Section 9.

3
C09S03.001: Given:
dy
dx
= 2x

y. Then
y
1/2
dy = 2x dx;
2y
1/2
= x
2
+ C;
y
1/2
=
x
2
+ C
2
;
y(x) =
_
x
2
+ C
2
_
2
.
C09S03.002: Given:
dy
dx
= 2xy
2
. Then
y
2
dy = 2x dx;
y
1
= C x
2
;
y(x) =
1
C x
2
.
C09S03.003: Given:
dy
dx
= x
2
y
2
. Then
y
2
dy = x
2
dx;
y
1
= C
1
3
x
3
;
y(x) =
1
C
1
3
x
3
=
3
K x
3
where K = 3C is a constant.
C09S03.004: Given:
dy
dx
= (xy)
3/2
. Then
y
3/2
dy = x
3/2
dx;
2y
1/2
=
2
5
x
5/2
+ C
1
;
y
1/2
=
1
5
x
5/2
+ C
2
(C
2
is a constant);
y
1/2
=
1
1
5
x
5/2
C
2
=
5
x
5/2
+ C
3
(C
3
is a constant);
y(x) =
_
5
x
5/2
+ C
3
_
2
.
C09S03.005: Given:
dy
dx
= 2x(y 1)
1/2
. Then
1
(y 1)
1/2
dy = 2x dx;
2(y 1)
1/2
= x
2
+ C;
(y 1)
1/2
=
x
2
+ C
2
;
y 1 =
_
x
2
+ C
2
_
2
;
y(x) = 1 +
_
x
2
+ C
2
_
2
.
C09S03.006: Given:
dy
dx
= 4x
3
(y 4)
2
. Then
(y 4)
2
dy = 4x
3
dx; (y 4)
1
= C x
4
;
y 4 =
1
C x
4
; y(x) = 4 +
1
C x
4
.
C09S03.007: Given:
dy
dx
=
1 +

x
1 +

y
. Then
(1 +

y ) dy =
_
1 +

x
_
dx; y +
2
3
y
3/2
= x +
2
3
x
3/2
+ C.
It is possible to solve explicitly for y(x). To see the explicit form, enter the Mathematica command
DSolve[ y

[x] == (1 + Sqrt[x])/(1 + Sqrt[y[x]]), y[x], x ]

and be prepared for about 32 lines of output.
C09S03.008: Given
dy
dx
=
x + x
3
y + y
3
. Then
(y + y
3
) dy = (x + x
3
) dx;
1
2
y
2
+
1
4
y
4
=
1
2
x
2
+
1
4
x
4
+ C.
It is possible to solve explicitly for y(x), but probably better not to do so, as there are ambiguities of sign
involving the square roots. The result is, however, not too complicated:
y(x) =
_
1
_
(x
2
+ 1)
2
+ C .
C09S03.009: Given:
dy
dx
=
x
2
+ 1
x
2
(3y
2
+ 1)
. Then
(3y
2
+ 1) dy =
_
1 +
1
x
2
_
dx; y
3
+ y = x
1
x
+ C.
It is possible to solve explicitly for y(x). Enter the Mathematica command
DSolve[ y

[x] == (x

2 + 1)/(x

2(3(y[x])

2 + 1)), y[x], x ]
to see the result.
2
C09S03.010: Given:
dy
dx
=
(x
3
1)y
3
x
2
(2y
3
3)
. Then
2y
3
3
y
3
dy =
x
3
1
x
2
dx; (2 3y
3
) dy = (x x
2
) dx;
2y +
3
2
y
2
=
1
2
x
2
+
1
x
+ C; 4xy
3
+ 3x = x
3
y
2
+ 2y
2
+ 2Cxy
2
.
It is possible to solve explicitly for y(x) using any of various computer algebra programs, but the results are
rather complicated.
C09S03.011: Given:
dy
dx
= y
2
, y(0) = 1. Then
y
2
dy = dx; y
1
= C x; y(x) =
1
C x
.
Then the initial condition yields
1 = y(0) =
1
C
, and thus y(x) =
1
1 x
.
C09S03.012: Given:
dy
dx
= y
1/2
, y(0) = 4. Then
y
1/2
dy = dx; 2y
1/2
= x + C; y
1/2
=
x + C
2
;
y(x) =
_
x + C
2
_
2
.
The last equation and the initial condition tell us only that C
2
= 16, but the third equation tells us that

4 = [y(0)]
1/2
=
0 + C
2
, so that C = 4.
Therefore y(x) =
_
x + 4
2
_
2
.
C09S03.013: Given:
dy
dx
=
1
4y
3
, y(0) = 1. Then
4y
3
dy = dx; y
4
= x + C; 1
4
= [y(0)]
4
= 0 + C;
C = 1; [y(x)]
4
= x + 1; y(x) = (x + 1)
1/4
.
We take the positive root in the last step because y(0) > 0.
C09S03.014: Given:
dy
dx
=
1
x
2
y
, y(1) = 2. Then
2y dy = 2x
2
dx; y
2
= C
2
x
, 2
2
= [y(1)]
2
= C 2;
C = 6; y
2
= 6
2
x
; y(x) =
_
6 2x
1
.
3
We took the positive square root in the last step because y(1) > 0.
C09S03.015: Given:
dy
dx
=
_
xy
3
, y(0) = 4. Then
y
3/2
dy = x
1/2
dx; 3y
3/2
dy = 3x
1/2
dx; 6y
1/2
= C
1
2x
3/2
;
y
1/2
= C
2

1
3
x
3/2
; y
1/2
=
1
C
2

1
3
x
3/2
; y
1/2
=
3
C x
3/2
;
2 = [y(0)]
1/2
=
3
C
; C =
3
2
; y(x) =
9
_
3
2
x
3/2
_
2
;
y(x) =
36
_
3 2x
3/2
_
2
.
C09S03.016: Given:
dy
dx
=
x
y
, y(3) = 5. Then
y dy = x dx; 2y dy = 2x dx; y
2
= x
2
+ C;
25 = [y(3)]
2
= 9 + C; C = 16; y
2
= x
2
+ 16;
y(x) =
_
x
2
+ 16.
We took the positive root in the last step because y(3) > 0.
C09S03.017: Given:
dy
dx
=
x
y
, y(12) = 5. Then
y dy = x dx; 2y dy = 2x dx; y
2
= C x
2
;
25 = [y(12)]
2
= C 144; C = 169; y
2
= 169 x
2
;
y(x) =
_
169 x
2
.
We took the negative root in the last step because y(12) < 0.
C09S03.018: Given: y
2
dy
dx
= x
2
+ 2x + 1, y(1) = 2. Thus
3y
2
dy = 3(x + 1)
2
dx; y
3
= (x + 1)
3
+ C; 2
3
= [y(1)]
3
= (1 + 1)
3
+ C;
C = 0; y
3
= (x + 1)
3
; y(x) = x + 1.
C09S03.019: Given:
dy
dx
= 3x
2
y
2
y
2
, y(0) = 1. Then
y
2
dy = (3x
2
1) dx; y
1
= x x
3
+ C; y =
1
x x
3
+ C
;
1 = y(0) =
1
C
; y(x) =
1
x x
3
+ 1
.
4
C09S03.020: Given:
dy
dx
= 2xy
3
(2x
2
+ 1), y(1) = 1. Then
y
3
dy = (4x
3
+ 2x) dx;
1
2
y
2
= x
4
+ x
2
+ C
1
; y
2
= C 2x
2
2x
4
;
y
2
=
1
C 2x
2
2x
4
; 1 = [y(1)]
2
=
1
C 4
; C = 5;
y
2
=
1
5 2x
2
2x
4
; y(x) =
1

5 2x
2
2x
4
.
We took the positive root in the last step because y(1) > 0.
C09S03.021: Given:
dy
dx
= y + 1, y(0) = 1.
_
dy
y + 1
=
_
1 dx; ln(y + 1) = x + C;
y + 1 = e
x+C
= Ae
x
; y(x) = Ae
x
1;
1 = y(0) = A1; A = 2.
Answer: y(x) = 2e
x
1.
C09S03.022: Given:
dy
dx
= 2 y, y(0) = 3.
_
dy
y 2
=
_
(1) dx; ln(y 2) = C x; y 2 = e
Cx
= Ae
x
;
y(x) = 2 + Ae
x
; 3 = y(0) = 2 + A; y(x) = 2 + e
x
.
C09S03.023: Given:
dy
dx
= 2y 3, y(0) = 2.
_
2 dy
2y 3
=
_
2 dx; ln(2y 3) = 2x + C; 2y 3 = e
2x+C
= Ae
2x
;
y(x) =
Ae
2x
+ 3
2
; 2 = y(0) =
A + 3
2
; y(x) =
e
2x
+ 3
2
.
C09S03.024: Given:
dy
dx
=
1
4

y
16
=
4 y
16
, y(0) = 20.
_
dy
y 4
=
_

1
16
dx; ln(y 4) = C
x
16
; y 4 = e
C(x/16)
= Ae
x/16
;
y(x) = 4 + Ae
x/16
; 20 = y(0) = 4 + A; y(x) = 4 + 16e
x/16
.
C09S03.025: Given:
dx
dt
= 2(x 1), x(0) = 0.
_
dx
x 1
=
_
2 dt; ln(x 1) = 2t + C; x 1 = e
2t+C
= Ae
2t
;
x(t) = 1 + Ae
2t
; 0 = x(0) = 1 + A; x(t) = 1 e
2t
.
5
C09S03.026: Given:
dx
dt
= 2 3x, x(0) = 4.
_
3 dx
3x 2
=
_
(3) dt; ln(3x 2) = C 3t; 3x 2 = e
C3t
= Ae
3t
;
x(t) =
1
3
_
2 + Ae
3t
_
; 4 = x(0) =
1
3
(2 + A) ; x(t) =
1
3
_
2 + 10e
3t
_
.
C09S03.027: Given:
dx
dt
= 5(x + 2), x(0) = 25.
_
dx
x + 2
=
_
5 dt; ln(x + 2) = 5t + C; x + 2 = e
5t+C
= Ae
5t
;
x(t) = Ae
5t
2; 25 = x(0) = A2; x(t) = 27e
5t
2.
C09S03.028: Given:
dx
dt
= 3 4x, x(0) = 5.
_
4 dx
4x + 3
=
_
(4) dt; ln(4x + 3) = C 4t; 4x + 3 = e
C4t
= Ae
4t
;
x(t) =
1
4
_
Ae
4t
3
_
; 5 = x(0) =
1
4
(A3) ; x(t) =
1
4
_
17e
4t
+ 3
_
.
C09S03.029: Given:
dv
dt
= 10(10 v), v(0) = 0.
_
dv
v 10
=
_
(10)dt; ln(v 10) = C 10t; v 10 = e
C10t
= Ae
10t
;
v(t) = 10 + Ae
10t
; 0 = v(0) = 10 + A; v(t) = 10
_
1 e
10t
_
.
C09S03.030: Given:
dv
dt
= 5(10 v), v(0) = 10.
_
dv
v 10
=
_
5 dt; ln(v 10) = 5t + C; v 10 = Ae
5t
;
v(t) = 10 + Ae
5t
; 10 = v(0) = 10 + A; v(t) = 10 20e
5t
.
C09S03.031: Let the population at time t (in years) be Q(t); t = 0 corresponds to the year 1990. From
the data given in the problem, we know that
dQ
dt
= (0.04)Q + 50000; Q(0) = 1,500,000.
25
dQ
dt
= Q + 1,250,000;
1
Q + 1,250,000

dQ
dt
=
1
25
;
ln(Q + 1,250,000) = (0.04)t + C;
Q(t) + 1,250,000 = Ke
t/25
.
6
Now from the condition Q(0) = 1,500,000 it follows that 1,500,000 + 1,250,000 = K, so
Q(t) + 1,250,000 = 2,750,000e
t/25
.
In the year 2010, we have Q(20) = 1,250,000 + 2,750,000e
0.8
4,870,238, so the population in the year
2010 will be approximately 4.87 million people.
C09S03.032: Let h(t) denote the temperature (in

F) of the cake at time t (in minutes). By Newtons law
of cooling, we have h

(t) = k [h(t) A] where k is a constant and A = 70 is the ambient temperature. Thus

_
dh
h A
=
_
k dt; ln(h A) = C
1
+ kt; h A = Ce
kt
.
Thus h(t) = 70 + Ce
kt
. The initial condition h(0) = 210 implies that 210 = 70 + C, so C = 140. Thus
h(t) = 70 + 140e
kt
. We are also given h(30) = 140, so that
140 = 70 + 140e
30k
; 70 = 140e
30k
;
1
2
= e
30k
; k =
ln2
30
.
Now h(t) = 100 when 100 = 70 + 140e
kt
, so that
e
kt
=
3
14
, and so t =
1
k
ln
3
14
=
30 ln(14/3)
ln2
66.67177264.
Answer: The cake will be at 100

F about one hour and seven minutes after it is removed from the oven.
C09S03.033: One eective way to derive a dierential equation is to estimate the changes that take place
in the dependent variable over a short interval [t, t+t] where t is the independent variable. In this problem
t is measured in months, and the change in the principal balance from time t to time t + t is
P(t + t) P(t) rP(t) t c t.
The reason is that the interest added to the principal balance is rP(t) t and the monthly payment decreases
the principal by c t. Thus
P(t + t) P(t)
t
rP(t) c. (1)
The errors in this approximation will approach zero as t 0, and when we evaluate the limits of both
sides of the approximation in (1) we obtain
dP
dt
= rP c, P(0) = P
0
.
C09S03.034: First we solve the initial value problem derived in Problem 33.
_
r dP
rP c
=
_
r dt; ln(rP c) = C + rt; rP c = Ae
rt
;
P(t) =
1
r
_
c + Ae
rt
_
; P
0
= P(0) =
1
r
(c + A) ; P(t) =
c + (rP
0
c) e
rt
r
.
7
In Problem 34, the loan is to be paid o in 36 months, and thus P(36) = 0. We use this information to solve
for the monthly payment c:
c + (rP
0
c)e
36r
r
= 0;
c
_
1 e
36r
_
+ rP
0
e
36r
= 0;
c =
rP
0
e
36r
e
36r
1
.
In part (a), we substitute P
0
= 3600 and r = 0.01 (the 12% annual rate converted to the monthly rate of
1%, then converted to a decimal) and nd that c = $119.08. In part (b), we substitute r = 0.015 instead
and nd that c = $129.42.
C09S03.035: Let P = P(t) denote the number of people who have heard the rumor after t days. Then
dP
dt
= k(100000 P);
_
dP
P 100000
=
_
(k) dt;
ln(P 100000) = C kt; P 100000 = Ae
kt
;
P(t) = 100000 Ae
kt
.
We assume that P(0) = 0, so that A = 100000 and thus P(t) = 100000
_
1 e
kt
_
. Next, P(7) = 10000, so
100000
_
1 e
7k
_
= 10000; 1 e
7k
=
1
10
;
e
7k
=
9
10
; e
7k
=
10
9
; k =
1
7
ln
10
9
.
Half the population of the city will have heard the rumor when P(T) = 50000, so that
100000
_
1 e
kT
_
= 50000; 1 e
kT
=
1
2
; e
kT
= 2; T =
ln2
k
46.05169435.
Therefore half the population will have heard the rumor 46 days after it begins.
C09S03.036: Here we have P
0
= 280 (million),
=
17
1000
= 0.017, =
7
1000
= 0.007,
k = = 0.01, I = 1.5, and I/k = 150. By Eq. (15),
P(t) = 280e
(0.01)t
+ 150
_
e
(0.01)t
1
_
.
In the year 2020 we therefore have P(20) 375.2 (million). The increase in the population is 95.2 million;
natural growth accounts for
280e
(0.01)(20)
280 62.0
million and immigration accounts for the remaining 33.2 million.
8
C09S03.037: Assuming that you begin with nothing, the value of the account P(t) (in thousands of
dollars, at time t in years) satises P(0) = P
0
= 0. If I is your yearly investment, which we assume is made
continuously (well approximated by equal monthly deposits), then
dP
dt
=
1
10
P(t) + I; 10
dP
dt
= P + 10I;
10
P + 10I
dP = dt;
1
P + 10I
dP =
1
10
dt;
ln(P + 10I) =
1
10
t + C; P + 10I = Aexp
_
t
10
_
;
P(t) = 10I + Aexp
_
t
10
_
. 0 = P(0) = A10I :
A = 10I. P(t) = 10I
_
1 + exp
_
t
10
__
.
Thus the account will grow to a value of 5,000,000 in 30 years when
5000 = 10I
_
1 + exp
_
30
10
__
: I =
500
e
3
1
26.19785.
Hence your monthly investment should be I/12 2.18315; that is, approximately $2183.15 per month. It
is of interest to note that your total investment will be $654,946.21 and that the accrued interest will be
$4,345,053.79. You should now recompute the real answer to this problem under the assumption that your
interest income will be subject to federal, state, and local taxes. Dont forget to compute the total tax you
expect to pay over the 30 years of investing.
C09S03.038: Set up your coordinate system with time t in hours and with t = 0 corresponding to the
time of death. Let T(t) denote the temperature of the body (in

F) at time t 0. Then the solution of the
initial value problem
dT
dt
= k(70 T), T(0) = 98.6
is
T(t) = 70 + (28.6)e
kt
.
If t = a at 12 noon, then
T(a) = 70 + (28.6)e
ka
= 80 and
T(a + 1) = 70 + (28.6)e
k(a+1)
= 75.
Hence
(28.6)e
ka
= 10 and (28.6)e
ka
e
k
= 5.
It follows that k = ln2, and the rst of the previous two equations then yields
a =
ln(2.86)
ln2
1.516
9
(in hours), so the death occurred at about 10:29 a.m.
C09S03.039: Given:
dN
dt
= k(10000 N), with time t measured in months.

dN
10000 N
= k dt;
ln(10000 N) = C
1
kt;
10000 N = Ce
kt
.
On January 1, t = 0 and N = 1000. On April 1, t = 3 and N = 2000. On October 1, t = 9; we want to
determine the value of N then.
9000 = Ce
0
= C, so N(t) = 10000 9000e
kt
.
2000 = 10000 9000e
3k
, so 8 = 9e
3k
.
Therefore k =
1
3
ln
_
9
8
_
. So N(9) = 1000 9000e
9k
= 1000
_
10 9e
3 ln(9/8)
_
3679.
C09S03.040:
dx
dt
= k(100000 x(t)):
dx
100000 x
= k dt;
ln(100000 x) = C
1
kt;
100000 x = Ce
kt
;
x(t) = 100000 Ce
kt
.
On March 1, t = 0 and x = 20000. On March 15, t = 14 and x = 60000.
20000 = 100000 C, so C = 80000.
x(t) = 10000
_
10 8e
kt
_
.
60000 = x(14) = 10000(10 8e
14k
). so 6 = 10 8e
14k
.
Solve for k =
1
14
ln2.
(a) x(t) = 10000(10 8e
kt
) where k =
1
14
ln2.
(b) x(T) = 80000: Solve 10 8e
kT
= 2 for T: T =
1
k
ln4 = 28. So 80000 people will be infected on
March 29.
(c) lim
t
N(t) = 100000: Eventually everybody gets the u.
C09S03.041: Let t = 0 when it began to snow, with t = t
0
at 7:00 a.m. Let x(t) denote the distance
traveled by the snowplow along the road, so that x(t
0
) = 0. If y = ct is the depth of the snow at time t, w
is the width of the road, and v = x

(t) is the velocity of the snowplow, then plowing at a constant rate

means that the product wyv is constant. Hence x(t) satises the dierential equation
k
dx
dt
=
1
t
10
where k is a positive constant. The solution for which x(t
0
) = 0 satises the equation
t = t
0
e
kx
.
We are given x = 2 when t = t
0
+ 1 and x = 4 when t = t
0
+ 3, and it follows that
t
0
+ 1 = t
0
e
2k
and t
0
+ 3 = t
0
e
4k
.
Elimination of t
0
yields the equation
e
4k
3e
2k
+ 2 = 0; that is,
_
e
2k
1
__
e
2k
2
_
= 0.
Thus it follows (because k > 0) that e
2k
= 2. Hence t
0
+1 = 2t
0
, and so t
0
= 1. Therefore it began to snow
at 6:00 a.m.
C09S03.042: Let t = 0 when it began to snow, with t = t
0
at 7:00 a.m. Let x(t) denote the distance
traveled by the snowplow along the road, so that x(t
0
) = 0. If y = ct is the depth of the snow at time t, w
is the width of the road, and v = x

(t) is the velocity of the snowplow, then plowing at a constant rate

means that the product wyv is constant. Hence x(t) satises the dierential equation
k
dx
dt
=
1
t
where k is a positive constant. The solution for which x(t
0
) = 0 satises the equation
t = t
0
e
kx
.
We are given x = 4 when t = t
0
+ 1 and x = 7 when t = t
0
+ 2, and it follows that
t
0
+ 1 = t
0
e
4k
and t
0
+ 2 = t
0
e
7k
(1)
at 8:00 a.m. and 9:00 a.m., respectively. Elimination of t
0
yields the equation
2e
4k
e
7k
1 = 0,
which we solve (using Newtons method) for k 0.08276. With this value of k we nally solve either of the
equations in (1) for t
0
2.5483 (h), about 2 h 33 min. Thus it began to snow at about 4:27 a.m.
C09S03.043: Substitution of v = dy/dx in the dierential equation for y = y(x) yields
a
dv
dx
=
_
1 + v
2
,
and separation of variables then yields
1

1 + v
2
dv =
1
a
dx;
sinh
1
v =
x
a
+ C
1
;
dy
dx
= sinh
_
x
a
+ C
1
_
.
Because y(0) = 0, it follows that C
1
= 0, and therefore
11
dy
dx
= sinh
_
x
a
_
;
y(x) = a cosh
_
x
a
_
+ C.
Of course the (vertical) position of the x-axis may be adjusted so that C = 0, and the units in which T
and are measured may be adjusted so that a = 1. In essence, then, the shape of the hanging cable is the
graph of y = cosh x.
12