You are on page 1of 24

November 10, 2008 19:17 ISM ET chapter 8 Sheet number 1 Page number 404 black

404
CHAPTER 8
Mathematical Modeling with Dierential
Equations
EXERCISE SET 8.1
1. y

= 9x
2
e
x
3
= 3x
2
y and y(0) = 3 by inspection.
2. y

= x
3
2 sinx, y(0) = 3 by inspection.
3. (a) rst order;
dy
dx
= c; (1 +x)
dy
dx
= (1 +x)c = y
(b) second order; y

= c
1
cos t c
2
sint, y

+y = c
1
sint c
2
cos t + (c
1
sint +c
2
cos t) = 0
4. (a) rst order; 2
dy
dx
+y = 2
_

c
2
e
x/2
+ 1
_
+ce
x/2
+x 3 = x 1
(b) second order; y

= c
1
e
t
c
2
e
t
, y

y = c
1
e
t
+c
2
e
t

_
c
1
e
t
+c
2
e
t
_
= 0
5. False. It is a rst-order equation, because it involves y and dy/dx, but not d
n
y/dx
n
for n > 1.
6. True. y = 1/2 is a solution.
7. True. As mentioned in the marginal note after equation (2), the general solution of an nth order
dierential equation usually involves n arbitrary constants.
8. False. Every solution of the rst order dierential equation y

x
= Ae
x+b
with b = 0.
9. (a) If y = e
2x
then y

= 2e
2x
and y

= 4e
2x
, so
y

+y

2y = 4e
2x
+ (2e
2x
) 2e
2x
= 0.
If y = e
x
then y

= e
x
and y

= e
x
, so y

+y

2y = e
x
+e
x
2e
x
= 0.
(b) If y = c
1
e
2x
+c
2
e
x
then y

= 2c
1
e
2x
+c
2
e
x
and y

= 4c
1
e
2x
+c
2
e
x
, so
y

+y

2y = (4c
1
e
2x
+c
2
e
x
) + (2c
1
e
2x
+c
2
e
x
) 2(c
1
e
2x
+c
2
e
x
) = 0.
10. (a) If y = e
2x
then y

= 2e
2x
and y

= 4e
2x
, so
y

6y = 4e
2x
(2e
2x
) 6e
2x
= 0.
If y = e
3x
then y

= 3e
3x
and y

= 9e
3x
, so y

6y = 9e
3x
3e
3x
6e
3x
= 0.
(b) If y = c
1
e
2x
+c
2
e
3x
then y

= 2c
1
e
2x
+ 3c
2
e
3x
and y

= 4c
1
e
2x
+ 9c
2
e
3x
, so
y

6y = (4c
1
e
2x
+ 9c
2
e
3x
) (2c
1
e
2x
+ 3c
2
e
3x
) 6(c
1
e
2x
+c
2
e
3x
) = 0.
11. (a) If y = e
2x
then y

= 2e
2x
and y

= 4e
2x
, so y

4y

+ 4y = 4e
2x
4(2e
2x
) + 4e
2x
= 0.
If y = xe
2x
then y

= (2x + 1)e
2x
and y

= (4x + 4)e
2x
, so
y

4y

+ 4y = (4x + 4)e
2x
4(2x + 1)e
2x
+ 4xe
2x
= 0.
(b) If y = c
1
e
2x
+ c
2
xe
2x
then y

= 2c
1
e
2x
+ c
2
(2x + 1)e
2x
and y

= 4c
1
e
2x
+ c
2
(4x + 4)e
2x
, so
y

4y

+4y = (4c
1
e
2x
+c
2
(4x +4)e
2x
) 4(2c
1
e
2x
+c
2
(2x +1)e
2x
) +4(c
1
e
2x
+c
2
xe
2x
) = 0.
12. (a) If y = e
4x
then y

= 4e
4x
and y

= 16e
4x
, so y

8y

+ 16y = 16e
4x
8(4e
4x
) + 16e
4x
= 0.
If y = xe
4x
then y

= (4x + 1)e
4x
and y

= (16x + 8)e
4x
, so
y

8y

+ 16y = (16x + 8)e

4x
8(4x + 1)e
4x
+ 16xe
4x
= 0.
(b) If y = c
1
e
4x
+c
2
xe
4x
then y

= 4c
1
e
4x
+c
2
(4x +1)e
4x
and y

= 16c
1
e
4x
+c
2
(16x +8)e
4x
, so
y

8y

+16y = (16c
1
e
4x
+c
2
(16x+8)e
4x
)8(4c
1
e
4x
+c
2
(4x+1)e
4x
)+16(c
1
e
4x
+c
2
xe
4x
) = 0.
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 2 Page number 405 black
Exercise Set 8.1 405
13. (a) If y = sin2x then y

= 2 cos 2x and y

= 4 sin2x, so y

+ 4y = 4 sin2x + 4 sin2x = 0.
If y = cos 2x then y

= 2 sin2x and y

= 4 cos 2x, so y

+ 4y = 4 cos 2x + 4 cos 2x = 0.
(b) If y = c
1
sin2x +c
2
cos 2x then y

= 2c
1
cos 2x 2c
2
sin2x and y

= 4c
1
sin2x 4c
2
cos 2x,
so y

+ 4y = (4c
1
sin2x 4c
2
cos 2x) + 4(c
1
sin2x +c
2
cos 2x) = 0.
14. (a) If y = e
2x
sin3x then y

= e
2x
(2 sin3x+3 cos 3x) and y

= e
2x
(5 sin3x12 cos 3x), so
y

+4y

+13y = e
2x
(5 sin3x12 cos 3x) +4e
2x
(2 sin3x+3 cos 3x) +13e
2x
sin3x = 0.
If y = e
2x
cos 3x then y

= e
2x
(3 sin3x 2 cos 3x) and y

= e
2x
(12 sin3x 5 cos 3x), so
y

+4y

+13y = e
2x
(12 sin3x 5 cos 3x) +4e
2x
(3 sin3x 2 cos 3x) +13e
2x
cos 3x = 0.
(b) If y = e
2x
(c
1
sin3x +c
2
cos 3x) then y

= e
2x
[(2c
1
+ 3c
2
) sin3x + (3c
1
2c
2
) cos 3x] and
y

= e
2x
[(5c
1
+ 12c
2
) sin3x (12c
1
+ 5c
2
) cos 3x], so
y

+ 4y

+ 13y = e
2x
[(5c
1
+ 12c
2
) sin3x (12c
1
+ 5c
2
) cos 3x]
+ 4e
2x
[(2c
1
+ 3c
2
) sin3x + (3c
1
2c
2
) cos 3x]
+ 13e
2x
(c
1
sin3x +c
2
cos 3x) = 0.
15. From Exercise 9, y = c
1
e
2x
+c
2
e
x
is a solution of the dierential equation, with
y

= 2c
1
e
2x
+c
2
e
x
. Setting y(0) = 1 and y

(0) = 4 gives c
1
+c
2
= 1 and 2c
1
+c
2
= 4.
So c
1
= 1, c
2
= 2, and y = e
2x
2e
x
.
16. From Exercise 10, y = c
1
e
2x
+c
2
e
3x
is a solution of the dierential equation, with
y

= 2c
1
e
2x
+ 3c
2
e
3x
. Setting y(0) = 1 and y

(0) = 8 gives c
1
+c
2
= 1 and 2c
1
+ 3c
2
= 8.
So c
1
= 1, c
2
= 2, and y = e
2x
+ 2e
3x
.
17. From Exercise 11, y = c
1
e
2x
+c
2
xe
2x
is a solution of the dierential equation, with
y

= 2c
1
e
2x
+c
2
(2x + 1)e
2x
. Setting y(0) = 2 and y

(0) = 2 gives c
1
= 2 and 2c
1
+c
2
= 2,
so c
2
= 2 and y = 2e
2x
2xe
2x
.
18. From Exercise 12, y = c
1
e
4x
+c
2
xe
4x
is a solution of the dierential equation, with
y

= 4c
1
e
4x
+c
2
(4x + 1)e
4x
. Setting y(0) = 1 and y

(0) = 1 gives c
1
= 1 and 4c
1
+c
2
= 1,
so c
2
= 3 and y = e
4x
3xe
4x
.
19. From Exercise 13, y = c
1
sin2x +c
2
cos 2x is a solution of the dierential equation, with
y

= 2c
1
cos 2x 2c
2
sin2x. Setting y(0) = 1 and y

(0) = 2 gives c
2
= 1 and 2c
1
= 2,
so c
1
= 1 and y = sin2x + cos 2x.
20. From Exercise 14, y = e
2x
(c
1
sin3x +c
2
cos 3x) is a solution of the dierential equation, with
y

= e
2x
[(2c
1
+ 3c
2
) sin3x + (3c
1
2c
2
) cos 3x]. Setting y(0) = 1 and y

(0) = 1 gives
c
2
= 1 and 3c
1
2c
2
= 1, so c
1
= 1 and y = e
2x
(sin3x + cos 3x).
21. y

= 24x so y =
_
(24x) dx = 2x
2
+2x+C. Setting y(0) = 3 gives C = 3, so y = 2x
2
+2x+3.
22. (y

= 6x so y

=
_
(6x) dx = 3x
2
+C. Setting y

(0) = 2 gives C = 2, so y

= 3x
2
+ 2 and
y =
_
(3x
2
+ 2) dx = x
3
+ 2x +D. Setting y(0) = 1 gives D = 1 so y = x
3
+ 2x + 1.
23. If the solution has an inverse function x(y) then, by equation (3) of Section 3.3,
dx
dy
=
1
dy/dx
= y
2
. So x =
_
y
2
dy = y
1
+C. When x = 1, y = 2, so C =
3
2
and x =
3
2
y
1
.
Solving for y gives y =
2
3 2x
. The solution is valid for x <
3
2
.
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 3 Page number 406 black
406 Chapter 8
24. If the solution has an inverse function x(y) then, by equation (3) of Section 3.3,
dx
dy
=
1
dy/dx
=
1
1 +y
2
. So x =
_
dy
1 +y
2
= tan
1
y +C. When x = 0, y = 0,
so C = 0, x = tan
1
y, and y = tanx. The solution is valid for /2 < x < /2.
25. By the product rule,
d
dx
(x
2
y) = x
2
y

+ 2xy = 0, so x
2
y = C and y = C/x
2
. Setting y(1) = 2
gives C = 2 so y = 2/x
2
. The solution is valid for x > 0.
26. By the product rule,
d
dx
(xy) = xy

+y = e
x
, so xy =
_
e
x
dx = e
x
+C. Setting y(1) = 1 +e
gives C = 1, so xy = e
x
+ 1 and y =
e
x
+ 1
x
. The solution is valid for x > 0.
27. (a)
dy
dt
= ky
2
, y(0) = y
0
, k > 0 (b)
dy
dt
= ky
2
, y(0) = y
0
, k > 0
28. (a) Either y is always zero, or y is positive and increases at a rate proportional to the square
root of y.
(b) Either y is always zero, or y is positive and decreases at a rate proportional to the cube of y,
or y is negative and increases at a rate proportional to the cube of y.
29. (a)
ds
dt
=
1
2
s (b)
d
2
s
dt
2
= 2
ds
dt
30. (a)
dv
dt
= 2v
2
(b)
d
2
s
dt
2
= 2
_
ds
dt
_
2
31. (a) Since k > 0 and y > 0, equation (3) gives
dy
dt
= ky > 0, so y is increasing.
(b)
d
2
y
dt
2
=
d
dt
(ky) = k
dy
dt
= k
2
y > 0, so y is concave upward.
32. (a) Both y = 0 and y = L satisfy equation (4).
(b) The rate of growth is
dy
dt
= k
_
1
y
L
_
y; we wish to nd the value of y which maximizes
this. Since
d
dy
_
k
_
1
y
L
_
y
_
=
k
L
(L 2y), which is positive for y < L/2 and negative for
y > L/2, the maximum growth rate occurs for y = L/2.
33. (a) Both y = 0 and y = L satisfy equation (6).
(b) The rate of growth is
dy
dt
= ky(L y); we wish to nd the value of y which maximizes this.
Since
d
dy
[ky(L y)] = k(L 2y), which is positive for y < L/2 and negative for y > L/2,
the maximum growth rate occurs for y = L/2.
34. If T is constant then
dT
dt
= 0, so equation (7) gives T = T
e
. Hence T = T
e
is the unique constant
solution of (7).
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 4 Page number 407 black
Exercise Set 8.2 407
35. If x = c
1
cos
_
_
k
m
t
_
+ c
2
sin
_
_
k
m
t
_
then
dx
dt
= c
2
_
k
m
cos
_
_
k
m
t
_
c
1
_
k
m
sin
_
_
k
m
t
_
and
d
2
x
dt
2
= c
1
k
m
cos
_
_
k
m
t
_
c
2
k
m
sin
_
_
k
m
t
_
=
k
m
x. So m
d
2
x
dt
2
= kx; thus x satises
the dierential equation for the vibrating string.
36. (a) From Exercise 35 we have x = c
1
cos
_
_
k
m
t
_
+c
2
sin
_
_
k
m
t
_
and
x

= c
2
_
k
m
cos
_
_
k
m
t
_
c
1
_
k
m
sin
_
_
k
m
t
_
. Setting x(0) = x
0
and x

(0) = 0 gives
c
1
= x
0
and c
2
_
k
m
= 0, so c
2
= 0 and x = x
0
cos
_
_
k
m
t
_
.
(b) From the discussion before Example 2 in Section 0.3, the amplitude is |x
0
|, the period is
2
_
k/m
= 2
_
m
k
, and the frequency is
_
k/m
2
. The amplitude is the maximum displacement
of the mass from its rest position. The period is the length of time the mass takes to move
back and forth once. The frequency tells how often the mass moves back and forth in one
unit of time.
EXERCISE SET 8.2
1.
1
y
dy =
1
x
dx, ln|y| = ln|x| +C
1
, ln

y
x

= C
1
,
y
x
= e
C1
= C, y = Cx
including C = 0 by inspection.
2.
dy
1 +y
2
= 2xdx, tan
1
y = x
2
+C, y = tan
_
x
2
+C
_
3.
dy
1 +y
=
x

1 +x
2
dx, ln|1 +y| =
_
1 +x
2
+C
1
, 1 +y = e

1+x
2
e
C1
= Ce

1+x
2
,
y = Ce

1+x
2
1, C = 0
4. y dy =
x
3
dx
1 +x
4
,
y
2
2
=
1
4
ln(1 +x
4
) +C
1
, 2y
2
= ln(1 +x
4
) +C, y =
_
[ln(1 +x
4
) +C]/2
5.
2(1 +y
2
)
y
dy = e
x
dx, 2 ln|y| +y
2
= e
x
+C; by inspection, y = 0 is also a solution.
6.
dy
y
= xdx, ln|y| = x
2
/2 +C
1
, y = e
C1
e
x
2
/2
= Ce
x
2
/2
, including C = 0 by inspection.
7. e
y
dy =
sinx
cos
2
x
dx = sec xtanxdx, e
y
= sec x +C, y = ln(sec x +C)
8.
dy
1 +y
2
= (1 +x) dx, tan
1
y = x +
x
2
2
+C, y = tan(x +x
2
/2 +C)
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 5 Page number 408 black
408 Chapter 8
9.
dy
y
2
y
=
dx
sinx
,
_ _

1
y
+
1
y 1
_
dy =
_
csc xdx, ln

y 1
y

= ln| csc x cot x| +C

1
,
y 1
y
= e
C1
(csc x cot x) = C(csc x cot x), y =
1
1 C(csc x cot x)
, C = 0;
by inspection, y = 0 is also a solution, as is y = 1.
10.
1
y
dy = cos xdx, ln|y| = sinx +C, y = C
1
e
sin x
11. (2y + cos y) dy = 3x
2
dx, y
2
+ siny = x
3
+C,
2
+ sin = C, C =
2
, y
2
+ siny = x
3
+
2
12.
dy
dx
= (x + 2)e
y
, e
y
dy = (x + 2)dx, e
y
=
1
2
x
2
+ 2x +C, 1 = C,
e
y
=
1
2
x
2
+ 2x 1, e
y
=
1
2
x
2
2x + 1, y = ln
_
1 2x
1
2
x
2
_
13. 2(y 1) dy = (2t + 1) dt, y
2
2y = t
2
+t +C, 1 + 2 = C, C = 3, y
2
2y = t
2
+t + 3
14.
dy
dx
cosh
2
x = y cosh2x, y
1
dy =
cosh2x
cosh
2
x
dx =
2 cosh
2
x 1
cosh
2
x
dx = (2 sech
2
x) dx,
lny = 2x tanhx +C, y = e
C
e
2xtanh x
, 3 = e
C
e
20tanh 0
= e
C
, y = 3e
2xtanh x
15. (a)
dy
y
=
dx
2x
, ln|y| =
1
2
ln|x| +C
1
,
|y| = C
2
|x|
1/2
, y
2
= Cx;
by inspection y = 0 is also a solu-
tion.
(b) 1
2
= C 2, C = 1/2, y
2
= x/2
2 2
2
2
x
y
x = 0.5y
2
x = 1.5y
2
x = y
2
x = 2y
2
x = 2.5y
2
y = 0 x = 3y
2
16. (a) y dy = xdx,
y
2
2
=
x
2
2
+C
1
, y =
_
C
2
x
2
(b) y =

25 x
2
3 3
3
3
x
y
y = 9 x
2
y = 6.25 x
2
y = 4 x
2
y = 1 x
2
y = 2.25 x
2
y = 0.25 x
2
17.
dy
y
=
xdx
x
2
+ 4
ln|y| =
1
2
ln(x
2
+ 4) +C
1
y =
C

x
2
+ 4
1.5
1
2 2
C = 1
C = 1
C = 0
C = 2
C = 2
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 6 Page number 409 black
Exercise Set 8.2 409
18. cos y dy = cos xdx
siny = sinx +C
y = 2n + sin
1
(sinx +C) or y = (2n + 1) sin
1
(sinx +C) for some integer n
For C = 0, the integral curves are lines of the form y =
2n+x and y = (2n+1)x for integers n. These divide
the xy-plane into squares rotated 45

from the axes. For

C = 0, 2 < C < 2, each integral curve stays within either
the top half or the bottom half of one of these squares. The
gure shows 5 such curves in the square x 2 < y < x,
x < y < 3 x. From top to bottom, their equations
are y = sin
1
(sinx + 0.5), y = sin
1
(sinx + 1.5),
y = sin
1
(sinx + 1.7), y = sin
1
(sinx + 1), and y =
sin
1
(sinx + 0.3).
x
y
!
!!
!
19. (1 y
2
) dy = x
2
dx,
y
y
3
3
=
x
3
3
+C
1
, x
3
+y
3
3y = C
2 2
2
2
x
y
20.
_
1
y
+y
_
dy = dx, ln|y| +
y
2
2
= x +C
1
,
ye
y
2
/2
= e
C1
e
x
= Ce
x
including C = 0.
5 5
3
3
x
y
21. True. The equation can be rewritten as
1
f(y)
dy
dx
= 1, which has the form (1).
22. False. The equation can be rewritten as
1
g(y)
dy
dx
=
1
h(x)
, which has the form (1).
23. True. After t minutes there will be 32 (1/2)
t
grams left; when t = 5 there will be 32 (1/2)
5
= 1
gram.
24. True. The population will quadruple in twice the doubling time.
25. Of the solutions y =
1
2x
2
C
, all pass through the point
_
0,
1
C
_
and thus never through (0, 0).
A solution of the initial value problem with y(0) = 0 is (by inspection) y = 0. The method of
Example 1 fails in this case because it starts with a division by y
2
= 0.
26. If y
0
= 0 then, proceeding as before, we get C = 2x
2

1
y
, C = 2x
2
0

1
y
0
, and
y =
1
2x
2
2x
2
0
+ 1/y
0
, which is dened for all x provided 2x
2
is never equal to 2x
2
0
1/y
0
; this
last condition will be satised if and only if 2x
2
0
1/y
0
< 0, or 0 < 2x
2
0
y
0
< 1. If y
0
= 0 then y = 0
is, by inspection, also a solution for all real x.
27.
dy
dx
= xe
y
, e
y
dy = xdx, e
y
=
x
2
2
+C, x = 2 when y = 0 so 1 = 2 +C, C = 1, e
y
= x
2
/2 1
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 7 Page number 410 black
410 Chapter 8
28.
dy
dx
=
3x
2
2y
, 2y dy = 3x
2
dx, y
2
= x
3
+C, 1 = 1 +C, C = 0,
y
2
= x
3
, y = x
3/2
passes through (1, 1).
2
0
0 1.6
29. (a)
dy
dt
= 0.02y, y
0
= 10,000 (b) y = 10,000e
t/50
(c) T =
1
0.02
ln2 34.657 h (d) 45,000 = 10,000e
t/50
,
t = 50 ln
45,000
10,000
75.20 h
30. k =
1
T
ln2 =
1
20
ln2
(a)
dy
dt
= ((ln2)/20)y, y(0) = 1 (b) y(t) = e
t(ln 2)/20
= 2
t/20
(c) y(120) = 2
6
= 64 (d) 1,000,000 = 2
t/20
,
t = 20
ln10
6
ln2
398.63 min
31. (a)
dy
dt
= ky, y(0) = 5.0 10
7
; 3.83 = T =
1
k
ln2, so k =
ln2
3.83
0.1810
(b) y = 5.0 10
7
e
0.181t
(c) y(30) = 5.0 10
7
e
0.1810(30)
219,000
(d) y(t) = (0.1)y
0
= y
0
e
kt
, kt = ln0.1, t =
ln0.1
0.1810
= 12.72 days
32. (a) k =
1
T
ln2 =
1
140
ln2 0.0050, so
dy
dt
= 0.0050y, y
0
= 10.
(b) y = 10e
0.0050t
(c) 10 weeks = 70 days so y = 10e
0.35
7 mg.
(d) 0.3y
0
= y
0
e
kt
, t =
ln0.3
0.0050
240.8 days
33. 100e
0.02t
= 10,000, e
0.02t
= 100, t =
1
0.02
ln100 230 days
34. y = 10,000e
kt
, but y = 12,000 when t = 5 so 10,000e
5k
= 12,000, k =
1
5
ln1.2. y = 20,000 when
2 = e
kt
, t =
ln2
k
= 5
ln2
ln1.2
19, in the year 2017.
35. y(t) = y
0
e
kt
= 10.0e
kt
, 3.5 = 10.0e
k(5)
, k =
1
5
ln
3.5
10.0
0.2100, T =
1
k
ln2 3.30 days
36. y = y
0
e
kt
, 0.7y
0
= y
0
e
5k
, k =
1
5
ln0.7 0.07
(a) T =
ln2
k
9.90 yr
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 8 Page number 411 black
Exercise Set 8.2 411
(b) y(t) y
0
e
0.07t
,
y
y
0
e
0.07t
, so e
0.07t
100 percent will remain.
38. (a) None; the half-life is independent of the initial amount.
(b) kT = ln2, so T is inversely proportional to k.
39. (a) T =
ln2
k
; and ln2 0.6931. If k is measured in percent, k

= 100k,
then T =
ln2
k

69.31
k

70
k

.
(b) 70 yr (c) 20 yr (d) 7%
40. Let y = y
0
e
kt
with y = y
1
when t = t
1
and y = 3y
1
when t = t
1
+ T; then y
0
e
kt1
= y
1
(i) and
y
0
e
k(t1+T)
= 3y
1
(ii). Divide (ii) by (i) to get e
kT
= 3, T =
1
k
ln3.
41. From (19), y(t) = y
0
e
0.000121t
. If 0.27 =
y(t)
y
0
= e
0.000121t
then t =
ln0.27
0.000121
10,820 yr, and
if 0.30 =
y(t)
y
0
then t =
ln0.30
0.000121
9950, or roughly between 9000 B.C. and 8000 B.C.
42. (a)
1
0
0 50000
(b) t = 1988 yields
y/y
0
= e
0.000121(1988)
79%.
43. (a) Let T
1
= 5730 40 = 5690, k
1
=
ln2
T
1
0.00012182; T
2
= 5730 +40 = 5770, k
2
0.00012013.
With y/y
0
= 0.92, 0.93, t
1
=
1
k
1
ln
y
y
0
= 684.5, 595.7; t
2
=
1
k
2
ln(y/y
0
) = 694.1, 604.1; in
1988 the shroud was at most 695 years old, which places its creation in or after the year 1293.
(b) Suppose T is the true half-life of carbon-14 and T
1
= T(1 +r/100) is the false half-life. Then
with k =
ln2
T
, k
1
=
ln2
T
1
we have the formulae y(t) = y
0
e
kt
, y
1
(t) = y
0
e
k1t
. At a certain
point in time a reading of the carbon-14 is taken resulting in a certain value y, which in the
case of the true formula is given by y = y(t) for some t, and in the case of the false formula
is given by y = y
1
(t
1
) for some t
1
.
If the true formula is used then the time t since the beginning is given by t =
1
k
ln
y
y
0
. If
the false formula is used we get a false value t
1
=
1
k
1
ln
y
y
0
; note that in both cases the
value y/y
0
is the same. Thus t
1
/t = k/k
1
= T
1
/T = 1 + r/100, so the percentage error in
the time to be measured is the same as the percentage error in the half-life.
44. If y = y
0
e
kt
and y = y
1
= y
0
e
kt1
then y
1
/y
0
= e
kt1
, k =
ln(y
1
/y
0
)
t
1
; if y = y
0
e
kt
and
y = y
1
= y
0
e
kt1
then y
1
/y
0
= e
kt1
, k =
ln(y
1
/y
0
)
t
1
.
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 9 Page number 412 black
412 Chapter 8
45. (a) If y = y
0
e
kt
, then y
1
= y
0
e
kt1
, y
2
= y
0
e
kt2
, divide: y
2
/y
1
= e
k(t2t1)
, k =
1
t
2
t
1
ln(y
2
/y
1
),
T =
ln2
k
=
(t
2
t
1
) ln2
ln(y
2
/y
1
)
. If y = y
0
e
kt
, then y
1
= y
0
e
kt1
, y
2
= y
0
e
kt2
,
y
2
/y
1
= e
k(t2t1)
, k =
1
t
2
t
1
ln(y
2
/y
1
), T =
ln2
k
=
(t
2
t
1
) ln2
ln(y
2
/y
1
)
.
In either case, T is positive, so T =

(t
2
t
1
) ln2
ln(y
2
/y
1
)

.
(b) In part (a) assume t
2
= t
1
+ 1 and y
2
= 1.25y
1
. Then T =
ln2
ln1.25
3.1 h.
46. (a) In t years the interest will be compounded nt times at an interest rate of r/n each time. The
value at the end of 1 interval is P + (r/n)P = P(1 + r/n), at the end of 2 intervals it is
P(1 +r/n) +(r/n)P(1 +r/n) = P(1 +r/n)
2
, and continuing in this fashion the value at the
end of nt intervals is P(1 +r/n)
nt
.
(b) Let x = r/n, then n = r/x and
lim
n+
P(1 +r/n)
nt
= lim
x0
+
P(1 +x)
rt/x
= lim
x0
+
P[(1 +x)
1/x
]
rt
= Pe
rt
.
(c) The rate of increase is dA/dt = rPe
rt
= rA.
47. (a) A = 1000e
(0.08)(5)
= 1000e
0.4
\$1, 491.82
(b) Pe
(0.08)(10)
= 10, 000, Pe
0.8
= 10, 000, P = 10, 000e
0.8
\$4, 493.29
(c) From (11), with k = r = 0.08, T = (ln2)/0.08 8.7 years.
48. Let r be the annual interest rate when compounded continuously and r
1
the eective annual
interest rate. Then an amount P invested at the beginning of the year is worth Pe
r
= P(1 + r
1
)
at the end of the year, and r
1
= e
r
1.
49. (a) Given
dy
dt
= k
_
1
y
L
_
y, separation of variables yields
_
1
y
+
1
L y
_
dy = k dt so that
ln
y
L y
= lny ln(L y) = kt + C. The initial condition gives C = ln
y
0
L y
0
so
ln
y
L y
= kt + ln
y
0
L y
0
,
y
L y
= e
kt
y
0
L y
0
, and y(t) =
y
0
L
y
0
+ (L y
0
)e
kt
.
(b) If y
0
> 0 then y
0
+ (L y
0
)e
kt
= Le
kt
+ y
0
(1 e
kt
) > 0 for all t 0, so y(t) exists for
all such t. Since lim
t+
e
kt
= 0, lim
t+
y(t) =
y
0
L
y
0
+ (L y
0
) 0
= L.
(Note that for y
0
< 0 the solution blows up at t =
1
k
ln
y
0
L y
0
, so lim
t+
y(t) is undened.)
50. The dierential equation for the spread of disease can be rewritten as
dy
dt
= kL
_
1
y
L
_
y, which is
the logistic equation with k replaced by kL. Making this replacement in the solution from Exercise
49 gives y(t) =
y
0
L
y
0
+ (L y
0
)e
kLt
.
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 10 Page number 413 black
Exercise Set 8.2 413
51. (a) k = L = 1, y
0
= 2
2
0
0 2
(b) k = L = y
0
= 1
2
0
0 2
(c) k = y
0
= 1, L = 2
2
0
0 5
(d) k = 1, y
0
= 0.5, L = 10
0
0
10
10
52. y
0
400, L 1000; since the curve y =
400,000
400 + 600e
kt
passes through the point (200, 600),
600 =
400,000
400 + 600e
200k
, 600e
200k
=
800
3
, k =
1
200
ln2.25 0.00405.
53. y
0
2, L 8; since the curve y =
2 8
2 + 6e
kt
passes through the point (2, 4), 4 =
16
2 + 6e
2k
,
6e
2k
= 2, k =
1
2
ln3 0.5493.
54. This is the logistic equation (Equation (4) of Section 8.1) with
k = 0.98, L = 5, and y
0
= 1. From Exercise 49, the solution is
given by y(t) =
5
1 + 4e
0.98t
.
5
0
0 12
55. (a) y
0
= 5 (b) L = 12 (c) k = 1
(d) L/2 = 6 =
60
5 + 7e
t
, 5 + 7e
t
= 10, t = ln(5/7) 0.3365
(e)
dy
dt
=
1
12
y(12 y), y(0) = 5
56. (a) y
0
= 1 (b) L = 1000 (c) k = 0.9
(d) 750 =
1000
1 + 999e
0.9t
, 3(1 + 999e
0.9t
) = 4, t =
1
0.9
ln(3 999) 8.8949
(e)
dy
dt
=
0.9
1000
y(1000 y), y(0) = 1
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 11 Page number 414 black
414 Chapter 8
57. (a) Assume that y(t) students have had the u t days after the break. If the disease spreads
as predicted by equation (6) of Section 8.1 and if nobody is immune, then Exercise 50 gives
y(t) =
y
0
L
y
0
+ (L y
0
)e
kLt
, where y
0
= 20 and L = 1000. So y(t) =
20000
20 + 980e
1000kt
=
1000
1 + 49e
1000kt
. Using y(5) = 35 we nd that k =
ln(193/343)
5000
. Hence y =
1000
1 + 49(193/343)
t/5
.
(b)
t
y(t)
0
20
1
22
2
25
3
28
4
31
5
35
6
39
7
44
8
49
9
54
10
61
11
67
12
75
13
83
14
93
(c)
3 6 9 12
25
50
75
100
t
y
58. If T
0
< T
a
then
dT
dt
= k(T
a
T) where k > 0. If T
0
> T
a
then
dT
dt
= k(T T
a
) where k > 0;
both cases yield T(t) = T
a
+ (T
0
T
a
)e
kt
with k > 0.
59. (a) From Exercise 58 with T
0
= 95 and T
a
= 21, we have T = 21 + 74e
kt
for some k > 0.
(b) 85 = T(1) = 21 + 74e
k
, k = ln
64
74
= ln
32
37
, T = 21 + 74e
t ln(32/37)
= 21 + 74
_
32
37
_
t
,
T = 51 when
30
74
=
_
32
37
_
t
, t =
ln(30/74)
ln(32/37)
6.22 min
60.
dT
dt
= k(70 T), T(0) = 40; ln(70 T) = kt +C, 70 T = e
kt
e
C
, T = 40 when t = 0, so
30 = e
C
, T = 70 30e
kt
; 52 = T(1) = 70 30e
k
, k = ln
70 52
30
= ln
5
3
0.5,
T 70 30e
0.5t
61. (a)
dv
dt
=
ck
m
0
kt
g, v = c ln(m
0
kt) gt +C; v = 0 when t = 0 so 0 = c lnm
0
+C,
C = c lnm
0
, v = c lnm
0
c ln(m
0
kt) gt = c ln
m
0
m
0
kt
gt.
(b) m
0
kt = 0.2m
0
when t = 100 so
v = 2500 ln
m
0
0.2m
0
9.8(100) = 2500 ln5 980 3044 m/s.
62. (a) By the chain rule,
dv
dt
=
dv
dx
dx
dt
=
dv
dx
v so m
dv
dt
= mv
dv
dx
.
(b)
mv
kv
2
+mg
dv = dx,
m
2k
ln(kv
2
+mg) = x +C; v = v
0
when x = 0 so
C =
m
2k
ln(kv
2
0
+mg),
m
2k
ln(kv
2
+mg) = x +
m
2k
ln(kv
2
0
+mg), x =
m
2k
ln
kv
2
0
+mg
kv
2
+mg
.
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 12 Page number 415 black
Exercise Set 8.2 415
(c) x = x
max
when v = 0 so
x
max
=
m
2k
ln
kv
2
0
+mg
mg
=
3.56 10
3
2(7.3 10
6
)
ln
(7.3 10
6
)(988)
2
+ (3.56 10
3
)(9.8)
(3.56 10
3
)(9.8)
1298 m
63. (a) A(h) = (1)
2
= ,
dh
dt
= 0.025

h,

h
dh = 0.025dt, 2

h = 0.025t + C; h = 4 when
t = 0, so 4 = C, 2

h = 0.025t + 4,

h = 2
0.025
2
t, h (2 0.003979 t)
2
.
(b) h = 0 when t 2/0.003979 502.6 s 8.4 min.
64. (a) A(h) = 6
_
2
_
4 (h 2)
2
_
= 12
_
4h h
2
,
12
_
4h h
2
dh
dt
= 0.025

h, 12

4 hdh = 0.025dt,
8(4 h)
3/2
= 0.025t +C; h = 4 when t = 0 so C = 0,
(4 h)
3/2
= (0.025/8)t, 4 h = (0.025/8)
2/3
t
2/3
,
h 4 0.021375t
2/3
ft
h 2
2
24 (h 2)
2
h
(b) h = 0 when t =
8
0.025
(4 0)
3/2
= 2560 s 42.7 min
65.
dv
dt
=
1
32
v
2
,
1
v
2
dv =
1
32
dt,
1
v
=
1
32
t +C; v = 128 when t = 0 so
1
128
= C,

1
v
=
1
32
t
1
128
, v =
128
4t + 1
cm/s. But v =
dx
dt
so
dx
dt
=
128
4t + 1
, x = 32 ln(4t + 1) +C
1
;
x = 0 when t = 0 so C
1
= 0, x = 32 ln(4t + 1) cm.
66.
dv
dt
= 0.02

v,
1

v
dv = 0.02dt, 2

v = 0.02t +C; v = 9 when t = 0 so 6 = C,

2

v = 0.02t + 6, v = (3 0.01t)
2
cm/s. But v =
dx
dt
so
dx
dt
= (3 0.01t)
2
,
x =
100
3
(3 0.01t)
3
+C
1
; x = 0 when t = 0 so C
1
= 900, x = 900
100
3
(3 0.01t)
3
cm.
67. Suppose that H(y) = G(x) + C. Then
dH
dy
dy
dx
= G

(x). But
dH
dy
= h(y) and
dG
dx
= g(x), hence
y(x) is a solution of (1).
68. Suppose that y(x) satises y(x
0
) = y
0
and
_
y(x)
y
0
h(r) dr =
_
x
x
0
g(s) ds for all x in some interval
containing x
0
. Dierentiate with respect to x; by the Fundamental Theorem of Calculus (Theorem
5.6.3) and the chain rule, we have h(y(x))
dy
dx
= g(x).
69. If h(y) = 0 then (1) implies that g(x) = 0, so h(y) dy = 0 = g(x) dx. Otherwise the slope of L
is
dy
dx
=
g(x)
h(y)
. Since (x
1
, y
1
) and (x
2
, y
2
) lie on L, we have
y
2
y
1
x
2
x
1
=
g(x)
h(y)
. So h(y)(y
2
y
1
) =
g(x)(x
2
x
1
); i.e. h(y) dy = g(x) dx.
70. It is true that the method may not give a formula for y as a function of x. But sometimes such a
formula (in terms of familiar functions) does not exist. In such cases the method at least gives a
relationship between x and y, from which we can nd as good an approximation to y as we want.
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 13 Page number 416 black
416 Chapter 8
71. Suppose that y = f(x) satises h(y)
dy
dx
= g(x). Integrating both sides of this with respect to x
gives
_
h(y)
dy
dx
dx =
_
g(x) dx, so
_
h(f(x))f

(x) dx =
_
g(x) dx. By equation (2) of Section
5.3 with f replaced by h, g replaced by f, and F replaced by
_
h(y) dy, the left side equals
F(f(x)) = F(y). Thus
_
h(y) dy =
_
g(x) dx.
EXERCISE SET 8.3
1.
x
y
-2 -1 1 2
-2
-1
1
2
2.
x
y
1 2 3 4
1
2
3
4
3.
5
1
2
x
y
y(0) = 1
y(0) = 2
y(0) = 1
4.
dy
dx
+y = 1, = e
_
dx
= e
x
,
d
dx
[ye
x
] = e
x
, ye
x
= e
x
+C, y = 1 +Ce
x
(a) 1 = 1 +C, C = 2, y = 1 2e
x
(b) 1 = 1 +C, C = 0, y = 1
(c) 2 = 1 +C, C = 1, y = 1 +e
x
5. lim
x+
y = 1
6. (a) IV, since the slope is positive for x > 0 and negative for x < 0.
(b) VI, since the slope is positive for y > 0 and negative for y < 0.
(c) V, since the slope is always positive.
(d) II, since the slope changes sign when crossing the lines y = 1.
(e) I, since the slope can be positive or negative in each quadrant but is not periodic.
(f ) III, since the slope is periodic in both x and y.
7. y
0
= 1, y
n+1
= y
n
+
1
2
y
1/3
n
n
x
n
y
n
0
0
1
1
0.5
1.50
2
1
3 4
2
5
2.07
1.5
2.71 3.41
2.5
4.16
6
3
4.96
7
3.5
5.81
8
4
6.71
2 4 1 3
9
x
y
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 14 Page number 417 black
Exercise Set 8.3 417
8. y
0
= 1, y
n+1
= y
n
+ (x
n
y
2
n
)/4
n
x
n
y
n
0
0
1
1
0.25
0.75
2 3 4 5
0.50
0.67
0.75
0.68
1.00
0.75
1.25
0.86
6
1.50
0.99
7
1.75
1.12
8
2.00
1.24
0.5 1 1.5 2
0.5
1
1.5
2
x
y
9. y
0
= 1, y
n+1
= y
n
+
1
2
cos y
n
n
t
n
y
n
0
0
1
1
0.5
1.27
2
1
3 4
2
1.42
1.5
1.49 1.53
3
3
t
y
10. y
0
= 0, y
n+1
= y
n
+e
yn
/10
n
t
n
y
n
0
0
0
1
0.1
0.10
2 3 4 5
0.2
0.19
0.3
0.27
0.4
0.35
0.5
0.42
6
0.6
0.49
7
0.7
0.55
8
0.8
0.60
9
0.9
0.66
10
1.0
0.71
1
1
t
y
11. h = 1/5, y
0
= 1, y
n+1
= y
n
+
1
5
sin(n/5)
n
t
n
y
n
0
0
0
1
0.2
0.12
0.4
0.31
2 3 4
0.6
0.50
0.8
0.62
1.0
0.62
5
12. False. This is only true if the slope at (x, y) does not depend on y.
13. True.
dy
dx
= e
xy
> 0 for all x and y. So, for any integral curve, y is an increasing function of x.
14. True.
d
2
y
dx
2
=
d
dx
(e
y
) = e
y
dy
dx
= e
2y
> 0 for all y.
15. True. Every cubic polynomial has at least one real root. If p(y
0
) = 0 then y = y
0
is an integral
curve that is a horizontal line.
16. (a) By inspection,
dy
dx
= e
x
2
and y(0) = 0.
(b) y
n+1
= y
n
+e
x
2
n
/20 = y
n
+e
(n/20)
2
/20 and y
20
= 0.7625. From a CAS, y(1) = 0.7468.
17. (b) y dy = xdx, y
2
/2 = x
2
/2 +C
1
, x
2
+y
2
= C; if y(0) = 1 then C = 1 so y(1/2) =

3/2.
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 15 Page number 418 black
418 Chapter 8
18. (a) Yes.
(b) y = 0 is an integral curve of y

= 2xy which is a horizontal line.

19. (b) The equation y

= 1 y is separable:
dy
1 y
= dx, so
_
dy
1 y
=
_
dx, ln|1 y| = x +C.
Substituting x = 0 and y = 1 gives C = ln2, so x = ln2ln|1y| = ln

2
1 y

. Since the
integral curve stays below the line y = 1, we can drop the absolute value signs: x = ln
2
1 y
and y = 1 2e
x
. Solving y = 0 shows that the x-intercept is ln2 0.693.
20. (a) y
0
= 1, y
n+1
= y
n
+ (

y
n
/2)x
x = 0.2 : y
n+1
= y
n
+

y
n
/10; y
5
1.5489
x = 0.1 : y
n+1
= y
n
+

y
n
/20; y
10
1.5556
x = 0.05 : y
n+1
= y
n
+

y
n
/40; y
20
1.5590
(b)
dy

y
=
1
2
dx, 2

y = x/2 +C, 2 = C,

y = x/4 + 1, y = (x/4 + 1)
2
,
y(1) = 25/16 = 1.5625
21. (a) The slope eld does not vary with x, hence along a given parallel line all values are equal
since they only depend on the height y.
(b) As in part (a), the slope eld does not vary with x; it is independent of x.
(c) From G(y) x = C we obtain
d
dx
(G(y) x) =
1
f(y)
dy
dx
1 =
d
dx
C = 0, i.e.
dy
dx
= f(y).
22. (a) Separate variables:
dy

y
= dx, 2

y = x + C, y = (x/2 + C
1
)
2
is a parabola that opens up,
and is therefore concave up.
(b) A curve is concave up if its derivative is increasing, and y

y is increasing.
23. (a) By implicit dierentiation, y
3
+ 3xy
2
dy
dx
2xy x
2
dy
dx
= 0,
dy
dx
=
2xy y
3
3xy
2
x
2
.
(b) If y(x) is an integral curve of the slope eld in part (a), then
d
dx
{x[y(x)]
3
x
2
y(x)} = [y(x)]
3
+3xy(x)
2
y

(x) 2xy(x) x
2
y

(x) = 0, so the integral curve

must be of the form x[y(x)]
3
x
2
y(x) = C.
(c) x[y(x)]
3
x
2
y(x) = 2
24. (a) By implicit dierentiation, e
y
+xe
y
dy
dx
+e
x
dy
dx
+ye
x
= 0,
dy
dx
=
e
y
+ye
x
xe
y
+e
x
(b) If y(x) is an integral curve of the slope eld in part (a), then
d
dx
{xe
y(x)
+ y(x)e
x
} = e
y(x)
+ xy

(x)e
y(x)
+ y

(x)e
x
+ y(x)e
x
= 0 from part (a). Thus
xe
y(x)
+y(x)e
x
= C.
(c) Any integral curve y(x) of the slope eld above satises xe
y(x)
+ y(x)e
x
= C; if it passes
through (1, 1) then e +e = C, so xe
y(x)
+y(x)e
x
= 2e denes the curve implicitly.
25. (a) For any n, y
n
is the value of the discrete approximation at the right endpoint, that, is an
approximation of y(1). By increasing the number of subdivisions of the interval [0, 1] one
might expect more accuracy, and hence in the limit y(1).
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 16 Page number 419 black
Exercise Set 8.4 419
(b) For a xed value of n we have, for k = 1, 2, . . . , n, y
k
= y
k1
+y
k1
1
n
=
n + 1
n
y
k1
. In partic-
ular y
n
=
n + 1
n
y
n1
=
_
n + 1
n
_
2
y
n2
= . . . =
_
n + 1
n
_
n
y
0
=
_
n + 1
n
_
n
. Consequently,
lim
n+
y
n
= lim
n+
_
n + 1
n
_
n
= e, which is the (correct) value y = e
x

x=1
.
26. Eulers Method is repeated application of local linear approximation, each step dependent on the
previous step.
27. Visual inspection of the slope eld may show where the integral curves are increasing, decreasing,
concave up, or concave down. It may also help to identify asymptotes for the integral curves.
For example, in Exercise 3 we see that y = 1 is an integral curve that is an asymptote of all other
integral curves. Those curves with y < 1 are increasing and concave down; those with y > 1 are
decreasing and concave up.
EXERCISE SET 8.4
1. = e
_
4 dx
= e
4x
, e
4x
y =
_
e
x
dx = e
x
+C, y = e
3x
+Ce
4x
2. = e
2
_
x dx
= e
x
2
,
d
dx
_
ye
x
2
_
= xe
x
2
, ye
x
2
=
1
2
e
x
2
+C, y =
1
2
+Ce
x
2
3. = e
_
dx
= e
x
, e
x
y =
_
e
x
cos(e
x
)dx = sin(e
x
) +C, y = e
x
sin(e
x
) +Ce
x
4.
dy
dx
+ 2y =
1
2
, = e
_
2dx
= e
2x
, e
2x
y =
_
1
2
e
2x
dx =
1
4
e
2x
+C, y =
1
4
+Ce
2x
5.
dy
dx
+
x
x
2
+ 1
y = 0, = e
_
(x/(x
2
+1))dx
= e
1
2
ln(x
2
+1)
=
_
x
2
+ 1,
d
dx
_
y
_
x
2
+ 1
_
= 0, y
_
x
2
+ 1 = C, y =
C

x
2
+ 1
6.
dy
dx
+y =
1
1 e
x
, = e
_
dx
= e
x
, e
x
y =
_
e
x
1 e
x
dx = ln(1e
x
)+C, y = e
x
ln(1e
x
)+Ce
x
7.
dy
dx
+
1
x
y = 1, = e
_
(1/x)dx
= e
ln x
= x,
d
dx
[xy] = x, xy =
1
2
x
2
+C, y =
x
2
+
C
x
,
2 = y(1) =
1
2
+C, C =
3
2
, y =
x
2
+
3
2x
8. Divide by x to put the dierential equation in the form (3):
dy
dx
x
1
y = x. We have p(x) =
1
x
and
q(x) = x, so
_
p(x) dx = ln|x|. So we may take e
ln |x|
= |x
1
| as an integrating factor. Since
integrating factors are only determined up to a constant factor, we may drop the absolute value
signs and simply take = x
1
. We have
d
dx
(x
1
y) = x
1
dy
dx
x
2
y = x
1
_
dy
dx
x
1
y
_
= 1, so
x
1
y = x +C and y = x
2
+Cx. Since y(1) = 1, C = 2 and y = x
2
2x.
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 17 Page number 420 black
420 Chapter 8
9. = e
2
_
x dx
= e
x
2
, e
x
2
y =
_
2xe
x
2
dx = e
x
2
+C,
y = 1 +Ce
x
2
, 3 = 1 +C, C = 4, y = 1 + 4e
x
2
10. = e
_
dt
= e
t
, e
t
y =
_
2e
t
dt = 2e
t
+C, y = 2 +Ce
t
, 1 = 2 +C, C = 1, y = 2 e
t
11. False. If y
1
and y
2
both satisfy
dy
dx
+p(x)y = q(x) then
d
dx
(y
1
+y
2
)+p(x)(y
1
+y
2
) = 2q(x). Unless
q(x) = 0 for all x, y
1
+y
2
is not a solution of the original dierential equation.
12. True. If y = C is a solution, then dy/dx = 0, so p(x)C = q(x).
13. True. The concentration in the tank will approach the concentration in the solution owing into
the tank.
14. False. Although equation (18) implies that v

=
mg
c
, where mg is the weight of the object, the
model does not specify how c depends on the object.
15.
-2 2
10
10
x
y
y(0) = 1
y(1) = 0
y(1) = 1
16.
dy
dx
2y = x, = e
2
_
dx
= e
2x
,
d
dx
_
ye
2x

= xe
2x
,
ye
2x
=
1
4
(2x + 1)e
2x
+C, y =
1
4
(2x + 1) +Ce
2x
(a) 1 = 3/4 +Ce
2
, C = 1/(4e
2
), y =
1
4
(2x + 1) +
1
4
e
2x2
(b) 1 = 1/4 +C, C = 5/4, y =
1
4
(2x + 1)
5
4
e
2x
(c) 0 = 1/4 +Ce
2
, C = e
2
/4, y =
1
4
(2x + 1) +
1
4
e
2x+2
17. It appears that lim
x+
y =
_
+, if y
0
1/4;
, if y
0
< 1/4.
To conrm this, we solve the equation using the method of integrating factors:
dy
dx
2y = x, = e
2
_
dx
= e
2x
,
d
dx
_
ye
2x

= xe
2x
, ye
2x
=
1
4
(2x + 1)e
2x
+ C,
y =
1
4
(2x + 1) +Ce
2x
. Setting y(0) = y
0
gives C = y
0

1
4
, so y =
1
4
(2x + 1) +
_
y
0

1
4
_
e
2x
.
If y
0
= 1/4, then y =
1
4
(2x + 1) + as x +. Otherwise, we rewrite the solution as
y = e
2x
_
y
0

1
4
+
2x + 1
4e
2x
_
; since lim
x+
2x + 1
4e
2x
= 0, we obtain the conjectured limit.
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 18 Page number 421 black
Exercise Set 8.4 421
18. (a) y
0
= 1, y
n+1
= y
n
+ (2y
n
x
n
)(0.1) = (12y
n
x
n
)/10
n 0 1 2 3 4 5
x
n
0 0.1 0.2 0.3 0.4 0.5
y
n
1 1.2 1.43 1.696 2.0052 2.36624
(b) Less. The integral curve appears to be concave up, so each y
n
is an underestimate of the
actual value of y(x
n
).
(c) From Exercise 16, we have y =
1
4
(2x + 1) + Ce
2x
for some constant C. Since y(0) = 1 we
nd C =
3
4
, so y =
3e
2x
+ 2x + 1
4
and y
_
1
2
_
=
3e + 2
4
2.53871.
19. (a) y
0
= 1,
y
n+1
= y
n
+ (x
n
+y
n
)(0.2) = (x
n
+ 6y
n
)/5
n
x
n
y
n
0
0
1
1
0.2
1.20
2 3 4 5
0.4
1.48
0.6
1.86
0.8
2.35
1.0
2.98
(b) y

y = x, = e
x
,
d
dx
_
ye
x

= xe
x
,
ye
x
= (x + 1)e
x
+C, 1 = 1 +C,
C = 2, y = (x + 1) + 2e
x
x
n
y(x
n
)
abs. error
perc. error
0
1
0
0 3
0.2
1.24
0.04
6 9 11 13
0.4
1.58
0.10
0.6
2.04
0.19
0.8
2.65
0.30
1.0
3.44
0.46
(c)
0.2 0.4 0.6 0.8 1
3
x
y
20. h = 0.1, y
n+1
= (x
n
+ 11y
n
)/10
n
x
n
y
n
0
0
1.00
1
0.1
1.10
2 3 4 5
0.2
1.22
0.3
1.36
0.4
1.53
0.5
1.72
6
0.6
1.94
7
0.7
2.20
8
0.8
2.49
9
0.9
2.82
10
1.0
3.19
With x = 0.2, Eulers method gives y(1) 2.98; with x = 0.1, it gives y(1) 3.19. The true
value is y(1) = 2e 2 3.44; so the absolute errors are approximately 0.46 and 0.25 respectively.
21.
dy
dt
= rate in rate out, where y is the amount of salt at time t,
dy
dt
= (4)(2)
_
y
50
_
(2) = 8
1
25
y, so
dy
dt
+
1
25
y = 8 and y(0) = 25.
= e
_
(1/25)dt
= e
t/25
, e
t/25
y =
_
8e
t/25
dt = 200e
t/25
+C,
y = 200 +Ce
t/25
, 25 = 200 +C, C = 175,
(a) y = 200 175e
t/25
oz (b) when t = 25, y = 200 175e
1
136 oz
22.
dy
dt
= (5)(20)
y
200
(20) = 100
1
10
y, so
dy
dt
+
1
10
y = 100 and y(0) = 0.
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 19 Page number 422 black
422 Chapter 8
= e
_
(1/10)dt
= e
t/10
, e
t/10
y =
_
100e
t/10
dt = 1000e
t/10
+C,
y = 1000 +Ce
t/10
, 0 = 1000 +C, C = 1000;
(a) y = 1000 1000e
t/10
lb (b) when t = 30, y = 1000 1000e
3
950 lb
23. The volume V of the (polluted) water is V (t) = 500 +(20 10)t = 500 +10t; if y(t) is the number
of pounds of particulate matter in the water, then y(0) = 50 and
dy
dt
= 010
y
V
=
y
50 +t
. Using
the method of integrating factors, we have
dy
dt
+
1
50 +t
y = 0; = e
_
dt
50+t
= 50+t;
d
dt
[(50+t)y] = 0,
(50 + t)y = C, 2500 = 50y(0) = C, y(t) = 2500/(50 + t). (The dierential equation may also be
solved by separation of variables.)
The tank reaches the point of overowing when V = 500 + 10t = 1000, t = 50 min, so
y = 2500/(50 + 50) = 25 lb.
24. The volume of the lake (in gallons) is V = 264r
2
h = 264(15)
2
3 = 178,200 gals. Let y(t)
denote the number of pounds of mercury salts at time t; then
dy
dt
= 0 10
3
y
V
=
y
178.2
lb/h
and y
0
= 10
5
V = 1.782 lb;
dy
y
=
dt
178.2
, lny =
t
178.2
+ C
1
, y = Ce
t/(178.2)
, and
C = y(0) = y
0
= 1.782, y = 1.782e
t/(178.2)
lb of mercury salts.
t 1 2 3 4 5 6 7 8 9 10 11 12
y(t) 5.588 5.578 5.568 5.558 5.549 5.539 5.529 5.519 5.509 5.499 5.489 5.480
We assumed that the mercury is always distributed uniformly throughout the lake, and doesnt
settle to the bottom.
25. (a)
dv
dt
+
c
m
v = g, = e
(c/m)
_
dt
= e
ct/m
,
d
dt
_
ve
ct/m
_
= ge
ct/m
, ve
ct/m
=
gm
c
e
ct/m
+C,
v =
gm
c
+Ce
ct/m
, but v
0
= v(0) =
gm
c
+C, C = v
0
+
gm
c
, v =
gm
c
+
_
v
0
+
gm
c
_
e
ct/m
(b) Replace
mg
c
with v

and ct/m with gt/v

in (16).
(c) From part (b), s(t) = C v

t (v
0
+v

)
v

g
e
gt/v
; s
0
= s(0) = C (v
0
+v

)
v

g
,
C = s
0
+ (v
0
+v

)
v

g
, s(t) = s
0
v

t +
v

g
(v
0
+v

)
_
1 e
gt/v
_
26. (a) Let t denote time elapsed in seconds after the moment of the drop. From Exercise 25(b),
while the parachute is closed
v(t) = e
gt/v
(v
0
+v

) v

= e
32t/120
(0 + 120) 120 = 120
_
e
4t/15
1
_
and thus
v(25) = 120
_
e
20/3
1
_
119.85, so the skydiver is falling at a speed of 119.85 ft/s
when the parachute opens. From Exercise 25(c), s(t) = s
0
120t +
120
32
120
_
1 e
4t/15
_
,
s(25) = 10000 120 25 + 450
_
1 e
20/3
_
7449.43 ft.
(b) If t denotes time elapsed after the parachute opens, then, by Exercise 25(c),
s(t) 7449.4324t+
24
32
(119.85+24)(1e
32t/24
). A calculating utility nds that s(t) = 0
for t 307.4 s, so the skydiver is in the air for about 25 + 307.4 = 332.4 s.
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 20 Page number 423 black
Review Exercises, Chapter 8 423
27.
dI
dt
+
R
L
I =
V (t)
L
, = e
(R/L)
_
dt
= e
Rt/L
,
d
dt
(e
Rt/L
I) =
V (t)
L
e
Rt/L
,
Ie
Rt/L
= I(0) +
1
L
_
t
0
V (u)e
Ru/L
du, I(t) = I(0)e
Rt/L
+
1
L
e
Rt/L
_
t
0
V (u)e
Ru/L
du.
(a) I(t) =
1
5
e
2t
_
t
0
20e
2u
du = 2e
2t
e
2u
_
t
0
= 2
_
1 e
2t
_
A.
(b) lim
t+
I(t) = 2 A
28. From Exercise 27 and Endpaper Table #42,
I(t) = 15e
2t
+
1
3
e
2t
_
t
0
3e
2u
sinudu = 15e
2t
+e
2t
e
2u
5
(2 sinu cos u)
_
t
0
= 15e
2t
+
1
5
(2 sint cos t) +
1
5
e
2t
.
29. (a) Let y =
1

[H(x) + C] where = e
P(x)
,
dP
dx
= p(x),
d
dx
H(x) = q, and C is an arbitrary
constant. Then
dy
dx
+p(x)y =
1

(x)

2
[H(x) +C] +p(x)y = q
p

[H(x) +C] +p(x)y = q

(b) Given the initial value problem, let C = (x
0
)y
0
H(x
0
). Then y =
1

[H(x)+C] is a solution
of the initial value problem with y(x
0
) = y
0
. This shows that the initial value problem has
a solution.
To show uniqueness, suppose u(x) also satises (3) together with u(x
0
) = y
0
. Following the
arguments in the text we arrive at u(x) =
1

[H(x) + C] for some constant C. The initial

condition requires C = (x
0
)y
0
H(x
0
), and thus u(x) is identical with y(x).
30. (a) y = x and y = x are both solutions of the given initial value problem.
(b)
_
y dy =
_
xdx, y
2
= x
2
+ C; but y(0) = 0, so C = 0. Thus y
2
= x
2
, which is
impossible.
REVIEW EXERCISES, CHAPTER 8
2. (a) yes (b) yes (c) no (d) yes
3.
dy
1 +y
2
= x
2
dx, tan
1
y =
1
3
x
3
+C, y = tan
_
1
3
x
3
+C
_
4.
1
tany
dy =
3
sec x
dx,
cos y
siny
dy = 3 cos xdx, ln| siny| = 3 sinx +C
1
,
siny = e
3 sin x+C1
= e
C1
e
3 sin x
= Ce
3 sin x
, C = 0,
y = sin
1
_
Ce
3 sin x
_
, as is y = 0 by inspection
5.
_
1
y
+y
_
dy = e
x
dx, ln|y| +y
2
/2 = e
x
+C; by inspection, y = 0 is also a solution
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 21 Page number 424 black
424 Chapter 8
6.
dy
y
2
+ 1
= dx, tan
1
y = x +C, /4 = C; y = tan(x +/4)
7.
_
1
y
5
+
1
y
_
dy =
dx
x
,
1
4
y
4
+ ln|y| = ln|x| +C;
1
4
= C, y
4
+ 4 ln(x/y) = 1
8.
dy
y
2
= 4 sec
2
2xdx,
1
y
= 2 tan2x +C, 1 = 2 tan
_
2

8
_
+C = 2 tan

4
+C = 2 +C, C = 3,
y =
1
3 2 tan2x
9.
dy
y
2
= 2xdx,
1
y
= x
2
+C, 1 = C, y = 1/(x
2
+ 1)
1 1
1
x
y
10. 2y dy = dx, y
2
= x + C; if y(0) = 1 then C = 1, y
2
= x + 1, y =

x + 1; if y(0) = 1 then
C = 1, y
2
= x + 1, y =

x + 1.
1 1
1
1
x
y
1 1
1
1
x
y
11.
x
y
1 2 3 4
1
2
3
4
12.
dy
y
=
1
8
xdx, ln|y| =
1
16
x
2
+C, y = C
1
e
x
2
/16
13. y
0
= 1, y
n+1
= y
n
+

y
n
/2
n 0 1 2 3 4 5 6 7 8
x
n
0 0.5 1 1.5 2 2.5 3 3.5 4
y
n
1 1.50 2.11 2.84 3.68 4.64 5.72 6.91 8.23
2 4 1 3
9
x
y
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 22 Page number 425 black
Review Exercises, Chapter 8 425
14. y
0
= 1, y
n+1
= y
n
+
1
2
siny
n
n 0 1 2 3 4
t
n
0 0.5 1 1.5 2
y
n
1 1.42 1.92 2.39 2.73
3
3
t
y
15. h = 1/5, y
0
= 1, y
n+1
= y
n
+
1
5
cos(2n/5)
n 0 1 2 3 4 5
t
n
0 0.2 0.4 0.6 0.8 1.0
y
n
1.00 1.20 1.26 1.10 0.94 1.00
16. y
n+1
= y
n
+ 0.1(1 + 5t
n
y
n
), y
0
= 5
n 0 1 2 3 4 5 6 7 8 9 10
t
n
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2
y
n
5.00 5.10 5.24 5.42 5.62 5.86 6.13 6.41 6.72 7.05 7.39
17. From formula (19) of Section 8.2, y(t) = y
0
e
0.000121t
, so 0.785y
0
= y
0
e
0.000121t
, t = ln0.785/0.000121
2000.6 yr
18. (a)
d
dt
y(t) = 0.01y, y(0) = 5000
(b) y(t) = 5000e
0.01t
(c) 2 = e
0.01t
, t = 100 ln2 69.31 h
(d) 30,000 = 5000e
0.01t
, t = 100 ln6 179.18 h
19. = e
_
3 dx
= e
3x
, e
3x
y =
_
e
x
dx = e
x
+C, y = e
2x
+Ce
3x
20.
dy
dx
+y =
1
1 +e
x
, = e
_
dx
= e
x
, e
x
y =
_
e
x
1 +e
x
dx = ln(1 +e
x
) +C, y = e
x
ln(1 +e
x
) +Ce
x
21. = e

_
x dx
= e
x
2
/2
, e
x
2
/2
y =
_
xe
x
2
/2
dx = e
x
2
/2
+C,
y = 1 +Ce
x
2
/2
, 3 = 1 +C, C = 4, y = 1 + 4e
x
2
/2
22.
dy
dx
+
2
x
y = 4x, = e
_
(2/x)dx
= x
2
,
d
dx
_
yx
2

= 4x
3
, yx
2
= x
4
+C, y = x
2
+Cx
2
,
2 = y(1) = 1 +C, C = 1, y = x
2
+ 1/x
2
23. By inspection, the left side of the equation is
d
dx
(y coshx), so
d
dx
(y coshx) = cosh
2
x =
1
2
(1 + cosh2x) and y coshx =
1
2
x +
1
4
sinh2x +C =
1
2
(x + sinhxcoshx) +C.
When x = 0, y = 2 so 2 = C, and y = 2 sech x +
1
2
(x sech x + sinhx).
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 23 Page number 426 black
426 Chapter 8
24. (a) = e

_
dx
= e
x
,
d
dx
_
ye
x

= xe
x
sin3x,
ye
x
=
_
xe
x
sin3xdx =
_

3
10
x
3
50
_
e
x
cos 3x +
_

1
10
x +
2
25
_
e
x
sin3x +C;
1 = y(0) =
3
50
+C, C =
53
50
, y =
_

3
10
x
3
50
_
cos 3x +
_

1
10
x +
2
25
_
sin3x +
53
50
e
x
(c)
10 2
2
4
x
y
25. (a) linear (b) both (c) separable (d) neither
26.
dy
dx
4xy = x
(a) IF: e
_
(4x)dx
= e
2x
2
,
d
dx
[ye
2x
2
] = xe
2x
2
, ye
2x
2
=
_
xe
2x
2
dx =
1
4
e
2x
2
+ C,
y =
1
4
+Ce
2x
2
(b)
dy
dx
= 4xy +x,
dy
4y + 1
= xdx,
1
4
ln(4y + 1) =
1
2
x
2
+C, ln(4y + 1) = 2x
2
+C
1
,
4y + 1 = C
2
e
2x
2
, y =
1
4
(C
2
e
2x
2
1) = C
3
e
2x
2

1
4
, same as in part (a)
27. Assume the tank contains y(t) oz of salt at time t. Then y
0
= 0 and for 0 < t < 15,
dy
dt
= 5 10
y
1000
10 = (50 y/100) oz/min, with solution y = 5000 +Ce
t/100
. But y(0) = 0 so
C = 5000, y = 5000(1 e
t/100
) for 0 t 15, and y(15) = 5000(1 e
0.15
). For 15 < t < 30,
dy
dt
= 0
y
1000
5, y = C
1
e
t/200
, C
1
e
0.075
= y(15) = 5000(1e
0.15
), C
1
= 5000(e
0.075
e
0.075
),
y = 5000(e
0.075
e
0.075
)e
t/200
, y(30) = 5000(e
0.075
e
0.075
)e
0.15
646.14 oz.
28. (a) Assume the air contains y(t) ft
3
of carbon monoxide at time t. Then y
0
= 0 and for
t > 0,
dy
dt
= 0.04(0.1)
y
1200
(0.1) = 1/250 y/12000,
d
dt
_
ye
t/12000
_
=
1
250
e
t/12000
,
ye
t/12000
= 48e
t/12000
+C, y(0) = 0, C = 48; y = 48(1 e
t/12000
). Thus the percentage
of carbon monoxide is P =
y
1200
100 = 4(1 e
t/12000
) percent.
(b) 0.012 = 4(1 e
t/12000
), t = 36.05 min
MAKING CONNECTIONS, CHAPTER 8
1. (a) u(x) = q p y(x) so
du
dx
= p
dy
dx
= p(q py(x)) = (p)u(x).
If p < 0 then p > 0 so u(x) grows exponentially.
If p > 0 then p < 0 so u(x) decays exponentially.
November 10, 2008 19:17 ISM ET chapter 8 Sheet number 24 Page number 427 black
Making Connections, Chapter 8 427
(b) From (a), u(x) = 4 2y(x) satises
du
dx
= 2u(x), so equation (14) of Section 8.2 gives
u(x) = u
0
e
2x
for some constant u
0
. Since u(0) = 4 2y(0) = 6, we have u(x) = 6e
2x
;
hence y(x) = 2 3e
2x
.
2. (a)
du
dx
=
d
dx
(ax+b y(x) +c) = a+b
dy
dx
= a+b f(ax+by +c) = a+b f(u), so
1
a +b f(u)
du
dx
= 1.
(b) From (a) with a = b = 1, c = 0, f(t) = 1/t, we have
1
1 + 1/u
du
dx
= 1, where u = x + y. So
u
u + 1
du = dx,
_
u
u + 1
du =
_
dx, u ln|u +1| = x +C, x +y ln|x +y +1| = x +C, and
y ln|x +y + 1| = C.
3. (a)
du
dx
=
d
dx
_
y
x
_
=
x
dy
dx
y
x
2
=
xf
_
y
x
_
y
x
2
. Since y = ux,
du
dx
=
xf(u) ux
x
2
=
f(u) u
x
and
1
f(u) u
du
dx
=
1
x
.
(b)
dy
dx
=
x y
x +y
=
1 y/x
1 +y/x
has the form given in (a), with f(t) =
1 t
1 +t
. So
1
1 u
1 +u
u
du
dx
=
1
x
,
1 +u
1 2u u
2
du =
dx
x
,
_
1 +u
1 2u u
2
du =
_
dx
x
,
1
2
ln|1 2u u
2
| = ln|x| + C
1
, and
|1 2u u
2
| = e
2C1
x
2
. Hence 1 2u u
2
= Cx
2
where C is either e
2C1
or e
2C1
.
Substituting u =
y
x
gives 1
2y
x

y
2
x
2
= Cx
2
, and x
2
2xy y
2
= C.
4. (a)
du
dx
= (1 n)y
n
dy
dx
= (1 n)y
n
[q(x)y
n
p(x)y] = (1 n)q(x) (1 n)p(x)y
1n
=
(1 n)q(x) (1 n)p(x)u. Hence
du
dx
+ (1 n)p(x)u = (1 n)q(x).
(b)
dy
dx

1
x
y = 2y
2
, so this has the form given in (a) with p(x) = 1/x, q(x) = 2, and n = 2.
So u = y
1
satises
du
dx
+
1
x
u = 2. An integrating factor is given by = e
_
dx/x
= e
ln x
= x.
So
d
dx
(xu) = x
du
dx
+u = 2x, xu = x
2
+C, u = x+Cx
1
, and y = u
1
=
1
x +Cx
1
=
x
x
2
+C
.
Since y(1) =
1
2
, C = 1 and y =
x
x
2
+ 1
.