Professional Documents
Culture Documents
10 DIFFERENTIAL
EQUATIONS
10.1 Solving Differential Equations
Preliminary Questions
1. Determine the order of the following differential equations:
(a) x 5 y = 1 (b) (y )3 + x = 1
(c) y + x 4 y = 2
SOLUTION
(a) The highest order derivative that appears in this equation is a first derivative, so this is a first order equation.
(b) The highest order derivative that appears in this equation is a first derivative, so this is a first order equation.
(c) The highest order derivative that appears in this equation is a third derivative, so this is a third order equation.
2. Is y = sin x a linear differential equation?
SOLUTION Yes.
3. Give an example of a nonlinear differential equation of the form y = f (y).
SOLUTION One possibility is y = y 2 .
4. Can a nonlinear differential equation be separable? If so, give an example.
SOLUTION Yes. An example is y = y 2 .
5. Give an example of a linear, nonseparable differential equation.
SOLUTION One example is y + y = x.
Exercises
1. Which of the following differential equations are first-order?
(a) y = x 2 (b) y = y 2
(c) (y )3 + yy = sin x
(d) x 2 y e x y = sin y
y
(e) y + 3y = (f) yy + x + y = 0
x
SOLUTION
(a) The highest order derivative that appears in this equation is a first derivative, so this is a first order equation.
(b) The highest order derivative that appears in this equation is a second derivative, so this is not a first order equation.
(c) The highest order derivative that appears in this equation is a first derivative, so this is a first order equation.
(d) The highest order derivative that appears in this equation is a first derivative, so this is a first order equation.
(e) The highest order derivative that appears in this equation is a second derivative, so this is not a first order equation.
(f) The highest order derivative that appears in this equation is a first derivative, so this is a first order equation.
2. Which of the equations in Exercise 1 are linear?
SOLUTION
In Exercises 39, verify that the given function is a solution of the differential equation.
S E C T I O N 10.1 Solving Differential Equations 1017
3. y 8x = 0, y = 4x 2
SOLUTION Let y = 4x 2 . Then y = 8x and
y 8x = 8x 8x = 0.
4. y + 8y = 0, y = 4e8x
SOLUTION Let y = 4e8x . Then y = 32e8x , and
y + 8y = 32e8x + 8(4e8x ) = 0.
5. yy + 4x = 0, y = 12 4x 2
SOLUTION Let y = 12 4x 2 . Then
4x
y = ,
12 4x 2
and
4x
yy + 4x = 12 4x 2 + 4x = 4x + 4x = 0.
12 4x 2
6. y + 4x y = 0, y = 3e2x
2
y + 4x y = 12xe2x + 4x(3e2x ) = 0.
2 2
8. y 2x y + 8y = 0, y = 4x 4 12x 2 + 3
SOLUTION Let y = 4x 4 12x 2 + 3. Then y = 16x 3 24x, y = 48x 2 24, and
9. y 2y + 5y = 0, y = e x sin 2x
SOLUTION Let y = e x sin 2x. Then
and
10. Which of the following equations are separable? Write those that are separable in the form y = f (x)g(y) (but do
not solve).
(a) x y + y 2 = 0 (b) 1 x 2 y = e y sin x
(c) y = x 2 + y 2 (d) y = 9 y 2
SOLUTION
1018 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
(a) x y + y 2 = 0 is separable:
x y + y2 = 0
x y = y 2
1
y = y2
x
(b) 1 x 2 y = e y sin x is separable:
1 x 2 y = e y sin x
sin x
y = e y .
1 x2
(c) y = x 2 + y 2 is not separable; y is already isolated, but is not equal to a product f (x)g(y).
(d) y = 9 y 2 is separable: y = (1)(9 y 2 ).
11. Consider the differential equation y = e y cos x.
(a) Write it as ey d y = cos
x d x.
(b) Integrate both sides of ey d y = cos x d x.
(c) Show that y = ln |C sin x| is the general solution.
(d) Find the particular solution satisfying y(0) = 0.
SOLUTION
(a) As y = dd xy , we regroup:
dy
= e y cos x
dx
ey d y = cos x d x
ey + C1 = sin x + C2
ey = sin x + C
y = ln | sin x + C|
y = ln |C sin x|
0 = ln(C)
y = x 2 e y .
ey d y = x 2 d x.
ey + C1 = x 1 + C2
ey = x 1 + C
S E C T I O N 10.1 Solving Differential Equations 1019
(c) Solving the last expression from part (b) for y, we find
y = lnx 1 + C
y = lnC x 1 .
13. y = x y 2
SOLUTION Rewrite
dy dy
= x y2 as = x d x.
dx y2
Integrating both sides of this equation yields
dy
= x dx
y2
1 1
= x 2 + C.
y 2
Thus,
1
y=1 ,
2x +C
2
dy 1 dy 1
= xy as = x d x.
dx 2 y 2
Integrating both sides of this equation yields
dy 1
= x dx
y 2
1 2
ln |y| = x + C.
4
Solving for y, we find
1 2 1 2
|y| = e 4 x +C = eC e 4 x
1 2 1 2
y = eC e 4 x = Ae 4 x ,
dy dy
= 9y as = 9 d x.
dx y
Integrating both sides of this equation yields
dy
= 9 dx
y
ln |y| = 9x + C.
16. y = 2(4 y)
SOLUTION Rewrite
dy dy
= 2(4 y) as = 2 d x.
dx 4 y
Integrating both sides of this equation yields
dy
= 2 dx
4y
ln |4 y| = 2x + C.
ln |4 y| = 2x C
|4 y| = e2xC = eC e2x
y = 4 eC e2x = 4 + Ae2x ,
dy
2 + (6y + 4) = 0
dx
as
dy dy
= (3y + 2) and then = d x.
dx 3y + 2
Integrating both sides of this equation yields
dy
= dx
3y + 2
ln |3y + 2| = x + C.
|3y + 2| = ex+C = eC ex
3y + 2 = eC ex
1 2 2
y = eC ex = Aex ,
3 3 3
dy dy
=2 y as = 2 dt.
dt y
y = (t + C)2 ,
dy
19. te y = 0
dt
SOLUTION Rewrite
dy
= te y as ey d y = t dt.
dt
Integrating both sides of this equation yields
ey d y = t dt
1 2
ey = t + C.
2
Solving for y, we find
1
y = ln t 2 + C
2
1
y = ln C t 2 ,
2
where C is an arbitrary constant.
20. 1 x 2 y = x y
SOLUTION Rewrite
dy dy x
1 x2 = xy as = d x.
dx y 1 x2
Integrating both sides of this equation yields
dy x
= dx
y 1 x2
ln |y| = 1 x 2 + C.
y = eC e 1x = Ae 1x ,
2 2
dy dy
= y 2 (1 x 2 ) as = (1 x 2 ) d x.
dx y2
Integrating both sides of this equation yields
dy
= (1 x 2 ) d x
y2
1 3
y 1 = x x + C.
3
Solving for y, we find
1 3
y 1 = x x +C
3
1
y= 1 ,
3x x + C
3
SOLUTION Rewrite
dy
y =x as y d y = x d x.
dx
Integrating both sides of this equation yields
y dy = x dx
1 2 1
y = x 2 + C.
2 2
Solving for y, we find
y 2 = x 2 + 2C
y = x 2 + A,
dx 1 1
(t 2 + 1) = (x 2 + 1) as dx = 2 dt.
dt x2 + 1 t +1
Integrating both sides of this equation yields
1 1
dx = dt
x +1
2 t2 + 1
tan1 x = tan1 t + C.
We can simplify this expression by applying the sum formula for the tangent function:
dy 1 x3
(1 + x 2 ) = x3 y as dy = d x.
dx y 1 + x2
Integrating both sides of this equation yields
1 x3
dy = d x.
y 1 + x2
3
To integrate x 2 , note
1+x
x3 (x 3 + x) x x
= =x .
1 + x2 1 + x2 1 + x2
Thus,
1 2 1
ln |y| = x ln |x 2 + 1| + C
2 2
e x /2
2
|y| = eC
x2 + 1
e x /2 e x /2
2 2
y = eC = A ,
x2 + 1 x2 + 1
where A = eC is an arbitrary constant.
S E C T I O N 10.1 Solving Differential Equations 1023
25. y = x sec y
SOLUTION Rewrite
dy
= x sec y as cos y d y = x d x.
dx
Integrating both sides of this equation yields
cos y d y = x dx
1 2
sin y = x + C.
2
Solving for y, we find
1 2
y = sin1 x +C ,
2
where C is an arbitrary constant.
dy
26. = tan y
dt
SOLUTION Rewrite
dy
= tan y as cot y d y = dt.
dt
Integrating both sides of this equation yields
cos y
dy = dt
sin y
ln |sin y| = t + C.
|sin y| = et+C = eC et
sin y = eC et
y = sin1 Aet ,
dy 1
= y tan t as d y = tan t dt.
dt y
Integrating both sides of this equation yields
1
dy = tan t dt
y
ln |y| = ln |sec t| + C.
dx
= t tan x as cot x d x = t dt.
dt
1024 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
1 2
ln |sin x| = t + C.
2
Solving for y, we find
1 2 1 2
| sin x| = e 2 t +C = eC e 2 t
1 2
sin x = eC e 2 t
1 2
x = sin1 Ae 2 t ,
29. y + 2y = 0, y(ln 2) = 3
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy 1
+ 2y = 0 as d y = 2 d x,
dx y
and then integrate to obtain
ln |y| = 2x + C.
Thus,
y = Ae2x ,
where A = eC is an arbitrary constant. The initial condition y(ln 2) = 3 allows us to determine the value of A.
1
3 = Ae2(ln 2) ; 3= A ; so 12 = A.
4
Finally,
y = 12e2x .
30. y 2y + 4 = 0, y(1) = 4
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy 1
2y + 4 = 0 as d y = 1 d x,
dx 2y 4
and then integrate to obtain
1
ln |2y 4| = x + C.
2
Thus,
y = Ae2x + 2,
where A = 12 e2C is an arbitrary constant. The initial condition y(1) = 4 allows us to determine the value of A.
Finally,
31. yy = xey ,
2
y(0) = 1
S E C T I O N 10.1 Solving Differential Equations 1025
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy
= xey
2 2
y as ye y d y = x d x,
dx
and then integrate to obtain
1 y2 1
e = x 2 + C.
2 2
Thus,
y = ln(x 2 + A),
where A = 2C is an arbitrary
constant. The initial condition y(0) = 1 allows us to determine the value of A. Since
y(0) < 0, we have y = ln(x 2 + A), and
1 = ln( A); 1 = ln( A); so e = A.
Finally,
y = ln(x 2 + e).
dy
32. y 2 = x 3 , y(2) = 0
dx
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy
y2 = x 3 as y 2 d y = x 3 d x,
dx
and then integrate to obtain
1 3 1
y = x 2 + C.
3 2
Thus,
3 2 1/3
y= A x ,
2
where A = 3C is an arbitrary constant. The initial condition y(2) = 0 allows us to determine the value of A.
1/3
3 3 1/3 3
0 = A 22 ; 0= A ; so A = .
2 8 8
Finally,
3 3 2 1/3
y= x .
8 2
y = Ae(1/2)x x + 2,
2
where A = eC is an arbitrary constant. The initial condition y(0) = 3 allows us to determine the value of A.
3 = Ae0 + 2 = A + 2 so A = 1.
Finally,
y = e(1/2)x x + 2.
2
1026 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
dy
34. (1 t) y = 0, y(2) = 4
dt
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy 1 1
(1 t) =y as dy = dt,
dt y t 1
and then integrate to obtain
ln |y| = ln |t 1| + C.
Thus,
A
y= ,
t 1
where A = eC is an arbitrary constant. The initial condition y(2) = 4 allows us to determine the value of A.
A
4 = = A.
21
Finally,
4
y= .
t 1
dy
35. = yet , y(0) = 1
dt
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy 1
= yet as d y = et dt,
dt y
and then integrate to obtain
ln |y| = et + C.
Thus,
t
y = Aee ,
where A = eC is an arbitrary constant. The initial condition y(0) = 1 allows us to determine the value of A.
1 = Ae1 so A = e.
Finally,
t t
y = (e)ee = e1e .
dy
36. = tey , y(1) = 0
dt
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy
= tey as e y d y = t dt,
dt
and then integrate to obtain
1 2
ey = t + C.
2
Thus,
1 2
y = ln t +C ,
2
where C is an arbitrary constant. The initial condition y(1) = 0 allows us to determine the value of C.
1 1 1
0 = ln + C ; 1 = + C; so C = .
2 2 2
Finally,
1 2 1
y = ln t + .
2 2
S E C T I O N 10.1 Solving Differential Equations 1027
dy
37. t 2 t = 1 + y + t y, y(1) = 0
dt
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy
t2 = 1 + t + y + t y = (1 + t)(1 + y)
dt
as
1 1+t
d y = 2 dt,
1+y t
and then integrate to obtain
ln |1 + y| = t 1 + ln |t| + C.
Thus,
t
y = A 1/t 1,
e
where A = eC is an arbitrary constant. The initial condition y(1) = 0 allows us to determine the value of A.
1
0= A 1 so A = e.
e
Finally,
et
y = 1/t 1.
e
38. 1 x 2 y = y 2 + 1, y(0) = 0
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy 1 1
1 x2 = y2 + 1 as dy = d x,
dx y2 + 1 1 x2
and then integrate to obtain
tan1 y = sin1 x + C.
Thus,
y = tan(sin1 x + C),
where C is an arbitrary constant. The initial condition y(0) = 0 allows us to determine the value of C.
0 = tan sin1 0 + C = tan C so 0 = C.
Finally,
y = tan sin1 x .
39. 1 x 2 y = y2, y(0) = 1
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy dy dx
1 x2 = y2 as 2
= ,
dx y 1 x2
and then integrate to obtain
1
= sin1 x + C.
y
Thus,
1
y= ,
sin1 x + C
1028 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
where C is an arbitrary constant. The initial condition y(0) = 1 allows us to determine the value of C.
1 1
1= = so C = 1.
sin1 0 + C C
Finally,
1 1
y= = .
sin1 x 1 1 sin1 x
40. y = tan y, y(ln 2) =
2
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy dy
= tan y as = d x,
dx tan y
and then integrate to obtain
ln |sin y| = x + C.
Thus,
y = sin1 ( Ae x ),
where A = eC is an arbitrary constant. The initial condition y(ln 2) = 2 allows us to determine the value of A.
1
= sin1 (2 A); 1 = 2A so A= .
2 2
Finally,
1 x
y = sin1 e .
2
dy
= y 2 sin x as y 2 d y = sin x d x,
dx
and then integrate to obtain
y 1 = cos x + C.
Thus,
1
y= ,
A + cos x
where A = C is an arbitrary constant. The initial condition y(0) = 3 allows us to determine the value of A.
1 1 1 2
3= ; A+1= so A= 1= .
A+1 3 3 3
Finally,
1
y= .
cos x (2/3)
y 6x 2 y = 0
a 2 a 6 = (a 3)(a + 2) = 0.
y + 2y 8y = 0
y + 2y 8y = eax (a 2 + 2a 8).
dy y1 2
t (y 1) = 2y as d y = dt,
dt y t
and then integrate to obtain
y ln |y| = 2 ln |t| + A.
eA e y = |y|t 2
t 2 y = Ce y
where C = eA is an arbitrary constant. The initial condition y(1) = 1 will gives us the value of C.
1(1) = Ce so C = e1 .
t 2 (2) = e21
e
t2 =
2
t = e/2.
In Exercises 4548, use Eq. (4) and Torricellis Law [Eq. (5)].
45. A cylindrical tank filled with water has height 10 ft and a base of area 30 ft2 . Water leaks through a hole in the
bottom of area 13 ft2 . How long does it take (a) for half of the water to leak out and (b) for the tank to empty?
SOLUTION Because the tank has a constant cross-sectional area of 30 ft2 and the hole has an area of 13 ft2 , the differ-
ential equation for the height of the water in the tank is
1v
dy v
= 3 = .
dt 30 90
By Torricellis Law,
v = 2gy = 8 y,
46. A conical tank filled with water has height 12 ft [Figure 7(A)]. Assume that the top is a circle of radius 4 ft and that
water leaks through a hole in the bottom of area 2 in.2 . Let y(t) be the water level at time t.
4
40 ft
12
(a) Show that the cross-sectional area of the tank at height y is A(y) = 9 y 2 .
(b) Find the differential equation satisfied by y(t) and solve for y(t). Use the initial condition y(0) = 12.
(c) How long does it take for the tank to empty?
SOLUTION
(a) By similar triangles, the radius r at height y satisfies
r 4 1
= = ,
y 12 3
so r = y/3 and
2
A(y) = r 2 = y .
9
1 ft2 , so the differential equation for the height of the water in the tank
(b) The area of the hole is B = 2 in2 = 72
becomes:
dy 8B y 1
= = y 3/2 .
dt A(y)
S E C T I O N 10.1 Solving Differential Equations 1031
1 8 y
dy 1
= 48 = .
dt 80 20y y 2 480 20 y
2 t 2
(20 y)3/2 = 103/2
3 480 3
2/3
t
y(t) = 20 + 103/2 .
320
The tank is empty when y = 0. Thus, te satisfies the equation
2/3
te
20 + 103/2 = 0.
320
It follows that
48. A cylindrical tank filled with water has height h ft and a base of area A ft2 . Water leaks through a hole in
the bottom of area B ft2 .
(a) Show that the time required for the tank to empty is proportional to A h/B.
(b) Show that the emptying time is proportional to V h 1/2 , where V is the volume of the tank.
(c) Two tanks have the same volume and same-sized hole, but different heights and bases. Which tank empties first: the
taller or shorter tank?
1032 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
SOLUTION Torricellis law gives the differential equation for the height of the water in the tank as
dy B y
= 2g
dt A
Separating variables and integrating then yields:
B
y 1/2 d y = 2g dt
A
Bt
2y 1/2 = 2g +C
A
Bt
y 1/2 = g/2 +C
A
When t = 0, y = h, so C = h 1/2 and
Bt
y 1/2 = h g/2 .
A
(a) When the tank is empty, y = 0. Thus, the time required for the tank to empty, te , satisfies the equation
Bte
0= h g/2 .
A
It follows that
A A h
te = 2h/g = 2/g ;
B B
that is, the time required for the tank to empty is proportional to A h/B.
(b) The volume of the tank is V = Ah; therefore
A h 1 V
= ,
B B h
and
A h 2/g V
te = 2/g = ;
B B h
that is, the time required for the tank to empty is proportional to V h 1/2 .
(c) By part (b), with V and B held constant, the emptying time decreases with height. The taller tank therefore empties
first.
49. Figure 8 shows a circuit consisting of a resistor of R ohms, a capacitor of C farads, and a battery of voltage V . When
the circuit is completed, the amount of charge q(t) (in coulombs) on the plates of the capacitor varies according to the
differential equation (t in seconds)
dq 1
R + q=V
dt C
(a) Solve for q(t).
(b) Show that lim q(t) = C V .
t
(c) Find q(t), assuming that q(0) = 0. Show that the capacitor charges to approximately 63% of its final value C V after
a time period of length = RC ( is called the time constant of the capacitor).
V C
FIGURE 8 An RC circuit.
SOLUTION
S E C T I O N 10.1 Solving Differential Equations 1033
where K = ek .
(b) Using the result from part (a), we calculate
lim q(t) = lim C V + K e RC t = C V + K lim e RC t = C V + K 0 = C V .
1 1
t t t
(c) Using the result from part (a), the condition q(0) = 0 determines K = C V . Thus,
1
q(t) = C V 1 e RC t ,
and
q( ) = q(RC) = C V (1 e1 ) 0.632C V .
50. Assume in the circuit of Figure 8 that R = 100 , C = 0.01 F, and V = 10 V. How many seconds does it take for
the charge on the capacitor plates to reach half of its limiting value?
SOLUTION From Exercise 49, we know that
q(t) = C V 1 et/(RC) = 0.1(1 et ),
and the limiting value of q(t) is C V = 0.1. If the charge on the capacitor plates has reached half its limiting value, then
0.1
= 0.1(1 et )
2
1 et = 1/2
et = 1/2
t = ln 2
Therefore, the charge on the capacitor plates reaches half of its limiting value after ln 2 .693 seconds.
dV
51. One hypothesis for the growth rate of the volume V of a cell is that is proportional to the cells surface
dt
area A. Since V has cubic units such as cm3 and A has square units such as cm2 , we may assume roughly that A V 2/3 ,
dV
and hence = kV 2/3 for some constant k. If this hypothesis is correct, which dependence of volume on time would
dt
we expect to see (again, roughly speaking) in the laboratory?
(a) Linear (b) Quadratic (c) Cubic
SOLUTION Rewrite
dV
= kV 2/3 as V 2/3 dv = k dt,
dt
and then integrate both sides to obtain
3V 1/3 = kt + C
V = (kt/3 + C)3 .
52. We might also guess that the rate at which a snowball melts is proportional to its surface area. What is the differential
equation satisfied by the volume V of a spherical snowball at time t? Suppose the snowball has radius 4 cm and that it
loses half of its volume after 10 min. According to this model, when will the snowball disappear?
SOLUTION Because the snowball is spherical, V = 43 r 3 , A = 4 r 2 , and the differential equation
dV
= k A
dt
becomes
dr dr
4 r 2 = k(4 r 2 ) or = k.
dt dt
Thus, r (t) = kt + C, for some arbitrary constant C. Initially, the snowball has radius 4 cm, so C = 4 and r (t) = 4 kt.
Because the initial radius is 4 cm, the initial volume of the snowball is 256
3 cm . After ten minutes, the volume has
3
128
3
been reduced to 3 cm , so the radius has been reduced to 2 4 cm. Therefore,
3
3
2 4 = 4 k(10),
and
2 3 4
k= cm/min.
5
The snowball has disappeared when r = 0. This will occur when 4 kt = 0, or when
4 20
t= = 48.5 minutes.
k 2 34
53. In general, ( f g) is not equal to f g , but let f (x) = e2x and find a function g(x) such that ( f g) = f g . Do the
same for f (x) = x.
SOLUTION If ( f g) = f g , we have
g (x) 2e2x
= 2x = 2.
g(x) 2e e2x
Integrating and solving for g(x), we find
dg
= 2 dx
dg
ln |g| = 2x + C
g(x) = Ae2x ,
g (x) 1
= .
g(x) 1x
Thus,
dg 1
= dx
g 1x
ln |g| = ln |1 x| + C
A
g(x) = ,
1x
54. A boy standing at point B on a dock holds a rope of length attached to a boat at point A [Figure 9(A)]. As the
boy walks along the dock, holding the rope taut, the boat moves along a curve called a tractrix (from the Latin tractus
meaning to pull). The segment from a point P on the curve to the x-axis along the tangent line has constant length .
Let y = f (x) be the equation of the tractrix.
y
(a) Show that y 2 + (y/y )2 = 2 and conclude y = . Why must we choose the negative square root?
y2
2
(b) Prove that the tractrix is the graph of
+ 2 y 2
x = ln 2 y 2
y
A A
P = (x, f(x))
f (x)
x x
B x
Dock f (x)
f ' (x)
FIGURE 9
SOLUTION
(a) From the diagram on the right in Figure 9, we see that
f (x) 2
f (x)2 + = 2 .
f (x)
If we let y = f (x), this last equation reduces to y 2 + (y/y )2 = 2 . Solving for y , we find
y
y = ,
2 y 2
where we must choose the negative sign because y is a decreasing function of x.
(b) Rewrite
dy y 2 y 2
= as d y = d x,
dx 2 y 2 y
Therefore,
2 y 2 y
x = ln 2 y 2 + C = ln 2 y 2 + C
y 2 y 2
+ 2 y 2
= ln 2 y 2 + C
y
55. If a bucket of water spins about a vertical axis with constant angular velocity (in radians per second), the water
climbs up the side of the bucket until it reaches an equilibrium position (Figure 10). Two forces act on a particle located
at a distance x from the vertical axis: the gravitational force mg acting downward and the force of the bucket on the
y = f(x)
m 2x
mg
x
x
FIGURE 10
particle (transmitted indirectly through the liquid) in the direction perpendicular to the surface of the water. These two
forces must combine to supply a centripetal force m 2 x, and this occurs if the diagonal of the rectangle in Figure 10 is
normal to the waters surface (that is, perpendicular to the tangent line). Prove that if y = f (x) is the equation of the
curve obtained by taking a vertical cross section through the axis, then 1/y = g/( 2 x). Show that y = f (x) is a
parabola.
SOLUTION At any point along the surface of the water, the slope of the tangent line is given by the value of y at that
point; hence, the slope of the line perpendicular to the surface of the water is given by 1/y . The slope of the resultant
force generated by the gravitational force and the centrifugal force is
mg g
= 2 .
m2 x x
Therefore, the curve obtained by taking a vertical cross-section of the water surface is determined by the equation
1 g 2
= 2 or y = x.
y x g
Performing one integration yields
2 2
y = f (x) = x + C,
g
where C is a constant of integration. Thus, y = f (x) is a parabola.
3y x
56. Show that the differential equations y = and y = define orthogonal families of curves; that is, the
x 3y
graphs of solutions to the first equation intersect the graphs of the solutions to the second equation in right angles (Figure
11). Find these curves explicitly.
x0
has a slope of 3y . The slopes are negative reciprocals of one another; hence the tangent lines are perpendicular.
0
Separation of variables and integration applied to y = 3y
x gives
dy dx
=3
y x
ln |y| = 3 ln |x| + C
y = Ax 3
S E C T I O N 10.1 Solving Differential Equations 1037
3y d y = x d x
3y 2 /2 = x 2 /2 + C
y = C x 2 /3
57. Find the family of curves satisfying y = x/y and sketch several members of the family. Then find the differential
equation for the orthogonal family (see Exercise 56), find its general solution, and add some members of this orthogonal
family to your plot.
SOLUTION Separation of variables and integration applied to y = x/y gives
y dy = x dx
1 2 1
y = x2 + C
2 2
y = x2 + C
If y(x) is a curve of the family orthogonal to these, it must have tangent lines of slope y/x at every point (x, y). This
gives
y = y/x
"
1 1
1
58. A 50-kg model rocket lifts off by expelling fuel at a rate of k = 4.75 kg/s for 10 s. The fuel leaves the end of the
rocket with an exhaust velocity of b = 100 m/s. Let m(t) be the mass of the rocket at time t. From the law of conservation
of momentum, we find the following differential equation for the rockets velocity v(t) (in meters per second):
dm
m(t)v (t) = 9.8m(t) + b
dt
(a) Show that m(t) = 50 4.75t kg.
(b) Solve for v(t) and compute the rockets velocity at rocket burnout (after 10 s).
SOLUTION
(a) For 0 t 10, the rocket is expelling fuel at a constant rate of 4.75 kg/s, giving m (t) = 4.75. Hence, m(t) =
4.75t + C. Initially, the rocket has a mass of 50 kg, so C = 50. Therefore, m(t) = 50 4.75t.
dm
(b) With m(t) = 50 4.75t and = 4.75, the equation for v becomes
dt
dm
dv b dt (100)(4.75)
= 9.8 + = 9.8 +
dt 50 4.75t 50 4.75t
and therefore
4.75 dt
v(t) = 9.8t 100 = 9.8t + 100 ln(50 4.75t) + C
50 4.75t
1038 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
59. Let v(t) be the velocity of an object of mass m in free fall near the earths surface. If we assume that air resistance is
dv
proportional to v 2 , then v satisfies the differential equation m = g + kv 2 for some constant k > 0.
dt
(a) Set = (g/k)1/2 and rewrite the differential equation as
dv k
= ( 2 v 2 )
dt m
Then solve using separation of variables with initial condition v(0) = 0.
(b) Show that the terminal velocity lim v(t) is equal to .
t
SOLUTION
b
Area of cross section
y + y is A(y)
y Volume of slice is
V(y + y) V(y) A(y)y
y0 = a
FIGURE 12
SOLUTION
(a) If y is very small, then the slice between y and y + y is very similar to the prism formed by thickening the
cross-sectional area A(y) by a thickness of y. A prism with cross-sectional area A and height y has volume Ay.
This gives
61. In most cases of interest, the general solution of a differential equation of order n depends on n arbitrary constants.
This exercise shows there are exceptions.
(a) Show that (y )2 + y 2 = 0 is a first-order equation with only one solution y = 0.
(b) Show that (y )2 + y 2 + 1 = 0 is a first-order equation with no solutions.
SOLUTION
and
63. A spherical tank of radius R is half-filled with water. Suppose that water leaks through a hole in the bottom of area
B ft2 . Let y(t) be the water level at time t (seconds).
1040 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
dy 8B y
(a) Show that = .
dt (2Ry y 2 )
(b) Show that for some constant C,
2 3/2 1 5/2
Ry y =C t
4B 3 5
(c) Use the initial condition y(0) = R to compute C and show that C = te , the time at which the tank is empty.
(d) Show that te is proportional to R 5/2 and inversely proportional to B.
SOLUTION
(a) At height y above the bottom of the tank, the cross section is a circle of radius
r = R 2 (R y)2 = 2Ry y 2 .
The cross-sectional area function is then A(y) = (2Ry y 2 ). The differential equation for the height of the water in
the tank is then
dy 8B y
= .
dt (2Ry y 2 )
(b) Rewrite the differential equation as
2Ry 1/2 y 3/2 d y = dt,
8B
and then integrate both sides to obtain
2 3/2 1 5/2
Ry y = C t,
4B 3 5
where C is an arbitrary constant.
(c) The initial condition y(0) = R allows us to determine the value of C:
2 5/2 1 5/2 7 5/2
C= R R = R .
4B 3 5 60B
Moreover, note that y = 0 when t = C, C = te , the time at which the tank is empty.
(d) From part (c),
7 5/2
te = R ,
60B
from which it is clear that te is proportional to R 5/2 and inversely proportional to B.
Exercises
1. Figure 8 shows the slope field for y = sin y sin t. Sketch the graphs of the solutions with initial conditions y(0) = 1
and y(0) = 1. Show that y(t) = 0 is a solution and add its graph to the plot.
y
3
2
1
0 t
1
2
3
3 2 1 0 1 2 3
FIGURE 8 Slope field for y = sin y sin t.
If y(t) = 0, then y = 0; moreover, sin 0 sin t = 0. Thus, y(t) = 0 is a solution of y = sin y sin t.
2. Figure 9 shows the slope field of y = y 2 t 2 . Sketch the integral curve passing through the point (0, 1), the curve
through (0, 0), and the curve through (0, 2). Is y(t) = 0 a solution?
y
3
2
1
0 t
1
2
3
3 2 1 0 1 2 3
FIGURE 9 Slope field for y = y 2 t 2 .
0 t
0.5
1
1 0.5 0 0.5 1 1.5 2
FIGURE 10 Slope field for y = t 2y.
1042 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
y + 2y = + t = t,
1 1
2 2
so f (t) = 12 (t 12 ) is a solution to y = t 2y. The slope field with the four required solutions is shown below.
y = 2Ce2t ,
1
2
and
y + 2y = 2Ce2t + t
1 1
+ 2Ce2t = t.
2 2
0 t
1
1 0 1 2 3
t y = t (t 1 + Cet ) = 1 Cet .
Thus, y = t 1 + Cet is a solution of y = t y. On the slope field, we can see that the isoclines of 1 all lie along
the line y = t 1. Whenever y > t 1, y = t y < 1, so the solution curve will converge downward towards the
line y = t 1. On the other hand, if y < t 1, y = t y > 1, so the solution curve will converge upward towards
y = t 1. In either case, the solution is approaching t 1.
5. Show that the isoclines of y = 1/y are horizontal lines. Sketch the slope field for 2 t, y 2 and plot the
solutions with initial conditions y(0) = 0 and y(0) = 1.
1 1
SOLUTION The isocline of slope c is defined by y = c. This is equivalent to y = c , which is a horizontal line. The
slope field and the solutions are shown below.
S E C T I O N 10.2 Graphical and Numerical Methods 1043
6. Show that the isoclines of y = t are vertical lines. Sketch the slope field for 2 t, y 2 and plot the integral
curves passing through (0, 1) and (0, 1).
SOLUTION The isocline of slope c for the differential equation y = t has equation t = c, which is the equation of a
vertical line. The slope field and the required solution curves are shown below.
y
2
0 t
2
2 1 0 1 2
7. Sketch the slope field of y = t y for 2 t, y 2. Based on the sketch, determine lim y(t), where y(t) is a
t
solution with y(0) > 0. What is lim y(t) if y(0) < 0?
t
SOLUTION The slope field for y = t y is shown below.
y
2
0 t
2
2 1 0 1 2
With y(0) > 0, the slope field indicates that y is an always increasing, always concave up function; consequently,
limt y = . On the other hand, when y(0) < 0, the slope field indicates that y is an always decreasing, always
concave down function; consequently, limt y = .
8. Match the differential equation with its slope field in Figures 11(A)(F).
(i) y = 1
(ii) y =
y
t
(iii) y = t 2 y
(iv) y = t y 2
(v) y = t 2 + y 2
(vi) y = t
y y
3 3
2 2
1 1
0 t 0 t
1 1
2 2
3 3
3 2 1 0 1 2 3 3 2 1 0 1 2 3
FIGURE 11(A) FIGURE 11(B)
1044 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
y y
3 3
2 2
1 1
0 t 0 t
1 1
2 2
3 3
3 2 1 0 1 2 3 3 2 1 0 1 2 3
FIGURE 11(C) FIGURE 11(D)
y y
3 3
2 2
1 1
0 t 0 t
1 1
2 2
3 3
3 2 1 0 1 2 3 3 2 1 0 1 2 3
FIGURE 11(E) FIGURE 11(F)
SOLUTION
(i) Every segment in the slope field for y = 1 will have slope 1; this matches Figure 11(C).
(ii) The segments in the slope field for y =
y
will have positive slope in the first and third quadrants and negative
t
slopes in the second and fourth quadrant; this matches Figure 11(B).
(iii) The segments in the slope field for y = t 2 y will have positive slope in the upper half of the plane and negative
slopes in the lower half of the plane; this matches Figure 11(F).
(iv) The segments in the slope field for y = t y 2 will have positive slope on the right side of the plane and negative slopes
on the left side of the plane; this matches Figure 11(D).
(v) Every segment in the slope field for y = t 2 + y 2 , except at the origin, will have positive slope; this matches Figure
11(A).
(vi) The isoclines for y = t are vertical lines; this matches Figure 11(E).
9. One of the slope fields in Figures 12(A) and (B) is the slope field for y = t 2 . The other is for y = y 2 . Identify which
is which. In each case, sketch the solutions with initial conditions y(0) = 1, y(0) = 0, and y(0) = 1.
y y
3 3
2 2
1 1
0 t 0 t
1 1
2 2
3 3
3 2 1 0 1 2 3 3 2 1 0 1 2 3
FIGURE 12(A) FIGURE 12(B)
SOLUTION For y = t 2 , y only depends on t. The isoclines of any slope c will be the two vertical lines t = c.
This indicates that the slope field will be the one given in Figure 12(A). The solutions are sketched below:
y
2
0 t
2
2 1 0 1 2
For y = y 2 , y only depends on y. The isoclines of any slope c will be the two horizontal lines y = c. This indicates
that the slope field will be the one given in Figure 12(B). The solutions are sketched below:
S E C T I O N 10.2 Graphical and Numerical Methods 1045
y
2
0 t
2
2 1 0 1 2
0 t
2
2 1 0 1 2
(b) Rewrite
dy t
= as y d y = t dt,
dt y
and then integrate both sides to obtain
1 2 1
y = t 2 + C.
2 2
Solving for y, we find that the general solution is
y = t 2 + C.
0 t
2
2 1 0 1 2
11. Sketch the slope field of y = t 2 y in the region 3 t 3, 3 y 3 and sketch the solutions satisfying
y(1) = 0, y(1) = 1, and y(1) = 1.
SOLUTION The slope field for y = t 2 y, together with the required solution curves, is shown below.
y
3
2
1
0 t
1
2
3
3 2 1 0 1 2 3
1046 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
12. Let F(t, y) = t 2 y and let y(t) be the solution of y = F(t, y) satisfying y(2) = 3. Let h = 0.1 be the time step
in Eulers Method and set y0 = y(2) = 3.
(a) Calculate y1 = y0 + h F(2, 3).
(b) Calculate y2 = y1 + h F(2.1, y1 ).
(c) Calculate y3 = y2 + h F(2.2, y2 ) and continue computing y4 , y5 , and y6 .
(d) Find approximations to y(2.2) and y(2.5).
SOLUTION
(a) With y0 = 3, t0 = 2, h = 0.1, and F(t, y) = t 2 y, we find
y3 = y2 + h F(t2 , y2 ) = 3.3919;
y4 = y3 + h F(t3 , y3 ) = 3.58171;
y5 = y4 + h F(t4 , y4 ) = 3.799539;
y6 = y5 + h F(t5 , y5 ) = 4.0445851.
n tn yn
0 0 0
1 0.1 y0 + h F(t0 , y0 ) = 0
2 0.2 y1 + h F(t1 , y1 ) = 0.01
3 0.3 y2 + h F(t2 , y2 ) = 0.029801
4 0.4 y3 + h F(t3 , y3 ) = 0.058920
5 0.5 y4 + h F(t4 , y4 ) = 0.096631
6 0.6 y5 + h F(t5 , y5 ) = 0.142026
7 0.7 y6 + h F(t6 , y6 ) = 0.194082
8 0.8 y7 + h F(t7 , y7 ) = 0.251733
9 0.9 y8 + h F(t8 , y8 ) = 0.313929
10 1.0 y9 + h F(t9 , y9 ) = 0.379681
(b) Rewrite
dy
= tey as e y d y = t dt,
dt
and then integrate both sides to obtain
1 2
ey = t + C.
2
Thus,
1
y = ln t 2 + C .
2
Applying the initial condition
y(0) = 0 yields 0 = ln |C|, so C = 1. The exact solution to the initial value problem is
then y = ln 12 t 2 + 1 .
S E C T I O N 10.2 Graphical and Numerical Methods 1047
In Exercises 1419, use Eulers Method with h = 0.1 to approximate the given value of y(t).
14. y(0.5); y = y 2 , y(0) = 0
SOLUTION Let F(t, y) = y 2 . With t0 = 0, y0 = 0, and h = 0.1, we compute
y(0.0) = y0 =0
y(0.1) y1 = y0 + 0.1(y0 )2 = 0
y(0.2) y2 = y1 + 0.1(y1 )2 = 0
y(0.3) y3 = y2 + 0.1(y2 )2 = 0
y(0.4) y4 = y3 + 0.1(y3 )2 = 0
y(0.5) y5 = y4 + 0.1(y4 )2 = 0
Clearly, y(t) = 0 for all t is the exact solution to the initial value problem.
15. y(1); y = y, y(0) = 0
SOLUTION Let F(t, y) = y. With t0 = 0, y0 = 0, and h = 0.1, we compute
y(0.0) = y0 =0
y(0.1) y1 = y0 + 0.1(y0 ) =0
y(0.2) y2 = y1 + 0.1(y1 ) =0
y(0.3) y3 = y2 + 0.1(y2 ) =0
y(0.4) y4 = y3 + 0.1(y3 ) =0
y(0.5) y5 = y4 + 0.1(y4 ) =0
y(0.6) y6 = y5 + 0.1(y5 ) =0
y(0.7) y7 = y6 + 0.1(y6 ) =0
y(0.8) y8 = y7 + 0.1(y7 ) =0
y(0.9) y9 = y8 + 0.1(y8 ) =0
y(1.0) y10 = y9 + 0.1(y9 )=0
y(0.0) = y0 =1
y(0.1) y1 = y0 + 0.1(y0 t0 ) = 1
y(0.2) y2 = y1 + 0.1(y1 t1 ) = .99
y(0.3) y3 = y2 + 0.1(y2 t2 ) = .9702
y(0.4) y4 = y3 + 0.1(y3 t3 ) = .941094
y(0.5) y5 = y4 + 0.1(y4 t4 ) = .903450
y(0.6) y6 = y5 + 0.1(y5 t5 ) = .858278
y(0.7) y7 = y6 + 0.1(y6 t6 ) = .806781
y(1.0) = y0 =2
y(1.1) y0 + 0.1(t0 sin(y0 )) = 2.09093
y(1.2) y1 + 0.1(t1 sin(y1 )) = 2.18638
y(1.3) y2 + 0.1(t2 sin(y2 )) = 2.28435
y(1.4) y3 + 0.1(t3 sin(y3 )) = 2.38264
y(1.5) y4 + 0.1(t4 sin(y4 )) = 2.47898
1048 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
y(0.0) = y0 =2
y(0.1) y1 = y0 + 0.1(t0 /y0 ) =2
y(0.2) y2 = y1 + 0.1(t1 /y1 ) = 2.005
y(0.3) y3 = y2 + 0.1(t2 /y2 ) = 2.014975
y(0.4) y4 = y3 + 0.1(t3 /y3 ) = 2.029864
y(0.5) y5 = y4 + 0.1(t4 /y4 ) = 2.049569
..
.
y(2.4) y24 = y23 + 0.1(t23 /y23 ) = 3.095655
y(2.5) y25 = y24 + 0.1(t24 /y24 ) = 3.173183
y(2.6) y26 = y25 + 0.1(t25 /y25 ) = 3.251968
y(0.0) = y0 =1
y(0.1) y1 = y0 + 0.1(t0 y0 ) = 0.9
y(0.2) y2 = y1 + 0.1(t1 y1 ) = 0.82
y(0.3) y3 = y2 + 0.1(t2 y2 ) = 0.758
y(0.4) y4 = y3 + 0.1(t3 y3 ) = 0.7122
y(0.5) y5 = y4 + 0.1(t4 y4 ) = 0.68098
yk = yk1 + h f (tk1 ).
y1 = y0 + h f (t0 ) = h f (t0 )
y2 = y1 + h f (t1 ) = h f (t0 ) + f (t1 )
y3 = y2 + h f (t2 ) = h f (t0 ) + f (t1 ) + f (t2 )
..
.
N 1
y N = y N1 + h f (t N 1 ) = h f (t0 ) + f (t1 ) + f (t2 ) + . . . + f (t N 1 ) = h f (tk )
k=0
b
Observe this last expression is exactly the N th left-endpoint approximation to y(b) = f (u) du.
a
In Exercises 2122, use a modification of Eulers Method, Eulers Midpoint Method, that gives a significant improvement
in accuracy. With time step h and initial value y0 = y(t0 ), the values yk are defined successively by the equation
yk = yk1 + hm k1
h h
where m k1 = F tk1 + , yk1 + F(tk1 , yk1 ) .
2 2
21. Apply both Eulers Method and the Euler Midpoint Method with h = 0.1 to estimate y(1.5), where y(t) satisfies
y = y with y(0) = 1. Find y(t) exactly and compute the errors in these two approximations.
S E C T I O N 10.2 Graphical and Numerical Methods 1049
SOLUTION Let F(t, y) = y. With t0 = 0, y0 = 1, and h = 0.1, fifteen iterations of Eulers method yield
The exact solution to y = y, y(0) = 1 is y(t) = et ; therefore y(1.5) = 4.481689. The error from Eulers method
is |4.177248 4.481689| = 0.304441, while the error from Eulers midpoint method is |4.471304 4.481689| =
0.010385.
t
22. If f (t) is continuous on [a, b], then the solution to y = f (t) with initial condition y(a) = 0 is y(t) = f (u) du.
a
ba
Show that the Euler Midpoint Method with time step h = for N steps yields the N th midpoint approximation to
b N
y(b) = f (u) du.
a
SOLUTION For a differential equation of the form y = f (t), the equations for Eulers midpoint method reduce to
h h
m k1 = f tk1 + and yk = yk1 + h f tk1 + .
2 2
With a step size of h = (b a)/N , y(b) = y N . Starting from y0 = 0, we compute
h h
y1 = y0 + h f t0 + = h f t0 +
2 2
h h h
y2 = y1 + h f t1 + = h f t0 + + f t1 +
2 2 2
h h h h
y3 = y2 + h f t2 + = h f t0 + + f t1 + + f t2 +
2 2 2 2
..
.
h h h h h
y N = y N1 + h f t N 1 + = h f t0 + + f t1 + + f t2 + + . . . + f t N 1 +
2 2 2 2 2
N1
h
=h f tk +
k=0
2
b
Observe this last expression is exactly the N th midpoint approximation to y(b) = f (u) du.
a
In Exercises 2326, use Eulers Midpoint Method with h = 0.1 to approximate the given value of y(t).
k tk mk yk
0 0.0 0 0
1 0.1 0 0
2 0.2 0 0
3 0.3 0 0
4 0.4 0 0
5 0.5 0 0
1050 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
Thus, y(0.5) y5 = 0.
24. y(1); y = y 2 , y(0) = 0
SOLUTION Let F(t, y) = y 2 . With t0 = 0, y0 = 0, and h = 0.1, we compute
k tk mk yk
0 0.0 0 0
1 0.1 0 0
2 0.2 0 0
3 0.3 0 0
4 0.4 0 0
5 0.5 0 0
6 0.6 0 0
7 0.7 0 0
8 0.8 0 0
9 0.9 0 0
10 1.0 0 0
k tk mk yk
0 0.0 0 0
1 0.1 0.05 0.005
2 0.2 0.16025 0.021025
3 0.3 0.282076 0.0492326
4 0.4 0.416694 0.0909021
5 0.5 0.565447 0.147447
6 0.6 0.729819 0.220429
7 0.7 0.911450 0.311574
8 0.8 1.112152 0.422789
9 0.9 1.333928 0.556182
10 1.0 1.578991 0.714081
k tk mk yk
0 0.0 0 1
1 0.1 1.0525 1.10525
2 0.2 1.198685 1.225118
3 0.3 1.414524 1.366571
4 0.4 1.737871 1.540358
5 0.5 2.236298 1.763988
Exercises
1. Find the general solution of the logistic equation
y = 3y(1 y/5)
5
y= .
1 e3t /C
The initial condition y(0) = 2 allows us to determine the value of C:
5 1 5 2
2= ; 1 = ; so C = .
1 1/C C 2 3
The particular solution is then
5 10
y= = .
1 + 32 e3t 2 + 3e3t
3. Let y(t) be a solution of y = 0.5y(1 0.5y) such that y(0) = 4. Determine lim y(t) without finding y(t) explicitly.
t
SOLUTION The equation is a logistic equation with k = 0.5 and A = 2. Let F(y) = 0.5y(1 0.5y), and take y0 = 4.
Then, F(y0 ) = 0.5(4)(1 0.5(4)) < 0, so, y(t) is strictly decreasing and
lim y(t) = A = 2.
t
4. Let y(t) be a solution of y = 5y(1 y/A) satisfying y(0) = 10. Find lim y(t), assuming that
t
(a) A = 15 (b) A = 5 (c) A = 10
SOLUTION Let F(y) = 5y(1 y/A).
(a) In this case, F(y0 ) = 5(10)(1 10/15) > 0, so y(t) is strictly increasing with
(b) In this case, F(y0 ) = 5(10)(1 10/5) < 0, so y(t) is strictly decreasing with
lim y(t) = 5.
t
(c) In this case, F(y0 ) = 5(10)(1 10/10) = 0, so the solution is constant with y = 10; consequently,
5. A population of squirrels lives in a forest with a carrying capacity of 2,000. Assume logistic growth with growth
constant k = 0.6 yr1 .
(a) Find a formula for the squirrel population P(t), assuming an initial population of 500 squirrels.
(b) How long will it take for the squirrel population to double?
SOLUTION
(a) Since k = 0.6 and the carrying capacity is A = 2000, the population P(t) of the squirrels satisfies the differential
equation
8. Spread of a Rumor A rumor spreads through a small town. Let y(t) be the fraction of the population that has heard
the rumor at time t and assume that the rate at which the rumor spreads is proportional to the product of the fraction y of
the population that has heard the rumor and the fraction 1 y that has not yet heard the rumor.
(a) Write down the differential equation satisfied by y in terms of a proportionality factor k.
(b) Find k (in units of days1 ), assuming that 10% of the population knows the rumor at t = 0 and 40% knows it at
t = 2 days.
(c) Using the assumptions of part (b), determine when 75% of the population will know the rumor.
SOLUTION
1054 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
(a) y (t) is the rate at which the rumor is spreading, in percentage of the population per day. By the description given,
the rate satisfies:
1 1 1 1
= ; 1 = 10; so C = .
10 1 1/C C 9
2 1 5 1
= ; 1 + 9e2k = ; so k = ln 6 0.896 days1 .
5 1 + 9e2k 2 2
The particular solution of the differential equation for y is then
1
y(t) = .
1 + 9e0.896t
(c) If 75% of the population knows the rumor at time t, we have
3 1
=
4 1 + 9e0.896t
4
1 + 9e0.896t =
3
ln 27
t= 3.67839
0.896
Thus, 75% of the population knows the rumor after approximately 3.67 days.
9. A rumor spreads through a school with 1,000 students. At 8 AM, 80 students have heard the rumor, and by noon, half
the school has heard it. Using the logistic model of Exercise 8, determine when 90% of the students will have heard the
rumor.
SOLUTION Let y(t) be the proportion of students that have heard the rumor at a time t hours after 8 AM. In the logistic
model of Exercise 8, we have a capacity of A = 1 (100% of students) and an unknown growth factor of k. Hence,
1
y(t) = .
1 ekt /C
The initial condition y(0) = 0.08 allows us to determine the value of C:
2 1 1 25 2
= ; 1 = ; so C = .
25 1 1/C C 2 23
so that
2
y(t) = .
2 + 23ekt
The condition y(4) = .5 now allows us to determine the value of k:
1 2 1 23
= ; 2 + 23e4k = 4; so k = ln 0.6106 hours1 .
2 2 + 23e4k 4 2
90% of the students have heard the rumor when y(t) = .9. Thus
9 2
=
10 2 + 23e0.6106t
20
2 + 23e0.6106t =
9
1 207
t= ln 7.6 hours.
0.6106 2
Thus, 90% of the students have heard the rumor after 7.6 hours, or at 3:36 PM.
S E C T I O N 10.3 The Logistic Equation 1055
10. A simpler model for the spread of a rumor assumes that the rate at which the rumor spreads is proportional
(with factor k) to the fraction of the population that has not yet heard the rumor.
(a) Compute the solutions to this model and the model of Exercise 8 with the values k = 0.9 and y0 = 0.1.
(b) Graph the two solutions on the same axis.
(c) Which model seems more realistic? Why?
SOLUTION
(a) Let y(t) denote the fraction of a population that has heard a rumor, and suppose the rumor spreads at a rate propor-
tional to the fraction of the population that has not yet heard the rumor. Then
y = k(1 y),
for some constant of proportionality k. Separating variables and integrating both sides yields
dy
= k dt
1y
ln |1 y| = kt + C.
Thus,
y(t) = 1 Aekt ,
where A = eC is an arbitrary constant. The initial condition y(0) = 0.1 allows us to determine the value of A:
0.1 = 1 A so A = 0.9.
0.8
0.6
0.4
0.2
2 4 6 8
(c) The model from Exercise 8 seems more realistic because it predicts the rumor starts spreading slowly, picks up speed
and then levels off as we near the time when the entire population has heard the rumor.
11. Let k = 1 and A = 1 in the logistic equation.
(a) Find the solutions satisfying y1 (0) = 10 and y2 (0) = 1.
(b) Find the time t when y1 (t) = 5.
(c) When does y2 (t) become infinite?
SOLUTION The general solution of the logistic equation with k = 1 and A = 1 is
1
y(t) = .
1 et /C
1 10
y1 (t) = = .
1 10 e t 10 9et
9
1
y2 (t) = .
1 2et
1056 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
1 2et = 0 or t = ln 2.
12. Reverse Logistic Equation Consider the logistic equation (with k, B > 0)
dP P
= k P 1 7
dt B
(a) Sketch the slope field of this equation.
(b) The general solution is P(t) = B/(1 ekt /C), where C is a nonzero constant. Show that P(0) > B if C > 1 and
0 < P(0) < B if C < 0.
(c) Show that Eq. (7) models an extinctionexplosion population. That is, P(t) tends to zero if the initial population
satisfies 0 < P(0) < B and it tends to after a finite amount of time if P(0) > B.
(d) Show that P = 0 is a stable equilibrium and P = B an unstable equilibrium.
SOLUTION
(a) The slope field of this equation is shown below.
y
3
2
1
0 t
1
2
3
3 2 1 0 1 2 3
1
(b) Suppose that C > 0. Then 1 C1 < 1, 1 C1 > 1, and
B
P(0) = > B.
1 C1
1
On the other hand, if C < 0, then 1 C1 > 1, 0 < 1 C1 < 1, and
B
0 < P(0) = < B.
1 C1
(c) From part (b), 0 < P(0) < B when C < 0. In this case, 1 ekt /C is never zero, but
ekt
1
C
as t . Thus, P(t) 0 as t . On the other hand, P(0) > B when C > 0. In this case 1 ekt /C = 0 when
t = k1 ln C. Thus,
1
P(t) as t ln C.
k
S E C T I O N 10.3 The Logistic Equation 1057
(d) Let
P
F(P) = k P 1 .
B
Then, F (P) = k + 2kBP . Thus, F (0) = k < 0, and F (B) = k + 2k = k > 0, so P = 0 is a stable equilibrium
and P = B is an unstable equilibrium.
13. A tissue culture grows until it has a maximum area of M cm2 . The area A(t) of the culture at time t may be modeled
by the differential equation
A
A=k A 1 8
M
where k is a growth constant.
(a) By setting A = u 2 , show that the equation can be rewritten
2
u = 1 k 1 u
2 M
SOLUTION
so that Eq. (8) becomes:
(a) Let A = u 2 . This gives A = 2u u,
u2
2u u = ku 1
M
2
u = k 1 u
2 M
Now, rewrite
du k u2 du
= 1 as = 12 k dt.
dt 2 M 1 u 2 /M
The partial fraction decomposition for the term on the left-hand side is
1 M 1 1
= + ,
1 u /M
2 2 M +u M u
so after integrating both sides, we obtain
M M + u 1
ln = kt + C.
2 M u 2
Thus,
M +u
= Ce(k/ M)t
M u
u(Ce(k/ M)t + 1) = M(Ce(k/ M)t 1)
and
Ce(k/ M)t 1
u= M .
Ce(k/ M)t + 1
14. Use the model of Exercise 13 to determine the area A(t) (t in hours) of a tissue culture with initial size
A(0) = 0.2 cm2 , assuming that the maximum area is M = 5 cm2 and the growth constant is k = 0.06. Graph the
solution using a graphing utility.
SOLUTION Substituting M = 5 and k = 0.06 into the result of Exercise 13(b),
2
Ce(0.06/ 5)t 1
A(t) = 5 .
Ce(0.06/ 5)t + 1
The initial condition A(0) = 0.2 gives
C 1 2
0.2 = 5
C +1
so C = 1.5. Hence
2
1.5e(0.06/ 5)t 1
A(t) = 5 .
1.5e(0.06/ 5)t + 1
A
5
4
3
2
1
t
50 100 150 200
15. Show that if a tissue culture grows according to Eq. (8), then the growth rate reaches a maximum when A = M/3.
A
SOLUTION According to Equation (8), the growth rate of the tissue culture is k A(1 M ). Therefore
d A 1 3 1 3A
k A 1 = k A1/2 k A1/2 /M = k A1/2 1 =0
dA M 2 2 2 M
when A = M/3. Because the growth rate is zero for A = 0 and for A = M and is positive for 0 < A < M, it follows
that the maximum growth rate occurs when A = M/3.
16. In 1751, Benjamin Franklin predicted that the U.S. population P(t) would increase with growth constant k =
0.028 yr1 . According to the census, the U.S. population was 5 million in 1800 and 76 million in 1900. Assuming
logistic growth with k = 0.028, what is the predicted carrying capacity for the U.S. population? Hint: Use Eqs. (3) and
(4) to show that
P(t) P0
= ekt
P(t) A P0 A
P(t)
= Cekt .
P(t) A
But
P0
C= ,
P0 A
so
P(t) P0
= ekt .
P(t) A P0 A
Now, let t = 0 correspond to the year 1800. Then the year 1900 corresponds to t = 100, and with k = 0.028, we have
76 5
= e(0.028)(100) .
76 A 5 A
Solving for A, we find
5(e2.8 1)
A= 5 943.07.
2.8 1
76 e
Thus, the predicted carrying capacity for the U.S. population is approximately 943 million.
S E C T I O N 10.3 The Logistic Equation 1059
d2 y
= k 2y 1 y 1
2y
dt 2 A A
(b) Show that y(t) is concave up if 0 < y < A/2 and concave down if A/2 < y < A.
(c) Show that if 0 < y(0) < A/2, then y(t) has a point of inflection at y = A/2 (Figure 5).
(d) Assume that 0 < y(0) < A/2. Find the time t when y(t) reaches the inflection point.
SOLUTION
(a) The derivative of Eq. (10) with respect to t is
2kyy 2y y 2y y 2y
y = ky = ky 1 =k 1 ky 1 = k2 y 1 1 .
A A A A A A
18. Let
A
y=
1 ekt /C
y
A
A
2
Inflection point
y(0)
t
FIGURE 5 Inflection point in a logistic curve occurs at y = A/2.
be the general nonequilibrium solution of y = ky(1 y/A) with k > 0. If y(t) has a vertical asymptote at t = tb , that
is, if lim y(t) = , we say that the solution blows up at t = tb .
ttb
(a) Show that if 0 < y(0) < A, then y does not blow up at any time tb .
(b) Show that if y(0) > A, then y blows up at a time tb , which is negative (and hence does not correspond to a real
time).
(c) Show that y blows up at some positive time tb if and only if y(0) < 0 (and hence does not correspond to a real
population).
1060 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
SOLUTION
(a) Let y(0) = y0 . From the general solution, we find
A 1 A y0
y0 = ; 1 = ; so C = .
1 1/C C y0 y0 A
If y0 < A, then C < 0, and the denominator in the general solution, 1 ekt /C, is always positive. Thus, when
0 < y(0) < A, y does not blow up at any time.
(b) 1 ekt /C = 0 when C = ekt . Solving for t we find
1
t = ln C.
k
Because C = y yA0
and y0 > A, it follows that C > 1, and thus, ln C > 0. Therefore, y blows up at a time which is
0
negative.
(c) Suppose that y blows up at some tb > 0. From part (b), we know that
1
tb = ln C.
k
Thus, in order for tb to be positive, we must have ln C < 0, which requires C < 1. Now,
y0
C= ,
y0 A
so tb > 0 if and only if
y0 y0 A A
<1 or equivalently =1 > 1.
y0 A y0 y0
This last inequality holds if and only if y0 = y(0) < 0.
Exercises
1. Consider y + x 1 y = x 3 .
(a) Verify that (x) = x is an integrating factor.
(b) Show that when multiplied by (x), the differential equation can be written (x y) = x 4 .
(c) Conclude that x y is an antiderivative of x 4 and use this information to find the general solution.
(d) Find the particular solution satisfying y(1) = 0.
SOLUTION
(a) The equation is of the form
y + A(x)y = B(x)
x y + y = x 4.
(x y) = x 4 .
5
(c) Since (x y) = x 4 , (x y) = x5 + C and
x4 C
y= +
5 x
(d) If y(1) = 0, we find
1 1
0= +C so = C.
5 5
The solution, therefore, is
x4 1
y= .
5 5x
dy
2. Consider + 2y = e3t .
dt
(a) Verify that (t) = e2t is an integrating factor.
(b) Use Eq. (4) to find the general solution.
(c) Find the particular solution with initial condition y(0) = 1.
SOLUTION
(a) The equation is of the form
y + A(t)y = B(t)
1 = 1 + C so 2 = C.
y = e3t + 2e2t .
2
3. Let (x) = e x . Verify the identity
( (x)y) = (x)(y + 2x y)
y + 2x y = x.
2
SOLUTION Let (x) = e x . Then
(x)(y + 2x y) = x (x) = xe x .
2
1062 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
y + A(x)y = B(x)
When multiplied by the integrating factor, the original differential equation becomes
ex y ex y = e x or (ex y) = e x .
Therefore,
y(x) = e2x + Ce x .
1 = 1+C so 0 = C.
Therefore,
y = e2x .
In Exercises 518, find the general solution of the first-order linear differential equation.
5. x y + y = x
SOLUTION Rewrite the equation as
1
y + y = 1,
x
which is in standard linear form with A(x) = 1x and B(x) = 1. By Theorem 1, the integrating factor is
(x) = e A(x) d x = eln x = x.
When multiplied by the integrating factor, the rewritten differential equation becomes
x y + y = x or (x y) = x.
6. x y y = x 2 x
S E C T I O N 10.4 First-Order Linear Equations 1063
1
y y = x 1,
x
which is in standard linear form with A(x) = 1x and B(x) = x 1. By Theorem 1, the integrating factor is
(x) = e A(x) d x = e ln x = x 1 .
When multiplied by the integrating factor, the rewritten differential equation becomes
1 1 1 y 1
y 2y =1 or =1 .
x x x x x
Integration of both sides now yields
y
= x ln x + C.
x
Finally,
y(x) = x 2 x ln x + C x.
7. 3x y y = x 1
SOLUTION Rewrite the equation as
1 1
y y = 2,
3x 3x
which is in standard form with A(x) = 13 x 1 and B(x) = 13 x 2 . By Theorem 1, the integrating factor is
(x) = e A(x) d x = e(1/3) ln x = x 1/3 .
When multiplied by the integrating factor, the rewritten differential equation becomes
1 4/3 1 1 7/3
x 1/3 y x = x 7/3 or (x 1/3 y) = x .
3 3 3
Integration of both sides now yields
1
x 1/3 y = x 4/3 + C.
4
Finally,
1
y(x) = x 1 + C x 1/3 .
4
8. y + x y = x
SOLUTION This equation is in standard form with A(x) = x and B(x) = x. By Theorem 1, the integrating factor is
(x) = e x d x = e(1/2)x .
2
When multiplied by the integrating factor, the original differential equation becomes
e(1/2)x y = e(1/2)x + C.
2 2
Finally,
y(x) = 1 + Ce(1/2)x .
2
9. y + 3x 1 y = x + x 1
1064 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
SOLUTION This equation is in standard form with A(x) = 3x 1 and B(x) = x + x 1 . By Theorem 1, the integrating
factor is
1
(x) = e 3x = e3 ln x = x 3 .
When multiplied by the integrating factor, the original differential equation becomes
x 3 y + 3x 2 y = x 4 + x 2 or (x 3 y) = x 4 + x 3 .
10. y + x 1 y = cos(x 2 )
SOLUTION This equation is in standard form with A(x) = x 1 and B(x) = cos(x 2 ). By Theorem 1, the integrating
factor is
1
(x) = e x d x = eln x = x.
When multiplied by the integrating factor, the original differential equation becomes
11. x y = y x
SOLUTION Rewrite the equation as
1
y y = 1,
x
which is in standard form with A(x) = 1x and B(x) = 1. By Theorem 1, the integrating factor is
(x) = e (1/x) d x = e ln x = x 1 .
When multiplied by the integrating factor, the rewritten differential equation becomes
1 1 1 1 1
y 2y = or y = .
x x x x x
Integration on both sides now yields
1
y = ln x + C.
x
Finally,
y(x) = x ln x + C x.
3y
12. x y = x 3
x
S E C T I O N 10.4 First-Order Linear Equations 1065
3
y + 2 y = x 2,
x
which is in standard form with A(x) = 3x 2 and B(x) = x 2 . By Theorem 1, the integrating factor is
2
(x) = e 3x d x = e3/x .
When multiplied by the integrating factor, the rewritten differential equation becomes
3
e3/x y + 2 e3/x y = x 2 e3/x or (e3/x y) = x 2 e3/x .
x
Integration on both sides now yields
e3/x y = x 2 e3/x d x + C.
Finally,
y(x) = e3/x x 2 e3/x d x + Ce3/x .
13. y + y = e x
SOLUTION This equation is in standard form with A(x) = 1 and B(x) = e x . By Theorem 1, the integrating factor is
(x) = e 1 d x = e x .
When multiplied by the integrating factor, the original differential equation becomes
When multiplied by the integrating factor, the original differential equation becomes
or
Finally,
x cos x + C
y(x) = .
sec x + tan x
SOLUTION This equation is in standard form with A(x) = tan x and B(x) = cos x. By Theorem 1, the integrating
factor is
(x) = e tan x d x = eln sec x = sec x.
When multiplied by the integrating factor, the original differential equation becomes
y sec x = x + C.
Finally,
16. e2x y = 1 e x y
SOLUTION Rewrite the equation as
y + ex y = e2x ,
which is in standard form with A(x) = ex and B(x) = e2x . By Theorem 1, the integrating factor is
x
x
(x) = e e d x = ee .
When multiplied by the integrating factor, the rewritten differential equation becomes
x x x x x
ee y + exe y = e2x ee or (ee y) = e2x ee .
Finally,
x
y(x) = 1 + ex + Cee .
x x e x y = e x + C.
Finally,
y(x) = x x + C x x ex .
18. y + y = cos x
S E C T I O N 10.4 First-Order Linear Equations 1067
SOLUTION This equation is in standard form with A(x) = 1 and B(x) = cos x. By Theorem 1, the integrating factor
is
(x) = e 1 d x = e x .
When multiplied by the integrating factor, the original differential equation becomes
Integration on both sides (integration by parts is needed on the right-hand side of the equation) now yields
1 x
ex y = e (sin x + cos x) + C.
2
Finally,
1
y(x) = (sin x + cos x) + Cex .
2
(x) = e3x .
When multiplied by the integrating factor, the original differential equation becomes
20. x y + y = e x , y(1) = 3
SOLUTION First, we find the general solution of the differential equation. Rewrite the equation as
1 1
y + y = ex ,
x x
which is in standard form with A(x) = x 1 and B(x) = x 1 e x . By Theorem 1, the integrating factor is
1
(x) = e x d x = eln x = x.
When multiplied by the integrating factor, the rewritten differential equation becomes
(x y) = e x .
x y = e x + C;
hence,
1 x C
y(x) = e + .
x x
1068 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
1
21. y + y = x 2 , y(1) = 2
x +1
SOLUTION First, we find the general solution of the differential equation. This linear equation is in standard form with
1 and B(x) = x 2 . By Theorem 1, the integrating factor is
A(x) = x+1
(x) = e 1/(x+1) d x = eln(x+1) = x + 1.
When multiplied by the integrating factor, the original differential equation becomes
((x + 1)y) = x 1 + x 2 .
(x + 1)y = ln x x 1 + C;
hence,
1 1
y(x) = C + ln x .
x +1 x
The initial condition y(1) = 2 allows us to determine the value of C:
1
2= (C 1) so C = 5.
2
The solution to the initial value problem is therefore
1 1
y(x) = 5 + ln x .
x +1 x
When multiplied by the integrating factor, the original differential equation becomes
(e x y) = e x sin x.
Integration on both sides (integration by parts is needed on the right-hand side of the equation) now yields
1 x
(e x y) = e (sin x cos x) + C;
2
hence,
1
y(x) = (sin x cos x) + Cex .
2
The initial condition y(0) = 1 allows us to determine the value of C:
1 3
1= +C so C= .
2 2
The solution to the initial value problem is therefore
1 3
y(x) = (sin x cos x) + ex .
2 2
SOLUTION First, we find the general solution of the differential equation. Rewrite the equation as
which is in standard form with A(x) = cot x and B(x) = csc x. By Theorem 1, the integrating factor is
(x) = e cot x d x = e ln sin x = csc x.
When multiplied by the integrating factor, the rewritten differential equation becomes
hence,
When multiplied by the integrating factor, the original differential equation becomes
hence,
C
y(t) = 1 + .
sec t + tan t
The initial condition y(0) = 1 allows us to determine the value of C:
C
1=1+ so C = 0.
1+0
The solution to the initial value problem is therefore
y(x) = 1.
We find the same particular solution with the initial condition y( /4) = 1.
25. y + (tanh x)y = 1, y(0) = 3
SOLUTION First, we find the general solution of the differential equation. This equation is in standard form with
A(x) = tanh x and B(x) = 1. By Theorem 1, the integrating factor is
(x) = e tanh x d x = eln cosh x = cosh x.
When multiplied by the integrating factor, the original differential equation becomes
(cosh x y) = sinh x + C;
1070 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
hence,
3 = C.
x 1
26. y + y= , y(1) = 0
1 + x2 (1 + x 2 )3/2
SOLUTION First, we find the general solution of the differential equation. This equation is in standard form with
A(x) = x and B(x) = 1 . By Theorem 1, the integrating factor is
1+x 2 (1+x 2 )3/2
(x) = e (x/(1+x )) d x = e(1/2) ln(1+x ) = 1 + x 2 .
2 2
When multiplied by the integrating factor, the original differential equation becomes
1
1 + x2 y = .
1 + x2
Integration on both sides now yields
1 + x 2 y = tan1 x + C;
hence,
tan1 x C
y(x) = + .
1+x 2 1 + x2
The initial condition y(1) = 0 allows us to determine the value of C:
1
0= +C so C = .
2 4 4
The solution to the initial value problem is therefore
1
y(x) = tan1 x .
1 + x2 4
27. Find the general solution of y + ny = emx for all m, n. Note: The case m = n must be treated separately.
SOLUTION For any m, n, Theorem 1 gives us the formula for (x):
(x) = e n d x = enx .
When multiplied by the integrating factor, the original differential equation becomes
If m
= n, integration on both sides yields
1
enx y = e(m+n)x + C,
m+n
so
1
y(x) = emx + Cenx .
m+n
However, if m = n, then m + n = 0 and the equation reduces to
(enx y) = 1,
so integration yields
28. A 200-gal tank contains 100 gal of water with a salt concentration of 0.1 lb/gal. Water with a salt concentration of
0.4 lb/gal flows into the tank at a rate of 20 gal/min. The fluid is mixed instantaneously, and water is pumped out at a
rate of 10 gal/min. Let y(t) be the amount of salt in the tank at time t.
(a) Set up and solve the differential equation for y(t).
(b) What is the salt concentration when the tank overflows?
SOLUTION Because water flows into the tank at the rate of 20 gal/min but flows out at the rate of 10 gal/min, there is
a net inflow of 10 gal/min. Therefore, at any time t, there are 100 + 10t gallons of water in the tank.
(a) The net flow of salt into the tank at time t is
dy gal lb gal y lb 1
= salt rate in salt rate out = 20 0.4 10 =8 y.
dt min gal min 100 + 10t gal 10 + t
Rewriting this linear equation in standard form, we have
dy 1
+ y = 8,
dt 10 + t
1 and B(t) = 8. By Theorem 1, the integrating factor is
so A(t) = 10+t
1
(t) = e (10+t) dt = eln(10+t) = 10 + t.
When multiplied by the integrating factor, the rewritten differential equation becomes
hence,
80t + 4t 2 + C
y(t) = .
10 + t
The initial condition y(0) = 10 allows us to determine the value of C:
C
10 = so C = 100.
10
The solution to the initial value problem is therefore
80t + 4t 2 + 100
y(t) = .
10 + t
(b) The tank overflows when t = 10. The amount of salt in the tank at that time is
800 + 400 + 100
y(10) = = 65 lb,
20
so the concentration of salt is
65 lb
= 0.325 lb/gal.
200 gal
29. Repeat Exercise 28(a), assuming that water is pumped out at a rate of 20 gal/min. What is the limiting salt concen-
tration for large t?
SOLUTION Because water flows into the tank at the same rate as water flows out of the tank, the amount of water in
the tank remains a constant 100 gallons. Now, let y(t) be the amount of salt in the tank (in pounds) at any time t (in
minutes). The net flow of salt into the tank t is
dy gal lb gal y lb 1
= salt rate in salt rate out = 20 0.4 20 = 8 y.
dt min gal min 100 gal 5
Rewriting this linear equation in standard form, we have
dy 1
+ y = 8,
dt 5
so A(t) = 15 and B(t) = 8. By Theorem 1, the integrating factor is
(t) = e (1/5) dt = et/5 .
1072 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
When multiplied by the integrating factor, the rewritten differential equation becomes
et/5 y = 40et/5 + C;
hence,
y(t) = 40 + Cet/5 .
10 = 40 + C so C = 30.
y(t) = 40 30et/5 .
From here, we see that y(t) 40 as t , so that the limiting concentration of salt in the tank is 100 40 = 0.4 pounds
per gallon. This is completely intuitive; as time goes on, the concentration in the tank should resemble the concentration
pouring into it.
30. Repeat Exercise 28(a), assuming that water is pumped out at a rate of 25 gal/min. What is the limiting salt concen-
tration for t large?
SOLUTION Because water flows into the tank at the rate of 20 gal/min but flows out at the rate of 25 gal/min, there is
a net outflow of 5 gal/min. Therefore, at any time t, there are 100 5t gallons of water in the tank. The net flow of salt
into the tank t is then
dy gal lb gal y lb 5
= salt rate in salt rate out = 20 0.4 25 =8 y.
dt min gal min 100 5t gal 20 t
Rewriting this linear equation in standard form, we have
dy 5
+ y = 8,
dt 20 t
5 and B(t) = 8. By Theorem 1, the integrating factor is
so A(t) = 20t
1
(t) = e5 (20t) dt = e5 ln(20t) = (20 t)5 .
When multiplied by the integrating factor, the rewritten differential equation becomes
hence,
20
31. Water flows into a tank at the variable rate Rin = gal/min and out at the constant rate Rout = 5 gal/min. Let
1+t
V (t) be the volume of water in the tank at time t.
(a) Set up a differential equation for V (t) and solve it with the initial condition V (0) = 100.
(b) Find the maximum value of V .
(c) Plot V (t) and estimate the time t when the tank is empty.
SOLUTION
(a) The rate of change of the volume of water in the tank is given by
dV 20
= Rin Rout = 5.
dt 1+t
Because the right-hand side depends only on the independent variable t, we integrate to obtain
V (t) = 20 ln(1 + t) 5t + C.
100 = 20 ln 1 0 + C so C = 100.
Therefore
when t = 3. Because ddtV > 0 for t < 3 and ddtV < 0 for t > 3, it follows that
120
100
80
60
40 32 34 36 38
20
10 20 30 40
32. A stream feeds into a lake at a rate of 1,000 m3 /day. The stream is polluted with a toxin whose concentration is
5 g/m3 . Assume that the lake has volume 106 m3 and that water flows out of the lake at the same rate of 1,000 m3 /day.
Set up a differential equation for the concentration c(t) of toxin in the lake and solve for c(t), assuming that c(0) = 0.
What is the limiting concentration for large t?
SOLUTION Let M(t) denote the amount of toxin, in grams, in the lake at time t. The rate at which toxin enters the lake
is given by
g m3 g
5 3 1000 = 5000 ,
m day day
while the rate at which toxin exits the lake is given by
M(t) g m3 M(t) g
1000 = ,
106 m3 day 1000 day
where we have assumed that any toxin in the lake is spread uniformly throughout the lake. The differential equation for
M(t) is then
dM M
= 5000 .
dt 1000
1074 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
M(t)
The concentration of the toxin in the lake is given by c(t) = , so c (t) = 16 M (t), giving
106 10
dc 1 1
= c.
dt 200 1000
Rewriting this linear equation in standard form, we have
dc 1 1
+ c= ,
dt 1000 200
1 and B(t) = 1 . By Theorem 1, the integrating factor is
so A(t) = 1000 200
(t) = et/1000 .
When multiplied by the integrating factor, the rewritten differential equation becomes
1 t/1000
(et/1000 c) = e .
200
Integration on both sides now yields
et/1000 c = 5et/1000 + A;
hence,
c(t) = 5 + Aet/1000 .
0=5+ A so A = 5.
Therefore
c(t) = 5 1 et/1000 grams/m3 .
grams
As t , c(t) 5, so the limiting concentration of pollution is 5 .
m3
In Exercises 3335, consider a series circuit (Figure 4) consisting of a resistor of R ohms, an inductor of L henries, and
variable voltage source of V (t) volts (time t in seconds). The current through the circuit I (t) (in amperes) satisfies the
differential equation:
dI R 1
+ I = V (t) 12
dt L L
V(t) L
FIGURE 4 RL circuit.
33. Find the solution to Eq. (12) with initial condition I (0) = 0, assuming that R = 100 , L = 5 H, and V (t) is
constant with V (t) = 10 V.
SOLUTION If R = 100, V (t) = 10, and L = 5, the differential equation becomes
dI
+ 20I = 2,
dt
which is a linear equation in standard form with A(t) = 20 and B(t) = 2. The integrating factor is (t) = e20t , and
when multiplied by the integrating factor, the differential equation becomes
(e20t I ) = 2e20t .
hence,
1
I (t) = + Ce20t .
10
The initial condition I (0) = 0 allows us to determine the value of C:
1 1
0= +C so C = .
10 10
Finally,
1
I (t) = 1 e20t .
10
0.006
0.004
0.002
t
0 0.5 1 1.5 2 2.5 3
is a linear equation in standard form with A(t) = LR and B(t) = L1 V (t). By Theorem 1, the integrating factor is
(t) = e (R/L) dt = e(R/L) t .
When multiplied by the integrating factor, the original differential equation becomes
V
(e(R/L) t I ) = e(R/L) t .
L
Integration on both sides now yields
V (R/L) t
(e(R/L) t I ) = e + C;
R
hence,
V
I (t) = + Ce(R/L) t .
R
The initial condition I (0) = 0 allows us to determine the value of C:
V V
0= +C so C = .
R R
Therefore the current is given by
V
I (t) = 1 e(R/L) t .
R
36. Tank 1 in Figure 5 is filled with V1 gallons of water. Water flows into the tank at a rate of R gal/min and
out through the bottom at the same rate R. Suppose that I gallons of blue ink are dumped into the tank at time t = 0 and
mixed instantaneously. Let y1 (t) be the quantity of ink in the tank at time t.
dy R
(a) Explain why y1 satisfies the differential equation 1 = y1 .
dt V1
(b) Solve for y1 (t) with V1 = 100, R = 10, and I = 2.
SOLUTION
(a) After the initial I gallons of blue ink are dumped into Tank 1, no additional blue ink is added. The rate at which blue
ink enters the tank is therefore zero. The rate at which ink leaves the tank is the product of the rate R at which water
leaves the tank and the concentration y1 /V1 of the ink in the water. Thus,
d y1 R
= y1 .
dt V1
S E C T I O N 10.4 First-Order Linear Equations 1077
(et/10 y1 ) = 0.
Integration yields
y1 (t) = 2et/10 .
37. Continuing with the previous exercise, let Tank 2 be another tank filled with V2 gallons of water. Assume
that the inky water from Tank 1 empties into Tank 2 as in Figure 5, mixes instantaneously, and leaves Tank 2 at the same
rate R. Let y2 (t) be the amount of ink in Tank 2 at time t.
R (gal/min)
Tank 1
R (gal/min)
Tank 2
R (gal/min)
FIGURE 5
(b) With V1 = 100, R = 10 and I = 2, we know from the previous exercise that y1 (t) = 2et/10 . Substituting this
expression and the given parameter values into the differential equation obtained in part (a), we have
d y2 2 t/10 1 1 1
= 10 e y2 = et/10 y .
dt 100 200 5 20 2
Hence,
d y2 1 1
+ y2 = et/10 .
dt 20 5
The integrating factor for this linear equation is (t) = et/20 so we get
1
et/20 y2 = et/20 ,
5
and
et/20 y2 = 4et/20 + C.
1078 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
Thus,
t = 20 ln 2.
(a) Remember that (x) = A(x) (x). Now, let y(x) = ( (x))1 . Then
1 A(x) 1 A(x)
y + A(x)y = (x) + = A(x) (x) + = 0.
( (x))2 (x) ( (x)) 2 (x)
(b) Suppose f (x) satisfies f (x) + A(x) f (x) = B(x). Now, let y(x) = f (x) + C/ (x), where C is an arbitrary
constant. Then
C C A(x)
y + A(x)y = f (x) (x) + A(x) f (x) +
( (x))2 (x)
C (x)
= f (x) + A(x) f (x) + A(x) = B(x) + 0 = B(x).
(x) (x)
39. Use the Fundamental Theorem of Calculus and the Product Rule to verify directly that for any x0 , the function
x
f (x) = (x)1 (t)B(t) dt
x0
40. Transient Currents Suppose the circuit described by Eq. (12) is driven by a sinusoidal voltage source V (t) =
V sin t (where V and are constant).
(a) Show that
V
I (t) = (R sin t L cos t) + Ce(R/L) t
R2 + L 2 2
(b) Let Z = R 2 + L 2 2 . Choose so that Z cos = R and Z sin = L . Use the addition formula for the sine
function to show that
V
I (t) = sin( t ) + Ce(R/L) t
Z
This shows that the current in the circuit varies sinusoidally apart from a DC term (called the transient current in
electronics) that decreases exponentially.
SOLUTION
(a) With V (t) = V sin t, the equation
dI R 1
+ I = V (t)
dt L L
becomes
dI R V
+ I = sin t.
dt L L
This is a linear equation in standard form with A(t) = LR and B(t) = VL sin t. By Theorem 1, the integrating factor is
(t) = e A(t) dt = e(R/L) t .
(a) y 5 is a nonlinear term involving the dependent variable, so this is not a linear equation; the highest order derivative
that appears in the equation is a first derivative, so this is a first-order equation.
(b) This is linear equation; the highest order derivative that appears in the equation is a first derivative, so this is a
first-order equation.
(c) y is a nonlinear term involving the dependent variable, so this is not a linear equation; the highest order derivative
that appears in the equation is a third derivative, so this is a third-order equation.
(d) This is linear equation; the highest order derivative that appears in the equation is a second derivative, so this is a
second-order equation.
2. Find a value of c such that y = x 2 + ecx is a solution of 2y + y = x.
SOLUTION Let y = x 2 + ecx . Then
y = 1 + cecx ,
and
For this to equal x, we must have 2c + 1 = 0, or c = 12 (remember that ecx is never zero).
y 3 d y = t 2 dt.
y4 t3
= + C.
4 3
Thus,
1/4
4 3
y= t +C ,
3
where C is an arbitrary constant.
4. x yy = 1 x 2
SOLUTION Rewrite the equation
dy 1
xy = 1 x2 as y dy = x d x.
dx x
Upon integrating both sides of this equation, we obtain
1
y dy = x dx
x
y2 x2
= ln |x| + C.
2 2
Thus,
y = ln x 2 + A x 2 ,
dy
5. x y=1
dx
SOLUTION Rewrite the equation as
dy dx
= .
1+ y x
upon integrating both sides of this equation, we obtain
dy dx
=
1+y x
ln |1 + y| = ln |x| + C.
Thus,
y = 1 + Ax,
dy x y2 dy x
= 2 as = 2 d x.
dx x +1 y2 x +1
Upon integrating both sides of this equation, we obtain
dy x
= dx
y2 x2 + 1
1 1
= ln x 2 + 1 + C.
y 2
Thus,
1
y=1
,
2 ln x + 1 + C
2
In Exercises 710, solve the initial value problem using separation of variables.
7. y = cos2 x, y(0) =
4
SOLUTION First, we find the general solution of the differential equation. Because the variables are already separated,
we integrate both sides to obtain
1 1 x sin 2x
y = cos2 x d x = + cos 2x d x = + + C.
2 2 2 4
The initial condition y(0) = 4 allows us to determine C = 4 . Thus, the solution is:
x sin 2x
y(x) = + + .
2 4 4
8. y = cos2 y, y(0) =
4
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy dy
= cos2 y as = d x.
dx cos2 y
Upon integrating both sides of this equation, we find
tan y = x + C;
thus,
y = tan1 (x + C).
1082 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
dy dy
= x y2 as = x d x.
dx y2
Upon integrating both sides of this equation, we find
dy
= x dx
y2
1 1
= x 2 + C.
y 2
Thus,
1
y=1 .
2x + C
2
1 + 2C = 1
C = 1
10. x yy = 1, y(3) = 2
SOLUTION First, we find the general solution of the differential equation. Rewrite
dy dx
xy =1 as y dy = .
dx x
Next we integrate both sides of the equation to obtain
dx
y dy =
x
1 2
y = ln |x| + C.
2
Thus,
y = 2(ln |x| + C).
To satisfy the initial condition y(3) = 2 we must choose the positive square root; moreover,
2 = 2(ln 3 + C) so C = 2 ln 3.
11. Figure 1 shows the slope field for y = sin y + t y. Sketch the graphs of the solutions with the initial conditions
y(0) = 1 , y(0) = 0, and y(0) = 1.
2
1
0 t
1
2
2 1 0 1 2
FIGURE 1
SOLUTION
12. Which of the equations (i)(iii) corresponds to the slope field in Figure 2?
(i) y = 1 y 2
(ii) y = 1 + y 2
(iii) y = y 2
SOLUTION From the figure we see that the the slope is positive even for y > 1, thus, the slope field does not correspond
to the equation y = 1 y 2 . Moreover, the slope at y = 0 is positive, so the slope field also does not correspond to the
equation y = y 2 . The slope field must therefore correspond to (ii): y = 1 + y 2 .
13. Let y(t) be the solution to the differential equation with slope field as shown in Figure 2, satisfying y(0) = 0. Sketch
the graph of y(t). Then use your answer to Exercise 12 to solve for y(t).
2
1
0 t
1
2
2 1 0 1 2
FIGURE 2
SOLUTION As explained in the previous exercise, the slope field in Figure 2 corresponds to the equation y = 1 + y 2 .
The graph of the solution satisfying y(0) = 0 is:
1084 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
To solve the initial value problem y = 1 + y 2 , y(0) = 0, we first find the general solution of the differential equation.
Separating variables yields:
dy
= dt.
1 + y2
Upon integrating both sides of this equation, we find
y1 = y0 + h F(t0 , y0 ) = 1.0375
y2 = y1 + h F(t1 , y1 ) = 1.076580
y3 = y2 + h F(t2 , y2 ) = 1.117318
y4 = y3 + h F(t3 , y3 ) = 1.159798
y5 = y4 + h F(t4 , y4 ) = 1.204112
y6 = y5 + h F(t5 , y5 ) = 1.250361
15. Let y(t) be the solution of (x 3 + 1) y = y satisfying y(0) = 1. Compute approximations to y(0.1), y(0.2), and
y(0.3) using Eulers Method with time step h = 0.1.
SOLUTION Rewriting the equation as y = 3
y
we have F(x, y) = 3y . Using Eulers Method with x0 = 0, y0 = 1
x +1 x +1
and h = 0.1, we calculate
1
y(0.1) y1 = y0 + h F(x 0 , y0 ) = 1 + 0.1 3 = 1.1
0 +1
y(0.2) y2 = y1 + h F(x 1 , y1 ) = 1.209890
y(0.3) y3 = y2 + h F(x 2 , y2 ) = 1.329919
y y = t 2,
which is in standard form with A(t) = 1 and B(t) = t 2 . The integrating factor is
(t) = e 1 dt = et .
Chapter Review Exercises 1085
When multiplied by the integrating factor, the rewritten differential equation becomes
(et y) = t 2 et .
Integration on both sides (integration by parts is needed for the integral on the right-hand side of the equation) now yields
et y = et (t 2 + 2t + 2) + C;
hence,
y(t) = Cet t 2 2t 2.
4 = 2 + C so C = 6.
y = 6et t 2 2t 2.
dy y
17. = + x, y(1) = 3
dx x
SOLUTION First, we find the general solution of the differential equation. Rewrite the equation as
1
y y = x,
x
which is in standard form with A(x) = 1x and B(x) = x. The integrating factor is
1 1
(x) = e x d x = e ln x = .
x
When multiplied by the integrating factor, the rewritten differential equation becomes
1
y = 1.
x
Integration on both sides now yields
1
y = x + C;
x
hence,
y(x) = x 2 + C x.
3 = 1+C so C = 2.
y = x 2 + 2x.
dy
18. = y 3t, y(1) = 2
dt
SOLUTION First, we find the general solution of the differential equation. Rewrite the equation as
y y = 3t,
which is in standard form with A(t) = 1 and B(t) = 3t. The integrating factor is
(t) = e A(t) dt = e dt = et .
When multiplied by the integrating factor, the rewritten differential equation becomes
Integration on both sides (integration by parts is needed for the integral on the right-hand side of the equation) now yields
et y = (3t + 3)et + C;
1086 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
hence,
y(t) = 3t + 3 + Cet .
y = 2e et + 3t + 3 = 2et+1 + 3t + 3.
19. y + 2y = 1 + ex , y(0) = 4
SOLUTION The equation is already in standard form with A(x) = 2 and B(x) = 1 + ex . The integrating factor is
(x) = e 2 d x = e2x .
When multiplied by the integrating factor, the original differential equation becomes
20. x 2 y = x 2 + 1, y(1) = 10
SOLUTION First, we find the general solution of the differential equation. Rewrite the equation as
1
y = 1 + 2 .
x
Because the variables have already been separated, we integrate both sides to obtain
1 1
y= 1 + 2 d x = x + C.
x x
The initial condition y(1) = 10 allows us to determine the value of C:
10 = 1 1 + C so C = 10.
When multiplied by the integrating factor, the original differential equation becomes
1
y = cos x.
cos x
Integration on both sides now yields
1
y = sin x + C;
cos x
hence,
1
y(x) = sin x cos x + C cos x = sin 2x + C cos x.
2
The initial condition y( ) = 2 allows us to determine the value of C:
2 = 0 + C(1) so C = 2.
22. x y = 2y + x 1, y( 32 ) = 9
SOLUTION First, we find the general solution of the differential equation. This is a linear equation which can be
rewritten as
2 1
y y =1 .
x x
3 1 9 40
9= + + C so C= .
2 2 4 9
The solution to the initial value problem is then
40 2 1
y= x x+ .
9 2
(y 1) d y = t dt.
y2 1
y = t2 + C
2 2
y 2 2y + 1 = t 2 + C
(y 1)2 = t 2 + C
y(t) = t 2 + C + 1
To satisfy the initial condition y(1) = 3 we must choose the negative square root; moreover,
3 = 1 + C + 1 so C = 15.
y + 1 y = ytet ,
2
24. y(0) = 1.
SOLUTION First, we find the general solution of the differential equation. This is a separable equation that we rewrite
as
1 1 2
+ d y = tet dt.
y y
dw 1 + w2
25. =k , w(1) = 1
dx x
SOLUTION First, we find the general solution of the differential equation. This is a separable equation that we rewrite
as
dw k
= d x.
1 + w2 x
Upon integrating both sides of this equation, we find
dw k
= dx
1+w 2 x
tan1 w = k ln x + C
w(x) = tan(k ln x + C).
Because the initial condition is specified at x = 1, we are interested in the solution for x > 0; we can therefore omit the
absolute value within the natural logarithm function. The initial condition w(1) = 1 allows us to determine the value of
C:
1 = tan(k ln 1 + C) so C = tan1 1 = .
4
The solution to the initial value problem is then
w = tan k ln x + .
4
3y 1
26. y + =t +2
t
Chapter Review Exercises 1089
3 1
y + y =t +2+ .
t t
When multiplied by the integrating factor, the rewritten differential equation becomes
(t 3 y) = t 4 + 2t 3 + t 2 .
y
27. y + = sin x
x
SOLUTION This is a first order linear equation with A(x) = 1x and B(x) = sin x. The integrating factor is
(x) = e A(x) d x = eln x = x.
When multiplied by the integrating factor, the original differential equation becomes
(x y) = x sin x.
Integration on both sides (integration by parts is needed for the integral on the right-hand side) now yields
x y = x cos x + sin x + C;
hence,
sin x C
y(x) = cos x + + .
x x
28. State whether the differential equation can be solved using separation of variables, the method of integrating factors,
both, or neither.
(a) y = y + x 2 (b) x y = y + 1
(c) y = y 2 + x 2 (d) x y = y 2
SOLUTION
(a) The equation y = y + x 2 is a first order linear equation; hence, it can be solved by the method of integration factors.
However, it cannot be written in the form y = f (x)g(y); therefore, separation of variables cannot be used.
(b) The equation x y = y + 1 is a first order linear equation; hence, it can be solved using the method of integration
factors. We can rewrite this equation as y = 1x (y + 1); therefore, it can also be solved by separating the variables.
(c) The equation y = y 2 + x 2 cannot be written in the form y = f (x)g(y); hence, separation of variables cannot be
used. This equation is also not linear; hence, the method of integrating factors cannot be used.
(d) The equation x y = y 2 can be rewritten as y = 1x y 2 ; therefore, it can be solved by separating the variables. Since
it is not a linear equation, the method of integrating factors cannot be used.
dy
29. Let A and B be constants. Prove that if A > 0, then all solution of + Ay = B approach the same limit as t .
dt
SOLUTION This is a linear first-order equation in standard form with integrating factor
(t) = e A dt = e At .
When multiplied by the integrating factor, the original differential equation becomes
(e At y) = Be At .
1090 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
8 ft
4 ft
6 ft
12 ft
FIGURE 3
dy Bv(y)
= ,
dt A(y)
where v(y) denotes the velocity of the water flowing through the hole when the trough is filled to height y, B denotes the
area of the hole and A(y) denotes the area of the horizontal cross section of the trough at height y. By Torricellis Law,
2
v(y) = 8 y ft/s. The area of the hole is B = 12 3 = 16 ft2 . The horizontal cross section of the trough at height
y is a rectangle of length 12 and width w(y). As w(y) varies linearly from 6 when y = 0 to 8 when y = 4, it follows that
y
w(y) = + 6;
2
therefore, the area of horizontal cross section at height y is
y
A(y) = 12w(y) = 12 + 6 = 6y + 72 = 6(y + 12).
2
The differential equation for y(t) then becomes
dy y
= .
dt 12(y + 12)
Rewrite the above equation as
y + 12
d y = dt,
y 12
(w(y) + 6)y
12 = 168
2
y
+ 12 y = 28
2
y 2 + 24y 56 = 0
The roots of this quadratic
equation are y = 12 10 2. We eliminate the negative root, and thus the trough is half
empty when y = 10 2 12. The height of the water takes on this value when
3/2 1/2
10 2 12 + 36 10 2 12 = t + 80.
8
Solving for t, we find t 61.56 seconds.
31. Find the solution of the logistic equation y = 0.4y(4 y) satisfying y(0) = 8.
SOLUTION We can write the given equation as
y
y = 1.6y 1 .
4
This is a logistic equation with k = 1.6 and A = 4. Therefore,
A 4
y(t) = = .
1 ekt /C 1 e1.6t /C
The initial condition y(0) = 8 allows us to determine the value of C:
4 1 1
8= ; 1 = ; so C = 2.
1 C1 C 2
Thus,
4 8
y(t) = = .
1 e1.6t /2 2 e1.6t
32. Let y(t) be the solution of y = 0.3y(2 y) with y(0) = 1. Determine lim y(t) without solving for y explicitly.
t
SOLUTION We write the given equation in the form
y
y = 0.6y 1 .
2
This is a logistic equation with A = 2 and k = 0.6. Because the initial condition y(0) = y0 = 1 satisfies 0 < y0 < A,
the solution is increasing and approaches A as t . That is, lim y(t) = 2.
t
33. Suppose that y = ky(1 y/8) has a solution satisfying y(0) = 12 and y(10) = 24. Find k.
SOLUTION The given differential equation is a logistic equation with A = 8. Thus,
8
y(t) = .
1 ekt /C
The initial condition y(0) = 12 allows us to determine the value of C:
8 1 2
12 = ; 1 = ; so C = 3.
1 C1 C 3
Hence,
8 24
y(t) = kt
= .
1 e /3 3 ekt
Now, the condition y(10) = 24 allows us to determine the value of k:
24
24 =
3 e10k
1092 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
3 e10k = 1
ln 2
k= 0.0693.
10
34. A lake has a carrying capacity of 1,000 fish. Assume that the fish population grows logistically with growth constant
k = 0.2 days1 . How many days will it take for the population to reach 900 fish if the initial population is 20 fish?
SOLUTION Let y(t) represent the fish population. Because the population grows logistically with k = 0.2 and A =
1000,
1000
y(t) = .
1 e0.2t /C
The initial condition y(0) = 20 allows us to determine the value of C:
1000 1 1
20 = ; 1 = 50; so C = .
1 C1 C 49
Hence,
1000
y(t) = .
1 + 49e0.2t
The population will reach 900 fish when
1000
= 900.
1 + 49e0.2t
Solving for t, we find
35. A rabbit population on a deserted island increases exponentially with growth rate k = 0.12 months1 . When
the population reaches 150 rabbits (say, at time t = 0), hunters begin killing the rabbits at a rate of r rabbits per month.
(a) Find a differential equation satisfied by the rabbit population P(t).
(b) How large can r be without the rabbit population becoming extinct?
SOLUTION
(a) The rabbit population P(t) obeys the differential equation
dP
= 0.12P r,
dt
where the term 0.12P accounts for the exponential growth of the population and the term r accounts for the rate of
decline in the rabbit population due to hunting.
(b) Rewrite the linear differential equation from part (a) as
dP
0.12P = r,
dt
which is in standard form with A = 0.12 and B = r . The integrating factor is
(t) = e A dt = e 0.12 dt = e0.12t .
When multiplied by the integrating factor, the rewritten differential equation becomes
sin1 y = tan1 x + C
Thus,
To express the solution in the required form, we use the addition formula
This yields
1 + x2 x
tan1 x
37. A tank contains 100 gal of pure water. Water is pumped out at a rate of 20 gal/min, and polluted water with
a toxin concentration of 0.1 lb/gal is pumped in at a rate of 15 gal/min. Let y(t) be the amount of toxin present in the
tank at time t.
1094 C H A P T E R 10 I N T R O D U C T I O N TO D I F F E R E N T I A L E Q U ATI O N S
5
4
3
2
1
5 10 15 20
Chapter Review Exercises 1095
To determine the time at which the amount of toxin is maximal, we note that
dy d 20 t (20 t)4 1 (20 t)3
= = + =0
dt dt 2 16000 2 4000
when
3
t = 20 10 2 7.4 minutes.