D 1=2; and yp D


.1

x C x 2 / sin 2x.

ex
.1 C x/ cos 2x C
2

9.3.34. If y D ue x , then y 000 y 00 C 2y D e x .u000 C 3u00 C 3u0 C u/ .u00 C 2u0 C u/ C 2u D


e x .u000 C 2u00 C u0 C 2u/. Since cos x and sin x satisfy u000 C 2u00 C u0 C 2u D 0, let up D x.A0 C

Section 9.3 Undetermined Coefficients for Higher Order Equations

177

A1 x/ cos x C .B0 C B1 x/ sin x where


4A1 C 8B1
8A1 4B1
2A0 C 4B0 C 4A1 C 6B1
4A0 2B0 6A1 C 4B1

D 4
D
12
D 20
D
12:

Then A1 D 1, B1 D 1; A0 D 1, B0 D 3; and yp D xe x .1 C x/ cos x C .3 C x/ sin x.


9.3.36. If y D ue 3x , then D e 3x .u000 C 9u00 C 27u0 C 27u/ 6.u00 C 6u0 C 9u/ C 18.u0 C 3u/ D
e 3x .u000 C 3u00 C 9u0 C 27u/. Since cos 3x and sin 3x satisfy u000 C 3u00 C 9u0 C 27u D 0, let up D
x.A0 C A1 x/ cos 3x C .B0 C B1 x/ sin 3x where
36A1 C 36B1
36A1 36B1
18A0 C 18B0 C 6A1 C 18B1
18A0 18B0 18A1 C 6B1
Then A1 D 1=12, B1 D 0; A0 D, B0 D 1=12; and yp D

D
D
D
D

3
3
2
3:

xe 3x
.x cos 3x C sin 3x/.
12

9.3.38. If y D ue x , then y .4/ 3y 000 C 2y 00 C 2y 0 4y D e x .u.4/ C 4u000 C 6u00 C 4u0 C u/


3.u000 C 3u00 C 3u0 C u/ C 2.u00 C 2u0 C u/ C 4.u0 C u/ C u D e x .u.4/ C u000 u00 C u0 2u/. Let
up D A cos 2x C B sin 2x where 18A 6B D 2 and 6A C 18B D 1. Then A D 1=12, B D 1=12,
ex
and yp D
.cos 2x sin 2x/.
12
9.3.40. If y D ue x , then y .4/ C 6y 000 C 13y 00 C 12y 0 C 4y D e x .u.4/ 4u000 C 6u00 4u0 C
u/ C 6.u000 3u00 C 3u0 u/ C 13.u00 2u0 C u/ C 12u0 u/ C 4u D e x .u.4/ C 2u000 C u00 /. Let
up D .A0 C A1 x/ cos x C .B0 C B1x/ sin x where

2B0 6A1
2A0 C 2A1

2B1
2A1
2B1
6B1

D
D
D
D

Then A1 D 1=2, B1 D 1=2, A0 D 1=2, B0 D 1, and yp D

1
1
4
5:
x

e
2

.1 C x/ cos x C .2

x/ sin x.

9.3.42. If y D ue x , then y .4/ 5y 000 C 13y 00 19y 0 C 10y D e x .u.4/ C 4u000 C 6u00 C 4u0 C u/ 5.u000 C
3u00 C 3u0 C u/ C 13.u00 C 2u0 C u/ 19.u0 C u/ C 10u D e x .u.4/ u000 C 4u00 4u0 /. Since cos 2x
and sin 2x satisfy u.4/ u000 C 4u00 4u0 D 0, let up D x.A cos 2x C B sin 2x/ where 8A 16B D 1
xe x
and 16A C 8B D 1. Then A D 3=40, B0 D 1=40, and yp D
.3 cos 2x sin 2x/.
40
9.3.44. If y D ue x , then y .4/ 5y 000 C 13y 00 19y 0 C 10y D e x .u.4/ C 4u000 C 6u00 C 4u0 C u/ 5.u000 C
3u00 C 3u0 C u/ C 13.u00 C 2u0 C u/ 19.u0 C u/ C 10u D e x .u.4/ u000 C 4u00 4u0 /. Since cos 2x
and sin 2x satisfy u.4/ u000 C 4u00 4u0 D 0, let up D x.A0 C A1 x/ cos 2x C .B0 C B1 x/ sin 2x/
where
16A1 32B1 D
8
32A1 C 16B1 D
4
8A0 16B0 40A1 12B1 D
7
16A0 C 8B0 C 12A1 40B1 D
8:

178 Chapter 9 Linear Higher Order Equations


Then A1 D 0, B1 D 1=4; A0 D 0, B0 D 1=4, and yp D

xe x
.1 C x/ sin 2x.
4

9.3.46. If y D ue 2x , then y .4/ 8y 000 C 32y 00 64y 0 C 64y C 4y D e 2x .u.4/ C 8u000 C 24u00 C 32u0 C
16u/ 8.u000 C6u00 C12u0 C8u/C32.u00 C4u0 C4u/ 64.u0 C2u/C64u D e 2x .u.4/ C8u00 C16u/. Since
cos 2x, sin 2x, x cos 2x, and x sin 2x satisfy u.4/ C 8u00 C 16u D 0, let up D x 2 .A cos 2x C B sin 2x/
x 2e 2x
.cos 2x sin 2x/.
where 32A D 1 and 32B D 1. Then A D 1=32, B D 1=32, and yp D
32
9.3.48. Find particular solutions of (a) y 000 4y 00 C 5y 0 2y D 4e x , (b) y 000 4y 00 C 5y 0 2y D e 2x ,
and (c) y 000 4y 00 C 5y 0 2y D 2 cos x C 4 sin x.
(a) If y D ue x , then y 000 4y 00 C5y 0 2y D e x .u000 C3u00 C3u0 Cu/ 4.u00 C2u0 Cu/C5.u0 Cu/ 2u D
x 000
e .u
u00 /. Let u1p D Ax 2 where 2A D 4. Then A D 2, and y1p D 2x 2 e x .
(b) If y D ue 2x , then y 000 4y 00 C 5y 0 2y D e 2x .u000 C 6u00 C 12u0 C 8u/ 4.u00 C 4u0 C 4u/ C
5.u0 C 2u/ 2u D e 2x .u000 C 2u00 C u0 /. Let u2p D x. Then y2p D xe 2x .
000
00
0
(c) If y3p D A cos xCB sin x, then y3p
4y3p
C5y3p
2y3p D .2AC4B/ cos xC. 4AC2B/ sin x D
2 cos x C 4 sin x if A D 1 and B D 0, so y3p D cos x.
From the principle of superposition, yp D 2x 2 e x C xe 2x cos x.
9.3.50. Find particular solutions of (a) y 000 y 0 D 2.1 C x/, (b) y 000 y 0 D 4e x , (c) y 000 y 0 D 6e x ,
and (d) y 000 y 0 D 96e 3x
000
0
(a) Let y1p D x.A C Bx/. Then y1p
y1p
D A 2Bx D 2.1 C x/ if A D 2 and B D 1; therefore
2
y1p D 2x C 2x .
(b) If y D ue x , then y 000 y 0 D e x .u000 C 3u00 C 3u0 C u/ .u0 C u/ D e x .u000 C 3u00 C 2u0 /. Let
u2p D 4x. Then y2p D 4xe x .
(c) If y D ue x , then y 000 y 0 D e x .u000 3u00 C 3u0 u/ .u0 u/ D e x .u000 3u00 C 2u0 /. Let
u2p D 3x. Then y2p D 6xe x .
000
0
(d) Since e 3x does not satisfy the complementary equation, let y4p D Ae 3x . Then y4p
y4p
D
3x
4x
24Ae . Let A D 4; then y4p D 4e .
From the principle of superposition, yp D 2x C x 2 C 2xe x 3xe x C 4e 3x
9.3.52. Find particular solutions of (a) y 000 C 3y 00 C 3y 0 C y D 12e x and (b) y 000 C 3y 00 C 3y 0 C y D
9 cos 2x 13 sin 2x.
(a) If y D ue 2x , then y 000 C 3y 00 C 3y 0 C y D e 2x .u000 3u00 C 3u0 u/ C 3.u00 2u0 C u/ C
3.u0 u/ C u D e x u000 . Let u000
1p D 12. Integrating three times and taking the constants of integration
to be zero yields u1p D 2x 3 . Therefore,y1p D 2x 3.
(b) Let y2p D A cos 2x C B sin 2x where 11A 2B D 9 and 2A 11B D 13. Then A D 1,
B D 1, and y2p D cos 2x C sin 2x.
From the principle of superposition, yp D 2x 3 e 2x cos 2x C sin 2x.
9.3.54. Find particular solutions of (a) y .4/ 5y 00 C 4y D 12e x , (b) y .4/ 5y 00 C 4y D 6e x , and (c)
y .4/ 5y 00 C 4y D 10 cos x.
(a) If y D ue x , then y .4/ 5y 00 C 4y D e x .u.4/ C 4u000 C 6u00 C 4u0 C u/ 5.u00 C 2u0 C u/ C 4u D
x .4/
e .u C 4u000 C u00 6u0 /. Let u1p D 2x. Then y1p D 2xe x .
(b) If y D ue x , then y .4/ 5y 00 C 4y D e x .u.4/ 4u000 C 6u00 4u0 C u/ 5.u00 2u0 C u/C 4u D
e x .u.4/ 4u000 C u00 C 6u0 /. Let u2p D x. Then y2p D xe x .
(c) Let y3p D A cos xCB sin x where 10A D 10 and 10B D 0. Then A D 1, B D 0, and y3p D cos x.
From the principle of superposition, yp D 2xe x C xe x C cos x.
9.3.56. Find particular solutions of (a) y .4/ C 2y 000
3y 00 4y 0 C 4y D e 2x .

3y 00

4y 0 C 4y D 2e x .1 C x/ and (b) y .4/ C 2y 000

Section 9.3 Undetermined Coefficients for Higher Order Equations

179

(a) If y D ue x , then y .4/ C 2y 000 3y 00 4y 0 C 4y D e x .u.4/ C 4u000 C 6u00 C 4u0 C u/ C 2.u000 C


3u00 C 3u0 C u/ 3.u00 C 2u0 C u/ 4.u0 C u/ C 4u D e x .u.4/ C 6u000 C 9u00 /. Let u1p D x 2 .A C Bx/
x2
where .18A C 36B/ C 54Bx D 2 C 2x. Then B D 1=27, A D 1=27, and y1p D
.1 C x/e x .
27
(b) If y D ue 2x , then y .4/ C 2y 000 3y 00 4y 0 C 4y D e 2x .u.4/ 8u000 C 24u00 32u0 C 16u/ C
2.u000 6u00 C 12u0 8u/ 3.u00 4u0 C 4u/ 4.u0 2u/ C 4u D e 2x .u.4/ 6u000 C 9u00 /. Let
x 2 2x
e
.
u2p D Ax 2 where 18A D 1. Then A D 1=18 and yp D
18
2
x
.2 C 2x/e x C 3e 2x .
From the principle of superposition, yp D
54
9.3.58. Find particular solutions of (a) y .4/ C 5y 000 C 9y 00 C 7y 0 C 2y D e x .30 C 24x/ and (b)
y .4/ C 5y 000 C 9y 00 C 7y 0 C 2y D e 2x .
(a) If y D ue x , then y .4/ C 5y 000 C 9y 00 C 7y 0 C 2y D e x .u.4/ 4u000 C 6u00 4u0 C u/ C 5.u000
3u00 C 3u0 u/ C 9.u00 2u0 C u/ C 7.u0 u/ C 2u D e x .u.4/ C u000 /. Let u1p D x 3 .A C Bx/ where
.6A C 24B/ C 24Bx D 30 C 24x. The B D 1, A D 1, and y1p D x 3 .1 C x/e x .
(b) If y D ue 2x , then y .4/ C 5y 000 C 9y 00 C 7y 0 C 2y D e 2x .u.4/ 8u000 C 24u00 32u0 C 16u/ C
5.u000 6u00 C 12u0 8u/ C 9.u00 4u0 C 4u/ C 7.u0 2u/ C 2u D e 2x .u.4/ 3u000 C 3u00 u0 /. Let
u2p D x. Then y2p D xe 2x .
From the principle of superposition, yp D x 3 .1 C x/e x C xe 2x .
9.3.60. If y D ue 2x , then y 000 y 00 y 0 Cy D e 2x .u000 C6u00 C12u0 C8u/ .u00 C4u0 C4u/ .u0 C2u/C
u D e 2x .u000 C 5u00 C 7u0 C 3u/. Let up D A C Bx, where .3A C 7B/ C 3x D 10 C 3x. Then B D 1,
A D 1 and yp D e 2x .1 C x/. Since p.r / D .r C 1/.r 1/2 , y D e 2x .1 C x/ C c1 e x C e x .c2 C c3x/
9.3.62. If y D ue 2x , then y 000 6y 00 C11y 0 6y D e 2x .u000 C6u00 C12u0 C8u/ 6.u00 C4u0 C4u/C11.u0 C
2u/ 6u D e 2x .u000 u0 /. Let up D x.ACBx CCx 2 / where . AC6C / 2Bx 3Cx 2 D 5 4x 3x 2.
Then C D 1, B D 2, A D 1, and yp D xe 2x .1 C x/2 . Since p.r / D .r 1/.r 2/.r 3/,
y D xe 2x .1 C x/2 C c1 e x C c2 e 2x C c3 e 3x .
9.3.64. If y D ue x , then y 000 3y 00 C 3y 0 y D e x .u000 C 3u00 C 3u0 C u/ 3.u00 C 2u0 C u/ C
3.u0 C u/ u D e x u000 . Let u000 D 1 C x. Integrating three times and taking the constants of integration
x3
x3ex
to be zero yields u D
.4 C x/. Therefore, yp D
.4 C x/. Since p.r / D .r 1/3 , y D
24
24
x3ex
.4 C x/ C e x .c1 C c2 x C c3 x 2 /.
24
9.3.66. If y D ue 2x , then y 000 C 2y 00 y 0 2y D e 2x .u000 6u00 C 12u0 8u/ C 2.u00 4u0 C 4u/
.u0 2u/ 2u D e 2x .u000 4u00 C 3u0 /. Let up D .A0 C A1 x/ cos x C .B0 C B1 x/ sin x where
4A1 C 2B1
2A1 C 4B1
4A0 C 2B0 8B1
2A0 C 4B0 C 8A1

D
D
D
D

2
9
23
8:




x
3
2x
Then A1 D 1=2, B1 D 2; A0 D 1, B0 D 3=2, and yp D e
1C
cos x C
2x sin x .
2


 2

x
3
Since p.r / D .r 1/.r C 1/.r C 2/, y D e 2x 1 C
cos x C
2x sin x C c1 e x C c2 e x C
2
2
c3 e 2x .

180 Chapter 9 Linear Higher Order Equations


9.3.68. If y D ue x , then y .4/ 4y 000 C 14y 00 20y 0 C 25y D e x .u.4/ C 4u000 C 6u00 C 4u0 C u/ 4.u000 C
3u00 C 3u0 C u/ C 14.u00 C 2u0 C u/ 20.u0 C u/ C 25u D e x .u.4/ C 8u00 C 16u/. Since cos 2x, sin 2x,
x cos 2x, and x sin 2x satisfy u.4/ C8u00 C16u D 0, let up D x 2.A0 CA1x/ cos 2x C.B0 CB1 x/ sin 2x
where
96A1 D 6
96B1 D 0
32A0 C 48B1 D 2
48A1 32B0 D 3:
Then A1 D
.r

1=16, B1 D 0; A0 D

1/2 C 12 , y D

1=16, B0 D 0, and y D

x2ex
.1 C x/ cos 2x. Since p.r / D
16

x2ex
.1 C x/ cos 2x C e x .c1 C c2 x/ cos 2x C .c3 C c4 x/ sin 2x.
16

9.3.70. If y D ue x , then y 000 y 00 y 0 C y D e x .u000 3u00 C 3u0 u/ .u00 2u0 C u/ .u0


u/ C u D e x .u000 4u00 C 4u0 /. Let up D x.A C Bx/, where .4A 8B/ C 8Bx D 4 C 8x.
Then B D 1, A D 1, and yp D x.1 C x/e x . Since p.r / D .r C 1/.r 1/2 the general solution is
y D x.1 C x/e x C c1 e x C c2 e x C c3 xe x . Therefore,
2
3 2
3 2
32
3
y
x.1 C x/e x
e x ex
xe x
c1
4 y 0 5 D 4 e x .x 2 x 1/ 5 C 4 e x e x e x .x C 1/ 5 4 c2 5 :
y 00
e x .x 2 3x/
e x e x e x .x C 2/
c3
Setting x D 0 and imposing the initial conditions yields
32
3
3 2
2 3 2
c1
2
0
1 1 0
4 0 5 D 4 1 5 C 4 1 1 1 5 4 c2 5 ;
c3
0
0
1 1 2

so c1 D 1, c2 D 1, c3 D 1, and y D e

.1 C x C x 2 / C .1

x/e x .

9.3.72. If y D ue x , then y 000 2y 00 5y 0 C 6y D e x .u.4/ 4u000 C 6u00 4u0 C u/ C 2.u000


3u00 C 3u0 u/ C 2.u00 2u0 C u/ C 2.u0 u/ C u D e x .u.4/ 2u000 C 2u00 /. Let up D x 2 .A C Bx/,
where .4A 12B/ C 12Bx D 20 12x. Then B D 1, A D 2, and yp D x 2.2 x/e x . Since
p.r / D .r C 1/2 .r 2 C 1/, the general solution is y D x 2 .2 x/e x C e x .c1 C c2x/C c3 cos x C c4 sin x.
Therefore,
2
3 2
3 2
32
3
y
x 2 .2 x/e x
e x
xe x
cos x
sin x
c1
6 y0 7 6
7 6 e x .1 x/e x
6
7
x.x 2 5x C 4/e x
sin x
cos x 7
6 00 7 D 6
7C6
7 6 c2 7
x
x
3
2
x
4 y 5 4 .x
5 4 e
8x C 14x 4/e
.x 2/e
cos x
sin x 5 4 c3 5
y 000
.x 3 11x 2 C 30x 18/e x
e x .3 x/e x
sin x
cos x
c4

Setting x D 0 and imposing the initial conditions yields


2
3 2
3 2
3
0
1
6
7 6
7 6 1
4
0
6
7D6
7C6
4
7 5 4
4 5 4 1
22
18
1

so c1 D 2, c2 D 1, c3 D 1, c4 D 1, and y D .2

0
1
2
3

1
0
1
0

x/.x 2 C 1/e

32
0
c1
6 c2
1 7
76
0 5 4 c3
1
c4

C cos x

7
7;
5

sin x.

9.3.74. If y D ue x , then y .4/ 3y 000 C 5y 00 2y 0 D e x .u.4/ C 4u000 C 6u00 C 4u0 C u/ 3.u000 C 3u00 C
3u0 C u/ C 4.u00 C 2u0 C u/ 2.u0 C u/ D e x .u.4/ C u000 C u00 C u0 /. Since cos x and sin x satisfy

Section 9.4 Variation of Parameters for Higher Order Equations

181

u.4/ C u000 C u00 C u0 D 0, let up D x.A cos x C B sin x/ where 2A 2B D 2 and 2A 2B D 2.


Then A D 1, B D 0, and yp D e x cos x. Since p.r / D r .r 1/.r 1/2 C 1 the general solution is
y D e x cos x C c1 C e x .c2 C c3 cos x C c4 sin x/. Therefore,
2
3 2
3 2
y
xe x cos x
1 ex
e x cos x
e x sin x
x
x
6 y 0 7 6 e x ..x C 1/ cos x x sin x/ 7 6 0 e x
e .cos x sin x/
e .cos x C sin x/
6 00 7 D 6
7 6
4 y 5 4 e x .2 cos x 2.x C 1/ sin x/ 5C4 0 e x
2e x sin x
2e x cos x
000
x
x
x
x
y
e .2x cos x C 2.x C 3/ sin x/
0 e
e .2 cos x C 2 sin x/ e .2 cos x 2 sin x/

Setting x D 0 and imposing the initial conditions yields


2
3 2
3 2
2
0
1 1
6 0 7 6 1 7 6 0 1
6
7 6
7 6
4 1 5 D4 2 5C4 0 1
5
0
0 1

1
1
0
2

32
0
c1
6 c2
1 7
76
2 5 4 c3
2
c4

so c1 D 2, c2 D 1, c3 D 1, c4 D 1, and 2 C e x .1 C x/ cos x

sin x

7
7;
5

1.

9.4 VARIATION OF PARAMETERS FOR HIGHER ORDER EQUATIONS

2
2
2

e x
xe x
x2e x

2
2
2
9.4.2. W D
D 2e 3x ;
2xe x
e x .1 2x 2 /
2xe x .1 x 2 /

e x2 .4x 2 2/ 2xe x2 .2x 2 3/ 2e x2 .2x 4 5x 2 C 1/

2
2

x2
2
x2
xe x
x2e x
2
e
x
e

2
2x
W1 D
D
x
e
;
W
D

D
2
2
2
2
2
e x .1 2x 2 / 2xe x .1 x 2 /
2xe x 2xe x .1 x 2 /

2
2

1
F W2
F W1
2
xe x
2
e x

2xe 2x ; W3 D
D x 5=2 ; u02 D
D
D e 2x ; u01 D
2
2
x
x
2
2xe
P0 W
2
P0 W
e
.1 2x /
p
F W2
2 5=2
x 3=2 ; u03 D
D x=2; u1 D x 7=2=7; u2 D
x ; u3 D x 3=2=3; yp D u1 y1 Cu2 y2 Cu3 y3 D
P0 W
5
8
2
e x x 7=2.
105

ex
e x

e x
ex

x
x

x
x

e
.x
1/
e
.x
C
1/
x
D 2=x 2 ; W1 D x x
9.4.4. W D 0

2
2
e x .x C 1/
e .x 1/
x
x

e x .x 2 2x C 2/ e x .x 2 C 2x C 2/
x2
x2
0

3
3
x
x

e x
ex

x
1
1
2
e
.x
C
1/
e x .x 1/ 0

x
x
;
W
D
D
;
W
D
D
; u1 D

x
x
2
3
e .x C 1/
e .x 1/

x2
x2
x2
0

2
2
x
x
F W1
F W2
F W2
x
x
0
0
D 2; u2 D
D e .x C 1/; u3 D
D e .x 1/; u1 D 2x; u2 D e x .x C 2/;
P0 W
P0 W
P0 W
u3 D e x .x 2/; yp D u1 y1 C u2 y2 C u3 y3 D 2.x 2 C 2/=x.

1
x
e

1
e
x

2
x
x
2.x
2/

e
x D e .x 1/ ; W2 D
9.4.6. W D e x
e x
1=x 2 D
;
W
D

1
1

e x
x3
x2
2
x

x
2
e
e

x
x3

32

3
c1
7 6 c2 7
76
7
5 4 c3 5
c4

182 Chapter 9 Linear Higher Order Equations

x
e

ex

1
x
1
x2

x
e
e x .x C 1/
x
;
W
D
3
e
x2

e x
D
e x

2; u01 D

F W1
D e
P0 W

.x

1/; u02 D

F W2
F W2
x2
D e x .x C 1/; u03 D
D 2x 2 ; yp D u1 y1 C u2 y2 C u3 y3 D 2 .
P0 W
P0 W
3

1
p

x
p
x2
p
1

2
p
x x2

p
x

5
15
x
1
1

; W1 D
; W2 D 1
9.4.8. W D
p
2x D
D
1

2 x

4x
2
2x 3=2
2px 2x

2x

3=2
1
3

2x
2

4x 5=2
4x 3=2
p

x
p
3=2

p
3x
x

D 1 ; u0 D F W1 D 5 x; u0 D F W2 D 3x 3=2; u0 D
; W3 D 1
1
2
3

1
2
x
P
W
P0 W
0
p

2 x

3=2
2x
F W2
2
10 3=2
6
D ; u1 D
x ; u2 D x 5=2 ; u3 D 2 ln jxj; yp D u1 y1 C u2 y2 C u3 y3 D 2x 2 ln jxj
P0 W
x
3
5
32 2
32 2
x . Since
x satisfies the complementary equation we take yp D ln jxj.
15
15

ex

x 1=x
1

ex

x
x

ex
2e
.1
x/
e
.x
1/
1

D
x
x
D
9.4.10. W D 1
;
W
D
;

x
1

3
1
e .x 1/

x
x2
x2
x2

x
2

2
e .x
2x C 2/
x2
x2

0
3
3
x
x

ex
1

x
e .x 2/
x

x
x D 2 ; u0 D F W1 D 1; u0 D F W2 D
W2 D
D
;
W
D

x
3
2
1
e .x 1/

1
x
x 1
P0 W
P0 W
1

2
2
x
x
F W2
x 2 .3 x/
x.2 x/; u03 D
D 2xe x ; u1 D x; u2 D
; u3 D 2e x .x C 1/; yp D u1 y1 C
P0 W
3
2x 3 C 3x 2 C 6x C 6
2
u2 y2 C u3 y3 D
. Since x C
satisfies the complementary equation we take
3x
x
2
2x C 6
yp D
.
3

x ex e x
x

e
x e x
e x
x
x

D e x .x C
9.4.12. W D 1 e
e
D 2; W2 D
D 2x; W1 D e x
e x
1
e x
0 ex e x

x ex

D e x .x 1/; u0 D F W1 D 1; u0 D F W2 D e x .x C 1/=2; u0 D F W2 D
1/; W3 D
1
2
3
1 ex
P0 W
P0 W
P0 W
x
x
x
e .x 1/=2; u1 D x; u2 D e .x C2/=2; u3 D e .x 2/=2; yp D u1 y1 Cu2 y2 Cu3 y3 D x 2 2.

Section 9.4 Variation of Parameters for Higher Order Equations


p
x

183

p
1= x

x 3=2
1
1

3=2
x
p
p

x 3=2
3=2
1
1
3 x
3

x
x
p

p
3=2
5=2
12

3
x
3
2
2 x
2x
2x
D 6 ; W1 D
D

1
3
3
15
5=2
2x 3=2

x
2
2x
p

3
15
3

4x 3=2
4x 5=2
4 x
4x 7=2
p

5=2
7=2
3
15
3
105
4x
4 x
4x

5=2
8x 7=2
8x 3=2 8x 9=2
8x p

1
1
1

3=2

x
x

x
p
3=2
3=2

x
x
x
p

1
3 x
3
6
6
1
1
3

p
;
W
D
D
;
W
D

D
p
2
3
5=2
7=2
5=2
3=2
5=2

2
x
2x
x
x
2 x
2x
2x

3
15
1
1
3
15

3=2
7=2
3=2
5=2
7=2
4x
4
x
4x
4x
4x
4x

x
p
x 3=2

2
F W2
F W1
2
1
1
3 x
0
0

; W4 D
p
D x 3=2 ; u1 D P W D 3x; u2 D P W D
3=2
x 9=2
2
2
x
2x
0
0

3
3
1

p
4x 3=2
4x 5=2
4 x
F W2
F W4
3x 2
x4
3x 2 ; u03 D
D 1; u04 D
D x 3 ; u1 D
; u2 D x 3 ; u3 D x; u4 D
;
P0 W
P0 W
2
4
x 5=2
yp D u1 y1 C u2 y2 C u3 y3 D
.
4

x x2 x3
x 4
x
x x3

x3
x 4
x4

1 2x 3x 2 4x 3
2
3
2
4
6

4x D 2x ; W2 D 1 3x
4x 3
9.4.16. W D
2 D 12x ; W1 D 2x 3x
0
2
6x
12x
2

2
0 6x 12x 2
6x 12x
0 0

6
24x

x x2
x x2 x3
x 4

F W1
x2
6x 5 ; W3 D 1 2x 4x 3 D 6x 4 ; W4 D 1 2x 3x 2 D 2x 3; u01 D
D
;
P0 W
6
0 2 12x 2
0 2
6x

9.4.14. W D

x
F W2
1 0
F W4
1
x3
x2
x
F W2
D ; u03 D
D
; u4 D
D
; u1 D
; u2 D
; u3 D
;
P0 W
2
P0 W
2
P0 W
6x
18
4
2
4
4
4
x ln jxj 11x
11x
ln jxj
u4 D
; yp D u1 y1 C u2 y2 C u3 y3 D
. Since
satisfies the complementary
6
6
36
36
4
x ln jxj
equation we take yp D
.
6

x x2
2

1=x
1=x 2

x
1=x
1=x 2
2
3

1 2x
1=x
2=x
D 72=x 6; W1 D 2x
1=x 2
2=x 3 D 12=x 4;
9.4.18. W D

2=x 3
6=x 4
3
0 2

2
2=x
6=x 4
0 0
6=x 4
24=x 5

x
x x 2 1=x 2
x x2
1=x
1=x 2
1=x

2
3
5
3
2
W2 D 1
1=x
2=x D 6=x ; W3 D 1 2x
2=x D 12=x ; W4 D 1 2x
1=x 2
3
4
4
0 2=x
0 2
0 2
6=x
6=x
2=x 3
F W1
F W2
F W2
F W4
6=x; u01 D
D 2; u02 D
D 1=x; u03 D
D 2x 2 ; u04 D
D x 3 ; u1 D 2x;
P0 W
P0 W
P0 W
P0 W
u2 D ln jxj; u3 D 2x 3 =3; u4 D x 4 =4; yp D u1 y1 C u2 y2 C u3 y3 D x 2 ln jxj 19x 2=12. Since
19x 2 =12 satisfies the complementary equation we take yp D x 2 ln jxj.

u02 D

184 Chapter 9 Linear Higher Order Equations

e 2x
x
e x e 2x

e =x

x
2x
x

e
.x
1/
e
.2x
1/
2x
e

2e
e 6x

2
2
x
x
9.4.20. W D
;
D
x 2
2x
2
e .x
2x C 2/
2e .2x
2x C 1/
x

x4
e

4e 2x

x3
x3

x 3
3x 2 C 6x 6/ 2e 2x .4x 3 6x 2 C 6x 3/
e x 8e 2x e .x

x4
x4

x
2x

2x
e
e

x
x

x
2x
e 5x

e
.x
1/
e
.2x
1/
W1 D 2e 2x
D 3 ;

x
x2
x2

x 2
2x
2
e
.x
2x
C
2/
2e
.2x
2x
C
1/

4e 2x
x3
x3

x
2x

e
e
ex

x
x

x
2x
e 4x

x
e .x 1/
e .2x 1/
W2 D e
;
D
2
2

x3
x
x

x 2
2x C 2/ 2e 2x .2x 2 2x C 1/
x e .x

e
3
x
x 3

2x

e
e x e 2x

2x
e 5x .2 x/
x
e
.2x
1/

2x
W3 D e
;
D
2e

x3
x2

2x
2
2e .2x
2x C 1/
x
4e 2x
e

3
x

x
e

x
e 2x

x
4x
x

e .x 1/
D e .x C 2/ ;
W4 D e x 2e 2x

2
x3
x

x 2

x
e
.x
2x
C
2/

e
4e 2x
x3
F W2
F W2
F W4
F W1
u01 D
D 3; u02 D
D 3e x ; u03 D
D 3.2 x/; u04 D
D 3e x .x C 2/;
P0 W
P0 W
P0 W
P0 W
u1 D 3x; u2 D 3e x ; u3 D 3x.4 x/=2; u4 D 3e x .x C 3/; yp D u1 y1 C u2 y2 C u3 y3 D
3e x .x 2 C 4x C 6/
3e x .2x C 3/
. Since
is a solution of the complementary equation we take yp D
2x
x
x
3xe
.
2

x x3
3

x ln x

x
x ln x
3
9.4.22. W D 1 3x 2 1 C ln x D 4x 2 ; W1 D
2x 3 ln x; W2 D
2
D x
3x
1
C
ln
x

0 6x
1=x

x x3
F W1
1 0
F W2
1
x
ln
x
3
0

D x; W3 D

1 1 C ln x
1 3x 2 D 2x ; u1 D P W D 2 ln x=x x ; u2 D P W D x 3 ;
0
0
F W2
2
1
2
0
u3 D
D
; u1 D .ln x/
ln x; u2 D
; u3 D 2 ln x; yp D u1 y1 C u2 y2 C u3 y3 D
P0 W
x
2x 2
x
x
x.ln x/2 x ln x
. Since x ln x
satisfies the complementary equation we take yp D x.ln x/2
2
2

Section 9.4 Variation of Parameters for Higher Order Equations

185

The general solution is y D x.ln x/2 C c1 x C c2 x 3 C c3 x ln x, so


3 2
32
3
2
3 2
x x3
x ln x
x.ln x/2
y
c1
2
2
1 C ln x 7 4 c 5
.ln x/
2 ln x 7 C 6 1 3x
4 y0 5 D 6
5 4
5
4
2
2 ln x
2
1
00
y
c3
0 6x
x
x
x
Setting x D 1 and imposing the initial conditions yields
2
3 2
3 2
32
3
4
0
1 1 0
c1
4 4 5 D 4 0 5 C 4 1 3 1 5 4 c2 5 :
2
2
0 6 1
c3

Solving this system yields c1 D 3, c2 D 1, c3 D 2. Therefore,y D x.ln x/2 C 3x C x 3 2x ln x.


x

2x
e

2x

xe x
x e 2x

xe x
x
2x

D e x .1 3x/;
9.4.24. W D e
2e
e .1 x/ D e .6x 5/; W1 D
2x
x
2e
e .1 x/
e x 4e 2x e x .x 2/
x

e
F W1
xe x
e 2x

W2 D x
D 1 2x; W3 D x
D e 3x ; u01 D
D 1 3x; u02 D
e
e x .1 x/
e
2e 2x
P0 W
F W2
x.2 3x/
e 2x
F W2
D e x .2x 1/; u03 D
D e 2x ; u1 D
; u2 D e x .2x C 1/; u3 D
;
P0 W
P0 W
2
2
x
e
yp D u1 y1 C u2 y2 C u3 y3 D e x .3x 2 C x C 2/=2. Since
is a solution of the complementary
2
x
e .3x C 1/x
equation we take yp D
.
2 x
e .3x C 1/x
The general solution is y D
C c1 e x C c2 e 2x C c3 xe x , so
2
2
3
e x .3x C 1/x
2
2
3 6
3
7 x
c1
2
e 2x
xe x
y
6
7 e
x
2

e
.3x
C
7x
C
1/
7
6
4 y0 5 D 6
7 C e x 2e 2x e x .1 x/ 4 c2 5 :
7
6
2
00
y
4 e x .3x 2 C 13x C 8/ 5 e x 4e 2x e x .x 2/ c3
2
Setting x D 0 and imposing the initial conditions yields
2
3 2
3 2
32
3
4
0
1 1
0
c1
3 5
4
1 5 4 c2 5 :
D 4 12 5 C 4 1 2
2
1 4
2
c3
19
4
Solving this system yields c1 D

4xe

x x2 ex

9.4.26. W D 1 2x e x
0 2 ex

x x2
e x .x 1/; W3 D
1 2x

3, c2 D

1, c3 D 4. Therefore, y D

e x .3x C 1/x
2

3e x

e 2x C

x ex
e x
x 2
D e x .x 2 2xC2/; W1 D x

2x e x D e .x 2x/; W2 D 1 e x D

D x 2 ; u0 D F W1 D x.x 2/; u0 D F W2 D 1 x; u0 D F W2 D
1
2
3

P0 W
P0 W
P0 W
4
2
2
x
C
6x
C
12x
C
12
6x
C
12x
x 2 e x ; yp D u1 y1 C u2 y2 C u3 y3 D
. Since
is a solution of the
6
6
4
x C 12
complementary equations we take yp D
.
6

186 Chapter 9 Linear Higher Order Equations


x 4 C 12
C c1x C c2 x 2 C c3 e x , so
6
3
2 4
3
x x2 ex
.x C 12/=6

3
5D 4
5 C 1 2x e x
2x =3

0 2 ex
2x 2

The general solution is y D


2

y
4 y0
y 00

Setting x D 0 and imposing the initial conditions yields


2 3
2
3 2
0
2
0 0
4 5 5 D 4 0 5C4 1 0
0
0
0 2

2
3
c1

4 c2 5 :

c3

32
3
1
c1
1 5 4 c2 5 :
1
c3

Solving this system yields c1 D 3, c2 D 1, c3 D 2. Therefore, y D

x 4 C 12
C 3x
6

x 2 C 2e x .

e x e 3x
e
x C 1 e 3x
e 3x
4x

e x 3e 3x D 2e 4x .3x 1/; W1 D x
D
2e
;
W
D
2
3x
1
e
3e
3e 3x
x
3x
e
9e

x C 1 ex
F W2
F W1
x
0

e 3x .3x C 2/; W3 D
D 2e x ; u02 D
D 3x 2;
x D xe ; u1 D
1
e
P0 W
P0 W
x.3x C 4/
e 2x .2x C 1/
F W2
D xe 2x ; u1 D 2e x ; u2 D
; u3 D
; yp D u1 y1 C
u03 D
P0 W
2
4
x
2
x
e .6x C 2x 7/
7e
u2 y2 C u3 y3 D
. Since
is a solution of the complementary equation we take
4
4
x
xe .3x C 1/
yp D
2
xe x .3x C 1/
The general solution is y D
C c1 .x C 1/ C c2 e x C c3e 2x , so
2

2
3 2
3
x C 1 e x e 3x
y
xe x .3x C 1/=2

4 y 0 5 D 4 e x .3x 2 C 7x C 1/=2 5 C 1
e x 3e 3x :

0
y 00
e x .3x 2 C 13x C 8/=2
e x 9e 3x

xC1

9.4.28. W D 1
0

Setting x D 0 and imposing the initial conditions yields


2 3 3 2
3 2
32
3
0
1 1 1
c1
4
4 5 5 D 4 1 5 C 4 1 1 3 5 4 c2 5 :
4
2
1
0 1 9
c3
4
4
Solving this system yields c1 D
e 2x
.
2

9.4.30. W D

2x

1
x
1
x2
2
x3
6
x2

1
, c2 D
2

1
1
, c3 D . Therefore, y D
4
2

x ln x

ln x C 1
D

x2

2
x

12
2x
;
W
D
1

x3

xe x .3x C 1/ x C 1
C
2
2

1
x
1
x2
2
x3

ln x
ln x C 1 D 6
x

ex
C
4

x ln x

3
;
x

Section 9.4 Variation of Parameters for Higher Order Equations

187

1
1

x ln x
2
x
x
x x
x ln x

x
x

4
1
1

; W3 D 1 2x ln x C 1 D x; W4 D 1 2x
D
ln x C 1 D
2
2
1

x
x2
x

0
2
2
1
2

0 2

x3
x
x3
F W1
9 0
F W2
3
F W2
3x 2 0
F W4
9
6 0
; u1 D
D 9 ln x=2
; u2 D
D ; u03 D
D
; u4 D
D
;
x
P0 W
4
P0 W
x
P0 W
4
P0 W
2
9x ln x 27x
x3
9x
u1 D
; u2 D 3 ln x; u3 D
; u4 D
; yp D u1 y1 Cu2 y2 Cu3 y3 D 3x 2 ln x 7x 2 .
2
4
4
2
Since 7x 2 satisfies the complementary equation we take yp D 3x 2 ln x.
c3
C c4 x ln x, so
The general solution is y D 3x 2 ln x C c1x C c2 x 2 C
x

1
2

x ln x 2
3 x x
2
3 2 3x 2 ln x
3
x

c1
y
1

6
7
ln x C 1 6 c 7
6 y 0 7 6 6x ln x C 3x 7 1 2x
x2
6 00 7 D 6 6 ln x C 9 7 C
6 2 7:

4 c3 5
4 y 5 4
2
1
5
0 2

6
000

3
c4
y
x
x

1
6
x
0 0

2
2
x
x

W2 D 1

Setting x D 1 and imposing the initial conditions yields


2
3 2 3
1 1
7
0

6 11 7 6 3 7 1 2
6
7D 6 7C
4 5 5 4 9 5 0 2

0 0
6
6

1
1
2
6

0
1
1
1

6
6
4

3
c1
c2 7
7:
c3 5
c4

Solving this system yields c1 D 0, c2 D 7, c3 D 0, c4 D 0. Therefore, y D 3x 2 ln x 7x 2.

p
p

p
p
x
1=x
1= x

p
x
1=x
1= x
2
3=2
1

p
1=2 x
1=x
1=2x
D 9=8x 6; W1 D 1=2 x
1=x 2
1=2x 3=2
9.4.32. W D
3=2
3
5=2

0
1=4x
2=x
3=4x
3=2
3

1=4x
2=x
3=4x 5=2
0 3=8x 5=2
6=x 4
15=8x 7=2

p
p
p
x
x
1=x
1= x
x
1= x

p
1=2x 3=2 D 3=4x 2;
3=4x 4; W2 D 1
1=x 2
1=2x 3=2 D 3=2x 7=2; W3 D 1
1=2 x
0 2=x 3
0
3=4x 5=2
1=4x 3=2 3=4x 5=2

p
x

x
1=x

p
p
F W2
F W1
1=2 x
1=x 2 D 3=2x 5=2; u01 D
W4 D 1
D 1=x; u02 D
D 2= x;
P0 W
P0 W
0
1=4x 3=2 2=x 3
p
p
F W2
F W4
u03 D
D x; u04 D
D 2 x; u1 D ln x; u2 D 4 x; u3 D x 2 =2; u4 D 4x 3=2=3;
P0 W
P0 W
yp D u1 y1 C u2 y2 C u3 y3 D x ln x 19x=6. since 19x=6 satisfies the complementary equation we
take yp D x ln x.
p
p
The general solution is y D x ln x C c1x C c2 x C c3 =x C c4= x, so
2
p
p
2
3 2
3
3
x
c1
x
1=x
1= x
y
x ln x

p
6 y 0 7 6 ln x C 1 7 1
7
1=2 x
1=x 2
1=2x 3=2 6
6 00 7 D 6
7C
6 c2 7
3=2
3
5=2
4 c3 5 :
4 y 5 4 1=x 5 0
1=4x
2=x
3=4x

0 3=8x 5=2
y 000
1=x 2
6=x 4
15=8x 7=2 c4

188 Chapter 9 Linear Higher Order Equations


Setting x D 1 and imposing the initial conditions yields
2
3 2
3
1
2
0
1

6 0 7 6 1 7 1 1=2
6
7 6
7
4 4 5 D4 1 5C 0
1=4

37
0 3=8
1
4

Solving this system yields c1 D 1, c2 D

1
1
2
6

1
1=2
3=4
15=8

6
6
4

3
c1
c2 7
7:
c3 5
c4

1, c3 D 1, c4 D 1. Therefore, y D x ln x Cx

p
1
1
xC C p .
x
x

F Wj
(1  j  n), the argument used in the derivation of the method of
P0 W
variation of parameters implies that yp is a solution of (A).
(b) Follows immediately from (a), since uj .x0 / D 0, j D 1; 2; : : : ; n.
(c) Expand the determinant in cofactors of its nth row.
(d) Just differentiate the determinant n 1 times.
@j G.x; t/
has two identical rows, and is therefore zero, while
(e) If 0  j  n 2, then
@x j xDt

@n 1 G.x; t/
D W .t/

@x j
xDt
Z x
Z x
@G.x; t/
(f) Since yp .x/ D
G.x; t/F .t/ dt, yp0 .x/ D G.x; x/F .x/C
F .t/ dt. But G.x; x/ D
@x
x0 Z
x0
x
@G.x; t/
0 from (e), so yp0 .x/ D
F .t/ dt. Repeating this argument for j D 1; : : : ; n and invoking
@x
x0
(e) each time yields the conclusion.
9.4.34. (a) Since u0j D . 1/n

9.4.36.

y1 .t/
0
y .t/
1
y1 .x/

y1 .t/
y10 .t/
y1 .x/

y2 .t/
y20 .t/
y2 .x/

t t2
1=t

D 1 2t
1=t 2

x x
1=x
2

t
1=t

D x
2t
1=t 2
D

3
Since P0 .t/ D t and W .t/ D 1
x
Z x
2
.x t/ .2x C t/
F .t/ dt .
6xt 3
x0
9.4.38.

y1 .t/
0
y .t/
1
y1 .x/

y1 .t/
y10 .t/
y1 .x/

y2 .t/
y20 .t/
y2 .x/

t
x
1
2

t
1=t
e t =t

2
t
1
1=t
e .1=t 1=t 2 /

x
1=x
e x =x

1=t

e t =t

x
2
t
2
1=t
e .1=t 1=t /
xe t
t2

1=t 1
C
1=t 2 x
t/2 .2x C t/
:
xt

x2
t2
.x
C
D
t
x

1=t
6
.x
1=t 2 D , G.x; t/ D
t
1=x

3x C 2

t2
2t
x2

e t .t 2/
xt

t t2

1 2t

t/2 .2x C t/
, so yp D
6xt 3

1 t
e t =t
t

x 1 e .1=t 1=t 2 /
2e x
x 2 e t e t t.t 2/ 2te x
D
xt
xt 2

ex t
C

x 1

1=t
1=t 2

Section 9.4 Variation of Parameters for Higher Order Equations

t
1=t
e t =t

2
t

Since P0 .t/ D t.1 t/ and W .t/ D 1


1=t
e .1=t 1=t 2 /
0 2=t 3 e t .1=t 2=t 2 C 2=t 3/
Z x 2
x
t.t 2/ 2te .x t /
x 2 t.t 2/ 2te .x t /
,
so
y
D
F .t/ dt .
p
2x.t 1/2
2x.t 1/2
x0

189

D 2e .1 t/ , G.x; t/ D

t3

9.4.40.

y1 .t/
y10 .t/
y100 .t/
y1 .x/

y1 .t/
y10 .t/
y100 .t/
y1 .x/

y2 .t/
y20 .t/
y200 .t/
y2 .x/

y3 .t/
y30 .t/
y300 .t/
y3 .x/

D 0

Since P0 .t/ D t and W .t/ D

9.4.42.

y1 .t/
0
y .t/
100
y .t/
1
y1 .x/

y1 .t/
y10 .t/
y100 .t/
y1 .x/

y2 .t/
y20 .t/
y200 .t/
y2 .x/

1
0
0
0

y3 .t/
y30 .t/
y300 .t/
y3 .x/

t t2
1 2t
0 2
0 0

x0

1
0
0
0
t/
.1 C 2t/ C e

6
C
t

6x
t2

t2
2t
2
0

.x t/3
12
, G.x; t/ D
, so yp D
4
t
6x

8/ C 16x 2 e 2.x

e 2t
2e 2t
D 128t,
4e 2t
2t
8e
2t/ 4x 2 C 4t 2 2
, so
4x 2 C 4t 2

2x 2
2
2.t x/3
C
D
:
t3
x
xt 3

.16t 2

e 2t
2e 2t
4e 2t
8e 2t
2.x t /
.1
32t 2
t/
.1 C 2t/ C e 2.x t /.1 2t/
32t 2

Since P0 .t/ D t and W .t/ D

1=t
1=t 2
2=t 3
1=x

1 t2
t2
1=t
1=t

2t
1=t 2 C x 0 2t
1=t 2
0 2
2
2=t 3
2=t 3

1 t t2
1 t
1=t

1
0 1
1=t 2 C 0 1 2t
0 0 2=t 3 x 0 0 2

1 t 2 e 2t
e 2t

2t

2e 2t
D 0 2t 2e2t

0 2 4e
4e 2t

1 x 2 e 2x
e 2x
2

t
1
e 2t
e 2t

2
2t
2t

2e
Cx 0
D
2t 2e 2t

2 4e
0
4e 2t

1 t2

e 2t

2x
2x
2t
e 0 2t
2e
Ce

0 2

4e 2t
D

e 2.x
G.x; t/ D
Z x 2.x
e
yp D

1=t
1=t 2
2=t 3
6=t 4

t
1
0

x 2

t2
2t
2
x2

t
1
0
x

t /t

e 2t
2e 2t
4e 2t
1
0
0

t2
2t
2

.8t C 4/ C e

F .t/ dt.

.x

x0

e 2t
2e 2t
4e 2t

e 2t
2e 2t
4e 2t
2.x t /

t/3
F .t/ dt.
6x

.8t

4/:

CHAPTER 10
Linear Systems of Differential
Equations

10.1 INTRODUCTION TO SYSTEMS OF DIFFERENTIAL EQUATIONS


10.1.2. Q10 D .rate in/1 .rate out/1 and Q20 D .rate in/2 .rate out/2 .
The volumes of the solutions in T1 and T2 are V1 .t/ D 100 C 2t and V2 .t/ D 100 C 3t, respectively.
T1 receives salt from the external source at the rate of (2 lb/gal)  (6 gal/min) D 12 lb/min, and from
1
T2 at the rate of (lb/gal in T2 /  (1 gal/min) D
Q2 lb/min. Therefore, (A) (rate in)1 D 12 C
100 C 3t
1
Q2 . Solution leaves T1 at 5 gal/min, since 3 gal/min are drained and 2 gal/min are pumped to
100 C 3t
5
1
Q1  5 D
Q1 . Now (A)
T2 ; hence (B) .rate out/1 D . lb/gal in T1 /  (5 gal/min) D
100 C 2t
100 C 2t
5
1
and (B) imply that (C) Q10 D 12
Q1 C
Q2 .
100 C 2t
100 C 3t
T2 receives salt from the external source at the rate of (1 lb/gal)  (5 gal/min) D 5 lb/min, and from
1
1
T1 at the rate of (lb/gal in T1 /  (2 gal/min) D
Q1  2 D
Q1 lb/min. Therefore, (D)
100 C 2t
50 C t
1
(rate in)2 D 5 C
Q1 . Solution leaves T2 at 4 gal/min, since 3 gal/min are drained and 1 gal/min is
50 C t
1
4
pumped to T1 ; hence (E) .rate out/2 D . lb/gal in T2 / (4 gal/min) D
Q2  4 D
Q2 .
100 C 3t
100 C 3t
1
4
Now (D) and (E) imply that (F) Q20 D 5 C
Q1
Q2 . Now (C) and (F) form the desired
50 C t
100 C 3t
system.
10.1.4. mX00 D
10.1.8.

X0

mgR2

X
; see Example 10.1.3.
kXk3

I1i
J1i
I2i

J2i
y1;i C1
y2;i C1

D g1 .ti ; y1i ; y2i /;


D g2 .ti ; y1i ; y2i /;
D g1 .ti C h; y1i C hI1i ; y2i C hJ1i / ;
D g2 .ti C h; y1i C hI1i ; y2i C hJ1i / ;
h
D y1i C .I1i C I2i /;
2
h
D y2i C .J1i C J2i /:
2
191

192 Chapter 10 Linear Systems of Differential Equations


10.2 LINEAR SYSTEMS OF DIFFERENTIAL EQUATIONS
10.2.6. Let yi D y .i 1/ , i D 1; 2; : : : ; n; then yi0 D yi C1 , i
   C Pn .t/y1 D F .t/, so
2
0
1
0
 0
6 0
0
1

 0
1 6
6 ::
:
:
:
::
::
: : :::
AD
6
P0 6 :
4 0
0
0
 1
Pn Pn 1 Pn 2    P1

D 1; 2; : : : ; n
3
7
7
7
7
7
5

and

1 and P0 .t/yn0 C P1 .t/yn C

1
fD
P0

0
0
::
:

6
6
6
6
6
4 0
F

7
7
7
7:
7
5

If P0 ; P1 ; : : : ; Pn and F are continuous and P0 has no zeros on .a; b/, then P1 =P0 ; : : : ; Pn =P0 and F=P0
are contnuous on .a; b/.
10.2.7. (a) .c1 P Cc2 Q/0ij D .c1 pij Cc2 qij /0 D c1 pij0 Cc2 qij0 D .c1 P 0 Cc2Q0 /ij ; hence .c1 P Cc2Q/0 D
c1 P 0 C c2 Q 0 .
r
X
(b) Let P be k  r and Q be r  s; then PQ is k  s and .PQ/ij D
pi qj . Therefore, .PQ/0ij D

r
X
D1

pi0 qj C

r
X
D1

D1

pi qj0 D .P 0 Q/ij C .PQ0 /ij . Therefore, .PQ/0 D P 0 Q C PQ0 .

10.2.10. (a) From Exercise 10.2.7(b) with P D Q D X, .X 2 /0 D .XX/0 D X 0 X C XX 0 .


(b) By starting from Exercise 10.2.7(b) and using induction it can be shown if P1 ; P2 ; : : : ; Pn are
square matrices of the same order, then .P1 P2    Pn /0 D P10 P2    Pn C P1 P20    Pn C   C P1P2    Pn0 .
Taking P1 D P2 D    D Pn D X yields (A) .Y n /0 D Y 0 Y n 1 C Y Y 0 Y n 2 C Y 2 Y 0 Y n 3 C    C
n
X1
Y n 1Y 0 D
Y r Y 0Y n r 1.
r D0

(c) If Y is a scalar function, then (A) reduces to the familiar result .Y n /0 D nY n

Y 0.

10.2.12. From Exercise 10.2.6, the initial value problem (A) P0 .x/y .n/ C P1 .x/y .n 1/ C   C Pn .x/y D
F .x/, y.x0 / D k0 ; y 0 .x0 / D k1 ; : : : ; y .n 1/ .x0 / D kn 1 is equivalent to the initial value problem (B)
y0 D A.t/y C f.t/, with
2
3
2
3
2
3
0
1
0
 0
0
k0
6 0
7
6 0 7
0
1



0
7
6
7
6
7
1 6
6 ::
::
::
:: 7 ; f D 1 6 :: 7 ; and k D 6 k1 7 :
::
AD
6 :
7
6
7
6
: 7
: : 7
:
:
P0 6
P0 6 : 7
4 :: 5
4 0
5
4
5
0
0
 1
0
kn 1
Pn Pn 1 Pn 2    P1
F

Since Theorem 10.2.1 implies that (B) has a unique solution on .a; b/, it follows that (A) does also.

10.3 BASIC THEORY OF HOMOGENEOUS LINEAR SYSTEM






1
0
1
y
0
10.3.2. (a) The system equivalent of (A) is (B) y D
y, where y D
.
y0
P0 .x/ P2 .x/ P1 .x/




y1
y2
Let y1 D
and y1 D
. Then the Wronskian of fy1; y2 g as defined in this section is
y10
y20


y1 y2
D W.
y10 y20

Section 10.3 Basic Theory of Homogeneous Linear System

193

(b) The trace of the matrix in (B) is P1 .x/=P0 .x/, so Eqn. 10.3.6 implies that W .x/ D W .x0 / exp
10.3.4. (a) See the solution
of Exercise 9.1.18.

0
y0
a11 y11 C a12 y21 a11 y12 C a12 y22
y
11
12
D
(c)
y21 y22
y21
y

 22
y11 y12
a11 W C a12 0 D a11 W . Similarly,
D a22 W .
0
0
y21
y22

D a11 y11

y21

y21
y12

Ca
12

y22
y21

 Z

x
x0

y22
D
y22

10.3.6. (a) From the equivalence of Theorem 10.3.3(b)and (e), Y .t0 / is invertible.
(b) From the equivalence of Theorem 10.3.3(a)and (b), the solution of the initial value problem is
y D Y .t/c, where c is a constant vector. To satisfy y.t0 / D k, we must have Y .t0 /c D k, so c D Y 1 .t0 /k
and y D Y 1 .t0 /Y .t/k.




e 4t
2e 3t
c
2c2 D 10
10.3.8. (b) y D
C
c
where 1
; so c1 D 6, c2 D 2, and
2
e 4t
5e 3t
c1 C 5c2 D
4


6e 4t C 4e 3t
yD
.
6e 4t 10e 3t






1
e 4t
2e 3t
1
2
5 2
1
(c) Y .t/ D
; Y .0/ D
; Y .0/ D
; y D Y .t/Y 1 .0/k D
e 4t
5e 3t
1 5
1 1
7


1 5e 4t C 2e 3t 2e 4t 2e 3t
k.
7 5e 4t 5e 3t 2e 4t C 5e 3t
 3t 


e
et
c C c2 D 2
10.3.10. (b) y1 D c1
C
c
, where 1
; so c1 D 5, c2 D 3, and
2
e 3t
et
c1 c2 D 8


5e 3t 3e t
yD
.
5e 3t C 3e t
 3t



e
et
1
1
(c) Y .t/ D
; Y .0/ D
;
3e 3t
et
1
1




1 1
1 e 3t C e t e 3t e t
1
1
1
Y .0/ D
; y D Y .t/Y .0/k D
k.
1
2 1
2 e 3t e t e 3t C e t
2
3
2
3
2 4t 3
e 2t
e 2t
e
c1 c2 C c3 D
0
5 C c2 4 e 2t 5 C c3 4 e 4t 5, where
10.3.12. (b) y D c1 4
0
c2 C c3 D
9 ; so
e 2t
0
e 4t
c1 C c3 D 12
2
3
e 2t C e 4t
1
c1 D 11, c2 D 10, c3 D 1, and y D 4 10e 2t C e 4t 5.
3
11e 2t C e 4t
2
3
2
3
2
3
2t
2t
4t
e
e
e
1
1 1
1
1
2
1
(c) Y .t/ D 4
0
e 2t e 4t 5; Y .0/ D 4 0
1 1 5; Y 1 .0/ D 4 1
2
1 5;
3
2t
4t
1
1
1
e
0
e
1
0 1
2
3
2e 2t C e 4t
e 2t C e 4t
e 2t C e 4t
1
e 2t C e 4t 5 k.
y D Y .t/Y 1 .0/k D 4 e 2t C e 4t 2e 2t C e 4t
3
2t
4t
2t
4t
e
Ce
e
Ce
2e 2t C e 4t

10.3.14. If Y and Z are both fundamental matrices for y0 D A.t/y, then Z D C Y , where C is a constant
invertible matrix. Therefore, Z Y 1 D C and Y Z 1 D C 1 .

10.3.16. (a) The Wronskian of fy1 ; y2; : : : ; yn g equals one when t D t0 . Apply Theorem 10.3.3.
(b) Let Y be the matrix with columns fy1; y2 ; : : : ; yng. From (a), Y is a fundamental matrix for
y0 D A.t/y on .a; b/. From Exercise 10.3.15(b), so is Z D Y C if C is any invertible constant matrix.


P1 .s/
ds .
P0 .s/

194 Chapter 10 Linear Systems of Differential Equations


10.3.18. (a) 10 .t/ D Z 0 .t/Z.s/ D AZ.t/Z.s/ D A.t/ and 1 .0/ D Z.s/, since Z.0/ D I . 20 .t/ D
Z 0 .t C s/ D AZ.t C s/ D A2 .t/ (since A is constant) and 2 .0/ D Z.s/. Applying Theorem 10.2.1 to
the columns of 1 and 2 shows that 1 D 2 .
(b) With s D t, (a) implies that Z.t/Z. t/ D Z.0/ D I ; therefore .Z.t// 1 D Z. t/.
(c) e 0A D I is analogous to e oa D e 0 D 1 when a is a scalar, while e .t Cs/A D e tAe sA is analagous
to e .t Cs/a D e t a e sa when a is a scalar.
10.4

CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS I

1 5 4
3
D . C 1=2/. C 2/. Eigenvectors associated with 1 D 1=2 satisfy
10.4.2.

4   3  5 4




3 3
x1
0
1
D
, so x1 D x2 . Taking x2 D 1 yields y1 D
e t =2. Eigenvectors
4 4
x2
0
1
 3 3 
 

x1
0
D
, so x1 D x2 . Taking x2 D 1 yields
associated with 2 D 2 satisfy 4 4
x2
0
1 1






1
1
1
y2 D
e 2t . Hence y D c1
e t =2 C c2
e 2t .
1
1
1

1 
4

10.4.4.
D . 1/. C 3/. Eigenvectors associated with 1 D 3 satisfy
1
1 


 



2
4
x1
0
2
D
, so x1 D 2x2. Taking x2 D 1 yields y1 D
e 3t . Eigenvectors
1
2
x2
0
1


 

2
4
x1
0
associated with 2 D 1 satisfy
D
, so x1 D 2x2 . Taking x2 D 1 yields
1
2
x2
0


 


2
2
2
y2 D
e t . Hence y D c1
e 3t C c2
et .
1
1
1



 

4 
3
2
3
x1
0

10.4.6.
D . 2/. 1/. Eigenvectors associated with 1 D 2 satisfy
D
,
2
1 
2
3
x2
0


3
3
so x1 D x2 . Taking x2 D 2 yields y1 D
e 2t . Eigenvectors associated with 2 D 1 sat2
2


  


3
3
x1
0
1
isfy
D
, so x1 D x2. Taking x2 D 1 yields y2 D
e t . Hence y D
2
2
x2
0
1




3
1
c1
e 2t C c2
et .
2
1

1 
1
2

10.4.8. 1
2 
3 D . C 3/. C 1/. 2/. The eigenvectors associated with

4
1
1 
2
3
::
:
0
4
1
2
6
7
6
7
::
with 1 D 3 satisfy the system with augmented matrix 6 1
, which is row
1
3 : 0 7
4
5
::
4
1
2 : 0
2
3
::
1 : 0 7
6 1 0
6
7
:
equivalent to 6 0 1
. Hence x1 D x3 and x2 D 2x3. Taking x3 D 1 yields y1 D
2 :: 0 7
4
5
::
0 0
0 : 0

Section 10.4 Constant Coefficient Homogeneous Systems I


2

3
1
4 2 5e
1
2

3t

:
2 ::
:
3 ::
:
0 ::

1 satisfy the system with augmented ma2

3
::
1 : 0 7
7
:
. Hence x1 D
4 0:: 0 7
5
::
0 : 0

0 7
6 1 0
7
6
, which is row equivalent to 6 0 1
x3
1
1
0 7
5
4
4
1
0
0 0
2
3
1
and x2 D 4x3. Taking x3 D 1 yields y2 D 4 4 5e t . The eigenvectors associated with with
1
2
3
::
1
1
2
:
0
6
7
6
7
::
3 D 2 satisfy the system with augmented matrix 6 1
, which is row equivalent
4
3 : 0 7
4
5
::
4
1
3 : 0
2
3
::
2
3
1
6 1 0 1 : 0 7
6
7
to 6 0 1 1 ::: 0 7. Hence x1 D x3 and x2 D x3 . Taking x3 D 1 yields y3 D 4 1 5e 2t .
4
5
1
:
0 0 0 :: 0
2 3
2
3
2
3
1
1
1
Hence y D c1 4 2 5 e 3t C c2 4 4 5 e t C c3 4 1 5 e 2t .
1
1
1

3 

5
8

1 
2 D . 1/. C 2/. 2/. The eigenvectors associated with
10.4.10. 1

1
1
1 
2
3
::
2
5
8
:
0
6
7
6
7
::
with 1 D 1 satisfy the system with augmented matrix 6 1
, which is row equiv2
2 : 0 7
5
4
::
1
1
2 : 0
2
3
:
1 0 23 :: 0
6
7
6
7
alent to 6 0 1 4 ::: 0 7. Hence x1 D 32 x3 and x2 D 43 x3 . Taking x3 D 3 yields y1 D
4
5
3
::
0 0 0 : 0
2
3
2
4 4 5e t . The eigenvectors associated with with 2 D 2 satisfy the system with augmented ma3
2
3
2
3
::
::
5
5
8
:
0
1
1
0
:
0
6
7
6
7
6
7
6
7
::
::
trix 6 1
,
which
is
row
equivalent
to
7
6
7. Hence x1 D x2
1
2 : 0 5
4
4 0 0 1 : 0 5
::
::
1
1
1 : 0
0 0 0 : 0
2
3
1
and x3 D 0. Taking x2 D 1 yields y2 D 4 1 5e 2t . The eigenvectors associated with with
0
6
6
trix 6
4

. The eigenvectors associated with with 2 D

195

196 Chapter 10 Linear Systems of Differential Equations


2

6
6
3 D 2 satisfy the system with augmented matrix 6
4

:
8 ::
:
2 ::
:
3 ::

3
0 7
7
, which is row equivalent to
0 7
5

1
1
0
3
:
2
3
1 0 74 :: 0
7
6
7
6
7
6 0 1 5 ::: 0 7. Hence x1 D 47 x3 and x2 D 54 x3 . Taking x3 D 4 yields y3 D 4 5 5e 2t .
4
5
4
4
::
0 0 0 : 0
2
3
2
3
2
3
2
1
7
Hence y D c1 4 4 5 e t C c2 4 1 5 e 2t C c3 4 5 5 e 2t .
3
0
4

4 
1
4

3 
2 D . 3/. C 2/. C 1/. The eigenvectors associated with
10.4.12. 4
1
1
1 
3
2
::
1
1
4
:
0
7
6
7
6
::
1 D 3 satisfy the system with augmented matrix 6 4
, which is row equiva6
2 : 0 7
5
4
::
1
1
4 : 0
2
3
::
11 : 0 7
6 1 0
6
7
:
lent to 6 0 1
. Hence x1 D 11x3 and x2 D 7x3. Taking x3 D 1 yields y1 D
7 :: 0 7
4
5
:
0 0
0 :: 0
2
3
11
4 7 5e 3t . The eigenvectors associated with 2 D 2 satisfy the system with augmented matrix
1
2
3
2
3
::
::
1
4 : 0 7
1 : 0 7
6 6
6 1 0
6
7
6
7
::
:
,
which
is
row
equivalent
to
. Hence x1 D x3 and x2 D
6 4
6 0 1
1
2 : 0 7
2 :: 0 7
4
5
4
5
:
:
1
1
1 :: 0
0 0
0 :: 0
3
2
1
2x3 . Taking x3 D 1 yields y2 D 4 2 5e 2t . The eigenvectors associated with 3 D 1 satisfy the sys1
3
3
2
2
::
::
1
0
1
:
0
:
0
5
1
4
6
7
7
6
7
6
6
7
::
::
.
,
which
is
row
equivalent
to
tem with augmented matrix 6 4
7
6
1 : 0 7
2
2 : 0 5
5
4 0 1
4
::
::
0 0
0 : 0
1
1
0 : 0
2
3
2
3
2
3
1
11
1
Hence x1 D x3 and x2 D x3 . Taking x3 D 1 yields y3 D 4 1 5e t . Hence y D c1 4 7 5 e 3t C c2 4 2 5 e
1
1
1

3 
2
2

10.4.14.
2
7 
2 D . C 5/. 5/2 . The eigenvectors associated with 1 D

10
10
5 
2

2t

3
1
C c3 4 1 5 e t .
1

Section 10.4 Constant Coefficient Homogeneous Systems I


2

6
6
5 satisfy the system with augmented matrix 6
4

1
6
6
6 0
4

::
1
:
5
::
1
:
5
:
0 ::

2 12

10 10

197

3
::
: 0 7
7
::
, which is row equivalent to
: 0 7
5
::
: 0

2
3
1
7
1
1
7
. Hence x1 D x3 and x2 D x3 . Taking x3 D 5 yields y1 D 4 1 5e 5t . The
1
0 7
5
5
5
5
0 0
0
2
3
::
2 2
2 : 0 7
6
6
7
:
eigenvectors associated with 2 D 5 satisfy the system with augmented matrix 6
,
2 2
2 :: 0 7
4
5
:
10 10
10 :: 0
2
3
:
1 1 :: 0 7
6 1
6
7
:
which is row equivalent to 6 0
. Hence x1 D x2 x3 . Taking x2 D 0 and x3 D
0 0 :: 0 7
5
4
:
0
0 0 :: 0
2
3
2
3
1
1
1 yields y2 D 4 0 5e 5t . Taking x2 D 1 and x3 D 0 yields y3 D 4 1 5e 5t . Hence y D
1
0
3
2 3
2
2 3
1
1
1
c1 4 1 5 e 5t C c2 4 0 5 e 5t C c3 4 1 5 e 5t .
0
5
1



 

7 
4
12 4
x1
0
10.4.16.
D
.
5/.C5/.
Eigenvectors
associated
with

D
5
satisfy
D
,
1
6
7 
6 2
x2
0


x2
1
so x1 D
. Taking x2 D 3 yields y1 D
e 5t . Eigenvectors associated with 2 D 5 satisfy
3
3


 



2 4
x1
0
2
D
, so x1 D 2x2. Taking x2 D 1 yields y2 D
e 5t . The general
6 12
x2
0
1








  

1
2
2
1
2
2
5t
5t
solution is y D c1
e C c2
e . Now y.0/ D
) c1
C c2
D
,
3
1
4
3
1
4




2
4
so c1 D 2 and c2 D 2. Therefore, y D
e 5t C
e 5t .
6
2

21 
12

10.4.18.
D . 9/. C 3/. Eigenvectors associated with 1 D 9 satisfy
24
15 


 



12
12
x1
0
1
D
, so x1 D x2 . Taking x2 D 1 yields y1 D
e 9t . Eigenvec24
24
x2
0
1


 

1
24
12
x1
0
tors associated with 2 D 3
D
, so x1 D x2 . Taking x2 D 2 yields
24
12
x2
0





 2


1
1
1
5
3t
9t
y2 D
e . The general solution is y D c1
e C c2
e 3t . Now y.0/ D
)
2
1
2
3



 

 


1
1
5
7
2
c1
C c2
D
, so c1 D 7 and c2 D 2. Therefore, y D
e 9t
e 3t .
1
2
3
7
4

198 Chapter 10 Linear Systems of Differential Equations

10.4.20.

1 D

1
6

1
3


2
3

1
6


0

0 D


0

1
2

1=2/2 . The eigenvectors associated with with

. C 1=2/.
2

6
6
1=2 satisfy the system with augmented matrix 6
4

2
3

1
3

2
3

1
3

:
0 ::
:
0 ::
:
1 ::

7
7
, which is row equivalent to
0 7
5

0 0
0
3
:
2
3
1 21 0 :: 0
1
6
7
x
6
7
2
and x3 D 0. Taking x2 D 2 yields y1 D 4 2 5e t =2.
6 0 0 1 ::: 0 7. Hence x1 D
4
5
2
0
:
0 0 0 :: 0
The eigenvectors associated with with 2 D 3 D 1=2 satisfy the system with augmented matrix
3
2
3
2
::
::
1
1
0
:
0
1
1
0
:
0
3
7
6
7
6 3
7
6
7
6
::
::
,
which
is
row
equivalent
to
. Hence x1 D x2 and
7
6
6 2
2
0 0 : 0 7
0 : 0 5
5
4 0
4 3
3
:
:
0
0 0 :: 0
0
0 0 :: 0
2
3
1
x3 is arbitrary. Taking x2 D 1 and x3 D 0 yields y2 D 4 1 5e t =2. Taking x2 D 0 and x3 D 1 yields
0
2
3
2
3
2
3
2 3
0
1
1
0
y3 D 4 0 5e t =2. The general solution is y D c1 4 2 5 e t =2 C c2 4 1 5 e t =2 C c3 4 0 5 e t =2.
1
0
0
1
2
2
2
2 3
2
3
3
3
3
4
1
1
0
4
Now y.0/ D 4 7 5 ) c1 4 2 5 C c2 4 1 5 C c3 4 0 5 e t =2 D 4 7 5, so c1 D 1, c2 D 5, and
1
0
0
1
1
2
3
2 3
2
3
1
5
0
c3 D 1. Hence y D 4 2 5 e t =2 C 4 5 5 e t =2 C 4 0 5 e t =2.
0
0
1

6 
3
8

1 
2 D . 1/. C 2/. 3/. The eigenvectors associated with
10.4.22. 2
3
3
5 
3
2
::
3
8 : 0 7
6 5
6
7
:
1 D 1 satisfy the system with augmented matrix 6 2
, which is row equivalent to
0
2 :: 0 7
5
4
::
3
3
6 : 0
2
3
::
2
3
1 : 0 7
1
6 1 0
6
7
:
. Hence x1 D x3 and x2 D x3 . Taking x3 D yields y1 D 4 1 5e t . The
6 0 1
1 :: 0 7
4
5
1
:
0 0
0 :: 0
2
3
::
8
3
8
:
0
6
7
6
7
::
eigenvectors associated with 2 D 2 satisfy the system with augmented matrix 6 2
,
3
2 : 0 7
4
5
::
3
3
3 : 0
2

Section 10.4 Constant Coefficient Homogeneous Systems I


2

6 1 0
6
which is row equivalent to 6 0 1
4
0 0

3
1
y2 D 4 0 5e
1
2
3
8
6 3
6
6 2
2
2
4
3
3
8

2t

1
0
0

199

3
::
: 0 7
7
::
. Hence x1 D x3 and x2 D 0. Taking x3 D 1 yields
: 0 7
5
::
: 0

. The eigenvectors associated with 3 D 3 satisfy the system with augmented matrix

::
:
::
:
::
:

:
1 0 ::
:
0 1 ::
:
0 0 ::

0 7
0 7
6 1
6
7
7
, which is row equivalent to 6 0
. Hence x1 D x2 and x3 D
0 7
0 7
5
4
5
0
0
0
2
3
2
3
2
3
1
1
1
0. Taking x2 D 1 yields y3 D 4 1 5e 3t . The general solution is y D c1 4 1 5e t C c2 4 0 5e 2t C
0
1
1
2
3
2
3
2
3
2 3
2
3
2
3
1
0
1
1
1
0
c3 4 1 5e 3t . Now y.0/ D 4 1 5 ) c1 4 1 5 C c2 4 0 5 C c34 1 5 D 4 1 5, so c1 D 2,
0
1
1
1
0
1
2
3
2 3
2
3
2
3
1
c2 D 3, and c3 D 1. Therefore, y D 4 2 5 e t 4 0 5 e 2t C 4 1 5 e 3t .
2
3
0

3 
0
1

2 
7 D . 2/. C 2/. 4/. The eigenvectors associated with
10.4.24. 11
1
0
3 
3
2
::
:
0
1
0
1
6
7
6
7
::
with 1 D 2 satisfy the system with augmented matrix 6 11
7, which is row equiv4
7
:
0
5
4
::
1
0 1 : 0
2
3
::
1
0
1
:
0
6
7
6
7
alent to 6 0 1 1 ::: 0 7. Hence x1 D x3 and x2 D x3. Taking x3 D 1 yields y1 D
4
5
:
0 0 0 :: 0
2
3
1
4 1 5e 2t . The eigenvectors associated with with 2 D 2 satisfy the system with augmented ma1
2
3
2
3
::
::
5
0
1
:
0
1
0
0
:
0
6
7
6
7
6
7
6
7
trix 6 11 0 7 ::: 0 7, which is row equivalent to 6 0 0 1 ::: 0 7. Hence x1 D x3 D 0
4
5
4
5
::
::
1 0 5 : 0
0 0 0 : 0
2 3
0
and x2 is arbitrary. Taking x3 D 1 yields y2 D 4 1 5e 2t . The eigenvectors associated with with
0

200 Chapter 10 Linear Systems of Differential Equations


3
::
:
0
6
7
6
7
::
3 D 4 satisfy the system with augmented matrix 6 11
, which is row equivalent
6
7 : 0 7
4
5
:
1
0
1 :: 0
2
3
:
2
3
1 :: 0 7
1
6 1 0
6
7
:
to 6 0 1
. Hence x1 D x3 and x2 D 3x3. Taking x3 D 1 yields y3 D 4 3 5e 4t .
3 :: 0 7
4
5
1
:
0 0
0 :: 0
2
3
2 3
2
3
2 3
1
0
1
2
The general solution is y D c14 1 5e 2t C c2 4 1 5e 2t C c3 4 3 5e 4t . Now y.0/ D 4 7 5 )
1
0
1
6
2
3
2
3
2 3
2
3
1
0
1
2
c1 4 1 5 C c24 1 5 C c3 4 3 5 D 4 7 5, so c1 D 2, c2 D 3, and c3 D 4. Hence y D
1
0
1
6
2
3
2
3
2
3
2
0
4
4 2 5 e 2t 4 3 5 e 2t C 4 12 5 e 4t
2
0
4

3 
1
0

2 
0 D .C1/. 2/2 . The eigenvectors associated with 1 D 1 sat10.4.26. 4
4
4
2 
2
3
2
::
1
1 0 : 0 7
4
6 4
6 1 0
7
6
6
:
:
isfy the system with augmented matrix 6 4
,
which
is
row
equivalent
to
6 0 1
1 0 : 0 7
1
4
5
4
::
4
4 3 : 0
0 0
0
2
3
1
Hence x1 D x2 =4 and x2 D x3 . Taking x3 D 4 yields y1 D 4 4 5e t . The eigenvectors associated
4
2
3
::
1 0 : 0 7
6 1
6
7
:
, which is row
with with 2 D 3 D 2 satisfy the system with augmented matrix 6 4
4 0 :: 0 7
4
5
::
4
4 0 : 0
2
3
::
1 0 : 0 7
6 1
6
7
:
equivalent to 6 0
. Hence x1 D x2 and x3 is arbitrary. Taking x2 D 1 and x3 D 0
0 0 :: 0 7
4
5
::
0
0 0 : 0
2
3
2
3
1
0
yields y2 D 4 1 5e 2t . Taking x2 D 0 and x3 D 1 yields y3 D 4 0 5e 2t . The general solution is
0
1
2
3
2
3
2 3
2
3
2 3
2 3
1
1
0
7
1
1
y D c1 4 4 5e t C c24 1 5e 2t C c3 4 0 5e 2t . Now y.0/ D 4 10 5 ) c1 4 4 5 C c2 4 1 5 C
4
0
1
2
4
0
2
3 2
3
2 3
2
3
0
7
1
6
c3 4 0 5 D 4 10 5, so c1 D 1, c2 D 6, and c3 D 2. Hence y D 4 4 5 e t C 4 6 5 e 2t .
1
2
4
2
2

::
:
::
:
::
:

0 7
7
0 7
5
0

Section 10.5 Constant Coefficient Homogeneous Systems II

201

10.4.28. (a) If y.t0 / D 0, then y is the solution of the initial value problem y0 D Ay; y.t0 / D 0. Since
y  0 is a solution of this problem, Theorem 10.2.1 implies the conclusion.
(b) It is given that y01 .t/ D Ay1 .t/ for all t. Replacing t by t  shows that y01.t / D Ay1 .t / D
Ay2 .t/ for all t. Since y02.t/ D y01 .t / by the chain rule, this implies that y02 .t/ D Ay2 .t/ for all t.
(c) If z.t/ D y1 .t /, then z.t2 / D y1 .t1 / D y2 .t2 /; therefore z and y2 are both solutions of the initial
value problem y0 D Ay; y.t2 / D k, where k D y2 .t2 /.
10.4.42. The characteristic polynomial of A is p./ D 2 .a C b/ C ab , so the eigenvalues
 of
p
aCb
aCb C
b
a
C

A are 1 D
and 1 D
, where D .a b/2 C 4; x1 D
and
2
2
2


b a
x2 D
are associated eigenvectors. Since > jb aj, if L1 and L2 are lines through
2
the origin parallel to x1 and x2 , then L1 is in the first and third quadrants and L2 is in the second and
2
> 0. If Q0 D P0 there are three possibilities:
fourth quadrants. The slope of L1 is  D
b aC
(i) if D ab, then 1 D 0 and P .t/ D P0 , Q.t/ D Q0 for all t > 0; (ii) if < ab, then
1 > 0 and limt !1 P .t/ D limt !1 Q.t/ D 1 (monotonically); (iii) if > ab, then 1 < 0 and
limt !1 P .t/ D limt !1 Q.t/ D 0 (monotonically). Now suppose Q0 P0 , so that the trajectory
cannot intersect L1 , and assume for the moment that (A) makes sense for all t > 0; that is, even if
one or the other of P and Q is negative. Since 2 > 0 it follows that either limt !1 P .t/ D 1 or
limt !1 Q.t/ D 1 (or both), and the trajectory is asymptotically parallel to L2 . Therefore,the trajectory
must cross into the third quadrant (so P .T / D 0 and Q.T / > 0 for some finite T ) if Q0 > P0 , or into
the fourth quadrant (so Q.T / D 0 and P .T / > 0 for some finite T ) if Q0 < P0 .
10.5

CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS II


10.5.2.
1



 


1
x1
0
D .C1/2 . Hence 1 D 1. Eigenvectors satisfy 1
D
,
2 
1
1
x2
0


1
so x1 D x2 . Taking x2 D 1 yields y1 D
e t . For a second solution we need a vector u such that
1


 





1
1
u1
1
1
1
t
D
. Let u1 D 1 and u2 D 0. Then y2 D
e C
te t . The general
1
1
u2
1
0
1







1
1
1
t
t
t
solution is y D c1
e C c2
e C
te
.
1
0
1



 

3 
1
1
1
x1
0
2
10.5.4. y0 D
D
.
2/
.
Hence

D
2.
Eigenvectors
satisfy
D
,
1
1
1 
1
1
x2
0


1
so x1 D x2 . Taking x2 D 1 yields y1 D
e 2t . For a second solution we need a vector u such that
1


 





1
1
u1
1
1
1
2t
D
. Let u1 D 1 and u2 D 0. Then y2 D
e C
te 2t .
1
1
u2
1
0
1







1
1
1
The general solution is y D c1
e 2t C c2
e 2t C
te 2t .
1
0
1



 

10 
9
6 9
x1
0
2
10.5.6.
D
.C4/
.
Hence

D
4.
Eigenvectors
satisfy
D
,
1
4
2 
4 6
x2
0


3
so x1 D 32 x2 . Taking x2 D 2 yields y1 D
e 4t . For a second solution we need a vector u such that
2


 





6 9
u1
3
1 e 4t
3
1
D
. Let u1 D 2 and u2 D 0. Then y2 D
C
te 4t . The
4 6
u2
2
0
2
2
1

202 Chapter 10 Linear Systems of Differential Equations


general solution is y D c1
2


10.5.8. 4 4
0

2
6

3
2

1
1


4

4t

C c2

5D

.



1
0

4t

3
2

te

::
:
::
1
:
2
:
0 ::


.

4/2 . Hence 1 D 0 and 2 D 3 D 4. The eigenvectors

6
6
associated with 1 D 0 satisfy the system with augmented matrix 6
4
2

4t

0 2
4 6
0 4

:
1 ::
:
1 ::
:
2 ::

3
0 7
7
, which is
0 7
5
0

1
2

0 7
6 1 0
1
1
7
6
. Hence x1 D
row equivalent to 6 0 1
x3 and x2 D
x3. Taking x3 D 2 yields
0 7
5
4
2
2
0 0
0
2
3
1
y1 D 4 1 5. The eigenvectors associated with 2 D 4 satisfy the system with augmented matrix
2
3
2
3
2
::
::
1
: 0 7
1 : 0 7
2
6 4 2
6 1 0
1
7
6
7
6
::
::
1
. Hence x1 D x3 and
, which is row equivalent to 6 0 1
6 4 2
: 0 7
1 : 0 7
2
5
2
4
5
4
:
:
0 0
0 :: 0
0 4
2 :: 0
2
3
1
1
x2 D x3. Taking x3 D 2 yields y2 D 4 1 5e 4t . For a third solution we need a vector u such
2
2
32
3
2 3
2
4 2
1
u1
1
that 4 4 2
1 5 4 u2 5 D 4 1 5. The augmented matrix of this system is row equivalent to
0 4
2
2
u
2
3 3
:
2
3
2 3
1
:: 0
0
1
2
7
6 1 0
4t
1
e
6
:: 1 7. Let u D 0, u D 0, and u D . Then y D 4 1 5
4
1
1 5te 4t . The
C
7
6 0 1
3
1
2
3
: 2 5
2
4
2
2
0
2
:
0 0
0 :: 0
general solution is
2
3
2 3
02
3
2 3
1
1
1
0
1
4t
e
y D c1 4 1 5 C c2 4 1 5 e 4t C c3 @4 1 5
C 4 1 5 te 4t A:
2
2
2
0
2
2

10.5.10. 4


2
1

1

3

1
2
1

5D

.

2/. C 2/2 . Hence 1 D 2 and 2 D 3 D


2

6
6
eigenvectors associated with 1 D 2 satisfy the system with augmented matrix 6
4

2. The

:
1 ::
:
2 ::
:
3 ::

0 7
7
,
0 7
5
0

Section 10.5 Constant Coefficient Homogeneous Systems II


2

6 1 0
6
which is row equivalent to 6 0 1
4

0
1

0 0

3
0
y1 D 4 1 5e 2t . The
1
2
:
1 ::
6 1 1
6
:
trix 6 2 2
2 ::
4
:
1 3
1 ::

203

3
::
: 0 7
7
::
. Hence x1 D 0 and x2 D x3 . Taking x3 D 1 yields
: 0 7
5
::
: 0

eigenvectors associated with 2 D


3

2 satisfy the system with augmented ma:


1 ::
:
0 ::
:
0 ::

0 7
0 7
6 1 0
6
7
7
, which is row equivalent to 6 0 1
. Hence x1 D x3 and
0 7
0 7
5
4
5
0
0 0
0
2 3
1
x2 D 0. Taking x3 D 1 yields y2 D 4 0 5e 2t . For a third solution we need a vector u such
1
2
32
3
2 3
1 1
1
u1
1
that 4 2 2
2 5 4 u2 5 D 4 0 5. The augmented matrix of this system is row equivalent to
1 3
1
u3
1
3
2
:: 1
2 3
2 3
1 : 2 7
1
1
6 1 0
2t
1
1
e
6
:: 1 7. Let u D 0, u D , and u D . Then y D 4 1 5
4
0 5te 2t .
C
7
6 0 1
3
1
2
3
0 : 2 5
2
2
2
4
0
1
:
0 0
0 :: 0
The general solution is
2
3
1
2
3
2
3
02 3
0
1
1
1
2t
e
2t
2t
2t
y D c1 4 1 5 e C c2 4 0 5 e
C c 3 @4 1 5
C 4 0 5 te A:
2
1
1
0
1
2

10.5.12. 4


2
2

5
1

3
3


1

5D

. C 2/.

4/2 . Hence 1 D

2 and 2 D 3 D 4. The
2

3
::
: 0 7
7
::
,
: 0 7
5
::
: 0

5 3
6 8
6
eigenvectors associated with 1 D 2 satisfy the system with augmented matrix 6 2
1 3
4
2
1 3
2
3
::
6 1 0 1 : 0 7
6
7
which is row equivalent to 6 0 1 1 ::: 0 7. Hence x1 D x3 and x2 D x3. Taking x3 D 1
4
5
::
0 0 0 : 0
2
3
1
yields y1 D 4 1 5e 2t . The eigenvectors associated with 2 D 4 satisfy the system with augmented
1
2
3
2
3
::
::
2
5
3
:
0
1
0
1
:
0
6
7
6
7
6
7
6
7
::
::
matrix 6 2
,
which
is
row
equivalent
to
. Hence x1 D x3
7
6
5
3 : 0 5
1 : 0 7
4
4 0 1
5
::
::
2
1
3 : 0
0 0
0 : 0

204 Chapter 10 Linear Systems of Differential Equations


2

3
1
and x2 D x3 . Taking x3 D 1 yields y2 D 4 1 5e 4t . For a third solution we need a vector u such
1
2
32
3
2 3
2
5
3
u1
1
5
3 5 4 u2 5 D 4 1 5. The augmented matrix of this system is row equivalent to
that 4 2
2
1
3
u3
1
2
3
:: 1
2
3
2
3
1 : 2 7
1
1
6 1 0
4t
1
e
6
7
:
. Let u3 D 0, u1 D , and u2 D 0. Then y3 D 4 0 5
C 4 1 5te 4t . The
6 0 1
1 :: 0 7
4
5
2
2
0
1
:
0 0
0 :: 0
2 3
1
2
3
2
3
02 3
1
1
1
1
4t
e
C 4 1 5 te 4t A.
general solution is y D c1 4 1 5 e 2t C c2 4 1 5 e 4t C c3 @4 0 5
2
1
1
1
0



 
15 
9
12
9
x1
2
10.5.14.
D
.
3/
.
Hence

D
3.
Eigenvectors
satisfy
D
1
16
9 
16
12
x2


3
3
so x1 D x2 . Taking x2 D 4 yields y1 D
e 3t . For a second solution we need a vector u such
4
4


 





1
12
9
u1
3
1 e 3t
3
that
D
. Let u1 D and u2 D 0. Then y2 D
C
te 3t .
16
12
u2
4
0
4
4
4






  3t 
3
5
3
1 e
C
te 3t . Now y.0/ D
)
The general solution is y D c1
e 3t C c2
4
8
4
0
4


 1  





3
5
5
12
c1
C c2 4 D
, so c1 D 2 and c2 D 4. Therefore, y D
e 3t
te 3t .
4
0
8
8
16



 
7 
24
12 24
x1
2

D
.
5/
.
Hence

D
5.
Eigenvectors
satisfy
D
10.5.16.
1
6
17 
6 12
x2


2
so x1 D 2x2. Taking x2 D 1 yields y1 D
e 5t . For a second solution we need a vector u such
1









1
2
12 24
u1
1 e 5t
2
D
. Let u1 D and u2 D 0. Then y2 D
that
C
te 5t .
6 12
u2
1
0
1
6
6


  5t 




2
1 e
2
3
The general solution is y D c1
e 5t C c2
C
te 5t . Now y.0/ D
)
1
0
1
1
6


 1  





2
3
3
12
c1
C c2 6 D
, so c1 D 1 and c2 D 6. Therefore, y D
e 5t
te 5t .
1
1
1
6
0
2
3
1 
1
0
10.5.18. 4
1
1 
2 5 D . 1/. C 2/2 . Hence 1 D 1 and 2 D 3 D 2. The
1
1
1 
2
3
::
2
1
0
:
0
6
7
6
7
::
eigenvectors associated with 1 D 1 satisfy the system with augmented matrix 6 1
,
2
2 : 0 7
4
5
::
1
1
2 : 0
2
3
:
2
:
6 1 0 3 : 0 7
2
4
6
7
which is row equivalent to 6 0 1 4 ::: 0 7. Hence x1 D
x3 and x2 D
x3 . Taking
3
4
5
3
3
::
0 0 0 : 0

0
0


,

0
0


,

Section 10.5 Constant Coefficient Homogeneous Systems II

205

x
23 D 3 yields
1
6 1
6
6 1
1
4
1
1

The eigenvectors
associated with 2 D 2 2 satisfy the system
3
3 with augmented matrix
::
::
0 : 0 7
6 1 1 0 : 0 7
7
6
7
::
, which is row equivalent to 6 0 0 1 ::: 0 7. Hence x1 D x2 and
2 : 0 7
5
4
5
::
::
1 : 0
0 0 0 : 0
2
3
1
x3 D 0. Taking x2 D 1 yields y2 D 4 1 5e 2t . For a third solution we need a vector u such
0
3
2
32
3
2
1
1
0
u1
1
1
2 5 4 u2 5 D 4 1 5. The augmented matrix of this system is row equivathat 4 1
1
1
1
u3
0
2
3
::
2
3
1 7
1
6 1 1 0 :
6
7
lent to 6 0 0 1 :::
. Let u2 D 0, u1 D 1, and u3 D 1. Then y3 D 4 0 5e 2t C
1 7
4
5
1
:
0 0 0 ::
0
3
2
3
02
3
2
3
3
2
2
1
1
1
1
2
4 1 5te 2t . The general solution is y D c14 4 5e t Cc24 1 5e 2t Cc3 @4 0 5e 2t C 4 1 5te
0
0
1
0
3
2
3
2
3
2
3
2
3 2
3
6
2
1
1
6
Now y.0/ D 4 5 5 ) c14 4 5 C c2 4 1 5 C c34 0 5 D 4 5 5, so c1 D 2, c2 D 3,
7
3
0
1
7
2
3
2
3
2
3
4
2
1
and c3 D 1. Therefore, y D 4 8 5 e t C 4 3 5 e 2t C 4 1 5 te 2t .
6
1
0
2
3
7 
4
4
10.5.20. 4
1
0 
1 5 D . C 3/. 1/2 . Hence 1 D 3 and 2 D 3 D 1. The
9
5 6 
2
3
::
4 4 : 0 7
6 4
6
7
:
,
eigenvectors associated with 1 D 3 satisfy the system with augmented matrix 6 1
3 1 :: 0 7
4
5
:
9
5 9 :: 0
2
3
:
1 :: 0 7
6 1 0
6
7
:
which is row equivalent to 6 0 1
. Hence x1 D x3 and x2 D 0. Taking x3 D 1 yields
0 :: 0 7
4
5
:
0 0
0 :: 0
2
3
1
y1 D 4 0 5e 3t . The eigenvectors associated with 2 D 1 satisfy the system with augmented ma1
2
3
2
3
::
::
8
4
4
:
0
1
0
0
:
0
6
6
7
7
6
7
6
7
::
::
trix 6 1
,
which
is
row
equivalent
to
7
6
7. Hence x1 D 0
1
1
:
0
0
1
1
:
0
4
5
4
5
::
::
9
5
5 : 0
0 0
0 : 0

2t A

206 Chapter 10 Linear Systems of Differential Equations


2

3
0
and x2 D x3 . Taking x3 D 1 yields y2 D 4 1 5e t . For a third solution we need a vector u such
1
2
32
3
2 3
8
4 4
u1
0
1 1 5 4 u2 5 D 4 1 5. The augmented matrix of this system is row equivalent to
that 4 1
9
5 5
u
1
2
3 3
::
2
3
2 3
0 :
1 7
1
0
6 1 0
6
7
::
4 2 5e t C 4 1 5te t .
.
Let
u
D
0,
u
D
1,
and
u
D
2.
Then
y
D
6 0 1
7
3
1
2
3
1 :
2 5
4
0
1
::
0 0
0 :
0
2
3
2 3
02
3
2 3 1
0
1
0
1
The general solution is y D c14 0 5e 3t C c2 4 1 5e t C c3 @4 2 5e t C 4 1 5te t A. Now
1
1
0
1
2 3
2
3
2
3
2
3
2
3
6
1
0
1
6
y.0/ D 4 9 5 ) c1 4 0 5 C c24 1 5 C c3 4 2 5 D 4 9 5, so c1 D 2, c2 D 1, and
1
1
1
0
1
2 3
2
3
2 3
4
0
2
c3 D 4. Therefore, y D 4 0 5 e 3t C 4 9 5 e t 4 4 5 te t .
1
4
2
2
3
4 
8
4
3
1 
3 5 D . C 4/. 8/2 . Hence 1 D 4 and 2 D 3 D 8. The
10.5.22. 4
1
1
9 
2
3
::
8
8
4
:
0
6
7
6
7
::
,
eigenvectors associated with 1 D 4 satisfy the system with augmented matrix 6 3
3
3 : 0 7
4
5
::
1
1 13 : 0
2
3
::
1 0 : 0 7
6 1
6
7
:
which is row equivalent to 6 0
. Hence x1 D x2 and x3 D 0. Taking x2 D 1 yields
0 1 :: 0 7
4
5
::
0
0 0 : 0
2 3
1
y1 D 4 1 5e t . The eigenvectors associated with 2 D 8 satisfy the system with augmented matrix
0
2
3
2
3
::
::
8
4 : 0 7
6 4
6 1 0 1 : 0 7
6
7
6
7
::
:
,
which
is
row
equivalent
to
6 3
6 0 1 0 :: 0 7. Hence x1 D x3 and
9
3 : 0 7
4
5
4
5
::
::
1
1
1 : 0
0 0 0 : 0
2
3
1
x2 D 0. Taking x3 D 1 yields y2 D 4 0 5e 8t . For a third solution we need a vector u such that
1
2
32
3
2
3
4
8
4
u1
1
4 3
9
3 5 4 u2 5 D 4 0 5. The augmented matrix of this system is row equivalent to
1
1
1
u3
1

Section 10.5 Constant Coefficient Homogeneous Systems II


2

6 1
6
6 0
4
0

:
0 1 ::
:
1 0 ::
:
0 0 ::

3
4
1
4

3
2
3
3
1
8t
1
e
. Then y3 D 4 1 5
C 4 0 5te 8t .
4
4
0
1
3
02
3
2
3
1
3
1
8t
e
5e 8t C c3 @4 1 5
C 4 0 5te 8t A. Now y.0/ D
4
0
1
3 2
3
3
4
4
1 5
D 4 1 5, so c1 D 1, c2 D 3, and c3 D 8.
4
3
0
3
2
8
C 4 0 5 te 8t .
8

7
3
7
7. Let u3 D 0, u1 D , and u2 D
5
4

2
3
1
1
The general solution is c1 4 1 5e t C c24 0
0
1
2
3
2
3
2
3
2
4
1
1
4 1 5 ) c1 4 1 5 C c2 4 0 5 C c3 4
3
0
1
3
3
2
2
1
3
Therefore, y D 4 1 5 e 4t C 4 2 5 e 8t
0
3
2
3
5 
1
1
1
9 
3 5 D .
10.5.24. 4
2
2
4 
2
:
1
1 ::
6 1
6
:
with augmented matrix 6 1
3
3 ::
4
:
2
2
2 ::
2

207

6/3 . Hence 1 D 6. The eigenvectors satisfy the system


3

3
::
: 0 7
7
::
.
: 0 7
5
::
: 0

0 7
0
6 1 0
7
6
, which is row equivalent to 6 0 1
1
0 7
5
4
0 0
0
0
2 3
0
Hence x1 D 0 and x2 D x3 . Taking x3 D 1 yields y1 D 4 1 5e 6t . For a second solution we need a
1
2
32
3 2
3
0
1
1
1
u1
vector u such that 4 1
3
3 5 4 u2 5 D 4 1 5. The augmented matrix of this system is row
1
2
2
2
u3
2
3
::
2
3
1
0 :
1
4 7
6 1 0
1
1
e 6t
6
7
::
1 7. Let u3 D 0, u1 D
equivalent to 6 0 1
, and u2 D . Then y2 D 4 1 5
C
1
:
4 5
4
4
4
4
0
:
0 0
0 ::
0
2
3
2
32
3 2 1 3
4
0
1
1
1
v1
6 1 7
4 1 5te 6t . For a third solution we need a vector v such that 4 1
3
3 5 4 v2 5 D 4 4 5.
1
2
2
2
v3
0
2
3
:: 1
0 : 8 7
6 1 0
6
:: 1 7. Let v D 0, v D 1 ,
The augmented matrix of this system is row equivalent to 6 0 1
3
1
1
: 8 7
4
5
8
::
0 0
0 : 0
2
3
2
3
2 3
1
1
0
6t
6t
2 6t
1
e
te
t e
and v2 D . Then y3 D 4 1 5
C4 1 5
C4 1 5
. The general solution is y D
8
8
4
2
0
0
1

208 Chapter 10 Linear Systems of Differential Equations


2

3
02
0
c1 4 1 5 e 6t C c2 @4
1

3
2
3
1
1
0
e 6t
1 5
C 4 1 5 te 6t A
4
0
1

2
3
1
e 6t
C4
Cc3 @4 1 5
8
0
02

3
2
3
1
1
0
te 6t
t 2 e 6t
A:
1 5
C4 1 5
4
2
0
1

3
4
2
1 
1 5 D .C2/3 . Hence 1 D 2. The eigenvectors satisfy the sys10.5.26. 4
2
3
1 
2
3
2
3
::
::
4
4
4
:
0
1
0
0
:
0
6
7
6
7
6
7
6
7
::
::
tem with augmented matrix 6 2
,
which
is
row
equivalent
to
7
6
7.
1
1 : 0 5
4
4 0 1 1 : 0 5
:
:
2
3
3 :: 0
0 0 0 :: 0
2
3
0
Hence x1 D 0 and x2 D x3 . Taking x3 D 1 yields y1 D 4 1 5e 2t . For a second solution we need a
1
3 2
3
32
2
0
4
4
4
u1
vector u such that 4 2
1
1 5 4 u2 5 D 4 1 5. The augmented matrix of this system is row
1
2
3
3
u3
2
3
::
2
3
1 7
1
6 1 0 0 :
6
7
::
equivalent to 6 0 1 1 :
. Let u3 D 0, u1 D 1, and u2 D 1. Then y2 D 4 1 5 e 2t C
1 7
4
5
0
:
0 0 0 ::
0
3
2
32
3
2
v1
4
4
4
0
4 1 5 te 2t . For a third solution we need a vector v such that 4 2
1
1 5 4 v2 5 D
v3
2
3
3
1
2
3
:
2
3
:
3
1
4 7
6 1 0 0 :
7
:
4 1 5. The augmented matrix of this system is row equivalent to 6
1 7. Let v3 D 0,
6 0 1 1 ::
2 5
4
0
:
0 0 0 ::
0
2
3
2
3
2
3
3
1
0
3
1
e 2t
t 2 e 2t
v1 D , and v2 D
. Then y3 D 4 2 5
C 4 1 5 te 2t C 4 1 5
. The general
4
2
4
2
0
0
1
2
3
02
3
2
3
1
0
1
0
solution is y D c1 4 1 5 e 2t C c2 @4 1 5 e 2t C 4 1 5 te 2t A
1
0
1
02
3
2
3
2
3
1
3
1
0
2t
2
2t
e
t
e
A
Cc3 @4 2 5
C 4 1 5 te 2t C 4 1 5
4
2
0
0
1
10.5.28. 4

2
2

12
24 
24

10
11
8

5D

. C 6/3 . Hence 1 D

6. The eigenvectors satisfy the

Section 10.5 Constant Coefficient Homogeneous Systems II


:
12 10 ::
:
18 11 ::
:
24 14 ::

6 4
6
system with augmented matrix 6 2
4

3
2
0 7
6 1
7
6
, which is row equivalent to 6 0
0 7
5
4

1
2

209
3
::
: 0 7
7
::
.
: 0 7
5
::
: 0

0 0
0
3
2
D 2 yields y1 D 4 1 5e 6t . For a second solu2
32
3
2
3
10
u1
2
11 5 4 u2 5 D 4 1 5. The augmented matrix
14
u3
2
3
::
1 :
1 7
7
:
::
1 7. Let u3 D 0, u1 D
1
1, and u2 D
2
6 5
::
0 :
0

x3
. Taking x3
2
2
4
12
tion we need a vector u such that 4 2
18
2
24
2
6 1 0
6
of this system is row equivalent to 6 0 1
4
0 0
2 3
2
3
6
2
1
e 6t
. Then y2 D 4 1 5
C 4 1 5 te 6t . For a third solution we need a vector v such that
6
6
0
2
3
2
2
32
3
4
12 10
v1
1
4 2
18 11 5 4 v2 5 D 4 16 5. The augmented matrix of this system is row equivalent to
2
24 14
v3
0
2
3
::
2
3
1
1 :
12
3 7
6 1 0
1
1
e 6t
6
7
:
1
::
1 7. Let v3 D 0, v1 D
, and v2 D
. Then y3 D 4 1 5
6 0 1
2
36 5
4
3
36
36
0
::
0 0
0 :
0
2
3
2
3
2
3
0 2 3
2
6
2
2
6
6t
te 6t
t 2 e 6t
e
4 1 5
C4 1 5
. The general solution is y D c1 4 1 5 e 6t C c2 @ 4 1 5
C4
6
2
6
0
2
2
0
Hence x1 D

x3 and x2 D

0 2

3
12
e 6t
Cc3 @ 4 1 5
36
0

10.5.30. 4


1
1

0
3

1
1


0

2
2

3
2
6
te 6t
4 1 5
C4
6
0

5D

:
1 ::
:
1 ::
:
1 ::

3
1
2
t 2 e 6t
A:
1 5
2
2

.C3/3 . Hence 1 D 3. The eigenvectors satisfy the sys3

3
::
: 0 7
7
::
.
: 0 7
5
::
: 0

0 7
6 1 0 1
7
6
,
which
is
row
equivalent
to
6 0 0 0
1 0
0 7
5
4
1 0
0
0 0 0
2
3
1
Hence x1 D x3 and x2 is arbitrary. Taking x2 D 0 and x3 D 1 yields y1 D 4 0 5 e 3t . Taking x2 D
1
2
3
0
1 and x3 D 0 yields y2 D 4 1 5 e 3t . For a third solution we need constants and and a vector u such
0

6
6
tem with augmented matrix 6
4

3
2
1 5 te
2

6t A

210 Chapter 10 Linear Systems of Differential Equations


2

32
3
2
3
2 3
1
u1
1
0
that 4
1 5 4 u2 5 D 4 0 5 C 4 1 5. The augmented matrix of this system is row
1
u3
1
0
3
::

6 1 0 1 :
7
6
7
:
:
equivalent to 6 0 0 0 : C 7; hence the system has a solution if D D 1, which yields
4
5
::
0 0 0 :
0
2
3
2 3
1
1
the eigenvector x3 D 4 1 5. Taking u1 D 1 and u2 D u3 D 0 yields the solution y3 D 4 0 5 e 3t C
1
0
2
3
2
3
2
3
02 3
2
1
0
1
1
4 1 5 te 3t . The general solution is y D c1 4 0 5 e 3t C c2 4 1 5 e 3t C c3 @4 0 5 e 3t C 4
1
1
0
0
2
3
3 
1
0
5 D . C 2/3 . Hence 1 D 2. The eigenvectors satisfy the
1
1 
0
10.5.32. 4
1
1
2 
2
3
2
3
::
::
1 0 : 0 7
6 1
6 1 1 0 : 0 7
6
7
6
7
::
:
system with augmented matrix 6 1
,
which
is
row
equivalent
to
6 0 0 0 :: 0 7.
1 0 : 0 7
4
5
4
5
::
::
1
1 0 : 0
0 0 0 : 0
2
3
1
Hence x1 D x2 and x3 is arbitrary. Taking x2 D 1 and x3 D 0 yields y1 D 4 1 5 e 2t . Taking x2 D
0
2
3
0
0 and x3 D 1 yields y2 D 4 0 5 e 2t . For a third solution we need constants and and a vector u such
1
32
3
2
3
2 3
2
1
0
u1
1
1 0
5
4
5
4
5
4
4
1
u
1
1
0
D
C 0 5. The augmented matrix of this system is row
that
2
0
1
u3
1
1 0
2
3
::

6 1 1 0 :
7
6
7
:
equivalent to 6 0 0 0 :: C 7; hence the system has a solution if D D 1, which yields
4
5
::
0 0 0 :
0
2
3
2
3
1
1
the eigenvector x3 D 4 1 5. Taking u1 D 1 and u2 D u3 D 0 yields the solution y3 D 4 0 5 e 2t C
1
0
2
3
2
3
2
3
02
3
2
1
1
0
1
4 1 5 te 2t . The general solution is y D c1 4 1 5 e 2t C c2 4 0 5 e 2t C c3 @4 0 5 e 2t C 4
1
0
1
0
1
1
1

0
0
0
2

10.5.34.

y03

Ay3

D .1 I
C.1 I
D

ue 1 t

A/ve 1 t C .1 I

A/ute 1 t C ue 1 t

t 2 e 1 t
C xte 1 t
2
xte 1 t C ue 1 t C 0 C xte 1 t D 0:

A/x

3
1
1 5 te
1

3
1
1 5 te
1

3t A

2t A

Section 10.6 Constant Coefficient Homogeneous Systems III

211

Now suppose that c1 y1 C c2 y2 C c3y3 D 0. Then




t2
c1x C c2 .u C tx/ C c3 v C tu C x D 0:
2

.A/

Differentiating this twice yields c3 x D 0, so c3 D 0 since x 0. Therefore,(A) reduces to (B) c1x C


c2 .u C tx/ D 0. Differentiating this yields c2 x D 0, so c2 D 0 since x 0. Therefore,(B) reduces to
c3 x D 0, so c1 D 0 since x 0. Therefore,y1 , y2 , and y3 are linearly independendent.
10.6

CONSTANT COEFFICIENT HOMOGENEOUS SYSTEMS III

10.6.2.
2
10
is 4

D . C 1/2 C 4. The augmented


9 
3
2
::
2i
4
: 0 5
, which is row equivalent to 4
:
26
10 2i :: 0

5 i
.5 i /x2 =13. Taking x2 D 13 yields the eigenvector x D
13


5 i
t
yields
e .cos 2t C i sin 2t//
13
11 
26

y D c1 e

5 
10.6.4.
3

5 cos 2t C sin 2t
13 cos 2t

D .


C c2 e

matrix of .A
1

5Ci
13

::
:
::
:

. 1 C 2i / I / x D 0
3

0 5
. Therefore,x1 D
0

0
0

. Taking real and imaginary parts of

5 sin 2t cos 2t
13 sin 2t

2/2 C 9. Hence,  D 2 C 3i is an eigenvalue of A. The associated


2
3 3i
6
eigenvectors satisfy .A .2 C 3i / I / x D 0. The augmented matrix of this system is 4
3
3 3i
2
3
::
1
1 i : 0 5
which is row equivalent to 4
. Therefore,x1 D .1 C i /x2 . Taking x2 D 1 yields
::
0
0
: 0


1Ci
x1 D 1 C i , so x D
is an eigenvector. Taking real and imaginary parts of e 2t .cos 3t C
1






sin 3t C cos 3t
cos 3t sin 3t
1Ci
C c2 e 2t
.
i sin 3t/
yields y D c1 e 2t
1
cos 3t
sin 3t

3 
3
1


1
5 
3 D . C 1/ . C 2/2 C 4 . The augmented matrix of .A C
10.6.6.

3
7
3 
3
2
3
2
::
::
2
3
1
:
0
1
0
1
:
0
6
7
6
7
6
7
6
7
::
::
I /x D 0 is 6 1
,
which
is
row
equivalent
to
7. Therefore,x1 D
7
6
4
3 : 0 5
4
4 0 1 1 : 0 5
::
::
3
7
4 : 0
0 0 0 : 0
2
3
1
x2 D x3 . Taking x3 D 1 yields y1 D 4 1 5 e t . The augmented matrix of .A . 2 C 2i /I / x D 0
1
1

3
::
: 0 5
,
::
: 0

212 Chapter 10 Linear Systems of Differential Equations


2

6
6
is 6
4

2i

1
2i

3
5

.1 C i /
x3 and x2 D
2

.1

2i
i/

::
:
::
:
::
:

3
2
0 7
6 1 0
7
6
, which is row equivalent to 6 0 1
0 7
5
4
0

0 0

x3 . Taking x3 D 2 yields the eigenvector x2 D


2

3
::
: 0 7
7
::
1 i
. Therefore,x1 D
: 0 7
2
5
::
0
: 0
2
3
1Ci
4 1 C i 5. The real
2
3
cos 2t sin 2t
cos 2t sin 2t 5 and
2 cos 2t
1Ci
2

3
2
1Ci
and imaginary parts of e 2t .cos 2t C i sin 2t/ 4 1 C i 5 are y2 D e 2t 4
2
3
2
sin 2t C cos 2t
y3 D e 2t 4 sin 2t C cos 2t 5. Therefore,
2 sin 2t
2
3
2
3
2
3
1
cos 2t sin 2t
sin 2t C cos 2t
y D c1 4 1 5 e t C c2e 2t 4 cos 2t sin 2t 5 C c3 e 2t 4 sin 2t C cos 2t 5:
1
2 cos 2t
2 sin 2t

10.6.8.

2
6 4
6
is 6 4
4
4

3

4
1 
2 D . 1/ . C 1/2 C 4 . The augmented matrix of .A I /x D 0
4
2
3 
3
2
3
::
::
1
3 : 0 7
6 1 0 1 : 0 7
7
7
6
::
:
,
which
is
row
equivalent
to
6 0 1 1 :: 0 7. Therefore,x1 D x2 D
2
2 : 0 7
5
4
5
::
::
2
2 : 0
0 0 0 : 0
2
3
1
x3 . Taking x3 D 1 yields y1 D 4 1 5 e t . The augmented matrix of .A . 1 C 2i /I / x D 0 is
1
2
3
3
2
::
::
1 i
1
3
: 0 7
: 0 7
2
6 2 2i
6 1 0
6
7
7
6
::
:
. Therefore,x1 D
, which is row equivalent to 6 0 1
6
4
2i
2
1 :: 0 7
: 0 7
4
5
5
4
::
::
4
2 4 2i : 0
0 0
0 : 0
2
3
1Ci
1 i
5. The real and
2
x3 and x2 D x3 . Taking x3 D 2 yields the eigenvector x2 D 4
2
2
2
3
2
3
1Ci
sin 2t cos 2t
5 are y2 D e t 4
5 and y3 D
imaginary parts of e t .cos 2t C i sin 2t/ 4
2
2 cos 2t
2
2 cos 2t
2
3
cos 2t sin 2t
5. Therefore,
2 sin 2t
e t4
2 sin 2t
2
3
2
3
2
3
1
sin 2t cos 2t
cos 2t sin 2t
5 C c3 e t 4
5:
2 cos 2t
2 sin 2t
y D c1 4 1 5 e t C c2 e t 4
1
2 cos 2t
2 sin 2t
3

Section 10.6 Constant Coefficient Homogeneous Systems III

213

1 7 3
5
10.6.10.
D . 2/2 C 1. The augmented matrix of .A .2 C i /I /x D 0 is

2
5 3
3
2
3
2
3
::
::
1C3i
1
5
: 0 5
1
: 0 5
2
4 1 3i
, which is row equivalent to 4
. Therefore,x1 D
::
::
3
2
1 3i : 0
0
0
: 0


1 C 3i
1 C 3i
x2 . Taking x2 D 2 yields the eigenvector x D
. Taking real and imaginary parts of
2
2


1 C 3i
e 2t .cos t C i sin t/
yields
2




cos t 3 sin t
sin t C 3 cos t
y D c1 e 2t
C c2 e 2t
:
2 cos t
2 sin t

34 

52

D . 2/2 C 16. The augmented matrix of .A .2 C 4i / I / x D 0


10.6.12.
20
30 
2
3
2
3
::
::
8Ci
:
0
32
4i
52
:
0
1
5
5, which is row equivalent to 4
5. Therefore,x1 D
is 4
::
::
20
32 4i : 0
0
0
: 0


.8 C i /
8 i
x2 . Taking x2 D 5 yields the eigenvector x D
. Taking real and imaginary parts of
5
5






8 i
sin 4t 8 cos 4t
cos 4t 8 sin 4t
2t
2t
2t
e .cos 4tCi sin 4t/
yields y D c1 e
C c2 e
.
5
5 cos 4t
5 sin 4t

3 
4
2


10.6.14.
5
7 
8 D . C 2/ . 2/2 C 9 . The augmented matrix of .A C

10
13
8 
2
3
2
3
::
::
5
4
2 : 0 7
2 : 0 7
6
6 1 0
6
7
6
7
::
:
2I /x D 0 is 6
,
which
is
row
equivalent
to
. Therefore,x1 D
6 0 1
5
9
8 : 0 7
2 :: 0 7
5
5
4
4
:
:
10 13
6 :: 0
0 0
0 :: 0
2
3
2
x2 D 2x3. Taking x3 D 1 yields y1 D 4 2 5 e 2t . The augmented matrix of .A .2 C 3i /I / x D 0 is
1
2
3
2
3
::
::
1
3i
4
2
:
0
1
0
1
i
:
0
6
7
6
7
6
7
6
7
::
::
,
which
is
row
equivalent
to
6
7
6
7. Therefore,x1 D
5
5
3i
8
:
0
0
1
i
:
0
4
5
4
5
::
::
10
13
10 3i : 0
0 0
0
: 0
2
3
1Ci
5. The real
.1 i /x3 and x2 D ix3 . Taking x3 D 1 yields the eigenvector x2 D 4
i
1
3
2
3
2
cos 3t sin 3t
1Ci
5 are y2 D e 2t 4
5 and
sin 3t
i
and imaginary parts of e 2t .cos 3t C i sin 3t/ 4
1
cos 3t

214 Chapter 10 Linear Systems of Differential Equations


2

3
sin 3t C cos 3t
5. Therefore,
y3 D c3 e 2t 4
cos 3t
sin 3t
2 3
2
3
2
2
cos 3t sin 3t
2t
2t
2t
5 C c3 e 4
y D c1 4 2 5 e
C c2 e 4
sin 3t
1
cos 3t

1 

10.6.16. 0
1
2
2
6 0 2
6
is 6 0 1
1
4
1 0
1

2
2

2
1



0

::
:
::
:
::
:

.

2/ .


1/2 C 1 . The augmented matrix of .A

3
2
0 7
6 1
7
6
, which is row equivalent to 6 0
0 7
5
4
0

0
1
0

1
Taking x3 D 1 yields y1 D 4 1 5 e t . The augmented matrix of .A
1
2

6 1 0
6
which is row equivalent to 6 0 1
4
0 0

::
:
::
:
::
:

3
sin 3t C cos 3t
5:
cos 3t
sin 3t

:
1 ::
:
1 ::
:
0 ::

I /x D 0

3
0 7
7
. Therefore,x1 D x2 D 1.
0 7
5
0

6
6
.1 C i /I / x D 0 is 6 0
4
1

i
0

1
1

0 7
.1 C i /
7
1Ci
. Therefore,x1 D .1 C i /x3 and x2 D
x3 .
0 7
2
5
2
0
0
2
3
2 C 2i
Taking x3 D 2 yields the eigenvector x2 D 4 1 C i 5. The real and imaginary parts of e 4t .cos t C
2
2
3
2
3
2
3
2 C 2i
2 cos t 2 sin t
2 sin t C 2 cos t
i sin t/ 4 1 C i 5 are y2 D e t 4 cos t sin t 5 and y3 D c3 e t 4 cos t C sin t 5. Therefore,
2
2 cos t
2 sin t
2 3
2
3
2
3
1
2 cos t 2 sin t
2 sin t C 2 cos t
y D c1 4 1 5 e t C c2e t 4 cos t sin t 5 C c3e t 4 cos t C sin t 5:
1
2 cos t
2 sin t

7 
15
10.6.18.
3 1 
2
::
3 6i
15
:
4
is
::
3
3 6i :

D . 4/2 C 36. The augmented matrix of .A .4 C 6i / I / x D 0

3
3
2
::
1
1
C
2i
:
0
0 5
5. Therefore,x1 D
, which is row equivalent to 4
::
0
0
0
: 0


1 2i
.1C2i /x2 . Taking x2 D 1 yields the eigenvector x D
. Taking real and imaginary parts of
1






1 2i
2 sin 6t cos 6t
2 cos 6t sin 6t
e 4t .cos 6tCi sin 6t/
yields y D c1e 4t
C c2 e 4t
.
1
cos 6t
sin 6t




 



5
1
2
c1
5
5 cos 6t C 5 sin 6t
Now y.0/ D
)
D
, so c1 D 1, c2 D 3, and y D e 4t
.
1
1
0
c2
1
cos 6t 3 sin 6t

::
:
::
:
::
:

3
0 7
7
,
0 7
5
0

Section 10.6 Constant Coefficient Homogeneous Systems III

215





1 4 6
1 2 1
1Ci
2
10.6.20.
D 
C . The augmented matrix of A
I x D 0

5
2 6
2
4
2 6
3
2
3 2
::
::
1C3i
1
1 3i
2
: 0 5
1
: 0 5
5
4
4
, which is row equivalent to
. Therefore,x1 D
is
::
::
6
5
1 3i : 0
0
0
: 0


1 C 3i
1 C 3i
x2 . Taking x2 D 5 yields the eigenvector x D
. Taking real and imaginary parts of
5
5






1 C 3i
cos t=2 3 sin t=2
sin t=2 C 3 cos t=2
t =2
t =2
t =2
e .cos t=2Ci sin t=2/
yields y D c1 e
C c2 e
.
5
5 cos t=2
5 sin t=2




 



1
2
cos.t=2/ C sin.t=2/
1
1 3
c1
1
, c2 D , and y D e t =2
.
Now y.0/ D
)
D
, so c1 D
cos.t=2/ C 2 sin.t=2/
1
5 0
c2
1
5
5

4 

4
0

10.6.22. 8
10 
20 D . 8/ . 2/2 C 4 . The augmented matrix of .A
2
3
2 
2
3
2
3
::
::
0 : 0 7
2 : 0 7
6 1 0
6 0 4
7
6
7
6
:
:
:
. Therefore,x1 D
8I /x D 0 is 6 8 6
, which is row equivalent to 6 0 1
2 :: 0 7
20 : 0 7
4
5
4
5
::
::
2 3
6 : 0
0 0
0 : 0
2
3
2
x2 D 2x3 . Taking x3 D 2 yields y1 D 4 2 5 e 8t . The augmented matrix of .A .2 C 2i /I / x D 0 is
1
2
3
2
3
::
::
2
2i
4
0
:
0
1
0
2
C
2i
:
0
6
7
6
7
6
7
6
7
::
::
,
which
is
row
equivalent
to
. Therefore,x1 D
6
7
6
8
8 2i
20
: 0 5
2i
: 0 7
4
4 0 1
5
::
::
2
3
4 2i : 0
0 0
0
: 0
2
3
2 2i
.2 2i /x3 and x2 D 2ix3 . Taking x3 D 1 yields the eigenvector x2 D 4 2i 5. The real and
1
2
3
2
3
2 cos 2t C 2 sin 2t
2 2i
5 and y3 D
2 sin 2t
imaginary parts of e 2t .cos 2t C i sin 2t/ 4 2i 5 are y2 D e 2t 4
2 cos 2t
1
2
3
2 3
2
3
2 sin 2t C 2 cos 2t
2
2 cos 2t C 2 sin 2t
5, so the general solution is y D c1 4 2 5 e 8t Cc2 e 2t 4
5C
e 2t 4
2 cos 2t
2 sin 2t
sin 2t
1
2 cos 2t
2
3
2 3
2
32
3 2
3
2 sin 2t C 2 cos 2t
8
2 2
2
c1
8
5. Now y.0/ D 4 6 5 ) 4 2 0
c3 e 2t 4
2 cos 2t
2 5 4 c2 5 D 4 6 5, so c1 D 2,
sin 2t
5
1 1
0
c3
5
2
3
2
3
4
4 cos 2t C 8 sin 2t
c2 D 3, c3 D 1, and y D 4 4 5 e 8t C e 2t 4 6 sin 2t C 2 cos 2t 5.
2
3 cos 2t C sin 2t

4 
4
4

10 3 
15 D . 8/.2 C 16/. The augmented matrix of .A 8I /x D 0 is
10.6.24.
2
3
1 

216 Chapter 10 Linear Systems of Differential Equations


3
2
3
::
0
1
0
2
:
0
6
7
6
7
6
7
6
7
::
,
which
is
row
equivalent
to
. Therefore,x1 D 2x3
6 10
7
6
5
0 5
1 : 0 7
4
4 0 1
5
:
2
3
0
0 0
0 :: 0
2
3
2
and x2 D x3 . Taking x3 D 1 yields y1 D 4 1 5 e 8t . The augmented matrix of .A 4iI / x D 0 is
1
2
3
2
3
::
::
4
4
: 0 7
1Ci : 0 7
6 4 4i
6 1 0
6
7
6
7
::
:
, which is row equivalent to 6 0 1
. Therefore,x1 D
6
10 3 4i
15
: 0 7
1 C 2i :: 0 7
4
5
4
5
::
::
2
3
1 4i : 0
0 0
0
: 0
2
3
1 i
.1 i /x3 and x2 D .1 2i /x3 . Taking x3 D 1 yields the eigenvector x2 D 4 1 2i 5. The
1
2
3
2
3
1 i
cos 4t C sin 4t
real and imaginary parts of .cos 4t C i sin 4t/ 4 1 2i 5 are y2 D 4 cos 4t C 2 sin 4t 5 and y3 D
1
cos 4t
3
2
2
3
2
3
2
cos 4t C sin 4t
sin 4t cos 4t
4 sin 4t 2 cos 4t 5, so the general solution is y D c1 4 1 5 e 8t C c2 4 cos 4t C 2 sin 4t 5 C
1
cos 4t
sin 4t
2
3
2
3
2
32
3 2
3
sin 4t cos 4t
16
2 1
1
c1
16
c3 4 sin 4t 2 cos 4t 5. Now y.0/ D 4 14 5 ) 4 1 1
2 5 4 c2 5 D 4 14 5, so c1 D 3,
sin 4t
6
1 1
0
c3
6
2
3
2
3
6
10 cos 4t 4 sin 4t
c2 D 3, c3 D 7, and y D 4 3 5 e 8t C 4 17 cos 4t sin 4t 5.
3
3 cos 4t 7 sin 4t
2

:
4 ::
:
15 ::
:
7 ::

10.6.28. (a) From the quadratic formula the roots are


p
kuk2 kv2k C .kuk2 kv2 k/2 C 4.u; v/2
k1 D
2.u; v/
p
2
2
kuk
kv k
.kuk2 kv2k/2 C 4.u; v/2
k2 D
:
2.u; v/

Clearly k1 > 0 and k2 < 0. Moreover,


k1 k2 D
(b) Since k2 D

.kuk2

kv2k/2



.kuk2 kv2k/2 C 4.u; v/2
D 1:
4.u; v/2

1= k1,
.2/

u1

D u

k2 v D u C

v.2/
1

D v C k2 u D v

1
1
1 .1/
vD
.v C k1 u/ D
v
k1
k1
k1 1
1
1
1 .1/
uD
.u k1 v/ D
u :
k1
k1
k1 1

Section 10.6 Constant Coefficient Homogeneous Systems III

217

2
3

::
10
15
5i
10
:
0
5,
D 2 C25. The augmented matrix of .A 5iI /x D 0 is 4
::
15 
25
15 5i : 0
2
3
:
3Ci
:: 0
1
5
5. Therefore,x1 D .3 i / x2 . Taking x2 D 5 yields the
which is row equivalent to 4
::
5
0
0
: 0






3 i
3
1
eigenvector x D
, so u D
and v D
. The quadratic equation is 3k 2 33kC3 D
5
5
0




:5257
:8507
0, with positive root k  :0902. Routine calculations yield U 
,V
.
:8507
:5257
2
3

::
3 

15
3
6i
15
:
0
D 2 C36. The augmented matrix of .A 6iI /x D 0 is 4
5,
10.6.32.
::
3
3 
3
3 6i : 0
2
3
::
1 1 2i : 0 5
which is row equivalent to 4
. Therefore,x1 D .1 2i /x2 . Taking x2 D 1
::
0
0
: 0






1 C 2i
1
2
yields the eigenvector x D
, so u D
and v D
. The quadratic equation is
1
1
0


:9732
2k 2 C 2k C 2 D 0, with positive root k  1:6180. Routine calculations yield U 
,
:2298


:2298
V
.
:9732

5 
12

10.6.34.
D . C 1/2 C 36. The augmented matrix of .A . 1 C 6i /I /x D 0
6
7 
2
3
2
3
::
::
6
6i
12
:
0
1
.1
C
i
/
:
0
5, which is row equivalent to 4
5. Therefore,x1 D
is 4
::
::
6
6 6i : 0
0
0
: 0

 



1Ci
1
1
.1 C i /x2 . Taking x2 D 1 yields the eigenvector x D
, so u D
and v D
.
1
1
0
2
The quadratic
equation
is
k  1 D 0, with positive root k  1:6180. Routine calculations yield


k
:5257
:8507
U
,V
.
:8507
:5257

4 
9

10.6.36.
D . C 1/2 C 36. The augmented matrix of .A . 1 C 6i /I /x D 0
5
2 
2
3
2
3
::
::
3 6i
3
6i
9
:
0
1
:
0
5
5, which is row equivalent to 4
5. Therefore,x1 D
is 4
::
::
5
3 6i : 0
0
0
: 0


 


3 6i
3 6i
3
6
x2 . Taking x2 D 5 yields the eigenvector x D
, so u D
and v D
. The
5
5
0
5
2
quadratic
 equation
 is 18k C 2kC 18 D 0, with positive root k  1:0571. Routine calculations yield
:8817
:4719
U
,V
.
:4719
:8817

1 
5
10.6.38.
D . C 1/2 C 100. The augmented matrix of .A . 1 C 10i /I /x D 0
20
1 

10.6.30.

15 
25

218 Chapter 10 Linear Systems of Differential Equations


3
3
2
::
i
i
:
0
0
1
2
5. Therefore,x1 D x2 .
5, which is row equivalent to 4
is 4
::
2
0 0 : 0
20
10i
0






i
0
1
Taking x2 D 2 yields the eigenvector x D
, so u D
and v D
. Since .u; v/ D 0 we
2
2
0
 
 
0
1
just normalize u and v to obtain U D
,VD
.
1
0

7 
6

D . C 1/2 C 36. The augmented matrix of .A . 1 C 6i /I /x D 0


10.6.40.
12
5 
2
3
2
3
::
::
1 i
6 6i
6
: 0 5
1
: 0 5
2
is 4
, which is row equivalent to 4
. Therefore,x1 D
::
::
12
6 6i : 0
0
0
: 0


 


1 i
1 i
1
1
x2 . Taking x2 D 2 yields the eigenvector x D
, so u D
and v D
. The
2
2
0
2
2
quadratic
4k C
 equation
 is k
 1 D 0, with positive root k  :2361. Routine calculations yield
:5257
:8507
U
,V
.
:8507
:5257
2

10.7

10i

::
:
::
:

VARIATION OF PARAMETERS FOR NONHOMOGENEOUS LINEAR SYSTEMS




1
3e t
e 2t
2e t
et
0
1
;
u
D
Y
f
D
e t
2e 2t
e 2t
3e 2t
5



4t
3t
3t
5e
13e C 3e
; yp D Y u D
.
6e t
e 3t 11e 3t




e 2t
e t
e 2t e 2t
0
1
10.7.4. Y D
; u D Y f D
2e 2t
e t
2e t
et




t
2t
t
2e
e
5 3e
uD
; yp D Y u D
.
e 2t 4e t
5e t 6



sin t
cos t
sin t cos t
0
1
10.7.6. Y D
; u D Y f D
cos t
sin t
cos t sin t




t sin t
t
uD
; yp D Y u D
.
t cos t
0
2
3
2
e 3t e 2t
e t
3e 3t
1
3t
t 5
0
1
4
4
10.7.8. Y D
e
0
3e
;u D Y f D
4e 2t
6
3t
2t
t
e
e
7e
et
2
3
2
3
3t
2t
2t
3t
3e
9e
9e
2e
14
1 4
8e t C 4e 2t 5; u D
16e t 4e 2t 5; yp D Y u D
6
12
e 2t e t
e 2t 2e t
10.7.2. Y D

e 2t
uD
2e 5t

50e 37
10e 3t

2
2e t



1
t

20e 4t 2e
10e 5t C 6e

2e 2t
2e 2t

t cos t C sin t
t sin t cos t

2e
4e t

2t
t

32
3
3e 3t
1
2t
2e 2t 5 4 e t 5 D
0
et
et
2
3
t
3e C 4
14
6e t 4 5.
6
10
2
3
2
32 t 3
et e t
te t
9e t
0
9e t
e
1 4
e t 3e t te t 5; u0 D Y 1 f D
3e t .3 2t/
6te t 9e t 5 4 e t 5 D
10.7.10. Y D 4 e t
18
et
e t
te t
6e t
6e t
0
et
2
3
2
3
2
3
0
0
2e t
1
1
1
4 e 2t .3 2t/ 5; u D 4 e 2t .2 t/ 5; yp D Y u D 4 e t 5.
3
3
3
2e 2t
e 2t
2e t
6e
6e

3t

Section 10.7 Variation of Parameters for Nonhomogeneous Linear Systems

219







1
1 e t .t C 1/
1
e t e t
t
e t .t C 2/
10.7.12. u0 D Y 1 f D
D
;
u
D
t
t
2
t
e
t
e .1 t/
e t .2 t/
2t  e
2
 t

 22 
t
t
t e
e
e .t C 2/
t
D
.
yp D Y u D
e t
e t .2 t/
2t
2 et

  2t 



1
1 2e 2t e 3t
1
2
e t
e
3e 2t C e 3t
10.7.14. u0 D Y 1 f D
D
;
u
D
t
2t
2t
3t
e
2
e
2e
e
e 3t 3e 2t
3 
9


 3 2t

2t
3t
3t
1 2 e t
1
3e C e
5e
e
D
.
yp D Y u D
2
e 3t 3e 2t
9 et
9 e 3t 5e 2t

 2




1
1 2e t C t 2
t
e t
t
1
t e t
0
1
D
;u D
10.7.16. u D Y f D 2
et
t
t2 1
t et
t 2 2e t
t  1
2




t
t
2
t
3
1
1 te .t C 2/ C t
t e
2e C t
2
yp D Y u D
D
.
t
2
t
t
3
e
t
t
2e
te
.t
2/
C
t
C
2
2
2
2
32 t 3
2
3
2
e 5t
e 4t e 3t
e
e 4t
e 4t
1
1
1
0
1
2t
t
t 5
4
5
4
5
4
4
10.7.18. u D Y f D
2e
e
1
0
D
2e
;uD
8e t
3
3
12
2t
t
t
e
2e
1
0
e
4e t
2 5t
32
3
2
3
2t
4t
t
e
e
0
e
3e
1 4 4t
1
e
0
e t 5 4 8e t 5 D 4 1 5.
yp D Y u D
12
4
e 3t
1
1
4e t
e t
10.7.20. u0 D Y

f D

1 4
2t

2 t
e
1
4 et
yp D Y u D Y D
4t
et

5;

e 2t
1
2

e t
e t
e t

32 t 3
2
3
2
3
te t te t C e 2t
e
1
2t
1
te t
te t 1 5 4 0 5 D 4 e 2t 5; u D 4 e 2t 5;
4
t
0
2
e
0
0
32
3
2
3
t
2t
2t C 1
et
4 2t 1 5.
e t 5 4 e 2t 5 D
4t
0
0
2t C 1

10.7.22. (c) If yp D Y u, then y0p D Y 0 u C Y u0 D AY u C Y u0 , so (E) y0p D Ayp C Y u0 . However, from


the derivation of the method of variation of parameters in Section 9.4, Y u0 D f as defined in the solution
of (a). This and (E) imply the conclusion.
(d) Since Y u0 D f with f as defined in the solution of (a), u1 ; u2 ; : : : ; un satisfy the conditions required
in the derivation of the method of variation of parameters in Section 9.4; hence, yp D c1y1 C c2 y2 C
   C cn yn is a particular solution of (A).

CHAPTER 11
Boundary Value Problems and Fourier
Expansions
11.1

EIGENVALUE PROBLEMS FOR y 00 C y D 0

11.1.2. From Theorem 11.1.2 with L D  , n D n2 , yn D sin nx, n D 1; 2; 3; : : :


1/2

.2n

11.1.4. From Theorem 11.1.4 with L D  , n D

, yn D sin

.2n

1/x
2

, n D 1; 2; 3; : : : ;

11.1.6. From Theorem 11.1.6 with L D  , 0 D 0, y0 D 1, n D n2 , y1n D cos nx, y2n D sin nx,
n D 1; 2; 3; : : :
11.1.8. From Theorem 11.1.5 with L D 1, n D

.2n

.2n 1/ x
1/2  2
, yn D cos
, n D 1; 2; 3; : : :
4
2

11.1.10. From Theorem 11.1.6 with L D 1, 0 D 0, y0 D 1, n D n2  2 , y1n D cos n x, y2n D


sin n x, n D 1; 2; 3; : : :
11.1.12. From Theorem 11.1.6 with L D 2, 0 D 0, y0 D 1, n D
n x
sin
, n D 1; 2; 3; : : :
2
11.1.14. From Theorem 11.1.5 with L D 3, n D
11.1.16. From Theorem 11.1.3 with L D 5, n D

.2n

n x
n2  2
, y1n D cos
, y2n D
4
2

1/2  2
.2n 1/ x
, yn D cos
, n D 1; 2; 3; : : :
36
6

n2  2
n x
, yn D cos
, n D 1; 2; 3; : : :
25
5

11.1.18. From Theorem 11.1.1, any eigenvalues of Problemp11.1.4 must beppositive. If  > 0, then
every solution of y 00 C y D
cos  x C c2 sin  x where c1 and c2 are
p 0 is of thepform y D c1p
0
constants. Therefore,
y
D

.
c
sin

x
C
c
cos
 x/. Sincep
y 0 .0/ D 0, c2 D 0. Therefore,
1
2
p
p
y D c1 cos  x. Since y.L/ D 0, c1 cos  L D 0. To make c1 cos  L D 0 with c1 0 we must
p
.2n 1/
.2n 1/2  2
choose  D
, where n is a positive integer. Therefore,n D
is an eigenvalue
2L
4L2
.2n 1/ x
and yn D cos
is an associated eigenfunction.
2L

221

222 Chapter 11 Boundary Value Problems and Fourier Expansions

rx
L
r  x L
dx D
sin
D 0, so y0 D 1 is orthogonal
11.1.20. If r is a positive integer, then
cos
L
r
L L
L
Z L
m x
n x
to all the other eigenfunctions. If m and n are distinct positive integers, then
cos
cos
dx D
L
L
0
Z L
1
m x
n x
cos
cos
dx D 0, from Example 11.1.4.
2 L
L
L
Z

11.1.22. Let m and n be distinct positive integers. From the identity cos A cos B D

cos.A C B/ with A D .2m


Z

cos
0

.2m

1/ x=2L and B D .2n

1/ x
.2n 1/ x
1
cos
dx D
2L
2L
2

L
0

1/ x=2L,


.m
cos

1
cos.A
2

n/ x
.m C n 1/ x
C cos
L
L

11.1.24. If y D c1 C c2 x, then y 0 .0/ D 0 implies that c2 D 0, so y D c1 . Now


Z L
c1
dx D c1 L D 0 only if c1 D 0. Therefore,zero is not an eigenvalue.

B/ C

dx D 0:

L
0

y.x/ dx D

If y D c1 cosh kx C c2 sinh kx, then y 0 .0/ D 0 implies that c2 D 0, so y D c1 cosh kx. Now
Z L
Z L
sinh kL
y.x/ dx D c1
cosh kx dx D c1
D 0 with k > 0 only if c1 D 0. Therefore, there are no
k
0
0
negative eigenvalues.
If y D c1 cos kx C c2 sin kx, then y 0 .0/ D 0 implies that c2 D 0, so y D c1 cos kx. Now
Z L
Z L
sin kL
n
y.x/ dx D c1
cos kx dx D c2
D 0 if k D
, where n is a positive integer. Therefore,
k
L
0
0
n x
n2  2
n D
and yn D cos
, n D 1; 2; 3; : : : .
2
L
L
Z L
11.1.26. If y D c1 C c2 .x L/, then y 0 .L/ D 0 implies that c2 D 0, so y D c1 . Now
y.x/ dx D
0
Z L
c1
dx D c1 L D 0 only if c1 D 0. Therefore,zero is not an eigenvalue.
0

If y D c1 cosh k.x L/ C c2 sinh k.x L/, then y 0 .L/ D 0 implies that c2 D 0, so y D c1 cosh k.x
Z L
Z L
sinh kL
L/. Now
y.x/ dx D c1
cosh k.x L/ dx D c1
D 0 with k > 0 only if c1 D 0.
k
0
0
Therefore,there are no negative eigenvalues.
If y D c1 cos k.x L/ C c2 sin k.x L/, then y 0 .L/ D 0 implies that c2 D 0, so y D c1 cos k.x L/.
Z L
Z L
sin kL
n
Now
y.x/ dx D c1
cos k.x L/ dx D c2
D 0 if k D
, where n is a positive integer.
k
L
0
0
n2  2
n.x L/
n x
Therefore, n D
and yn D cos
, or, equivalently, yn D cos
, n D 1; 2; 3; : : : .
2
L
L
L

Section 11.2 Fourier Expansions I


11.2 FOURIER EXPANSIONS I
11.2.2.
1
2
Z

a0

an

bn

.2
1

x/ dx D

.2
1
1

.2
1
"

2 dx D 2I

x/ cos n x dx D 4

4
sin n x D 0I
cos n x dx D
n
0

1
0

x sin n x dx
#

1
2 X . 1/n
sin n x. From Theorem 11.2.4,
 nD1 n

8
<

11.2.4. Since f is even, bn D 0 for n  1; a0 D


D
D
D
F .x/ D

x/ sin n x dx D 2

F .x/ D

an

1 Z 1

2
x cos n x
cos n x dx
n
0
0
"
1 #

2
1
2
cos n
sin n x D . 1/n
I
n
n
n
0

F .x/ D 2 C

.1

3x 2 / cos n x dx D

2;
x;
2;
Z

x D 1;
1 < x < 1;
x D 1:

.1
0

"

2
.1
n

3x 2 / dx D .x

x 3 / D 0; if n  1, then
0

#
1
Z 1

3x / sin n x C 6
x sin n x dx
0
0
#

"
1 Z 1

12

x
cos
n
x
cos n x dx

2
2
n 
0
0
"
1 #

12
1
12
cos n
sin n x D . 1/nC1 2 2 I
2
2
n 
n
n 
0

1
12 X
cos n x
. 1/n
. From Theorem 11.2.4, F .x/ D 1
2
 nD1
n2

11.2.6. Since f is odd, an D 0 if n  0;


Z
Z
2 
1 
b1 D
x cos x sin x dx D
x sin 2x dx
 0
 0
 Z 




1
1
D
x cos 2x
cos 2x dx D

2
2
0
0

3x 2 , 1  x  1.


sin 2x 
D
2 0

1
:
2

223

224 Chapter 11 Boundary Value Problems and Fourier Expansions


if n  2, then
bn

Z
1 
x cos x sin nx dx D
xsin.n C 1/x C sin.n 1/x dx
 0
0
 



Z 
1
cos.n C 1/x
cos.n 1/x 
cos.n C 1/x
cos.n 1/x
x
C
C
dx


nC1
n 1
nC1
n 1
0
0




1
1
1 sin.n C 1/x
sin.n 1/x 
2n
. 1/n
C
C
C
D . 1/n 2
I

2
2
nC1
n 1

.n C 1/
.n 1/
n
1
0
2


D
D
D

X
1
n
sin x C 2
. 1/n 2
sin nx. From Theorem 11.2.4, F .x/ D x cos x,   x   .
2
n
1

F .x/ D

nD2

 Z 


Z

1 
1
11.2.8. Since f is even, bn D 0 if n  1; a0 D
x sin x dx D
x cos x
cos x dx D
 0

0

0 Z 


Z
Z

2 
1 
1
sin x 

D 1; a1 D
x sin x cos x dx D
x sin 2x dx D
x cos 2x
cos 2x dx D
1C
 0
 0
 0
2
0
0

1
sin 2x
1
C
D
; if n  2, then

2
4 0
2
Z
Z
1 
2 
an D
x sin x cos nx dx D
xsin.n C 1/x sin.n 1/x dx
 0
 0
 



Z 
cos.n 1/x cos.n C 1/x 
cos.n 1/x cos.n C 1/x
1
D
x
dx


n 1
nC1
n 1
nC1
0
0




1
1
1 sin.n 1/x sin.n C 1/x 
2
D . 1/nC1
D . 1/nC1 2
I

2
2
n 1 nC1

.n 1/
.n C 1/
n
1
0
1
X
. 1/n
2
cos nx. From Theorem 11.2.4, F .x/ D x sin x,   x   .
n2 1

1
cos x
2

F .x/ D 1

nD2

1
sin  x 1=2
11.2.10. Since f is even, bn D 0 if n  0; a0 D
cos  x dx D
D ; a1 D



0
0

Z 1=2
Z 1=2
1 sin 2 x 1=2
1
2
2
cos  x dx D
.1 cos 2 x/ dx D
D ; if n  2, then
2
2 0
2
0
0
Z

an

D 2

D
D

F .x/ D

1


1=2

cos  x cos n x dx D

1=2

1=2

cos.n C 1/x C cos.n

1/x dx




sin.n C 1/ x
sin.n 1/ x 1=2
1
1
1
n
C
D
cos

nC1
n 1

n
C
1
n
1
2
0
8
2
<
2
n
. 1/mC1
if n D 2m;
2
cos
D
.4m
1/
2
:
.n
1/
2
0
if n D 2m C 1I


1
1
C cos  x

2

1
2 X . 1/n
cos 2n x. From Theorem 11.2.4, F .x/ D f .x/, 1  x  1.

4n2 1
nD1

Section 11.2 Fourier Expansions I


11.2.12. Since f is odd, an D 0 if n  0; b1 D 2

1 sin 4 x 1=2
1
D ; if n  2, then
2
4 0
2
D 2

bn

1


F .x/ D

1
sin  x
2

sin 2 x dx D

1=2

sin  x sin n x dx D

1=2

sin.n
n

1/ x
1

cos.n

D
if n  2, then
an

D 2
D
D
D

1/ x

<

cos 4 x/ dx D

cos.n C 1/ x dx

if n D 2m;

if n D 2m C 1I

1=2

x sin  x cos  x dx D

"

1=2

1
x cos 2 x
2
0

1=2

x sin  x cos n x dx D

1  x < 12 ;
x D 21 ;
1
1
2 < x < 2;
1
x D 2;
1
< x  1:
2

0;
1
;
2
sin  x;
1
;
2
0;

11.2.14. Since f is even, bn D 0 if n  1;


"
1=2
Z 1=2

1
a0 D
x sin  x dx D
x cos  x

0
0
D 2

.1
0

1
4 X
n
. 1/n 2
sin 2n x. From Theorem 11.2.4,
 nD1
4n
1

F .x/ D

a1

1=2

1=2


sin.n C 1/ x 1=2

nC1
0
(
4m
mC1
2n
n
. 1/
cos
D
4m2 1
.n2 1/
2
0


225

1=2
0

1=2
0

sin  x 1=2
1
cos  x dx D
D 2I
 2 0


1=2
0

x sin 2x dx
#

1
sin 2 x 1=2
1
cos 2 x dx D
C
D
I
4
4 2 0
4

1=2
0

xsin.n C 1/ x

" 

1
cos.n 1/ x cos.n C 1/ x
x

n 1
nC1


1 cos.n 1/=2 cos.n C 1/=2
2
n 1
nC1

1=2

1=2 

sin.n

1/ x dx

cos.n C 1/ x
nC1
0
0


1 sin.n 1/=2 sin.n C 1/=2
2
.n 1/2
.n C 1/2
8
2
2 4m C 1

< . 1/m 2
if n D 2m;
1 n
n
2 n2 C 1
n
 .4m2 1/2
sin
C
cos
D
1 2m C 1

 n2 1
2
 2 .n2 1/2
2

if n D 2m C 1I
: . 1/m
4 m.m C 1/
Z

cos.n
n

1/ x
1

dx

226 Chapter 11 Boundary Value Problems and Fourier Expansions

F .x/ D

1
1
2
1
1
2 X
1 X
2n C 1
n 4n C 1
C
cos

xC
.
1/
cos
2n
xC
. 1/n
cos.2nC1/ x:
2
2
2
2

4

.4n
1/
4
n.n C 1/
nD1

nD1

From Theorem 11.2.4,

F .x/ D

0
2

11.2.16. Note that


x g.x/ dx D
1
Z 1
1
1
dx D , and if n  1, then
2 0
2
an D

0
2

x cos n x dx C

<

0;
1
;
4
x sin  x;
1
;
4
0;

1
x g. x/ dx; therefore, a0 D
2
2

1
2

x / cos n x dx D

.1
0

and
bn

D
D
D
D
D

1
0

Z

0
2

x dx C

.1
0

From Theorem 11.2.4,

1
1
1
1 X1
8 X
1
C
sin 2n x C 3
sin.2n C 1/ x:
2
 nD1 n
 nD0 .2n C 1/3

<

1
2;
2

x ;
1
F .x/ D
2;

1
x2;

:
1
;
2

x D 1;
1 < x < 0;
x D 0;
0 < x < 1;
x D 1:

x / dx D

sin n x 1
cos n x dx D
D0
n 0

x sin n x dx C
.1 x / sin n x dx D
.1 2x 2 / sin n x dx
1
0
0
"
#
1
Z 1

1
2
.1 2x / cos n x C 4
x cos n x dx
n
0
0
"
#
1 Z 1

4
1 C cos n
x sin n x
sin n x dx
n
n2  2
0
0

1 C cos n
4.1 cos n /
1 C cos n 4 cos n x 1
D
C

3
3
n
n 
n
n3  3
0
8
1

<
if n D 2m;
m
8

:
if n D 2m C 1I
.2m C 1/3  3

F .x/ D

1  x < 12 ;
x D 12 ;
1
1
2 < x < 2;
1
x D 2;
1
< x  1:
2

Section 11.2 Fourier Expansions I


11.2.18.

an

D
D

bn

227

Z 2

Z 2
Z 3
1
5
2 dx C
3 dx C
1 dx D . If n  1, then
a0 D
6
2
3
2
2
Z 3
1
n x
f .x/ cos
dx
3 3
3
Z 2

Z 2
Z 3
1
n x
n x
n x
3
2n
2 cos
dx C
3 cos
dx C
cos
dx D
sin
;
3
3
3
3
n
3
3
2
2
Z 2



Z 2
Z 3
1
n x
n x
n x
1
2n
2 sin
dx C
3 sin
dx C
sin
dx D
cos n cos
I
3
3
3
3
n
3
3
2
2

1
1
5
3 X1
2n
n x
1 X1
F .x/ D C
sin
cos
C
cos n
2
 nD1 n
3
3
 nD1 n

2n
cos
3

Z 
Z
1
e x 
sinh 
1
11.2.20. (a) a0 D
e x dx D
D
.
If
n

1
then
(A)
a
D
n
2 


Z  2 
1
(B) bn D
e x sin nx dx. Integrating (B) by parts yields
 



Z 

1 x
bn D
e sin nx
n
e x cos nx dx D nan :



Integrating (A) by parts yields


Z

1 x

an D
e cos nx C n




x

e sin nx dx D . 1/n

sin

n x
:
3

e x cos nx dx and


2 sinh 
2 sinh 
C nbn D . 1/n



.C/

n2 an ;

2 sinh  . 1/nC1 n
2 sinh  . 1/n
. Now (C) implies that bn D
. Therefore,
2

n C1

n2 C 1
!
1
1
X
X
sinh 
. 1/n
. 1/n n
F .x/ D
1C2
cos nx 2
sin nx :

n2 C 1
n2 C 1
nD1
nD1

from (C). Therefore, an D

!
1
X
1
sinh 
(b) From Theorem 11.2.4, F . / D cosh  , so
1C2
D cosh  , which implies

n2 C 1
nD1
the stated result.

Z
1 
sin kx 
sin k
11.2.24. Since f is even, bn D 0, n  1, a0 D
cos kx dx D
D
; if n  1 then
 0
k 0
k
Z
Z
2 
1 
an D
cos kx cos nx dx D
cos.n k/x C cos.n C k/x dx
 0
 0

 
1 sin.n k/x
sin.n C k/x
D
C


n k
nCk

 0
cos n sin kx
1
1
2k sin k
D
D . 1/nC1 2
I

nCk n k
.n
k 2 /
"
#
1
X
sin k 1
. 1/n
2k
cos nx .
F .x/ D

k
n2 k 2
nD1

228 Chapter 11 Boundary Value Problems and Fourier Expansions


11.2.26. Since f is continuous on L; L and f . L/ D f .L/, Theorem 11.2.4 implies that
f .x/ D a0 C
1
if a0 D
2L

L  x  L;

L
L

f .x/ dx, and, for n  1,

"
#

Z L
n x
1
n x L
n x
0
f .x/ cos
dx D
f .x/ sin
f .x/ sin
dx
L
n
L L
L
L
L
"
#
L
Z L
Z L

L
n
x
n
x
L
n x
0
00

f .x/ cos
f .x/ cos
dx D
f 00 .x/ cos
dx
2 2
n2  2
L L
L
n
L
L
L
1
L

an

1 
X
n x
n x 
an cos
C bn sin
;
L
L
nD1

(since f 0 . L/ D f 0 .L/), and


"
#

Z
Z L
n x
1
n x L
n x
1 L
0
bn D
f .x/ sin
dx D
f .x/ cos
f .x/ cos
dx
L L
L
n
L L
L
L
Z L
1
n x
D
f 0 .x/ cos
dx (since f . L/ D f .L/)
n L
L
"
#

Z L
Z L
L
n x L
n x
L
n x
00
0
D
f
.x/
sin
f
.x/
sin
dx
D
f 00 .x/ sin
dx:
2 2
n2  2
L L
L
n
L
L
L
If f 000 is integrable on L; L, then
Z L
n x
L
an D
f 00 .x/ cos
dx D
n2  2 L
L
Z L
n x
L2
D
f 000 .x/ sin
dx:
n3  3 L
L
1 
X

"
L2
n x
f 000 .x/ sin
n3  3
L

n x
f 000 .x/ sin
dx
L
L

n x 
n x
C bn sin
where
L
L
nD1
"Z
#
Z L
Z L
0
1
1
a0 D
f .x/ dx D
f .x/ dx C
f .x/ dx :
2L L
2L
L
0

11.2.28. The Fourier series is a0 C

Since

f .x/ dx D

1
an D
L
Since
Z

an cos

0
L

f .x C L/ dx D

n x
1
f .x/ cos
dx D
L
L
L

n x
f .x/ cos
dx
L
L

"Z

.A/

L
0

f .x/ dx, (A) implies that a0 D 0. If n  1, then

n x
f .x/ cos
dx C
L
L

n x
D
f .x C L/ cos
dx D
L
L
Z L
n x
D . 1/nC1
f .x/ cos
dx;
L
0

#
n x
f .x/ cos
dx :
L

f .x/ cos
0

n.x C L/
dx
L

.B/

Section 11.3 Fourier Expansions II


(B) implies that a2n
b2n D 0, n  1.

D An and a2n D 0, n  1. A similar argument showsthat b2n

229

D Bn and

m x
m x
11.2.30.(b) Let 0 D 1 and 2m D cos
, 2m 1 D sin
, m  1. Then c0 D a0 and c2m D am ,
L
L

Z L
Z L
Z 
1 L
2mx
2
2
c2m 1 D bm , m  1. Since
0 .x/ dx D 2L and
2m 1 .x/ dx D
1 cos
dx D
2 L
L
L

Z L
Z L L
1
2mx
2
1 C cos
dx D L, m  1, Exercise 12.2.29(d) implies the concluL,
2m
.x/ dx D
2 L
L
L
sion.
11.3 FOURIER EXPANSIONS II
Z 1
.1
11.3.2. a0 D
.1 x/ dx D

x/2 1
1
D ; if n  1,

2
2
0
0
"
#
1 Z 1
Z 1

D 2
.1 x/ cos n x dx D
.1 x/ sin n x C
sin n x dx
n
0
0
0
8
1
4
<

2
2
if n D 2m 1;
n
D
D
cos
n
x
1
.
1/

D
.2m
1/2  2

2 2
:
n2  2
n
0
0
if n D 2mI

an

C.x/ D

D
D
D

C.x/ D

D
D

cos k
k

1
L

.x 2

L2 / dx D

C.x/ D

1
L

x3
3

1
2k X 1

nD1

. 1/n cos k
cos nx:
n2 k 2

 L

2L2
; if n  1,
3
0
0
"
#

Z
Z L
2 L 2
n x
2
n x L
n x
2
2
2
.x
L / cos
dx D
.x
L / sin
2
x sin
dx
L 0
L
n
L 0
L
0
"
#

Z L
2
2
4L
n x L
n x
4L2
n x L
n 4L
n 4L
x
cos
cos
dx
D
.
1/
sin
D
.
1/
I

n2  2
L 0
L
n2  2 n3  3
L 0
n2  2
0

11.3.6. a0 D
an

1/ x:

cos kx 
1 cos k
sin kx dx D
; if n  1, then
D
k
k
0
0
Z
Z
1 
2 
sin kx cos nx dx D
sin.n C k/x sin.n k/x dx
 0
 0

 

1 cos.n k/ 1 cos.n C k/
1 cos.n k/x cos.n C k/x
D

n k
nCk
n k
nCk
0
2k1 . 1/n cos k
I
.n2 k 2 /

1
11.3.4. a0 D

an

1
4 X
1
1
C 2
cos.2n
2
 nD1 .2n 1/2

L2 x

1
2L2
4L2 X . 1/n
n x
C 2
cos
:
2
3

n
L
nD1

230 Chapter 11 Boundary Value Problems and Fourier Expansions




e 1
; if n  1, then
11.3.8. a0 D
e dx D e D

0
0



Z
Z 

2  x
2 x
x

an D
e cos nx dx D
e cos nx
n
e sin nx dx
 0

0
0



Z 

2
2
n 
x
2
x

D
. 1/ e
1 ne sin nx
n
e cos nx dx D . 1/n e 


0
0


.1 C n2 /an D

2
. 1/n e 


1; an D

C.x/ D

1
11.3.10. a0 D
L
an

.x

D
D

"
4L
.x
n2  2

e

1


1
2Lx/ dx D
L

.x 2

C.x/ D

11.3.12. bn D 2

1
2 X . 1/n e  1
cos nx:
 nD1 .n2 C 1/

x3
3

 L

D
Lx

x/ sin n x dx D

2
n

"

.1

2
L

L=2

sin
0

n x
dx D
L

S.x/ D

.x

n x
L/ sin
dx
L

1 Z

x/ cos n x C

2
n x L=2
2 h
cos
D
1

n
L 0
n

1
2 X1h
1
 nD1 n

cos

cos

cos n x dx
0

n i
;
2

n i
n x
sin
:
2
L

11.3.16.
b1

D
D

2


2

Z
Z
1 
x 2  1 
x sin2 x dx D
x.1 cos 2x/ dx D
x cos 2x dx
 0
2 0  0
0
 Z 




1

sin 2x 


x sin 2x
cos 2x dx D C
D 2I
2
2
4
0
0
0


4L3
n x L
4L2
sin
D
I
n3  3
L 0
n2  2

1
1
X

2
1
D 2 ; S.x/ D 2
sin
n
x
sin n x.

n2  2
n

n
0
nD1

11.3.14. bn D

2L2
; if n  1,
3

1
4L2 X 1
n x
2L2
C 2
cos
:
2
3
 nD1 n
L

.1

n2 an I

1;

n x
2
n x
2Lx/ cos
dx D
.x 2 2Lx/ sin
L
n
L
#
L Z L
n x
n x
4L2
L/ cos
cos
dx
D
L 0
L
n2  2
0

"

2
L

2
. 1/n e 
.n2 C 1/

2
C
n

Section 11.3 Fourier Expansions II


if n  2, then
bn

D
D
D
D

Z
Z
1 
2 
x sin x sin nx dx D
xcos.n 1/x cos.n C 1/x dx
 0
 0
 



Z 
1
sin.n 1/x sin.n C 1/x 
sin.n 1/x sin.n C 1/x
x
dx


n 1
nC1
n 1
nC1
0
0






1 cos.n 1/x cos.n C 1/x 
1
1
1
D
. 1/nC1 1

2
2
2
2

.n 1/
.n C 1/
 .n 1/
.n C 1/
8 0
0
if n D 2m 1;

 <
4n
nC1
16m
.
1/
1
D
if n D 2mI
:
.n2 1/2 
.4m2 1/
1
16 X
n
sin 2nx:

.4n2 1/2


S.x/ D sin x
2

2
11.3.18. cn D
L

cos

.2n

nD1

1/ x
4
.2n 1/ x L
4
dx D
sin
D . 1/nC1
;

2L
.2n 1/
2L
.2n 1/
0
1
4 X . 1/n
.2n 1/ x
cos
:

2n 1
2L

CM .x/ D

nD1

11.3.20.
dn

D 2
D
D
D

x cos

.2n

1/

"

1/

"

4
.2n
4
.2n

4
.2n

CM .x/ D

1/

1/ x
dx
2
.2n

x sin

nC1

. 1/


1/ x 1

2
0

. 1/n C

.2n
2

.2n

1/


1/

sin

.2n

cos
I

.2n

1/ x
2

dx

#
1/ x 1

2
0


1 
4 X
2
.2n 1/ x
. 1/n C
cos
:
 nD1
.2n 1/
2

11.3.22.
cn


Z
Z 
2 
.2n 1/x
1  cos.2n C 1/x
cos.2n 3/x
cos x cos
dx D
C
dx
 0
2
 0
2
2


2 sin.2n C 1/x=2 sin.2n 3/x=2 
D
C


2n C 1
2n 3
0


1
1
4.2n 1/
n 2
n
D . 1/
C
D . 1/
I
 2n C 1
2n 3
.2n 3/.2n C 1/
D

231

232 Chapter 11 Boundary Value Problems and Fourier Expansions

CM .x/ D

1
4 X
2n 1
.2n 1/x
. 1/n
cos
:

.2n 3/.2n C 1/
2
nD1

11.3.24.
2
L

cn

x 2 / cos

.Lx

.2n

"

1/ x
dx
2L

Z L
1/ x L
.2n 1/ x
.Lx x / sin
.L 2x/ sin
dx

.2n 1/
2L
2L
0
0
"
#

Z L
8L
.2n 1/ x L
.2n 1/ x
.L 2x/ cos
cos
dx
C2
.2n 1/2  2
2L
2L
0
0

8L2
32L2
.2n 1/ x L
32L2
.2n 1/ x
C
sin
D
sin

2
2
3
3
3
3
.2n 1/ 
.2n 1/ 
2L
.2n 1/ 
2
0
2
8L2
32L
C . 1/n 1
I
.2n 1/2  2
.2n 1/3  3
4

D
D
D
D

.2n



1
1
4. 1/n
.2n 1/ x
8L2 X
1
C
cos
:
2
.2n 1/2
.2n 1/
2L

CM .x/ D

nD1

11.3.26.
2
L

dn

x 2 sin

.2n

Z L
1/ x L
.2n 1/ x
x cos
2
x cos
dx

.2n 1/
2L
2L
0
0
"
#

Z L
.2n 1/ x L
.2n 1/ x
16L
x sin
sin
dx

.2n 1/2  2
2L
2L
0
0

16L2
32L2
.2n 1/ x L
. 1/nC1
C
cos

.2n 1/2  2
.2n 1/3  3
2L
0
16L2
32L2
nC1
. 1/
I
.2n 1/2  2 .2n 1/3  3
4

D
D
D
D

SM .x/ D

"

1/ x
dx
2L

.2n



1
16L2 X
1
2
.2n 1/ x
n
. 1/ C
sin
:
2
.2n 1/2
.2n 1/
2L
nD1

11.3.28.
dn

D
D
D


Z 
1  sin.2n C 1/x
sin.2n 3/x
cos x sin
dx D
C
dx
2
 0
2
2
0

 
2 cos.2n C 1/x=2
cos.2n 3/x=2
C


2n 1
2n 3
0


2
1
1
4.2n 1/
C
D
I
 2n C 1
2n 3
.2n 3/.2n C 1/
2


.2n

1/x

Section 11.3 Fourier Expansions II

SM .x/ D

233

1
4 X
2n 1
.2n 1/x
sin
:

.2n 3/.2n C 1/
2
nD1

11.3.30.
dn

2
L

x 2/ sin
"

.Lx
0

.2n

1/ x
dx
2L

Z L
1/ x L
.2n 1/ x
D
.Lx x / cos
.L 2x/ cos
dx

.2n 1/
2L
2L
0
0
"
#

Z L
8L
.2n 1/ x L
.2n 1/ x
.L 2x/ sin
sin
dx
D
C2
.2n 1/2  2
2L
2L
0
0

8L2
32L2
.2n 1/ x L
n
D . 1/
cos

.2n 1/2  2 .2n 1/3  3


2L
0
4

D . 1/n

.2n

32L2
8L2
C
I
2
2
.2n 1/ 
.2n 1/3  3



1
8L2 X
1
4
.2n 1/ x
n
SM .x/ D 2
. 1/ C
sin
:

.2n 1/2
.2n 1/
2L
nD1

Z
1 L
11.3.32. a0 D
.3x 4
L 0
and f 000 .x/ D 24.3x L/,
an

D
D

1
4Lx / dx D
L
3

3x 5
5

 L

D
Lx

"
48L3
.3x
.3x
n4  4
0

48L3
144L4
n x L
n
.
1/
2L
C
L
C
sin
D
n4  4
n5  5
L 0

48L2
n3  3

n x
L/ sin
dx D
L

C.x/ D

2L4
5

2L4
. Since f 0 .0/ D f 0 .L/ D 0
5

n x L
L/ cos
L 0

48L4
1 C . 1/n 2 ;
n4  4

L
0

n x
cos
dx
L

n  1I

1
48L4 X 1 C . 1/n 2
n x
cos
:
4
4
 nD1
n
L



Z
1 x 5 Lx 4
L2 x 3 L
L4
1 L 4
.x
2Lx 3 C L2 x 2 / dx D
C
D
. Since f 0 .0/ D

L 0
L 5
2
3
30
0
f 0 .L/ D 0 and f 000 .x/ D 12.2x L/,
"
#

Z
Z L
24L2 L
n x
24L3
n x L
n x
an D
.2x L/ sin
dx D
.2x L/ cos
2
cos
dx
n3  3 0
L
n4  4
L 0
L
0

24L3
48L4
n x L
24L4
n
D
.
1/
L
C
L
C
sin
D
1 C . 1/n
n4  4
n5  5
L 0
n4  4
8
<
0
if n D 2m 1;
D
n  1:
3L4
:
if n D 2m;
4
4
m 
11.3.34. a0 D

234 Chapter 11 Boundary Value Problems and Fourier Expansions


1
3L4 X 1
2n x
cos
:
4
4

n
L

L4
30

C.x/ D

nD1

11.3.36. Since f .0/ D f .L/ D 0 and f 00 .x/ D 2,

Z L
4L
n x
4L2
n x L
bn D
sin
dx D
cos
D
n2  2 0
L
n3  3
L 0
8
8L2
<
; if n D 2m 1;
3 3
D
: .2m 1/ 
0;
if n D 2mI
S.x/ D

4L2
.cos n
n3  2

1/

1
1
.2n 1/ x
8L2 X
sin
:
3
3

.2n 1/
L
nD1

11.3.38. Since f .0/ D f .L/ D 0 and f 00 .x/ D 6x,


"
Z
12L L
12L2
n x
n x
bn D 2 2
x sin
dx D
x cos
3
3
n  0
L
n 
L
S.x/ D

L
0

#
n x
12L3
cos
dx D . 1/nC1 3 3 I
L
n 

1
n x
12L3 X . 1/n
sin
:
3
3

n
L
nD1

11.3.40. Since f .0/ D f .L/ D f 00 .0/ D f 00 .L/ D 0 and f .4/ D 360x,


"
#

Z
Z L
720L3 L
n x
720L4
n x L
n x
bn D
x sin
dx D
x cos
cos
dx
n4  4 0
L
n5  5
L 0
L
0

720L5
720L5
n x L
720L5
D . 1/nC1 5 5 C 6 6 sin
D . 1/nC1 5 5 I

n 
n 
L 0
n 
S.x/ D

1
720L5 X . 1/n
n x
sin
:
5
n5
L
nD1

11.3.42. (a) Since f is continuous on 0; L and f .L/ D 0, Theorem 11.3.3 implies that
1
X
.2n 1/ x
f .x/ D
cn cos
, L  x  L, with
2L
nD1
cn

D
D
D
D
D

2
L

4
.2n

f .x/ cos
"

1/
4

.2n

1/ x
dx
2L

f .x/ sin
L

.2n

1/ x L

2L
0

L
0

f .x/ sin
0

.2n

1/ x
dx
2L

1/ x
dx (since f .L/ D 0)
2L
0
"
#

Z L
8L2
.2n 1/ x L
.2n 1/ x
0
00
f .x/ cos
f .x/ cos
dx

.2n 1/2  2
2L
2L
0
0
Z L
8L
.2n 1/ x
f 00 .x/ cos
dx (since f 0 .0/ D 0):
2
2
.2n 1/  0
2L
.2n

1/

f 0 .x/ sin

.2n

Section 11.3 Fourier Expansions II


(b) Continuing the integration by parts yields
"
#

Z L
16L2
.2n 1/ x L
.2n 1/ x
00
000
f .x/ sin
f .x/ sin
dx
cn D

.2n 1/3  3
2L
2L
0
0
Z L
16L2
.2n 1/ x
D
f 000 .x/ sin
dx:
.2n 1/3  3 0
2L
11.3.44. Since f 0 .0/ D f .L/ D 0 and f 00 .x/ D 2,
Z L
16L
.2n 1/ x
cn D
cos
dx
.2n 1/2  2 0
2L

32L2
.2n 1/ x L
32L2
nC1
D
sin
D
.
1/
I

.2n 1/3  3
2L
.2n 1/3  3
0

CM .x/ D

1
.2n 1/ x
32L2 X . 1/n
cos
:
3
3

.2n 1/
2L
nD1

11.3.46. Since f 0 .0/ D f .L/ D 0 and f 00 .x/ D 6.2x C L/,


Z L
48L
.2n 1/ x
cn D
.2x C L/ cos
dx
2
2
.2n 1/  0
2L
"
#

Z L
.2n 1/ x L
.2n 1/ x
96L2
D
.2x C L/ sin
2
sin
dx

.2n 1/3  3
2L
2L
0
0
"
#
96L2
4L
.2n 1/ x L
nC1
D
. 1/
3L
cos

.2n 1/3  3
.2n 1/
2L
0


4
96L3
. 1/n 3 C
I
D
.2n 1/3  3
.2n 1/


1
96L3 X
1
4
.2n 1/ x
n
CM .x/ D
. 1/ 3 C
cos
:
3
3
 nD1 .2n 1/
.2n 1/
2L
11.3.48. Since f 0 .0/ D f .L/ D f 00 .L/ D 0 and f 000 .x/ D 12.2x L/,
Z L
192L2
.2n 1/ x
cn D
.2x L/ sin
dx
3
3
.2n 1/  0
2L
"
#

Z L
384L3
.2n 1/ x L
.2n 1/ x
.2x L/ cos
2
cos
dx
D

.2n 1/4  4
2L
2L
0
0
"
#
384L3
4L
.2n 1/ x L
D
L
sin

.2n 1/4  4
.2n 1/
2L
0


4
n
384L
. 1/ 4
D
1C
I
.2n 1/4  4
.2n 1/


1
384L4 X
1
. 1/n 4
.2n 1/ x
CM .x/ D
1
C
cos
:
4
4

.2n 1/
.2n 1/
2L
nD1

235

236 Chapter 11 Boundary Value Problems and Fourier Expansions


11.3.50. (a) Since f is continuous on 0; L and f .0/ D 0, Theorem 11.3.4 implies that
1
X
.2n 1/ x
f .x/ D
dn sin
, L  x  L, with
2L
nD1
dn

D
D
D
D
D

2
L

.2n 1/ x
f .x/ sin
dx
2L
"
#

Z L
4
.2n 1/ x L
.2n 1/ x
0
f .x/ cos
f .x/ cos
dx

.2n 1/
2L
2L
0
0
Z L
.2n 1/ x
4
f 0 .x/ cos
dx (since f .0/ D 0)
.2n 1/ 0
2L
"
#
L Z L

8L
.2n
1/
x
.2n
1/
x

f 0 .x/ sin
f 00 .x/ sin
dx

.2n 1/2  2
2L
2L
0
0
Z L
.2n 1/ x
8L
f 00 .x/ sin
dx since f 0 .L/ D 0:
.2n 1/2  2 0
2L
0

(b) Continuing the integration by parts yields


"
#

Z L
16L2
.2n 1/ x L
.2n 1/ x
000
00
dn D
f .x/ cos
f .x/ cos
dx

.2n 1/3  3
2L
2L
0
0
Z L
.2n 1/ x
16L2
D
f 000 .x/ cos
dx:
3
3
.2n 1/  0
2L
11.3.52. Since f .0/ D f 0 .L/ D 0, and f 00 .x/ D 6.L
dn

D
D
D
D

2x/

Z L
48L
.2n 1/ x
.L 2x/ sin
dx
.2n 1/2  2 0
2L
"
#

Z L
.2n 1/ x L
.2n 1/ x
96L2
.L 2x/ cos
cos
dx
C2
.2n 1/3  3
2L
2L
0
0
"
#
96L2
4L
.2n 1/ x L
LC
sin

.2n 1/3  3
.2n 1/
2L
0


3
96L
4
1 C . 1/n
I
.2n 1/3  3
.2n 1/

SM .x/ D



1
96L3 X
1
4
.2n 1/ x
n
1 C . 1/
sin
:
3
3
 nD1 .2n 1/
.2n 1/
2L

11.3.54. Since f .0/ D f 0 .L/ D f 00 .0/ D 0 and f 000 .x/ D 6,


dn

D
D

96L2
.2n 1/3  3

1/ x
dx
2L
0

192L3
.2n 1/ x L
192L3
n
sin
D
.
1/
I

.2n 1/4  4
2L
.2n 1/4  4
0
cos

.2n

Section 11.3 Fourier Expansions II


1
192L3 X . 1/n
.2n 1/ x
sin
:
4
4

.2n 1/
2L

SM .x/ D

nD1

11.3.56. Since f .0/ D f 0 .L/ D f 00 .0/ D 0 and f 000 .x/ D 12.2x


D

dn

D
D
D
D

192L2
.2n 1/3  3

237

L/,

1/ x
dx
2L
0
"
#

Z L
384L3
.2n 1/ x L
.2n 1/ x
.2x L/ sin
2
sin
dx

.2n 1/4  4
2L
2L
0
0
"
#
384L3
4L
.2n 1/ x L
nC1
. 1/
LC
cos

.2n 1/4  4
.2n 1/
2L
0


4L
384L3
. 1/nC1 L
.2n 1/4  4
.2n 1/


4
384L
4
n
.
1/
C
I
.2n 1/4  4
.2n 1/
Z

.2x

L/ cos

.2n



1
1
4
.2n 1/ x
384L4 X
n
SM .x/ D
. 1/ C
sin
:
4
.2n 1/4
.2n 1/
2L
nD1

11.3.58. The Fourier sine series of f4 on 0; 2L is

1
X

Bn sin

nD1

1
Bn D
L

2L
0

n x
1
f4 .x/ sin
dx D
2L
L

Replacing x by 2L

x yields

"Z

L
0

n x
f .x/ sin
dx C
2L

2L

f .2L

n x
, where
2L

x/ sin

n x
dx D
2L

2L

f .2L
L

f .x/ sin
0

#
n x
x/ sin
dx :
2L

n.2L x/
dx. Since
2L

n.2L x/
n x
sin
D . 1/nC1 sin
,
2L
2L
Z 2L
Z L
n x
n x
f .2L x/ sin
dx D . 1/nC1
f .x/ sin
dx;
2L
2L
L
0
so
1 C . 1/nC1
Bn D
L

L
0

8
Z
.2m 1/ x
< 2 L
n x
f .x/ sin
dx
f .x/ sin
dx D
2L
: L 0
2L
0

Therefore,the Fourier sine series of f4 on 0; 2L is

1
X

dn sin

nD1

dn D

2
L

f .x/ sin
0

.2n

.2n

1/ x
with
2L

1/ x
dx:
2L

if n D 2m
if n D 2m:

1;

238 Chapter 11 Boundary Value Problems and Fourier Expansions

11.3.60. The Fourier cosine series of f4 on 0; 2L is A0 C


1
A0 D
2L

2L
0

"Z

1
f4 .x/ dx D
2L

L
0

f .x/ dx C

1
X

n x
, where
2L

An cos

nD1

2L

x/ dx D

f .2L
L

1
L

f .x/ dx
0

and
1
An D
L

2L
0

n x
1
f4 .x/ cos
dx D
2L
L

Replacing x by 2L
Z

2L

f .2L
L

"Z

L
0

n x
f .x/ cos
dx C
2L

2L

f .2L

#
n x
x/ cos
dx :
2L

x yields

n x
dx D
x/ cos
2L

n.2L x/
f .x/ cos
dx D
2L

f .x/ cos
0

n.2L x/
n x
n x
D cos n cos
D . 1/n cos
,
2L
2L
2L
8
Z
0
<
n x
1 C . 1/n L
Z
f .x/ cos
dx D
2 L
m x
An D
:
f .x/ cos
dx
L
2L
0
L 0
L

n.2L x/
dx:
2L

Since cos

Therefore,the Fourier cosine series of f4 on 0; 2L is A0 C

1
X

nD0

A2n cos

if n D 2m

if n D 2m:

1
X
n x
n x
= a0 C
an cos
.
L
L
nD0

CHAPTER 12
Fourier Solutions of Partial Differential

12.1

THE HEAT EQUATION

12.1.2. X.x/T .t/ satisfies ut D a2 uxx if X 00 C X D 0 and (A) T 0 D a2 T for the same value
of . The product also satisfies the boundary conditions u.0; t/ D ux .L; t/ D 0; t > 0, if and only
if X.0/ D X 0 .L/ D 0. Since we are interested in nontrivial solutions, X must be a nontrivial solution
of (B) X 00 C X D 0; X.0/ D 0; X 0 .L/ D 0. From Theorem 11.1.4, n D .2n 1/2  2=4L2 is an
.2n 1/ x
eigenvalue of (B) with associated eigenfunction Xn D sin
, n D 1; 2; 3; : : : . Substituting
2L
2 2
2
0
2 2 2
2
 D .2n 1/  =4L into (A) yields T D ..2n 1/  a =4L /T , which has the solution Tn D
2 2 2
2
e .2n 1/  a t =4L .
.2n 1/ x
2 2 2
2
, n D 1; 2; 3; : : :
We have now shown that the functions un .x; t/ D e .2n 1/  a t =4L sin
2L
2
satisfy ut D a uxx and the boundary conditions u.0; t/ D ux .L; t/ D 0; t > 0. Any finite sum
m
X
.2n 1/ x
2 2 2
2
dn e .2n 1/  a t =4L sin
also has these properties. Therefore,it is plausible to expect
2L
nD1
1
X
.2n 1/ x
2 2 2
2
that that this is also true of the infinite series (C) u.x; t/ D
dn e .2n 1/  a t =4L sin
2L
nD1
1
X
.2n 1/ x
under suitable conditions on the coefficients fdn g. Since u.x; 0/ D
dn sin
, if fdn g are
2L
nD1
the mixed Fourier sine coefficients of f on 0; L, then u.x; 0/ D f .x/ at all points x in 0; L where the
mixed Fourier sine series converges to f .x/. In this case (C) is a formal solution of the initial-boundary
value problem of Definition 12.1.3.
12.1.8. Since f .0/ D f .1/ D 0 and f 00 .x/ D 2, Theorem 11.3.5(b) implies that
n

D
D

S.x/ D

1
Z 1

4
4
D
sin
n
x
dx
D
cos
n
x

n2  2 0
n3  3
0
8
8
<
; if n D 2m 1;
.2m 1/3  3
:
0;
if n D 2mI

4
.cos n
n3  2

1
8 X
1
.2n 1/ x
sin
. From Definition 12.1.1,
3
3
 nD1 .2n 1/
L

u.x; t/ D

1
8 X
1
e
3
 nD1 .2n 1/3

239

.2n 1/2  2 t

sin.2n

1/ x:

1/

240 Chapter 12 Fourier Solutions of Partial Differential


12.1.10.
1

2


2

D
D

if n  2, then
n

D
D
D
D

S.x/ D

Z
Z
1 
x 2 
1 
x sin x dx D
x.1 cos 2x/ dx D
x cos 2x dx
 0
2 0  0
0



Z 

1

sin 2x 

x sin 2x
cos 2x dx D C
D I

2
2
4 0
2
0
0

Z
Z
2 
1 
x sin x sin nx dx D
xcos.n 1/x cos.n C 1/x dx
 0
 0
 



Z 
1
sin.n 1/x sin.n C 1/x 
sin.n 1/x sin.n C 1/x
x
dx


n 1
nC1
n 1
nC1
0
0

 




1 cos.n 1/x cos.n C 1/x
1
1
1
D
. 1/nC1 1

2
2
2
2

.n 1/
.n C 1/
 .n 1/
.n C 1/
8 0
0
if n D 2m 1;

 <
4n
nC1
16m
.
1/
1
D
if n D 2mI
:
.n2 1/2 
.4m2 1/


sin x
2

1
16 X
n
sin 2nx. From Definition 12.1.1,
 nD1 .4n2 1/2

u.x; t/ D


e
2

3t

sin x

1
16 X
n
e
 nD1 .4n2 1/2

12n2 t

sin 2nx:

12.1.12. Since f .0/ D f .L/ D 0 and f 00 .x/ D 6x, Theorem 11.3.5(b) implies that
"
#

Z 3
Z 3
36
n x
108
n x 3
n x
n D
x sin
dx D
x cos
cos
dx
n2  2 0
3
n3  3
3 0
3
0

108
n x 3
324
108
D . 1/nC1 3 3 C 4 4 sin
D . 1/nC1 3 3 I
n 
n 
3 0
n 
S.x/ D

1
324 X . 1/n
n x
sin
. From Definition 12.1.1, u.x; t/ D
3
3
 nD1 n
3

1
324 X . 1/n
e
 3 nD1 n3

4n2  2 t =9

sin

n x
.
3

12.1.14. Since f .0/ D f .1/ D f 00 .0/ D f 00 .L/ D 0 and f .4/ D 360x, Theorem 11.3.5(b) and
Exercise 35(b) of Section 11.3 imply that
"
#
1 Z 1
Z 1

720
720
n D
x sin n x dx D
x cos n x
cos n x dx
n4  4 0
n5  5
0
0

720
n x 1
720
nC1 720
D . 1/
C 6 6 sin
D . 1/nC1 5 5 I
n5  5
n 
L 0
n 
S.x/ D

1
720 X . 1/n
sin n x. From Definition 12.1.1, u.x; t/ D
5
n5
nD1

1
720 X . 1/n
e
5
n5
nD1

7n2  2 t

sin n x.

Section 12.1 The Heat Equation

241

12.1.16. Since f .0/ D f .1/ D f 00 .0/ D f 00 .L/ D 0 and f .4/ D 120.3x 1/, Theorem 11.3.5(b) and
Exercise 35(b) of Section 11.3 imply that
"
#
1
Z 1
Z 1

240
240
n D
.3x 1/ sin n x dx D
.3x 1/ cos n x
3
cos n x dx
n4  4 0
n5  5
0
0
1

240
720
240
n
D
. 1/ 2 C 1 C 6 6 sin n x D
1 C . 1/n 2 I
5 5
n5  5
n 
n
0
1
240 X 1 C . 1/n 2
sin n x. From Definition 12.1.1,
 5 nD1
n5

S.x/ D

1
240 X 1 C . 1/n 2
e
 5 nD1
n5

u.x; t/ D

12.1.18. 0 D
n

.x 2
0

4x/ dx D

1
L

x3
3

2x 2

 2

"

sin n x:

8
; if n  1,
3

n x
2
n x 2
2
2
.x
4x/ cos
dx D
.x
4x/ sin
2
n
2 0
0
"
#

Z 2
n x 2
n x
16
8
.x 2/ cos
cos
dx D 2 2

2
2
n 
2 0
2
n 
0

D
D

C.x/ D

1
2

2n2  2 t

1
8 16 X 1
n x
C 2
cos
. From Definition 12.1.3, u.x; t/ D
2
3  nD1 n
2

#
n x
2
.x 2/ sin
dx
2
0

32
n x 2
16
sin
D 2 2I

3
3
n 
2 0
n 
Z

1
8 16 X 1
C 2
e
3  nD1 n2

n2  2 t

cos

n x
.
2

1
1
.2n 1/ x
384 X
12.1.20. From Example 11.3.5, C.x/ D 4
cos
. From Definition 12.1.3,
 4 nD1 .2n 1/4
2
1
1
.2n 1/ x
384 X
2 2
u.x; t/ D 4
e 3.2n 1/  t =4 cos
.
 4 nD1 .2n 1/4
2


 1
Z

1 1
1 3x 5
2
4
3
4
12.1.22. 0 D
.3x
4Lx / dx D
x
D
. Since f 0 .0/ D f 0 .1/ D 0 and

L 0
L
5
5
0
f 000 .x/ D 24.3x 1/, Theorem 11.3.5(a) implies that
"
#
1
Z 1
Z 1

48
48

n D
.3x 1/ sin n x dx D
.3x 1/ cos n x
3
cos n x dx
n3  3 0
n4  4
0
0
1

48
144
48
n
D
D
.
1/
2
C
1
C
sin
n
x
1 C . 1/n 2 ; n  1I

4
4
5
5
4
4
n 
n 
n

0

C.x/ D

2
5

1
48 X 1 C . 1/n 2
cos n x. From Definition 12.1.3,
 4 nD1
n4

u.x; t/ D

2
5

1
48 X 1 C . 1/n 2
e
4
n4
nD1

3n2  2 t

cos n x:

242 Chapter 12 Fourier Solutions of Partial Differential




Z
1  4
1 x5  x4
 2x3
3
2 2
12.1.24. 0 D
.x
2 x C  x / dx D
C
 0
 5
2
3
0
000
f . / D 0 and f .x/ D 12.2x  /, Theorem 11.3.5(a) implies that


Z 

24
24

.2x

/
sin
nx
dx
D
.2x

/
cos
nx
n D

3
4
n  0
n 
0


24
48
24
n
n
D
. 1/  C  C 5 sin nx D
1 C . 1/
n4 
n 
n4
0
(
0
if n D 2m 1;
3
D
n  1I
if n D 2m;
m4
C.x/ D

4
30

1
X
1
4
cos
2nx.
From
Definition
12.1.3,
u.x;
t/
D
n4
30
nD1


4

D  . Since f 0 .0/ D

30
0
2

cos nx dx

1
X
1
e
n4
nD1

4n2 t

cos 2nx.

12.1.26.
2


D
D
D
D
SM .x/ D 8

.2n 1/x
x 2/ sin
dx
2
0



Z 
.2n 1/x 
.2n 1/x
4
2
. x x / cos
. 2x/ cos
dx

.2n 1/
2
2
0
0



Z 
8
.2n 1/x 
.2n 1/x
.
2x/
sin
C
2
sin
dx

.2n 1/2 
2
2
0
0


8
32
.2n 1/x
. 1/n
cos

.2n 1/2 .2n 1/3 


2
0
8
32
n
. 1/
C
I
.2n 1/2
.2n 1/3 

1
X

nD1

. x

1
.2n

1/2

u.x; t/ D 8

1
X

nD1

. 1/ C

.2n

1
.2n

1/2

1/

sin

.2n

12.1.28. Since f .0/ D f 0 .1/ D 0, and f 00 .x/ D 6.1


n

D
D
D
D

1/x
2

. 1/ C

.2n

1/

. From Definition 12.1.4,

3.2n 1/2 t =4

sin

.2n

1/x
2

2x/, Theorem 11.3.5(d) implies that

Z 1
48
.2n 1/ x
.1 2x/ sin
dx
2
2
.2n 1/  0
2
"
#

Z 1
96
.2n 1/ x 1
.2n 1/ x
.1 2x/ cos
cos
dx
C2
.2n 1/3  3
2
2
0
0
"
#
96
4
.2n 1/ x 1
1C
sin

.2n 1/3  3
.2n 1/
2
0


96
4
1 C . 1/n
I
.2n 1/3  3
.2n 1/

Section 12.1 The Heat Equation

SM .x/ D

243



1
96 X
1
4
.2n 1/ x
n
1
C
.
1/
sin
. From Definition 12.1.4,
3
3

.2n 1/
.2n 1/
2
nD1

u.x; t/ D



1
96 X
1
4
n
1
C
.
1/
e
 3 nD1 .2n 1/3
.2n 1/

.2n 1/2  2 t

sin

.2n

1/ x
:
2

12.1.30. Since f .0/ D f 0 .L/ D f 00 .0/ D 0 and f 000 .x/ D 6, Theorem 11.3.5(d) and Exercise 11.3.50(b)
imply that
Z 1
96
.2n 1/ x
cos
dx
n D
.2n 1/3  3 0
2
1
192
.2n 1/ x
192
D
sin
D . 1/n
I

4
4
.2n 1/ 
2
.2n 1/4  4
0

SM .x/ D

1
X

. 1/n
.2n 1/ x
192
sin
. From Definition 12.1.4,
4
 nD1 .2n 1/4
2
u.x; t/ D

192 X . 1/n
e
 4 nD1 .2n 1/4

.2n 1/2  2 t

sin

.2n

1/ x
:
2

12.1.32. Since f .0/ D f 0 .1/ D f 00 .0/ D 0 and f 000 .x/ D 12.2x 1/, Theorem 11.3.5(d) and Exercise 11.3.50(b) imply that
Z 1
192
.2n 1/ x
n D
.2x 1/ cos
dx
3
3
.2n 1/  0
2
"
#

Z 1
384
.2n 1/ x 1
.2n 1/ x
D
.2x 1/ sin
2
sin
dx

.2n 1/4  4
2
2
0
0
"
#
4
.2n 1/ x 1
384
nC1
D
. 1/
1C
cos

.2n 1/4  4
.2n 1/
2
0




384
384
4
4
nC1
n
D
. 1/
1
D
. 1/ C
I
.2n 1/4  4
.2n 1/
.2n 1/4  4
.2n 1/


1
384 X
1
4
.2n 1/ x
n
SM .x/ D 4
. 1/ C
sin
. From Definition 12.1.4,
4

.2n 1/
.2n 1/
2
nD1



1
384 X
1
4
n
u.x; t/ D 4
. 1/ C
e

.2n 1/4
.2n 1/

.2n 1/2  2 t

sin

.2n

nD1

12.1.36. From Example 11.3.3, CM .x/ D


u.x; t/ D

1
8 X
1
.2n 1/ x
cos
. From Definition 12.1.5,
2
2

.2n 1/
2
nD1

1
8 X
1
e
2

.2n 1/2
nD1

1/ x
:
2

3.2n 1/2  2 t =4

cos

.2n

1/ x
:
2

244 Chapter 12 Fourier Solutions of Partial Differential


12.1.38. Since f 0 .0/ D f . / D 0 and f 00 .x/ D 2, Theorem 11.3.5(c) implies that

Z 
16
.2n 1/x
32
.2n 1/x 
32
cos
dx
D
sin
D . 1/nC1
I
n D

2
3
.2n 1/  0
2
.2n 1/ 
2
.2n 1/3 
0
CM .x/ D

1
32 X . 1/n
.2n 1/x
cos
. From Definition 12.1.5,
 nD1 .2n 1/3
2

u.x; t/ D

1
32 X . 1/n
e
 nD1 .2n 1/3

7.2n 1/2 t =4

cos

.2n

1/x
2

12.1.40. Since f 0 .0/ D f .1/ D 0 and f 00 .x/ D 6.2x C 1/, Theorem 11.3.5(c) implies that
Z 1
48L
.2n 1/ x
n D
.2x C 1/ cos
dx
.2n 1/2  2 0
2
"
#

Z 1
.2n 1/ x 1
.2n 1/ x
96
D
.2x C 1/ sin
2
sin
dx

.2n 1/3  3
2
2
0
0
"
#
4
.2n 1/ x 1
96
nC1
D
. 1/
3
cos

.2n 1/3  3
.2n 1/
2
0


96
4
D
. 1/n 3 C
I
.2n 1/3  3
.2n 1/


1
1
4
.2n 1/ x
96 X
n
CM .x/ D 3
. 1/ 3 C
cos
. From Definition 12.1.5,

.2n 1/3
.2n 1/
2
nD1



1
96 X
1
4
n
u.x; t/ D 3
. 1/ 3 C
e

.2n 1/3
.2n 1/

.2n 1/2  2 t =4

cos

.2n

nD1

1/ x
:
2

12.1.42. Theorem 11.3.5(c) and Exercise 11.3.42(b) imply that Since f 0 .0/ D f .1/ D f 00 .1/ D 0 and
f 000 .x/ D 12.2x 1/,
Z 1
192
.2n 1/ x
n D
.2x 1/ sin
dx
3
3
.2n 1/  0
2
"
#

Z 1
384
.2n 1/ x 1
.2n 1/ x
D
.2x 1/ cos
2
cos
dx

.2n 1/4  4
2
2
0
0
"
#


384
4
.2n 1/ x 1
384
. 1/n 4
D
1
sin
D .2n 1/4  4 1 C .2n 1/ I
.2n 1/4  4
.2n 1/
2
0
CM .x/ D



1
384 X
1
. 1/n 4
.2n 1/ x
1C
cos
. From Definition 12.1.5,
4
4

.2n 1/
.2n 1/
2
nD1

u.x; t/ D



1
384 X
1
. 1/n 4
1
C
e
4
.2n 1/4
.2n 1/
nD1

.2n 1/2  2 t =4

cos

.2n

1/ x
:
2

Section 12.1 The Heat Equation

Z
2 L=2
n x
2
n x L=2
2 h
sin
dx D
cos
D
1
12.1.44. n D
L 0
L
n
L 0
n
1
2 X1h
n i
n x
S.x/ D
1 cos
sin
. From Definition 12.1.1,
 nD1 n
2
L
1
2 X1h
1
 nD1 n

u.x; t/ D

cos

n i
e
2

n2  2 t 2 =L2

cos

sin

245

n i
;
2

n x
:
L

12.1.46.
n

.2n 1/ x
4
.2n 1/ x L=2
sin
dx D
cos

2L
.2n 1/
2L
0
0


4
.2n 1/ /
1 cos
I
.2n 1/
4

D
SM .x/ D

2
L


1
4 X 1
1
 nD1 2n 1

u.x; t/ D

L=2

cos

.2n

1/ /
4


1
4 X 1
1

2n 1

cos

nD1

sin

.2n

.2n

1/ /
4

1/ x
. From Definition 12.1.4,
2L


.2n 1/2  2 a 2 t =4L2

sin

.2n

1/ x
:
2L

12.1.48. Let u.x; t/ D v.x; t/ C q.x/; then ut D vt and uxx D vxx C q 00 , so


vt D 9vxx C 9q 00 54x; 0 < x < 4; t > 0;
v.0; t/ D 1 q.0/; v.4; t/ D 61 q.4/; t > 0;
v.x; 0/ D 2 x C x 3 q.x/; 0  x  4:

.A/

We want q 00 .x/ D 6x; q.0/ D 1; q.4/ D 61; q.x/ D x 3 C a1 C a2 x; q.0/ D 1 ) a1 D 1;


q.x/ D x 3 C 1 C a2 x; q.4/ D 61 ) a2 D 1; q.x/ D x 3 C 1 x. Now (A) reduces to
vt D 9vxx ; 0 < x < 4; t > 0;
v.0; t/ D 0; v.4; t/ D 0; t > 0;
v.x; 0/ D 1; 0  x  4;
which we solve by separation of variables.
n

D
D

S.x/ D

n x
2
n x 4
sin
dx D
cos
4
n
4 0
0
8
4
<
2
if n D 2m 1;
1 . 1/n D
.2m 1/
:
n
0
if n D 2mI

1
2

1
4 X
1
.2n 1/ x
sin
. From Definition 12.1.1,
 nD1 .2n 1/
4

v.x; t/ D

1
4 X
1
e

.2n 1/
nD1

9 2 n2 t =16

sin

.2n

1/ x
:
4

246 Chapter 12 Fourier Solutions of Partial Differential


Therefore,
x C x3 C

u.x; t/ D 1

2 2
1
4 X e 9 n t =16
.2n 1/ x
sin
:
 nD1 .2n 1/
4

12.1.50. Let u.x; t/ D v.x; t/ C q.x/; then ut D vt and uxx D vxx C q 00 , so

vt D 3uxx C 3q 00 18x; 0 < x < 1; t > 0;


vx .0; t/ D 1 q 0 .0/; v.1; t/ D 1 q.1/; t > 0;
v.x; 0/ D x 3 2x q.x/; 0  x  1:

.A/

We want q 00 .x/ D 6x; q 0 .0/ D 1; q.1/ D 1; q 0 .x/ D 3x 2 C a2 ; q 0 .0/ D 1 ) a2 D 1;


q 0 .x/ D 3x 2 1; q.x/ D x 3 x C a1 x; q.1/ D 1 ) a1 D 1; q.x/ D x 3 x 1. Now (A) reduces
to
vt D 3uxx ; 0 < x < 1; t > 0;
vx .0; t/ D 0; v.1; t/ D 0; t > 0;
v.x; 0/ D 1 x; 0  x  1:
From Example 11.3.3, CM .x/ D
1
1
8 X
e
 2 nD1 .2n 1/2

nD1

1/2  2 t =4

3.2n

u.x; t/ D

1
8 X
1
.2n 1/ x
cos
. From Definition 12.1.5, v.x; t/ D
2
2

.2n 1/
2

cos

.2n

x C x3 C

1/ x
. Therefore,
2

1
8 X
1
e
2
 nD1 .2n 1/2

3.2n 1/2  2 t =4

.2n

cos

1/ x
:
2

12.1.52. Let u.x; t/ D v.x; t/ C q.x/; then ut D vt and uxx D vxx C q 00 , so

vt D vxx C q 00 C  2 sin  x; 0 < x < 1; t > 0;


v.0; t/ D q.0/; vx .1; t/ D  q 0 .1/; t > 0;
v.x; 0/ D 2 sin  x q.x/; 0  x  1:

.A/

We want q 00 .x/ D  2 sin  x; q.0/ D 0; q 0 .1/ D  ; q 0 .x/ D  cos  x Ca2; q 0 .1/ D  ) a2 D 0;


q 0 .x/ D  cos  x; q.x/ D sin  x C a1 ; q.0/ D 0 ) a1 D 0; q.x/ D sin  x. Now (A) reduces to
vt D vxx ; 0 < x < 1; t > 0;
v.0; t/ D 0; vx .1; t/ D 0; t > 0;
v.x; 0/ D sin  x; 0  x  1:
n

D 2

sin  x sin

.2n

1/ x
dx D
2

1
0


sin.2n 3/ x=2 sin.2n C 1/ x=2 1

.2n 3/
.2n C 1/
0


1
1
1
n 2
n8
D . 1/
D . 1/
 2n 3 2n C 1
 .2n C 1/.2n
D

SM .x/ D

2


1
8 X
. 1/n
 nD1 .2n C 1/.2n

v.x; t/ D

3/

sin

.2n

3/

1/ x
. From Definition 12.1.4,
2

1
8 X
. 1/n

.2n C 1/.2n
nD1

cos.2n C 1/ x
2

3/ x

cos.2n

3/

.2n 1/2  2 t =4

sin

.2n

1/ x
:
2

dx

Section 12.2 The Wave Equation


1
8 X
. 1/n

.2n C 1/.2n

Therefore, u.x; t/ D sin  x C

nD1

3/

.2n 1/2  2 t =4

sin

.2n

247

1/ x
.
2

12.1.54. (a) Since f is piecewise


smooth of 0; L, there is a constant K such that jf .x/j  K, 0  x 
Z
Z

2 L
n x
2 L
2 2 2

L. Therefore, jn j D
f .x/ sin
dx 
jf .x/j dx D 2K. Hence, jn e n  a t =L j 
L 0
L 0
L
2

2 2

2Ke n  a t =L , so u.x; t/ converges for all x if t > 0, by the comparison test.


n x
2 2
2
(b) Let t be a fixed positive number. Apply Theorem 12.1.2 with D x and wn .x/ D n e n  t =L sin
.
L

n x
2K
2 2
2
2 2
2
, so jwn0 .x/j 
ne n  t =L , 1 < x < 1. Since
Then wn0 .x/ D nn e n  t =L cos
L
L
L
1
X
n2  2 a 2 t =L2
ne
converges if t > 0, Theorem 12.1.1 (with 1 D x1 and 2 D x2 arbitrary) implies

nD1

the concclusion.
1
X
(c) Since
n2 e

n2  2 a 2 t =L2

nD1
n2  2 t =L2

also converges if t > 0, an argument like that in (b) with wn .x/ D

n x
yields the conclusion.
L
(d) Let x be arbitrary, but fixed. Apply Theorem 12.1.2 with D t and
n x
 2a2 2
n x
2 2 2
2
2 2 2
2
wn .t/ D n e n  a t =L sin
. Then wn0 .t/ D
n n e n  a t =L sin
, so jwn0 .t/j 
L
L2
L
1
X
2K 2 a2 2 n2  2 a 2t0 =L2
2 2 2
2
n e
if t > t0 . Since
n2 e n  a t0 =L converges, Theorem 12.1.1 (with 1 D
L
nD1
t0 > 0 and 2 D t1 arbitrary implies the conclusion for t  t0 . However, since t0 is an arbitrary positive
number, this holds for t > 0.

nn e

cos

12.2 THE WAVE EQUATION


hR
R1
1=2
12.2.1. n D 2 0 x sin n x C 1=2 .1
Z

1=2

x sin n x dx
0

D
D

1=2

.1
0

x/ sin n x dx

i
x/ sin n x dx ;

"
1=2

1
x cos n x
n
0

1=2

cos n x dx
0

1=2

n
1
1
cos
C 2 2 sin n x
D
2n
2
n 
0
D
D

1
n

"

.1

1
n
cos
2n
2

x/ cos n x

1=2

1
n
1
n
cos
C 2 2 sin
I
2n
2
n 
2

1=2

cos n x dx
0

1
1
n
1
n

sin
n
x
D
cos
C 2 2 sin
I

2
2
n 
2n
2
n 
2
1=2

8
4
<
4
n
. 1/mC1
bn D 2 2 sin
D
.2m 1/2  2
:
n 
2
0

if n D 2m

if n D 2mI

248 Chapter 12 Fourier Solutions of Partial Differential Equations

Sg .x/ D

1
4 X . 1/nC1
sin.2n
2
.2n 1/2

1/ x. From Definition 12.1.1,

nD1

1
4 X . 1/nC1
u.x; t/ D
sin 3.2n
3 3 nD1 .2n 1/3

1/ t sin.2n

1/ x:

12.2.2. Since f .0/ D f .1/ D 0 and f 00 .x/ D 2, Theorem 11.3.5(b) implies that
1
Z 1

4
4
4
sin n x dx D
cos n x D
.cos n
n D
3 2
n2  2 0
n3  3
n
0
8
8
<
; if n D 2m 1;
D
.2m 1/3  3
:
0;
if n D 2mI

Sf .x/ D

1
1
8 X
sin.2n
3
.2n 1/3

1/ x. From Definition 12.1.1,

nD1

u.x; t/ D

1
8 X
1
cos 3.2n
3
 nD1 .2n 1/3

1/ t sin.2n

1/ x:

12.2.4. Since g.0/ D g.1/ D 0 and g00 .x/ D 2, Theorem 11.3.5(b) implies that
1
Z 1

4
4
4
n D
sin n x dx D
cos n x D
.cos n
2
2
3
3
3
n  0
n 
n 2
0
8
8
<
; if n D 2m 1;
D
.2m 1/3  3
:
0;
if n D 2mI

Sg .x/ D

1/

1
8 X
1
sin.2n
3

.2n 1/3

1/

1/ x. From Definition 12.1.1,

nD1

1
1
8 X
u.x; t/ D
sin 3.2n
3 4 nD1 .2n 1/4

12.2.6. From Example 11.2.6, Sf .x/ D


u.x; t/ D

12.2.8. From Example 11.2.6 Sg .x/ D


u.x; t/ D

1/ t sin.2n

1/ x:

1
324 X . 1/n
n x
sin
. From Definition 12.1.1,
3
3

n
3
nD1

1
324 X . 1/n
8 nt
n x
cos
sin
:
3
3
 nD1 n
3
3
1
324 X . 1/n
n x
sin
. From Definition 12.1.1,
3
n3
3
nD1

1
81 X . 1/n
8 nt
n x
sin
sin
:
4
4
2
n
3
3
nD1

Section 12.2 The Wave Equation


12.2.10.
1

2


2

D
D

if n  2, then
D

D
D
D
Sf .x/ D

Z
Z
1 
x 2 
1 
x sin x dx D
x.1 cos 2x/ dx D
x cos 2x dx
 0
2 0  0
0
 Z 




1

sin 2x 

x sin 2x
cos 2x dx D C
D I

2
2
4 0
2
0
0

Z
Z
2 
1 
x sin x sin nx dx D
xcos.n 1/x cos.n C 1/x dx
 0
 0
 



Z 
1
sin.n 1/x sin.n C 1/x 
sin.n 1/x sin.n C 1/x
x
dx


n 1
nC1
n 1
nC1
0
0

 





1 cos.n 1/x cos.n C 1/x
1
1
1
D
. 1/nC1 1

2
2
2
2

.n 1/
.n C 1/
 .n 1/
.n C 1/
8 0
0
if n D 2m 1;

 <
4n
nC1
16m
.
1/
1
D
if n D 2mI
:
.n2 1/2 
.4m2 1/
1
16 X
n
sin 2nx. From Definition 12.1.1,
2
 nD1 .4n
1/2


sin x
2

u.x; t/ D

p

cos 5 t sin x
2

1
p
16 X
n
cos 2n 5 t sin 2nx:
2
2
 nD1 .4n
1/

12.2.12.
1

2


2

D
D

if n  2 then
n

D
D
D
D

Sg .x/ D

Z
Z
1 
x 2  1 
x sin x dx D
x.1 cos 2x/ dx D
x cos 2x dx
 0
2 0  0
0



Z 

1

sin 2x 

x sin 2x
cos 2x dx D C
D I

2
2
4 0
2
0
0


Z
Z
1 
2 
x sin x sin nx dx D
xcos.n 1/x cos.n C 1/x dx
 0
 0
 



Z 
1
sin.n 1/x sin.n C 1/x 
sin.n 1/x sin.n C 1/x
x
dx


n 1
nC1
n 1
nC1
0
0






1 cos.n 1/x cos.n C 1/x 
1
1
1
D
. 1/nC1 1

2
2
2
2

.n 1/
.n C 1/
 .n 1/
.n C 1/
8 0
0
if n D 2m 1;

 <
4n
nC1
16m
.
1/
1
D
if n D 2mI
:
.n2 1/2 
.4m2 1/


sin x
2

1
16 X
n
sin 2nx. From Definition 12.1.1,
 nD1 .4n2 1/2

p

u.x; t/ D p sin 5 t sin x
2 5

1
p
8 X
1
p
sin 2n 5 t sin 2nx:
2
2
1/
 5 nD1 .4n

249

250 Chapter 12 Fourier Solutions of Partial Differential Equations


12.2.14. Since f .0/ D f .1/ D f 00 .0/ D f 00 .L/ D 0 and f .4/ D 360x, Theorem 11.3.5(b) and
Exercise 35(b) of Section 11.3 imply that
"
#
1 Z 1
Z 1

720
720
n D
x sin n x dx D
x cos n x
cos n x dx
n4  4 0
n5  5
0
0

720
n x 1
720
nC1 720
C 6 6 sin
D . 1/nC1 5 5 I
D . 1/
n5  5
n 
L 0
n 
Sf .x/ D

1
720 X . 1/n
sin n x. From Definition 12.1.1, u.x; t/ D
 5 nD1 n5

720 X . 1/n
cos 3n t sin n x.
 5 nD1 n5



mL
1 L
12.2.16. (a) t must be in some interval of the form mL=a; .m C 1/L=a. If
 t  mC
,
a
2 a


1 L
.m C 1/L
then (i) holds with 0    L=2a. If m C
t 
, then (ii) holds with 0    L=2a.
2 a
a
(b) Suppose that (i) holds. Since


mL
.2n 1/ a
.2n 1/ a
.2n 1/ a
cos
C
D cos
cos.2n 1/m D . 1/m cos
;
L
a
L
L
(A) implies that u.x; t/ D . 1/m u.x; /.
Suppose that (ii) holds. Since


.2n 1/ a
.m C 1/L
cos
C
L
a

1/ a
cos.2n 1/.m C 1/
L
.2n 1/ a
D . 1/mC1 cos
;
L
D cos

.2n

(B) implies that that u.x; t/ D . 1/mC1 u.x; /.


12.2.18. Since f 0 .0/ D f .2/ D 0 and f 00 .x/ D 2, Theorem 11.3.5(c) implies that
Z 2
32
.2n 1/ x
n D
cos
dx
2
2
.2n 1/  0
4

.2n 1/ x 4
128
128
nC1
D
sin
I
D . 1/
.2n 1/3  3
4
.2n
1/3  3
0
CMf .x/ D

1
128 X . 1/n
.2n 1/ x
cos
. From Exercise 12.2.17,
3
.2n 1/3
4
nD1

u.x; t/ D

128 X . 1/n
3.2n 1/ t
.2n 1/ x
cos
cos
:
3
.2n 1/3
4
4
nD1

12.2.20. Since g0 .0/ D g.2/ D 0 and g00 .x/ D 2, Theorem 11.3.5(c) implies that
Z 2
32
.2n 1/ x
n D
cos
dx
.2n 1/2  2 0
4

128
.2n 1/ x 4
128
D
sin
D . 1/nC1
I

3
3
.2n 1/ 
4
.2n 1/3  3
0

Section 12.2 The Wave Equation


1
128 X . 1/n
.2n 1/ x
cos
. From Exercise 12.2.17,
3
3

.2n 1/
4

CMf .x/ D

nD1

u.x; t/ D

1
512 X . 1/n
3.2n 1/ t
.2n 1/ x
sin
cos
:
3 4 nD1 .2n 1/4
4
4

12.2.22. Since f 0 .0/ D f .1/ D 0 and f 00 .x/ D 6.2x C 1/, Theorem 11.3.5(c) implies that
n

D
D
D
D

48L
.2n 1/2  2

1/ x
dx
2
0
"
#

Z 1
96
.2n 1/ x 1
.2n 1/ x
.2x C 1/ sin
2
sin
dx

.2n 1/3  3
2
2
0
0
"
1 #

4
.2n
1/
x
96
nC1

.
1/
3
cos

.2n 1/3  3
.2n 1/
2
0


4
96
. 1/n 3 C
I
.2n 1/3  3
.2n 1/
.2x C 1/ cos

.2n



1
96 X
1
4
.2n 1/ x
n
CMf .x/ D 3
. 1/ 3 C
cos
:
 nD1 .2n 1/3
.2n 1/
2

From Exercise 12.2.17,



1
1
4
.2n
96 X
n
u.x; t/ D 3
. 1/ 3 C
cos
 nD1 .2n 1/3
.2n 1/

p
1/ 5  t
.2n 1/ x
cos
:
2
2

12.2.24. Since g0 .0/ D g.1/ D 0 and g00 .x/ D 6.2x C 1/, Theorem 11.3.5(c) implies that
n

D
D
D
D

CMg .x/ D

48L
.2n 1/2  2

1/ x
dx
2
"0
#

Z 1
.2n 1/ x 1
.2n 1/ x
96
.2x C 1/ sin
2
sin
dx

.2n 1/3  3
2
2
0
0
"
#
96
4
.2n 1/ x 1
nC1
. 1/
3
cos

.2n 1/3  3
.2n 1/
2
0


96
4
. 1/n 3 C
I
.2n 1/3  3
.2n 1/
Z

.2x C 1/ cos

.2n



1
96 X
1
4
.2n 1/ x
n
.
1/
3
C
cos
. From Exercise 12.2.17,
3
3
 nD1 .2n 1/
.2n 1/
2



1
192 X
1
4
.2n
n
u.x; t/ D
p
.
1/
3
C
sin
.2n 1/
 4 5 nD1 .2n 1/4

p
1/ 5  t
.2n 1/ x
cos
:
2
2

251

252 Chapter 12 Fourier Solutions of Partial Differential Equations


12.2.26. Since f 0 .0/ D f .1/ D f 00 .1/ D 0 and f 000 .x/ D 24.x
cise 42(b) of Section 11.3 imply that
n

D
D
D
D

CMf .x/ D

384
.2n 1/3  3

1/ x
dx
2
0
"
#

Z 1
768
.2n 1/ x 1
.2n 1/ x
.x 1/ cos
cos
dx

.2n 1/4  4
2
2
0
0
"
#
768
2
.2n 1/ x 1
1
sin

.2n 1/4  4
.2n 1/
2
0


768
. 1/n 2
1C
I
.2n 1/4  4
.2n 1/
.x

1/ sin

.2n



1
1
. 1/n 4
.2n 1/ x
384 X
1C
cos
. From Exercise 12.2.17,
 4 nD1 .2n 1/4
.2n 1/
2

u.x; t/ D



1
384 X
3.2n 1/ t
.2n 1/ x
1
. 1/n 4
1C
cos
cos
:
4
4
 nD1 .2n 1/
.2n 1/
2
2

12.2.28. Since g0 .0/ D g.1/ D g00 .1/ D 0 and g000 .x/ D 24.x
cise 11.2.42(b) imply that
n

1/, Theorem 11.3.5(c) and Exer-

D
D
D
D

384
.2n 1/3  3

.x
0

1/ x
dx
2

Z
.2n 1/ x 1
1/ cos

2
1/ sin

.2n

"
#
1
.2n 1/ x
768
.x
cos
dx
.2n 1/4  4
2
0
0
"
#
2
.2n 1/ x 1
768
1
sin

.2n 1/4  4
.2n 1/
2
0


768
. 1/n 2
1C
I
.2n 1/4  4
.2n 1/

CMg .x/ D
From Exercise 12.2.17,
u.x; t/ D

1/, Theorem 11.3.5(c) and Exer-



1
384 X
1
. 1/n 4
.2n 1/ x
1
C
cos
:
4
4
 nD1 .2n 1/
.2n 1/
2



1
768 X
1
. 1/n 4
3.2n 1/ t
.2n 1/ x
1
C
sin
cos
:
5
5
3 nD1 .2n 1/
.2n 1/
2
2

12.2.30. Since f 0 .0/ D f .1/ D f 00 .1/ D 0 and f 000 .x/ D 24.x

1/, Theorem 11.3.5(c) and Exer-

Section 12.2 The Wave Equation

253

cise 42(b) of Section 11.3 imply that


Z 1
384
.2n 1/ x
.x 1/ sin
dx
n D
.2n 1/3  3 0
2
"
#

Z 1
768
.2n 1/ x 1
.2n 1/ x
D
.x 1/ cos
cos
dx

.2n 1/4  4
2
2
0
0
"
#
768
2
.2n 1/ x 1
D
1
sin

.2n 1/4  4
.2n 1/
2
0


768
. 1/n 2
1C
I
D
.2n 1/4  4
.2n 1/
CMf .x/ D



1
768 X
1
. 1/n 2
.2n 1/ x
1
C
cos
. From Exercise 12.2.17,
 4 nD1 .2n 1/4
.2n 1/
2

u.x; t/ D



1
768 X
1
. 1/n 2
.2n 1/ t
.2n 1/ x
1C
cos
cos
:
4
.2n 1/4
.2n 1/
2
2
nD1

12.2.32. Setting A D .2n 1/ x=2L and B D .2n 1/ at=2L in the identities
1
1
cos A cos B D cos.A C B/ C cos.A B/ and cos A sin B D sin.A C B/ sin.A B/ yields
2
2


.2n 1/ at
.2n 1/ x
1
.2n 1/.x C at/
.2n 1/.x at/
cos
cos
D
cos
C cos
.A/
2L
2L
2
2L
2L
and
sin

.2n

.2n 1/ x
1/ at
cos
2L
2L

D
D

Since CMf .x/ D


1
X

nD1

1
X

n cos

nD1

n cos

.2n

.2n


1
.2n 1/.x C at/
.2n 1/.x
sin
sin
2
2L
2L
Z
.2n 1/ xCat
.2n 1/
cos
d :
4L
2L
x at

at/

.B/

1/ x
, (A) implies that
2L

.2n 1/ x
1
1/ at
cos
D CMf .x C at/ C CMf .x
2L
2L
2

at/:

.C/

Since it can be shown that a mixed Fourier cosine series can be integrated term by term between any two
limits, (B) implies that
1
X

nD1

2Ln
.2n 1/ at
.2n 1/ x
sin
cos
.2n 1/ a
2L
2L

D
D
D

Z xCat
1
1 X
.2n 1/
n
cos
d
2a
2L
x at
nD1
!
Z xCat X
1
1
.2n 1/
n cos
d
2a x at
2L
nD1
Z xCat
1
CMg ./ d :
2a x at

254 Chapter 12 Fourier Solutions of Partial Differential Equations


This and (C) imply that
u.x; t/ D

1
CMf .x C at/ C CMf .x
2

at/ C

1
2a

xCat

CMg ./ d :
x at

12.2.34. We begin by looking for functions of the form v.x; t/ D X.x/T .t/ that are not identically zero
and satisfy vt t D a2 vxx , v.0; t/ D 0, vx .L; t/ D 0 for all .x; t/. As shown in the text, X and T must
satisfy X 00 C X D 0 and (B) T 00 C a2 T D 0 for the same value of . Since v.0; t/ D X.0/T .t/
and vx .L; t/ D X 0 .L/T .t/ and we dont want T to be identically zero, X.0/ D 0 and X 0 .L/ D 0.
Therefore, must be an eigenvalue of (C) X 00 C X D 0, X.0/ D 0, X 0 .L/ D 0, and X must be
.2n 1/2
, integer), with
a -eigenfunction. From Theorem 11.1.4, the eigenvalues of (C) are n D
 2=4L2
.2n 1/ x
.2n 1/2  2
, n D 1, 2, 3,. . . . Substituting  D
into
associated eigenfunctions Xn D sin
2L
4L2
(B) yields T 00 C ..2n 1/2  2a2 =4L2/T D 0, which has the general solution
Tn D n cos

.2n

1/ at
2n L
.2n 1/ at
C
sin
;
2L
.2n 1/ a
2L

where n and n are constants. Now let



.2n
vn .x; t/ D Xn .x/Tn .t/ D n cos

1/ at
2n L
.2n 1/ at
C
sin
2L
.2n 1/ a
2L

sin

.2n

1/ x
:
2L

Then
@vn
.x; t/ D
@t

.2n

1/ a
.2n 1/ at
.2n 1/ at
n sin
C n cos
2L
2L
2L

so

sin

.2n

1/ x
;
2L

@vn
.2n 1/ x
.x; 0/ D n sin
:
@t
2L
.2n 1/ x
.2n 1/ x
Therefore,vn satisfies (A) with f .x/ D n sin
and g.x/ D n sin
. More gen2L
2L
erally, if 1; 2 ; : : : ; m and 1 ; 2 ; : : : ; m are constants and
vn .x; 0/ D n sin

um .x; t/ D

.2n

m 
X
.2n
n cos

nD1

1/ x
2L

and

1/ at
2n L
.2n 1/ at
C
sin
2L
.2n 1/ a
2L

sin

.2n

1/ x
;
2L

then um satisfies (A) with


f .x/ D

m
X

n sin

nD1

This motivates the definition.

.2n

1/ x
2L

and

g.x/ D

m
X

nD1

n sin

.2n

1/ x
:
2L

Section 12.2 The Wave Equation

255

12.2.36. Since f .0/ D f 0 .1/ D 0, and f 00 .x/ D 6.1 2x/, Theorem 11.3.5(d) implies that
Z 1
48
.2n 1/ x
n D
.1 2x/ sin
dx
.2n 1/2  2 0
2
"
#

Z 1
96
.2n 1/ x 1
.2n 1/ x
.1 2x/ cos
cos
dx
D
C2
.2n 1/3  3
2
2
0
0
"
#
96
4
.2n 1/ x 1
D
1C
sin

.2n 1/3  3
.2n 1/
2
0


96
4
1 C . 1/n
I
D
.2n 1/3  3
.2n 1/


1
96 X
1
4
.2n 1/ x
n
SMf .x/ D
1
C
.
1/
sin
. From Exercise 12.2.34,
3
3
 nD1 .2n 1/
.2n 1/
2


1
1
4
3.2n 1/ t
.2n 1/ x
96 X
n
1 C . 1/
cos
sin
:
3
.2n 1/3
.2n 1/
2
2

u.x; t/ D

nD1

12.2.38. Since g.0/ D g0 .1/ D 0, and g00 .x/ D 6.1 2x/, Theorem 11.3.5(d) implies that
Z 1
48
.2n 1/ x
.1 2x/ sin
dx
n D
.2n 1/2  2 0
2
"
#

Z 1
96
.2n 1/ x 1
.2n 1/ x
D
.1 2x/ cos
cos
dx
C2
.2n 1/3  3
2
2
0
0
"
#
96
4
.2n 1/ x 1
D
1C
sin

.2n 1/3  3
.2n 1/
2
0


4
96
D
1 C . 1/n
I
.2n 1/3  3
.2n 1/


1
96 X
.2n 1/ x
1
4
n
SMg .x/ D
1 C . 1/
sin
. From Exercise 12.2.34,
3
3
 nD1 .2n 1/
.2n 1/
2


1
64 X
1
4
3.2n 1/ t
.2n 1/ x
n
1
C
.
1/
sin
sin
:
 4 nD1 .2n 1/4
.2n 1/
2
2

u.x; t/ D

12.2.40. Since f .0/ D f 0 . / D f 00 .0/ D 0 and f 000 .x/ D 6, Theorem 11.3.5(d) and Exercise 11.3.50(b)
imply that

Z 
96
.2n 1/x
192
.2n 1/x 
192
n D
cos
dx D
sin
D . 1/n
I

3
4
.2n 1/  0
2
.2n 1/ 
2
.2n 1/4 
0
1

192 X . 1/n
.2n 1/x
SMf .x/ D
sin
. From Exercise 12.2.34,
 nD1 .2n 1/4
2
1
192 X . 1/n
.2n
u.x; t/ D
cos
4

.2n 1/
nD1

p
1/ 3 t
.2n 1/x
sin
:
2
2

256 Chapter 12 Fourier Solutions of Partial Differential Equations


12.2.42. Since g.0/ D g0 . / D g00 .0/ D 0 and g000 .x/ D 6, Theorem 11.3.5(d) and Exercise 50(b) imply
that

Z 
96
.2n 1/x
192
.2n 1/x 
192
n
n D
cos
dx D
sin
D . 1/ .2n 1/4  I
.2n 1/3  0
2
.2n 1/4 
2
0
SMg .x/ D

.2n 1/x
192 X . 1/n
sin
. From Exercise 12.2.34,
4
 nD1 .2n 1/
2
p
1
384 X . 1/n
.2n 1/ 3 t
.2n 1/x
u.x; t/ D p
sin
sin
:
5
.2n 1/
2
2
3
nD1

12.2.44. Since f .0/ D f 0 .1/ D f 00 .0/ D 0 and f 000 .x/ D 12.2x 1/, Theorem 11.3.5(d) and Exercise 11.3.50(b) imply that
Z 1
192
.2n 1/ x
.2x 1/ cos
dx
n D
.2n 1/3  3 0
2
"
#

Z 1
384
.2n 1/ x 1
.2n 1/ x
D
.2x 1/ sin
2
sin
dx

.2n 1/4  4
2
2
0
0
"
1 #

4
.2n
1/
x
384
nC1

D
.
1/
C
cos

.2n 1/4  4
.2n 1/
2
0




4
384
4
384
nC1
n
D
.
1/
D
.
1/
C
I
.2n 1/4  4
.2n 1/
.2n 1/4  4
.2n 1/


1
384 X
1
4
.2n 1/ x
n
SMf .x/ D 4
.
1/
C
sin
. From Exercise 12.2.34,

.2n 1/4
.2n 1/
2
nD1



1
384 X
1
4
n
u.x; t/ D 4
. 1/ C
cos.2n
 nD1 .2n 1/4
.2n 1/

1/ t sin

.2n

1/ x
:
2

12.2.46. Since g.0/ D g0 .1/ D g00 .0/ D 0 and g000 .x/ D 12.2x 1/, Theorem 11.3.5(d) and Exercise 11.3.50(b) imply that
Z 1
.2n 1/ x
192
n D
.2x 1/ cos
dx
3
3
.2n 1/  0
2
"
#

Z 1
384
.2n 1/ x 1
.2n 1/ x
D
.2x 1/ sin
2
sin
dx

.2n 1/4  4
2
2
0
0
"
#
384
4
.2n 1/ x 1
nC1
D
. 1/
C
cos

.2n 1/4  4
.2n 1/
2
0




384
4
384
4
nC1
n
D
. 1/
D
. 1/ C
I
.2n 1/4  4
.2n 1/
.2n 1/4  4
.2n 1/


1
384 X
1
4
.2n 1/ x
n
SMg .x/ D 4
. 1/ C
sin
. From Exercise 12.2.34,
4
 nD1 .2n 1/
.2n 1/
2
u.x; t/ D



1
384 X
1
4
n
.
1/
C
sin.2n
5
.2n 1/5
.2n 1/
nD1

1/ t sin

.2n

1/ x
:
2

Section 12.2 The Wave Equation

257

12.2.48. Since f is continuous on 0; L and f 0 .L/ D 0, Theorem 11.3.4 implies that SMf .x/ D f .x/,
0  x  L. From Exercise 11.3.58, SMf is the odd periodic extension (with period 2L) of the function
f .x/;
0  x  L;
r .x/ D
which is continuous on 0; 2L. Since r .0/ D r .2L/ D f .0/ D
f .2L x/; L < x  2L;

f 0 .x/;
0 < x < L;
0, SMf is continuous on . 1; 1/. Moreover, r 0 .x/ D
r 0 .0/ D
0
f .2L x/; L < x < 2L; C
fC0 .0/, r 0 .2L/ D fC0 .0/, and, since f 0 .L/ D 0, r 0 .L/ D 0. Hence, r is differentiable on 0; 2L.
Since r .0/ D r .2L/ D f .0/ D 0, Theorem 12.2.3(a) with h D r , p D SMf , and L replaced by 2L
implies that SMf is differentiable
 on 00. 1; 1/. Similarly, SMg is differentiable on . 1; 1/.
f .x/;
0 < x < L;
00
00
Now we note that r .x/ D
r 00 .L/ D f 00 .L/, and rC
.0/ D r 00 .2L/ D
f 00 .2L x/; L < x < 2L;
0
0
fC00 .0/ D 0. Since SMf
is the even periodic extension of r 0 , Theorem 12.2.3(b) with h D r 0 , q D SMf
,
0
and L replaced by 2L implies that SMf is differentiable on . 1; 1/. Now follow the argument used to
complete the proof of Theorem 12.2.4.
12.2.50. From Example 11.3.5, Cf .x/ D 4
u.x; t/ D 4

1
.2n 1/ x
768 X
cos
. From Exercise 12.2.49,
4
4

.2n 1/
2
nD1

768 X
1
cos
 4 nD1 .2n 1/4

p
5.2n 1/ t
.2n 1/ x
cos
:
2
2

 

3x 5
2 4
 x 4 D
. Since f 0 .0/ D f 0 . / D 0 and
5
5
0
0
f 000 .x/ D 24.3x  /, Theorem 11.3.5(a) implies that



Z 
Z 

48
48

n D
.3x

/
sin
nx
dx
D
.3x

/
cos
nx
3
cos
nx
dx

n3  0
n4 
0
0


48
144
48
D
. 1/n 2 C  C 5 sin nx D
1 C . 1/n 2 ; n  1I
4
n4 
n 
n
0
12.2.52. 0 D

Cf .x/ D

.3x 4

2 4
5

48

4Lx 3 / dx D

1
X
1 C . 1/n 2
cos nx. From Exercise 12.2.49,
n4

nD1

u.x; t/ D

2 4
5

48

1
X
1 C . 1/n 2
cos 2nt cos nx:
n4
nD1


 

1 3x 5
2 4
4
x
D
. Since g0 .0/ D g0 . / D 0 and
12.2.54. 0 D
.3x
4Lx / dx D


5
5
0
0
g000 .x/ D 24.3x  /, Theorem 11.3.5(a) implies that



Z 
Z 

48
48

n D
.3x  / sin nx dx D
.3x  / cos nx
3
cos nx dx
n3  0
n4 
0
0


48
144
48
D
. 1/n 2 C  C 5 sin nx D
1 C . 1/n 2 ; n  1I
4
n 
n 
n4
0
Z

258 Chapter 12 Fourier Solutions of Partial Differential Equations

Cg .x/ D

2 4
5

48

1
X
1 C . 1/n 2
cos nx. From Exercise 12.2.49,
n4

nD1

u.x; t/ D

2 4t
5

24

1
X
1 C . 1/n 2
sin 2nt cos nx:
n5

nD1


4

D  . Since f 0 .0/ D

30
0



Z
1  4
1 x5  x4
 2x3
.x
2 x 3 C  2x 2 / dx D
C
 0
 5
2
3
f 0 . / D 0 and f 000 .x/ D 12.2x  /, Theorem 11.3.5(a) implies that


Z 

24
24

n D
.2x

/
sin
nx
dx
D
.2x

/
cos
nx

3
4
n  0
n 
0


24
48
24
D
. 1/n  C  C 5 sin nx D
1 C . 1/n
n4 
n 
n4
0
(
0
if n D 2m 1;
3
D
n  1I
if n D 2m;
m4
12.2.56. 0 D

Cf .x/ D

4
30

1
X
4
1
cos
2nx.
From
Exercise
12.2.49,
u.x;
t/
D
n4
30
nD1

4
30

1
X
1
 4t
cos
2nx.
From
Exercise
12.2.49,
u.x;
t/
D
n4
30
nD1

cos nx dx

1
X
1
cos 8nt cos 2nx.
n4
nD1



Z
1 x5  x4
 2x 3
1  4
12.2.58. 0 D
.x
2 x 3 C  2 x 2/ dx D
C
 0
 5
2
3
g0 . / D 0 and g000 .x/ D 12.2x  /, Theorem 11.3.5(a) implies that


Z 

24
24

n D
.2x

/
sin
nx
dx
D
.2x

/
cos
nx

3
4
n  0
n 
0


24
48
24
n
n
D
. 1/  C  C 5 sin nx D
1 C . 1/
n4 
n 
n4
0
(
0
if n D 2m 1;
3
 1I
D
if n D 2m;
m4
Cg .x/ D


4

D  . Since g0 .0/ D

30
0
2

cos nx dx
0

1
3X 1
sin 8nt cos 2nx.
8 nD1 n5

1
12.2.60. Setting A D n x=L and B D n at=L in the identities cos A cos B D cos.AC B/C cos.A
2
1
B/ and cos A sin B D sin.A C B/ sin.A B/ yields
2


n at
n x
1
n.x C at/
n.x at/
cos
D
cos
C cos
.A/
cos
L
L
2
L
L
and

n at
n x
sin
cos
L
L

D
D



1
n.x C at/
n.x at/
sin
sin
2 Z
L
L
xCat
n
n
cos
d :
2L x at
L

.B/

Section 12.2 The Wave Equation

Since Cf .x/ D 0 C

1
X

n sin

nD1

0 C

1
X

n x
, (A) implies that
L

n cos

nD1

259

n at
n x
1
cos
D Cf .x C at/ C Cf .x
L
L
2

at/:

.C/

Since it can be shown that a Fourier sine series can be integrated term by term between any two limits,
(B) implies that
0 t C

1
X
n L
n at
n x
sin
cos
n a
L
L
nD1

Z xCat
1
1 X
n
n
cos
d
2a nD1
L
x at
!
Z xCat
1
X
1
n
0 C
n cos
d
2a x at
L
nD1
Z xCat
1
Cg ./ d :
2a x at

D 0 t C
D
D

This and (C) imply that


1
u.x; t/ D Cf .x C at/ C Cf .x
2

at/ C

1
2a

xCat

Cg ./ d :
x at

12.2.62.(a). Since jpn .x/j  1 and jqn .t/j  1 for all t, jkn pn .x/qn .t/j  jkn j for all .x; t/, and the
comparison test implies the conclusion.
(b) If t is fixed but arbitrary, then jkn pn0 .x/qn .t/j  jjnjkn j, so Theorem 12.1.2 with D x and
wn .x/ D kn pn .x/qn .t/ justifies term by term differentiation with respect to x on . 1; 1/. If x is fixed
but arbitrary, then jkn pn .x/qn0 .t/j  jjnjkn j, so Theorem 12.1.2 with D t and wn .t/ D kn pn .x/qn .t/
justifies term by term differentiation with respect to t on . 1; 1/.
(c) The argument is similar to argument use in (b).
1
1
X
X
n x
n at
n x
n at
n L
(d) Apply (b) and (c) to the series
n cos
sin
and
sin
sin
, recallL
L
n a
L
L
nD1
nD1
ing that the individual terms in the series satisfy ut t D a2 uxx for all .x; t/.
Z xCat
f .x C ct/ C f .x ct/
1
(d) u.x; t/ D
C
g.u/ du.
2
2a x at

12.2.64.
u.x; t/

.x C at/ C .x
2
2

D x C .x C at/

at/
.x

1
2a
2

xCat
x at

4au du D x C 2

at/ D x.1 C 4at/:

xCat
x at

xCat

u du D x C u2
x at

12.2.66.
u.x; t/

D
D

sin.x C at/ C sin.x


2
sin.x C at/ C sin.x
2

Z xCat
1
C
a cos u du
2a x at
at/
sin.x C at/ sin.x at/
C
D sin.x C at/:
2
at/

260 Chapter 12 Fourier Solutions of Partial Differential Equations


12.2.68.
u.x; t/

D x sin x cos at C at cos x sin at C

Z xCat
.x C at/ sin.x C at/ C .x at/ sin.x at/
1
C
sin u du
D
2
2a x at
xsin.x C at/ C sin.x at/
atsin.x C at/ sin.x at/
D
C
2
2
cos.x at/ cos.x C at/
C
2a

sin x sin at
:
a

12.3 LAPLACES EQUATION IN RECTANGULAR COORDINATES


12.3.2. Since f .0/ D f .1/ D 0 and f 00 .x/ D 2 6x, Theorem 11.3.5(b) implies that
"
#

Z 1
Z 2
n x
8
n x 2
n x
4
.4 6x/ sin
dx D 3 3 .4 6x/ cos
C6
cos
dx
n D
n2  2 0
2
n 
2 0
2
0

32
96
n x 2
32
n
D
.1 C . 1/ 2/ C 4 4 sin
D
1 C . 1/n 2 I

3
3
3
n 
n 
2 0
n 3

S.x/ D

1
32 X 1 C . 1/n 2
n x
sin
. From Example 12.3.1,
3
n3
2
nD1

u.x; y/ D

1
32 X 1 C . 1/n 2 sinh n.3
 3 nD1
n3 sinh 3n=2

y/=2

sin

n x
:
2

12.3.4.
1

2


2

D
D

if n  2, then
n

D
D
D
D

S.x/ D

Z
Z
1 
x 2 
1 
x sin x dx D
x.1 cos 2x/ dx D
x cos 2x dx
 0
2 0  0
0
 Z 




1

sin 2x

x sin 2x
cos 2x dx D C
D I

2
2
4 0
2
0
0

Z
Z
2 
1 
x sin x sin nx dx D
xcos.n 1/x cos.n C 1/x dx
 0
 0
 



Z 
1
sin.n 1/x sin.n C 1/x 
sin.n 1/x sin.n C 1/x
x
dx


n 1
nC1
n 1
nC1
0
0

 





1 cos.n 1/x cos.n C 1/x
1
1
1
nC1
D
.
1/
1


.n 1/2
.n C 1/2
 .n 1/2 .n C 1/2
8 0
0
if n D 2m 1;

 <
4n
nC1
16m
.
1/
1
D
if n D 2mI
:
.n2 1/2 
.4m2 1/


sin x
2

1
16 X
n
sin 2nx. From Example 12.3.1,
2
 nD1 .4n
1/2

u.x; y/ D

 sinh.1 y/
sin x
2
sinh 1

1
16 X n sinh 2n.1 y/
sin 2nx:

.4n2 1/2 sinh 2n
nD1

Section 12.3 Laplaces Equation in Rectangular Coordinates

261

x/2 1
1
12.3.6. 0 D
.1 x/ dx D
D ; if n  1,

2
2
0
0
"
#
1 Z 1
Z 1

2
.1 x/ sin n x C
sin n x dx
n D 2
.1 x/ cos n x dx D
n
0
0
0
8
1
4
<

2
if n D 2m 1;
D 2 1 . 1/n D
cos
n
x
D
.2m 1/2  2

2 2
:
n2  2
n
0
0
if n D 2mI
Z

C.x/ D

.1

1
1
4 X
1
C 2
cos.2n
2

.2n 1/2

1/ x. From Example 12.3.3,

nD1

u.x; y/ D

1
y
4 X
sinh.2n 1/y
C 3
cos.2n
2

.2n 1/3 cosh 2.2n 1/

1/ x:

nD1

1/3 1
1
12.3.8. 0 D
.x 1/ dx D
D ; if n  1, then

3
3
0
0
"
#
1
Z 1
Z 1

2
2
2
n D 2
.x 1/ cos n x dx D
.x 1/ sin n x
2
.x 1/ sin n x dx
n
0
0
0
"
#
1 Z 1
1

4
4
4
4

D
.x 1/ cos n x
cos n x dx D 2 2
sin n x D 2 2 I
2
2
3
3
n 
n 
n 
n 
0
0
0
Z

C.x/ D

.x

1
1
4 X 1
y
4 X sinh ny
1
C 2
cos
n
x.
From
Example
12.3.3,
u.x;
y/
D
C
cos n x.
3
 nD1 n2
3
 3 nD1 n3 cosh n

12.3.10. Since g.0/ D g0 .1/ D 0, and g00 .y/ D 6.1 2y/, Theorem 11.3.5(d) implies that
Z 1
48
.2n 1/y
n D
.1 2y/ sin
dy
2
2
.2n 1/  0
2
"
#

Z 1
96
.2n 1/y 1
.2n 1/y
D
.1 2y/ cos
cos
dy
C2
.2n 1/3  3
2
2
0
0
"
#
4
.2n 1/y 1
96
D
1C
sin

.2n 1/3  3
.2n 1/
2
0


96
4
D
1 C . 1/n
I
.2n 1/3  3
.2n 1/
SM .y/ D



1
96 X
1
4
.2n 1/y
n
1
C
.
1/
sin
. From Example 12.3.5,
3
3
 nD1 .2n 1/
.2n 1/
2

u.x; y/ D


1 
96 X
4
cosh.2n 1/.x 2/=2
.2n 1/y
n
1
C
.
1/
sin
:
3
3

.2n 1/ .2n 1/ cosh 2.2n 1/=2
2
nD1

262 Chapter 12 Fourier Solutions of Partial Differential Equations


12.3.12. From Example 11.3.8.3,
SM .y/ D



1
96 X
1
4
.2n 1/y
n
3
C
.
1/
sin
:
3
3

.2n 1/
.2n 1/
2
nD1

From Example 12.3.5,



1 
96 X
4
cosh.2n 1/.x 3/=2
.2n 1/y
n
3 C . 1/
sin
:
u.x; y/ D 3
3
 nD1
.2n 1/ .2n 1/ cosh 3.2n 1/=2
2
12.3.14.
cn

D
D

2
3

.3y
0

4
.2n

1/

y 2 / cos
"
.3y

"
24
.3
.2n 1/2  2

.2n

1/y
6

y 2 / sin

.2n

dy

1/y 3

6
0

.3

2y/ sin

.2n

1/y
6

dy

Z 3
1/y 3
.2n 1/y
D
2y/ cos
cos
dy
C2
6
6
0
0

72
288
.2n 1/y 3
288
.2n 1/y
D
C
sin
D .2n 1/3  3 sin
.2n 1/2  2
.2n 1/3  3
6
2
0
288
72
D
C . 1/n 1
I
.2n 1/2  2
.2n 1/3  3


1
72 X
.2n 1/y
1
4. 1/n
CM .y/ D
1C
cos
. From Example 12.3.7,
2
2
 nD1 .2n 1/
.2n 1/
6
u.x; y/ D

.2n


1 
4. 1/n
cosh.2n 1/ x=6
.2n 1/y
432 X
1
C
cos
:
3
.2n 1/ .2n 1/3 sinh.2n 1/=3
6
nD1

12.3.16. Since g0 .0/ D g.1/ D 0 and g00 .y/ D 6y, Theorem 11.3.5(c) implies that
Z 1
48
.2n 1/y
n D
y cos
dy
2
2
.2n 1/  0
2
"
#

Z 1
96
.2n 1/y 1
.2n 1/y
D
y sin
sin
dy

.2n 1/3  3
2
2
0
0
"
#
96
2
.2n 1/y 1
nC1
D
. 1/
C
cos

.2n 1/3  3
.2n 1/
2
0


96
2
D
. 1/n C
dyI
.2n 1/3  3
.2n 1/


1
96 X
1
2
.2n 1/y
n
CM .y/ D
.
1/
C
cos
. From Example 12.3.7,
 3 nD1 .2n 1/3
.2n 1/
2
u.x; y/ D



1
192 X
cosh.2n 1/ x=2
2
.2n 1/y
n
.
1/
C
cos
:
4
4

.2n 1/ sinh.2n 1/=2
.2n 1/
2
nD1

Section 12.3 Laplaces Equation in Rectangular Coordinates

263

12.3.18. The boundary conditions require products v.x; y/ D X.x/Y .y/ such that (A) X 00 C X D 0,
X 0 .0/ D 0, X 0 .a/ D 0, and (B) Y 00 Y D 0, Y .0/ D 1, Y .b/ D 0. From Theorem 11.1.3, the eigenvaln2  2
ues of (A) are  D 0, with associated eigenfunction X0 D 1, and n D
, with associated eigenfunca2
n x
, n D 1, 2, 3,. . . . Substituting  D 0 into (B) yields Y000 D 0, Y0 .0/ D 1, Y0 .b/ D 0,
tions Yn D cos
a
y
n2  2
. Substituting  D
into (B) yields Yn00 .n2  2=a2 /Yn D 0, Yn .0/ D 1, Yn .b/ D
so Y0 .y/ D 1
b
a2
sinh n.b y/=a
sinh n.b y/=a
n x
0, so Yn D
. Then vn .x; y/ D Xn .x/Yn .y/ D
cos
, so
sinh n b=a
sinh n b=a
a
n x
n x
vn .x; 0/ D cos
. Therefore,vn is solution of the given problem with f .x/ D cos
. More genera
a
m


X sinh n.b y/=a
y
n x
ally, if 0 ; : : : ; m are arbitrary constants, then um .x; y/ D 0 1
C
n
cos
b
sinh n b=a
a
nD1
m
X
n x
. Therefore, if f is an arbiis a solution of the given problem with f .x/ D 0 C
n cos
a
nD1
trary piecewise smooth function on 0; a we define the formal solution of the given problem to be
1
1

X
y  X sinh n.b y/=a
n x
n x
u.x; y/ D 0 1
C
n
cos
, where C.x/ D 0 C
n cos
is
b
sinh n b=a
a
a
nD1
nD1
Z
Z
2 a
n x
1 a
the Fourier cosine series of f on 0; a; that is, 0 D
f .x/ dx and n D
f .x/ cos
dx,
a 0
a 0
a
n  1.


Z
1 x5
4x 3 2
6
1 2 4
3
2
4
Now consider the special case. 0 D
.x
4x C 4x / dx D
x C
D .

2 0
2 5
3
5
0
Since f 0 .0/ D f 0 .2/ D 0 and f 000 .x/ D 12.2x 2/, Theorem 11.3.5(a) implies that
"
#

Z 2
Z 2
96
n x
192
n x 2
n x
n D
.2x 2/ sin
dx D
.2x 2/ cos
2
cos
dx
n3  3 0
2
n4  4
2 0
2
0

768
n x 2
384
192
n
D
.
1/
2
C
2
C
sin
D
1 C . 1/n
n4  4
n5  5
2 0
n4  4
(
0
if n D 2m 1;
48
D
n  1:
if n D 2m;
m4  4
C.x/ D

8
15

1
48 X 1
8.1 y/
cos n x; u.x; y/ D
4
4

n
15
nD1

1
48 X 1 sinh n.1 y/
cos n x.
4
n4
sinh n
nD1

12.3.20. The boundary conditions require products v.x; y/ D X.x/Y .y/ such that (A) X 00 C X D 0,
X.0/ D 0, X 0 .a/ D 0, and (B) Y 00 Y D 0, Y .0/ D 1, Y .b/ D 0. From Theorem 11.1.4, the
.2n 1/2  2
.2n 1/ x
, with associated eigenfunctions Yn D sin
,
eigenvalues of (A) are n D
2
4a
2a
2 2
.2n 1/ 
n D 1, 2, 3,. . . . Substituting  D
into (B) yields Yn00 ..2n 1/2  2=4a2 /Yn D 0,
4a2
sinh.2n 1/.b y/=2a
Yn .0/ D 1, Yn .b/ D 0, so Yn D
. Then vn .x; y/ D Xn .x/Yn .y/ D
sinh.2n 1/ b=2a
sinh.2n 1/.b y/=2a
.2n 1/ x
.2n 1/ x
sin
, so vn .x; 0/ D sin
. Therefore,vn is solution of
sinh.2n 1/ b=2a
2a
2a

264 Chapter 12 Fourier Solutions of Partial Differential Equations


1/ x
. More generally, if 1 ; : : : ; m are arbitrary constants,
2a
m
X sinh.2n 1/.b y/=2a
.2n 1/ x
cos
is a solution of the given problem
then um .x; y/ D
n
sinh.2n
1/
b=2a
2a
nD1
m
X
.2n 1/ x
with f .x/ D
n sin
. Therefore, if f is an arbitrary piecewise smooth function on 0; a
2a
nD1
1
X
sinh.2n 1/.b y/=2a
.2n 1/ x
sin
,
we define the formal solution of the given problem to be u.x; y/ D
n
sinh.2n 1/ b=2a
2a
nD1
1
X
.2n 1/ x
is the mixed Fourier sine series of f on 0; a; that is, n D
where Sm .x/ D
n sin
2a
nD1
Z
2 a
.2n 1/ x
f .x/ sin
.
a 0
2a
Now consider the special case. Since f .0/ D f 0 .L/ D 0 and f 00 .x/ D 2, Theorem 11.3.5(d) implies
that

Z 3
.2n 1/ x
288
.2n 1/ x 3
288
48
sin
dx
D
cos
D
I
n D

2
2
3
3
.2n 1/  0
6
.2n 1/ 
6
.2n 1/3  3
0
the given problem with f .x/ D sin

.2n

1
288 X
.2n 1/ x
1
SM .x/ D 3
sin
I
 nD1 .2n 1/3
6

u.x; y/ D

1
288 X sinh.2n 1/.2
 3 nD1 .2n 1/3 sinh.2n

y/=6
.2n 1/ x
sin
:
1/=3
6

12.3.22. The boundary conditions require products v.x; y/ D X.x/Y .y/ such that (A) X 00 C X D 0,
X 0 .0/ D 0, X 0 .a/ D 0, and (B) Y 00 Y D 0, Y 0 .0/ D 0, Y .b/ D 1. From Theorem 11.1.3, the
n2  2
eigenvalues of (A) are  D 0, with associated eigenfunction X0 D 1, and n D
, with associated
a2
n x
eigenfunctions Yn D cos
, n D 1, 2, 3,. . . . Substituting  D 0 into (B) yields Y000 D 0, Y00 .0/ D 0,
a
n2  2
Y0 .b/ D 1, so Y0 D 1. Substituting  D
into (B) yields Yn00 .n2  2 =a2 /Yn D 0, Yn0 .0/ D
a2
cosh ny=a
sinh ny=a
n x
0, Yn .b/ D 1, so Yn D
. Then vn .x; y/ D Xn .x/Yn .y/ D
cos
, so
cosh n b=a
cosh n b=a
a
n x
n x
vn .x; b/ D cos
. Therefore,vn is solution of the given problem with f .x/ D cos
. More
a
a
m
X cosh ny=a
n x
generally, if 0 ; : : : ; m are arbitrary constants, then um .x; y/ D 0 C
n
cos
is
cosh
n
b=a
a
nD1
m
X
n x
a solution of the given problem with f .x/ D 0 C
n cos
. Therefore, if f is an arbitrary
a
nD1
piecewise smooth function on 0; a we define the formal solution of the given problem to be u.x; y/ D
1
1
X
X
cosh ny=a
n x
n x
0 C
n
cos
, where C.x/ D 0 C
n cos
is the Fourier cosine series of
cosh
n
b=a
a
a
nD1
Z a
Z a nD1
1
2
n x
f on 0; a; that is, 0 D
f .x/ dx and n D
f .x/ cos
dx, n  1.
a 0
a 0
a

Section 12.3 Laplaces Equation in Rectangular Coordinates


Now consider the special case.

Z
1  4
1 x5
0 D
.x
2 x 3 C  2x 2 / dx D
 0
 5

 x4
 2x 3
C
2
3

265

 
4

D  :

30
0

Since f 0 .0/ D f 0 . / D 0 and f 000 .x/ D 12.2x  /, Theorem 11.3.5(a) implies that



Z 
Z 

24
24

.2x  / sin nx dx D
.2x  / cos nx
2
cos nx dx
n D
n3  0
n4 
0
0


24
48
24
D
. 1/n  C  C 5 sin nx D
1 C . 1/n
4
4
n 
n 
n
0
(
0
if n D 2m 1;
3
n  1I
D
if n D 2m;
4
m
C.x/ D

4
30

1
X
1
4
cos 2nx; u.x; y/ D
4
n
30

nD1

1
X
1 cosh 2ny
cos 2nx
n4 cos 2n

nD1

12.3.24. The boundary conditions require products v.x; y/ D X.x/Y .y/ such that (A) X 00 X D
0, X 0 .0/ D 0, X.a/ D 1, and (B) Y 00 C Y D 0, Y .0/ D 0, Y .b/ D 0. From Theorem 11.1.2,
n2  2
ny
the eigenvalues of (B) are n D
, with associated eigenfunctions Yn D sin
, n D 1, 2,
b2
b
2 2
n 
3,. . . . Substituting  D
into (A) yields Xn00 .n2  2=b 2 /Xn D 0, Xn0 .0/ D 0, Xn .a/ D 1, so
b2
cosh n x=b
cosh n x=b
ny
ny
Xn D
. Then vn .x; y/ D Xn .x/Yn .y/ D
sin
, so vn .a; y/ D sin
.
cosh n a=b
cosh n a=b
b
b
ny
Therefore,vn is solution of the given problem with g.y/ D sin
. More generally, if 1 ; : : : ; m
b
m
X
ny
cosh n x=b
are arbitrary constants, then um .x; y/ D
n
sin
is a solution of the given problem
cosh
n
a=b
b
nD1
m
X
ny
with g.y/ D
n sin
. Therefore, if g is an arbitrary piecewise smooth function on 0; b we
b
nD1
1
X
cosh n x=b
ny
define the formal solution of the given problem to be u.x; y/ D
n
sin
, where
cosh n a=b
b
nD1
Z
1
X
2 b
ny
ny
S.y/ D
n sin
is the Fourier sine series of g on 0; b; that is, n D
g.y/ sin
dy.
b
b 0
b
nD1

Now consider the special case. Since g.0/ D g.1/ D g00 .0/ D g00 .L/ D 0 and f .4/ .y/ D 24,
Theorem 11.3.5(b) and Exercise 35(b) of Section 11.3 imply that
1
Z 1

48
48
n D
sin ny dy D
cos ny
4
4
5
5
n  0
n 
0
8
96
<
48
if n D 2m 1
D
. 1/n 1 D
.2m 1/5  5
5
5
:
n 
0
if n D 2mI

S.y/ D

1
96 X
1
sin.2n
5

.2n 1/5
nD1

1/y; u.x; y/ D

1
96 X
cosh.2n 1/ x
sin.2n
5

.2n 1/5 cosh.2n 1/
nD1

1/y.

266 Chapter 12 Fourier Solutions of Partial Differential Equations


12.3.26. The boundary conditions require products v.x; y/ D X.x/Y .y/ such that (A) X 00 X D 0,
X 0 .0/ D 0, X 0 .a/ D 1, and (B) Y 00 C Y D 0, Y .0/ D 0, Y .b/ D 0. From Theorem 11.1.2, the
n2  2
ny
eigenvalues of (B) are n D
, with associated eigenfunctions Yn D sin
, n D 1, 2, 3,. . . .
2
b
b
2 2
n 
into (A) yields Xn00 .n2  2 =b 2/Xn D 0, Xn0 .0/ D 0, Xn0 .a/ D 1, so Xn D
Substituting  D
b2
b cosh n x=b
b cosh n x=b
ny
@vn
ny
. Then vn .x; y/ D Xn .x/Yn .y/ D
sin
, so
.a; y/ D sin
.
n sinh n a=b
n sinh n a=b
b
@x
b
ny
Therefore,vn is solution of the given problem with g.y/ D sin
. More generally, if 1; : : : ; m are
b
m
b X
cosh n x=b
ny
arbitrary constants, then um .x; y/ D
n
sin
is a solution of the given problem
 nD1 n sinh n a=b
b
m
X
ny
with g.y/ D
n sin
. Therefore, if g is an arbitrary piecewise smooth function on 0; b we
b
nD1
1
cosh n x=b
ny
b X
n
sin
, where
define the formal solution of the given problem to be u.x; y/ D
 nD1 n sinh n a=b
b
Z
1
X
ny
2 b
ny
S.y/ D
n sin
is the Fourier sine series of g on 0; b; that is, n D
g.y/ sin
dy.
b
b 0
b
nD1
Z 2

Z 4
ny
ny
1
y sin
C
.4 y/ sin
dy ;
Now consider the special case. n D
2 0
4
4
2
"
#

Z 2
Z 2
ny
4
ny 2
ny
y sin
dy D
y cos
cos
dy
4
n
4 0
4
0
0

n
4
ny 2
2
n
4
n
2
cos
C 2 2 sin
D
cos
C 2 2 sin
I
D
n
2
n 
4 0
n
2
n 
2
Z

.4
0

Z 2
ny 4
ny
D
y/ cos
C
cos
dy
4 2
4
0

2
n
4
ny 4
2
n
4
n
D
cos
sin
D
cos
C 2 2 sin
I
n
2
n2  2
4 2
n
2
n 
2
8
16
<
n
16
. 1/mC1
if n D 2m 1
n D 2 2 sin
D
.2m 1/2  2
:
n 
2
0
if n D 2mI

ny
y/ sin
dy
4

"
2
.4
n

S.y/ D

u.x; y/ D

1
16 X . 1/nC1
.2n 1/y
sin
I
2
2
 nD1 .2n 1/
4

1
64 X
cosh.2n 1/ x=4
.2n 1/y
. 1/nC1
sin
:
3
3
 nD1
.2n 1/ sinh.2n 1/=4
4

12.3.28. The boundary conditions require products v.x; y/ D X.x/Y .y/ such that (A) X 00 X D 0,
X 0 .0/ D 1, X.a/ D 0, and (B) Y 00 C Y D 0, Y 0 .0/ D 0, Y 0 .b/ D 0. From Theorem 11.1.3, the

Section 12.3 Laplaces Equation in Rectangular Coordinates

267

n2  2
, with associeigenvalues of (B) are 0 D 0, with associated eigenfunction Y0 D 1, and n D
b2
ny
ated eigenfunctions Yn D cos
, n D 1, 2, 3,. . . . Substituting 0 D 0 into (A) yields X000 D 0,
b
n2  2
into (A) yields Xn00 .n2  2=b 2 /Xn D
X00 .0/ D 1 X.a/ D 0, so X0 D x a. Substituting  D
b2
b sinh n.x a/=b
0, Xn0 .0/ D 1, Xn .a/ D 0, so Xn D
. Then vn .x; y/ D Xn .x/Yn .y/ D
n cosh n a=b
ny
@vn
ny
b sinh n.x a/=b
cos
, so
.0; y/ D cos
. Therefore,vn is solution of the given probn cosh n a=b
b
@x
b
ny
. More generally, if 0 ; : : : ; m are arbitrary constants, then um .x; y/ D
lem with g.y/ D cos
b
m
b X sinh n.x a/=b
ny
0 .x
a/ C
n
cos
is a solution of the given problem with g.y/ D
 nD1
n cosh n a=b
b
m
X
ny
n cos
. Therefore, if g is an arbitrary piecewise smooth function on 0; b we define the formal
b
nD1
1
b X sinh n.x a/=b
ny
solution of the given problem to be u.x; y/ D 0 .x a/ C
n
cos
where
 nD1
n cosh n a=b
b
Z
1
X
1 b
ny
ny
C.y/ D 0 C
n cos
is the Fourier cosine series of g on 0; b; that is, 0 D
g.y/ cos
dy,
b
b 0
b
nD1
Z
2 b
ny
n D
g.y/ cos
dy, n  1.
b 0
b
Now consider the special case. From Example 11.3.1,
C.y/ D
u.x; y/ D

.x

2/
2


2

1
4 X
1
cos.2n
 nD1 .2n 1/2

1/yI

1
4 X sinh.2n 1/.x 2/
cos.2n
 nD1 .2n 1/3 cosh 2.2n 1/

1/y:

12.3.30. The boundary conditions require products v.x; y/ D X.x/Y .y/ such that (A) X 00 C X D 0,
X 0 .0/ D 0, X.a/ D 0, and (B) Y 00 Y D 0, Y .0/ D 1, and Y is bounded. From Theorem 11.1.5, the
.2n 1/2  2
.2n 1/ x
eigenvalues of (A) are n D
, with associated eigenfunctions Yn D cos
, n D 1,
2
4a
2a
2 2
.2n 1/ 
2, 3,. . . . Substituting  D
into (B) yields Yn00 ..2n 1/2  2=4a2/Yn D 0, Yn .0/ D 1, so
4a2
.2n 1/ x
Yn D e .2n 1/y=2a . Then vn .x; y/ D Xn .x/Yn .y/ D e .2n 1/y=2a cos
, so vn .x; 0/ D
2a
.2n 1/ x
.2n 1/ x
cos
. Therefore,vn is solution of the given problem with f .x/ D cos
. More
2a
2a
m
X
.2n 1/ x
generally, if 1 ; : : : ; m are arbitrary constants, then um .x; y/ D
n e .2n 1/y=2a cos
2a
nD1
m
X
.2n 1/ x
is a solution of the given problem with f .x/ D
n cos
. Therefore, if f is an ar2a
nD1
bitrary piecewise smooth function on 0; a we define the formal solution of the given problem to be

268 Chapter 12 Fourier Solutions of Partial Differential Equations


1
X

1
X
1/ x
.2n 1/ x
, where Cm .x/ D
n cos
is the mixed
2a
2a
nD1
nD1
Z
2 a
.2n 1/ x
f .x/ cos
.
Fourier cosine series of f on 0; a; that is, n D
a 0
2a
Now consider the special case. Since f 0 .0/ D f .L/ D 0 and f 00 .x/ D 2, Theorem 11.3.5(d) implies
that
Z 3
16L
.2n 1/ x
n D
cos
dx
2
2
.2n 1/  0
6

288
.2n 1/ x 3
288
nC1
D
sin
I
D . 1/
.2n 1/3  3
6
.2n 1/3  3

u.x; y/ D

n e

.2n 1/y=2a

cos

.2n

CM .x/ D

u.x; y/ D

1
X

288
. 1/n
.2n 1/ x
cos
I
 3 nD1 .2n 1/3
6

1
288 X . 1/n
e
3
.2n 1/3

.2n 1/y=6

cos

nD1

.2n

1/ x
:
6

12.3.32. The boundary conditions require products v.x; y/ D X.x/Y .y/ such that (A) X 00 C X D
0, X.0/ D 0, X.a/ D 0, and (B) Y 00 Y D 0, Y 0 .0/ D 1, and Y is bounded. From Theon2  2
n x
rem 11.1.2, the eigenvalues of (A) are n D
, with associated eigenfunctions Yn D sin
,
2
a
a
2 2
n 
n D 1, 2, 3,. . . . Substituting  D
into (B) yields Yn00 .n2  2=a2 /Yn D 0, Yn0 .0/ D 1, so
a2
a
n x
@vn
n x
a
Yn D
e ny=a . Then vn .x; y/ D Xn .x/Yn .y/ D
e ny=a sin
, so
.x; 0/ D sin
.
n
n
a
@y
a
n x
Therefore,vn is solution of the given problem with f .x/ D sin
. More generally, if 1 ; : : : ; m are
a
m
a X n ny=a
n x
arbitrary constants, then um .x; y/ D
e
sin
is a solution of the given problem

n
a
nD1
m
X
n x
with f .x/ D
n sin
. Therefore, if f is an arbitrary piecewise smooth function on 0; a we
a
nD1
1
n x
a X n ny=a
define the formal solution of the given problem to be u.x; y/ D
e
sin
, where
 nD1 n
a
Z
1
X
n x
2 a
n x
C.x/ D
n sin
is the Fourier sine series of f on 0; a; that is, n D
f .x/ sin
dx.
a
a
a
0
nD1
Now consider the special case. Since f .0/ D f . / D 0 and f 00 .x/ D 2 6x, Theorem 11.3.5(b)
implies that



Z 
Z 

2
2

n D
.2 6x/ sin nx dx D 3
.2 6x/ cos nx C 6
cos nx dx
n2  0
n 
0
0


4
12
4
D
1 C . 1/n 2 C 4 sin nx D
1 C . 1/n 2 I
3
n
n 
n3
0
S.x/ D 4

1
1
X
X
1 C . 1/n 2
1 C . 1/n 2
sin
nx;
u.x/
D
4
e
n3
n4

nD1

nD1

ny

sin nx;

Section 12.3 Laplaces Equation in Rectangular Coordinates

269

12.3.34. The boundary conditions require products v.x; y/ D X.x/Y .y/ such that (A) X 00 C X D 0,
X.0/ D 0, X 0 .a/ D 0, and (B) Y 00 Y D 0, Y 0 .0/ D 1, and Y is bounded. From Theorem 11.1.4, the
.2n 1/2  2
.2n 1/ x
eigenvalues of (A) are n D
, with associated eigenfunctions Yn D sin
, n D 1,
2
4a
2a
2 2
.2n 1/ 
into (B) yields Yn00 ..2n 1/2  2=4a2/Yn D 0, Yn0 .0/ D 1, so
2, 3,. . . . Substituting  D
4a2
2a
2a
.2n 1/ x
e .2n 1/y=2a . Then vn .x; y/ D Xn .x/Yn .y/ D
e .2n 1/y=2a sin
,
Yn D
.2n 1/
.2n 1/
2a
@vn
.2n 1/ x
.2n 1/ x
so
.x; 0/ D sin
. Therefore,vn is solution of the given problem with f .x/ D sin
.
@y
2a
2a
m
2a X n
.2n 1/ x
More generally, if 1 ; : : : ; m are arbitrary constants, then um .x; y/ D
e .2n 1/y=2a sin
 nD1 2n 1
2a
m
X
.2n 1/ x
. Therefore, if f is an aris a solution of the given problem with f .x/ D
n sin
2a
nD1
bitrary piecewise smooth function on 0; a we define the formal solution of the given problem to be
1
1
X
2a X n
.2n 1/ x
.2n 1/ x
u.x; y/ D
e .2n 1/y=2a sin
,where Sm .x/ D
n sin
is
 nD1 2n 1
2a
2a
nD1
Z
.2n 1/ x
2 a
the mixed Fourier sine series of f on 0; a; that is, n D
f .x/ sin
.
a 0
2a
Now consider the special case.
n

D
D
D
D
D

2
5

x 2/ sin
"

.5x
0

.2n

1/ x
dx
10

Z 5
1/ x 5
.2n 1/ x
.5x x / cos
.5 2x/ cos
dx

.2n 1/
10
10
0
0
"
#

Z 5
40
.2n 1/ x 5
.2n 1/ x
.5 2x/ sin
sin
dx
C2
.2n 1/2  2
10
10
0
0

200
800
.2n 1/ x 5
n
. 1/
cos

.2n 1/2  2 .2n 1/3  3


10
0
200
800
. 1/n
C
I
.2n 1/2  2
.2n 1/3  3
4

SM .x/ D
u.x; y/ D

.2n



1
200 X
1
4
.2n 1/ x
n
.
1/
C
sin
I
2
2
 nD1 .2n 1/
.2n 1/
10



1
2000 X
1
4
n
.
1/
C
e
 3 nD1 .2n 1/3
.2n 1/

.2n 1/y=10

sin

.2n

1/ x
:
10

12.3.36. Solving BVP.1; 1; 1; 1/.f0; 0; 0; 0/ requires products X.x/Y .y/ such that
X 00 C X D 0;
Hence, Xn D cos

X 0 .0/ D 0;

n x
, Yn D
a

X 0 .a/ D 0I

Y 00

Y D 0;

a cosh n.y b/=a


, and c1
n sinh n b=a

Y 0 .b/ D 0;

Y 0 .0/ D 1:

1
a X
cosh n.y b/=a
n x
An
cos
is

n sinh n b=a
a
nD1

270 Chapter 12 Fourier Solutions of Partial Differential Equations


1
X
n x
a formal solution of BVP.1; 1; 1; 1/.f0; 0; 0; 0/ if c1 is any constant and
An cos
is the Fourier
a
nD1
Z a
cosine expansion of f0 on 0; a, which is possible if and only if
f0 .x/ dx D 0.
0

1
a X
cosh ny=a
n x
Bn
cos
is a formal solution of BVP.1; 1; 1; 1/.0; f1; 0; 0/ if
Similarly, c2 C

n sinh n b=a
a
nD1
1
X
n x
c2 is any constant and
Bn cos
is the Fourier cosine expansion of f1 on 0; a, which is possible
a
nD1
Z a
if and only if
f1 .x/ dx D 0.
0

1
b X cosh n.x a/=b
ny
Cn
cos
is a formal so nD1
n sinh n a=b
b
1
X
ny
is the Fourier cosine expanlution of BVP.1; 1; 1; 1/.0; 0; g0; 0/ if c3 is any constant and
Cn cos
b
nD1
Z b
1
b X
cosh n x=b
ny
sion of g0 on 0; b, which is possible if and only if
g0 .x/ dx D 0, and c4 C
Dn
cos
 nD1
n sinh n a=b
b
0
1
X
ny
is a formal solution of BVP.1; 1; 1; 1/.0; 0; 0; g1/ if c4 is any constant and
Dn cos
is the Fourier
b
nD1
Z b
cosine expansion of g1 on 0; b, which is possible if and only if
g1 .x/ dx D 0.

Interchanging x and y and a and b shows that c3

Adding the four solutions yields


u.x; y/

CC
C

1
a X Bn cosh ny=a An cosh n.y
 nD1
n sinh n b=a

1
b X Dn cosh n x=b Cn cosh n.x
 nD1
n sinh n a=b

b/=a
a/=b

cos

n x
a

cos

ny
;
b

where C is an arbitrary constant.


12.4 LAPLACES EQUATION IN POLAR COORDINATES
12.4.2. v.r; / D R.r /./ where (A) r 2R00 C rR0 R D 0 and 00 C  D 0, .0/ D 0,
n2  2
n 
n2  2
,

D
sin
,
n
D
1,
2,
3,.
.
.
.
Substituting

D
. / D 0. From Theorem 11.1.2, n D
n
2

2
2 2
n 
into (A) yields the Euler equation r 2 Rn00 C rRn0
Rn D 0 for Rn . The indicial polynomial is
2



n
n
sC
, so Rn D c1 r n= C c2 r n= , by Theorem 7.4.3. We want Rn ./ D 1 and
s


n= n=
n=

r
0 r n=
Rn .0 / D 0, so Rn .r / D 0 n=
;
0
 n= 0n=  n=
vn .r; / D

0

n= n=

0

0

n= n=


0n=  n=

n=

n=

sin

n 
I

Section 12.4 Laplaces Equation in Polar Coordinates

1
X
n 
n 
,
where
S./
D
n sin
is the Fourier
n= n=
n=


0

0  n=
nD1
nD1
Z
1
n 
sine series if f on 0; ; that is, n D
f ./ sin
d, n D 1, 2, 3,. . . .
0

u.r; / D

1
X

271

0

n= n=

n=

0

n=

sin

12.4.4. v.r; / D R.r /./ where (A) r 2R00 C rR0 R D 0 and 00 C  D 0, 0 .0/ D 0, . / D 0.
.2n 1/2  2
.2n 1/ 
From Theorem 11.1.5, n D
, n D cos
, n D 1, 2, 3,. . . . Substituting
4 2
2
2 2
.2n 1/ 
.2n 1/2  2
2 00
0
into
(A)
yields
the
Euler
equation
r
R
C
rR
Rn D 0 for Rn . The
D
n
n
4 2
4 2



.2n 1/
.2n 1/
indicial polynomial is s
sC
, so Rn D c1r .2n 1/=2 C c2 r .2n 1/=2 ,
2
2
by Theorem 7.4.3. We want Rn to be bounded as r ! 0C and Rn ./ D 1, so we take Rn .r / D
1
X
r .2n 1/=2
r .2n 1/=2
.2n 1/ 
r .2n 1/=2
.2n 1/ 
;
v
.r;
/
D
cos
;
u.r;
/
D

cos
,
n
n
.2n 1/=2
2
2
 .2n 1/=2
 .2n 1/=2

nD1
1
X
.2n 1/ 
where CM ./ D
n cos
is the mixed Fourier cosine series of f on 0; ; that is, n D
2
nD1
Z
2
.2n 1/ 
f ./ cos
d, n D 1, 2, 3,. . . .
0
2
12.4.6. v.r; / D R.r /./ where (A) r 2R00 C rR0 R D 0 and 00 C  D 0, 0 .0/ D 0, 0 . / D 0.
n 
n2  2
, n D cos
, n D 1, 2, 3,. . . . Substituting
From Theorem 11.1.3, 0 D 0, 0 D 1; n D
2


 D 0 into (A) yields the equation r 2R000 C rR00 D 0 for R0 ; R0 D c1 C c2 ln r . Since we want R0 to be
bounded as r ! 0C and R0 ./ D 1, R0 .r / D 1; therefore v0 .r; / D 1.
n2  2
n2  2
2 00
0
Substituting  D
into
(A)
yields
the
Euler
equation
r
R
C
rR
Rn D 0 for Rn . The
n
n
2
2


n
n
indicial polynomial is s
sC
, so Rn D c1 r n= C c2r n= , by Theorem 7.4.3. Since


r n=
r n=
n 
we want Rn to be bounded as r ! 0C and Rn ./ D 1, Rn .r / D n= ; vn .r; / D n= cos
,



1
1
X r n=
X
n 
n 
n D 1, 2, 3,. . . ; u.r; / D 0 C
n n= cos
, where F ./ D 0 C
n cos
is the



nD1
nD1
Z
Z
2
n 
1
Fourier cosine series of f on 0; ; that is, 0 D
f ./ d and n D
f ./ cos
d,
0
0

n D 1, 2, 3,. . . .

CHAPTER 13
Boundary Value Problems for Second
Order Ordinary Differential Equations

13.1 BOUNDARY VALUE PROBLEMS


x
x; y D  x C c1e xC
 c2 e  ; y.0/
 D  2 H) c1 C c2 D 2;
2e
2
1 1
c1
2
y.1/ D 1 H)
1 C c1 e C c2 =e D 1;
D
; c1 D
; c2 D
;
e 1=e
c2
2
e 1
1 e

2 e x e .x 1/
yD xC
e 1

13.1.2. By inspection, yp D

13.1.4. By inspection, yp D
3 H)

x; y D

1 C 2c1 D 3; c1 D 2; y.1/

x C c1 e x C c2 e

x
; y 0 D 1 C c1 e x c2 e x ; y.0/ C y 0 .0/ D
2c
2
y 0 .1/ D 2 H)
D 2; c2 D e; y D x C 2e x C e .x 1/
e

13.1.6. yp D Ax 2e x ; yp0 D A.x 2 C 2xe x /; yp00 D A.x 2 e x C 4xe x C 2e x /; yp00 2yp0 C yp D 2Ae x D e x
if A D 1; yp D x 2 e x ; y D .x 2 C c1 C c2 x/e x ; y 0 D .x 2 C 2x C c1 C c2 C c2 x/e x ; B1 .y/ D 3 and
B2 .y/ D 6e H) c1 2c2 D 3, 2c1 C 3c2 D 24; c1 D 13; c2 D 8; y D .x 2 8x C 13/e x .
13.1.8. B1.y/ D y.0/; B2 .y/ D y.1/ y 0 .1/. Let y1 D x, y2 D 1; B1 .y1 / D B2.y1 / D 0. By
variation of parameters, if yp D u1 x C u2 where u01 x C u02 D 0 and u01 D F , then yp00 D F .x/. Let
Z 1
Z x
Z 1
Z x
u01 D F , u02 D xF ; u1 D
F .t/ dt, u2 D
tF .t/ dt; yp D x
F .t/ dt
tF .t/ dt;
x
0
x
0
Z 1
yp0 D
F .t/ dt; y D yp C c1x C c2 . Since B1 .yp / D 0, B1 .x/ D 0 and B1 .1/ D 1, B1 .y/ D
x
Z 1
0 H) c2 D 0; hence, y D yp C c1x. Since B2 .yp / D
tF .t/ dt and B2 .x/ D 0, B2.y/ D 0 H)
0
Z 1
tF .t/ dt D 0: There is no solution if this conditions does not hold. If it does hold, then the solutions
0

are y D yp C c1x, with c1 arbitrary.

13.1.10. (a) The condition is b a .kC1=2/ (k D integer). Let y1 D sin.x a/ and y2 D cos.x b/.
Then y1 .a/ D y20 .b/ D 0 and fy1 ; y2 g is linearly independent if b a .k C 1=2/ (k D integer),
since

sin.x a/
cos.x b/

D cos.b a/ 0:
cos.x a/
sin.x b/
273

274 Chapter 13 Boundary Value Problems for Second Order Ordinary Differential Equations
Now Theorem 13.1.2 implies that (A) has a unique solution for any continuous F and constants k1 and
k2 . If y D u1 sin.x a/ C u2 cos.x b/ where
u01 sin.x a/ C u02 cos.x
u01 cos.x a/ u02 sin.x
then y 00 C y D F .

u1 D

u01 D F .x/

1
cos.b a/

yD

u02 D

F .t/ cos.t
Z

sin.x
cos.b

a/
;
a/
Z x
F .t/ sin.t

1
cos.b a/ a
Z
cos.x b/ x
F .t/ sin.t
cos.b a/ a

F .t/ cos.t

F .x/

u2 D

b/ dt;

a/
a/

sin.x
cos.b

b/
;
a/

cos.x
cos.b

D 0
D F;

b/
b/

b/ dt

b/ dtI

a/ dt:

(b) If b a D .k C 1=2/ (k D integer), then y1 D sin.x a/ satisfies both boundary conditions


y.a/ D 0 and y 0 .b/ D 0. Let y2 D cos.x a/. If yp D u1 sin.x a/ C u2 cos.x/ where
u01 sin.x
u01

cos.x

then yp00 C yp D F ; u01 D F cos.x


Z

u1 D
yp D

sin.x

yp0 D

cos.x

a/ C u02 cos.x
a/

a/; u02 D

u02

sin.x

a/

F sin.x

a/;

F .t/ cos.t

a/ dt;

a/
a/

u2 D

F .t/ cos.t

a/ dt

D 0;

a/

D F;
x

F .t/ sin.t

cos.x

a/

F .t/ cos.t
x

a/ dtI

a/ dt C sin.x

a/

F .t/ sin.t

a/ dtI

F .t/ sin.t

a/ dt:

x
a

The general solution of y 00 Cy D F is y D yp Cc1 sin.x a/Cc2 cos.x a/. Since b a D .k C1=2/ ,
Z b
Z b
y 0 .b/ D 0 H) yp0 .b/ D . 1/k
F .t/ sin.t a/ dt D 0; therefore, F must satisfy
F .t/ sin.t
a

a/ dt D 0. In this case, the solutions of the boundary value problem are y D yp C c1 sin.x
arbitrary.
a/ and y2 D sinh.x b/. Then y1 .a/ D 0, y2 .b/ D 0, and

sinh.x a/ sinh.x b/

D sinh.b a/ 0:
W .x/ D
cosh.x a/ cosh.x b/

a/, with c1

13.1.12. Let y1 D sinh.x

(Since W is constant (Theorem 5.1.4), evaluate it by setting x D b.) From Theorem 13.1.2, (A) has a
unique solution for any continuous F and constants k1 and k2 . If y D u1 sinh.x a/ C u2 sinh.x b/
where
u01 sinh.x
u01 cosh.x
then y 00

y D F.

u01 D F .x/

a/ C u02 sinh.x
a/ C u02 cosh.x

sinh.x
sinh.b

b/
;
a/

b/ D 0
b/ D F;

u02 D F .x/

sinh.x
sinh.b

a/
;
a/

Section 13.1 Boundary Value Problems


1
u1 D
sinh.b

a/

F .t/ sinh.t
x

a/
a/

sinh.x
yD
sinh.b

b/ dt;

1
u2 D
sinh.b

b/ dt C

F .t/ sinh.t
x

sinh.x
sinh.b

b/
a/

a/
Z x

275

F .t/ sinh.t

a/ dtI

F .t/ sinh.t

a/ dt:

a/ and y2 D cosh.x b/. Then y10 .a/ D y20 .b/ D 0 and

cosh.x a/ cosh.x b/

D sinh.b a/ 0:
W .x/ D
sinh.x a/ sinh.x b/

13.1.14. Let y1 D cosh.x

(Since W is constant (Theorem 5.1.40, evaluate it by setting x D b.) If


y D u1 cosh.x a/ C u2 cosh.x b/ where
u01 cosh.x
u01
then y 00

y D F.

u1 D

1
sinh.b

yD

sinh.x

u01 D F .x/
a/

cosh.x
sinh.b

a/ C u02 cosh.x
a/ C u02 sinh.x
b/
;
a/

cosh.x
sinh.b

u02 D

F .t/ cosh.t

b/ dt;

a/
a/

u2 D

F .t/ cosh.t
x

b/ dt

b/ D 0

b/ D F;
F .x/

1
sinh.b

cosh.x
sinh.b

a/
;
a/

cosh.x
sinh.b

b/
a/

a/
Z x

F .t/ cosh.t

F .t/ cosh.t

a/ dt:

13.1.16. Let y1 D sin !x, y2 sin !.x  /; then y1 .0/ D 0, y2 . / D 0,

sin !x
sin !.x  /

W .x/ D
D ! sin ! 0
! cos !x ! cos !.x  /

if and only if ! is not an integer. If this is so, then y D u1 sin !x C u2 sin !.x
u01 sin !x C u02 sin !.x  /
!.u01 cos !x C u02 cos !.x  //

a/ dtI

 / if

D 0
D FI

sin !.x  /
sin !x
u01 D F
; u02 D F
I
! sin !
! sin !
Z 
Z x
1
1
u1 D
F .t/ sin !.t  / dtI u2 D
F .t/ sin !t dtI
! sin ! x
! sin ! 0


Z 
Z 
1
yD
sin !x
F .t/ sin !.t  / dt C sin !.x  /
F .t/ sin !t dt :
! sin !
x
0

If ! D n (positive integer), then y1 D sin nx is a nontrivialsolution of y 00 C y D 0, y.0/ D 0, y. / D


sin nx
cos nx
0. Let y2 D cos nx; then W .x/ D
D n, and yp D u1 sin nx C u2 cos nx
n cos nx
n sin nx

satisfies yp00 C n2 yp D 0 if

u01 sin nx C u02 cos nx


nu01 cos nx nu02 sin nx

D 0
D FI

276 Chapter 13 Boundary Value Problems for Second Order Ordinary Differential Equations
u01 D

Z
Z
1
1 
1 x
F sin nx; u1 D
F .t/ cos nt dt; u2 D
F .t/ sin nt dt;
n
n x
n 0


Z 
Z x
1
sin nx
F .t/ cos nt dt C cos nx
F .t/ sin nt dt I
n
x
0

1
F cos nx, u02 D
n
yp D

y D yp CZc1 sin nx C c2 cos nx. Since yp D 0, y.0/ D 0, so c2 D 0; y D yp C c1 sin nx. Since



y. / D 0,
F .t/ sin nt dt D 0 is necessary for existence of a solution. If this hold, then the solutions
0

are y D yp C c1 sin nx, with c1 arbitrary.

13.1.18. Let y1 D cos !x; y2 D sin !.x  /; then y10 .0/ D y2 . / D 0, and

cos !x
sin !.x  /

W .x/ D
D ! cos ! 0
! sin !x ! cos !.x  /

if and only if ! n C 1=2 (n D integer). If this is so, then y D u1 cos !x C u2 sin !.x
y 00 C ! 2 y D F .x/ if
!.
then

u01

u01 cos !x C u02 sin !.x


sin !x

C u02

cos !.x

 / satisfies

/ D 0

 //!

D FI

F cos !x
;
! cos !
Z 
Z x
1
1
u1 D
F .t/ sin !.t  /; dt; u2 D
F .t/ cos !t dt;
! cos ! x
! cos ! 0


Z 
Z x
1
yD
sin !x
F .t/ sin !.t  /; dt C sin !.x  /
F .t/ cos !t dt :
! cos !
x
0
F sin !.x  /
;
! cos !

u01 D

u02 D

If ! D n C 1=2 (n D integer), then y1 D cos.n C 1=2/x is a nontrivial solution y 00 C y D 0,


y .0/ D y. / D 0. Let y2 D sin.n C 1=2/x; then

cos.n C 1=2/x
sin.n C 1=2/x

D n C 1=2;
W .x/ D
.n C 1=2/ sin.n C 1=2/x .n C 1=2/ cos.n C 1=2/x
0

so yp D u1 cos.n C 1=2/x C u2 sin.n C 1=2/x satisfies yp00 C .n C 1=2/2 yp D F if


u01 cos.n C 1=2/x C u02 sin.n C 1=2/x
.n C 1=2/u01 sin.n C 1=2/x C .n C 1=2/u02 cos.n C 1=2/x
u01 D

yp D

1
n C 1=2

yp0 D

u02 D

F cos.n C 1=2/x
I
n C 1=2

Z x
F .t/ sin.n C 1=2/t
F .t/ cos.n C 1=2/t
u1 D
dtI u2 D
dtI
n
C
1=2
n C 1=2
x
0


Z 
Z x
cos.n C 1=2/x
F .t/ sin.n C 1=2/t dt C sin.n C 1=2/x
F .t/ cos.n C 1=2/t dt I
Z

F sin.n C 1=2/x
I
n C 1=2

D 0
D F:

sin.n C 1=2/x

F .t/ sin.n C 1=2/t dt C cos.n C 1=2/x

F .t/ cos.n C 1=2/t dtI

Section 13.1 Boundary Value Problems

yp0 .0/

y D

y D yp C c1 cos.n C 1=2/x C c2 sin.n C 1=2/xI

yp0

C .n C 1=2/. c1 sin.n C 1=2/x C c2 cos.n C 1=2//x:

D 0, y .0/ D 0 H) c2 D 0; y D yp C c1 cos.n C 1=2/x;


Z
0
0
y D yp C .n C 1=2/c1 .n C 1=2/x; y. / D 0 H) yp . / D 0. Hence,
Since

277


0

F .y/ cos.n C 1=2/t dt D 0

is a necessary condition for existence of a solution. If this holds, then the solutions are y D yp C
c1 cos.n C 1=2/x with c1 arbitrary.

13.1.20. Suppose y D c1 1 C c2 2 is a nontrivial solution of the homogeneous boundary value problem. Then B1 .y/ D c1 B1 .1 / C c2 B1 .2 / D 0. From Theorem 13.1.1 we may assume without
B1 .1 /
loss of generality that B1 .2 / 0. Then c2 D
c1 . Therefore, y is constant multiple of
B1 .2 /
y0 D B1 .2 /1 B1.1 /2 0. To that check y satisfies the boundary conditions, note that B1 .y0 / D
B1 .2 /B1 .1 / B1 .1 /B1 .2 / D 0 B2 .y0 / D B1 .2 /B2 .1 / B1 .1 /B2 .2 / D 0, by Theorem 13.1.2.
13.1.22. y1 D a1 C a2 x; y1 .0/ 2y10 .0/ D a1 2a2 D 0 if a1 D 2, a2 D 1; y1 D 2 C x. y2 D b1 C b2x;
y2 .1/ D 2y10 .1/ D b1 C 3b2 D 0 if b1 D 3, b2 D 1; y2 D 3 x.
8

.2 C t/.3 x/ ;
0  t  x;
<
2Cx 3 x
5
D 5I G.x; t/ D
W .x/ D
1
1

.2 C x/.3 t/

:
;
x  t  1:
5


Z 1
Z x
1
yD
.2 C x/
.3 t/F .t/ dt C .3 x/
.2 C t/F .t/ dt :
(B)
5
x
0
(a) With F .x/ D 1, (B) becomes


Z 1
Z x
1
y D
.2 C x/
.3 t/ dt C .3 x/
.2 C t/ dt
5
x
0

 2

 2

x
6x C 5
x C 4x
1
D
.2 C x/
C .3 x/
5
2
2
2
x
x 2
D
:
2
(b) With F .x/ D x, (B) becomes


Z 1
Z x
1
y D
.2 C x/
.3t t 2 / dt C .3 x/
.2t C t 2 / dt
5
x
0

 3

 3

2
1
2x
9x C 7
x C 3x 2
D
.2 C x/
C .3 x/
5
6
3
3
5x
7x 14
D
:
30
(b) With F .x/ D x 2 , (B) becomes


Z 1
Z x
1
.2 C x/
.3t 2 t 3 / dt C .3 x/
.2t 2 C t 3 / dt
y D
5
x
0

 4

 4

2
1
x
4x C 3
3x C 8x 3
D
.2 C x/
C .3 x/
5
4
12
4
5x
9x 18
D
:
60

278 Chapter 13 Boundary Value Problems for Second Order Ordinary Differential Equations
2
x
x, y2 D x 2 2x; then y1 .1/ D 0, y2 .2/ D 0; W .x/ D
2x
8
.t
1/x.x
2/

;
1  t  x;
<
t3
2
:
Since P0 .x/ D x , G.x; t/ D

x.x 1/.t 2/

:
;
x

t

2:
t3
Z 2
Z x
t 2
y D x.x 1/
F
.t/
dt
C
x.x
2/
F .t/ dt:
t3
x
1
13.1.24. y1 D x 2

(a) With F .x/ D 2x 3 , (B) becomes


Z 2
y D 2x.x 1/
.t
x

x.x

1/.x

2/ dt C 2x.x
2/2 C x.x

(b) With F .x/ D 6x 4, (B) becomes


Z 2
y D 6x.x 1/
.t 2/t dt C 6x.x
x

2x.x

1/.x C 1/.x

2/2 C x.x

2/

2/

2/.x

x
1

x 2 2x
D x2.
2x 2

(B)

1
2

.t

1/ dt

1/ D x.x

1/.x

2/:

.t

1/t dt

2/.x

1/2 .2x C 1/ D x.x

1/.x

2/.x C 3/:

13.1.26. y1 D a1 C a2 x; y10 D a2 ; B1 .y1 / D a1 C a2 D 0 if a1 D , a2 D ; y1 D x.


0
y2 D b1 C
 b2 x; y2 D b2 ; B2.y2 / D b1 C . C /b2 D 0 if b1 D  C , b2 D ; y2 D  C x;
x  C x
W .x/ D
D . C / . From Theorem 13.1.2, (A) has a unique solution


if and only if . C /  0. Then
8
. t/. C x/

; 0  t  x;
<
. C / 
G.x; t/ D
. x/. C t/

:
; x  t  1:
. C / 

13.1.28. y1 D a1 cos x Ca2 sin x; y10 D a1 sin x Ca2 cos x; B1 .y1 / D a1 Ca2 D 0 if a1 D , a2 D
. y1 D cos x sin x. y2 D b1 cos x C b2 sin x; y20 D b1 sin x C b2 cos x; B2 .y2 / D b2 b1 D
cos x sin x
 cos x C sin x
0 if b1 D , b2 D ; y2 D  cos x C sin x; W .x/ D
Since
sin x cos x
 sin x C cos x



W is constant, we can evaluate it with x D 0: W D
D  C . From Theorem 13.1.2, (A)

has a unique solution if and only if  C 0. Then
8
. cos t sin t/. cos x C sin x/

; 0  t  x;
<
 C
G.x; t/ D
. cos x sin x/. cos t C sin t//

:
; x  t  :
 C
13.1.30. y1 D e x .a1 cos x C a2 sin x/; y10 D e x a1 .cos x sin x/ C a2 .sin x C cos x/;
B1 .y1 / D . C /a1 C a2 D 0 if a1 D , a2 D . C /.

Section 13.2 Sturm-Liouville Problems


y1 D e x . cos x . C / sin x. y2 D e x .b1 cos x C b2 sin x/;
y20 D e x .b1 .cos x sin x// C b2 .sin x C cos x/;
B2 .y2 / D e =2. C /b2 b1 D 0 if b1 D , b2 D . C /;
y2 D e x . C / cos x C sin x/;
To evaluate W .x/, we write y1 D e x v1 and y2 D e x v2 , where
v1 D cos x . C / sin x and v2 D . C / cos x C sin x.
Then y10 D y1 C e x v10 and y2 D y2 C e x y20 ,

y1

y1
y2
y2
v2
2x v1

W .x/ D
D
D
e
x 0
x 0
x 0
0

v 0 xv 0
y1 C e v1 y2 C e v2
e v1 xv2
1
2

. cos x . C / sin x

. C / cos x C sin x
D
:
. sin x . C / cos x . . C / sin x C cos x

279

Since vi00 C vi D 0, i D 1, 2, Theorem 5.1.4 implies that W .x/ D Ke 2x , where is a constant that can be
determined by setting x D 0 in the determinant:

 C
W .x/e 2x
D C . C /. C /:

From Theorem 13.1.2, the boundary value problem has a unique solution if and only if C . C /. C
/ 0. In this case the Greens function is
8
cos t . C / sin t C / cos x C sin x

; atx
<e x t
C . C /. C /
G.x; t/ D
cos x . C / sin x C / cos t C sin t

; x  t  =2:
:e x t
C . C /. C /
13.1.32. Let yp D

G.x; t/F .t/ dt. From Theorem 13.1.3, Lyp D F , B1 .yp / D 0, and B2 .yp / D 0.

The solution of Ly D F , B1 .y/ D k1 , and B2 .y/ D k2 is of the form y D yp C c1 y1 C c2y2 . Since


k1
B1 .yp / D 0 and B1 .y1 / D 0, B1 .y/ D k1 H) k1 D c2 B1 .y2 / H) c2 D
. Since
B1 .y2 /
k2
B2 .yp / D 0 and B2 .y2 / D 0, B2 .y/ D k2 H) k2 D c1B2 .y1 / H) c1 D
.
B2 .y1 /
13.2 STURM-LIOUVILLE PROBLEMS



1
2
p0
1
00
0
13.2.2. y 00 C y 0 C 1
y
D
0;
D
;
ln
jpj
D
ln
jxj;
p
D
x;
xy
C
y
D
x
x2
p
x
 x 2

.xy 0 /0 C x
y D 0.
x
b 0
c
p0
b
y C 2 y D 0;
D ; ln jpj D b ln jxj; p D x b ;
x
x
p
x
x b y 00 C bx b 1y 0 C cx b 2 y D 0; .x b y 0 /0 C cx b 2y D 0.


1

p0
1
00
0
00
13.2.6. xy C .1 x/y C y D 0; y C
1 y 0 C y D 0;
D
x
x
p
x
p D xe x ; xe x y 00 C .1 x/y 0 C e x y D 0; .xe x y 0 /0 C e x y D 0.


2
y D 0;
x

13.2.4. y 00 C

1; ln jpj D ln jxj

x;

280 Chapter 13 Boundary Value Problems for Second Order Ordinary Differential Equations
13.2.8. If  is an eigenvalues of (A) and y is a -eigenfunction, multiplying the differential equation in

(B) by y yields .xy 0 /y C y 2 D 0;
x
2 Z 2
Z 2 2
Z 2

y .x/
0
0
0

dx D
.xy .x// y.x/ dx D xy .x/y.x/ C
x.y 0 .x//2 dxI
2
x
1
1
1
1
2
Z 2 2
Z 2

y .x/
0
y.1/ D y. 2/ D 0 H) xy .x/y.x/ D 0I 
dx D
x.y 0 .x//2 dx:
x
1
1
1

Therefore   0: We must still show that  D 0 is not an eigenvalue. To this end, suppose that .xy 0 /0 D 0;
c1
; y D c1 ln jxj C c2 ; y.1/ D 0 H) c2 D 0; y D c1 ln jxj; y.2/ D 0 H)
then xy 0 D c1 ; y 0 D
x
c1 D 0; y  0; therefore  D 0 is not an eigenvalue.
p
13.2.10. Characteristic equation: r 2 C 2r C 1 C  D 0; r D 1
.
 D 0: y D e x .c1 C c2x/; y 0 D e x .c1 c2 C c2 x/; y 0 .0/ D 0 H) c1 D c2 ; y 0 D c2 xe x ;
y 0 .1/ D 0 H) c2 =e D 0 H) c2 D 0;  D 0 is not an eigenvalue.
 D k 2 , k > 0: r D 1 k; y D e x .c1 cosh kx C c2 sinh kx/;
y 0 D c1e x . cosh kx C k sinh kx/ C c2 e x . sinh kx C k cosh kx/.
The boundary conditions require that
c1 C c2 k D 0 and . cosh k C k sinh k/c1 C . sinh k C k cosh k/c2 D 0.
This system has a nontrivial solution if and only if .1 k 2 / sinh k D 0. Let k D 1 and c1 D c2 D 1; then
 D 1 is the only negative eigenvalue, with associated eigenfunction y D 1.
 D k 2 , k > 0: r D 1 i k; y D e x .c1 cos kx C c2 sin kx/;
y 0 D c1 e x . cos kx k sin kx/ C c2 e x . sin x C k cos kx/. The boundary conditions require that
c1 C c2 k D 0 and . cos k k sin k/c1 C . sin k C k cos k/c2 D 0.
This system has a nontrivial solution if and only if .1 C k 2 / sin k D 0. Let k D n (k a positive
integer)and c1 D n , c2 D 1; then n D n2  2 is an eigenvalue, with associated eigenfunction yn D
e x .n cos n x C sin n x/.
13.2.12. Characteristic equation: r 2 C  D 0.
 D 0 W y D c1 C c2x. y.0/ D 0 H) c1 D 0, so y D c2x. Now y.1/ 2y 0 .1/ D 0 H) c2 D 0.
Therefore  D 0 is not an eigenvalue.
 D k 2 , k > 0: y D c1 cosh kx C c2 sinh kx; y 0 D k.c1 sinh kx C c2 cosh kx/. y 0 .0/ H) c2 D 0,
so y D c1 cosh kx. Now y.1/ 2y 0 .1/ D 0 H) c1.cosh k 2k sinh k/ D 0, which is possible with
1
. Graphing both sides of this equation on the same axes show that it
c1 0 if and only if tanh k D
2k
has one positive solution k0 ; y0 D cosh k0 x is a k02 -eigefunction.
 D k 2 , k > 0: y D c1 cos kx C c2 sin kx; y 0 D k. c1 sin kx C c2 cos kx/. y 0 .0/ H) c2 D 0, so
y D c1 cos kx. Now y.1/ 2y 0 .1/ D 0 H) c1.cos k C 2k sin k/ D 0, which is possible with c1 0
1
if and only if tan k D
. Graphing both sides of this equation on the same axes shows that it has a
2k
solution kn in ..2n 1/=2; n /, n D 1, 2, 3, . . . ; yn D cos kn x is a kn2 -eigenfunction.
13.2.14. Characteristic equation: r 2 C  D 0.
 D 0 W y D c1 C c2x. The boundary conditions require that c1 C 2c2 D 0 and c1 C  c2 D 0, which
imply that c1 D c2 D 0, so  D 0 is not an eigenvalue.
 D k 2 , k > 0: y D c1 cosh kx C c2 sinh kx; y 0 D k.c1 sinh kx C c2 cosh kx/. The boundary
conditions require that
c1 C 2kc2 D 0 and c1 cosh k C c2 sinh 2k D 0.
This system has a nontrivial solution if and only if tanh k D 2k. Graphing both sides of this equation

Section 13.2 Sturm-Liouville Problems

281

on the same axes shows that it has a solution k0 in .0;  /; y0 D 2k0 cosh k0 x sinh k0 x is a k02 eigenfunction.
 D k 2 , k > 0: y D c1 cos kx C c2 sin kx; y 0 D k. c1 sin kx C c2 cos kx/. The boundary conditions
require that
c1 C 2kc2 D 0 and c1 cos k C c2 sin k D 0.
This system has a nontrivial solution if and only if tan k D 2k. Graphing both sides of this equation on
the same axes shows that it has a solution kn in .n; n C 1=2/, n D 1, 2,3, . . . ; yn D 2kn cos kn x sin kn x
is a kn2 -eigenfunction.
13.2.16. Characteristic equation: r 2 C  D 0.
 D 0 W y D c1 C c2x. The boundary conditions require that c1 C c2 D 0 and c1 C 4c2 D 0, so
c1 D c2 D 0. Therefore  D 0 is not an eigenvalue.
 D k 2 , k > 0: y D c1 cosh kx C c2 sinh kx; y 0 D k.c1 sinh kx C c2 cosh kx/. The boundary
conditions require that
c1 C kc2 D 0 and .cosh 2k C 2k sinh 2k/c1 C .sinh 2k C 2k cosh 2k/c2 D 0.
k
This system has a nontrivial solution if and only if tanh 2k D
. Graphing both sides of this
1
p 2k 2
equation on the same axes shows that it has a solution k0 in .1= 2/; y0 D k0 cosh k0 x sinh k0 x is a
k02 - eigenfunction.
 D k 2 , k > 0: y D c1 cos kx C c2 sin kx; y 0 D k. c1 sin kx C c2 cos kx/. The boundary conditions
require that
c1 C kc2 D 0 and .cos 2k 2k sin 2k/c1 C .sin 2k C 2k cos 2k/c2 D 0.
k
This system has a nontrivial solution if and only if tan 2k D
. Graphing both sides of this
1 C 2k 2
equation on the same axes shows that it has a solution kn in ..2n 1/=4; n=2/, n D 1, 2,3, . . . ;
yn D kn cos kn x sin kn x is a kn2 -eigenfunction.
13.2.18. Characteristic equation: r 2 C  D 0.
 D 0 W y D c1 C c2 x. The boundary conditions require that 3c1 C 2c2 D 0 and 3c1 C 4c2 D 0, so
c1 D c2 D 0. Therefore  D 0 is not an eigenvalue.
 D k 2 , k > 0: y D c1 cosh kx C c2 sinh kx; y 0 D k.c1 sinh kx C c2 cosh kx/. The boundary
conditions require that
3c1 C kc2 D 0 and .3 cosh 2k 2k sinh 2k/c1 C .3 sinh 3k 2k cosh 2k/c2 D 0.
9k
This system has a nontrivial solution if and only if tanh 2k D
. Graphing both sides of this equa9 C 2k 2
tion on the same axes shows that it has solutions y1 in .1; 2/ and y2 in .5=2; 7=2/; yn D kn cosh kn x
3 sinh kn x is a kn2 -eigenfunction, k D 1, 2.
 D k 2 , k > 0: y D c1 cos kx C c2 sin kx; y 0 D k. c1 sin kx C c2 cos kx/. The boundary conditions
require that
3c1 C kc2 D 0 and .3 cos 2k C 2k sin 2k/c1 C .3 sin 2k 2k cos 2k/c2 D 0.
9k
This system has a nontrivial solution if and only if tan 2k D
. Graphing both sides of this
9
p 2k 2
equation on the same axes shows that it has solutions k0 in .3= 2;  / and kn in ..2n C 3/=4; .n C
2/=3/, n D 1, 2,3, . . . ; yn D kn cos kn x 3 sin kn x is a kn2 -eigenfunction.
13.2.20. Characteristic equation: r 2 C  D 0.
 D 0 W y D c1 C c2 x. The boundary conditions require that 5c1 C 2c2 D 0 and 5c1 C 3c2, so
c1 D c2 D 0. Therefore  D 0 is not an eigenvalue.
 D k 2 , k > 0: y D c1 cosh kx C c2 sinh kx; y 0 D k.c1 sinh kx C c2 cosh kx/. The boundary
conditions require that
5c1 C 2kc2 D 0 and .5 cosh k 2k sinh k/c1 C .5 sinh k 2k cosh k/c2 D 0.

282 Chapter 13 Boundary Value Problems for Second Order Ordinary Differential Equations
20k
. Graphing both sides of this equa25 C 4k 2
tion on the same axes shows that it has solutions k1 in .1; 2/ and k2 in .5=2; 7=2/; yn D 2kn cosh kn x
sinh kn x is kn -eigenfunction, n D 1, 2.
 D k 2 , k > 0: y D c1 cos kx C c2 sin kx; y 0 D k. c1 sin kx C c2 cos kx/. The boundary conditions
require that
5c1 C 2kc2 D 0 and .5 cos k C 2k sin k/c1 C .5 sin k 2k cos k/c2 D 0.
20k
. Graphing both sides of this
This system has a nontrivial solution if and only if tan k D
25 4k 2
equation on the same axes shows that it has a solution kn in
..2n C 1/=2; .n C 1/ /, n D 1, 2,3, . . . ; yn D 2kn cos kn x 3 sin kn x is a kn2 -eigenfunction.
This system has a nontrivial solution if and only if tanh k D

13.2.22.  D 0: x 2 y 00 2xy 0 C 2y D 0 is an Euler equation with indicial equation r .r 1/ 2r C 2 D


.r 1/.r 2/ D 0. y D x.c1 Cc2x/; y.1/ D y.2/ D 0 H) c1 Cc2 D c1 C2c2 D 0 H) c1 D c2 D 0,
so  D 0 is not an eigenvalue.
 D k 2 ; k > 0: y D x.c1 cosh k.x 1/ C c2 sinh k.x 1//; y.1/ D 0 H) c1 D 0; y D
c2 x sinh k.x 1/; y.2/ D 0 H) 2c2 sinh k D 0 H) c2 D 0;  is not an eigenvalue.
 D k 2 ; k > 0: y D x.c1 cos k.x 1/Cc2 sin k.x 1//; y.1/ D 0 H) c1 D 0; y D c2 x sin k.x 1/;
y.2/ D 0 with c2 0 if k D n (n a positive integer); n D n2  2 ; yn D x sin n.x 1/ is a kn2 eigenfunction.
13.2.24.  D 0: x 2 y 00 2xy 0 C 2y D 0 is an Euler equation with indicial equation r .r 1/ 2r C 2 D
.r 1/.r 2/ D 0. y D x.c1 C c2 x/; y 0 D c1 C 2c2x; y.1/ D y 0 .2/ D 0 H) c1 C c2 D c1 C 4c2 D
0 H) c1 D c2 D 0, so  D 0 is not an eigenvalue.
 D k 2 ; k > 0: y D x.c1 cosh k.x 1/ C c2 sinh k.x 1//; y.1/ D 0 H) c1 D 0; y D
c2 x sinh k.x 1/; y 0 D c2 .sinh k.x 1/Ckx cosh k.x 1//; y 0 .2/ D 0 H) c2 .sinh k Ck cosh k/ H)
c2 D 0;  is not an eigenvalue.
 D k 2 ; k > 0: y D x.c1 cos k.x 1/Cc2 sin k.x 1//; y.1/ D 0 H) c1 D 0; y D c2 x sin k.x 1/;
0
y D c2.sin k.x 1/ C kx cos k.x 1//; y 0 .2/ D 0 with c2 0 if and only if sin k C 2k cos k D 0
or, equivalently, tan k D 2k. Graphing both sides of this equation on the same axes shows that it has a
solution kn in ..2n 1/=2; n /, n D 1, 2,3, . . . ; yn D x sin kn .x 1/ is a kn2 -eigenfunction.
13.2.26.  D 0: y D c1 C c2x. The boundary conditions require that c1 C c2 D 0 and c1 C . C /c2 D
0, so c1 D c2 D 0. Therefore  D 0 is not an eigenvalue of (A).
 D k 2 , k > 0: y D c1 cosh kx C c2 sinh kx; y 0 D k.c1 sinh kx C c2 cosh kx/. The boundary
conditions require that
c1 C kc2 D 0
.cosh k C k sinh k /c1 C .sinh k C k cosh k /c2 D 0:

(D)

This system has a nontrivial solution if and only if .1 k 2 2/ sinh k D 0, which holds with k > 0 if and
only if k 2 D 1=. Therefore  D 1= 2 is the only negative eigenvalue. We can choose k D 1=.
Either way, the first equation in (D) implies that e x= is an associated eigenfunction.
 D k 2 , k > 0: y D c1 cos kx C c2 sin kx; y 0 D k. c1 sin kx C c2 cos kx/. The boundary conditions
require that
c1 C kc2 D 0
(E)
.cos k k sin k /c1 C .sin k C k cos k /c2 D 0:
This system has a nontrivial solution if and only if .1 C k 2 2 / sin k D 0. Choosing k D n produces
eigenvalues n D n2  2. Setting k D n in the first equation in (E) yields c1 C nc2 D 0, so yn D
n cos nx sin nx.

Section 13.2 Sturm-Liouville Problems

283

13.2.28. y D c1 C c2 x. The boundary conditions require that


c1 C c2 D 0 and c1 C .L C /c2 D 0.
This system has a nontrivial solution if and only if .L C /  D 0.
13.2.30. (a) y D c1 cos kx C c2 sin kx; y 0 D k. c1 sin kx C c2 cos kx/. The boundary conditions
require that
c1 C kc2 D 0 and . cos kL k sin kL/c1 C . sin kL C k cos kL/c2 D 0.
This system has a nontrivial solution if and only if its determinant is zero. This implies the conclusion.
(b) If  D 0, (A) reduces to
. C k 2 / sin kL D 0:

(B)

From the solution of Exercise 13.2.29(b),  C k 2 > 0 for all k > 0. Therefore the positive zeros of
(B) are kn D n=L, n D 1, 2, 3, . . . , so the positive eigenvalues (SL) are n D n2  2=L2 , n D 1, 2, 3,
....
13.2.32. Suppose  is an eigenvalue and y is an associated eigenfunction. From the solution of Exercise 13.2.31,


b
2

r .x/y .x/ dx D p.a/y.a/y .a/

p.b/y.b/y .b/ C

p.x/.y 0 .x//2 dx:

If D 0 then either y.a/ D 0 or y 0 .a/ D 0, so y.a/y 0 .a/ D 0. If < 0 then y.a/ D


y.a/y 0 .a/ D

(A)

0
.y .a//2 :

0
y .a/, so

(B)

Moreover, y 0 .a/ 0 because if y 0 .a/ D 0 then y.a/ D 0, from (B), and y  0, a contradiction. Since

> 0 if < 0, we conclude that if  0, then

p.a/y.a/y 0 .a/  0;

(C)

with equality if and only if  D 0. A similar argument shows that if   0, then


p.b/y.b/y 0 .b/  0;

(D)

with equality if and only if D 0. Since ./2 C ./2 > 0, the inequality must hold in at least one of
(C) and (D). Now (A) implies that  > 0.