Ecuaciones Diferenciales - D. Zill - 7° Ed. (Solucionario)

COMPLETE SOLUTIONS
MANUAL FOR
ZILL'S
A FIRST COURSE IN
DIFFERENTIAL
EQUATIONS
WITH MODELlNG ApPLlCATIONS
7TH EOITION
ANO
ZILL& CULLEN'S
DIFFERENTIAL
EQUATIONS
WITH BOUNDARyVALUE PROBLEMS
5TH EOITION
BROOKS/COLE
THOMSON
LEARNING
Australia Canada Mexico Singapore Spain United Kingdom United States
Table of Contents
1
Introduction to Differential Equations
First-Order Differential Equations
22
Modeling with First-Order Differential Equations
71
Higher-Order Differential Equations
104
Modeling with Higher-Order Differential Equations
194
Series Solutions of Linear Equations
240
The Laplace Transform
308
Systems of Linear First-Order Differential Equations
370
Numerical Solutions of Ordinary Differential Equations
430
10
Plane Autonomous Systems and Stability
458
11
Orthogonal Functions and Pourier Series
491
12
Partial Dfferential Equations and Boundary-Value

Problems in Rectangular Coordinates
538
13
Boundary-Value Problems in Other Coordinate Systems
616
14
Integral Transform Method
654
15
Numerical Solutions of Partial Differential Equations
695
Gamma function
717
Introduction to Matrices
718
Appendix 1
Appendix II
Introduction to Differential Equations

Exercises 1.1
1. Second-order ; linear.
2. Third-order; nonlinear because of (dy/dx)4
+ y2 = 1, we see that it is
1) (dx / dy) + x = O, we see that
3. The differential equation is first-order. Writing it in the form x(dy/dx)
nonlinear in y because of y2 However, writing it in the form (y2 it is linear in x.
4. The differential equation is first-order. Writing it in the form u(dv/du)
+ (1 + u)v =
that it is linear in v. However, writing it in the form (v + uv - ue'U) (du/dv)
ue'U we see
+ u = O, we see that it
is nonlinear in u.
5. Fourth-order; linear
6. Second-order; nonlinear because of cos(r
7. Second-order; nonlinear because of
+ u)
JI + (dy/dx)2
8. Second-order; nonlinear because of 1/ R2

9. Third-order; linear
10. Second-order: nonlinear because of j;2

11. From y
= e=t? we obtain yl = _~e-x/2.
12. Frorn y
= ~ - ~e-20t we obtain dy/dt

dy
&
13. From y
= e3x cos 2x we obtain yl
that y" - 6y'

14. From y
+ 20y =
=-
y" = tanx
+ 13y
+y
= _Cx/2
+ e-x/2
= O'.
= 24e-20t, so that
24e-20t
+ 20 (~
- ~e-20t)
= 24.
= 3e3x cos 2x - 2e3x sin 2x and y" = 5e3T cos 2x - 12e3T sin 2x, so
= O.
+ tan x)
ln(secx + tanx).
cos x ln(sec x
+ cosx
Then 2y'
w.e obtain yl = -1 + sin x ln(sec x
Then y"
+ Y = tanx.
1
+ tan x)
and
Exercises
1. 1
15. Writing in(2X - 1) - In(X - 1)
=t
nd differentiating implicitly .
we obtain
2
dX
2X - 1 dt
1 dX
X - 1 dt
---------=1
( 2X 2-
1)
dX
dt
1- X - 1
---~4----~2--~\--~2--~4-t
=1
2X - 2 - 2X + 1 dX
-------=1
(2X - l)(X - 1) dt
dX
dt
-2
-4
= -(2X
- l)(X
- 1) = (X - 1)(1 - 2X).
Exponentiating both sides of the implicit solution we obtain

2X - 1
---
X-1
Solving et
= et
==:> 2X - 1
O we get t
= Xet
et ==:> (et
- 1)
= (et - 2)X
==:> X
et - 1
-t---.
e-2
or on (ln2, (0).
In2. Thus, the solution is defined on (-oo,ln2)
The graph of the solution defined on (-00, in 2) is dashed, and the graph of the solution defined on
(ln 2, (0) is solid.
16. Implicitly differentiating the solution we obtain
2 dy
- 2x -- - 4xy
dx
+ 2y --dy
dx
= O==:>-
x dy - 2xy dx
==:> 2xy dx
+ (x2
+ y dy
=O
- y)dy = O.
Using the quadratic formula to solve y2 - 2x2y - 1 = Ofor y, we get
V 4x4 + 4) /2 = x2 J.X4+l. Thus, two explicit

are YI = x2 + vx4 + 1 and Y2 = x2 - VX4 + l. Both
-2
solutions are defined on (-00,00). The graph of YI (x) is solid and
-4
y = (2x2
solutions
the graph of Y2 is dashed.

17. Differentiating P
= el et /
(1 + el et) we obtain
dP
(1 + elet) elet
dt
= ~----~----,,-----
elet. elet
(1 + elet)2
Exercses
Substituting into the differential equation, we have
2 {X
2
y' + 2xy = 1 - 2xe-X Jo et dt - 2clxe-x
19. Fromy
= cle 2x+c2xe
weobtain -dy
dx
+ 2xe-X
?
= (2cl+c2)e-x+2c2xe-x
(X 2
Jo el dt + 2clxe-x
d2y
and -0
dx-
1. 1
= l.
= (4Cl+4C2) e 2x +4C2xe 2x ,
so that
20. Frorn y
= qx-l
+ C2x + C3Xlnx + 4x2 we obtain
-dy = -C1X-
dx
+ c2 + c3 + c31n x +
8x,
d2y
-3
-1
dx2 = 2CIX + C3X + 8,
and
so that
3 d3y
x dx3
+ 2x
d2y
dy
dx2 - x dx
+y
= (-6q + 4Cl + Cl + cl)x-l
+ (-C3 + c3)xlnx
+ (-C3 + 2C3 - c2 - c3 + C2)X

+ (16 - 8 + 4)x2
21. (a) From rPl = x2 we obtain rP'l = 2x, so
xrP~ - 2rPl = x(2x) - 2x2 = O.

From rP2 = _x2 we obtain rP2 = -2x,
so
XrP2- 2rP2 = x(-2x)
- 2(-x2)
= O.
Thus, rPl and rP2are solutions of the differential equation on (-00, (0).
(b) Frorn y
-x2
2'
{ z,
22. The function y(x)
does not exist at x
x <O
{-2X.
we obtain y' =
'
x ;:: O
2x,
is not continuous at x
= Osince x~olim
x <O
so that xy' - 2y = O.
x ;:: O
y(x)
= 5 and
lim y(x)
x~o+
= O.
23. (a) The domain of the function, found by solving x +
3;:: O, is [-3, (0).
= -5. Thus, y'(x)
Exercises
(b) From
1. 1
= 1 + (x
+ 3)-1/2 we have .
(y - x)y' - y + X - 2 = [x + 2VX+3 - x][l + (1 + (x - 3)-1/2]
y!
- [X + 2Vx+3] + x - 2
= 2vX+3 + 2 - x - 2Vx+3 + x - 2 = O.
Since y(x) is not differentiable at x
-3, y is a solution of the differential equation on (-3,00).
< 5t < K/2, so 5 tan 5t will be a solution on
24. (a) An interval on which tan 5t is continuous is -n)2

(-K/10,K/1O).
(b) For (1- sin t)-l/2 to be continuous we must have 1- sin t > Oor sin t < 1. Thus, (1- sin t)-l/2
will be a solution on (K/2, 5K/2).
25. (a) From y
= memt.
emt we obtain y'
Then y'
+ 2y =
O implies
+ 2emt = (m + 2)emt = O.
memt
Since emt > O for all t, m = -2. Thus y = e-2t is a solution.
= me'?' and y" = m2emt Then y" - 5y' + 6y = Oimplies
(b ) From y = emt we obtain y'
m2emt _ 5memt
+ 6emt = (m -
2)(m - 3)emt = O.
Since emt > O for all t, m = 2 and m = 3. Thus y = e2t and y = e3t are solutions.
26. (a) From y
= t'" we obtain y' = mtm-1 and y" = m(m - 1)tm-2 Then ty" + 2y' = Oimplies
tm(m - 1)tm-2 + 2mtm-1 = [m(m - 1) + 2m]tm-l = (m2 + m)tm-1
= m(m + 1)tm-1
Since tm-1 > O for t > O, m

(b) From y = t'" we obtain y'
= O and
-1. Thus y
= mtm-1 and y"
= m(m
= O.
1 and 7J =
- 1)tm-2
el
are solutions.
Then t2y" - 7t7J'
+ 15y =
implies
t2m(m
+ 15tm =
- 1)tm-2 - 7tmtm-1
[m(m - 1) - 7m
= (m2
- 8m
+ 15]tm
+ 15)tm
= (m - 3)(m - 5)tm
Since t'" > O for t > O, m = 3 and m = 5. Thus y = t3 and y = t5 are solutions.
27. From x = e-2t + 3t and y = _e-2t + 5e6t we obtain
dx
dt
_2e-2t
+ 18e6t
and
dy = 2e-2t
dt
+ 30e6t
Then
x
+ 3y = (e-2t + 3e6t) + 3( _e-2t + 5t)

= _2e-2t
+ 18e6t =
and
dx
dt
= O.
Exercises
+ 3y
5x
28. From x
+ 3e6t) + 3( _e-2t + 5e6t)
= 5(e-2t
=
1. 1
+ 30e6t
2e-2t
= dy .
dt
= cos 2t + sin 2t + ~et and y = - cos 2t - sin 2t - ~et we obtain

dx
-d = - 2 sin 2t
and
d2x
dt
-2
1t
+ 2 cos 2t + - e
-4cos2t
1
5
+ _et
- 4sin2t
dy.
1t
and
- = 2sm2t - 2cos2t --e

dt
5
and
-2
d2y
= 4 cos 2t
dt
+ 4 Sll12t - - el .
5
Then
4y
+ et = 4(- cos 2t - sin 2t - ~ et) + et

=
-4 cos 2t - 4 sin 2t
+ Se
d2x
dt2
and
4x - et
= 4(cos 2t
+ sin 2t + ~et) - et
d2y
dt2 .
1 t
= 4 cos 2t + 4 sin 2t - Se
29. (y')2
+ 1 = O has no real solution.
30. The only solution of (y')2

31. The first derivative of
+ y2 = Ois y = O, since if y -1 O, y2 > O and (y')2 + y2 ~ y2 > '0.
f(t)
equations are y' = y and y'

32. Any function of the form y
= et is el. The first derivative of

= kV,
differential equation is y" - Y
cet or y
f(t)
t ,is ki. The differential
respectively.
ce:' is its own second derivative.
O. Functions of the form y
c sin t or y
derivatives that are the negatives of themselves. The differential equation is y"
The corresponding
c cos t have second
+ Y = O.
33. Since the nth derivative of 4>(x) must exist if 4>(x) is a solution of the nth order differential equation,
alllower-order derivatives of 4>(x) must exist and be continuous. [Recall that a differentiable function
is continuous.]
34. Solving the system
for
and
C2
Cy(O)
+ C2Y2(0) = 2
Cy~ (O)
+ C2Y2(0)
=O
we get
and
C2
2y~ (O)
= - ---_,.:.~-'-::--_-y (0)Y2(0) -
yi (0)Y2(0)
Exercises
1. 1
Thus, a particular solution is
where we assume that YI(O)Y2(O) - y~(O)'Y2(O) =1- O.

35. For the first-order differential equation integrate f(x).
For the second-order differential equation
+ clt + C2
integrate twice. In the latter case we get y = f(J f(t)dt)dt
36. Solving for yl using the quadratic formula we obtain the two differential equations
so the differential equation cannot be put in the form dy/dt

37. The differential equation yyl-ty
= O has normal form dy/dt
= f(t,y).
= t. These are not equivalent because
y = O is a solution of the first differential equation but not a solution of the second.
38. Differentiating we get yl = CI
+ 3C2t2
Y -
and y" = 6c2t. Then

ty" )
(yll )
2
t + 6t
and the differential equation is t2y" - 3tyl
+ 3y
C2
= y" /6t and
= ty I -
el
= yl - ty" /2, so
1 2 11
3't y
= O.
39. When g(t) = O, y = O is a solution of a linear equation.

40. (a) Solving (10 - 5y)/3x
= O we see that y = 2 is a constant solution.
(b) Solving y2 + 2y - 3 = (y + 3)(y - 1) = O we see that y = -3 and y = 1 are constant solutions.

(e) Since l/(y
- 1) = O has no solutions, the differential equation has no constant solutions.
(d) Setting yl = O we have V" = O and 6y = 10. Thus V = 5/3 is a constant solution.
41. One solution is given by the upper portion of the graph with domain approximately (0,2.6). The
other solution is given by the lower portion of the graph, also with dornain approximately (0,2.6).
42. One solution, with dornain approximately (-00,1.6)
is the portion of the graph in the second
quadrant together with the lower part of the graph in the first quadrant.
A second solution, with
domain approximately (0,1.6) is the upper part of the graph in the first quadrant.
solution, with domain (0,00), is the part of the graph in the fourth quadrant.
The third
Exercises
+ y3)/xy = 3c w
xy(3x2 + 3y2y') _
43. Differentiating (x3
:;J'?J
3x3
obtain
t;
1. 1
x3
+ y3)(xy' + y)
=O
+ 3xy3y' -9'-x3_y_---:xy3y'
1:;:;,y3 _ x4
_ y4
xy3)y'
=O
= -3x3y
+ x3y + y4
44. A tangent line will be vertical whor y' is undefined, or in this case, where x(2y3 - x3) = O. This
gives x
= O and 2y3 = x3 Subst.itut.ing y3 = x3/2 into x3 + y3 = 3xy we get

x3
+ ~x3
= 3x
C;/3
3 3
3
-x = --x
2
21/3
x)
x3 = 22/3x2
Thus, there are vertical tangent Iines at x = O and x = 22/3, or at (O,O) and (22/3,21/3) _ Since
22/3 :::::1.59, the estimates of the dornains in Problem 42 were close,
45. Since q;'(X) > O for all x in J, q;(x) is an increasing function on J. Hence, it can have no relative
extrema on J.
46. (a) When y
(b) When y
= 5, y' = O, so y = 5 is a solution of y' = 5 - y.

> 5, y' < O, and thc solution must be decreasing. When y < 5, y' > O, and the
solution must be increasing. Thus, none of the curves in color can be solutions.
(c)
-5
47. (a) y = Oand y = albo

(b) Since dy/dx
dy/dx
y(a - by)
> O [01' O < y < afb, y = q;(x) is increasing on this interval. Since
< Ofor y < Oor y> albo
.11
= q;(x)
is decreasing on these intervals.
Exercises 1. 1
(c) Using implicit differentiation we compute
~:; = y( -by')
+ y'(a
- by) = y'(a - 2by).
= Owe obtain y = a/2b. Since d2y / dx2 > Ofor O < y < a/2b and d2y / dx2 < O
< y < a/b, the graph of y = cp(x) has a point of infiection at y = a/2b.
Solving d2y / dx2

for a/2b
(d)
48. The family of parabolas is plotted using

ellipses are plotted using
C2
Cl
1, 4, lO. The
= 1, 2, 3. It appears from the figure
that the parabolas and ellipses intersect at right angles. To verify

this we note that the first differential equation can be written in the
form dy/dx = y/2x and the second in the form dy/dx = -2x/y.
Thus, at a point of intersection, the slopes of tangent lines are
negative reciprocals of each other, and the two tangent lines are
perpendicular.
49. In Mathematica
use
Clear[y]
y[x_]:= x Exp[5x] Cos[2x]
y[x]
Y"" [x] - 20 Y'" [x]
50. In Mathematica
158 Y" [x] - 580 Y' [x] + 841 y[x] / / Simplify
use
Clear[y]
y[x_]:= 20 Cos[5 Log[x]]/x - 3 Sin[5 Log[x]]/x
y[x]
x-3 Y'" [x]
2x-2 Y" [x]
20 x Y'[x] - 78 y[x] / / Simplify
Exercses
1.2
Exercises 1.2
1
1. Solving --3 =
--
1 + Cl
_
1
we get Cl = -4. The solution is y =
1 - 4e-t
.11
2. Solvmg 2 = ---
we get q = __e-l. The solution is y = 2 _ e-(t+l) .

1 + cj e
2
3. Using x' = -Cl sint + C2cost we obtain Cl = -1 and C2= 8. The solution is x
4. Using
Xl
5. Using x'
= -Cl sin t + C2cos t we obtain C2= O and -Cl = l.
= -cost + 8sint.
The solution is x = - cos t.
= -Cl sin t + C2cos t we obtain
J3 + -C2
1
1
= -2
2
-Cl
2
1
--Cl
2
.
S o1vmg
we fin d Cl =
6. Using x'
J3
4
J3 = O.
+ -C2
2
an d C2= 4:1 . Th e soluti

ution
IS
x =
J3 cos t + 4:l. sm t .
4
-Cl sin t + C2cos t we obtain
V2
-Cl
2
V2
= V2
+-C2
2
V2 Cl + -V2 C2= 2V2 .
- -
Solving we find Cl = -1 and C2= 3. The solution is x = - cos t + 3 sin t.

7. Frorn the initial conditions we obtain the systern
Cl + C2= 1
Cl - C2= 2.
Solving we get Cl =
~ and C2= -
~. A solution of the initial-value problem is y
~ex - ~e-x.
8. Frorn the initial conditions we obtain the systern

cj e + C2e-l = O
cj e - C2e-l
Solving we get el
= ~ and
C2=
- ~e2
= e.
A solution of the initial-value problern is y = ~ex
- ~e2-x.
9. From the initial conditions we obt.ain
Cle-l -
C2e =
-5.
Solving we get Cl = O and C2= 5e-l. A solution of the initial-value
problern is y
= 5e-x-l.
Exercises
1.2
10. From the initial conditions we obtain
Cl -
Solving we get
Cl
= C2 = O. A solution
11. Two solutions are y

12. Two solutions are y
= Oand y = x3
= Oand y = x2
C2
= O.
of the initial-value problem is y
= O.
(AIso, any constant multiple of x2 is
EL
solution.)
13. Por f(x, y) = y2/3 we have ~f = ~y-l/3 Thus the differential equation will have a unique solution
uy
3
in any rectangular region of the plane where y
i= o.
14. Por f(x, y) = .xy we have ~f = ~ ~. Thus the differential equation will have a unique solution
2y y
uy
in any region where x> Oand y > Oor where x < O and y < O.
15. Por f(x, y) = Ji. we have ~f = ~. Thus the differential equation will have a unique solution in any
x
uy
x
region where x i= o.
af
16. Por f(x, y) = x + y we have ay = l. Thus the differential equation will have a unique solution in
the entire planeo
17. Por f(x, y)
2 2
af
X Y '). Thus the differential equation will have a unique
we have -;:;- =
uy
(4 - y2
x2
--2
4- y
solution in any region where y < -2, -2 < y < 2, or y > 2.

18. Por f(x, y)
x2
af
_3x2y2
we have -;:;- =
2' Thus the differential equation will have a unique
uy
(1 + y3)
--3
1+y
solution in any region where y
i=
-l.
y2
af
2x2y
we have -;:;- =
2' Thus the differental equation will have a unique
x +y
uy
(x2 + y2)
solution in any region not containing (O,O).
19. Por f(x,y)
20. Por f(x, y)
= --
y +x
8f
-2x
we have -;:;- = (
)2'
y-x
uy
y-x
Thus the differential equation will have a unique
solution in any region where y < x or where y > X.

21. The differential equation has a unique solution at (1,4).
22. The differential equation is not guaranteed to have a unique solution at (5,3).
23. The differential equation is not guaranteed to have a unique solution at (2, -3).
10
Exercises
1.2
24. The differential equation is not guaranteed to have a unique solution at (-1,1).
25. (a) A one-parameter family of solutions is y = ex. Since y' = e, xy' = xe = y and y(O) = e 0= O.
(b) Writing the equation in the form y' = y/x we see that R cannot contain any point on the y-axis.
Thus, any rectangular region disjoint from the y-axis and containing (xo, Yo) will determine an
interval around Xo and a unique solution through (xo, yo). Since zo = O in part (a) we are not
guaranteed a unique solution through (O,O).
(e) The piecewise-defined function which satisfies y(O)
= O is not a solution since it is not differ-
entiable at x = O.
d
26. (a) Since dx tan(x
+ e) =
sec2(x
+ e)
= 1 + tan2(x
+ e),
we see that y = tan(x
+ e) satisfies
the
differential equation.
(b) Solving y(O) = tan e = O we obtain e = O and y = tan x.
Since tan x is discontinuous at
x = 7l' /2, the solution is not defined on (-2,2) because it contains 7l' /2.
(e) The largest interval on which the solution can exist is (-7l' /2, 7l'/2).
27. (a) Since
!!_ ( __
dt
equation.
1_)
+e
( t+e )
(b) Solving y(O) = -l/e
t+c
is a solution of the differential
= 1 we obtain e = -1 and y = 1/(1 - t). Solving y(O) = -l/e
we obtain e = 1 and y = -1/(1

of existence of y
= Y , we see that y =
+ t).
= -1
Being sure to inelude t = O, we see that the interval
= 1/(1- t) is (-00,1), while the interval of existence of y = -1/(1 + t) is
(-1,00).
(e) Solving y(O)
= -l/e
= Yo we obtain e = -l/yo
1
Y = - -l/yo
+t =
and
Yo
1 _ yo t '
yO =1 O.
Since we must have -l/yo
+ t =1 O, the
either (-00, l/yo)
> Oor (l/yo, 00) when yo < O.
when yo
(d) By inspection we see that y

28. (a) Differentiating 3x2 - y2
= Ois a
largest interval of existence (which must contain O) is
solution on (-00,00).
= e we get 6x - 2yy' = Oor yy' = 3x.

11
Exercises
1.2
(b) Solving 3x2 - y2
= 3 for y
Y = <P2(X) = -J3(x2
= <P4(X) = -J3(x2
< 00,
-oo<x<-l,
-4
- 1) ,
Only y = <P3(X) satisfies y( -2)
= 2 and y
< 00,
1< x
- 1) ,
- 1) ,
Y = <P3(X) = J3(x2
(e) Setting X
1< x
-1) ,
y = <Pl(X) = J3(x2
we get
-00
< X < -1.
= 3.
= -4 in 3x2 _y2
= e we get 12-16 = -4 = e,
so the explicit solution is

y = -J3x2
+4,
-00
< X < oo.
-4
-2
-4
(d) Setting e = O we have y = V3x and y

29. When x = O and y = ~, y'
-V3x, both
defined on (-00,00).
-1, so the only plausible solution curve is the one with negative slope
at (O, ~ ), or the black curve.

30. The value ofy'
x
= dyjdx
f(x,y)
is determined by f(x,y),
so it cannot be arbitrarily specified at
= xo
31. If the solution is tangent to the x-axis at (xo, O), then y'
these values into y'
+ 2y
=O
when x
= Xo
and y
= O.
Substituting
= 3x - 6 we get O + 0= 3xo - 6 or Xo = 2.
32. We look for the curve containing the point corresponding to the initial conditions.
(a) D
(b) A
(e)
(d)
(e) B
(f) A
33. The theorem guarantees a unique (meaning single) solution through any point. Thus, there cannot
be two distinct solutions through any point.
34. The functions are the same on the interval (0,4), which is all that is required by Theorem 1.1.
12
Exercises
35. Frorn y'
=x-
O, for y
1.3
x - 1, we see that solutions intersecting the line
1 have horizontal tangent lines at the point of intersection. From
y' > O, for y > x - 1, we see that solutions aboye the line y
are increasing. From y' < O, for y < x -
1,
=x - 1
we see that solutions are
decreasing below the line y = x - 1. From y" = O, for y
x, we see
that solutions have possible infiection points on the line y = x. Actually,

y = x is easily seen to be a solution of the differential equation, so the
solutions do not have infiection points. From y" > O for y > x we see
that solutions aboye the line y = x are concave up. From y" < O for y < x we see that solutions
below the line y = x are concave down.
Exercises 1.3
dP
dP
= kP+r'
- = kP-r
dt
'
dt
2. Let b be the rate ofbirths and d the rate of deaths. Then b = k1P and d = k2P. Since d.Pf dt = b-d,
1. -
the differential equation is dP/dt = klP - k2P.

3. Let b be the rate of births and d the rate of deaths. Then b = klP and d
the differential equation is dP / dt
= k2P2. Since dP/dt = b=d,
= kl P - k2P2.
4. Let P(t) be the number of owls present at time t. Then dP/dt
= k(P - 200 + 10t).
5. From the graph we estimate To = 180 and Tm = 75. We observe that when T = 85, dT/dt ~ -1.
Frorn the differential equation we then have
k = dT/dt
T-Tm
-1
85-75
= -0.1.
6. By inspecting the graph we take Tm to be Tm(t) = 80 - 30cos7ft/12.
Then the temperature of the
body at time t is determined by the differential equation
~ =
k [T -
(80 - 30 cos ;2
t) ], t > O.
7. The number of students with the flu is x and the number not infected is 1000 - x, so dxf dt. =
kx(1000 - x).
8. By analogy with differential equation modeling the spread of a disease we assurne that the rate at
which the technological innovation is adopted is proportional to the number of people who have
adopted the innovation and also to the number of people, y( t), who have not yet adopted it. If one
person who has adopted the innovation is introduced into the population then x
dx
dt = kx(n + 1 - x), x(O) = 1.
13.
+y
=n
+ 1 and
Exercises
1.3
9. The rate at which salt is leaving the tank is

(3 gal/min) .
C~O
lb/gal) = 1~ lb/mino
Thus dA/dt = A/lOO.

10. The rate at which salt is entering the tank is
Rl = (3 gal/rnin) . (2 lb/gal) = 6 lb/mino
Since the solution is pumped out at a slower rate, it is accumulating at the rate of (3 - 2)gal/min
1 gal/min. After t minutes there are 300 + t gallons of brine in the tank. The rate at which salt is
leaving is
R2
= (2 gal/min) ( -- A) lb/gal
300 + t
2A
lb/mino
300 + t
The differential equation is

dA = 6 _ 2A
dt
300 + t
11. The volume of water in the tank at time t is V
dh
dt
Using.Ao
_1 dV
Aw dt
= Awh.
_1 (-cAofiih)
The differential equation is then
= _ cAo fiih.
Aw
Aw
= 7r 22)2 = ~ , Aw = 102 = 100, and g = 32, this becomes

dh
dt
= _ C7r/36 J64h = _~ v'h.

100
12. The volume of water in the tank at time t is V
450
~7rr2h = ~Awh. Using the formula from
Problem 11 for the volume of water leaving the tank we see that the differential equation is
dh
dt
Using Ah
= _2_ dV = _2_ ( -CAh fiih)

Aw dt
= 7r(2/12)2 = 7r/36,
= - 3cAh
Aw
Aw
fiih.
= 32, and e = 0.6, this becomes
dh = _ 3(0.6)7r/36
dt
Aw
J64h = _ 0.47rh 1/2.

Aw
To find Aw we let r be the radius of the top of the water. Then r / h

Aw
= 7r(2h/5f
= 47rh2/25. Thus
dh = _ 0.47r h1/2 = -2.5h-3/2

dt
47rh2/25
dq
d2q
13. Since i = dt and L dt2
dq
+ R dt
= E(t) we obtain
di
L dt + Ri = E(t).
14. By Kirchoff's second law we obtain Rdq + el q = E(t).

dt
14
8/20, so r
2h/5 and
Exercises
dv
= -kv
dt
15. From Newton's second law we obtain m-
1.3
+ mg.
16. We have from Archimedes' principIe
upward force of water on barrel = weight of water displaced
(62.4) x (volume of water displaced)
= (62.4)7T(S/2)2y = 15.67Ts2y.
w d2y
--2
It then follows from Newtori's second law that

9
9 dt
-15.67Ts 'Y
= 32 and w is the weight of the barrel in pounds.
17. The net force acting on the mass is
d2x
F = ma = m dt2 = -k(s
Since the condition of equilibrium is mg
+ x) + mg
+ mg
= -kx
- ks.
= ks, the differential equation is

d2x
= +ks.
m dt2
18. From Problem 17, without a damping force, the differential equation is md2x/dt2
= -kx.
With a
damping force proportional to velocity the differential equation becomes

d2x
m dt2
dx
d2x
= -kx - (3 dt
or
m dt2
dx
+ (3 dt + kx = O.
19. Let x(t) denote the height of the top of the chain at time t with the positive direction upward. The
weight of the portion of chain off the ground is W = (x ft) . (1 lb/ft) = x. The mass of the chain is
m = W/g = x/32. The net force is F = 5 - W = 5 - x. By Newton's second law,
d x
dt (32 v) = 5 - x
dv
x dt
or
dx
+ v dt
= 160 - 32x.
Thus, the differential equation is

x-d2x
dt
+ (dx)2
+ 32x =
dt
160.
d
20. The force is the weight of the chain, 2L, so by Newton's second law, -d [mv] = 2L. Since the mass
of the portion of chain off the ground is m = 2(L - x)/g, we have
!!_ 2(L
dt
- x) v] = 2L
9
(L-x)-+vdv
dt
or
Thus, the differential equation is

(L - x)-d2x - (dx)
dt2
dt
15
= Lg.
(dX)
-dt
=Lg.
Exercises
21. From g
1.3
kj R 2 we find k
gR~.0 Using a
d 2 r j dt~0 and the fact that the positive direction is
upward we get
or
22. The gravitational force on mis F = -klVIrmjr2.
u, = r3 M j R3
Since NIr
= 41fr3j3 and M
= 41fR3j3 we have
and
F = -k Mrm = -k r3VImjR3 = -k mM r.
r2
Now from F
r2
R3
= ma = d2r j dt2 we have
23. The differential equation is dA = k(M _ A).
dt
dA
24. The differential equation is - = kl(M - A) - k2A.
dt
25. The differential equation is x'(t) = r - kx(t) where k > O.
26. By the Pythagorean Theorem the slope of the tangent line is y'
27. We see from the figure that 28
y
-x
= tan
+ o: = tt,
o: = tan(7T- 28)
-y
-r====
Js2 - y2
Thus
=-
tan 28 =
2 tan 8
1 _ tan2 8 .
Since the slope of the tangent line is y' = tan 8 we have y j x

or y - y(y')2 = 2xy', which is the quadratic equation y(y')2
= 2y' [1- (y')2]

+ 2xy' - y = O
in yl Using the quadratic formula we get

,
Y =
-2x
J4x2
2y
+ 4y2
-x
Jx2
+ y2
.x
Since dy j dx > O, the differential equation is

dy
dx
- x
+ J x2 + y2
or
28. The differential equation is dPjdt
y ~~ - Jx2
+ y2 + X
= O.
= kP, so from Problem 31 in Exercises 1.1, P = ekt, and a
one-parameter family of solutions is P = ct.

29. The differential equation in (3) is dTjdt
= k(T - Tm).
When the body is cooling, T > Tm, so
< O and k < O. When the body is warrning, T < Ti, so

T - T m < O. Since T is increasing, dT j dt > O and k < O.
T - Tm > O. Since T is decreasing, dTjdt
16
Exercses
30. The differential equation in (8) is dA/dt
at this time and A

for t
> O and
1.3
If A(t) attains a maximum, then dA/dt = O
= 6 - A/I00.
= 600. If A(t) continues to increase without reaching a maximum then
A cannot exceed 600. In this case, if A'(t)
A'(t)
> O
approaches O as t increases to infinity, we
see that A(t) approaches 600 as t increases to infinity.

31. The input rate of brine is ri gal/rnin and the concentration of salt in the infiow is e lb/gal, so the
input rate of salt is riCi lb/mino The output rate of brine is ro gal/min and the concentration of
salt in the outfiow is colb/gal, so the output rate of salt is rocolb/min.
The solution in the tank
< ro). After
is accumulating at arate of (ri - ro) gal/rnin (or decreasing if ri
Vo + (ri - r o)t gallons of brine in the tank, and the output rate of salt is r oA/[Vo
t minutes there are
+ (ri -
ro)t]lb/min.
The differential equation is

dA
roA
- = riCi dt
Vo+(ri-ro)t
32. This differential equation could describe a population that undergoes periodic fiuctuations.
dP
33. (1):
di =
(3):
di =
(6):
dX
-dt
(10):
dT
kP
is linear
(2):
k(T - Tm) is linear
(5):
= k(a - X)(f3 - X) is nonlinear
dh
dt
=-
Ah ~
dA
di
dx
dt
= kx(n + 1 - x) is nonlinear
dA
A
=6- is linear
100
(8): -
dt
d2q
Aw V 2gh is nonlinear
(11):
is linear
= kA
L dt2
dq
+ R dt +
Cq
= E(t)
is linear
dv
(14): m- = mg - kv is linear
(12):
dt
d2s
d2x
ds
(15): m dt2 + k dt = mg is linear

34. From Problem 21, d2r / dt2
= -g
(16): -
dt2
64
- -x = O is linear
L
R2 / r2. Since R is a constant, if r
+ s,
then d2r / dt2
= d2s / dt2
and, usng a Taylor series, we get

d2s
R2
2
dt2 = -g (R + s)2 = -gR (R
+ s)
-2
;:::;-gR
[R
-2
- 2sR
-3
+ ... ] = -g +
2gs
R3
+ ....
Thus, for R much larger than s, the differential equation is approximated by d2s/dt2 = -g.
35. If p is the mass density of the raindrop, then m = pV and
dm = p dV
di
dt
If dr l dt is a constant, then dm/dt

decreasing, k
r = kt] p
< O. Solving
dr/dt
= p !!:_ [~1T3l= P (47rr2 dr) =

dt 3
dt
pS dr .
dt
= kS where p dr l dt = k or dr l dt = k/p. Since the radius is

= k/p we get r = (k/p)t + Co. Since r(O) = ro, Co = ro and
+ ro.
17
Exercises 1.3
d
Frorn Newton's second law, -[mv]
dt
dv
m dt
= mg, where
dm
vis the velocity of the raindrop. Then
or
+vdi = mg
Dividing by 4p1fr3/3 we get

dv
3k
dt
pr
-+-v=g
dv
or
3k/ p
+ k t / P + ro v
-d
= g,
< O.
36. We assume that the plow clears snow at a constant rate of k cubic miles per hour. Let t be the
time in hours after noon, x(t) the depth in miles of the snow at time t, and y(t) the distance the
plow has moved in t hours. Then dy/dt is the velocity of the plow and the assumption gives
dy
wx-=k
dt
where w is the width of the plow. Each side of this equation simply represents the volume of snow
plowed in one hour. Now let to be the number of hours before noon when it started snowing and
let s be the constant rate in miles per hour at which x increases. Then for t
> -to, x
= s(t
The differential equation then becomes
dy
dt
-=---
ws t + to
Integrating we obtain
k
y = - [In (t + to) + e]
ws
where e is a constant. Now when t = 0, y = so e = -In to and
y = ~ In (1
ws
..!.)
.
to
Finally, frorn the fact that when t = 1, Y = 2 and when t = 2, Y
(1 + ~)
Expanding and simplifying gives
ta + to -
= 3, we obtain
(1 + t~f
1 = O. Since to
> 0, we find to ~ 0.618 hours ~
37 minutes. Thus it started snowing at about 11:23 in the morning.
18
+ to).
Chapter
1 Review Exercises
Chapter 1 Review Exercises

= ky
dy
dx
d
2. dx (5 + cle-2x)
d
3. -(q eos kx
dx
= -2cle-2x
+ C2sin kx)
d2
-2 (Cleos kx
dx
+ cle-2x
= -2(5
= -kCl sin kx
+ C2sin kx)
dy
dx
- 5)
= -2(y - 5) or -dy = -2y + 10

dx
+ kC2 eos kx
= -k2cl eos kx - k2c2 sin kx = -k2(Cl eos kx
d2y
d2y
= _k2y
or + k2y = O
dx2
dx2
d
4. - (Cleosh kx + C2sinh kx) = kc sinh kx
dx
+ C2sin kx)
+ C2sinh kx)
d22 (Cleosh kx
dx
d2y
2
dx2 = k y
5. y'
or
= k2 Clcosh kx
+ k2 C2sinh kx
= k2 (ci eosh kx
+ C2sinh kx)
d2y
2
dx2 - k y = O
= C1ex + C2xex + C2ex
y"
+ kC2 eosh kx
y"
+ C2xex + 2C2ex;
= 2(C1eX + C2xex + C2eX) = 2y'
= C1ex
+ Y = 2(C1eX + C2xeX) + 2C2ex
+ CleX eos x + C2ex eos x + C2ex sin x;

el eX sin x + q e" eos x = - 2CleX sin x + 2c2ex eos x;
y" - 2y'
+y
=O
6. y' = -C1ex sin x
y" = -Cl eX eos x - C1eX sin x y" - 2y' = -2clex

7. a,d
eos x - 2c2ex sinx = -2y;
8. e
13. A few solutions are y
y" - 2y'
+ 2y
eos x
+ C2ex eos x + C2ex sin x
=O
10. a,e
9. b
= O, y
+ c2ex
C2ex sin x
11. b
12. a,b,d
= e, and y = eX.
14. Easy solutions to see are y = Oand y = 3.

15. The slope of the tangent line at (x, y) is y', so the differential equation is y' = x2
16. The rate at which the slope changes is dy' / dx
y" +y'
+ y2.
y", so the differential equation is y"
-y'
or
= O.
17. (a) The domain is all real numbers.

(b) Sinee y'
= 2/3xl/3, the solution y
= x2/3 is undefined at x
= O. This funetion is a solution of
the differential equation on (-00, O) and also on (0,00).

18. (a) Differentiating y2 - 2y = x2 -
+ Cwe obtain
19
2yy' - 2y' = 2x - 1 or (2y - 2)y' = 2x - 1.
1 Review Exercises
Chapter
(b) Setting x = O andoy = 1 in the solution we have 1 - 2 = O - O + e or e

of the initial-value problem is y2 - 2y = x2 - x - l.
-l. Thus, a solution
= O by the quadratic formula we get y = (2V4 + 4(x2 - x - 1) )/2
(c) Solving y2-2y-(x2-x-1)
-JX2=X = 1 vx(x - 1). Since x(x - 1) ;:: O for x S O or x ;:: 1, we see that neither
=1
= 1+
vx(x - 1) nor y
= 1 - vx(x - 1) is differentiable at x = O. Thus, both functions are
solutions of the differential equation, but neither is a solution of the initial-value problem.
19. (a)
y
3
-3
(b) When y
= x2 + el. y'
= 2x and (y')2
')
(c) Pasting together x-, x ;::O, and -x
= 4x2. When y
{_x
6V4 +
5( -1)3
x<O
2,
2
, x S O, we get y =
20. The slope of the tangent line is y' 1(-1,4)=
21. Differentiating y
= -x2 + e2, y' = -2x and (y')2 = 4x2.
x,
-.
= 7.
= sin(lnx) we obtain y' = cos(lnx)/x and y" = -[sin(lnx) +cos(lnx)1/x2
2 '"
x y + xy + y
= x 2( -
sin(lnx) + cos(lnx))
+x
cos(lnx)
. (1 )
+ sm n x
Then
= O.
22. Differentiating y = cos(lnx) ln(cos(lnx)) + (lnx) sin(lnx) we obtain
= cos
(1)
nx
1
(Sin(lnx))
(1 ) cos nx
x
=- ln(cos(ln x)) sin(ln x)

x
1 (
+ neos
(1
nx
))(
sin(lnx))
1 cos(lnx)
+ nx
x
x
+ sin(lnx)
x
(ln x) cos(ln x)
x
+~--'---'-----'-
and
y
,,[
= -x ln(cos(lnx)) cos(lnx) + sin(lnx)

x
1
(Sin(lnX))]
(1 ) cos nx
x
1
[
(Sin(lnx))
+ ln( cos(ln x)) sin(ln x)"2 + x (In x) x
x
2
1
"2
x
COS(lnX)] 1
?
x
z-
1[
sin (ln x)
.
="2 - ln( cos(ln x) cos(ln x) +
(1 ) + ln(cos(ln x) sm(ln x)
x
cos nx
- (lnx) sin(lnx) + cos(lnx) - (lnx) cos(lnx)].
20
1
(In x) cos(ln x)"2
x
1 Review Exercises
Chapter
Then
2
x 2 y 1/
+ XY / + y =
+ cos(ln x)
+ (In x)
=
+ sin ((ln X)) + ln(cos ())lnx
-In ( cos(lnx))cos(lnx)
sin2(lnx)
cos(ln x)
. ()lnx
sm
cos lnx
- (In x) cos(ln x) - ln( cos(ln x)) sin(ln x)

cos(ln x)
+ cos(ln x)
ln(cos(ln x))
+ (In x)
sin2(lnx) + cos2(ln x)
cos(ln x)
+ cos ()In x =
sin (In x)
1
cos(ln x)
(This problem is easily done using M athematica.)

23. From the graph we see that estimates for YO and Yl are YO = -3 and Yl
24. The differential equation is
dh
dt
Using Ao
= _ cAo
= O.
fi;h.
Aw
= 11"(1/24)2 = 11"/576, Aw = 11"(2)2 = 411", and g = 32, this becomes

dh = _ c11"/576
dt
411"
J64h
V3
= -mg - {J-mg
=~
288
25. From Newton's second law we obtain

dv
mdt
. ( lnx )
- (lnx ) sm
or
21
vh.
= sec (1)n x
Flrst-Order Differential Equations

Exercises 2.1
1.
7.
2.
8.
\ I \ 1\1 \
....._'"'""
\ I
\ I
\ 1\ 1\ I \
\ I \ lit
\
\1 \ 1II \
\ \ ,-""
\ \ , .._'"'"
\\"
..-~,
\ I \
\ \"
I I I \
\ I
\ I
\ I \ I I I \
\ I \ I I I \
\ I
\I \ I I I \
.........
......... ~
~
.........
.........
22
.-.........
\ I
__
~,
~,
__
<lo ..
"
,
\ 1\1 \ .....- ...

\ 1\1 \ \ , ...
_'"'
\ t \ I \ \ \ \ ..-
,~
,~
,~
,~
......
-_
__
......
Exercises 2. 1
9.
10.
y
II I 11I 11I I
I I I
1111/111'
I
1111/1'
1/.
111"
1'1'
I
y
111
11
11
111
,.,
II
II1
II1
III
II
II
II
III
~~~~~.f-+~~~~'~'
1X
1""
..... "
I ,
I , 1" I , .._"
I , l' I " .._,
l'
,\ I ",
...
11.
1\ 1, I ",
I I 11I '1
III
11
11
12.
y
II
I l'
"11/111111
"11/111111
I III
1"
J"
11""
13.
1I
"
I .. ,.,."
.,
II
11
11
,,,
'"
,,,
111
1111
1111
III
III
III
1111
111.
11
1111
111I
1111
1II1
'11111
1111"
II
, 1II
111.
1111'
1111'
1111
111I
.. ,,"
"
"
"
1\ 1 1 1\1
'1"
I
I
1 ........
,"
_
_
-----
III
III
III
1'11
I1II
11II
I l' , , l'
1111'111'
I1I1
1111'
11II
111"
'"
1111 111
" I
"
I
11I
,"
"
""""'\
""
'"
.., -1
.. \ - 1
..\-1
..\-1
11
I
I
I I
I
II
," ,
"11
11- \-
11
...
11
11- \11_ \_
\J
11- \11 ... ' _
II
II
,,-
"III I
1-'-
I I
11- \-
111111
11- \-
11-,
, , ,
I
I
I
"
II
1 1-\-
11
= -1, y = O,and y =
1'"
1""
y
_1
I1 - \ - I
I 1- \-1
11-' -.
1 1-'
11-'
11-'
"
15. Writing the differential equation in the form dyJdx = y(llocated at y
I
I
, I
" "'\\'\,
"1
""
", I "'1 ""
-- .. _ .. ""
, I r r r 1"
II 11 I 1'"
I
11 I
I 11 ,
II11 I 1""
1 .........
---------, " " '"

11"
I
I
I
I
"""""
.. ........
"""""
X
---------,\ ""
,,,
"
,,,
1" , , 1" , , ,
,"
11""
11""
\,,,
\ "
.........
_- ......
\ \
--- __ "".. "
----....."
'"'_.._ ...
,'"
14.
y
1111
"
.. I
II
II
II
....
I "1
, I111
I ",,
'11111/111
'1//1
I I11I
l'
11
II
II
y
, 1'"
y)(l
+ y) we see that critical points are
1. The phase portrait is shown below.

-1
y
(a)
(b)
(c)
(d)
23
-+------------~------------~2x
Exercises 2. 1
16. Writing the differential equation in the form dyjdx
are located at y = -1, Y = 0, and y
= y2(1 - y)(l
+ y)
we see that critical points
1. The phase portrait is shown below.
-1
(a)
(b)
iL
-2
-1
2 x
(d)
(c)
-2
-3
-.
-5
17. Solving y2 - 3y = y(y - 3) = Owe obtain the critical points O and 3.
From the phase portrait we see that Ois asymptotically stable and 3 is unstable.
18. Solving y2 - y3
= y2(1 - y)
we obtain the critical points Oand 1.
o
From the phase portrait we see that 1 is asymptotically stable and Ois semi-stable.
19. Solving (y - 2)4
= O we obtain the critical point
2.
From the phase portrait we see that 2 is semi-stable.

20. Solving 10 + 3y - y2 = (5 - y)(2
+ y) = Owe obtain
-2
the critical points -2 and 5.
From the phase portrait we see that 5 is asymptotically stable and -2 is unstable.
21. Solving y2(4 - y2)
= y2(2 - y)(2 + y)
-2
= Owe
obtain the critical points -2, O, and 2.
24
Exercses 2. 1
From the phase portrait we see that 2 is asymptotically stable, Ois semi-stable, and -2 is unstable.
22. Solving y(2 - y)(4 - y) = Owe obtain the critical points 0,2, and 4.
From the phase portrait we see that 2 is asymptotically stable and O and 4 are unstab!e.
23. Solving y ln(y
+ 2) = Owe obtain
-2
the critical points -1 and O.

o
-1
From the phase portrait we see that -1 is asymptotically stable and O is unstab!e.
24. Solving yeY
9y = y(eY
9) = Owe obtain the critical points O and In 9.

o
In 9
From the phase portrait we see that O is asymptotically stable and In 9 is unstable.
25. (a) Writing the differential equation in the form
dvdt = mk (mkg _ v)
we see that a critical point is mg j k.
rng/k
From the phase portrait we see that mg j k is an asymptotical!y stable critica! point. Thus,
limt->oov
= mg j
k.
(b) Writing the differential equation in the form
dvdt = 'm~ (mkg _ v2) = mk ( VT

mg _ v) ( fi'i'Lg+ v)
VT
we see that a critical point is
Jmgjk.
--J rng/k
J mg j k is an asymptotically
From the phase portrait we see that

limt~oo v
stable critical point. Thus,
= Jmgjk.
26. (a) From the phase portrait we see that critica! points are a and (3. Lct X(O)
= Xo.
If Xo
< a, we see that X
-t
a as t
-t
oo. If a < Xo < (3, we see that X
25
-r+
a as t
-r+
oo.
Exercses 2. 1
If Xo
> (3, we see that X(t) increases in an unbounded manner, but more specific behavior of
X (t) as t
---+ 00
is not known.
(b) When a = (3 the phase portrait is as shown.
If Xo
< a, then X(t)
---+
a as t ---+ co. If Xo > a, then X(t) increases in an unbounded manner.
This could happen in a finite amount of time. That is, the phase portrait does not indicate
that X becomes unbounded as t
---+
oo.
(e) When k = 1 and a = (3 the differential equation is dX / dt = (a - X) 2 Separating variables

and integrating we have
dX
(a _ X)2
= dt
--=t+c
a- X
a-X=--
t+c
X=a---.
For X(O)
For X (O)
= a/2
t+c
we obtain
1
X(t)=a-
/
t+2 a
X(t)=a-
/
t- 1a
2a we obtain
1
For Xo > a, X(t) increases without bound up to t

X ---+ a as t
co.
-jo
26
= l/a. For t > l/a, X(t) increases but
Exercises 2. 1
27. Critical points are y = O and y
= c.
y
1111111111
1 r I I 11 I I I I
I 11 J 1I I
I JI
1 11 111
1 1 111 1 11 11
1111
11 1/ 1 11/1
III~IIIII
\""""
\ \ , \ \ \ I \,
""'"
'\\"1'"
\ \ , \ \ \ 1"
"""""
-2.2,
,\
, I \ \ \ I \ \ \
, I \ \ \ I \ \ \
\1 1111111
\1 1111111
1111111111
1111111111
points are y
11
III
II
I 1I II
1"1
I111
/1/11/111/
111111111'
\
\111111\11
\ \ \ 1 \ \ 1 \ \ \
28. Critical
1 111111
\ \ \\ l' \\
""""
1111111111
111
I 11 I 11
III~IIIII
II111
I 11
111111111
11111111
I I
0.5, and
y = 1.7.
0.5
-2.2
1.7
29. At each point on the circle of radius c the lineal elernent has slope c2.
30. (a) When x = Oor y = 4, dy/dx
dy/dx
= -2 so the lineal elernents have slope -2.
= x - 2, so the lineal elernents at (x,3)
When y
= 3 or y
= 5,
or (x,5) have slopes x - 2.
(b) At (O,YO) the solution curve is headed down. As x increases, it will eventually turn around and
head up, but it can never cross y
O. Thus, y cannot approach
31. When y
< ~x2, y'
00
= 4 where a tangent line to a solution curve rnust have slope
as x approaches oo.
= x2 - 2y is positive and the portions of solution
curves "inside" the nullcline parabola are increasing. When y
> ~x2,
y' = x2 - 2y is negative and the portions of the solution curves

"outside" the nullcline parabola are decreasing.
I
I I ,
11/
//11
1111
11 I
I f 1/
II
I
/'.1.
I I I
I , I
I I
I 1
I I
I I I
I I
I I I
J
J
I
JI'
I
III
l'
I
I I
I I
I I
I I
I
I I
-3~~~~~~O~/~/~~~I~I~;~I~I~~~
32. For dxl dt = Oevery real nurnber is a critical point, and hence all critical points are nonisolated.
33. Recall that for dy / dx
f(y)
we are assurning that f and
f'
are continuous functions of y on
sorne interval J. Now suppose that the graph of a nonconstant solution of the differential equation
27
Exercses 2. 1
crosses the line y
= c. If the point of intersection is taken .as an initial condition we have two distinct
solutions of the initial-value problem. This violates uniqueness, so the graph of any nonconstant
f is continuous it can only
solution must lie entirely on one side of any equilibrium solution. Snce
change signs around a point where it is O. But this is a critical point. Thus, f(y) is completely
positive or completely negative in each region R; If y(x) is oscillatory or has a relative extremum,
then it must have a horizontal tangent lne at sorne point (xo, Yo). In this case yO would be a critical
point of the difIerential equation, but we saw aboye that the graph of a nonconstant solution cannot
intersect the graph of the equilibrium solution y = yo.
34. By Problem 33, a solution y(x) of dyjdx
monotone. Since y'(x)
= f(y)
cannot have relative extrema and hence must be
= f(y) > O, y(x) is monotone increasing, and since y(x) is bounded aboye
by C2, limx-+ooy(x) = L, where L ::; C2. We want to show that L = C2. Since L is a horizontal
asymptote of y(x), limx-+ooy'(x)
f(L)
= O. Using the fact that f(y) is contnuous we have
= f( x-+oo
lim y(x)) = Iim
x-oo
But then L is a critical point of f. Since

C2, L = C2
Cl
f(y(x))
= lim
x-oo
y'(x)
= O.
< L ::; C2, and f has no critical points between
Cl
and
35. (a) Assuming the existence of the second derivative, points of infiection of y(x) occur where
y"(x) = O. Frorn dyjdx = g(y) we have d2yjdx2 = g'(y) dyjdx. Thus, the y-coordinate of a
point of infiection can be located by solving g' (y) = O. (Points where dy j dx = Ocorrespond to
constant solutions of the difIerential equation.)
(b) Solving y2_y_6
= (y-3)(y+2)
points. Now d2yjdx2
= Owe see that 3 and -2 are critical
(2y - 1) dyjdx
(2y - l)(y - 3)(y + 2),
so the only possible point of infiection is at y
i,
although the
concavity of solutions can be difIerent on either side of y = -2 and
y = 3. Since y"(x) < O for y < -2 and
i < y < 3, and y"(x)
>O
< y < ~ and y > 3, we see that solution curves are concave
down for y < -2 and i < y < 3 and concave up for -2 < y 3. Points of inflection of solutions of autonomous difIerential
-s
for -2
-5
equations will have the same y-coordinates because between critical points they are horizontal
translates of each other.
28
Exercises 2.2
Exercises 2.2
In many of the following problems
solve for g(y) we exponentiate
implies g(y)
ecef(x).
wc will encounter
Letting
3. Frorn dy
4. From (
el
dx we obtain In
Iyl =
1
6. From - dy = -2x dx we obtain In
y
41n
From
+ 2 + ;)
(y
10. Frorn (2y
11.
dy =
cscy
- cos y = - ~x -
ll. From --
---2-
sec x
+ 5)2 dx
or
+ 2y + In Iyl
2
we obtain
dx or sin y dy
t sin 2x + e
+ 2e3x = e.
dy = (4x
1 - --.
x+e
Ixl + e or y = elx4
dy = x21n x dx we obtain y;
+ 3)2
= x + e or y =
Iyl = -x 2 + e or y = ele-x.
7. From e-2Ydy = e3xdx we obtain 3e-2y
9.
= elef(x)
+ e.
~e-3x
=-
4 cos y
2y
+3 =
cos2x dx
= x33In
Ixl -
(eY
eY
+ 1) 2
dy -
+ 5 + e.
-~(1 + cos 2x) dx we obtam
= 2x + sin 2x + Cl
_ex 3 dx we obtain - (eY

(eX + 1)
+ 1)-1
= ~(eX
y
x
(
2)1/2
(
2)1/2
1/2 dy =
1/2 dx we obtain
1+Y
= 1+x
(1 + y2)
(1 + x2)
15. From
S dS
16. Frorn -Q 1
-70
= k dr we obtain S = er.
dQ
= kdt
we obtain In
IQ -
701 = kt
29
+ e or Q -
70
+ e.
+ 1)-2 + e.
14. FrOl11
+ e.
4x
sin 3x
2
.
2
1
2
12. From 2y dy = - cos3 3x dx = - tan 3x sec 3x dx we obtain y = -'6 sec 3x
13. From
~x3
= qt.
+ e.
To
= eCef(x) which
= ~(x + 1)3 + e.
y-1
Ig(y)1 = f(x) + e.
I = ef(x)+c
t cos 5x + e.
)2 dy = dx we obtain -_-
= -x
5. From - dy
y
of the form In
This yields Ig(y)
ec we obtain g(y)
we obtain y
_e-3x dx we obtain y
y-1
+ 1)2 dx
an expression
both sides of the equation.
1. From dy = sin 5x dx we obtain y = 2. From dy = (x
Exercses
2.2
17. From P _1 p2dP

P
or -P
1-
= (1P + 1 _1)P
t
dP = dt we obtain In IPI-In 11-PI = Hc so that In 1 _P P = Hc

C1et
t
1 + C1e
= C1e . Solving for P we have P =
1
18. From N dN = (tet+2 - 1) dt we obtain In INI = tet+2 - et+2 - t + c.
19. From y - 23dy = x - 1 dx or (1 __ 5_) dy = (1 __ 5_) dx we obtain
y+
x+4
y+3
x+4
y-5Inly+31=x-5Inlx+41+c
20. From y + 1 dy = x + 2 dx or (1 + _2_)

y-1
x-3
y-1
or
dy = (1 + _5_)
x-3
Y +2Inly-ll=x+5Inlx-31+c
21. From x dx =
or
n
1 - y2
(X+4)5
y+3
=C1ex-y.
dx we obtain
(y_1)2 _
x-y
(x _ 3)5 - C1e
dy we obtain ~x2 = sin-1 y + c or y = sin (x2 + e).

2
1
1
eX
1
1
22. From -2 dy =
dx = ()2
dx we obtain -- = tan-1 eX + c or y = .
y
eX + e-x
eX + 1
y
tan-1 eX + c
23. From
-2-1-
dx = 4dt we obtain tan-1 x = 4t + c. Using x(7r/4) = 1 we find c = -37r/4.

x +1
solution of the initial-value problem is tan -1 x = 4t - 3; or x = tan ( 4t _ 3;).
24. From -')--1 dy

y- - 1
The
1 dx or -1( -- 1 - -- 1) dy = -1( -- 1 - -- 1) dx we obtain

.
x2 - 1
2 Y- 1 y + 1
2 x - 1 x + 1
= --
In Iy - 11-ln Iy + 11= In Ix - 11-ln Ix + 11+ In e or y - 1 = c(x - 1) . Using y(2) = 2 we find

y+1
x+1
y-l
c = 1. The solution of the initial-value problem is -y+1
x-l
= --
x+1
or y = x.
1 - x dx = ( 2"
1 - -1) dx we obtain
. In Iyl = -- l' -In Ixl = c or xy = c1e- 1/x. Using
25. From -1 dy = -2y
x
x
x
x
y(-l) = -1 we find C1= e-l. The solution of the initial-value problem is xy = e-1-1/x.
1
26. From 1 _ 2y dy = dt we obtain -~ In 11- 2yl = t + c or 1 - 2y = ee-2t. Using y(O) = 5/2 we find
C1= -4. The solution of the initial-value problem is 1 - 2y = _4e-2t
27. Separating variables and integrating we obtain
dx
Vf=X2
dy
J1- y2
=0
and
30
sin-1x-sin-1y=c.
or y = 2e-2t + ~ .
Exercises 2.2
Setting x = O and y =
V3/2
we obtain e =
-1r
/3. Thus, an implicit solution of the initial-value
problem is sin-l x - sin-l y = 1r/3. Solving for y and using a trigonometric identity we get
1
y = sin ( sin- x
28. From
1r)
+ '3
= x cos
1r
V3 V1=X'2
'31r + Vr.:":
1 - x2 sin '3 = '2 +
2
.
X
-x
dy dx we obtain
1 + (2y)2
- 1 + (x2)2
1
+e
- tan-l 2y = -- tan-l x2
2
2
Using y(l)
= O we find
C}
or
tan-l 2y
+ tan-l
x2 = el.
= 1r/4. The solution of the initial-value problem is
tan-l 2y
+ tan-l
x2
= ~.
4
= 2 and y(x) = -2 satisfy the initial conditions y(O) = 2 and

= -2, respectively. Setting x = ~ and y = 1 in y = 2(1 + eX)/(l - eX) we obtain
29. (a) The equilibrium solutions y(x)

y(O)
1 = 21 + ee
1-ee'
1-
ee
= 2 + 2ee,
and e=--.
-1 = 3ee,
3e
The solution of the corresponding initial-value problem is

y =2
1- lx-l
3
1 + le4x-l
3_ e4x-l
2---;----:3 + e4x-l
(b) Separating variables and integrating yields

1
In Iy -
21 -
In Iy -
1
In Iy
21 -
+ 21 + In el
In Iy
=x
+ 21 + In e = 4x
In I e(y - 2)
y+2
I = 4x
y - 2
4
e-- = e x
y+2
Solving for y we get y = 2(e+X)/(e-X).
The initial condition y(O)
= -2 can be solved for,
= -2. The initial condition y(O) = 2 does not correspond to a value

= 2 is a solution of the initial-value problem.
Setting x = ~ and y = 1 in y = 2(e + X)/(e - X) leads to e = -3e. Thus, a solution of the
yielding e = O and y(x)
of e, and it must simply be recognized that y(x)

initial-value problem is
y =2
~3e + x
-3e - e4x
3 - x-l
= 2 ---:----:3 + e4x-l
31
Exercises
30. From (_1._
2.2
Y - 1
+ -1)
y
.!.x dx
dy =
we obtain ln jy - 11-ln Iyl = ln Ixl + C or y =
1-
1
ClX
Another solution is y = O.
(a) If y(O) = 1 then y = 1.

(b)
If y(O) = O then y = O.
1
(e) If y(1/2) = 1/2 then y = 1 + 2x .
i we obtain
(d) Setting x = 2 and y =

1
1-
1
Thus,y=--=--.
1 + ~x
31. Singular solutions of dy/dx
(eX + e-X)dy/dx
= y2 is y
32. Differentiating ln(x2
Cl (2)
1-
2Cl
= 4,
and
Cl
-2'
2
2 + 3x
= x~
are y = -1 and y = 1. A singular solution of
= O.
+ 10) + cscy
= C we get
2x
x2
dy
+ 10 - coty csey dx = O,
2x
x2
+ 10
_ eosy_l_dy
siny sinydx
'
or
2xsin2ydx
- (x2
+ 10) eosydy
= O.
Writing the differential equation in the form

dy
dx
(x2
2xsin2 y
+ 10) eosy
we see that singular solutions oecur when sin2 y = O, or y = k7r, where k is an integer.
32
Exercises 2.2
33.
The singular solution
1 satisfies the initial-value problem.
y
1. 01
-0.004-0.002
0.0020.004
0.002 0.004
0.98
0.97
34. Separating variables we obtain

---
y-1
=x + c
dy
y-1
and
)2 = dx. Then
y
1.02
x+c-1
x+c
= ----
Setting x = O and y = 1.01 we obtain c = -100. The solution is

y=
0.0040.002
x - 101
.
x-lOO
0.99
0.98
dy
)2
y - 1
+ 0.01
= dx. Then
1
x+c
y=l+-tan--.
10
10
Setting x = O and y = 1 we obtain c = O. The solution is
10tan-110(y-1)=x+c
1.0004
and
1. 0002
1
x
y = 1 + 10 tan 10 .
0.9998
0.9996
dy
)2
y - 1
- 0.01
= dx. Then
1.0004
51n 10Y 1
111
10y- 9
= x + c.
1.0002
Setting x = O and y = 1 we obtain c = 51n 1 = O. The solution is

5 In 11Oy -
111 = x.
0.9998
10y- 9
0.9996
33
Exercises 2.2
37. Separating variables, we have
dy
Y - y3
dy
y(l _ y)(l
+ y) =
(1
1/2
y + 1_ y
1/2
- 1 + )dy
dx.
Integrating, we get
In Iyl -
"2 In 11 - yl - "2 In 11 + yl = x + c.
> 1, this becomes
When y
1
1
lny - -ln(y - 1) - -ln(y
2
2
Letting x
x
= Oand
+ 1) = In
y
~
Vy2 - 1
= x + c.
= 2 we find c = In(2/ V3). Solving for y we get Yl (x) =
2ex /
v'4e2x
3, where
> In( V3/2).
When O < y
< 1 we have
1
In y - -ln(l
2
Letting
-00
1
- y) - -ln(l
2
+ y) = In
x = O and y = ~ we find c = In(l/V3).
y
~
VI - y2
= x + c.
Solving for y we get Y2(X)
= eX/v'e2x + 3, where
< x < oo.
When -1
< y < O we have

1
ln( -y) - -ln(l
2
Letting x = O and y
where -00 < x < oo.
When y
1
- y) - -ln(l
2
= -~ we find c = In(l/V3).
+ y)
= In ~
-y
VI _ y2
=x
+ c.
= -ex/v'e2x+3,
Solving for y we get Y3(X)
< -1 we have
1
ln(-y) - -ln(l2
1
y) - -ln(-l2
y) = In
-y
#-=1 = x + c.
= O and y = -2 we find c = In(2/ V3). Solving for y we get

where x > ln( V3/2).
Letting x
34
Y4(X)
-2ex /
v'4e2x
- 3,
Exercises 2.2
y
SX
2
2
-2
-2
-2
-4
-4
-4
SX
l/(y - 3) _
(y - 3)2 -
dy/dx_
(y-1)2-
-I
38. (a) The second derivative of y is

d2y
dx2
1
(y-3)3'
The solution curve is concave up when d2y/dx2
> O or y > 3,
< Oor y < 3. From the phase

<3
and concave down when d2y/dx2
portrait we see that the solution curve is decreasing when y

and increasing when y
> 3.
-2
(b) Separating variables and integrating we obtain
y
8
(y - 3) dy = dx
'21Y 2 y2 _ 6y
3y = x
+e
+ 9 = 2x + el
(y - 3)2 = 2x
+ el
y = 3 ..j2x
-2
+ el.
The initial condition dictates whether to use the plus or minus signo
When Yl(O) = 4 we have q = 1 and Yl(x) = 3 + ..j2x
When Y2(O)
= 2 we have q = 1 and Y2(X) = 3 - ..j2x
+ l.
+ l.
When Y3(1) = 2 we have q = -1 and Y3(X) = 3 - ..j2x - l.

When Y4(-1)
= 4 we have el
= 3 and Y4(X) = 3 + ..j2x + 3.
39. (a) Separating variables we have 2ydy

y( -2)
= -1 we find e = -1, so y2
= (2x
= x2
+ l)dx.
+x
Integrating gives y2 = x2
- 1 and y = -vx2
root is chosen because of the initial condition.
35
+x
+ x + e.
When
- l. The negative square
Exetcises 2.2
y
(b) The interval of definition appcars to be approximately (-00, -1.65).
-7-
4 -3 -
2"
-1_1
-2
-3
-4
-5
+x (-00, -~ - h/5).
(e) Solving x2
O we get x
- ~ ~J5, so the exact interval of definition is
+ 2e3x
40. (a) From Problem 7 the general solution is 3e-2y

3e-2y
+ 2e3x = 5. Solving for y
we get y
= c. When y(O) = O we find c = 5, so
= -~ In 1(5 -
2e3x).
y
(b) The interval of definition appears to be approximately (-00,0.3).
-1
-2
(e) Solving 1(5 - 2e3x) = Owe get z = 1In(~), so the exact interval of definition is (-00, ~ ln(~)).
41. (a) While Y2(X)
= -V25
- x2 is defined at x
= -5
and x
= 5, y~(x) is not defined at these values,
and so the interval of definition is the open interval (-5,5).

(b) At any point on the x-axis the derivative of y( x) is undefined, so no solution curve can cross
the x-axis. Since
-xlv
is not defined when y
= O, the
initial-value problem has no solution.
42. (a) Separating variables and integrating we obtain x2 - y2 = c. Por c
i= O the
graph is a square
= Oand y = X. Since the

differential equation is not defined for y = O, solutions are y = x, x < Oand y = x, x > O.
The solution for y(a) = a is y = x, x> O; for y(a) = -a is y = -x; for .y( -a) = a is y = -x,
x < O; and for y( -a) = -a is y = x, x < O.
hyperbola centered at the origino All four initial conditions imply c
(b) Since x/y is not defined when y = O, the initial-value problem has no solution.
(e) Setting x
= 1 and y = 2 in x2 - y2 = c we get c = -3, so y2 = x2 + 3 and y(x)
= vx2
+ 3,
where the positive square root is chosen because of the initial condition. The domain is all real
numbers since x2
+ 3 > Ofor all X.

36
Exercises 2.2
43. Separating variables we have dy/(
I+Y2 sin2 y)
= dx
y
3.5
which is not readily integrated (even by a CAS). We

note that dy / dx
2': O for all values of x and y and that
dy/dx = O when y = O and y = 7f, which are equilibrium

solutions.
-6
44. Separating variables we have dy / (.jY
+ y) = dx / ( .x + x).
-4
-2
To integrate
J dt / ( Vi + t)
8 x
we substitute
u2 = t and get
J u+u
du
~2
J l+u
_2_
du = 2ln 11+ ul
+ C = 2ln(1 + vX) + c.
Integrating the separated differential equation we have

2ln(1 +.jY) = 2ln(1
Solving for y we get y
+ vX) + C
or
ln(l +.jY) = ln(l
+ vX) + ln cj.
= [C1(1+.x) - 1J2.
45. We are looking for a function y(x) such that

y2
(~~r
= 1.
Using the positive square root gives

dy
dx
=~
Thus a solution is y = sin(x
~ dy
=:}
v=
= dx
=:}
. -1
sm y
= x + C.
_y2
+ c). If we use the negative square root we obtain
y = sin(c - x)
=-
sin(x - c)
=-
sin(x
+ Cl)'
= -1 are solutions.
Note also that y = 1 and y

46. (a)
(b) Por Ixl > 1 and Iyl > 1 the differential equation is dy / dx = ~
variables and integrating, we obtain
dy
dx
and cosh-l y = cosh-1 x
vX2=l
37
/.JX2=l.
+ c.
Separating
Exercises 2.2
Setting x
= 2 and y = 2 we find c = cosh-1 2 - cosh-1 2 = O and cosh-1 y = cosh'"! x. An
explicit solution is y = x.
47. (a) Separating variables and integrating, we have
(3y2
+ l)dy =
-(8x
+ 5)dx
and
y3
+ Y = -4x2
Using a CAS we show various contours of f(x, y)
- 5x
= y3
+ c.
+y+4x2
5x. The plots shown on [-5,5] x [-5,5J correspond to c-values

of O, 5, 20, 40; 80, and 125.
(b) The value of c corresponding to y(O) = -1 is f(O, -1) = -2;

to y(O)
= 2 is f(0,2)
and to y(-l)
= 10; to y(-l)
= 4 is f(-1,4)
-3 is -31.
(-2y
+ y2)dy
= (x - x2)dx
and
-y
2131213
+ 3'y
= 2"x - 3'x
+ c.
Using a CAS we show sorne contours of f(x, y)

2y3 - 6y2
+ 2x3
= 67;
- 3x2. The plots shown on [- 7, 7] x
[-5,5J correspond to e-vales of -450, -300, -200,

-120, -60, -20, -10, -8.1, -5, -0.8, 20,60, and
120.
38
Exercises 2.3
(b) The value of e corrcsponding to y(O) = ~ is f(O, ~) =

- 247.
The portion of the graph between the dots cor-
responds to the solution curve satisfying the intial condition.
To determine the interval of definition we find
dy/dx
for 2y3 - 6y2
+ 2x3
differentiation we get y'
- 3x2
= -~.
(x - x2)/(y2
Using implicit
-2
- 2y), which
-4
is infinite when y = O and y = 2. Letting y = O in

2y3 - 6y2 + 2x3 - 3x2 = - 247 and using a CAS to solve
for x we get x = -1.13232. Similarly, letting y
-2
= 2, we find x
= 1.71299. The largest interval
of definition is approximately (-1.13232,1. 71299).

(c) The value of e corresponding to y(O) = -2 is f(O, -2) =
-40. The portion of the graph to the right of the dot

corresponds to the solution curve satisfying the initial
condition. To determine the interval of definition we find
dy / dx for 2y3 - 6y2 + 2x3 - 3x2 = -40. Using implicit
-2
-4
-6
= (x - x2)/(y2 - 2y), which

is infinite when y = O and y = 2. Letting y = O in
differentiation we get y'
+ 2x3
2y3 - 6y2
for x we get x
-8
-4
-2
10
- 3x2 = -40 and using a CAS to solve
= -2.29551. The largest interval of definition is approximately (-2.29551, (0).
Exercises 2.3
1. For y' - 5y = O an integrating factor is e-J 5 dx

-00
+ 2y =
= e2x so that
O an integrating factor is eJ2dx
+y
= e3x an integrating factor is eJ dx
= ce-2x for
dx
+ 4y
-00
1 an integrating
factor is eJ 4 dx
-00
x so that
d~ [e4Xy]
= 1x and y =
1 + ce-4x
< x < oo. The transient term is ce-4x.
5. For y' +3x2y = x2 an integrating factor is eJ 3x2 dx

for
d~ [e2Xy] = O and y
= eX so that ..!!:_ [eXy] = e4x and y = -41e3x+ ce?" for
< x < oo. The transient term is ce-x.
4. For y'
for
= O and y = ce5x for
< x < oo. The transient term is ce-2x.
3. For y'
-00
d~ [e-5Xy]
< x < oo.
2. For y'
-00
= e-5x so that
< x < oo. The transient term is ce-x3.

39
ex3 so that ddx [ex3 y] = x2 ex3 and y = i+ce-x3
Exercses 2.3
x3 an integrating factor is eJ 2xdx
+ 2xy =
6. For y'
~x2 - ~ + ee-x2 for
for O < x
< x < oo. The transient term is ce-x2
+ -11y =
7. For y'
-00
2" an integrating factor is e (l/x)dx
< oo.
= eX so that
1
= x so that -d [xy] = -1 and y = -lnx
+ -e
dx
= x2 + 5 an integrating factor is e- J2dx = e-2x so that d~ [e-2Xy] = x2e-2x + 5e-2x

and y = _~x2 - ~x -1} + ee2x for -00 < x < oo.
8. For y' _ 2y
= ~ so that .!!:_ [~y]
9. For y' _ ~y = xsinx an integrating factor is e- J(l/x)dx

x
y = ex - x cos x for O < x < oo.
+ ~y = ~
10. For y'
for O < x
y = ~x3 - ~x
12. For y'_
13. For y'

y
1 eX
=- ?
2 z-
14. For y'
x4 so that
ddx [x4y]
+~
x
y
x6
x4 and
< oo.
+ --ce"
x+1
for O < x
for -1
< x < oo.
= x2ex
factor is eJ[1+(2/x)]dx
< oo. The transient term is
so that
ce:"
x
-2- .
~e-x sin 2x an integrating factor is eJ[1+(l/x)]dx = xe" so that .!!:_ [xeXy]

x
~
1
ee-x
__ e-x cos 2x + -- for O < x < oo. The entire solution is transient.
2x
x
16. For dx
dy
2
4y5 an integrating factor is e- J(4/y)dy = y-4 so that
for O < y
=
< oo.
eY an intezratinz factor is eJ(2/y)dy

o
= y2 so that
-eY + 2" + 2" for O < y < oo. The transient term is
Y
= 3x and y = ~ + ex-2
[x2y]
+ (1 + ~) Y =
+ cy4
dx
= sinx and
dx
ce:"
2y6
eY
for O < x
2x + 3
and y = -x - -x+1
15. For ~~ - ; x
.!!:_
= x an integrating factor is e-f[x/(1+x)]dx = (x+ l)e-X so that .!!:_ [(x + l)e-Xy] =
+ -2x
sin 2x and y
+ ex-4
( 1 + -x2) y = -x2eX an integrating
so that
x2 - 1 an integrating factor is eJ(4/x)dx
x
y
(1 + x)
+ l)e-
x(x
= x2
dx
< oo.
+ ~y
11. For y'
an integrating factor is eJ(2/x)dx
40
2+e
-2- .
:y [y-4x]
4y and
Exercses 2.3
= sec x an integr ating factor is eJ

y = sin x + e cos x for -7f /2 < x < 7f/2.
17. For y' + (tan x)y
tan
x dx
18. For y' + (cot x)y = sec2 x ese x an integrating factor is eJ

and y = sec x + e ese x for O < x
= 2xe-x
19. For y' + x + 2 y

x+1
x2
2x and y = _x+1
20. For y' + _4_
x+2
dr
de
+r
e-x + -x+1
sec e
ce-3x
y = e-3x + -x
for O < x
= (x+1)eX,
y] = sec2 x
so'!!:_ [(x + l)eXy]

dx
--.iy
x
-1
< oo. The entire solution is transient.
= cos e an integrating factor is eJ secadO
(3 +~)
25. For y ,1+ - y

x
e-x for -1 < x
dx
(4t - 2)et2-t and P
and (x - l)y
< 7f/2.
.!!:_ [( sin x)
= 3(x + 2)-1 + c(x + 2)-4 for -2 < x < oo. The entire solution is transient.
22. For ~~ + (2t - l)P
24. For y' +
x dx = sin x so that
Y = ( 5 )2 an integrating factor is eJ[4/(x+2)]dx = (x + 2)4 so that dd [(x + 2)4y]

x+2
x
1 + sine and r(sece + tan ')
23. For y' +
cot
an integrating factor is eJ[(x+2)/(x+1)]dx
x+1
5(x + 2)2 and y

21. For
= sec x so that :x [(sec x) y] = sec2 X and
= sec e+tan
e so that ded
[r(sec e + tan
= e - cose + e for -7f/2 < e < 7f/2 .
4t - 2 an integrating factor is eJ(2t-1)dt
= 2 + cet-t2 for
-00
et2-t so that :t [pet2-t]
< t < oo. The transient term is c-t2.
= e~3X an integrating factor is ef[3+(1/x)]dx = xe3x so that ~ [xe3Xy]

for O < x
e)] =
1 and
< oo. The transient term is ce-3x [.
= x + 1 an integrating factor is eJ[2/(x2-1)Jdx = x - 1 so that
x-1
x(x + 1) + c(x + 1) for -1
x+1
< x < l.
= -1 ex an integrating factor is e J(1/ x )dx

x
< oo. If y(l) = 2 then e = 2 - e and y
1
x
so that
2- e
.
x
.!!:_ [X dx
x+1
1y] - 1
1
-d []xy = e x and y = -e
dx
x
+- e
x
= 2 and x = 2y2
+ cy
= _ex + --
26. For dx _ ~ x = 2y an integrating factor s e- J(1/y)dy = ~ so that

dy
y
y
49
for -00 < y < oo. If y(l) = 5 then e = -49/5 and x = 2y2 -
.!!:_
dy
[~x]
y
s-Y'
27. For di + !.!:. i = E an integrating factor is eJ (R/ L) dt

dt
L
L
i = ~ + ce-Rt/L
for
-00
< t < oo. If i(O)
= eRt/ L so that
= io then e = io
!!_
dt
[ieRt/ L]
E eRt/ L and
L
- E/ R and i = ~ + (io - ~) CRt/L.
28. For dT - kT = -Tmk an integrating factor is eJ(-k)dt = e-kt so that dd [Te-kt] = -Tmke-kt
dt
t
T = Tw. + cekt for -00 < t < oo. If T(O) = To then e = To - Tm and T = Tm + (To - Tm)t.
41
and
Exercises 2.3
= Inx an integrating factor is ef[l/(x+l))dx = x + 1 so that ~[(x + l)y] = ln z

29. For y' + _l_y
x+1
x+1
~
x
x
C
and y = -In x - -+ -for O < x < oo. If y(l) = 10 then C = 21 and y =
x+1
x+1
x+1
x
x
21
--Inx---+--.
x+1
x+1
x+1
= cos2 x an integrating factor is eftanxdx = secx so that ~ [(secx) y] = cosx and
dx
y = sinxcosx + ccosx for -1f/2 < x < 1f/2. If y(O) = -1 thenc = -1 and y = sinxcosx - cosx.
30. For y' + (tanx)y
31. For y' + 2y =
f (x) an integrating factor is e2x so that

ye
2x
={
-2le2x
+ ci ,
C2,
If y(O) = O then
C2
-1/2
Cl
O:::; x < 3;
x> 3.
and for continuity we must have
1"x
= ~- ~so that
O :::; x :::;3;
x> 3.
32. For y' + y =
f (x) an integrating factor is e" so that

O:::;x:::;l;
x>l.
If y(O)
= 1 then
Cl
Oand for continuity we must have
so that
ye
= 2 then
Cl
2e
x>l.
an integrating factor is eX so that

x2
{~ex2
C2,
If y(O)
O :::; x :::;1;
1,
y= { 2el-x-1,
33. For y' + 2xy = f(x)
C2
+Cl,
O:::; x:::;
1;
x> l.
= 3/2 and for continuity we must have
C2
1e + ~ so that
O :::; x :::;1;
x>l.
42
Exercises 2.3
~'
1+ x
- { -x
1 + x2'
,2x
Y +--y1 + x2
34. For
O:::;x:::;l;
1
x
an integrating factor is 1 + x2 so that
( 1 +x
If y(O)
= O then
Cl
= O and
-1
O:::; x:::;
2) y = { ~x2 + ci,
_~x2+C2,
x>l.
for continuity we must have

1
- -
_
y -
C2 =
2 (1, + x2) ,
3
1
2 (1 + x
1;
2) - -,
< x < l'
x>
1 so that
,
l.
35. We need
P(x)dx
2X,
O:::;x:::;l
-21nx,
x>
20
15
An integrating factor is
eJ
P(x)dx =
2x
e ,
10
O:::;x:::;l
{ 1/x2,
and
d [{ ye2x,
dx
y / x2 ,
O:::; x < 1] = { 4xe2x,

x >1
4/ x,
O:::; x < 1
>1
Integrating we get
ye2x,
{ y / x2 ,
Using y(O)
= 3 we find
Cl =
O:::; x :::;1 = { 2xe2x - e2x

x > 1
41n x + C2,
+ Cl,
4. For continuity we must have
2x - 1 + 4e-2x, .
{
Y = 4x21nx + (1 + 4e-2)x2,
36. An integrating factor for y' - 2xy
C2
O:::; x :::;1 .
x > 1
= 2 - 1 + 4e-2 = 1 + 4e-2.
O :::; x :::;1
x >1
= 1 is e-x2. Thus
-d [_X2
e y 1 =e
_x2
dx
2
[X
e-X y = Jo e-t
dt
erf(x)
and
2
Y = eX erf(x)
. 43
+ ce" .
+c
Then
Exercises
Frorn y(l)
2.3
= 1 we get 1= e erf(l)
Y
37. For y'
+ eXy =
+ ce,
so that c = e-l - erf(l). Thus
1 an integrating factor is
d
When y'
= 1 we get e = e, so y
+ eXy
= e-e
= eX
-1
+ eX (erf(x) - erf(l)).
and
t
fox dt
ee y
t
= lox
o dt + c.
+ e 1-e x .
= O we can separate variables and integrate:
dy
y
= _ex dx
In jyj
and
Thus y = Cle-e". From y(O) = 1 we get Cl = e, so y

When y'
Thus
- lee y] = ee
dx
Frorn y(O)
= eX erf(x) + (e-l - erf(l))eX
+ eXy = e"
we can see by inspection that y
_ex
+ c.
= el-e".
= 1 is a solution.
38. We want 4 to be a critical point, so use y' = 4 - y.

39. (a) Al! solutions of the form y = x5ex - x4ex
since 4/ x is discontinuous at x
+ cx4
satisfy the initial condition.
In this case,
= O, the hypotheses of Theorem 1.1 are not satisfied and the
initial-value problem does not have a unique solution.

(b) The differential equation has no solution satisfying y(O) = Yo, yO =j:. O.
(c) In this case, since xo =j:. O, Theorem 1.1 applies and the initial-value problem has a unique
solution given by y
40. On the interval (-3,3)
= x5ex
- x4ex
+ cx4
where c = yo/x~ - xoexQ
+e
XQ
the integrating factor is

efxdxj(x2-9)
= e- fxdxj(9-x2)
= e1In(9-x2)
)9 _ x2
and so
_:!_ [) 9 - x2
dx
41. We want the general solution to be y

approaches O as x
-> 00
y]
=O
and
y = ";9 c-
= 3x - 5 + ce:", (Rather than e-x, any function that
could be used.) Differentiating we get

y' = 3 - ce-x
so the differential equation y'
+y
=3-
(y - 3x
+ 5) = -y + 3x
= 1 is l/e and the right-hand derivative at x = 1 is
1 - l/e. Thus, y is not differentiable at x = 1.

= c/x3
- 2,
= 3x - 2 has solutions asymptotic to the line y = 3x - 5.
42. The left-hand derivative of the function at x

43. (a) Differentiating Yc
"
x2
we get
,
3c
3 c
3
y - -_ - -- _ - --y
e x4 x x3 x
44
Exercises
+ 3y = O. Now
so a differential equation with general solution Yc = c/x3 is xy'

xY~
+ 3yp = x(3x2) + 3(x3)
2.3
= 6x3
so a differential equation with general solution y = c/x3
+ x3
is xy'
+ 3y
= 6x3. This will be a
general solution on (O, (0).

(b ) Since y(l)
y(l)
13
= 13 - 1/13
+ 2/13
O, an initial condition is y(l)
= 3, an initial condition is y(l)
= O. Since
= 3. In each
case the interval of definition is (O, (0). The initial-value problem

xy'
+ 3y = 6x3,
= O has
y(O)
solution y
= x3 for
-00
< x < oo.
-3
(e) The first two initial-value problems in part (b) are not unique. For example, setting y(2) =
23 - 1/23 = 63/8, we see that y(2) = 63/8 is also an initial condition leading to the solution
y = x3 - 1/x3.
44. Since eJ P(x)dx+c
= eJ P(x)dx
C1eJ P(x)dxy
C2
= cleJ
r=.
we would have
C1eJ P(x)dx f(x) dx
and
eJ P(x)dxy
C3
eJ P(x)dx f(x) dx,
which is the same as (6) in the texto

45. We see by inspection that y = O is a solution.
46. The first equation can be solved by separation of variables. We obtain x
we obtain
C1 = Xo
and so x
= xoe-Alt.
cle-Alt.
The second equation then becomes

or
which is linear. An integrating factor is eA2t; Thus

.!!._ [eA2ty 1 = xOA1e-AlteA2t
= XOA1e(A2-Al)t
dt
e A2ty = \ XOA1\ e (A2-Al)t + C2

/\2 - /\1
From y(O) = Yo we obtain
C2
y=
= /\2
\ XOA1
\
- /\1
-Alt
+ C2e -A2t
= (YOA2 - YOA1 - XOA1)/(A2 - Al)' The solution is

XOA1 -Alt
e
A2 - Al
+ YOA2 - YOA1 - XOA1 e -A2t

A2 - Al
45
Frorn
x(O) = Xo
Exercises 2.3
47. (a) Letting y = YI
+ Y2
+ Y2(XO)
we have y(xo) = YI(XO)
+ O= a
=a
and
yl + P(x)y = (YI + Y2)1 + P(X)(YI + Y2)

= (Yi + P(X)YI) + (y; + P(X)Y2) = O + (x) = (x).
Thus y = YI
+ Y2
is a solution of the initial-value problem yl
+ P(x)y
= (x),
y(xo) = a.
(b) By Theorem 1.1 the initial-value problem
yl + P(x)y = h(x)
+ h(x),
(1)
y(xo)=a+,6
has a unique solution. Consider y = YI +Y2 Since y(xo) = YI(XO) +Y2(XO) = a+,6,
the initial condition of (1). Also
yl + P(x)y = (YI + Y2)1 + P(x)(Yl

= (y~
y satisfies
+ Y2)
+ P(X)YI) + (y~ + P(X)Y2) = h(x) + f2(x),
so y satisfies the differential equation in (1). Thus y = Yl +Y2 is the unique solution of (1) and
y(xo) = a +,6.
Since
CIYI
yl + P(x)y
is a solution of yl
+ P(x)y
e-X2 y
ft/2,
Y = eX 2
=_
_x2
(X
Jo
e-t2 dt = _ ft erf(x)
2
and noting that erf(O) = O, we get

--
ft erf(x) + -J1F)
2
CIa,
yl - 2xy = -1 is e-x2. Thus
d [ _x2 1
dx e y =-e
From y(O) =
y(xo)
and C2Y2 is a solution of

= C2f2(X), y(xo) = C2,6, we see by the aboye argument that y(xo) = CIa + c2,6.
48. (a) An integrating factor for
(b)
= Clil(X),
+ c.
ft /2.
Thus
= -ft eX 2 (1 - erf(x)) = -ft eX 2 erfc(x).

2
2
Using Mathematica we find y(2) ~ 0.226339.
49. (a) An integrating factor for

I
2
y +-Y=
x
10sin x
3
x
46
Exercses 2.3
is x2. Thus
d [ 2 1
-xy=l-dx
Osinx
X
x 2 y = 10
y
From y(l)
= Owe get e = -lOSi(l).
la
sin t
-dt
+C
10x-2Si(x)
+ cx-2.
Thus
y = lOx-2Si(x) - lOx-2Si(1) = 10x-2(Si(x) - Si(l)).

y
2
(b)
x
-1
-2
-3
-4
-5
(e) From the graph in part (b) we see that the absolute maximum occurs around x = 1.7. Using
the root-finding capability of a CAS and solving y'(x) = O for x we see that the absolute
maximum is (1.688,1.742).
dy
. 2
Y
= sin x dx
Now, letting t
and
(X
2
In Iyl = Jo sin t dt
+ c.
;tiu we have
rrr2, Jo(y2;; X sin( 2,u2)du,
(X
Jo sin t2 dt = V
tt
so
y
cleJ;sint2dt = CIe#Jo.2!;x
Using S(O) = O and y(O) = 5 we see that
Cl
sin(1ru2/2) du
= 5 and y
(b)
-JO
-5
JO
47
= cle#S(y2;;x).
= 5e#S(y2;;x).
Exercises 2.3
(e) Frorn the graph we see that as x
9.35672. Since limx_oo 5(x)
-~, limx__:_ooy(x) = 5e-.;;s
-+ 00,
y(x) oscillates with decreasing amplitude approaching
= ~, limx_oo y(x) = 5e.;;s
~ 9.357, and since lim.,__
oo
S(x) =
~ 2.672.
(d) Frorn the graph in part (b) we see that the absolute maximum occurs around x
absolute minimum occurs around x = -1.8.
= 1.7 and the
Using the root-finding capability of a CAS and
solving y'(x) = O for x, we see that the absolute mximum is (1.772,12.235) and the absolute
minimum is (-1.772,2.044).
Exercises 2.4
1. Let M = 2x-1
h'(y)
and N = 3y+7 so that My = O = Nx. From fx
= 3y + 7, and h(y)
= ~y2 + 7y. The solution is x2 -
= 2x-1
we obtain f
+ ~y2 + 7y
= x2-x+h(y),
= c.
2. Let NI = 2x + y and N = -x - 6y. Then My = 1 and Nx = -1, so the equation is not exacto
= 5x + 4y and N = 4x - 8y3 so that My = 4 = Nx. From fx = 5x + 4y we obtain

f = ~x2 + 4xy + h(y), h'(y) = _8y3, and h(y) = _2y4. The solution is ~x2 + 4xy - 2y4 = C.
4. Let j1I = siny - ysinx and N = cosx + xcosy - y so that My = cosy - sinx = Nx' From
3. Let M
fx = sin y - y sin x we obtain f
= x sin y + y cos x + h(y),
h'(y)
= -y,
and h(y) =
h2. The solution
is x sin y + y cos x - ~y2 = C.
= 2y2x - 3 and N
5. Let M
1=
= 2yx2 + 4 so that NIy
x2y2 - 3x + h(y), h'(y) = 4, and h(y)
4xy
Nx'
Frorn fx = 2y2x - 3 we obtain
= 4y. The solution is x2y2 - 3x + 4y = C.
6. Let M = 4x3 - 3y sin 3x - y/x2 and N = 2y - l/x + cos 3x so that My

N = 1/x2 - 3sin3x.
= -3 sin 3x - 1/x2 and
The equation is not exacto
7. Let M = x2 - y2 and N
= x2
- 2xy so that My = -2y and Nx
2x - 2y. The equation is not
exacto
8. Let M = 1 + lnx+y/x
1=
and N = -1 +lnx so that NIy = l/x = Nx' From Iy = -1 +lnx we obtain
-y + ylnx + h(y), h'(x) = 1 + lnx, and h(y) = x lnx. The solution is -y + ylnx + xlnx = C.
9. Let M = y3 - y2sinx
- x and N = 3xy2 + 2ycosx
Ix = y3 - y2sinx - x we obtain I = xy3 + y2cosx

solution is xy3 + y2 cos X - ~x2 = C.
ix
11. Let M
= 3y2 - 2ysinx = Nx' From
- ~x2 + h(y), h'(y) = O, and h(y) = O. The
= x3 + y3 and N = 3xy2 so that My = 3y2 = Nx.
10. Let NI
1=
so that My
From Ix = x3 + y3 we obtain
+ xy3 + h(y), h'(y) = O, and h(y) = O. The solution is x4 + xy3 = C.
= ylny - e-xy and N = l/y + xlny so that My = 1 + lny + ye-xy and Nx = lny. The
equation is not exacto
48
Exercises 2.4
12. Let M
obtain
= 3x2y + eY and N =
f = x3y + xeY + h(y),
=
13. Let M
f
= xy
14. Let M
+ 2ex + h(y),
= 1-
3/x
= 1 = Nx. Frorn fx = y - 6x2 = O. The solution is xy - 2x3 -
x so that My
= O,and
h' (y)
y and N = 1 - 3/y
x - 31n Ixl + xy
f =
+ y + xy
+ h(y),
h(y)
+ x so that
h'(y)
= 1=
My
1 - ~, and h(y)
From
Nx'
2xex we obtain
2xex
+ 2ex = c.
= 1-
fx
+y
3/x
y - 31n Iyl. The solution is
- 31n Ixyl = c.
15. Let M
y - 6x2 - 2xex and N
- 2x3 - 2xex
we obtain
x
+ xeY - 2y so that My = 3x2 + eY = Nx' From fx = 3x2y + eY we

h'(y) = -2y, and h(y) = _y2. The solution is x3y + xe" - y2 = c.
x3
1/ (1 + 9x2)
x2y3 -
1/ (1 + 9x2)
we obtain
= c.
and N
= !x3y3
x3y2 so that My
-!arctan(3x)
+ h(y),
3x2y2
Nx'
h'(y) = O, and h(y)
From
= x2y3 -
fx
= O. The solution is
x3y3 - arctan(3x)
16. Let M = -2y and N
so that
5y-2x
My
= -2
+ ~y2
h'(y) = 5y, and h(y) = ~y2. The solution is -2xy
= tanx -
17. Let M
and N
sinxsiny
tan x - sin x sin y we obtain

is In Isec xl
18. Let M
From
fx
+ cos x sin y
= In Isec xl
-2xy+h(y),
= c.
so that
cosxcosy
-2y we obtain f
My
+ cos x sin y + h(y),
-sinxcosy
h' (y) = O, and h(y)
Nx'
From
= O. The
fx
solution
= c.
2y sin x cos x - y
= Nx. From fx
2ysinxcosx
+ 2y2exy2 and
- 1+ 4xy eXY
My
- y
+ 2y2exy2 we obtain
2sinxcosx
h(y) = O. The solution is ySin2 x - xy
+ sin2 x + 4xyexy2 so that
N = -x
+ 2exy2
f =
+ 4yeXY
ysin2 x - xy
= Nx'
+ 2exy2 + h(y),
h'(y)
= O, and
= c.
= 4t3Y-15t2-y
and N = t4+3y2-t
so that My = 4t3-1 = Nt. From ft = 4t3Y-15t2-y
we
obtain f = t4y - 5t3 - ty + h(y), h' (y) = 3y2, and h(y) = y3. The solution is t4y - 5t3 - ty + y3 = c.
19. Let M
20. Let M = l/t+

Nt Frorn
ft =
1/t2-y/
l/t
and h(y) = yeY
(t2
+ 1/t2
+ y2) and
- y/ (t2
N = yeY +t/
= x2+2xy+y2
and N
we obtain f = !x3 + x2y +
!x3
My = (y2 _ t2) / (t2
(f) +
+ y2)2
h(y), h'(y)
= yeY,
eY. The solution is
21. Let M
+ x2y + xy2
+ x2y + xy2
+ y2) so that
+ y2) we obtain f = In Itl- ~ - aretan
In Itl- ~ - aretan
!x3
(t2
= 2xy+x2-1
xy2
h(y),
(f) +
so that
h'(y)
My
yeY
eY
= 2(x+y)
-1, and h(y)
= c.
= Nx' From fx = x2+2xy+y2
= -y. The general solution is
- Y = c. lf y(l) = 1 then c = 4/3 and the solution of the initial-value problem is

- Y=
1.
49
Exercses 2.4
= eX+y
22. Let M
and N
= 2+x+yeY
h(y), h'(y) = 2 + yeY, and h(y)

If y(O)
= 1 then e = 3 and
so that My
= 2y + yeY
= 1 = Nx'
From Ix
= eX+y
we obtain I = eX+xy+
y. The general solution is eX + xy
the solution of the initial-value problem is eX + xy
+ 2y + ue" + 2y + yeY -
eY = c.
eY
= 3.
+ 2t - 5 and N = 6y + 4t - 1 so that My = 4 = Ni, Prom!t = 4y + 2t - 5 we

obtain I = 4ty + t2 - 5t + h(y), h'(y) = 6y - 1, and h(y) = 3y2 - y. The general solution is
4ty + t2 - 5t + 3y2 - y = c. If y( -1) = 2 then e = 8 and the solution of the initial-value problem is
4ty + t2 - 5t + 3y2 - y = 8.
23. Let M = 4y
24. Let M
= t/2y4 and N = (3y2 - t2) /y5 so that My = -2t/y5
3
1=-4
+
h(y), h'(y) = 3' and h(y)
4y
y
t2
3
--2
2y
Ni. From
It
= t/2y4 we obtain
3
. The general solution is -4 --2
4y
2y
t2
then e = -5/4 and the solution of the initial-value problem is ~ - -;.

4y
2y
= c.
If y(l)
=1
= -~
- 3x2y - 2x and N ='2y sinx .; x3 + lny so that My = 2ycosx

- 3x2 = Nx. From
Ix = y2 cos X -3x2y -2x we obtain 1= y2 sinx _x3y _x2 +h(y), h'(y) = In y, and h(y) = y In y-y.
25. Let M = y2 cos X
The general solution is y2 sin x - x3y - x2 + y In y - y = c. If y(O)

of the initial-value problem is y2 sin x - x3y - x2 + y In y - y = O.
= y2 + Y sin x and N = 2xy - cos x-1/
26. Let M
y2
+ ysinx
we obtain
= xy2 - Y cos x
(1
+ h(y),
+ y2)
=
+ 4kxy3
and N
= 3y2 + 40xy3
-1
--2
l+y
1 then e
initial-value problem is xy2 - Y cosx - tan-1 y = -1 27. Equating My = 3y2
so that My = 2y
h'(y) =
solution is xy2 - Y cos x - tan-1 y = c. If y(O)
From
Iy = 2x2y
+ 2xy2
cos x and N
cos x we obtain
I = x2y2
+ sin x = Nx'
From Ix =
'
and h(y) = - tan-1 y. The general
-1 - tt/ 4 and the solution of the
we obtain k
= 10.
28. Equating My = 18xy2 - sin y and N = 4kxy2 - sin y we obtain k

29. Let M = _x2y2 sin x
= e then e = O and the solution
= 9/2.
= 2x2y cos X so that My = _2x2y sin x + 4xy cos x

cos X
+ h(y),
h' (y)
= O, and
h(y)
= O. The
= Nx'
solution of the
differential equation is x2y2 cos x = c.

30. Let M
and N = (y2+2xy-x2)/(y2+2xy+x2)
(x2+2xy_y2)/(x2+2xy+y2)
My = -4xy/(x
+ y)3 = Nx'
h(y), h'(y) = -1, and h(y)
From Ix = (x2
+ 2xy + y2
- 2y2) /(x +y)2 we obtain I
so that
2y2
= x + --
= -y. The solution of the differential equation is x2 + y2 =
x+y
c(x + y).
+ 3x2
+ h(y),
31. We note that (My - Nx)/ N = s.], so an integrating factor is eJ dx/x = x. Let M = 2xy2
and N = 2x2y so that My
h'(y)
= O, and h(y)
= 4xy =
Nx' Frorn Ix
= 2xy2 + 3x2
we obtain I = x2y2
= O. The solution of the differential equation is x2y2
+ x3
= c.
32. We note that (My - Nx)/N = 1, so an integrating factor is eJ dx = e", Let M = xyeX
50
+ x3
+ y2ex + yeX
Exercises 2.4
and N = xex + 2yeX so that My = xex + 2yeX + eX = N.
I = xyeX + y2ex + h(x),
xyeX
+ y2ex
h'(y) = O, and h(y) = O. The solution of the differentia1 equation is
= c.
33. We note that (Nx - My)/M
= 2/y, so an integrating factor is eJ2dy/y = y2. Let M = 6xy3 and
N = 4y3 + 9x2y2 so that My = 18xy2

h' (y)
From Iy = xe" + 2yeX we obtain
Nx'
From Ix = 6xy3 we obtain I = 3x2y3 + h(y),
= 4y3, and h(y) = y4. The solution of the differentia1 equation is 3x2y3 + y4
34. We note that (My - Nx)/N

cosxcscx
cotx,
= -
so an integrating factor is e- Jcotxdx
= cotx and N = (1 + 2/y) sinxcscx
we obtain I = 1n(sinx)+h(y),
h'(y)
equation is 1n(sin x) + y + ln y2
cscx.
Let M
= 1 + 2/y, so that My = 0= Nx. From Ix = cotx
= 1+2/y,
and h(y) = y+1ny2. The solution ofthe differentia1
= c.
= 3, so an
= c.
integrating factor is eJ3dx
= e3x.
Let
= (10 - 6y + e-3x)e3x = 10e3x - 6ye3x + 1
M
and
so that My
h' (y)
= O, and
_6e3x + Nx. From Ix
10e3x - 6ye3x + 1 we obtain I == 130e3x- 2ye3x + x + h(y),
h(y) = O. The solution of the differential equation is
36. We note that (Nx - My)/M
= -3/y,
13 e3x
so an integrating factor is e-3Jdy/y
M = (y2
+ xy3)/y3
= l/y
- 2ye3x + x
= 1/y3.
= c.
Let
+x
and
N = (5y2 - xy + y3 sin y)/y3 = 5/y - x/y2 + sin y,
so that My = _1/y2
= Nx. From Ix = l/y+x
we obtain I
= x/y+~x2+h(y),
h'(y) = 5/y+siny,
and h(y) = 51n Iyl- cos y. The solution of the differentia1 equation is x/y + ~x2 + 51n Iyl- cos y = c.
Let M = x/(4+x2)
- 2x/(4 + x2), so an integrating factor is e-2 J xdx/(4+x2) = 1/(4 + x2).
and N = (x2y+4y)/(4+x2)
= y, so that My = O = Nx. From Ix = x(4+x2)
we obtain I = ~ 1n(4 + x2) + h(y), h'(y) = y, and h(y) = ~y2. The solution of the differentia1
equation is ~ 1n(4 + x2) + ~y2
= c.
38. We note that (My - Nx)/ N = -3/(1 + x), so an integrating factor is e-3 J dX/(l+x) = 1/(1 +x)3.
M = (x2+y2-5)/(1+x)3
From Iy = -y/(l
and N = -(y+xy)/(1+x)3
+ x)2 we obtain I =
= -y/(1+x)2,
-h2/(1 + x)2
+ h(x),
so that My = 2y/(1+x)3
h'(x)
2
+ (1+x)2
2
+ (l+x)
51
+lnI1+xl
= Nx.
= (x2 - 5)/(1 + x)3, and
h(x) = 2/(1 + x)2 + 2/(1 + x) + ln 11+ z]. The solution of the differential equation is
y2
-2(1+x)2
Let
= c.
Exercises 2.4
39. (a) Implicitly differentiating x3
3x2
+ 2x2
dy
dx
+ 2x2y + y2 = c and
+ 4xy + 2y
we obtain
dy = _ 3x2 + 4xy .
dx
2x2 + 2y
dy = O and
dx
+ 3x2)dx + (2y + 2x2)dy = O.

+ 2x2y + y2 = c we find c = 4, and
Separating variables we get (4xy
= O and y = -2 in x3
(b) Setting x
solving for dy/dx
setting x
= y = 1 we also
find c = 4. Thus, both initial conditions determine the same implicit solution.
(e) Solving x3
+ 2x2y + y2 = 4 for y
,.y
we get
and
40. To see that the equations are not equivalent consider dx
.t(x, y)
=y
resulting in y dx
+ x dy = O. A solution
= (x/y)dy
= O. An integrating factor is
of the latter equation is y
= O, but this is not a
solution of the original equation.

41. The explicit solution is y = .r-(3-+-c-o-s2-x-)-/-(-1---x-2-).
Since 3 + cos2 X
> O for al! x we must have
> O or -1 < x < 1. Thus, the interval of definition is (-1, 1).

(a) Since fy = N(x, y) = xexy + 2xy + l/x we obtain f = eXY + xy2 + JI... + h(x) so that fx
1 - x2
42.
y
yexy + y2 - 2
(b) Since fx
+ h'(x).
= M(x, y)
so that fy
Let M(x, y)
+x
(x2
= yl/2x-l/2
= y-l/2xl/2
y
yexy + y2 - 2'
+ y r1
+ ~ (x2 + yr1 + g'(x).
we obtain f
= 2yl/2xl/2 + ~ In Ix2 + yl + g(y)
= y-l/2xl/2
Let N(x,y)
+ ~ (x2 +yr1.
43. First note that
d( /x2 + y2) =
x
/x2
Then x dx
+ y dy
= /
x2
+ y2 dx
x
+ y2
dx
+
/ x2
The left side is the total differential of / x2

Thus / x2
+ y2 = X + c is a
dx
y
'x2
+ y2
dy.
becomes
-===
/ x2
+ y2
+ y2
+ y2
dy
= d( V/ x2 + y2 )
= dx.
and the right side is the total differential of x
solution of the differential equation.
44. To see that the statement is true, write the separable equaton as -g(x)
M = -g(x)
and N
= l/h(y),
+ c.
dx
+ dy/h(y).
Identifying
we see that l\l/y = O = Nx, so the differential equation is exacto
52
Exercises 2.5
Exercises 2.5
l. Letting y = ux we have
+ x(udx + xdu) = O
(x - ux)dx
+ xdu
dx
dx
x
- + du
In Ixl
=O
=O
+u =e
x In Ixl
+y
= ex.
2. Letting y = ux we have
(x
+ ux)
dx
+ x(udx + xdu) = O
+ 2u) dx + x du = O
(1
dx
du
-+--=0
x
1 + 2u
1
In [z] + 2 In 11+ 2ul = e
(1 + 2~) =
x2
x2
3. Letting x
+ 2xy
el
= el.
= vy we have
vy(v dy
+ y dv) + (y
- 2vy) dy
+ (v2
+ 1)
vy dv
- 2v
+ dy = O
vdv
(vIn Iv In
1.::11y
1)2
11- --v -
dy
+ In Iyl = e
x/y - 1
(x - y) In Ix -
53
+ In y
yl -
=e
y = e(x - y).
Exercses
4.. Letting x
2.5
vy we have
y(vdy
+ydv)
- 2(vy
y dv - (v
+ y)
dy
+ 2)
dy = O
~_dy=O
v +2
=O
In Iv + 21- In Iyl = C
In
I~
+ 2-ln Iyl = c
x
5. Letting
+ 2y
= cly2.
= ux we have
(U2X2 + ux2)
+ xdu) = O
dx - x2(udx
u2 dx - xdu = O
1
=c
u
x
ln ]r] + - = c
y
In Ixl
y In
6. Letting
y = ux
+-
Ixl + x
= cy.
we have
(U2X2 + ux2)
dx
+ x2(u
(u2
dx
+ 2u)
dx
dx
-+
x
1
In [z] + 2'ln
lul -
+ x du) = O
+ x du = O
du
=0
u(u+2)
1
2'ln
lu + 21 = c
x2u
--=Cl
u+2
2Y = ci
X ;;
54
(y
;;
+2
Exercises 2.5
7. Letting y
= ux we have
+ x) (u dx + x du)
(ux - x) dx - (ux
+ 1)
(u2
+ x( u + 1) du = O
dx
dx
In Ixl + ~ In
=O
u + 1 du = O
u2 + 1
(u2 + 1) + tan-l u = C
(x
+ 3ux)
dx - (3x
+ ux) (u dx + x du) = O
(u2
1) dx + x(u + 3) du = O
dx
x
u +3
(u-1)(u+1)
In Ixl + 21n lu -
11 -
du
=O
In lu + 11 = C
x(u - 1)2
u+1
Cl
(y - x)2 = Cl(Y
+ x).
-uxdx
+ (x + y'ux)(udx + xdu) = O
(x
+ xy'u)
(u-3/2
du
+ u3/2
dx = O
+ ~) du + d:
=O
_2u-l/2 + In lul + In Ixl = C

In Iy/xl
+ In Ixl =
2NY +
y(1n Iyl - c)2 = 4x.
55
Exercises 2.5
10. Letting
y = ux
we have
(ux
+ Vx2 + u2x2
) dx - x(udx
+ x du) = O
xVI +'u2dx
- x2du = O
dx
du
-x
In JxJ - In
VI + u2
lu + VI + u
u + VI + u
=0
=e
2 =
11. Letting
y = ux
we have
(X3 - u3x3) dx
+ u2x3(udx + xdu)
=O
+ u2xdu = O
dx
dx
2
- +u du
In JxJ
12. Letting
= 2 we find el
y = ux
=O
+ "3u3 = e
3x31n JxJ
Using y(l)
elx
+ y3
elx3.
= 8. The solution of the initial-value problem is 3x31n JxJ
+ y3 = 8x3.
we have
(X2
+ 2u2x2)
dx - ux2(u dx
(1 + u2)
+ x du) = O
dx - ux du =
dx
udu
- 1 +u2
=O
+ u2)
=e
In IxJ - ~ In (1
x2
1 + u2 = el
x4
Using y( -1) = 1 we find el
= 1/2. The
= el
(y2
+ x2)
solution of the initial-value problem is 2x4
56
= y2 + x2.
Exercses 2.5
dx - z e" du = O
dx
ud u=O
--e
x
In Ixl-
eU = e
In Ixl - eY/x = e.
Using y(l)
= O we find
14. Letting x
= vy we have
e = -l. The solution of the initial-value problem is In Ixl
y( v dy
+ y dv) + vy (1n vy
eY/x - l.
- In y - 1) dy = O
+ v In u dy
y dv
=O
~+dy=O
vlnv
y
In 11nIvll
+ In Iyl = e
I~I
= el.
yln
Using y(l)
15. From y
= e we find el = -e.
1
+ -y1 = -yx
x3w = x3
+ e or
and w
The solution ofthe initial-value problem is Ylnl~1
= y 3 we obtain.
y-l we obtain dw
dx
eXw = _~e2x + e or y-l = _~ex + ce:",
e-3x
+y
= xy
+w
= _ex.
and w = y- 3 we obtain -dw - 3w = -3x.
= xe-3x + --e-3x + e or y-3 = x + --
18. From y' -
3
3
3
+ -w
= -.
An integrating factor is x so that
dw
dx
y3 = 1 + cx-3.
16. From y' - Y = eXy2 and w
17. From y I
dx
+ ee3x.
(1 + 2:.)
y = y2 and w = y-l we obtain
x
xe" so that xexw = =xe"
+ eX + e or
1
1 2
19. From y - -y = --2Y
and w
I
= -e.
= y-
y-
= -1 +
dw
dx
An integrating factor is e- 3z so that
(1 + 2:.)
w = -l. An integrating
x
factor is
e
-1 + _e-x.
x
dw
we obtain dt
57
An integrating factor is eX so that
+ -w
t
-2'
An integrating factor is t so that
Exercises 2.5
tw
t
Y
= In t + e or y- = -In t + - _ Writing this in the form - = In x + e, we see that the solution
can also be expressed in the form

,2
20. From y + 3 (
1+ t
2) Y =
et/y
= elX.
2t
4
dw
2t
- 2t
.
2) Y and w = y-3 we obtain - - --2 w = --2
. An mtegrating
3 1+ t
dt
1+ t
1+ t
(
. --21 so that --2w = --21 + e or y-3 = 1 + e ( 1 + t2 ) .

factor IS
l+t
l+t
l+t
,2
3 4
3
. dw
6
9
.
6
21. From y - -y = -2Y and w = y- we obtam - + -w = --2 . An integratmg factor is x so that
x
x
dxx
x
X6W = _*x5 + e or y-3
22. From y' + y
= _*x-l + ex-6. If y(l)
1 then
e=
~ and y-3 = _*x-l + ~x-6.
= y-l/2 and w = y3/2 we obtain dw + ~w = ~. An integrating factor is e3x/2 so that
dx
2
2
e3x/2w = e3x/2 + e or y3/2 = 1 + ee-3x/2. If y(O) = 4 then e = 7 and y3/2 = 1 + 7e-3x/2.
du
= x + y + 1 so that du/dx = 1 + dy/dx. Then - - 1 = u or --2

du = dx. Thus
dx
1 +u
tan-l u = x + e or u = tan(x + e), and x + y + 1 = tan(x + e) or y = tan(x + e) - x - 1.
23. Let u
24. Let u
or u2
du
dx
1-u
or udu = dx. Thus !u2
u
= x + y so that duf d = 1 + dy/dx. Then - - 1 = -= 2x
= 2x +
+ el, and (x + y)2
25. Let u = x + y so that du/dx
el.
du
2
2
Then dx - 1 = tan u or cos udu = dx. Thus
= 1 + dy/dx_
~u+i sin 2u = x+e or 2u+sin2u = 4x+el, and 2(x+y)+sin

2x
=x+e
2(x+y) = 4X+Cl or 2y+sin 2(x+y) =
+ Cl.
26. Let u = x + y so that du/dx
du
dx
. u )/(1 - sin
in uu) we h ave
by (1 - sin
tan u - sec u
=x
= 1 + dy/dx. Then - - 1 = sm u or
= dx. Multiplying
1 - sin
2 ud u = dx or (2sec u - tan u sec u ) du = dx. Th us
cos u
+ C or tan( x + y) - sec(x + y)
27. Let u = y - 2x + 3 so that duf d
1
.
du
1 + smu
=x
+ c.
du
= dy/dx - 2. Then dx + 2 = 2 +
Fu or
1
yU du = dx. Thus
2yU = x + C and 2y1y - 2x + 3 = x + e_

28. Let u = y - x + 5 so that du/dx
_e-u
29. Let u
du
= dy/dx - 1. Then dx + 1
1 + en or e-udu
= dx. Thus
= x + C and _ey-x+5 = x + e.
=x+
y so that duf d
du
= 1 + dy/dx. Then dx - 1 = cos u and
1
1 + cos u
1 - cos u
1 - cos U
= 1 _ cos2 U = sin2 u
58
= ese
1 + cosu
2
u - ese u cot u
du = dx. Now
Exercises
so we have J(csc2 u-cscucotu)du
x + c. Setting x
= O and
= J dx
and - cot u--csc u
= 7r / 4 we obtain
= v2 -
= x+c.
= 3x+2y
so that duf dx
du
Then -d
x
= 3+2dy/dx.
h- 1.
= 3+
Now
u+ 2
5u + 6
1
5
--=-+--so we have
J (~+
5
and
tu + .g ln 125u+ 301 = x + c. Thus
25u
2u
-u +2
5u + 6
u +2
and --6 du
u +2
5u+
= --
+ 30
) du = dx
1
4
5(3x + 2y) + 25 ln 175x + 50y + 301 = x +
Setting x = -1 and y
= -1 we obtain e = t ln 95.
= dx.
25u
+ 30
1. The solution is
csc(x + y) - cot(x + y) = x +
30.Let u
Thus - cot(x+y)+csc(x+y)
2.5
c.
The solution is
144
5(3x + 2y) + 25 ln 175x + 50y + 301 = x + 5ln 95

or
5y - 5x + 2ln 175x + 50y + 301 = 10ln95.

31.
We write the differential equation M(x, y)dx + N(x, y)dy
= O as dy/dx = f(x,
y) where
M(x,y)
f(x, y)
= - N(x, y) .
The function f(x, y) must necessarily be homogeneous of degree O when M and N are homogeneous
of degree a. Since Mis homogeneous of degree a, M(tx, ty) = tO: M(x, y), and letting t
= l/x we
have
1
M(l, y/x) = - M(x, y)
xO:
or
M(x, y)
= xO:M(l,
y/x).
Thus
dy = f(x y) = _ xO:M(l,y/x)
dx
'
xO:N(l, y/x)
= _ M(l,y/x)
N(l,yjx)
= F (J!._).
x
To show that the differential equation also has the form
~~ = G (~)
we use the fact that M(x, y) = yO:M(x/y,
the substitutions u
1). The forms F(y/x)
= y/x or y = ux and v = x/y or x = vy.

59
and G(x/y)
suggest, respectively,
Exercises
32. As x
2.5
----+. -00,
Then, as x
e6x
----+ 00,
----+
O and y
e-6x
33. (a) The substitutions
----+
----+
2x
O and y
= Yl
+u
+ 3.
----+
Now write (1 + eX)/(l
- ee6x) as (e-6x
+ e)/(e-6x
e).
2x - 3.
and
dy
dx
dYl
dx
-=-+-
du
dx
lead to
dYl
du
-;;
+ dx
=P
+ Q(Yl + u) + R(Yl + u)
=P
+ QYl + RYf + Qu + 2YIRu + Ru2
or
du
dx -(Q+2YlR)u=Ru.
This is a Bernoulli equation with n = 2 which can be reduced to the linear equation
dw
dx
+ (Q + 2YlR)w =
-R
by the substitution w = u-l.

(b) Identify P(x) = -4/x2,
Q(x) = -l/x,
and R(x) = 1. Then dw

dx
+ (_2.x +~)
w = -1.
x
+ e or u = [-~x + ex-3
integrating factor is x3 so that x3w = _~x4
1.
An
Thus, y = ~ +u.
34. Write the differential equation in the form x(Y' / y) = In x + In y and let u = In y. Then du] dx = y' / y
and the differential equation becomes x(du/dx)
= ln z
+u
or duf d - u/x = (lnx)/x,
which is
first-order, linear. An integrating factor is e- J dx] = l/x, so that (using integration by parts)
!!._ [2.u]
= ln
dx x
and
x2
~ =_
2. _ ln x
The solution is
In y = -1 - ln x
+ ex
+ e.
e-cx-1
or
= --
Exercises 2.6
1. We identify (x, y) = 2x - 3y
+ 1. Then,
_
for h = 0.1,
Yn+l = Yn + 0.1(2xn - 3Yn + 1) = 0.2xn
+ 0.7Yn + 0.1,
and
y(l.1) ~ Yl = 0.2(1)
+ 0.7(5) + 0.1 = 3.8
y(1.2) ~ Y2 = 0.2(1.1)
+ 0.7(3.8) + 0.1 = 2.98.
60
Exercises 2.6
For h = 0.05
Yn+l = Yn
+ 0.05(2xn
- 3Yn
+ 1) = O.lxn + 0.85Yn + 0.1,
and
y(1.05) ;:::::
Yl = 0.1(1)
y(1.1) ;:::::
Y2
+ 0.85(5) + 0.1 = 4.4
= 0.1(1.05) + 0.85(4.4) + 0.1 = 3.895
y(1.15) ;:::::
Y3 = 0.1(1.1)
y(1.2) ;:::::
Y4
2. We identify f(x, y) = x
+ y2.
+ 0.85(3.895) + 0.1 = 3.47075
= 0.1(1.15) + 0.85(3.47075) + 0.1 = 3.11514
Then, for h = 0.1,
and
y(O.l) ;:::::
Yl
= 0.1(0) + O + 0.1(0)2 = O
y(0.2) ;:::::
Y2
= 0.1(0.1) + O + 0.1(0)2 = 0.01.
For h = 0.05
Yn+l
= Yn + 0.05(xn + y~) = 0.05xn + Yn + 0.05y~,
and
y(0.05) ;:::::
Yl
= 0.D5(0) + O+ 0.05(0)2 = O
y(O.l) ;:::::
Y2 = 0.05(0.05)
y(0.15) ;:::::
Y3
y(0.2) ;:::::
Y4
+ O+ 0.05(0)2 = 0.0025
= 0.05(0.1) + 0.0025 + 0.05(0.0025)2 = 0.0075

= 0.05(0.15) + 0.0075 + 0.05(0.0075)2 = 0.0150.
3. Separating variables and integrating, we have

dy
= dx
y
and
In Iy I
=x
+ c.
Thus y = Clex and, using y(O) = 1, we find c = 1, so y = e" is the solution of the initial-value
problem.
61-
Exercises 2.6
h-O
- 1
h-O
- 05
xn
Yn
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1. 00
1.0000
1.1000
1.2100
1.3310
1.4641
1.6105
1.7716
1.9487
2.1436
2.3579
2.5937
True
Value
1.0000
1.1052
1.2214
1.3499
1.4918
1.6487
1.8221
2.0138
2.2255
2.4596
2.7183
Abs.
Error
% Rel.
0.0000
0.0052
0.0114
0.0189
0.0277
0.0382
0.0506
0.0650
0.0820
0.1017
0.1245
0.00
0.47
0.93
1.40
1.86
2.32
2.77
3.23
3.68
4.13
4.58
Error
xn
Yn
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.0000
1.0500
1.1025
1.1576
1.2155
1.2763
1.3401
1.4071
1.4775
1.5513
1.6289
1.7103
1.7959
1.8856
1.9799
2.0789
2.1829
2.2920
2.4066
2.5270
2.6533
True
Value
1.0000
1.0513
1.1052
1.1618
1.2214
1.2840
1.3499
1.4191
1.4918
1.5683
1.6487
1.7333
1.8221
1.9155
2.0138
2.1170
2.2255
2.3396
2.4596
2.5857
2.7183
Abs.
Error
% Re1.
0.0000
0.0013
0.0027
0.0042
0.0059
0.0077
0.0098
0.0120
0.0144
0.0170
0.0198
0.0229
0.0263
0.0299
0.0338
0.0381
0.0427
0.0476
0.0530
0.0588
0.0650
0.00
0.12
0.24
0.36
0.48
0.60
0.72
0.84
0.96
1.08
1. 20
1.32
1.44
1.56
1.68
1.80
1.92
2.04
2.15
2.27
2.39
Error
4. Separating variables and integrating, we have
dy
y
Thus y = qeX
and, using y(l)
= 2xdx
and
In Iyl = x2
= 1, we find c = e-1,
+ c.
2
so y = e" -lis the solution of the initial-value
problem.
h-O
- 1
h=O 05
xn
Yn
1.00
1.10
1.20
1.30
1.40
1.50
1.0000
1.2000
1.4640
1.8154
2.2874
2.9278
True
Value
1.0000
1.2337
1.5527
1.9937
2.6117
3.4903
Abs.
Error
% Rel.
Error
xn
Yn
0.0000
0.0337
0.0887
0.1784
0.3243
0.5625
0.00
2.73
5.71
8.95
12.42
16.12
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.45
1.50
1.0000
1.1000
1.2155
1.3492
1.5044
1.6849
1.8955
2.1419
2.4311
2.7714
3.1733
62
True
Va1ue
1.0000
1.1079
1.2337
1.3806
1.5527
1.7551
1.9937
2.2762
2.6117
3.0117
3.4903
Abs.
Error
% Rel.
0.0000
0.0079
0.0182
0.0314
0.0483
0.0702
0.0982
0.1343
0.1806
0.2403
0.3171
0.00
0.72
1.47
2.27
3.11
4.00
4.93
5.90
6.92
7.98
9.08
Error
Exercises 2.6
h-O
- 05
h=O.l
5.
h=O.l
Xn
Yn
xn
Yn
0.00
0.10
0.20
0.30
0.40
0.50
0.0000
0.1000
0.1905
0.2731
0.3492
0.4198
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.0000
0.0500
0.0976
0.1429
0.1863
0.2278
0.2676
0.3058
0.3427
0.3782
0.4124
h=O.l
h=O 05
Xn
Yn
Xn
Yn
0.00
0.10
0.20
0.30
0.40
0.50
0.5000
0.5250
0.5431
0.5548
0.5613
0.5639
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.5000
0.5125
0.5232
0.5322
0.5395
0.5452
0.5496
0.5527
0.5547
0.5559
0.5565
7.
h=O.l
9.
6.
Xn
Yn
Xn
Yn
1.0000
1.0000
1.0191
1.0588
1.1231
1.2194
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.45
1.50
1.0000
1.0000
1.0049
1.0147
1.0298
1.0506
1.0775
1.1115
1.1538
1.2057
1.2696
10.
I 1
Eul e r
~
-10
1.0000
1.0500
1.1053
1.1668
1.2360
1.3144
1.4039
1.5070
1.6267
1.7670
1.9332
h=0.05
Xn
Yn
Xn
Yn
0.00
0.10
0.20
0.30
0.40
0.50
1.0000
1.1000
1.2159
1.3505
1.5072
1.6902
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
1.0000
1.0500
1.1039
1.1619
1.2245
1.2921
1.3651
1.4440
1.5293
1.6217
1.7219
h-O
- 05
Xn
Yn
Xn
Yn
0.00
0.10
0.20
0.30
0.40
0.50
0.5000
0.5250
0.5499
0.5747
0.5991
0.6231
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.5000
0.5125
0.5250
0.5375
0.5499
0.5623
0.5746
0.5868
0.5989
0.6109
0.6228
ARUnge-KUlta
f\
\
,
!.
h;O.1
A RungC.Kjlt.
Yn
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
1\
Xn
1.0000
1.1000
1.2220
1.3753
1.5735
1.8371
h=O.l
1.00
1.10
1.20
1.30
1.40
1.50
. 1
Yn
0.00
0.10
0.20
0.30
0.40
0.50
h=O.l
8.
h=O 05
11.
h=0.05
Xn
Eulcr
\.
10
63
h=O.05
~RUngC-Kutta
~uler
5
Exercises 2.6
12.
h=O.25
h=O.1
h=O.05
Run g e-Kut t a
Runge-Kutta
Euler
Euler
Eul e r
-2
-2
13. Using separation of variables we find that the solution of the differential equation is y = 1/(1- x2),
which is undefined at x = 1, where the graph has a vertical asymptote.
h=O 1
Euler
h=O 05
Euler
h-O
1
R-K
h=O 05
R-K
Xo
yo
Xo
yo
Xo
yo
Xo
Yo
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1. 00
1.0000
1.0000
1. 0200
1. 0616
1.1292
1. 2313
1. 3829
1. 6123
1. 9763
2.6012
3.8191
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1. 00
1.0000
1.0000
1.0050
1. 0151
1. 0306
1.0518
1. 0795
1.1144
1.1579
1.2115
1.2776
1.3592
1.4608
1. 5888
1.7529
1. 9679
2.2584
2.6664
3.2708
4.2336
5.9363
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1. 00
1.0000
1.0101
1.0417
1.0989
1.1905
1. 3333
1. 5625
1.9607
2.7771
5.2388
42.9931
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1. 00
1.0000
1.0025
1.0101
1.0230
1.0417
1.0667
1.0989
1.1396
1.1905
1.2539
1. 3333
1.4337
1. 5625
1.7316
1.9608
2.2857
2.7777
3.6034
5.2609
10.1973
84.0132
Because the actual solution of the differential equation becomes unbounded at x approaches 1, very
small changes in the inputs x will result in large changes in the corresponding outputs y. This can
be expected to have a serious effect on numerical procedures.
64
Chapter 2 Revew Exercses

14. (a)
0.6
.._______
--~.~~
-01
"J4
.o
(b) For y
+ 2xy
= 1 an integrating factor is e
Y = e-x 210X eX22ft

dx
+ ce-x = -
If y(O) = Othen e = Oand y = ft
(e) Using FindRoot in Mathematica
2
d
2
2
x x = e" so that - [eX y 1 = eX and
dx
J2 d
erf (x)
+ ce-x. 2
erf (x).
we see that the solution has a relative minimum at
(-0.924139, -0.541044) and a relative maximum at (0.924139,0.541044).

1. Writing the differential equation in the form
is a repeller for k
> O and an attractor for
y' = k(y + Aj k) we see that the critical point - Aj k
< O.
2. Separating variables and integrating we have

dy
y
i dx
x
In y = 41n x
We see that when x = O, y
= O, so the
+ e = In x4 + e
initial-value problem has an infinite number of solutions for
k = Oand no solutions for k =1 O.

3. dy = (y _ 1)2(y _ 3)2
dx
4. dy = y(y _ 2)2(y - 4)
dx
65

x
5.
x even
For dx/dt = xn,when
ti
is even, Ois serni-stable; when n is odd, Ois unstable. For daf d: = -xn,when
n is even, O is serni-stable; when n is odd, O is stable.

6. The zero of f occurs at approximately 1.3. Since P'(t)
for P
> 1.3, limt_oo P(t)

y
7. ,._,~"",
----"""
-.",
_.....................
'
= 1.3.
"",.,_"
"",,---.",
,,_
...
_... ..._
----"\,,, "",,--_.
,.-
....."",
","
......
- ...1
8. (a) linear in y, homogeneous, exact

(b) linear in x
(e) separable, exact, linear in x and y

(d) Bernoulli in x
(e) separable
(f) separable, linear in x, Bernoulli
(g) linear in x
(h) homogeneous
(i) Bernoulli
(j) homogeneous, exact, Bernoulli
(k) linear in x and y, exact, separable, homogeneous
(1) exact, linear in y

(m) homogeneous
(n) separable
66
= f(P) > Ofor P < 1.3 and P'(t) = f(P) > O

cos x dx =
Y
-2--
y +1
dy
1 + -1sin 2x = -In
1
-x
2
4
2
===}
(2y
+ 1 + c
===}
(2
2x + sin 2x = 21n y + 1 + c.
10. Write the differential equation in the form y In; dx = (x In ; - y) dy. This is a homogeneous
equation, so let x = uy. Then dx = u dy + y du and the differential equation becomes
ylnu(udy+ydu)=(uylnu-y)dy
or
ylnudu=-dy.
Separating variables we obtain

lnudu=_dy
===}
ulnlul-u=-lnlyl+c
===}
x(lnx -lny)
===}
:::'lnl:::.I-:::'=-ln1Y1+c
y
y
y
- x = -yln Iyl + cy.
. 1 equation
.
2 IS Bernou 11'
U'smg w
11. The diff
1 erentia
-dy + --y 2
---y 3x2
1.
dx
6x + 1
6x + 1
dw
6
9x2
-d + -6--w
= ---.
An integrating factor is 6x + 1, so
x
x+1
6x + 1
d
2
[(6x + l)w] 9x
dx
- -
===}
3x3
c
w +
- - 6x + 1 6x + 1
===}
(6x + 1)y3
y 3 we obtai
tam
== -3x3 + c.
(Note: The differential equation is also exact.)

12. Write the differential equation in the form (3y2 + 2x)dx + (4y2 + 6xy)dy = O. Letting M = 3y2 + 2x
and N = 4y2 + 6xy we see that My = 6y = N so the differential equation is exacto From
Ix
= 3y2 + 2x we obtain
so h(y) =
3.
= 3xy2 + x2 + h(y). Then
Iy
= 6xy + h'(y) = 4y2 + 6xy and h'(y) = 4y2
The general solution is
13. Write the equation in the form

1
3
-dQ + -Q
= t lnt.
dt
An integrating factor is e1n t
= t,
so
1
dt [tQ] = t In t
===}
tQ = - 25 t + 5t In t + c
===}
Q = - 25 t + 5t In t + t'
14. Letting u = 2x + Y + 1 we have

du _ 2 dy
dx - + dx'
67

and so the given differential equation is transformed into
du
dx
2u
+1
u
Separating variables and integrating we get

u
--du=dx
2u + 1
du = dx
(~_~_1_)
2 2 2u + 1
1
1
"2u- ln 12u + 11 = x
+c
2u - ln 12u + 11 = 2x
+ cj ,
Resubstituting for u gives the solution

4x
+ 2y + 2 -
ln 14x + 2y
+ 31 = 2x + CI
2x
+ 2y + 2 -
ln 14x + 2y
+ 31 = CI.
or
dy
15. Write the equation in the form -d
x
8x
+ zrr:
x +4
2x
x +4
= -2--.
An integrating factor is
x2
+ 4)4 ,so
= 2r2 cos e sin e + r cos e and N = 4r + sin e - 2r cos2 e we see that M; = 4r cos e sin e +
= Ne so the differential equation is exacto From fe = 2r2 cos e sin e + r cos e we obtain
f = _r2cos2e + r sin ' + h(r). Then fr = -2rcos2e + sine + h'(r) = 4r + sine - 2rcos2e and
h'(r) = 4r so h(r) = 2r2. The general solution is
16. Letting M
cos e
_r2 cos2 e + r sin () + 2r2 = c.

d
17. The differential equation has the form dx [(sinx)y] =
y
c/sinx.
7r < x
The initial condition implies c
< 27ris chosen to include x
o.
Integrating we have (sinx)y = c or
= -2sin(71r/6)
= 77r/6.
68
= l. Thus, y = 1/sinx,
where

18. Separating variables and integrating we have
d; = _ 2 (t + 1) dt
y
1
-- = -(t
+ 1) 2 + e
The initial condition implies e = -9, so the solution of the initial-value problem is
1
y
t2
+ 2t -
where
< t < 2.
- 4
19. (a) For y < O, y'Y is not a real number.

(b) Separating variables and integrating we have
dy
= dx
y'Y
Letting y(xo) = YO we get e = 2#0

2y'Y = x
Since y'Y> O for y
-=1=
+ 250
and
- xo, so that
- Xo
and
1
= 4(x
+ 2#0
- xo) .
= ~(x + 2#0 - xo) must be positive. Thus, the
O, we see that dy/dx
interval on which the solution is defined is (xo - 2#0,

20. (a) The differential equation is homogeneous and we let y
(x2 - y2) dx
(x2 - u2x2) dx
+ c.
2y'Y = x
+ xy
dy
(0).
= ux.
=
Then
+ ux2(udx + x du) = O
dx
+ uxdu = O
dx
udu =-x
1
'2u2 = ln jxj + e
=
y2
2" = - 2ln j x j
+ q.
The initial condition gives q = 2, so an implicit solution is y2 = x2(2 - 2ln jxj).
69
(b) Solving for y in part (a) and being sure that the initial condition is still satisfied, we have y = -V2lxl(l
y
2
- In Ixl)1/2, where
-e:::; x:::; e so that 1-1n [z] ;:::O. The graph ofthis function indicates that the derivative is not defined at x = Oand x = e. Thus,
the solution of the initial-value problem is y = -V2 x(1-1n x)1/2,

for O < x
-2
< e.
21. The graph of Yl (x) is the portion of the closed black curve lying in the fourth quadrant. Its interval
of definition is approximately (0.7,4.3). The graph of Y2(X)
is the portion of the left-hand black
curve lying in the third quadrant. Its interval of definition is (-00, O).
22. The first step of Euler's method gives y(1.1) ~ 9 + 0.1(1
more time gives y(1.2) ~ 9.4 + 0.1(1
+ l.lV9.4) ~
70
+ 3) = 9.4.
9.8373.
Applying Euler's method one
Modeling with First-Order

Differential
Equations
Exercises 3.1
1. Let
P = P(t) be the population at time t, and Po the initiaI population. From dP/dt = kP we
Using P(5) = 2Po we find k = ~ In 2 and P = Poe(ln2)t/5. Setting P(t) = 3Po
obtain P = pot.
we have
3 = e(ln2)t/5
==:::}
In3 = (1n2)t
5
==:::}
t = 5In3 ::::::
7.9 years.
In 2
Setting P(t) = 4Po we have

4 = e(ln2)t/5
2. Setting P =10,000 and t
=3
==:::}
In4 = (ln2)t
5
==:::}
t = 10 years.
in ProbIem 1 we obtain
10,000 = Poe(ln2)3/5
==:::}
Po = 10,000e-O.61n2::::::
6597.5.
Then P(10) = Poe21n2 = 4Po ::::::

26,390.
P = P(t) be the population at time t. From dP/dt = kt and P(O) = Po = 500 we obtain
P = 500t. Usng P(10) = 575 we find k = 1~In 1.15. Then P(30) = 500e31n1.15::::::
760 years ..
3. Let
N = N(t) be the number ofbacteria at time t and No the initiaI number. From dN/dt = kN we
obtain N = Not. Using N(3) = 400 and N(10) = 2000 we find 400 = Noe3k or = (400/No)1/3.
From N(10) = 2000 we then have
4. Let
2000 = y, e10k = y, (400)10/3

O
O No
==:::}
2000 = y,-7/3
40010/3
o
==:::}
y, =
O
= l(t) be the intensity, t the thickness, and 1(0) = 10. If dl/dt

1 = 10t, k = k lnO.25, and 1(15) = 0.0009810.
5. Let 1
2000 )-3/7 ::::::

201.
40010/3
kl and 1(3) = 0.2510 then
d.Sl d: = t S we obtain S = Soert where S(O) = So.

= 5.75% then S(5) = $6665.45.
(b) If S(t) =$10,000 then t = 12years.
(e) S:::::: $6651.82
6. From
(a) If So = $5000 and r
N(t) be the amount of lead at time t. From d.Nf dt = kN and N(O) = 1 we obtan
ekt. Using N(3.3) = 1/2 we find k =
In 1/2. When 90% of the lead has decayed, 0.1 grams
7. Let N
71
Exercises
3. 1
will remain. Setting N(t)
= 0.1 we have
et(1/3.3)ln(1/2) = 0.1
==}
_t_In ~
3.3
= In 0.1
==}
= 3.3 In/o.1 ;:::j10.96 hours.
In 1 2
8. Let N = N(t) be the amount at time t. From dN/dt = kt and N(O) = 100 we obtain N = 100t.
Using N(6) = 97 we find k = tlnO.97. Then N(24) = 100e(1/6)(ln097)24= 100(0.97)4;:::j88.5 mg.
9. Setting
N(t)
= 50
in Problem 8 we obtain
1
50 = 100e t
==}
kt = In 2'
10. (a) The solution of dA/dt = kA is A(t)

half-Iife T
In 1/2
t = (1/6) In 0.97 ;:::j136.5 hours.
==}
= Aoekt. Letting A = ~Ao and solving for t we obtain the
= -(ln2)/k.
(b ) Since k = -(ln2)/T
we have
A(t) = Aoe-(ln2)t/T = AoTt/T.

(e) Writing gAo
= Ao2-t/T as 2-3 = 2-t/T and solving for t we get t
3T. Thus, an initial
amount Ao wil! decay to kAo in three half lives.

11. Assume that
A = Aot and k = -0.00012378. If A(t) = 0.145Ao then t ;:::j15,600years,
12. From Example 3, the amount of carbon present at time t is A(t) = Aoe-O.00012378t.Letting t = 660
and solving for Ao we have A(660) = Aoe-O.0001237(660)
= 0.921553Ao. Thus, approximately 92%
of the original amount of C-14 remained in the cloth as of 1988.
13. Assume that dT/dt
= k(T-10)
so that T
10 + cekt. IfT(O)
and k = 2In(2/3) so that T(l) = 36.670 IfT(t)

= k(T-5)
and c = 59.4611 so that T(O)
so that T
= 150 then
= 70
and T(1/2) = 50 then c

0
= 60
t = 3.06 minutes.
= 5+cekt. IfT(l) = 550 and T(5) = 300 then k = -:!-ln2
= 64.4611.
k(T - 100) so that T = 100 + ce": If T(O) = 200 and T(l) = 220 then
c = -80 and k = In(39/40) so that T(t) = 90 implies t = 82.1 seconds. If T(t) = 98 then
t = 145.7seconds.
16. Using separation of variables to solve dT/dt = k(T -Tm) we get T(t) = Tm +cekt. Using T(O) = 70
we find c = 70 - Tm, so T(t) = Tm + (70 - Tm)ekt. Using the given observations, we obtain
T(~) = Tm + (70 - Tm)ek/2 = 110

T(l) = Tm + (70 - Tm)
72
145.
Exercises
Then
/2
3. 1
= (110 - Tm)/(70 - Tm) and
ek
= (ek/2)2 = (110-Tm)2
= 145-Tm
70 - Tm
70 - Tm
(110 - Tm)2
70 - Tm
12100 - 220Tm
= 145 _ T
+ T'/'n=
10150 - 250Tm
+ T'/'n
Tm = 390.
The temperature in the oven is 390
0.
17. From dA/dt
4 - A/50 we obtain A = 200 + ce-t/50
If A(O) = 30 then c = -170 and A =
200 - 170e-t/50.
18. Frorn dA/dt
19. Frorn dA/dt
= 0- A/50 we obtain A
= ce-t/5D If A(O)
= 30 then c = 30 and A = 30e-t/50.
= 10 - A/100 we obtain A = 1000 + cct/100.
= 1000 - 1000e-t/100.
If A(O) = O then c = -1000 and
At t = 5, A(5) ;:::;
48.77 pounds.
dA
lOA
2A
= 10 = 10 we obtain A = 1000 - 10t
dt
500-(10-5)t
100-t
A(O) = O then c = -110' The tank is empty in 100 minutes.
20. Frorn
dA
21. From -d =3-
00
4A
2A
(
) =3---weobtainA=50+t+c(50+t)-2.
6- 4 t
50 + t
t)2. If
IfA(0)=10then
e = -100,000 and A(30)
+ c(100 -
= 64.38 pounds.
22. (a) Initially the tank contains 300 gallons of solution.
Since brine is pumped in at arate
of
3 gal/rnin and the solution is pumped out at arate of 2 gal/mn, the net change is an increase
of 1 gal/min. Thus, in 100 minutes the tank will contain its capacity of 400 gallons.
(b) The differential equation describing the amount of salt in the tank is A'(t)
=6
- 2A/ (300 + t)
with solution
A(t) = 600 + 2t - (4.95 x 107)(300
+ t)-2,
O < t :::;100,
as noted in the discussion following Example 5 in the texto Thus, the amount of salt in the
tank when it overflows is
A(100) = 800 - (4.95
107)(400)-2
= 490.625
lbs.
(c) When the tank is overflowing the amount of salt in the tank is governed by the differential
73
Exercses
3. 1
equation
dA
di
= (3 gal/min)(2 lb/gal) - (400 lb/gal)(3 gal/min)
A(100) = 490.625.
Solving the equation we obtain A(t)
= 800
+ ee-3tI400.
The initial condition yields
e = -654.947, so that
A(t) = 800 - 654.947e-3tI400
When t = 150, A(150) = 587.37 lbs.
(d) As t
-4
00, the
amount of salt is 800 lbs, which is to be expected since
(400 gal)(2lbs/gal)=
(e)
800 lbs.
8oor---------~~~---600
~----2-00~--4-0~0----6-0-0--t
23. Assume Ldi/dt
+ Ri =
E(t), L
then e = -3/5 and limt~ooi(t)

24. Assume Ldi/dt
+ Ri = E(t),
,;=
'(0)' =
Smee
2
20
we obtai
tain e
= 0.1, R = 50, and E(t) = 50 so that i = ~ + ee-500t. If i(O) = O

= 3/5.
= Eosinwt, and i(O) = io so that
E(t)
EoR
. t
EoLw
t+
-RtIL
L2w2 + R2 smw - L2w2 + R2 cosw
ce
EoLw
= 20. + L2w2
+ R2
25. Assume Riiql dt. + (l/e)q = E(t), R = 200, e

If q(O) = Othen e = -1/100 and i = ~e-50t_
26. Assume Rdq/dt+
and i
(l/e)q
= _200ee-200t.
0.1472 amps. As t
-4
= 10-4, and E(t) = 100 so that q = 1/100 + ee-50t.
= 1000, e = 5 x 10-6, and E(t) = 200. Then q = 10100

+ee-200t
0.4 then e = - 560' q(0.005) = 0.003 coulombs, and i(0.005)
= E(t), R
If i(O) =
00
we have q
-4
10~0'
27. For O~ t ~ 20 the differential equation is 20d'i/dt

:t [etllOi]
= 6etl10 and i = 60 + ele-tilO.
+ 2i =
If i(O) = Othen
For t > 20 the differential equation is 20 di/ dt
120. An integrating factor is etilO, so

el
+ 2i = Oand
74
= -60 and i
i = e2e-tl10.
= 60 - 60e-tllO.
Exercises
At t
= 20
we want C2e-2
= 60 - 60e-2 so that
i(t)=
C2 = 60
3. 1
(e2 - 1). Thus
< 20;
O ::s t
t>20.
60 - 60e-t/la
{ 60(e2-1)e~t/1O,

E
dq
dt
a - q /C = k~l + k-2t
=}
Setting q(O) = qa we find
C2
qI
-C In I Ea - C
(Ea - qo/C)-C
1/k2
k1
' so
(Ea - q/C)-C
(kl + k2t)1/k2
+ k2t[ + Cl
-k In [kl
-2
q)-C
( Ea - C
(
=
=}
qa)-C
Ea - C
q
(
qa) ( kl )
Ea - - - Ea -C C
k + k2t
q = EaC
29. (a) From mdv/dt
= mg-
+ (qa -
kv we obtain v = gm/k+ce-kt/m.
EaC)
k)
( k + lk2t
kl
+ k2t
)-1/k
1/ C k2
1/Ck2
Ifv(O) = va then c = va-gm/k
and
the solution of the initial-value problem is

gm
v=T+
(b) As t
-t 00
va-T
gm)
-kt/m
the limiting velocity is gm/ k.
(e) Frorn ds/dt = v and s(O) = Owe obtain

gm - m ( va - -gm) e-ktf
s = -t
k
k
k
m.
+ c.
30. (a) Integrating d2s/dt2 = -g we get v(t) = ds/dt = -gt

the velocity is v(t) = -32t
+ -m ( va -
-gm) .
k
From v(O) = 300 we find c = 300, so
+ 300.
= O we get s(t) = -16t2 + 300t. The maximum height is

attained when v = O,that is, at ta = 9.375. The maximum height will be s(9.375) = 1406.25 ft.
(b) Integrating again and using s(O)
31. When air resistance is proportional to velocity, the model for the velocity is m dv / dt = -mg
kv (using the fact that the positive direction is upward.) Solving the differential equation using
separation of variables we obtain v(t) = -mg/k
v(t) = _ mg
+ ce:"!":
From v(O)
+ (300 + mg)e-kt/m.
k
75
= 300 we get
Exercises
3. 1
Integrating and using s(O) = Owe find

s(t) = _ mg t
k
Setting k
= 0.0025, m = 16/32
+ m (300 + mg)
k
(1 _ e-kt/m).
= 0.5, and 9 = 32 we have
s(t)
1,340,000 - 6,400t - 1,340,000e-O.005t
and
v(t) = -6,400
+ 6,700e-0005t.
The maximum height is attained when v = O, that is, at ta = 9.162. The maximum height will be
s(9.162) = 1363.79 ft, which is less than the maximum height in part (a).
32. Assuming that air resistance is proportional to velocity and the positive direction is downward, the
model for the velocity is m dv / dt = mg - kv. Using separation of variables to solve this differential
equation we obtain v(t) = mg/k
+ ce-kt/m.
From v(O) = O we get v(t)" = (mg/k)(l
- e-kt/m).
= 5, and 9 = 32 we have v(t) = 320(1 - e-o.lt). Integrating,

we find s(t) = 320t + 3200e-o.lt.
At t = 15, when the parachute opens, v(15) = 248.598 and
s(15) = 5514.02. At this point the value of k changes to k = 10 and the new initial velocity
Letting k
0.5, m
160/32
is Vo = 248.598. Her velocity with the parachute open (with time measured from the instant of
= 16 + 232.598e-2t.
opening) is vp(t)
Integrating, we find sp(t) = 16t - 116.29ge-2t.
seconds after leaving the plane is five seconds after the parachute opens.
Twenty
Her velocity at this
time is vp(5) = 16.0106 ft/sec and she has fallen s(15) + sp(5) = 5514.02 + 79.9947 = 5594.01 ft.
Her terminal velocity is limt->oovp(t) = 16, so she has very nearly reached her terminal velocity
five seconds after the parachute opens. When the parachute opens, the distance to the ground is
15,000 - 5514.02
= 9485.98 ft.
Solving sp(t)
= 9485.98 we get t = 592.874 s = 9.88 mino Thus,
it
will take her approximately 9.88 minutes to reach the ground after her parachute has opened and
a total of (592.874 + 15)/60 = 10.13 minutes after she exits the planeo
33. (a) The differential equation is frst-order, linear.
eJ3bdt/(bt+ro) = (ro + bt)3. Then
!((ro
+ bt)3v]
= g(ro
+ bt)3
and
Letting b = k/ p, the integrating factor is
(ro
+ bt)3v
= :b (ro
+ bt)4 + c.
The solution of the differential equation is v( t) = (g/ 4b) (ro + bt) + c(ro + bt) -3. Using v(O) = O
we find c = -gr~/4b,
so that
9
grt
gP(
k )
v(t) = 4b (ro + bt) - 4b(ro + bt)3 = 4k ro + -}
(b) Integratingdr/dt
= k/pweget
= kt/p+c.
Usingr(O)
76
gpr~
4k(ro + kt/ p)3 .
= ro wehavec = ro,
sor(t)
= kt/p+ro
Exercises
3. 1
= 0.007ft when t = lOs, then solving r(lO) = 0.007 for k/p, we obtain k/p = -0.0003 and
(e) Ifr
r(t) = 0.01 - 0.0003t. Solving r(t) = O we get t = 33.3, so the raindrop will have evaporated
completely at 33.3 seconds.
dP
p = k cos t di =>
If P(O)
= Po then
35. (a) Frorn
= k sin t + e ==> P =
= Po and P = po
el ek
sin
t.
sin t:
dP/dt = (kl - k2)P we obtain P = Poe(kl-k2)t where Po = P(O).
(b) If kl > k2 then P

as t
In IP I
-7
00
as t
-7
oo. If
k: = k2 then P = Po for every t. If kl < k2 then P
-7
oo.
-7
36. The first equation can be solved by separation of variables. We obtain x = ele-Alt.
we obtain el
= Xo
From x(O) = xo
and so x = xoe-Alt. The second equation then becomes

or
which is linear. An integrating factor is eA2t. Thus
Frorn y(O) = YO we obtain
e2
(YOA2 - YOAl - XOAl)/(A2
- A1). The solution is
k(M - A) - k2A = Ofor A we find the equilibrium solution A = kM/(k1 + k2)' From
the phase portrait we see that lim,.....oo A(t) = k1M /(kl + k2).
37. (a) Solving
------~-------+------~~-----A
Since k2
> O, the material will never be completely memorized and the larger k2 is, the less
the amount of material will be memorized over time.
77
Exercses 3. 1
(b) Write the differential equation in the form dA/dt+(kl

k1iVI. Then an integrating factor is e(k+k2)t, and
!!_
dt
[e(k+k2)t
Al
=:}
=:}
= k M e(k +k2)t
1
e(k+k2)tA=
A -- k k1Mk
1
UsingA(O)=Owefindc=A_,
+k2)A =
k1M
k1 + k2
2
e(k+k2)t+c
+ ce-(k+k2)t .
k1M andA=
k1 + k2
kliV~ (1_e-(k+k2)t).
k1 + 2
As t _,
00,
k1M .
k1 + k2
38. (a) Solving r - kx = O for x we find the equilibrium solution x = r/k.

and when x
When x
< r/k, dxf d: > O
> r/k, dx/dt < o. From the phase portrait we see that limt-+oox(t) = r/k.
------__------~------~~------x
r
(b) From dxf d: =

so that x_,
r -
kx and x(O) = Owe obtain x = r j k - (r/k)e-kt
r/k as t _, oo. If x(T) = r/2k then T = (ln2)/k.
39. It is necessary to know the air temperature from the time of death until the medical examiner
arrives. We will assume that the temperature of the air is a constant 65F. By Newton's law of
cooling we then have
dT
- 65)
T(O) = 82.
dt
'
Using linearity or separation of variables we obtain T = 65+ct.
From T(O) = 82 we find c = 17, so
- = k(T
that T = 65 + 17ekt. To find k we need more information so we assume that the body temperature
at t = 2 hours was 75F. Then 75 = 65 + 17e2k and k = -0.2653 and
T(t)
= 65 + 17e-O.2653t.
At the time of death, to, T(to) = 98.6F, so 98.6 = 65 + 17e-O.2653t, which gives t = -2.568. Thus,
the murder took place about 2.568 hours prior to the discovery of the body.
78
Exercses
40. We will assume that the temperature
3.2
of both the room and the cream is 72F, and that the
of the coffee when it is first put on the table is 175F. If we let TJ(t) represent the
temperature
temperature of the coffee in Mr. Jones' cup at time t, thcn

dTI
dt
k(TI
72),
which implies TI = 72 + CI t. At time t = O Mr. Jones adds cream to his coffee which immediately
reduces its temperature by an amount a, so that TI (O)
175 - a. Thus 175 - a
= 5, TI (5) = 72 + (103 - a)k

Now we let T2(t) represent the temperature of the coffee in Mrs. Jones' cupo From T2 = 72 + c2t
and T2(0) = 175 we obtain C2 = 103, so that T2(t) = 72 + 103t. At t = 5, T2(5) = 72 + 103e5k.
which implies
CI =
103 - a, so that TI (t) = 72 + (103 - a)t
= TI (O) = 72 + CI,
Whcn cream is added to Mrs. Jones' coffee the temperature
fact that k
At t
is reduced by an amount a. Using the
< O we have
T2(5) - a
72 + 103e5k
< 72 + 103k
ak
= 72 + (103 - a)e5k = TI (5).
Thus, the temperature
of the coffee in Mr. Jones' cup is hotter.
41. Drop an object from a great height and measure its terminal velocity,
that
Vt
= gm/ k,
so k
Vt.
In Problem 29(b) we saw
= gm/vi.
42. We saw in part (a) of Problem 30 that the ascent time is t = 9.375. To find when the cannonball
hit s the ground we solve s(t)

the time of descent is td
-16t2
+ 300t = O, getting
a total time in flight of t
18.75 - 9.375 = 9.375. The impact velocity is
Vi
= 18.75. Thus,
= v(18.75) = -300,
which has the same magnitude as the initial velocity.

We saw in part (b) of Problem 30 that the ascent time in the case of air resistance is
Solving s(t)
= Owe see that
1,340,000- 6,400t-1,340,OOOe-0.005t
Thus, the descent time is td
= 18.466 - 9.162 = 9.304. The impact velocity is
-290.91, compared to an initial velocity of Vo
Vi
When O < N
9.162.
= v(18.466) =
300.
Exercises 3.2
1. (a) Solving N(1 - 0.0005N)
t =
the total time of flight is 18.466.
= O for N we find the equilibrium solutions N = O and N = 2000.
< 2000, dN/dt > O. From the phase portrait we see that 1imi-<oo.N(t)
o
2000
A graph of the solution is shown in part (b).
79
= 2000.
Exercises 3.2
(b) Separating variables and integrating we have
2000
dN
N(l - 0.0005N)
= (~_
= dt
1
)dN
N - 2000
1500
1000
and
500
In N -ln(N
- 2000)
Solving for N we get N(t)
= t + c.
= 2000+t
/(1
10
+ +t) =
2000et /(1
solving for we find = 1/1999 and so N(t) = 2000et/(1999
dN
di =
N(a - bN) and N(O) = 500 we obtain N = 500b
=50,000 and N(l)
3. From
dP
di
= 1000 we have a
= P (10-1
10-7
p)
20t
+ et). Using N(O) = 1 and

+ et). Then N(lO) = 1833.59,
so 1834 cornpanies are expeeted to adopt the new teehnology when t

2. From
15
= 10.
500a
500b)e-at'
+ (a _
= 0.7033, b = 0.00014, and N
Sinee t~~ N =
50,000
= 1 + 9ge-O.7033t .
500
and P(O) = 5000 we obtain P = 0.0005 + 0.0995e-o.lt so that
P __. 1,000,000 as t __. oo. If P(t) = 500,000 then t = 52.9 months.

4. (a) We have dP/dt
= P(a - bP) with P(O) = 3.929 million.
Using separation of variables we
obtain
P
(t) = 3.929b
a/b
1 + (a/3.929b - l)e-at
3.929a
- 3.929b)e-at
+ (a
c
1 + (c/3.929 - l)e-at .
At t = 60(1850) the population is 23.192 million, so
23 92 =
.1
or c = 23.192 + 23.192(c/3.929
1+ (c/3.929 _ 1)e-60a
- 1)e-60a. At t = 120(1910)
c
91 972 = --,-..,------:---=;,-.
1 + (c/3.929 - 1)e-120a
or c = 91.972 + 91.972(c/3.929
- 1)(e-60a)2. Combining the two equations for c we get
23.192)/23.192)2 (_c __
( (c- c/3.929
- 1
3.929
1)
= c - 91.972
91.972
or
91.972(3.929)(c - 23.192)2
(23.192)2(c - 91.972)(c - 3.929).
The solution of this quadratic equation is c = 197.274. This in turn gives a = 0.0313. Therefore
P( )
t
197.274
= 1 + 49.21e-o.0313t
80
Exercises 3.2
dX
150
8. From -d = k(150 - X)2, X(O) = O, and X(5) = 10 we obtain X = 150 k
where
t
150 t + 1
k = .000095238. Then X(20) = 33.3 grams and X -; 150 as t -; 00 so that the amount of A -; O
and the amount of B -; Oas t -; oo. If X (t) = 75 then t = 70 minutes.
9. (a) The initial-value problem is dh/dt
h
10
= -8AhVh/Aw,
h(O) = H. Separating variables and integrating we have
!!:!!._
Vh
= _ 8Ah dt
and
Aw
the initial-value problem is
= (AwVH-
Aw
Using h(O) = H we find e

h(t)
2Vh = _ 8Ah t + c.
~--~~~=-~~-500
= 2VH, so the solution of
lhW = (AwVHfor O:S t
4Aht)2/A;
4Aht)/Aw,
where AwVH-
:s AwH/4Ah
(h) Identifying H = 10, Aw = 47l',and Ah = 7l'/576 we have h(t) = t2/331,776 Solving h(t)
= Owe
4Aht ;:: O. Thus,
(fs72 /144)t
+ 10.
see that the tank empties in 576/10 se~onds or 30.36 minutes.
10. To obtain the solution of this differential equation we use h(t) from part (a) of Problem 11 in
Exercises 11.3 with Ah replaced by cAh. Then h(t) = (AwVH-
4cAht)2/A;.
Solving h(t) = O
with c = 0.6 and the values from Problem 11 we see that the tank ernpties in 3035.79 seconds or
50.6 minutes.
11. (a) Separating variables and integrating we have
6h3/2dh
Using h(O)
20 we find c
= -5t
and
12h5/2
5
= -5t + c.
1920.;5, so the solution of the initial-value problem is h(t) =
(800.;5 - i~t)2/5. Solving h(t) = Owe see that the tank empties in 384.;5 seconds or
14.31 minutes.
(b) When the height of the water is h, the radius of the top of the water is r = h tan 30 = h] J3
and Aw
= 7l'h2/3. The differential equation is

dh
dt
Ah r;;:;:.
7l'(2/12)2
= -cAw V2gh = -0.6 7l'h2 /3 v64h = - 5h3/2 .
Separating variables and integrating we have

5h3/2dh
Using h(O)
= -2 dt and 2h5/2 = -2t + c.
= 9 we find c = 486, so the solution
Solving h(t)
= Owe
of the initial-value problem is h(t)
= (243-t)2/5.
see that the tank empties in 24.3 seconds or 4.05 minutes.
12. When the height of the water is h, the radius of the top of the water is g(20 - h) and Aw =
47l'(20- h)2/25. The differential equation is
dh
dt
=_
Ah r;;:;:.
eAw V2gh
=_
7l'(2/12)2
~4h =
0.647l'(20_ h)2 /25 v 64f1
82
_~
Vh
6 (20 _ h)2 .
Exercises 3.2
(b)
Year
Census
Population
Predicted
Population
1790
1800
1810
1820
1830
1840
1850
1860
1870
1880
1890
1900
1910
1920
1930
1940
1950
3.929
5.308
7.240
9.638
12.866
17.069
23.192
31.433
38.558
50.156
62.948
75.996
91.972
105.711
122.775
131.669
150.697
3.929
5.334
7.222
9.746
13.090
17.475
23.143
30.341
39.272
50.044
62.600
76.666
91.739
107.143
122.140
136.068
148.445
Error
Error
0.000
-0.026
0.Q18
-0.108
-0.224
-0.406
0.049
1.092
-0.714
0.112
0.348
-0.670
0.233
-1.432
0.635
-4.399
2.252
0.00
-0.49
0.24
-1.12
-1.74
-2.38
0.21
3.47
-1.85
0.22
0.55
-0.88
0.25
-1.35
0.52
-3.34
1.49
The modeI predicts a population of 159.0 million for 1960 and 167.8 million for 1970. The
census populations for these years were 179.3 and 203.3, respectively. The percentage errors
are 12.8 and 21.2, respectiveIy.
5. (a) The differential equation is dP/dt = P(l-InP),
solution P
which has equilibrium
= e. When Po > e, dP/dt < O, and when Po < e, dP/dt >
o.
= P(l + In P), which has equilibriurn

soIution P = l/e. When Po > l/e, dP/dt > O, and when Po < l/e,
dP/dt < o.
(b) The differential equation is dP / dt
6. From
then
dP
di
C
= P(a-bInP)
a
= -b - In Ro .
7. Let X = X(t)
we obtain
be the amount of
-1
b
In la-bInPI
e at
= t+Cl so that
P = ea e-ce
time t and dX = k(120 - 2X)(150 - X).
dt
-w
. If P(O) = Po
If X(O) = O and
150 - 150e180kt
X(5) = 10 then X = 1 _ 2.5e180kt where k = .0001259, and X(20) = 29.3 grams. Now X ~ 60
as t -;
00,
so that the amount of A -; O and the amount of B -; 30 as t -; oc.
81
Exercises 3.2
14. We solve
dv
dt
'
m- = -mg-kv
v(O) = 300
using separation of variables. This gives
Integrating and using s(O)
= O we find
s(t) = ; In leos
fi
t - tan-l 300{Jg
I +~
In (1
+ 90~~Ok)
Solving v(t) = O we see that ta = 6.60159. The maximum height is s(ta) = 823.84 ft.
15. (a) Let p be the weight density of the water and V the volume of the object. Archimedes' principle
states that the upward buoyant force has magnitude equal to the weight of the water displaced.
Taking the positive direction to be down, the differential equation is
dv
2
m dt = mg - kv -
pv.
(b) Using separation of variables we have

mdv
= dt
(mg - pV) - kv2
.22!:._
Vk
Vk dv
(Vmg - pV)2 - (Vkv)2
1
Vk --===~
Vmg - pV
tanh
-1
Vkv
Vmg - pV
= dt
=t
+ c.
Thus
v ()t =
(e) Since tanh t
->
1 as t
-> 00,
Jmg - pV
k
tan
h (vkmg
the terminal velocity is
16. (a) Writing the equation in the form (x - Jx2
- kpV
m
t
J (mg -
+ y2 )dx + y dy
+ Cl )
pV) / k .
we identify iVI = x - Jx2
+ y2
and
N = y. Since M and N are both homogeneous of degree 1 we use the substitution y = ux. It
84
Exercises
3.2

(20-h)2
-'---=""':- dh
vh
5
dt
6
and
= --
r: 80
800y h - -h
3
3/2
2 5/2
+ -h
5
5
6
= --t
+ e.
Using h(O) = 20 we find e = 2560v5j3, so an implicit solution of the initial-value problem is
To find the time it takes the tank to empty we set h
= O and solve for
t. The tank empties in
1024v5 seconds or 38.16 minutes. Thus, the tank empties more slowly when the base of the cone
is on the bottom.
13. (a) Separating variables we obtain
mdv
= dt
mg - kv2
~
dv
= dt
9 1- (kvjmg)2
y'ri0
Jkjmgdv
-Vkg 1- (Vkvjy'ri0)2
0itanh-1
V kg
=dt
Vkv = t + e
y'ri0
[g"g
tanh -1 Vkv
;;;:n;:; =
- t + el
ymg
Thus the velocity at time t is
v(t) =
Setting t
= O and
'mg tanh ( Vfkg

VT
-:;;;t+ el )
v = Vo we find el = tanh-1(Vkvojy'ri0).
(b) Since tanh t _, 1 as t _,
00,
we have v _, Jmgjk as t _, oo.
(e) Integrating the expression for v(t) in part (a) we obtain
Setting t = O and s = So we find e2 = So -
ln( cosh el).
83
Exercises
3.2
follows that
(x - Vx2
x
+ u2x2)
[(1 - ,jI +
dx
2)
+ ux(udx + xdu)
+ U2]
dx
=O
+ x211.du=
udu
dx
1 + u2 - ..j 1 + u2
udu
x
dx
..JI + u2 (1 - ..JI + u2
Letting w = 1 - ..JI + u2 we have dw = -u du] ,11 + u2 so that

- In
(1 - VI + u
2 ) = In x
+C
1
--==:o:=CIX
1 - ..JI + u2
r:-:-;;
C2
1 - VI -t- u~ = --
SoIving for y2 we have

y2
= 2C2X+ c~ = 4 (~)
(x
+ ~)
which is a famiIy of paraboIas symmetric with respect to the x-axis with vertex at (-c2/2, O)
and focus at the origino
(b) Let u
x2 + y2 so that
Then
dy
1 du
y-=---x
dx
2 dx
and the differentiaI equation can be written in the form
1 du
- - - x
2 dx
= -x+ yU or ~ du = yU.
2 dx
85
Exercises
3.2

du
2y'u = x
.jU=x+c
17. (a) From 2W2
W3
W2(2 - W) = Owe see that W
= O and
= 2 are constant solutions.
(b) Separating variables and using a CAS to integrate we get

dW
V
W 4-2W
=dx
-tanh-1(-2V4-2W)=x+c.
and
Using the facts that the hyperbolic tangent is an odd function and 1 - tanh2 x
= sech2 x we
have
1
2"V4 -
2W
= tanh( -x - c) = - tanh(x + c)
4(4 - 2W) = tanh2(x + c)

1-
2" W = tanh (x + c)
1
2"W
= 1 - tanh2(x
+ c) = sech2(x + c).
Thus, W(x) = 2sech2(x + c).

(e) Letting x
= O and W = 2 we find that sech2(c) = 1 and c = O.
= 102 for r2 we see that r2 = 20h - h2. Combining the rate of input of
with the rate of output due to evaporation, k11"r2 = k11"(20h - h2), we have dV/dt =
18. (a) Solving r2 + (10 - h)2

water, 11",
11" - k11"(20h - h2).
Using V = lO11"h2
-111"h3,
we see also that dV/dt
= (20nh - 11"h2)dh/dt.
Thus,
and
86
dh
dt
1 - 20kh
+ kh2
20h - h2
Exercises
(b) Letting k = 1/100, separating variables and integrating (with

the help of a CAS), we get
100h(h - 20)
(h _ 10)2 dh = dt
and
100(h2 - lOh + 100)

-~-10---h----'= t + c.
3.2
h
10
8
Using h(O) = O we find c = 1000, and solving for h we get

h(t) = 0.005( Jt2
+ 4000t -
t), where the positive square root is
chosen because h 2: O.
2000
(e) The volume of the tank is V = ~7I(1O)3 feet, so at arate of
71
4000 6000 80001000b
cubic feet per minute, the tank
will fill in ~(10)3 ~ 666.67 minutes ~ 11.11 hours.

(d) At 666.67 minutes, the depth of the water is h(666.67) = 5.486 feet. From the graph in (b) we
suspect that limt_oo h(t) = 10, in which case the tank will never completely fill. To prove this
we compute the limit of h(t):
.
11mh(t) = 0.005 lim
t-oo
t-oo
(yt/ 2 + 4000t -
= 0.005 11m
t-oo
19. (a)
t
u
10
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
170
P(t)
';.n~
5.308
7.240
9.638
12.866
17.069
23.192
31.433
38.558
50.156
62.948
75.996
91.972
105.711
122.775
131.669
150.697
179.300
t2 + 4000t - t2
t) = 0.005 lim -./!=.2;===:===::=::=-----:t-oo v t + 4000t + t
4000t
tJ1
+ 4000/t + t
4000
= 0.005 lirn --
Q (t)
u.O.;:,
0.036
0.033
0.033
0.033
0.036
0.036
0.023
0.030
0.026
0.021
0.021
0.015
0.016
0.007
0.014
0.019
87
t-oo 1
+1
= 0.005(2000)
= 10.
Exercses 3.2
(b) The regression line is Q = 0.0348391 - 0.000168222P.
Q
..
0.02
0.015
0.01
0.005
--'-~2 0:;---'-'4
0:;--::":6
0:;--~80;--:1'"70--:1
7072--:1-:-4::-'""0
70
p
(e) The solution of the logistic equation is given in equation (5) in the texto Identifying a
0.0348391 and b
= 0.000168222 we have
P(t) =
aPo
.
bPo + (a - bpo)cat
(d ) With Po = 3.929 the solution becomes
P( ) _
0.136883
t - 0.000660944 + 0.0341781e-o.034839lt.
(e)
175
150
125
100
75
50
25
75
100
125
150
(f) We identify t = 180 with 1970, t = 190 with 1980, and t = 200 with 1990. The model predicts
P(180) = 188.661, P(190) = 193.735, and P(200) = 197.485. The actual population figures
for these years are 203.303, 226.542, and 248.765 millions. As t -; 00, P(t) -; a/b = 207.102.
20. (a) Using a CAS to solve P(l - P)
+ 0.3e-P = Ofor P
we see that P
= 1.09216 is an equilibrium
solution.
(b) Since f(P)
> Ofor O 1.09216, the

> 1.09216. Thus P = 1.09216 is
solution P(t) is decreasing for Po

an attractor.
-1
-2
88
Exercises 3.2
p
(e) The curves for the second initial-value problem are thicker. The
= 1. Comparing
equilibrium solution for the logic model is P
1.5
1.09216 and 1, we see that the percentage increase is 9.216%.
1D t
21. To find td we solve

m
dv
2
dt = mg - kv )
using separation of variables. This gives
v(t) =
If
=O
v(O)
ffi
tanh
t.
Integrating and using s(O) = Ogives

m In ( cosh
s(t) = k
yfk9)
-;;:;t
To find the time of descent we solve s(t) = 823.84 and find td = 7.77882. The impact velocity is
V(td) = 182.998, which is positive because the positive direction is downward.

22. (a) Solving
Vt
= Jmg/k
for k we obtain k = mg/v'
dv = mg - -2
mg
dt
Vt
m-
The differential equation then becomes

1v
-dv = 9 ( 1 - 2"
or
dt
2)
Vt
Separating variables and integrating gives

Vt tanh
-1 v
= gt
+ ci.
Vt
The initial condition v(O)
= Oimplies
el
= O, so
gt
v(t) = Vt tanh - .
Vt
We find the distance by integrating:
s(t) =
Vt tanh
The initial condition s(O) = Oimplies
gt dt =
V[
Vt
e2
= O, so
In (cosh gt )
+ e2.
Vt
s(t) = V[ In (cosh gt ) .
9
Vt
In 25 seconds she has fallen ~O,OOO

- 14,800 = 5,200 feet. Solving
5200 = (v;/32) In (cosh 32(25))
Vt
89
Exercises 3.2
for
Vt
gives
271.711 ft/s. Then
Vt ~
v2
gt
s(t) = _J_ln(cosh-)
9
= 2307.08 ln(cosh 0.1l7772t).
Vt
(b) At t = 15, s(15) = 2,542.94 ft and v(15) = s'(15) = 256.287 ft/sec.

23. While the object is in the air its velocity is modeled by the linear differential equation mdv/dt
= 160, k
mg - kv. Using m
t, and 9 = 32, the differential equation
becomes dv/dt + (1/640)v
32. The integrating factor is eJ dt/640 = et/640 and the solution of the differential equation is
et/640v = J 32et/640dt
= 20,480et/640 + c. Using v(O) = O we see that e = -20,480 and v(t) =
20,480 - 20,480e-t/640.
Integrating we get s( t) = 20,480t + 13,107,200e-t/640
e = -13,107,200 and s(t) = -13,107,200+20,480t+13,107,200e-t/640.
+ c. Since s(O)
= O,
To find when the object hits
the liquid we solve s(t) = 500-75 = 425, obtaining ta = 5.16018. The velocity at the time of impact
with the liquid is
Va
m dv/dt = mg-kv2.
= v(ta) = 164.482. When the object is in the liquid its velocity is modeled by
Using m = 160, 9 = 32, and k = 0.1 this becomes dv/dt
= (51,200-v2)/1600.

dv
51,200 - v2
Solving v(O) =
Va
= __!!:!_
and
1600
J2 In 1 v - 160J2
640
v + 160J2
1=
_l_t + c.
1600
= 164.482 we obtain e = -0.00407537. Then, for v < 160J2

v - 160J2I=
v + 160J2
eV2t/5-1.8443 or
_ v - 160J2
v + 160J2
= 226.274,
= eV2t/5-1.8443
Solving for v we get

13964.6 - 2208.2geV2t/5
v( t) = 61.7153 + 9.75937eV2t/5 .
Integrating we find
s(t)
Solving s(O)
= O we
= 226.275t - 1600In(6.3237 + eV2t/5) + c.
see that e = 3185.78, so

s(t) = 3185.78 + 226.275t - 1600In(6.3237 + eV2t/5).
To find when the object hit s the bottom of the tank we solve s(t) = 75, obtaining tb = 0.466273.
The time from when the object is dropped from the helicopter to when it hits the bottom of the
tank is ta + tb = 5.62708 seconds.
90
Exercises 3.3
Exercises 3.3
1. The equation dxf dt.
dx]
-A1dt
= -A1X can be solved by separation of variables. Integrating both sides of
+ c from which
we obtain In [z] = -A1t
C1XO so that x =
xoe-Alt.
= Xo
we find
Substituting this result into the second differential equation we have

dy
-- +A2Y=
dt
which is linear. An integrating factor is
.5._
dt
Y
we get x = C1e-Alt. Using x(O)
= A1XO
eA2t
y]
[eA2t
e(A2-AI)te-A2t
A2 - Al
Using y(O) = O we find c2 = -A1XO/(A2
_
y-,
A1xOe
_A t
1
so that
= A 1x Oe(A2-AI)t + c2
+ C2e-A2t
A1XO
A2 - Al
e-Alt
+ c2e-A2t
- Al). Thus
A1XO (-Alt
e
/\2 - Al
e -A2t) .
Substituting this result into the third differential equation we have

dz = A1A2XO (e-Alt
dt
_ e-A2t)
A2 - Al
Integrating we find
2. We see from the graph that the half-life of A is ap-
proximately 4.7 days. To determine the half-life of B

we use t
x,
y,
20
= 50 as a base, since at this time the amount
of substance A is so small that it contributes very little to substance B. Now we see from the graph that
y(50) ~ 16.2 and y(191)
~ 8.1. Thus, the half-life of
B is approximately 141 days.
~~2~5--~5~0--~7~5--1~0~0--1-2~5--15~0~-
t = 5 days. The amounts of x and z are the same

at about t = 20 days. The amounts of y and z are the same at about t = 147 days. The time when
y and z are the same makes sense because most of A and half of B are gone, so half of e should
3. The amounts of x and y are the same at about
have been formed.
91
Exercses 3.3
4. Suppose that the series is described schematically by W ==> -AlX ==> -A2Y
==> -A3Z where
-Al, -A2, and -A3 are the decay constants for VV, X and Y, respectively, and Z is a stable elemento
Let w(t), x(t), y(t), and z(t) denote the amounts of substances W, X, Y, and Z, respectively. A
model for the radioactive series is
dw
= -AlW
dt
dx
- = AlW - A2X
dt
dy
dt = A2X - A3Y
dz
dt = A3Y
5. The system is
11121
Xl = 2 . 3 + 50 X2 - 50 Xl . 4 = - 25Xl + 50 X2 + 6
I
1
1
1
2
2
x2 = 50XI 4- 50X2 - 50X2' 3 = 25xl - 25x2.
6. Let Xl, X2, and X3 be the amounts of salt in tanks A, B, and

I
1
Xl = 100x2 2-
11
x2 = -Xl'
100
e, respectively,
so that
1
1
3
100Xl 6= 50X2 - 50Xl
1
6 + -X3
100
1
- -X2
100
1
. 2 - -X2
100
I
1
1
1
x3 = 100x2' 5 - 100x3 - 100x3 4=
. 5 = -Xl
50
1
+ -X3
100
- -X2
100
1
1
20X2 - 20X3.
7. (a) A model is
dXl = 3 . X2 _ 2 . Xl
dt
100 - t
100 + t '
Xl(O) = 100
dX2 = 2 . Xl _ 3 . X2
dt
100 + t
100 - t '
X2(0)
= 50.
(b) Since the system is closed, no salt enters or leaves the system and Xl(t) +X2(t) = 100+50 = 150
for all time. Thus Xl = 150 - X2 and the second equation in part (a) becomes
dX2
2(150 - X2)
=
dt
100 + t
3X2
100 - t
300
100 + t
=---
2X2
100 + t
or
dX2
(2
3)
+ 100 + t + 100 - t
which is linear in X2. An integrating factor is
di"
X2 =
300
100 + t '
e2ln(IOO+t)-3ln(IOO-t)
= (100 + t)2(100 - t)-3
92
3X2
100 - t
Exercises 3.3
so
Using integration by parts, we obtain

(100
+ t)2(100
- t)-3X2 = 300 [~(lOO + t)(100 - t)-2 - ~(100 - t)-l
+ e] .
Thus
X2
1
2
1
]
= (100300
+ t)2 [c(100 - t) 3 - "2(100t) + "2(100 + t)(100 - t)
300
3
1
(100 + t)2 [c(100 - t) + t(lOO - t) .
Using X2(0) = 50 we find e = 5j3000.
At t
= 30, X2 = (300j1302)(703c
+ 30
70) ~ 47.4 lbs.
8. A model is
dXI
di
dX2 =
di
1
(4 galjmin)(O lbjgal) - (4 galjmin) ( 200XI
lbjgal )
.
(4 gal/min)
dX3 = (4 galjmm)
.
di
(1200XI
(1150X2
)lbjgal ) lbjgal
. (1150X2 lbjgal )
. . ( 100X3
1 lbjgal )
(4 galjmm)
(4 galjmm)
or
dXI
1
= --Xl
dt
50
dX2
1
= -XIdt
50
dX3
2
- = -X2dt
75
-
2
-X2
75
1
-X3
25
Over a long period of time we would expect Xl, X2, and X3 to approach Obecause the entering pure
water should flush the salt out of all three tanks.
9. From the graph we see that the populations are first egual
at about t = 5.6. The approximate periods of
x,y
and y are
both 45.
5D
93
IDD
Exercises
3.3
10. (a) The population y(t) approaches 10,000, while the population
x(t) approaches extinction.
k,
l~k
1 Jo
10
(b) The population x(t) approaches 5,000, while the population
y(t) approaches extinction.
20
; 1
1
10
(e) The population y(t) approaches 10,000, while the population
x(t) approaches extinction.
(d) The population x(t) approaches 5,000, while the population

y(t) approaches extinction.
1:}::1
II
5
11. (a)
E:
t..
I"'t
20
:1
10
20
10
20
1'" t
1'" t
(b)
1:
1 1
20
(e)
20
1 I"'t
40
20
40
(d)
1 1 1'"
40
In each case the population x(t) approaches 6,000, while the population y(t) approaches 8,000.
12. By Kirchoff's first law we have il = i2 + i3. By Kirchoff's second law, on each loop we have E(t) =
Li~ + R1i2 and E(t) =
u; + R2i3 + ~q
so that q = CRli2 - CR2i3. Then i3
so that the system is

Li;
-Rli;
+ Li~ + Rli2 =
1
E(t)
+ R2i3 + Ci3 = O.
94
= q'
= CRli2 - CR2i3
Exercses
13. By Kirchoff's first law we have il = i:
+ i3.
3.3
Applying Kirchoff's second law to each loop we obtain
di2.
E(t) =
2lRl
E(t) =
2lRl
+ i; di + 22R2
and
.
di3.
+ L2di + 23R3.
Combining the three equations, we obtain the system
di2
..
Lldi + (Rl + R2)22 + Rl23 = E
L2 di3 + Rli2 + (Rl + R3)i3 = E.
dt
14. By Kirchoff's first law we have il = i2
E(t) =
u; + Ri2
and E(t)
+ i3.
By Kirchoff's second law, on each loop we have
= u; + Cq so that q = CRi2. Then
i3
= q'
CRi~ so that system is
Li' + Ri2 = E(t)
15. We first note that s(t)
+ i(t) + r(t)
n. Now the rate of change of the number of susceptible
persons, s(t), is proportional to the number of contacts between the number of people infected and
the number who are susceptible; that is, ds/dt
= -klSi.
We use -kl because s(t) is decreasing.
Next, the rate of change of the number of persons who have recovered is proportional to the number
infected; that is, dr / dt = k2i where k2 is positive since r is increasing. Finally, to obtain di/ dt we
use
d
d
-(s+i+r)=-n=O.
dt
dt
This gives
di
ds
.
.
-dt
= --dr
=
-k22 + klSZ.
dt dt
The system of equations is then
ds
.
= -klS2
dt
-di = - k''22
+ k lS2.
dt
dr
- = k2i.
dt
A reasonable set of initial conditions is i(O) = io, the number of infected people at time O, s(O) =
n - io, and r(O) = O.
16. (a) If we know s(t) and i(t) then we can determine r(t) from s
95
+ i + r = n.
Exercses 3.3
(b) .In this case the system is
ds
= -0.2si
dt
di
O.7z+ 02'
z=:
. sz.
s(O) = 10 - io since r(O) = O and i(t) + s(t) + r(t) = O for
all values of t. Now k2/kl = 0.7/0.2 = 3.5, so we consider initial conditions s(O) = 2, i(O) = 8;
s(O) = 3.4, i(O) = 6.6; s(O) = 7, i(O) = 3; and s(O) = 9, i(O) = 1.
We also note that when i(O)
io,
s.i
s.i
s.i
10
10
10
10
s.i
101
101
We see that an initial susceptible population greater than k2/ kl results in an epidemic in the
sense that the number of infected persons increases to a maximum before decreasing to
the other hand, when s(O)
<
k2/kl,
and there is no epidemic.

17. Since xo > YO > O we have x(t) > y(t) and Y - x < O. Thus dx/dt < O
and dy/dt > O. We conclude that x(t) is decreasing and y(t) is
increasing.
As t
--+ 00
o.
On
the number of infected persons decreases from the start
we expect that x(t)
--+
C and y(t)
where C is a constant common equilibrium concentration.
18. We write the system in the form
dx
= kl(y - x)
dt
dy
dt = k2(X - y),
96
--+
C,
x,y
x(O)
Exercises 3.3
where kl
= KjVA and k2 = KjVB. Letting z(t) = x(t) - y(t) we have

dx
dt-
dy
dt =kl(y-x)-k2(X-y)
dz
dt = kl(-Z)
- k2Z
This is a first-order linear differential equation with solution z(t)
CIe-(k+k2)t.
Now
and
Since y(t) = x(t) - z(t) we have
The initial conditions x(O) = Xo and y(O) = YO imply

and
The solution of the system is
x(t)
= (xo - yo)k1
k1
y(t)
e-(k+k2)t
+ k2
= (YO - XO)k2 e-(k+k2)t

kl
+ k2
+ XOk2 + yok1
k1 + k2
+ XOk2 + YOkl
kl + k2
As t ___.00, x(t) and y(t) approach the common limit
This makes intuitive sense because the limiting concentration is seen to be a weighted average of
the two initial concentrations.
97
3.3
Exercises
19. Since there are initially 25 pounds of salt in tank A and
xI,x2
none in tank B, and since furthermore only pure water is
30
being pumped into tank A, we would expect that Xl(t)
20
would steadily decrease over time. On the other hand,
10
since salt is being added to tank B from tank A, we would

expect X2(t) to increase over time. However, since pure
50
100
water is being added to the system at a constant rate

and a mixed solution is being pumped out of the system, it makes sense that the amount of salt in
both tanks would approach O over time.
1. From
dP
di
2. Let A
= A(t) be the volume of C02 at time t. From ddA = 1.2 - ~ and A(O) = 16ft3 we obtain
= 0.018P
and P(O)
= 4 billion we obtain P = 4eO.018t so that P(45) = 8.99 billion.
t
4
A = 4.8 + 11.2e-t/4. Since A(10) = 5.7ft3, the concentration is 0.017%. As t ......00 we have
A ......4.8 ft3 or 0.06%.
< 4, 6 :::;t < 10, and 12 :::;t < 16, no voltage is applied to the heart and E(t) = O. At
the other times the differential equation is dE/ dt = - E / RC. Separating variables, integrating,
and solving for E, we get E = ke-t/RC, subject to E(4) = E(10) = E(16) = 12. These initial
conditions yield, respectively, k = 12e4/RC, k = 12elO/RC, and k = 12e16/RC. Thus
O :::; t < 4, 6:::; t < 10, 12:::; t < 16
O,
3. (a) For O:::; t
E ( t ) --
12e(4-t)/RC
12e(lO-t)/ RC,'
12e(16-t)/ RC,
(b)
4:::;t<6
10 :::;t < 12
16:::; t < 18.
10
\\\
4
4. From V dC/dt = kA(Cs
10 12
16 18 '
C) and C(O) = Co we obtain C = Cs
98
+ (Co - Cs)e-kAt/V,

5. (a) The differential equation is
dT
-dt = k[T -
T2 - B(TI - T))
dT
(1 + B)T - (BTI

_2__BIn 1(1 + B)T - (BTI
1+
+ T2)1= kt + e
Since T(O) = TI we must have

T(t)
(b) Since k
< O, lim (1+B)t
e3
di
40 - 2i
___!!:!__.
10 -
+ T2 .
1+ B
140 - 2il
BTI + T2
l+B
.
variables we obtain
el
= 2/JIO.
= 4 or i
10 the equation for the eurrent becomes 0.2i
=k
we obtain dx = ~
dy = 2k sin e eos e de,

Othen c
dx
= 2k
= q(lO -
Solving for i we get i(t) = 4t -
tt
2,
OS t
t2:10
20,
= Owhen e =
BTI + T2)
l+B
= -In 110 - ti + e or J40 - 2i
'( ) _ { 4t It
-
If x
Then
< 10. For t 2:
[1 + (y')2]
T(t) = BTI + T2 + e (1+B)t.

l+B
3
and
t~~Ts=T2+BTI-B
Sinee i(O) = O we must have
7. From y
= k dt. Then
BTI + T2 + TI - T2 (1+B)t.
l+B
l+B
1tO) ~: + 0.2i = 4. Separating
1
--In
2
O:::;t
+ T2)J.
TI - T2
B and so
1+
t--+oo
(e) SineeTs=T2+B(TI-T),
(1 -
+ T2)
- (BTI
= Oand lim T(t) = BTI
i=ecx:
6. We first solve
= k[(l + B)T
dy. If y
t).
10
20
tt
2,
= 20. Thus
< 10 .
= k sin2 e then
G- ~
eos 2e) de,
and
x = ke -
.
'2k sin
2e + c.
= O.
8. (a) From y = - x - 1+ el e" we obtain y' = y + x so that the differential equation of the orthogonal
. dy = --- 1 or -dx + x = -y. A n integratmg
. f actor 1Se
. y , so
f am1'1y 1S-d
x
y+x
dy
~[eYxl
dy
= -yeY
==:;.
eYx
= -yeY + eY + c
99
==:;.
= -y + 1 + ce-Y.
(b) Differentiating the family of curves, we have

,
1
y =-(X+CI)2
1
=-y2'
The differential equation for the family of orthogonal trajectories

is then y'
= y'2.
Separating variables and integrating we get

dy = dx
y2
1
-- = x + e
y
y=---.
x+e
9. From ~~
dy
dt
= klx(a - x) we obtain C~a + a1~ax) dx
= k 2XY we obtam.
lnlyl=
k2 I l+Cle
klln
k ti +c
(k
or
y=C2 l+Cle
v"
10. In tank A the salt input is

(7
!~~)
( + ( !~~)
(:o~
2~~l)
~~l)
= (14 + 1~0x2) ~~n .
The salt output is
(3~) (2~)+ (5~) (2~) = _2_

min
100 gal
min
gal)
( 5 min
(Xl lb)
100 gal
100 gal
25
Xl ~ .
min
In tank B the salt input is

=
1
lb
20 Xl min .
The salt output is
(1~) (2~)+
min
100 gal
(2~) =
(4 gal)
min
100 gal
The system of differential equations is then

dx
dt
dX2
dt
1
2
14 + 100x2 - 25xI
-X-
1
20
1
20
-X2.
100
]_X2~.
20
min
Chapter 3 Related Exercises
1. (a) The differential equation is dP / dt = P (5 - P) - 4. Solving P (5 - P) - 4 = O for P we obtain
equilibrium solutions P = 1 and P = 4. The phase portrait is shown below and solution curves
are shown in part (b).
We see that for Po

O
>
4 and 1
< Po <
4 the population approaches 4 as t increases.
For
< Po < 1 the population decreases to O in finite time.
(b) The differentiaI equation is
~ = P(5 - P) -
= _(P2
5P
+ 4) =
-(P - 4)(P - 1).
Separating variables and integrating, we obtain
dP
= -dt
(P - 4)(P - 1)
( ~-~)dP=-dt
P-4
P-1
~ In 1P - 41
3
P-1
= =t + c
P- 4
-3t
P _ 1 = ci e
Setting t
= O and P
= Po we find CI = (Po - 4)/(Po - 1). Solving for P we obtain
P(t) = 4(Po - 1) - (Po - 4)e-3t

(Po - 1) - (Po - 4)e-3t
(e) To find when the population becomes extinct in the case O < Po < 1 we set P = O in
P - 4
--=--e
P -1
Po - 4
Po-1
-3t
from part (a) and solve for t. This gives the time of extinction
t
= _ ~ In 4(Po - 1) .
3
Po - 4
2. (a) Solving P(5 - P) - ~ = O for P we obtain the equilibrium solution P = ~. For P =F ~,

dP / dt
< O. Thus, if Po < ~, the population becomes extinct (otherwise there would be another
equilibrium solution.) Using separation of variables to solve the initial-value problem we get
101

P(t) = [4Po + (10Po - 25)t]/[4 + (4Po - 10)t]. To find when the population becomes extinct
for Po < ~ we solve P(t) = for t. We see that the time of extinction is t = 4Po/5(5 - 2Po).
(b) Solving P(5 - P) - 7 = for P we obtain complex roots, so there are no equilibrium solutions.
< for all values of P, the population becomes extinct for any initial condition.
Since dP / dt
Using separation of variables to solve the initial-value problem we get
O-5)
v'3 tan [tan-I (2P v'3

P(t) = "25 + 2
=
Solving P(t)
3. (a) Without
v'3]t
- 2
for t we see that the time of extinction is
harvesting, the population is governed by the
logistic equation dP/dt
P(r - r P] K).
With initial
population Po, the population was shown in Section 3.2

to be
P( )
rPo
t = rPo/K+(r-rPo/K)e-rt
Po
50
100
50
100
Poi K + (1 - Poi K)e-rt .

To find when P(t) = ~K we solve
Po
= ~K
Poi K + (1 - Poi K)e-rt
2
or
2Po
= 1.
Po + (K - Po)e-rt
This gives
= _~ ( Po ) = __1_
(
70,000
) ~ 1 4 rs
r In K _ Po
0.08 In 400,000 _ 70,000
9. y .
(b) Letting K = 400,000, Po = 70,000, and r = 0.08, and

solving
PI =
for h, we get h
K - JK2 - 4Kh/r
2
= Po
= ha = 4620.
102

(c) The MSY is obtained when Ea = r /2 = 0.04, and the
yield in this case is EPI = KE(1 - E/r) = 8,000. The

limiting population is K(1 - E/r)
= 200,000.
50
4. (a) Letting 1959 be year O and 1973 year 14, we obtain the
100
400
graph shown.
300
200
100
(b) Taking P(O)
= 1.91/(c 414 . 294) ~ 78.4613 and ex-
10
12
14 t
10
12
14 t
perimenting with the other pararneters, we find that the
400
graph of the logistic function fits the data points reason-
300
ably well for K = 500 and r
.. ..
= 0.11.
(e) The graph is shown with Po = 78.4613, K = 500,

r = 0.11, and h = irK ~ 6.875.
400
300
200
(d) The graph is shown with Po = 78.4613, K
= 500,
r = 0.11, and E = ~r = 0.055.
20
40
60
80
100 t
20
40
60
80
100 t
250
200
150
100
50
103
Hlqher-Order DifferentiaJ Equations

Exercises 4.1
_
Thcn y(O) = CI + C2 = o, y'(O) = CI - C2 = 1 so
1. From y = CleX + C2e-X we find y' = CleX - C2CX.
that CI = 1/2 and C2 = -1/2. The solution is y = ~ex - ~e-x.
2. From y = clx
so that cI
+ C2e-x we find y' = 4clx
= 3/5
+ ge-x.
we find y' = CI + c2(1 + ln z ). Then y(l)
that CI = 3 and C2 = -4.
= c +
Then y(O) = CI + C2 = 1, y'(O) = 4CI - C2 = 2
and C2 = 2/5. The solution is y = ~x
3. From y = CIX+C2xInx
4. From y
- c2e-x.
The solution is y
= 3x - 4x
= CI = 3, y'(l)
= CI +C2 = -1 so
In x.
C2cos x + C3sin x we find y' = -C2 sin x + C3cos x and yl/ = -C2 cos X - C3sin x. Then
Y(7!") = CI - C2 = O, Y'(7!") = -C3 = 2, yl/(7!") = C2 = -1 so that CI = -1, C2 = -1, and C3 = -2. The
solution is y = -1 - cosx - 2sinx.
5. From y = c + C2x2 we find y' = 2C2X. Then y(O) = CI = O, y'(O) = 2C2 . O = O and y'(O) = 1 is not
possible.
Since a2(x)
= x is O at x = O, Theorem
6. In this case we have y(O)

are y
= x2 and y
= CI = O,
4.1 is not violated.
y'(O) = 2C2 . O
= O so
CI
=O
and C2 is arbitrary, Two solutions
= 2x2.
7. From x(O) = Xo = CI we see that x(t) = xocoswt

Then x'(O) = Xl = C2W implies C2 = xjw.
+ C2Sinwt and x'(t)
= -xosinwt
Thus
.
+ -Xl smwt.
x ( t ) = Xo coswt
8. Solving the system
x(to) = CI coswto + C2sinwto
=x
+ C2Wcoswto
x'(to) = -CIW sinwto
= Xl
for Cl and C2 gives

CI =
wXo cos wto - Xl sin wto

w
and
C2 =
Xl cos wto + wXo sin wto

w
Thus
X ()t =
wXo cos wto - Xl sin wto

coswt
w
= xo(coswtcoswto
= Xo cosw(t
+ sinwtsinwto)
- to) + Xl sin w(t - to).

w
104
Xl cos wto + wxo sin wto
w
+ Xl (sinwtcoswto
w
sinwt
- coswtsinwto)
+ C2wcoswt.
Exercises 4. 1
9. Since a2(x) = x - 2 and Xo
10. Since ao(x)
= tan
x and Xo
= Othe problem has a unique solution for
-00
= Othe
=t: /2
problem has a unique solution for
11. We have y(O) = el +cs = O, y'(l) = ele+e2e-l
(e2 -1)
= 1 so that el = e/
< x < 2.
< x < 7r /2.
and e2 = -e/
(e2 -1).
The solution is y = e (ex - e-x) / (e2 - 1).

12. In this case we have y(O)
= el
= 1,
y'(l)
= 2e2
= 6 so that el
= 1 and e2 = 3. The solution is
y=1+3x2.
13. From y
= el eX cos x
+ e2ex sin x we find y'
= eleX( -
sin x + cos x) + e2eX(coSx + sinx).
(a) We have y(O) = el = 1, y'(O) = el + e2 = O so that el

y = e" cosx - eX sinx.
= el = 1, Y(7r)
(b) We have y(O)
= -ele7T
(e) We have y(O) = el = 1, Y(7r/2)

y = eX cosx + e-7T/2ex sinx.
(d) We have y(O)
= el
1 and e2
-l. The solution is
= -1, which is not possible.
= e2e7T/2= 1 so that el = 1 and e2
= O, Y(7r) = -ele7T
= e-7T/2. The solution is
= O so that el = O and e2 is arbitrary. Solutions are
y = e2ex sin x, for any real numbers e2.
= el + e2 + 3 = O, y(l) = el + e2 + 3 = 4, which is not possible.
14. (a) We have y(-l)
(b) We have y(O) = el . O+ e2 . 0+3 = 1, which is not possible.

(e) We have y(O)
= el . O+ e2 . 0+3
e2 = -3 - el. Solutions are

(d) We have y(l)
solution is y
15. Since (-4)x
= 3, y(l) = el + e2 + 3 = O so that el is arbitrary and
y = CIX2 -
(el + 3)x4 + 3.
= el + e2 + 3 = 3, y(2) = 4el + 16e2+ 3 = 15 so that el

= _x2 + x4 + 3.
= -1 and e2 = l. The
+ (3)x2 + (1)(4x - 3x2) = Othe functions are linearly dependent.
16. Since (1)0 + (O)x + (O)eX
= Othe functions are linearly dependent. A similar argument shows that
any set of functions containing f(x)
= Owill be linearly dependent.
17. Since (-1/5)5 + (1) cos2 x + (1) sin2 x = Othe functions are linearly dependent.
18. Since (1) cos 2x + (1)1 + (-2) cos2x = Othe functions are linearly dependent.
19. Since (-4)x
+ (3)(x - 1) + (l)(x + 3) = Othe functions are linearly dependent.
20. From the graphs of h(x)
=2+
x and h(x)
= 2 + [z]
we see that the funetions are linearly independent since

they cannot be multiples of each other.
3X
105
3X
Exercises 4.1
1+ x
21. The functions are linearly independent since W (1 + x,
22. Since (-1/2)eX
+ (1/2)e-X
+ (1) sinh x = Othe
x2)
X,
x2
1 2x
functions are linearly dependent.
23. The functions satisfy the differential equation and are linearly independent since
W (e-3x,e4X)
for
-00
i- O
= 7ex
< x < oo. The general solution is

y
= cle-3x + C2x.
=2
W(cosh2x,sinh2x)
for
-00
< x < co. The general solution is
+ C2 sinh 2x.
y = Cl cosh 2x
W (eX cos 2x, eX sin 2x)
for
-00
< x < oo. The general solution is y
= 2e2x i- O
= clex cos 2x
+ C2ex sin 2x.
W (eX/2, xex/2)
for
-00
= eX
i- O

y = clex/2
+ C2xex/2.
W (x3,x4)
for O < x
= x6
i- O
< oo. The general solution is
W(cos(lnx),sin(lnx))
for O < x
= l/x
i- O

y
= Cl cos(lnx) + C2 sin(lnx).
106
= 2 i- O.
Exercises 4. 1
= 9x-6 =J. O
W (x,x-2,x-2Inx)
for O < x

y = CIX
+ C2X-2 + C3x-21n
x.
W (1, x, cos x, sin x)

for
-00

=
Cl
+ C2X + C3 cos x + C4 sin x.
= e2x and Y2 = x form a fundamental set of solutions of the homogeneous

equation, and Yp = 6ex is a particular solution of the nonhomogeneous equation.
31. The functions Yl
32. The functions Yl =
COS x
and Y2 = sin x form a fundamental set of solutions of the homogeneous
equation, and Yp = x sin x+ (cos x) ln(cos x) is a particular solution of the nonhomogeneous equation.
33. The functions Yl
equation, and Yp
e2x and Y2
x2e2x
34. The functions Yl = x-l/2

equation, and Yp
35.
+x
- 2 is a particular solution of the nonhomogeneous equation.
and Y2
= lsx2 - ~x
xe2x form a fundamental set of solutions of the homogeneous
x-1
form a fundamental set of solutions of the homogeneous
is a particular solution of the nonhomogeneous equation.
(a) We have y'PI = 6e2x and y"PI' = 12e2x so

" - 6y'PI
YPI
Also, Y~2 = 2x
+ 3 and
+ 5y PI =
12e2x - 36e2x
+ 15e2x =
_ge2x .
Y~2 = 2, so
Y~2 - 6Y~2 + 5YP2 = 2 - 6(2x
+ 3) + 5(x2 + 3x)
= 5x2
+ 3x
- 16.
(b) By the, superposition principle for nonhomogeneous equations a particular solution of y" - 6y' +
5y = 5x2
+ 3x
- 16 - ge2x is Yp
= x2 + 3x + 3e2x. A particular solution of the second equation
is
36.
(a) YPI = 5
(b) YP2 = -2x
(e) Yp = YPI + YP2 = 5 - 2x
(d) Yp = ~YPI - 2YP2 = ~ + 4x

37. (a) Since D2x = O, x and 1 are solutions of y" = O. Since they are linearly independent, the
general solution is y = cj z
+ C2.
107
Exercises
(b)
4. 1
Since D3x2
= o, x2,
= O. Since they
x, and 1 are solutions of ylII
general solution is y = C1x2
are linearly independent, the
+ C2X + C3.
(e) Since D4x3 = O, x3, x2, x, and 1 are solutions of y(4) = O. Since they are linearly independent,
= C1x3 + C2x2 + C3X + C4.

(d ) By part (a), the general solution of y" = O is Yc. = C1X + C2. Since D2x2 = 2! = 2, Yp = x2 is a
particular solution of y" = 2. Thus, the general solution is y = C1X + C2 + x2.
the general solution is y
(e) By part (b), the general solution of y'" = O is Yc. = C1x2
+ C2X + C3.
Since D3x3
= 3! = 6,
yp = x3 is a particular solution of ylll = 6. Thus, the general solution is y = C1x2 +C2X+C3 +x3.
(f) By part (e), the general solution ofy(4) = Ois Yc. = C1x3+C2X2+C3X+C4.
Since D4x4 = 4! = 24,
yp = x4 is a particular solution of y(4) = 24. Thus, the general solution is y

C3X + C4
= cx3 + C2x2 +
+ x4.
38. By the superposition principle, if Y1 = eX and Y2 = e-x are both solutions of a homogeneous linear
differential equation, then so are
1
2(Y1
+ Y2)
eX + e-x
2
= coshx
1
2(Y1 - Y2)
and
39. (a) From the graphs of Y1 = x3 and Y2 = !x!3 we see
eX - e-x
2
= sinhx.
y
that the functions are linearly independent since

they cannot be multiples of each other. It is easily
shown that Y1
= x3 solves x2y" - 4xy' + 6y
3x
= O.
To show that Y2 = !x!3 is a solution let Y2 = x3

for x ~ O and let Y2 = _x3 for x
< O.
(b) If x ~ Othen Y2 = x3 and W(Y1, Y2) =

If x
< O then
92 = -x3
and W(Y1,Y2)
13:
x31
3x2
= 1 x32
3x
= o.
_x312
-3x
= O.
= x2 is zero at x = O.
(e) The functions Y1 = x3 and Y2 = x2 are solutions of x2y" - 4xy' + 6y = O. They are linearly
This does not violate Theorem 4.3 since a2(x)
independent since W (x3, x2) = x4
(d) The function y
=1=
Ofor
-00
< x < oo.
= x3 satisfies y(O) = O and y'(O) = O.
(e) Neither is the general solution since we form a general solution on an interval for which
a2 (x)
=1=
O for every x in the interval.
40. Since ex-3 = e-3eX = (e-Se2)ex
= e-Sex+2,
we see that ex-3 is a constant multiple of ex+2 and the
functions are linearly dependent.
108
Exercises 4.2
41. Since OYI + OY2 +
... + OYk + 1Yk+1 =
O, the set of solutions is linearly dependent.
42. The solutions are linearly dependent. Suppose n of the solutions are linearly independent (if not,
+ 1 solutions is linearly dependent).

Then y = CIYI+ C2Y2 +
+ CnYn is the
then the set of n

Y2, ... , Yn
equation and for sorne choice, ci, c2,
Without loss of generality, let this set be YI,

general solution of the nth-order differential
, c~, of the coefficients Yn+l = ciYI + c2Y2 + ... + c~Yn. But
then the set YI. Y2, ... , Yn, Yn+l is linearly dependent.
Exercises 4.2
In Problems 1-8 we use reduction of order to find a second solution. In Problems 9-16 we use formula
(5) from the texto
1. Define y = u(x )e2x so
y'
= 2ue2x + u' e2x,
Therefore u"
= e2xu" + 4e2xu' + 4e2xu,
y"
and
y" - 4y'
= CIX+ C2. Taking CI = 1 and C2 =
O and u
+ 4y = 4e2xu" = O.
O we see that a second solution is
Y2 = xe2x.
2. Define y
= u(x)xe-X
so
y' = (1 - x)e-Xu
+ xe-xu',
y" = xe-xu"
+ 2(1 -
x)e-Xu'
- (2 - x)e-Xu,
and
y"
If w
= u'
we obtain the frst-order equation w'
Therefore w
3. Define y
+ 2y' + y = e-X(xu" + 2u') = O
or
+~ w = Owhich
x
u"
+ ~ u' = O..
x
has the integrating factor e2 J dx]
= u' = c/x2 and u = ci]. A second solution is Y2 = - xe-x = e-x.

x
= u(x) cos4x so
y'
= -4u
sin 4x
+ u' cos 4x,
y"
= u" cos 4x
- 8u' sin 4x - 16u cos 4x
and
y"
+ 16y =
(cos4x)u"
= O or u" - 8(tan4x)u' = O.
- 8(sin4x)u'
If w = u' we obtain the first-order equation w' - 8(tan 4x)w = O which has the integrating factor
e-SJtan4xdx = cos24x. Now
d
dx [(cos2 4x)w]
= O gives (cos2 4x)w = c.

109
Exercises 4.2
Therefore w = u' = C sec2 4x and u
4. Define y
= C! tan 4x.
= sin 4x.
A second solution is Y2 = tan 4x cos 4x
= u(x) sin 3x so
y' = 3u cos 3x
+ u' sin 3x,
= u" sin 3x + 6u' cos 3x - 9u sin 3x,
y"
and
y"
If w
+ 9y
+ 6(cos 3x)u' = O or u" + 6(cot 3x)u' = O.

equation w' + 6(cot 3x)w = O which has the integrating
= (sin 3x)u"
u' we obtain the first-order

= sin23x. Now
factor
e6Jcot3xdx
d
.
dx[(sm23x)w]
Therefore w = u'
= c csc2
=O
gives
(sin2 3x)w
= c.
3x and u = Cl cot 3x. A second solution is Y2
= cot 3x sin 3x
= cos 3x.
5. Define y = u(x) coshx so

y' = usinhx
+ u' coshx,
y" = u" coshx
+ 2u' sinhx + ucoshx
and
y" - y = (coshx)u"
If w
= u' we obtain the first-order
e2Jta.nhxdx
= cosh/
+ 2(sinhx)u' = O or
equation w' + 2(tanhx)w
= csech2 x and u
6. Define y = u(x)X
+ 2(tanhx)u'
= O.
= O which has the integrating factor
x. Now
d
dx [(cosh2 x)w]
Therefore w = u'
u"
Cl
= O gives (cosh'' x)w
= c.
tanhx. A second solution is Y2 = tanh x cosh x = sinhx.
so
and
y" - 25y
If w
= X(u" + lOu')
= u' we obtain the first-order equation
w'
= O or
u" + lOu' = O.
+ lOw = Owhich
has the integrating factor elO J dx
elOx. Now
..!!:.... [elOXw] = O gives
Therefore w = u' = ce-lOx

7. Define y = u(x)e2x/3
dx
and u = cle-lOx.
elOxw = c.
A second solution is Y2 = e-loxx
so
and
9y" - 12y'
+ 4y = ge2x/3u" = O.
110
= e-5x.
Exercises 4.2
Therefore u"
= O and u =
+ C2.
CIX
Taking q = 1 and
O we see that a second solution is
C2 =
Y2 = xe2x/3.
8. Define y = u(x )ex/3 so
and
+ y'
6y"
- y
=e
(6u"
X/3
If w = u' we obtain the first-order equation
X/6.
+ 5u')
w' +iw
=O
or
u"
+ ~u' = O.
= Owhich has the integrating factor e(S/6)
Now
Therefore w
= u'
d
- [eSx/6w] = O gives x/6w = C.
dx
and u = cle-Sx/6. A second solution is Y2 = e-Sx/6ex/3
= ce-Sx/6
9. Identifying P(x) = -7/x
we have
Y2=X4
e- J -(7/x)dx
x8
dx=x4
~dx=x41nlxl.
A second solution is Y2 = x41n [z].

10. Identifying P(x) = 2/x we have
Y2 = x2
A second solution is Y2
11. Identifying P(x)
e- J(2/x) dx
x4
dx
x-6dx
1
= -"5x-3.
= x-3.
= L] we have
Y2
= x2
= lnx
e- J dx/x
(In z )? dx = lnx
dx
x(lnx)2
= lnx
(1-lnx ) =-1.
A second solution is Y2 = 1.
12. Identifying P(x) = O we have
e- JOdx
=xl/21nx
x(lnx)2
Y2=x1/21nx!
(1__ ) =_x 2.
1/
lnx
A second solution is Y2 = xl/2.

13. Identifying P(x) = -l/x
we have
. !
Y2 = xsm(1nx)
e- J =dx]
.
2'
2
dx = xsm(1nx)
x sin (lnx)
= [xsin(lnx)] [-cot(lnx)]
= -xcos(1nx).
A second solution is Y2 = x cos(ln x).
111
x
2(
x sin lnx
2'
dx
= e-x/2.
J dx =
Exercses 4.2
14. Identifyng P(x) = -3/x
Y2 =
we have
X2
le
cos(lnx)
- J -3dx/x
4
x cos
2(1
nx
dx
= x 2 cos(lnx)
3
X2(
x cos In x
dx
= x2 cos(ln x) tan(ln x) = x2 sin(ln x).

A second solution is Y2 = x2sin(lnx).
15. Identifying P(x) = 2(1+ x)/ (1- 2x - x2) we have
Y2 = (x
+ 1)
= (z + 1)
e- J2(1+x)dx/(1-2x-x2)
(x + 1)2
dx
2x - x2
+ 1)2 dx = (x
+ 1)
l
l 1-
(x
= (x + 1) [__
(x
+ 1)
l [2 +
(x
e1n(1-2x-x2)
(x+ 1)2 dx
1) 2 - 1 dx
2_ - x] = -2 - x2 - x.
x+l
A second solution is Y2 = x2 +
16. Identifying P(x) = -2x/
Y2 =
(1 -
+ 2.
x2) we have
e- J -2xdx/(1-x2)dx
A second solution is Y2 = In 1(1 + x)/(I17. Define y = u(x)e-2x
e-1n(1-x2)dx
1_11 - x2
dx
= ~ In 1 + xI
2
1- x
x)l.
so
+ u'e-2x,
y' = _2ue-2x
y" = u"e-2x _ 4u'e-2x + 4ue-2x
and
y" - 4y
= e-2xu" - 4e-2xu' = O or u" - 4u' = O.
If w = u' we obtain the first order equation w' - 4w = O whch has the integrating factor e-4
J dx =
e-4x. Now
d
dx[e-4Xwj
Therefore w
u' = cx and u
c1x.
observation that a particular solution is Yp

y
18. Define y = u(x)1
=O
gives e-4xw = c.
A second solution is Y2
e-2xe4x = e2x. We see by
= -1/2. The general solution is

1
= cle-2x + C2e2x - -
so
y' = u',
y"
= u"
and
y"
112
+ y' = u" + u'
= O.
Exercises 4.2
If w = u' we obtan the first order equaton w' + w = Owhch has the ntegrating factor eJ dx = e",
Now
d
.
dx [eXw] = O gives eXw = c.
Therefore w
u'
ce-x and u
cle-x.
A second solution is Y2 = 1 . e-x = e-x.
We see by
observation that a particular solution is Yp = x. The general solution is
19. Define y = u(x)eX so
and
y" - 3y'
+ 2y =
eXu" - eXu' = O
If w = u' we obtain the first order equation w'-w
or
u" - u' = O.
= Owhich has the integratng factor e- J dx = e-x.
Now
d
dx [e-Xw] = O
gives
e-xw = c.
= u' = ce" and u = ce". A second solution is Y2 = eXex = e2x. To find a particular
solution we try yp = Ae3x. Then y' = 3Ae3x, y" = 9Ae3x) and 9Ae3x - 3 (3Ae3x) + 2Ae3x = 5e3x.
Therefore w
Thus A
= 5/2
and yp
20. Define y = u(x)eX
= ~e3x.
so
and
y" - 4y'
+ 3y
= eXu" - 2exu'
=O
or
u" - 2u'
= O.
If w = u' we obtain the first order equation w' - 2w = Owhich has the integrating factor e-2 J dx =
e-2x. Now
d
dx [e-2xw] = O gives e-2xw = c.
= ce2x and u = cle2x. A second solution is Y2 = eXe2x = e3x. To find a particular

solution we try yp = ax + b. Then y~ = a, y~ = O, and 0- 4a + 3(ax + b) = 3ax - 4a + 3b = x.
Then 3a = 1 and -4a + 3b = O so a = 1/3 and b = 4/9. A particular solution is YP = ~x + ~ and
Therefore w = u'
the general solution is
113
Exercises 4.2
2L (a) For mI constant, let YI
= emx. Then
Yl
= mIemX and
y't
= myemX Substituting into the
dfferential equation we obtain
+ by + CYI = amremX + bmIemX + cemX
ay~
Thus, YI = emX will be a solution of the differential equation whenever amI
+ bm; + e = o.
Since a quadratic equation always has at least one real or complex root, the differential equation
must have a solution of the form YI
= emx.
(b) Write the differential equation in the form

y"
+ -yb, + -ye = O,
a
and let YI = emx be a solution. Then a second solution is given by

Y2 = emX
= emX
J --e
J
e-bx/a
2mX dx
e-(b/a+2mtlxdx
emxe-(b/a+2m)
bja
+ 2mI
bja
+ 2mI
Thus, when mI =1- -bj2a,

When mI = -bj2a
(mI =1- -bj2a)
e-(b/a+m)
.
a second solution is given by Y2
= em2x where m2 = -bja - mI
a second solution is given by

Y2 = emX
dx = xemx.
(e) The functions

sin x
sinhx
= 2i (etX
=
e -tX)
2(eX - e-X)
are all expressible in terms of exponential functions.

22. We have Yl = 1 and Y't = 0, so xY't - xYl + YI = O-x + x =
differential equation. Letting y = U(X)YI(X) = xu(x) we get
y' = xu'(x)
Then xy" - xY'
substitution w
+y =
x2u"
+ 2xu'
+ u(x)
and
- x2u' - xu
y" = xu"(x)
+ xu =
and YI(X)
= x is a solution of the
+ 2u'(x).
x2u" - (x2 - 2x)u'
O. If we make the
u', the second-order linear differential equation becomes x2w' - (x2 - x)w
114
= 0,
Exercises 4.2
which is separable:
dw =
dx
(1- ~)w
x
dw
-:;;
= ( 1In w
=x
1)
~ dx
+e
- In x
eX
Then u'
= el-
x
el eX/ x and u = el J e" dx / x o To integrate eX/ x we use the series representation for e" o
Thus, a second solution is

Y2 = xu(x) = elX
= el x
= elX
1(
eX
-;- dx
- 1+x
x
el
+ -x
x2
2!
3!
(~
+ 1 + 2.x + 2.x
x
2!
3!
=elx ( lnx+x+2(2!)x
+-
+ ooo) dx
o o
o)dX
1
+ 3(3!)x
4)
( x In x + x 21+31--x
2(2!)
+ --x
+000
+ ooo
3(3!)
An interval of definition is probably (0,00) because of the In x termo

23. (a) We have y' = y" = eX, so
xy" - (x
and y
+ lO)y' + 10y
= xe" - (x
+ 10)eX + IOe"
= O,
= eX is a solution of the differential equation.
(b) By (5) a second solution is

Y2 = Yl
J P(x)dx
X
dx = e
Yl
x+lnx10
= eX
e2x
dx = eX
x=t
eJ
10
dx
X
dx = e
eJ(1+l0/x)dx
e
dx
2
X
x10e-x dx
= eX( -3,628,800 - 3,628,800x - 1,814,400x2 - 604,800x3

- 30,240x5
5,040x6
= -3,628,800 - 3,628,800x
- 30,240x5
5,040x6
720x7
90x8
10x9
720x7
115
90x8
10x9
151,200x4
xl0)e-x
- 1,814,400x2 - 604,800x3
-
xlOo
151,200x4
Exercses 4.2
(e) By Corollary (A) of Theorem 4.2, --
1101
Y2
lO!
L -n! xn
is a solution.
n=O
Exercises 4.3
= Oand m = -1/4 so that y = Cl + C2e-x/4.
2. From m2 - 36 = O we obtain m = 6 and m = -6 so that y = cle6x + C2e-6x.
1. From 4m2
3.
+ m = O we obtain
= 3 and m = -2 so that y = cle3x + C2e-2x.

m = 1 and m = 2 so that y = clex + C2e2x.
From m2 - m - 6 = O we obtain m
+ 2 = O we obtain
m2 + 8m + 16 = O we obtain m = -4 and m = -4 so that
m2 - 10m + 25 = O we obtain m = 5 and m = 5 so that y
4. From m2 - 3m
5. From
6. From
7. From 12m2 - 5m - 2 = O we obtain m
= Cle-4x + c2xe-4x.
+ c2xx.
= clx
= -1/4 and m = 2/3 so that y
= cle-x/4
- 1 = O we obtain m == -2 V5 so that y = cle( -2+v's)x
8. From m2
+ 4m
9. From m2
+ 9 = O we obtain m = 3i and m = -3i so that y = Clcos 3x + C2sin 3x.
+ C2e(-2-v's)x.
+ 1 = O we obtain m = i/vIJ and m = -i/vIJ so that y = Clcosx/vIJ + C2sinx/vIJ.

m2 - 4m + 5 = O we obtain m = 2 iso that y = e2x(cl cos x + C2sin x).
2m2 + 2m + 1 = O we obtain m = -1/2 i/2 so that
10. From 3m2

11. From
12. From
y = e-x/2(cl cos x/2

13. From 3m2
+ 2m + 1 = O we obtain
14. From 2m2
3m
+ 4 = O we obtain
y
From m3 - 4m2 - 5m
m = 3/4
= e3x/4
+ C2sin x/2).
= -1/3 V2i/3 so that
y = e-x/3 (Cl COS J2 x/3
15.
+ C2e2x/3.
+ C2sin J2 x/3) .
V23i/4 so that
(Cl COS V23 x/4
+ C2sin v'23 x/4)
= O we obtain m = O, m = 5, and m = -1 so that
16. From m3 - 1 = O we obtain m = 1 and m = -1/2 vlJi/2 so that

y = qeX
17. From m3 - 5m2
+ e-x/2
+ 3m + 9 = O we obtain
(C2cos J3 x/2
m = -1, m
y = Cle-x
+ C3sin J3 x/2)
= 3, and m
+ C2e3x + C3xe3x.
116
= 3 so that
Exercises 4.3
18. From m3 + 3m2
4m - 12 = O we obtain m
= -2, m = 2, and m = -3 so that
+ C2e2x + C3e-3x.
y = cle-2x
19. From m3 + m2
2 = O we obtain m = 1 and m
= clet + e-t(c2cost
u
20. From m3 - m2
4 = Owe obtain m
= -1
= 2 and m
iso that
+ C3sint).
= -1/2 V7i/2 so that
x = Ce2t + e-t/2 (C2 cos J7 t/2 + C3sin J7 t/2) .

21. From m3 + 3m2 + 3m + 1 = O we obtain m = -1, m = -1, and m = -1 so that
y
= Ce -x + C2xe -x + C3X2e -x .
22. From m3 - 6m2 + 12m - 8 = Owe obtain m = 2, m = 2, and m = 2 so that

y = CIe2x
23. From m4 + m3 + m2
= Owe obtain m
+ C2xe2x + c3x2 e2x.
= O, m
= O, and m = -1/2 V3 i/2 so that
y = CI + C2X + e-x/2 (C3 cos V3x/2

24. From m4 - 2m2 + 1 = Owe obtain m
+ C4sin V3x/2)
= 1, m = 1, m = -1, and m = -1 so that
25. From 16m4 + 24m2 + 9 = Owe obtain m = V3 i/2 and m

y = CIcos V3x/2
26. From m4 - 7m2
= V3 i/2 so that
+ C2sin V3 x/2 + C3Xcos V3 x/2 + C4Xsin V3 x/2.
18 = Owe obtain m
= 3, m = -3,
and m = J2 i so that
y = CIe3x + C2e-3x + C3cos J2 x + C4sin J2 x.

27. From mS + 5m4
2m3
10m2 + m + 5
= Owe obtain m
= -1, m = -1, m
= 1, and m = 1, and
m = -5 so that
U=Cle
28. From 2ms - 7m4 + 12m3 + 8m2
x
+C2re
-r
+C3e
+C4re
= Owe obtain m = O, m
+cse
= O, m
-Sr
= -1/2, and m
= 2 2i so that
= CI+ C28 + C3e-s/2 + e2s(c4 cos 28 + Cssin 28).
29. From m2 + 16 = Owe obtain m

then Cl = 2, C2= -1/2,
-r
= 4i so that y = CICOS 4x + C2sin 4x. If y(O) = 2 and y' (O) = -2
and y= 2cos4x-
~sin4x.
30. From m2 + 1 = O we obtain m = i so that y = CIcose + C2sine. If Y(7f/3) = O and y'(7f/3)

1
V3
V3
1
.
then "2C+ 2C2 = O, -2CI
+ "2C2= 2, so CI = -V3, C2= 1, and y = -V3 cose + sm e.
117
=2
Exercises 4.3
31. From m2 - 4m - 5
and y'(l)
y =
-tel-x
O we obtain m
+ c2es =
2, then cle-l
-1 and m
+ 5c2es =
O, -cle-l
5, so that y
2, so Cl
+ c2x.
= -e/3, C2 =
cle-x
If Y(l) = O
e-s /3, and
+ tx-s.
= O we obtain m = -1/2 and m
32. From 4m2-4m-3
= 5 then
= 1, -~Cl
= 3/2 so that y
= 5, so ci =
= cle-x/2+c2e3x/2.
Ify(O)
=1
-7/4, C2= 11/4, and y
= _e-x/2+ lie3x/2.
33. From m2+m+2 = Owe obtain m = -1/2/7
i/2 so that y = e-x/2 (c cos /7 x/2 + C2sin /7 x/2).
If y(O) = O and y'(O) = O then Cl = O and C2= O so that y = O.
34. From m2 - 2m + 1 = O we obtain m = 1 and m = 1 so that y = cex + c2xex. If y(O) = 5 and
y'(O) = 10 then Cl = 5, Cl + C2= 10 so Cl = 5, C2= 5, and y = 5ex + 5xex.
35. From m3+12m2+36m
= Owe obtain m = O,m = -6, and m = -6 so that y = Cl+c2e-6x+c3xe-6x.
and y'(O)
If y(O) = O, y'(O)
c +C2
+~C2
1, and y"(O) = -7 then

Cl + C2= O,
-6C2
+ C3= 1,
36c2 - 12c3 = -7,
+ gxe-6x.
m = 2, and m =
so ci = 5/36, C2= -5/36, C3= 1/6; and y = 356 - 3s6e-6x

36. Frorn m3
+ 2m2
5m - 6 = Owe obtain m
y
If y(O)
= O, y'(O)
= O, and y"(O)
-1,
-3 so that
= cle-x + C2e2x + c3e-3x.
= 1 then
so Cl = -1/6, C2= 1/15, C3= 1/10, and

1
y = __e-x
6
1
1
+ _e2x
+ _e-3x.
15
10
37. From m2 - 10m + 25

38.
= Owe obtain m = 5 and m = 5 so that y = clx + c2xx. If y(O) = 1 and

y(l) = O then Cl = 1, cles + C2es = O, so Cl = 1, C2= -1, and y = x - xesx.
Frorn m2 + 4 = O we obtain m = 2i so that y = Clcos2x + c2sin2x.
If y(O) = O and Y(7r) = O
then c = O and
39. Frorn m2
y = C2sin
2x.
+ 1 = O we obtain
+ C2sinx.
m = i so that y = Cl COSX
If y'(O) = O and y'(7r/2) = 2
then Cl = -2, C2= O, and y = -2cosx.
+ 2 = O we obtain m = 1 iso
ci = 1 and Y(7r) = e1l'COS7r = =e",
40. Frorn m2 - 2m
Y(7r)
1 then
that y
eX(cl cosx
+ C2sinx).
If y(O)
1 and
Since _e1l' =f 1, the boundary-value problem has
no solution.
41. The auxiliary equation is m2 - 3 = O which has roots -V3
= cleV3x + C2e-V3x.
and V3. By (10) the general solution
= clcoshV3x + c2sinhV3x. For

y = cleV3x + c2e-V3x the initial conditions imply Cl + C2= 1, V3Cl - V3C2 = 5. Solving for Cl + C2
is y
By (11) the general solution is y
118
Exercises
4.3
we find CI = i(3 + 5V3) and C2 = i(3 - 5V3) so y = i(3 + 5V3)ev'3x + i(3 - 5V3)ev'3x. For
y = CIcosh V3x + C2sinh V3x the initial conditions imply CI = 1, V3C2 = 5. Solving for CI and C2
we find CI = 1 and C2= V3 so y = cosh V3x +
iV3 sinh V3x.
42. The auxiliary equation is m2 - 1 = O which has roots -1 and 1. By (10) the general solution is
y
= qeX
+ c2e-x.
By (11) the general solution is y = CIcoshx + C2sinhx. For y
boundary conditions imply CI+C2 = 1, cle-c2e-1
= clex
+ c2e-x the
= O. Solving for CI and C2we find CI = 1/(1+e2)
and C2= e2/ (1+ e2) so y = eX/(1 + e2) + e2e-x /(1 + e2). For y = CIcosh x + C2sinh x the boundary
conditions imply CI = 1, C2= - tanh 1, so y = cosh x - (tanh 1) sinh x.
43. The auxiliary equation should have two positive roots, so that the solution has the form y =
cl1X + c22X. Thus, the differential equation is (f).
44. The auxiliary equation should have one positive and one negative root, so that the solution has the
= cl1X + c2e-k2X. Thus, the differential equation is (a).

45. The auxiliary equation should have a pair of complex roots a bi where
form y
has the form
eax (CIcos bx
< O, so that the solution
+ C2sin bx). Thus, the differential equation is (e).
46. The auxiliary equation should have a repeated negative root, so that the solution has the form
= cle-x + C2xe-x. Thus, the differential equation is (c).

47. The differential equation should have the form y" + k2y = O where k
y
= 1 so that the period of the
solution is 27r. Thus, the differential equation is (d).

48. The differential equation should have the form y" + k2y
= O where k
= 2 so that the period of the
solution is n. Thus, the differential equation is (b).

49. (a) The auxiliary equation is m2 - 64/L = O which has roots 8/JI.
of the differential equation is x = CIcosh(8t/JI)
(b) Setting x(O)
= Xo and x'(O)
Thus, the general solution
+ C2sinh(8t/JI).
= O we have CI = xo, 8C2/JI
= O. Solving for CI and C2we get
CI = Xo and C2= O, so x(t) = xocosh(8t/JI).

(e) When L = 20 and Xo
= 1, x(t)
= cosh(4tv's).
The chain will last touch the peg when
x(t) = 10. Solving x(t) = 10 for t we get tI = tv'scosh-I10::::::

chain at this instant is X'(tl) = 12il75
50. Both -e[l]
1.67326. The velocity of the
::::::
17.7989 ft/s.
and CI represent arbitrary constants, and each may take on any real value.
51. Since (m- 4) (m+5)2
= m3 +6m2
-15m
-100 the differential equation is ylll +6y" -15y' -100y
= O.
The differential equation is not unique since any constant multiple of the left-hand side of the
differential equation would lead to the auxiliary roots.
52. A third root must be m3 = 3 - i and the auxiliary equation is
(m + 21) [m -
(3 + i)][m - (3 - i)] = (m +
1 (m2
2)
119
11
- 6x + 10) = m3 - 2m2
+ 7m + 5.
Exercises 4.3
T4e differential equation is
11
y'" - -;y" + 7y' + 5y = O.

53. From the solution YI = e-4x cos x we conclude that mI = -4 + i and m2 = -4 - i are roots of the
auxiliary equation. Hence another solution must be
m3
+ 6m2 + m -
34 by [m - (-4
+ i)][m
Y2 =
- ( -4 - i) 1 = m2
e-4x sin x. Now dividing the polynomial
+ 8m + 17 gives m
- 2. Therefore m3 = 2
is the third root of the auxiliary equation, and the general solution of the differential equation is
54. Since l/x
-+
O as x
-+ 00,
to the solutions of y"
+y
we would expect the solutions of y"
+ (l/x)y' + y
= Oto behave similar
O; that is, like sin x and cos x for large values of x.
Solutions of
xy" + y' + xy = Oare obtained using an ODE solver and are shown below with the indicated initial
conditions.
y
y
1
1
0.5
-0.5
-1
-1
y(l)
= O,
y'(l)
=2
y(l)
= 2,
y'(l)
=O
55. Factoring the difference of two squares we obtain

m4
+ 1 = (m 2 + 1) 2 -
2m2 = (m2
+ 1 - V2 m) (m 2 + 1 + V2 m)
Using the quadratic formula on each factor we get m
= O.
= V2/2 V2i/2. The solution of the
differential equation is
y(x)
= eV2x/2
(CI cos
V; x +
56. (a) The auxiliary equation m2
C2 sin
V; x) +
+ bm + C = Ohas
e-V2x/2
solutions m
(C3
cos
V;
+ C4 sin
V; x) .
= (-b Jb2 - 4c )/2. If b :::;O, then
> O, and the solution cannot approach

O as x -+ oo. Thus, for the solution to approach O we must have b > O. Now, if c < O then
Jb2 - 4c > b and -b + Jb2 - 4c > O. Thus y(x) cannot approach O. Finally, if e > O then
Jb2 - 4c Oand e > O.
(b) If b2 - 4c > O, then y = cIemX + C2em2x and the only solution satisfying y(O) = O, y(l) = Ois
the solution will contain a term of the form ef3x for (3
= O.
120
Exercises 4.3
If b2 - 4c
= o, then
= eemx + C2temx.
Again, the only solution satisfying y(O)
= o, y(l) = O
is y = O.
If b2
4c
Now y(O)
< O then
= O implies
CI = O and
y
= c2e-bx/2 sin 4c - b2 x.
If we are to have a nontrivial solution, the condition y(l)

4c - b2
O implies ~
= nm or
= n27l'2 for n a positive integer.
57. The auxiliary equation is m2 + A = O and we consider three cases.

Case I When A
O the general solution of the differential equation is y = CI + C2X. The boundary
conditions imply O = y(O)
= CI and
= y( 7l'/2)
= C27l'
/2, so that c
= C2 =
O and the problem
possesses only the trivial solution.
< O the general solution of the differential equation is y = cleV-"Xx+ C2e-V-"Xx,

or alternatively, y = CIcosh H
x + C2sinh H
x. Again, y(O) = O implies CI = O so y =
c2sinhHx.
The second boundary condition implies O = Y(7l'/2) = c2sinhH7l'/2
or C2= O. In
Case II When A
this case also, the problem possesses only the trivial solution.
Case III
When A
> O the general solution of the differential equation is y
C2sin Vf.. z: In this case also, y(O)
= O yields
condition implies O = C2sin Vf.. 7l'/2.
Vf.. = 2k for k
CI = O, so that y
When
Vf.. 7l'/2
= C2sin Vf.. x.
= CICOS Vf.. x
The second boundary
is an integer multiple of
tt ,
that is, when
a nonzero integer, the problem wiJI have nontrivial solutions. Thus, for A = 4k2 the
boundary-value problem wiJI have nontrivial solutions y

On the other hand, when
Vf.. is not
= C2sin 2kx, where k is a nonzero integer.
an even integer, the boundary-value problem will have only the
trivial solution.
58. Applying integration by parts twice we have
J eax f(x) dx
= ~ eax f(x) - ~ J eax f'(x) dx

= ~eaxf(x)
a
- ~ [~eaxf'(x)
a a
- ~Jeax!"(x)dx]
a
= ~ eax f(x) - 2_ eax f'(x) + 2_ J eax f"(x) dx.

a
a2
a2
dx
= ~ eax f(x) -
Collecting the integrals we get

J eax (f(x)
:2
!"(x))
121
:2
eax f'(x).
4.3
Exercises
In order for the technique to work we need to have
eax (f(x)
:2
f"(x))
dx
eax f(x) dx
or
f(x)
- ~ f"(x)
a
= kf(x),
where k =1- O. This is the second-order differential equation

f"(x)
If k
< 1, k
+ a2(k
- l)f(x)
= O.
=1- O, the solution of the differential equation is a pair of exponential functions, in which
case the original integrand is an exponential function and does not require integration by parts for
its evaluation. Similarly, if k
= 1, f" (x) = O and
f (x) has the form f (x)
ax
+ b.
In this case a
single application of integration by parts will suffice. Finally, if k > 1, the solution of the differential
equation is
f(x)
= Clcos av'k=!
+ C2sin a..Jk=l
x,
and we see that the technique will work for linear combinations of cos ax and sin ax.
59. Using a CAS to solve the auxiliary equation m3 - 6m2
m2
= -0.270534,
= 0.658675, and m3 = 5.61186. The general solution is
60. Using a CAS to solve the auxiliary equation 6.11m3

ml
+ 2m + 1 we find ml
+ 8.59m2 + 7.93m + 0.778 = Owe find
= -0.110241, m2 = -0.647826 + 0.857532i, and m3 = -0.647826 - 0.857532i. The general
solution is
y = Cle-O.l1024lx+ e-O.647826x
(C2
cos 0.857532x
61. Using a CAS to solve the auxiliary equation 3.15m4

ml
-1.74806, m2
- 5.34m2
+ C3sin 0.857532x).
+ 6.33m
- 2.03
= O we find
= 0.501219, m3 = 0.62342 + 0.588965i, and m4 = 0.62342 - 0.588965i. The
general solution is
y = Cle -1.74806x
+ c2eO.50l2l9x+ eO.62342x (C3 cos 0.588965x + C4sin 0.588965x).
62. Using a CAS to solve the auxiliary equation m4

m2
= 1/2 - /3i/2,
m3 = -1/2
+ V7i/2,
+ 2m2
and m4
- m + 2 = O we find ml = 1/2 + /3
= -1/2 - V7i/2. The general solution is
122
i/2,
Exercises
+ 3m3 - 16m2 + 15m - 4 = O we obtain m = -4, m = ~, m = 1, and m = 1, so that

+ c2ex/2 + C3ex + C4xex. If y(O) = -2, y'(O) = 6, y"(O) = 3, and y"'(O) = ~, then
63. From 2m4

y
4..4
= cle-4x
CI + C2 + C3 = -2
1
-4CI
+ "2C2+ C3+ C4 = 6
16cI
+ C2 + C3+ 2C4 = 3
-64cI
_
116
_ 918
+ SC2 + C3+ 3C4 = "2 '
58
so CI - -'75 ' C2 - -"""3 ' C3- 25 ' C4- -"5 ' an
4 -4x
116 x/2
y = - 75 e
- Te
64.
From m4 - 3m3
y
+ 3m2
- m
= O we obtain m
= CI + C2ex + c3xex + C4x2ex.
918 x
58 x
25 e - 5xe
= O, m = 1, m = 1, and m
If y(O) = O, y'(O)
so CI = 2, C2 = -2, C3 = 2, C4 = -1/2,
= O,
= 1 so that
y"(O) = 1, and y"'(O) = 1 then
and
Exercises 4.4
1. From m2
yp
+ 3m + 2 = O we find
mI = -1 and m2 = -2. Then Yc = cle-x
+ C2e-2x
+ C2e-2x + 3.
y = cle~x
2. From 4m2
+ 9 = O we find mI = -~
i and m2 = ~i. Then Yc = CI COS ~x + C2sin ~x and we assume
yp = A. Substituting into the differential equation we obtain 9A = 15. Then A

3
y = CI COS "2 x
3.
and we assume
= A. Substituting into the differential equation we obtain 2A = 6. Then A = 3, yp = 3 and
From m2 - 10m
+ 25
. 3
= ~, yp
= ~ and
+ C2SIn "2 x + "3 .
= O we find mI = m2 = 5. Then Yc = clx
+ c2xx
and we assume
yp = Ax + B. Substituting into the differential equation we obtain 25A = 30 and -lOA

- 6 ' B --:5 6 ' yp -- 6:5x + 6:5' an d
Th en A -:5
123
+ 25B
= 3.
Exercises 4.4
4. From m2
+ m - 6 = O we find m1 = -3 and m2 = 2. Then Yc =
Ax
-3x
y=C1e
+C2e
2x
1
--x--
+ C2e2x and we assume

= 2 and A - 6B = O. Then
C1e-3x
+ B.
Substituting into the differential equation we obtain -6A

1 YP -- -"J1x - 18'
1 an d
A -- -3'1 B-- -18'
Yp
18
+ m + 1 = O we find m1 = m2 = O. Then Yc = C1e-2x + C2xe-2x and we assume

Yp = Ax2 + Bx + C. Substituting into the differential equation we obtain A = 1, 2A + B = -2,
and ~A + B + C = O. Then A = 1, B = -4, C = ;, Yp = x2 - 4x +;, and
5. From tm2
y = C1e-2x
6. Frorn m2 - 8m
+ c2xe-2x + x2
+ 2"'
- 4x
+ 20 = Owe find m1 = 2 + 4i and m2 = 2 - 4i. Then Yc = e2x(c1 cos4x + C2sin4x)

+ Bx + C + (Dx + E)ex. Substituting into the differential equation we
and we assume Yp = A.x2

obtain
2A - 8B
+ 20C
13E = O
-6D+
-16A
=O
+ 20B
=O
13D = -26
20A = 100.
Then A = 5, B
i,
= 4, C =
y
D = -2, E
-H, Yp = 5x2 + 4x + i + (-2x
+ C2sin4x) + 5x2 + 4x + ~~+ ( -2x
e2x(c1 cos 4x
iD eX and
- ~~) eX.
7. Frorn m2 + 3 = O we find m1 = Y3i and m2 = -Y3i.

Then Yc = c1cosY3x + C2sin Y3x
and we assume Yp = (Ax2 + Bx + C)e3x. Substituting into the differential equation we obtain
+ 6B + 12C = O, 12A + 12B =

Yp = ( -4x2 + 4x - 1) e3x and
2A
+B
Then A
= Clcos Y3 x + C2sin Y3 x + ( -4x2 + 4x
8. Frorn 4m2 - 4m - 3 = Owe find ml

YP = A cos 2x
O, and 12A = -48.
= -4,
= 4, C = -j,
- ~) e3x.
= ~ and m2 = -~ . Then Yc = cle3x/2 + C2e-x/2 and we assume
sin 2x. Substituting into the differential equation we obtain -19 - 8B = 1 and
8A - 19B = O. Then A
-lis,
y=cle
=-
4~5 '
YP =
- 41iscos 2x
3x/2 +C2e -x/2 --cos

19
425
124
4~5
2x--sm 8.
425
sin 2x, and

2x.
Exercises 4.4
= clex + C2 and we assume Yp = Ax.
Substituting into the differential equation we obtain -A = -3. Then A = 3, Yp = 3x and
9. From m2 - m = O we find mI = 1 and m2 = O. Then Yc
y =
Cl e"
+ C2 + 3x.
+
10. From m2
2m
O we find
+ Ex + Cxe-2x.
Yp = Ax2
mI
and
-2
m2 =
O. Then Yc
4A = 2, and -2C = -l. Then A = ~, E = 2, C = ~, Yp = ~x2 + 2x

Y = cle-2x
C2
and we assume
+ 2E =
5,
+ ~xe-2x, and
+ C2 + _x2 + 2x + _xe-2x.
+ ~ = O we find mI = m2 = ~.
+ Ex2ex/2. Substituting into the differential
11. From m2 - m
Yp = A
cle-2x
Substituting into the differential equation we obtain 2A
Then Yc
= Clex/2 + C2xex/2 and we assume

= 3 and 2E = 1. Then
equation we obtain ~A
A = 12, E = ~, Yp = 12 +~x2ex/2, and
12. From m2 - 16
Yp = Axx.
and
13. From
m2
= O we find
mI
+4
= O we find mI = 2i and m2 = -2i.
Yp = Ax cos 2x
+ Ex
-4A = 3. Then A
sin 2x. Substituting
= -~,
E = O, Yp
y
14. From m2
Yp = (Ax3
= 4 and m2 = -4.
Then Yc = clx
+ C2e-4x
and we assume
Substituting into the differential equation we obtain 8A = 2. Then A = ~, YP = ~xx
Then Yc =
Cl COS2x
C2 sin
2x and we assume
into the differential equation we obtain 4E = O and
= -ixcos2x,
ci COS2x
and
+ C2sm 2x - '4 x cos 2x.
+ 4 = O we find mI = 2i and m2 = -2i. Then

+ Ex2 + Cx) cos 2x + (Dx3 + Ex2 + Fx) sin 2x.
we obtain
2E
+ 4F
6A+8E
=O
=O
12D = O
-4C
+ 2E
-8E+6D
=-3
=O
-12A = l.
125
Yc = Cl COS2x
+ C2 sin 2x
and we assume
Substituting into the differential equation
Exercises 4.4
Then A
=-
/2 ' B
0, C
= ~ , D = 0,
-h, F = 0, YP = ( -112x3 + ~~x) cos 2x + l6x2 sin 2x,
and
+ C2sin 2x + ( -112x3 + ~~x)
y = C cos 2x
15. From m2
(Ax2
+ 1=
+ Bx)
cos x
we find mi
= i and
+ (Cx2 + Dx)
m2
+ 2C =
into the differential
O. Then A
-4,
YP = -lOx
14 3
+ 75x +
5x
= ce
1 4
+ C2-
10x
we obtain -20A
equation
and 2G - 5D = 6. Then A = -l~'

53 2 697
d
250x - 625x, an
0, C
0, D
= ~, YP =
= O. Then Yc = cle5x + C2 and we assume YP =
we find mi = 5 and m2
Substituting
= -~, B =
= -1,
6B - 10G
1 4
we obtain 4C = 0,
equation
1 2
l.
2'x cosx+2xsmx.
Y=C1COSX+C2sinx16. From m2 - 5m =
Ax4+Bx3+Cx2+Dx.
2x.
= -i. Then Yc = Cl COS X + C2sin x and we assume YP =
sin x. Substituting
2A + 2D = 0, -4A = 2, and -2B

- ~x2 COS X + ~x sin x, and
+ 116x2sin
cos 2x
14 3
53
= 25io',D = -~~~,
697
+ 75x + 250x
i:,
= 2, 12A-15B
- 625x.
17. Prom m2 - 2m + 5 = we find m = 1+ 2i and m2 = 1- 2i. Then Yc = eX (Cl COS 2x + C2sin 2x) and
we assume YP = Axex cos 2x + Bxe" sin 2x. Substituting
into the differential equation we obtain
4B = 1 and -4A
= O. Then A
= 0, B =
i, YP = ixex
y = eX(c cos 2x
18. From m2 - 2m
+2
+ 2B
= 1 and - 2A
+ C2sin 2x) + xex
we find mI = 1 + i and m2
and we assume YP = Ae2x cos x
sin 2x, and
+B
+ Be2x
= 1-
sin x. Substituting
sin 2x.
i. Then Yc
= eX(c cosx + C2sinx)
equation
we obtain
= -3. Then A = ~, B = - i ' YP = ~e2x cos x - ie2x sin x and
y = eX(C cos x
7 2
+ C2sin. x) + Se
X cos x
l.
- Se 2x sin x.
19. Prom m2 + 2m + 1 = we find mI = m2 = -l. Then Yc = ce-x + C2xe-x and we assume

YP = A cos x + B sin x + C cos 2x + D sin 2x. Substituting into the differential equation we obtain
2B
= 0, -2A = 1, -3G + 4D = 3,
YP =
-"21 cos
x -
9 COS 2 X
25
and -4G - 3D
2
2
25 sin x, an d
Y = ce20.
From m2
+ 2m
assume YP = A
- 24 =
+ C2xe-
2'1 cos x
we find m = -6
+ (Bx2 + Cx)x.
Substituting
= O.
Then
A = -~, B = 0, C = -295' D =
9
- 25 cos 2x
and m2
4.
12
+ 25 sin 2x.
Then
126
~,
Yc
equation
cle-6x
+ C2x
we obtain -24A
and we
= 16,
Exercises 4.4
2E+10C
and
= -2, and 20E = -1. Then A = -~, E =

y = CIe
21. From m3 - 6m2
-6x
-io'
2 (120 x
+ C2e4x - 3" -
C=
2
-lgo' Yp = _~_(iox2
+ I~OX)x,
19)
4x
100 x e .
= O we find mI = m2 = O and m3 = 6. Then Yc = CI + C2X+ c3x and we assume
Yp = Ax2 + E cos x + C sin x. Substituting into the differential equation we obtain -12A
6E - C
and
= -1, and E + 6C = O. Then A = -~, E =

6x
y = CI + C2X+ C3e
22. From m3 -2m2 -4m+8
= O. Then A = ~, E
1 2
6
1 .
x - 37 cos x + 37 sin x.
-"4
= cle2x+c2xe2x+c3e-2x
Substituting into the differential equation we obtain 24A = 6
= -136' YP = (~x3 - 136X2)

e2x, and
= Cle2x +
3,
= 3\' Yp = _~x2 - 367cosx+ 3\ sinx,
= Owe find mI = m2 = 2 and m3 = -2. Then Yc
and we assume yp = (Ax3 + Ex2)e2x.

and 6A + 8E
-:f7,
C2xe2x + C3e-2x + (~x3 _ 2.x2) e2x.

4
16
= O we find mI = m2 = m3 = 1. Then Yc = Clex + C2xex + C3x2ex and

we assume Yp = Ax + E + Cx3ex. Substituting into the differential equation we obtain -A = 1,
23. From m3 - 3m2 + 3m - 1

3A - E
= O, and
6C = -4. Then A = -1, E

y
= CIe
+ C2xex + C3X2xe
= Owe find mI
24. From m3-m2-4m+4
and we assume Yp = A+Exex+Cxe2x.

-3E
= -3,
= 1, m2
C =
-, Yp = -x
- 3 - ~x3ex, and
23x e .
- x - 3 - 3"x
= 2, and m3 = -2. Then Yc = clex+c2e2x+c3e-2x
Substituting into the differential equation we obtain 4A = 5,
= -1, and 4C = 1. Then A = ~, E
t, C = ~, Yp = ~+ txex
+ ~xe2x, and
5 1
1
y = clex + C2e2x + C3e-2x + - + -xex + _xe2x
4
25. From m4 + 2m2 + 1 = O we find mI = m3 = i and m2 = m4 = -i. Then Yc = CIcosx + C2sinx +

C3Xcos x + C4Xsin x and we assume Yp = Ax2 + Ex + C. Substituting into the differential equation
we obtain A
= 1, E
= -2, and 4A + C
= 1. Then A = 1, E
= -2, C
= -3, Yp = x2 - 2x - 3, and
y = C cosx + c2sinx + C3Xcosx + c4xsinx + x2 - 2x - 3.

26.
From m4 - m2
= O we find mI
= m2 = O, m3 = 1, and m4
and we assume Yp = Ax3 + Ex2 + (Cx2 + Dx)e-x.

obtain -6A
= 4, -2E
O, lOC - 2D
= -1. Then
Yc
= CI+ C2X+ C3ex+ C4e-x
Substituting into the differential equation we
= 0, and -4C = 2. Then A = -~, E = O, C = -~,
D = -~, Yp = _~x3 - Gx2 + ~x) e-x, and
127
Exercises
4.4
27. We have Ye
we find A
= Cl COS2x + C2sin 2x and
= - ~.
we assume Yp
= A.
Thus y = ClCOS2x + C2sin 2x - ~. From the initial conditions we obtain c
and C2= j2, so y
= j2 sin 2x - ~ .
28. We have Ye = cle-2x + c2ex/2 and we assume Yp = Ax2 + Ex + C. Substituting
equation we find A
=O
-7, E
= -19, and
From the initial conditions we obtain Cl =

1 -2x
y = -Se
e=
-37. Thus y
= cle-2x
+ c2ex/2 - 7x2 - 19x - 37.
-! and C2= l~6 , so
186 x/2
+5
e -
7x - 19x - 37.
= cle-x/5+c2 and we assume Yp = Ax2+Ex. Substituting into the differental equation

we find A = -3 and E = 30. Thus y = cle-x/5 + C2- 3x2 + 30x. From the initial conditions we
29. We have Ye
obtain Cl = 200 and C2= -200, so

y
= 200-x/5
200 - 3x2
+ 30x.
30. We have Ye = cle-2x + c2xe-2x and we assume YP = (Ax3 + Ex2)e-2x.
differential equation we find A = ~ and E = ~. Thus y
Substituting
= cle-2x + C2xe-2x +
(ix3
into the
+ ~x2) e-2x.
From the initial conditions we obtain Cl = 2 and C2= 9, so

y
= 2e-2x + 9xe-2x + (~x3 + ~x2) e-2x.
= e-2x(cl COSX+C2sinx) and we assume YP = Ae-4x. Substituting into the differential

equation we find A = 5. Thus y = e- 2x(ClCOSX + C2sin x) + 7e-4x. From the initial conditions we
31. We have Ye
obtain Cl = -10 and C2= 9, so

y = e-2X(-10cosx + 9sinx + 7e-4x).
32. We have Ye
= Cl coshx
+ C2sinhx and we assume Yp
= Ax cosh x + Ex
sinh z. Substituting into the
differential equation we find A = O and E = ~. Thus

Y
c coshx + c2sinhx + '2xsinhx.
From the initial conditions we obtain Cl = 2 and C2= 12, so

y
= 2coshx + 12sinhx + ~xsinhx.
33. We have x = Cl coswt + C2sinwt and we assume xp
= Atcoswt + Etsinwt.
Substituting into the
differential equation we find A = -Fo/2w and E = O. Thus x = Clcoswt+C2 sinwt-(Fo/2w)tcoswt.

From the initial conditions we obtain Cl = O and C2= Fo/2w2, so
x
= (Fo/2w2)sinwt - (Fo/2w)tcoswt.
128
Exercises 4.4
34. We have x = q cos wt
+ C2sin wt
equation
X
+B
and we assume xp = A cos ,t
we find
A = Fo/(w2
= C1cos wt
sin ,t, where , =J w. Substituting
B = O. Thus
- ,2) and
Fo
+ c2 sin wt + (2w -,
2) cos ,t.
From the initial conditions we obtain Cl = Fo/(w2 _,2) and C2 = O, so

Fo
Fo
X
35. We have Yc
differential
Cl
= (w 2 -,'
+ C2ex + C3xex
we find A
equation
2, B
and
+ Bx2ex + c-.
Ax
Substituting
into the
C = ~. Thus
+ C3xe x + 2X + 9xex + 2x
12X 2ex +"2e

1 5x
equation
we find A
= C1e-2x + eX(c2 cos j3 x + C3sin
1
2
+ _x.
- 12x2ex
+ eX(c2 cos j3 X + C3sin j3 x)
36. We have Yc = c1e-2x
-12,
2) cos-rz.
-,
we obtain C1 = 11, C2 = -11, and C3 = 9, so
y = 11 - llr.
Substituting
+ (2w
and we assume yp
y = C1 + C2ex
Frorn the initial conditions
2) coswt
and we assume
yp = Ax
t ' B = - ~, and
J3 x) + 4"x -
= ~.
+ B + Cxe-2x.
Thus
8" + "3xe-2X.
Fr om thee iun't'la 1 con diti

. C1 = -12
23 ' C2 = -"24'
59 an d C3 = 72
17 v;:;3
1 ions we o b tam
o , so
2x +e x (59
r:
17 r: :
--cosv3x+-v3smv3x
24
72
y=--e- 23
12
37. We have Yc = C1COS X + C2sin x and we assume yp = A 2 + Bx
equation
we find A
1, B
and y(l)
= O we obtain
= O, and
1
5
+-x--+-xe-.
4
8
;:;)
+ C.
C = -1. Thus y = C1COSX
Substituting
2x
3
+ C2sinx + x2 - 1.
From y(O)
=5
C1- 1 = 5
(cos 1)C1 + sin(1)c2 = O.
Solving this system we find C1 = 6 and C2 = -6 cot 1. The solution of the boundary-value problem
is
y = 6 cos x - 6( cot 1) sin x

38. We have Yc = eX(c1 cosx
equation
we find A
+ C2sin x)
+ x2
and we assume yp = Ax
1 and B = O. Thus y
- 1.
+ B.
Substituting
= eX(c1 COSX+C2 sinx)+x.
we obtain
C1 = O
129
From y(O)
= O and Y(7r) = 7r
Exercises 4.4
Solving this system we find Cl = O and C2 is any real number.
The solution
of the boundary-valus
problem is
39. We have Yc
Cl COS2x
+ C2sin 2x
and we assume Yp
we find A
equation
O and B =
we have y = C3cos 2x
[O,7r/2]. On (7r/2,00)
=A
cos x
1 Thus
+ C4sin 2x.
+B
sin x on [O,7r/2]. Substituting
+ C2sin 2x + 1sin x
y = q cos 2x
on
Frorn y(O) = 1 and y'(O) = 2 we obtain
q =1
1
Solving this system

Now continuity
we find Cl
1 and
1= -C3
Hence C3 =
. Continuity
-i = -2C4
differential
have y
= eX(q
cos3x
={
cos 3x
Solving this system,
7r
+ c2sin3x)
+ C4sin3x).
is
O:::; x :::;tt /2
x> 7r/2.
Frorn y(O) = O and y'(O)
= -2,
+ 3q
~.
Thus y
+ C2sin 3x) + 2 on
=
into the
[O,ttJ. On (7r,00)' we
O we obtain
= O.
= eX(
-2 cos3x
+ ~sin3x) + 2 on
[O,7r].
= 7r implies
+ ~Sin37r) + 2 = e1l'(c3cos37r + c4sin37r)
or C3 = -2e-1l'(1
+ e"}.
continuity
of y' at 7r implies
230e1l'sin 37r = e1l'[(c3 + 3C4) cos 37r + (-3C3

or -c3e1l' - 3C4e1l'= O. Since C3 = -2e-1l'(1
problem
problem
and we assurne yp = A on. [O,7rJ. Substituting
we find q = -2 and C2 =
of y at x
+ 2e1l' = -c3e1l'
initial-value
on [O,7r/2J.
+ ~sin 2x + 1sin x,
~ cos 2x + ~sin2x,
e1l'(-2cos37r
or 2
cos 2x
Now, continuity
+ 1sin x
'3 cos 2' = - 2C3 sin 7r + 2C4 COS7r
we find A = 2. Thus, y = e" (Cl cos 3x
equation
= eX(c3
2x
of y' at x = 7r/2 implies
Then C4= ~ and the solution of the initial-value
y(x)
40. We have Yc
+ ~sin
'65 sin. 7r + '31 sin. 2'7r = C3cos 7r + C4sin. 7r
- 2 sin 7r + '3 cos 7r +

or
= cos 2x
of y at x = 7r/2 implies
COS7r +
or -1 +
'3 + 2C2 = 2.
C2 = ~. Thus y
+ e")
+ C4) sin 37rJ
we have C4 = ~e-1l'(1
+ e"}.
is
y(x) =
ex (- 2 cos 3x
{ (1
+ sin 3x),
+ e1l')eX-1l' (-2
cos 3x
130
+ ~ sin 3x),
z > x,
The solution of the
Exercises
41. (a) From Yp = Ax
we find Y~ = Akx
and Y~ = Ak2x.
Substituting
4.4
equation we get
aAk2ekx
so (ak2
+ bk + c)A =
(b) From Yp = Axx
+ bAkx + cAx
= (ak2
+ bk + c)A
+ bm + e = O, A
1. Since k is not a root of am2
we find Y~ = Akxx
= ekx,
= lj(ak2
+ Ax and Y~ = Ak2xx + 2Akx.
+ bk + e).
Substituting into
the differential equation we get

aAk2xx
+ 2aAkekx + bAkxx + bAx + cAxekx

= (ak2 + bk + c)Ax
+ (2ak + b)Aekx
+ (2ak + b)A
= (O)AxX
= (2ak + b)AX = x
+ bk + e = O because k is a root of the auxiliary equation. Now, the roots of the

auxiliary equation are -bj2a Jb2 - 4ac, and since k is a root of multiplicity one, k =J:. -bj2a
and 2ak + b =J:. o. Thus (2ak + b)A = 1 and A = lj(2ak + b).
where ak2
(e) If k is a root of multiplicity two, then, as we saw in part (b), k = -bj2a

From Yp = Ax2x we find Y~ == Akx2x + 2Axx
Substituting into the differential equation, we get
aAk2x2ekx
and Y~ = Ak2x2x
and 2ak
+ 4Akxx
+ b = O.
= 2Ax.
+ 4aAkxx + 2aAx + bAkx2ekx + 2bAxekx + cAx2x

= (ak2 + bk + c)Ax2ekx + 2(2ak + b)Axekx + 2aAx
= (0)Ax2ekx
+ 2(0)Axekx + 2aAekx = 2aAekx =
ekx.
Since the differential equation is second-order, a =J:. O and A = lj(2a).

42. Using the double-angle formula for the cosine, we have
sin x cos 2x = sinx(cos2 x - sin2 x) = sinx(l-
2sin2 x) = sinx - 2sin3 x.
Since sin x is a solution of the related homogeneous differential equation we look for a particular
solution of the form YP = Ax sin x
+ Ex
cos x
+ e sin3 x.
we obtain
2a cos x
+ (6c -
2b) sinx - 8csin3 x = sinx - 2sin3 x.
Equating coefficients we find a = O, e = ~, and b = ~. Thus, a particular solution is

1
YP = -x cos x
43. (a)
f(t) = et sin t. We see that Yp --*
(b ) f(t)
= e-t. We see that Yp --*
00
00
as t
as t
1 . 3
+ -4
sm x.
--* 00
--* 00
and Yp --* O as t
and Yp --*
131
00
as t
--*
--*
-oo.
-oo.
Exercises 4.4
(e) f(t)
= sin 2t. We see that Yp is sinusoidal.
(d) f(t)
= 1. We see that Yp is constant and simply translates Yc vertically.
= e2x(Cl cos 2x + C2 sin 2x). We assume a particular solution of

= (Ax3 + Bx2 + Cx)e2x cos 2x + (Dx3 + Ex2 + F)e2x sin 2x. Substituting into the
44. The complementary function is Yc

the form Yp
differential equation and using a CAS to simplify yields
[12Dx2 + (6A + SE)x + (2B + 4F)Je2x cos 2x

+ [-12Ax2 + (-SB + 6D)x + (-4C + 2E)]e2x sin 2x
= (2x2 - 3x)e2x cos 2x
+ (lOx2 -
X -
1)e2Xsin2x.
This gives the system of equations
12D = 2,
6A+SE
-12A = 10,
-SB+6D=-1,
i,
2B
= -3,
from which we find A = -~, B =

C = ~, D
solution of the differential equation is
+ 4F
= 0,
-4C+2E=-1,
= ~, E =
i, and
F =
-k.
Thus, a particular
YP = ( _~x3 + ~x2 + ~x )e2X cos 2x + (~x3 + ~x2 - ~x )e2X sin 2x.

45. The complementary function is Yc =
Cl
cos X
+ C2 sin x + C3X
COS X
+ C4X
sin x. We assume a particular
solution of the form Yp = Ax2 cos X + Bx3 sin x. Substituting into the differential equation and using
a CAS to simplify yields
(-Sa
This implies -Sa
+ 24b) cos x + 3bx sin x = 2 cos x - 3x sin x.
+ 24b = 2 and -24b =
-3. Thus b = ~, a = ~, and Yp = ~x2 cos X
Exercises 4.5
1.
(9D2 - 4)y = (3D - 2)(3D + 2)y = sinx
2. (D2 - 5)y = (D - V5)(D + V5)y = x2 - 2x

3. (D2 - 4D - 12)y = (D - 6)(D + 2)y = x - 6
4. (2D2 - 3D - 2)y
= (2D + l)(D
- 2)y
+ ~x3 sin x.
=1
5. (D3 + lOD2 + 25D)y = D(D + 5)2y = e"

6. (D3 + 4D)y = D(D2 + 4)y = eX cos 2x
7. (D3 + 2D2 - 13D + lO)y = (D - l)(D - 2)(D + 5)y = xe":"
8. (D3 + 4D2 + 3D)y = D(D + l)(D + 3)y = x2 COS X
132
3x
Exercises 4.5
+ 8D)y
9. (D4
10. (D4 - 8D2

11. D4y
+ 2)(D2
= D(D
+ 16)y =
(D - 2)2(D
+ 4)y
=4
+ 2)2y =
(x3 - 2x)X
D4(lOx3 - 2x) = D3(30x2 - 2) = D2(60x) = D(60) = O
= 8Dex/2
12. (2D - l)y = (2D - 1)4ex/2

13. (D-2)(D+5)(e2x+3e-SX)
+ 64)(2cos8x
14. (D2
- 2D
= 4ex/2
- 4ex/2
- 4ex/2 = O
= (D-2)(2e2x-15e-sx+5e2x+15e-SX)
- 5sin8x) = D(-16sin8x
= -128 cos 8x
- 40 cos8x)
= (D-2)7e2x
+ 64(2cos8x
+ 320 sin 8x + 128 cos 8x
= 14e2x-14e2x
- 5sin8x)
- 320 sin 8x = O
15. D4 because of x3
16. DS because of x4
17. D(D - 2) because of 1 and e2x
18. D2(D - 6)2 because of x and xe6x
19. D2
+ 4 because of cos 2x
21. D3(D2
22. D2(D2
23. (D
20. D(D2
+ 16) because o x2 and sin 4x

+ 1)(D2 + 25) because of x, sin x,
+ l)(D
+ 1) because of 1 and sinx
and cos 5x
- 1)3 because of e-x and x2ex
24. D(D - l)(D - 2) because of 1, eX, and e2x

25. D(D2 - 2D
+ 5)
because of 1 and eX cos 2x
+ 2D + 2) (D2
26. (D2
- 4D
+ 5) because
of e-x sin x and e2x cos x
27. 1, x, x2, x3, x4

28. D2
+ 4D = D(D + 4); 1, e-4x
29. e6x, e-3x/2
30. D2 - 9D - 36 = (D - 12)(D
+ 3);
e12x, e-3x
31. cos y'5 x, sin y'5 x
+ 10 = D2 - 2(3)D + (32 + 12); e3x cosx,

10D2 + 25D = D(D - 5)2; 1, x, xx
32. D2 - 6D
33. D3 -
e3x sinx
35. Applying D to the differential equation we obtain
D(D2
9)y
= O.
Then
y = ..._,___..,
ce3x + C2e -3x +
Yc
133
C3
=O
Exercises 4.5
and YP = A. Substituting Yp into the differential equation yields -9A = 54 or A = -6. The general
solution is
D(2D2 -7D + 5)y = O.

Then
Yc
and YP
A. Substituting Yp into the differential equation yields 5A
-29 or A
-29/5.
The
general solution is
D(D2 + D)y
= D2(D + l)y = O.
Then
Yc
and yp = Ax. Substituting YP into the differential equation yields A = 3. The general solution is
y
= Cl + c2e-3x + 3x.
38. Applying D to the dfferential equation we obtain
Then
Yc
and Yp = Ax. Substituting YP into the differential equation yields A
39. Applying D2 to the differential equation we obtain
Then
Yc
134
10. The general solution is
Exercises 4.5
and Yp = Ax + B. Substituting Yp into the differential equation yields 4Ax
Equating coefficients gives
4A
+ (4A + 4B)
= 2x
+ 6.
=2
4A +4B = 6.
Then A = 1/2, B = 1, and the general solution is
Then
Yc
and Yp = Ax2
+ Bx.
Substituting Yp into the differential equation yields 6Ax
+ (2A + 3B) = 4x -

6A=4
2A+3B
= -5.
Then A = 2/3, B = -19/9, and the general solution is

Y = q + C2e
-3x
+ 'Jx
19
2
-
gx.
Then
Y
= C! + C2X + C3e-x + C4X4 + CSX3 + C6X 2

Yc
and Yp = Ax4
+ Bx3 + Cx2.
Substituting Yp into the differential equation yields

12Ax2
+ (24A + 6B)x + (6B + 2C)

12A = 8
24A+6B
6B
+ 2C
135
= O.
= 8x2.
5.
Exercises 4.5
Then A = 2/3, B
= -8/3, C = 8, and the general solution is

y = ci
+ C2X + c3e-x + ~x4

3
- ~x3
3
+ 8x2.
Then
!jc
and YP = Ax3
+ Bx2 + Cx + D.
Ax3
+ (B
- 6A)x2
Substituting Yp nto the differential equation yields
+ (6A
- 4B
+ C)x + (2B
- 2C
+ D) = x3 + 4x.
Equatng coefficients gives

A=1
B - 6A
6A - 4B+C
2B - 2C
Then A
= 1, B
+D
=O
=4
= O.
= 6, C = 22, D = 32 , and the general solution is
= clex + C2xex + x3 + 6x2 + 22x + 32.
43. Applying D - 4 to the differential equation we obtain

(D - 4)(D2 - D - 12)y = (D - 4)2(D
+ 3)y
= O.
Then
!jc
and YP = Axe4x.
Substituting
Yp into the differential equation yields 7Ax
coefficients gives A = 1/7. The general solution is
44. Applying D - 6 to the differential equation we obtain

(D - 6)(D2
+ 2D + 2)y
Then
!jc
136
= O.
=:
x.
Equating
Exercises 4.5
and Yp = Ax.
Substituting
coefficients gives A
Yp into the differential equation yields 50Ax
= 5x.
Equating
= 1/10. The general solution is

y = e -x
(Cl COS X
+ C2 sin x) + lOe 6x.
45. Applying D(D - 1) to the differential equation we obtain

D(D - 1)(D2 - 2D - 3)y
= D(D - l)(D + l)(D - 3)y =
O.
Then
Yc
and yp = Aex
+ B.
Substituting Yp into the differential equation yields -4Aex
Equating coefficients gives A
= -1
y
46. Applying D2(D
+ 2) to
= 3. The general solution

= cle3x + C2e-x - eX + 3.
and B
is
the differential equation we obtain
D2(D
+ 2)(D2 + 6D + 8)y
+ 2)2(D + 4)y
= D2(D
= O.
Then
Yc
and YP = Axe-2x
+ Bx + C.
2Ae-2x
+ 8Bx + (6B + 8C)
= 3e-2x
+ 2x.

2A= 3
8B
6B+8C=
Then A = 3/2, B = 1/4, C
O.
= -3/16 , and the general solution is

3 -2x
y = cre- 2x + C2e- 4x + -2xe
47. Applying D2
+ 1 to
+ -x1 - -.3
4
16

(D2
+ 1)(D2 + 25)y =
O.
Then
y
= Cl cos 5x + C2 sin 5x + C3 cos x + C4 sin x

Yc
137
- 3B = 4ex - 9.
Exercises
4.5
and Yp = A cos x. + B sin x. Substituting Yp into the differential equation yields

24A cosx
+ 24B sinx = 6 sin x.
Equating coefficients gives A = O and B = 1/4. The general solution is

y = Cl
+ 1) to
48. Applying D(D2
COS 5x
+ C2 sin 5x + 4 sin x.
D(D2 + 1)(D2 + 4)y = O.

Then
y = Cl
COS 2x
+ C2 sin 2x + C3 cos X + C4 sin x + Cs

Yc
+ B sin x + C.
and Yp = A ~os x
3A cos x
= 4/3,

y
+ 3B sin x + 4C = 4 cos x + 3 sin x
= -2.
B = 1, and C
- 8.
= Cl COS 2x + C2 sin 2x + ~ cos x + sin x
- 2.
49. Applying (D - 4)2 to the differential equation we obtain
Then
Yc
and Yp = Axx
+ Bx.

49Axx
+ (14A + 49B)e4x
= -xx.

49A = -1
14A
Then A = -1/49,
+ 49B
= O.
B = 2/343, and the general solution is

y
= cle-3x + C2xe-3x
-xx
49
50. Applying D2(D - 1)2 to the differential equation we obtain
138
+ _e4x
343
Exercises
Then
Yc
and Yp = AxeX
+ Be + Cx + D.
Substituting Yp into the differentialequation yields
+ (5A
-6Axex
- 6B)eX
10Cx
+ (3C
-6A
=1
- 10D)
= xe" + x.
5A - 6B
=1
-10C
3C - lOD = O.
Then A
= -1/6, B = -5/36, C = -1/10, D

= ce2x
+C2e
-Sx
= -3/100, and the general solution is
1
-"6xe
5
-36e
1
3
-10x-100'
51. Applying D(D - 1)3 to the differential equation we obtain

D(D - 1)3(D2 - l)y
= D(D
- 1)4(D
+ l)y = O.
Then
y = cl e x
+ C2e-x + C3X3e x + C4X2e x + csxe x + C6
'----v---'
Yc
and yp = Ax3ex
+ Bx2ex + Cxe" + D.
6Ax2
Substituting yp into the differential equation yields
+ (6A + 4B)xe
+ (2B + 2C)e
+ 5.
D = x2ex

6A = 1
6A+ 4B
2B
=O
+ 2C = O
-D= 5.
Then A = 1/6, B
= -1/4, C = 1/4, D = -5, and the general solution is

y = cj e
52. Applying (D
+ 1)3 to the
1 3 x
+ C2e-x + "6x
e -
1 2 x
4x e
1 x
+ 4xe
differential equation we obtain

(D
+ 1)3(D2 + 2D + l)y
139
= (D
+ l)Sy
= O.
4.5
Exercises
4.5
Then
Yc
+ Bx3e-x + Cx2e-x.
and Yp = Ax4e-x
+ 6Bxe-X + 2Ce-X = x2e-x.
12Ax2e-x
1~'
B = O, and C
+ 2 to
2D
general solution is
1
+ c~xe-x
+ _x
4e-x
~
12
Y = c 1e-x
53. Applying D2
= O. The

(D2
2D
+ 2)(D2
+ 5)y =
2D
O.
Then
y
= eX(C1cos 2x + C2 sin 2x) + eX( C3 cos x + C4 sin x)

Yc
and Yp = Aex cos x
+ Be"
sin x. Substituting Yp into the differential equation yields

3Aex cos x
+ 3Bex
sin x = eX sin x.
Equating coefficients gives A = O and B = 1/3. The general solution is

y = eX(c1 cos 2x
54. Applying D2
+ 10 to
2D
(D2
2D
+ C2 sin 2x) + ~ex
sin x.
+ 10)
(D2
+D +
y = (D2
+ 10)
2D
(D
~f
y
= O.
Then
Yc
and Yp = Aex cos 3x
+ Be"
sin 3x. Substituting Yp into the differential equation yields
(9B - 27A/4)eX cos 3x - (9A
+ 27B /4)eX
sin 3x
= _ex cos 3x + eX sin 3x.

27
--A+9B=-1
4
-9A - 27 B
4
1.
Then A = -4/225, B = -28/225, and the general solution is

Y
= c e-x/2 + c xe-x/2
1
4
_ex
225
140
28
cos 3x - _ex
225
sin 3x.
Exercises 4.5
55. Applying D2
+ 25 to

(D2 + 25)(D2
+ 25) = (D2 + 25)2 = O.
Then
y = Cl cos 5x
+ C2 sin 5x + C3X
cos 5x
+ C4X
COS 5x
Yc
and Yp = Ax cos 5x
+ Bx
sin 5x. Substituting Yp into the differential equation yields

10B cos 5x - lOA sin 5x = 20 sin 5x.
= -2 and B = O. The general solution is

=
y
56. Applying D2
+ 1 to
Cl COS 5x
+ C2 sin 5x -
2x cos 5x.

(D2
+ 1)(D2 + 1) = (D2 + 1)2 = O.
Then
y = Cl
COS X
+ C2 sin x + C3X
cos x
+ C4X
COS X
Yc
and Yp = Ax cos x
+ Bx
sin x. Substituting Yp into the differential equation yields

2B cos x - 2A sin x = 4 cos x - sin x.
Equating coefficients gives A = 1/2 and B = 2. The general solution is

y = Cl
57. Applying (D2
+ 1)2 to
COS X
. x + '2x
1 cos x
+ C2 sin
- 2'
x sm x.

(D2
+ 1)2(D2 + D + 1)
= O.
Then
y = e.-x/2
[Cl
/3 x
cos 2
/3 x ] + C3 cos x + C4 sin. x + C5X

+ C2 Slll. 2
cos x
.x
+ C6X sin
Yc
and Yp = A cos x+B sin z-l-Cz cos x+ Dx sinx. Substituting YPinto the differential equation yields
(B
+ C + 2D) cos x + Dx
cos x
+ (-A -
2C
+ D) sin x -
=O
B+C+2D
D=O
-A -2C+D
=O
-C = 1.
141
Cx sin x = x sin x.
Exercises 4.5
Then A
= 2, B = 1, C = -1, and D = 0, and the general solution is

y = e -xj2
58. Writing cos2x
[Cl
=!(1 +cos2x)
/3
cos -
. /3] + ')~
+ C2 sm -
+ sm x
cos x
- x cos x.
and applying D(D2 +4) to the differential equation we obtain

D(D2
+ 4)(D2 + 4)
COS 2x
+ C2 sin 2x + C3Xcos 2x + qx
= D(D2
+ 4)2
= O.
Then
y = Cl
sin 2x
+ C5
Yc
and YP = Ax cos 2x
+ Bx
+ C.
sin 2x
-4A sin 2x
Substituting Yp into the differential cquation yields
+ 4B
cos 2x
+ 4C = 2" + 2" cos 2x.
= 0, B = 1/8, and C = 1/8. The general solution is
Cl COS 2x
+ C2 sin 2x + "8x
+ "8'
sin 2x

D3(D3
+ 8D2)
= D5(D
+ 8)
= O.
Then
Yc
and Yp = Ax2
+ Bx3 + Cx4.
16A
+ 6B + (48B + 24C)x + 96Cx2
=2
+ 9x
- 6x2.
+ 6B
16A
=2
+ 24C =
48B
96C = -6.
Then A
= 11/256, B = 7/32, and C = -1/16, and the general solution is

Y=
11 ')
+ C2X + C3e- + 256x~ + 32x
+ 1) to the
1)2(D + 1)(D3
60. Applying D(D - 1)2(D

D(D -
Sx
Cl
1 4
- 16x .
differential equation we obtain

- D2
+D
- 1) = D(D - 1)3(D
+ 1)(D2 + 1)
Then
y=cle
+C2COsx+c3smx
+C4+cSe
Yc
142
-x
+C6xe
2 x
+C7X e
= O.
Exercises 4.5
and
YP
= A + Be-X + Cxe" + Dx2ex. Substituting
Yp
into the differential equation yields
4DxeX + (2C + 4D)eX - 4Be-X - A = xe" - e-x + 7.

4D = 1
2C
+ 4D = O
-4B = -1
-A=
Then A = -7, B
1/4, C
y=qe
= -1/2,
x
and D
= 1/4,
7.
and the general solution is
.
+c2COsx+c3smx-7+e
1 -x
-"2xe
+x
e.
61. Applying D2 (D - 1) to the differential equation we obtain
Then
Yc
and
YP = A
+ Bx + Cx3ex.
Substituting
YP
(-A + 3B) - Ex + 6Cex = 16 - x + e",

-A+3B=16
-B=
-1
6C = 1.
Then A = -13, B = 1, and
e = 1/6,
62. Writing (eX + e-x)2 = 2 + e2x + e-2x and applying D(D - 2)(D
+ 2) to
the differential equation we
obtain
D(D - 2)(D + 2)(2D3
3D2 - 3D + 2) = D(D - 2)2(D + 2)(D + 1)(2D - 1) = O.
Then
Yc
143
Exercises 4.5
and YP = A
+ Bxe2x + Ce-2x.
Substituting.p,
2A
+ 9Be2x
= 2 + e2x + e-2x.
- 20Ce-2x
Equating coefficients gives A = 1, B = 1/9, and C = -1/20. The general solution is
63. Applying D(D - 1) to the differential equation we obtain
Then
Yc
and YP = 'Ax2
+ Bx2ex.
Substituting Yp into the differential equation yields 2A
= 1/2 and
= 1/2. The
+ 2Bex
= 1 + eX.
general solution is
64. Applying D3(D - 2) to the differential equation we obtain
Then
Yc
and YP = Ax2
+ Bx3 + Cx4 + Dxe2x.

(-8A
+ 24C)
Substituting YP into the differential equation yields

- 24Bx - 48Cx2
+ 16De2x
= 5x2 - e2x.

-8A
+ 24C
=O
-24B
=O
-48C
=5
16D = -1.
Then A
= -5/16, B = O,C = -5/48, and D = -1/16, and the general solution is

2x
y = q + C2X + C3e
-2x
+ C4e
144
5
- 16x
5
- 48x
1 2x
- 16xe
Exercises 4.5
+ C2e-8x.
cle8x
Using D to annihilate 16 we find yp
A.
Substituting Yp into the differential equation we obtain -64A = 16. Thus A = -1/4 and
y = Cle
8x
+ C2e-8x
1
- 4
The initial conditions imply

5
+ C2 = "4
ci
8Cl - 8C2 = O.
Thus e = C2 = 5/8 and
5
y = _e8x
8
66. The complementary function is Yc = Cl
5
+ _e-8x
8
+ C2e-x.
1
4
Using D2 to annihilate x we find yp = Ax
Substituting Yp into the differential equation we obtain (A
+ 2B) + 2Bx
+ Bx2.
= x. Thus A = -1 and
B = 1/2, and
1
= Cl + C2e-x - x + -x
2
y' = -C2e-x - 1 + x.
Y

e+c2=1
= 1.
-C2
Thus
ci
= 2 and C2 = -1, and

Y =2- e
-x
1 2
x + "2x
.
= Cl +C2e5x. Using D2 to annihilate
Substituting YP into the differential equation we obtain (-5A+2B)-10Bx

and
B= -1/10, and
Y
= Cl + C2e5x + -x9 - -x1
1
= 5c2e5x + -9 - -x.
25
25

Cl
+ C2 = O
41
5C2 = 125
145
10
x-2
we find yp
= -2+x.
= Ax+Bx2.
Thus A = 9/25
Exercises 4.5
Thus CI = -41/125
and C2= 41/125, and

41
41
5x
= -125 + 125e
9
1 2
+ 25x - 10x .
68. The complementary function is Yc = clex + C2e-6x.

Yp = Ae2x. Substituting
Using D -
:2 to annihilate
lOe2x we find
Yp into the differential equation we obtain 8Ae2x = lOe2x. Thus A = 5/4
and

c
CI -
+ C2= -4
6C2
-"2.
Thus CI = -3/7 and C2= 5/28, and

3
y = __ex
7
5
+ _e-6x
28
5
+ _e2x
4
69. The complementary function is Yc = CICOSX+ C2sinx.

8cos2x
- 4sinx
we find Yp = Axcosx
+ Bxsinx
Using (D2 + 1)(D2 + 4) to annihilate
+ Ccos2x
+ Dsin2x.
differential equation we obtain 2B cos x - 3C cos 2x - 2A sin x - 3D sin 2x

A
= 2, B = O, C
= -8/3,
y
and D
Substituting Yp into the
= 8 cos 2x
- 4 sin x. Thus
= O, and
= CIcos x + C2sin x +
y ,. =-csmx+c2coSX+
2x cos x - ~ cos 2x
2 cosx-
2xsmx+3sm
16 . 2x.

8
3
c2+-=-1
-CI - 7f = O.
Thus Cl =
+t:
and C2= -11/3, and

y
-7fCOSX- -11.smx + 2x cosx - -8 cos 2x.

3
3
70. The complementary function is Yc = Cl + C2ex + C3xex. Using D(D - 1)2 to annihilate
xe" + 5 we find Yp = Ax + Bx2ex + Cx3ex.
Substituting Yp into the differential equation
146
Exercises 4.5
we obtain A + (2B + 6C)eX + 6Cxex = xeX + 5. Thus A = 5, B = -1/2,
y
= Cl +
and C = 1/6, and
1
1
c2ex + C3xex + 5x - _x2ex + _x3ex
y' = C2ex + c3(xeX + eX) + 5 - xe" + ~x3eX

1
1
y" = C2ex + c3(xeX + 2eX) - eX - xe" + _x2ex + _x3ex.
C2+ C3+ 5 = 2
Thus q
= 8, C2=
-6, and C3= 3, and
71. The complementary function is Yc = e2x(cl cos 2x + C2sin 2x). Using D4 to annihilate x3 we
find yp = A + Ex + Cx2 + Dx3. Substituting Yp into the differential equation we obtain

(8A - 4B + 2C) + (8B - 8C + 6D)x + (8C - 12D)x2 + 8Dx3 = x3. Thus A = O, B = 3/32,
C = 3/16, and D = 1/8, and
y=e2x(clcOs2x+c2sin2x)+
3
32x+
3
1
16x2+gX3
3
3
3
y' = e2x [cl(2cos2x - 2sin2x) + c2(2cos2x + 2sin2x)] + 32 + gX + gX2.
Thus Cl = 2, C2= -3/64,
and
y
72. The complementary
3
64
3
3
+ _x2
32
16
= e2X(2cos2x - - sin2x) + -x
function is Yc = Cl + C2X + C3x2 + c4ex.
1
+ _x3.
8
Using D2(D - 1) to annihilate
x + eX we find yp = Ax3 + Bx4 + Cxe". Substituting yp into the differential equation we obtain
147
Exercses 4.5
(-6A + 24B) - 24Bx + CeX = x + eX. Thus A = -1/6, B = -1/24, and C = 1, and
y = Cl
+ C2X + c3X2 + c4e - -x1314- -x

X
I
x
12
13
y = C2 + 2C3X + C4e - 2'x - '6x
12
y "2= c3 + C4ex - x - 2'x
24.
+ xe x
+ ex + xe x
+ 2xe + xe x .
2
Thus
Cl
= 2, C2 = 1, C3 = O, and
C4
+ C4 = O.
-2, and
x
y = 2 + x - 2e -
'61x 3 -
1 4
24 x
+ xe x .
73. To see in this case that the factors of L do not commute consider the operators
and (D
+ 4)(xD -
(xD - 1)(D + 4)
1). Applying the operators to the function x we find
(xD - 1)(D + 4)x = (xD2 + 4xD - D - 4)x
+ 4xDx - Dx - 4x
= xD2x
= x(O)
+ 4x(1) -
1-
4x = -1
and
(D + 4)(xD - 1)x = (D + 4)(xDx - x)

=
(D + 4)(x 1 - x) = O.
Thus, the operators are not the same,
Exercises 4.6
The particular solution, Yp = UlYl + U2Y2, in the following problems can take on a variety of forms,
especially where trigonometric functions are involved. The validity of a particular form can best be
checked by substituting it back into the differential equation.
148
Exercses 4.6
1. The auxiliary equation is m2
+ 1 = O,
so Yc
W=
Identifying
f (x)
= Cl COS X + C2 sin x
cosx
sinx
cosx
- sinx
and
I =1.
= sec x we obtain
I
Ul
=-
U~
sin x sec x
=-
cos x sec x
tanx
1.
Then
Ul
= ln I cos z],
U2
= x, and
y = Cl
COS X
+ C2 sin x + cos x In I cos z] + x sin x.
+ 1 = O, so Yc = Cl COS X + C2 sin x
W=
Identifying
f (x)
xl
cos
I sinx
sin
cosx
and
l.
= tan x we obtain
l.
ul
I
U2
= -smxtanx
cos2 X _. 1
= cosx - secx
cosx
= smx.
Then u 1 = sin x - ln I sec x

y
+ tan z], U2 = - cos x, and

= el cos x + C2 sin x + cos x (sin x - ln I sec :r + tan
+ 1 = O, so Yc =
W=
Identifying
f (x)
= sin x
Cl COS X
+ e2
xl) - cos x sin x.
sin x and
cosx
sinx =1.
cos z
I -sinx
we obtain
U~
I
U2
Then
= - sin2 x
.
= cos x sin x.
1
1
u 1 = - sin 2x - - x
4
2
U2
1.
1
sm x cos x - - x
2
2
=-
-2"1 cos2 x.
and
Y = el cos x
= el cos x
.
1.
2
+ C2 sm
x + - sm x cos x
222
+ e2 sin
1
x - 2"xcos x.
149
1
12.
- -x cos x - - cos x sm x
4.6
Exercises
1 = O, so Yc
cosx
w=
Identifying
f (x)
= sec x tan
= Cl COSX + C2sin x
= - tan2
= - sinx(secxtanx)
U2 = cosx(secxtanx)
Ul
=x
- tanx,
U2
= -In I cos z],
y = Cl COSX
+ C2sin x +
1 = O, so Yc
= tanx.
x cos x - sin x - sin x In I cos xl

x cos x - sin x In I cos x l
Cl COSX + C2sin x
cosx
w-
sinx
cosx
I -SInX
I =l.
=, -
SIn X COS x
U2 = cos3 x = cos x (1 Ul =
and
= cos2 x we obtain
Ul
Then
x = 1 - sec2 x
and
= Cl cos X + C3sin x +
Identifying f(x)
I =l.
x we obtain
U~
Then
sinx
cosx
I -sinx
and
sin 2 x) .
icos3 x, U2 = sinx - isin3 x, and

y = Cl cos x
14
.2
+ C2SIn X + '3 cos x + SIn X -
Cl COSX + C2sin x + '3
(cos2 X
1.4
'3 SIn X
sin2 x) (cos2 X
1. 2
+ '3 SIn X.
= Cl COSX + C2sin x + '3 cos2 X + '3 sin x
ClCOSX+C2SInX+'3
1 = O, so Yc = Cl COSX
w=
Identifying f(x)
= sec2
cos x
+ C2 sin
sinx I =l.
cos x
I -sinx
x we obtain
I
u ---1-
sinx
cos? x
U2 = sec x.
150
x and
sin2 x)
sin2 x
Exercises 4.6
Then
U =
1
---cosx
= -secx
U2 = In I sec x
+ tan
z]
and
y = c cos x
cos x
Identifying f(x)
+ sin x In I sec x + tan
+ C2 sin x
- cos x sec x
+ C2 sin x
- 1 + sin x In I sec x
1 = 0, so Yc
= cex + c2e-x
+ tan x l.
and
= coshx = !(e-X + eX) we obtain

,
1 2x
1
+-4
u = -e
,
1 1 2x
u2 = -- --e
Then
U
1 -2x
1
= --e
+-x
1 2x 1
U2 = --e --x
and
y = cie
= c3e
+ c2e -x
1 -x
- Se
1 (Xe
+ C4e-x + x
1
+ xe
1 x
1 -x
- Se - xe
- e -X)
= C3ex + C4e-x + 2"xsinhx.

8. The auxiIiary equation is m2
1 = 0, so Yc
= cex + C2e-x and
e-x
= -2.
-e-x
Identifying
f (x)
= sinh 2x we obtain
1
u~ = __e-3x
1
+ _ex
,
1 -x
1 3x
u2 = -e
--e
151
z]
Exercises 4.6
Then
Ul
1 -3x
12
1 x
+-e
= -e
1 -x
1 3x
--e
--e
U2 =
12
and
= cex + C2e-x + 3" sinh2x.

W =
Identifying f(x)
+ c2e-2x
4 = O, so Yc = ce2x
I 2e2x
e2x
I = -4.
e-2x
_2e-2x
= e2x/x we obtain u = 1/4x and
and
u2 = _x /4x.
Then
Ul
= -lnlxl,
4
U2 = --
X -dt
4 xo t
and
xQ
Identifying f(x)
+ c2e-3x
9 = O, so Yc = cle3x
I ;::
Ul
=-
I = -6.
_~~~3X
= 9x/e3x we obtain u = ~xe-6x and

1 -6x
24 e
U2 = -X
U2 =
1
- xe
and
152
and
-~x.
-6x
Then
> O.
Exercises
+ 3m + 2 =
W
Identifying f(x)
= 1/(1
+ eX)
(m
+ l)(m + 2) =
~:~x
_~~~2X
0, so Yc
= cle-x + C2e-2x and
= _e-3x.
we obtain
eX
u, - __
1 - 1 + eX
e2x
Then
Ul
= ln(l + eX),
y
U2 = ln(l
eX
= - 1 + eX = 1 + eX
u2
+ eX)
- e
- eX, and
= Cle-x + C2e-2x + e-x ln(l + eX) + e-2x ln(l + eX) - e-x

= C3e-x + C2e-2x + (1 + e-X)e-X ln(l + eX).
12. The auxiliary equation is m2 - 2m
+1=
w=
Identifying f(x)
= eX /
(1 + x2)
we obtain
,
u' 2 -
Ul
= -!In (1 + x2),
y
xexex
e2x (1+ x2)
+ 3m + 2 =
=
(m
z.
+ l)(m + 2) = 0, so Yc = cle-x + C2e-2x
1_e-x
e-x
e-2x
_2e-2x
1--
e-3x
= sin e" we obtain
,
u2
Ul = -
eXeX
1
-_e2x (1 + x2) - 1 + x2'
= clex + c2xex - ~ex In (1 + x2) + xe" tan-1
vV
Then
+ x2
= -1
U2 = tan-1 x, and
Identifying f(x)
= 0, so Yc = qeX + C2xex and
I eX
ul = -
Then
(m - 1)2
eX
e-x sin e"
= _e-3x
= _e2x sin e"
cos e", U2 = e" cosx - sin e", and

y = cle-x
+ c2e-2x
- e-x cos e"
153
+ e-x
cos e" - e-2x sin e"
and
4.6
Exercises 4.6
+1=
(m - 1)2
= O, so Yc = clet + c2tet and
= jet
et
Identifying f(t)
= tan-l t we obtain
I
=-
Ul
tetettan-lt
2t
e
Then
Ul
= -ttan
1 + t2
= --2-tan-
-1
t
t +"2
U2 = t tan -1t - ~ In ( 1 + t2)

and
= clet + C3tet + ~et [(t2 15. The auxiliary equation is m2
(1 + t2)].
t -In
+ 2m + 1 = (m + 1)2 = O, so Yc = qe-t + c2te-t and
w=
Identifying f(t) = e-tInt
1) tan-l
-t
_e-t
t e -t
-te-t
+ e-t
I -e -2t
-
we obtain
Then
ul
= --t1 2 In t + -t1 2
2
U2 = z ln t - t
and
Y = qe-t
= Clet
1
1
2
+ c2te-t - _t2e-t
In t + _t2e-t + t2e-t In t - t e:'
2
+ c2te-t + ~t2e-t
16. The auxiliary equation is 2m2
In t - ~t2e-t
4'
+ 2m + 1 = O, so Yc = e-x/2(
154
cl
cos x /2
+ C2 sin x /2) and
Exercises 4.6
w=
e-x/2 cos:!:
2
I _le-x/2
cos :!:- le-x/2
222
Identifying f(x)
=-
u2 = -
e-x/2 sin(x/2)2.jX
/2
e-x
= -4e
e-x/2 cos(x/2)2JX
/2
e-x
= 4e
Then
Ul
= -4
U2 = 4
y = e-x / 2 ( Cl cos -x
2
= jexsecx
+ 6 = O, so Yc = e" (Cl
6m
I e" cos x
+ C2 sin x)
e X smx
eX cos x
- e" sin x
= jx,
y = clex cosx
(eX sinx)(eX secx)/3

e2x
=-
u2 =
(eX cosx)(eX secx)/3

e2x
and
2x
+ e" sin x = e
= - -3 tan x
1
:3'
and
l'
+ C2ex cosx + :31n(cosx)eX
18. The auxiliary equation is 4m2 - 4m
+ 1 = (2m
w-= tex/2~
cos X
sin -x x et/2Jt cos -t dt.

2 xo
2
+ 4e-x/2
we obtain
I
U2
et/2Jt cos - dt
xo
2
eX cos x
ul
= j ln(cosx),
x/2 r:
x
v x cos -.
2
- 4e-x/2 cos -x x et/2Jt sin -t dt

2 xo
2
w=
x/2 r:: . x
v x sm 2
+ C2 sin -X)
Identifying f(x)
et/2Jt sin - dt
2
and
Ul
1 -x
Ze .
Xo
Then
= 2.jX we obtain
u}
Identifying f(x)
e-x/2 sin f2
I
cos :!:- lex/2 sin:!:
le-x/2
sin :!:
2
2
ex/2 .
I !ex/2
cosx
- 1)2 = O, so Yc
xex/2
!xex/2 + ex/2
we obtain
155
+ :3xeX
sinx.
= Clex/2 + C2xex/2
I = e",
and
Exercises 4.6
Then
1 (
2)3/2
= 12 1 - x
Ul
x r:--; l.
1
U2 = -V 1 - x- + - sin " x
and
19. The auxiliary equation is 4m2 - 1 = (2m - 1)(2m

W =
+ 1) = O, so Yc = clex/2 + C2e-x/2 and
I !eex/22 _!ee='? I = -1
2
.
x/
Identifying f(x)
U2 = -xex/4+
x/
= xex/2/4 we obtain ui = x[A and ui = -xex/4.
Then
Ul
= x2/8 and
eX/4. Thus
+ c2e-x/2 + ~x2ex/2
y = clex/2
_ ~xex/2
+ ~ex/2
4
1
1
= C3ex/2 + C2e-x/2 + _x2ex/2
__ xex/2
and
yf = ~c3eX/2 _ ~c2e-X/2
+ ~x2ex/2
16
+ ~xex/2
8
_ ~ex/2.
4

C3
1
1
- -C2
224
-C3
Thus
C3
= 3/4 and
C2
=1
C2
1
-
= O.
= 1/4, and
+m -
1 = (2m - l)(m
ex/2
W =
Identifying f(x)
I !ex/2
= (x + 1)/2 we obtain
156
+ 1) = O, so Yc = clex/2 + C2e-x
and
Exercises 4.6
Then
Ul
_e-x/2
U2
--xe
1
3
Gx - 2)
Thus
-x- 2
x/2 +C2e -x
y=cle
and
,
x/2
y =-Cle
-x
-c2e
-.

Cl -
C2 -
2= 1
C2 -
1
-Cl
Thus
Cl
= 8/3 and
C2
= 1/3, and
y =
+ 2m -
8 x/2
1 -x
3"e
+ 3"e
- x - 2.
8 = (m X
2e-2x - e-x
+ 4) = O, so Yc = cle2x + C2e-4x
2)(m
I=
e-4x
- 4 e -4x
w = 12e::
Identifying f(x)
O.
_6e-2x.
we obtain
,
ul =
1 -4x
1 -3x
-e
- -e
1 3x
u2= --e
6
1 2x
--e
3
Then
Thus
y = cle2x
= cle2x
+ C2e-4x
_ ~e-2x
+ C2e-4x
__ e-2x
+ ~e-x
12
and
y' = 2ee2x _ 4C2e-4x
18
1
+ _e-x
1
+ _e-2x
2
157
__
1
e-x.
9
+ ~e-x
18
_ ~e-2x
and
Exercises 4.6
q
5
36 = 1
C2 -
7
2q - 4C2 + 18 = O.
Thus
q = 25/36
C2 = 4/9,
and
and
25
y = _e2x
36
Identifying f(x)
4
+ _e-4x
9
__ 1e-2x
4
+ 4 = (m - 2)2
1
+ _e-x
9
= 0, so Yc = qe2x
+ C2xe2x and
= (12x2 - 6x) e2x we obtain
u = 6x2
- 12x3
U2 = 12x2 - 6x.
Then
= 2x3
3x4
U2 = 4x3
3x2.
Ul
Thus
y
q e2x
+ C2xe2x + (2x3
3x4) e2x
+ (4x3
3x2) xe2x
3x2)
+ 2e2x
= qe2x + C2xe2x + e2x (x4 _ x3)

and
y'
= 2qe2x + C2 (2xe2x + e2x) + e2x ( 4x3
(x4 _ x3)
=1
ci
2q
Thus
q =
1 and
C2 =
+ C2 = O.
-2, and
y = e2x _ 2xe2x
+ e2x (x4 _ x3) = e2x (x4 - x3
2x
1) .

y"
and identify f(x)
= x-1/2. From
Yl
+ .!:.y' +
x
(1 __4x21_) y = x-1/2
= x-1/2cosx
and Y2
158
= x-1/2sinx
we compute
Exercises 4.6
Now
I
= COSX,
.
= smx
so
Ul
u; = cosx
so
U2 = sinx.
COSX
+ c2x-1/2
sinx
+ x-1/2
COS X
+ C2X-l/2
sin x
+ x-l/2.
Ul
and
Thus
y = CIX-1/2
=
ClX-1/2
cos2 X
+ x-l/2
sin2

11
y
and identify f(x)
= sec(lnx)jx2.
sec(lnx)
+ ~Y + x2Y
From
Yl
= --x""'2--'-
= cos(lnx) and
sin(ln x)
cos(ln x)
w=
= sin(lnx) we compute
Y2
sin (In x)
cos(ln x)
1
x
Now
I
ul
=-
tan(ln x)
x
so
Ul
= In I cos(lnx)'
and
I
u2
= -1 so
x
U2 = In x.
Thus, a particular solution is

Yp = cos(lnx) In I cos(lnx)1
+m
= m(m2
1
w=
+ (lnx)
+ 1) = O, so Yc
cosx
-sinx
- cosx
159
sin(lnx).
= Cl
+ C2 COSX + C3 sinx
sinx
cosx
-sinx
=1.
and
Exercises
4.6
Identifying f(x)
= tanx we obt~in
cosx
sinx
- sinx
cosx
tan x
- cos x
- sin x
sinx
cosx
tanx
- sinx
= - sinx
cosx
u~ = W3 = O
O
-sinx
O
tan x
-cosx
= tanx
= - sin x tan x
cos2 X - 1
= cos x - sec x.
cosx
Then
ln I cos z]
Ul
U2
= cosx
U3
= sin x - in I sec x + tan z]
and
y
= Cl + C2 cos X + C3 sin x - in I cos z] + cos2 X
+ sin 2 x
= C4
for
-00
- sin x in I sec x
+ C2 COSX + C3 sinx
+ tan z]
ln I cos rl
>
sinx in I secx
+ tan
z]
< x < oo.
+ 4m = m (m2 + 4) = O, so Yc = Cl + C2 COS 2x + C3 sin 2x

1
W
cos 2x
= O -2sin2x
O -4 cos 2x
160
sin2x
2cos2x
-4 sin 2x
= 8.
and
Exercses
Identifying
f (x) = sec 2x
we obtain
,1
u1
4.6
= -W
1 =
8
1
-8
cos 2x
sin 2x
-2 sin 2x
2 cos 2x
sec 2x
-4 cos 2x
-4 sin 2x
sin2x
2 cos 2x
O sec 2x
,1
u3 = 8"W3
1
4
-4 sin 2x
1
cos 2x
1
8" O -2sin2x
O -4 cos 2x
= -4 sec 2x
= -- tan 2x.
sec 2x
Then
U1
1
= 8"In I sec 2x
U2
U3
1
= 8"lnlcos2xl
+ tan
2xl
-'4x
and
Y = CI + C2 cos 2x
for
+t:
/4
<
+ C3 sin 2x + 8"In I sec 2x + tan
2x I -
'4 x cos 2x + 8"sin 2x In I cos 2x I
<.7f /4.
6m+ 30 = O,which has roots 1 + 3i, so Yc
= eX(e cos 3x+
+ 30y = 15 sin x, which

+ B sin x and substituting
We consider first the differential equation 3y" - 6y'

undetermined coefficients. Letting
YPl
= A cos x
C2
sin 3x).
can be solved using

equation we get
(27A - 6B)cosx
+ (6a+
27b) sinx = 15sinx.
Then
27A - 6B = O and
6a
+ 27b = 15,
so A = ~ and B = 197 . Thus, YPl = (7 cos X + f.7 sin x. Next, we consider the differential equation
3y" - 6y' + 30y, for which a particular solution YP2 can be found using variation of parameters. The
Wronskian is
W=
Identifying f(x)
eX sin 3x
I = 3e2x.
3ex cos 3x + eX sin 3x
eX cos 3x
I eX cos 3x - 3ex sin 3x
= ~ex tan x we obtain
, = -'9sm
l.
3xtan 3x
u1
and
161
u2
l.
= '9 sin 3x.
Exercses
4.6
Then
Ul
. -3X) - In ( cos -3x + sm

. -3X)]
= -1.sm 3x + -1 [(In cos -3x - sm
27
27
1
u') = -- cos3x.
27
Thus
YP2
21eX7 cos3x [In( cos 3; _ sin 3;) _ In (cos 3; + sin 3;) ]
and the general solution of the original differential equation is
= (m-1)2 = 0, which has repeated root 1, so Yc = clex+c2xex.
28. The auxiliary equation is m2-2m+1
We consider first the differential equation y" - 2y' + y = 4x2 - 3, which can be solved using
undetermined coefficients. Letting
= Ax2 + Bx +
YPl
and substituting into the differential
equation we get
= 4x2 - 3.
AX2 + (-4A + B)x + (2A - 2B + e)

Then
A = 4,
-4A + B = 0,
and
2A - 2B +
e = -3,
so A = 4, B = 16, and e = 21. Thus, YPl = 4x2 + 16x + 21. Next we consider the differential
equation y" - 2y' + y = x-1ex, for which a particular solution YP2 can be found using variation of
parameters. The Wronskian is
w=
Identifying f(x)
= eX/x
we obtain u~
eX
xe"
eX xe" + eX
I = e2x
-1 and u2 = l/x.
Then
Ul
= -x and U2 = lnx, so that
and the general solution of the original differential equation is

y
= Yc +
YPl
YP2
= clex + c2xex + 4x2 + 16x + 21 - xe" + xe" lnx.
29. The interval of definition for Problem 1 is (-7r /2, 7r/2), for Problem 7 is (-00,00),
is (0,00), and for Problem 18 is (-1,1).

y
for -7r/2
for Problem 9
In Problem 24 the general solution is
= cj costln z ) +c2sin(lnx) + cos(lnx) In I cos(lnx)1 + (Inx)sin(1nx)
< lnx < 7r/2 or e-1r/2 < x < e1r/2. The bounds on lnx are due to the presence of sec(lnx)
in the differential equation.

30. We are given that Yl = x2 is a solution of x4y" + x3y' - 4x2y = O. To find a second solution we use
reduction of order. Let y = x2u(x). Then the product rule gives

y' = x2u' + 2xu
and
y"
162
= x2u" + 4xu' + 2u,
Exercises 4.7
so
X4y"
Letting w
= u',
this becomes xw'

dw
Thus, w = x-5
+ x3y' - 4x2y = x5(xu" + 5u')

+ 5w = O. Separating variables
5
= -- dx
In Iwl
and
solution of the homogeneous differential equation is Yc
and integrating we have
+ c.
= -5lnx
A second solution is then Y2
and u = -tx-4.
= O.
= x2x-4
= 1/x2, and the general
= CX2 + C2/x2. To find a particular solution,
YP' we use variation of parameters. The Wronskian is

'(IV =
I x2 1/x2 I = _. .
2x
Identifying f(x)
U2 =
-t lnx,
tam u =;jX
1/ X4 we obtai
_2/x3
-5
an d' u2 = -;jX
-1
Then u
i -4
-I6x
an d
so
Y
1
= - -x16
x 2 - -1 (ln x )x -2
4
-4
1
= - -x16
-2 - -1_2
x In x.
4

y = cx
C2
1
1
+ -x2
- -- -4x2
16x2
In x.
Exercises 4.7
1. The auxiliary equation is m2 - m - 2 = (m

2. The auxiliary equation is 4m2 - 4m
+1=
+ l)(m
- 2) = Oso that y
= CIX-l + C2x2.
(2m - 1)2 = O so that y = Clxl/2
+ C2xl/2Inx.
3. The auxiliary equation is m2 = Oso that y = CI + c2ln x.

4. The auxiliary equation is m2 - 4m = m(m - 4) = Oso that y = CI + C2x4.
5. The auxiliary equation is m2 + 4 = Oso that y
= CIcos(2ln x) + C2sin(2ln x).

6. The auxiliary equation is m2 + 4m + 3 = (m + l)(m + 3) = Oso that y = CIX-l + C2X-3.
+ C2x2+v'6.
is m2 + 2m - 4 = O so that y = CjX-I+v's + C2X-I-v's.
is 25m2 + 1 = Oso that y = Ccos (i In x) + C2(! In x)
7. The auxiliary equation is m2 - 4m - 2 = Oso that y = clx2-v'6

8. The auxiliary equation
+ 1) = O so that y = CIX/2 + C2X-1/2

11. The auxiliary equation is m2 + 4m + 4 = (m + 2)2 = O so that y = CIX-2 + c2x-2Inx.
12. The auxiliary equation is m2 + 7m + 6 = (m + l)(m + 6) = Oso that y = CIX-l + C2X-6.
10. The auxiliary equation is 4m2 - 1 = (2m - 1)(2m
13. The auxiliary equation is 3m2+3m+
1 = Oso that y
163
= x-I/2
[CIcos
(V;
In x)
+ C2sin
(V;
lnx ) ].
Exercses 4.7
14. The auxiliary equation is m2 - 8m+ 41 = O so that y
= x4
[C1cos(51nx) + C2sin(51nx)]:
15. Assuming that y = xm and substituting into the differential equation we obtain
m(m - l)(m - 2) - 6 = m3
3m2 + 2m - 6 = (m - 3)(m2 + 2)
= O.
Thus
y
16. Assuming that y
= C1x3 + C2cos (hin x) + C3sin ( h In x) .
= xm and substituting into the differentiaI equation we obtain
m(m -l)(m
- 2) + m - 1 = m3
3m2 + 3m - 1 = (m - 1)3 = O.
Thus
17. Assuming that y
= Xffi and substituting into the differential equation we obtain

= m4 - 7m2 + 6m = m(m - l)(m - 2)(m + 3) = O.
m(m - l)(m - 2)(m - 3) + 6m(m - l)(m - 2)

Thus
18. Assuming that y
m(m-1)(m-2)(m-
Xffi
and substituting into the differential equation we obtain
3) +6m(m-1)(m-2)
+3m+ 1 = m4 +2m2+ 1 = (m2 + 1)2 = O.
+9m(m-1)
Thus
y
= C1cos(lnx) + C2sin(lnx) + c3In x cos(ln x) + c4ln x sin(ln x).
= m(m
- 5)
= O so that
Yc
= C1+
C2x5 and
x5 = 5x 4 .
W(l, x 5 ) = O1 5x4
Identifying f(x) = x3 we obtain
= -ix4 and U2 = 1/5x. Then u1 = -Ax5,
51515
y=C1+C2X
-25x
+Sx
515
lnx=c1+c3X
+Sx
U2 = ilnx,
and
lnx.
20. The auxiliary equation is 2m2 + 3m + 1 = (2m + l)(m + 1) = O so that Yc = C1X-1 + C2X-1/2 and
W(x-1
Identifying f(x) = ~-
x-1/2) =
-1
X
_x-2
we obtain u = x -
-1/2
X
_~x-3/2
I = _x-5/2.
1
2
x2 and U2 = x3/2 - x1/2. Then U1 = ~x2 -1x3,
U2 = ~x5/2 - ~x3/2, and

Y
1
1 2 2 2 2
= C1x-1 + C2x-1/2 + -2x
- -3x + -5x - -x
3
164
-1
= C1x
-1/2
1
1 2
+ C2x
- -x + -x .
6
15
4.7
Exercises
+1=
(m - 1)2
w (x, x ln x)
Identifying f(x)
and
= 2/x
we obtain
ul
= -21nx/x
y = C1X
Identifying f(x)
= x2ex we obtain
I1
= x
ul
x ln x = x.
1 + ln x
and U2
+ C2X ln x
+ 2 = (m
= O so that Yc = ClX + C2X ln x and
- x(ln x)2
- l)(m - 2)
= -x2ex
= z. Then
+ 2x(ln
Ul
U2
= 21nx,
x)2
= Oso that Yc
and U2 = xe".
= -(lnx?,
Then
= ClX
Ul
+ C2x2
and
= _x2ex + 2xex
2ex,
U2 = xe" - eX, and
+ 2m
y = Cl + C2X-2
= m(m
and
+ 2)
= O, so that
y' = -2C2X-3.

-10
-2C2 = 4.
Thus, Cl = 2, C2 = -2, and y
= 2 - 2x-2.
-20
y

y
= ClX2 + C2x4
+ 8 = (m
and
y'
- 2)(m - 4) = O, so that
2ClX
+ 4C2X3.

4 x
-4
4Cl + 32c2 = O.
Thus, Cl = 16, C2 = -2, and y
= 16x2
20
2x4.
30
165
Exercses
4.7
25. The auxiliary
equation
+ 1 = 0,
is m2
so that
y
3
+ C2sin(lnx)
y = C cos(lnx)
and
. (ln z )
y , = -CI ~1 sm
x
The initial conditions
+ C2-1 cos(in
x
lOO"
50
x ).
-3
irnply CI = 1 and C2 = 2. Thus
+ 2sin(lnx).
y = cos(inx)
+ 4 = (m
is m2 - 4m
Y=Clx2+c2x2inx
and
- 2)2 = 0, so that
y'=2CIX+C2(x+2xinx).
imply CI = 5 and C2 + 10
= 3.
Thus y = 5x2 - 7x2lnx.

-10
-20
-30
= O so that
Yc = CI + c2in x and
15
W(l, ln z) = 11 in/x I = ~.
O 1 x
x
Identifying
f(x)
UI = tx2 - !x2inx,
equation
so that Yc = CIX2
UI
f(x)
u2 = x.
10
Then
+8
in x = c
C2+ ~ =
12
+ C2in x + - x
4
-!. Thus,
= (m - 2)(m - 4)
CI =
~, C2 = -1,
0,
y
0.05
and
= I x2
2x
X43]
4x
= 8x4 we obtain
= -x4, U2 = 2x2,
12
+ -x
t = 1 and
is m2 - 6m
+ C2x4
W
Identifying
- - x in x
+ t x2.
in x
28. The auxiliary
1212
+ -x
imply CI +
= ~-
ln z and
U2 = ~x2, and
y = CI + C2in x
and y
u~ = +z
1 we obtain
= 2x5.
u~ = -4x3
and y = CX2
and
+ C2X4 + x6.
u2 = 4x.
Then
166
1 x
-1
imply
Exercises 4.7
1
-Cl
4
1
1
=-16
64
1
3
Cl + 2"C2 = -16 .
1
1
d
Th us Cl = 16'
C2= -"2' an y
+ -C2
1 2 1 4
= l6X
-"2X
+ X 6.
29. Substituting into the differential equation we obtain
d2y
dt2
+8
The auxiliary equation is m2 + 8m - 20 = (m

y
dy
dt - 20y = O.
+ 10)(m
- 2) = O so that
= cle-lOt + C2e2t = qx-lO + C2x2.
d2y
dy
dt2 - 10 dt
+ 25y
= O.
The auxiliary equation is m2 - 10m + 25 = (m - 5)2 = O so that

y
= clt + C2tt = C1X5+ c2x51n x.
d2y
dy
-dt2 + 9+ 8y
dt
The auxiliary equation is m2
= e2t
+ 9 + 8 = (m + l)(m + 8)
= O so that
undetermined coefficients we try YP = Ae2t. This leads to 30Ae2t

y=Cle-
+C2e-t+30e
1 2t
=C1x
-1
+C2X
Yc = cle-t
= e2t, so that A
-8
1
+30x.
+ C2e-8t. Using
= 1/30 and

d2y
dy
dt2 - 5 dt
The auxiliary equation is m2 - 5m
+6
undetermined coefficients we try YP
+ 6y = 2t.
= (m - 2)(m - 3) = O so that Yc = cle2t
+ C2e3t. Using
+ B. This leads to (-5A + 6B) + 6At
= 2t, so that
At
A = 1/3, B = 5/18, and

y
= cle2t + c2e3t + 3"t + 18 = C1X2+ C2x3 + 3"lnx + lB'
d2y
dy
dt2 - 4 dt
+ 13y
167
=4
+ 3e .
Exercses
4. 7
The auxiliary equation is m 2- 4m + 13 = Oso that Yc = e2t (ci cos 3t + C2sin 3t). Using undetermined
coefficients we try Yp = A+Bet.
This leads to 13A+ 10Bet
= 4 +3et, so that A = 4/13, B = 3/10,
and
.
Y=e2t(clcos3t+c2sm3t)+
=X2[Clcos(3lnx)
4
3
13 + 10et
+c2sin(3lnx)1
4
3
+ 13 + 10x.
34. From
it follows that
Substituting into the differential equation we obtain

-d3y - 3 -d2y + 2 -dy - 3 (d2y
- -dY) + 6 -dy - 6y = 3 + 3t
dt3
dt2
dt
dt2
dt
dt
or
d3y
d2y
dt3 - 6 dt2
The auxiliary equation is m3-6m2+11m-6
dy
+ 11 dt
- 6y
= 3 + 3t.
= (m-1)(m-2)(m-3)
= Oso that Yc = clet+c2e2t+
C3e3t. Using undetermined coefficients we try YP = A + Bt. This leads to (11B - 6A) - 6Bt = 3 + 3t,
so that A
= -17/12, B = -1/2, and

t
2t
3t 17 1
2
3
17 1
y = cle + C2e + C3e - - - -t = C1X+ C2X + C3X - - - -lnx.
12 2
12 2
168
Exercises 4.7
In the next two problems we use the substitution t = -x since the initial conditions are on the interval
(-00, O). In this case
dy
dy dx
dt = dx dt =
dy
dx
and
35. The differential equation and initial conditions become

d2y
4t2-+y=0
dt2
The auxiliary equation is 4m2 - 4m
y = cl tl/2
+ c2t1/21n
y(t)
'
+ 1 = (2m
I
t=l
=2,
- 1)2 = O, so that
+ C2
y' = ~CICl/2
and
It=l = -4.
y'(t)
(Cl/2 + ~cl/2In t) .
The initial conditions imply CJ = 2 and 1 + C2= -4. Thus

y
= 2tl/2 - 5tl/2ln t
2(_x)1/2 - 5(_x)1/2In( -x),
< O.
36. The differential equation and initial conditions become

t2 ~:; - 4t ~~ + 6y = O;
y(t)
It=2 =
y'(t)
8,
t=2
= O.
+ 6 = (m - 2)(m - 3) = O, so that
= clt2 + C2t3 and y' = 2clt + 3c2t2.
The auxiliary equation is m2 - 5m

y
4Cl + 12c2 = O
from which we find c
= 6 and
C2= -2. Thus

y
37. Letting u = x
+ 2 we obtain
ddy
x
d2y
dx2
= 6t2 -
2t3
= 6x2 + 2x3,
< O.
dy and, using the chain rule,

du
d (dY)
dx du
d2y du
= du2 dx
d2y
du2(1)
d2y
du2
Substituting into the differential equation we obtain

2 d2y
dy
u -+u-+y=O.
du2
du
The auxiliary equation is m2 + 1 = O so that
= C1cos(ln u) + C2sin (In u)
= c cos[ ln(x
169
+ 2)) + C2sin[ ln(x + 2)).
Exercises
4.7
38. If 1 - i is a root of the auxiliary equation then so is 1 + i, and the auxiliary equation is
(m - 2)[m - (1 + i)][m - (1 - i)]

We need m3 -4m2+6m-4
= m3 -
to have the form m(m-1)(m-2)
last expression and equating coefficients we get b
-1, c
4m2 + 6m - 4
= O.
+bm(m-1)
+cm+d. Expanding this
-4. Thus, the differential
3, and d
equation is
+ 3xy'
X3y'" - x2y"
39. For x2y"
- 4y = O.
= O the auxiliary equation is m(m - 1) = O and the general solution is y = Cl + C2X. The
initial conditions imply Cl = YO and C2 = Yl, so Y
Yo + Y1X. The initial conditions are satisfied
for all real values of YO and Yl.
= (m - l)(m - 2) = O and the

general solution is Y = C1X + c2x2. The initial condition y(O) = Yo implies O = Yo and the condition
For x2y" - 2xy' + 2y = O the auxiliary equation is m2 - 3m + 2
y'(O) = Yl implies Cl = Yl. Thus, the initial conditions are satisfied for YO = O and for all real values
of Yl.
For x2y" - 4xy' + 6y = O the auxiliary equation is m2 - 5m + 6
= (m - 2)(m - 3) = O and the
= C1X2 + c2x3. The initial conditions imply y(O)

the initial conditions are satisfied only for YO = Yl = O.
general solution is Y
40. The function y(x)
-v'xcos(lnx)
is defined for x>
for kan integer or where x = e1r/2+krr.
= O = YO and y' (O) = O. Thus,
O and has x-intercepts where lnx = 7r/2 + kr:
Solving 7r/2 + ktc
= 0.5 we get k::::; -0.34, so e1r/2+k1r < 0.5
for all negative integers and the graph has infinitely many x-intercepts in (0,0.5).
41. The auxiliary equation is 2m(m - l)(m - 2) - 10.98m(m - 1) + 8.5m + 1.3
= O, so that ml
-0.053299, m2 = 1.81164, m3 = 6.73166, and

Y
= C1X-O.053299
+ C2X1.81164 + C3x6.73l66.
42. The auxiliary equation is m(m - l)(m - 2) + 4m(m - 1) + 5m - 9 = O, so that ml

the two complex roots are -1.20409
Y
2.22291i.
1.40819 and
The general solution of the differential equation is
= ClX1.408l9+ x-1.20409[c2cos(2.22291ln x) + C3sin(2.22291ln x)].
43. The auxiliary equation is m(m - l)(m - 2)(m - 3) + 6m(m - l)(m - 2) + 3m(m -1) - 3m + 4
so that ml
= O,
= m2 = v'2 and m3 = m4 = -v'2. The general solution of the differential equation is

y = C1XV2 + c2xV2lnx
+ C3X-v'2+ c4x-v'2ln x.
44. The auxiliary equation is m(m-1)(m-2)(m-3)-6m(m-1)(m-2)
so that ml = m2
+33m(m-1)
-105m+ 169 = O,
= 3 + 2i and m3 = m4 = 3 - 2i. The general solution of the differential equation
is
Y
= X3[Cl cos(2ln x) + C2sin(2ln x)] + x3ln X[C3 cos(2ln x) + C4sin(2ln x)].

170
Exercises 4.8
Exercises 4.8
1. From
Dx = 2x - y and Dy = x we obtain y
= 2x -
Dx, Dy = 2Dx - D2x, and (D2 - 2D + 1)x = O.
Then
X
= Clet +
)t + C2tet .
C2tet an d Y = (Cl - C2e
2. From Dx = 4x + 7y and Dy = x - 2y we obtain y = ~Dx - ~x, Dy
(D2 - 2D - 15)x = O. Then

1
5t
Y = "7cle
-3t
- C2e
3. From Dx = -y + t and Dy = x - t we obtain y = t - Dx, Dy = 1 - D2x, and (D2 + 1)x = 1 + t.

Then
x = Cl COS t + C2sin t + 1 + t
and
y = Clsin t - C2cos t + t - 1.
4. Frorn Dx - 4y
= 1 and
x + Dy
= 2 we obtain y = iDx
-..t,
Dy
= iD2x,
and (D2 + 1)x
= 2.
Then
x = Cl COS t + C2sin t + 2
and
1
1.
1.
1
y = C2cost - Cl smt - Cl smt - .
(D2 + 2)y = O we obtain y = ~(D2 + 5)x, D2y = ~(D4 + 5D2)x,
5. Frorn (D2 + 5)x - 2y = O and -2x+

and (D2 + 1)(D2 + 6)x = O. Then
x = Cl COS t + C2sin t + C3cos v'6 t + C4sin v'6 t

and
y = 2Cl COS t + 2C2sin t - ~C3 cos v'6 t - ~C4 sin v'6 t.
-i - i(D
6. Frorn (D + 1)x + (D - 1)y = 2 and 3x + (D + 2)y = -1 we obtain x
Dx = -i(D2
+ 2D)y, and (D2 + 5)y = -7. Then

y
and
= Cl COS /5 t + C2sin /5 t - ~
( 32
/5)
x = --Cl - -C2
3
cosv'5t+
171
(/5
-Cl
3
2).
- -C2 smv'5t+
3
-.
5
+ 2)y,
Exercises 4.8
(D2
+ 4)(D
+ 2)x
- 2)(D
= -3et.
Then
and
.
Y = -e cos 2t 8. From (D2
(D - 5)(D2
+ 5)x + Dy = O and
+ 4)y = O. Then
2t
C2 SIn
+ l)x + (D
(D
+ C3e2t + C4e -2t
- 4)y
1 t
Se
.
= O we obtain (D - 5)(D2 + 4)x = O and
x = cle5t
+ C2 cos 2t + C3 sin 2t
y = c4e5t
+ C5 cos 2t + C6 sin 2t.
and
Substituting into (D
(6e
so that
C4
+ l)x + (D -
4)y = Ogives
+ c4)e5t + (C2 + 2C3 -
-6Cl,
= ~C3,
C5
C6
4C5 + 2C6) cos 2t
+ D2y =
2C5 - 4C6) sin 2t = O
= -~C2, and
-t
y = -6cle"
9. Frorn Dx
+ (-2C2 + C3 -
+ 2'C3 cos 2t
+ l)x + (D
e3t and (D
- l)y
- 2'C2 sin 2t.
= 4e3t we obtain D(D2 + l)x = 34e3t and
D(D2 + l)y = _8e3t. Then

y = e
+ C2 SIn. t + C3 cos t -
and
x
+ l)x + (D
(C4 so that
C4
ci)
4 3t
15e
17
C4
+ C5 sin t + C6 cos t + 15e3t.
- l)y = 4e3t gives
+ (C5 -
C6 -
+ (C6 + C5 + C2 -
C3 -
C2) sin t
.
cos t + C3 SIn t
C3) cos t
=O
= cj , C5 = C3, C6 = -C2, and

x = Cl
10. Frorn D2x - Dy = t and (D + 3)x

D(D + l)(D + 3)y = -1 - 3t. Then
C2
+ (D + 3)y =
and
172
17 3t
15e
2 we obtain D(D
+ l)(D + 3)x = 1 + 3t
and
Exercises
+ 3)x + (D + 3)y = 2 and D2x - Dy = t

3(Cl
+ C4) + 2(C2 + c5)e-t =
4.8
gives
2
and
Y = -q
-t
C2e - 3c3e
-3t
+ t - "21t 2 .
(D2 - l)x - y = O and (D - l)x + Dy = Owe obtain y = (D2 - l)x, Dy = (D3 - D)x, and
(D - 1)(D2 + D + l)x = O. Then
11. Frorn
J3 + C3sin
. =:'
J3]
= cie t + e-t/2 [C2cos 2t
and
y =
(-~C2
2
J3 C3) e-t/2cos J3t + (J3 C2 -
12. Frorn (2D2-D-1)x-(2D+1)y

and (2D
+ l)(D + l)y
= 1 and
(D-1)x+Dy
~C3) e-t/2 sin J3t.

2
2
= -1 we obtain
(2D+1)(D-1)(D+1)x
= -1
= -2. Then
x = cle-t/2 + C2e-t + c3et + 1

and
y = C4e-t/2 + C5e-t - 2 .
Substituting into (D - l)x
+ Dy
= -1 gives
( -~Cl - ~C4) e-t/2

so that C4 = -3Cl,
+ (-2C2
- c5)e-t
=O
C5 = -2C2, and
y = - 3Cle -t/2 - 2 C2e-t - 2 .
13. From (2D-5)x+Dy
= et and
(D-1)x+Dy
= 5et we obtain
Dy = (5-2D)x+
and (4-D)x
4et.
Then
and Dy
= -3q t + 5et so that

y=--Cle3
4
14. From Dx+Dy
= et and
and D2(D - l)x
4t +C2+5e.
(_D2+D+ l)x+y = Owe obtain y = (D2-D-1)x,
= et. Then
and
173
Dy = (D3 _D2 -D)x,
Exercises
4.8
y = -Cl - C2- C2t- C3et - tet + et.
15. Multiplying the first equation by D + 1 and the second equation by D2 + 1 and subtracting we
obtain (D4 - D2)x
1. Then
X
= Cl + C2t + C3et + C4e-t -"2t.1 2
Multiplying the first equation by D + 1 and subtracting we obtain D2(D + 1)y = 1. Then
12
Y = C5+ C6t+ c-e" -"2t .

Substituting into (D - 1)x + (D2 + 1)y = 1 gives
C2+ C6)t = 1
(-Cl + C2+cs -1) + (-2C4 + 2c7)e-t + (-1-
so that C5= Cl - C2+ 2, C6= C2+ 1, and C7= C4. The solution of the system is
.X
= Cl + C2t +
12
-t
C3e + C4e -"2t
y = (Cl - C2+ 2) + (C2+ 1)t + C4e-t - ~t2.

16. From D2x - 2(D2 + D)y = sin t and
+ Dy
we obtain x
= -Dy,
D2x
D(D2 + 2D + 2)y = - sin t. Then

1
y = Cl + C2e-t cos t + C3e-t sin t + 5" cos t + 5" sin t

and
17. From Dx
1
2
x = (C2+ C3)e - t sin t + (C2- C3)e - t cos t + 5" sin t - 5" cos t.
= y, Dy
= z. and
Dz = x we obtain x = D2y = D3x so that (D - 1)(D2 + D + 1)x = O,
. Tt
J3 + C3cos TtJ3 ] ,
x = cj et + e -t/2 [C2sin
y = Clet +
and
z=qe
t+
(1 J3)
(1-"2c2+Tc3J3)
--C2 - -C3
22
. -t
J3 +
e -t/2 sm
2
. J3
e -t/2 smTt+
1)
(J3
-TC2-"2C3 1)
(J3
-C2
22
-t/2
- -C3 ecos-t
J3
2'
J3
e -t/2 cosTt.
18. From Dx + z = et, (D - 1)x + Dy + Dz = O, and x + 2y + Dz = et we obtain z = -Dx + ,

Dz = -D2x + et, and the system (_D2 + D - 1)x + Dy = _et and (_D2 + 1)x + 2y = O. Then
y = 1(D2 - 1)x, Dy = 1D(D2 - 1)x, and (D - 2)(D2 + 1)x = -2et so that

x = Cle2t
y =
+ C2COS t + C3sin t +
3 2t
.
"2Cle - C2cost - C3smt,
and
174
et,
Exercises
=-
2Cl e2t - C3 cos t
4.8
+ C2 sin t.
19. Write the system in the form
Dx - 6y = O
x - Dy + z = O
x
+y
- Dz = O.
Multiplying the second equation by D and adding to the third equation we obtain
(D + l)x - (D2 - l)y = O. Eliminating y between this equation and Dx - 6y = Owe find
(D3
D - 6D - 6)x = (D + l)(D + 2)(D - 3)x = o.
Thus
x = cle-t
+ C2e-2t + C3e3t,
and, successively substituting into the first and second equations, we get
1 -t
1 -2t
Y = -"6cle - 3"c2e
5
6
-t
Z = --Cle
1
3
-C2e
-2t
3t
+ "2c3e
1 3t
+ -C3e
2
20. Write the system in the form
(D+l)x-z=O
(D
+ l)y
(D2+D-l)z
= O. Eliminating
+ 1 and
z =O
+ Dz
x - y
Multiplying the third equation by D
= O.
adding to the second equation we obtain (D
z between this equation and (D+l)x-z
+ l)x +
= Owe find
D(D+l)2x
so that Dy
= _(D2 + 5D)x.
= O.
Thus
x
Cl
+ C2e-t + c3te-t,
and, successively substituting into the first and third equations, we get
y = Cl
+ (C2 -
+ c3te-t
c3)e-t
= Oand 4x-
(D+ l)y = Owe obtain y = -(D+5)x

Then 4x + (D2 + 5D)x + (D + 5)x = O and (D + 3)2x = O. Thus
21. From (D +5)x+y
and
Y
=-
(2 Cl
+ C2) e -3t
175
2 C2t e -3t
Exercises
Using x(1)
4.8
O and y(l)
1 we obtain
or
Cl
2Cl
Thus
Cl
= e3 and
22. From Dx - y
Then _!(D2
C2
+ C2 = O
+ 3C2 =
_e3.
= _e3. The solution of the initial value problem is
= -1 and
- 2D)y
3x
+ (D
- 2)y
= e-3t+3
_ te-3t+3
_e-3t+3
= Owe obtain
= y - 1 and (D2 - 2D + 3)y
y = et
( Cl COS
+ 2te-3t+3.
x = -!(D
- 2)y so that Dx = _!(D2
- 2D)y.
= 3. Thus
v2 t + C2 sin v2 t) + 1
and
x = ~et
[( ci -
v2 C2) cos v2 t + (v2 ci + C2) sin v2 t 1 + ~.
Using x(O) = y(O) = O we obtain

ci
+1
=O
~ (c - v2 C2) + ~ = O.
Thus
Cl
= -1 and
C2
V2/2. The
='
solution of the initial value problem is
= et(2 - - cos vr:2 t - -V2.sm v;:;)

2 t +-2
= et
(-
cos v2 t
V2.)sin v2 t + 1.
+2
23. Equating Newton's law with the net forces in the x- and y-directions gives m d2X2 = Oand m d2Y2=
dt
dt
-mg, respectively. From mD2x = Owe obtain x(t) = clt+C2, and frorn mD2y = -mg or D2y = -g
we obtain y(t) = _~gt2
+ C3t + ca.
24. From Newton's second law in the x-direction we have

d2x
1dx
dx
mdt2 = -kcose =
dt = -lcl dt .
+:
In the y-direction we have

~y
m-
dt2
.
1~
-mg - ksme = -mg - k-- = -mg
v dt
176
-Icl-.~
dt
Exercises 4.9
From mD2x + IclDx = Owe have D(mD + Icl)x = Oso that (mD + Icl)x = Cl. This is a first-order
linear equation. An integrating factor is ellcldt/melclt/m so that
!!_ [eic1t/mx]= c
dt
elclt/m
1
and e!cltx = (clm/lcl)elclt/m+C2. The general solution ofthis equation is x(t)
= c3+c2elclt/m. From
(mD2 + IclD)y = -mg we have D(mD + lel)y = -mg so that (mD + Icl)y = -mgt
first-order linear equation with integrating factor elclt/m. Thus
d
- [elclt/my]= (-mgt
dt
and
+ Cl This is a
+ cl)elclt/m
y(t) = - m 9 t + m 9 + C3+ c2e-lclt/m.

[e]
c2
25. Multiplying the first equation by D + 1 and the second equation by D we obtain
D(D + l)x - 2D(D + l)y = 2t + t2

D(D + l)x - 2D(D + l)y = O.
This leads to 2t + t2 = O, so the system has no solution.
Exercises 4.9
= eX, so
1. We have y~ = yI
(yn2 = (ex)2 = e2x = yf.

Also,
Y2 = -
sin x and y~ =
cosx, so
(y~) 2 =
However, if y
(y")2
=1=-
f = cos2x = y~.
qYl + C2Y2, we have (y")2 = (cleX - C2cos x)2 and y2 = (ci eX + C2cos xf
y~
= yI = O, so
YlYl"
Y2 = 2x
1 . O = O = -1()2
O
2
However, if y
= -21(Yl')2 .
and y~ = 2, so
Y2Y2" = x 2(2 )
~(y/)2
cos x
y2.
2. We have
Also,
(-
= ~[Cl . O+
= 2x 2 = '21(2x )2 = '21(Y2')2
qYl + c2Y2, we have yy" = (el . 1 + C2x2)(Cl. 0+ 2C2)
c2(2x)]2
= 2c~x2.
Thus yy"
=1=-
~(y/)2.
177
Thus
Exercses 4.9
3. Let u :::;:
y' so that u' = y", The equation becomes u' = -u - 1 which is separable. Thus
du
-2--
+1
= -dx ==> tan-l u = -x + Cl ==> y' = tan(cl - x) ==> y
du
l+u
--2
= 1 + u2. Separating variables we obtain
= y". The equation becomes u'
4. Let u = y' so that u'
= dx ==> tan-
u = x + Cl ==>
= tan(x
= In 1COS(Cl- x)1 + c2
==> y = +In 1cos(x + cl)1 + C2
cj )
= y' so that u' = y", The equation becomes x2u' + u2 = O. Separating variables we obtain
5. Let u
du
u2
dx
= - x2 ==>
-;
= ;; + Cl =
ClX+ 1
X
1
1
Y = 2" In IClx + 11- -x
cl
Cl
1 (
==>
= - Cl
+ l/Cl
1 (1
= Cl qx + 1 - 1
+ C2
,
"du
.
6. Let u = y so that y = u dy' The equation becomes (y
+ 1)u du
dy =
u . Separating variables we
obtain
du
dy
y+1
==> In lul = In Iy + 11+ ln c ==> u = Cl(y + 1)

dy
==> - = Cl(y
dx
+ 1)
dy
==> -= Cldx
y +1
==> In Iy + 11 = ClX+ C2 ==> Y + 1 =
C3eCx.
3
= y , so that y "du= u dy' The equation becomes u du
dy + 2yu = O. Separating variables we
7. Let u
obtain
du
2"
u
+ 2y dy = O==>-
+ y2 = C
==> u=
==>
8. Let u
= y , so that
"du
= u dy'
dy
du = - ==> u
y2
q
If Cl = O, then y dy
= -dx
y2
1
Y
) dy
qy - 1
,1
==>y=---
+ ci
y2
+ ci
"3 y + qy = x + C2
.
2 du
The equaton becomes y u dy
= -- + Cl ==> Y
==> _!:_ (1 +
= u.
qy - 1
Y
==>
dy
Y
ClY- 1
= dx
= dx (for Cl f. O) ==> _!:_y+ 12 In Iy - 11 = x + C2.

Cl
and another solution is
h2 = -x
178
+ c2.
cl
Exercises 4.9
9. (a)
Iy
l.
-10
(b ) Let u
du
= y ' so that y 11 = u du
dy . The equation becomes u dy + yu = O. Separating variables we
obtain
When x
= 0, y =
1 and y'
dy
1
-=--y
dx
When x
= 0, y
=
1
-1 so -1
= -~ + Cl and
dy
1
= -- dx
===?
--
===?
Y = tan ( -~x
y2
+1
= 1 so 1 = tan e, and C2
Cl
tan-
===?
+ C2)
= -~. Then
1
y = --2x
+ C2
= 1[/4. The solution of the initial-value problem is
The graph is shown in part (a).

(e) The interval of defintion is -1[/2
< 1[/4 - x/2 < 1[/2 or -1[/2 < x < 31[/2.
179
Exercises 4.9
10. Let u = y' so that u' = y". The equation becomes (u')2
which results in u' =
~.
+ u2 =
To solve u' = ~
we
separate variables:
du
Jl="U2
==> sin-l u = x + el ==> u = sin(x + c)
= dx
-27t
27t
==> y' = sin(x + el).

When x
Thus
7r ,
j3
(7r)
j3
-"67r .
= "2 ' y = 2"" ' so 2"" = sin "2 + c and el =

y' = sin
When x =
7r
"2'
(x -~)
(7r
7r)
(x - ~) + e2.
= 2"' so 2" = - cos "2 -"6 + e2 = - 2" + e2 and
initial-value problem is y = 1 - cos

To solve u'
==> y = - cos
= -~
(x - ~).
e2 = l. The solution of the
we separate variables:
du
-v'l=U2 = -dx
==> cos-l u =
+ el
==> u = cos(x + el) ==> y' = cos(x + ci).

When x
7r ,
j3
(7r)
j3
= "2 ' y = ""2 ' so ""2 = cos "2 + el and

y'
= cos (x
- i)
==> y
el
7r
= - 3". Thus
= sin (x
. (7r
= 7r
"2 ' y = 2"1 ' so 2"1 = sm
"2 - 3"7r) +e2 = 2"1+cz and
When x
problem is y = sin
- i) + e2
e2
..
..
= O. The solution of the initial-value
(x - i)1
11. Let u = y' so that u' = y". The equation becomes u' - -u = _u3, which is Bernoulli. Using
W
=U
.,
-
2
2
.
.
. 2
we obtam - + -w = - - _An mtegratmg factor lS x , so
dx
x
x
d 2 1
2
2
el
dx[x w = -2x ==> x w = -x + c ==> w = -1 + x2
.
dw
-2
==>
==>
dy
dx
ci
= -1 + 2"
x
Vel - x2
==> u=
Vel - x2
==> y = -
J el -
x2
+ e2
==> el - x2 = (e2 _ y) 2 ==> x2 + (e2 - y)2 = el-
180
Exercses
12. Let u
y' so that u'
y". The equation becomes u' - ~u
dw
su bstituti
itution w = u- 1 we obtai
tain -d
d
-[xw]
dx
= -x
==> w
1
= --X+-C
2
u2, which is Bernoulli. Using the
. factor IS
. x, so
= -.1 An mtegratmg
1
+ -w
1
1
==> x
u
4.9
Cl -
x2
2x
==> u =
2x
Cl -
x2
==> y
= -In ICl
x2 +C2.
In Problems 13-16 the thinner curve is obtained using a numerical solver, while the thicker curve is the
graph of the Taylor polynomiaJ.
13. We look for a solution of the form
1
1
y(x) = y(O) + y'(O) + 2Y"(0) + 3TY"'(O)
From y"(x)
+ 4Ty(4)
(x)
+ 5iy(5)
(x).
40
= x + y2 we compute
y'''(x)
y(4)(x)
= 1 + 2yy'
= 2yy" + 2(y')2
y(5)(x) = 2yy'"
Using y(O)
+ 6y'y".
= 1 and y'(O) = 1 we find

y"(O) = 1,
y"'(O) = 3,
y(4)(0) = 4,
y(5)(0) = 12.
An approximate solution is
y(x)
14. We look for a solution of the forrn

1
1
y(x) = y(O) + y'(O) + -2Y"(0) + --y"'(O)
3.
Frorn y"(x)
1
10
= 1 +x + _x2 + _x3 + _x4 + _x5.
1
+ --y(4)
4.
(x)
1 (
+ -51Y
5)(x).
= 1 - y2 we compute
y"'(x) = -2yy'
Using y(O)
y(4)(x) = -2yy"
_ 2(y')2
y(5)(x) = -2yy'"
- 6y'y".
= 2 and y'(O)
y"(O) = -3,
= 3 we find
y"'(O) = -12,
-5
y(4)(0)
= -6,
y(5)(0)
102.
-10
181
Exercises 4.9
3
y(X) = 2 + 3x - -x
2
1
1
y(x) = y(O) + y'(O) + "2Y"(O) + 3iylll(O)
3
1
- 2x - -x
4
+ -x17
20
.
y
4T1y(4) (x)
5T1y(5)(x).
= x2 + y2 - 2y' we compute
From y"(X)
+ 2yy'
ylll(X) = 2x
= 1 and y'(O)
y"(O)
= -1,
+ 2yy"
- 2ylll
20
- 2y(4l.
= 1 we find
y'"(O) = 4,
30
+ 2yy'"
y(5) (x) = 6y'y"
40
- 2y"
y(4)(x) = 2 + 2(y')2
Using y(O)
10
= -6,
y(4)(0)
y(5)(0)
= 14.
12231475
y(x) = 1 + x - -x + -x
+ -x
- -x
60

y(x)
= y(O) +y' (O) + ~y" (O) + ~ylll (O) + ~y(4) (x) + ~y(5) (x) + ~y(6) (x).
From y"(X)
ylll(X)
y(4)(x)
2
3!
eY we compute
4!
5!
61
= eYy'
= eY(y')2 + eYy"
y(5)(x) = eY(y')3
+ 3eYy'y" + eYylll
y(6)(x) = eY(y')4
+ 6eY(y')2y" + 3eY(y")2 + 4eYy'ylll + eYy(4).
Using y(O) = Oand y'(O) = -1 we find

y"
(O) = 1,
10
-2
-4
y 111(0) =-,1
y(x)
17. We need to solve
(1 + u2) 3/2
= u' or
-x
+ -x1213-14-x1516+ -x
2
[1 + (y')2f/2
(1 + u2) 3/2
= y".
12
+ -x
24
Let u
+ -x
45
= y' so that u' = y".
The equation becomes
= ~~. Separating variables and using the substitution
182
tan
Exercises 4.9
we have
(
du
1 + u2)3/2 = dx
==:;.
==:;.
18. Let u =
dx
di
==:;.
==:;.
y'
sec2 e
(
)3/'> de = x
e -
1 + tan2
cos e de = x
y'
1+
sin e = x
==:;.
==:;.
sec2 e
sec3 e de = x
==:;.
~
[]
(y')2
=x
x2
= x ==:;. (y')2 = x2 1+ (y')2 = -1--2
- x
vIl=X2
(for x
> O)
==:;.
=-~
du
_k2
The equation becomes u dx = ~.
d2x
du
so that dt2 = u dx
Separating variables we
obtain
k2
udu=--2dx
x
When t
==:;.
-u
2
k2
x
=-+c
k2
k2
Xo
Xo
= O, x = Xo and v = O so O = - + e and e = --.

~V2
= k2
(~
2.)
Xo
Then
dx
~/xo--o
- x
-=-kv2
dt
xXo
and
Separating variables we have

dx = kv2 dt
- ~
V~
==:;.
Po
kV"2
t=-~
J/
x
xo-x
Using Mathematica to integrate we obtain
t = - -1 ro- [-x
k
(xo - x) - -Xo tan -1 (xo - 2x)

2
2x
[V
=-1 ro
-x(xo-x)+-tan-
Xo
2
Xo - 2x
2Jx(xo - x)
dx.
{x!;]
-Xo - x
1.
19.
2
zp'
20 '
-2
d2x
For dt2
xl = o
+ sin x =
-2
xl
O the motion appears to be periodic with amplitude 1 when
amplitude and period are larger for larger magnitudes of
183
Xl.
Xl
O. The
Exercises 4.9
lO'
-1
d2x
For dt2
dx
dt
+ sin x = Othe
lO'
10 '
-1
xl = O
-1
xl= I
x l = -2.5
motion appears to be periodic with decreasing amplitude. The dx I dt
term could be said to have a damping effect.

20. When y = sin x, y'
= cos x,
y" = - sin x, and

(y")2 _ y2 = sin2 x - sin2 x = O.
From (y")2 - y2 = O we have y" = y, which can be treated as two linear equations. Since linear
combinations of solutions of linear homogeneous differential equations are also solutions, we see that
y
clex
+ C2e-x and
C3 cos x
+ C4 sin x must satisfy the differential equation. However, linear
combinations that involve both exponential and trigonometric functions will not be solutions since
the differential equation is not linear and each type of function satisfies a different linear differential
equation that is part of the original dfferential equation.
21. Letting u
= y", separating variables, and integrating we have
f1+2
1i- u~,
VI du+ u2
du _
dx - V
and
- d
- x,
sinh-l u = x
+ CI.
Then
u = y"
= sinh(x + CI),
y' = cosh(x
+ CI) + C2,
and
= sinh(x + e) + C2x + C3.

1. Y
=O
2. Since Yc
= cI + C2e-x, a particular solution for y"
- Y = 1 + e" is yp
= A + Bxe".
3. True
4. True
5. They are linearly independent over (-00,00)
6. (a) Since h(x)
= 21nx
and linearly dependent over (0,00).
= 2f(x), the functions are linearly dependent.

184

(b) Since xn+l is not a constant multiple of z", the functions are linearly independent.
(c) Since x
+ 1 is not
(d) Since h(x)
a constant multiple of x, the functions are linearly independent.
= cosxcos(7f/2) - sinxsin(7f/2) = -sinx
= -fz(x),
the functions are linearly
dependent.
(e) Since h(x)
= O fz(x), the functions are linearly dependent.
(f) Since 2x is not a constant multiple of 2, the functions are linearly independent.
(g) Since 3(x2)
(h) Since xex+l
+ 2(1 + 0(4x
x2) - (2 + x2) = O, the functions are linearly dependent.

- 5)eX
exe" = O, the functions are linearly dependent.
7. (a) The auxiliary equation is (m - 3)(m

equation is ylll
+ y"
- 17y'
+ 5)(m
- 1) = m3
+ m2 -
17m
+ 15 = O,so the
differential
+ 15y = O.
(b) The form of the auxiliary equation is

m(m - l)(m - 2) + bm(m - 1) + cm
Since (m - 3)(m
+ 5)(m
- 1) = m3
+ d = m3 + (b -
+ m2 -
17m
+ 15 =
3)m2
+ (c - b + 2)m + d = O.
O, we have b - 3 = 1, c - b + 2 = -17,
and d = 15. Thus, b = 4 and c = -15, so the differential equation is ylll + 4y" -15y'
+ 15y
= O.
8. Variation of pararneters will work for all choices of g(x), although the integral involved may not
always be able to be expressed in terms of elementary functions.
The method of undetermined
coefficients will work for the functions in (b), (c), and (e).
9. From m2 - 2m - 2 = O we obtain m = 1 y'3 so that
y
10. From 2m2
11. From m3
+ 2m + 3 = Owe obtain
+ 10m2 + 25m
Cl e(l+V3)x
+ c2e(l-v'3)x.
m = -1/2 V5/2 so that
= Owe obtain m = O, m = -5, and m = -5 so that
12. From 2m3
+ 9m2 + 12m + 5 = Owe obtain
13. From 3m3
+ 10m2 + 15m + 4 = Owe obtain

y = cle-x/3
m = -1, m
+ e-3x/2
= -1, and m = -5/2 so that
m = -1/3 and m = -3/2 0/2

(c2 cos
185
V;
+ C3 sin
V;
x) .
so that

14. From 2m4
+ 3m3 + 2m2 + 6m -
4 = O we obtain m
= 1/2, m = -2, and m = v'2 i so that
+ C2e-2x + C3cos h x + C4sin h x.
y = crex/2
15. Applying D4 to the differential equation we obtain D4(D2 - 3D

y
VTI
2
. VTI)
2
+ C2sin --x
= e3x/2 Clcos --x
+ 5) = O. Then
+ C3+ C4X + C5x2 + C6x3
Yc
and Yp
= A + Bx + Cx2 + Dx3.
(5A - 3B
Substituting YP into the differential equation yields
+ 2C) + (5B
+ 6D)x + (5C
- 6C
- 9D)x2
Equating coefficients gives A = -222/625, B = 46/125, C

solution is
3x/,)
222
Y = e - ClCOS--x
+ C2Sin--x
- 2
2
625
(VTI
VTI)
+ 5Dx3 =
36/25, and D
36
+ -x46 + -x
125
+ 4x3.
-2x
25
16. Applying (D - 1)3 to the differential equation we obtain (D - 1)3(D - 2D
4
+ -x
5
4/5. The general

3
.
+ 1) = (D
Then
Yc
and YP = Ax2ex
+ Bx3ex + Cx4ex.
+ 6Bxex + 2Aex = x2ex.
12Cx2ex
Equating coefficients gives A = O, B
= O,
and C = 1/12. The general solution is

1
+ _x
124ex
Y = C 1eX + C2xe"
17. Applying D(D2
+ 1) to
+ 1)(D3
D(D2
+ 6D) = D2(D2 + l)(D
- 5D2
- 2)(D - 3)
= O.
Then
Yc
and YP = Ax
+B
cos x
+ C sin x.
6A
+ (5B + 5C)
cosx
Equating coefficients gives A = 4/3, B

Y = Cl + c2e
+ (-5B
+5C) sinx = 8 + 2sinx.
= -1/5, and C
2x
3x
+ C3e +"3x
186
= 1/5. The general solution is
1 .
"5 cos x + 5" sin x.
- 1)5 = O.

18. Applying D to the differential equation we obtain D(D3
D2) = D3(D - 1) = O. Then
Yc
= Ax2. Substituting Yp into the differential equation yields -2A

gives A = -3. The general solution is
and Yp
y = Cl
+ C2X + C3ex
- 3x2.
2m+ 2 = [m - (1 + i)][m - (1- i)]
= 0, so Yc = clex sinx + C2ex cos x
and
w=
Identifying
eX sinx
I e" cos x
+ eX sin x
(eX cos x) (eX tan x)

2
-e X
=-
,
(eX sinx)(eX tan x)
u2 =
2
-e X
Ul
eX cosx
-exsinx+excosx
2x
= -e
=-
cosx,
U2
y = Cl eX sin x
=
Cl eX
sin x
= sinx
-In
= Slnx
sin2x
= ---cosx = cosx
I secx + tan z],
and
+ C2ex cos x
- eX cos x In I sec x
1 = 0, so Yc = qeX
+ tan x l.
+ C2e-x
e -X
-e-x
Identifying f(x)
- secx.
+ C2ex cos x - e" sin x cos x + eX sin x cos x - eX cos x In I sec x + tan z]
= 2ex /(eX
+ e-X)
eX
1-
and
I = -2.
we obtain
1
u' -
+ e-x
eX
- ----0'-- 1 + e2x
,
e2x
e3x
X
eX
u2 = - eX + e-X = - 1 + e2x = -e + 1 + e2x'
Then
Ul
f (x) = eX tan x we obtain

ul
Then
= 6. Equating coefficients
= tan-l eX, U2 = =e" + tan-l eX, and
m - 1=
y
so that
= qXl/2 + C2X-l/3.
187

22. The auxiliary equation is 2m3 + 13m2 + 24m + 9 = (m + 3)2(m + 1/2) = O so that
Y=ClX
23. The auxiliary equation is m2 - 5m + 6
x4 - x2lnx so that
-3
+C2X
-3
Inx+4:x
(m - 2)(m - 3)
O and a particular solution is Yp
y = ClX2 + C2x3 + x4 - x2lnx.

24. The auxiliary equation is m2 - 2m + 1 = (m _1)2
y
= Oand
= Clx+c2xlnx+
a particular solution is YP = ~x3 so that
1
4:x3.
25. (a) The auxiliary equation is m2 + w2 = O, so Yc = Clcoswt + C2sinwt.

A cos oi + B sin at and
When w
i-
a, Yp =
y = Clcoswt + C2sinwt + A cos at + B sin cd:

When w = a, Yp = Ateos at + Bt sin cd: and
y = Clcos wt + C2sin wt + At cos at + Bt sin cd:
(b) The auxiliary equation is m2 - w2 = O, so Yc = clewt + c2e-wt.
y = clewt
When w
i- a, YP = Al<t and
+ c2e-wt + Aeat.
When w = a, Yp = Ateat and
26. (a) If y = sinx is a solution then so is y = cosx and m2 + 1 is a factor of the auxiliary equation
m4 + 2m3 + 11m2 + 2m + 10 = o. Dividing by m2 + 1 we get m2 + 2m + lO, which has roots
-1
3i. The general solution of the differential equation is

y = Cl COS X + C2sin x + e-x (C3 COS 3x + C4sin 3x).
(b) The auxiliary equation is m(m + 1) = m2 + m = O, so the associated homogeneous differential

equation is y" + y' = O. Letting y
= Cl + C2e-x + ~x2 - x and computing y" + y' we get x.
Thus, the differential equation is y" + y' = x.

27. (a) The auxiliary equation is m4 - 2m2 + 1
= (m2 - 1)2 = O, so the general solution of the
y
= Clsinhx
+ C2coshx
+ C3Xsinhx + C4Xcosh x.
(b ) Since both sinh x and x sinh x are solutions of the associated homogeneous differential equation,
a particular solution of y(4) - 2y" + y = sinhx has the form YP = Ax2 sinhx + Bx2 coshx.
188

28. Since
yi
= 1 and yI = O, x2yi' - (x2 + 2x)yi + (x + 2)Yl
= _x2 - 2x
+ x2 + 2x = O, and
Yl
= x is
a solution of the associated homogeneous equation. Using the method of reduction of order, we let
y
= ux. Then y' = xu' + u and y" = xu" + z, so

x2y" - (x2 + 2x)y' + (x + 2)y = x3u" + 2x2u' - x3u' - 2x2u' - x2u - 2xu + x2u + 2xu
= x3u" - x3u' = x3 (u" - u').
To find a second solution of the homogeneous equation we note that u

u" - u' = O. Thus, Yc =
CX
+ C2xex.
= eX is a solution of
To find a particular solution we set x3(u" - u') = x3 so
that u" - u' = 1. This differential equation has a particular solution of the form Ax. Substituting,
we find A = -1, so a particular solution of the original differential equation is YP = _x2 and the
general solution is y
= cx
+ C2xex

Y(7f/2)
- x2.
2m + 2 = Oso that m = 1 i and y = eX(cl cos x
= Oand Y(7f) = -1 we obtain
= e-7r and
Cl
+ 2m + 1 =
C2
= O. Thus, y = eX-7r
+ C2 sin x).
Setting
cosx.
+ 1)2 = O, so that y = cle-x + C2xe-x. Setting

C2e = Oand -Cl + C2 = O. Thus c = C2 and y = ce-x + cxe:"
y( -1) = Oand y'(O) = Owe get cj e -
(m
is a solution of the boundary-value problem for any real number c.

31. The auxiliary equation is m2 - 1 = (m - l)(m
Assuming YP
+ 1) =
O so that m
1 and y
clex
+ c2e-x.
Ax + B + e sin x and substituting into the differential equation we find A = -1,
= O, and e = --!.
Thus YP = -x - -!sinx and

Y=
Cl
+ C2e-x -
.
"21 sin
x.
x -
Setting y(O) = 2 and y'(O) = 3 we obtain

Cl
Cl -
Solving this system we find
Cl
and
li
C2 -
-i.
3
= 3.
2
The solution of the initial-value problem is
-
13 x 5 -x
1 .
= -e
- -e - x - - smx.
C2
+ C2 = 2
442
+ 1 = O, so Yc = Cl cos X + C2 sin x and

W=
cosx
sin x =1.
cosx
-sinx
Identifying f(x) = sec3 x we obtain

,
ul
= -smxsec
x =
sin x
---3COS X
u; = cos x sec3 x = sec2 x.
189

Then
Ul
1
1
= --sec
2 cos-' X
2
= ----
u2 = tanx.
Thus
Y=
cr cos x + C2 sin
X
= Cl
=
and
2'1 cos x sec2.x + sin X tan x
2' sec x +
COS X
C3 COS X
,.
Yp =
-C3
+ C2 sin X
1-
COS2 X
cos x
+ C2 Slll X + 2'1 sec x.
sin x
+ C2 cos z + 2' sec x tan
x.

1
C3
+ 2'
=1
C2 =-.
Thus
C3
= C2 = 1/2 and
Y =
33. Let u
1.
2' cos x + 2' sin X + 2' sec x.

du
= y' so that u' = y". The equation becomes u dx = 4x. Separating variables we obtain
u du
When x
= 4x dx
===:;,.
1
_u2 = 2x2 +
2
= 1, y' = u = 2, so 4 = 4 + C2 and
C2
C1 ===:;,.
u2 = 4x2
+ C2.
= O. Then
or
dy = -2x
dx
When x = 1, y = 5, so 5 = 1 + C3 and 5 = -1 + C4. Thus C3 = 4 and C4 = 6. We have y = x2 + 4

and y = -x2 + 6. Note however that when y = -x2 + 6, y' = -2x and y'(l) = -2 =f 2. Thus, the
solution of the initial-value problem is y
34. Let u
=y
so that y
"du
= u -.dy
= x2 + 4.
.
du
The equation becomes 2u dy
2u dy = 3y2 dy
===:;,.
190
= 3y 2 .
u2 = y3 + Cl.

When x =
o, y
= 1 so
= y3
==}
(~~)
==}
_2y-l/2
u2
When x
1 = 1 + Cl and Cl
= 1 and y' = u
= O, y = 1, so 1
however that when y = (x
y3
==}
=X
+ C2
2 =
~~
= O. Then
= y3/2
==}
==}
Y = (x
y-3/2 dy
= dx
+ C2)2 .
4
4
4
Note
= 2" ==} C2 = 2. Thus, y = (x + 2)2 and y =
(x
2)2 .
c2
4
,
8
+ 2)2 , y = - (x + 2)3 and y'(O) = -1 =11. Thus, the solution of the
initial-value problem is y = (x ~ 2)2 .

35. (a) The auxiliary equation is 12m4
+ 64m3 + 59m2
- 23m -12
= Oand has roots -4, -3/2,
-1/3,
and 1/2. The general solution is
(b) The system of equations is

Cl + C2 + C3 + C4 = -1
3
1
-4e - -C2 - -C3
+ -C4
232
16c
-64cl
=2
+ C2 + 9C3 + 4C4 =
27
8
- -C2'-
1
27
-C3
+ -C4
= O.
8
. a CAS we fin d Cl -- - 495

73 ' C2 -- 35'
109 C3 -- - 3726
257 . Th e so1u t'ron of th e
Usmg
385 ' an d C4 -- 45
191
36.
Consider xy"
+ y' =
O and look fr a solution of the form y = z?'.
'f
Substituting into the differential equation we have

xy"
+ y'
= m(m - l)xm-l
-1
+ mxm-l
= m2x.
-2
Thus, the general solution of xy"
+ y'
particular solution of xy"
+ y'
= O is Yc = Cl + C2lnx. To find a
-IX
we use variation of parameters.
The Wronskian is
Identifying f(x) = _x-l/2

,
x-l/2Inx
= y'Xlnx
l/x
so that
ul
= x3/2(
-5
we obtain
ul =
Then yp
-q
~~:I=~
w=l~
-3
= x 3/2(2-Inx
393
~ In x - ~) - ~x3/2In x
is y = Cl + c2Inx - ~x3/2
u; =
and
4)
- -
= - ~x3/2
and
u2
_x-l/2
= -VI,
s.]
= --x2
3/2 .
and the general solution of the differential equation
The initial conditions are y(l) = O and y'(l) = O. These imply that
Cl = ~ and C2= ~. The soIution of the initiaI-value problem is y = ~+ ~In x - ~x3/2.

37. From (D - 2)x
+ (D
- 2)y = 1 and Dx
+ (2D
- l)y = 3 we obtain (D - l)(D - 2)y = -6 and
Dx = 3 - (2D - l)y. Then
Substituting into (D - 2)x
+ (D
- 2)y = 1 gives C3= 5/2 so that

t
3 2t
x = -C2e - -Cle
2
38. From (D - 2)x - y
=t-
2 and -3x
x = cr e
+ (D
- 4)y
5
+ -.
2
= -4t we obtain (D - l)(D - 5)x = 9 - 8t. Then
8
3
+ C2e5t - -t
- 5
25
and
y = (D - 2)x - t
39. From (D - 2)x - y
+ 2 = -cle
5t
16
11
+ 3C2e + 25 + 25 t.
= _et and -3x + (D - 4)y = -7et we obtain (D - l)(D - 5)x = -4et so that
Then
192

40. From
= sin2t
(D+2)x+(D+l)y
and 5x+(D+3)y
= cos2t
Then
y
and
Cl
cos t
1
. t - 3'
2 cos 2t + 3'
7.sm 2t
+ C2 sin
1
x = -"5(D + 3)y + "5 cos 2t
193
we obtain (D2+5)y
= 2cos2t-7sin2t.
Modeling with Higher-Order

Differential
Equations
Exercises 5.1
----_
+ 16x = O we obtain
1. From ~x"
x = ct cos s)2 t
+ C2 sin s)2 t
so that the period of motion is 21f/s)2 = V21f/S seconds.

2. Frorn 20x"
+ kx
= O we obtain
2V'S0. t + C2 sin ~2V'S0. t
x = ct cos ~
and k = 320 N/m. If SOx" + 320x = O then x = ct cos 2t
so that the frequency 2/1f = ~[kf51f

C2 sin 2t
so that the frequency is 2/21f
3. Frorn ~x"
+ 72x = O, x(O) =
4. From ~x"
+ 72x
5. Frorn ix"
+ 40x = O, x(O) =
(a) x(1f/12)
(b) x'
= 1/1f vibrations/second.
-1/4, and x'(O)
= Owe obtain x = -~ cos4V6t.
== 2 we obtain x == ~ sin 4V6 t.
= O, x(O) = O, and x'(O)
1/2, and x'(O) == Owe obtain x
= ~cos St.
= -1/4, x(1f/S) = -1/2, x(1f/6) == -1/4, x(1f/S) = 1/2, x(91f/32) = V2/4.
= 4 ft/s directed downward.

= (2n + 1)1f/16 for n = 0,1,2, ....
-4 sin St so that x'(31f/16)
(e) If x = ~cosSt = Othen t

6. Frorn 50x"
7. From 20x"
+ 200x = O, x(O) = O, and x' (O) = -10 we obtain x = -5

+ 20x = O, x(O) = O, and x'(O) = -10 we obtain x = -10
sin 2t and x' = -10 cos 2t.

sin t and x'
= -lOcost.
(a) The 20 kg mass has the larger amplitude.
= -5V2 mis, x'(1f/2) = O mis; 50 kg: x'(1f/4) = O mis, x'(1f/2) = 10 mis

(e) If -5 sin 2t = -10 sin t then 2 sin t( cos t -1) = Oso that t = n1f for n = O, 1, 2, ... , placing both
(b) 20 kg: x'(1f/4)
masses at the equilibrium position.
The 50 kg mass is moving upward; the 20 kg mass is
moving upward when n is even and downward when n is odd,

8. Frorn x"
+ 16x =
O, x(O)
= -1,
and x'(O)
x = -cos4t
= -2
we obtain
J5
-'- 2'sin4t = Tcos(4t
194
- 3.6).
Exercises 5. 1
The period is 7f/2 seconds and the amplitude is .)5/2 feet. In 47f seconds it will make 8 complete
vbrations.
9. From
;tX" + x
= O, x(O) = 1/2, and x'(O) = 3/2 we obtain

x
10. From 1.6x"
+ 40x
= O, x(O)
= -1/3,
and x'(O)
x = - ~ cos 5t
3
If x = 5/24 then t
+ 200x
11. From 2x"
(a) x =
Vl3.sm(2t + 0.588).
3.
= "2 cos 2t + 4" sm 2t = -4= 5/4
we obtain
+ ~sin 5t = ~
4
12
sin(5t
+ 0.927).
t (~+ 0.927 + 2n7f) and t = t e: + 0.927 + 2n7f)
for n
= 0,1,2, ....
= -2/3, and x'(O) = 5 we obtain
= O, x(O)
-cos 10t + ~sin 10t = ~sin(10t
- 0.927).
(b) The amplitude is 5/6 ft and the period is 27f/10 = 7f/5
(e) 37f = 7fk/5 and k = 15 cycles.

(d) If x = O and the weight is moving downward for the second time, then 10t - 0.927 = 27f or
t=0.721s.
(e) If x' =
235
cos(10t - 0.927)
n = O, 1, 2, '"
= O then 10t - 0.927 = 7f/2 + n7f or t
= (2n
+ 1)7f/20 + 0.0927 for
(f) x(3) = -0.597 ft
(g) x'(3) = -5.814 ft/s

(h) x"(3) = 59.702 ft/s2
(i) If x
= O then
(j) If x = 5/12 then

(k)
If x = 5/12 and
12. From x"
+ 9x
+ n7f) for n = O, 1,2, ... and x'(t) = 2; ft/s.

t = ro (7f/6 + 0.927 + 2n7f) and t = lo (57f/6 + 0.927 + 2n7f) for n
x' < O then t = ro (57f/6 + 0.927 + 2n7f) for n = O, 1,2, ....
t = fo(0.927
= O, x(O) = -1, and x'(O)

X
and x' = 2V3 cos(3t

n = 1,2,3, ....
13. From k1 = 40 and k2
= - cos 3t -
+ 47f/3).
If x'
= -V3 we obtain
V3 sin
. 3t
""3
= 3 then t
2.
= V3 sin 37 +
= -77f /18
47f)
3
+ 2n7f/3
and t = -7f /2
+ 2n7f/3
for
= 120 we compute the efIective spring constant k = 4(40)(120)/160 = 120.
Now, m = 20/32 so k/m = 120(32)/20 = 192 and x"

obtain x(t)
= O, 1, 2, ....
+ 192x
= O. Using x(O) = O and x'(O) = 2 we
:.fJ sin 8V3 t.
14. Let m denote the mass in slugs of the first weight. Let k1 and k2 be the spring constants and
k = 4k1k2/(k1
+ k2)
the efIective spring constant of the system. Now, the numerical value of the
195
Exercises 5. 1
first weight is W
= mg = 32m, so
32m
= klG)
and
32m
= k2G)
From these equations we find 2k1 = 3k2. The given period of the combined system is 27r/w
so w
= 30.
Since the mass of an 8-pound weight is 1/4 slug, we have from w2

302
.s:
1/4
= 4k
or
= 7r/15,
= k/m
k = 225.
We now have the system of equations
Solving the second equation for k1 and substituting in the first equation, we obtain
4(3k2/2)k2
3k2/2 + k2
12k = 12k2
5k2
5
= 225.
Thus, k2 = 375/4 and k1 = 1125/8. Finally, the value of the first weight is
W =.32m = k1 = 1125/8
3
3
= 375 ~ 46.88 lb.

8
15. For large values of t the differential equation is approximated by x" = O. The solution of this
equation is the linear function x = clt
+ C2.
Thus, for large time, the restoring force will have
decayed to the point where the spring is incapable of returning the mass, and the spring will simply
keep on stretching.
16. As t beco mes larger the spring constant increases; that is, the spring is stiffening. It would seem
that the oscllations would become periodic and the spring would oscllate more rapidly. It is likely
that the amplitudes of the oscillations would decrease as t increases.
17. (a) above
(b) heading upward
18. (a) below
(b) from rest
19. (a) below
(b) heading upward
20. (a) above
(b) heading downward
iX" + x' + 2x = O, x(O) = -1, and X'(O) = 8 we obtain x = 4te-4t - e-4t and x' =
8e-4t - 16te-4t. If x = O then t = 1/4 second. If x' = O then t = 1/2 second and the extreme
displacement is x = e-2 feet.
21. From
196
Exercises 5. 1
= O, x(O) = O, and XI(O) = 5 we obtain x = 5te-2v'2t and

x' = 5e-2v'2t (1- 2V2t).
If Xl = O then t = V2/4 seeond and the extreme displacement is
22. From ~x" + V2XI + 2x

X = 5V2e-1/4
feet.
23.
(a) From
24.
(b) From x" + x' +

(a) x = ~e-8t ( 4t
x" + 10xl + 16x

16x
= O, x(O) = 1, and XI(O) = O we obtain x = 1e-2t - ~e-8t.
= O, x(O) = 1, and XI(O) = -12 then x = _e-2t
- 1) is never zero; the extreme displacernent is x(O)
+ ie-8t.
= 1 meter.
~e-8t (5 - 2t) = O when t = ~In a ::::: 0.153 seeond; if x' = te-8t (t -10)
t = ~In 10 :::::0.384 seeond and the extreme displacernent is x = -0.232 meter.
(b) x =
25.
(a) From
(b) x
O.lx" + O.4xl + 2x
= e-2t
V; [- jg
eos4t - ~
(e) If x = O then 4t + 4.25

t = 1.294 seeonds.
26.
(a) From
(b) x
= O, x(O) = -1, and x' (O) = O we obtain

sin4t] =
= 27r,37r, 47r, ...
V;
(~
eos4t + ~
e-2t [- eos 4t - ~ sin 4t].
e-2t sin(4t + 4.25).
so that the first time heading upward is
ix" + x' + 5x = O, x(O) = 1/2, and Xl (O)
= e-2t ~
= O then
sin4t) = ~e-2t
= 1 we obtain x
sin (4t +
= e-2t (~eos 4t + ~ sin 4t).
).
(e) If x = O then 4t+7r /4 = 7r,27r,37r, ... so that the times heading downward are t = (7 +8n)7r /16
for n=O, 1, 2, ....
(d)
1
\
\
\
.5
,,
--_-----"'2
-.5
/
/
-1
27. From
156 x"
+ 3xl+ 5x = Owe find that the roots of the auxiliary equation are m = - ~13 h/432 - 25 .
> Othen 13> 5/2.

(b) If 4132- 25 = O then 13= 5/2.
Ce) If 4132- 25 < Othen O < 13< 5/2.
(a) If 4132- 25
197
Exercises 5. 1
28. From 0.75x"
m=
+ (3x! + 6x = Oand
- ~(3 ~
(3
18 and
(32 -
= e-2(3t/3 (C1 cosh ~J(32 -
If x(O) = O and x'(O) = -2 then

29. If ~x"
+ ~x' + 6x
C1 = O and C2 =
= 10cos3t, x(O)
Xc
and xp =
13 ( cos 3t
> 3vZ we find that the roots of the auxiliary equation are
+ sin 3t)
18t
+ C2 sinh ~J(32
-3/
(32 -
18.
(V47
= e-t/2 C1 cos -2- t
V47 t )
+ C2 sin -2-
so that the equation of motion is

10
+ -(cos3t
.
+ sm3t).
= 12cos 2t+3 sin 2t, x(O) = -1, and x'(O) = 5 then
Xc
= e-t(c1 cos 2t+C2 sin 2t)
= 3 sin 2t so that the equation of motion is
and xp
x = e-t cos 2t
x
(b)
18t] .
= -2, and x'(O) = Othen
V47 t - --64. sm --V47)t

x = e-t /2 (4-- cos -3
2
3V47
2
30. (a) If x" +2x' +5x
+ 3 sin 2t.
x
(e)
steady-state
-3 transJ.ent
.
-3
31. From x" + 8x' + 16x = 8 sin 4t, x(O) = O, and x' (O)
xp = - cos 4t so that the equation of motion is
= O we obtain
Xc
= C1e-4t + c2te-4t and
C1 e-4t
1
x = _e-4t
32.
+ te-4t
As t
-> 00
= 625e-4t(24 + 100t) - 625e-t(24cos4t + 7sin4t).
the displacement x
->
O.
33. From 2x" + 32x = 68e-2t cos 4t, x(O)

xp = ~e-2t cos4t - 2e-2t sin4t so that
x
and
1
- - cos4t.
From x" + 8x' + 16x = e-tsin4t, x(O) = O, and x'(O) = O we obtain

24 -t cos 4t - 62Se
7 -t sin
. 4t so th a t
xp = -625e
x
+ c2te-4t
Xc
= O, and x'(O) = O we obtain x =
1
-21 cos 4t + 49 sin 4t + 2e-2t
198
cos 4t - 2e-2t sin 4t.
C1 cos 4t
+ C2 sin 4t
and
Exercises 5. 1
34. Since x = ~ sin(4t - 0.219) -
sin(4t - 2.897), the amplitude approaches v85/4 as t _, oo.
~e-2t
35. (a) By Hooke's law the external force is F(t) = kh(t) so that mx" + f3x' + kx = kh(t).
(b)
From ~x" + 2x' + 4x

and xp
= 20 cos t, x(O) = O, and x'(O) = O we obtain
e-2t(
el cos 2t + e2sin 2t)
= ~~cos t + ~ sin t so that

x = e-
36. (a) From 100x" + 1600x

and xp =
2 (
t
56
72)
-13 cos 2t - 13 sin 2t
56
32 .
+ 13 cos t + 13 SIn t.
= 1600sin8t, x(O) = O, and x'(O) = O we obtain
Xc
= el cos4t + e2sin4t
-l sin 8t so that
x
(b) If x
Xc
. 4t - -l. SIn 8t.

= -2 SIn
3
= lsin4t(2 - 2cos4t) = Othen t = mr/4 for n = 0,1,2, ....
(e) If x' = ~ cos 4t- ~ cos 8t = ~(l-cos 4t)(1 +2 cos 4t) = Othen t = 1f /3+mr /2 and t = 1f /6+n1f /2
for n = O, 1, 2, ... at the extreme values. Note: There are many other values of t for which
x' = O.
(d ) x(1f/6 + n1f/2) = V3/2 cm. and x(1f/3 + n1f/2) = -V3/2cm.
(e)
x
1
37. From x" + 4x
= -5 sin 2t + 3 cos 2t, x(O) = -1, and x'(O) = 1 we obtain
Xc
= el cos 2t + e2sin 2t,
xp = ~t sin 2t + ~t cos 2t, and

3.
x = - cos 2t - -1 sin 2t + -t
844
SIn
38. From x" +9x = 5 sin3t, x(O) = 2, and x'(O) = Owe obtain
2t + -t
5 cos 2t.
Xc
= el cos3t+e2 sin3t,
xp = -~tcos3t,
and
5.
x = 2 cos 3t + 18 SIn 3t - 6't cos 3t.

39. (a) Frorn
xp =
x'' + w2x = Focos"'(t, x(O) = O, and x'(O)

(Fo cos "'(t)/
(w2
= O we obtain
",(2)so that
x =
Fo
w -"'(
2 cos wt +
199
Fo
w -"'(
2 cos "'(t.
Xc
= el coswt + e2sinwt and
Exercises 5. 1
(b)
lim
"(-+w
Fo
2
2(cos 1t - coswt) = lim
w - 1
"(-+w
40. From x" + w2x = Focoswt,

xp
= (Fot/2w)sinwt
- Fot sin ryt

-
21
Fo
= -t sinwt.
2w
x(O) = 0, and x'(O) =
= (Fot/2w)sinwt
so that x
we obtain Xc = Cl coswt + C2sinwt and
and
lim FOtsinwt = FOtsinwt.

2w
2w
41. (a) From cos( u - v) = cos u cos v + sin u sin v and cos( u + v) = cos u cos v - sin u sin v we obtain
"(-+w
sin u sin v = ~[cos(u - v) - cos(u + v)]. Letting u
~(1 - w)t and v = ~(1 + w)t, the result
follows.
(b) If 10 = 1(1 - w) then 1
;::::o
w so that x = (Fo/2q) sin l sin 1t.
42. See the article "Distinguished Oscillations of a Forced Harmonic Oscillator" by T.C. Procter in The
College Mathematics Journal, March, 1995. In this article the author illustrates that for Fo = 1,
>. = 0.01, 1
22/9, and w
2 the system exhibits beats oscillations on the interval [0,97r], but
that this phenomenon is transient as t .......oo.
-1
43. (a) The general solution of the homogeneous equation is
xc(t) = cle-,\t cos( jw2 - >.2t) + c2e-,\t sin( jw2 - >.2t)

= Ae-,\t
where A
sin[jw2 - >.2t + 4>],
V cy + c , sin 4> = ci] A, and cos 4> = C2/A.
Fo(w2 - 12).
xp(t) = (w2-12)2+4>.212
sm1t+
Now
Fo( -2>'1)
(W2-12)2+4>.212
where
Fo( -2>'1)
and
200
.
COS1t= Asm(1t+B),
Exercises 5. 1
= O then v
+ 2).2 -
= O so that. v = Oor , = VW2 - 2).2. The first derivative

test shows that 9 has a maximum value at , = vw2 - 2).2. The maximum value of gis
(b) If g'(I)
(,2
W2)
,1 = vw
and the resonance curve grows without bound at ,1 =
(c) We identify w2
= k/m =
4, ).
{3/2, and
2 - 2).2
= J4 -
{32/2. As {3---; O,
,1 ---;
2. That is, the system approaches pure
resonance.
(3
yI
2.00
1. 00
0.75
0.50
0.25
1. 41
1. 87
1. 93
1. 97
1. 99
0.58
1. 03
1. 36
2.02
4.01
,t
44. (a) For n = 2, sin2

when
'1 = w /2.
1(1- cos2,t).
The system is in pure resonance when 2,d27r = w/27r, or
(b) Note that
,t
sin3
= sin ,t sin2 ,t
= '2 [sin ,t - sin ,t cos 2,t].
Now
sin(A
+ B) + sin(A
- B) = 2sinAcosB
so
sin ,t cos 2,t = ~[sin 3,t - sin ,t]
and
sin3
,t
'4 sin ,t
'4 sin 3,t.
Thus
x"
+ w 23.
x = '4 sm v t - 'l'3t
4 sm
,.
The frequency of free vibration is w/27r. Thus, when ,d27r = w/27r or

3'2/27r = w /27r or 3,2 = w or
,3
,1
= w /3, the system will be in pure resonance.
201
= w, and when
Exercises
(e)
5. 1
yl=1/2
yl=l
y2=1I3
10
n=3
10
-5
-10
-10
+ 2q' + 100q = O we obtain q(t) = e-20t(cr cos 40t + C2 sin 40t). The initial conditions
q(O) = 5 and q'(O) = O imply cr = 5 and C2 = 5/2. Thus
45. Solving 2~q"
q(t)
= e-20t (5cos40t+
~sin40t)
:::::)25
+ 25/4e-20tsin(40t
+ 1.1071)
and q(O.Ol) :::::4.5676 coulombs. The charge is zero Ior the first time when 40t
+ 0.4636 =
7f
or
t::::: 0.0509 second.
46. Solving
h" + 20q' + 300q

= O imply
and q'(O)
= O we obtain q(t) = ce-2t + C2e-60t. The initial conditions q(O) = 4

= 6 and C2 = -2. Thus
q(t) = 6e-20t _ 2e-60t.
= O we find Ot = 1/3 which implies t < O. Therefore the charge is never O.

Solving %q" + 10q' + 30q = 300 we obtain q( t) = e-3t (C cos 3t + C2 sin 3t) + 10. The initial
Setting q
47.
q(O) = q'(O) = O imply
conditions
= C2 = -10. Thus
q(t)=10-10e-3t(cos3t+sin3t)
and
i(t)=60e3tsin3t.
Solving i( t) = O we see that the maximum charge occurs when t
= 7f /3
and q( 7f /3) :::::10.432.
+ 100q' + 2500q = 30 we obtain q(t) = ce-50t + c2te-50t + 0.012. The initial conditions
q(O) = O and q'(O) = 2 imply C = -0.012 and C2 = 1.4. Thus
48. Solving q"
q(t) = -0.012e-50t
+ 1.4te-50t + 0.012 and i(t)
Solving i(t) = Owe see that the maximum charge occurs when t
= 2e-50t
=
- 70te-50t
1/35 and q(1/35) :::::0.01871.
+ 2q' + 4q = Owe obtain Yc = e-t (cos v'3 t + sin v'3 t). The steady-state charge has the
form Yp = A cos t + B sin t. Substituting into the differential equation we find
49. Solving q"
(3A
Thus, A
+ 2B) cos t + (3B - 2A) sin t = 50 cos t.
= 150/13 and B = 100/13. The steady-state charge is

qp(t)
150
100
13cos t + 13 sin t
and the steady-state current is
202
Exercises 5. 1
50. From
Ea
ip(t) = Z
and Z
= JX2 + R2
(RZ sn-rz - ZX cos,t )
we see that the amplitude of ip(t) is
A = jE'6R2
E'6X2 = Ea. /R2 X2 = Ea
\ Z4 + Z4
Z2 V
+
Z .
51. Thc differential equation is ~q" + 20q' + 1000q
= 100sin t. To use Example 10 in the text we identify
Ea = 100 and , = 60. Then

111
X = L, - ~ = 2(60) - 0.001(60) ~ 13.3333,
Z = ,j X2 + R2 =
,j X2 + 400 ~
24.0370,
and
Eo = 100 ~ 4 1603.
From Problem 50, then
+ cp)
ip(t) ~ 4.1603(60t
where sin
angle. Now
-X/Z
and coscp = R/Z. Thus tan
~ -0.6667 and cp is a fourth quadrant
-X/R
cp ~ -0.5880 and
ip(t) ~ 4.160~(60t - 0.5880).
52. Solving
h" + 20q' + 1000q = O we obtain
qc(t) = (ci cos 40t
+ B cos 60t + e sin 40t + Deos
has the form qp(t) = A sin 60t
+ C2 sin 40t).
The steady-state charge
40t. Substituting
equation we find
(-1600A
- 2400B) sin 60t
+ (400C
= 200 sin 60t
Equating coefficients we obtain A
+ (2400A
- 1600B) cos 60t
- 1600D) sin 40t
+ (1600C + 400D)
cos40t
+ 400 cos 40t.
-1/26,
-3/52,
e = 4/17,
and D
state charge is
qp(t)
1
3
26 sin 60t - 52 cos 60t
=-
4
17 sin 40t
1
17 cos 40t
and the steady-state current is

30
cos60t
13
ip(t) = --
45
sin60t
13
+-
203
160
40
cos40t - - sin40t.
17
17
+-
1/17. The steady-
Exercses 5. 1
53. Solving ~q"
+ 10q' + 100q =
150 we obtain q(t) = e-lOt(cl cos 10t
conditions q(O) = 1 and q'(O) = O imply Cl = C2= -1/2.
+ C2sin 10t) + 3/2-.
The initial
Thus
q(t) = -~e-lOt(cos 10t + sin 10t) + ~.

2
As t
-+ 00,
q(t)
-+
3/2.
where Z = JX2 + R2 and

X = L"( - l/C"(. Since Eo is constant the amplitude will be a maximum when Z is a minimum.
Since R is constant, Z will be a minimum when X = O. Solving L"( - l/C"( = O for "( we obtain
"(= 1/..JLC. The maximum amplitude will be Eo/ R.
54. By Problem 50 the amplitude of the steady-state current is Eo/Z,
55. By Problem 50 the amplitude of the steady-state current is Eo/Z,

X
L"( -
l/C"(.
+ R2
and
Since Eo is constant the amplitude will be a maximum when Z is a minimum.
Since R is constant, Z will be a mnimum when X.

C
where Z = JX2
O. Solving L"( - 1/ c"( = O for C we obtain
= 1/L,,(2.
56. Solving O.lq"
+ 10q =
Asin"(t + Bcos"(t.
100sin"(t we obtain q(t)
c cos
10t + C2sin 10t + qp(t) where qp(t) =
Substituting qp(t) into the differential equation we find

(100 - "(2)A sin "(t + (100 - "(2)B cos "(t = 100sin "(t.

initial conditions q(O)
= 100/(100
= q'(O) = O imply
q(t) =
- ,,(2)and B
Cl = O and C2=
= O. Thus, qp(t) =
100 2 sin "(t. The

100 - "(
-10"(/(100 - "(2). The charge is
102(lOsin"(t-"(sin10t)
100 - "(
and the current is
i(t) =
100"(
2(cos"(t-coslOt).
100 - "(
57. In an L-C series circuit there is no resistor, so the differential equation is

d2q
L dt2 + C q = E(t).
Then q(t) = Clcos (t/..JLC) +C2 sin (t/..JLC)
qp(t) into the differential equation we find
(~-
+qp(t) where qp(t) = A sin "(t+ B cos "(t. Substituting
L"(2) Asin"(t + (~ - L"(2) Bcos"(t = Eocos"(t.
Equating coefficients we obtain A = O and B
= EoC/(l - LC"(2). Thus, the charge is
1
1
EoC
q(t) = Cl COS ..JLC t + C2sin ..JLC t + 1 _ LC"(2 cos vr.
204
Exercises
The initial conditions q(O) = qo and q'(O) = io imply
The current is
Cl
= qo - EoC/(l-
.
Cl.
1
C2
1
EoC,.
2(t) = - ..JLC sin ..JLC t + ..JLC cos ..JLC t - 1 _ LC,2
= 20 COS ..JLC t -..JLC
EOC).
qO- 1_ LC,2
SIn
LC,2) and
t - 1_
58. When the circuit is in resonance the form of qp(t) is qp(t) = At cos kt
= io..JLC.
,t
EoC,.
LC,2
1
SIn..;re
C2
5.2
+ Bt
SIn
,t.
sin kt where k = 1/..JLC .
Substituting qp(t) into the differential equation we find

2
Eo
q~ + k q = -2kA sin kt + 2kB cos kt = Leos kt.
= O and B
= Eo/2kL.
The charge is
q(t) = C! cos kt + C2 sin kt + 2~~ t sin kt.

The initial conditions q(O) = qo and q'(O) = io irnply C!
= qo and
C2
= io/k. The current is
i(t) = -ek sin kt + C2kcos kt + 2~~ (kt cos kt + sin kt)

EO
) ..
Eo
= ( 2kL - qok SIn kt + 20 cos kt + 2L t cos kt.
Exercises 5.2
1. (a) The general solution is

(
YX)=C!+C2X+C3X
The boundary conditions are y(O)

conditions give C! = Oand
C2 =
Wo
+C4X + 24EIX .
= O, y'(O) = O, y"(L) = O, y"'(L) = O. The first two
O. The conditions at x = L give the system

Wo
2C3+ 6C4L + 2EIL
=O
WO
6C4+ EIL = O.
Solving, we obtain
C3
= woL2/4EI
and
C4
= -woL/6EI.
y(x) = ~(6L2x2
24EI
205
The defiection is
- 4Lx3 + x4).
Exercises 5.2
(b)
~~------------x
20.40.60.81
1

(
YX)=e+C2X+C3X
+C4X
Wo
+ 24EIX
The boundary conditions are y(O) = O,y"(O) = O,y(L) = O,y"(L)

give e
= O and
C3
= O. .The conditions at x
C2L
C2
Wo
Wo
+ C4L + 24EIL
6C4L
Solving, we obtain
= L give the
= WOL3/24EI
and
C4
+ 2EIL
system
=O
= O.
= -woL/12EI.
y(x) = 2:;I(L3x
- 2Lx3
= O. The first two conditions
The defiection is
+ x4).
(b)
+-------------~x
0.2 0.4 0.6 0.8
1 y
3. (a) The general soluton is

2
y (x ) = C + C2X + C3X
Wo
+ C4X + 24EIX
The boundary conditions are y(O) = O,y'(O) = O,y(L) = O,y"(L)

give Cl = O and C2 = O. The conditions at x = L give the system
C3L2
+ C4L3 + 2:;IL4
Wo
2C3
+ 6C4L + 2EIL
206
=O
2
= O.
= O. The first two condtons
Exercises 5.2
Solving, we obtain
C3
= woL2/16EJ
and
C4
= -5woL/48EJ.
y(x) = ~(3L2x2
48EJ
- 5Lx3
The defiection is
+ 2x4).
(b)

y(x)
Cl
el
o, y'(O)
1r
sm LX'
= O,y"(L) = O. The first two conditions

= Oand e2 = -woL3 / EJ1r3. The conditions at X = L give the system
The boundary conditions are y(O) =

give
wOL4.
+ C2X + C3X + C4X + EJ1r4
C3L
Solving, we obtain
C3
wO
+ C4L + El7r3L
= 3woL2/2El7r3
woL
y(x) = 2El7r3
= O,y(L)
(2-2L
and
X
=O
2e3
+ 6C4L = O.
C4
= -woL/2El7r3.
+ 3Lx
The defiection is
2 - X 3 + ----;2L3
1r)
sin LX
(b)
1 Y
(e) Using a CAS we find the maximum defiection to be 0.270806 when
207
= 0.572536.
Exercises 5.2
The boundary conditions are y(O) = o, y"(O) = O, y(L)
give C1= Oand
C3
= O,y"(L) = O. The
first two conditions
= O. The conditions at x = L give the system

3
Wo C2L + C4L + 120EIL" = O
Wo
6C4L + 6EIL
Solving, we obtain
C2
= 7woL4/360EI
and
C4
= O.
= -woL2/36EI.
The defiection is
y(x) = 36~~I(7L'x - lOL2x3 + 3x5).

(b)
(e) Using a CAS we find the maximum defiection to be 0.234799 when x = 0.51933.
6.
(a) Ymax
wOL4
= y(L) = 8EI
(b) Replacing both L and x by L/2 in y(x) we obtain woL4/128EI,
which is 1/16 ofthe maximum
defiection when the length of the beam is L.
5woL4
(e) Ymax = y(L/2) = 384EI
(d) The maximum defiection in Example 1 is y(L/2) = (wo/24EI)L4/16
1/5 of the maximum displacement of the beam in Problem 2.
WOL4/384EI, which is
7. The general solution of the differential equation is
{P
(P Wo 2 woEI
El x + C2 sinh V El x + 2P x + ---;r .
y = C1cosh V
Setting y(O) = O we obtain
Cl
-woEI / p2, so that
woEI
{P
. (P wo 2 woEI
y = ----;r cosh V El x + C2 sinh V El x + 2P x + ---;r'
Setting y' (L) = O we find
C2
(fu
w~~I sinh
fu
L - W;L) /
208
fu fu
cosh
L.
Exercises 5.2
8. The general solution of the differential equation is
y
/P
. /P
Wo
woEI
= el cos V El x + e2sin V El x + 2P x + ---p2 .
Setting y(O) = O we obtain el = -woEI / p2, so that
/P
. /P
El x + cz sin V El x +
'WoEI
y = - ---:>2 cos V
Setting y' (L)
:s: O the
'WoEI
= O we find
e2 =
9. For A
Wo
2P x + --p2 .
/P woEI. /P WOL)
(- V El ----p2 Slll V El L - P
only solution of the boundary-value
V/P
El cos V/P
El L.
problem is y = O. For A
> O we have
y = el cos V>-x + e2sin V>-x.

Now y(O) = O implies el = O, so
gives
V>-n=nn
or
A=n2,
n=1,2,3,
....
The eigenvalues n2 correspond to the eigenfunctions sin nx for n = 1, 2, 3, ....

10. For A :::;O the only solution of the boundary-value problem is y
= O. For A > O we have

Now y(O)
= O implies el = O, so
y
() = e2sin V>- = O
gives
,n
VA=nn
or
A=16n,
n=1,2,3,
....
The eigenvalues 16n2 correspond to the eigenfunctions sin 4nx for n = 1, 2, 3, ....
11. For A
:s: O the
only solution of the boundary-value problem is y = O. For A

Now
y'(x) = -el V>- sinV>-x+e2V>-cosV>-X

and y'(O) = O implies e2 = O, so
y( L) = el cos V>-L = O
209
> Owe have
Exercises 5.2
gives
r:
VA
L =
(2n - 1)7f
2
or
A=
(2n - 1)27f2
' n = 1, 2, 3, ....
4L2
.
2
.
(2n - 1)7f
The eigenvalues (2n - 1) 7f2/4L2 correspond to the eigenfunctions cos
L
x for
n
= 1,2,3, '" .
12. For A:::; O the only solution of the boundary-value problem is y

y
Now y(O)
= O. For A > O we have
= el cos,f).. x + e2sin ,f).. x.
= O implies CI = O, so
y' (~)
= e2,f).. cos,f)..
=O
gives
,f).. ~ = (2n;
1)7f or
A = (2n - 1)2, n = 1, 2, 3, ....
The eigenvalues (2n - 1)2 correspond to the eigenfunctions sin(2n - l)x.
< O the only solution of the boundary-value problem is y = O. For A = O we have y = CIX +cz.
= CI and y'(O) = O implies el = O. Then y = e2 and y'(7f) = O. Thus, A = O is an eigenvalue
with corresponding eigenfunction y = 1.
13. For A
Now y'
For A > O we have

y
= el cos,f).. x + e2sin ,f).. x.
Now
y' (x) = -el ,f)..sin ,f)..x
and y'(O)
+ e2,f).. cos ,f).. x
= O implies e2 = O, so
gives
J>: 7f = n7f
or
A = n2, n
= 1, 2,
3, .. , .
The eigenvalues n2 correspond to the eigenfunctions cos nx for n
= O, 1, 2, ....
14. For A :::;O the only solution of the boundary-value problem is y = O. For A > O we have
y
= el cos J>: x + cz sin J>: x.
Now y( -7f) = Y(7f) = O implies

el cos J>: n - e2sin ,f).. n = O
el cos J>: n + e2sin ,f).. n = O.
210
(1)
Exercises
5.2
This homogeneous system will have a nontrivial solution when

cos ~ tt
1\
I cos V A tt
sin ~ tt I
.
1\
= 2 sin ~
sin V A tt
1\
tt COS V
1\
A tt = sin 2v A tt = O.
Then
n2
A = 4;
or
n = 1, 2, 3, ....
When n = 2k - 1 is odd, the eigenvalues are (2k - 1)2/4. Since cos(2k - 1)7f/2 = O and
sin(2k - 1)7f/2 =1- O, we see from either equation in (1) that C2 = O. Thus, the eigenfunctions
corresponding to the eigenvalues (2k - 1f /4 are y = cos(2k - l)x /2 for k = 1, 2, 3, ....
Similarly,
when n = 2k is even, the eigenvalues are k2 with corresponding eigenfunctions y = sin kx for
k
= 1, 2, 3, ....
15. The auxiliary equation has solutions
m=~(-2J4-4(A+1))
For A
=-1!=1.
< O we have
y = e -x
(q cosh R x + C2sinh R x) .
The boundary conditions imply

y(O)
y(5) =
ci
sinh
C2e-5
5R =
so C1= C2= O and the only solution of the boundary-value problem is y = O.

For A = O we have
and the only solution of the boundary-value problem is y = O.

For A
> O we have
y = e -x
(C1
COS ~
or
A =
X + C2sin ~
x) .
Now y(O) = O implies q = O, so

gives
n27f2
5~
The eigenvalues n27f2/25
16. For A
= nit
---:25'
n = 1,2,3, ....
n7f
correspond to the eigenfunctions e-x sin 5x for n
< -1 the only solution of the boundary-value problem is
y = O. For A = -1 we have
y = CIX+ C2. Now y' = C1and y'(O) = O implies C1= O. Then y = C2and y'(l)
is an eigenvalue with corresponding eigenfunction y
211
1.
1, 2, 3, ....
= O. Thus, A = -1
Exercises 5.2
For A
> -1 we.have
y = ClCOS~x+c2sin~x.
Now
y' = -cl~sin~x+c2Jr+lcosJr+lx
and y'(O) = O implies C2= O, so
y'(l)
= -clVI+!sinJr+l
gives
~
= n1f
or
/\ = n21f2 -1,
n = 1,2,3, ....
The eigenvalues n21f2 - 1 correspond to the eigenfunctions cos titcx for n = O, 1, 2, ....
17. For A = O the only solution of the boundary-value problem is y = O. For A
=f. O we have
y = ClCOSAX + C2sin AX.

Now y(O)
=O
implies Cl = O, so
y(L) = c2sinAL
gives
AL = tit: or
A=
=O
n1f
L'
ti
= 1,2,3, ....
The eigenvalues n1f/ L correspond to the eigenfunctions sin

18. For A = O the only solution of the boundary-value
n1f
LX
for
ti
= 1, 2, 3, ....
problem is y = O. For A
=f. O we have
y = ClCOSAX + C2sin AX.

Now y(O) = O implies Cl = O, so
gives
.
(2n - 1)1f
31fA=
or
A=--,
2n - 1
n=l,2,3,
The eigenvalues (2n - 1)/6 correspond to the eigenfunctions sin 2n;

19. For A
....
1x for n = 1, 2, 3, ....
> O a general solution of the given differential equation is

y
= q cos(~
lnx) + c2sin(~
lnx).
Since In 1 = O, the boundary condition y(l) = O implies Cl = O. Therefore

y = c2sin(~
Using In e"
= 1f we find that y (e")
lnx).
= O implies
C2Sin~1f
212
=O
Exercises 5.2
or v0."'Jr= n'Jr, n = 1, 2, 3, ....
The eigenvalues and eigenfunctions are, in turn,
A = n2,
= 1, 2, 3,
...
and
For A ::; O the only solution of the boundary-value
= sin (n In x).
problem is y = O.
To obtain the self-adjoint form we note that the integrating factor is (1/x2)ef
dx]
l/x.
That is,
the self-adjoint form is

A = O.
- d [xy '] + -y
dx
x
Identifying the weight function p(x) = l/x we can then write the orthogonality relation
r
)1
20.
e"
1
-sin(nlnx)sin(mlnx)dx
x
For A = O the general solution is y
m =J n.
Cl + C2lnx. Now y' = C2/ z, so y' (e-l)
C2e
O implies
= O gives Cl = O. Thus y(x) = O.
C2= O. Then y = Cl and y(l)

For A < O, y
= O,
= C1X-vY + C2XvY.
The initial conditions give C2 = cle2vY
and Cl = O, so that
C2= O and y (x) = O.

For A > O, y
= Clcos( v0."In x) + C2sin( v0."In x). From y(l)
y = c2sin(v0."lnx).
Now y'
c2(v0."/x)cos(v0."lnx),
cos v0."= O or A = (2n - 1)2'Jr2/4for n
so y'(e-l)
1, 2, 3, ....
= c2ev0."cosv0." = O implies
The corresponding eigenfunctions are
2n - 1
y = sin ( -2-'Jrlnx
21. For A = O the general solution is y = el +C2Inx.
= O we obtain Cl = O and
Now y' = C2/X, so y'(l)
= C2 = O and y
el.
Since y'(e2) = O for any Cl we see that y(x) = 1 is an eigenfunction corresponding to the eigenvalue
A = O.
For A < O, y
= C1X-vY + C2XvY. The initial conditions imply el = C2= O, so y(x) = O.

For A > O, y = Clcos( v0."In x) + C2sin] v0."In x). Now
,
~.
Y = -cl-sm(VA
and y'(l)
= e2v0." = O implies C2 = O.
A = n2'Jr2/4 for n = 1, 2, 3, ....
r:
v0."
r:
lnx) +C2-COS(VA lnx),
x
Finally,
y'(e2)
= -(elv0."/e2)sin(2v0.")

y = cos (~'Jr In x ) .
22.
For A> 1/4 a general solution of the given differential equation is

y
C1X-
1/'">
-cos
(/4A 2 - 1)lnx
+ C2X- 1/2 sin
213
(/2
1 lnx. )
O implies
Exercises 5.2
Since In 1 = O, the boundary condition Y(l) = Oimplies cr= O. Therefore
y = C2X-l/2sin (~
lnx).
Using In e2 = 2 we find that y( e2) = O implies

c2e-lsin(V4A-1)
or
J4I-=1 = n-lr, n = 1, 2, 3, ....

A=
(n27r2 + 1) /4,
=0
The eigenvalues and eigenfunctions are, in turn,

n = 1,2,3,
...
For A < O the only solution of the boundary-value
and
= x-l/2sin
(n27rlnx).
problem is y = O.
For A = 1/4 a general solution of the differential equation is

y = ClX-l/2
From y(l)
O we obtain Cl
+ C2x-l/2In
x.
= C2x-l/2In x. From y(e2) = O we obtain 2c2e-l = O or
O, so y
C2= O. Thus, there are no eigenvalues and eigenfunctions in this case.

To obtain the self-adjoint form we note that the integrating factor is
= (1/x2).
(1/x2)eJ(2/x)dx
x2 = 1.
That is, the self-adjoint form is

d~
[x2y'] + Ay
= O.
Identifying the weight function p(x) = 1 we can then write the orthogonality relation
le
or
i .x-l/2 sin
(~7r
he x-l sin (~7r
In
x) x-l/2 sin (n;
lnx) sin
lnx)
(n27rlnx) dx = O,
23. If restraints are put on the column at x = L/4, x = L/2, and x
criticalload
dx = O,
m
m i=
n,
i= n.
= 3L/4, then the
will be P4.
24. (a) The general solution of the dfferential equation is
= Cl COS
ffi
+ C2sin
214
ffi
+ O.
Exercises 5.2
= o, the initial conditions are y(O) = y'(O) = O. If = O

this implies that CI = C2= O and y(x) = O. That is, there is no deflection.
Since the column is embedded at x
(b)
If =1- O, the initial conditions give, in turn,
q =
- and C2= O. Then
In order to satisfy the condition y(L) = we must have

or
This gives )P/El
L = mr/2 for n = 1, 2, 3, ....
cos
--:;
L = O.
The smallest value of Pn, the Euler load, is
then
25. The general solution is
I"'if-
y = CICOS
From y(O)
+ C2sin
I"'if-
= O we obtain CI = O. Setting y(L) = O we find )pw2/T
critical speeds are
Wn
= mrVT /
Lv-P,
n = 1, 2, 3, ....
n1l'
y (x) = C2sin T x,
x.
L = mr, n = 1, 2, 3, ....
n = 1, 2, 3, .. . ,
where C2=1- O.
26. (a) When T(x) = x2 the given differential equation is the Cauchy-Euler equation
+ 2xy' + pw2y = O.
x2y"
The solutions of the auxiliary equation

m( m - 1)
+ 2m + pw2 = m2 + m + pw2 = O
are
mI =
/
-"21 - "21y4pw
- h,
when pw2 > 0.25. Thus

y = CIX-I/2 COS(A
In x)
where A = )4pw2
+ C2X-I/2sin(A In x)
1/2. Applying y(l) = O gives CI = O and consequently

y = C2X-I/2sin(Alnx).
215
Thus,
The corresponding deflection curves are
Exercises 5.2
An
y(e)
The condition
= n7r, n = 1, 2,3,
O requires
....
O. We obtain
a nontrivial
solution
when
But
An
Solving for
c2e-l/2sinA
= J4pw~
- 1j2
= n7r.
gives
Wn
The corresponding
solutions
are
Yn(x) = C2X-1/2 sin(n7r lnx).

(b)
1
n=l
n=3
x
eX
-1
27. The auxiliary
1
n=2
conditions
-1
equation
is m2
u(a) = Uo and u(b)
+m =
= Ul
-1
+ 1) =
m(m
O so that
+C2
yield the system cla-l
_ (UO - Ul) b
b
a
- a
Cl -
an
u(r)
= uo,
+ C2.
clr-1
The boundary
= Ul.
clb-l +C2
Solving gives
_ ulb - uoa
b
.
- a
C2 -
Thus
UO - Ul) ab
ulb - uoa
ur() = (
-+
b-a
r
b-a
28.
The auxiliary
equation
is m2 = O so that u(r)
and u(b) = Ul yield the system Cl
+ C21n a
+ C21n r. The
Cl + C21n b = Ul.
= Cl
= uo,
- uolnb
ln(ajb)
ullna
C =
Thus
u(r) = ullna - Uo lnb

ln(ajb)
+ Uo -
and
.
boundary
conditions
u(a) ~ uo
Solving gives
Uo - Ul
C2= ln(ajb)'
Ul 1
_ Uo ln(rjb) - ulln(rja)
nr ln(ajb)
ln(ajb)
29. (a) The general solution ofthe differential equation is y = Cl cOS4X+C2 sin4x.
From yo = y(O) = Cl
+ c2sin4x.
any solution must
we see that y
satisfy yo
= yocos4x
Yl We also see that when
boundary-value
= Y(7rj2) = yO we see that

yO = Yl, Y = yO cos 4x + C2sin 4x
From Yl
problem for any choice of C2. Thus, the boundary-value
a unique solution for any choice of yO and Yl.

(b)
Whenever
(e) When yo
yO = Yl there are infinitely many sol utions.
=f. Yl there
will be no solutions.
216
is a solution of the
problem does not have
Exercises 5.2
(d) The boundary-value problem will have the trivial solution when YO
will not be unique.
30.
(a)
Yl
O. This solution
COS 4x + C2 sin 4x. From 1 = y(O) = Cl

+ C2 sin 4x. From 1 = y(L) = cos 4L + c2 sin 4L we see that C2 =
The general solution of the differential equation is Y = Cl

we see that Y = cos 4x
(1 - cos 4L) / sin 4L. Thus,
Y = cos 4x
will be a unique solution when sin 4L
(1-.sm4CO~4L) sin 4x
i- O; that is, when L i- kt: /4 where k = 1, 2, 3, ....
(b) There will be infinitely many solutions when sin 4L = O and 1 - cos 4L
O; that is, when
L = br/2 where k = 1, 2, 3, ....

(e) There will be no solution when sin 4L
i- O and 1 - cos 4L i- O; that is, when L
= br /4 where
k = 1, 3, 5, ....
(d) There can be no trivial solution since it would fail to satisfy the boundary conditions.
31. (a) A solution curve has the sarne y-coordinate at both ends of the interval
[-7r,
1 and the tangent
7r
lines at the endpoints of the interval are parallel.

(b)
For A
or
=O
2Cl7r
= O is Y = ClX + C2.
the solution of y"
= O. Thus,
Cl
From the first boundary condition we have
= O and Y = C2. This constant solution is seen to satisfy the boundary-
value problem.
For A
< O we have Y
= Cl cosh AX
y(
-7r)
+ C2 sinh AX.
In this case the first boundary condition gives
= Cl cosh( -An)
= Cl cosh A7r -
+ C2 sinh( C2
sinh A7r
= Y(7r) = Cl cosh A7r
or
2C2
sinh A7r = O. Thus
a similar fashion that
Cl
C2
= O and Y = Cl cosh AX.
= O. Thus, for A
A7r)
+ C2 sinh A7r
The second boundary condition implies in
< O, the only solution of the boundary-value problem
is Y = O.
For A
> O we have Y
= Cl
COS AX
y(
-7r)
+ C2 sin AX.
=
Cl
The first boundary condition implies
cos( - A7r)
= Cl COS A7r
= Y(7r)
= Cl
+ C2 sin( C2
sin A7r
COS A7r
217
An)
+ C2 sin A7r
Exercises 5.2
or 2C2 sin A71" = O. Similarly, the second boundary condition implies 2eA sin A71"
C2
= O the solution is y = O. However, if ci =1- O or
must be an integer, n. Therefore, for Cl and
C2
C2
= O. If e =
=1- O, then sin A71" = O, which implies that A
not both O, y = e cos nx+c2 sin nx is a nontrivial
solution of the boundary-value problem. Since cos( -nx)
= cos nx and sin( -nx)
= - sin nx,
we may assume without loss of generality that the eigenvalues are An = n, for n a positive
integer. The corresponding eigenfunctions are Yn = cos nx and Yn
= sin nx.
(e)
-3
y = 2 sin 3x
y = sin 4x - 2 cos 3x
> O the general solution is y = Cl COS V).. X + C2 sin V).. x. Setting y(O)
so that y = C2 sin V).. x. The boundary condition y(l) + y'(l) = O implies
32. (a) For A
C2 sin
roots.
C2
=1- O, this equation is equivalent to tan
Since An =
< O. For A
general solution y
Cl
= O,
J).. + C2V).. COS V).. = O.
(b) We see from the graph that tan x = -x has infinitely many
Xn
Cl
V).. = -V)... Thus, the eigenvalues are

x;;, n = 1, 2, 3, ... , where the Xn are the consecutive positive roots of tan V).. = -V)...
Taking
An
= O we find
C2
= O, so y
x;;,
there are no new eigenvalues when
= O, the differential equation y"
tan x
5
= O has
= ClX + C2. The boundary conditions imply

= o.
(e) Using a CAS we find that the first four nonnegative roots of tan x = -x are approximately
2.02876,4.91318,7.97867, and 11.0855. The corresponding eigenvalues are 4.11586,24.1393,
63.6591, and 122.889, with eigenfunctions sin(2.02876x), sin( 4.91318x), sin(7.97867x), and
sin(11.0855x).
218
Exercises 5.3
Exercises 5.3
1. The period corresponding to x(O)
_
= 1 is approxi-
= 1, x'(O)
mately 5.6. The period corresponding to x(O) = 1/2, x'(O) = -1

is approximately 6.2.
-2
2. The solutions are not periodic.
x
10
8
6
4
2.___ --+------------t
-2
3. The period corresponding to x(O) = 1, x'(O)
= 1 is approx-
imately 5.8. The second initial-value problem does not have
10
8
a periodic solution.
6
4
-2
4. Both solutions have periods o approximately 6.3.
219
Exercises 5.3
5. Frorn the graph we see that [xII ~ 1.2.
.:c
3
2
-1
6. Frorn the graphs we see that the interval is approximately

(-0.8,1.1).
-1
7. Since
xeO.01x
for small values of
X,
a linearization is d2X2
.:c
8.
= x[l + O.Olx + 2! (0.01x)2 + ...

dt
+X
J ~ x
= O.
-3
For x(O)
= 1 and x'(O) = 1 the oscillations are symmetric about the line x = O with amplitude
slightly greater than 1.

For x(O) = -2 and x'(O) = 0.5 the oscillations are symmetric about the line x = -2 with small
amplitude.
For x(O) =
vI2 and
x'(O) = 1 the oscillations are symmetric about the line x = Owith amplitude a
220
Exercises 5.3
little greater than 2.
For x(O) = 2 and x' (O)
0.5 the oscillations are symrnetric about the line x
2 with small
amplitude.
For x(O) = -2 and
Xl
(O) = Othere is no oscillation; the solution is constant.
For x(O) = -.;2 and XI(O) = -1 the oscillations are symmetric about the line x = Owith amplitude
a little greater than 2.
9. This is a damped hard spring, so al! solutions should be oscillatory with x -) O as t -) oo.
10. This is a damped soft spring, so so we expect no oscillatory

solutions.
x
5
4
3
2
-Jo
-2
221
Exercises
11. .
5.3
k l,
0.01
15
k1
-3
-15
x
kl
20
k1
100
-2
-2
-3
-3
When kl is very small the effect of the nonlinearity is greatly diminished, and the system is close
to pure resonance.
12. (a)
= -0.000471
-0.000472
The system appears to be oscillatory for -0.000471 :::;kl
< O and nonoscillatory for
. :::;-0.000472.
(b)
-0.078
= -0.079
The system appears to be oscillatory for -0.077 :::;k1 < O and nonoscillatory for k1 :::;0.078.
222
Exercses 5.3
13. For,\2 -
W2
> O we choose ,\
2 and w
1 with
= 1 and x'(O) = 2. For,\2 - w2 < Owe choose

,\ = 1/3 and w = 1 with x(O) = -2 and z'(O) = 4.
x(O)
In both cases the motion corresponds to the overdamped and underdamped
cases for spring/rnass
systems.
-3
14. (a) Setting dy/dt
= v, the differential equation in (13) becomes duf d: = _gR2/y2.
chain rule, dv/dt
= (dv/dy)(dy/dt)
= vdv/dt,
so vdv/dy
= _gR2/y2.
But, by the
Separating variables
and integrating we obtain

vdv = -gR
Setting v
2dy
y2
and
gR2
y
=-+c.
= va and y = R we find e = -gR + ~V6 and

R2
v2 = 2g(b) As y --;
1 2
-v
00
- 2gR+v6.
we assume that v --; 0+. Then v6
= 2gR
and va
= V29R.
(e) Using 9 = 32 ft/s and R = 4000(5280) ft we find

va
(d)
VD
= V2(32)(4000)(5280) ::::::
36765.2 ft/s::::::25067 mi/hr.
= V2(0.165)(32)(1080):::::: 7760 ft/s::::::5291 mi/hr
15. (a) Intuitively, one might expect that only half of a 10-pound chain could be lifted by a 5-pound
force.
(b) Since x = O when t
= O, and v = dx/dt = V160 - 64x/3, we have v(O) = v'I65:::::: 12.65 ft/s.
(e) Since x should always be positive, we solve x(t) = O,

getting t = O and t
~j5fi ::::::
2.3717.
Since the
x
10
8
graph o x(t) is a parabola, the maximum value occurs
at t- = ~j5fi. (This can also be obtained by solving
x'(t)
x(tm)
= O.) At this time the height of the chain is

::::::
7.5
0.5
ft. This is higher than predicted because
1.5
2.5t
of the momentum generated by the force. When the chain is 5 feet high it still has a positive
velocity of about 7.3 ft/s, which keeps it going higher for a while.
16. (a) Setting dxf d:
= v, the differential equation becomes (L-x)dv/dt-v2
rule, duf d: = (dv/dx)(dx/dt)
= vdv/dx,
so (L - x)vdv/dx
223.
- v2
= Lg. But, by the chain

= Lg. Separating variables
Exercises
5.3
and integrating we obtain

vIl
v
so VV2
+ Lg
L dV=-L-dx
9
-x
and
= c/(L - X). When X
-21n(v2+Lg)=-ln(L-x)+lnc,
= O,V =
O, and c
Ly'Eg.
Solving for V and simplifying
we get
dx
dt
= V(X)
V Lg(2Lx
- X2)
L - X
.
Again, separating variables and integrating we obtain

L-x
-=====:==dx
Lg(2Lx - x2)
= dt
Since z (O) = O, we have
=L
x(t)
= O and
V 12 -
y'2Lx - x2
y'Eg
=t+cr
and
y'2Lx - x2/y'Eg
= t.
Lgt2
= dx = -=VL=L=g=t=
and
v(t)
Solving for X we get
dt
/L - gt2
(b) The chain will be completely on the ground when x(t) = L or t =
;;.
(e) The predicted velocity of the upper end of the chain when it hits the ground is infinity.
17. (a) The weight of x feet of the chain is 2x, so the corresponding mass is m
= 2x/32 = x/16.
The
only force acting on the chain is the weight of the portion of the chain hanging over the edge
of the platform. Thus, by Newton's second law,
!!:_(mv)
dt
and x dv/dt
xvdv/dx
= !!:_ (.3:_v) =
+ v2 =
dt
16
32x.
]_(x
16
dv
dt
+ V dX) = ]_ (x
dt
16
Now, by the chain rule, dv/dt
dv
dt
+ v2)
= 2x
(dv/dx) (dx/dt)
v dv/dx,
so
dv
= O. This
is
+ v2 = 32x.
(b) We separate variables and write the differential equation as (v2 - 32x) dx
+ xv
not an exact form, but J.L(x) = x is an integrating factor. Multiplying by x we get

(xv2 - 32x2)dx
+ x2vdv =
O. This form is the total differential of u
3i x3 = c. Letting
implicit solution is ~x2v2 -
3 and v
= O we find c = -288. Solving for
v we get
dx
8y'x3 - 27
=v =
dt
vf3x
3 ::; x ::;8.
(e) Separating variables and integrating we obtain
x
y' 3
x - 27
dx =
8
M dt
y3
and
(X
13
224
~x2v2 - 2:fx3, so an
s
ds = ~ t + c.
y's3 - 27
vf3
Exercses 5.3
= O and
t = J3 (X
s
8 13 vs3 -
Since x = 3 when t = O, we see that e
ds.
27
We want to find t when x = 7. Using a CAS we find t(7) = 0.576 seconds.
18. (a) There are two forces acting on the chain as it falls from the platform. One is the force due to
gravity on the portion of the chain hanging over the edge of the platform.
This is FI = 2x.
The second is due to the motion of the portion of the chain stretched out on the platform. By
Newtori's second law this is
Po = _r!_[
]
2
dt mv
= _r!_[(8-X)2
dt
32
8 - x dv
1 dx
dt - 16v dt
=~
] = _r!_[~v]
dt
16
1 [
dv
2]
= 16 (8 - x) dt - v .
d
From dt [mv] = FI - F2 we have
_r!_ [2X v]
dt 32
dv
x-+v
dt
16
[(8 _ x) dvdt _ v2]

dv
dt
=32x-(8-x)-+v
dv
x dt
= 32x - 8 - + x -
dv
dt
dv
8 dt
= 32x.
dv
dt
By the chain rule, dv/dt
(b ) Integrating v dv
= 2x _ 2_
(dv/dx)(dx/dt)
= v dv/dx, so
dv
dv
dv
8 dt = 8v dx = 32x and v dx
= 4x.
= 4x dx we get ~v2 = 2x2 +e. Since v = O when x
v2 = 4x2 - 36 and v =
V4x2
= 3, we have e
= -18. Then
- 36. Using v = dx] dt, separating variables, and integrating we
obtain
dx
v'X2.=9
= 2 dt
and
x
cosh-I'3
= 2t + q.
+ el). Since x = 3 when t = O, we have cosh el

el = O. Thus, x(t) = 3 cosh2t. Differentiating, we find v(t) = daf d: = 6sinh2t.
Solving for x we get x(t) = 3 cosh(2t
= 1 and
(e) To find when the back end of the chain will leave the platform we solve x (t) = 3 cosh 2t =
8. This gives tI = ~cosh-l
6sinh(cosh-1
~)
= 2V55
~ ~
0.8184 seconds.
~ 14.83 ft/s.
225
The velocity at this instant is V(tl)
Exercises 5.3
(d) Replacing 8 with L and 32 with g in part (a) we have Ldv/dt
dv
L dt
.12
Integrating we get:zv
for v we find
dv
= Lv dx = gx
dv
v dx
and
= gx. Then
= LX.
g
v(x)=
/!!'_x2-!!'_x5.
L
L
Then the velocity at which the end of the chan leaves the edge of the platform s
v(L)
j f(p - x5)
19. Let (x, y) be the coordinates of 82 on the curve C. The slope at (x, y) is then
dy/dx
= (Vlt
- y)/(O - x) = (y - vlt)/X
or
xy' - y
Differentiating with respect to x and usng r = VdV2 gves

xy "
+ y"-
dt
y = -vI dx
xy
"
dt ds
ds dx
= -Vl--
-JI + (y')2)
xy" = -VI }:_(

V2
xy"
= rJ1 + (y')2.
Lettng u = y' and separating variables, we obtain

x du = rJ1
dx
du
~
+ u2
'!._dx
x
snh -1 u = r In x
+ In e =
ln(cz")
u = snhln cz"]
!(ex
At t = O, dy/dx
If r > 1 or r
= 2L x + e. Setting x = Xo and v = O, we find e = - 2L xo Solving
1_) .
dy =
T __
dx
2
cz"
O and x = a, so O = ca" -l/eaT. Thus e = l/aT and
< 1, integrating gives

y _- -a [ -- 1
2 l+r
(X)l+T
a
- -- 1
1-r
226
(X)l-T]
a
+ el
= -vlt.
Exercses 5.3
When t =
o, y
+ r) -
= O and x = a, so 0= (a/2)[1/(1
1/(1 - r)l
+ CI.
Thus C1= ar/(l - r2) and
(~)l+r
__1 -1 r (~)l-r]
+ ~.
a
a
1 - r2
y = 9:_[_1
2 1 +r
To see if the paths ever interseet we first note that if r

In other words, S2 always lags behind
SI.
Next. if r
> 1, then
< 1, then
VI
> V2 and y_,
VI
< V2 and y
00
as x _, 0+.
= ar/(l-
r2) when
x = O. In other words, when the submarine's speed is greater than the ship's, their paths will
interseet at the point (O,ar/(l-
r2)).
Finally, if r = 1, then integration gives
y = ~ [x2 _ ~ In x] + C2.
2 2a a
When t
and
= O, y = Oand x = a, so O = (1/2)[a/2
y=~[X2
2 2a
Sinee
u : 00 as x
-~lnx]
- (l/a) In al + C2 Thus
= -(1/2)[a/2
=~[~(x2-a2)+~ln~].
2 2a
a
-~[9:_-~lna]
2 2 a
C2
_, 0+, S2 will never eateh up with
- (l/a) In al
SI.
20. (a) Let (r, e) denote the polar eoordinates of the destroyer
SI.
When
(9, O)to (3, O) it stands to reason, sinee S2 travels half as fast as
SI,
SI
travels the 6 miles from
that the polar eoordinates
of S2 are (3, e2), where e2 is unknown. In other words, the distanees of the ships from (O,O)
are the same and r(t)
SI
= 15t then gives the radial distanee of both ships. This is neeessary if
is to intereept S2.
(b) The differential of are length in polar eoordinates is (ds)2 = (r def
Using ds/dt
= 30 and
dr/dt
= 15 then
gives
2
900
= 225t2 ( ~~) + 225
675 = 225t2 (~;)
de
dt
-=-
e(t) =
When r
= 3, e = O, so C = -y'3
e(t) =
y'3
t
v'3 In t + C = v'3 In ;5 + c.
In(1/5) and
v'3 (In ~15 -ln~) 5

227
v'3 In:3 .
+ (dr )2, so that
Exercises 5.3
Thus r = 3eB/v'3, whose graph is a logarithmic spiral.
(c) The time for SI to go from (9, O) to (3, O) =
for sorne angle
i hour.
Now SI must intercept the path of S2
/3, where O < /3 < 27l'. At the time of interception t2 we have 15t2
= 3e/3/v'3 or
t = e/3/v'3/5. The total time is then

1
t =-
21. Since (dx/dt)2
1
1
+ _e/3/v3
< -(1 + e27r/v3).
;;
;;
is always positive, it is necessary to use Idx/dtl(dx/dt)
in order to account for the
fact that the motion is oscillatory and the velocity (or its square) should be negative when the
spring is contracting.
22. (a) The approximation is accurate to two decimal places for
el = 0.3, and
accurate to one decimal place for
el = 0.6.
0.8
0.6
0.4
0.2
0.20.40.60.8
(h)
0.3r---~
0.2--
0.11-----_
+--'0"."""2
---"-0-'.
40-
e
o.
O.
O.
o.
o.
o.
O.
0.1
O.
.2
0.4
0.3
1.21.4
O.
t
The thinner curves are solutions of the nonlinear differential equation, while the thicker curves
are solutions of the linear differential equation.
23. (a) Write the differential equation as
d2
dt2 +w2sine
= O,
where w2 = g/e. To test for differences between the

earth and the moon we take e = 3, B(O) = 1, and
e/(O) = 2. Using 9 = 32 on the earth and 9 = 5.5 on
the moon we obtain the graphs shown in the figure.
-2
Comparing the apparent periods of the graphs, we see that the pendulum oscillates faster on
the earth than on the moon.
228
Exercises
5.3
(b) The amplitude is greater on the moon than on the earth.

(e) The linear model is
d2e
dt2 + w2e = O,
where w2 = g/e. When
and
e' (O) = 2, the
= 32,
e = 3, e(O)
= 1,
general solution is
e(t) = cos 3.266t + 0.612 sin 3.266t.

When g
-2
5.5 the general solution is
e(t) = cos 1.354t + 1.477sin1.354t.

As in the nonlinear case, the pendulum oscillates faster on the earth than on the moon and
still has greater amplitude on the moon.
24. (a) The general solution of
d2
dt2 + e = O
is e(t)
= Cl cost + C2sint. From e(O) = 1r/12 and e'(O) = -1/3 we find

e(t) = (1r/12)cost - (1/3)sint.
Setting e(t)
= O we have tant = 1r/4 which implies tI
(b) We set e(t) = e(O)
+ e'(O)t + ~e"(O)t + iell/(O)t + ...
= tan-I(1r/4)
::::;:
0.66577.
and use e"(t) = - sine(t) together with
e(O) = 1r/12 and e'(O) = -1/3. Then

e"(O) = - sin(1r/12) =
-h (J3 -1)/4
and
= h (J3 + 1)/12.
ell/(O) = - cos e(O) . e' (O) = - cos(1r/12)( -1/3)

Thus
e(t) = _:__ ~t _
12
(e) Setting 1r/12 - t/3
J2 (J3 -
1) t2 +
J2 (J3 + 1) t3 + .. '.
= O we obtain
tI
72
= 1r/4::::;:0.785398.
(d) Setting
_:_-~t-
J2(J3-1)t2=0
12 3
8
and using the positive root we obtain tI ::::;:
0.63088.
(e) Setting
_:_ _ ~ t _
12
J2 (J3 -
1) t2 +
J2 (J3 + 1) t3 = O
72
229
Exercises 5.3
we find tl ::::::
0.661973 to be the first positive root.
(f) From the output we see that y(t) is an interpolating function on

the intervalO:::; t :::;5, whose graph is shown. The positive root
of y(t) = Onear t = 1 is tl = 0.666404.
5
-0.4
(g) To find the next two positive roots we change the interval used in
0.4
NDSolve and Plot from {t,O,5} to {t,O,lO}. We see from the

graph that the second and third positive roots are near 4 and 7,
respectively. Replacing {t,l} in FindRoot
with {t,4} and then
{t,7} we obtain t2 = 3.84411 and t3 = 7.0218.

25. From the table below we see that the pendulum first passes the vertical position between 1.7 and
1.8 seconds. To refine our estimate of tl we estimate the solution of the differential equation on
[1.7,1.8] using a step size of h = 0.01. From the resulting table we see that tl is between 1.76 and
1.77 seconds. Repeating the process with h = 0.001 we conclude that tl ::::::1.767. Then the period of
the pendulum is approximately 4tl = 7.068. The error when using tl = 271"is 7.068 - 6.283 = 0.785
and the percentage relative error is (0.785/7.068)100 = 11.1.
h=O.Ol
h=O.l
tn
en
tn
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1.70
1.80
1.90
2.00
0.78540
0.78523
0.78407
0.78092
0.77482
0.76482
0.75004
0.72962
0.70275
0.66872
0.62687
0.57660
0.51744
0.44895
0.37085
0.28289
0.18497
0.07706
-0.04076
-0.16831
-0.30531
1.70
1.71
1.72
1.73
1.74
1.75
1.76
1.77
1. 78
1.79
1. 80
0.07706
0.06572
0.05428
0.04275
0.03111
0.01938
0.00755
-0.00438
-0.01641
-0.02854
-0.04076
h=O 001
1. 763
1. 764
1.765
1.766
1.767
1.768
1.769
1.770
0.00398
0.00279
0.00160
0.00040
-0.00079
-0.00199
-0.00318
-0.00438
230
en
1. 8 ft., since k = 4.
2. 271'/5,since
~X"
+ 6.25x
= O.
3. 5/4 m, since x = -cos4t+
~sin4t.
4. True
5. False; since an external force may existo
6. False
7. overdamped
8. From x(O) = (V2/2)sin<t> = -1/2
we see that sin
so <t>is an angle in the third or
= -1/V2,
fourth quadrant. Since x'(t) = V2cos(2t+<t, x'(O) = V2cos<t>= 1 and cos<t>> O. Thus <t>
is in the
fourth quadrant and <t>= -71'/4.
9. The period of a spring mass system is given by T
= 271'/w where w2 = k/m = kg/W,
where k is the
spring constant, W is the weight of the mass attached to the spring, and 9 is the acceleration due
to gravity. Thus, the period of oscillation is T = (271'/..jJg)JW.
is W, then (271'/..jJg)JW
= 3 and
(271'/..jJg)y'W
If the weight of the original mass
- 8 = 2. Dividing, we get JW/JW
- 8 = 3/2
or W = ~(W - 8). Solving for W we find that the weight of the original mass was 14.4 pounds.
10. (a) Solving
iX"
+ 6x
= Osubject to x(O) = 1 and x'(O) = -4 we obtain

x = cos4t - sin4t = V2sin (4t + 371'/4).
(b) The amplitude is V2, period is 71'/2,and frequency is 2/71'.
= 1 then t = n71'/2 and t = -71'/8 + n71'/2 for n = 1, 2, 3, ....

(d ) If x = O then t = 71'/16+ n71'/4 for n = O, 1, 2, .... The motion is upward for neven and
(e) If x
downward for n. odd.
(e) x'(371'/16) = O
(f) If x' = O then 4t + 371'/4= 71'/2+ n71' or t = 371'/16+ n71'.
11. Frorn
~X"
+ ~x' + 2x
= O, x(O)
= 1/3, and
x'(O)
= Owe obtain
e-2t - !e-4t.
12. From x" + /3x' + 64x = Owe see that oscillatory motion results if /32
- 256 < O or O ::; 1/31 < 16.

13. From mx" + 4x' + 2x = Owe see that non-oscillatory motion results if 16 - 8m 2': Oor O < m ::; 2.
14. From
~X"
+ x' + x = O, x(O) = 4, and x'(O) = 2 we obtain x = 4e-2t + 10te-2t.
t = 1/10, so that the maximum displacement is x = 5e-O.2
:::::
If x'(t) = O, then
4.094.
= cos"(t + sin-yz in the form x" + 634x = Bcos"(t + Bsin"(t we identify .x = O and
w2 = 64/3. The system is in a state of pure resonance when "( = w = j6473 = B/-/3.
15. Writing ~x" + ~x
231 .

16. Clearly xp = A/w2 suffices.
17. From kx" + Xl + 3x = e-t, x(O) = 2, and X/(O) = Owe obtain Xc = e-4t (c cos 2v2 t + C2sin 2v2
8 -t
17e
,an
xp =
t),
d
26
2Sv2
x=e-4t ( -cos2v'2t+--sin2v'2t
17
17
) +_e8 t.
17
18. (a) Let k be the effective spring constant and Xl and X2 the elongation of springs kl and k2. The
restoring forces satisfy klXI = k2X2 so X2
= (kjk2)XI.
From k(XI + X2) = klXI we have
k (Xl + ~~X2) = klXI
k ( k2 ~ kl )
= kl
k = klk2
kl + k2
1
1
1
k = kl + k2 .
(b) From kl = 2W and k2 = 4W we find l/k = 1/2W +1/4W = 3/4W. Then k = 4W/3 = 4m9/3.
The differential equation mx" + kx = Othen becomes x" + (49/3)x = O. The solution is
x(t) = c cos 2~
t + C2sin 2~
t.
The initial conditions x(O) = 1 and X/(O) = 2/3 imply CI = 1 and C2 = 1/.,f39.
(e) To compute the maximum speed of the weight we compute
Xl (t)
= 100sin50t, q(O)
19. From q" + 104q

qp
= 2~ sin 2~ t + ~ cos 2~ t and IXI (t) I = 4 ~ + ~ = ~ 39 + 1 .

ql(t) = O we obtain qc = ccos100t+
= O, and
A sin 50t, and

A sin 50t,
(a) q =
-1~O sin
(b) i =
-~ cos 100t + 1cos 50t, and
(e) q
100t +
= O when sin50t(1-
cos50t)
= Oor t = mr/50
for n
= 0,1,2,
....
20. (a) By Kirchoff's second law,
d2q
dq
1
L dt2 + R dt + e q = E(t).
Using ql(t)
= i(t)
we can write the differential equation in the form

di
L dt + Ri +
e1 q = E (t).
232
C2sin 100t,

Then differentiating we obtain
d2i
L dt2
+ Ri(t) + (1/C)q(t)
(b ) Frorn Li'(t)
+R
di
dt
+ ei = E
(t).
E(t) we find Li'(O) + Ri(O) + (1/C)q(0)
E(O) or Li'(O) +
Rio + (1/C)qo = E(O). Solving for i'(O) we get

i'(O)
21. For .A
Cl
>
[E(O) - b qo - Rio] .
the general solution is y =
Cl COS
J>: x + C2 sin J>: x.
cos 21rJ>: + C2 sin 21rJ>: , so the condition y(O)

=
ci
which is true when

y'
Now y(O)
Cl
and y(21r)
= y(21r) implies
cos 21rJ>: + C2 sin 21rJ>:
ci
J>: = n or .A = n2 for n = 1, 2, 3, .... Since

= - J>: Cl sin J>: x + J>: C2 cos J>: x = -nCl sin nx + nC2cos nx,
we see that y'(O) = nC2 = y'(21r) for n = 1, 2, 3, ....
Thus, the eigenvalues are n2 for n =
1, 2, 3, ... , with corresponding eigenfunctions cos nx and sin nx. When X = O, the general solution
is y = ClX + C2 and the corresponding eigenfunction is y = 1.
For .A
< O the general solution is y
y(21r) =
Cl
Cl
cosh H x
+ C2 sinh 21rH,
cosh 21rH
are no eigenvalues corresponding to .A
so y(O)
+ c2e-wt.
In this case y(O) =
can only be valid for .A
Cl
= O. Thus,
and
there
< O.
22. (a) The differential equation is d2r/dt2-w2r

re = clewt
+ C2 sinh H x.
= y(21r)
= -gsinwt.
The auxiliary equation is m2_w2 = O,so
+ Bcoswt.
A particular solution has the form rp = Asinwt
Substituting
= - 9 sin wt. Thus, B =

= (g/2w2) sinwt. The general solution of the differential equation
into the differential equation we find - 2Aw2 sin wt - 2Bw2 cos wt

O, A = g/2w2,
and "
is r(t)
= clewt + C2e-wt + (g/2w2) sinwt. The initial conditions imply
g/2w -
WCl
+ WC2 = Vo.
C1 = (2w2ro
Solving for
+ 2wvo
Cl
and
- g)/4w2
C2
we get
and
C2
Cl
+ C2
= ro and
= (2w2ro - 2wvo + g)/4w2,
so that
r ()t =
2w2ro
+ 2wvo
4w2
- 9 wt 2w2ro - 2wvo
e +
4w2
+9
-wt
+ 2w2 SIn wt.
(b) The bead will exhibit simple harmonic motion when the exponential terms are missing. Solving
q = O, C2 = O for ro and vo we find ro = Oand vo = g/2w.
To find the minimum length of rod that will accommodate simple harmonic motion we determine the amplitude of r(t) and double it. Thus L = g/w2.
233

(e) As t increases, ewt approaches infinity and e-wt approaches O. Since sin wt is bounded, the
distance, r(t), of the bead from the pivot point increases without bound and the distance of
the bead from P will eventually exceed L/2.
(d)
14
(e) For each vo we want to find the smallest value of t for which r(t) = 20. Whether we look for
r(t) = -20 or r(t) = 20 is determined by looking at: the graphs in part (d). The total times
that the bead stays on the rod is shown in the table below.
Vo
When vo
10
16.1
15
17
-20
-20
-20
20
20
1.55007
2.35494
3.43088
6.11627
4.22339
= 16 the bead never leaves the rod.

1. (a) The auxiliary equation is m 2 + 4
= Oso Xc = el cos 2t + e2 sin 2t. Letting xp = A sin 4t + B cos 4t
and substituting into the differential equation, we get -12A sin 4t - 12B cos 4t = sin 4t. Thus,
= - l~ , B = O, and xp = - f2 sin 4t. The general solution of the differential equation is

x = el cos 2t + e2 sin 2t - l~ sin 4t. The initial conditions imply el = Oand e2 = ~ + 1- ). Thus
A
(o:
x(t)
~(o:
+~)
sin2t 2
3
= ~(o: +~)sin2t 2
3
=
~ sin4t
12
~(2sin2tcos2t)
12
tt
O<t<-.
- 2
For x"
+x =
sin 4t the auxiliary equation is m2
+1=
O, so
Xc
= el cos t
+ e2 sin t.
Letting
= A sin 4t + B cos 4t and substituting into the differential equation we get -15A sin 4t 15B cos 4t = sin 4t. Thus, A = -l5 ' B = O, and xp = -l5 sin 4t. The general solution of
xp
234

the differential equation is x = Cl COSt +
x'(7r/2)
= -(ex +~) imply
Cl
= ex+
C2 sin
and
C2
t-
!5
sin 4t. The initial conditions x( tt /2)
= O,
= O. Thus
x(t) = (ex+~) cost - 115sin4t

= (ex + ~) cos t - 2_ (2 sin 2t cos 2t)
5
15
= (ex+ ~) cost - 2_(2sintcostcos2t)
= cos t [( ex+
D-
15
1~ sin t cos 2t ] ,
37r
tt
-<t<-.
2 - - 2
(b ) The velocity at the start of the second cycle is x' (37r/2) = ex+
(c) Using results from part (a), x =
x(37r/2)
= O and x'(37r/2)
Cl
= ex+
cos 2t +
125'
C2
sin 2t -
which imply
Cl
125 .
112sin 4t.
O and
C2
The initial conditions are now

= -~(ex +
fs).
Thus
2.)
x (t) = - ~ (ex +
sin 2t - 2_ sin 4t
2
15
12
= -~(ex+
175)sin2t-
= sin2t[-~(ex+
For t 2: 27r we have x =
Cl
COSt +
C2
:2(2sin2tcos2t)
2.)
- ~coS2t]
15
6
sin t -
15
'
A sin 4t, as in part
x(27r) = O and x'(27r) = -(ex + ~), which imply
x(t) = -(ex +~)
37r
-2 <
- t<
- 27r.
Cl
= O and
C2
(a). The initial conditions are
= -(ex +
sin t - 2_ sin4t
15
= -(ex +~)
sint - 2_(2sin2tcos2t)
15
15
= - (ex+ ~) sin t - ~ (2 sin t cos t cos 2t)
15
15
= sin t[- (ex + ~) - 2_ cos t cos 2t]
15
15
'
(d)
-2
-2
ex=l
ex=2
235
..
l~).
Thus
Chapter 5 Related Exercses

2. (a) We have x = el cos t
+ e2 sin t
and we let xp = At cos t
+ Bt
sin t.
Substituting into the
differential equation we get 2B cos t - 2A sin t = cos t. Thus A = O, B = -~ , and xp = ~t sin t.

The general solution of the differential equation is x(t)
= el cos t + e2 sin t + ~t sin t. The initial
conditions imply el = Oand e2 = O,so the solution of the initial-value problem is x( t) = ~t sin t.
(b) We have
Xc
= el cos t + e2 sin t and we let xp
A cos 2t + B sin 2t. Substituting into the
= cos 2t. Thus A = - ~, B = O, and xp =

-1 cos 2t. The general solution of the differential equation is x(t) = el cos t + e2 sin t - ~ cos 2t.
The initial conditions imply c = ~ and e2 = O, so the solution of the initial-value problem is
x(t) = ~cost - 1cos2t.
differential equation we get -3A cos 2t - 3B sin t
3. (a) In the first case the differential equations are x"
+X
= sin 4t when
X ::::: O and
x"
+ 4x
= sin 4t
< O. Since the initial velocity is positive, we solve first for x :::::o. The solution of
x" + x = sin 4t is x( t) = el cos t + C2 sin t - t\ sin 4t, as seen in Problem 1(a). The initial
conditions x(O) = Oand x'(O) = 1 imply el = Oand c2 = ~~
, so the solution of the initial-value
when x
problem is
19
x(t) = 15 sint - 15 sin4t

= ~:sin t
19
15
= - sin t
=
- 1~(2 sin 2t cos 2t)

2.
15
- - (2 sin t COS t cos 2t)
19
4
)
(sint) ( 15 - 15 costcos2t ,
0S;tS;7L
The first positive value of t for which x(t) is zero is t = 7r, at which time x'(7r) =
so x
< O immediately after t
x(7r) = O, x'(7r)
= tt. The new initial-value problem is then
-i~.The solution
of this initial-value problem is
x (t) = - ~sin 2t - 1~ sin 4t = (sin 2t) ( - ~ - ~ cos 2t) ,

The next value of t for which x(t)
= O is t
7r S; t S; 3; .
= 37r/2, at which time
x'(37r/2) = ~~ > O, so
> O immediately after t = 37r/2. The next initial-value problem is then x" + x
x(37r/2) = O, x' (37r/2) = ~~. The solution of this initial-value problem is
x
17
-n < O,
x" + 4x = sin 4t,
x(t) = 15 cost - 15 sin4t = (cost)
( 17
4
)
15 - 15 sintcos2t,
37r
2'"
S; t
S;
= sin4t,
57r
2""
= O is t = 57r/2, at which time x'(57r/2) = -~ < O, so

x < O irnmediately after t = 57r/2. The next initial-value problem is then x" + 4x = sin 4t,
x(57r/2) = O, x'(57r/2) = -~. The solution of this initial-value problem is
The next value of t for which x(t)
x(t) = _!_ sin 2t - _!_ sin 4t

15
12
(sin 2t) (_!_

15
236
- ~ cos 2t),
6
57r < t < b

2 - -
Chapter
= O. In this case,
where b is the value of t for which x(t)
5 Related Exercises
i cos 2t is O for t = 8.84514,
115 -
which is between 51i/2 and 31i. Thus, for the first two cycles,
(sin t)(
i~- 1~ cos t cos 2t),
(sin 2t) (-
x(t) =
O::;t::;1i
* - i cos 2t),
1i <
t
-
17
4
(cos t) (15
- 15
sin t cos 2t) ,
(sin 2t) (
< 37r2
<
t < 57r
- 2
37r
5; < t ::;8.84514.
fs - t cos 2t),
The first part of the next cycle is the solution of the initial-value problem x" + x = sin 4t,
x(8.84514) = O, x'(8.84514) = 0.28. This solution is approxirnately x(t) = -0.013 cost 0.109 sin t - 0.067 sin 4t and is positive on (8.84514,9.42478) with amplitude 0.042.
= sin 4t when x 2: O and x" + 4x =

< O. Since the initial velocity is positive we solve first for x 2: O. [The following
In the second case, the differential equations are x" + 64x

sin 4t when x
computations are done with the aid of a CAS having a differential equation solver and graphing
capability.] The solution of x"
+ 64x = sin4t,
1(.
x(O) = O, x'(O) = 1 is
11.)
x(t) = 48 sm4t+2'"sm8t
where
TI
= 0.415458 is found using the root-finding capability of the CASo The next initial-
value problem is x" + 4x = sin4t,
X(T1)
= O, X'(T1)
-0.909091. The solution is
x(t) = 0.403 cos 2t - 0.255 sin 2t - 0.083 sin 4t,

where
T2
2.1403. The next initial-value
TI::;
t::;
T2,
problem is x" + 64x = sin4t,
X(T2)
O, X'(T2)
= O, X'(T3)
l.16206. The solution is
x( t) = 0.153 cos 8t + 0.021 sin 4t - 0.008 sin 8t,

where
T3 =
T2::;
t ::; T3,
2.53479. The next initial-value problem is x" + 4x = sin4t,
X(T3)
-1. 28086. The solution is
x(t) = -0.737cos2t
where
T4
- 0.217sin2t - 0.083sin4t,
T3::;
t::;
T4,
= 4.0453.
In the third case, the differential equations are x" + 36x

sin 4t when x
of x" + 36x
< O.
= sin4t,
sin 4t when x 2: O and x" + 25x
Since the initial velocity is positive, we solve first for
x(O) = O, x'(O) = 1 is
x(t) = 2_ sin 4t +
20
2.
sin 6t,
15
237
O::; t
< TI,
x 2: O. The solution

where
TI = 0.571447.
Subsequent solutions are
x(t) = 0.1l5cos5t + 0.1l1sin4t + 0.093sin5t,

x(t) = -0.068 cos 6t + 0.05sin4t+
x(t)
0.082 sin 6t,
= 0.159 cos 5t + 0.111 sin 4t +
0.072 sin 6t,
Tl
T2
T3
S;
S;
S;
t S;
t S;
t S;
T2
T3
T4
= 1.24203
= 1.73555
= 2.21355.
-1
-1
b = 1,
b = 64,
a =4
b = 36,
a =4
a = 25
In each case it appears that x(t) has increasing amplitude as t increases.

(c) When a
= 0.5
the following graphs are obtained.

x
10
-1
-1
b = 1,
a =4
= 64,
a =4
= 36,
= 25
= 36,
= 25
When a = 3 the following graphs are obtained.
-1
-2
-3
1,
a =4
b = 64,
a =4
In both situations, it appears that the amplitude decreases for b = 1 and a = 4, but increases
in the other two cases.
238
Chapter
5 Related Exercses
4. (a) The solutions are obtained numcrically in CAS and plotted below in part (b). Since x'(O) =
1
> O, in each case the first differential equation used is x" + !3x' + 4x
= sin 4t, followed by
x" + !3x' + x = sin 4t, and alternating thereafter.

(b)
2
-2
x
1
-1
-2
!3 = 0.01
!3 = 0.5
!3 = 0.1
As t increases, the amplitude appears to increase for
!3 = 0.5.
239
!3
= 0.01 and decrease for
!3 = 0.1 and
Series Solutions of Linear Equations

Exercises 6.1
1.
u1m lan+ll
-= u1m 12n+lxn+l/(n+1)1
n---+ooan
n---+oo
2nxn [ti
u1m _-2n I1x = 211

X
n---+oo
n+ 1
The series is absolutely convergent for 21xl
< 1 or Ixl < 1/2. At x = -1/2, the series
f:
n=l
(_l)n
n
00 1
I:- is the
converges by the alternating series test. At x = 1/2, the series
harmonic series which
n=l n
diverges. Thus, the given series converges on [-1/2,1/2).
2.
u1m lan+ll
-n---+ooan
= l'1m !100n+l(x+7)n+l/(n+1)!!
n---+oo
100n(x + 7)n/n!
u1m _-100 Ix + 7 1= O
n---+oo
n + 1
The series is absolutely convergent on (-00,00).

ak 1
3. lim _____
k---+ooak
l
I=
1
(x - 5)k+ / lOk+1
1
1
lim
= lim -Ix - 51= -Ix
k---+oo (x-5)k/10k
k---+0010
10
The series is absolutely convergent for 110Ix series

00
I:(-l)k
k=l
4.
00 (_l)k (-10)k
2::
10k
k=l
51 < 1, Ix -
51 < 10, or on (-5,15).
00
2::
1 diverges by the k-th term test.
- 51
At x
= 15, the series
k=l
At x = -5, the
00 (-1)k10k
2::
10k
k=l
diverges by the k-th term test. Thus, the series converges on (-5,15).
lim jak+1j= lim I(k\~t(x-)~)k+ll=

k---+ooak
k---+oo
.x - 1
lim(k+1)lx-11=00,
k---+oo
x=l-l
The radius of convergence is O and the series converges only for x = l.

5. sinxcosx
= (x - ~3 + 1x250_ 5~:0 + ... ) (1-
6. e-x cos x = ( 1 - x +
7.
1
cosx
x2
x3
'2
- "6
+
1
2
1-~+ir4
,
4
.
--fu-+'"
6.
x4
24 - .,.
x; + ~: _
1-
x2
'2
+
;6
x4)
24 - '"
x2
5x4
61x6
= 1 + -2 + -41 + -61 + ...
..
240
0 + ... ) = x- 2~3 + 21~5_ ~~~+ ...
= 1- x +
x3
x4
3
- "6
+
...
Exercises
Since cos( 7r/2) = cos(
1-x
1 3
= - - -x
2+x
2 4
+ -x3
8. --
/2)
T;
- -x
16
= 0, the
3
00
2ncnxn-1
(-7r
7r/2).
-2, the series converges on (-2,2).
00
00
00
+ 6cnxn+l = 21 CIXO + 2nenxn-l
n=l
/2,
+ ...
Since the function is undefined at x

9.
series converges on
6. 1
n=O
+ 6enxn+l
n=2
._,__..._.,
k=n-l
00
n=O
~
k=n+l
00
= 2Cl+ 2(k + 1)Ck+lXk + 6Ck_lXk

k=l
k=l
00
= 2Cl+ [2(k + 1)Ck+l + 6Ck_l]Xk

k=l
00
10.
n(n
00
- 1)enxn
+ 2 n(n
n=2
00
- 1)enxn-2
+ 3 nenxn
n=2
n=l
00
00
00
= 22 1C2XO+ 2 . 3 . 2C3X' + 31 CIX' + n(n - 1)enxn +2 n(n - 1)cnxn-2 +3 nenxn

n=2
n=4
k=n
00
= 4C2 + (12c3
n=2
k=n
'----v--'
k=n-2
00
00
+ (12c3 + 3Cl)X + k(k - 1)CkXk + 2 (k + 2)(k + 1)Ck+2Xk + 3 kCkXk

n=2
n=2
n=2
00
= 4C2 + (3Cl+ 12c3)X + ([k(k - 1)+ 3k]Ck + 2(k + 2)(k + 1)Ck+2)Xk

n=2
00
= 4C2 + (3Cl+ 12c3)X + [k(k + 2)Ck + 2(k + 1)(k + 2)Ck+21xk

n=2
00
11. y' =
00
(_l)n+lxn-l,
n=l
y" =
(-1t+1(n
- 1)xn-2
n=2
00
(x
+ 1)Y" + y'
= (x
00
+ 1) (_l)n+l(n - 1)xn-2 + (_l)n+lxn-l

~l
n~
00
00
00
(_l)n+l(n- 1)xn-1 + (-1t+1(n

n=2
n=2
00
= -xo+xo+
- 1)xn-2
+ (-1t+1xn-l
n=l
00
(_l)n+l(n_1)xn-l+
n=2
k=n-l
(_l)n+l(n_1)xn-2+
n=3
k=n-2
241
00
(_l)n+lxn-l
n=2
'----v---'
k=n-l
Exercises
6. 1
00
00
00
+ (_1)k+3(k
= (_1)k+2kxk
k=l
+ 1)xk + (_1)k+2xk
k=l
k=l
00
[(_1)k+2k
_ (_1)k+2k
+ (_1)k+2]
_ (_1)k+2
xk = O
k=l
,_ ~
12. y - ~
n=l
'"
xy
(-1)n2n
2n-l
22n(n!)2 x
,
" _ ~
y - ~
n=l
(-1)n2n(2n
- 1) 2n-2
22n(n!)2
X
(-1)n2n(2n
- 1) 2n-l
~
22n(n!)2
X
+ ~l
_ ~
+ y + xy - ~l
(-l)n2n
2n-l
~
22n(n!)2 x
+~
k=n
13. Substituting y
y" - xy
k=n
(_1)n
2n+l
22n(n!)2 x
k=n+l
L~o cnxn into the differential equation we have

00
n(n
00
- 1)Cnxn-2 -
n=2
Cnxn+l
00
n=O
k=n+l
00
+ 1)Ck+2Xk
k=O
'-.r---'
k=n-2
(k + 2)(k
00
= 2C2 + [(k
+ 2)(k + 1)Ck+2 - Ck_tlxk = O.
k=l
Thus
C2= O
(k
+ 2)(k + 1)Ck+2 -
Ck+2 = (k
+ 2)(k + 1) Ck-l,
Ck-l = O
and
1
Choosing co = 1 and q = Owe find

1
6
C3
=-
C4
= Cs = O
C6 -
1
180
242
k = 1,2,3, ....
Ck_lXk
k= 1
Exercises
and so on. For
= O and
Co
Cl
= 1 we obtain
C3
=O
C5
= C6 = O
C7
= 504
and so on. Thus, two solutions are

1 3
1
= 1 + -x
+ -x
Yl
14. Subst.ituting y =
180
+ ...
+ x2y
Y2 =
1 4
1 7
+ -x
+ -x
+ ....
12
504
I.:~=O Cnxn into the differential equation we have

00
y"
and
.L n(n
00
_ l)cnxn-2
n=2
+ .L cnxn+2
00
n=O
k=n+2
00
+ 2)(k + l)ck+2Xk + .L Ck_2Xk
k=O
'-v----'
k=n-2
.L (k
k=2
00
= 2C2 + 6C3X +
.L [(k + 2)(k
+ I)Ck+2 + Ck_2]Xk
= O.
k=2
Thus
C2
(k
and
C3
=O
+ 2)(k + I)Ck+2 + Ck-2

1
Ck+2 = _ (k
Choosing
Co
+ 2)(k + 1) Ck-2,
= 1 and CI = Owe find

C4
= _-
12
CS=C6=C7=0
1
C8 = 672
and so on. For
Co
= O and
Cl
= 1 we obt ain
C4 =
Cs
= _-
20
C6=C7=C8=0
cg
1
= 1440
243
=O
k = 2,3,4, ....
6. 1
Exercises
6. 1

Yl
15. Substituting y
1 4
1
= 1 - -x
+ -x
12
672
- ...
Y2 = x - -x
.
20
and
1 9
+ --x
1440
= .L~=o cnxn into the differential equation we have

00
y" - 2xy'
+y
n(n
00
- 1)cnxn-2
- 2
n=2
ncnxn
n=l
~
k=n-2
k=n
00
(k
k=O
+ 2)(k + 1)Ck+2Xk -
00
+ cnxn
n=O
k=n
'---v-"
00
00
k=l
k=O
kCkXk + Ckxk
00
= 2C2 + Co +
[(k + 2)(k
+ 1)Ck+2 -
(2k - l)Ck]Xk = O.
k=l
Thus
2C2 + Co = O
(k
+ 2)(k + 1)Ck+2 - (2k - l)Ck = O
and
e
= --Co2
2k - 1
Ck+2 = (k
Choosing
Co
= 1 and
CI
+ 2)(k + 1) ci,
k=1,2,3,
= Owe find
C2
1
=-2
C3
= Cs = C7 = ... = O
1
C4 =--
8
7
C6=-and so on. For
Co
= O and
Cl
336
= 1 we obtain
C2
= C4 = C6 = ... = O
1
C3 =-
6
1
C5
= 24
C7
1
= 112
244
....
Exercises
Yl
= 1 - -x1
1
- -x
8
--x
7
336
and
16. Substituting y = ~=o cnxn into the differential equation we have

y" - xy'
+ 2y =
n(n
ncnxn
- 1)cnxn-2 -
n=2
+ 2 cnxn
n=l
k=n
00
kCkXk + 2 Ckxk
+ 2)(k + 1)Ck+2Xk -
= (k
k=O
n=O
k=n
'--v--'
'----v-----'
k=n-2
k=O
k=l
00
= 2C2 + 2co + [(k
+ 2)(k + 1)Ck+2 -
(k - 2)qJxk
k=l
Thus
2C2 + 2co = O
(k
+ 2)(k + 1)Ck+2 -
(k - 2)Ck = O
and
C2 =
-Co
k-2
q+2 = (k
Choosing
Co
1 and
C}
+ 2)(k + 1) q,
k=1,2,3,
....
= O we find
C2 =-1
Por
Co
= O and
Cl
C3
C4
=O
C6
= Cs = ClO = ... = O.
Cs
= C7
= ...
=O
= 1 we obtain
C2
C4
= C6 = ... = O
1
6
1
C5=-120
C3 =--

Yl = 1 - x2
and
1
Y2 = x - -x
6
245
--x
1
120
= O.
6. 1
Exercises 6. 1
y = ;~=o Cnxn into the differential equation we have
17. Substituting
00
y"
00
L n(n
+ x2y' + xy =
n=2
+L
ncnxn+l
n=l
"-----v---'
k=n-2
k=n+l
00
00
+L
- 1)cnxn-2
Cnxn+l
n=O
~
k=n+l
00
L (k + 2)(k
+ 1)Ck+2Xk + L (k
k=O
00
- 1)Ck_1Xk
k=2
+L
00
= 2C2 + (6C3 + Co)x
+ L [(k + 2)(k + 1)Ck+2 + kCk_l]xk = O.

k=2
Thus
C2
= 06C3 + Co
=O
+ 2)(k + 1)Ck+2 + kCk-l
(k
=O
and
1
C3
= --co
6
Ck+2 = - (k
Choosing
Co
+ 2)(k + 1) Ck-l,
k = 2,3,4, ....
= 1 and q = Owe find
and so on. Por
Co
= O and
Cl
C3
1
=--
C4
= C5 = O
C6
=-
c3
=O
C4
1
=--
C5
= C6 = O
1
45
= 1 we obtain
C7=-
252

1 3
Yl = 1- -x
6
1 6
+-x
45
Ck_1Xk
k=l
and
246
1
Y2 = x - -x
6
5
232
+-x
Exercises 6. 1
18. Substituting y = 2:;;":=0 cnxn into the differential equation we have
= n(n
y" + 2xy' + 2y =
- 1)Cnxll-2 + 2
= nc.ix"
n=l
n=2
.____,
k=n-2
k=n
oo
+2
00
= Cnxn
n=O
k=n
'--v-'"
oc
(k + 2)(k + 1)Ck+2Xk + 2
kCkXk + 2
Ckxk
k=O
k=l
k=O
oo
= 2C2 + 2co
+ [(k
+ 2)(k + 1)q+2 + 2(k + l)qlxk
k=l
Thus
2C2 + 2co
=O
(k + 2)(k + 1)Ck+2 + 2(k + l)ck = O

and
C2 =
Ck+2 =
Choosing
co
= 1 and
C1
-Co
-
+ 2 Ck,
k = 1,2,3, ....
= Owe find
C2 =-1
C3
= Cs = C7 = ... = O
C4
C6
1
=--
and so on. For
Co
= O and
C1
"2
= 1 we obtain
= C6 = ... = O
C2
C3
=--
C4
2
3
4
e5 =-
15
8
c-- -_
c
105
r:

Y1 = 1 - x 21+41-x6
- -x
+ ...
247
= O.
Exercises 6.1
19. Substituting y
= ~='o Cnxn into the differential equation we have

00
(x - 1)y"
+ y'
00
= L
n(n - 1)cnxn-l
n=,2
k=n-l
00
- L n(n - 1)cnxn-2
n=2
k=n-2
00
+L
ncnxn-l
n=l
k=n-l
00
= L (k
+ 1)kck+lXk
00
+ 2)(k + 1)q+2Xk + L
- L (k
k=O
k=l
(k
k=O
+ 1)Ck+lXk
00
= -2C2
+ Cl + L[(k + 1)kck+l
+ 2)(k + 1)Ck+2 + (k + 1)Ck+lJXk
- (k
= O.
k=l
Thus
+ Cl
-2C2
(k
+ 1)2Ck+l
=O
+ 2)(k + 1)Ck+2 = O
- (k
and
k+1
Ck+2=k+2ck+l'
Choosing
Co
= 1 and
Cl
= Owe find
C2
k=1,2,3,
= C3 = C4 = ... = O. For Co = O and

1
C2
....
= 2'
C3
1 we obtain
3'
C4
= 4'

and
Yl = 1
20. Substituting y
Y2
1 2 1 3 1 4
= x + -x
+ -x + -x + ....
2
= ~=o cnxn into the differential equation we have

00
(x
+ 2)y" + xy'
00
- y = L n(n - 1)cnxn-l
n=2
k=n-l
00
= L(k
+L
00
2n(n - 1)cnxn-2
n=2
k=n-2
00
+L
ncnxn - L Cnxn
n=l
n=O
~~
k=n
k=n
00
00
+ l)kck+lXk + L
2(k
k=O
k=l
+ 2)(k + 1)Ck+2Xk +
00
kCkXk - L Ckxk
k=l
k=O
00
= 4C2 -
Co
+L
[(k
+ 1)kck+l + 2(k + 2)(k + 1)Ck+2 + (k
- 1)cklxk = O.
k=l
Thus
4C2 (k
Co
=O
+ 1)kCk+l + 2(k + 2)(k + 1)Ck+2 + (k

248
- 1)Ck = O,
1, 2, 3, ...
Exercises
6. 1
and
C2 =
-ca
4
(k
+ l)kck+l + (k
ck+2=Choosing
Ca
1 and
Cl
and so on. For
ca
Cl
- l)Ck
2(k+2)(k+1)
k = 1, 2, 3, ....
= Owe find
= O,
= O and
C3
Cl
=-
c-0--480
= O,
C4
24'
1 we obtain
C2 = O
C3
=O
C4
= C5
= C6 = ... = O.
Tbus, two solutions are

and
Y2 = ClX.
21. Substituting y = L~=a cnxn into the differential equation we have

00
y" - (x
+ l)y'
- y
L n(n
00
- 1)cnxn-2 -
n=2
n=l
k=n
00
L (k + 2)(k
'---v----"
k=n-2
=
00
L ncnxn
L ncnxn-l
n=l
'-----v--'
k=n-l
00
+ 1)Ck+2Xk
k=a
00
L Cnxn
n=a
k=n
00
L kCkXk ...,...L (k + l)Ck+lXk
k= 1
k=a
00
L Ckxk
k=a
00
= 2C2 -
Cl -
Ca
+ L [(k + 2)(k + 1)Ck+2 -
(k
+ l)Ck+l
k=l
Thus
(k
+ 2)(k + 1)Ck+2 -
and
Cl
C2= -2Ck+2 =
Choosing
ca =
1 and
Cl
(k - l)(Ck+l
+ Ck) = O
+ ca
Ck+l + Ck
k + 2 Ck,
= 2,3,4,
....
= O we find
C2 =
2'
C3
6'
249
C4
=6
- (k
+ l)Ck]Xk = O.
Exercses 6. 1
and so on-, Por
Co
= O and c = 1 we obtain
C3
2'
C4
1
=4

1
u: = 1 + 2x + 6x + 6x + ...
and
Y2
= x + 2x + 2x + 4x + ....
22. Substituting y = :L~=o cnxn into the differential equation we have

(x2
+ 1) y"
00
- 6y
00
00
= L n(n - l)enxn + L n(n - 1)cnxn-2

n=2
- 6 L
n=2
k=n-2
k=n
00
cnxn
n=O
'--v---'
k=n
00
00
= L k(k - l)CkXk + L(k

k=2
+ 2)(k + 1)Ck+2Xk -
k=O
00
= 2C2 - 6co + (6C3 - 6Cl)X + L [(k2 - k -
6L
Ckxk
k=O
6) Ck + (k + 2)(k + 1)Ck+2] xk
k=2
Thus
2C2 - 6co = O
(k - 3)(k
+ 2)Ck + (k + 2)(k + 1)Ck+2 = O
and
C2
= 3co
C3
= Cl
k-3
Ck+2 = - k + 1 Ck,
Choosing
Co
= 1 and
Cl
k = 2,3,4, ....
= O we find
C2 = 3
C3
= Cs = C7 = ... = O
C4
=1
C6
1
=--
250
= O.
Exercises 6. 1
and so on. For
Co
= O and
ci
1 we obtain
C4
C3 =
C7
C2
C5
= C6 = ...
=O
= O.
Cg
= ...
Thus, two solutions are

Yl = 1 + 3x
+x4 -
1 6
+ ...
-x
5
and
Y2 = X
+ X 3.
23. Substituting y = L~=o Cnxn into the differential equation we have

00
(x2
+ 2)yll + 3xy'
00
2:
- y =
00
+ 22:
n(n - l)cnxn
n=2
n=2
k=n
00
2:
00
+ 22:
(k
k=O
k=2
00
nCnxn -
n= 1
~
k=n
k=n-2
00
k(k -l)CkXk
+ 32:
n(n - 1)Cnxn-2
+ 2)(k + 1)Ck+2xk + 32:
(4C2 - co)
2:
Ckxk
k=O
k=l
+ (12c3 + 2Cl)X + 2:
+ 2)(k + 1)ck+2 + (k2 + 2k
[2(k
k=2
Thus
4C2 12c3
2(k
Co =
+ 2c
=O
+ 2)(k + 1)Ck+2 + (k2 + 2k
1)
Ck =
and
1
C2
= -Co
C3
Ck+2 =
Choosing
Co
1 and
Cl
1
--Cl
6
k2
+ 2k
C2
1
=4
C3
- 1
- 2(k + 2)(k + 1) Ck,
k =
= Owe find
C5
C4 =--
C7
= ...
96
251
=O
cnxn
n=O
'-v----'
k=n
00
kCkXk -
00
2:
2,3,4,....
1) Ck] xk = O.
Exercises
6. 1
and so on. Por
Co
= O and
CI =
1 we obtain
C2=C4=C6==0
1
C3 =--
6
7
Cs
120

YI
= 1 + 4: x - 96 x + ...
1
Y2 = x - -x
6
and
7
+-x
120
24. Substituting y = ~=a cnxn into the differential equation we have

(x2 -
1) y" + xy'
00
- y =
n(n
00
n(n
- l)cnxn -
n=2
+ ncnxn
n=2
n=l
k=n
k(k
~~
k=n-2
CXJ
00
- 1)cnxn-2
00
-l)CkXk
cnxn
n=O
k=n
00
+ 2)(k + I)Ck+2Xk + kCkXk
- (k
k=O
k=2
k=n
00
k=l
00
Ckxk
k=O
00
[-(k
= (-2C2 - ca) - 6C3X +
+ 2)(k + 1)Ck+2 + (k2 -1)
k=2
Thus
-2C2 -
-(k
Co = O
+ 2)(k + 1)Ck+2 + (k
and
ca
= 1 and
CI = Owe
+ I)Ck
C2
= --Ca
C3
=O
k-l
Ck+2 = k + 2 Ck,
Choosing
- l)(k
= 2,3,4,
find
1
C2 =--
C3
= Cs =
C7
= ... = O
C4 =--
252
....
=O
Ck] xk
= O.
Exercses
and so on. For co
= Oand C} = 1 we obtain
C2
= C4
C6
= ... = O
C3
= Cs
C7
= ...
= O.
Thus, two solutions are
Yl
25. Substituting y =
1 2
1 4
1 - -x --x
and
Y2
= x.
I:~=o c,.,xn into the differential equation we have

00
(x - l)y" - xy'
00
+ y = L n(n - 1)cnxn-1
n=2
k=n-l
00
00
00
L(k+1)kck+1Xkk=l
-2C2
00
L n(n - 1)c,.,xn-2 - L ncnXn

n=2
n= 1
'-v--'
k=n- 2
k=n
00
L(k+2)(k+1)Ck+2Xkk=O
+ L cnxn
n=O
k=n
'---v-"'
00
LkckXk+
k=l
LCkXk
k=O
00
+ Co + L[-(k
+ 2)(k + 1)Ck+2 + (k + l)kck+l
- (k - l)Ck]Xk
k=l
Thus
+ Co =
-2C2
-(k
+ 2)(k + 1)Ck+2 + (k
- l)kck+l - (k - l)Ck = O
and
C2
1
= "2co
kCk+1
(k - l)Ck
ck+2 = k+2 - (k+2)(k+
1)'
Choosing
Co =
1 and
C1
Co
= 1,2,3, ....
= Owe find
C2
and so on. For
= O and
Cl
="2'
C4 = O
6'
C2 = C3 = C4 = ... = O. Thus,
C3
= 1 we obtain
Y = el ( 1 +"2x
1 3
+ 6x
+... ) + e2X
and
y' =
y =
el =
el
(x
+ ~x2 + ...) + C2.
-2 and Cz = 6, so
-2 (1 + ~x2 + ~x3 + ...) + 6x = 8x 253
2ex
= O.
6. 1
Exercises 6. 1
26. Substituting y = L~=O cnxn into the differential equation we have
(x+1)y"
- (2 - x)y'
+y
00
L n(n
00
n=2
00
L n(n
- l)cnxn-l
- 1)cnxn-2
n=2
k=n-2
00
00
L(k + l)kck+lxk
k= 1
n=l
k=n-l
00
L ncnxn-l
- 2
L ncnxn
n=O
k=n
n=l
'-..---'
k=n-l
'--....---'
k=n
00
L(k + 2)(k
+ 1)Ck+2Xk - 2 L(k + l)Ck+lXk +
k=O
00
k=O
L cnxn
00
00
k= 1
k=O
L kCkXk + L Ckxk
00
= 2C2 -
+ Co +
2Cl
L[(k + 2)(k
+ 1)ck+2 - (k + l)Ck+l + (k + l)Ck]Xk = O.
k=l
Thus
2C2 - 2Cl
(k
+ Co = O
+ 2)(k + 1)Ck+2 - (k + l)Ck+l + (k + l)Ck
=O
and
C2 =
2'co
ci -
Ck+2 = k
+ 2 Ck+l
+ 2 Ck,
= 1,2,3, ....
Choosing eo = 1 and el = O we find

C2 =
and so on. For eo = O and
Cl =
-2"
C4
1
12
=-
1 we obtain
C2 =
1,
C3
= O,
e4
1
=-4
and so on. Thus,

y
el
1213-14-x
( 1 - -x
2
6
y'
el
(-x
+ -x
+...
12
+ e2 ( x + x 21-4 -x
+ ... )
and

y =
- ~x2
2
el
+ ~x3 + ... ) + e2
3
= 2 and
2 (1 - ~x2
2
')
~x3
6
1
3
= 2 - x - 2x- - -x
e2 =
(1 + 2x
- x3
+ ... ) .
-1, so
+ "'!_x4 + ... ) - (x + x2 _ ~x4 + ... )

12
+ -x5
12
+ ....
254
Exercses
27. Substituting y
= 2:;::='=0 Cnxn into the differential equation we have

00
+ 8y
y" - 2xy'
00
=L
n(n - 1)Cnxn-2
- 2L
n=2
00
nCnxn
+8 L
cnxn
n=O
~
k=n
n=l
'-----'
k=n-2
k=n
00
00
= L (k
+ 2)(k + 1)Ck+2Xk
- 2L
k=D
00
+8L
kCkXk
Ckxk
k=O
k=l
00
= 2C2 + 8co
+ L[(k + 2)(k + 1)Ck+2 + (8 -
2k)Ck]Xk = O.
k=l
Thus
2C2 + 8co
(k
=O
+ 2)(k + 1)Ck+2 + (8
- 2k)Ck = O
and
C2 =
-4co
2k - 8
Ck+2 = (k
Choosing
Co
= 1 and
Cl
+ 2)(k + 1) Ck,
k = 1,2,3, ....
= Owe find
C2 =-4
Por
Co
= O and
Cl
C3
C4
4
=-
C6
= ca =
C5
C7
= ... =
3
ClO
= ... = O.
= 1 we obtain
C5 =-
10
and so on. Thus,

Y = el
( 1-
and
Y , = el ( -8x
4 x 4)
4x 2 + '3
+ e2 ( x
16 x 3) + e2 ( 1 +3
255
1
- x 3 + 10x
1
3x +:2x
+ ... )
4 +... ) .
6. 1
Exercises 6. 1
The initial conditions imply Cl = 3 and C2 = O, so
y
28. Substituting
= 3 (1-4x2
+ ~x4) =
+4x4.
L~=a cnxn into the differential equation we have

00
(x2
3 -12x2
+ l)y" + 2xy' =
L n(n
00
- l)cnxn
n=2
+L
00
n(n - 1)cnxn-2
n=2
k=n
L k(k
2ncnxn
n=l
'--v--'
k=n-2
00
+L
k=n
00
00
k=a
k=l
+ L (k + 2)(k + 1)Ck+2Xk + L
- l)ckxk
k=2
2kCkXk
00
= 2C2 + (6C3 + 2Cl)X
+ L [k(k + l)Ck + (k + 2)(k + 1)Ck+21xk = O.

k=2
Thus
k(k
+ l)Ck + (k + 2)(k + 1)Ck+2
=O
and
Ck+2 =
Choosing ca = 1 and
Cl
= Owe find
C3
- k + 2 Ck, k = 2, 3, 4, ....
= C4 = Cs
C3
= ... = O. Por ca = O and
1
=-_
1
Cs =--;:
C7
=_
7
and so on. Thus
and
Y
, = Cl (1 -
x2 + x4 - x6
256
+... ) .
Cl
= 1 we obtain
Exercises
imply ca = O and
Cl
71x 7 + ....
y = :Z:::~=acnxn into the differential equation we have
29. Substituting
Y 11
= 1, so
31x 3 +"51x
Y =x -
+ (sism x ) y
= ~
L..- n ( n - 1) Cnx n-2
n=2
~ [2C2 + 6C3X
= 2C2 + (6C3
+ (x
1
- -x
6
3 + -x1
s _...
) ( ca + ClX + C2X2)+ ...
120
+ 12c4x2 + 20csx3 + ...] + [cax + ClX2 + (C2
+ ca)x + (12c4 + Cl)X 2 + ( 20cs + C2 -
1 ) x3
"6ca
- ~ca) x3
+ ... = O.
Thus
6C3 + ca = O
+ ci
12c4
20cs
and
=O
1
-Ca = O
6
+ C2 -
C2 = O
C3
1
= --Ca
C4
1
= -12Cl
1
Cs = - 20C2
Choosing
ca
1 and
Cl
= O we
1
120ca.
find
C4
and so on. For ca
= O and
Cl
C2 =
and so
011.
6. 1
Thus, two solutions
Yl
1 3
1 - -x
6
1
Cs = 120
= O,
1 we obtain
O,
C3
= O,
+ ...
1
C4
= -12'
Cs = O
are
1
+ -x
120
and
257
Y2 =
X -
-x + ....
12
+ ...]
Exercises 6. 1
30. Substituting y = I:;:;O=o
cnxn into the differential equation we have
00
y"
+ eXy'
- y =
I:n(n
- 1)Cnxn-2
n=2
(1 + x + ~x2
2
= [2C2 + 6C3X
+ ~x3 + ...)
(Cl
+ 2C2X + 3C3X2 + 4C4X3 + ...) -
+ 12c4x2 + 20C5X3 + ...]
+ [Cl + (2C2 + cl)x + (3C3 + 2C2 + ~Cl)

= (2C2
+ ci
- ce)
x2
+ ...] - [co + ClX + C2x2 + ...]
+ (6C3 + 2C2)X + (12C4 + 3C3 + C2 + ~Cl)
x2
+ ... = O.
Thus
2C2 + Cl
Co
=O
6C3 + 2C2 = O
1
12c4 + 3C3 + C2 + 2Cl = O
and
C2 =
-Co -
-Cl
1
--C2
3
1
C4 = --C3
4
C3 =
Choosing
Co =
1 and
Cl
+ -C2
12
1
24
-Cl
= Owe find
1
C2 =
and so on. For
f Cnxn
n=O
Co = O and Cl =
2'
C4 = O
1 we obtain
1
C3
= 6'
1
24
C4 =--

1 2 1 3
Yl = 1 + -x - -x + ...
2
6
and
1
Y2 = x - -x
2
258
+ -1x 3
6
1 4
- - x
24
+ ....
Exercses
31. Substituting y
= :;:o=o Cnxn into the first differential equation leads to

00
y" - xy =
00
L n(n
- 1)Cnxn-2 -
n=2
L Cnxn+l
n=O
~
k=n+l
k=n-2
00
L (k + 2)(k
00
+ 1)Ck+2Xk
k=O
L Ck_lXk
k=l
00
= 2C2 +
L [(k + 2)(k
+ 1)Ck+2 - Ck_l]Xk =
l.
k=l
Thus
(k
+ 2)(k + 1)Ck+2 -
Ck-l
=O
and
1
C2 =-
k = 1,2,3, ....
Ck+2= (k+2)(k+1)'
Let Co and Cl be arbitrary and iterate to find

1
C2 =
2'
C3 = '6CO
1
C4 = 12cl
1
Cs = 20C2 = 40
and so on. The solution is
Y
1213141
= Co + ClX + 2'X + '6COX + 12 ClX +

13) +...
= Co ( 1 + '6x
Substituting y =
40 Cs +
14) +...
+ Cl ( X + 12x
...
ls + ....
+ '212x + 40x
:;:O=o cnxn into the second differential equation leads to

00
y" - 4xy' - 4y =
L n(n
00
- 1)Cnxn-2 -
n=2
L 4ncnxn
n=l
...._,..._.
k=n-2
259
k=n
00
L 4Cnxn
n=O
k=n
'-----v-'
6. 1
Exercses
6. 1
00
00
+ 2)(k + 1)Ck+2Xk
= (k
00
4kckXk
k=O
4CkXk
k=l
k=O
00
= 2C2 - 4co
+ [(k
+ 2)(k + 1)Ck+2
+ l)cklxk
- 4(k
k=l
=e
= 1 + L..... -k'x
k=l
k
o
Thus
2C2 - 4co = 1
(k
+ 2)(k + 1)Ck+2 -
4(k
+ l)Ck =
1
k!
and
C2 =
"2 + 2co
1
Ck+2 = (k
+ 2)! + k + 2 ck,
= 1, 2, 3,
Let Co and ct be arbitrary and iterate to find

1
C2 =
"2 + 2co
1
3!
4
3
1 4
+ -ct
3! 3
C3
= - + -ct
C4
=-
C5
C6
= 6! + "6C4 =
C7
=-
1 4
1
1
13
+
C2 = - + - + 2co = 4! 4
4! 2
4!
1
51 + 5"C3
1
= 5! + 5 3!
o
16
+ 15c1
4 13 8
4! + "6co
51 + 6
1
+ 2co
4 17
o
64
17
16
261
51 + 15C
51 + 3co
409
64
7! + "7C5 = 7! + 7.5! + 105Cl = T! + 105c1
and so on. The solution is
260
Exercises
Y = Ca + CIX + ( -1 + 2ca ) x 2
2
(261
6!
+ ~CO)X6 +
3
4
= Ca [ 1 + 2x 2 + 2x 4 + -x
(409
7!
6
+ -x21 2 + -x3!1
32. We identify P(x)
x4/5! -"',
= Oand
for [z]
+ -x13
4!
Q(x)
1 + -CI
4)
+ ( "1
3. 3
X 3 - (13
,+
4.
+ ~CI)Xi
+ (17
,+
5.
+ ...] + CI [ X + -x4 3 + -x16 5 + -x64

5
261
+ -x17
+
-x
5!
6!
= sin x [.
16)
-CI
15
X5
+ ".
105
2ca ) x 1
6. 1
15
+ -x409
7!
105
+".
+" ..
The Taylor series representation for sin x/x is 1- x2 /3! +
< oo. Thus, Q(x) is analytic at x
O and x = O is an ordinary point o the
differential equation.
33. The differential equation xy" = O has a singular point at x = O. It also has two solutions, YI = 1
and Y2
34.
If x
= z, that
are analytic at x
> O and y > O, then y"
= O.
< O and the graph o a solution curve is concave down. Thus,
= -xy
whatever portion o a solution curve lies in the first quadrant is concave clown. When x
y < O, y"
= -xy
> O, so whatever portion o a solution curve lies in the fourth quadrant is concave
up.
35. (a) Substituting y = L~=a Cnxn into the differential equation we have
00
y"
> O and
+ xy' + y
L n(n
00
- 1)cnxn-2
n=2
n=l
k=n-2
00
00
+ L ncnxn + L Cnxn
L (k + 2)(k
k=a
n=a
k=n
'---v-'
k=n
00
00
k=l
k=
+ 1)Ck+2Xk + L kCkXk + L Ckxk

00
= (2C2 + ca) + L [(k + 2)(k + 1)Ck+2 +" (k + l)Cklxk = O.

k=l
Thus
2C2 + Ca = O
(k
+ 2)(k + 1)Ck+2 + (k + l)Ck
and
1
C2 = --Ca
2
Choosing Ca = 1 and CI = Owe find
261
=O
Exercises 6. 1
C2 =--
C3
and so on. For
Co
= Oand
Cl
= Cs = C7 = ... = O
= 1 we obtain
C2=C4=C6==O
C3
= --
1
2
=-3
3!
es = -~( -~) = 5~3 =

C7
_~(\!2)
=_
\!2
6.;!. 2

~
= k=O
L......t
Yl
(_l)k
2k
2k . k' x
and
4
2
4
2
N=2
4
2 N=10
N=6
4 x
-4
-4
-4
-4
262
Exercises
(e)
yl
y2
4/
2'
6. 1
-2
-2
-4
-4
The graphs of Yl and Y2 obtained from a numerical solver are shown. We see that the partial
sum representations indicate the even and odd natures of the solution, but don't really give a
very accurate representation of the true solution. Increasing N to about 20 gives a much more
accurate representation on [-4,4].
(d) From e" = 2:.'{'=oxk/k!
we see that e-x2/2 = 2:.~o(-x2/2)k/k!
= Lk=O(-1)kx2k/2kk!.
Frorn
(5) of Section 4.2 we have
= (
~ (_l)k
6
2kk!
2k) (~
x
6 (2k + 11)2kk!
2k+l)
x
36. (a) We have

y"
+ (cos x)y
= 2C2 + 6C3X
+ 12qx2 + 20csx3 + 30C6X4 + 42c7XS

2
+ (1 - .:._ +.:._ - .:._+ ... )(co + CJx + C2x2 + C3x3 + C4x4 + csxS + ... )
2!
(2C2
4!
6!
+ ce) + (6C3 + Cl)X + (12q + C2 - ~co )x2 + (20cs + C3 - ~CJ)x3
+ (30C6 + C4 + ~co
24
- ~C2)x4
2
263
+ (42c7 + Cs + ~Cl - ~C3)XS + ....

24
Exercses 6. 1
Then
and
which gives
C6
= -co/80 and
C7
= -19cr!5040.
Thus
Yl(X) = 1 -
2X + 12X
1
- 80X
Y2(X) = X -
6X + 30X
19
- 5040X
+ ...
and
7
+ ....
(b) From part (a) the general solution of the differential equation is 'Y =
y(O) = Cl + C2' 0=
Y = Yl
(e)
and y/(O) =
Cl
+ Y2
= 1+ x
-"2x
Cl .
2
CIYl
6x + 12x + 30x
1
- 80x
19
- 5040x
+"'.
4
2
x
-2
-4
(d)
Then
O+ C2 = C2, so the solution of the initial-value problem is
+ C2Y2.
-6
4 -2
-2
-2
-4
-4
264
Exercises 6.2
Exercises 6.2
1. Irregular singular point: x = O
2. Regular singular points: x = O, -3
= 3; regular
3. Irregular singular point: x
singular point: x
-3
4. Irregular singular point: x = 1; regular singular point: x = O

5. Regular singular points: x = O, 2i
6. Irregular singular point: x = 5; regular singular point: x = O
7. Regular singular points: x = -3, 2
8. Regular singular points: x = O, i
= O;regular singular points: x = 2, 5
= -1;
regular singular points: x
= 0,3
we see that Xo
= 1 and
Xo
, +--y=O
x
y " +--y 5
x-1
x+1
-1 are regular singular points. For Xo
= 1 the differential equation
can be put in the form

(x - 1)2y"
+ 5(x
- l)Y'
In this case p(x) = 5 and q(x) = x(x - 1)2/(x
x(x - 1)2
Y = O.
x+1
1). For Xo = -1 the differential equation can be
put in the form

(x
+ 1)2y" + 5(x + 1) x + 1 y' + x(x + l)y = O.
x-1
In this case p(x) = (x + l)/(x - 1) and q(x) = x(x
+ 1).
12. Writing the differep.tial equation in the form

x+3
+ --
y' + 7xy = O
x
we see that Xo = O is a regular singular point. Multiplying by x2, the differential equation can be
y"
put in the form

x2y"
+ x(x + 3)y' + 7x3y
= O.
= x + 3 and q(x) = 7x3.

We identify P(x) = 5/3x + 1 and Q(x) = -1/3x2, so that p(x) = xP(x)
x2Q(x) = -j. Then ao = ~, be = -j ,and the indicial equation is
We identify p(x)
13.
r(r - 1)
+ ~r
3
- ~ = r2 + ~r - ~ = ~(3r2
3
3
3 3
265
+ 2r - 1) = ~(3r
3
- l)(r
~+ x
and q(x) =
+ 1) = O.
Exercises 6.2
The indicial roots are
k and
-1. Since these do not differ by an integer we expect to find two series
solutions using the method of Frobenius,
= l/x and Q(x) = lO/x, so that p(x) = xP(x)
= 1 and q(x) = x2Q(x)
= 10x.
Then ao = 1, bo = O, and the indicial equation is
+r
r(r - 1)
= r2 = O.
The indicial roots are Oand O. Since these are equal, we expect the method of Frobenius to yield a
single series solution.
15. Substituting y
= L~=o cnxn+,.
into the differential equation and collecting terms, we obtain

00
2xy" - y'
+ 2y
+ L [2(k + r
= (2r2 - 3r) cox,.-l
- l)(k
+ r)ck
- (k
+ r)ck + 2Ck_l]Xk+,.-1
k=l
which implies
2r2 - 3r
= r(2r - 3)
=O
and
(k
The indicial roots are r = Oand r
+ r)(2k + 2r
- 3)Ck + 2Ck-l
= O.
= 3/2. For r = Othe recurrence relation is
Ck =
2Ck-l
- k(2k _ 3)'
k = 1,2,3, ... ,
and
= 2co,
Cl
For r
= 3/2
C2
= -2co,
the recurrence relation is

2Ck-l
Ck = - (2k + 3)k '
and
ci
= -Sco,
C2 =
= 1,2,3,
... ,
2
35 Co,
The general solution on (0,00) is

Y
4 3 +... ) + C2X 3/2 ( 1 - -x

2 + -x2
= Cl ( 1 + 2x - 2x 2 + -x
9
- -x4
945
L~=o cnxn+,. into the differential equation and collecting

+ 5y' + xy = (2r2 + 3r) cox,.-l + (2r2 + 7r + 5) cx,.
2xy"
35
+... ) .
terms, we obtain
00
+ L [2(k + r)(k + r
k=2
= O,
266
- l)Ck
+ 5(k + r)ck + Ck_2]Xk+,.-1
O,
Exercises 6.2
which implies
2r2
+ 3r = r(2r + 3) = O,
(2r2 + 7r + 5)
= O,
Cl
and
+ r)(2k + 2r + 3)Ck + Ck-2 = O.

and r = O, so Cl = O . Por r = -3/2
(k
The indicial roots are r = -3/2
Ck-2
Ck = - (2k - 3)k '
k = 2,3,4, ... ,
and
1
C2
C3 = O,
= -"2co,
Por r = Othe recurrence relation is

Ck =
Ck-2
- k(2k + 3) , k
= 2,3,4, ... ,
and
1
C3 = O,
C2 = ~ 14co,
C4 =
1
616co.

Y
17. Substituting y
4xy"
Clx
1
1 - -x
2
= :;~=o Cnxn+r
+ ~yl + Y =
2
-3/2
+ -x1
40
+... ) + C2 ( 1 -
-x1
14
+ -x1
616
+... ) .
(4r2 - ~r) coxr-l

2
f [4(k + r)(k + r - l)Ck + ~(k2 + r)ck +
k=l
=0,
which implies
4r2 - ~r = r (4r -
+ r)(8k + 8r
+ Ck-l
=O
and
1
"2(k
- 7)q
= O.
The indicial roots are r = O and r = 7/8. Por r = Othe recurrence relation is
Ck =
2Ck_l
k(8k _ 7)'
1,2,3, ... ,
and
Cl
C3= --Co
= -2co,
459
267
Ck-l]
xk+r-l
Exercises
6.2
For r = 7/8 the recurrence relation is

2Ck-l
ck=-(8k+7)k'
1,2,3, ... ,
and
2
Cl
= -15co,
345co,
C2 =
C3
4
= - 32,085 Co

Y = Gl ( 1 - 2x
18. Substituting y
2
+ -x
9
- -x4
459
+... ) + G2X7/8
1 - -x2
15
+ -x2
345
- --x 4
32,085
+... ) .
= :~=o Cnxn+r into the differential equation and collecting terms, we obtain
2x2y" - xy'
+ (x2 + 1) y
= (2r2 - 3r
+ 1) coxr + (2r2 + r)
ClXr+l
00
L [2(k + r)(k
+r -
+ r)ck + Ck + Ck_2]Xk+r
l)Ck - (k
k=2
= 0,
which implies
2r2 - 3r
+ 1=
(2r - l)(r - 1) = 0,
(2r2 +r)
Cl =
+ 2r -
+ l]Ck + Ck-2 = O.
0,
and
[(k + r)(2k
The indicial roots are r = 1/2 and r = 1, so
3)
= O. For r = 1/2 the recurrence relation is
Cl
k=2,3,4,
... ,
and
1
C2
= -"6co,
C3
= 0,
1
C4
= 168 Co
For r = 1 the recurrence relation is

Ck-2
Ck = -k(2k+1)'
k = 2,3,4, ... ,
and
C3
= 0,

Y = Glx
1/2 (
1
1 - -x
6
+ -x1
168
+... ) + G2X (11 268
-x
10
+ -x1
360
+... ) .
Exercises 6.2
19. Substituting y
3xy"
= L:;:O;o Cr,xn+r into the differential equation and collecting terrns, we obtain
+ (2 -
x)y' - y
(3r2
r) coxr-l
00
+ 2:J3(k + r
+ r)ck + 2(k + r)ck
- l)(k
- (k
+ r)ck_1Jxk+r-l
k=l
= O,
which implies
3r2
r = r(3r -1) = O
and
(k
The indicial roots are r
+ r)(3k + 3r
- l)Ck - (k
+ r)ck-l
= O.
= Oand r = 1/3. For r = Othe recurrence relation is

Ck-l
Ck = (3k - 1) )
k=1,2,3,
... ,
and
1
ci
For r
= 2'co,
= 1/3 the recurrence relation is

k
Ck-l
Ck=
3k'
2 3
'=1"
, ... ,
and
1
C]
= '3co,
C2 =
1
18co,
C3 =
1
162co.

Y
1
Cl ( 1 + -x
2
+ -x1
10
+ -x1
80
+... ) + C2X 1/3
1
1 + -x
3
+ -x1
18
+ -x1
162
+".
L:;:O=o Cr,xn+r into the differential equation and collecting terms, we obtain
= O,
which implies
and
[(k
+ r)(k + r
- 1)
269
~J
q-
ck-l
= O.
Exercses 6.2
The indicial roots are r = 2/3 and r = 1/3. For r = 2/3 the recurrence relation is
3Ck-l
Ck = 3k2 + k '
k = 1,2,3, ... ,
and
3
Cl = 4co,
C3
= 560co.

3Ck-l
Ck = 3k2 _ k'
k = 1,2,3, ... ,
and
3
Cl = 2'co,
C3
= 160co.
The general solution on (O,(0) is

Y = Clx
2/3 ( 1
+ -x3 + -x9
56
+ -x9
560
+... ) + C2X 1/3 ( 1 + -x3 + -x9
20
+ -x9
160
+... ) .
21. Substituting y = :L~""ocnxn+r into the differential equation and collecting terms, we obtain
00
2xy" - (3
+ 2x)y' + y
= (2r2 - 5r) coxr-l
+ "L[2(k + r)(k + r
- l)Ck
k=l
- 2(k
+r
5r = r(2r - 5)
=O
- 3(k
+ r)ck
- l)Ck-l
+ ck_dxk+r-l
=0,
which implies
2r2
and
(k
+ r)(2k + 2r
- 5)Ck - (2k
= O and r = 5/2. For r
+ 2r
- 3)Ck-l
= Othe recurrence relation is
(2k - 3)Ck-l
Ck = k(2k _ 5) ,
k = 1,2,3, ... ,
and
1
Cl = 3co,
Ck =
= O.
2(k + l)Ck-l
k(2k + 5) ,
and
270
k = 1,2,3,. " ,
Exercises
6.2

Y
= Gl ( 1 + -x1
3
22. Substituting y
1 2 - -x
1 3
- -x
6
+... ) + G2X 5/2
4
1 + -x
7
+ -x4
21
+ -x32
693
+... ) .
= ~=a cnxn+r into the differentia! equation and collecting terms, we obtain
f:
[(k
+ r)(k + r
- l)Ck
+ (k + r)ck
iq +
Ck-2]
xk+r
k=2
=0,
which implies
r2 - ~ = (r
(r2
D D
= O,
(r -
+ 2r + ~) ci
= O,
and
[(k
The indicia! roots are r
= -2/3
and r
Ck =
+ r)2
- ~] q + Ck-2
= 2/3,
so
= O.
= O. For r = -2/3
9Ck-2
3k(3k - 4) ,
k = 2,3,4, ... ,
and
C3
= O,
C4 =
-Ca
128

Ck =
3k(3k
+ 4) ,
k = 2,3,4, ... ,
and
C3
= O,
- --Ca
1280
.
,
C4 -
The general so!ution on (O,(0) is

Y
3 2 + -x9
Glx -2/3 ( 1 - -x
4
128
+... ) + G2X 2/3

271
1 - -x3
20
+ --x9
1,280
+... ) .
Exercises
6.2
23. Substituting y = L~=O cnxn+r into the differential equation and collecting terms, we obtain
9x2y"
+ 9x2y' + 2y
= (9r2 - 9r
+ 2) Coxr
00
+ r)(k + r - l)Ck + 2Ck + 9(k +r - 1)Ck_1Jxk+r
+ ~[9(k
k=l
= O,
which implies
9r 2
9r
+2 =
(3r - 1)( 3r - 2)
=O
and
[9(k
+ r)(k + r - 1) + 2lck + 9(k + r = 2/3. For r
The indicial roots are r = 1/3 and r
and
ci
For r
= O.
(3k - 2)Ck-l
k(3k - 1) ,
Ck = -
l)Ck-l
= -2"co,
= 1,2,3, ... ,
C3
7
120
= --Co.

(3k - l)Ck-l
k(3k+1)
,
Ck=and
k = 1,2,3, ... ,
ci =
-2"co,
C3
1
21
= --Co

Y = Clx
24. Substituting
2x2y"
1/3 (
1
1 - -x
2
1 2 - -x7
+ -x
5
120
+...
+ C2X 2/3 ( 1- -x1 + -x5

2
28
- -x1
21
+...
= L~=O cnxn+r into the differential equation and collecting terms, we obtain
+ 3xy' + (2x -l)y
= (2r2
+r
-1) coxr
00
+ ~ [2(k + r)(k + r - l)ck + 3(k + r)ck - Ck + 2Ck_1Jxk+r

k=l
= O,
which implies
2r2
+r
- 1 = (2r - l)(r
+ 1)
=O
and
[(k
+ r)(2k + 2r + 1) - l]Ck + 2Ck_l

272
= O.
Exercises 6.2
= -1
= 1/2. For r = -1
and r
Ck =
2Ck-l
k(2k _ 3)'
= 1,2,3, ...
and
ci
= 2co,
C2
= -2co,

2Ck-l
- k(2k + 3) , k
Ck =
and
= 1,2,3, ... ,
ci
= -"5co,
C3 = --Co
945

Y
= Clx -1 (
25. Substituting y
xy"
4
1 + 2x - 2x 2 + -x
9
+... ) + C2X 1/2 ( 1 -
2
-x
5
+ -x2
35
- -x4
945
+ ....)
= ~=o Cnxn+r into the differential equation and collecting terms, we obtain
+ 2y'
- xy = (r2
+ r)
+ (r2 + 3r + 2) Cl xr
coxr-l
00
+ :L [(k + r)(k + r -
l)Ck
+ 2(k + r)ck
- ck_21xk+r-l
k=2
= 0,
which implies
r2
+ r = r(r + 1) = O,
(r2
+ 3r + 2)
Cl
= O,
and
(k
rl
= O and r2
+ r)(k + r + l)Ck
= -1,
so
= O. For
Cl
Ck-2
Ck = k(k
- Ck-2 = O.
+ 1) ,
rl
k=2,3,4,
and
1
C2 =
3fco
C3 = es = C7 =
C2n = (2n
... = O
+ l)!co.
273
= Othe
... ,
recurrence relation is
Exercises
For
1"2
6.2
= -1 the recurrence relation is

Ck-2
Ck = k(k - 1) ,
k = 2,3,4, ... ,
and
C3
= Cs = C7 = ...
C4
4TCQ
1
C2n = (2n)! Co .
Y=
e ~
2n
l;:O(2n+1)!x
[e1;:0~
(2n
= -;;;
1
x
+ l)!x
1
;:O(2n)!x
-1 ~
2X
2n+ 1
ef
+ 2 n=O
2n
1 2n]
(2n)!x
= -[el smh x + e2cosh z].

26. Substituting y = z.:::;:o=o
cnxn+r into the differential equation and collecting terms, we obtain
x2y"
+ xy' + ( x2
- ~) y = (1"2 - ~) cQxr
+ (1"2 + 21" + ~) C1Xr+1
f [(k + 1")(k +
1" -
l)Ck
+ (k + 1")ck -
~2
~Ck
+ Ck-2]
xk+r
=0,
which implies
(1"2 + 21"
and
C1
= O,
[(k+T)--4 .) 1] Ck+Ck-2=0.
1"1
= 1/2 and
1"2
= -1/2, so
Cl
Ck-2
Ck = - k (k + 1) ,
274
= O. For
1"1 =
= 2,3,4, ... ,
1/2 the recurrence relation is
Exercses
and
1
C2
= --CO
3!
C3
Cs
C7
= ... = O
(_1)n
C2n = (2n + 1)! en.
For r2
= -1/2 the recurrence relation is

Ck-2
k = 2,3,4, ... ,
q=-k(k-1)'
and
C2 =
C3
1
--cO
21
= Cs
C7
= ...
=O
1
C4 =
4!cO
(_1)n
C2n = -(2n.)1 Ca
y
CIxl/2
f (2n(_1)n+ 1)!x2n + C2X-I/2 f (_1)n

x2n
(2n)!
n=O
= CIX-I/2
n=O
f (2n(_l)n+ 1)!x2n+1 + C2
X-I/2
n=O
f (_l)n
(2n)!
x2n
n=O
= x-I/2[CI sinx + C2 cosx].

I::~=o Cnxn+r into the differential equation and collecting terms, we obtain
00
xy" - xy'
+ y = (r2 - r) coxr-l +
.L [(k + r + l)(k
+ r)ck+1 - (k + r)ck + q]xk+r = O
k=O
which implies
r2 - r = r(r - 1) = O
and
(k + r + l)(k + r)ck+1 - (k + r - l)Ck = O.

rl
= 1 and r2 = O. For
rl
= 1 the recurrence relation is
kci.
Ck+1 = (k+2)(k+l)'
k=0,1,2, ... ,
275
6.2
Exercises 6.2
and one solution is YI = CoX. A second solution is
Y2 = X
J -J
-dx
e x2
dx = x
J :2
x
(2_x2 +.!. +.!.2 +!..x3!
= xJ
1 (
dx = x J x2 1 + x
+ !"x2 + ...) dx
1 ')
+ -x1
2
12
= x ln x - 1 + -x-
=x
4!
+ -x1
72
+ 2x2 + 3!x3 + ...

[_.!.+ lnx
x
+.!.x
2
dx
+ 2_x2 + 2_x3 + ...]

12
72
+ ....
28. Substituting y
y"
= I:~="o cnxn+r into the differential equation and collecting terms, we obtain
3
+ -;yl
- 2y = (r2
+ 2r)
coxr-2
+ (r2 + 4r + 3)
CIXr-1
00
L [(k + r)(k
+r
- l)Ck
+ 3(k + r)ck
- 2ck_21xk+r-2
k=,,2
= O,
which implies
+ 2r
r2
(r2
(k
The indicial roots are TI = O and
+ 2) = O
= r(r
+ 4r + 3)
CI
+ r)(k + r + 2)Ck -
r2 =
2Ck-2 = O.
-2, so c = O. For
2Ck-2
Ck = k(k + 2)'
rl = O the
k = 2,3,4, ... ,
and
1
C2
= ::tco
C3= C5
= C7= ... = O
1
C4
= 48co
1
1,152
C6= --Co
The result is
Yl
1214+
1 -x
= Co( 1 + -x
4
48
276
+ --C6
+
1,152
oo'
Exercises 6.2
A second solution is
Y2 = YI J
= YI
= YI J
e- J(3jx)dx
2
dx = YI
YI
dx
x3 (1 + Ix2
+
~x4
2
48
+ _!_x6
+ ...) = YI
576
(2_x3 - .2_2x + _2'_x48
~X3
576
+y
= -~YIlnx
[__
dx
x3
(1 + Ix2
+ _'!_x4
+ ...)
4
48
+ ...) = YI
+ _2'_x2 _
1_
2x2
[ __
1 (
1 2
x3 1 - -2x
1_ - ~
2x2
lnx
7 4
19 6
)
+ 48
x + 576x + ...
+ _2'_x2 96
~x4
2,304
+ ...]
+ ...] .
~x4
2,304
96
29. Substituting y = :;;:o=o
Cr,xn+r

00
xY"
+ (1-
x)Y' - y
= T2cOXr-1 + L[(k
+ T)(k + T -
l)Ck
+ (k + T)Ck -
(k
+ T)ck_Ilxk+r-1
k=O
which implies
T2 = O
and
TI =
T2 = O and the recurrence relation is

Ck =
Ck-I
k
'
k = 1, 2, 3, ....
One solution is
YI = Co ( 1 + x
+ -x21213+ IX
+... )
3.
= coe
Y2 = YI
e- J(ljx-l)dx
e2x
= e xJ1( -
1- x
dx = eX
+ -x1213 2
= eX[l nx - x + -x1
J eXIx
J 1
e2x dx = e" -;;e-xdx
-x
3!
2 - --x1
2 2
3 .31
+ ... )
dx
= e xJ(l - - 1 + -x1 - -x12 + ... ) dx

x
3!
3 + ... ] = e xl nx - e x~(_l)n+1
L...,
x n.
n=1 n n!
277
= O,
Exercises 6.2
30. Substituting y
= I:~=,QCnxn+r into the differential equation and collecting terrns, we obtain

00
xY"
+ y' + y
+ [(k + r)(k + r
= r2cQxr-1
- 1)Ck + (k
+ r)Ck + Ck_l]Xk+r-l = O
k=l
= O and
which implies r2
"i
= T2 = O and the recurrence relation is

Ck-l
Ck = - k2'
k = 1,2,3, ....
One solution is
Yl
= ca
( 1- x
+ 22x
1
- (3!)2x
Y2
e- f(l/x)dx
= u.
= Yl
J x ( 1 - 2x + 2x2 -
dx
Yl
J
u: J (~ +
1(
5
;; 1+2x+2'x
= Yl lnx
= y1ln x
+ (4!)2x
2 + ~x
ca
i: (-1)n
(n!)2
n
x .
dx
(1 - x + ix2
- 3~x3
+ ...)
35
+ 288x4
5
gX3
23
+gX
- ...
4
)
+ 677
288x + ...
dx
+ 23x2 + 677 x3 + ...) dx
288
27
+ --x677
<1
1,152
5 2 23
+ -x
+ -x
( 2x
-...
dx
5 2 23
+ 2x + -x
+ -x
+ Yl
J
X
=Yl
=
2
Yl
27
+ --x677
+ ... ]
4
1,152
+... ) .
31. Substituting y
= I:~=a
cnxn+r into the differential equation and collecting terrns, we obtain

00
xy"
+ (x
- 6)y' - 3y = (r2 - 7r)caxr-1
+ [(k + r)(k + r k=l
- 6(k
+ r)ck
- 3Ck_dxk+r-1
which implies
r2
7r
= r(r
278
- 7)
=O
O,
1)Ck + (k
+r -
1)Ck-l
Exercises 6.2
and
(k + r)(k + r - 7)Ck + (k + r - 4)Ck-1 = 00
The indicial roots are r1
= 7 and r2 = 00 For r1 = 7 the recurren ce relation is

(k+ 7)kck
+ (k + 3)Ck-1
=0,
k=l,2,3,ooo,
or
k
Taking co
= 1,2,3,
=f. O we obtain
1
C1 = --Co
C2
5
18cO
1
C3 = -6"co,
and so on. Thus, the indicial root r1 = 7 yields a single solution.
relation is
k(k - 7)Ck
+ (k
- 4)Ck-1 = O,
Now, for r2 = O the recurrence
k = 1,2,3,
Then
-6C1 - 3co = O
-12c3-C2=0
-12c4 + OC3 = O
===}
C4
=O
+ C4 = O
===}
Cs
=O
=O
===}
C6
=O
OC7 + 3C6 = O
===}
C7 is arbitrary
-lOcs
-6C6 + 2cs
and
k-4
Ck = - k(k _ 7) Ck-1,
279
k = 8,9, lO,
Exercises 6.2
Taking Co 1= O and C7 = Owe obtain
1
Cl = --Co
2
1
C2= lOcO
1
C3 = --Co
120
C4
Taking
= es = C6 = ... = O.
Co = O and C7 1= O we obtain
ca = --C7
2
Cg =
ClO
1
= - 36c7'
In this case we obtain the two solutions

1
1 2
1 3
Yl = 1 - "2x + 10x - 120x
32. Substituting y
= ~=o
x(x - l)y"
+ 3y' -
cnxn+r
5
36c7
Y2 = X
and
1 a
-x
2
+ -x5
36
1
- -x
36
10
+ ....
2y
00
= (4r - r2) coxr-l
+ [(k + r
- l)(k
+r
- 12)Ck_l - (k
+ r)(k + r
k=l
= O,
which implies
4r - r2 = r( 4 - r)
and
-(k
+ r)(k + r
rl
= 4 and
-(k
- 4)Ck + [(k
r2
+r
- l)(k
= O. For rl = 4 the
+T
or
k=1,2,3,.
280
2) - 2]Ck-l
= O.
+ 4)kck + [(k + 3)(k + 2) ck=(k+1)Ck_l,
2]Ck-l = O
- l)Ck
Exercises 6.2
Taking
Co =1=
Owe obtain
CI =
and so on. Thus, the indicial root
2co
C2
= 3!co
C3
= 4!co,
= 4 yields a single solution. For
TI
T2
= O the recurren ce relation
is
-k(k
- 4)Ck
+ k(k
- 3)Ck-1
= O, k = 1,2,3, ... ,
or
-(k-4)Ck+(k-3)Ck_1
=0,
k=
1,2,3, ....
Then
3CI -
2co = O
2C2 - CI = O
C3
+ OC2
OC4
=O
+ C3 = O
=:>
C3 =
=:>
C4
is arbitrary
and
Ck =
Taking
Co =1=
O and
C4 =
(k-3)Ck_1
c-4
k = 5,6, 7, ....
Owe obtain
2
C2
= -Co
3
1
= "3co
C3 =
Taking
Co =
O and
C4 =1=
C4 = C5 = ... = O.
Owe obtain
CI = C2 = C3 =
In this case we obtain the two solutions

2
1 2
YI = 1 + "3x + "3x
and
33. (a) Frorn t
= l/x we have dt/dx
= _1/x2
5364
Y2 = x 42+
x7
= _t2
Then
281
+ ....
Exercises 6.2
dy
dx
dy di
dt dx
_=
-_
dy
dt
-t -
and
Now
x4 ~d2
dx
+ Ay
1 (d2
t4 __J!_2
dt
+ 2t3 ddY ) + Ay
= 4"
d2 + -2 _}j_
d + Ay = O
-i
dt
t dt
becomes
d2y
t -d
(b) Substituting y
d2y
t -d 2
dy
+ 2 -dt + >..ty
= O.
= L~=Ocntn+r into the differential equation and collecting terms, we obtain

dy
+ 2 - + Aty = (r2 + r)cotr-l

dt
+ (r2 + 3r + 2)Cltr
00
+ L [(k + r)(k + r -
l)Ck
+ 2(k + r)ck + ACk_2]tk+r-1
k=2
= O,
which implies
r2
+ r = r(r + 1)
(r2
+ 3r + 2)
= O,
CI = O,
and
(k
+ r)(k + r + l)Ck + ACk-2
The indicial roots are r 1 = Oand

Ck =
r2 =
-1, so
ACk-2
k( k + 1) ,
CI = O.
For
C3
O.
rl = Othe
k = 2,3,4, ... ,
and
C2 =
A
3!
--ca
= C5 = C7 = ...
282
=O
Exercises
For
r2 =
6.2
-1 the recurrence relation is

.ACk-2
Ck = - k( k - 1) ,
= 2,3,4, ... ,
and
.A
C2
= --ca
2!
C3
= Cs =
C7
= ... = O
~[CI
t
f (2n(-lt+ 1)! (V>:t)2n+1 + C2 f (_l)n

(2n)!
n=
(V>:t)2n]
n=
= ~[CI sin v>: t + C2 cos v>: t l.

t
(e) Using t = l/x, the solution of the original equation is

v>:
sin -
y(X)CIX
34.
+ C2X
v>:
cos -.
x
(a) Frorn the boundary conditions y(a) = O, y(b) = O we find

v>:
CI sin -
+ C2 cos -v>:
a
v>:
= O
v>:
CI sin -b- + C2 cos -b- = O.

Since this is a homogeneous system of linear equations, it will have nontrivial solutions if
. v>:
Slll-
.~
sm-
v>:
cosa
~
cosb
. v>:
v>:
v>: . ~
= sm - cos - cos - Sllla
b
a
b
=sin(~
- ~)
283
=sinv>:C~ba)
=0.
Exercises 6.2
This will be the case if
v0.
or
n ,'\ab
=b_
nAab
= L'
n = 1, 2, ... ,
or, if
The criticalloads
we have
are then Pn = n27f2(a/b)2 E/o/ L2.

y
= Clx[sin
v0. _
x
= C3X [Sin
and
Using C2
sin( v0./a) cos v0.]

cos( v0./ a)
x
v0. cos v0. _ cos v0. sin v0.]

x
(1 1) = C4xsin--n7fab ( 1-
Yn(x) = C3xsin--n7fab - - L
(b) When n
= 1,
= 11, and
= -Cl sin( v0./a)/ cos(v0./a)
1, we have,
-a) .
x
for C4 = 1,
Yl(x) = x sin 1.l7r
(1 - ~) .
35. Express the differential equation in standard form:
ylII + P(x)y" + Q(x)y' + R(x)y = O.

Suppose Xo is a singular point of the differential equation. Then we say that Xo is a regular singular
point if (x - xo)P(x),
(x - xo)2Q(x), and (x - xo)3R(x) are analytic at x = xo.
36. Substituting y = 2:~=o cnxn+r into the first differential equation and collecting terms, we obtain
00
x3y" + y = coxr +
L [Ck + (k +
7" -
l)(k + 7"
2)Ck_lJxk+r
k=l
It follows that Co
= Oand
Ck =
-(k + 7"
l)(k + 7"
The only solution we obtain is y(x) = O.
284
2)Ck_l'
= O.
Exercses
Substituting y
6.3
= L~=O Cnxn+r into the second differential equation and collecting terms, we obtain
00
x2y"
+ (3x
- l)y'
+y
+ 2: [(k + r + 1)2Ck -
= -rco
(k
+ r + l)Ck+dxk+r
= 0,
k=O
which implies
+ r + 1)2ck
(k
If en
-rco
- (k
+ r + l)Ck+l = O.
= 0, then the solution of the differential equation is y
= O. Thus, we take r = 0, from which
we obtain
Ck+l = (k+ l)Ck,
Letting
Co
is then y
= 1 we get
Cl
= 2,
= L~=O(n + l)!xn,
C2
= 3!,
C3
k = 0,1,2,
....
= 4!, and so on. The solution of the differential equation
which converges only at x
37. We write the differential equation in the form x2y"
= O.
+ (b/a)xy' + (c/a)y
= O and identify ao
= b/a
and bo = e] a as in (12) in the texto Then the indicial equation is

r(r - 1)
+ -a r + -a =
or
ar2
+ bxy' + cy
which is also the auxiliary equation ofax2y"
+ (b -
a)r
= O.
Exercises 6.3
1. Since v2
= 1/9 the general solution is y = clJl/3(X)
+ C2Ll/3(X).
3. Since v2
= 25/4 the general solution is y
+ c2 Yl (x).
= qJS/2(x)
+ c2Ls/2(x).
4. Since v2
= qJl/4(X)' + c2L1/4(X).
5. Since v2
2. Since v2 = 1 the general solution is y = Cl JI (x)
6. Since v2
7. Since v2
8. Since v2
1/16 the general solution is y
+ C2Yo (x).
= 4 the general solution is y = clJ2(X) + C2Y2(X).
= 2 the general solution is y = qh(3x)
+ C2Y2(3x).
= 1/4 the general solution is y = qJl/2(6x)
+ C2J_l/2(6x).
9. If y = x-l/2v(x)
+ c = 0,
the general solution is y = q Jo (x)
then
'21x-3/2v(x),
y' = x-l/2v'(x)
y" = X-l/2v"(x)
_ x-3/2v'(x)
285
+ ~x-S/2v(x),
4
Exercises
6.3
and
x2y"
+ 2xy' + x 2x2y
= x3/2v"
+ X 1/2v' + ( x 2x3/2
- ~X
-1/2) v.
Multiplying by xl/2 we obtain

X2V" +xv'
whose solution is
10. From y
= xnJn(x)
+ ()..2x2
+ c2J_1/2()..X).
= c1J1/2()..X)
-~)
V = O,
Then y = cx-1/2J1/2()..X)
+ C2X-1/2J_l/2()..X).
we find
and

xn+l J~
+ 2nxn
+ n(n
J~
= xn-l
- 1)xn-1 Jn
.O
+ (1 -
2n)(xn J~
+ nxn-1
Jn)
+ xn+l
Jn
(sinee Jn s a solution of Bessel's equation)
= O.
Therefore, xn Jn is a solution of the original equation.

11. From y = x-n Jn we find
,
y=x
-nJ' n-nx
-n-lJ

xy"
+ (1 + 2n)y' + xy
= x-n-1
[x2J~
= x-n-1
+ xJ~ + (x2
(sinee Jn is a solution of Bessel's equation)
= O.
Therefore, x-n Jn is a solution of the original equation.

12. From y = Vx Jv()..x) we find
y'
= )..v'xJ~()"x)
- n2) Jn]
+ ~x-l/2Jv()..x)
and
286
Exercises 6.3
x2y" + (A2X2 - v2 +
y = JX [A2x2J~(AX) + AxJ~(AX) + (A2x2 - v2) Jv(AX)]
JXo
(since Jn is a solution of Bessel's equation)
= o.
Therefore,
JX
Jv(AX) is a solution of the original equation.
13. From Problem 10 with n = 1/2 we find y
= x1/2J1/2(X).
Frorn Problem 11 with n
= xJ1(x).
Frorn Problem 11 with n
= -1/2
we find
y = x1/2J_1/2(X).
14. From Problem 10 with n = 1 we find y
y = xJ-l(X)
From Problem 11 with n = 1 we find
-X-1j_1(X).
JX Jo(2x).
we find y = JX J3/2(X)
16. Frorn Problem 12 with A = 2 and v = O we find y

17. From Problem 12 with A = 1 and v = 3/2
and y =
JX J-3/2(X).
= 3 we find y = x3 J3 (x). From Problem 11 with n
18. From Problem 10 with n
-1 we find
-xJl(X).
15. From Problem 10 with n = -1 we find y = X-1j_1(X).
y = X-1Jl(X)
= -3 we find
= X3j_3(X) = -x3]s(x).
19. (a) The recurrence relation follows from
fa
00
-vJv(x)
+ xJv-l(X)
=-
=_
(-l)nv
(x)2n+v
+ v + n) '2
(::.)2n+v +
00
(-l)nv
fan!r(l+v+n)
+ v) (::.)2n+v _
(-1)n(2n
- n=O
L n.Ir( 1 + v + n )
00
00
(_l)n
(x)2n+v-l
+ x ~o nlr(v + n) '2
nlf(l
00
(_l)n(v
fan!f(l+v+n)
+ n)
.2 (::') (::.)2n+v-l
2
2
- xJv(x).
(b) The formula in part (a) is a linear first-order differential equation in Jv(x).
An integrating
factor for this equation is z", so
ddx[xvJv(x)] = XVJv-l(X).
20. Subtracting the formula in part (a) of Problem 19 from the formula in Example 4 we obtain
0= 2vJv(x)
- xJv+1(x) - xJv-1(x)
or
21. Letting v = 1 in (15) in the text we have
xJo(x)
dx[xJl(X)]
so
287
2vJv(x) = xJv+l(X) + xJv-l(X).
Exercises 6.3
22. From (14) we obtain Jb(x) = -JI (x), and from (15) we obtain J(x) = Ll(X).
LI(X)
= -JI
(x).
23. Since
1
r ( 1 - "2 + n
(2n - 1)!
(n _ 1)!22n-I
we obtain
24. (a) By Problem 20, with
1/
= 1/2, we obtain
JIj2(X) = xJ3j2(X) + xJ_Ij2(X)
hj2(X)=Y;;; [2 (SinX
-;--cosx
with
1/
= -1/2
we obtain -J_1j2(X)
1/
= xJ1j2(X) + xJ_3j2(X)
J_3j2(X) =
and with
-ff
so that
so that
(CO;X + sinx);
= 3/2 we obtain 3J3j2(X) = xJSj2(X) + xJ1j2(X) so that
(h)
-0.5
-1
-1
y
'1
o.5
-0.5
-1
'1
L\ v ~
C>.-3/2C">
5 'Cl10
""B
0.5 /""'\
20 x
v C>.
~ 5/2
C">
25. Letting
we have
dx = dx ds = dx [~ {k (_~) e-o.tj2] = dx (_
dt
ds dt
dt a. Y -;;;,
2
ds
288
(k e-o.tj2)
Y -;;;,
Thus Jb(x)
Exercises
and
d2x = .!!:._ (dX) = dx (::
dt2
dt dt
ds 2
(k e-at/2) + .!!:._
Y -:;
(dX)
ds
dt
(_
(k e-at/2)
Y -:;
= dx (:: (k e-at/2) + d2x ds (_ (k e-at/2)

ds
Y -:;
Y -:;
ds2 dt
= dx (:: (k e-at/2) + d2x (!5_ e-at) .

ds
Y -:;
ds2
Then
m -d2x
dt2
+ ke-atx
= ke-at -d2x
ds2
+ -ma
2
If
-
e-at /2 -dx
dt
+ ke-a
t x = O.
Multplying by 22/ a2m we have

22 k e-at d2x
a2 m
ds2
+ ~ {k ="? dx + .2_ !5_ e-atx

a
Y -:;
di
=O
a2 m
or, since s = (2!a)Nme-a.t/2,

2 d2x
s ds2
dx
+ s ds + s
x = O.
26. Differentiating y = x1/2w (ax3/2) with respect to ax3/2 we obtain

y' = x1/2w' (~ax3/2) ax1/2
+ ~x-1/2w
(~ax3/2)
and
+ aw'
y" = axw" (~ax3/2) ax1/2
+ ~aw'
Gax3/2)
(~ax3/2)
- ~x-3/2w Gax3/2)
Then, after combining terms and simplifying, we have

y"
+ a2xy
= a [~x3/2wll
+ ~w' + (ax3/2
- 4a~3/2) w] = O.
Letting t = ~ax3/2 or ax3/2 = ~t this differential equation becomes

~ ~ [t2wlI(t)
+ tw'(t) + (t2
- ~) w(t)]
= O,
t O.
27. The general solution of Bessel's equation is
Thus, the general solution of Airy's equation for x

Y
=x
1/2w
> O is
(2"3ax 3/2) = C1X1/2J1/3 (2s=3/2) + C2X1/2J-1/3 (2"3ax 3/2) .

289
6.3
Exercises 6.3
28. Setting y = ft JI (2ft)
and differentiating we obtain
y' = ft J~(2ft)
2e
2yx
le JI (2ft)
2yx
= J{(2y'X)
le JI (2y'X)
2yx
and
"
"(
"(
1,
y = JI 2y X ) 2y'X + 2y'X JI (2y X ) 2ft

=
ftJ1
el,
2yx)+2xJ1(2yx)-
1
(e
- 4x3/2 JI 2y X )
4x3/2Jl(2yx).
Substituting into the differential equation and letting t = 2y'X we have
xy" + y = y'X J~'(2y'X) + ~J{(2Vx)

2
Jx
t2 [t2"
JI
[XJ~'(2y'X) +
le h(2y'X)
4yx
V; J;(2y'X)
+ (x -
+ y'X Jl(2y'X)
JI (2y'X)]
1) (]
(t) + 4t, JI (t) + (t2 - 4 JI t)
2..
[ t2 J;' (t) + tJU t) + (t2 2t
1) JI
(t )l.
Since h(t) is a solution of t2y" + ty' + (t2 - l)y = O, we see that the last express ion aboye is Oand
y = ft JI (2ft) is a solution of xy" + y = O.
29. (a) Using the expressions for the two linearly independent power series solutions, Yl(X) and Y2(X),
given in the text we obtain
P6(X) = 116(231x6 - 315x4 + 105x2 - 5)

and
P7(X) = 1~ (429x7 - 693x5 + 315x3 - 35x) .

(b) H(x)
satisfies
(1- x2)
y" - 2xy' + 42y = Oand P7(X) satisfies
30. The recurrence relation can be written

2k + 1
k
Pk+l (x) = -k-- XPk(X) - -k- Pk-l (x),
+1
+1
290
(1- x2)
y" - 2xy' + 56y = O.
k = 2, 3, 4, ....
Exercises
k = 6: P7(X) = 13x (231 x6 _ 315x4 + 105x2 _ ~) _ ~ (63 x5 _ 35x3

7
16
16
16
16
6 8
4
429 7 693 5 315 3 35
= -x
- -x
+ -x
--x
16
16
16
16
6.3
15X)
8
31. If x = cos then

dy
. dy
= - smede
dx'
d2y
. 2 d2y
dy
de2 = sm e dx2 - cos dx '
and
d2y
sine de2
+ cose dy
de + n(n + l)(sine)y
[(
= sine
2 ) d2y
dy
1- cos e dx2 - 2cose dx
+ n(n + l)y ] = O.
That is,
d2y
( 1 - x 2) -dx2
32. Frorn
y = u/vx
dy
2xdx
+ n(n + l)y
= O.
we find
2xu' - u
y' = -2-x3;C-/""'2-
and
Then
2 "
x y
+ xy +
(2
2)
4x5/2u" - 4x3/2u'
x - u y=
4x
+ 3xl/2u
2xu' - u
2x1/2
(x2;2)u
xl/2
= x3/2u" + ~x-l/2u + x3/2u - ;2X-l/2u = O.

4
Multiplying by xl/2 we obtain

x 2"u
1 + x 2u + -u
4
2u =
or
d2u
dx2
+ (1-
;2
-1/4)
x2
u= O.
For large values of x this equation can be approximated by u"
+u
= O whose solutions are sin x
and cos x. Thus, we expect the solutions to oscillate with increasing frequency for larger x.
33. Rolle's theorem states that for a differentiable function f(x),
exists a number e between a and b such that f'(c)
JO(x) = h(x).
for which h(x)
for which f(a)
= O. Frorn Problem 20 with
f(b) = O, there
;
= O we have
Thus, if a and b are successive zeros of Jo(x), then there exists a e between a and b
= J6(x) = o.
291
Exercises 6.3
34. Since
!J; f(x) dx! ::; J; If(x)1 dx, we have

11n(x)l::; ~ rlcos(xsint-nt)ldt::;
Kh
35. (a) We identify m
= 4, k = 1, and
o:
= 0.1.
r 1dt=
K)O
l.
Then
x(t) = cllo(lOe-o.05t) + c2Yo(10e-005t)

and
x'(t) = -0.5Cll(10e-0.05t) - 0.5e2Y(lOe-0.05t).

Now x(O)
= 1 and x'(O) = -1/2 imply

cllo(lO)
+ c2Yo(10) =
ql(lO)
+ c2Y(10) =
1.
Using Cramer's rule we obtain

Y(10) - Yo(10)
el = lo(10)Y(1O) - l(10)Yo(1O)
and
10(10) - lb(10)
C2= lo(10)Y(10) - lb(lO)Yo(lO) .
Using Y = -Yr and l = -11 and Table 6.1 we find el
= -4.7860 and C2= -3.1803. Thus
x(t) = -4.78601o(lOe-o.05t) - 3.1803Yo(lOe-0.05t).

(b)
10
5
-5
50
36. (a) Identifying o:
100
150
200
1, the general solution of x" + ttx

x(t) = Clxl/2 11/3 (~x3/2)
Solving the system x(O.l)

(b)
= O is
+ C2xl/2]_1/3 (~x3/2) .
= 1, x'(O.l) = -~ we find
-1
292
= -0.809264 and C2= 0.782397.
Exercises 6.3
37. (a) Letting t
= L - x, the boundary-value problem beeomes
d2
dt2
+ ohe = O,
e/(o) = O,
e(L) = O,
= Og/E l. This is Airy's differential equation, so by Problem 27 its solution is
where 0:2
= cltl/2Jl/3
+ C2tl/2J_l/3
Go:t3/2)
Go:t3/2)
= CJel(t) + C2e2(t).
(b) Looking at the series forms of el and e2 we see that el (O)

boundary eondition e/rO)
= o implies
= O, and
el
e(t) = C2VtJ-l/3
=1=
O, while e2(0) = o. Thus, the
so
Go:t3/2).
From e(L) = O we have

c2VLJ-l/3
so either
C2 =
O, in whieh
case
GO:L3/2)
= O,
e(t) = O, or J_l/3(~o:L3/2)
= O. The eolumn will just start
to bend when L is the length eorresponding to the smallest positive zero of J-l/3.
Mathematica,
the first positive root of Ll/3(X)
is Xl ~ 1.86635. Thus ~o:L3/2
Using
= 1.86635
implies
L =
(
3(1.86635)
)
20:
2/3
= [ 9El (1.86635)2 ] 1/3
40g
= [9(2.6
X 107)7l'(0.05)4/4(1.86635)2] 1/3
4(0.28)7l'(0.05)2
38. (a) Writing the differential equation in the form xy"
+ (PL/M)y
76.9 in.
= O, we identify A
PL/M.
From Problem 28 the solution of this differential equation is

y = Cl.xJl
Now
h(O)
+ C2.xYl
(2VPLX/M)
(2VPLX/M).
= O, so y(O) = O implies C2 = O and
y = Cl.xJl
(b) From y(L) = Owe have y = JI (2L..fPM)
(2VPLX/M).
= O. The first positive zero of JI is 3.8317 so, solving
2LV PI! M = 3.8317, we find PI = 3.6705M / L2. Therefore,

Yl(X) = Cl.xJl
2 J3.6705X
L
) = Cl.xJl
293
(3.8317)
-L-.x
Exercises 6.3
(e) For
el
= 1 and L = 1 the graph of Y1 = y'X J1 (3.8317 y'X)
is shown.
0.2
0.1
x
0.2
39.
Zeros and differences of Jo, J1, and
0.4
0.6
0.8
Jo
.Jo
JI
.J1
J2
.J2
2.4048
5.5201
8.6537
11.7915
14.9309
3.1153
3.1336
3.1378
3.1394
0.0000
3.8317
7.0156
10.1735
13.3237
3.8317
3.1839
3.1579
3.1502
0.0000
5.1356
8.4172
11.6198
14.7960
5.1356
3.2816
3.2026
3.1762
Successive zeros of Jn for n = O, 1, 2 appear to be approximately equally spaced for larger values
of x. Furthermore, this spacing appears to be approximately
can be used to determine that successive roots of Jo near x
it .
To check this conjecture a CAS
= 200 are 200.277156 and 203.418739,
with difference 3.141583. This is consistent with the observation in Problem 32 that the frequency
of the solutions for large x is approximately 1/27r.
40. (a) Using a CAS we find
1 d1
P1(x) =
"2 dx1
P2(x)
1d22
212
222! dx2 (x - 1) = "2(3x - 1)
P3(X)
1 d3
2
3
233! dx3 (x - 1)
= "2(5x - 3x)
P4(x)
1 d4
244! dx4 (x2 - 1)4
= S(35x4
(x - 1) = X
d5
30x2
+ 3)
P5(X) = 255' dx5 (x2 _1)5 = S(63x5 - 70x3 + 15x)

1d62
616
P6(X) = 266! dx6 (x - 1) = 16 (231x - 315x + 105x - 5)

1
d7
P7(X) = 277' dx7 (x - 1) = 16 (429x - 693x + 315x - 35x)

294
Exercises 6.3
(b)
P1
1
P2
1
0.5
-1
0.5
1 x
-1
-0.5
-1
P5
1
-1
-1
P7
1
P6
1
0.5
0.5
-0.5
-1
-1
-1
(e) Zeros of Legendre polynomials
H(x) : O
P2(X) : 0.57735
P3(X) : O, 0.77460
P4(X) : 0.33998, 0.86115
P5(X) : O, 0.53847, 0.90618
P6(X) : 0.23862, 0.66121, 0.93247
P7(X) : O, 0.40585, 0.74153 ,0.94911
PlO(X):
0.14887, 0.43340, 0.67941, 0.86506, 0.097391
The zeros of any Legendre polynomial are in the interval (-1, 1) and are symmetric with respect
to O.
295
Chapter 6 Review Exerc;ses
1. The interval of convergence is centered at 4. Since the series converges at -2, it converges at least
on the interval [-2,10).
Since it diverges at 13, it converges at most on the interval [-5,13). Thus,
at -7 it does not converge, at O and 7 it does converge, and at 10 and 11 it might converge.
2. We have
x3
sin x
x-
f ()x =--=
cosx
3. Solving x2 - 2x
+ 10 =O we obtain
radius of convergence is
4. Setting 1 - sin x
1-
x5
6 + 120 x
'2 + 24 -
x3
=x+-+-+
2x5
...
J11, which
x =1
11 - J111 = J11-
...
15
.. .
are singular points. Thus, the minimum
1.
= O we see the singular points closest to O are -311'/2 and 11'/2. Thus, the minimum
radius of convergence is 11'/2.

5. The differential equation (x3
+ y' + y
x2)y"
= O has a regular singular point at x = 1 and an
irregular singular point at x = 1.

6. The differential equation (x - l)(x
+ 3)y" + y = O has
regular singular points at x = 1 and x
= -3.
7. Substituting y = I:~o cnxn+r into the differential equation we obtain

00
2xy"
+ y' + y =
(2r2 - r) coxr-l
+ L [2(k + r)(k + r -
l)Ck + (k
+ r)ck + ck_llxk+r-l = O
k=l
which implies
2r2 - r = r(2r - 1) = O
and
(k
+ r)(2k + 2r -
The indicial roots are r = O and r = 1/2. For

Ck =
l)Ck + ck-l
= O the recurrence relation is
- k(2k - 1) , k= 1,2,3, ... ,
so
Cl
= O.
= -Co,
C3
= --Co
90

k(2k
+ 1) ,
= 1,2,3, ... ,
so
C3
296
1
630
= --Co

Two linearly independent solutions are
1
( l-x+6x
Y1=Cl
-90X1
+ ... )
and
Y2 = C2X
1/2 (
'31x + 301 x 2 -
1-
1 3
630x +...
8. Substituting y = L~=o cnxn into the differential equation we have

00
y" - xy' - y
00
00
L n(n - 1)cnxn-2 - L nenxn - L cnxn

n=2
n=l
n=O
~
"'--"
k=n-2
k=n
k=n
00
00
+ 2)(k + 1)Ck+2Xk -
= L(k
k=O
00
L kCkXk - L Ckxk
k=l
k=O
00
= 2C2 - Co
+ L[(k + 2)(k + 1)Ck+2 k=l
Thus
2C2 - Co = O
(k
+ 2)(k + 1)Ck+2 -
(k
+ l)Ck = O
and
1
C2
ck+2
Choosing ca = 1 and
C1
= '2co
= k + 2 Ck,
k = 1, 2, 3, ....
= Owe find
1
2
C2
=-
C3
= C5
C4
=-
C6
=-
= C7
= ... = O
1
8
1
48
297
(k
+ l)CkJXk = O.

and so on. For
ca = O and Cl =
1 we obtain
C2 = C4 = C6 = ... = O
1
3
1
=15
C3 = -
C5
1
= 105
C7

Yl
= 1 + 2x + S"X + 48 x + ...
and
Y2=X+3X
1
+15x
1
+105x
+ ....
= L~=a Cnxn into the differential equation we obtain
9. Substituting y
00
+ 3y =
(x - l)y"
(-2C2
+ 3ca) + L (k
- l)(k -
2)Ck-l
k(k - l)Ck
+ 3Ck_2]xk-2 = O
k=3
which implies
C2 = 3ca/2
and
ce=
Choosing
ca =
1 and
Cl
(k - l)(k - 2)Ck-l
k(k-1)
k = 3,4,5, ....
= O we find
3
C2 =
and so on. For
+ 3Ck-2
ca = O and c
2'
C3 =
2'
C3 =
2'
5
8
C4 =-
= 1 we obtain
C2
= O,
C4 = -

Yl
301354
= Cl ( 1 + -x~ + -x
2
2
+ -x + ....)
8
and
Y2 = C2 (x
10. Substituting y
+ ~x3 + ~x4 + ...) .
= L~=a cnxn into the differential equation we obtain

00
y" - x2y'
+ xy = 2C2 + (5C3 + ca)x + L [(k + 3)(k + 2)Ck+3

k=l
which implies
C2 = O, C3 = -ca/5,
and
k-1
k = 1,2,3,
298
....
- (k - l)cklxk+l
=O

Choosing
Co
= 1 and
and so on. For
co
Cl
= Owe find
= O and
Cl
C3
1
=--
C4
= C7 = ClO = ... = O
Cs
= Cs = Cll
C6
1
=-_
= ...
=O
90
= 1 we obtain
C3
= C6 = Cg = ... = O
C4
C7
Cs
Cs
= Cll = ...
ClO
= ... = O
=O

Yl = Co
11. Substituting y
= :;~=oCnxn+r
(1 - ~x3 -
:ox6 - ... )
and
Y2 = ClX.
into the differential equation, we obtain

00
xy" - (x + 2)y' + 2y = (r2
3r)coxr-l
.L [(k + r)(k
+ r - 3)Ck
k=1
- (k
+r -
3)Ck_Ilxk+r-l
= O,
which implies
r2
3r = r(r - 3) = O
and
3)Ck-1 =
(k + r)(k + r - 3)Ck - (k + r The indicial roots are
rl
O.
= 3 and r2 = O. For r2 = O the recurrence relation is
k(k - 3)Ck - (k -
3)Ck-l
= O,
= 1, 2, 3, ....
Then
Cl -
OC3 -
Co
OC2
=O
=O
==>
C3
is arbitrary
and
k = 4,5,6,
299
....

Taking Co =j:. O and C3 = Owe obtain
C2 = 2"CO
C3 = C4 = Cs
= ... = O.
1
C4 = C3
6
4!C3
Taking Co = O and C3 =j:. Owe obtain
Cs
1
6
5 . 4 C3 = 5! C3
1
C6 = ~C3
= 6!C3,
and so on. In this case we obtain the two solutions

Yl
1 2
= 1 + x + 2"x
and
Y2 = x 3 + -x6 4 + -x6
4!
5!
+ -x6 6 + ... = 6e x - 6 ( 1 + x + -x1 2) .

6!
12. Substituting y = ~=o cnxn into the differential equation we have

(cosx)y "
+y
(11 - -x 2 + -x1 4 - -x1 6 +

2
24
720
... ) (2C2 + 6C3X + 12c4X 2 + 20C5X 3 + 30C6X4 + ... )
[2C2 + 6C3X + (12c4 - C2)X2 + (20cs - 3C3)X3
+ [co + C1X + C2x2 + C3x3 + C4x4 + ...
= O.
Thus
Co + 2C2 = O
300
+ (30C6 - 6C4 + 112C2) x4 + ... ]

20cs - 2C3 = O
1
30C6 - 5C4 + 12c2 = O
and
c? = --Co
C3
= -"6Cl
C4
=O
Cs =
C6 =
Choosing
Co
= 1 and
Cl
Co
1
1
-C4 - -C2
6
360
= Owe find
= O and
-"2'
Cl
= 1 we find
C2
and so on. Por
1
10 C3
C2=0,
C3
= O,
C3=-"6'
C4
= O,
= O,
Cs
C4=0,
Cs
= - 60'
es
=O

Yl
13. Substituting y
1 2
1
= 1 - -x
+ -x
+...
= x - -1x 3 - -x1
2
720
6
L~=O Cnxn into the differential equation we have
00
y"
Y2
and
+ xy' + 2y
2:= n(n
00
- 1)Cnxn-2
+ 2:=
n=2
00
2:= (k + 2)(k
k=O
60
+ ....
00
ncnxn
+ 2 2:= Cnxn
n=l
k=n-2
n=O
k=n
'--v---"
k=n
00
00
k=l
k=O
+ 1)Ck+2Xk + 2:= kCkXk + 2 2:=
Ckxk
00
= 2C2 + 2co
+ 2:= [(k + 2)(k + 1)Ck+2 + (k + 2)Cklxk

k=l
Thus
2C2 + 2co = O
(k
+ 2)(k + 1)Ck+2 + (k + 2)Ck =
and
C2
= -Co
ck+2 = - k
+ 1 Ck,
301
k = 1,2,3, ....
= O.

Choosing ca = 1 and
Cl
= O we find
C2
= -1
C3
C4
1
=-
C6
=--
Cs
C7
= ... =
C6
= ... =
15
and so on. For ca = O and q = 1 we obtain

C2
C4
C3 =--
Cs = -
8
1
C7=
--
48
and so on. Thus, the general solution is

Y
ca ( 1 - x 21416)
+ 3"x - 15x
+... + cl
(131S
X -"2x
+ B"x
17
- 48x
+ ...)
and
Y,
= ca (
-2x
+ -x4332S -
Setting y(O) = 3 and y'(O)

= -2 we find
y = 3 - 2x - 3x
14. Substituting y
+... ) + Cl
-x
Co
3254
1 - -x
2
= 3 and
Cl
16
+ x + x 41S-:x
23
+ -x
8
= -2.
- 5"x
+ ...)
- -x76
48
Therefore, the solution of the
+ 2417 x + ....
= L~=a cnxn into the differential equation we have

00
(x
+ 2)y" + 3y =
L n(n -
CXJ
+2
l)cnxn-l
n=2
L n(n
+3
n=2
k=n-l
L (k + l)kck+lXk
k=l
L cnxn
n=O
k=n
'---v----'
k=n-2
00
00
- 1)cnxn-2
CXJ
CXJ
+ 2 L (k + 2)(k + 1)Ck+2Xk + 3 L
Ckxk
k=O
k=O
CXJ
= 4C2 + 3co + L [(k + 1)kck+l + 2(k + 2)(k + 1)Ck+2 + 3Cklxk

k=l
Thus
4C2
(k
+ l)kq+l
+ 3ea
=O
+ 2(k + 2)(k + 1)Ck+2 + 3q

302
=O
= O.

and
C2
3
4
--Co
k
Ck+2 = -
Choosing
Co
1 and
Cl
2(k
+ 2) Ck+l
2(k
k = 1,2,3, ....
+ 2)(k + 1) cu,
= Owe find
3
C2
=-4
C3
=-
1
8
1
C4 =-
16
9
320
Cs=-and so on. For
Co
= O and
Cl
=1
we obtain
C3
=-4
C4 =-
16
es = O
and so on. Thus, the general solution is
co
3213
( 1 -:x + S"X
+ 1614x -
9
320x
s)+...
+ Cl X -:x13 + 1614x + ... )
and
y,
Setting y(O)
21394)
= ce (33
-2"x
+ SX +:x - 64x +...
= O and y'(O) = 1 we find
Co
= O and
+ Cl
(31213
-:x
+:x +... ) .
= 1. Therefore, the solution ofthe initial-value
Cl
problem is
Y
1 3
= x - -x
+ -x1
16
+ ....

y
11
(1- cos x ) y , + xy
x
= O,
and noting that
is analytic at x
= O, we conclude
that x
= Ois an
ordinary point of the differential equation.
303

y" +(eX - x1 - x )y
=O
and noting that

x
2
2
x
x2
...
eX - 1 - x
x
3 18 270
we see that x = O is a singular point of the differential equation. Since
----=---+-+-_
?(
X)
x: eX _ 1 _ x
2x2
x4
3
+ x3
18 + 270 -
= 2x -
... ,
we conclude that x = O is a regular singular point.

I.:~=Ocnxn
into the differential equation we have
00
y"
+ x2y' + 2xy
2: n(n
00
n=2
2: (k + 2)(k
k=O
Cnxn+l
n=O
k=n+ 1
n=l
k=n00
00
+ 2: ncnxn+l + 22:
- 1)cnxn-2
~~
2
k=n+ 1
00
+ 1)Ck+2xk + 2: (k
00
+ 22: Ck_lXk
- l)Ck_lXk
k=2
k=l
00
= 2C2 + (6C3 + 2co)x + 2:[(k + 2)(k + 1)Ck+2 + (k + l)Ck_lJXk = 5 - 2x + 10x3.

k=2
Thus
6C3 + 2co=-2
20C5
(k
and
+ 4C2 = 10
+ 2)(k + 1)Ck+2 + (k + l)Ck-l = O,
= 4,5,6, ... ,
5
C2 = 2
1
C3=--CO--
C4
1
--Cl
C5 = ~ -
~C2 = ~ - ~
1
ck+2 = - k
+ 2 Ck-l
304
G)
=O

Using the recurrence relation, we find
1
1
C6 = -6"C3 = 3.6 (co
+ 1) = 32.
1
2! co + 32.2'
1
1
C7 = --C4 = -Cl
7
47
Cs = Cll = C14 = ... = O
1
1
1
C9 = -gC6 = - 33. 3! co - 33.3'
ClO =
1
10C7 =
1
4 . 7 . 10 Cl
C12 = - 12 C9 = 34 . 4! Co + 34 . 4!
Cl3 =
1
1
13 Co = 4. 7 . 10 . 13 Cl
and so on. Thus

1 3
1
[ l-'3x +32.2!x
y=co
+ci [x-'4x
1 4
1
-33.3!x
+ 4.7x
+ [~x2 _ ~x3
2
3
18. (a) From y
1
+34.4!x
1
- 4.7.10x
+ _1_x6
10
__
32 . 2!
12
- ...
1
+ 4.7.10.13x
1_x9
33 . 3!
+ _1_x12
34 . 4!
1 du
u dx
= -- - we obtain
Then dy / dx
= x2 + y2 becomes
-;
1d u
dx2
+ u2
()2dudx
=x
+ u2
(b) If u = xl/2w( ~x2) then

u' = x3/2w' (~x2)
+ ~x-l/2w
and
305
(~x2)
( dudx )2 '
13
]
-oo.
_ ... ].

so
u" + x2u
Letting t
= xl/2
[x2w" Gx2)
+ 2w' Gx2)
+ (x2 - ~x-2) W Gx2)]
= O.
~x2 we have
Vii [2tw"(t) + 2w'(t) +
(2t - 4 \t)
W(t)] = O
or
t2w"(t) + tw'(t) + (t2 - 116)w(t) = O.

This is Bessel's equation with
1/
= 1/4, so
(e) We have
y = _]:,du = _
u dx
xl/2w(t)
..!!:_xl/2w(t)
dx
= __ 1_
[xl/2dW dt + ~x-l/2w]
dt dx
2
= __ 1_
[x3/2dW + _1_ w]
dt
2xl/2
xl/2w
xl/2w
=-
1 [2x 2 di
dW]+ w = - 2xw
1 [dW
2xw
4t di + w ] .
Now
4t
dw
di + w = 4t dt [C)l/4(t) + C2J_l/4(t)] + clJl/4(t) + C2J-l/4(t)

= 4t [Cl (J_3/4(t) - :t Jl/4(t))
+ C)l/4(t) + C2J_l/4(t)
= 4CltJ_3/4(t) - 4C2th/4(t)
so
306
+ C2( -
:/-1/4(t)
- h/4(t))]

Letting e
= el / e2 we have
19. Let
1
Y2 = -x[ln(1
2
+ x)
- ln(1 - x)]-
so that
Y2,1[1
= -x
2
1
+ -- 1] + -[ln(1
+ x)
-l+x
l-x
-ln(1 - x)]
and
"
1 [1
Y2 = 2"x - (1 + x)2
1[
+ (1 -
1]
x)2
1[
1
1+ x
+ 2"
1]
1]
1- x
1[
+ 2"
1
1+ x
1]
x
+ 1-
= 2"x - (1 + x)2 + (1 - x)2 + 1 + x + 1 - x'

Then
+ x)y~ - 2xy~ + 2m = o,
(1 - x)(l
20. (a) By the binomial theorem we have
[1
+ (t2 -
2xt
(b) Letting x
)] -1/2
= 1 - 2"1 (t 2 -
2xt
+ '83 (t 2 -
= 1 in (1- 2xt+t2)-1/2,
we have
+ t2)-1/2
=-
(1 - 2t
= (1 - t)-l
2xt
)2
+ ... = 1 + xt + 2"1 (23x
= 1+ t
1-t
+ t2 + t3 + ",
- 1) t 2 + ....
(Itl < 1)
n=O
From part (a) in the text we have
00
:L Pn(l)tn
00
= (1 - 2t
+ t2)-1/2 = :L e
n=O
n=O
Equating the coefficients of corresponding terms in the two series, we see that Pn(l) = 1.
Similarly, letting x = -1 we have
(1 + 2t
+ t2)-1/2
= (1 + t)-l
1
l+t
=-
= 1- t
:L Pn(-I)tn,
00
= :L(-I)ntn
n=O
00
n=O
so that Pn( -1) = (-I)n,
307
+ t2
- 3t3
+ ".
(Itl < 1)
The Laplace Transform

Exercises 7.1
-s -
(2
4e-8tdt = _~e-st
s
= -;e
2.
o'i'{f(t)} =
Jo
1
= ( - -e
s
1 (
1
-; - O- -;e
6.
s2
f(
s>O
s>O
_~(e-2S - 1),
s
-1 ) - -1(O- e -8)
s2
= (-~e-stsint
1 ( 1 - e -8) , s>O
= -s2
_ -2-1-e-stcost)
s +1
s +1
+ -2-1-e-stsint)
o'i'{f(t)} = t'" (cost)e-8tdt = (-~e-stcost

J1</2
8 +1
(O
+ _1_e-1<S/2)
82
+1
= __
82
1_e-1<s/2,
+1
+1
11<
O
s> O
100
1</2
8> O
t)={O,O<t<l
t, i > 1
o'i'{f(t)} =
l te-si dt =
OO
8.
1
- -; ,
(O + _2_1_e-1<s) - (O - -2-1_) = _2_1_(e-1<S + 1),

8+1
s+l
8+1
=O-
7.
12 =
-8
-s - -e
1 -s) - ( 0--;-2 - -1) = -1( 1 - 3e-S) + -2 ( 1 - e -S) ,
S
s2 S
S
s2
o'i'{f(t)} = r(sint)e-stdt
Jo
=
-s) = -;e2
-s - -es21 -s) - ( O-
2 -s - -e
2
= ( --e
5.
__1te-st __ 1)e-st
8
82
100 = _1 e-s + -1 e-s,

I
s2
O,
0< t < 1
f(t) = { 2t _ 2, t> 1
o'i'{f(t)} = 2 (00 (t _ 1)e-8t dt = 2 (-~(t

JI
S
_ l)e-st _ ~ e-sI) 100 = 2 e-s,

s2
I
s2
308
8> O
s>O
Exercses 7. 1
1 - i,
:t'{f(t)}
10. f(t) =
12. :t'{f(t)}
13. :t'{f(t)}
0< t < 1
t> O
9. f(t) = { O
,
(1(1 _ t)e-st di = (-~(1 _ t)e-st + ~ e-st)
Jo
O,
O<t<a
e,
a < t < b;
O,
11=
82
~ e-s
82
8> O
82
-5
= ( 00 e-2t-5e-stdt = e-5 ( 00 e-(s+2)tdt = __e_e-(s+2)t
Jo
+ ~ - ~,
Jo
+2
= (00 te(4-s)tdt = (_1_te(4-S)t
(00 tte-stdt
Jo
Jo
4- 8
100-5= _e_,
e(4-S)t)
(4 - 8)2
8> -2
+2
100
o
1
= (4-8)2'
14. :t'{f(t)}
1000 t2e-2te-st
>4
di =
1000t2e-(s+2)tdt
= ( __ 1_t2e-(S+2)t _
8+2
15. :t'{f(t)}
1000 e-t(sin
= (
-(8+1)
(5+1)2+1
2
te-(s+2)t _
2
e-(S+2)t) 100=
2
(8+2)2
(8+2)3
O (8+2)3'
t)e-stdt =
1000(sin t)e-(s+l)tdt
e-(s+l)tsint-
- (8 + 1)2 + 1 - 82 + 28 + 2 '
1 - ,e(1-s)t
8
( (1-8)2+1
1-8
100
e-(S+l)tcost)
(8+1)2+1
cos t
8> -1
1
(1-8)2+1
8- 1
52 _ 25 + 2'
>1
309
e(1-s)t sin t )
100
o
8> -2
Exercises 7. 1
17. Y{f(t)}
18. Y{f(t)}
= 10':;<) t(cost)e-stdt
=
st
[( --2-1
8 +
82 - 1
(82+1)2'
28). ?
(82+ 1)-
(smt)e -st]
28
(82+1)2'
28) 2 (cost)e-st(82+1)
(8t-2--+
82- 12) (sint)e-st ] 00

(82+1)
O
8 +1
8>0
= 24~
20.
= 5!
Y{t5}
86
7 3
22. Y{7t + 3} = 2" + 8
8
2
2
6
3
23. Y {t + 6t - 3} = - + - - 83 82 8
24. Y{-4t
3!
2
3
1
25. Y{t3 + 3t2 + 3t + 1} = """"4 + 33" + 2" +8
8
8
8
26. Y{8t
t}
1
1
= - + -8 8- 4
29. Y{l +2e2t +e4t}
28. Y{t
121
+-8 8-2
8-4
= - + --
33. Y {sinhkt}
+ 16t + 9} = -4-
-12t
8 - k
CP
-..L-
cosht} = Y {e-t et ~ e-t}
2
+6t-1}
{ 2t -2+e
= 2
8 + 25
8
2
8 - k
= 2
1}_
{1
1
-e 2t - - - -1
2
2
2(8- 2) 28
= y {~+ ~e-2t} = 218+ 2(8~ 2)
310
3!
84
2 6 1
83 82 8
= 8--12-+---
-2t }=----+-1
8-2
3D. Ye
37. Y {sin2t cos2t} = y {~sin4t} = 2 2

2
8 + 16
2
16 9
+ - +83 82 8
-9t
2
1
5
- e
+ 5} = - - +83 8 + 9 8
34. Y {coshkt}
= 2k 2
(/){et sm
. h t} _
(/){ et et - e-t}
35. ..L-..L2
.82
32. Y {cos5t + sin 2t}
31. Y{ 4t2- 5sm3t} = 4-2 - 5-- 3

83
82+ 9
36. Y{e-t
8>0
4
10
21. Y {4t - lO} = - - 82 8
27. Y {1+
00
= 1000 t(sint)e-stdt
= [( --2--t
8 +1
19. Y{2t4}
82-1) 2 (cost)e-st + (t-2-1

(82+ 1)
8 +
2
8
1
8+2
+ -28 +4
Exercises
38. ~ {cos t}
=~
{1"2 + "21cos 2t } =
lIs
28
39. (a) Using integration by parts for
(b) Let u = st so that du
Q:
+ "2
82
+4
> O,
= 8 dt. Then
Q:
= r(1/2)
40. (a) ~{cl/2}
81/2
41. Identifying
For n = 1,
7. 1
> -l.
f!_
V-;
f(t) = tn we have f'(t) = ntn-1, n = 1, 2, 3, .. , . Then, since f(O) = O,
~{t}=-~{1}=2'
For n = 2,
For n
= 3,
42. Let F(t)
= t1/3.
Then F(t) is of exponential order, but f(t)
t = O and hence is not of exponential order.
311
F'(t)
iC2/3 is unbounded near
Exercises 7. 1
Let
f (t)
2tet cos et
d
2
sin et . This function is not' of exponential order, but we can show that
dt
=:_
its Laplace transform exists. Using integration by parts we have

:;t {2tet
cos et 2 }
= 1000 e-st
(00
= s)o
-d sin et 2) dt
dt
2
lim [ e-st sin et 2

a->oo
laO + s loaO e-st
sin et 2 dt ]
= s :;t{sin et }.
e-st sin et dt
Since sin et2 is continuous and of exponential order, :;t {sin et2} exists, and therefore :;t {2tet2 cos et2}
exists.
43. The relation will be valid when s is greater than the maximum of C1 and
44. Since et is an increasing function and t2
C2.
> In M + ct for N! > O we have et2 > e1n M+ct = N!ect for
2
t sufficiently large and for any c. Thus, et is not of exponential order.

45. By part (c) of Theorem 7.1
:;t {e(a+ib)t} =
1
s - (a + ib)
1
(s-a)+ib
(s - a) - ib (s - a)
s - a + ib
(s-a)2+b2
+ ib
By Euler's formula, eie = cos () + i sin (), so

:;t {e(a+ib)t} = :;t {eateibt.}:;t {eat (cos bt
= :;t { eat cos bt}
+ i :;t { eat sin bt}
s- a
= (s - a)2
+ i sin bt)}
+ b2 + i (s
b
- a)2
+ b2
Equating real and imaginary parts we get

at
:;t { ecos
46. We want f(ax
bt} = (
s-a
S -
r-+a b')-
+ (3y) = af(x) + (3f(y)

m(ax
and
:;t{eatsinbt}
= (
s-a
~2
+ b2'
or
+ (3y) + b = a(mx + b) + (3(my + b) = m(ax + (3y) + (a + (3)b
for all real numbers a and (3. Taking a = (3 = 1 we see that b = 2b, so b = O. Thus, f(x) = mx
will be a linear transformation when b
= O.
312
+b
Exercises
Exercises 7.2
3. ~
-1 {1
48}
s2 - s5
4!}
4
= ~ -1 {1s2 - 48
24 . S5 = t - 2t
6. ~-1{(S+32f}=~-1{~+42.2+2.
s
s
s
23}=1+4t+2t2
s
7. ~-1{12_~+_1_}=t_1+e2t
s
s
s- 2
9.
~-l
10.
{_l_}
4s + 1
~-l
~-l
{_l_}
_
5s - 2 -
~-l
13.
~-l
14.
~-l
1 } = ~e-t/4
4
{~.
4 s + 1/4
1 }_- ~e2t/5
5
{~ .
5 s - 2/5
4s } = ~-l
4s2 + 1
1 }=
4s2 + 1
{~ .
~-l
15. ~-1{2s2-6}=~-1{2_2s
s+9
s2 + 1/4
= cos ~t
1/2 } = ~ sin ~t
2 s2 + 1/4
2
2
s+9
7.2
-2.~}=2COS3t-2sin3t
s+9
313
Exercises
7.2
/2}
{s.
/2-
S + 1 }
16. X' -1{ ~2
= X' -1 -2-- + 1! -2-- = cosvr:
2t + - sin vr:
2t
s +
s + 2 v2 s + 2
2
17.
18.
19.
20.
1 } = X'-1
X'-1 { S + 1 } = X'-1
s2 _ 4s
_1_}
{_~ . ~ + ~ . _1_} = _~ + ~e4t
4 s
4 s- 4
4 4
X'-1 {
S
} = X'-1 {~ . _1_ + ~ . _1_} = ~et + ~e-3t
s2 + 2s - 3
4 s - 1 4 s +3
4
4
X'-1 {
X'-1 {
s2 + 3s
s-O.l
O)~s
)} = X'-1 {(0.3). _1_
s+0.2.
s-O.l
X'-1 {(s _ vi~s3 +

-1 {
23. X'
26. X'
27.
} _ X'-1 {~ . _1 __ ~ . _1_} _ ~e4t _ ~e-5t

9 s - 4 9 s +5 - 9
9
s2 + S - 20 -
21. X'-1 {(
22.
-1{
X'-1 {
{~ . ~ _ ~ .
= ~ _ ~e-3t
3 s
3 s+3
3 3
)3)}
= X'-1 {s2 ~ 3
(s2+4)(s+2)
28 - 4
-1 {
(s-2)(s-3)(s-6)
=X'
=X'
-1 {
= -4 +
( -1{ -- 1}
28. X'
s4 - 9
=X'
-1 {1
-.----.--
6V3
s+0.2
-)3 . s2~
1
2"s-2-s-3+2"s-6
1
s
1
2
's2+4+'s2+4-'s+2
} _ X'-1 {
(s2 + s)(s2 + 1) -
+ (0.6). _1_}
0.3eo.lt + 0.6e-O.2t
3} = coshv'3t - v'3 sinh v'3t

1
1}
1}
=2'e
2t
3t
-e
+2'e
1
1 .
=cos2t+sm2t-4e
2s - 4 } _ X'-1 {4
3
s
3}
s(s2 + 1)2 -~ + s + 1 + s2 + 1 + s2 + 1
3e-t + cost + 3 sin t
)3
s2 - 3
6V3
V3}
s2 - 3
314
1.
r,;
=-smhv3t--smv3t
6)3
1.
6/3
r:
6t
1 -2t
Exercises
30.
68 + 3}
.:t'-l { (82 + 1)(82 + 4)
= .:t'-l
2}
{8
1
8
1
2 82 + 1 + 82 + 1 - 2 82 + 4 - '2 . 82 + 4
1
= 2 cos t + sin t - 2 cos 2t - '2 sin 2t

31. The Laplace transform of the differential equation is
1
8.:t{y}
- y(O) - .:t{y} = -.
8
Solving for .:t' {y} we obtain

1
.:t'{y} = -- + -.
8
1
1
8-
Thus
28.:t'{y} - 2y(0)
= .:t'{y} = O.

6
.:t'{y} = 28+ 1 = 8+ 1/2

Thus
8.:t{y}
- y(O) + 6.:t{y}
1
8_
4.

1
.:t'{y} = (8-4)(8+6)
2
+ 8+6
= 10' 8-4
19
19
1
+ 10' 8+6'
Thus
Y
4t
= 10e + 10e
-6t

8.:t' {y} - .:t' {y}
28
= 82 + 25 .
Solving for .:t'{y} we obtain

2
Y{y} = (8-1)(82+25)
1118
= 13' 8-1
- 1382+25
Thus
1
y = -e
13
1
- -cos5t+
13
315
5
-sin5t.
13
55
+ 13' 82 +25'
7.2
Exercises 7.2
S2 ~ {y} - sy( O) - y' (O) + 5 [s ~ {y} - y (O)] + 4 ~ {y} = O.
Solving for ~ {y} we obtain
s+5
~{y}
= s2
+ 5s + 4 = "3 s + 1 -"3 s + 4
Thus
y = ~e-t _ ~e-4t.
3
3
s2 ~{y}
_ sy(O) - y'(O) - 4[s ~{y}
_ y(O)] = _6
s-3
3_.
s+l

<o
oL-{y}-
6
- (s - 3)(s2 - 4s) - (s
5 1
2
= _. - - -2 s
s - 3
3
+ 1)(s2 - 4s)
s-5
s2 - 4s
+-=--
3
1
11
1
+ -.
--o
5 s + 1 10 s - 4
- _. --
Thus
y
5
2
3
5
= __ 2e3t __ e-t
11
+ _e4t
10

s2~{y}
Solving for ~{y}
- sy(O) +~{y}
~2'
we obtain
~
{y}
+~=~+_2
(s2+1)(s2+2)
s2+1
2_
s2+1
s2+1
s2+2'
Thus
y = lOcos t + 2 sin t -
h sin ht.

s2 ~{y}
+ 9~{y}
= -.
s -1

~{y}
1
= (8-1)(s2+9)
11111
= 10' s-l
- 10' s2+9
Thus
y
1 tI.
= -e
- -sm3t
10
30
316
1
- -cos3t.
10
-10'
s
s2 +9'
7.2
Exercises
2 [s3.:t'{y} - s2(0) - sy'(O) - y"(O)] +3[s2 .:t'{y}-sy(0)-y'(0)]-3[s
.:t'{y}-y(0)]-2
.:t'{y} = _1_.
s+l
.:t'
{y}
2s + 3
(s+1)(s-1)(2s+1)(s+2)
= ~ "_1_ + ~
2s+1
_1 __
18s-1
+ ~ _1_
9s+1/2
9s+2"
Thus
1
5 __ 8e-t/2 + _e-2t
1
Y = _e-t
+ _et
2
18
s3 .:t'{y} - s2(0) - sy' (O)- y" (O)+ 2[s2 .:t'{y} - sy(O) - y' (0)]- [s .:t'{y} - y(O)]- 2 .:t'{y} = +-9
s
"
.:t'{}s2+12
y - (s - 1)(s + 1)(s + 2)(s2 + 9)
13
= 60
13
16
s - 1 - 20 s + 1 + 39 s + 2 + 130 s2 + 9 - 65 s2 + 9 .
Thus
13 t
Y = -e
60
41. Let f(t)
= 1 and g(t) = {
13 -t
- -e
20
16 -2t
3
+ -e
+ -cos3t
39
130
1 "
- -sm3t.
65
t > O, t =f. 1
. Then .:t'{f(t)} =.:t'{g(t)}
O, t = 1
1,
= 1, but f(t) =f. g(t).
.:t'{}
.:t' { }
s +3
s +3
y +
y = (s + 3)2 + 4 = s2 + 6s + 13 .
.:t'{ }_
Y
s+3
(s+1)(s2+6s+13)
4 .s +1 - 4
=4'
1 (S+3
s+1-4'
s+l
s2+6s+13
2)
(s + 3)2 + 4 - (s + 3)2 + 4 .
Thus
1
y = _e-t - e-3t cos 2t + _e-3t sin 2t.

4
4
43. For y" - 4y' = 6e3t - 3e-t the transfer function is W(s) = 1/(s2 - 4s). The zero-input response is
yo(t)=.:t'-l{
s-5
}=.:t'-l{~.~_~._l_}=~_~t,
s2 - 4s
4 s
4 s - 4
317
Exercises 7.2
and the zero-state response is
Yl (t) =:L
-1
{6
=:L-l{27.
3}
(3 _ 3)(32 _ 43) - (3 + 1)(32 - 43)
_1
20 3 - 4
2_ + ~. ~ _ ~. _1_}
3- 3 4 3 5 3+ 1
Exercises 7.3
1.
:L {telOt} =
2. :L {-6t}
te
3.
1
(3 - 10)2
1
= (3 + 6)2
:L{t3e-2t} =
4. :L{t10e-7t}
5
3!
(3 + 2)4
lO!
(3 + 7)11
:L{t(et+e2t)2}=:L{te2t+2te3t+tt}=
1
(3-2)2
2
6.
2
(3-3)2
2
-2t
(3_13)2+9
3+2
= (3+2)2+16
8.
:L { e
9.
:L{(1 - et + 3e-4t) cos 5t} =:L {cos 5t - et cos5t + 3e-4t cos 5t}
cos4t
3
- 32+25
11.
1
(3-4)2
:L{ e2t(t - 1)2} =:L {t2e2t - 2te2t + e2t} = (3 _ 2)3 - (3 _ 2)2 + 3 _ 2
7. :L{etsin3t}=
10.
t) } =:L {3tge
( { 3t(
.
:L e 9 - 4t + 10 sm"2
:L-l{
3-1
(3-1)2+25
- 4te
3t
3(3+4)
+ -,-----'---:-,,-(3+4)2+25
t}
3t .
9
4
5
+ lOe sm"2 = 3 _ 3 - (3 _ 3)2 + (3 _ 3)2 + 1/4
1
} _ :L-l{~
2
} _ ~t2e-2t
(3+2)3
2 (3+2)3
- 2
318
7.3
Exercises
{(S ~ 1)4}
= ~-l
(S ~!1)4}
= ~t3et
12.
~-l
13.
~-l {2s - 6s1 + 10} = ~-l {(s - 3)\ + 12}= e3t sin t
14.
~-l{ s2+ 2s1 + 5 }=~-l{~2 (s + 1)22+ 22}=~e-tsin2t

2
15.
~-l{ s2 + 4ss + 5 }=~-l{
16.
~-l{s2 +2s+5
}=~-1{2 (s +(s+3)
_~
5
}=2e-3tcos5t-~e-3tsin5t
6s + 34
3)2 + 52 5 (s + 3)2 + 52
5
17.
18.
19.
20.
{~
(s +22) 2-2
12 2}=e-2tcost-2e-2tsint
(s + 2) + 1
(s + 2) + 1
~-l {(s: 1)2} =~-l {s(:: 0;}=~-l {s ~ 1 - (s: 1)2} = e-t - u:'
~-l {(s-2)2
5s } = ~-1 {5(S - 2) + lO} =~-1 {_5_ +
10 } = 5e2t + 10te2t
(s-2)2
s-2
(s-2)2
2s - 1}
{5; - s21 - S +5 1 - (s +4 1)2 - 2"3 (s +2}1)3 = 5-t-5e -4te
~
s2(s + 1)3 = ~
-1 {
~-1
-t
-1
{(S + 1)2} =
(s+2)4
~-1
}=
1
_
2
+~
3!
(s+2)2
(s+2)3
6(s+2)4
u:? _ t2e-2t
+ ~t3e-2t
6
s~{y}
Solving for ~{y}
we obtain ~{y}
= (s
- y(O) + 4~{y}
1
+ 4)2 +
y
= -4'
s+
+ 4' Thus
= te-4t + 2e-4t.
s ~{y} - ~{y}
Solving for ~{y}
1
=;
+ (s _ 1)2 .
we obtain
~{y}
s(s - 1) + (8 - 1)3 =
Thus
319
-;
+ -s---1 + -(8---1-=-)3
.
-t
2 -t
-2"t e
Exercises 7.3
S2 :t'{y} - sy(O) - y'(O) + 2[s:t'{y}

Solving for :t'{y}
- y(O)] + :t'{y} = O.
we obtain
s
+3
:t'{y} = (s + 1)2 = s + 1 + (s + 1)2 .

Thus
s2 :t'{y} - sy(O) - y'(O) - 4 [s:t'{y} - y(O)] + 4:t'{y}

Solving for :t'{y}
we obtain
(s ~ 2)4'
:t'{y} = 2~ (s ~!2)6 . Thus, y = 2~t5e2t.

1
s
s2 :t'{y} - sy(O) - y'(O) - 6 [s :t'{y} - y(O)] + 9 :t'{y} = 2"'

Solving for :t' {y} we obtain
:t' y} = 1 + s2
{
s2(s - 3)2
= 2 ~+ ~~ _ 2 _1_ + 10
27 s
9 s2
27 s - 3
1
9 (s - 3)2 .
Thus
2
y=27+9t-27e
2 3t
10 3t
+gte
s2 :t'{y} - sy(O) - y'(O) - 4 [s:t'{y} - y(O)] + 4:t'{y}
= 4"'
s
Solvng for :t' {y} we obtain
:t' y = s5 - 4s4 + 6 = ~ ~ + ~~ + ~!
+ ~ ~ + ~ _1 __
{}
s4(S - 2)2
4 s
8 s2
4 s3
4 s4
4 s - 2
13
8
Thus
Y=
3 2
1 2t
+ st + t + t + e -
13 2t
ste
S2 :t'{y} - sy(O) - y'(O) - 6 [s:t'{y} - y(O)] + 13 :t'{y} = O.

Solving for :t' {y} we obtain
:t'{y} = - s2 _ 63s+ 13 = -~ (s _ 3~2+ 22 .
320
(s - 2)2 .
Exercises
7.3
Thus
y
3
.
= - 2e3t
SIl1 2t.
2[82.:t'{y}
- 8Y(0)] + 20[8 .:t'{y} - y(O)] + 51.:t'{y}
= O.

.:t"
{y}
48+40
282+208+51
28+20
(8+5)2+1/2
2(8+5)
(8+5)2+1/2
10
(8+5)2+1/2'
Thus

2
.:t' {y} - 8y (O) - y' (O) -:- [8.:t' {y} - y (O)] =
8 -
(8 _
1) 2 + 1 .

.:t'
{y}
=~~_~
1
8(82-28+2)
8-1
+~
282(8-1)2+12(8-1)2+1'
Thus
1 1t
1t.
y = - - -e cos t + -e sm t.
2 2
2
82 .:t'{y} - 8Y(0) - y'(O) - 2 [8.:t'{y} - y(O)] + 5.:t'{y}

88
= ~ + 12,
.:t'{y}
2. ~
482 + 8 + 1 =
+ ~ _.!:_ + -78/25 + 109/25
82(82 - 28 + 5)
25 8 5 82
82 - 28 + 5
7 1 1 1
7
8- 1
25858225(8-1)2+2225(8-1)2+22'
= - - +- - - -
51
+ - ----,..."....---.",-
Thus
7
Y=25
+ -t1
5
7 t
- -e cos 2t
25
51 t .
+ -e
sm 2t.
25
31. Taking the Laplace transform of both sides of the differential equation and letting e = y(O) we
321
Exercses 7.3
obtain
se {y"} + ~ {2y'} + ~ {y}

- sy(O) - y'(O) + 2s ~{y} - 2y(0) + ~{y}
s2~{y}
=O
=O
s2~ {y} - es - 2 + 2s ~{y} - 2e + ~ {y} = O

(s2 + 2s + 1) ~{y}
~{
es + 2e + 2
}_
y
es
(s
=e
+2
+ 1)2
2e
+ 1)2 +
(s
s +1- 1
2e + 2
+ ..,.---=
(s+1)2
(s+l)2
e+2
+ ..,.------,'"
(s+1)2'
= -s+l
Therefore,
To find e we let y(l)
2. Then 2
ee-1
+ (e + 2)e-1 = 2(e + 1)e-1
and e
e - l. Thus
y(t) = (e - l)e-t + (e + l)te-t.

32. Taking the Laplace transform of both sides of the differential equation and letting e
y' (O) we
obtain
~ {y"} + ~ {8y'} + ~ {20y} = O

s2~{y}
- y'(O) + 8s~{y}
s2~{y}
- e+8s~{y}
+ 20~{y}
=O
+20~{y}
=O
(i+8s+20)~{y}=e
~{
} _
y - s2 + 8s + 20 - (s + 4)2 + 4 .
Therefore,
y (t )
<..0-1 {
oL-
s+4
e)2
+4
To find e we let y'(;r) = O. Then O = y'(;r) = ee-4
1l'
= ce -4t
and e
sin 2 t.
= O. Thus, y(t) = O. (Since the
differential equation is homogeneous and both boundary conditions are O, we can see immediately
that y(t)
= O is a solution. We have shown that it is the only solution.)

322
Exercises
= -kx - f3x'. Now m = W/g = 4/32 = ~ slug, and 4 = 2k so
33. Recall from Section 5.1 that mx"
that k = 2 lb/ft. Thus, the differential equation is x"
x(O) = -3/2 and x'(O)
= O.
+ 7x' + 16x
= O. The initial conditions are
The Laplace transform of the differential equation is
82:t'{X} +
~8
+ 78:t'{X} + 21 + 16:t'{x} = O.
2
Solving for :t' {x} we obtain
:t' x _ -38/2 - 21/2 _

{ }-
82 + 78 + 16 -
8 + 7/2
7V15
-2 (8 + 7/2)2 + (V15/2)2
10
Vl5/2
(8 + 7/2)2 + (V15/2)2 .
Thus
3 -7t/2 cos-t
Vl5 - --e
7Vl5
x = --e
2
2
10
-7t/2
. Vl5
sm-t.
2
d2q
dq
dt2 + 20 dt + 200q = 150,
q(O) = q'(O) = O.
The Laplace transform of this equation is

150
82:t'{q} + 208:t'{q} + 200:t'{q} = -.
Solving for :t' {q} we obtain
:t'{ } _
150
q - 8(82 + 208 + 200)
3 1
4' ~ - 4'
8 + 10
3
10
(8 + 10)2 + 102 - 4' (8 + 10) 2 + 102 .
Thus
3 -lOt
3 -lOt .
q(t) = -3 - -e
cos lOt - -e
sm 10t
444
and
i(t) = q'(t) = 15e-10t sin 10t.

d2q
dq
2
Eo
dt2 + 2'\ dt + w q = L'
q(O) = q'(O) = O.

2
2
Eo 1
8 :t'{q} + 2'\8:t'{q} + w :t'{q} = L ~
or
(82 + 2'\8
+ w2):t'{q}
= ~o ~.
Solving for :t' {q} and using partial fractions we obtain
:t'{q} = Eo (1/W2
L
7.3
(1/w2)8 + 2'\/W2) = Eo (~_

82 + 2'\8
+ w2
323
Lw2
+ 2,\
82 + 2'\8
+ w2
) .
Exercises
For A
7.3
> w we write s2 + 2AS + w2
(s + A)2 - (A2 - w2), so (recalling that w2 = 1/ LC)
1
S+A
A)
,,;t'{q} = EaC ( ~ - (s + A)2 _ (A2 _ w2) - (s + A)2 _ (A2 _ w2)
Thus for A
> w,
q(t) = EaC(l-
For A
= w,
VA2
w2
sinh
VA2 -
w2 t) .
so
A)
s+
(S+A)2+(W2-A2)
= (s + A)2
s2 + 2A + w2
{q}
Thus for A
w2 t -
- (S+A)2+(W2-A2)
< w,
,,;t'
VA2 -
+ A)2 + (w2 - A2),
= (s
1
( ~-
,,;t'{q}=Ea
For A
e-At cosh
< w we write s2 + 2AS +w2
Thus for A
and
= Ea (l/A
Ea
1
L S(S+A)2
l/A
S+A
l/A ) = Ea (~ __ 1_ _
A )
(S+A)2
LA2 s S+A
(S+A)2
= w,
dq
R dt + C q = Ece
-kt
= O.
,q(O)

R ,,;t'{ q}
+ C ,,;t'{ q}
Ea s + k .
Solving for ,,;t' {q} we obtain
EaC
Ea/ R
,,;t'{q} = (s + k)(RCs + 1) = (s + k)(s + 1/ RC) .
When 1/ RC =1- k we have by partial fractions
::t { }
= Ea (1/ (1/RC - k) _ 1/(1/RC - k))

R
s + 1/RC
s +k
Ea
R
1
1/RC -
Thus
q(
When 1/ RC
=k
t)
EaC
(-kt
-ti Re)
kRC e - e
.
1_
we have
,,;t'{q} =
Ea
(s + k)2 .
324
(_1_ _
k
s +k
s + 1/RC
).
Exercises
7.3
Thus
q(t) = Ea te-kt = Ea te-tiRe.
37 . .:t{ (t - 1)OU(t _ 1)} = e~s
38. .:t{ e2-t OU(t - 2)}
39 .
= .:t {e-(t-2)
OU(t _ 2)}
-2s
= _e _
s+1
e-2s
2e-2s
.:t{tOU (t - 2)} = .:t {(t - 2)OU(t - 2) +20U (t - 2)} = -2 +-s
e-s
40. .:t { (3t + 1)OU(t - 1)} = 3 .:t { (t - 1)OU(t - 1)} + 4.:t { OU(t - 1)} = -2 +-
4e-s
41. .:t { cos 2t OU(t - 71')}
45 . .:t-1 L~-:s1}
46 .
= cos2 (t -
e-s
s(8+1)
48. .:t-1 { 2r2S
)}
s8-1
49. (c)
55.
56.
57.
se-7rS
.:t { cos 2 (t - 71') OU(t - 71')} = 82 + 4
= sin(t - 71') OU(t - 71')
.:t-1 {8:;:S~2}
47 .. .:t-1 {
= .:t-1
= .:t-1
%) OU(t _ %)
{e-S _ e-S}
8
8+1
= OU(t
_ 1) _
{- e-2s _ e-22S + e-2s }

8
8
s-1
50. (e)
51. (f)
e-(t-l)
OU(t - 1)
= _ OU(t _
2) - (t - 2) OU(t - 2) + -2ou (t - 2)
53. (a)
52. (b)
54. (d)
.:t {2 - 4 OU(t - 3)} = ~ _ ~e-3s

s
1 e-4s
e-5s
.:t {1 - OU(t - 4) + OU(t - 5)} = - - +8
S
s
.:t {t2 OU(t - 1) } = .:t { ((t - 1)2+ 2t - 1]OU(t - 1)} = .:t { ((t
325
1)2+ 2(t - 1) - 1]OU(t - 1)}
Exercises
7.3
2
2
= ( -+-+s3 s2
1)_s
e
s
1 e-2s
s2 -~--s-
59. 2{t-tOU(t-2)}=2{t-(t-2)OU(t-2)-2OU(t-2)}=
60. 2{sint-sint
61. 2{f(t)}=2{OU(t-a)-OU(t-b)}=--62. 2{f(t)}
OU(t-27f)} = 2{sint-sin(t-27f)OU(t-27f)}
e-as
e-bs
= 2{OU(t -1) + OU(t -2) + OU(t-3)
= s2+1
e-s
e-2s
2e-2s
e-27rs
- s2+1
e-3s
e-s
+ ...} = -;- + -s- + -s- + ... = ~ 1- e-S
63. The Laplace transforrn of the differential equation is

s2{y}
-y(O) +2{y}
5
s
= _e-s.
Solving for 2 {y} we obtain
2{ } = 5e-S
y
s(s+1)
= 5e-s [~ __
1_]
s+1
Thus
y = 5OU(t- 1) - 5e-(t-l)OU(t -1).
64. The Laplace transforrn of the differential equation is
s 2{y} - y(O) +2 {y}
Solving for 2{y}
= ~ - ~e-s.
s
we obtain
1
2{y}
= s(s
+ 1) -
2e-s
1
1
-8 [1
1]
s(s + 1) = ~ - s + 1 - 2e
~- s +1
Thus

s 2{y} - y(O) + 2 2{y} =
1
s-
? -
s +1
e-s-2-.
s
Solving for 2 {y} we obtain

2{y}
1
s2(s+2)
-8
-e
s+1
s2(s+1)
11 1 1
1
=-4~+2s2+4s+2-e
326
-s
[1 1 1 1
1
4~+2s2-4s+2'
1 ]
Exercses
7.3
Thus
y = -~ + ~t + ~e-2t - [~+ ~(t - 1) - ~e-2(t-l)]q(t
4 2
4
4 2
4
- 1).
52 .:t'{y} - 5y(O) - y'(O)
+ 4.:t'{y}
-s
e -.
5

1- 5
-s
.:t'{y} = 5(52+4) -e
1
5(52+4)
11
=4~-4
5
52+4
-2'
2
-s [1 1
1 5
52+4 -e
4~-452+4'
Thus
y = ~ - ~ cos 2t - ~ sin 2t - [~ - ~ cos 2(t - 1)] q(t - 1).
4 4
2
4 4
52.:t'{y}
- 5Y(O) - y'(O) + 4.:t'{y}
= e-27rS~1
5
Thus
y = cos 2t + [~sin (t - 27r)- ~ sin 2(t - 27r)]q (t - 27r).
68. The Laplace transform of the differental equation is
52.:t'{y}
e-s
- 5Y(O) - y'(O) - 5 [5.:t'{y} - y(O)] + 6.:t'{y}
= -.
5

1
.:t'{y} = e-s 5(5 _ 2)(5 - 3) + (5 - 2)(5 - 3)
= e-s
[~ ~ - ~ 5 ~ 2 + ~ 5 ~
3] -
5~ 2 + 5~
_ 1)
+ e3t
3.
Thus
y = [~_ ~e2(t-l)
6 2
+ ~e3(t-l)]q(t
3
_ e2t.

-7r5
52 .:t'{y} _ 5y(O) _ y'(O) +.:t'{y}
= _e
-27r5
e _.
5
.:t'{y} = = [~__
5
5_] _ e-27rs [~5 __ 52 5_]

+
+ 1
52 + 1
_1_.
52 + 1
327
Exercses 7.3
Thus
y
[1 - cos(t - 7r)]OU(t - -r) - [1 - cos(t - 27r)]OU(t - 27r) + sin t.
s2 ~{y}
_ sy(O) _ y'(O)
+ 4[s~{y}
_ y(O)]
+ 3~{y}
-28
-48
e _
s
-68
+ _e
++".
~{y} = ~ ~ _ ~ _1_ + ~ _1__ e-28 [~~ _ ~ _1_ + ~ _1_]

3s
2s+1
6s+3
3s
2s+1
+ ~ _1_] + e-68
_ e-48 [~~ _ ~ _1_

3s
2s+1
6s+3
+ ~ _1_]
[~~ _ ~ _1_
3s
2s+1
6s+3
6s+3
Thus
1 1 -t
y = 3" - 2e
- [~_
3
+ "61e -3t
~e-(t-4)
+ ~e-3(t-4)]OU(t
[1
1 -(t-2)
3" - 2e
_ 4)
71. Recall from Section 5.1 that mx" = -kx

so that k
+ "61e -3(t-2)]
(
OUt - 2)
+ [~3
+ f(t).
+ ~e-3(t-6)]OU(t
~e-(t-6)
2
Now m = W/g = 32/32 = 1 slug, and 32 = 2k
= 16 lb/ft. Thus, the differential equation is x" + 16x = f(t).
x(O) = O, x'(O)
= O.
20t,
f(t)
The initial conditions are
Also, since
f(t)
and 20t = 20(t - 5)
- 6).
+ 100 we can
={
O,
O:::; t
<5
t >: 5
write
= 20t - 20tOU(t - 5) = 20t - 20(t - 5)OU(t - 5) -100OU(t
- 5).

s2~{x} + 16~{x}
= 20 20 e-58 __ 100e-58.
s2
s2
S
Solving for ~ {x} we obtain
~{x}
20
_
20
e-58 _
100
e-58
- s2(s2 + 16) s2(s2 + 16)
s(s2 + 16)
=(~.2__~.
4 s2
16
s2
+ 16
)(1_e-58)_(25.~_25.
.5
s2
+ 16
)e-58.
Thus
x(t) = ~t - 156sin4t - [~(t - 5) - 156sin4(t - 5)] OU(t - 5) - [~5 - 245cos4(t
5
25
- 5)] OU(t - 5)
= t - 16 sin 4t - tOU (t - 5) + 16 sin4(t - 5) OU(t - 5) + "4 cos 4(t - 5) OU(t - 5).

328
Exercises 7.3
72. Recall from Section 5.1 that mx"
-kx + f(t).
= W/g = 32/32 = 1 slug, and 32 = 2k

+ 16x = f(t). The initial conditions are
Now m
so that k = 16 lb/ft. Thus, the differential equation is x"

x(O) = O, x' (O) = o. Also, since
< 27T
O:::;t
sin i,
f(t)
={
O,
t 2: 27T
and sin t = sin( t - 27T) we can write
f(t) = sin t - sin(t - 27T)OU(t - 27T).

82X{X}
+ 16x{x}
= -2--1
1
- -2--1e-27rS.
8
Solving for X {x} we obtain
VJ{ X }
oL.
= -1/15
82 + 16
e -27rS
- (82+16)(82+1)
- (82+16)(82+1)
+ 1/15 _ [-1/15 + 1/15] e-27rs.

82 + 1
82 + 16 82 + 1
Thus
x(t) = -~ sin4t + ~ sin t + ~ sin4(t - 27T)OU(t - 27T) - ~ sin(t - 27T)OU(t - 27T)

60
15
60
= { -lo sin 4t + A sin t,
15
O:::;t
< 27T
t 2: 27T.
O,
2.5 dq
dt
+ 12.5q
= 5OU(t
- 3).
8X{q}
+ 5x{q}
=-
e-3s.
Solving for X {q} we obtain
x{q}
e-3s = (~ . ~ _ ~ .
8(8+5)
5 8
_1_)
5 8+5
Thus
q(t) = ~OU(t - 3) 5
~e-5(t-3)
OU(t - 3).
74. The differcntial equation is

10 ~~ + 10q = 30et
329
30etOU(t - 1.5).
e-3s.
Exercises 7.3
3
3e1.5
8.;t'{q} - qO +.;t' {q} = 8 _ 1 - 8 _ 1.5 e-1.5s.
Solving for .;t'{q} we obtain
.;t'{q} = (qO _ ~) . _1_ + ~. _1_3e1.5
2
8+ 1 2 8- 1
(-2/5 +~)
e-1.55.
8 + 1 8 - 1.5
Thus
(t _ 37r)OU(t_
37r)
22'
= sin z } cos
~
di + 10i
1
8.;t'{i} + 10.;t'{i}
i(O) = O.
8e-31rs/2
= 82 + 1 + 82 + 1 .
Solving for .;t'{i} we obtain

.;t'{i} =
1
(82 + 1)(8 + 10)
__ 1_ (_1
- 101 8 + 10
e-31rs/2
(82 + 1)(8 + 10)
8_ + ~)
+ _1_ (~
+ ~
+ _1_)
e-31rs/2
82 + 1 82 + 1
101 8 + 10 82 + 1 82 + 1
.
Thus
1
i(t) = 101 (e-10t-cost+10sint)

+ 1~1 [_10e-10(t-31r/2)
(b)
+ lOcos (t - 3;) + sin (t _ 3;) ]OU(t _ 3;).
0'2f
--+--~""'---"--~..__.=::::--,._~._.__;:",.,_~::::----"'~~~--..__
1
3~
t
6
-0.2
The mximum
value of i(t) is approximately 0.1 at t = 1.7, the minimum s approximately
-0.1 at 4.7.
76.
(a) The differential equation is

50 ~~ + 0.~1 q = Eo[OU(t -1) -OU(t - 3)],
or
330
q(O) = O
Exercises
50 ~;
+ 100q = Eo[VU(t -
q(O) = O.
1) -VU(t - 3)],
+ 100:t{q} = Eo
50s:t{q}
Ge-s
- ~e-3S).
Solving for :t {q} we obtain
:t
{q}-
_ Eo [ e-s
50 s(s+2)
e-3s
s(s+2)
] _ Eo [~ (~ __ 1 ) e-s _ ~ (~ __ 1 ) e-3S]
- 50 2 s s+2
2 s s+2
.
Thus
(b)
The maximum value of q(t) is approximately 1 at t = 3.

d4y
El dx4 = wo[l - VU(x - Lj2)].
Taking the Laplace transform of both sides and using y(O)
= y'(O) = O we obtain
s4:t {y} _ sy" (O) - ylll(O) = Wo ~

El s
(1 _ e-Ls/2)
Letting y"(O) = Cl and y"'(O) = C2 we have
:t{y} = 2 +
s3
C2
s4
+ Wo ~
El
s5
(1 _ e-Ls/2)
so that
1
= -ClX
y(X)
To find Cl and
C2
2 + -C2X
1 3 + -1 -Wo [4x
6
24 El
(x
L)2 4 VU( x - -L)2 ] .
- -
we compute
y
"(x) = Cl
Wo [2x + C2X + "21 El
( x -"2L)2
1/(
u x -
and
y'" (x) = C2 +
~~ [x - (x -
~)vu ( x -
331
~) ] .
L)]
"2
7.3
Exercises 7.3
Then y" (L)
=;
v" (L)
= Oyields the system.

Cl
+ C2L + -1 -wo [2L - (L)2]

2 El
wo (L)
Solving for
and
Cl
we obtain
C2
Cl
"2
+ El
C2
wOL2 = O
= Cl + C2L + -3S -El
= C2
+ '2 El
1 woL
~wOL2/ El and
C2
=-
= O.
iwoL / E l. Thus
wo (1 2 2
1 3
1 4
1 (
y(x) = El 16L x - 12Lx + 24 x - 24 x
-"2L)4
(
L))
/Jx - "2
d4y
El dx4 = wo[/J(x- L/3) - /J(x- 2L/3)].

Letting y"(O)
= Cl
= y' (O) = O we obtain
s4:t'{y} _ sy"(O) _ ylll(O) = wo ~ (e-Ls/3

El s
and ylll(O) = C2 we have
e-2Ls/3).
:t'{y} = ~ + C2 + wo ~ (e-Ls/3 _ e-2Ls/3)

s3 s4 El s5
so that
1
y(x) = '2ClX
To find
Cl
and
C2
1 3 + 24
1 El
wo [( x - '3
L)4 /Jx
( - '3
L) - ( x - '3
2L)4 /Jx
( - '3
2L)]
+ 6C2X
we compute
and
Then y" (L)
= u" (L) = O yields the system

Cl
+ c2L + -1 -Wo [(2L)2

2 El
- (L)2]
-
Cl
Wo [2L
C2
Solving for q and
C2
we obtain
Cl
Wo (112L 2 x 2 - lS1Lx
y(x) = El
=
3
+ El
twoL2/EI
wOL2
+ C2L + -1 -6 El
L]
'3- '3
and
C2
=O
1 woL
C2
+ '3 El = O.
-~woL/EI.
Thus
1 [( x - '3
L)4 /Jx
( - '3
L) - ( x - 3
2L)4 /Jx
( - '3
2L)])
+ 24
332
Exercises
El d4y = 2wo
dx4
L
[!:_
x +(x - !:) a (x _ !:)] .
2
2
2

s4 .:t {y} - sy" (O) _ ylll (O)
= Cl
Letting y"(O)
= y'(O) = Owe obtain
= 2wo [.!:__ _ ~ + ~ e-LS/2] .

ElL
2s
s2
s2
= C2 we have
and y"'(O)
UJ{ } _--+-+-Cl
C2
2wQ [ L
1
1 -LS/2]
----+-e
s3
s4
ElL
2s5
s6
s6
""-Y
so that
y(X) = ~Clx2
2
=
To find
Cl
and
+ ~C2x3 + 2wQ ['!:__x4__ 1_x5 +

El L 48
120
'21ClX 2 + 6'1C2X3 + 60ElWQ L [5L

2X 4 -
!:)
2
_1_ (X -
120
L)
+ ( X - "2
5 (i'1
+ C2X + 60~~L [30LX2 - 20x3 + 20 (X -~)
=u.
(X _
lJij
!:)]
2
"2L)] .
X-
we compute
C2
y"(x) = Cl
o (X - ~)]
and
Y /1' (X)=C2+60ElL WQ
Then y" (L) = u" (L)
ci
= Oyields the
[60Lx-60x
Cl
and
y(x)
+ c2L + 60El L 30L - 20L + '2L

Wo
Solving for
C2
we obtain
wQL2
= 48ElX
2-
C2
+ 60El L[60L
Cl
= woL2/24El
woL
24El
+60
L)
x-"2
2 o
( x-"2'
L)]
system
[3
WQ
WQ
+ 60ElL
3] =
- 60L
and
[5L
2X
C2
5woL2
Cl
+ C2L + 24El
woL
+ 15L ] =
C2
+ 4El
= -woL/4El.
4 - X5'(+
X-
=O
= O.
Thus
L)5
"2
(
L)]
o X - "2

d4y
El dx4
= wo[l - o (x - L/2)J.
Letting y"(O)
= ci
S4.:t{y} _ sy"(O) - y"'(O) = Wo ~

El s
and y'''(O) = C2 we have
.:t{y} =
3. +
s3
C2
s4
+ Wo ~
El s5
333
= y'(O) = Owe obtain
(1 _ e-Ls/2).
(1- e-Ls/2)
7.3
Exercises 7.3
so that
1 2 + -C2X
1 3 + -1 -Wo [4X y(X) = -C1X
2
6
24 El
To find
Cl
and
C2
= y"(L)
1
-clL
2
ci
( - "2
L)] .
+ C2X+ "21WO[2
El X - ( X -"2L)2 llUx
= Oyields the system
2 +-C2L
1
3 +-1 Wo [4L
6
24 El
C
Solving for
-L)4 llU( X - -L)] .
we compute
y " (x) = Cl
Then y(L)
(X-
and
C2
we obtain
4]
- (L)
2
1
2
= -clL
Wo [2L -"2(L)
+ C2L + "21 El
Cl
1~8
woL2/EI and
2 +-C2L
1
3 +--L5wo
2]
C2
128EI
4 =0
3wo L 2 = O.
= ci + C2L + 8EI
= -lS278
Wo (9
2 2
19 3
1 4
1 (
y(x) = El 256L x - 256Lx + 24x - 24 x
wo~/EI.
-"2L)4
r711(
'u,
Thus
-"2L)) .
81. In order to apply Theorem 7.7 we need the function to have the form f(t-a)llU(t-a).
To accomplish
this rewrite the functions given in the forms shown below.

(a) 2t
+ 1 = 2(t -
1 + 1) + 1 = 2 (t - 1) + 3
(b) et = et-s+s = et-S

(c) cost=-cos(t-7r)
(d) t2-3t=(t-2)2+(t-2)-2
82. (a) Prom Theorem 7.6 we have :t{tekti}
= 1/(8 -
ki)2. Then, using Euler's formula,
:t {tekti} = :t {t cos kt + it sin kt} = :t {t cos kt} + i :t {t sin kt}

=
(8 - ki)2
(8+ki)2
82_k2
.2k8
=
+ Z ---=----,,-,-;:(82 + k2)2
(82 + k2)2
(82 + k2)2 .
Equating real and imaginary parts we have

82 _
:t{tcoskt}
= (82
k2
+ k2)2
and
2k8
:t{t sin kt} = (82 + k2)2'
(b) The Laplace transform of the differential equation is
82 :t {x} + w2 :t {x} =
Solving for :t{x}
we obtain :t{x}
2 8
8
+w
2'
= 8/(82 +w2)2 Thus x

334
= (1/2w)tsinwt.
Exercises
Exercises 7.4
8.::tt {
9.
ie
t}
=s3(s-1)2
s-l
(s+1)[(s-1)2+1J
::t{e-t*etcost}=
(s-2)(s2+1)
10. ::t{e2t*sint}
11. ::t{fot erdT}
= ~::t{et} = S(s~ 1)
12. ::t{t
COSTdT} = ~::t{cost}
t
13. ::t{fo CrcosTdT
14. ::t{l
TSinTdT}
15. ::t{fot Tet-r dT}
= S(S2S+ 1)
s2 ~ 1
1
-t
1
S + 1
= ~ ::t{e
cost} = ~ (s+1)2+1
= ~ ::t{tsint}
= ::t{t}::t{et}
1(
=~ -
7.4
dI)
+1
s(s2+2s+2)
1 - 2s
_
2
ds s2 + 1 = -~ (s2 + 1)2 - (s2 + 1)2
S2(Sl_1)
335
Exercises
7.4
16. ::t{ rtsinrCOS(t-r)dr}
=.:t'{sint}::t{cost}
Jo
17.
::t{t
r Sinrdr}
= -!!:_::t{ rt Sinrdr}
Jo
= - :S .:t'{t
18. ::t{t fat re-Tdr}

19. (a) ::t-1{
(b) ::t-1{
(e) ::t-1{
s(s - 1)
21
s3(s - 1)
re-Tdr}
{l/(S
} = ::t-1
s (s - 1)
Jo
ds
-dd
(~-2_+11_) = s2(s2+1)
3s2 + 1 2
-1)} = Jort eTdr
(S: 1)2) = S2~;: ~)3

= et-1
rt(eT_1)dr=et_t_1
Jo
=::t_1{1/S2(S-1)}
s
20. (a) The result in (4) is
S ss
= - :S
}=::t_1{1/S(S-1)}=
s
}
s
(s2+1)
::t-1{F(s)G(s)}
=f
(t(eT-r_1)dr=et_~t2_t_1
Jo
* g, so identify
2k3
F(s) = (s2 + k2)2
and
4s
G(s) = s2 + k2 .
Then
f(t)
= sinkt
- ktcoskt
g(t) = 4coskt
and
so
::t-1 {(S2S:3;2)2 } = ::t-1 {F(s)G(s)}
=f
*9 =4
= 4 fat (sin kr - kr cos kr) cos k(t
f(r)g(t
- r)dt
- r)dr.
Using a CAS to evaluate the integral we get
::t
-l{
(s2Sk+3S}.
k2)3 . = tsm kt - kt cos kt.
(b) Observe from part (a) that

::t{t(sinkt - ktcoskt)}
and from Theorem 7.8 that ::t{tf(t)}
2k3/(s2+k2)2, so
-F'(s).
336
Sk3s
(s2 + k2)3'
We saw in (5) that ::t{sinkt - ktcoskt}
Exercses 7.4
22.
:f{f(t)}
23.
:f{f(t)}
25.
:f{f(t)}
26. :f{f(t)}
12
(a e-stdt
1 - e- as Jo
1-
= 8 (1 +1e-as )
e- bs Jo(b~bte-stdt = ~8 (b18 _ -be s 1-)

- 1
io
1f
1 - e-1fS o
1 2
1- e- 1fS
e-st sintdt = --. 1
+1
82
+ e-1fsj2 = -- 1 coth- 7f8

- e-1fsj2
82 + 1
2
e1fsj2
e1fsj2
r e-st sin t dt = -21+ 1 . 1- 1e-
Jo
1fS

28
8:f{y} + :f{y} = (82 + 1)2'

Solving for :f {y} we obtain
28
:f{y} = (8 + 1)(82 + 1)2 =
Thus
1
Y (t) = - - e- t
2
=
1.1
l.
l.
- - sm t + - cos t + - (sin t - t cos t) + -2 t sm t

2
+ -l' cost
1t
--e-
1
- -tcost
2
+ -l.r sm t.
2

2(8 - 1)
8:f{y} - :f{y} = ((82 _ 1)2 + 1)2 .

Solving for :f {y} we obtain
2
:f{y} = ((8 _ 1)2 + 1)2 .

Thus
y = et sin t - tet cos t.

822{y}
-2 8 + 1 - 2 82 + 1 + 2 82 + 1 + (82 + 1)2 + (82 + 1)2'
- 8y(O) - y'(O) + 92{y}
337
= ~9'
8
Exercises 7.4
Letting y(O)
= 2 and
y'(O)
;t
= 5 and
solving for ;t{y}
we obtain
= 253 + 552 + 195 - 45 = ~
{y}
(52+9)2
+ _5_
+
5
52+9
(52+9)2'
52+9
Thus
5 ,
1,
Y = 2 cos 3t + "3 sin 3t + i3t sin 3t,

52;t{y}
- 5Y(0) -y'(O) +;t{y}
= -2-1-,
5 +1
Solving for ;t {y} we obtain

53 - 52 + 5
.i'{y} = (52 + 1)2 = 52 + 1 - 52 + 1 + (52 + 1)2 '

Thus
y
= cos t - 2' sin t - 2' t cos t,

52;t{y}
- 5y(0) - y'(O) + 16;t{y}
=;t {cos 4t - cos 4t Ilfl(t - 7f)}
or
= 1+ 25
(52+ 16);t{y}
+ 16
-e-7fS;t{cos4(t+7f)}
= 1 + 52 + 16 - e-7fS ;t{cos4t}
= 1+
52 + 16
e-7fS
52 + 16
Thus
;t{ }
1
5
y =52+16+(52+16)2-(52+16)2e
-7fS
and
y = ~sin 4t + ~ t sin 4t - ~ (t - 7f)sin 4 (t - 7f)OU (t - 7f),
4
8
8
52 ;t {y} - 5y(0) - y' (O) +;t {y} =;t { 1 -Ilfl (t - ~) + sin t OU( t - ~) }
or
(52 + 1) ;t{y}
= 5+ ~ - ~
e-7fs/2 + e-7fS/2;t{ sin (t +~)}
= 5 + ~ - ~e-7fs/2 + e-7fs/2 ;t{cost}

5
+ ~ _ ~ e-7fs/2 + _5_ e-7fs/2

5
52 + 1
338
Exercses 7.4
Thus
.,'i'{ }
s
s2
+ 1 + s(s2 + 1)
S
= s2
+ 1 + ~-
1 (1~ -
= ~-
s
s2
+1-
S)
s2
1
- s(s2
+1
+ 1)
(1~ -
-1rs/2
e-1fs/2
S)
s2
+1
s
(s2
(s2
-1fs/2
+ 1)2 e
+ 1)2
+
e-1fs/2
S
(s2
+ 1)2 e
-1fs/2
-1fs/2
and
(t - ~) ] VU (t - ~) + ~ (t - ~) sin (t - ~) VU (t - ~)
(1 - sin t) VU (t - ~) + ~ (t - ~) cos t VU (t - ~) .
y = 1 - [1 - cos
= 1-

2
s .,'i'{y}
2s
+ .,'i'{y}
= (s2
+ 1) + (s2 + 1) 2 .
2s
50
Thus
~{y}
(s2
+ 1)2 + (s2 + 1)3
-50
and, using Problem 20,

Y
34. (a)
= - (sin t - t cos t) + - (t sin t - t2 cos t).
y
1
-1
(b)
y
4
2
1
-2
-4
339
Exercises 7.4
35. The Laplace transform of the given .equation is
~{f}
Solving for ~{f}
+~{t}~{f}
1
= -2--.
we obtain ~{f}
Thus, f(t)
+1
= ~{t}.
= sin t.
36. The Laplace transform of the given equation is
= ~{2t}
~{f}
- 4~{sint}~{f}.
Solving for ~ {f} we obtain
= 2s2 + 2
= ~ 2_
s2(s2+5)
{f}
5s2
+ _8_
V5
5V5s2+5
Thus
28.
= -t +
f (t)
~ sin Vs t.
5v5
Solving for ~ {f} we obtain
~{
}_
_ ~ _1_ + ~
s2
1
+ ~ 2 _ ~ _1_
4 (s - 1)2 4 (s - 1)3 88+ 1 .
f - (s - 1)3(s + 1) - 8 s - 1
Thus
f(t)
1 t 3 t 12 t
= -e
+ -te + -t e --el_t
8
~{f}
Solving for ~{f}
+ 2~{cos
t}~{f}
= 4~{ e-t} + ~{sin t}.
we obtain
~{
} _ 4s2 + s + 5 __ 4_ _
7
+4
2
(s + 1)3 - S + 1 (s + 1)2
(s + 1)3.
f -
Thus
~{f}
Solving for ~{f}
we obtain ~{f}
+~{1}~{f}
1
= --.
s+l
= ~{1}.
Thus, f(t) = e-t
340
Exercses 7.4
Solving for .:t'{f} we obtain
.:t'{f} - _-
82
+1
+ _-1
82 + 1 '
Thus
f(t) = cost + sin t.

41. The Laplace transforrn of the given equation is
.:t'{f} =.:t' {1}+.:t' {t} - ~ .:t'{t3}.:t'{f},

+ 1)
.:t'{f} =
82(8
84 +
16 =
83
84 + 16
82
84 + 16
'
Thus
f (t)
= cos
J2 t cosh .J2 t +
1M(sin .J2 t cosh
.J2 t + cos .J2 t sinh .J2 t) ,
2v2

82 -
.:t'{f}
1 1
= 2T = 2 82
1 31
12 84
'
Thus
1
2
1 3
.
12
f(t) = -t - -t
8 .:t' {y}
- y ( O) = .:t' { 1} - .:t' { sin t} - .:t'{ 1} .:t'{y} .
.:t'
= 83 - 82
{y}
+8 =
8(82+1)2
_1_
+~
82+1
28
2(82+1)2'
Thus
.
1,
Y = sin t - 2 t sin t.
8.:t'{y} - y(O) + 6.:t'{y} + 9.:t'{1}.:t'{y}

341
.:t'{1}.
Exercises 7.4
Solving for ~{f}
we obtain ~{y}
= (
8+3
)2' Thus, y = te-3t.

i
30

di
0.1 dt
+ 3i + 0.05 lo
20
i(T)dT = 100[)L(t -1) -)L (t - 2)]
10
or
si
.5 3
= 1000[)L(t - 1) -)L (t - 2)],
di + 30i + 200 (t i(T)dT
Jo
-10
-20
where i(O) = O. The Laplace
transform of the differential equation is
.
200
8~{2} - y(O) + 30~{i} + -~{i}
1000
= _(e-S
-3 O
_ e-2S).
Solving for ~ {i} we obtain

~{i}
= 1000e-s
-1000e-2s
82+308+200
~)(e-S
8+108+20
_ e-2s).
= (~-
Thus
i(t) = 100(e-10(t-l) - e-20(t-l)) )L(t - 1) - 100(e-10(t-2) - e-20(t-2))
i

di
1
0.005 dt + i + 0.02
)L(t - 2).
lo i(T)dT
t
= 100[t - (t -l))L(t
r------.,
-1)]
1.5
or
d'
t
__:.+ 200i + 10,000 { i(T)dT
8t
Jo
= 20,000[t - (t - 1) )L(t - 1)],

0.5
where i(O) = O. The Laplace transforrn of the differential equation is

8~{i}
+ 200~{i} + 10,000 ~{i}

8
0.5
1.5
= 20,000( 12- 12e-S).

8
Solving for ~ {i} we obtain

~{i}
20,000 (1 _ e-S) = [~_

2
_
200
] (1 _ e-S)
8(8 + 100)2
8 8 + 100 (8 + 100)2
.
Thus
i(t) = 2 - 2e-100t - 200te-100t
- 2)L(t - 1) + 2e-100(t-l) UiL(t- 1) + 200(t - 1)e-100(t-l) '11 (t - 1).
342
Exercises
di
dt
where i(O)
7.4
i
1
+ i = E(t),
= O, The Laplace transform of this

4
equation is
s:t{i}
-0,5
+ :t{i}
= ~{E(t)},
-1
From Problem 21 we have

:t{E(t)}
1- e-s
= s (1 + e-S )'
Thus
(s
+ 1) :t{i}
1 -e -s
s + e- S)
= (1
and
1 -e -s
~ = s(s + 1)(1 +
:t{"}
= (~ __
1_) (1_ e-S)(l
s+ 1
s
=
e-S)
1 -e -s
s(s+l)l+e-s
( -1s - --s+l1)
(1 - 2e -s
_ e-s
+ 2e -2s
+ e-2s
_ e-3s
- 2e -3s
+ e-4s
+ 2e -4s
_ " ,)
- , , , ),
Therefore
i(t) = [1- 2OU(t -1)
- [e-t
+ 2OU(t - 2) -
+ 2e-(t-l)
+ 2e-(t-3)
2OU(t -'3)
OU(t - 1) - 2e-(t-2)
OU(t- 3) - 2e-(t-4)
+ 2OU(t - 4) -" ,]
OU(t - 2)
OU(t - 4)
00
= 1- e-t +2 2:::(-lt(1-
e-(t-n))OU(t
n=l
343
- n).
+ ... ]
Exercises 7.4
di .
dt + 2 = E(t),
1
0.5
where i(O) = O. The Laplace transforrn of this

equation is
-0.5
+ :t{i}
s:t{i}
:t{E(t)}.
-1
Frorn Problern 23 we have
:t{E(t)}
= -1(1
- - - 1)
= -1 - -1 -. 1
s s eS - 1
s2 s eS - 1
Thus
(s + 1) :t{i}
and
:t{"}
2
= s2(s + 1)
Therefore
1 1
( ---+-s2 S
1 1 1
- --s2 s eS - 1
=-
1
1
s(s+1)es-1
1) - (1
---- 1) (-S
e
S+1
s s +1
i(t) = (t -1 + e-t) - (1- e-(t-l))r7U(t

- (1- e-(t-3))r7U(t
=
(t - 1+ e-t) -
+e
-2s
+e
-38 +e -48
-1) - (1- e-(t-2))r7U(t
- 3) - (1- e-(t-4))r7U(t
- .... )
- 2)
- 4) - ...
00
L (1- e-(t-n))r7U
(t - n).
n=l
49. The differential equation is x"
+ 2x' + lOx = 20f(t),
where f(t) is the rneander function with a
Using the initial conditions x(O) = x'(O) = Oand taking the Laplace transforrn we obtain
(S2 + 2s + 10) :t{x(t)}
= 20 (1 _ e-7rs)__ l_
1 + e-7rS
20
= -(1 - 2e-7r8+ 2e-27rs_ 2e-37rs+ ... )
344
= tt .
Exercses 7.4
Then
.:t'{ (t)}
X
. 20
40
~ ( l)n -n7rS
= 5(52+ 25 + 10) + 5(52+ 25+ 10) ~l e
2
=~-
25 + 4
~ ( )n [4
45+ 8 ] -n7rS
52+25+1O+~
-1
~- 52+25+10 e
=~_2(5+1)+2+4f.
5 (5+1)2+9
(_l)n[~_
(5+1)+1]e-n7rS
(5+1)2+9
11=1
and
x(t) = 2 (1- e-tcos3t
+ 4~(-l)n[l-
- ~e-tsn3t)
e-(t-n7r)
cos3(t - mr)
- ~e-(t-n7J") sn3(t - mr)] lJU(t- mr).

The graph of x(t) on the interval [O,2n) is shown below.
x
-3
50. The differential equation is x"
+ 2x' + x
= 5f(t),
where f(t) is the square wave functon with a = tt .
Using the initial conditions x(O) = x'(O) = O and taking the Laplace transform, we obtain
(52+ 25 + 1).:t'{x(t)} = ~
5
1 + e-7rS
~(1 _ e-7rS + e-27rs _ e-37rs + e-47rs _ ... )

5
Then
.:t'{x(t)} =
5
f (_l)"e-n7rs
5(5 + 1)2 n=O
= 5 f (-1)" (~ __ 1_ _
n=O
5+ 1
1 ) e-n7rs
(5 + 1)2
and
00
x(t) = 5
L (-1)"(1
e-(t-n7r) -
n=O
345
(t - nn)e-(t-n7r))
lJU(t- nn).
Exercises
7.4
The graph of x(t) on the interval [O,4Jr) is shown below.

x
-5
51. f(t)
= -~
.,;t'-l{ ~(ln(s
ds
3)- ln(s + l)l} = _~t .,;t'-l {_ls - 3
l_} = _~t (
+1
t -
e-t)
52. By definition, t *q(t - a) = IJ(t - 7)q(7 - a) dr . We consider separately the cases when a
and when a
> t. When a < t, then

= (t (t _ 7) dr = _ (t - 7) 2 It = ~ (t - a f
(t (t _ 7) q (7 _ a) dr
Jo
When a
> t, then q (7 - a)
l;
7 < t < a and
= O since
(t - 7) q (7 - a) dr = O.
Therefore
hq(t-a)
> a = ~(t-a)2q(t-a).
t <a 2
{i(t-a)2,
O,
53. First method: By the definition of the Laplace transform and integration by parts we have
{l f (7)
.,;t'
dr }
1000 e =st (l f (7) d7)
dt
= _ e-st (t f(7) dr 100 +~ (00 e-st f(t) dt
s Jo
= ~ (00 e-st f(t)
s Jo
dt = F(s) .
Jo
Second Method: Let g(t) = Id f(7) dr ; then g'(t) = f(t) and g(O)
= O. By Theorem
.,;t'{g'(t)} = s .,;t'{g(t)} - g(O) = s .,;t'{g(t)},

so
and
.,;t'{f(t)} = s.,;t'
.,;t'{l
f(7) d7}
{la
f(7) d7} ,
= ~.,;t'{f(t)}
346
F;S) .
7.8
<t
Exercses
54. Let u
7.4
= t - T so that du = dr and
* 9 = 10t f(T)g(t
- T) dr = -
iD f(t
- u)g(u)du = 9 * f.
55. (a) Using Theorem 7.8, the Laplace transforrn of the differential equation is
d
- ds [s2y - sy(O) - y'(O)] + sY - y(O) + ds [sY - y(O)] + nY
= -_
dx
[s2y] + sY + - [sY] + nY
ds
=:
(s - s2)
y) - 2sY + sY + s(:s y) + y + nY
(!
y) + (1 + n - s)Y = O.
Separating variables, we find
dY = 1 + n-s ds = (_n __
y
s2 - S
S - 1
In Y = n ln(s ~ 1) - (1 + n) In s
Y
_
- CI
1 + n) ds
+c
(s - l)n
sl+n
Since the differential equation is homogeneous, any constant multiple of a solution will still
be a solution, so for convenience we take
el
= 1. The following polynornials are solutions of
Laguerre's differential equation:
n=O:
n = 1:
n = 2:
n
= 3:
n = 4:'
(b) Letting f(t)
= tne-t we note that f(k)(O)
= O for k = 1, 2, 3, ... ,
347
n - 1 and f(n)(o) = n!.
Exercises 7.4
Now, by the first translation theorem,
where Y
=::t {Ln(t)}.
Thus
56. (a) The output for the first three lines of the prograrn are
9y[t] + 6y'[t] + y"[t] == t sin[t]

1 - 28 + 9Y
Y-->-
+ 8 Y + 6(-2 + 8Y)
==
28
(1 + 82)2
-11 - 48 - 2282 - 483 - 1184 - 285)

(1 + 82)2(9 + 68 + 82)
The fourth line is the same as the third line with Y
-->
removed. The finalline of output shows
a solution involving complex coefficients of eit and e:". To get the solution in more standard
form write the last line as two lines:
euler = {E~(I t)-> Cos[t]+1
InverseLaplaceTransform
Sin[t], E~(-I t)-> Cos[t]-I
Sin[t]}
[Y,s,t] j .euler j jExpand
We see that the solution is

487
247)
y(t) = ( 250 + 5t e-3t + 250 (13 cos t - 15t cos t - 9 SIn t + 20t sin t) .
(b) The solution is
348
Exercises
7.5
(c) The solution is

q( t)
= 1 - cos t + (6 - 6 cos t) q,(t - 37f)- (4 + 4 cos t) q,(t - 7f).

q
Exercises 7.5
l. The Laplace transform of the differential equation yields
.:t'{y} = ~3e-2S
8-
so that
y
= e3(t-2)q,(t - 2).
2. The Laplace transform of the differential equation yields

2
e-s
8+1
8+1
.:t'{y} = -+so that

y = 2e-t
+ e-(t-l)q, (t -
1).
.:t'{y} = 82~1 (1+e-2TfS)

so that
= sin t + sin t q,(t - 27f).
.:t'{ } = ~
y
82
+ 16
e-2Tfs
so that
y
.
41 sm4(t
- 27f)q,(t - 27f).
.:t'{y} = _1_ (e-Tfs/2 + e-3TfS/2)

82
+1
349
Exercises 7.5
so that
y
= sin (t - ~)OU (t - ~) + sin (t - 3;)OU (t - 321T)

= - cos t
OU(t - ~) + cos t OU(t - ~) .
.s:'{y} = _8_ + _1_(e-271'S

82
+1
82
+1
+ e-471'S)
so that
y
= cos t + sin t [OU(t - 21T) + OU(t - 41T)J .
.s:'{y} =
1 (1 + e-S) = [~~_ ~ _1_] (1 + e-S)

82+ 28
2 8 2 8+ 2
so that
y
1 -2t + [1
1 -2(t-l)]IilI(
= -1 - -e
- - -e
'u t - 1).
2
.s:' y =
{}
8+ 1 +
1
e-2s = ~_1 __
82(8-2)
8(8-2)
48-2
~ ~ _ ~ ~ + [~ _1 __
48
282
28-2
~~] e-2s
28
so that
y = ~e2t _ ~ _ ~t + [~e2(t-2) - ~] OU(t - 2).
4
4 2
2
2
.s:'{ }
1
y = (8+2)2+1e
-271's
so that
y = e-2(t-271') sin tOU(t - 21T).
.s:'{y} =
(8:
1)2e-s
so that
y = (t - l)e-(t-l)OU(t
- 1).

4+8
e-71's + e-371's
.s:'{y} = 82+48+13 + 82+48+13
2
3
8 +2
3
(-71'S
+
+- 1
e
+e -371'S)
3 (8 + 2)2+ 32 (8 + 2)2+ 32 3 (8 + 2)2+ 32
=-
350
Exercses 7.5
so that
y
= ~e-2t sin 3t + e-2t cos 3t + ~e-2(t-7l")sin 3(t - 71fU (t - ?T)

3
+ ~e-2(t-37l")sin3(t
- 3?T)OU(t - 3?T).
3
1
e-2s
+ e-4s
.z:{y} = (s _ 1)2(s - 6) + (s - l)(s - 6)

1
= - 25
s- 1-
'5 (s -
1 1
1)2 + 25 s - 6
[1
+ -'5 s -
1 1 ] (-2S
1 + '5 s _ 6 e
so that
y
_~et
25
_ ~tet + ~e6t
5
+ [_~et-4
25
+ [_~et-2
+ ~e6(t-2)] OU(t - 2)
+ ~e6(t-4)] OU(t - 4).
.z:{}
1 2 "(O)
1 3! /1'(0)
1 Po 3!
y ='2s3Y
+6"s4Y
+6"Els4e
so that
-Ls/2
2'L) 3 OU( x - 2'L) .
1 /1 ) 2 1 /1,( 3 1 Po (
y = '2Y (O x + 6"Y O)x + 6" El X Using y/l(L) = O and ylll(L) = Owe obtain
1::)3 OU(x _ 1::)
y = ~ PoL x2 _ ~ Po x3 + ~ Po (x _
4 El
6 El
6 El
2
Po (!:"x2 _ ~x3)
O < x < !:..
El 4
6
' 2
- { POL2 (~x _ !:___)
4El 2
12'
1:: < x < L.

2 -
14. From Problem 13 we know that

1 /1 ) 2 1 111 ) 3 1 Po (
L ) 3 OU(
L)
y = '2y (O x + 6"Y (O x + 6" El X - 2'
x - 2'
Using y(L)
= Oand
y'(L)
= Owe obtain
1::)3 ou(x _ 1::)
y = ~ PoL x2 _ 2._ Po x3 + ~ Po (x _
16 El
12 El
6 El
2
O<
-x
L
-<x<L.
2 351
<-2
+e
-4s)
Exercises 7.5
1
~{Y}=82+w2
so that y(t) = sinwt. Note that y'(O) = 1, even though the initial condition was y'(O) = O.
Exercises 7.6
1. Taking the Laplace transform of the system gives
8~{X}
+ ~{y}
= -~{x}
8~{Y} -1 = 2~{x}
so that
1
~{x} = (8 -1)(8 + 2) = 3" -8---1 - 3" -8-+-2

and
Then
x
1
_et
3
1
_e-2t
3
and
2
3
1
+ _e2t
Y = _et

1
8~{x}-1=2~{y}+
8-1
8~{Y} - 1 = 8~{x}
- 2"
8
so that
~
{y}
83 + 782 - 8 + 1
8(8 - 1)(82 - 16)
1 1
8
1
173 1
53 1
16 -; - 158 - 1 + % 8 - 4 - 160 8 + 4
and
1
8 t
y = - - -e
16 15
173 4t
+ -e
96
53
-e
160
-4t
Then
x
1,
sy
st1 = st lIt
173 4t
15e + 192e
8~{X}
+ 1 = ~{x} - 2~{y}
8~{Y} - 2 = 5~{x}
352
-~{y}
53
-4t
+ 320e
Exercises 7.6
so that
-8 -
3 s2 + 9
::t{x} = s2 + 9 = - s2 + 9 and
5 .
3
x = - cos 3t - - sm 3t.
Then
1
1
223
7 .
y = -x - -x = 2cos3t - - sm3t.
1
s
(8 + 3) ::t{x} + s::t{y} = (8 -l)::t{x}
+ (8 -l)::t{y}
= 8 -1
so that
5s - 1
11
::t{y} = 38(S _ 1)2 = -3
~+ 3 s -
3 (s -
1+
1
1)2
and
::t{x} _ 1 - 28
11 1 1
1
1
- 3S(8 - 1)2 - 3 ~ - 3 s - 1 - 3 (s - 1)2 .
Then
X
1 t
= -lIt - -e - -te
and
3
Y = --
lIt
4 t
+ -e + -te.
3

1
s
(2s - 2) ::t{x} + 8::t {y} = -
(s - 3)::t{x} + (8 - 3) ::t{y} = -
so that
-s - 3
11
::t{x} = s(8-2)(8-3)=-2~+2s-2and
38 - 1
8-3
11
::t{y} = S(8 _ 2)(s - 3) = -6
~- 2 s -
+ '3 s -
3'
Then
x
5 2t
3t
1 5 2t 8 3t
= -- 1 + -e
- 2e
and y = -- - -e +-e
2
(8 + 1)::t{x} - (8 - l)::t{y} = -1
8::t{X} + (s+2)::t{y}
353
=1
Exercises 7.6
so that
:L{ } _
Y
8 + 1/2 _
8 + 1/2
82 + 8 + 1 - (8 + 1/2)2 + (V3/2)2
and
:L{x} =
-3/2
82 + 8 + 1
-3/2
(8 + 1/2)2 + (V3/2)2'
Then
y =e
_/
t 2
V3
cos -t
. V3
x = e -t/2 Sln-t
2 .
and
+ l):L{x} - :L{y}
(82
=-2
+ (82 + 1) :L{y} =
-:L{x}
so that
-282 - 1
1 1
:L{x} = 84+282 = -"2 82 -"2 82+2

and
x =-
1
2
3.
2V2
+t - -
Then
y =x
11
Slll V 2 t.
+ x = --t1 +
2
3
r:
In sin v 2 t.
2v2
(8 + 1) :L{x} + :L{y} = 1
4:L {x} - (8 + 1) :L {y} = 1
so that
and
-8
+3
8+1
2
(8 + 1)2 + 22 + 2 (8 + 1)2 + 22 .
:L {y} = 82 + 28 + 5
Then
x
= e-t cos 2t + 2e-t sin 2t and y = _e-t cos 2t + 2e-t sin 2t.
9. Adding the equations and then subtracting them gives
d2x
dt2
"2t2+ 2t
d2y
1
= _t2 - 2t.
dt2
2
354
Exercises
Taking the Laplace transform of the system gives
1
:t'{ x}
= 8~
and
1 4'
1 3!
+ 24 s5 + '3 s4
1 4!
1 3!
3 s4
:t'{y}=----24 s5
so that
(s - 4) :t'{x} + s3:t' {y} = -2-1

s
(S + 2) :t'{x} - 2s3:t'{y} = O
so that
4
414
:t'{x} = (s _ 2)(s2 + 1) =
and
2s + 4
:t'{y}
5s-
2-
= s3(s-2)(s2+1)
=;-
81
5 s2 + 1 - 5 s2 + 1
2
s2-2s3+5s-2-5s2+1
Then
4 2t
x = -e
5
- - cos t - - sin t
5
5
and
y = 1 - 2t - 2t
,
+ 51 e2t - 56 cos t + 58 sin
i.
S2:t'{X} + 3(s + l):t'{y} = 2

1
s :t'{x} + 3:t'{y} = (s + 1)2

so that
2s + 1
1
1
1 2
1
:t'{x} = - s3(s + 1) = ; + s2 + '2 s3 - S + 1 '
Then
X
= 1+ t +
= '3te
-t - e
-t
and
1
-t
'3x
11
= '3te
355
-t
+ '3e
+5s2+1'
-t
1
3
7.6
Exercises 7.6
2e-s
(8 - 4) ..:t'{x} + 2..:t'{y} = -3..:t'{x}
1 e-s
= 2 +-8
+ (8+ l)..:t'{y}
so that
(J){ x } =
-1/2
(8-1)(8-2)
1
= [28-1-28-2
+e -s
1
1]
1
(8-1)(8-2)
+e
-s
[1
-8-1
and
..:t'
e-s
8/4-1
=45-1-28-2+e
-8/2+2
-s
{y} = -;- + (5 - 1)(5- 2) + e

1
-s
1]
+ 8-2
(5 - 1)(8- 2)
[l 3 1
~-28-1
1]
+5-2
Then
x = ~et - ~e2t
+ [_et-l + e2(t-ll]
y = ~et _ ~e2t
4
2
+ [1 -
VU (t - 1)
and
~et-l
+ e2(t-ll] au (t - 1).
13. The system is
+ 2(X2 -
X'{
-3XI
X~
-2(X2 - Xl)
Xl)
XI(O)=O
X~
(O) = 1
X2(0) = 1
x;(O)
= O.
Taking the Laplace transform of the system gives
(82 + 5) ..:t'{xl} - 2..:t'{X2} = 1

-2..:t'{xl}
+ (82 + 2) ..:t'{X2}
=8
so that
..:t'
82+ 28+ 2 _ ~ _8_
{ l} - 84+ 782+ 6 - 5 82+ 1
+ ~ _1 __
5 82+ 1
and
356
~ _8_ + _4_ V6
5 82+ 6 5V6 52+ 6
Exercses
:L x
{2}
83 + 58 + 2 = ~ _8_
+ ~ _1_ + ~ _8
(82 + 1)(82 + 6)
5 82 + 1 5 82 + 1 5 82 + 6
7.6
2_ V6
5V6 82 + 6 .
Then
Xl
2
5
+-
1
2
sin t - - cos V6 t
5
5
4
5
2
5
= - cos t
and
1
5
X2 = - cos t + - sin t + - cos V6 t 14. In this system
Xl
4
re sin v'6 t
5y 6
2
;; sin v'6 t.
5y6
and X2 represent displacements of masses mI and m2 from their equilibrium
positions. Since the net forces acting on mI and m2 are
respectively, Newton's second law of motion gives
mlX{ = -kIXI + k2(X2 - xl)

m2X~ = -k2(X2 - Xl) - k3X2
Using kl
= k2 = k3 = 1, mI = m2
= 1,
XI(O) = O, XI(O) = -1, X2(0) = O, and x2(0) = 1, and
taking the Laplace transform of the system, we obtain

(2 + 82):L{Xl}
- :L{X2} =-1
:L{xl} - (2 + 82):L{X2} = -1
so that
Then
Xl
=-
1.
r:
v'3
sin y 3 t
15. (a) By Kirchoff''s first law we have il
E(t)
= Ril
+Lli2 and E(t)
= Ril
and
= iz + i3.
X2 =
1.
v'3
sin h t.
By Kirchoff's second law, on each loop we have
= E(t)
+L2i~ or Lli2+Ri2+Ri3
(b) Taking the Laplace transform of the system

0.01i2
+ 5i2 + 5i3 = 100
0.0125i~ + 5i2 + 5i3 = 100

gives
(8 + 500) :L{i2}
10,000
+ 500:L{i3} = _-
8,000
400:L{i2} + (8 + 400) :L{i3} = --
357
and L2i~+Ri2+Ri3
= E(t).
Exercises 7.6
so that
:t: .
8,000
80 1 80
{Z3} = s2 + 900s = 9" ~ - 9" s
+ 900
Then
.
Z3
(e) il
= i2 + i3
80 80 -900t.
9"
- 9"e
and
Z2
./
= 20 - 0.0025z3 -
Z3
100 100
9
-ge
-900t
= 20 - 20e-900t
16. (a) Taking the Laplace transform of the system
i; + i~ + 10i2 = 120 -
120OU(t - 2)
+ 5i~ + 5i3 = O
-10i;
gives
(s
. + S:t:{Z3}
. = -120 ( 1 + 10) :t:{Z2}
e _)28
+ 5(s + 1) :t:{i3}
-10s:t:{i2}
=O
so that
(LJ
ol.-
{z?J --
120(s
(3s2
+ 1)
+ lls + 10)s
( 1 - e -2s) - [48
- -- 60
- s + 5/3
s+2
+ -12]
s
( 1 - e -28)
and
:t:{i}=
3
3s2
240
+ lIs + 10
=[
(1_e-2S)
240
+ 5/3
_ 240] (1_e-2S)
s +2
Then
iz
= 12 + 48e-St/3 - 60e-2t -. [12 + 48e-S(t-2)/3
i3
= 240e-St/3 - 240e-2t - [240e-S(t-2)/3
- 60e-2(t-2)]
OU(t- 2)
and
(b) il
= i2 + i3 = 12 + 288e-St/3
- 240e-2(t-2)]OU(t
300e-2t - [12 + 288e-S(t-2)/3
- 300e-2(t-2)]
17. Taking the Laplace transform of the system

i;+ 11i2 + 6i3 = 50 sin t
i~ + 6i2 + 6i3 = 50 sin t
gives
(s
+ 11) :t:{z2} + 6:t:{Z3}
6:t:{i2}
+ (s + 6):t:{i3}
358
= s2
50
+1
= _250
- 2).
+1
OU(t - 2)
Exercises
so that
::t i
{2}
50s
(s+2)(s+15)(s2+1)
= _ 20 _1_
13s+2
375 _1_
1469s+15
145 _s_
113s2+1
+~
_1_
113s2+1'
Then
and
i2 = - 20 e-2t
13
375 e-1St
1469
+ 145 cos t + ~
113
113
,
25.
1,/
11 ,
30 -2t
23 = 3sin t - '622- 622 = 13e
250
+ 1469e
sin t
280
- 113 cos t
-1St
810 ,
113 sin t,
0,5i~ + 50i2 = 60
0,005i~ + i2 - il = O
gives
,
,120
s ::t{2} + 100::t{22} = -200::t{i}
+ (s + 200) ::t{i2}
=O
so that
::t{i } _
24,000
2 - s(s2
s+ 100
cos 100t -
~e-100t
= 6 1
100
S; - S (s + 100)2 + 1002 - S (s + 100)2 + 1002 '
+ 200s + 20,000)
Then
i2 = ~ 555
and
~e-l00t
, =e 0005
,/ + 22
' =S
6
21
u.
22
sin 100t
6 -100t cos 100t ,

- Se

2i~ + 50i2 = 60
0,005i;
+ i2 -
il = O
gives
,
,60
2S::t{2} + 50::t{22} = -
-200::t{iI}
+ (s + 200) ::t{i2}
=O
so that
::t{' }
6,000
22 = s(s2
61
+ 2008 + 5,000)
6
s + 100
- S; - S (s + 100)2 -
(50)2)2
359
6)2
- -5- (s
50)2
+ 100)2 -
(50)2)2
'
7.6
Exercises 7.6
Then
6 6
6J2.
i2 = - - _e-lOOt cosh 50J2 t - _e-lOOt
5 5
5
sinh 50J2 t
and
20. (a) Using Kirchoff's first law we write il
= i2+i3.
Since i2 = dqjdt we have il-i3
= dqjdt.
Using
Kirchoff's second law and summing the voltage drops across the shorter loop gives
1
E(t) = iRl + Cq,
(1)
so that
.
1
Il = -E(t)
- -q.
RlC
Rl
Then
.
-dq
=
Il dt
and
l.
I3 = -E(t)
- --q - I3
RC
Rl
dq 1
Rl dt + Cq + Ri3
= E(t).
Summing the voltage drops across the longer loop gives
E(t) = ilRl + L ~: + R2i3.

Combining this with (1) we obtain
di3
dt
ilRl + L- + R2i3
or
di3
L- + R2I3
(b) Using L
= iR
+ Cq
Cq = O.
= Rl = R2 = C = 1, E(t) = 50e-tOU(t -1) = 50e-1e-(t-l)OU(t -1), q(O) = i3(0) = O,
and taking the Laplace transform of the system we obtain

(8 + 1) ~{q}
+~
{i3} =
5Ce-1
__
8+1
e-s
(8 + 1)~{i3} - ~ {q} = 0,
so that
and
q(t) = 50e-1e-(t-l)
sin(t - l)OU(t - 1) = 50e-t sin(t - l)OU(t - 1).
360
Exercises
7.6
21. (a) Taking the Laplaee transform of the system
4e~ + e~ + 8el = O
e~ + e~ + 2e2 = O
gives
4 (s2 + 2).:t{el}
+ s2.:t{e2}
+ (s2 + 2) .:t {e2}
s2 .:t {el}
= 3s
=o
so that
or
Then
123
e2 = - eos ""t - - eos 2t
2
vs
and
e~ = -e~ - 2e2
so that
el =
(b)
1
4
2
y3
3
4
- eos ""t + - eos 2t.
91
92
-1
-2
Mass
m2
has extreme displaeements o greater magnitude.
equilibrium position at about t = 0.87, and mass
m2
Mass
m1
first passes through its
first passes through its equilibrium
position at about t = 0.66. The motion of the pendulums is not periodie sinee eos 2t/
period V3 1[", eos 2t has period
number.
1[",
and the ratio of these periods is
361
J3 , which
J3 has
is not a rational
Exercises 7.6
(e) The Lissajous curve is plotted for O
(d1/
t= 1
:s t :s 30.
t= 2
t= 4
92
81
82
1
2
3'
4
5
6
7
8
9
10
-0.2111
-0.6585
0.4830
-0.1325
-0.4111
0.8327
0.0458
-0.9639
0.3534
0.4370
0.8263
0.6438
-l. 9145
0.1715
1.6951
-0.8662
-0.3186
0.9452
-l. 2741
-0.3502
t= 7
t=
(e) Using a CAS to solve el(t) = e2(t) we see that el = e2 (so that the double
pendulum is straight out) when t is about 0.75 seconds.
362
10
t=0.75

(f) To make a movie of the pendulum it is necessary to locate the mass in the plane as a function
of time. Suppose that the upper arm is attached to the origin and that the equilibrium position
lies along the negative y-axis. Then mass mI is at (x, (t), Y1(t)) and mass m2 is at (X2(t), Y2(t)),
where
Xl(t) = 16sinBI(t)
and
Yl(t) = -16cosBl(t)
and
Y2(t) = Yl(t) - 16cosB2(t).
and
X2(t) = Xl(t) + 16sinB2(t)
A reasonable movie can be constructed by letting t range from O to 10 in increments of 0.1

seconds.

1.
.:t'{f(t)}
= rlte-stdt+
(X)(2_t)e-stdt=
Jo
3. False; consider f(t)

4. False, since
Jl
22e-s
8
t-I/2.
f(t) = (et)IO = elOt.
5. True, since lims->ooF(8)

6. False; consider
7 . .:t'{e-7t}
128
= 1 i- O. (See Theorem 4.5 in the text.)
f(t) = 1 and g(t) = 1.
= _1_
8+7
8. .:t'{te-7t}
9. .:t'{sin2t} =
(8:
7)2
-i+4
8
10. .:t'{e-3tsin2t}
= (8+3~2+4
2]
.
11. .:t'{t sin
2t} = -- d [-2ds 8 + 4
= (2 48 )2
8
+4
2
12. .:t'{sin2tOU(t - 7T)} = .:t'{sin2(t - 7T)OU(t- 7T)} = -2-e-1fS
8 +4
363
17.
:::t-1{
18. :::t-1
19.
20.
s
} =:::t-1
s2 - lOs + 29
12e-5S}
:::t-1{
(s -
s-5
5)2
+ 22
+~
(s -
2
5)2
+ 22
=tcos2t+~tsin2t
= (t - 5)OU(t - 5)
+ 1f 2 e -S}
(1)-1 { S
oL
(1)-1 {
2
s+1f
L2s2 + n21f2
2e
-s
1f
+ s+7i
2
2 e
-S}
=oL
2
s+1f
= COS 1f (t
- 1) OU(t - 1) + sin 1f (t - 1) OU(t - 1)
} _ ~ !:_:::t-1{
- L2 n1f
21.
:::t{e-5t} exists for s > -5.
22.
:::t{teStf(t)}
23.
:::t{eat f( t - k) OU(t - k)}
24.
:::t{l
n1f/L
}
s2 + (n21f2)/ L2
_l-sin~t
Lnat
= - :sF(S-S).
= e-ks
:::t{ea(t+k) f( t)} = e-ks eak :::t{eat f( t)}
= ~:::t{eatf(t)}
eaTf(T)dT}
F(ss-a),
= e-k(s-a)
F( s - a)
whereas
:::t{eat (t f(T) dT} =:::t { (t f(T) dT} I

= F(s) I
= F(s_- a) .
Jo
Jo
s-->s-a
s s-->s-a
S
a
25.
f(t)OU(t-to)
26.
f(t) - f(t)OU(t - to)
27.
f(t - to) OU(t - to)
28.
f(t) - f(t)ilU(t - to) + f(t)OU(t - ti)
29.
f(t) = t - [(t -1) + l]OU(t - 1) +OU(t - 1) -OU(t - 4) = t - (t - l)OU(t - 1) -qj(t - 4)

:::t{f(t)}
2_ - 2_e-s _ ~e-4s
s2
s2
364
Chapter
:L{etf(t)}
30. f(t)
e-(s-l) __ 1_e-4(s-1)
s-l
(s-l)2
= sint"i1(t - 7T) - sintOU(t - 3T) = - sin(t - 7T)"i1(t- 7T) + sin(t - 37T)"i1(t- 37T)
:L{f(t)}
= 2-
:L{f(t)}
:L{etf(t)}
+ _1_e-37rs
s2 + 1
l_e-7rs
S2 + 1
= __
:L{et f(t)} = _
31. f(t)
(s-1)2
7 Review Exercises
e-7r(s-l) +
(s-1)2+1
2"i1(t - 2) + [(t - 2) + 2]"i1(t- 2)

2
s
=-
e-37r(s-1)
(s-1)2+1
= 2 + (t -
2)au(t - 2)
1
+ 2e-2s
s
_2_
s-l
e-2(s-1)
(s-1)2
32. f(t) = t - t"i1(t -1) + (2 - t)"i1(t -1) - (2 - t)"i1(t - 2) = t - 2(t -l)"i1(t
-1)
+ (t - 2)"i1(t -
<-P{f(t )} = -1 - -e
2 -s +-e1 -2s
s2 s2
s2
eL-
:L{et f(t)} =
1
.
(s-1)2
e-(s-l) +
(s-1)2
e-2(s-1)
(s-1)2
33. Taking the Laplace transform of the differential equation we obtain

5
:L{y} = (s _ 1)2 + 2 (s - 1)3

so that

1
:L{y}
(s _ 1)2(s2 - 8s + 20)
6
s-4
= 169 s - 1 + 13 (s - 1)2 - 169 (s - 4)2 + 22 + 338 (s - 4)2 + 22

so that
6
169
_et
1
+ -tet
13
6
169
- _e4t
365
cos 2t
+ _t5.
338
sin 2t.
2)

35. Taking.the Laplace transform of thegiven differential equation we obtain
~ {}
s3 + 6s2 + 1
1
-2s
2
-2s
+ l)(s + 5) - s2(s + l)(s + 5) e - s(s + 1)(8 + 5) e
y = 82(S
6
13
= - 25 . --;+ 5" . s2 + "2 . s + 1 - 50 . s + 5
_ (_~ . ~ + ~ . ]_ + ~ . _1
1_ . _1_) e-2s
25 s 5 s2 4 s + 1 100 s + 5
+ _]_.
e-2s
5 5 2 s + 1 10 s + 5
_ (~.~_ ~._1_
_1_)
so that
y = _~ + ~t2
25 5
+ ~e-t _
13e-5t
50
i_OU(t - 2) - ~(t - 2)2OU(t - 2)

25
+ ~e-(t-2)OU(t - 2) - _2_ e-5(t-2JOU(t - 2).

4
100
~{y} =
s3 + 2 _ 2 + 2s + s2 e-s
s3(s - 5)
s3(s - 5)
=-
2 1
2 1
1 2
125 --;- 25 s2 - 5" s3
127
+ 125 s -
5-
[37
1 12 1
1 2
- 125 --;- 25 s2 - 5" s3
37
l]_s
+ 125 s _
5 e
so that
y
127
= -- 2 - -t2 - -t1 2 + -e
125
25
125
5t
-- 37)2- -(t
125 25
1 - 1)2 + -e
37
- 1) - -(t
5
125
37. Taking the Laplace transform of the integral equation we obtain

1
1 2
~ {y} = --;+ s2 + "2 s3

so that
38. Taking the Laplace transform of the integral equation we obtain
~{f}
so that f(t)
= 6t.
s~{x}
4~{x}
+~{Y}
1
s
2" + 1
+ s~{Y} = 2
366
1
s
= 6 2"
5(t-1J]
OU(t- 1).

so that
.:t'{x}
S2 - 2s + 1
s(s _ 2)(s + 2)
11
= -4 ~ + 8 s -
+ 8 s + 2'
Then
x
119
= -- + _e2t + _e-2t
4
and
Y = -x
9 -2t
1 2t
+ t = -e
- -e + t.
40. Taking the Laplace transforrn of the systern gives

1
s2 .:t'{x}
+ s2 .:t'{y} = s _ 2
2s.:t'{x}
+ s2.:t'{y}
1
=- s _ 2
so that
2
11
1
.:t'{x} = s(s-2)2=2~-2s-2+(s-2)2
and
.:t'
-s - 2 __ ~ ~ _ ~ 2_ + ~ _1_ _
{y} - s2(s - 2)2 -
4 s
2 s2
4 s - 2
(s - 2)2 .
Then
x
1
2
1
2
= - - _e2t + te2t and y
= -- - -t
4
2
3 2t
2t
+ -e
- te .
4
41. The integral equation is

10i + 2 fat i(T) dr = 2t2
+ 2t.
Taking the Laplace transforrn we obtain

.
.:t'{t}=
( 4
s+2
= s2(5s+2)
2)
s3
+ s2 10s+2
45
=-~+
s2
+ 5s+1
=-~+
Thus
i(t) = -9 + 2t
+ ge-t/5.

1 d2q
dq
2 dt2 + 10 dt + 100q =
10 - 10OU(t - 5).
Taking the Laplace transforrn we obtain
.:t'{ }
q == 2 (s2
20
+ 20s + 200)
_ [2_ ~ _ 2_
-
10 s
10 (s
(1
- e
-5S)
s + 10 _ 2_
10
] (1 _
+ 10)2 + 102 10 (s + 10)2 + 102
367
s2+ s+1/5'
e-5S)

so that
111
_e-10t cos 10t - _e-10t sin 10t
10 10
10
q(t) = - -
- [~
10
- ~e-10(t-5)
10
cos 10(t - 5) -
~e-10(t-5)
10
sin 10(t - 5)] OU(t- 5).
43. Taking the Laplace transform of the given differential equation we obtain
.:t'{y} = 2wo
ElL
(.!:_.
4885
4! __ 1_ . ~ + _1_ . ~ e-SL/2)
12086
12086
+ Cl . ~ + C2. 3!
283
684
so that
2wo [L 4
1 5
1 (
y = El L 48 x - 120x + 120 x
where y"(O) = Cl and y"'(O)
= C2. Using y"(L)
ci 2
C2 3]
-"2L) 5 OU( x -"2L) +"2
x +"6x
= O and
. Cl = woL2/24El,
y"'(L) = Owe find
C2= -woL/4El.
Hence
44. (a) In this case the boundary conditions are y(O)
= y"(O) = O and Y(7r) = y"(7r) = O. If we let
Cl = y' (O) and C2= y'" (O) then
84.:t'{y} -'- 83y(0) - 82y'(0) - 8Y(0) - y"'(O) + 4.:t'{y} =.:t' {wo/El}

and
.:t'{ } = Cl . ~
y
284+4
+ wo (~_
+ C2. _4_
484+4
8El8
From the table of transforms we get

Cl .
.
c2 .
y = "2(smxcoshx + cosxsmhx) + -(smxcoshx
Using Y(7r)
8- 1
(8-1)2+1
.
- cosxsmhx)
8+ 1
(8+1)2+1
wo
+ 4E/1-
)"
.
cosxcoshx)
= O and y"(7r) = Owe find

wo
Cl = 4El (1 + cosh 7r)csch zr,

Hence
y = 8~/1
+ cosh 7r)csch 7r(sinx coshx + cos x sinh x)
wo
- 8E/1
+ cosh 7r)csch 7r(smxcoshx
wo
- cos x sinh z ) + 4El(1
- cos xcoshx).
(b) In this case the boundary conditions are y(O) = y'(O) = O and Y(7r) = y'(7r) = O. If we let
Cl = y"(O) and C2= y"'(O) then
84 .:t'{y} - iy(O) - 82y'(0) - 8y(0) - y"'(O) + 4.:t'{y} =.:t' {(t - 7f/2)}
368

and
:t{ } =
y
Cl . ~
2 54
From the table of transforms
+ C2. _4_ +
4
Y(7T)
[Sin
= O and
54
+4
Wo . _4_ e-srr/2.
4El
54
cos
. 11X )
SlD
(x - ~) cosh (x - ~) - cos (x - ~)
Y'(7T)
+4
we get
cl.
. h
+ 4"
C2( .
= "2
SlDX SlD
X
SlDX COSh X
+ 4V;l
Using
+4
sinh
(x - ~) ] q (x - ~)
= O we find
Wo sinh
El sinh
11"
Cl=---
Wo cosh
El sinh
11"
C2=----
Hence
wo
sinh!I.
y = 2El ~h
sin
2 sinxsinhx
11"
+ 4V;l [sin
Wo
cosh!I.
El ~h
4
sin
2 (sinxcoshx
11"
- cosxsinhx)
(x - ~) cosh (x - ~) - cos (x - %) sinh (x - %)] ou (x - %) .
369
Systems of Linear First-Order

Differential
____
Equations
Exercises 8.1
1. Let X = (~). Then
x' =
2. Let X
(! -5)
8
= (~). Then
x' = (:
3. Let X ~
G)
4. Let X ~
G)
-7)
o
X.
Then
x' =
4 -9)o
-1
103
o
o
x' ~ ( :
-1
x~ (~)
X.
Then
-1
5 Lec
. X.
Dx
Then
-1
x'~U -Dx+
1
1
(-t;t2)
370
U)
+ .:
Exercises 8. 1
6. Let X ~
7.
Then
dx
di = 4x + 2y + et;
dx
8. dt
9.
G)
= ~IX + 5y -
dx
di
=X
9z - 8e- t;
Y + 2z
+ e-
- 3t;
dz
= -2y + 3z + e5t - 3e-2t
dt
dy
-t
dz
t
dt = 3x - 4y + z + 2e '+ t; dt = -2x + 5y + 6z + 2e- - t
dy
dt
= 4x + y + z + 2e5t;
dx
10. -d =3x-7y+4sint+(t-4)e4t;
dy
-=x+y+8sint+(2t+1)e4t
11. Since
X' = ( -5 ) e-5t
-10
and
we see that
X'=
12. Since
I
X =
5 cos t - 5 sin t ) t
2 cos
t - 4 sin t
and
we see that
X t :=
13. Since
( 3 -4)
( 3 -4)
4 -7
4 -7
X=
XI
-10
-5t
X.
t - 5 sin t)
( -- 22 45) X = (52 cos
cos t - 4 sin t
(-2-2 45)
( -1 1/4) X = (3/2)
= (-1 1/4)
1 -1
et
X.
1 -1
we see that
(-5)
e-3t/2
-3
X,
14. Since
( -12 O1) X = ( -15)

we see that
X' = (_~
371
~) X.
et
+(
4)
-4
te'
Exercises 8. 1
15. Since
we see that
16. Since
X'=
cos,
)
~ sin t - ~ cos t
- cost - sint
we see that
and
1)
(
COS,
)
O X = ~ sin t - ~ cos t
- cos t - sin t
O -1
1
X'~ ( : 1 ~) X
-2 O -1
17. Yes, since W(XI, X2) = _2e-8t =1- Oand X and X2 are linearly independent on
-00
< t < oo.
18. Yes, since W(XI, X2) = 8e2t =1- Oand X and X2 are linearly independent on
-00
< t < oo.
= Oand XI, X2, and X3 are linearly dependent on
-00
< t < oo.
19. No, since W(XI,X2,X3)
20. Yes, since W(XI, X2, X3) = -84e-t

-00
=1- O and Xj , X2, and X3 are linearly independent on
< t < oo.
21. Since
we see that
22. Since
we see that
23. Since
372
Exercises
8. 1
we see that
24. Since
X' =
p
3 cos
O
3t)
and
-3 sin 3t
1
-4
2 3)
2 O X,
-6
1 O
( -1)
4 sin 3t
3
( 3 cos 3t )
-3sin3t
we see that
25. Let
Then
x; ~
(=D ,-" ~AX"
X3~ m'~AX3'
and W(XI,X2,X3)
-00
20
i- O so that
Xj , X2, and X3 form a fundamental set for X' = AX on
< t < oo.
26. Let
1 r ) e v'2t ,
Xl = (
-1- v2
= (
-1 +
j2
) e-v'2t
373
Exercises 8. 1
and
Then
x:2 = ( -2-}2
-}2
and W(X1, X2)

set on
-00
) e-v'2t
= AX 2,
= 2}2 i- Oso that Xp is a particular solution and X and X2 form a fundamental
< t < oo.
Exercises 8.2
l. The system is
x'=
and det(A - >"1) = (>..- 5)(>..+ 1)
( -: _~ I ~)
e ~)x
= O. For >"1 = 5 we obtain
-1 ~2 ~)
so that
K1 ~
e).
For >"2 = -1 we obtain
Then
X = ci
( 21)
2. The system is
x'=
e5t
(-1)
+ C2
e-.t
e ~)x
and det(A - >"1) = (>..- 1)(>" - 4) = O. For >"1 = 1 we obtain
e ~I~)=C~I~)
so that
Kl~(-~)
For >"2 = 4 we obtain

(
-2
1
2 ':
==> (-1
1
-1
O
O o
lO)
Io
374
so that
K2
(11).
Exercises 8.2
Then
X =
elle( -2)
+ e2
(1)1 e .
4t
3. The system is
X' - (-~~2 ~) X
and det(A - Al) = (A - l)(A
+ 3) =
O. For Al = 1 we obtain
-5 21 O):=::;. (-5 210)

( -5/2
1 1O
O O 1O
For A2
so that
Kl
so that
K2 = (21) .
(25)'
-3 we obtain
-1 2 1 O) :=::;. (-1 2 1 O)
( -5/2
5 1O
O O 1O
Then
X =
el ( 52) et + e2 (2)1 e- 3t .
4. The system is
X' =
and det(A - Al)
=
(
For A2
(A
(-5/2
3/4
+ 1)(2A + 7) = O. For Al
2)
-2
= -7/2
1
210):=::;.(-1210)
3/4 3/2 1 O
O O 1O
we obtain
Kl=(-21)'
sothat
= -1 we obtain
( -3/23/4
2 lO):=::;. (-3 410)

1O
O O 1O
K2 = (43)'
so that
-1
Then
X
ci ( - 2)1 e- 7t/2 + e2 ( 4)
3 e-.
5. The system is
X' = (10
-5) X
-12
and det(A - Al)
For A2 =
-10
= (A - 8)(A + 10) = O. For Al = 8 we obtain

-5
O):=::;. (1 -5/210)
( 82 -20
O
O O 1O
we obtain
-1/4 lO)
( 208 -5-2 1 OO):=::;. (1O O O
1
375
so that
Kl = (52)'
Exercises 8.2
Then
6. The system is
X' = (-6 2) X
-3
and det(A - Al)
= A(A + 5) = O. For Al = O we obtain
210)
( -6
-3
1 1 O
(1 -1/310)
==>
so
that
Kl =
(31).
For A2 = -5 we obtain
( -3-1 6210)
I O
==>
(1 -210)
O
so that
K2 =
(21),
Then
7. The system is
XI=(~ ~ -~)X
O 1 -1
and det(A - Al)
= (A - 1)(2- A)(A + 1) = O. For Al = 1, A2 = 2, and A3 = -1 we obtain
. ~
so that
m, m, m'
m m m,-'
x ~ '1
KF
andKF
e"+~
"+02
8. The system is
X' = (~ ~~ :) X
052
and det(A - Al) = (2 - A)(A - 5)(A - 7)
= O. For Al
376
= 2, A2 = 5, and A3 = 7 we obtain
Exercses 8.2
so that
9. We have det(A - Al)
= -(A + l)(A - 3)(A + 2) = O. For Al
= -1, A2
= 3, and A3
= -2 we obtain
so that
= -A(A - l)(A - 2) = O. For Al = O, A2 = 1, and A3 = 2 we obtain
so that
11. We have det(A - Al) = -(A
+ l)(A + 1/2)(A + 3/2)
= O. For Al
= -1, A2
= -1/2,
and A3 = -3/2
we obtain
Kl~(_D, K2~C1D,
so that
X ~ c
(-D
12. We have det(A - Al) = (A - 3)(A
' + '2
+ 5)(6
en
and
,-'/2 + '3
- A) = O. For Al
377
K3~(_D,
(-D
,-31/2
3, A2 = -5, and A3 = 6 we obtain
Exercises 8.2
so that
13. We have det(A - Al) = (A
+ 1/2)(A
- 1/2)
K1=
= O. For Al
(~)
= -1/2
K2 =
and
and A2
= 1/2 we obtain
e) ,
so that
If
X(O)
then
C1
== 2 and
C2
= (:)
= 3.
= (2 - A)(A - 3)(A + 1) = O. Por Al
cn'
m cn,-'
H), m,
KI ~
K, ~
so that
x ~CI (
-~)
,"
+02
X(O) ~
C1
15. X = ci
= -1,
C2
= 5/2, and
(~:~:~~:~)
e8.58979t
C3 =
C1
~~4~;~~:3)
e2.25684t
+ C3
(=~:~~~~~~)
e-O.046632lt
0.359815
0.615458
0.35995
0.0312209
-0.280232
0.262219
0.949058
16. X
and K3 ~
-1/2.
+ C2 (
0.239535
-0.836611
e5.05452t
= -1 we obtain
a" + C3
If
then
= 3, and A3
= 2, A2
+ C2
-0.275304
-0.162664
+ C3
0.195825
0.176045
-0.826218
-0.346439
0.0508861
0.338775
0.31957
0956lt
e-2.92362t
continued
378
Exercises 8.2
0,313235
-0,301294
0,64181
0.466599
0,31754
+C4
+ C5
e2,02882t
e -o, l55338t
0,222136
0,173787
0,0534311
-0,599108
-0,799567
17. (a)
(b) Letting
Cl
1 and
C2
= O we get x = 5e8t, y
2e8t, Eliminating the parameter we find
y = tx, x > O. When Cl = -1 and C2 = Owe find y = tx, x < O. Letting Cl = Oand C2
get x = e-lOt, y = 4e-lOt.
Eliminating the pararneter we find y = 4x, x > O. Letting
and
C2 =
-1 we find y = 4x, x
< O.
(e) The eigenvectors K = (5,2) and K2 = (1,4) are shown in the figure in part (a).
= Owe get x = -2et, y = et.

Eliminating the parameter we find y = -!x, x < O. When
C1 = -1 and C2 = O we find y = -!x, x > O. Letting C1 = O
18. In Problem 2, letting
and
C2
= 1 and
C2
= 1 we get x = t, y = t. Eliminating the parameter
we find y
x
Cl
= x, x > O. When
Cl
= Q and
C2
= -1 we find y = z ,
< O.
= O we get x = - 2e-7tj2,
y = e-7tj2. Elimnatng the parameter we find y = -~x, x < O.
When Cl = -1 and C2 = O we find y = -~x, x > O. Letting
ci = O and C2 = 1 we get x = 4e-t, y = 3e-t. Eliminating the
parameter we find y = ~x, x O, When Cl = Oand C2 = -1 we
find y = ~x, x < O.
In Problem
4,
letting
Cl
1 and
C2
379
= 1 we
Cl
=O
Exercises 8.2
19. We have det(A - ,Al)
= ,A2 = O. For ,Al = Owe obtain

K= (~).
A solution of (A - ,A1I)P
= K is
P = (~)
so that
= (,A + 1)2
= O. For ,Al = -1
=K
is
P=
so that
X = ci
we obtain
C).
K=
A solution of (A - ,A11)P
(~)J.
+C2 [C)t+
X=q (~)
C)
:'
+ C2
C~5)
[C) u:' + C~5)
e-t]
= (,A - 2)2 = O. For ,Al = 2 we obtain

K=
=K
is
C).
P (-1~3)
=
so that
X C)
= C1
= (,A - 6)2
e2t
+ C2
= O. For ,Al
[C)
=K
+(
= 6 we obtain
K=
te2t
(~) .
is
so that
380
-1~3)
e2t]
Exercises 8.2
23. We have det(A - Al) = (1 - A)(A - 2)2 = O. For Al = 1 we obtain
K~(J
For A2 = 2 we obtain
K, ~
and K3 ~
(J
Then
24. We have det(A - Al) = (A - 8)(A
+ 1)2
= O. For Al = 8 we obtain
K~m
For A2
-1 we obtain
K, ~ ( -:)
and
K3 ~
H)
Then
= -A(5 - A)2 = O. For Al = o we obtain
K~(=n
For A2
= 5 we obtain
K~CD
A solution of (A - All)P
= K is
5/2)
P = 1/;
381
Exercises 8.2
so that
26. We have det(A - Al) = (1 - A)(A - 2? = O. For Al = 1 we obtain
KI~m
For A2
=2
we obtain
K~(-D
A solution of (A - A2I)P = K is
(-n
p~
so that
X ~ '1
el +02 ( -
D ,"
- 1)3 = O. For Al
+03 [( - ~ )
te"
+(
-n
e"].
= 1 we obtain
K~(D
Solutions of (A - A11)P = K and (A - A11)Q = Pare
p~m Q~n)
m (D m [(D m n) "~l
and
so that
X ~ '1
+02 [
te' +
e'] +O,
(A - 4)3 = O. For Al = 4 we obtain
K~m
382
t;"
te' +
Exercises 8.2
Solutions of (A - )l}l)P = K and (A - )l}l)Q
= Pare
p~m Q~m
and
so that
X~
'1
m [m m e"] [m t;e" m m e"].

t'+
'+'2
+~
t'+
= (A - 4)2 = O. For Al = 4 we obtain

K=
C).
P=
C)
A solution of (A - All)P = K is
so that
X
= el (~) e4t + e2 [
G)
te"
C)
e4t]
If
then el
= -7
and e2 = 13.
+ l)(A -
1)2 = O. For Al = -1 we obtain
KI~cn
For A2 = 1 we obtain
KF(D andKF(D
so that
x ~ '1 ..
e-l
+ '2
If
X(O)~
m m
m
383
e' + '3
e'.
Exercises 8.2
then
el
= 2,
e2
= 3, and
e3 = 2.
= 2 is an eigenvalue of multiplicity 5. Linearly
31. In this case det(A - Al) = (2 - A)5, and Al

independent eigenvectors are
Kl=
K2=
and
K3=
32. Letting el = 1 and e2 = Owe get x = et, y = e', Eliminating
the parameter we find y = x, x
> O. When
el =
-1 and
e2 =
we find y = x, x < O.
Letting c = 1 and
When c
= -1 and
e2 = Owe
get x = e2t, y = e2t. Eliminating the parameter we find y = x, x
e2 = Owe
find y
= x,
> O.
x < O.
In Problems 33-46 the form of the answer will vary according to the choice of eigenvector. For exarnple,
in Problem 33, if Kl is chosen to be (
1 .) the solution has the form

2-2
cos t
) e4t + e? (
sin t
) e4t .
X = el (
2 cos t + sin t
- 2 sin t - cos t
= A2 -
8A
+ 17 = O. For Al = 4 + i we obtain
384
Exercises
i)
KI= ( 2 +
5 '
so that
x, = (2+5
i)e(4+i)t=
e4t+i(cost+2sint)
(2Cost-sint)
5 cos t
e4t,
5 sin t
Then
X = el (
2 cos t - sin t ) 4t
( 2 sin t + cos t)
4t
e + e2
e ,
5 cos t
5 sin t
34. We have det(A - Al) = A2 + 1 = O, For Al = i we obtain

Kl=
( -l-i)
2
so that
XI=
-1 - i ) it
e = ( sin t - eos t ). +z.' (- cos t - sin t ) ,
2
2 cos t
2 sin t
Then
X
35. We have det(A - Al) = A2
= el (
sin t - cos t )
( - cos t - sin t )
+ e2
'
2 cos t
2 sin t
8A + 17 = O, For Al = 4 + i we obtain
K =
( -l-i)
2
so that
Xl = ( -1 - i ) e (4+ ~')t = ( sin t - cos t ).. e4t + z' ( - sin t - cos t ) e4t,
2
2 cos t
2 sin t
Then
X = el (
sin t - cos t) 4t
( - sin t - cos t ) 4t
e + e2
e ,
2 cos t
2 sin t
lOA + 34 = O, For Al = 5 + 3i we obtain

KI = ( 1 - 3i)
2
so that
Xl = ( 1 - 3i) e (5+3 ~')t = ( cos 3t + 3 sin 3t ) e 5t + z' (Sin 3t - 3 cos 3t ) e 5t ,

2
2 cos 3t
2 cos 3t
Then
X = el (
COs3t+3sin3t)
2 cos 3t
5t
e + e2
385
(Sin3t-3cos3t)
2 cos 3t
5t
e ,
8.2
Exercises 8.2
= A2 + 9 = O. For Al = 3i we obtain
Kl= ( 4 + 3i)
5
so that
Xl =
( 4 +5 3i)
Then
X = el
3it
(4 cos 3t - 3 sin 3t ) + . (4 sin 3t + 3 cos 3t )

2
5 cos 3t
4 cos 3t - 3 sin 3t)
+ e2
5 cos 3t
5 sin 3t
(4 sin 3t + 3 cos 3t ) .
5 sin 3t
= A2 + 2A + 5 = O. For Al = -1 + 2i we obtain
Kl =
e~
2i)
so that
=(2cos2t-2sin2tcos2t)e-t+i
Then
X
el (
2 cos 2t - 2 sin 2t ) -t
e
cos 2t
39. We have det(A - Al) = -A (A2
For A2
=i
+ 1) = O.
2 cos 2t + 2 sin 2t )
.
e-t.
sm2t
+ e2 ( 2 cos 2t + 2 sin 2t)

sin 2t
For Al = O we obtain
we obtain
so that
Then
X = el
1)
O
O
+ e2
(Sin
t)
- sin t
cost
386
+ e3
( - cos t )
c~s t .
smt
-t
Exercises
For A2
= -(A + 3)(A2
2A
+ 5) = O. For
Al
= -3 we obtain
1 + 2i we obtain
so that
X2=
Then
X=C1
( - cos 2t - 2 sin 2t)

-3cos2t
et.
2 cos 2t
O)
-2
For A2
- 2 cos 2t + sin 2t )
3sin2t
et+i
2 sin 2t
( - 2 cos 2t + sin 2t )
e-3t+c2
3sin2t
et+c3
( - cos 2t - 2 sin 2t )
-3cos2t
e',
2 cos 2t
= (1 - A)(A2
2A
2 sin 2t
+ 2) = O. For
Al
= 1 we obtain
= 1 + i we obtain
so that
Then
X = C1
O)
(
2 et +
(1
42. We have det(A - Al) = -(A - 6)(A2
C2
8A
cos t )
s~nt et
- SIn t
+ 20) =
+ C3
( sin t )
COS t
et.
cost
O. For Al = 6 we obtain
387
8.2
Exercises 8.2
For '\2 = 4 + 2i we obtain
en
K2~
so that
X2 = (-~)
e(4+2i)t
= ( Si~2t )
(-C~'2t)
t+i
2 cos 2t
e4t.
2 sm 2t
Then
X=Cl
O)1
(O
43. We have det(A -:-'\1) = (2 - ,\)(,\2
For '\2
-2
+ 3i
e6t+c2
sin 2t )
O
2 cos 2t
e4t+c3
+ 4,\ + 13) = O. For
( - cos 2t )
_o
2 sin 2t
e4t.
'\1 = 2 we obtain
we obtain
4 + 3i)
K2 =
-5
O
so that
X2=
4 -5+ 3i)
.
e(-2+3t)t=
(4 cos-5cos3t
3t - 3 sin 3t)
(4 sin-5sin3t
3t + 3 cos 3t)
e-2t+i
e-2t.
Then
X
=q
28 )
5 e 2t + C2
25
44. We have det(A - '\1)
For '\2
(4 cos 3t5 - 33sin 3t)

-
COS
(
e -2t
+ C3
4 sin 3t5 +. 33tcos 3t)

-
sm
O
= -(,\ + 2)(,\2 + 4) = O. For '\1 = -2 we obtain
= 2i we obtain
388
e.-2t
Exercises
so that
X2 =
-2-2i)
1
e2tt =
(-2Cos2t+2sin2t)
cos 2t
cos 2t
+i
(-2Cos2t-2Si112t)
sin 2t
sin 2t
Then
O)
X
C1
( - 2 cos 2t + 2 sin 2t )
cos 2t
45. We have det(A - "\1)
Por "\2
e-2t + C2
-1
+ C3
( - 2 cos 2t - 2 sin 2t )
sin 2t
.
~~
~~
= (1 - ,,\)(,,\2 + 25) = O. Por "\1
= 1 we obtain
= 5i we obtain
1 + 5i)
1
(
1
K2=
so that
X2 =
1 + 5i )
1
tt
( cos 5t - 5 sin 5t )
cos 5t
+i
( sin 5t + 5 cos 5t)

sin 5t
.
sin 5t
+ C3
( sin 5t + 5 cos 5t )
sin 5t
.
sin 5t
cos 5t
Then
X = C1
25 )
7 et
+ C2
( cos 5t - 5 sin 5t )
cos 5t
cos 5t
If
U)
X(O)~
then
C1
C2
= -1 and
c3
46. We have det(A - "\1) =
= 6.
,,\2 -
10,,\+ 29 = O. Por
K1 =
so that
Xl = ( 1 ) e (5+2i)t =
1 - 2i
"\1 =
5 + 2i we obtain
e~
2i)
cos 2t
) e5t
cos 2t + 2 sin 2t
+ z. (
sin 2t
) e5t .
sin 2t - 2 cos 2t
and
.
cos 2t
)
X = C1 (
cos 2t + 2 sin 2t
5t
+ C3 (
389
sin 2t
)
sin 2t - 2 cos 2t
5t
8.2
Exercises 8.2
If
X(O) = (-:)
then
CI =
-2 and
C2 = 5.
47.
Phase portrait for Problem 37
48. (a) Letting
Xl
= YI, X~ = Y2, X2 = Y3, and X2 = Y4 we have
The corresponding linear system is
YI = Y2
Y2
-10YI
+ 4Y3
Y3 = Y4
y~ = 4YI - 4Y3
or
o
y'=
-10
O
4 O
-4
Y.
Using a CAS, we find eigenvalues v'2i and 2y'3i with corresponding eigenvectors
:r=v'2i/4
1/2
:r=v'2i/2
1
1/2
O
390
:r=v'2/4
i
:r=v'2/2
O
Exercises
and
V3i/3
-2
::::V3i/6
-2
-V3/3
O
V3/6
8.2
Thus
-v'2/4
O
-v'2/2
O

C3
-i ' and
C4
YI
(O) = O, Y2(0) = 1, Y3(0) = O, and Y4(0) = -1 imply
CI
-g,
C2
= O,
= O. Thus,
(t) = -
y;
sin ht
X2(t) = Y3(t) = -
V;
sin ht
- ~
Xl
(t) =
YI
V;
sin 2v'3t
sin2v'3t.
(b) The second-order system is
X~ = -10XI + 4X2
X~ = 4XI - 4X2
or
X" = [-10
We assume solutions of the form X
and -12,
WI
= .)-(
-2)
= v'2 and
4] X.
4 -4
= V coswt and X
W2
)-(-12)
391
V sinwt. Since the eigenvalues are -2
= 2V3.
The corresponding eigenvectors
Exercises 8.2
are
and
Then, the general solution of the system is
X = ci
cos ht
[~]
+ C2[~]
+ C3[ -~]
sin ht
cos 2v'3t
+ C4[ -~]
sin 2v'3t.

and
X(O) = [~]
imply
ci =
O, C2= -h/10, C3= O, and C4= -/3/10.

Xl (t)= -
V2 sin ht + /3
10
V2.
Thus
sin 2v'3t
)3.
X2(t) = --smht
- -sm2/3t.
10
49. (a) From det(A - Al) = A(A - 2) = O we get Al = O and A2 = 2. For Al = Owe obtain
1 1
1110
2 we obtain
(
For A2 =
I O)
=:}
(1 1
0010
I O)
so that
(-: _~l~)= (-~~l~)
Then
X=ct
so that
( -1) (1)
1
+C2
1 e.2t
The line y = -x is not a trajectory of the system.

Trajectories are
x =
-Cl
+ C2e2t, Y = Cl + C2e2t or
+ 2Cl. This is a family of lines perpendicular

to the line y = -x. AH of the constant solutions
y =x
of the system do, however, lie on the line y = -x.

(b) From det(A - Al) = A2 = Owe get Al = O and
K = (-~).
A solution of (A - All)P = K is
P = (-~)
392
K 1 = (-11)
K2~ (:)
Exercises
so that
x = C! ( - ~ ) + C2 [ (
- ~ )
All trajectories are parallel to y = -x, but

y = -x is not a trajectory. There are constant
solutions of the system, however, that do lie
on the line y
= -x.
50. The system of differential equations is
and
393
t + ( - ~) ] .
8.2
Exercises 8.2
The general-solution of the system is
C2te2t
+ e e2t
C2e2t
x=
C3e2t
C5te2t
+ C4e2t
C5e2t
=e
e2t
+ C2
O
1
te2t
+ C3
e2t
O
1
+ C4
e2t
+ C5
e2t
te2t
e2t
O
1
= eKle2t
+ C2
Klte2t
e2t
O
O
O
O
+ C3K2e2t + C4K3e2t + C5
K3te2t
e2t
O
1
There are three solutions of the form X
the form X
= Kte2t + Pe2t
Ke2t,
where K is an eigenvector, and two solutions of
See (12) in the texto From (13) and (14) in the text
(A - 2I)K1 = O
and
394
Exercises
This implies
O O O
PI
O O O O O
O O O O O
P2
O O O O 1
O O O O O
P4
P5
O 1
so
P2 =
1 and
P5
= O, while
P3
Choosing
PI, P3, and P4 are arbitrary.
PI
= P3 = P4 = O we
1
O
p=
O
O
O
Therefore
a solution
is
x=
Repeating
O te2t
e2t.
for K3 we find
O
O
p=
O
O
1
so another
solution
is
x=
O te2t
+ 2y)2 = 4 sin2 2t
51. From x = 2 cos 2t - 2 sin 2t, y = - cos 2t we find x

(x
+ 2y
e2t.
= -2 sin 2t. Then
= 4(1 - cos2 2t) = 4 - 4 cos2 2t
=4-
4y2
and
or
395
X2
+ 4xy + 8y2
= 4.
have
8.2
Exercises 8.2
This is arotated conic section and, from the discriminant b2
4ac
= 16 - 32 < O, we see that the
curve is an ellipse.
if3, f3 > O. In Problem 36 the eigenvalues are 5 3i, in

Problem 37 they are 3i, and in Problem 38 they are -1 2i. Frorn Problem 47 we deduce that
52. Suppose the eigenvalues are a
the phase portrait will consist of a family of closed curves when a

origin will be a repellor when a
=O
and spirals when a =1- O. The
> O, and an attractor when a < O.
Exercises 8.3
1. From
X'-
( 3 -3) X+ (4)
2 -2
-1
we obtain
Then
and
so that
U=
q,-lF dt
and
Xp
= <llU
q,-l
-2
= ( e-t
J ( -11) = (-l1t)
= ( -11) t + (-15) .
-11 -10
=
5e-t
and
q,-l
dt
-5e-t
2. From
we obtain
Then
so that
U
q,- I F dt
(-2te-t)
2tet
and
396
3 -t
-e
2
_let
dt = (2te-t
2tet
+ 2e-t)
-
2et
Exercses 8.3
3. From
X' (3 -5) X + ( 1)
=
3/4 -1
et/2
-1
we obtain
Then
so that
U
<p-1F dt
J ( 3 -t)
~e1i
dt
and
Xp = <pU =
4.
-13/2)
( -13/4
te
t/2
+(
(3~4~t-t)
(-15/2)
/
-9 4
t/2
e.
From
X'
= (2
-1)
2
sin 2t )
2 cos2t
we obtain
-sin 2t )
Xc = el (
2 cos 2t
+ e2 (COS
2t ) 2t
e .
2 sin 2t
2t
Then
<p-
_e2t
(
sin 2t
2e2t cos2t
cos 2t)
2e2t sin 2t
e2t
and
so that
U=
<p-1Fdt =
and
Xp = <pU =
J (~
1
<p-1
_le-2t
~e-2t
cos 4t ) dt =
"2 sin 4t
e-2t
cos 2t
1
e-2t sin 2t
4
(i
sin 4t )
1
-;o; cos 4t
-k sin 2.tcos 4t - icos 2t cos 4t)

( icos 2t sin 4t - isin 2t cos 4t
5. From
X'
sin 2t
= (
O 2) X + ( -11)
-1 3
we obtain
X, = elle
( 2)
(1)
+ e2 1
Then
and
-1
<p.
397
et
e 2t .
-t
e
-2t
-e
2t
e .
cos 2t)
Exercises
8.3
so that
and
6. From
X'=(_~
~)X+C~3t)
we obtain
x, = el
et
(~)
+ e2
e)
e2t.
Then
cp -_ (2et t
e
e2t)
2t
e
and
cp-1
e-t)
~-2t
-t
(
_:-2t
so that
U=
cP -1 Fdt --
J(
2e-t - e-4t
_2e-2t + 2e-5t ) dt
( 4)
-2e
-t
1 -4t)
+ ".te
= (e -2t - se
2 -5t
and
7. From
we obtain
X,
el
1 e3t + e2
(-2)
1 e-
Then
4e3t
cP
= ( e
3t
_2e-3t)
-3t
e
so that
3t.
and
cP
-1
-3t
6"e
_ e3t
3t
(2t- e -3t - 3"2e-3t)
3t
)
dt
=
2t e 3t -:3 e3t
6te
6te-
and
Xp
= cpU =
O t + (-4/3)
-4/3 .
( -12)
8. From
398
Exercses 8.3
we obtain
X
= CI
( 4)
1
e3t
+ C2
Then
q,
= (
4e3t
e
3t
so that
and
9. From
we obtain
Then
so that
and
10. From
we obtain
Then
so that
399
(-2)
1
e- 3t .
Exercises 8.3
and
3).
xp = eu = ( -5
11. From
XI _
- (O
( sec t )
O
O X+
-1 )
we obtain
cos t )
+ e2
= el ( sin t
Xc
(sin t )
_ cos t .
Then
<P=
cost
( sin t
sin t )
-cost
<P
and
-1
sin t )
- cost
(cos t
. t
Slll
so that
and
t cos t
Xp
+ sin t In I sec ti)
= <P U = ( t sin t _ cos t In I sec ti'
12. From
( 1 -1) X+ (3)
X' -
et
we obtain
X = el (
c
-sin t ) t
e
cos t
t)
+ e2 ( cos
. te.
sin
Then
<P = (
- sin
') t
cos t
COS t
. t
sin
<P -1 = ( - sin
an
cos t
t)
.
cos
sin t
e -t
so that
U =
J <P
-1F
dt
-3 sin t
3 cos t
+ 3 cos t)
+ 3 sin t
and
Xp = <pU
13. From
I
XII
dt
( -3)
(3 cos t + 3 sin t)
3 sin t - 3 cos t
e.t
(1 -1) X + (c~St) e
sin t
we obtain
X=cr
c
-sin t ) t
e+c2.
cos t
400
cos t )
te.
sin
Exercises 8.3
Then
= (
-sin
t)
cos
sin t
cos t
so that
u=
_]
-] F dt =
J (~)
and
cos
Xp = U = ( .
smt
14. From
XI =

and
( -
cost) e-t
sin t
dt = (~)
t)
te .
(2 -2) X+-e(1) 1
8 -6
sin
cos t
-2t
we obtain
Then
1 2t+1)
= ( 2
4t
+1
-2t
_]

and
(-(4t+1)
2-1
2t
+ 2 In t )
_ In t
dt
2t+1)e2t
so that
U =
-] F dt =
J(
and
2t
X
p
= U = (
4t
+ In t -
2t In t)
+ 3In t - 4t In t
_ 2t
15. From
we obtain
Xc = CI ( cos t ) + C2 ( sin t )
- sin t
cos t
Then
cos t
= (
- sin t
sin t )
t
cos t
so that
U =
and
Xp=U=
- I F dt =
and
J (-
tan2
tan t
t ) t+. (-
cos
.
( - sin
t
-
t)
t)
sin
sin t tan t
401
= ( cos t - sin
sin t
dt =
t)
cos t
(t - tan t)
In I sec ti
( sin t ) ln l sec t],

cos t
Exercises 8.3
16. From
we obtain
cos t )
Xc = C1 (
.
-sm t
Then
cp=
cos
( - sin t
U=
t)
sin
cos t
+ C2 ( sin t )
cost
1 (COS t
and
cp- =
O)
dt =
- sin
t)
cost
sin t
so that
and
cp-1F dt =
J(
ese t
In I ese t - cot ti
sin t In I ese t - cot ti.)

Xp = cpU = (
.
costIn I csct - cot r]
17. From
XI =
(1-1/21 2) X+ ( sect
ese t ) e
we obtain
2 sin
Xc = cr (
t)
cost
+ C2 ( 2 cos
.
-smt
t)
e.
Then
cp= (
2 sin t
cos t
2 cos t) t
e
- sin t
Md
cP
-1
= ( ~sin t
~ cos t
cos t) -s
e
- sin t
so that
U =
cP -1 F dt =
J ~
(
J2
) dt = (
2
cos t - tan t
~ In I sin ti - In I sec ti
and
Xp = cpU = ( 33 sin t ) tet +

'2 cos t
18. From
I
X=
(COS1.
t ) In I sin ti + ( - 2 cos
. t ) In I sec r],
- '2 sin t
sm t
(1
we obtain
Xc = C1 (
Then
cp =
(COS t -
sin t
cos t
cos t - sin t )
cost
cos t + sin t )
sin t
+ C2
and
402
( cos t + sin t )
.
sin t
cP-
= (-
sin t
cost
t)
cos + sin
sin t - cost
Exercises
so that
U J cp F dt J (2 cos t + sin t - sec t)

=
-1
dt =
2 sin t - cos t
and
Xp = cpU = (
(2 sin t - cos t - In I sec t + tan ti)

- 2 cos t - sin t
3 sin t cos t - cos2 t - 2 sin2 t + (sin t - cos t) In I sec t

sin2 t - cos2 t - cos t(ln I sec t + tan ti)
+ tan ti)
19. From
we obtain
Then
1
cp =
e2t
-1 e2t
so that
and
20. From
x' =
(~
-:
1 -1
=~) X + ( ~ )
1
2et
we obtain
Then
403
-t
e-2t
-e
-e-2t
8.3
Exercises 8.3
so that
U
.
q>-lF dt =
(te-t +
-2e-t
2) dt = (-te-
- e-t
2e-t
21t e -2t
- t e -2t
+ 2t)
+ ;re
1 -2t
and
Xp=q>U=
-1/2)
-1
( -1/2
21. Frorn
,(3
X=
we obtain
q> (_e
=
-1) X +
-1
e2t)
2t
e
t
4t
(-3/4)
-1
-3/4
t+
(2)2
O
et+
(2)2
2
teto
(4e2t)
4e4t
'
and
X = q>q>-l(O)X(O)+ q>
r t q>-lFds
Jo
= q>.
(O) + (eq>.
2t
2t
+ 2t + 2t -
1)
= (~) te2t + ( -: ) e2t + ( -~) te4t + (~) e4t.

22. From
X' (1 -1) X + (1/l/t t )

=
1 -1
we obtain
q>=(11
l+t)
t
q>_l=(-t
'
l+t)
1
-1
'
and
X = q>q>-l(l)X(l)
+ q>
q>-IFds =
e (-:)
+ e C~t)
= (~)
t - (~)
23. Solving
2- A
3
1 = A2 - 1 = (A - l)(A
+ 1) = O
-1 -2 - A
we obtain eigenvalues Al = -1 and A2 = 1. Corresponding eigenvectors are
1
Thus
404
C)
lnt.
Exercises
8.3
Substituting
into the system yields
-al
from which we obtain
al =
-1 and
X(t)
2b1 = -5,
h = 3. Then
= el (-:
) e-t
+ e2 (-;)
et
+ ( -~) .
24. Solving
6- A
3- A
+ 14
= A2 - 9A
we obtain the eigenvalues Al = 2 and A2
Thus
X, = el
(A - 2)(A - 7)
=O
= 7. Corresponding eigenvectors are
( 1)
-4
e 2t
+ e2
(1)
1 e 7t .
Substituting
4al
+ 3bl
b2
+ 4 = O.
= 6. Substituting these two
give al = -~ and bl = ~. Then
Solving the first two equations simultaneously yields a2 = -2 and b2

values into the second pair of equations and solving for al and b:
X(t) = el ( _~)
e2t + e2 (~)
25. Solving
405
e7t + ( -:) t + (
-J) .
Exercises 8.3
we obtain the eigenvalues
)11
= 3 and
A2
7, 'Corresponding eigenvectors are
Thus
1)
x c = c1 ( -3
+ c2
e3t
(1)
9 e 7t ,
Substituting
X, =
b
( al)

1
from which we obtain
3a1
+ 3'b1
'9a1
+ 5b1 = -10
=3
al = 55/36 and bl = -19/4,

X(t) =
1)
( -3 e3t
el
Then
+ C2 ( 91)
e7t
( 55/36 )
-19/4
et,
26. Solving
-1-A
1
we obtain the eigenvalues Al
=A2+4=0
1- A
-1
= 2i and A2 = -2i, Corresponding eigenvectors are

K _ (
1-
1 + 2i
and
K2
= (
5 ),
1 - 2i
Thus
5 cos 2t
)
Xc = Cl (
cos 2t - 2 sin 2t
+ C2 (
5 sin 2t
)
2 cos 2t + sin 2t
Substituting
-al
+ 5bl + a2 + 1 =
-al
+ b: + oz = O
406
'
Exercises
from which we obtain a2

X(t)
27. Let I =
= -3,
oz
= -2/3,
al
= -1/3,
and
br = 1/3. Then
5cos2t
) + e2 (
5sin2t
) + (-3)
= el (
cos 2t - 2 sin 2t
2 cos 2t + sin 2t
-2/3
C~)
cost
+ (-1/3)
1/3
so that
and
Then
3e-12t )
-e-12t '
U = IJ> - 1Fdt
= (lOe2tSint)
30e12tsint
dt = (2e2t(2sint-cost))
26ge12t(12sint - cost)
and
Ip = IJ> U =
so that
1= (1) ee1
If 1(0) = (~) then el = 2 and e2 =
fg.
332sin t 29
( 276sin t 29
2t
+ e2
76cos t )
29
168cos t
29
3) e- + I .
( -1
12t
28. (a) Solving

1- A -1 1
1- A
1 -1
= A2-
2A = A(A- 2)
= O.
we obtain the eigenvalues Al = O and A2 = 2. Corresponding eigenvectors are
Kl = [~]
Thus
and
[ 1] [-1]
X, = el 1 + e2
If we substitute
407
1 e 2t .
sint.
8.3
Exercises
8.3
into the system, we get
[0] [1 -1]
-1
[al]
bl
[3]
-5
-br+3]
+ bl -
[al
-al
5 '
which cannot be solved for al and bi. The difference in the systems is that in Problem 23 the
homogeneous system does not have a constant solution, whereas in this problem it does.
(b) Trying a solution of the form
fails, so we instead try
Substituting into the system, we obtain

(al(-al
and note that a2
hence b2
-1.
= a2
+ bl - 5) + (-a2 + b2)t = b:
= bz. Adding the equations, we find -2 = a2 + b: = 2a2, so a2 = -1 and

From the first equation, we have al - bl + 3 = -1, and from the second
equation, we have -al

al
bl +3) + (a2 - b2)t
+ bl
and bl = 4. Thus
- 5 = -1. These are both equivalent to al - bl = -4, so we take
29. (a) The eigenvalues are O, 1,3, and 4, with corresponding eigenvectors
and
(b)
e=
( ~6
-4
2et
3e3t
et
e3t
2t
e3t
~e")
<p-l
-e
1
3
_,
1 -4t
-"3e
(e) <P-l(t)F(t) =
408
1 -t
"3e
O
2
"3e
-4t
-"3
-2e-t
2
"3e
-3t
"3
2t
e-
_\ e-3t
1 -4t
"3e
Exercises 8.3
4>(t)C + 4>(t)
J 4>-l(t)F(t)dt
409
Exercises 8.4
Exercises 8.4
1. For A
= (~ ~) we have
A2=(~ ~)(~ ~)=(~
~),
A3=AA2=(~
~)(~ ~)=(~
~),
A =AA3=C
~)(~ ~)=(~
16)'
and so on. In general

Ak
= (~
20k)
for
= 1, 2, 3,
o o o
Thus
t2
t3
2!
3!
l+t+-+-+o.o
(
and
(2t)2
(2t)3
2!
_At=(e-
2. For A
l+t+--+--+
0)
e-2t
3!
= (~ ~) we have
A2=(~ ~)(~ ~)=(~ ~)=l

A3 = AA2 = (~ ~) 1
=A
A (A2)2= 1
A = AA Al = A
= (~
410
~)
) (e
=
Exercises 8.4
and so on. In general
k
{A,
1,
k
k
= 1, 3, 5,
= 2, 4, 6,
o o o
o
o o o
Thus
e
At
A
A2
= 1+ -t
+ -t
1!
2!
= 1+ At
A3 3
+ -t
+ooo
3!
+ 2.It2 + 2.At3 + o o o
2!
3!
= 1 (1 + 2.
t2 + 2.t4 + o o o) + A (t + 2.t3 + 2.t5 + o o o)
21
4!
3!
51
cosh t
sinh t )
= Icosht + Asinht = (
sinh t cosh t
and
e-At = (COSh(-t)
sinh(-t)
sinh( -t))
cosh(-t)
= (
cosh t
-sinht
- sinh t)
cosht
3. For
A=( ~ ~ ~)
-2
-2
-2
A2 = (~
-2
-2
we have
Thus, A 3 = A 4 = A 5
eAt = 1 + At =
o o o
~) (~
-2
-2
-2
~ ~)
O O O
O and
O o1
O) + (
O
O O 1
A~U
O
O
1
') ('+1
tt
t
t
-2t -2t -2t
4. For
we have
~) = (~
-2
n
411
t
-2t
t
t+1
-2t -2t + 1
Exercises 8.4
Thus, A 4
= AS = A 6 = ... = O and
eAt
= 1 + At + -21A 2t2
e 0) e
= o
o o
3t
5t
o
o o
t o
0) +
o
o o o =
~t2 o o
0) ( 1
3t
~t2 + 5t
5. Using the result of Problem 1,
x= ( eot
G~)=
e~t )
ci (~)
+ C2
(~
) .
x = (COSht
sinh t
t)
sinh
cosh t
(q)
t)
= c (COSh
+ C2 (sinh
sinh t
C2
t) .
cosh t
X =
t +t
-2t
t)
t +t 1
t
-2t -2t + 1
(Cl)
C2
= C!
( t +t 1) +
-2t
C3

10
X =
3t
( ~t2 + 5t
9. To solve
412
C2
(t)t
+ 1 + C3 (t)
-2t
-2t
+1
Exercises 8.4
we identify to
= 0, F(s) = (_~), and use the results of Problem 1 and equation (5) in the texto
X(t) = eAtC+eAt
{t e-ASF(s)ds
Jto
) (e- e-2s0) ( 3) ds
{t
= (~
e2t Jo
{t ( 3e-s
e2t Jo
_e-2s
-1
ds
10. To solve
we identify to
= 0, F(s) = (e~t)'
X(t)
= eAtC + eAt
and use the results of Problem 1 and equation (5) in the texto
(t e-ASF(s) ds
i;
413
Exercises 8.4
11. To solve
= O, F( s) = (~),
we identify to
= (COSht
sinh t
(Cl
Cl
(Cl
Cl
sinh t
+ (COSht
(Cl)
C2
cosh t + C2 sinh t)
sinh t + C2 cosh t
= (Cl cosh t
Cl
sinh t)
cosh t
and use the results of Problem 2 and equation (5) in the texto
sinh t)
cosh t
sinh t
+ (COSht
sinh t)
cosh t
sinh t
+ C2 sinh t) + (COSht
+ C2 cosh t
sinh t
cosh t + C2 sinh t)
sinh t + C2 cosh t
cosh S
(t (
Jo
cosh s
sinh s
-'- sinh s
- sinh s) (1) ds
cosh s
1
s)
- sinh
coshs
sinh t) ( sinh s - cosh s )

cosh t
- cosh s + sinh s
+ (COSht
ds
t
O
sinh t) ( sinh t - cosh t )

cosh t
- cosh t + sinh t
sinh t
2
+ C2 sinh t) + (Sinh t Cl sinh t + C2 cosh t

sinh2 t -
= (Cl cosh t
(t (
Jo
cosh t) = ci (COSht)
cosh2 t
sinh t
+ C2 (Sinh t) _
(1)
cosh t
12. To solve
X' =
we identify to
= O, F(s)
(C~Sh t
sinh t
Cl
( Cl
sinh t)
cosh t
sinh t
1) X
O
+ (COSht )
sinh t
t)
COSh
, and use the results of Problem 2 and equation (5) in the texto
smht
(Cl)
C2
cosh t + C2 sinh t )
sinh t + C2 cosh t
= (Cl cosh t
Cl
=. (
(O1
+ (C~Sh t
sinh t
sinh t)
cosh t
(t (
Jo
cosh s
sinh s
( cosh t sinh t) (t (1 ) ds
sinh t cosh t Jo O
+.
+ C2 sinh t) + (COSht
+ C2 cosh t
sinh t
(Os)
cSlonshhtt)
414
Ito
S)
- sinh
coshs
S)
(C~Sh
sinh s
ds
Exercses 8.4
Cl cosh
= ( Cl sinh
t + C2sinh t )
t + C2cosh t +
Cl cosh
= (
Cl sinh
t + C2sinh t ) + ( t cosh t )
t + C2cosh t
t sinh t
(COSh t
sinh t
t) (t)
t
O
sinh
cosh
= Cl
(COSh t )
sinh t
+ C2 ( sinh t ) + t ( cosh t) .
cosh
13. We have
Thus, the solution
of the initial-value
problem is
X=(t:1)_4(t~1)+6(
-2t
-2t
~ ).
-2t + 1
14. We have
X(O) = C3 (~)
+ C4 (~) + (
~3)=
e:~)
= (~) .
Thus, C3 = 7 and C4 = ~, so
x = 7 ( ~ ) et + ~ (~)
S-
15. From sI - A =
4'
e2t
+ ( -~ ) .
-3 ) we find
s+4
3/2
(sI-A)-l
1/2
(S-2 - s+2
-1
1
--+-s-2
s+2
and
The general solution of the system is then
415
3/4 _ 3/4 )
s-2
-1/2
s-2
s+2
3/2
s+2
sinh
Exercises 8.4
16. From sI _ A
3/2)
e (
-1
(S - 4
.-1
2t
+ e
(-1/2)
-2t
+ C2
(3/4)
/
-1 2
(sI-A)-l
(S~3 ~ S~2
s-3
and
eAt
s-2
__2 +_2 )
s-3
s-2
-1
--+-s-3
s-2
2e3t _ e2t
- ( e3t _ e2t
_
= (e - C2) (~)
(S -1- 5
+ C2
2)
we find
s- 1
17. From sI _ A
2t
C3 (~)
e3t + (-e
e3t + C4
e)
e2t.
9 ) we find
8+1
(sI - A)-l =
_1
s-2
(
3
(s-2)2
1
(s - 2)2
416
+ 2C2)
e)
e2t
(-3/4)
/
32
-2t
Exercises 8.4
and
18. Frorn sI - A
= (~ s
q (
1+
t
3t) e
2t
+C2
(-9t)
1 - 3t
e 2t .
~12)
we find
(sI - A)-
and
At
(e-
s+l+l
(s+1)2+1
=
-2
(
e-
cos t + t sin t
-2e-t sint
(S+1)2+1)
s+l+l
(s+1)2+1
:' sin t
)
:' cos t - :' sin t .
X=e
At
C=
(e-
Cl
e-
cos t + t sin t
-2e-t sin t
cos t + sin t) -t
e
- 2 sin t
:' sin t
) (CC21)
:' cos t - e-t sin t
+ C2 (
sin t
cos t - sin t
417
)-t
Exercises 8.4
19. Solving
2-'\
-3
we find eigenvalues '\1
= 3 and
1
6 -,\
'\2
=,\2 _ 8,\ + 15= (,\ - 3)('\ - 5) = O
= 5.
Corresponding eigenvectors are
Then
1 1)
P = (1 3 '
-1
(3/2
-1/2
-1/2)
1/2 '
and
D = (~ ~),
so that
PDP-l = (
2 1).
-3 6
20. Solving
we find eigenvalues '\1
2-'\
1
= 1 and
1
2 -,\
'\2
= 3.
=,\2 _ 4A+ 3 = (,\ - 1)('\ - 3) = O

Corresponding eigenvectors are
Then
P= (-1 1)
p_l=(-1/21/2)
1 1 '
1/2
1/2 '
and
= (~ ~)
so that
21. From equation (3) in the text

etA
= etPDP- = 1 + t(PDP-l) + 2_t2(PDP-l)2 + 2_t3(PDP-l)3 + ...

2!
=P
3!
[1+ tD + i(tD)2 + i(tD)3 + ...] p-l = PetDp-l.

2.
3.
418
Exercises
22. From equation (3) in the text

O
etD =
f) +
(!
O
A2
O
O)
An
e
+2" O
1 2
O
A~
1
3!
+ -t
1 + A2t
+ ~(A2t)2 + ...
O
O
23. From Problems 19, 21, and 22, and equation (1) in the text
X = etAC = PetDp-1C
24. From Problems 20-22 and equation (1) in the text

X
= etAC = PetDp-1C
419
JJ
(,3
01 A~
:
jJ+
8.4
Exercses 8.4
25. If det(sI - A) = O, then s is an eigenvalue of A. Thus sI - A has an inverse if s is not an eigenvalue
of A. For the purposes of the discussion in this section, we take s to be larger than the largest
eigenvalue of A. Under this condition sI - A has an inverse.
26. Since A 3
= 0, A is nilpotent. Since
e At = 1 + At
+ A 2 -t2 + ... + A t -tk + ...
if A is nilpotent and A m = 0, then A k =

eAt = 1 + At
In this problem A 3
k!
2!
for k ::::m and
t2
+ A 2 - + ... + A m-l
2!
tm-l
(m - 1)!
= 0, so
2
eAt = 1+ At + A t; = [~ ~ ~] +
O O 1
t - t2/2
-t
[
-t - t2/2
1-
and the solution of X'
[=~ ~ ~] t + [- ~ ~ O1]2!_
-1
1 1
t + t2/2
~ 1+
-1
O 1
.; t2/2
= AX is
Cl]
C2
X(t) = eAtC = eAt

[
C3
[Cl(1-t-t2/2)+C2t+C3(t+t2/2)]
-Clt + C2 + C3t
Cl(-t_t2/2)+C2t+C3(1+t+t2/2)
27. (a) The following commands can be used in Mathematica:

A={ {4, 2},{3, 3}};
e={ el, e2};

m=MatrixExp[A
t];
soleeExpandjm.c]
Colleet[sol, {el, e2}] / / MatrixForm
The output gives
y(t) =
ci ( _~et
+ ~e6t) + C2 Get + ~e6t)
The eigenvalues are 1 and 6 with corresponding eigenvectors
420

so the solution of the system is
X(t) = bl (-~)
et
+ b:
C) e
6t
or
x(t)
= -2blet + b2e6t
y(t) = 3blet
If we replace bl with -iCl
+ iC2
+ b2t.
and b2 with tCl
+ ~C2, we obtain
the solution found using the
matrix exponential.
(b ) x(t) = cle-2tcost-
(Cl +c2)e-2tsint
y(t) = C2e-2t cos t
+ (2Cl + c2)e-2t
sin t
+ C3(-6e-2t + 6e-t)
= c2(4e-2t - 3e-t) + c4(4e-2t - 4e-t)
= CI (e-2t - e-t) + C3(_2e-2t + 3e-t)
28. x(t) = cI(3e-2t - 2e-t)

y(t)
z(t)
w(t) = C2(-3e-2t
+ 3e-t) + C4(_3e-2t + 4e-t)

1. If X
=k
( : ), then X'
We see that k
2. Solving for
CI
= O and
= ~.
and
C2
we find
CI = - ~
and
C2 =
i.
3. Since
( -1; -4~ -3~)( -1~)=(1:)=4(~)

-4 -1'
we see that A = 4 is an eigenvalue with eigenvector K3. The corresponding solution is X = K3t.
4. The other eigenvalue is A2
= 1 - 2i with corresponding eigenvector K2
solution is
2t ) t
( sin 2t)
t
X(t) = CI ( cos
.
e + C2
e.
- sin 2t
cos 2t
421
= (_~).
The general
5. We have det(A - Al) = (A - 1)2 = O and K = ( _~). A .solution to (A - AI)P = K is P =
e)
so that
= (A + 6)(A + 2)
X
Xl =
= ci
= Oso that
e-6t +e2
(_~)
C)
e-2t.
+ 5 = O. For A = 1 + 2i we obtain
2A
( 1) e
i
(1+2i)t =
(cos 2t) e t
- sin 2t
K = (~) and
+ 2. (sin 2t) e t .
cos 2t
Then
Xl = (
32
= el ( cos 2t)
- sin 2t
2A
+ 2 = O. For A = 1 +
i) e
(l+i)t
sin 2t) t
e.
- cos 2t
+ e? (
we obtain k, = (3 ~
= (3 cos t + sin t) e t + 2. (
2 cos t
i)
and
+ 3 sin t) e.t
2 sin t
- cos t
Then
X
= el (
-(A - 2)(A - 4)(A
10. We have det(A - Al) = -(8

and A2
3 cos t + sin t ) t
e
2 cos t
+ 2)(82 -
+ e2
( -
cos t + 3 sin t ) t
e .
2 sin t
+ 3) = Oso that
28 + 3) = O. The eigenvalues are Al = -2, A2 = 1 + V2i,
= 1 - V2i, with eigenvectors
422

Thus
x ~"
en [m ';'v2t - ( ~/2)
~+o, m sinv2t]
e-"
H2
v2t+
H3 [ (
=C1
- 7)
5 e-2t+c2
(COS
...;2t)
-...;2sin...;2tj2
v2t] e'
sin
e'
(Sin
...;2t )
...;2cos...;2tj2 et.
et+c3
cos ...;2t
11. We have
x c = c1
(1)
sin ...;2t
e2t + c2
(4)
1
e4t.
Then
cp = (e2t
O
and
U=
cp-1 F dt =
cp-1 =
4e4t)
e4t '
e-2t
(
_4e-2t)
<t '
2e-2t _ 64te-2t )
( 15e-2t + 32te-2t )
dt =
16te-4t
_e-4t - 4te-4t
'
so that
11 + 16t)
Xp = cpU = (
.
-1 - 4t
12. We have
Xc =
Then
cp=
and
2 cos t
( - sin t
-1
cp F dt
so that
Xp
Cl
t)
2 cos
( - sin t
t)
2 sin
cos t
+ C2 (
cp-1
'
(cost-sect)
sin t
dt =
2 sin
cost
t) e .
t
t - sin t)
(~C~S
~ sm t
(Sint-Inlsect+tantl)
- cos t
- 2 cos In I sect + tan
ti)
= cpU = (
.
- 1 + sin t In I sec t + tan ti
423
e-t,
cost
'

13. We have
X = Cl (
cos t + sin t )
2 cos t
+ C2 ( sin t -
cos t )
2 sin t
Then
q,
= (cos t + sin t sin t ~ cos t) , q, -1 =

2 cos t
2 sin t
and
q,-lFdt=
U=
J
= (-~ cos t
~ cos t -
sin t'
- cos t
!cos t - ~ sin t )
!cos t + ~sin t
'
1 sin t - 1 cos t + 1 ese t )

2
2
2
dt
(
J - ~ sin t - ~ cos t + ~ese t
i sin t + ~In I ese t -
i sin t + !In I ese t -
cot ti)
cot ti
'
so that
Xp = q,U = ( -1)
-1
+.(
sin t
) In I ese t - cot ti
sin t + cos t
14. We have
Then
q,-l=
t -2t -e -2t)
-e
-e
t -2t
e-u
e-u
'
and
t-1)
( -1
U=Jq,-lFdt=J
dt=
(lt2-t)
2_t
'
so that
X =q,U=
p
(-1/2)t2e2t+
1/2
(-2)te2t.
1
15. (a) Letting
we note that (A - 2I)K = O implies that 3kl

k2
= 0, k3 = 1 and then
k2
+ 3k2 + 3k3
= 1, k3 = we get
424
= 0, so kl
-(k2
+ k3)'
Choosing

respectively. Thus,
are two solutions.

(b) Frorn det(A - Al) = A2(3 - A)
= O we see that
Al
= 3,
and O is an eigenvalue of multiplicity
two. Letting
K = (~~) ,
k3
as in part (a), we note that (A - OI)K = AK
-(k2
+ k3).
Choosing k2
O implies that kl + k2 + k3 = O, so kl
= O, k3 = 1, and then k2 = 1, k3 = O we get
respectively. Since the eigenvector corresponding to Al
the general solution of the system is
16. For X = (~~)et we have X'
= X = IX.
425
= 3 is
Chapter .8 Related Exercises
1. (a) In operator notation we have
Xl
(t)
CI COS
- 2X2
=O
2Xl - (D2
+ 2)X2
= O.
+ 6D2 + 4)XI =
Eliminating X2 we have (D4

)3 - v5i and )3
(D2 +4)Xl
+ v5i,
O. The roots of the auxiliary equation are
so
)3 - v5 t +
C2
sin )3 - v5 t +
+ v5 t +
C3 cos )3
C4
sin )3 + v5 t.
Then
1
+ 4)Xl
X2(t) = 2(D2
(t)
= 1 + v5CI COS )3 - v5 t + 1 + v5C2 sin )3 - v5 t

2
1-v5
+ -2-C3
The frequencies are
WI =
/
r:
cos V 3 + V 5 t
)3 - v5 and
W2 =
)3
1-v5
2
C4 sin
/
~
V 3 - v 5 t.
+ v5.
(b) We write the system in the form

X" =
[-4 2] X
2 -2
and assume solutions of the form

X = Vcoswt
Since the eigenvalues are -3
+ v5
and
X = Vsinwt
and -3 - v5,
WI
)3 - v5 and
W2
corresponding eigenvectors are
VI = [(1 + ~)/2]
and
Then, the general solution of the system is

X(t)
=CI
[(1+~)/2]
[(1+~)/2]
cosV3-v5t+C2
2. If K is multiplied by a factor of 10, then
426
sinV3-v5t
)3
+ v5.
The
-10
A = M-1K=
-10
A
ol
T
-10
-10
-10
-10
2
-10
-10
The table below shows the eigenvalues

periods
Ai
-10
with the corresponding frequencies
-10
Wi =
r-xi
and
Ti = 21r/ Wi.
-13 .8117 -13.2639 -12.4064 -11.3175 -10.0974

3.7164
1.6907
3.6420
1.7252
3.5223
1.7838
3.1776
1.9773
3.3641
1.8677
-8.8604
-7.7256
-6.8082
-6.2089
-1.5000
2.9766
2.1108
2.7795
2.2605
2.6093
2.4080
2.4918
2.5216
1.2247
5.1302
We see that the frequencies are higher and consequently, the periods are smaller.
3. (a) For this exercise we assume a building of four stories, each with mass 10,000 kg and restoring
force constants each k, = 5,000 kg/s2 Then
1 O
M = 10,000
1 O
-10,000
5,000 -10,000
K=
O O 1 O
O O O 1
5,000
5,000
5,000 -10,000
5,000
5,000 -5,000
and
-1
A=
0.5 O
0.5 -1 0.5
O 0.5 -1
O
0.5
O
O
0.5
-0.5
(b) The table below shows the eigenvalues Ai with the corresponding frequencies Wi =
periods
Ti = 21r/ wl .
A
Ol
T,
-1.7660
-1.1737
-0.5000
-0.0603
1.3289
4.7280
1.0834
5.7998
0.7071
8.8858
0.2456
25.5855
427
r-xi
and
Chapter
8 Related Exercises
Since the periods are all larger than 4, the building should be safe in a period-2 earthquake.
(e) If k is multiplied by 10, then
-100
50
A=
50
-100
50
50
-100
50
50
-50
From the table below we see that the periods are generaUy in the range from 1.5 to 3 and the
building is unsafe from a period-2 earthquake.
A.
w
T
-17.660
-11.737
-5.000
-0.603
4.2024
1.4951
3.4259
1.8340
2.2361
2.8099
0.7766
8.0909
Experimenting, we see that if the building is .made stiffer by a factor of 8, the periods are
entering the danger zone.
A.
w
T
-14.128
-9.389
-4.000
-0.482
3.7588
1.67l6
3.0642
2.0505
2.0000
3.1416
0.6946
9.0459
4. The related homogeneous system is X" = M-1KX = AX, where A is shown in the solution of
Problem 3(a). We assume solutions of the form X
are -1. 7660, -l.1737,
= V cosw and X
= V sinwt. The eigenvalues
-0.5, and -0.0603 with eorresponding frequencies 1.3289, 1.0834, 0.7071,
and 0.2456. The eigenvectors are

-0.4285
0.6565
-0.6565
0.2280
0.5774
0.5774
0.4285
0.5774
0.5774
0.2280
V2=
-0.5774
0.2280
and
-0.4285
V4 =

X( t) = C1V 1 cos l.3289t
+ C7V4
+ C2V 1 sin 1.3289t + C3V 2 cos l.0834t
cosO.2456t
+ ca V4
sin 0.2456t.
Since
10000cos 3t
O
F(t) =
O
O
428
-0.5774
0.6565
Chapter
8 Related Exercises
to solve the nonhomogeneous system we look for a particular solution of the form
bl
al
a2
Xp =
cos3t
a3
a4
Substituting into X" = AX
+ M-IF(t)
-9al = 1 - al
b2
b3
b4
sin3t.
we obtain
+ 0.5a2
and
Solving these systems gives al
-0.12549,
a2 =
0.00787,
a3 =
0.00049,
a4
0.00003, and
= b: = b3 = o = O. Thus,
bl
-0.12549
X p -
0.00787
cos3t.
-0.00049
0.00003
= O and X' (O) = O, we find Cl = -0.0592, C2 = O, C3 = -0.0839,

= 0.0679, C6 = O, C7 = 0.0255, and C8 = O. Thus, the solution of the initial-value problem
Applying the initial conditions X(O)

C4
= O,
Cs
is
-0.4285
X(t)
-0.0592
-0.6565
0.5774
0.6565
cos 1.3289t - 0.0839
0.2280
0.5774
0.4285
0.2280
0.2280
-0.5774
+ 0.0679
0.5774
O
cos 1.0834t
cos 0.0679t
+ 0.0255
-0.4285
-0.5774
0.6565
0.5774
429
-0.12549
cos 0.2456t
0.00787
-0.00049
0.00003
cos 3t.
Numerical Solutions of
Ordinary Differential Equations
Exercises 9.1
AH tables in this chapter were constructed in a spreadsheet program which does not support subscripts.
Consequently,
1.
2.
h -
x(n)
1.00
1.10
1.20
1.30
1.40
1.50
h -
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
3.
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
Xn
and
0.1
y(n)
5.0000
3.9900
3.2546
2.7236
2.3451
2.0801
0.1
y(n)
2.0000
1.6600
1.4172
1.2541
1.1564
1.1122
0.1
y(n)
0.0000
0.1005
0.2030
0.3098
0.4234
0.5470
Yn
will be indicated as x(n) and y(n), respectively.

h = 0.05
x(n)
y(n)
1.00
5.0000
4.4475
1.05
3.9763
1.10
3.5751
1.15
3.2342
1.20
2.9425
1.25
2.7009
1.30
2.4952
1.35
2.3226
1.40
1.45
2.1786
1.50
2.0592
x(n)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.05
y(n)
2.0000
1.8150
1.6571
1.5237
1.4124
1.3212
1.2482
1.1916
1.1499
1.1217
1.1056
h = 0.05
x(n)
y(n)
0.00
0.0000
0.05
0.0501
0.10
0.1004
0.15
0.1512
0.20
0.2028
0.25
0.2554
0.30
0.3095
0.35
0.3652
0.40
0.4230
0.45
0.4832
0.50
0.5465
430
Exercses 9. 1
4.
x(n}
0.00
0.10
0.20
0.30
0.40
0.50
5.
0.1
y(n}
0.0000
0.0952
0.1822
0.2622
0.3363
0.4053
0.1
yen}
0.0000
0.0050
0.0200
0.0451
0.0805
0.1266
0.1
yen}
0.5000
0.5215
0.5362
0.5449
0.5490
0.5503
x(n}
0.00
0.10
0.20
0.30
0.40
0.50
6.
x(n}
0.00
0.10
0.20
0.30
0.40
0.50
7.
0.1
y(n}
1.0000
1.1110
1.2515
1.4361
1.6880
2.0488
x(n}
0.00
0.10
0.20
0.30
0.40
0.50
h -
x(n}
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
- 0.05
x(n}
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
y(n}
0.0000
0.0488
0.0953
0.1397
0.1823
0.2231
0.2623
0.3001
0.3364
0.3715
0.4054
0.05
y(n}
0.0000
0.0013
0.0050
0.0113
0.0200
0.0313
0.0451
0.0615
0.0805
0.1022
0.1266
0.05
yen}
0.5000
0.5116
0.5214
0.5294
0.5359
0.5408
0.5444
0.5469
0.5484
0.5492
0.5495
x(n}
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
x(n}
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.05
yen}
1.0000
1.0526
1.1113
1.1775
1.2526
1.3388
1.4387
1.5556
1.6939
1.8598
2.0619
431
Exercises 9. 1
8.
9.
10.
h = 0.1
x(n)
y(n)
0.00
1.0000
0.10
1.1079
0.20
1.2337
0.30
1.3806
0.40
1.5529
0.50
1.7557
x(n)
1.00
1.10
1.20
1.30
1.40
1.50
0.1
yen)
1.0000
1.0095
1.0404
1.0967
1.1866
1.3260
h x(n)
0.00
0.10
0.20
0.30
0.40
0.50
0.1
y(n}
0.5000
0.5250
0.5498
0.5744
0.5986
0.6224
11.
h
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
h = 0.05
x(n)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
x(n)
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1. 35
1.40
1. 45
1. 50
h =
x(n)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
y(n)
1.0000
1.0519
1.1079
1.1684
1.2337
1.3043
1.3807
1.4634
1.5530
1.6503
1.7560
0.05
y(n)
1.0000
1.0024
1.0100
1.0228
1.0414
1.0663
1.0984
1. 1389
1.1895
1.2526
1.3315
0.05
y(n}
0.5000
0.5125
0.5250
0.5374
0.5498
0.5622
0.5744
0.5866
0.5987
0.6106
0.6224
0.1
y(n}
2.0000
2.1220
2.3049
2.5858
3.0378
3.8254
exact
2.0000
2.1230
2.3085
2.5958
3.0650
3.9082
x(n)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
h - 0.05
y(n}
2.0000
2.0553
2.1228
2.2056
2.3075
2.4342
2.5931
2.7953
3.0574
3.4057
3.8840
432
exact
2.0000
2.0554
2.1230
2.2061
2.3085
2.4358
2.5958
2.7997
3.0650
3.4189
3.9082
Exercises
12. (a)
(b)
h=O.l
20
EULER
x{nl
1.00
1.10
1.20
1. 30
1.40
15
10
5
9. 1
IMPROVED
EULER
.v.uu.
_y (nl_
1.0000
1.2000
1.4938
1.9711
2.9060
1.0000
1.2469
1.6668
2.6427
8.7988
--~~====;=~~---x
1.1 1.2 1.3 1.4
13. (a) Using the Euler method we obtain y(O.l)
~ Yl = 1.2.
(b) Using y" = 4e2x we see that the local truncation error is
y"(c) h2 = 4e2c (0.1)2 = 0.02e2c.

2
Since e2x is an increasing function, e2c :::;e2(0.1) = eO.2 for O :::;e :::;0.1. Thus an upper bound
for the local truncation error is 0.02eO.2 = 0.0244.
(e) Since y(O.l)
= eO.2 = 1.2214, the actual error is y(O.l) - Yl =
(d) Using the Euler method with h = 0.05 we obtain y(O.l)
~ Y2
0.0214, which is less than 0.0244.
1.21.
(e) The error in (d) is 1.2214 - 1.21 = 0.0114. With global truncation error O(h), when the step
size is halved we expect the error for h
= 0.05 to be one-half the error when h =
0.1. Comparing
0.0114 with 0.214 we see that this is the case.
14. (a) Using the improved Euler method we obtain y(O.l) ~ Yl

(b) Using y/"
1.22.
= 8e2x we see that the local truncation error is

y"/(c) h3
6
= 8e2c
(0.1)3
= 0.001333e2c.
Since e2x is an increasing function, e2c :::;e2(0.1)= eO.2 for O :::;e:::; 0.1. Thus an upper bound
for the local truncation error is 0.001333eo.2 = 0.001628.
(e) Since y(O.l)
= eO.2 =
l.221403,
the actual error is y(O.l)
- Yl
0.001403 which is less than
0.001628.
(d) Using the improved Euler method with h

(e) The error in (d) is 1.221403 -1.221025
0.05 we obtain y(O.l)
~ Y2 = 1.221025.
= 0.000378. With global truncation error O(h2),
when
the step size is halved we expect the error for h = 0.05 to be one-fourth the error for h = 0.1.
Comparing 0.000378 with 0.001403 we see that this is the case.
15. (a) Using the Euler method we obtain y(O.l) ~ Yl
0.8.
(b) Using y" = 5e-2x we see that the local truncation error is
5e-~c (0.1)2
2
= 0.025e-2c.
433
Exercises 9. 1
Since e-2x is a decreasing function, e-2c :S ea = 1 for O-:S e :S 0.1. Thus an upper bound for
the local truncation error is 0.025(1)
(e) Since y(O.l)
= 0.025.
= 0.8234, the actual error is y(O.l) - Yl = 0.0234, which is less than 0.025.
(d) Using the Euler method with h = 0.05 we obtain y(O.l) ~ Y2 = 0.8125.
(e) The error in (d) is 0.8234 - 0.8125 = 0.0109. With global truncation error O(h), when the
step size is halved we expect the error for h
= 0.05 to be one-half the error when h. = 0.1.
Comparing 0.0109 with 0.0234 we see that this is the case.

16. (a) Using the improved Euler method we obtain y(O.l) ~ Yl
(b) Using
y'" =
= 0.825.
-10e-2x we see that the local truncation error is

lOe-2c (0.1)3 = 0.001667e-2c.
6
Since e-2x is a decreasing function, e-2c :S ea
= 1 for O :S e :S 0.1. Thus an upper bound for
the local truncation error is 0.001667(1) = 0.001667.

(e) Since y(O.l)
= 0.823413, the actual error is y(O.l) -Yl
= 0.001587, which is less than 0.001667.
= 0.05 we obtain y(O.l) ~ Y2 = 0.82378l.

(e) The error in (d) is 10.823413- 0.82371811 = 0.000305. With global truncation error O(h2),
when the step size is halved we expect the error for h = 0.05 to be one-fourth the error when
h = 0.1. Comparing 0.000305 with 0.001587 we see that this is the case.
(d) Using the improved Euler method with h
17. (a) Using y" = 38e-3(x-l)
we see that the local truncation error is

yi/(c) h2 = 38e-3(c-l) h2 = 19h2e-3(c-l).
2
2
is a decreasing function for 1 :S x :S 1.5, e-3(c-l) :S e-3(1-1)
(b) Since e-3(x-l)
1 for 1 :S c:S 1.5
and
y"(c) ~2 :S 19(0.1)2(1) = 0.19.
(e) Using the Euler method with h
= 0.1 we obtain y(1.5) ~ 1.8207. With h = 0.05
W(!
obtain
y(1.5) ~ 1.9424.
(d) Since y(1.5)
2.0532, the error for h = 0.1 is Ea.l = 0.2325, while the error for h
0.05 is
Ea.a5 = 0.1109. With global truncation error O(h) we expect Ea. 1/ Ea.a5 ~ 2. We actually have
Ea. 1/ Ea.a5 = 2.10.
18. (a) Using
y'"
= _114e-3(x-l)
434
Exercses 9. 1
(b) Since e-3(x-l)
is a decreasing function for 1 ::; x ::;1.5, e-3(c-l)
::; e-3(1-1) =
1 for 1 ::; e ::; 1.5
and
lylll(C) ~3 ::; 19(0.1)3(1) = 0.019.
(e) Using the improved Euler method with h
= 0.1 we obtain y(1.5) ~ 2.080108. With h
= 0.05
we obtain y(1.5) ~ 2.059166.

(d) Since y(1.5)
= 2.053216, the
error for h
= 0.1 is EO.l = 0.026892, while the
error for h = 0.05 is
Eo.os = 0.005950. With global truncation error O(h2) we expect Eo.1/ Eo.os ~ 4. We actuaJly
have Eo.1/ Eo.os = 4.52.
19. ()a
11
1
Usmg y = - (x + 1)2 we see that the local truncation error is
lyll(C) ~21 =
(e:
1)2 ~2.
(b) Since (x ~ 1)2 is a decreasing function for O::; x::; 0.5, (e: 1)2 ::; (O: 1)2 = 1 for O::; e::; 0.5
and
yll(C) ~21 ::; (1) (0.;)2 = 0.005.
(e) Using the Euler method with h = 0.1 we obtain y(0.5) ~ 0.4198. With h
y(0.5) ~ 0.4124.
= 0.05 we obtain
= 0.4055, the error for h = 0.1 is EO.l = 0.0143, while the error for h = 0.05 is
Eo.os = 0.0069. With global truncation error O(h) we expect Eo.1/ Eo.os ~ 2. We actuaJly have
Eo.1/ Eo.os = 2.06.
.
111
2
20. ( a ) U sing y = (x + 1)3 we see that the local truncaton error is
(d ) Since y(0.5)
111
y (e)
h3
6" =
1
h3
(e + 1)3 3"
(b) Since (x ~ 1)3 is adecreasing function for O::; x::; 0.5, (e: 1)3 ::; (O: 1)3 = 1 for O::; e::; 0.5
and
y"I(C) ~3 ::; (1) (0.;)3
(e) Using the improved Euler method with h
= 0.000333.
= 0.1 we obtain y(0.5) ~ 0.405281. With h = 0.05
we obtain y(0.5) ~ 0.405419.

(d) Since y(0.5)
= 0.405465, the
error for h = 0.1 is EO.l
= 0.000184, while the
error for h
= 0.05 is
Eo.os = 0.000046. With global truncation error O(h2) we expect Eo.1/ EO.05 ~ 4. We actually
have Eo.1/ EO.05 = 3.98.
435
Exercises 9. 1
21. Because Y~+l depends on Yn and is used to determine Yn+l, all of the Y~ cannot be computed at one
time independently of the corresponding 'Yn values. For example, the computation of Y4 involves
the value of Y3.
Exercises 9.2
1.
h = 0.1
y(n}
2.0000
2.1230
2.3085
2.5958
3.0649
3.9078
x(nJ
0.00
0.10
0.20
0.30
0.40
0.50
exact
2.0000
2.1230
2.3085
2.5958
3.0650
3.9082
2. Setting ex = 1/4 we find b
= 2,
= -1 and 3 = 1/4. The resulting second-order Runge-Kutta
method is
Yn+ 1 = Yn - k1
k1
= hf(xn, Yn)
k2 = hf(xn
h = 0.1
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
R-K
y(n)
2.0000
2.1205
2.3006
2.5759
3.0152
3.7693
+ 2k2
+ h/4,
Yn
+ kj4).
IMP EULER
y(n)
2.0000
2.1220
2.3049
2.5858
3.0378
3.8254
3.
x(n)
1. 00
1.10
1.20
1.30
1.40
1.50
v(n)
5.0000
3.9724
3.2284
2.6945
2.3163
2.0533
4.
x(n)
0.00
o . 10
0.20
0.30
0.40
0.50
v(n)
2.0000
1.6562
1.4110
1.2465
1.1480
1.1037
5.
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
v(n)
0.0000
0.1003
0.2027
0.3093
0.4228
0.5463
6.
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
v(n)
1.0000
1.1115
1.2530
1.4397
1.6961
2.0670
7.
x(nJ
0.00
0.10
0.20
0.30
0.40
0.50
v(n)
0.0000
0.0953
0.1823
0.2624
0.3365
0.4055
8.
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
v tn)
0.0000
0.0050
0.0200
0.0451
0.0805
0.1266
436
Exercises
9.
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
v(n)
0.5000
0.5213
0.5358
0.5443
0.5482
0.5493
ll.
x(n)
l. 00
1.10
1.20
1.30
1.40
1.50
1.0000
1.0101
1.0417
1.0989
1.1905
1.3333
(n)
10.
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
y(n)
1.0000
1.1079
1.2337
1.3807
1.5531
1.7561
12.
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
(n)
0.5000
0.5250
0.5498
0.5744
0.5987
0.6225
13. (a) Write the equation in the form

dv
dt
(b)
t n)
0.0
1.0
2.0
3.0
4.0
5.0
= 32 - 0.025v = f(t, v).
v(t)
v(n)
0.0000
25.2570
32.9390
34.9772
35.5503
35.7128
40
30
20
~0---1--~2--~3---4--~5t
(e) Separating variables and using partial fractions we have
--
1 (1
2V32 V32 -
JO.025 v
1)
V32 + JO.025
dv
= dt
and
2
V32~ 0.025 (ln 1V32 + JO.025

.
vi - ln
1V32 -
Since v(O) = O we find e = O. Solving for v we obtain

v(t) =
16v5 (ev'TIt - 1)
ev'TIt + 1
and v(5) ~ 35.7678.
437
JO.025 vi) = t
+ c.
9.2
Exercises
9.2
14. (a) See the table in part (e) of this problem.
(b) From the graph we estimate A(1) ~ 1.68, A(2) ~ 13.2, A(3) ~ 36.8, A(4) ~ 46.9, and
A(5) ~ 48.9.
A(t)
50
40
30
20
10
(e) Let a = 2.128 and (3 = 0.0432. Separating variables we obtain

dA
A(a - (3A)
1(1
-a -A + a
(3)
- (3A
1
-[In A -ln(a
a
dA
- (3A)]
A
In a _ (3A
= dt
= dt
= t +e
= a(t
+ e)
= eo(t+c)
a - (3A
A
= aeo(t+c)
- (3Aeo(t+c)
Thus
aeo(t+c)
a
a
- ....,--------;
- 1+ (3eo(t+c) - (3 + e-o(t+c) - (3 + e-oce-ot
A(t) -
From A(O) = 0.24 we obtain

a
0.24 = (3 + e-OC
so that
e-oc =
ajO.24 - (3 ~ 8.8235 and
438
Exercises 9.2
A( ) ~
2.128
t ~ 0.0432 + 8.8235e-2 128t .
t
A
A
A
(da s)
(observed)
(approximated)
(exact)
15. (a)
h - 0.05
x(n)
y(n)
1.00
1.05
1.10
1.15
1.20
1.25
1. 30
1.35
1.40
1.0000
1.1112
1.2511
1.4348
1.6934
2.1047
2.9560
7.8981
1.06E+15
1
2.78
1.93
1.95
2
13.53
12.50
12.64
3
36.30
36.46
36.63
4
47.50
47.23
47.32
5
49.40
49.00
49.02
(b)
h - 0.1
v(n)
1.0000
20
1.2511
15
1.6934
10
2.9425
--~========~------x
903.0282
1.1
16. (a) Using the fourth-order Runge-Kutta method we obtain y(O.l)
(b) Using
y(5)
(x)
;::::Yl
1.2
1.3
1.4
1.2214.
= 32e2x we see that the local truncation error is

y(5)(c)'!!_
120
= 32e2c (0.1)5
120
= 0.000002667e2c.
Since e2x is an increasing function, e2c :S e2(O.1) = eO.2 for O :S e :S 0.1. Thus an upper bound
for the local truncation error is 0.000002667eo.2
(e) Since y(O.l)
= eO.2
= 0.000003257.
= 1.221402758, the actual error is y(O.l) - tn
= 0.000002758 which is less
than 0.000003257.
(d) Using the Iourth-order Runge-Kutta
formula with h = 0.05 we obtain y(O.l)
;::::Y2 =
1.221402571.
(e) The error in (d) is 1.221402758 - 1.221402571 = 0.000000187. With global truncation error
O(h4), when the step size is halved we expect the error for h = 0.05 to be one-sixteenth the
= 0.1. Comparing 0.000000187 with 0.000002758 we see that this is the case.
(a) Using the fourth-order Runge-Kutta method we obtain y(O.l) ;::::Yl = 0.823416667.
(b) Using y(5) (x) = -40e-2x we see that the local truncation error is
error for h
17.
40e-2c (0.1)5 = 0.000003333.

120
Since e-2x is a decreasing function, e-2c :S
eO
the local truncation error is 0.000003333(1)
= 0.000003333.
439
= 1 for O :S e
:S 0.1. Thus an upper bound for
Exercises
9.2
(e) Since y(O.l) = 0.823413441, the actual error is ly(O.l) - Yll = 0.000003225, which is less than
0.000003333.
(d) Using the fourth-order Runge-Kutta method with h
= 0.05 we obtain y(O.l)
~ Y2
0.823413627.
(e) The error in (d) is 10.823413441- 0.8234136271 = 0.000000185. With global truncation error
O(h4),
when the step size is halved we expect the error for h
error when h
= 0.1. Comparing 0.000000185 with 0.000003225 we see that this is the case.
18. (a) Using y(5) = -1026e-3(x-l)
l
(b) Since e-3(x-l)
= 0.05 to be one-sixteenth the
y(5)(C).!i!_1 = 8.55h5e-3(c-l).
120
is a decreasing function for 1 :S x:s 1.5, e-3(c-l) :S e-3(1-1)
= 1 for 1 :S c:S 1.5
and
y(5) (c) 1~50:S 8.55(0.1)5 (1) = 0.0000855.
(e) Using the fourth-order Runge-Kutta method with h = 0.1 we obtain y(1.5) ~ 2.053338827.
With h = 0.05 we obtain y(1.5) ~ 2.053222989.
.
()
24
19. (a) Usmg y 5 =
(x + 1)5
(5)
y
(b) Since (x
+ 1)5 is a decreasing
c .!i!_ _
1
h5
() 120 - (c + 1)5 5
1
1
function for O :S x :S 0.5, (c + 1)5 :S (O + 1)5 = 1 for O :S c:S 0,5
and
y(5)(C) ~5 :S (1) (0.;)5
= 0.000002.
(e) Using the fourth-order Runge-Kutta method with h = 0.1 we obtain y(0.5) ~ 0.405465168.
With h = 0.05 we obtain y(0.5) ~ 0.405465111.
20. Each step of the Euler method requires only 1 function evaluation, while each step of the improved
Euler method requires 2 function evaluations - once at (xn, Yn) and again at (Xn+l, y~+l)'
The
second-order Runge-Kutta methods require 2 function evaluations per step, while the fourth-order
Runge-Kutta method requires 4 function evaluations per step. To compare the methods we approximate the solution of y' = (x
+y
- 1)2, y(O) = 2, at x = 0.2 using h = 0.1 for the Runge-Kutta
method, h = 0.05 for the improved Euler method, and h = 0.025 for the Euler method. For each
method a total of 8 function evaluations is required. By comparing with the exact solution we see
that the fourth-order Runge-Kutta method appears to still give the most accurate resulto
440
Exercises
EULER
IMP EUL
R-K
0.025 h = 0.05 h = 0.1
y(n)
y_(n)
y(n)
2.0000
2.0000
2.0250
2.0526
2.0553
2.0830
2.1165
2.1228
2.1230
2.1535
2.1943
2.2056
2.2395
2.2895
2.3075
2.3085
9.2
h
x(n)
0.000
0.025
0.050
0.075
0.100
0.125
0.150
0.175
0.200
exact
2.0000
2.0263
2.0554
2.0875
2.1230
2.1624
2.2061
2.2546
2.3085
y
21. (a) Por y'
+y =
10 sin 3x an integrating factor is eX so that
d
dx [eXy] = 10ex sin 3x ~
eXy
~
When x
3 cos 3x
= O, y
= O, so O = -3
+ 3e-x.
= eX sin 3x
+e
- 3ex cos 3x
y = sin 3x - 3 cos 3x
+ c and c =
+ ce-x.
3. The solution is y = sin 3x -
Using Newton's method we find that x = 1.53235 is the only
positive root in [0,2].

(b) Using the fourth-order Runge-Kutta method with h
= 0.1 we obtain the table ofvalues shown.
These values are used to obtain an interpolating function in Mathematica.

interpolating function is shown. Using Mathematica's
only positive root in [O,2Jis x
x(n)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
y(n)
0.0000
0.1440
0.5448
1.1409
1.8559
2.6049
3.3019
3.8675
4.2356
4.3593
4.2147
x(n)
1.0
1.1
1.2
1.3
1.4
1.5
1'.6
1.7
1.8
1.9
2.0
root finding capability we see that the
= 1.53236.
y
y(n)
4.2147
3.8033
3.1513
2.3076
1.3390
0.3243
-0.6530
-1.5117
-2.1809
-2.6061
-2.7539
441
The graph of the
Exercises 9.3
Exercises 9.3
1. For y' - y = x - 1 an integrating factor is e- J da: = e-x, so that
~[e-Xyl
dx
= (x
- l)e-X
and
y
=e
X(
-xe-x
+ e) =
-x
+ ce".
From y(O) = 1 we find e = 1 and y = -x + e". Comparing exact values with approximations
obtained in Example 1, we find y(0.2)
::::::1.02140276 compared to Yl = 1.02140000, y(O.4) ::::::
1.09182470 compared to Y2 = 1.09181796, y(0.6) ::::::1.22211880 compared to Y3 = 1.22210646, and

y(0.8) ::::::1.42554093 compared to Y4 = 1.42552788.
2. 100 REM ADAMS-BASHFORTHj ADAMS-MOULTON

110 REM METHOD TO SOLVE Y'=FNF(X,Y)
120 REM DEFINE FNF(X,Y) HERE
130 REM GET INPUTS
140 PRINT
150 INPUT "STEP SIZE=", H
160 INPUT "NUMBER OF STEPS (AT LEAST 4)=",N
170 IF N <4 GOTO 160
180 INPUT "XO =" ,X
190 INPUT "YO =" ,Y
200 PRINT
210 REM SET UP TABLE
220 PRINT "X", "Y"
230 PRINT
240 REM COMPUTE 3 ITERATES USING RUNGE-KUTTA
250 DIM Z(4)
260 Z(l)=Y
270 FOR I=1 TO 3
280 Kl=H*FNF(X,Y)
290 K2=H*FNF(X+Hj2,Y
300 K3=H*FNF(X+Hj2,Y
+K1j2)
+K2j2)
310 K4=H*FNF(X+H,Y+K3)
320 Y=Y+(Kl+2*K2+2*K3+K4)j6
330 Z(I+l)+ Y
442
Exercises
9.3
340 X=X+H
350 PRINT X,Y
360 NEXT 1
370 REM COMPUTE REMA1N1NG X AND Y VALUES
380 FOR 1=4 TO N
390 YP= Y+ H*(55*FNF(X,Z( 4))-59*FNF(X-H,Z(3) )+37*FNF(X-2*H,Z(2))
-9*FNF(X-3*H,Z(1)) )/24
400 Y= Y+ H* (9*FNF(X + H,YP)+ 19*FNF(X,Z( 4))-5*FNF(X-H,Z(3))+ FNF(X-2*H,Z(2))) /24
410 X=X+H
420 PRINT X,Y
430 Z(1)=Z(2)
440 Z(2)=Z(3)
450 Z(3)=Z(4)
460 Z(4)=Y
470 NEXT 1
480 END
3.
x (n)
0.00
0.20
0.40
0.60
0.80
4.
x iri)
0.00
0.20
0.40
0.60
0.80
y(n)
1.0000
0.7328
0.6461
0.6585
0.7332
0.7232
y(n)
2.0000
1.4112
1.1483
1.1039
1.2109
1.2049
initia1 condition
Runge-Kutta
Runge-Kutta
Runge-Kutta
predictor
corrector
initia1 condition
Runge-Kutta
Runge-Kutta
Runge-Kutta
predictor
corrector
443
Exercises
5.
x(n)
0.00
0.20
0.40
0.60
0.80
1.00
9.3
y(n)
0.0000
0.2027
0.4228
0.6841
1.0234
1.0297
1.5376
1.5569
initial condition
Runge-Kutta
Runge-Kutta
Runge-Kutta
predictor
corrector
predictor
corrector
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
6.
x(n)
0.00
0.20
0.40
0.60
0.80
1.00
y(n)
1.0000
1.4414
1.9719
2.6028
3.3483
3.3486
4.2276
4.2280
x (n)
initial condition
Runge-Kutta
Runge-Kutta
Runge-Kutta
predictor
corrector
predictor
corrector
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
444
'y(n)
0.0000
0.1003
0.2027
0.3093
0.4227
0.4228
0.5462
0.5463
0.6840
0.6842
0.8420
0.8423
1.0292
1.0297
1.2592
1.2603
1.5555
1.5576
initial condition
Runge-Kutta
Runge-Kutta
Runge-Kutta
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
y(n)
1.0000
1.2102
1.4414
1.6949
1.9719
1.9719
2.2740
2.2740
2.6028
2.6028
2.9603
2.9603
3.3486
3.3486
3.7703
3.7703
4.2280
4.2280
initial condition
Runge-Kutta
Runge-Kutta
Runge-Kutta
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
Exercises 9.3
7.
x(n)
0.00
0.20
0.40
0.60
0.80
1. 00
y(n)
0.0000
0.0026
0.0201
0.0630
0.1362
0.1360
0.2379
0.2385
initial condition
Runge-Kutta
Runge-Kutta
Runge-Kutta
predictor
corrector
predictor
corrector
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1. 00
8.
x tn)
0.00
0.20
0.40
0.60
0.80
1.00
y(n)
1.0000
1.2337
1.5531
1.9961
2.6180
2.6214
3.5151
3.5208
initial condition
Runge-Kutta
Runge-Kutta
Runge-Kutta
predictor
corrector
predictor
corrector
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
1.00
445
y(n)
0.0000
0.0003
0.0026
0.0087
0.0201
0.0200
0.0379
0.0379
0.0630
0.0629
0.0956
0.0956
0.1359
0.1360
0.1837
0.1837
0.2384
0.2384
initial condition
Runge-Kutta
Runge-Kutta
Runge-Kutta
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
v tn)
1.0000
1.1079
1.2337
1.3807
1.5530
1.5531
1.7560
1.7561
1.9960
1.9961
2.2811
2.2812
2.6211
2.6213
3.0289
3.0291
3.5203
3.5207
initial condition
Runge-Kutta
Runge-Kutta
Runge-Kutta
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
predictor
corrector
Exercises 9.4
Exercises 9.4
1. The substitution
y'
= u leads to the
Yn+l = Yn
The initial conditions are Yo
= -2
_
iteration formulas
+ hun,
Un+l = Un
+ h(4un
and Uo = 1. Then
u. = Yo + O.luo
= -2
+ 0.1(1) =
-1.9
Ul = Uo + 0.1(4uo - 4yo) = 1 + 0.1(4

Y2 = u. + O.lUl = -1.9
+ 0.1(2.2)
The general solution of the differential equation is y

we find C1
= -2 and
C2 = 5.
- 4Yn).
= -1.68.
= C1e2x + C2xe2x. From the initial conditions
= _2e2x + 5xe2x
Thus y
+ 8) = 2.2
and y(0.2) ;:::;1.4918.
2. The substitution y' = u leads to the iteration formulas
+ hun,
Yn+l = Yn
The initial conditions are yO
=4
Un+
1 = un + h (~ Un - :2 Yn) .
and Uo = 9. Then
Yl
= yo + O.luo
= 4 + 0.1(9)
Ul
= Uo + 0.1 iUo - iYO = 9 + 0.1[2(9) - 2(4)]
(2
Y2 = Yl
+ O.lUl
= 4.9
2)
= 4.9
+ 0.1(10) = 5.9.
The general solution of the Cauchy-Euler differential equation is y

conditions we find C1 = -1 and
3. The substitution
C2 = 5.
= 10
Thus y
= -x + 5x2 and y(1.2)
C1X
+ c2x2.
From the initial
= 6.
y' = u leads to the system

y' = u,
u'
= 4u - 4y.
Using formula (4) in the text with x corresponding to t, y corresponding to x, and u corresponding
to y, we obtain
RUllfLe-Kuttamethod with h=0.2
ml
m2
m3
m4
kl
k2
k3
k4
0.2000
0.4400
0.5280
0.9072
2.4000
3.2800
3.5360
4.8064
ml
m2
m3
m4
kl
k2
k3
k4
0.1000
0.2443
0.1600
0.3298
0.1710
0.3444
0.2452
0.4487
1.2000
1.7099
1.4200
2.0031
1.4520
2.0446
1.7124
2.3900
Runge-Kutta
446
0.00 -2.0000
0.20 -1.4928
1.0000
4.4731
method with h=O.l

x
0.00 -2.0000 1.0000

0.10 -1.8321 2.4427
0.20 -1.4919 4.4753
Exercises
4. The substitution y'
9.4
= u leads to the system

,2
2
u = -u - -y.
y' = u,
x2
Using formula (4) in the text with x corresponding to t, y corresponding to x, and u corresponding
to
y,
we obtain
Runge-Kutta
method with h=0.2
ml
m2
m3
m4
kl
k2
k3
k4
1.8000
2.0000
2.0017
2.1973
2.0000
2.0165
1.9865
1.9950
1.00
1.20
4.0000
6.0001
9.0000
11.0002
Runge-Kutta
method with h=O.l
ml
m2
m3
m4
kl
k2
k3
k4
0.9000
1.0000
0.9500
1.0500
0.9501
1.0501
0.9998
1.0998
1.0000
1.0000
1.0023
1.0019
0.9979
0.9983
0.9996
0.9997
1.00
1.10
1.20
4.0000
4.9500
6.0000
9.0000
10.0000
11.0000
5. The substitution y' = u leads to the system

u'
y' = u,
Using formula (4) in the text with
= 2u - 2y + etcost.
corresponding to x and u corresponding to

Runge-Kutta
ml
0.4000
m2
0.4600
m3
0.4660
m4
0.5320
kl
0.6000
k2
0.6599
k3
k4
0.6599
ml
0.2000
0.2315
m2
0.2150
0.2480
m3
0.2157
0.2487
m4
0.2315
0.2659
kl
k2
0.3000
0.3299
0.3150
0.3446
kl
k2
k3
0.3150
0.3444
6.
ml
10.0000
8.7500
10.1563
13.2617
17.9712
m2
m3
0.0000 12.5000
-2.500013.4375
-4.375017.0703
-6.3672 22.9443
-8.8867 31.3507
m4
0.0000
-20.0000
-28.7500 -5.0000
-40.0000 -8.7500
-55.1758 -12.7344
-75.9326 -17.7734
5.0000
4.3750
5.0781
6.6309
8.9856
447
k4
k3
0.3298
0.3587
we obtain
method with h=0.2

t
0.00
O .7170 0.20
Runge-Kutta
y,
1.0000
1.4640
2.0000
2.6594
method with h=O.l

t
0.00 1.0000
0.10 1.2155
0.20 1.4640
u
2.0000
2.3150
2.6594
Runge-Kutta method with h=O.l

i2
k4
t
il
0.00 0.0000 0.0000

22.5000
-5.0000
0.10 2.5000 3.7500
29.6875
0.20
-10.6250
2.8125 5.7813
-16.0156 40.3516 0.30 2.0703 7.4023
-22.5488 55.3076 0.40 0.6104 9.1919
-31.2024 75.9821 0.50 -1.5619 11.4877
Exercises 9.4
i1,i2
7
i1
i2
2
1
As t -;
00
5t
we see that il(t) -; 6.75 and i2(t) -; 3.4.
7.
Runge-Kutta
method with h=0.2
ml
m2
m3
m4
kl
k2
k3
k4
2.0000
2.2800
2.3160
2.6408
1.2000
1.4000
1.4280
1.6632
0.00
0.20
6.0000
8.3055
2.0000
3.4199
ml
m2
m3
m4
kl
k2
k3
k4
1.0000
1.1494
1.0700
1.2289
1.0745
1.2340
1.1496
1.3193
0.6000
0.7073
0.6500
0.7648
0.6535
0.7688
0.7075
0.8307
Runge-Kutta
x,y
0.511.52t
448
method with h=O.l
0.00 6.0000
0.10 7.0731
0.20 8.3055
y
2.0000
2.6524
3.4199
Exercises
8.
Runge-Kutta
ml
m2
m3
m4
0.6000
0.9400
1.1060
1.7788
kl
k2
k3
k4
method with h=0.2

t
0.00 1.0000
1.40002.06002.39403.72120.202.07853.3382
Runqe-Kutta
ml
m2
m3
m4
kl
k2
k3
k4
0.3000
0.5193
0.3850
0.6582
0.4058
0.6925
0.5219
0.8828
0.7000
1.1291
0.8650
1.4024
0.9068
1.4711
9.4
1.0000
method with h=O.l

t
0.00 1.0000
1.1343 0.10 1.4006
1.84740.202.0845
1.0000
1.8963
3.3502
x,y
50
40 y(t)
30
9.
Runge-Kutta
m1
m2
m3
m4
kl
k2
k3
k4
method with h=0.2

t
0.00 -3.0000
-1.0000 -0.9200 -0.9080 -0.8176 -0.6000 -0.7200 -0.7120 -0.8216 0.20 -3.9123
Runge-Kutta
m1
m2
m3
m4
kl
k2
k3
k4
5.0000
4.2857
method with h=O.l

t
0.00 -3.0000
-0.5000 -0.4800 -0.4785 -0.4571 -0.3000 -0.3300 -0.3290 -0.3579 0.10 -3.4790
-0.4571 -0.4342 -0.4328 -0.4086 -0.3579 -0.3858 -0.3846 -0.4112 0.20 -3.9123
449
5.0000
4.6707
4.2857
Exercises
9.4
x,y
10.
Runqe-Kutta
ml
m2
m3
m4
kl
k2
k3
k4
0.6400
1.2760
1.7028
3.3558
1.3200
1.7720
2.1620
3.5794
Runge-Kutta
ml
m2
m3
m4
kl
k2
k3
k4
0.3200
0.7736
0.4790
1.0862
0.5324
1.1929
0.7816
1.6862
0.6600
1.0117
0.7730
1.2682
0.8218
1.3692
1.0195
1.7996
x,y
2
1.5
0.1
0.2
0.3t
11. Solving for x' and y' we obtain the system
+ y + 5t
x'
y'
= 2x + y - 2t.
-2x
450
method with h=0.2

t
0.00 0.5000
0.20 2.1589
y
0.2000
2.3279
method with h=O.l

t
0.00 0.5000
0.10 1.0207
0.20 2.1904
y
0.2000
1.0115
2.3592
Exercses
Runge-Kutta
mI
m2
m3
m4
kl
-0.8000 -0.5400 -0.6120 -0.3888
0.0000
k2
k3
k4
mI
m2
m3
kl
m4
method with h=0.2

t
0.00
-0.2000 -0.1680 -0.3584 0.20
Runge-Kutta
k2
k3
k4
9.4
1.0000
0.4179
-2.0000
-2.1824
method with h=O.l

t
0.00
-0.4000 -0.3350 -0.3440 -0.2858 0.0000 -0.0500 -0.0460 -0.0934 0.10
-0.2866 -0.2376 -0.2447 -0.2010 -0.0929 -0.1362 -0.1335 -0.1752 0.20
1.0000
0.6594
0.4173
-2.0000
-2.0476
-2.1821
x,y
12. Solving for
X'
4 t
and y' we obtain the system

X
= -y - 3t
2
+ 2t -
Runge-Kutta
mI
m2
m3
m4
-1.1000 -1.0110 -1.0115 -0.9349
kl
k2
k3
k4
1.1000
1.0910
1.0915
1.0949
Runge-Kutta
mI
m2
m3
m4
-0.5500 -0.5270 -0.5271 -0.5056

-0.5056 -0.4857 -0.4857 -0.4673
method with h=0.2

t
0.00
0.20
3.0000
1.9867
-1.0000
0.0933
method with h=O.l
kl
k2
k3
k4
0.5500
0.5456
0.5470
0.5457
0.5471
0.5457
0.5456
0.5473
0.00
0.10
0.20
3.0000
2.4727
1.9867
-1.0000
-0.4527
0.0933
451
Exercises 9.4
x,y
60
40
y(t)
20
4
-20
8 t
x(t)
-40
-60
Exercises 9.5
1. We identify P(x)
= o, Q(x) = 9, f(x)
= O, and h
(2 - 0)/4
0.5. Then the finite difference
equation is
Yi+l
+ 0.25Yi + Yi-l = O.
The solution of the corresponding linear system gives

0.0
4.0000
0.5
-5.6774
2. We identify P(x)
1.0
-2.5807
1.5
6.3226
2.0
1.0000
= O, Q(x) = -1, f(x) = x2, and h = (1 -
0)/4
= 0.25. Then the finite difference
equation is
+ Yi-l
Yi+l - 2.0625Yi
= 0.0625x;'

0.00
0.0000
0.25
-0.0172
3. We identify P(x)
0.50
-0.0316
2, Q(x)
0.75
-0.0324
1. 00
0.0000
= 1, f(x) = 5x, and h = (1 -
0)/5
equation is
1.2Yi+l - 1.96Yi
+ 0.8Yi-l = 0.04(5xi).

y
0.0
0.0000
0.2
-0.2259
0.4
-0.3356
0.6
-0.3308
0.8
-0.2167
1.0
0.0000
452
0.2. Then the finite difference
Exercises
9.5
4. We identify P(x) = -10, Q(x) = 25, f(x) = 1, and h = (1 - 0)/5 = 0.2. Then the finite difference
equation is
-Yi
+ 2Yi-l = 0.04.

0.0
1.0000
0.2
1.9600
0.4
3.8800
0.6
7.7200
0.8
15.4000
1.0
0.0000
5. We identify P(x) = -4, Q(x) = 4, f(x) = (1 + x)e2x, and h = (1 - 0)/6 = 0.1667. Then the finite
difference equation is
0.6667Yi+l - 1.8889Yi+ 1.3333Yi_l = 0.2778(1
+ xi)e2xi.

x
0.0000
3.0000
0.1667
3.3751
0.3333
3.6306
0.5000
3.6448
6. We identify P(x) = 5, Q(x) = O, f(x)
0.6667
3.2355
0.8333
2.1411
1.0000
0.0000
= 4y1x, and h = (2 - 1)/6
= 0.1667. Then the finite
1.4167Yi+l - 2Yi
+ 0.5833Yi_l = 0.2778(4.JXi).

x
1.0000
1.0000
1.1667 1.3333 1.5000 1.6667 1.8333 2.0000

-0.5918 -1.1626 -1.3070 -1.2704 -1.1541 -1.0000
7. We identify P(x) = 3/x, Q(x) = 3/x2, f(x)
= O, and
h = (2 - 1)/8 = 0.125. Then the finite
0.1875)
(
0.0469)
(
0.1875)
Yi+l+ -2+--2- Yi+ 1---.- Yi-l=O.
( 1+--.-X,
xi
x,
1.000
5.0000
1.125
3.8842
1.250
2.9640
8. We identify P(x) = -l/x,
1.375
2.2064
1.500
1.5826
1.625
1.0681
Q(x) = x-2, f(x) = lnx/x2,
1.750
0.6430
1.875
0.2913
2.000
0.0000
and h = (2 - 1)/8 = 0.125. Then the finite
0.0625)
(
- Yi+l + -2
( 1 - --o
xt
0.0156)
+ --2-
xi
(
0.0625)
Yi + 1 + --o - Yi-l
xt
0.0156lnxi.

1.000
0.0000
1.125
1.250
1.375
1.500
1.625
1.750
1.875
2.000
-0.1988 -0.4168 -0.6510 -0.8992 -1.1594 -1.4304 -1.7109 -2.0000
453
Exercises 9.5
9. We identify P(x)
= 1-x,
Q(x)
= x, f(x)
= (1-0)/10
= x, and h
equation is
[1 + 0.05(1 - Xi)]Yi+l
+ [-2 + O.OlXi]Yi + [1 -
0.05(1 - Xi)]Yi-l
= O.Olxi.

0.0
0.0000
0.1
0.2660
0.2
0.5097
0.3
0.7357
0.4
0.9471
I
0.5
1.1465
0.6
1.3353
0.7
1.5149
0.8
1.6855
0.9
1.8474
1.O
2.0000
= x, Q(x) = 1, f(x) = x, and h = (1 - 0)/10 = 0.1. Then the finite difference
equation is

0.0
1.0000
0.1
0.8929

equation is
0.2
0.7789
0.3
0.6615
= O, Q(x) = -4, f(x)
0.4
0.5440
0.5
0.4296
0.6
0.3216
0.7
: 0.2225
0.8
0.1347
= O, and h = (1 - 0)/8
Yi+l -
2.0625Yi
0.9
0.0601
1.O
O .0000 :
+ Yi-l = O.

x
y
0.000
0.0000
0.125
0.3492
12. We identify P(T)
0.250
0.7202
0.375
1.1363
2/T, Q(T) = O, f(T)
0.500
1.6233
0.625
2.2118
= O, and h
0.750
2.9386
= (4 - 1)/6
0.875
3.8490
1.000
5.0000
equation is
0.5)
( 1 + G Ui+l - 2Ui
+ (0.5)
1- G
Ui-l
= O.

u
1.0
2.0
3.0
3.5
4.0
1.5
2.5
50.0000 72.2222 83.3333 90.0000 94.4444 97.6190 100.0000
13. (a) The difference equation
(1 +
%Pi)
Yi+l
+ (-2 + h2Qi)Yi +
(1 -
%Pi)
Yi-l
= h2fi
is the same as the one derived on page 434 in the texto The equations are the same because
the derivation was based only on the differential equation, not the boundary conditions. If we
454

allow i to range from
Yn-l
o to
n - 1 we obtain n equations in the n + 1 unknowns Y-l, YO, Yl, ... ,
Since Yn is one of the given boundary conditions, it is not an unknown.
(b) Identifying Yo = y(O), Y-l
= y(O -
h), and Yl
= y(O + h) we have from (5) in the tcxt
21h[Yl - Y-d = y'(O) = 1 or
Yl - Y-l = 2h.
The difference equation corresponding to i = O,
(1+
~PO) Yl
+ (-2 + h2Qo)YO +
(1-
~PO) Y-l = h2fo
becomes, with Y-l = Yl - 2h,
(1 +
~Po) Yl
+ (-2 + h2Qo)YO +
(1 - ~po)
(Yl - 2h)
h2 fo
or
2m + (-2
+ h2Qo)yo
= h2 fo
Alternatively, we may simply add the equation Yl -Y-l

obtaining n + 1 equations in the
(e) Using n
ti
+ 2h -
Po.
= 2h to the list of n difference equations
+ 1 unknowns Y-l, YO, Yl, ... , Yn-l
= 5 we obtain
0.0
0.2
0.4
0.6
0.8
1.0
-2.2755 -2.0755 -1.8589 -1.6126 -1.3275 -1.0000
14. Using h = 0.1 and, after shooting a few times, y'(O) = 0.43535 we obtain the following table with
the fourth-order Runge-Kutta method.
0.0
1.00000
0.1
1.04561
0.2
1.09492
0.3
1.14714
0.4
1.20131
0.5
1.25633
0.6
1.31096
0.7
1.36392
0.8
1.41388
0.9
1.45962

1. ~
x(n)
1. 00
1.10
1.20
1.30
1.40
1.50
EULER
2.0000
2.1386
2.3097
2.5136
2.7504
3.0201
IMPROVED
EULER
2.0000
2.1549
2.3439
2.5672
2.8246
3.1157
RUNGE
KUTTA
2.0000
2.1556
2.3454
2.5695
2.8278
3.1197
h=0.05
x(n)
1. 00
1. 05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.45
1.50
EULER
2.0000
2.0693
2.1469
2.2328
2.3272
2.4299
2.5409
2.6604
2.7883
2.9245
3.0690
455
IMPROVED
EULER
2.0000
2.0735
2.1554
2.2459
2.3450
2.4527
2.5689
2.6937
2.8269
2.9686
3.1187
1.0
1.50003
_
RUNGE
KUTTA
2.0000
2.0736
2.1556
2.2462
2.3454
2.4532
2.5695
2.6944
2.8278
2.9696
3.1197

2. ~
x(n)
0.00
0.10
0.20
0.30
0.40
0.50
3.
.h=iU.
x(n)
0.50
0.60
0.70
0.80
0.90
1.00
4.
EULER
0.0000
0.1000
0.2010
0.3049
0.4135
0.5279
IMPROVED
EULER
0.0000
0.1005
0.2030
0.3092
0.4207
0.5382
RUNGE
KUTTA
0.0000
0.1003
0.2026
0.3087
0.4201
0.5376
h-0.05
x(n)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
EULER
0.0000
0.0500
0.1001
0.1506
0.2017
0.2537
0.3067
0.3610
0.4167
0.4739
0.5327
IMPROVED
EULER
0.0000
0.0501
0.1004
0.1512
0.2027
0.2552
0.3088
0.3638
0.4202
0.4782
0.5378
RUNGE
KUTTA
0.0000
0.0500
0.1003
0.1511
0.2026
0.2551
0.3087
0.3637
0.4201
0.4781
0.5376
EULER
0.5000
0.6000
0.7095
0.8283
0.9559
1.0921
IMPROVED
EULER
0.5000
0.6048
0.7191
0.8427
0.9752
1.1163
RUNGE
KUTTA
0.5000
0.6049
0.7194
0.8431
0.9757
1.1169
h=0.05
x(n)
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
EULER
0.5000
0.5500
0.6024
0.6573
0.7144
0.7739
0.8356
0.8996
0.9657
1.0340
1.1044
IMPROVED
EULER
0.5000
0.5512
0.6049
0.6609
0.7193
0.7800
0.8430
0.9082
0.9755
1.0451
1.1168
RUNGE
KUTTA
0.5000
0.5512
0.6049
0.6610
0.7194
0.7801
0.8431
0.9083
0.9757
1.0452
1.1169
EULER
1.0000
1.2000
1.4760
1.8710
2.4643
3.4165
IMPROVED
EULER
1.0000
1.2380
1.5910
2.1524
3.1458
5.2510
RUNGE
KUTTA
1.0000
1.2415
1.6036
2.1909
3.2745
5.8338
h=0.05
x(n)
1.00
1.05
1.10
1.15
1.20
1.25
1.30
1.35
1.40
1.45
1.50
EULER
1.0000
1.1000
1.2183
1.3595
1.5300
1.7389
1.9988
2.3284
2.7567
3.3296
4.1253
IMPROVED
EULER
1.0000
1.1091
1.2405
1.4010
1.6001
1.8523
2.1799
2.6197
3.2360
4.1528
5.6404
RUNGE
KUTTA
1.0000
1.1095
1.2415
1.4029
1.6036
1.8586
2.1911
2.6401
3.2755
4.2363
5.8446
1EQ_J.
x(n)
1.00
1.10
1.20
1.30
1.40
1.50
5. Using
Yn+l = Yn
+ hun,
YO = 3
Uo
= 1
we obtain (when h = 0.2) Yl = y(0.2) = YO + huo = 3 + (0.2)1 = 3.2. When h = 0.1 we have
Yl
= YO + O.luo = 3 + (0.1)1 = 3.1
Ul = Uo
Y2
Yl
+ 0.1(2xo + l)yo = 1 + 0.1(1)3 = 1.3
+ O.lUl =
3.1 + 0.1(1.3) = 3.23.
456

6.
x(n)
y(n)
0.00
0.10
0.20
0.30
0.40
7. Using Xo
2.0000
2.4734
3.1781
4.3925
6.7689
7.0783
initial condition
Runge-Kutta
Runge-Kutta
Runge-Kutta
predictor
corrector
= 1, YO = 2, and h = 0.1 we have

Xl
= Xo + h(xo +yo) = 1 + 0.1(1 + 2) = 1.3
YI = YO
+ h(xo -
X2 =
+ h(XI + YI) =
YO)
= 2 + 0.1(1 - 2) = 1.9
and
Xl
+ h(XI
Y2 = YI
1.3 + 0.1(1.3
+ 1.9) =
- YI) = 1.9 + 0.1(1.3 - 1.9)
1.62
= 1.84.
Thus, x(0.2) ~ 1.62 and y(0.2) ~ 1.84.

8. We identify P(x) = O, Q(x) = 6.55(1
+ x),
= 1, and h
f(x)
= (1 - 0)/10
= 0.1. Then the finite
Yi+l
+ [-2 + 0.0655(1 + Xi)]Yi + Yi-l = 0.001
or
Yi+l
+ (0.0655xi - 1.9345)Yi + Yi-l = 0.001.

0.0
0.0000
0.1
4.1987
0.2
8.1049
0.5
0.4
0.3
11.3840 13.7038 14.7770
0.7
12.5396
457
0.6
14.4083
0.8
9.2847
0.9
4.9450
1. O
0.0000
10
Plane Autonomous Systems

and Stability
Exercises 10.1
1. The corresponding plane autonomous system is

X'
If (x,y)
is a critical point, y =
= y,
y' = -9sinx.
and -9sinx
o.
Therefore x = mf and so the critical points
are (mf, O)for n =: 0, 1,2, ....

x' = y,
If (x,y)
is a critical point, then y
y' = -2x _ y2.
and so -2x
- y2
= -2x = O. Therefore (0,0) is the sole
critical point.
X' = y,
If (x, y) is a critical point, y
y' = x2 - y(l - x3).
= and so x2 - y(l - x3) = x2 = O. Therefore (O,O) is the sole critical
point.
I
If (x, y) is a critical point, y =

sole critical point.
X = y,
y = -4--
1 + x2
x
and so -4 --2
l+x
- 2y.
2(0) = O. Therefore x
and so x = 0,
and so (O,O) is the

I
X = y,
+ EX3 .
+ EX3 = O. Hence x(-l
If (x,y)
is a critical point, y =
-#.
The critical points are (0,0), (#,0)
and -x
y = -x
and (-#,0).

x' = y,
y' = -x
458
+ Exlxl.
+ Ex2) =
#,
Exercises
+ Exlxl = x( -1 + Elxl) = O.
If (x, y) is a critical point, y = O and -x

critical points are (O, O), (l/E, O) and (-l/E,
7. From x
+ xy =
into -y - xy
+ y) =
O we have x(l
= O, we obtain
Hence x
10. 1
= O, l/E, -l/E.
The
O).
= O or
O. Therefore x
y = -1. If x = O, then,
= O, Likewise, if y = -1, 1 + x = O or
substituting
= -1. We may conclude
that (O,O) and (-1, -1) are critical points of the system.
8. From y2 - x
= O we have x
= y2. Substituting into x2 -y
= O, we obtain y4 -y
= O or y(y3 -1) = O.
It fo11owsthat y = O, 1 and so (O, O) and (1,1) are the critical points of the system.
9. From x - y = O we have y
= x. Substituting into
3x2 - 4y = O we obtain
3x2 - 4x
= x(3x
- 4)
= O.
It fo11owsthat (O,O) and (4/3,4/3) are the critical points of the system.
10. From x3 - y = O we have y = x3. Substituting into x - y3 = O we obtain x - x9 = O or x(l - x8).
Therefore x
= O, 1, -1 and so the critical points of the system are (O, O), (1,1), and (-1, -1).
+ h = 10.
11. From x(lO - x - ~y) = O we obtain x = O or x

y
= O or
+y =
16. We therefore have four cases. If x
may conclude that y(
-h + 6)
Likewise y(16 - y - x) = O implies that
= O, y
= O or y
= 16. If x + h = lO, we
= O and so y = O, 12. Therefore the critical points of the system are
(O, O), (0,16), (10, O), and (4,12).

= O. It fo11owsthat y = 10, and from -2x + y
y+5
10. Therefore (10,10) is the sole critical point of the system.
12. Adding the two equations we obtain 10 - 15 _y_
10 = Owe may conclude that x

13. From x2eY = O we have x
= O. Since
eX - 1
eO -
= O, the second equation is satisfied for an
arbitrary value of y. Therefore any point of the form (O,y) is a critical point.
14. From sin y
= Owe have
y = mL From eX-Y
1, we may conclude that x - y
= O or x
= y. The
critical points of the system are therefore (mr, mr) for n = O, 1,2, ., ..
15. From x(l - x2 - 3y2) = O we have x = O or x2
+ 3y2 =
1. If x = O, then substituting into
y(3 - x2 - 3y2) gives y(3 - 3y2) = O. Therefore y = O, 1, -1. Likewise x2 = 1 - 3y2 yields 2y = O
so that y
O and x2
1 - 3(0)2
1. The critical points of the system are therefore (O, O), (0,1),
(O, -1), (1, O), and (-1, O).

16. From -x(4 - y2) = O we obtain x = O, y = 2, or y = -2. If x = O, then substituting into 4y(1- x2)
yields y
= O. Likewise y = 2 gives 8(1-x2)
= O or x
= 1, -1. Finally y
= -2 yields
or x = 1, -1. The critical points of the system are therefore (O,O), (1,2), (-1,2),
-8(1-x2)
=O
(1, -2), and
(-1, -2).
17. (a) From Exercises 8.2, Problem 1, x = clt
(b) From X(O)
= (2, -1) it fo11owsthat
Cl
c2e-t and y = 2clt
= O and
459
C2
+ c2e-t.
= 2. Therefore x = -2e-t
and y = 2e-t.
Exercises 10. 1
(e)
y
(-2,2)
2
~
-2
-2
18. (a) From Exercises 8.2, Problem 6, x = Cl + 2C2e-5t and y = 3Cl + C2e-5t.
= (3,4) it follows that Cl = C2=
(b ) From X(O)
(e)
= 1+
2e-5t and y = 3 + e-5t.
4
2
-4
l. Therefore x
-2
/(3,4)
4 x
/~
~
~,-4
19. (a) From Exercises 8.2, Problem 37, x
= cl(4cos3t - 3sin3t) + c2(4sin3t + 3cos3t) and y =
Cl(5 cos 3t) + C2(5 sin 3t). Al! solutions are one periodic with p = 27f/3.
(b) From X( O) = (4, 5) it follows that Cl
= 1 and C2 = O. Therefore x = 4 cos 3t - 3 sin 3t and
y = 5 cos 3t.
(e)
6
(4,5)
6 x
-6
-6
20. (a) From Exercises 8.2, Problem 34, x = Cl(sin t - cos t) + C2(- cos t - sin t) and y = 2Clcos t
2C2sin t. Al! solutions are periodic with p = 27f.
(b ) Frorn X(O)
= (-2,2) it follows that Cl = C2=
l. Therefore x
460
= -2cost and y = 2cost+2sint.
Exercises
(e)
10. 1
y
4
-4
4 x
-4
21. (a) Frorn Exercises 8.2, Problem 35, x

2Cl(cos t) e4t
+ 2c2(sin t) t.
(b) Frorn X(O) = (-1,2)

y = 2(cos t) t.
Cl(sin t - cos t)t
+ C2(-
sin t - cos t)t
and y
Because of the presence of e4t, there are no periodic solutions.
it follows that Cl = 1 and C2 = O. Therefore x = (sin t - cos t)t and
(e)
,-s
-5
-5
22. (a) Frorn Exercises 8.2, Problem 38, x = cle-t(2cos2t

y = c e:' cos 2t
+ C2e-t sin 2t.
C2 =
y
2
(2,1)
-2
+ c2e-t(2cos2t
+ 2sin2t)
and
Because of the presence of e-t, there are no periodic solutions.
(b) From X(O) = (2,1) it follows that Cl = 1 and

y = e-t cos 2t.
(e)
- 2sin2t)
-2
461
O. Therefore x = e-t(2 cos 2t - 2 sin 2t) and
Exercises
10. 1
23. Switching to polar coordinates,

-dr = -1 (dX
x-
dt
+ y -dY)
dt
dt
de
1 (
dx
dt = r2 -y dt
+x
1 -xy - x 2 T 4
= -(
r
dY)
1 2
dt = r2 (y
+ xy
- y 2 T 4 ) = ==T 5
+ xyr 4 + x 2 -
xyr ) = 1.
dr
5
If we use separation of variables on - = ==T we obtain
dt
T
Since X(O) = (4, O), r = 4 and
= ( 4t
+ CI
)1/4
and
e=t+C2.
e = O when t = O. It follows that
C2
= O and
C1
1
= 256' The final
solution may be written as

4
r=
e=t
. -Y1024t + 1 '
and so the solution spirals toward the origin as t increases.
-dr = -1 (dX
x-
dt
dt
+ y -dY)
dt
de
1 (
dx
dt = r2 -y dt
+x
= -1(xy - x 2 r 2 - xy
dY)
1
dt = r2 (-y
If we use separation of variables, it follows that

1
r =
and
y'-2t + CI
Since X(O) = (4, O), r = 4 and
- xyr
+ y 2r 2 ) = r 3
- x
+ xyr 2 ) = -1.
e = -t + C2.
e = O when t = O. It follows that
C2
= O and
C1
The final

4
r
Note that r
-+
00 as t -+ (3\) -.
e = -t.
= -vr.;1=-=3;;';2~t
'
Because O::; t ::;
l2' the
curve is not a spiral.
dr
1 (dX
dY)
1
2
2
dt=-;
Xdt+Ydt
=-;[-xy+x(l-r)+xy+y
de
1 (
dx
dt = r2 -y dt
+x
dY)
1 2
2
2
dt = r2 [y - xy(l - r ) + x
(l-r
+ xy(l
)]=r(l-r)
- r )] = 1.
Now dr = r - r3 or dr - r = _r3 is a Bernoulli differential equation. Following the procedure in

dt
dt
2
/
3 dr
/
Section 2.5 of the text, we let w = r : so that w = -2rdt ' Therefore w + 2w = 2, a linear
462
Exercises
first order differential equation. It follows that w
= 1 + cle-2t and so r2 =
1
1 + Cle-
10. 1
2t' The general

1
e = t + C2.
r = -v!:I:=+=c=l=e ==2~t
'
If X(O) = (1, O), r = 1 and
y
= r sin t
= O. Therefore
e = O when t = O. Therefore
= sin t. This solution generates the cirele r =

Cl
= -3/4,
C2
= O = C2 and so x = r cos t = cos t and

1. If X(O) = (2, O), r = 2 and e = O when
ci
= O and so
r=
e = t.
VI - ~ e-2t
This solution spirals toward the cirele r = 1 as t increases.

-dr = -1 (dX
x - + y -dY)
dt
de
dt
dt
1 (
dx
r2 -y dt
dt
+x
= -1 [xy - -x2
dY)
1 [
dt = r2 -y
(4 -
T ) -
xY
+ 7(4
-
xy - -y2 (4 r
2 ]
T )
2
2
xY
2
r ) - x - 7(4 - r)
=-1.
From Example 3, Section 2.2,

r
If X(O) = (1, O), r
cle4t
= 2 11 +
- ele4t
and
= 1 and e = O when t = O. It
r
=2
1- le4t
1+
3
~e4t
e = -t + C2.
follows that
and
e2
= O and
el
= -~.
Therefore
e = -t.
= Owhen t = 3 or t = ln3 and r -+ -2 as t -+ oo. The solution therefore approaches

4
the cirele r = 2. If X(O) = (2, O), it follows that Cl = C2 = O. Therefore r = 2 and e = -t so
Note that r
that the solution generates the cirele r = 2 traversed in the elockwise direction. Note also that the
original system is not defined at (O,O)but the corresponding polar system is defined for r = O. If the
Runge-Kutta method is applied to the original system, the solution corresponding to X(O)
will stall at the origino
27. The system has no critical points, so there are no periodic solutions.
28. Frorn x(6y-l)
= Oand y(2-8x)
= Owe see that (O,O) and (1/4,1/6)
are
critical points. From the graph we see that there are periodic solutions
around (1/4,1/6).
463
= (1, O)
Exercises 10. 1
29. The only critical point is (O,O). There appears to be a single periodic
solution around (O,O).
30. The systern has no critical points, so there are no periodic solutions.
31. If X(t)
= (x(t),y(t))
is a solution,
d
of dx
of dy
- f(x(t) y(t)) = - - + - - = QP - PQ = O
dt
'
ox dt
ay dt
'
using the chain rule. Therefore f(x(t),y(t))
curve of f.
= c for sorne constant c, and the solution lies on a level
Exercises 10.2
1. (a) If X(O)
= Xo lies on the line y = 2x, then X(t) approaches (O, O) along this lineo For all other
initial conditions, X(t) approaches (O,O) frorn the direction determined by the line y = -x/2.
(b)
'-
<;
<,
<,
'-1
,-
<,
'-
-1
2. (a) If X(O)
'Y = -x/2
= Xo lies on the line y = -x, then X(t) becornes unbounded along this lineo For all
other initial conditions, X(t) becornes unbounded and y = -3x/2 serves as an asyrnptote.
464
Exercses 10.2
(b)
,,
\
,,
,
-4
(1,1)
-2
-2
4 x
2
\
v
\'yy3XI21
\,
-4
3. (a) AH solutions are unstable spirals which become unbounded as t increases.

(b)
(1,1)
-5
-5
4. (a) AH solutions are spirals which approach the origino

(b)
-1
-1
5. (a) AH solutions approach (O,O) from the direction specified by the line y = X.
465
Exercises
10.2
(b)
-1
-1
6. (a) AH solutions become unbounded and y = x/2 serves as the asymptote.

(b)
4
3
-4 -3 ___
-2 -t l-"-1
4 x
/':')XI2' ~:
-4
7. (a) If X(O)
= Xo
hes on the line y
= 3x,
then X(t) approaches (O,O) along this lineo For all other
initial conditions, X(t) becomes unbounded and y = x serves as the asymptote.

(b)
3
2
-3 -2 -1
/
1
-1
-2
-3
8. (a) The solutions are ellipses which encircle the origino
466
Exercses 10.2
(b)
3
2
(1,1)
-3
3 x
-2
-3
< 0, we may conclude from Figure 10.18 that (O,O) is a saddle point.
9. Since b. = -41
10. Snce b. = 29 and 7

11. Since b. = -19
= -12, 72 - 4b. > and so from Figure 10.18, (O,O) is a stable node.
< 0, we may conclude from Figure 10.18 that (0,0) is a saddle point.
= -3 and so from Figure 10.18, (O,O) is a stable spral point.
12. Since b. = 1 and 7 = -1, 72 - 4b.
13. Since b. = 1 and 7 = -2, 72 - 4b. =

14. Since b.
= 1 and 7 = 2, 72 -
4b.
and so from Figure 10.18, (O,O) is a degenerate stable node.
= and so from Figure 10.18, (O,O) is a degenerate unstable node.
15. Since b. = 0.01 and 7 = -0.03,72 - 4b.
< and so from Figure 10.18, (O,O) is a stable spiral point.
16. Since b. = 0.0016 and 7 = 0.08, 72 - 4b. =

unstable node.
17. b.
= 1-
J.L2, 7 = 0, and so we need b. = 1 - J.L2
and so from Figure 10.18, (O,O) is a degenerate
> for (O,O) to be a center. Therefore
IJ.LI
< 1.
18. Note that b. = 1 and 7 = J.L. Therefore we need both 7 = J.L < and 72 - 4b. = J.L2 - 4 < for (O,O)
to be a stable spiral point. These two conditions may be written as -2 < J.L < O.
19. Note that b. = J.L + 1 and 7 = J.L + 1 and so 72 - 4b. = (J.L + 1) 2 - 4 (J.L + 1) = (J.L + 1) (J.L - 3). 1t follows
20. 7
2a, b.
<
if and only if -1
< 3. We may conclude that (O,O) will be a saddle point

when J.L < -1. Likewise (O,O) will be an unstable spiral point when 7 = J.L + 1 > and 72 - 4b. < O.
This condition reduces to -1 < J.L < 3.
that 72 - 4b.
<
J.L
a2 + (32 > 0, and 72 - 4b. = -4(3 < O. If a < 0, (O,O) is a stable spiral point. If
a > 0, (O,O) is an unstable spiral point. Therefore (O,O) cannot be a node or saddle point.
-F or X = -A-IF. Since Xp(t) = -A-IF is a particular
solution, it follows from Theorem 8.6 that X(t) = Xc(t)+XI is the general solution to X' = AX+F.
If 7 < and b. > then Xc(t) approaches (O,O)by Theorem 10.1(a). It follows that X(t) approaches
21. AXI + F = O implies that AXI
X as t _, oo.
22. If be < 1, b. = adxy(l
stable node.
- be)
>
and 72 - 4b.
= (ax - dy)2
467
+ 4abedxy > O.
Therefore (O,O) s a
Exercises
10.2
23. (a) The critical point is X
= (-3,4).
(b) From the graph, X appears to be an unstable node or a saddle

point.
(e) Since zx = -1, (O,O) is a saddle point.

24. (a) ThecriticalpointisXI=(-1,-2).
(b) From the graph, X appears to be a stable node or a degenerate

stable node.
(e) Since 7 = -16, ,0,.= 64, and 72
4,0,.= O, (O,O) is a degenerate stable node.
25. (a) The critical point is X = (0.5,2).

(b) Prom the graph, X appears to be an unstable spiral point.
(e) Since 7 = 0.2, ,0,.= 0.03, and 72 - 4,0,.= -0.08, (O,O) is an unstable spiral point.
26. (a) The critical point is X = (1,1).
(b) Frorn the graph, X appears to be a center.
(e) Since 7 = O and ,0,.= 1, (O,O) is a center.
468
Exercises
10.3
Exercises 10.3
ddr = ~ (x ddx + y ddY) = ~(ax2 - f]xy + xy2 + f]xy + ay2 - xy2) = ~ ar2
t
r
t
t
r
r
Therefore r
ce'" and so r
-+
Oif and only if a
= aro
< O.
dr
2. The differential equation dt = ar(5 - r) is a logistic differential equation. [See Section 3.2, (4) and
(5).] It follows that

r =
e = -t + e2.
and
1 + qe-50t
If a > O, r -+ 5 as t -+ +00 and so the critical point (O,O) is unstable. If a

and so (O,O) is asymptotically stable.
3. The critical points are x = Oand x = n+ 1. Since g'(x)
g'(n + 1) = -k(n
= k(n+
< O, r
1) - 2kx, g'(O)
-+
Oas t
= k(n+
1)
-+
+00
> Oand
+ 1) < O. Therefore x = O is unstable while x = n + 1 is asymptotically stable.
See Theorem 10.2.

4. Note that x = k is the only critical point since ln(x/k)
-k - kln(x/k),
g'(k) = -k
is not defined at x = O. Since g'(x)
< O. Therefore x = k is an asyrnptotically stable critical point by
Theorem 10.2.
5. The only critical point is T
Theorem 10.2.
= To.
= k, g'(To) = k > O. Therefore T
Since g'(T)
6. The only critical point is v = mg/k.
Now g(v)
=9-
(k/m)v
= To is unstable by
and so g'(v) = -k/m
< O. Therefore
v = mg / k is an asymptotically stable critical point by Theorem 10.2.

7. Critical points occur at x = a, f]. Since g'(x)
k(f]-a).
Since a> f], g'(a)
> Oand so x
k(-a
- f]
+ 2x),
g'(a)
k(a - f]) and g'(f])
= a is unstable. Likewise x = f] is asymptotically stable.
8. Critical points occur at x = a, f], ,. Since
g'(x) = k(a - x)(-f]

g'(a) = -k(f] - a)(I - a)
g'(f])
-,-
2x)
+ k(f]
- x)(I - x)(-l),
< Osince a > f] > . Therefore x = a is asymptotically stable. Similarly
> O and g'(I) < O. Therefore x
= f] is unstable while x =r is asymptotically stable.
9. Critical points occur at P = o.[b, e but not at P = O. Since g'(P)
g'(a/b) = (a/b - e)(-b)

Since a < be, -a + be
asymptotically stable.
= -a
+ be
and
= (a - bP)
+ (P
- e)( -b),
g'(e) = a - be.
> O and a - be < O. Therefore P = a/b is unstable while P
469
e is
Exercises 10.3
10. Since A
> O, the only critical point is A
= K2. Since g'(A) = ~kKA-l/2
- k, g'(K2)
-k/2
< O.
Therefore A = K2 is asymptotically stable.

11. The sole critical point is (1/2,1) and
g'(X) = (-2Y
-2X)
2x - 1 .
2y
= -2 and 6. = 2 so that 72
Computing g'((1/2, 1)) we find that 7
46. = -4
< O. Therefore
(1/2,1) is a stable spiral point.

12. Critical points are (1,0) and (-1,0),
and
, (X) = (2XO -2Y2).
At X = (1, O), 7
= 4, 6. = 4, andso
72
46. = O. We may conclude that (1, O) is unstable but we
are unable to classify this critical point any further. At X = (-1, O), 6. = -4
< Oand so (-1, O) is
a saddle point.
13. y' = 2xy - y
x
= y(2x - 1). Therefore if (x, y) is a critical point, either x
1/2 and y - x2
+ 2 = O implies
that (x,y)
(1/2, -7/4).
= 1/2 or y = O. The case
The case y = O leads to the critical
points (J2, O) and (-J2, O). We next use the Jacobian matrix
g'(X) = (-~~
1)
2x ~
to classify these three critical points. For X = (J2,0) or (-J2,0),

both critical points are saddle points. For X
-13
14. y'
< O. Therefore (1/2,-7/4)
= -y + xy
= y( -1
y = O and 2x - y2
+ x).
= (1/2, -7/4),7
= -1,6.
= -1 and 6.
< O. Therefore
= 7/2 and so 72
46. =
is a stable spiral point.
Therefore if (x, y) is a critical point, either y = O or x = 1. The case
O implies that (x, y)
(O,O). The case x
= 1 leads to the critical points
(1, J2) and (1, -J2). We next use the Jacobian matrix
g'(X) =
(2
-2Y)
x - 1
to classify these critical points. For X = (O,O), 6. = -2

either (1, J2) or (1, -J2), 7
2,6. = 4, and so 72
< O and so (O,O) is a saddle point. For
46. = -12. Therefore (1, J2) and (1, -J2)
are unstable spiral points.

15. Since x2 - y2 = O, y2 = x2 and so x2 - 3x
+ 2 = (x
- l)(x - 2)
= O. It follows that the critical
points are (1,1), (1, -1), (2,2), and (2, -2). We next use the Jacobian
g'(X) =
(-3
2x
470
2Y)
-2y
Exercises 10.3
= (1,1),
to classify these four critical points. For X
Therefore (1,1) is a stable node. For X
= -5, 6 = 2, and so
(1, -1),6
T2 -
46
= 17 > O.
-2 < O and so (1, -1) is a saddle point.
= (2,2), 6 = -4 < O and so we have another saddle point. Finally, if X = (2, -2),
For X
6 = 4, and so
T2 -
= 1,
46 = -15 < O. Therefore (2, - 2) is an unstable spiral point.
= O, y = x or y = -x. The case y = x leads to (4,4) and (-1,1) but the case y = -x
leads to x2 - 3x + 4 = O which has no real solutions. Therefore (4,4) and (-1,1) are the only
16. From y2 - x2
critical points. We next use the Jacobian matrix
x-
g'(X) = ( y
2y
-2x
= (4,4),
to classify these two critical points. For X

(4,4) is an unstable spiral point. For X
= (-1,1),
= 12, 6 = 40, and so T2 - 46 < O. Therefore
T
T
3)
= -3, 6 = 10, and so x2 - 46
< O. It follows
that (-1, -1) is a stable spiral point.

17. Since x'
= -2xy = O, either x = Oor y = O. If x = O, y(l- y2) = Oand so (O,O), (0,1), and (O,-1)
are critical points. The case y = Oleads to x = O. We next use the Jacobian matrix
g'(X)
-2y
( -1 + y
-2X)
1+ x - 3y2
to classify these three critical points. For X = (O,O), T

For X
= (0,1), T =
-4, 6
= 4 and so T2
46
= 1 and
= O. We can
6 = Oand so the test is inconclusive.
conclude that (0,1) is a stable critical
point but we are unable to classify this critical point further in ths borderline case. For X = (O,-1),
6 = -4
< O and so (O,-1) is a saddle point.
18. We found that (O,O), (0,1), (O,-1), (1, O) and (-1, O) were the critical points in Exercise 15,
Section 10.1. The Jacobian is
g'(X) = (
For X
= (O,O), T = 4, 6
(0,1) and (O,-1) give
1 - 3x2 - 3y2
-2xy
= 3 and so T2 - 46 = 4
= -8, 6 = 12, and
therefore stable nodes. For X
= (1, O) or
T2
-6XY)
3 _ x2 _ 9y2 .
> O. Therefore (O,O) is an unstable node. Both

- 46 = 16 > O. These two critical points are
(-1, O), 6 = -4 < O and so saddle points occur.
19. We found the critical points (0,0), (10,0), (0,16) and (4,12) in Exercise 11, Section 10.1. Since the
Jacobian is
g'(X) =
10 - 2x - Iy
2
-y
471
_IX)
2
16 - 2y - x
Exercises
10.3
we may classify the critical points as follows:
Conclusion
unstable node
(O, O)
(10,0)
26
160
-4
(0,16)
-14
-60
-32
(4, 12)
-16
24
36
saddle point
saddle point
stable node
160
20. We found the sole critical point (10,10) in Exercise 12, Section 10.1. The Jacobian is
gl(X) = (
g'((10, 10)) has trace
= -46/15,6.
-22
= 2/15, and
15
- (y + 5)2
72 -
46.
> O. Therefore (O,O) is a stable node.

el = y,
Since
lel < 11",
yl = (cos e - ~) sin e.
it follows that critical points are (0,0), (11"/3,0) and (-11"/3,0).
The Jacobian matrix
is
gl(X) =
( cos 2e - ~ cos e
~)
and so at (O,0),7 = and 6. = -1/2. Therefore (O,O) is a saddle point. Por X = (11"/3, 0),7 =
and 6. = 3/4. It is not possible to classify either critical point in this borderline case.
Xl = y,
If (x, y) is a critical point, y =
+ (~_ 3y2)y
yl = -x
and so -x
+ x)
- x2 = -x(l
_ x2.
= O. Therefore (O,O) and (-1, O) are
the only two critical points. We next use the Jacobian matrix
gI (X)
(0
-1 - 2x
1)
"2 - 9y
to classify these critical points. Por X = (O,O), 7 = 1/2, 6. = 1, and
72 -
46.
< O. Therefore (O,O)
is an unstable spiral point. Por X = (-1, 0),7 = 1/2, 6. = -1 and so (-1, O) is a saddle point.
and the only critical point is (O,O). Since the Jacobian matrix is
gl(X) = ex
+03x2y x3
472
J'
Exercises
10.3
= -1 and 6. = O, and we are unable to classify the critical point in this borderline case.
4x
- 2y
1 + x2
and the only critical point is (O, O). Since the Jacobian matrix is
I
x = y,
= -2, 6. = 4,72 - 46. =
(-4
10_x2
_:),
(1 + x2)2
-12, and so (O,O) is a stable spiral point.
g'(X)
y = ---
25. In Exercise 5, Section 10.1, we showed that (O, O),
(flTt, O) and (- flTt ' O) are the
critical points.
We will use the Jacobian matrix

I
g (X) =
to classify these three critical points. Por X
this critical point.
Por
(flTt, O),
1)
-1 + 3tx2
= (O, 0),7 = O and 6. = 1 and we are unable to classify
= O and 6. = -2
and so both of these critical points are
saddle points.
26. In Exercise 6, Section 10.1, we showed that (0,0), (l/t,O),
Since Dxxlxl
21xl, the
6.
= (O, O),
are the critical points.
Jacobian matrix is
g (X) =
Por X
and (-1/10,0)
(O2Elxl _ 1 O1) .
= O, 6. = 1 and we are unable to classify this critical point. Por (l/E, O),
= O,
-1, and so both of these critical points are saddle points.

((3 + 0:2y2)x
x'
y,
1 + 0:2x2
y =-
and the Jacobian matrix is

g'(X)
= (
((3 + o:y2)~0:2x2 - 1)
(1 + 0:2x2)2
Por X = (O, O),
28. From x'
y
= 1/(3.
-(Xx
= O and 6. = (3. Since (3 < O, we may conclude that (O, O) is a saddle point.
+ xy
The case y
= x(
-o: +
y) = O, either x = O or y = 0:. If x = O, then 1 - (3y = O and so
= o: implies that
1 - (30: - x2
= O or x2 = 1 -
0:(3. Since 0:(3
> 1, this equation
has no real solutions. It follows that (0,1/(3) is the unique critical point. Since the Jacobian matrix
is
g'(X)
(-o: +
-2x
473
X)
-(3'
Exercises 10.3
T
= -a - (3 +
1- a(3
73 = -(3 + -(3-
< O and 6.
= a(3 - 1
> O. Therefore (0,1/(3) is a stable critical
point.
29. (a) The graphs of -x+y-x3
= Oand _x_y+y2
in the figure. The Jacobian matrix is
g'(X) =
( -1-
3x2
-1
For X = (O,O), T = -2, 6. = 2,
1)
-1+ 2y
T2 -
= O are shown
46. = -4, and so (O,O) is
a stable spiral point.
(b) For Xj , 6.
= -6.07 < O and so a saddle point occurs at Xj .
30. (a) The corresponding plane autonomous system is

1 3) Y I = E(Y - -y
= y,
X'
3
and so the only critical point is (O,O). Since the Jacobian matrix is
I
g (X) =
T
= E, 6. = 1, and so
T2 -
46. = E2
(O-1
1 ),
E(l _ y2)
4 at the critical point (O,O).
> O, (O,O) is an unstable critical point.

(e) When E < O and T2 - 46. = E2 - 4 < O, (O,O) is a stable spiral point. These two requirements
can be written as -2 < E < O.
(d) When E = O, X" + X = Oand so X = CI COS t + C2 sin t. Therefore all solutions are periodic (with
(b) When
period 271')and so (O,O) is a center.

31.
-2x3
may be solved by separating variables. It follows that y2 + x4 = C. If X(O) =
dx
X
y
(xQ, O) where XQ > O, then C = x~ so that y2 = x~ - x4. Therefore if -XQ < X < XQ, y2 > O and
dy
y'
--
so there are two values of y corresponding to each value of x. Therefore the solution X(t) with
X(O) = (xQ, O) is periodic and so (O,O) is a center.
32.
~ i
dX
x'
~-~may
y
since X(O) = (x(O), x'(O))
be solved by separating variables. It follows that
= (1, O), e = ~. Therefore

y2
x3-3x2+2
233
(x-1)(x2-2x-2)
474
~ ~
- = - - x2
2
3
+ C and
Exercises 10.3
But (x - 1)(x2 - 2x - 2)
> Ofor 1- J3 < x < 1 and so each x in this interval has 2 corresponding
values of y. therefore X(t) is a periodic solution.

33. (a) x'
2xy
O implies that either x
O or y
= O.
and there are no real solutions. If y = 0,1-
If x
= O, then
from 1 - x2
+ y2 = O, y2 = -1
x2 = O and so (1, O) and (-1, O) are critical points.
The Jacobian matrix is
g'(X) = (
and so
T =
O and 6.
=4
at either X
2y
-2x
2X)
2y
(1, O) or (-1, O). We obtain no information about these
critical points in this borderline case.

(b)
1 - x2 + y2
d
or 2xy.JI.. = 1 - x2 + y2. Letting
2xy
dx
dJ.L 1
1
J.L= - , it follows that - = - -1 and so J.L= - - - x + 2c.
dx
x2
X
x
y2
1
Therefore - = -- -x+2c
which can be put in the form
x
x
dy
dx
y'
x'
y2
(x - c)2
+ y2
= c2
l.
The solution curves are shown and so both (1, O) and (-1, O)
-2
are centers.
34. (a)
dy
dx
y'
x'
(b) Let w
-x - y2
X
dy
= - - - y and so y
y
dx
---
+ y = -xy
-1
= yl-n = y2. It follows that dw + 2w = - 2x, a linear first order differential equation
dx
whose solution is
y 2 = w = ce -2x.
(1 x ) .
+ "2 -
Since x(O) = ~ and y(O) = X'(O) = O, 0= c and so

y
="2 1 -
x,
a parabola with vertex at (1/2, O). Therefore the solution X(t)

periodic.
35.
dy
dx
y'
x'
y2
y
and so
x4
x2
"2 = ""4 -"2 + c
y(O) = x' (O) = vo, it follows that
Cl
v5 and so
. 475
or y
x4
"2
-
with X(O) = (1/2, O) is not
+ Cl
Since x(O)
O and
Exercises 10.3
2 (x 2 - 1) 2 + 2v5 - 1
Y ="2x - x + Vo =
2
.
2
The x-intercepts on this graph satisfy

X2
JI - 2v5
= 1
and so we must require that 1 - 2v6 2: O (or Ivo I :S ~ v2

1-
JI - 2v5 and -Xo
) for real
solutions to existo If x5
< x < xo, then (x2 - 1)2 + 2V6 - 1 > Oand so there are two corresponding
values of y. Therefore X(t)
with X(O) = (O,vo) is periodic provided that Ivol :S
1v2.

x' = y,
y' = X2
- x
+1
and so the critical points must satisfy y = O and

x=
Therefore we must require that
:S ~ for real solutions to existo We will use the Jacobian matrix
g (X)
to attempt to classify ((1 ~)/2,O)
For X
(O
2X _
1)
1 O
when :S 1/4. Note that
= ((1 + vT"=4f)/2, O) and
((1 - vT=4E)/2,
1 vT=4E
2
= O and
6. = =f~.
< 1/4, 6. < O and so a saddle point occurs. For X =
O), 6. 2: O and we are not able to classify this critical point using linearization.

x'
,
a
f] 3 R
y = -- x - - x - - y
L
L
L
= y,
= q and y = q'. If X = (x, y) is a critica! point, y = Oand -ax - f]x3 = -x(a + {3x2) = O.
> O, a + f]x2 = O has no real solutions and so (O,O) is the only critica! point. Since
where x
If f]
= ( -a _O3f]x2
g' (X)
= -R/ L < O and 6. = a/ L > O. Therefore (O,O) is a stab!e critica! point. If f] < O, (O,O) and
(X,O), where x2 = -a/{3 are critica! points. At X(X,O),
T = -R/L
< O and 6. = -2a/L < O.
Therefore both critical points are sadd!es.

38. If we !et dx/dt
= y, then dy/dt
= -x3
dy
dx
X. From this we obtain the first-order differentia! equation

dy/dt
dx/dt
476
Exercises
10.3
Separating variables and integrating we obtain
=-
y dy
(x3
+ x)
dx
and
1 2
1 4
1 2
"2Y = -:x -"2x + ci.

=
Completing the square we can write the solution as y2

C2 = ~(x5
+ 1)2
+ 1)2 + C2. If X(O) = (xQ,O), then
and so
2
1 (2
y = -"2 x
Note that y
_~(x2
= O when x
)2
1 (2
)2
+ 1 +"2 xQ + 1
x6
+ 2x6 + 1 -
x4
2x2 - 1
= -XQ. In addition, the right-hand side is positive for -XQ
< x < XQ,and

= X(t)
so there are two corresponding values of y for each x between -XQ and XQ.The solution X
that satisfies X(O)
39. (a) Letting x
= (xQ,O) is therefore periodic, and so (O,O) is a center.

and
= Xl
we obtain the system
Xl
= Y and
yl =
1/2 - sinx. Since sinn/6
sin 5n /6 = 1/2 we see that (n /6, O) and (5n /6, O) are critical points of the system.
(b) The Jacobian matrix is

g/(X) =
O 1)
( - cosx
O
and so
Al = gI = ((n/6,0))
(O
= -V3/2
01) and
A2
= gI =
((5n/6,0))
(OV3/2 O1) .
> O and the trace of Al is O, no conclusion can be drawn regarding the critical
point (n/6,O). Since det Aj, < O, we see that (5n/6,0) is a saddle point.
Since det Al
(e) From the system in part (a) we obtain the first-order differential equation
dy
dx
1/2 - sinx
y
Separating variables and integrating we obtain
J
and
y dy =
J (~-
sin x ) dx
1
1
_y2 = -x
2
2
+ cos x + C
y2 = X
+ 2 cos x + C2.
or
477
Exercises
10.3
For Xo near
7r /6,
= (xo, O) then
if X(O)
C2
= -xo - 2 cos Xo and y2 = X + 2 cos x - Xo - 2 cos xo.
Thus, there are two values of y for each x in a sufficiently small interval around
(7r /6,
40.
7r /6.
Therefore
O) is a center.
(a) Writing the system as x'

point. Setting x3 - 2y3
x(x3 - 2y3) and y' = y(2x3 - y3) we see that (O, O) is a critical
= O we have x3 = 2y3 and 2x3 - y3 = 4y3 - y3 = 3y3. Thus, (O, O) is
the only critical point of the system.

(b) From the system we obtain the first-order differential equation
dy
2x3y _ y4
dx = x4 - 2xy3
or
which is homogeneous. If we let y = ux it follows that

(2x4u - x4u4) dx
+ (2x4u3
x4u(1
+ u3)
+ x5(2u3
dx
2u3
- dx
x
-1 dx
x
+ xdu) = O
- x4)(udx
- 1) du
-
=O
+ u (3u + 1) du
+ (1-- - -1 + 2u - 1 )
u +1 u
u2 - u + 1
=O
du = O.
Integrating gives
In Ixl + In lu + 11-ln
lul + In lu2 -
u + 11=
or
x (u: 1) (u + 1)
(y ; x) (~~ _ ~ + 1) =
= C2
2 _ U
(xy
+ x2)(y2
- xy
xy3
x3
478
c2
+ x2)
= c2x2y
+ x4
= c2x2y
+ y2 = 3c3xy.
Cl
Exercises
10.4
(c) We see from the graph that (O,O) is unstable. It is not possible
to classify the critical point as a node, saddle, center, or spiral
point.
Exercises 10.4
= e(o) = !!: and y(O) = e'(O)
1. We are given that x(O)
2g
7f
cos '3
+ e = T + e and
so e
= Wo - T'
y2
Smce
y2
wo.
2g
T cosx
+ e,
Wo
Therefore
2t (cos x - ~
+ 2lg w5)
.
1
1 2
1
1 2
and the x-mtercepts occur where cos x = - - - Wo and so - - - Wo must be greater than -1 for
2
2g
solutions to existo This condition is equivalent to

.
2. (a) Since y
2g
= T cosx + e, x(O)
~
-l (cose - coseo).
When e
= e(O)
Iwol <
= eo and y(O)
2g
1Ff.
2g
= e'(O) = O, e = -Tcoseo
and so y
~
= -eo, y2 = -(cos(
-ea) - cos(eo)) = O. Therefore y
1
= -
&
= O
when e = eo.
(b) Since y = de and e is decreasing between the time when e = eo, t = O, and e = -eo, that is,
t
=T,
dt
de = _
Therefore -dt = de
dt
{Jg
T = _ (T
V 2g
fi9 J cos e VT
cos eo .
1
and so
2g ...jcos e - cos eo
r:
Jo=oo
1
de =
...jcos e - cos eo
(T rOo
V 2g

x' = y,
f'(x)
y' = -g 1 + [f'(x)J2
479
J-oo ...jcos e - cos eo

(J
- my
de
.
Exercises
10.4
and .
If X
~ (_
f'(X)
_ P_ ) __
(1 + [f'(x)J2)f"(x)
- f'(x)2f'(X)f"(X)
DX
9 1 + [f'(X)J2
my 9
(1 + [f'(x)]2)2
= (X1,Y1) is a critical point, Y1 = O and j'(xl)
therefore
I
g (Xj ) =
4. When {3
(O
-g
f"()
= O. The Jacobian at this critical point is
1)
(3.
-m
Xl
= O the Jacobian matrix is

(-9f~(X1)
which has complex eigenvalues A
~)
= J9f"(X1)i.
The approximating linear system with x' (O)
=O
has solution
x(t)
and period 21T/ J gf"(xl).
5. (a) If f(x)
x2
=
""2'
j'(x)
= x(O)
cos J9f"(X1)
Therefore p ~ 21T/ J gf"(X1)
for the actual solution.
= x and so
dy
dx
y'
x'
X 1
1 + x2 y.
-=-=-g---
We may separate variables to show that y2 = -g lnl l
+ x2) + c.
But x(O) = Xo and y(O)
X'(O) = VD. Therefore c = V5 + 9 lnf l + x5) and so

y
Now
V5 - gln
(1 ++ x~)Xo
1
Therefore, if Ixl :::;[ev5/9(1
= vD2 -
2)
1+ x
1 + Xo
9 In
--2
2: O if and only if x2:::; eV5/9(1 + x5) - 1.
+ X5) - 1J1/2, there are two values of y for a given value of x and so
the solution is periodic.

(b) Since z =
Xmax
x2
""2'
the maximum height occurs at the largest value of x on the cycle. From (a),
= [ev5/9(1
+ x5) - 1]1/2 and so

Zmax
6. (a) If f(x)
= coshx,
X2
max
= _[e
2
(1
vO/9
+ x5) - 1J.
f'(X) = sinhx and [f'(X)J2 + 1 = sinh2 x + 1 = cosh2 x. Therefore

dy
y'
sinhx 1
- - - -g
dx - x' cosh 2 x y .
480
Exercises 10.4
.
We may separate vanables to show that y
= -- 2g + e.
But x(O)
cosh x
Xo and
y(O) = x'(O) = vo.
2
2g
Therefore e = Vo - --and so
cosh Xo
2
2g
29
2
Y - -+vo
cosh x
cosh Xo
Now
2g
cosh x
-_
2g
cosh Xo
- ---
2g cosh Xo
- 2g cosh Xo
2
+ Vo > O if and only if cos11X < --'-___,__---
V6
and the solution to this incquality is an interval [-a,a].
Therefore each x in (-a,a)
has two
corresponding values of y and so the solution is periodic.

(b) Since z
Xmax
= cosh x, the maximum height occurs at the largest value of x on the cycle. Frorn (a),
= a where cosh a =
2g cosh Xo
2
2g - Vo cosh Xo
Zmax
7. If Xm
< Xl < Xn, then F(Xl) > F(xm)
Therefore
2g cosh zr,
2
2g - Vo cosh xo
F(xn).
Letting X = Xl,
G(y) = ~
= F(xm)G(alb)
F(Xl)
F(Xl)
eo
Therefore frorn (2) on page 474, G(y) = F(XI)
< G(alb).
has two solutions YI and Y2 that satisfy
u: < alb < Y2

8. From (1), when y
= alb, Xn is taken on at sorne time t. From (3), if z >
Xn
there is no corresponding
value of y. Therefore the maximum number of predators is Xn and Xn occurs when y = alb.
9. (a) In the Lotka-Volterra Model the average number of predators is die and the average number
of prey is alb. But
Xl = -ax + bxy -
EIX =
-(a + EI)X + bxy
y' = -cxy + dy - E2Y = -exy + (d - (2)y

and so the new critical point in the first quadrant is (di e - E21e, alb
+ EI/b).
(b) The average number of predators di e - E21e has decreased while the average number of prey
alb + EI/b has increased. The fishery science model is consistent with Volterra's principie.
481
Exercises 10.4
10. (a) Solving
+ 0.02y) = O
x( -0.1
y(0.2 - 0.025x) = O
in the first quadrant we obtain the critical
20
40
60
100
80
point (8,5). The graphs are plotted using

x(O) = 7 and y(O)
= 4.
(b) The graph in part (a)

is around 40.
was
obtained using NDSolve
Since x(O)
in M athematica.
= 7, we use the FindRoot
We see that the period
equation solver in Mathematica
to
approximate the solution of x(t) = 7 for t near 40. From this we see that the period is more
closely approximated by t = 44.65.
11. Solving
x(20 - 0.4x - 0.3y)
=O
y(10 - O.ly - 0.3x)
=O
we see that critical points are (O, O), (0,100), (50, O), and (20,40). The Jacobian matrix is
'(
X)=
(0.08(20 - 0.8x - 0.3y)

-0.018y
-0.024x
)
0.06(10 - 0.2y - 0.3x)
and so
Al = g, ((O, O)) = (1.6
A3
= g, ((50, O)) =
(-1.6
O
O)
0.6
-1.2)
-0.3
, ((0,100)) = (-0.8
-1.8
A2
=g
A4
= g'((20, 40))
-0.64
= (
-0.72
O)
-0.6
-0.48)
-0.24
= 6.1 = 0.96 > O, T = 2.2 > O, and T[ -46.1 = 1 > O, we see that (O, O) is an unstable
node. Since det(A2) = 6.2 = 0.48 > O, T = -1.4 < O, and Ti -46.2 = 0.04 > O, we see that (0,100)
is a stable node. Since det(A3) = 6.3 = 0.48 > O, T = -1.9 < O, and Tl- 46.3 = 1.69> O, we see
that (50, O) is a stable node. Since det(A4) = -0.192 < Owe see that (20,40) is a saddle point.
Since det(Al)
12. 6.
= rlr2,
= rl +r2 and T2-46.
= (rl +r2)2 -4rlr2
= (rl-r2)2.
Therefore when rl!
r2, (0,0)
is an unstable node.
13. For X
T2
= (Kl, O), T = -rl +rz (1 - ~~ Q2l) and 6. = -rlr2 (1- ~~ Q2l). Ifwe let e = 1- ~~ Q21,
- 46. = (cr2
+ rl)2
stable node. If Kl
> O. Now if kl > K2 , e

Q2l
< O and so T < O, 6. > O. Therefore (Kl, O) is a
< K2 , e> O and so 6. < O. In this case (Kl, O) is a saddle point.

Q2l
482
Exercises
14.
10.4
(x, f)) is a stable no de if and only if K1 > K2 and K2 > K1. [See Figure 10.38(a) in the text.]
012
021
From Problem 12, (0.0) is an unstable node and from Problem 13, since K1
< K2 , (K1, O) is a
021
saddle point. Finally, when K2
< K1 , (O,K2) is a saddle point. This is Problem 12 with the roles

012
of 1 and 2 interchanged. Therefore (O, O), (K1, O), and (O, K2) are unstable.
K1
15 . 012
< K 2 < K 1021 an d so 012021 >
1. T here fore
/\ = (1 -
t...l
T1T2
012021 )"xy K1K2
< O an d so ('z, y') is
a saddle point.
X'
= y,
I
-g.
y = -smx--y
(3
ml
and so critical points must satisfy both y = O and sinx = O. Therefore (mf,O)
points.
are critical
(b) The Jacobian matrix
_1(3)
( _E.:osx
Z
ml
has trace
= -!}_z and determinant
.6. =
2
T
E. > O at
9
(32
(O, O). Therefore
4.6. = m2Z2 - 4 T
(32 - 4gZm2
m2Z2
We may conclude that (O, O) is a stable spiral point provided
(32 - 4gZm2
< O or
(3 < 2m.gl.

x
= y,
(3
y = --yly
m
I --xk
and so a critical point must satisfy both y = O and x = O. Therefore (O, O) is the unique critical
point.
(b) The Jacobian rnatrix is
( _ k _
m
and so
/2IYI)
= O and .6. = k > O. Therefore (O, O) is a center, stable spiral point, or an unstable
m
spiral point. Physical considerations suggest that (O,O) must be asymptotically stable and so
(O,O) must be a stable spiral point.
483
Exercses 10.4
18. (a) The magnitude of the frictional force between the bead and the wire is .t(mgcosB)
for sorne
.t > O. The component of this frictional force in the x-direction is

(.tmg cos B) cos B = .tmg cos2 e.
But
1
cos B = ~=====
+ [f'(x)J2
.tmg
and so .tmg cos B = 1 + [f'(x)J2
/1
It follows from Newton's Second Law that
"
mx
f'(x)
.t
x + mg 1 + [f'(x)]2
(3 ,
1 + [f'(x)J2 -
= -mg
and so
"
x
.t-J'(x)
= 9 1 + [f'(x)J2
(3,
- mx .
(b) A critical point (x, y) must satisfy y = O and f'(x)
(Xl, O) where f'(XI)
= u:
= u: Therefore critical points occur at
The Jacobian matrix of the plane autonomous system is
g'(X) = (g (1 + [f'(x)]2)( - f"(x)) ~ (.t - f'(x))2f'(x)f"(x)

(1 + [f'(x)J2)2
-m~)
and so at a critical point X ,
g'(X) = (-9f~(XI)
-m~)'
1 + .t2
(3
Therefore r = --
occurs. When
< O and
,0.. =
9 f" (x )
~ . When
1+.t
f"(XI)
< O, ,0.. < O and so a saddle point
1"(Xl) > Oand

2
r -4,0..=
(32
-,
m2
-4g--
f"(xI)
< O,
1 + .t2
(Xl, O) is a stable spiral point. This condition may also be written as

(32
19.
< 4gm2 f"(XI) .

1 + .t2
and so using separation of variables, y2 = - (X f(.t) d.t + C or y2 + 2F(x) = C.

2
Jo
We may conclude that for a given value of X there are at most two corresponding values of y. If
ddy
y:
f(x)
Y
(O,O) were a stable spiral point there would exist an x with more than two corresponding values of
f(O) = Ois required for (O,O) to be a critical point of the corresponding

plane autonomous system x' = y, y' = - f(x).
y. Note that the condition
484
Exercises
20. (a) x'
= x( -a + by) = Oimplies that x = Oor y = afb. If x
10.4
= O, then, from
-exy + K y(K - y) = O,
y
= O or K.
Therefore (O,O) and (O,K) are critical points. If fj = o.[b, then
fj [-ex + ~(K - fj)] = O.

The corresponding value of x, x = X, therefore satisfies the equation
ex = K
(K - fj).
(b) The Jacobian matrix is
g'(X) = (
and so at X
-a + by
-ex + K(K - 2y)
-ey
> b' 6. < O in this case. Therefore (O,O) and (O,K) are each saddle
y
=- K
Y < Oand
6. = bcifj
y), we may write the Jacobian matrix
and ci: = ~ (K -
g'((x, fj))
T
-1)' Since
points. Por X = (x, fj) where

as
= (O,0),6. = -aT < O. Por X = (O,K), 6. = n(Kb - a) = -Tb (K
we are given that K
and so
bx
T
>
= (
,
-ey --
bX)
T,
K
O. Therefore (x, y) is a stable critical point and so it is
either a stable node (perhaps degenerate) or a stable spiral point.

(e) Write
using
ci= ~(K-y)=
Therefore
72 -
46.
~Y[(~+4b)Y-4bK].
< O if and only if

4bK
fj <
4bK2
K + 4b = T + 4bK
Note that
4bK2
T+4bK
where K is large, and y =
1< K.
4bK .K::::K
T+4bK
Therefore
72 -
point will resulto
485
46.
< Owhen K is large and a stable spiral
Exercises
10.4
21. The equation
=a
x'
implies that x
o or y
= _1_.
1~
Y x - x
When a
a-1
=x
y -
1) = o
> O, Y = _1_
= (3. On the other hand, if y =
equation for yl, y
e7
> O. If x = O, then from the differential

a-1
1
'1
1
1
_-,
~
= - and so - x - _+ (3 = O.
a-1
l+y
a
a
a-1
It follows that
x=
and if (3(a - 1)
= _a_[(a
-1)(3 -1]
a-1
a-1
1, x > O. Therefore (x,?)) is the unique critical point in the first quadrant. The
>
a ((3 - _1_)
Jacobian matrix is
:x
a y ~ 1- 1
(1 y) 2 )
-x
(
y
- 1
-l+y
(1+y)2
and for X = (x, y), the Jacobian can be written in the form
g/(X) =
(a ~ 1)2 x
).
(a - 1)2 _ 1
a2
It follows that
2
7
and so
+ 1).
= -(6
= _ [( a :21)
Therefore 72 - 46
x + 1] < O,
= (6 + 1f - 46
(a :21) 2 X
= (6 - 1)2 > O. Therefore
(x, y) is a
stable node.
22. Letting y = Xl we obtain the plane autonomous system
Xl
=Y
yl = -8x
Solving x5
+ 8x =
6x3
+ 6x3
x5.
x(x2 - 4)(x2 - 2)
O we see that
critical points are (O, O), (O, - 2), (0,2), (O, -y2), and (O, y2).
The Jacobian matrix is
I
g (X)
-8 + 18x2 _ 5x4
1)
and we see that detg'(Xj) = 5x1_18x2 +8 and the trace of g/{X) is O. Since det(g'((y2,
-8
O))) =
< O, (y2, O) are saddle points. For the other critical points the determinant is positive and
linearization discloses no information. The graph of the phase plane suggests that (O, O) and (2, O)
are centers.
486
Chapter
10 Review Exercises
1. True
2. llue
3. a center or a saddle point
4. complex with negative real parts
5. False; there are initial conditions for which lim X(t)
t=ecc:
(O,O).
6. llue
7. False; this is a borderline case. See Figure 10.25 in the texto
8. False; see Figure 10.29 in the texto
9. The system is linear and we identify 6. = -o: and
when 6.
> O and
= o: + 1. Since a critical point will be a center
= Owe see that for o: = -1 critical points will be centers and solutions will be
periodic. Note also that when o: = -1 the system is

Xl
= -x - 2y
yl = X
+ y,
which does have an isolated critical point at (O,O).

10. We identify g(x) = sin x in Theorem 10.2. Then
gl(mr)
= cos nzr < O when
Xl
= mf is a critical point for n an integer and
n is an odd integer. Thus, mr is an asymptotically
stable critical point
when n is an odd integer.

1 -xy - x 2 T 3 + xy - y 2 T 3 ) = -r
-dr = -1 (dX
x - + y -dY) = -(
dt
dt
di
-de = -1 ( -y -dx + x -dY) = -1 (y 2 + xyr 3 + x 2 - xyr 3 ) = 1.

dt
r2
di
di
r2
1
Using separation of variables it follows that r = ij3TtCl and
r
= 1 and
e = O,. It
3t + el
follows that
el
--
00,
12. (a) If X(O)
-+
Since X(O)
= (1, O),
= 1, e2 = O, and so
r=
As t
e = t + e2.
ij3t + 1 '
e = t.
O and the solution spirals toward the origino
= Xa lies on the line y = -2x, then X(t) approaches (O,O) along this lineo For all other
initial conditions, X(t) approaches (O,O) from the direction determined by the line y = X.
487
Chapter
10 Review Exercises
(b) If X(O) = Xo les on the lne y = -x, then X(t) approaches (O,O) along this lineo.For all other
initial conditions, X(t) becomes unbounded and y
> Oand so (O,O) is a center.
13. (a)
= O, 6. = 11
(b)
= -2, 6. = 1, 72
14. From x'
= 2x serves as an asymptote.
46.
= Oand so (O,O) is a degenerate stable node.
= x(l + y - 3x) = O,either x = Oor 1 + y - 3x = O. If x
= O, then, from y( 4 - 2x - y)
we obtain y( 4 - y) = O. It follows that (O,O) and (0,4) are critical points.
=O
If 1 + y - 3x = O,
then y(5 - 5x) = O. Therefore (1/3, O) and (1,2) are the remaining critical points. We will use the
Jacobian matrix
g' (X) =
(1 +
y - 6x
-2y
x
)
4 - 2x - 2y
to classify these four critical points. The results are as follows:

X
(O, O)
15. From x
(0,4)
4
-20
(t, O)
la
-3
(1,2)
saddle point
saddle point
10
-15
dx
= -rsine
dt
de
dt
-5
stable spiral point
r coet), y = r sin ' we have

-
dy = reose de
dt
dt
Then r'
Conclusion
unstable node
dr
cose
dt
+-
+ dr
sine.
dt
= or , e' = 1 gives
dx
= -r sine
dt
+ ar cose
. e.
-dy = T cos e + Cl.r sm
dt
= O, which corresponds to X = (O,O), is a critical point. Solving r' = Cl.r we have

= cleat. Thus, when CI. < O, limt->oor(t) = O and (0,0) is a stable eritical point. When a = O,
We see that r
r
r' = Oand r = Cl. In this case (O,O) is a center, which is stable. Therefore, (O,O) is a stable critical
. point for the system when CI. ::; O.
16. The corresponding plane autonomous system is x' = y, y' = t(1 - x2) - x and so the Jaeobian at
the critical point (O,O) is
g/((O,O))
= (_~ ~).
488
10 Review Exercises
Chapter
< Oif and only if -2 < J.-L< 2. We may

therefore conclude that (O,O) is a stable node for J.-L< -2, a stable spiral point for -2 < J.-L< O, an
unstable spiral point for O < J.-L< 2, and an unstable node for J.-L> 2.
Therefore
= J.-L,!:::. = 1 and
72 -
41:::.= J.-L2- 4. Now J.-L2- 4
17. Critical points occur at x = l. Since

g'(X) = -~ e-x/2(x2
2
g' (1)
18.
dy
> O and g' ( -1) < O. Therefore x

y'
-dx = ,x =
x(O)
= Xo
+1
_2xVy2
y
and y(O) = X'(O)
- 4x - 1),
= 1 is unstable and x = -1 is asymptotically stable.
= O. It
_x2
+ c.
But
follows that c = 1 + x5 so that

y2
= (1 + x6 - x2)2 -
l.
> 1 for -Xo < x < Xo and y = O for x
Note that 1 + X5 - x2
[Y2+l =
We may separate variables to show that
= xo.
has two corresponding values of y and so the solution X(t) with X(O)
Each x with -Xo < x < Xo

= (xo, O) is periodic.
19. The corresponding plane autonomous system

Y I = -- (3 Y - -k (s
x' = y,
+ x )3 + 9
and so the Jacobian is

I
g(X)=
Por X = (O,O), 7 = _f}_
O
-~(s+x)2
< O, !:::. = 3k s2 > O. Therefore

m
7 -
(32
12k 2
1
- s = -((3
m2
m
m2
41:::. = -
Therefore (O,O) is a stable node if (32 > 12kms2

where ks3
- 12kms
).
and a stable spiral point provided (32 < 12kms2,
= mg.
20. (a) If (x,y)
is a critical point, y = O and so sinx(w2cosx
= g/w2l.
o.
Either sinx = O (in which
> 1 and so the latter equation has no

real solutions. Therefore (O,O) is the only critical point if w2 < g/l. The Jacobian matrix is
case x
= O) of
COSX
But if w2
g'(X) = (
and so
= -(3/ml
< gil,
- gil)
O
w2 cos 2x -
g/w2l
1:)
T cos x -:%J
< O and 1:::. = gil - w2 > O for X
= (O,O). It follows that
asymptotically stable and so after a small displacement, the pendulum will return to
e' = o.
489
(O,O) is
e = O,
Chapter
(b) If w2
(x,
10 Review Exercises
> gjl, cosx
= gjw2l will have two solutions x =
that satisfy
O) are two additional critical points. If X = (O,O), 6.
saddle point. If X =
6. =
cos x
(x,
-
O), T
-7r
< X < tt . Therefore
= gjl - w2 < O and so (O,O) is a
= -f3jml < O and
w2 cos 2x
3:2 - (2 :':2 - 1) =
w2
Therefore (x, O) and (-x, O) are each stable. When e(O)
w2
3:2 >
= ea, e'(O) =
o and
expect the pendulum to reach one of these two stable equilibrium positions.
490
O.
ea is small we
11
Orthogonal Functions and

Fourier Series
Exercises 11.1
1.
2.
r2 xx2dx
= ~x412
r1 x3(x2
+ l)dx =
1-2
J-1
4-2
= 102(x - l)dx = Gx2
102 eX(xe-X
4.
71"cos x sin 2 x dx = -1 sin 3 x 171"= O

10O
3
o
1-71"/2
6.
571"
~x611 +~x411 =
6
-1
4
-1
3.
- e-X)dx
5. rr/2 xcos2xdx
= ~ (-21 cos2x
+ xSin2x)
1:=0
- x)
171"/2
-71"/2
=O
/4 eX sin x dx = (-21eX sin x _ ~ex cos x) 1571"/4= O
h~
7I"~
7. For m =1- n
rr/2
Jo
sin(2n + l)x sin(2m + l)x dx
= -1 1071"/2 [cos2(n
2 o
1
4 n-m)
= (
- m)x - cos2(n + m + l)x] dx
171"/2
sin2(n-m)x
(.
ti
+ m + 1 )sin2(n+m+1)x
= O.
For m = n
Jorr/2 sin2(2n+1)x
dx =
r= (12 - 21
Jo
1
=-x
2
11[/2
o
1T
=-
491
cos 2(2n + l)x
) dx
1
) sin2(2n+
4 n+1
(2
l)x
11[/2
o
171"/2
o
Exercises 11. 1
so that
II sin(2n
8. For m =1-
+ l)xll =
2"V1f.
ti
rr/2
Jo
cos(2n+ l)x cos(2m+ l)xdx

1 fn7r /2
[cos2(n - m)x + cos 2(n + m + l)x] dx
2 o
=-
1
(
4n-m
17r/2
)sin2(n-m)x
1
(
4n+m+l
17r/2
)sin2(n+m+l)x.
= O.
For m = ti
fo7r/2
cos2(2n + l)x dx = fo7r/2

1
1~2
= -x
2 o
+ ~ cos 2(2n + l)X) dx
1~2
(2
) sin 2(2n + l)x
4 n+l
o
so that
II cos(2n +
l)xll =
2"V1f .
9. For m =1- n
7r
lao
sin nx sinmxdx
la
= -1 7r [cos(n - m)x - cos(n + m)x] dx

2 o
= (
2n-m
17r -
)sin(n-m)x
( 1
2n+m
17r
)sin2(n+m)x
= O.
For m
=n
7r
lco
sin2nxdx
1
] dx
= lo7r [1
- - -cos2nx
o 2 2
so that
II sin nxll
492
1
= -x
2
;_.
7r o
1
-sin2nx
4n
17r=-7r
o
Exercises
10. For m
11. 1
=1p
n?f
lcO sin -xp
mit
P
sin -x
= -1 JP ( cos (77 - m)7r x - cos (n + m)7r x ) dx
dx
20
p
2(n-m)Tt
. (n-m)7r
p
S111
IP -
p
. (n+m)7r
sm
x
2(n+m)7r
p
IP
O
= O.
For m =
ti
(P
Jo
[1'2 - 1
2 n7r
(P
sin -;xdx = Jo
2n7r]
'2cospx
dx
= '2x
I0P
p
2n7r
4nTtsinpx
IP0= p'2
so that
11. For m =1- n

p
loO
lloP ( cos (n -
n7r
mt:
cos - x cos - x dx = p
20
m)7r
x + cos
p
. (n-m)7r
IP
sin
x
2(n-m)7r
p
o
(n
+ m)7r x )
p
dx
p
. (n+m)7r
sin
x
2(n+m)7r
p
= O.
For m = n
p
loo
n7r
cos2 -xdx=
p
loP (1-+-cos-x
1 2n7r)
O
dx=-x
Also
p
loO
nst
1 . cos -x
p
4n7r
IP = O
p
n7r
dx = - sin -x
n7r
p
IP +--sin--x
p
2n7r
p
IP =-.p
O
and
so that
12. For m
=1-
n, we use Problems 11 and 10:
j
j
p
-P
p
-P
n7r
cos-x
m7r
cos-xdx
n7r
sin -x
sin -x
n7r
cos-x
mt:
cos-xdx
p
= 2 lcP
n7r
sin -x
m7r
sin -x
tnat
= 2 JP
dx
=O
dx = O.
Also
nm
j -Pp sin -x
P
tnn
cos -x
P
1
dx = 2
jP ( sin (n -P
493
m)7r
P
x + sin
(71+ m)7r x )
P
dx = O,
IP
O
Exercises 11. 1
-P
-P
mr
1 . cos -x
P
mr
-P
n7r
sin -x
P
dx =
l sm -x
P
and
n7r
cos -x dx
P
dx =
P
-P
dx =
n7r
-Pp sin 2 -x
P
dx = P
-P
-P
and
IP = O,
mr
sin -x
p
mr
-p
ru:
-- p cos -x
IP
mr
= O,
-P
2n7r
p
2n7r
sin -x
dx = --- cos --x
p
4n7r
p
Por m = n
-x
-Pp cos2 n7r
P
-p
(1- + -1
(1 1
cos 2n7r)
-x
p
dx = p,
2n7r)
- - - cos -x
dx = p,
IP
-P
= O.
12dx= 2p
-P
so that
13. Since
CXle-x.2
l. 2xdx = _e-x
-~
OC e-x 2 1
-CXl
210
-e -x21CXl=
O
-00
(4x2 - 2) dx = 2
=2
OC
-CXlx ( 2xe-x2)
dx - 2
O
'
CXl
2
-oc e-x dx
21C CXl 2)
( -xe-x -CXl + -CXle-x dx
- 2 CXl e-x 2 dx
-oc
and
OC e-x 2 . 2x (4x~ -oc

?
2) dx = 4
OC
(
2)
-oc x2 2xe-x
= 4 ( _x2e-x 21CXl
-oc +
the functions are orthogonal.
494
dx - 4
OC
-CXl xe"?2dx
CXl
-CXl xe-x 2)dx
- 4
CXl
-CXl xe?" 2dx
Exercises 11. 1
14. Since
rOO e-X 2 1(1 -
Jo
foOO e-x
1 Gx2
- 2x
x) dx
+ 1) dx
= (x
OO
roo
O - Jo
- l)e-X
= (2X -1-
= 1 + (2
= O,
I~+ fooo
~x2) e-x
- x)e-X
e-xdx
I~+ fooo
e-xdx
e-X(x
- 2)dx
= O,
and
= e=
(1"2X -"2x5
3
+ 3x - 1) 100
o + Joroo e -x
1 + e-x Gx2 - 5x + 3) [ +
=1- 3
+ e-X(3x
- 5) [
-3
fooo
fooo
-23x 2 + 5x
3) d x
e-X(5 - 3x) dx
e-xdx
= O,
the functions are orthogonal.
Ji <PO(X)<Pn(x)dx = O for n = 1, 2, 3, ...
15. By orthogonality
n = 1,2,3, ....
16. Using the facts that <Poand
(o:x
; that is,
Ji <Pn(x)dx = O for
<Pl are orthogonal to <Pn for n > 1, we have
+ (J)<Pn(X)
dx
= o:
x<Pn(x) dx
+ (J l; t
{b
<Pn(X) dx
(b
= o: Ja <Pl(X)<Pn(x) dx + (J Ja <PO(x)<Pn(x) dx
=0:0+{J0=0
for n = 2, 3, 4, ....
17. Using the fact that <Pnand <Pmare orthogsnal for n

lI<Pm(x) + <Pn(x)112 = \Pm(X)
m we have
+ <Pn(x)J2dx b [<p~(x) + 2<Pm(X)<Pn (x) + <p;(x)] dx
= b <p~(x)dx + 2l
= lI<Pm(x)1I2
=1=
<Pm(X)<Pn(x)dx
+ 11 <Pn(x)
112 .
495
+ b
<p;(x) dx
Exercises 11. 1
18. Setting
(2
1-2
0=
(2
+ CX3 + C2X )
13(x)f (X) dx = 1-2 (x2
dx =
16
64
:3 + SC2
and
0= [:
we obtain
= O and
CI
13(x)h(x)
C2
dx
(X3
[22
+ CX4 + C2X5) dx
= 654CI
= -5/12.
19. Since sin nx is an odd function on
[-7r,
7r],
(1, sin nx) = ["Ir"lr sin nx dx = O
= 1 is orthogonal to every member of {sin nx}. Thus {sin nx} is not complete.
and f(x)
20. (h
(b
+ h, 13) = Ja
+ h(x)l13(x)
[h(x)
21. (a) The fundamental period is

(b) The fundamental period is
(b
= Ja h(x)13(x)
dx
27r / 27r
27r / ( 4/ L)
dx
(b
+ Ja
h(x)13(x)
dx
= (h, 13) + (h 13)
l.
= 17r L.
+ sin 2x is 27r.
period of sin 2x + cos 4x is 27r / 2 = tt .
period of sin 3x + cos 4x is 27r since the
(e) The fundamental period of sin x

(d) The fundamental
(e) The fundamental
and
27r /
smallest integer multiples of
4 = 'n/2 that are equal are 3 and 4, respectively.
(f) The fundamental period of f(x)
= 2p/n.
is 27r/(n7r/p)
Exercises 11.2
1.
ao =
an
I.
i" f (x)
7r J-"Ir
1
bn = -
-"Ir
j"lr
-"Ir
7r
f(x)
100
7r n=l
7r
I.
7r
= -(1- cos n-r) =
1
-[1(-ltl
n7r
tit:
sinnx
sinnxdx
=O
(0 -1 dx
J -"Ir
+ I.
7r
r 2dx = 1
Jo
= -1 j"lr f (x) cos nx dx = -1 JO - cos nx dx + -1 J"Ir 2 cos nx dx = O

7r
bn
1 lo"lr
7r
l-(-l)n
-11"
=-
7r
= 1
7r
n7r
f(x)sin-xdx
= - +-
= -1 J"Ir cosnxdx
7r
2. ao = I. j11" f (x) dx
an
I.
i" 1dx
7r Jo
dx =
= -1 j"lr f(x)cos-xdxn7r
7r
-"Ir
= -1 j11" f(x)sinnxdx
7r
-11"
27r /3
7r
-11"
= -1 JO -sinnxdx+
7r
-rr
7r
-1 lo"lr 2sinnxdx
7r
496
3
= -[1(-ltl
n7r
Exercises
1 3 00 1 - (_1)n
f(x) = - + - L
2 7r n=l
n
3. ao = JI f(x) dx
-1
SlD nx
= JO 1 dx +
-1
Jo
x dx
=~
2
an
1
= J I f(x) cosn7rX dx = JO coenxx dx + 101 x cosrcs x dx = 22"[(-lt
-1
-1
O
n 7r
bn
I f(x) sin n7rXdx

J -1
3
f(x)=-4+L
00
[(-1)n-1
n=1
4. ao = I f (x) dx =
an
b-, =
-1
J
J
1x
f(x)cosn7rxdx
-Ion
I
-1
5. ao = ~j7r
[(
f(x)dx
tt
i7r
x2 sin nx dx
f(x) = -7r2
101 x cos nr z dx = ~[(-lt

1
Jor
(_l)n+l
x sin n7rXdx = -'---'-n7r
cos n7rX +
(_1)n+l
[2(-l)n
n2
n=1
00
]
sin n7rX
n7r
r x2dx
1
= ~7r2
i7r
7r
7r
non
tt
cosnx
(x
1 -sinnx
2
x2cosnxdx=-
= -1 ( - -x2 cos nx
+L
-lJ
tt
7r
-7r
7r Jo
1 j7r f(x)cosnxdx=7r -7r
bn = -1
dx = ~
2
=~
n7r
n7r
_l)n - 1
2
1
cosn7rx--sinn7rx
7r
an=-
-1
f (x) sin n7rXdx =
3
00
f(x) = -4 + L
n=1
17r+ -2lo7r
17r -- 2lo7r
7r
x cos nx dx )
+ (7r'_(_l)n+l + 2[(-l)n-1J)]
n
n37r
nOn
sinnx
j7r f(x) dx = ~ JO 7r2 dx + ~ r (7r2 - x2) dx = ~7r2

-7r
n -7r
7r Jo
3
1 j7r f(x)cosnxdx=1 JO 7r2cosnxdx+- 1 lo7r (7r2-x2)cosnxdx
7r =r:
7r -7r
7r O
2
-1 (7r2 - x2 sinnx 17r + -2i7r xsinnxdx ) = 2(-lt+
1
7r
non
7r
bn=- 1 j7r f(x)sinnxdx=7r
-7r
1 JO 7r2sinnxdx+-
1 lo7r (7r2-x2)sinnxdx
7r
Ti
-7r
497
xsinnxdx
2( _l)n
=--2-
2 [(-1t
= -tt (-1)n+ 1+ -3-
6. ao = ~
an=-
-lJ
= JO sin n7rXdx + 101 x sin n7rXdx = -- 1
22
7r
Jo
11.2
tt
- 1J
Exercises 11.2
-1] +.!. (X2 -11"2 cosnx
= ~[(-lt
11"
J; [2n2 (_I)n+l
511"2
f(x) = (3 +
11"
11"
-11"
I.
an = -1
11"
00
I,
j1l"
11"
= 211"
=O
= -1
11"
I,
'(
(3 - 2x) cos nx dx
=O
-11"
4
n
(3-2x)sinnxdx=-(-lt
-11"
= 3 +4 L
=6
-11"
f (x) cos nx dx
n=l
_1)n
--
sin nx
r sin x dx = ~11"
ao = .!.j1l" f (x) dx =.!.
11" Jo
-11"
= -1 j1l" f(x) cosnx dx = -1111" sinx cosnx dx = -1 la1l" [sin(n + l)x + sin(1 - n)x] dx
11"
11"
-11"
1+ (-lt
= 1I"1-n
(
2)
al
-11"
(3 - 2x)dx
= .:
-11"
11"
(-lt]
00
f (x)
sin nx
n 11"
00
11"
bn=11"
(-In])
+ 2[1- n311"
+ -i-[I-
2
L -(
_l)n+1 sin nx
= 11" +
= -1
= ~(-lt
8. ao = .!.j1l" f(x)dx
11" -11"
an
11"
-11"
n=l
9.
+ (11"~(_I)n
r iCOSnXdX)
Jo
= -1 j1l" (x + 11") cos nx dx
f (x) cos nx dx
-11"
11"
-11"
= -1 j1l" f (x) sin nx dx = -2 (-1 t+
f(x)
an
cos nx
= .!.j1l" f (x) dx = !_ j1l" (x + 11") dx
7. ao
111"-
nOn
= 211"
Jor
for n
sin2xdx
211"
= 2,3,4, , , ,
=O
bn = -1
f (x) sin nx dx = -1 la1l" sin x sin nx dx
11" -11"
11" o
1 {11"
= 211" Jo [cos(1 - n)x - cos(1 + n)x] dx = O for n = 2,3,4, , , ,
b: = -21
(1 - cos 2x) dx = ~
11" k
2
f (x)
1
=-
11"
ool+(_I)n
n=211"1 - n
+ - sin x + L
2) cos nx
498
Exercses
10.
11.2
2 j1r/2
2 Jc1r/2
?
f (x) dx = cos x dx = ::
7f -1r/2
7f o
tt
ao = an
= -2 j1r
tt
/2
f(x) cos 2nx dx
-1r/2
= -2J1r
/2
= -2Jc1r
/2
7f
1
cosx cos 2nx dx = 7f
!n
1r/2
[cos(2n - l)x
+ cos(2n + l)x] dx
2(_1)n+1
7f(4n2 - 1)
2 j1r /2
bn = f(x) sin 2nx dx
7f -1r/2
7f
1 Jc1r/2
cos x sin 2nx dx = [sin(2n - l)x
7f O
+ sin(2n + l)x] dx
4n
7f(4n2 - 1)
1 L [2( _1)n+1
4n
00
f (x) = - +
(2
) cos 2nx
7f n=l 7f 4n - 1
11.
0,0
= -21j2
-2
an=-21 j2
-2
f (x) dx = ~ (JO - 2 dx
2 -1
f(x)
12.
ao
= --
;:2
t.
= - t.
=~
an=-21
bn
1
2
+L
00
n=l
Jo
(10
dX) = - ~
2
1
2
(JO [ -2sin-x n7f

2
-1
-2
ten
f(x)cos-xdx=2
1
2
-2
n7f
f(x)sin-xdx=
2
-1
2
ten
la
) =--sin-1
nr:
cos-xdx
2
n7f
sin-xdx
2
nat
=- 3
ru:
n7f
2
( 1-cos- n7f)
2
n7f) . nr: ]
1 - cos sm-x
2
2
2 cos-xdx
n7f
2
n7f
xcos-xdx+
2
O
l
O
+ -3
] dx+
la
\ dX) = ~
(!nI
(la
] dx+
-1
X dx
- 1
(JO [ -2cos-xn7f
1 . n7f
nr:
--n7f
sm - cos -x
2
2
f (x) dx = ~
11
) sin 2nx
1
2
nr:
f(x)cos-xdx=2
nr:
bn=- 1 j2 f(x)sin-xdx=2 -2
2
+ 7f(24n
) =2"2
2 ( cos--1
n7f
n tt
2
n7f
xsin-xdx+
2
2
1
nr:
sin-xdx
2
2 ( sin -n7f + -(
n7f -lt+l )
= -n27f2
2
2
f(x) = -3
8
+ '"
--2 ( cos-nr:
s: n27f2
2
00
n=l
n7f
an=-51 j5 f(x)cos-xdx=-5
5
n7f
-1 ) cos-x+
2
1
5
(JO
-5
n7f
cos-xdx+
5
nr:
-- 2 (.sm-n7f + -(-1)
n27f2
2
2
la (l+x)cos-xdxn7f)
5
499
n+l).
sm-xn7f ]
2
5
=2"2[(-1)n-1]
n 7f
Exercises
11.2
mf
5
(JO
1
5
-5
9
Loo' [ --5 [( -1)n -lcos-x+]
n1f
=-+
4
n21f2
5
fx()
t.
2 -2
n1f
f(x)cos-xdx=-
n1f
(JO
-2
(JO
-2
la
n1f
(2+x)cos-xdx+
2
O
la
n1f
(2+x)sin-xdx+
2
O
n1f
2cos-xdx
n1f
2sin-xdx
2 [ 1 - ( -1 )n] cos -x
mf + -2 ( -1 )n+l.sm -x
n1f ]
= -32 + Loo [ -2
n1f
2
n=l n21f2
bn =
16. ao =
an
1 j1l'
1 j1l'
f (x) sin nx dx = -
1f
-11'
~ j1l' f (x) dx = ~
1f
1f
-11'
bn =
f(x)
r(eX -
lo
1 j1l'
f (x) cos nx dx
1f -11'
=-
1 j1l'
1f
f (x) sin nx dx
-11'
e1l'-1f-1
21f
+L
00
n=1
eXsin nx dx =
-11'
1) dx
= ~ (e 11'1f
1111'
(eX - 1) cosnx dx
1 lo1l'
(eX - 1) sin nx dx
[e1l'(_l)n_1
2) cosnx
1f 1 + n
')
2 2[1-(-lt]
n 1f
w
) =-(-lt+1
2
n1f
1f - 1)
[e1l'(_l)n - 1]
(
2)
1f 1 + n
1f O
1f O
(_l)nn(e-1I' - e1l')
(
'))
1f 1 + n-
==-
f ()x
1f
5
=-(-lt+1
n1f
5
O
5 ( -1)n+l.sm-xn1f ]
n1f
5
n=l
an=- 1
la (l+x)cos-xdxmf)
5
n1f
sin-xdx+
. 5
1 (ne1l'(_l)n+l
1
2
+n
=-
1f
n
+ ( --2
1+ n
[
e1l'(_l)n+
n
(_l)n
+ --2
+ -- -1)
1+ n
n
n
+ 11 +
(_l)n_1)
n
sin nx
17. The function in Problem 5 is discontinuous at x = 1f, so the corresponding Fourier series converges
500
Exercses 11.2
to 7r2/2 at x = Ti. That is,
7r2 7r2
=- +
2
6
[2( _l)n
L
2
n= 1
n
00
cos n7r+
(7r
?[( -l)n - 1])
]
- (_l)n+l + 3
sin n7r
n
n tt
'
and
~2 = ~ (7r22_ ~2) = 1 + 212+ ;2 + ....
At x
= O the series converges to O and
so
7r2
1
--1--+---+
12 22
1
32
1
42
18. From Problem 17

7r2= ~ (7r2 + 7r2) = ~ (2 + _3_ + _3_ + ... ) = 1 + 2_ + 2_ + ....
8
2 6
12
2
32 52
32 52
19. The function in Problem 7 is continuous at x = tt /2 so
-37r =
2
and
f (7r)
- = 7r+
2
(i
L.._., -2 ( -1) n+l
sin -n7r = 7r+ 2 1 - - + -1 - -1 + ... )

2
3 5 7
n=l n
7r
111
-=1--+---+.
4
357
20. The function in Problem 9 is continuous at x = tt /2 so

7r)
1= f ( 2
1
2
tt
1 1
2
=-+-+----+--_
tt
2 37r
and
or
= - + - + L.._.,
n=2
1 + (_l)n
nt:
cos 7r(1- n2)
2
2
3 . 57r
Ti
2
2
7r=1+-+---+----'"
2 3 35
2
57
Ti
1
1
1
4=2+~-'3.5+57-
501
2
5 . 77r
...
Exercises
11.2
21. (a) Letting ca
= ao/2,
f(x)
Cn = (an - ibn), and C-n

=
L (nT
anCOS -x
ao
2"
+
00
= Co
+L
einrrx/p
+ ibn)/2 we have
nT )
+ bnsin-x
00
n=l
= (an
+ e-inrrx/p
2
an
+ bn
einrrx/p _ e-inrrx/p)
2i
- bn
ieinrrx/p _ ie-inrrx/p)
2
n=l
00
= Co
+L
einrrx/p
+ e-inrrx/p
2
an
n=l
~
=co+L.._.
n=l
(an - ibn inrrx/p

e
2
+ an + ib e -inrrx/p)
2
00
= Co + L (Cneinrrx/p + c_nei(-n)rrx/p)
n=l
00
eneinrrx/p.
n=-oo
(b) Multiplying both sides of the expression in (a) by e-imrrx/p and integrating we obtain
f(x)e-imrrx/pdx
: ( f:
L
00
jP
Cn
n=-oo
ei(n-m)rrx/pdx
ei(n-m)rrx/pdx
L Cn jP
ei(n-m)rrx/pdx
-p
ni'm
=
dx
-p
L Cn jP
ni'm
cneinrrx/pe-imrrx/p)
n=-oo
+ Cm jP
ei(m-n)rrx/pdx
+ Cm jP
dx
-p
-P
L Cn jP
ni'm
-P
ei(n-m)rrx/pdx
+ 2pcm.
-P
Recalling that
eiY=cosy+isiny
we have for n-m
j p ei(n-m)rrx/pdx
an integer and n
=
-P
and
f. m
P
ei(n-m)rrx/p
i(n - m)T
e-iY=cosy-isiny
\P
-P
(ei(n-m)rr _ e-i(n-m)rr)
i(n - m)T
=.(
p ) [cos(n-m)T+isin(n-m)T-cos(n-m)T+isin(n-m)T]
zn-mT
= Q.
502
Exercses
11.3
Thus
j -P f(x)e-im7rx/Pdx = 2pcm
. / "dx.
Cm = -1 jP f(x)e-,m7rx
p
and
2p
22. Identifying f(x)
= e-x
-P
and p = 7r, we have
+ 1)7r
2(in
+ 1)7r
=_
7r
-7r
e-(in+l)x
1
2(in
[e-(in+l)7r
_ e(in+l)7r]
e(in+l)7r _ e-(in+l)7r
= ---;----,---
2(in
e7r(cos n7r
+ 1)7r
+ i sin n7r)
e-7r
(cos n7r - i sin n7r)
==-~------:---..,:-------2(in + 1)7r
cosn7r
2(in + 1)7r
(e7r - e-7r
= -'----;------
(e7r - e-7r)(_I)n
2(in
Thus
f(x) =
P'-7r-
'L
00
(-lt e 2
n=-oo
2(in
+ 1)7r
e
einx.
+ 1)7r
Exercises 11.3
==-sin3x = -f(x), f(x) is an odd function.

cos( -x) = -x cos x = - f(x), f(x) is an odd function.
1. Since f(-x)
= sin(-3x)
2. Since f( -x)
= -x
= (_x)2 - X = x2 - x, f(x) is neither even nor odd.

4. Since f( -x) = (_x)3 + 4x = _(x3 - 4x) = - f(x), f(x) is an odd function.
5. Since f( -x) = el-xl = elxl = f(x), f(x) is an even function.
6. Since f( -x) = e-x - e" = - f(x), f(x) is an odd function.
3. Since f( -x)
< x < 1, f( -x)

For ::;x < 2, f( -x)
(_x)2 = x2 = - f(x), f(x) is an odd function.
7. For
8.
= -x + 5 = f(x),
9. Since f(x)
f(x) is an even function.
is not defined for x < 0, it is neither even nor odd.
503
Exercises 11.3
10. Sincej'( -x)
= I(_x)51 = Ix51= f(:1;), f(x) is an even function.
11. Since f(x) is an odd function, we expand in a sine series:
bn
=~
(1r 1. sin nx dx
~Jo
Thus
f(x)
00
n=l
rtt;
L-
[1 -
= 2. [1n~
(-1 t] .
(-ltl sin nx.
12. Since f(x) is an even function, we expand in a cosine series:

ao =
an =
1dX
=1
n~
2
nt:
cos -x dx = - - sin -2 .
2
nt:
Thus
1
f(x) = - +
2
-2
L
-sinn=l ni:
nt:
00
nt:
cos -x.

ao=an
.:
xdx=~
2 -1 t = -2lo1r x cos nx dx = -2-[(

~
n ~
Thus
~
f (x)
=- +
2
00
1].
L -2-[(-1 t
n=l n ~
- 1]cos nx.

bn = -2lo1r x sin nxdx
~
Thus
00
n=l
2 -lt+1.
= -(
L -(-1t+
f (x) =
sin nx.

l
2
2
x dx = -
ao = 2
inO
an=2
x2
lo l x2cosn~xdx=2 ( -sinn~x
n~
Thus
f(x)
=- +
3
00
1 -- n~2 lol xsinn~xdx )
L 22(-ltcosn~x.
n=l n ~
504
4
= 22"(-lt
n ~
Exercises
16. Since f(x)
11.3
s an odd function, we expand in a sine series:

b., = 2
x2
= 2 ( --cosn7rX
nt:
f 1 x2snmfxdx
)0
11
+ -2101 x coe tin x dx )

tvn
Thus
f(x) =
L (2(-lt+1
17. Since f(x)
n7r
- 1J) sin tinx.
tt
is an even function, we expand in a cosine series:

2101r (7r 2 -
aO = -
tt
an
4
+ 33"[(
_l)n
00
n=l
= -2101r (7r2
7r
2 dx
X )
4 2
= -7r
3
= -2
x2)cosnxdx
(7r2
7r
x2 sinnx 11r+ -2101r xsinnxdx

non
O
4
= 2(-lt+1.
n
Thus
f(x)
_7r2 +
3
18. Since f(x)

bn
4
L 2(-lt+1
00
cosnxdx.
n=l n
is an odd function, we expand in a sine series:
= -2101r x3sinnxdx
7r
211"2
n
= _(_1)n+1
211"2
= -(-1)
ti
= -2
tt
( --cosnx
x3
non
11r
+ -3101r x2cosnxdx
12 107r
xsinnxdx
n 11" O
-2-
n+1 - - 12 ( -- x cosnx 11r

n27r
non
+ -1
io
1r cosnxdx )
= -211"2(-1
n'
)n+1 + -12 ( -1 )n .
n3
Thus
12 ( -1)n). sin nx.
f (x ) = ~L.. (211"2(
-1)n+1 + "3
n=l
n
n
19. Since
f (x) is an odd function, we expand in a sine series:

bn
= -tt2101r
(x+
o
l)sinnxdx
2(7r + 1)
(_l)n+1
n7r
Thus
f(x) =
L
00
n=l
(2(11" + 1) (_l)n+l
n7r
505
+ -2)
n11"
.
smnx.
2
+-.
nr:
Exercises 11.3
b., = 2la1 (x - 1) sin mfX dx
= 2 [la1
x sin mfX dx - la1 sin mfX dX]
]1 = - -2
Ix
1
= 2 [ 22
sin mfX - - cos mfX + - cos n7rX
n 7r
n7r
n7r
n7r
n7r
)
cos - - 1 .
2
Thus
f(x)
00
n=l
n7r
L -sinn7rx.
=-

ao
an =
l
lao
l
x dx
1dx
= -3
n7r
lao xcos -xdx

2
nt:
cos -xdx
2
4
= 22
n--t:
Thus
4 (
nt:
)
nt:
f (x) = -43 + '"L00 -cos - - 1 cos -x.
n
27r2
2
2
n=l

b.;
= -1
7r
i7r
O
127r 7rsin -x
n dx = -24 sin -n7r + -(
2 _l)n+l.
7r
2
n 7r
2
n
n
2
x sin -x dx + .
Thus
f(x)
L (4-sin-
n7r + _(_l)n+
2
2
n
00
n=l n27r
1) sin-x.n
2
io7r sin x dx = -7r4
ao
= -2
an
= -2 j.7r sinx
7r o
7r o
cosnxdx
la
= -1 7r [sin(n + l)x + sin(l - n)x] dx

7r O
7r(1:n2)[1+(-ltl
al
forn=2,3,4,
...
io
= -1 7r sin2xdx = O.
7r O
Thus
f(x)
=~+
7r
f 2[17r(1+-(_l)n]
n
n=2
2)
506
cosnx.
Exercises
24. Since (x) is an even function,
we expand
Exercise 11.2 for the computation
in a cosine series.
[See the solution
of Problem
of the integrals.]
ao = - 2 la1f /2 cos x dx = -4
7r/2 o
7r
2
= -/7r 2
an
Thus
ln1f/2 cosx cos -/-x

rit:
tt 2
2
(x)=-+L
7r
dx =
00
4(_l)n+1
n=l
7r 4n - 1
(2
4(_l)n+1
(2
7r 4n - 1
)cos2nx.
f1/2
25. ao = 2 Jo
1 dx = 1
1/2
an = 2
bn
=2
lnO
1 . sin n 7rx dx
00
n=l
L -2
00
J1/2
(x)
cos -n7r)
-nt:2 sin -nr:2 cosn7rX
+L
26. ao = 2 f1
1
1
1/2
bn = 2
= -2 ( 1 tin
n=l nr:
an=2
tin
1/2
1
(x) = 2
(x)
2
n7r
1 . cos n7rXdx = - sin -
lnO
1/2
( 1- cos- n7r) sin nzrz

2
1 dx = 1
lcosn7rxdx=--sin-
trn
tu:
. n7rXd x = -2 ( cos -n7r + ( -1 )n+1)

1 . sm
ru:
1 L (2-- sin -ntt cosn7rX )

2
nm
2
=- +
00
n=l
(x) = L00 - 2 (n7r

cos n=l ru:
27.
+ (_1)n+1 )
sin n7rX
ao = -4lc1f /2 cosxdx =-4

7r O
7r
4lc1f/2
2lc1f/2
an = -
cos x cos 2nx dx = 7r O
7r O
bn =
4101f
-7r
O
/2
2la1f
cos x sin 2nx dx = 7r O
[cos(2n
/2
[sin(2n
+ l)x + cos(2n
+ l)x + sin(2n
507
- l)x] dx =
- l)x] dx =
4( _l)n
(
2)
7r 1 - 4n
(2
8n
7r 4n - 1
11.3
10 in
Exercises 11.3
2
4( _l)n
00
f(x)
= :;:+ ]; 7r(1 _ 4n2) cos 2nx
f (x)
8n
00
L
n=l7r
(2
) sin 2nx
4n - 1
28. ao = ~ (" sin x dx = ~

tt
an=- 2
tt
Jo
tt
lo'" sinxcosnxdx=-
tt
lo'" [sin(n+1)x-sin(n-1)x]dx=
2[( _l)n
+2) 1] for ti = 2,3,4, ...
(
la
bn = -2 11" sinx sin nx dx = -1 Jc1l" [cos(n - l)x - cos(n

7r O
7r o
+ l)x]
7r1-n
dx = O
for n = 2,3,4, ...
al = ~ (" sin 2x dx = O
7r
Jo
bl = ~ (" sin 2 x
tt
Jo
f(x)
dx = 1
= sin z
2
L (-1)n+1
2
1- n
00
f(x) = - + 7r 7r n=2
29. ao =
~(
7r
an = -2
7r
bn =
f (x)
("/2 X dx
Jo
(lo"'/2
o
+ i"
J"'/2
x cos nx dx
cos na:
(7r - x) dX)
'"
",/2
=~
2
(7r - x) cos nx dx
2 (n7r
= -22 cos -
~ ( ("/2 x sin nx dx + t" (7r - x) sin nx dX) =

tt Jo
J"'/2
00
2 (n7r
= -7r + L
-2- 2cos -
n=l
n 7r
1) 1
+ (-1 t+ -
ti
--j:-tt sin n7r2

ti
cos nx
4
n7r
L -2sin - sin nx
00
f (x) =
n=l
ti
2'"
= - 2'"
30. ao=- 1
tt
tt
(x-7r)dx=-
tt '"
an
tt r
(x - 7r) cos -2n x dx
bn = -1 211"( x 7r 11"
f (x)
7r +
=4
7r
)' sm
4 [ (-1
-2n.-t:
t -cos -n7r]
2
-n x dx = -2 ( -1)n+l - - 4. sm -n7r
2
n
n27r
2
4 [
L
(-1)
n=l n27r
00
- cos -n7r] cos-xn

2
2
508
tt
+ (-1) n+l
1)
Exercises
()x n=l~Z:: (2n- (-1 )n+1 - n4.
n1f)
-2- Slll
tt
2
+ 1,\dX
31. ao = (1 xdx
Jo
an =
. -n X
sm
2
=~
2
101 X COS n:; X dx = n2~2 (COS

1
~1f
1)
1,2.
. n1f
xsm-xdx+
4.
n1f
1sm-xdx=-2
bn=
ino
f (x)
3
00
4 (
n1f
)
nTi
= - + '" -cos - - 1 cos - X
4
e:
n21f2
2
2
n=l
--Slll-
n21f2
32. ao = (11 dx
Jo
~ (4
()
f X = Z::
n=l
n tt
ino
. n1f
l'Slll-xdx+
1,2
3
~
=4+~n21f2
n=l
fx()
f(x)
2 (
n+1
-1)
n1f
1,\2 - x) dx = ~2
Jo
n1f
2sm-+2
2 n+1]' Slll-Xn1f
+ -(-1)
n1f
2
n1f
2
an = (1 1 . cos n1fX dx + 1,2(2 - x) cos n1fX dx
bn=
11.3
tt
n1f
2
(2-x)sin-xdx=-+-22sm2
n1f
(n1f
= 242 (COS ~
cos2+(-1)
n+1)
+ (-1 t+
4.
n1f
1f
1)
n1f
cos2x
00 ( 2
4
n1f)
n1f
= '"
sin sin-x
e: -n1f + -n21f2
2
2
n=l
33. ao
=2
an =.2
bn = 2
+ x)
(\x2
Jo
(x2+x)
dx
=~
3
cos n1fX dx =
2(x2 + x)
11
2 'lo1
sin n1fX - (2x+ 1)sin n1fX dx
n1f
O
n1f O
lo (x2 + x) sin n1fX dx = O
4
+ _[(_l)n
= _(_l)n+1
n1f
f(x)
= -6+
f(x)
= '"
~
n=l
n31f3
00
cosn1fX
-1]
L 22[3(
n=l n tt
00(4 _(_l)n-r) .
n1f
2(x2+x)
n1f
-lt
- 1]cosnsrz
4
+ _[(_l)n
n31f3
1] sin n-rz
509
= 222
2lo1
(2x + 1) costin dx
n1f o
+ -
1f
[3(-l)n-1]
Exercises 11.3
3"4
34. ao = Jo (2x - x ) dx =
2
an
bn
= Jor2 (2x
nt:
x2) cos -xdx
2
la (2x O
2.
tu:
- x )sIn2xdx
lao
1 la
n7l"
bn
27r 2 .
=x SIn nx
7l" O
+ 2:
4
-37l"2
00
n=l
r xdx
=~
4
x 2 cos nx dx = 2
n
7l"
1]cos-x
= -1
f(x)
~7l"2
an
27r
16
1]
n37l"3[1- (-1) ]
7l"
=.!.7l" Jor27rx2 dx =
7l"
n7l"
sin-x
00
n=l
7l"
16
2: 33[1(_l)n]
f(x)
-2 2[(-lt+l
00
n=l
36. ao
= -3 + 2: 2'2[( -lt+l
f(x)
35. ao
(4
dx = --47l"
n
47l")
- -sinnx
n
2cosnx
n
= 7l"
7l" Jo
.:
an = -
7l"
bn
x cos 2nx dx = O
= -217r x sin 2nx dx

7l" O
00
= -7l" + 2:
f(x)
37. ao
(1
--
= -- 1
sin 2nx
n=l
= 2la\X + 1) dx = 3
an=2
.
rl(x+
Jo
1)cos2n7l"xdx=
bn = 2
f(x)
38. ao
la (x + 1) sin 2n7l"xdx
O
3001
=-2
= --
n7l'
2: -
n=l
sin 2n7l"x
n7l"
= 2la2 (2 - x) dx = 2
510
Exercises
an
= 2 102 (2 -
bn
= 2 10
f(x)
x) cos mrx dx
(2 - x) sin mrx dx
= 1+
00
n=l
n7r
=O
=-
n7r
L -sinn7rx
39. We have
bn
= -2lo1r
10
5sin ntdt = -[1(-In
ntt
f(t) =
tt
so that
10[1 - (-l)n]
00
sin nt.
n7r
n=l
00
Substituting the assumption xp(t)
L En sin nt
into the differential equation then gives
n=l
and so
E - 10[1 - (_l)n
n - nst (O
1 - n 2)'
Th
us
10
xp(t)
=-
7r
1 - (_l)n
00
L (
n=l n 10 -
2) SlD nt.
40. We have
b
= -7r2 10O1(1 -
t) sin n nt dt
2
= -n7r
so that
f(t) =
00
n=l
n7r
L-
sin nnt:
00
Substituting the assumption xp(t) =
L En sin nnt
n=l
and so En
= n7r (O
1
2
2 2)'
- n t:
Thus
2001
xp(t) = -
L n (10 -
7r n=l
511
2 2)
7r
sin ntit.
11.3
Exercises
11.3
41. We have
so that
21f2
f(t) =
(-4)
00
3 + L --;z
cosnt.
n=l
Substituting the assumption
-:fA + f1
00
xp(t) =
Ancosnt
4' -;
f1 (1
+ 12xp = 6Ao .+
) cos nt = 3
21f2
(-4)
+ J; --;z cos nt
00
An
00
+ 12
-4'n2
1f2
and Ao
9'
42. We have
r1/2
ao = (1/2) Jo
r1/2
f(t)
=- +
an = (1/2) Jo
= 2'
tdt
tcos2n1ftdt
= n21f2 [(_l)n
-1]
so that
1
L
n=l
00
(-1)n-1
n
2 2
cos 2n1ft.
1f
Substituting the assumption

1
- x~
1
and Ao = 24'
00
+ 12xp = 6Ao + L An(12
- n21f2) cos 2n1ft
n=l
00
=- +L
4
n=l
(-1)n-1
An = n21f2(12 _ n21f2) . Thus
1
xp(t) = 48
+ 1f2
f1
00
(-l)n-1
n2(12 _ n21f2) cos 2n1ft.
512
- 1
(_l)n
2 2
1f
cos 2n1ft
Exercises
11.3
43. (a) The general solution is x(t) = el cos vT6t + e2sin vT6t + xp(t), where
1 - (_l)n .
(
2) sin ni,
7["n=l n 10 - n
10
00
xp(t) = - 2::
The initial condition x(O) = O implies el + xp(O) = O. Since xp(O) = O, we have el = O and
x(t) = e2 sin vT6t + xp(t). Then x'(t) = e2vT6 cos vT6t + x~(t) and x'(O) = Oimplies
10 00 1- (_l)n
e2\l()"+ - 2::
2 cos 0= O.
7["n=l 10 - n
Thus
vT6 00 1- (_l)n
e2 = ---;- 2:: 10 _ n2
n=l
and
10 00 l-(-l)n[l
2::
2
7["n=l 10 - ti
1
]
- sin nt - 117\ sin vT6t .
n
y 10
x(t) = -
(b) The graph is plotted using eight nonzero terms in the series expansion of x(t).
x
4
2
-2
-4
44. (a) The general solution is x(t) = el cos 4V3t + e2sin 4V3t + xp(t), where
7["2
xp(t)=-+162::
18
00
2( 2
n=l n n -
48)cosnt.
The initial condition x(O) = O implies el + xp(O) = 1 or

7["2
el = 1 - xp(O) = 1 - 18 - 16
Now x'(t) = -4V3el sin4V3t+4e2 cos4V3t+x~(t),
E n2(n2 00
48) .
so x'(O) = Oimplies 4e2+X~(0) = o. Since
x~(O) = O, we have e2 = Oand

7["2
00
1)
x(t) = ( 1 - - - 162:: 2( 2
)
18
n=l n n - 48
7["2
00
1
+ 16 2:: 2( 2
8) cosnt
18
n=l n ti - 4
cos4..J3t + -
7["2 (
7["2)
00
1
=-+
1-cos4V3t+162::
2( 2 48)[cosnt-cos4..J3t1.
18
18
n=l n n -
513
Exercises
11.3
(b) The graph is plotted using five nonzero terms in the series expansion of x(t) ..
x
1.5
1
0.5
t
1
-0.5
-1
45. (a) We have
bn = ~ (L woX sin tut x dx = 2wo (-lt+1

L Jo
L
L
mr
so that
w(x)
mr
L -2wo (-lt+ 1 sin -L
x.
00
n=l
rttt
00
(b) If we assume y(x) =
n7r
L En sin L x
then
n=l
and so the differential equation EIy(4)

En
= w(x)
gives
2wo( _l)n+l L 4
EIn57r5
Thus
46. We have
[n7r
2n7r]
2L/3
tvn
2wo
bn = wo sin - x dx = cos - - cos -L L/3
L
tit:
3
3
so that
w(x)
If we assume y(x)
00
2wo [n7r
L
cos n=l nt:
3
n7r
00
L En sin L x
then
n=l
514
2mr]
- cos -
tit:
sin -
x.
Exercses
and so the differential equation Ely(4)
Thus
11.3
= w(x) gives
4 00
n7r
2n7r
_ 2 WoL '" cos 3 - COS-3- . ti n
y (x ) - --5 L
5
sin -L x.
El t: n=l
n
47. We note that w(x)is 2n-periodic and even. With p = tt we find the cosine expansion of
={
f(x)
\Ve have
ao = ~
r f(x)dx
n Jo
an
= -2
=~
r/
n Jo
wodx = Wo
= -2
7r
non
< x < n /2
n/2<x<n.
Wo
0,
f(x) cosnxdx
J7r /2
nn
= -2wo sin -2
.
Wocosnxdx
nn
Thus,
Wo
w(x) = 2
2wo
+-
1 . nn
L
-sm-cosnx.
n=l n
2
00
= Ao/2 + L~=l Ancosnx.
Now we assume a particular solution of the form yp(x)
L~=l Ann4 cos nx and substituting into the differential equation, we obtain
Ely~4)(x)
+ kyp(x) =
kAo
-2-
+ L An(Eln4 + k) cosnx
00
n=l
Wo
2wo
= - +-
1 . nn
L
sm n=l
2
00
cos nx.
ti
Thus
Ao
= Wo
and
_ 2wo sin(nn/2)
n-n
n(Eln4 + k) ,
and
Wo 2wo 00 sin(nn /2)
yp(x) = - + (El 4 k) cosnx.
2k
n n=l n
n + .
48. (a) If f and 9 are even and h(x) = f(x)g(x)

h(-x)
then
= f(-x)g(-x)
= f(x)g(x)
= h(x)
and h is even.
(e) If f is even and 9 is odd and h(x)
h(-x)
= f(x)g(x)
= f(-x)g(-x)
then
= f(x)[-g(x)]
515
-h(x)
Then y~4)(x) =
Exercises 11.3
and h is odd.
(d) Let h(x)
= j(x) g(x) where j and g are even. Then

h( -x) = j( -x) g( -x) = j(x) g(x),
and h is an even function.
(f ) If j is even then
fa
La
j(x) dx =
-10a j( -u) du + Jor j(x) dx = Jol" j(u) du + Jofa j(x) dx
=2
fa j(x)
Jo
dx.
(g) If j is odd then
_aaj(x)
dx = =-
:a
O j(u)
j( -x) dx + loa j(x) dx =
loa j(u)
du + loa j(x) dx
du + loa j(x) dx = O.
jex) is even then j( -x) = j(x). If j(x) is odd then j( -x) = - j(x). Thus, if j(x) is both even
and odd, j(x) = j( -x) = - j(x), and j(x) = O.
49. If
50. For Ely(4l+ky = Othe roots ofthe auxiliary equation are

and m4 = -a - ai, where a = (kj EI)I/4jV2. Thus
Yc
= eCY.x (CI cos a + C2 sin oz) + a -CY.X
We expect y(x) to be bounded as x

bounded as x
-+ -00,
-+ 00,
so we must have
C3
= a+ai,
mI
(C3
so we must have
cos ax
CI
= C4 = O. Thus, Yc
m2
= a-ai,
= C2 = O. We also expect
< 27f.
f (x)
2
1
2
-1
-2
-3
516
y(x) to be
= Oand the solution of the differential
51. The graph is obtained by summing the series from n = 1 to 20. It appears that
X,o<x<7f
-7r,
7r <
-a+ai,
+ C4 sin oz ).
equation in Problem 47 is yp(x).
j(x) = {
m3 =
14
Exercises
= 1 to
52. The graph is obtained by summing the series from n
11.4
10. It appears that
O< x < 1
1 < x < 2.
1- x,
f(x) = { O,
f (x)
1.5
1.25
1
0.75
0.5
0.25
10
-0.25
-0.5
53. (a) The function in Problem 51 is not unique; it could also be

O<x<7r
X,
f(x) =
x=7r
.,
7r
<
<
27r.
(b) The function in Problem 52 is not unique; it could also be

-2 < x < -1
O,
x+l,
-l<x<O
= { -x + 1,
f(x)
O< x
<1
1 < x < 2.
O,
Exercises 11.4
1. For A
:S O the only solution of the boundary-value problem is y = O. For A > O we have

y
CI
cos ~
+ C2 sin ~
x.
Now
y'(x) = -CI ~sin

and y'(O) = Oimplies
y(l)
C2 =
O, so
+ y'(l)
CI(COS~
The eigenvalues are An = x;;' where
v'>..
+ C2~COS~X
~x
Xl,
= O or
~sin~)
X2, X3, ...
cot ~
= V)...
are the consecutive positive solutions of cot V).. =
The corresponding eigenfunctions are cos ~
517
= COS xnx
for n
1, 2, 3, ....
Using a CAS
Exercises
11.4
we find that the first four eigenvalues are 0.7402, )1.7349,41.4388,
and 90.8082 with corresponding
eigenfunctions cos 0.8603x, cos 3.4256x, cos 6.4373x, and cos 9.5293x.
2. For A
< O the only solution of the boundary-value
problem is y = O. For A = O we have y = elX +cz
Now y' = el and the boundary conditions both imply el + e2 = O. Thus, A = O is an eigenvalue
with corresponding eigenfunction YO = x - 1.
For A
> O we have
y = el cos ~ x + e2sin ~
and
y' = -el~sin
~x
+ e2~cos~X.

el
+ e2~ = O
el cos ~
+ e2sin ~
= O
which gives
-e2~cos~+e2sin~=0
or
tan~=~.
The eigenvalues are An = x;; where Xl, X2, x3, ... are the consecutive positive solutions of
tan /).. = /)...
taking e2 = -1 in the first equation of the system.)
.;>:.;;, cos .;>:.;;, x
- sin';>:';;' x (obtained by
Using a CAS we find that the first four
positive eigenvalues are 20.1907, 59.6795, 118.9000, and 197.858 with corresponding eigenfunctions
4.4934 cos 4.4934x-sin 4.4934x, 7.7253 cos 7.7253x-sin 7.7253x, 10.9041 cos 10.9041x-sin 10.9041x,
and 14.0662 cos 14.0662x - sin 14.0662x.
3. For A = O the solution of y" = O is y = elX + e2. The condition y'(O) = O implies el = O, so A = O
is an eigenvalue with corresponding eigenfunction 1.
For A
< Owe have y = el cosh Nx+e2
sin Nx
and y' = el N sinh Nx+e2N
The condi tion y' (O) = O implies e2 = O and so y = el cosh
N x.
cosh Nx.
Now the condition y' (L) = O
implies el = O. Thus y = Oand there are no negative eigenvalues.
> O we have y = el cos ~ x + e2sin /).. x and y' = -el/).. sin ~ x + e2/).. cos ~ x. The
condition y' (O) = O implies e2 = O and so y = el cos /).. x. Now the condition y' (L) = O implies
-el/).. sin /).. L = O. For el -=1= O this condition will hold when ~ L = mr or A = n27f2/ L2, where
For A
n = 1, 2, 3,
n
1, 2, 3,
4. For A
These are the positive eigenvalues with corresponding eigenfunctions cos(n7f/L)x,

.
< Owe have

y =
ci
cosh N x + e2sinh
y' = el N sinh
Nx+
518
Nx
e2N
cosh
N x.
Exercises
11.4
Using the fact that cosh x is an even function and sinh x is odd we have
y(-L)
= clcosh(-~L)
= Clcosh ~
+c2sinh(-~L)
L - C2sinh ~
and
y'(-L)
= cl~sinh(-~L)
= -Cl ~
+c2~cosh(-~L)
L+
sinh ~
cosh ~
C2~
L.

Clcosh ~
L - C2sinh ~
L = Clcosh
v=>: L + C2sinh v=>: L
or
v=>: L = O
2C2 sinh
and
or
= o.
2Cl~sinh~L
Since
y
R L =J o, Cl = C2= O and
the only solution of the boundary-value problem in this case is
= O.
For A = O we have
y
= ClX
y'
From y(-L)
= Cl
= y(L) we obtain
=ci l.
Then Cl = O and y
For A
+ C2
+ C2= clL + C2
= 1 is an eigenfunction corresponding to the eigenvalue A
> O we have
y = ci
y' = -
COS
ci
,j)._ X + C2sin ,j)._ x
,j)._ sin ,j)._ x +
C2,j)._
cos ,j)._ x.
The first boundary condition implies

ci
cos ,j)._ L - C2sin ,j)._ L = Cl COS ,j)._ L + C2sin ,j)._ L
or
2C2 sin
,j)._ L = O.
519
= O.
Exercises 11.4
Thus, if
el
= O and
e2
i= O,
V).,L=mr
The corresponding eigenfunctions are sin
n; x, for n
= 1, 2,3, ., .. Similarly, the second boundary
condition implies
2q V)., sin V)., L
If
e2 = O and el
= O.
i= O,
n21f2
A = V' n
= 1,2,3, ... ,
n1f
and the corresponding eigenfunctions are cos LX, for n = 1, 2, 3, ....

5. The eigenfunctions are cos A
11cos
where cot
A = Fn.
Thus
>:;. xl1 = la cos >:;. x dx = ~la (1 + cos 2>:;' x)

2
=~
(x + 2y~An sin 2>:;' x)
=~
[1 + 2~
=~
[1 + ~
sin
=~
[1 + ~
(sinFn)
(2 sin
r:
11 = ~
cos
(1 +
~ sin 2>:;')
2y An
>:;.) ]
Fn cot Fn sin >:;.]
6. The eigenfunctions are sin Fn x where tan

11sin
dx
Fn (sin Fn)]
=~
= ~2
(x-
2y
=~
[1-
2~
=~
[1- ~
=~
[1- ~ (- Fn cos
o= ~
2
~sin2Fnx)
x,
11
(2 sin .:
tan
= -An. Thus
>:;. xl1 = la sin Fn x dx = ~ la (1 - cos 2>:;' x)

2
(1 + sin Fn) .
cos
(1-
dx
~sin2Fn)
2y An
>:;.) ]
Fn cos >:;. cos >:;.]

2 -:
520
= ~(
1 + cos Fn) .
2
Exercises
7. (a)
11.4
If,\::; O the initial conditions imply y = O. Por .\ > O the general solution of the Cauchy-Euler
differential equation is y = el cos(.J:\ lnx) + e2sin(.J:\ lnx). The condition y(l)
= O implies
el = O, so that y = e2sin(.J:\ lnx).
n1T,
2, 3, ....
The condition y(5) = O implies .J:\ ln5 =
n = 1,
Thus, the eigenvalues are n21T2/(ln5)2 for n = 1, 2, 3, ... , with corresponding
eigenfunctions sin[(n1T/ In 5) In xJ.

(b) The self-adjoint form is
-d [xy '] + -y
dx
x
= O.
(c) An orthogonali ty relation is

(5 sin (m1T lnx) sin (n1T lnx) dx = O.
In5
ln5
JI
8. (a) The roots of the auxiliary equation m2 + m + ,\ = O are

general solution of the differential equation is el +e2e-x.
i( -1
VI - 4,\). When ,\
= O the
The initial conditions imply el +e2 = O
and el + e2e-2 = O. Since the determinant of the coefficients is not O, the only solution of this
homogeneous system is el = e2 = O, in which case y = O. Similarly, if O
solution is
y' = ele~(-1+v'1-4A)x + e2e~(-I-v'1-4A)x.
< .\ <
!'the general
In this case the initial conditions again imply el = e2 = O, and so y = O. Now, for A
> ~, the
general solution of the differential equation is

y = cle-xj2 cos)4,\
- 1 x + C2e-xj2 sin V4,\ - 1x.
The condition y(O) = O implies el = O so y = c2e-xj2 sin V4,\ - 1x. From

y(2)
= c2e-1
sin 2)4'\ - 1 = O
we see that the eigenvalues are determined by 2V4'\ - 1
= n1T for n
= 1, 2, 3, .. , . Thus,
the eigenvalues are n21T2/42 + 1/4 for n = 1, 2, 3, .. , , with corresponding
eigenfunctions
Cxj2sin~x.
2
(b) The self-adjoint form is
(c) An orthogonality relation is

fo2 eX (e-Xj2 sin ~1T x) (e-Xj2 cos n21Tx) dx = fo2 (sin ~1T x) (cos ~1T x) dx = O.
9. To obtain the self-adjoint form we note that an integrating factor is (l/x)e](1-x)dxjx
the differential equation is
521
= e-x.
Thus,
Exercises
11.4
and the self-adjoint form is
+ ne-Xy
d~ [xe-Xy']
Identifying the weight function p(x)
= O.
O and
= e-X and noting that since r(x)
limX-->CXlr(x) = O, we have the orthogonality relation

foCXl
e-X Ln(x)Lm(x)
=1 n.
dx = O, m
= e-x2. Thus, the
10. To obtain the self-adjoint form we note that an integrating factor is eJ -2xdx
2
dxd [e-X 2 y ,]
+ 2ne-x
e-X y" - 2xe-x
+ 2ne-x
y'
and the self-adjoint form is
Identifying the weight function p(x) = 2e-x

limx-->CXl
r(x)
= O,
= O.
and noting that since r(x) = e-x,
we have the orthogonality relation

JCXlCXl
2e-x2 Hn(x)Hm(x)
dx
= O, m
=1 ~.
A
l+x
+ x2)y" + 2xy' + --2
(1
Letting x = tan () we have
e = tan
"!
Y = O.
x and
dy
dy de
1 dy
dx = de dx = 1 + x2 de
d2y
dx2
d [
dY]
(d2y de)
de2 dx
= dx 1 +x2 de = 1 +x2
d2y
2x
2x
- (1+x2)2
dy
de
dy
= (1 + x2)2 de2 - (1 + x2)2 de .

The differential equation can then be written in terms of y(e) as
(1 +x)
(1+x2)2
d2y
2x
de2 - (1+x2)2
d2y
1 + x2 de2
= _1_
+ _A_
1 + x2
dY]
de +2x
=O
or
522
[11 +x2
dY]
de
+ 1 +x2Y
limX-->-CXlr(x)
Exercses
11.4
= Y(1/4) = O. For A :::; O the only solution of the

boundary-value problem is y = O. For A > Othe general solution of the differential equation is
y = C1 COS v1 e + C2 sin v1 e. The condition y(O) = O implies C1 = O so y = C2 sin v1 e. Now
The boundary conditions become y(O)
the condition y( 1/ 4) = O implies

v1n / 4
C2
sin v11 / 4 = O. For
C2
=1- O this condition will hold when
= nn or A = 16n2, where n = 1,2,3, .... These are the eigenvalues with corresponding
eigenfunctions sin 4ne = sin (4n tan -1 x), for n = 1, 2, 3, ....

(b) An orthogonality relation is
1
10O
-2--
+1
sin( 4m tan-1 x) sin(4n tan-1 x) dx = O.
12. (a) This is the parametric Bessel equation with /) = 1. The general solution is
y = e1J1 (AX)C2Y1(AX).
Since y is bounded at Owe must have
defines the eigenvalues Al, A2, A3, ....
C2
= O, so that y = cI11(AX). The condition J1(3A) = O
(When A = Othe differential equation is Cauchy-Euler
and the only solution satisfying the boundary condition is y = O,so A = Ois not an eigenvalue.)
(b) Frorn Table 6.1 in Secti0!16.3 in the text we see that eigenvalues are determined by 3A1 = 3.832,
3A2
= 7.016, 3A3 = 10.173, and 3A4 = 13.323. The first four eigenvalues are thus Al = 1.2773,
A2 = 2.3387, A3 = 3.391, and A4 = 4.441.

13. When A = O the differential equation is r(x)y"
+ r'(x)y'
= O. By inspection we see that y
= 1 is a
solution of the boundary-value problem. Thus, A = Ois an eigenvalue.

14. (a) An orthogonality relation is
10
where
Xm
1
COS XmX COS XnX
=O
=1- Xn are positive solutions of cot x = x.
(b) Referring to Problem 1 we use a CAS to compute
101(cosO.8603x)(cos
3.4256x) dx = -1.8771 x 10-6.
15. (a) An orthogonality relation is
101(Xm cos XmX where
Xm
sin XmX)(xn COSXnX - sin xnx) dx = O
=1- Xn are positive solutions of tan x = x.
(b) Referring to Problem 2 we use a CAS to compute
101(4.4934 cos 4.4934x
- sin 4.4934x )(7.7253 cos 7.7253x - sin 7.7253x) dx = -2.5650
523
10-4.
Exercises 11.5
Exercises 11.5
1. Identifying b = 3, the first four eigenvalues are
3.832
3
Al = --
;:j
1.277
A2
;:j
2.339
7.016
3
A3 = 10.173 = 3.391
3
A4
= 13.323 ;:j 4.441.

3
2. We first note from Case
to JI (2A)
rrr in the
text that Ois an eigenvalue. Now, since J6(2A) = Ois equivalent
= O, the next three eigenvalues are

3.832
A2 = -2-
= 1.916
A3 = 7.016 = 3.508
2
A4 = 10.173 = 5.087.
2
3. The boundary condition indicates that we use (15) and (16) in the texto With b = 2 we obtain
dt
1
=
2J[(2/\)
= Ai dx I
1 (2)",
. A; Jo
tJo(t) dt
[From (4) in the text]
\2)..;
1
=
2A[J'f(2Ai) tJ1(t)O
1
Thus
524
Exercises
11.5
4. The boundary condition indicates that we use (19) and (20) in the texto With b = 2 we obtain
el
=~
r2 x dx = ~4 x2212o = 1,
4 Jo
e; =
2
2
4J6(2Ai) Jo XJO(AiX) dx
1
2)';
tJo(t) dt
102),;
-d tJ
dt [
t dt
1( )]
12),;
2AP6(2Ai) t) (t) o
)(2Ai)
A;)6(2Ai) .
= O we conclude
Now since Jb(2Ai) = O is equivalent to J1(2Ai)

the expansion of
dt = Ai dx
= 2J6(2Ai) . A; Jo
1
- 2AP6(2Ai)
t = AiX
ei
= O for i = 2, 3, 4,
....
Thus
f on O < x < 2 consists of a series with one nontrivial term:

f(x) =
= 1.
Cl
5. The boundary condition indicates that we use (17) and (18) in the texto With b = 2 and h = 1 we
obtain
- (4A[
ln
2),;
+ 1)J6(2Ai)
2
-d tJ t dt
dt[ 1()]
12),;
= (4A; + 1)J6(2Ai) tJ1(t) o

4A;)l
=
(4A;
(2Ai)
+ 1)J6(2Ai) .
Thus
00
AiJl (2Ai)
f(x) = 4 L (4A2 + 1)J2(2A') JO(AiX).
t=l
t
o t
525
AiX
dt = Ai dx
[Prom (4) in the text]
Exercises 11.5
6. Writing the boundary condition in the form
2Jo(2A) + 2AJ(2A) = O
we identify b = 2 and h
= 2.
Using (17) and (18) in the text we obtain

Ci =
2A2
(4A; + 4)J6(2Ai)
10
XJO(AiX) dx
r2)";
1
=
2(A; + 1)J6(2Ai) Jo
2(A; + 1)J6(2Ai) th(t)
t = AiX
[Frorn (4) in the text]
dt[th(t)]dt
dt = Ai dx
12)..;
Adl(2Ai)
= (Al + 1)J6(2Ai) .
Thus
7. The boundary condition indicates that we use (17) and (18) in the texto With n = 1, b = 4, and
h = 3 we obtain
r4
2A2
Ci
(16AT-
1 +'g)Jr(4Ai)
5AT
= 4(2AT + 1)Jr(4Ai)
Jo xJl(AiX)5x
4)..;
i(2AT + l)Jr(i)
4Ai(2AT + 1)Jf(4Ai) t h(t)
Jo
2
20Ad2( 4Ai)
= (2A; + 1)Jr(4Ai) .
Thus
526
= AiX
dt:::::Ai dx
4)";
. Ay Jo
dx
t h(t)dt
d
dt[t h(t)] dt
14)..;
Exercises
8. The boundary condition indicates that we use (15) and (16) in the texto With n = 2 and b
11.5
= 1 we
obtain
2
=
JlP.'i) . At
rA 3
Jo
t h(t) dt
= A{Il(Ai) t J3(t)
IAi
O
2
Thus
9. The boundary condition indicates that we use (19) and (20) in the texto With b
el
= ~ r3 xx2 dx = ~ x4
9
Jo
9 4
13
=~
= 3 we obtain
2
. - 1 3A;3
t J, t dt
- 9J6(3Ai) At o
o( )
2
-- 9A{I6(3A
i)
lco3Ai t 2 -dtd [tJ
dt
1( )]
u = t2
du = 2t dt
=
With
ri
9\4;(Jo
3Ai
) (t3Jl(t)
= O in equation (5) in the text we have
527
dv = ft [tJl(t)] dt
v = tJl(t)
[3Ai -2 3Ai t2Jl(t) dt)

o
Jo
J(x)
-Jl(X),
so the boundary condition
Exercises 11.5
Thus
la. The boundary condition indicates that we use (15) and (16) in the texto With b = 1 it follows that
Ci
= Jr~Ai)
101 x (1- x2)
JO(AiX) dx
= Jr~Ai)
[101 XJO(AiX)
dx -
2 [1Ar JofA; tJo(t)
Jr(Ai)
2 [1Ar JofA;
= Jr(Ai)
101 x3 JO(AiX)
fA;
dX]
dt - A; Jo t Jo(t) dt
fA;
dt (tJ1(t)) dt - A; Jo t dt [tJ1(t)]dt
u = t2
du = 2tdt
_3_ fAi!!:_ [2J ( )] d]
= _2_ [Jl(Ai) _ Jl(Ai)

Jr(Ai)
2 [2
Ai
= Jr (Ai) A; t 12(t)
+ A; Jo dt t
Ai
412(Ai)
IAi]
O
Ar Jr (Ai)
Thus
528
dv = -1t [tJ1(t)] dt
v = tJ1(t)
Exercises
11. (a)
11.5
y
4
x
-2
-4
(b) Using FindRoot

X4
in Mathematica we find the roots
11.9561, and Xs
Xl
= 2.9496, X2 = 5.84113, X3 =
15.0624.
(e) Dividing the roots in part (b) by 4 we find the eigenvalues Al

A3
8.87273,
0.7374, A2
146028,
= 2.21818, A4 = 2.98904, and AS = 3.76559.
(d ) The next five eigenvalues are A6
= 4.54508, A7
= 5.32626,
A8 = 6.1085, A9 = 6.89145, and
AlO= 7.6749.
12. (a) From Problem 7, the coefficients of the Fourier-Bessel series are
Using a CAS we find
Cl
26.7896,
C2
= -12.4624,
C3
7.1404,
-4.68705, and
C4
Cs = 3.35619.
(b)
SI
20
s2
15
'iD
15
S3
20
S5
20
15
15
10
(e)
S10
20
1 2 3 4 5x
S10
20
15
15
10
10
5
x
1
-5
3
4x
-10
529
Exercises
11.5
13. We compute
Co =
C1 =
f (x)
1in
~ o
-')
-23
1in
2 o
=-
xPo(x) dx
in xP1(x)dx
xdx
in x2dx
2 o
= -3
= -1
0.8
0.6
0.4
=-1
0.2
~-1-----~0~.5-=~~--~0~.5~--71x
C2
= -5
in XP2(X) dx = -5ia
1
-(3x3
2 o 2
- x)dx = -5
16
C3
= -27
in XP3(X) dx = -72 ino -(5x4

1
2
- 3x2)dx
2 o
C4=
9 (1
2' Jo
XP4(X) dx
=O
9 (1 1
= 2' Jo g(35xs - 30x3 + 3x)dx = - 32
11 (1
11 (1 1
C5
= 2 Jo xPs(x) dx = 2 Jo g(63x6 - 70x4 + 15x2)dx = O
C6
=-
13101
XP6(X) dx
2 O
1
= -13101 -(231x
7
2
16
315x5
+ 105x3 -
5x)dx
= -.13
256
Thus
14. We compute
1
2'31
51
Co = -21
1eXPo(x)dx
-1
C1 =
C2
-11
=-
1
2'31
= -51 -
= -1
2
eXP(x) dx =
1 exp. 2 ( x )
-1
dx
1eXdx
-1
-11
f (x)
1 = -(e
xe" dx
= 3e-1
1 1(3x 2 ex
-1 2
e- 1)
- eX)d x
= ~(e - 7e-1)
2
530
2.5
2
Exercises
Thus
f(x)
+ 3e-1 P1(X) + ~(e -7e-1)P2(X)
= ~(e - e-1)po(x)
2
+ ~(-5e + 37e-1)P3(X) + ~(36e

2
- 266e-1)p4(x)
+ ....
15. Using cos28 = ~(cos 28 + 1) we have
P2(cos8) = 2"(3cos 8 - 1) = 2"cos 8 - 2"

3
= 4'(cos 28 + 1) - 2" =
16. From Problem 15 we have
"4 cos 28 + "4
= 4'(3 cos 28 + 1).
P2(COS&) = 4'(3cos28 + 1)
or
4
1
cos 2&= '3P2(cos 8) - '3 .
Then, using Po (cos 8)
= 1,
F(8) = 1 - cos 28 = 1 - [~P2(COS 8) - ~]
4
= '3
17. If
'3P2(cos8)
'3 Po (cos8)
- '3P2(cos8).
f is an even function on (-1,1) then
J 1 f(x )P2n(X)
dx = 2
-1
Jo
f(x )P2n(X) dx
and
J 1 f(X)P2n+1(X)
dx = O.
-1
Thus
C2n = 2(2n) + 12
J1
-1
f(X)P2n(X)
4n + 1 ( 2
dx = -2
= (4n
+ 1) fo1f(x)P2n(X)
C2n+1 = O, and
00
f(x) =
L C2nP2n(X).
n=O
18. If
f is an odd function on (-1,1) then
II
f(x)P2n(x)
dx = O
and
531
IIo f(X)P2n(X) dx )
dx,
11.5
Exercises 11.5
r1-1 f (x )P2n+ 1( X) dx = 2 Jor 1 f (x )P2n+ 1( X) dx.

1
Thus
C2n+1 = 2(2n
+ 1) + 1 JI
2
-1
+3
4n
dx = --
f(X)P2n+1(X)
(2
la
f(X)P2n+1(X)
+ 1) 101 f(X)P2n+1
= (4n
dx )
(X) dx,
C2n = O, and
00
f(x)
L C2n+1P2n+I(X).
n=O
19. From (26) in Problem 17 in the text we find
l
lol
1
loo xPo(x)dx= o xdx=-, 2
l
lol 1
=5
loo XP2(X) dx = 5 028-(3x3 -
s
1
co=
C2
C4 =
t' XP4(X)
Jo
dx = 9
0.8
5
x)dx = - ,
0.6
0.4
t' ~(35x5 - 30x3

Jo 8
+ 3x)dx
= - 36 '
and
C6
= 13 1 XP6(X) dx = 13
Jo
l ~(231x7
Jo 16
+ 105x3
- 315x5
- 5x)dx = ~.
128
Hence, from (25) in the text,

1
2
f(x)
On the interval -1
= -Po(x)
5
8
+ -P2(X)
3
- -P4(X)
16
13
128
+ -P6
< x < 1 this series reprcsents the function f(x)
+ ....
=
[z].
20. From (28) in Problem 18 in the text we find

ci = 3
C3=7Jo
Jo
1 PI
(x) dx = 3
t' P3(x)dx=7Jo
r
r
Jo
l
x dx = ~,
2
1
2(5x3-3x)dx=-S'
C5
= 11 Jo P5(X) dx = 11 Jo
S1 (63x5
-70x3
+ 15x)
11
dx = 16
and
C7
= 15
Jo
P7(X) dx = 15
t' ~
Jo 16
(429x7 - 693x5
+ 315x3
532
- 35x) dx =
_2.?_.
128
Exercises
11.5
Hence, from (27) in the text,

3
11
75
f(x) = 2"H(x) - SP3(X) + 16P5(x) - 12SP7(x) + ....

On the interval -1 < x < 1 this series represents the odd function
-1 < x <
1,0<x<1.
-1,
f(x) = {
21. Since
f is expanded as a series of Bessel functions, JI (AiX) and JI is an odd function, the series
should represent an odd function.

22. (a) Since
Jo is an even function, a series expansin of a
function defined on (0,2) would converge to the even

extension ofthe functon on (-2,0).
':1
0.5
-2
2 x
= 2J2(X)jx - J3(X). Since h is even
(b ) In Section 6.3 we saw that J~(x)

and
-1
h is odd we see that

J2(-x)
2J2(-X)j(-x)
- h(-x)
+ h(x)
= -2h(x)jx
= -J~(x),
so that J~ is an odd function. Now, if f(x) = 3h(x)
+ 2xJ2(x),
we see
that
f( -x)
= 3h(
-x) - 2xJ~( -x)
= 3h(x) + 2xJ~(x) = f(x),

so that
f is an even function. Thus, a series expansion of a function
-1 -0.5
0.5
1x
defined on (0,1) would converge to the even extension of the function on
(-1,0).
23. Since there is a Legendre polynomial of every degree, any polynomial can be represented as a finite
linear combination of Legendre polynomials. For f(x) = x2 we have

eo
= ~ (1 x2. 1 dx = ~
el
= -32
21-1
e2=- 5
JI x x dx
JI x -(3x -l)dx=-,
2
-1
2 -1
so
x2
~Po(x) + ~P2(X).
3
3
533
2
3
Exercises 11.5
For j(x) = x3 we have
co=- 1 j1
2
-1
ldx=O
j1-1 x3. X dx = ~5
5 j1
3 1
2
C2 = x - (3x - 1) dx
C1 = ~
2
2
C3=-
7
2
-1
-1
1
-(5x
=O
2
-3x)dx=-,
so
1. True, since J~1I'(x2 - 1)x5 dx = O

2. Even, since if f and 9 are odd then h( -x)
3. Cosine, since
= f( -x)g(
-x)
= - f(x)[-g(x)]
= f(x)g(x)
= h(x)
f is even
4. True
5. False; the Sturm-Liouville problem,
d~[r(x)y']
y'(a) = o,
+ :>"p(x)y = O,
on the interval [a, b], has eigenvalue
:>..
y'(b) = o,
= O.
6. Periodically extending the function we see that at x = -1 the function converges to

at x = Oit converges to 1(0+ 1) = 1, and at x = 1it converges to 1(-1
1(-1+0) = -1;
+ O) = -1.
7. The Fourier series will converge to 1, the cosine series to 1, and the sine series to O at x = O.
Respectively, this is because the rule (x2
+ 1) defining
f(x)
determines a continuous function on
(-3,3), the even extension of f to (-3, O) is continuous at O, and the odd extension of f to (-3, O)
approaches -1 as x approaches Ofrom the left.
8. cos.5x, since the general solution is y = Cl
COS .,fi. X
+ C2 sin.,fi. x and y' (O) = O implies
C2
= O.
9. Since the coefficient of y in the differential equation is n2, the weight function is the integrating
factor
534
11 Review Exercises
Chapter
on the interval [-1, 1]. The orthogonality
j
10. Since Pn(x) is orthogonal
relation
Vf=X2 Tm(x)Tn(x)
-1
=
to Po(x)
m =1= n .
dx = O,
> O,
1 for n
j 1 Pn(x)
j1
dx =
-1
11. '.Ve know from a half-angle
is
PO(x)Pn(x) dx = O.
-1
formula in trigonometry
cos2 x = ~- ~cos 2x, which is a cosine
that
series.
12.
(a) Por m
lno
L .
sin
=1=
(2n
+ 1)71'x
2L
sin
(2m
+ 1)71'x dx
2L
1
=-
laL (
2 o
COS
n-m
-L-71'X
COS
+ m + 71'71'X) dx
L
(b) From
rL sin2 (2n 2L+ 1)71'x dx = )0rL (~2 _ ~2 cos (2n 2L+ 1)71'2:) dx = !::.2
)0
we see that
'
11
sin
(2n
+ 1)71'x 1I =
2L
ff
Vi'
13. Since
Aa = JO (- 2x) dx = 1,
-1
An =
Bn =
j (-2x)
and
-1
(-2x)
cos nsrz dx = 22'[(-l)n
n 71'
sin nzr r dx
-1
-1],
= _(_l)n
n71'
for n = 1, 2, 3, ... we have
f(x)=~+
f(
22 2[(-lt-1]coSn71'x+~(-1)nSinn71'x).
n=1 n 71'
n71'
14. Since
Aa =
An =
and
-. =
-1
(2x -l)dx
2
3
= --,
(2x2 -1)cosn71'xdx
-1
= 22'(_l)n,
n 71'
j1 (2x2 _ 1) sin n71'Xdx = O

-1
535
O.

for n
= 1, 2, 3, ... we have
15. (a) Since
and
An = 2
J-l
e-X
cosmrx dx =
22 2((1- (_l)ne-lJ
l+n 7f
for n = 1, 2, 3, ... we have

00
f(x)
= 1 - e-l
+2 L
l-(-l)ne-l
1+ n
n=l
cos n-rz.
2 2
tt
(b) Since
for n
16.
= 1, 2, 3, ... we have
f (x)
2
y
1
0.5
-1 -0.5
0.5
1x
-1 -0.5
-0.5
f(x)
0.5
1x
-0.5
-1
= Ixl - 1
f(x)
= 2x2
e-, ,
= e-Ixl
f(x)
{
f(x)
~ex,
O<x<l
x=O
-1 < x
17. For A > O a general solution of the given differential equation is

y = Cl cos(3v'\ In x)
+ C2 sin(3v'\
In x)
and
,
r:
3Cl';;
y = ---sin(3vA
x
Since In 1 = O, the boundary condition y'(l)
lnx)
3C2';;
+ --COS(3VA
= O implies
C2
y = Cl cos(3v'\ lnx).
536
r:
lnx).
= O. Therefore
<O
Chapter
Using In e = 1 we find that y(e)
n = 1, 2, 3, ....
cos (2n;
= O implies
Cl
cos 3v
r;
A
11 Review Exercises
2n - 1
r:
= O or 3v A = -2-7f,
for
The eigenvalues are A = (2n - 1)27f2/36 with corresponding eigenfunctions
1 tt In x) for n = 1, 2, 3, . . . .
18. To obtain the self-adjoint form of the differential equation in Problem 17 we note that an integratng
factor s (1/x2)ef
dx] =
l/x. Thus the weight function is 9/x and an orthogonality relation is
he ~ cos en;
17flnx) cos em2-17rInx)
dx = o, m -1-
n.
19. The boundary condition indicates that we use (15) and (16) of Secton 1l.5 in the texto With b = 4
we obtain
Ci
fn4 XJO(AiX)f(x)
= 16J2(12 4,A)
t":
Jo
8Jr(4Ai)
8Jr(4Ai)tJl(t)
[From (4) in 1l.5 in the text]
dt[tJl(t)]dt
\2)..i
dx
JI (2Ai)
=
4AiJr(4Ai)'
Thus
20. Since f(x)
x4 is a polynomial in x, an expansion of f in polynomials in x must terminate with

the terrn havng the same degree as f. Using the fact that X4Pl(X) and x4P3(X) are odd functions,
=
we see immediately that
Cl
ca ="2
C2=- 5
C4
= -9
= C3 = O. Now
JI x4po(x) dx ="2 JI x4dx = 5

JI x4P2(X)dx=- JI -(3x6-x4)dx=JI X4P4(X) dx = - JI -(35x8 - 30x6 + 3x4)dx
-1
-1
-1
5
2
-1
1
2
-1
9
2
-1
1
8
Thus
537
8
= -.35
12
Partial Differential
Equations and
Boundary-Value
Rectangular
____
Problems in
Coordinates
Exercises 12.1
1. If u = XY then
Ux
= X'Y,
Uy
= xv',
Xly=XY',
and
X'
Y'
-=-=A.
Then
so that
X
Ale,X2x,
y = A2e,X2y,
and
2. If u = XY then
Ux
= -,y'y,
uy
= Xy',
x'v
= -3XY',
and
X'
Y'
-=-=A.
-3X
y
Then
538
Exercses 12. 1
so that
= A1e'f3).2x,
y = A2e).2y,
and
3. If u = XY then
Ux
= Xly,
Uy
= Xyl,
Xly = X(Y - y/),

and
X'
-=
X
Y - Y'
=),
Then
so that
X
= Ale).2x,
y = A2e(l'f).2)y,
and
4. If u = XY then
Ux
= Xly,
Uy
= xv',
Xly = X(Y
+ yl),
and
Then
so that
X
= Ale).2x,
y = A2e(-1).2)y,
and
539.
Exercises 12. 1
5. If u = XY then
ux = X'Y,
uy
xX'y
and
= XY',
=yXY',
xx'
yY'
-=
-=,\
Then
so that
x = A1X>.2,
y = A2y>.2,
and
6. If u = XY then
uy
yX'y
and
x'
= XY',
= xXY',
y'
-X =-Y
x
-y
=,\.
Then
so that
x = Ale>.2x2/2,
y = A2e::::>'
2y2/2,
and
7. Ifu=XYthen
uxx = Xl/y,
"vv = Xyl/,
and
540
yx = X'y',
Exercises 12. 1
X"y + x'v' + XY" = O,
which is not separable.
8. If u =
XY then
uyx
= X'Y',
yX'Y' + XY = O,
and
x'
-X
yY'
-=-=A.
Then
so that
X
= Ale:::>.2x,
y = A2y1/>.2,
and
9. If u
= XT
then
Ut
uxx
= XT',
= X"T,
kX"T - XT
= XT',
and we choose
T'
kX" - X
so that
For A2 > O we obtain
so that
For _A2
< O we obtain
so that
541
-1 k);
Exercises
12. 1
If ).2 = Othen
T' + T = O,
X" = O and
and we obtain
In this case
10. If u = XT then
Ut
= XT',
uxx = X"T,
kX"T = XT',
and
X"
T'
2
= - = ).
X
kT
so that
For
).2
> O we obtain
so that
U
For
_).2
= XT = e-c~kt(
Cl
cos C3X +
') C3X .
C2 sm
< O we obtain
X
= Ale-AX + A2eAX,
T = A3eA2kt,
and
For
).2 = O we
obtain
X = AlX
11. If u
+ A2,
and
= XT then
uxx
= X"T,
Utt =
XT",
542
= XT
= ClX
+ C2.
Exercises 12. 1
a2X"T = XT"
and
so t.hat,
For ,\2
> O we obtain
= Al
COS
AX
+ A2
sin AX,
T = A3 cos aAt + A4 sin a.si;

and
u
For _A2
<O
= XT = (Al
COS
AX
+ A2
sin Ax)(A3 cos aAt
we obtain
= AleAx
+ A2e-AX,
T = A3eaAt + A4e-aAt,
and
For A 2
= O we
obtain
= AIX+A2,
T = A3t + A4,
and
12. If u = XT then
Ut
= XT',
Utt
= XT",
XX = X"T,
a2 X"T = XT"
and
X"
X
so that.
543
+ 2kXT',
+ A4
sin aAt).
Exercses 12. 1
For >,2
> O we obtain
x = Ale'\x + A2e-'\x,
T
= A 3e (-k+\fk2+a2,\2)t + A 4e (-k-v'k2+a2,\2)t
and
u = XT
For -,>,2
= (Ale'xx
+ A2e-'xx)
+ A4e(-k-v'k2+a2,X2)t).
(A3e(-k+v'k2+a'l,\2)t
< O we obtain
x = Al
= e-kt (A3 cos
COS >'X
Ja2>,2 -
+ A2
k2t
sin >'x.
+ A4sin
Ja2>,2 -
k2 t).
If k2 - a2>,2 = O then
so that
For >,2
u = XT = (Al
COS >'X
+ A2 sin >'x) ( A3e( -k+v'k2_a2,X2)t + A4e( -k-v'k2-a2,X2)t)
= (Al
COS >'X
+ A2
= (Al
COS ~x
sin>'x)e-kt
+ A2 sin ~x)
(A3
COS
(A3e-kt
J a2>,2-
+ A4te-kt)
= O we obtain
X=AlX+A2,
T
= A3 + A4e-2kt,
and
13. If u = XY then
uxx = X"Y,
"vv = XY",
X"Y
+xY"
and
544
= O,
k2 t
.
+ A4
sin
Ja2>,2 -
k2 t)
Exercses 12. 1
X"
y"
-X
_=_=,\2
so that
For
,\2
> O we obtain
X
= Al
y
cos Xz
+ A2 sin Xz,
= A3eAY + A4e-AY,
and
For
_,\2
< O we obtain
y = A3 cos'\y
+ A4 sin '\y,
and
For
14. If u
,\2
= O we obtain
= XY then
Uxx
= X"Y,
Uyy
= XY",
x2X"y + xY" = O,
and
x2X"
y"
Y
--=-=,\
-X
so that
For
,\2
> O we obtain
If 1 -
4,\2
> Othen
If 1 -
4,\2
< Othen
545
Exercises
12. 1
If 1 - 4A2
= O then
so that
U
For
_)..2
= XY =
+ A2X(1/2-~/2))
(AlX(1/2+~/2)
COS ~J4)..2
(A3eAY + A4e-AY)
- llnx + A2 sin ~J4)..2 - llnx)
xl/2 (Al
(AlXl/2 + A2xl/21n x) (A3ey/2 + A4e -y/2) .
(A3ey/2 + A4e-y/2)
< O we obtain
y = A3 cos)..y + A4 sin )..y,

and
U
= XY =
(Ale(1/2+v'1+4A2/2)X + A2e(1/2-v'1+4),2/2)x) (A3 cos)..y + A4 sin )..y).
For )..2 = O we obtain
x = AlX+A2,
y
= A3y+A4,
and
15. If U = XY then
xx
= X"Y,
yy = Xy",
X"y + XY" = XY,

and
X"
X
y - y"
Y
_-:-:-_= )..2
so that
X" =f )..2X = O and
y" + ()..2 - l)Y = O.
For )..2 > O we obtain

If )..2- 1 > Othen
y = A3COS~y
+ A4sin~y.
546
Exercises
If
1 < Othen
).,2 -
+ A4e-~Y.
y = A3e~Y
If
1 = Othen Y
).,2 -
= A3Y + A4 so that
= XY =
=
For
12. 1
_).,2
(AleAX +A2e-AX) (A3COS~y+A4Sin~y),

(AleAX + A2e-AX) (A3e~Y
+ A4e-~Y)
< Owe obtain
x = Al cos Xz + A2 sin ).,x,

y = A3e~Y+A4e-~Y,
and
u
For
).,2
= XY
(Al cos Xz + A2 sin Ar ) (A3e~Y
+ A4e-~Y)
= O we obtain
x
+ A2,
y = A3eY + A4e-Y,
= Alx
and
16. If u = XT then
Utt
= XT",
XX = X"T,
and
a2X"T - g = XT",
which is not separable.
= C = 1, we compute B2 - 4AC = -3 < O. The equation is elliptic.

18. Identifying A = 3, B = 5, and C = 1, we compute B2 - 4AC = 13> O. The equation is hyperbolic.
17. Identifying A = B
19. Identifying A = 1, B = 6, and C = 9, we compute B2 - 4AC = O. The equation is parabolic.

20. Identifying
1, B
= -1, and C
= -3, we compute
B2 - 4AC
13 > O. The equation is
hyperbolic.
21. Identifying
A = 1, B = -9, and C = O, we compute B2 - 4AC = 81 > O. The equation is
hyperbolic.
1, and C = O, we compute B2 - 4AC = 1 > O. The equation is hyperbolic.
23. Identifying
A = 1, B = 2, and C = 1, we compute B2 - 4AC = O. The equation is parabolic.
24. Identifying
A = 1, B = O, and C = 1, we compute B2 - 4AC = -4 < O. The equation is elliptic.
O, B
22. Identifying
547
Exercises 12.1
. 25. Identifying A
= a2, B = 0, and
C = -1, we compute B2 - 4AC
= 4a2 >
O. The equation is
hyperbolic.
26. Identifying A
= k > 0, B = 0, and
C =
0, we compute B2
4AC = -4k < O. The equation is
elliptic.
27. If u = RT then
u; = R'T,
= R"T,
Urr
Ut
= RT',
RT' = k (R"T + ~R'T) ,

and
If we use
r2R" + rR'
T'
2
--r"-2R=--- = -kT-= .x .
_.x2 <
then
r2R!' + rR' + .x2r2R= O and
T":: .x2kT = O
so that
R = A2Jo(.xr) + A3yo(.xr),
= A1e-k,X2t,
T
and
28. If u = Re then
Ur =
= R"e,
Urr
ii
R'e,
= Re',
Utt =
Re",
R"e + ~R'e + .2_Re"

r
and
r2
r2 R" + rR'
--R--=--e
e"
= O,
=.x.
Then
r2R!'+rR'-.x2R=0
and
548
e"+.x2e=0
Exercises
12.2
so that
8(j =
Cl
+ C2 sin ).(j,
cos).(j
and
u =
29. For u
R8 =
(Cl
cos).(j
+ C2 sin ).(j) (C3r'\ + C4r-'\).
= Al e,\2y cosh 2).x + B, e,\2y sinh 2).x we compute

02u
2
,\2
ox2 = 4). Al e y cosh 2).x
+ 4).2B; e,\2y sinh 2).x
and
ou
ay = ).2 Ale
Then 02ujox2
,\2
y cosh Zx.r
+).
2,\2
Ble
y sinh Zxz;
= 40ujoy.
For u = A2e-,\2y cos 2).x
+ B2e-,\2y
sin 2).x we compute
02u
2,\2
ox2 = -4). A2e-
y cos 2).x -
2,\2
4).
B2e-
y sin 2).x
and
ou
2
ay = -). A2 cos 2).x -).
Then 02ujox2
For u
x2
+ 4y2
B2 sm 2).x.
= 40ujoy.
= A3X + B3 we compute 02ujox2 = oujoy
30. We identify A
= xy + 1, B = x + 2y, and C
= O. Then 02ujox2
= 1. Then B2 - 4AC
= 40ujoy.
= x2 + 4y2 - 4. The equation
4 defines an ellipse. The partial differential equation is hyperbolic outside the ellipse,
parabolic on the ellipse, and elliptic inside the ellipse.
Exercises 12.2
1. k ~:~ = ~~,
u(O, t)
= 0,
<
ou
ox
u(x, O) = (x),
2. k ~:~
= ~~,
u(O,t)=uo,
u(x, O) = 0,
< L, t >
Ix=L = 0,
<x <L
< x < L, t >
u(L,t)=Ul,
<
t>O
<L
549
Exercises 12.2
3.
k [}2u = ou
ox2
Bt:'
OU I
a
x
u(O, t) = 100,
u(x, O) = f(x),
o2u
4. k ox2
< x < L, t > O

x=L
= -hu(L,
+ h( u-50)
2 o2u _
o2u
ox2 - ot2 '
OU
ot '
u(x, O)
- x),
oUI
OX
o2u
= ot2'
u(O, t) = O,
. a
2 o2u
ox2
OX
u(O, t)
. 7fX
= Sln-
L '
t=O
-ou
ot
= sin
I
t=O
o2u
+ Ax = ot2'
Bu
OX
02U o2u
ox2 + oy2 = O,
x=O
=0
'
t=O
O<x<L
< x < L, t > O
ni,
=0
'
t >O
O<x<L
< x < L, t > O, A a constant
u(L, t) = O,
u(x, O) = O,
oUI
OX
u(L, t)
= O,
O<x<L
t >O
_ 2/3 ou - o2u
ot - ot2 '
u(x, O) = f(x),
02U
-O
t=O-
< x < L, t > O
-ou
u(O, t) = O,
8. a2 ox2
t>O
u(L, t) = O,
u(x, O) = O,
9.
< x < L, t > O
= O,
202U
t>O
u(L,t)
6. a ox2
< x < L, t > O
u(O,t) = O,
= x(L
t>O
O<x<L
ou I = O ou I
ox x=O ' OX x=L
u(x, O) = 100,
O<x<L
5.
t),
=O
t>O
'
O<x<L
< x < 4, O < y < 2
u(4,y)=f(y),
0<y<2
550
Exercses
8u
=O
8y y=o
)
u(x,2)
82u 82u
10. 8x2 + 8y2 = O,
= O,
12.3
0<x<4
< x < tt ; y>

lOO,
u(-rr, y) = {
O,
O<y
:s 1
y> 1
u(x, O) = f(x),
Exercises 12.3
1. Using u = XT and
_).2
as a separation constant leads to
X" + ).2X = O,
X(O) = O,
X(L)
= O,
and
Then
for
ti
= 1, 2, 3, ... so that
Imposing
n-rr
00
u(x, O) =
L An sin LX
n=l
gives
n-rr dx= -2 ( 1 - cos -n-rr)

An = -2foL/2 sin -x
L o
L
n-rr
2
for
ti
= 1, 2, 3, ... so that
u(x, t) = -
-rr
2. Using u = XT and
_).2
00
L
n=l
n-rr
e -:r sin -L x.
1 - cos!!!!: _ kn2".2 t
2
X" + ).2X = O,
X(O) = O,
551
Exercises
12.3
X(L) = 0,
and
+ k).,2T =
T'
O.
Then
n7r
X = c sin LX
and
n7r
for n = 1, 2, 3, ... so that
Imposing
gives
An=ZJo
(L
4L2
n37r3[1-(-ltl
x(L-x)sinLxdx=
for n = 1, 2, 3, ... so that

_ 4L2
u(x,t) 3. Using u
= XT and
_).,2
1- (_l)n
3
n=l
n
00
-3
7r
. nat
=:
x.
-~t
L
X"+).,2X=0,
X'(O) = O,
X'(L) = 0,
and
Then
n7r
X = C cos LX
for n
and
= 0, 1, 2, ... so that
Imposing
00
u(x, O) = f(x) =
n7r
An cos LX
n=O
gives
1 (L
u(x,t)=ZJo
f(x)dx+Z"I;
00
((L
n7r)
Jo f(x)cosLxdx
552
kn2,,2 t
e-----y:-
=r=
n7r
Exercises
4. If L
=2
and f (x) is x for O < x
= -1 +4 [100
u(x,t)
5. Using u
< 1 and f (x) is O for 1 < x < 2 then
= XT and
_)...2
n=l
2n-rr
. tit:
1 (
tit:
sm- + -cos- -1
2
n2-rr2
2
)]
_ kn2,,2 t
4
tit:
COS-x.
2
X" + )...2X = O,
X' (O)
= O,
X'(L) = O,
and
T' + (h+ k)...2)T = O.

Then
and
for n = O, 1, 2, ... so that
00
u= Ane
T = C2e
2 2)
- ( h+!q.f-
2 2) t nst
( h+!q.fcosTx.
L
n=O
Imposing
u(x, O)
= f(x) =
tit:
00
An cos TX
n=O
gives
u(x, t) =
e-ht
rL
Jo f(x) dx
+L
E (rL
00
n-rr)
Jo f(x) cos Txdx
_(h+kn2,,2)t
=tr:
n-rr
cos TX'
6. In Problem 5 we instead find that X(O) = O and X(L) = O so that

n-rr
X = el sinTx
and
2 00 ( L
n-rr)
u=f(x) sin -x dx
L n=l Jo
L
_ (h+
kn2,,2 )
-;:-
n-rr
sin -L x.
7. (a) The solution is
where
An = - 2
100
[J
50
O
n-rr dx
0.8x sin -x
100
+ llOO
50
n-rr] dx = 22
320
0.8(100 - x) sin -x
sin -n-rr .
100
n -rr
x
Thus,
553
12.3
Exercises
12.3
= O for neven, the first five nonzero terms correspond to n = 1, 3, 5, 7, 9. In this

case sin(mr/2) = sin(2p - 1)/2 = (-l)p+l for p = 1, 2, 3,4,5, and
(b) Since An
_ 320 ~ (_l)p+l (-16352(2p-l)2rr2/1002)t . (2p-1)71"

( x, t) - -2 L. (
)2e
sin
00
71" p=l 2p - 1
u(x,
x.
ti
8.
100
x=3lT/4
x=5lT/6
x=11lT/12
X=1T
Exercises 12.4
1. Using 'u = XT and _,\2 as a separation constant leads to
X,,+,\2X=0,
X(O)
= O,
X(L) = O,
and
Then
554
Exercises 12.4
for n
= 1, 2, 3, ... so that
u=
L (mra
AnCosTt+BnsinTt
nTia )
00
no:1
nTi
sin LX.
Imposing
u(x, O)
nTi
00
= :x(L -
x)
= L An sin LX
n=1
and
nTia
00
Ut(x, O) = O =
L BnT
no:1
nTi
sin LX
gives
for n
= 1, 2, 3, ... so that
L2
u(x,t) = 3
Ti
00
L
n=l
1 - (_l)n
3
nTia.
nTi
=:r:
t sm-L x.
2. Using u = XT and _)-2 as a separation constant leads to
X" + )-2X = O,
= O,
X(O)
X(L) = O,
and
Then
. nTi
X = C1sm-x
L
for n = 1, 2, 3, ... so that
u=
and
nTia
= C2COS
Tt
L (nTia
AnCosTt+BnsinTt
00
nTia
+ C3sin Tt
nTia )
ntt
sin LX.
n=l
Imposing
nTi
00
u(x,O) = O =
L AnsinLx
n=1
and
Ut(x, O) = x(L - x) =
nTia
L BnT
00
n=l
gives
An = O and
555
ntt
sin LX
Exercises 12.4
for n = 1, 2, 3, ... so that
u(x, t)=
3. Using u = XT and
_.,\2
4L3
-4
a7l'
1- (_l)n . mw
4
sin -t
n=l
n
L
. mI'
sin -L x.
CXl

X"+.,\2X=0,
= O,
X(O)
X(L) = O,
and
Then
X
n7l'
= C1Sm-x and T
L
n7l'a
C2
cos Tt
n7l'a
+ C3 sin Tt
for n = 1, 2, 3, ... so that

u=
CXl
n=l
nat a
Ancos-t
+
. n7l'a)
. n7l'
BnsmL' t sm-x.
Imposing
n7l'
CXl
An sin LX
u(x, O) =
n=l
gives
2 ( (L/3 3
ntt
IX sin LX dx
An = I Jo
(2L/3
n7l'
+ lL/3
sin LX dx
(L
+ 12L/3
so that
6V3
1=7'
6V3
A5 = - 5271'2'
As
= O,
a7 =
6V3
7271'2
....
Imposing
Ut(x,O) = O
BnTn7l'a sinLxn7l'
CXl
n=l
556
3)
3 - IX
n7l'
sin LX dx
Exercises
for n = 1, 2, 3, ...
gives En
u(x t)
,
6V3 ( cos -t7fa
= --
4. Using u = XT and
_,\2
7f
57fa
sin -x - - cos -t
L
52
L
7f2
so that
57f
+ -721 cos -t77fa.

L
sin -x
L
X" + ,\2X = 0,
= 0,
X(O)
X(7f) = 0,
and
Then
X =
el
sin nx
and
T=
e2 cos nat
+ e3 sin nat
for n = 1, 2, 3, ... so that

00
u=
L (An cos nat + En sin nat) sin nx.

n=l
Imposing
u(x, O) = ~x(7f2- x2) =

6
An sin nx
and
Ut(x, O) = O
n=l
gives
En
for n
= 1, 2, 3, ... so that
u(x, t) = 2
.
5. Using u
= XT and
_,\2
an d An
00
(_l)n+l
n=l
= 23 (_l)n+l
n
cos nat sin nx.
X" + ,\2X = 0,
X(O)
= 0,
X(7f) = 0,
and
Then
X = el sin nx
and
T = e2 cos nat + e3 sin nat
557
77f
sm -x
L
- . .. .
12.4
Exercises 12.4
for n
= 1, 2, 3, ... so that
00
2:= (An
u =
cos nat
+ B sin nat)
sin nx.
n=l
Imposing
00
u(x, O) = O =
00
2:= An sin nx
Ut(x, O) = sin x =
and
n=l
2:= Bnna
sin nx
n=l
gives
1
a
B; = 2"'
An = O,
and
B = O
for n = 2, 3, 4, ... so that

1
u(x, t) = - sin at sin x.
a
6. Using u = XT and
_).2

X"
+ ).2X = O,
X(O) = O,
X(l)
= O,
and
Then
X
for n
Cl
sin mrx
and
= C2 cos ntt at + C3 sin nitat
1, 2, 3, ... so that
00
2:= (An
cos rnrcz
+ Bn
sin n7f"at) sin n7f"x.
n=l
Imposing
00
2:= An
u(x, O) = 0.01 sin 37f"x =
sin n7f"X
n=l
and
00
Ut(x, O) = O =
2:= Bnn7f"a sin n7f"X

n=l
gives Bn
= Ofor
= 1, 2, 3, ... , A3 = 0.01, and An = Ofor n = 1, 2, 4, 5, 6, ... so that

u(x, t)
7. Using u
= XT and
_).2
= 0.01sin37f"x cos Srrcz.

X"
+ ).2X =
X(O) = O,
558
O,
Exercises
X(L) = 0,
and
Then
n7r
X = clsin-x
T = C2 cos
and
n7ra
=:'
+
C3
. n7ra
sin
=:'
for n = 1, 2, 3, ... so that

u=
L
00
n7ra
AnCosTt+EnsinTt
tino. )
nt:
sin LX.
n=l
Imposing
nt:
00
u(x, O)
L An sin LX
n=l
and
Ut(x,O) = O =
n7ra
00
L EnT
tvn
sin LX
n=l
gives
En
and
8h
n7r
= """22
sin n 7r
2
An
for n = 1, 2, 3, ... so that

8h
u(x,t) = 2"
7r
00
L 2"sinn=ln
tit:
n7ra
tun
=rt:
t sin-L x.
8. Using u = XT and _,\2 as a separation constant leads to
X"+,\2X=0,
X'(O)
= O,
X'(L)
= 0,
and
Then
n7r
X = Cl cos LX
for n
= 1, 2, 3, ....
Since ,\
and
n7ra
T = C2 cos Tt
= O is also an eigenvalue with eigenfunction X(x) = 1 we have

u = Ao +
tina
00
ten
LX.
L Ancos Ttcos
n=l
Imposing
fo
00
u(x,O) = X =
559.
n7r
AncosLx
12.4
Exercises 12.4
gives
Ao = ~ (L x dx = !:_
L Jo
2
and
2 (L
mI'
An = L Jo xcos ;xdx
for n
2L
n
n27!'2[(-1) -1]
= 1, 2, 3, ... , so that
L 2L
u(x, t) = - + 2"
2
9. Using u
= XT and
7!'
(_l)n
.L
n=l
n
-1
00
nata
n7!'
cos -L t cos -L x.
_>.2 as a separation constant leads to
X"+>.2X=0,
X(O) = O,
X(7!') = O,
and
T" + 2(3T' + >.2T = O.

Then
so that
u=
f e-f3t (An cos Jn2 - t + B sin Jn2 - t) sin nx.

(32
(32
n=l
Imposing
00
.L An sin nx
u(x, O) = f(x) =
n=l
and
Ut(x, O) = O =
f (BnJn2
- (32 - (3An) sin nx
n=l
gives
u(x,
t) = e-f3t f An (e os Jn2 - (32 t + J n (3
where
211' f(x)
An = -
7!'
10. Using u
= XT
sin Jn2 - (32
2 - (32
n=l
sin nx dx.
and _>.2 as a separation constant leads to X"
560
t) sin nx,
+ >.2 X = O, X(O)
= O, X(7!')
= Oand
Exercises
TI/ + (1
that
+ )..2)T
= o.
= O, T'(O)
Then X =c2sinnx and T = c3cosvn2

00
u =
+ 1t
12.4
for n = 1, 2, 3, ... so
LEn cos n2 + 1 t sin nx.

nI
00
Imposing U(x, O) =
L En sin nx
gives
n=l
2la
1r
En = -
1T O
/2 X sin nx dx + -2
1r
1T 1r/2
4
n1T
(1T - x) sin nx dx = -2 sin 1Tn
2
neven
O,
= { :rr;:l1
4 (_1)(n+3)/2
n = 2k - 1, k = 1, 2, 3, ....
Thus with n = 2k - 1,
4
u(x,t)
00
sin
-;:;--:--;
n1r
L --{-COSVn2
1T
n
=-
00
(_l)k+1
)2 cosJ(2k
1
L (2k 1T
+ 1tsinnx
=-
n=1
k=1
-1)2
+ 1tsin(2k
-l)x.
XI/(L)
O.
11. Separating variables in the partial differential equation gives

X(4)
TI!
-=--=)..
X
a2T
so that
and
X = Clcosh )..x
+ C2sinh )..x + C3cos )..x + C4sin )..x
T = Cscos a)..2t
+ C6sin a)..2t.
The boundary conditions translate into X(O)

X(O)
= XI/(O) = O we find
= C3= O.
= X(L)
O and XI/(O)
From
From
X (L) = C2sinh )..L + C4sin)..L = O

Xl/ (L) = )..C2sinh
2
)..L - )..2C4sin )..L = O
we see by subtraction that C4sin)..L = O. This equation yields the eigenvalues ).. = n1TL for n = 1,
2, 3, ....

X
n1T
= c4sin-x.
. L
Thus
561
Exercises 12.4
From
u(x, O) = f(x)
00
n7r
L An sin LX
n=l
we obtain
An =
L Jo
nt:
f(x) sin LX dx.
From
and
we obtain
and
2L loL
. nat
e; = -2-2g(x)sm-L xdx.
n
a O
7r
12. (a) Using
x = CIeosh AX + C2sinh AX + C3eos AX + C4sin AX

=O
= O, XI(O) = O we find, in turn, C3= -C! and C4= -C2. The eonditions X(L)
and XI (L) = O then yield the system of equations for Cl and C2:
and X(O)
q (eosh AL
Cl(Asinh AL
- eos AL) + C2(sinh AL - sin AL) = O
+ Asin AL) + C2(Aeosh AL -
Aeos AL) = O.
In order that this system have nontrivial solutions the determinant of the eoefficients must be
zero:
A(eoshAL - cos AL)2 - A(sinh2 AL - sin2 AL)
= O.
A = O is not an eigenvalue sinee this leads to X = O. Thus the last equation simplifies to
eosh AL eos AL = 1 or eosh x eos x = 1, where x = AL.
,- '", ,
(b ) The equation eosh x eos x = 1 is the same as eos x
0.5
\ sech x
\
\
seeh x. The figure indieates that the equation has an
'!2
infinite number of roots.
-0.5
-1
562
,cos x
~
)
"
\.
/
I
Exercises
12.4
= 4.7300, X2 = 7.8532, X3 = 10.9956,

and X4 = 14.1372. Thus the first four eigenvalues are )'1 = xdL = 4.7300/L, A2 = X2/L =
7.8532/ L, A3 = X3/ L = 10.9956/ L, and A4 = 14.1372/ L.
(e) Using a CAS we find the first four positive roots to be Xl
13. From (5) in the text we have
u(x, t) =
L (n7T'a
An cos Tt
00
n=l
Since Ut(x, O) = g(x) = O we have En
00
= ];
n7T'a )
n7T'
+ En sin =t:
sin TX'
= O and
1[
An 2 sin
(n7T'
tix a )
(n7T'
TX + T
t + sin TX
n7T'a )]
T
t
00
n7T' + at) + sin -L
n7T'](x - at) .
= -1 L
An [sin -(x
2 n=l
From
u(x, O)
= f(x) =
00
L An sin
n7T'
TX
n=l
we identify
f(x + at) =
n7T'
00
L An sin T(x
+ at)
n7T'
T(x
- at),
n=l
and
00
f(x - at) =
L Ansin
n=l
so that
1
u(x, t) = 2[f(x + at) + f(x - at)].

14. (a) We note that ~x = rx = 1, ~t = a, and rt = -a. Then
ou
ou o~ ou or
ox = o~ ox + or ox = ue + uT)
and
02U = ~ (u + u ) = oue o~ + oue or + oUT)o~ + oUT)or

ox2
ox e
T)
o~ ox
or ox
o~ ox
or ox
= ue~ + 2ueT)+ uT)T)'

Similarly
563
Exercses
12.4
Thus
(b) Integrating
cPu
we obtain
[)
[)~[)r= [)r U~
=O
J ~ U~dr J
O dn
U~
= f(~).
Integrating this result with respect to ~ we obtain
J ~~d~ = J f(~)d~
u = F(~) + G(r).
Since ~ = x
+ at
and r = x - at, we then have
u = F(~) + G(r) = F(x + at) + G(x - at).

Next, we have
U(x, t) = F(x + at) + G(x - at)

u(x, O) = F(x) + G(x) = f(x)
Ut(x, O) = aF'(x) - aG'(x) = g(x)
Integrating the last equation with respect to x gives
l1 g(s) ds +
F(x) - G(x) = -
Xo
Substituting G(x) = f(x) - F(x) we obtain

1
F(x)=-f(x)+2
where c
2a
= cl/2. Thus
G(x) = -1 f(x) - - 1
2
2a
g(s)ds+C
g(s) ds - c.
x
Xo
x
xo
(e) From the expressions for F and G,

1
F(x+at)=-f(x+at)+2
2a
G(x - at) = -1 f(x - at) - - 1

2
2a
564
q.
x+at
g(s)ds+c
t:
xo
xo
g(s) ds - c.
Exercses
12.4
Thus,
u(x, t)
= F(x + at) + G(x
Here we have used -
x-at
xo
- at)
1
= -[f(x
+ at) + f(x
2
= Xo
g(s) ds
x-at
1 x+at
- at)] + g(s) ds.

2a x-at
g(s) ds.
1 x+at
15. u(x, t) = -2[sin(x + at) + sin(x - at)] + ds

2a x-at
1
= -2[sin x cos at
16. u(x, t)
+ cos x sin at + sin x cos at -
1 x+at
= -2sin(x + at) + sin(x - at)] + -
2a x-at
cos x sin at]
+ -2a1 s
x+at
= sin x cos at
x-at
+t
cos s ds
. x cos at + -1 ['sm (x + at )'- sm (x - at )]'= sm x cos at + -l. cos x sm at

= sm
2a
17.
u ( x, t )
=O
+ -1
x+at .
2a x-at
sm 2s ds
= -1
[-
cos(2x + 2at) + cos(2x - 2at)]
2a
2a
= -4 [- cos 2x cos 2at + sin 2x sin 2at + cos 2x cos 2at + sin 2x sin 2at] = - sin 2x sin 2at
18.
-4
-2
-4
-2
19. (a)
-6
(b) Since g(x) = O, d'Alembert's solution with a = 1 is

1
u(x, t) = 2[f(x + t) + f(x - t)].
565
-4
-2
Exercises
12.4
Sample plots are shown below.
t=
~
-6
-4
-2
o
4
-6
-4
t= 0.4
~
-6
-4
-2
-4
-2
-6
-4
-6
-4
/\ '[ /\
/\
-4
-2
'[
-2
-2
~t"
6
-6
-4
-2
/\ '[ /\
t= 3.
t= 2.
-6
~t.O
~t.oa
-6
-2
t= 0.2
-6
-4
-2
'[
t= 4.
/\
4
/\
-6
-4
-2
a =
1, d'Alembert's solution is
11
u(x, t) = -
x+l
2 x=t
g(s) ds
where
566
g(s)={l,
[s] S 0.1
O,
t= 5.
(e) The single peaked wave disolves into two peaks moving outward.
20. (a) With
[s] > 0.1
/\
Exercises
12.4
Sarnple plots are shown below.
'1
-6
-4
-2
-2
ffi
-4
-2
-6
-4
-6
-6
-6
-4
-4
-2
'1
O
'1
-2
t= 0.2
t= O
~
2
-4
-6
-2
-2
t= 0.4
-4
-6
t= 0.8
1<
2
-6
-4
-6
-4
'"'\
-6
-4
-2
m
O
'1
-2
t= 0.6
'=\'
t= l.
t= 5.
'1
-2
(b) Sorne frames of the movie are shown in part (a), The string has a roughly rectangular shape
with the base on the x-axis increasing in length.
21. (a) and (b)
With the given parameters, the solution is
801.n7T
.
=""2":L ""2"sm-cosntsmnx.
1T n=l n
2
For neven, sin(mr/2) = O,so the first six nonzero terms correspond to n = 1, 3, 5,7,9,
u(x,t)
this case sin(n1T/2) = sin(2p - 1)/2 = (_l)p+l for p = 1,2,3,4,5,6,

and
8 00 (_l)p+l
u(x, t) = ""2"
)2 cos(2p - l)t sin(2p - l)x.
1T p=l 2p - 1
:L (
567
1l. In
Exercises 12.4
= 0.5,
Frames of the movie corresponding to t
1.5, and 2 are shown.
1,
o:r
o:~
-0.5
0.5
1.5
2.5
0.5
-0.5
~x
2.5
~I
u
O~~~x
1.5
-1
-1
-0.5
0.5
1.5
2.5
O . +~~7:5~=:C=Jlt~5==~==~2:~5~~?x
-0.5
-1
-1
Exercises 12.5
1. Using u = XY
and
_),2

X"+),2X=0,
X(O)
= O,
X(a)
= O,
and
Y(O) = O.
Then
n'rr
X = Cl sin-x
a
and
C2 sinh
n'rr
-y
a
for n = 1, 2, 3, ... so that
u=
n7f
L An sin -x
a
00
nat
sinh -y.
a
n=l
Imposing
n7fb
DO
u(x, b) = f(x) = L An sinhn=l
gives
n7fb
2
=a
a
Ansinh-
n7f
sin-x
a
a
loa f(x)sm-xdx
. n7f
a
so that
00
u(x,y) = LAnsin-x
n7f
n=l
568
n7f
sinh-y
Exercises 12.5
where
2
nnb
An = - csch a
loa f (x) sin -xnn

o
dx,
2. Using u = XY and _).2 as a separation constant leads to
+ ).2X
X"
= O,
X(O) = O,
X(a) = O,
and
y" - ).2y = O,
Y'(O) = 00
Then
nn
X = c sin-x
a
for n = 1, 2, 3,
nn
y = C2 cosh-y
and
so that
u=
nn
L An sin -x
nn
cosh -yo
a
00
n=l
= f(x) =
Ancosh-
Imposing
u(x,b)
nnb
00
n=l
nn
sin-x
a
gives
Ancosh-tmb
a
nt:
= -2laa f(x)sin-xdx
a
so that
u(x, y) =
L An SiD-xnna
nn
cosh-y
00
n=l
where
2
natb
An = -secha
a
3. Using u = XY and
_).2
la
O
nm
f(x)sm-xdx.
a

X"
+ ).2X =
O,
X(O) = O,
X(a) = O,
and
y" - ).2y
= O,
Y(b) = 00
569

Ecuaciones Diferenciales - D. Zill - 7° Ed. (Solucionario)

Uploaded by

Copyright:

Available Formats

Ecuaciones Diferenciales - D. Zill - 7° Ed. (Solucionario)

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ecuaciones Diferenciales - D. Zill - 7° Ed. (Solucionario)

Uploaded by

Copyright:

Available Formats

COMPLETE SOLUTIONS

Australia Canada Mexico Singapore Spain United Kingdom United States

Introduction to Differential Equations

First-Order Differential Equations

Modeling with First-Order Differential Equations

Higher-Order Differential Equations

Modeling with Higher-Order Differential Equations

Series Solutions of Linear Equations

The Laplace Transform

Systems of Linear First-Order Differential Equations

Numerical Solutions of Ordinary Differential Equations

Plane Autonomous Systems and Stability

Orthogonal Functions and Pourier Series

Partial Dfferential Equations and Boundary-Value

Boundary-Value Problems in Other Coordinate Systems

Integral Transform Method

Numerical Solutions of Partial Differential Equations

Introduction to Differential Equations

2. Third-order; nonlinear because of (dy/dx)4

3. The differential equation is first-order. Writing it in the form x(dy/dx)

nonlinear in y because of y2 However, writing it in the form (y2 it is linear in x.

4. The differential equation is first-order. Writing it in the form u(dv/du)

that it is linear in v. However, writing it in the form (v + uv - ue'U) (du/dv)

8. Second-order; nonlinear because of 1/ R2

10. Second-order: nonlinear because of j;2

= e=t? we obtain yl = _~e-x/2.

= ~ - ~e-20t we obtain dy/dt

= e3x cos 2x we obtain yl

that y" - 6y'

w.e obtain yl = -1 + sin x ln(sec x

15. Writing in(2X - 1) - In(X - 1)

Exponentiating both sides of the implicit solution we obtain

In2. Thus, the solution is defined on (-oo,ln2)

Using the quadratic formula to solve y2 - 2x2y - 1 = Ofor y, we get

V 4x4 + 4) /2 = x2 J.X4+l. Thus, two explicit

solutions are defined on (-00,00). The graph of YI (x) is solid and

the graph of Y2 is dashed.

+ C2x + C3Xlnx + 4x2 we obtain

= (-6q + 4Cl + Cl + cl)x-l

+ (-C3 + 2C3 - c2 - c3 + C2)X

21. (a) From rPl = x2 we obtain rP'l = 2x, so

xrP~ - 2rPl = x(2x) - 2x2 = O.

XrP2- 2rP2 = x(-2x)

22. The function y(x)

does not exist at x

23. (a) The domain of the function, found by solving x +

3;:: O, is [-3, (0).

= -5. Thus, y'(x)

Since y(x) is not differentiable at x

-3, y is a solution of the differential equation on (-3,00).

< 5t < K/2, so 5 tan 5t will be a solution on

24. (a) An interval on which tan 5t is continuous is -n)2

25. (a) From y

emt we obtain y'

Since emt > O for all t, m = -2. Thus y = e-2t is a solution.

= me'?' and y" = m2emt Then y" - 5y' + 6y = Oimplies

(b ) From y = emt we obtain y'

Since tm-1 > O for t > O, m

= mtm-1 and y"