Professional Documents
Culture Documents
Ordinary Differential Equations - Solution Manual
Ordinary Differential Equations - Solution Manual
Davidson
ORDINARY DIFFERENTIAL
EQUATIONS
Student Solution Manual
January 23, 2012
Springer
Chapter 1
Solutions
Section 1.1
1. The rate of change in the population P (t) is the derivative P (t). The
Malthusian Growth Law states that the rate of change in the population is
proportional to P (t). Thus P (t) = kP (t), where k is the proportionality
constant. Without reference to the t variable, the differential equation
becomes P = kP
11. The highest order derivative is y so the order is 3. Solving for y gives
the standard form: y = 2y 3y + y.
3
4 1 Solutions
19.
y (t) = 3ce3t
3y + 12 = 3(ce3t 4) + 12 = 3ce3t 12 + 12 = 3ce3t .
21.
1 Solutions 5
cet
y (t) =
(1 cet )2
1 1 1 (1 cet ) cet
y 2 (t) y(t) = = = .
(1 cet )2 1 cet (1 cet )2 (1 cet )2
If c 0 then the denominator 1 cet > 0 and y(t) has domain R. If c > 0
then 1 cet = 0 if t = ln 1c = ln c. Thus y(t) is defined either on the
interval (, ln c) or ( ln c, ).
23.
cet
y (t) =
cet 1
t 1 cet
ey 1 = e ln(ce 1)
1= 1 = .
cet 1 cet 1
Since c > 0 then y(t) is defined if and only if cet 1 > 0. This occurs
if et > 1t which is true if t > ln 1c = ln c. Thus y(t) is defined on the
interval ( ln c, ).
25.
1
y (t) = (c t)2 (1) =
(c t)2
1
y 2 (t) = .
(c t)2
Section 1.2
1. y = t
5
0
y
5
5 4 3 2 1 0 1 2 3 4 5
t
3. y = y(y + t)
5
0
y
5
5 4 3 2 1 0 1 2 3 4 5
t
5. 5
0
y
5
5 4 3 2 1 0 1 2 3 4 5
t
1 Solutions 7
7. 5
y
1
5
5 4 3 2 1 0 1 2 3 4 5
t
9. 5
0
y
5
5 4 3 2 1 0 1 2 3 4 5
t
19. Differentiation gives y = 3ct2 + 2t. However, from the given function
2
we have ct3 = y t2 and hence ct2 = yt
t . Substitution gives y =
2
3 yt
t + 2t = 3y
t t.
8 1 Solutions
Section 1.3
13. The variables are already separated, so integrate both sides to get y 5 /5 =
t2 /2 + 2t + c, c a real constant. Simplifying gives y 5 = 25 t2 + 10t + c. We
leave the answer in implicit form
17. There are two equilibrium solutions; y = 0 and y= 4. Separating vari-
1 1 1
ables and using partial fractions gives 4 y + 4y dy = dt. Integrating
y y
and simplifying gives ln 4y = 4t+ k1 which is equivalent to 4y = ce4t ,
4t
4ce
c a nonzero constant. Solving for y gives y = 1+ce 4t . When c = 0 we get
33. Let m denote the number of Argon-40 atoms in the sample. Then 8m is
the number of Potassium-40 atoms. Let t be the age of the rock. Then
t years ago there were m + 8m = 9m atoms of Potassium-40. Hence
N (0) = 9m. On the other hand, 8m = N (t) = N (0)et = 9met . This
ln 8
implies that 89 = et and hence t = 9 = ln 8
2 ln 9 212 million years
old.
or
80 = ke5r
72 = ke4r .
41. Let us start time t = 0 at 1980. Then P (0) = 290. The Malthusian growth
model gives P (t) = 290ert. At t = 10 (1990) we have 370 = 290e10r
1
and hence r = 10 ln 37
29 . At t = 30 (2010) we have P (30) = 290e
30r
=
37
3
290 29 602.
ln 3
43. We have 3P (0) = P (5) = P (0)e3r . So r = 5 . Now we solve the equation
ln 2 5 ln 2
2P (0) = P (t) = P (0)ert for t. We get t = r = ln 3 3.15 years.
47. We have P (0) = P0 = 400, P (3) = P1 = 700, and P (6) = P2 = 1000. Us-
ing the result of the previous problem we get m = 700(700(400+1000)24001000)
(700)2 4001000 =
1, 400
Section 1.4
and the left hand side of this equation is a perfect derivative, namely,
(e3t y) . Thus, (e3t y) = e4t . Now take antiderivatives of both sides and
multiply by e3t . This gives
1 t
y= e + ce3t
4
for the general solution of the equation. To find the constant c to satisfy
the initial condition y(0) = 2, substitute t = 0 into the general solution
12 1 Solutions
1
to get 2 = y(0) = 4 + c. Hence c = 49 , and the solution of the initial
value problem is
1 t 9 3t
y= e e .
4 4
e2t y 2e2t y = 1,
the left hand side of which is a perfect derivative ((e2t )y) . Thus
((e2t )y) = 1 and taking antiderivatives of both sides gives
(e2t )y = t + c,
y = te2t + 4e2t .
et
5. The general solution from Problem 4 is y = t + ct . Now let t = 1 to get
t
0 = e + c. So c = e and y = et et .
ty + y = t cos(t2 ),
the left hand side of which is a perfect derivative (ty) . Thus (ty) =
t cos(t2 ) and taking antiderivatives of both sides gives ty = 21 sin(t2 ) + c
sin(t2 )
where c R is a constant. Now divide by t to get y = 2t + ct . for the
general solution.
R
9. In this case p(t) = 3 and the integrating factor is e 3 dt = e3t .
Now multiply to get e3t y 3e3t y = 25e3t cos 4t, which simplifies
to (e3t y) = 25e3t cos 4t. Now integrate bothR sides to get e3t y =
(4 sin 4t 3 cos 4t)e3t + c, where we computed 25e3t cos 4t by parts
twice. Dividing by e3t gives y = 4 sin 4t 3 cos 4t + ce3t .
11. In
R standard form we get z 2tz = 2t3 . An integrating factor is
2t dt t2 2 2
e = e . Thus (et z) = 2t3 et . Integrating both sides gives
1 Solutions 13
2 2
et z = (t2 + 1)et + c, where the integral of the right hand side is done
2 2
by parts. Now divide by the integrating factor et to get z = t2 +1+cet .
the left hand side of which is a perfect derivative ((e sin t )y) . Thus
(t2 y) = t2 + t3
2
and taking antiderivatives of both sides gives (t2 )y = t1 t 2 + c
where c R is a constant. Now multiply by t2 to and we get y = t
1
2 + ct
2
for the general solution. Letting t = 1 gives 3 = y(1) = 3
2 +c
3
so c = 2 and
1 3
y(t) = t t2 .
2 2
1 bt
y= e + ceat
a+b
for the general solution.
19. In standard form we get y (tan t)y = sec t. In this case p(t) =
tan t, an antiderivative is P (t) = ln cos t, and the integrating factor
is (t) = eP (t) = cos t. Now multiply by the integrating factor to get
(cos t)y (sin t)y = 1, the left hand side of which is a perfect deriva-
tive ((cos t)y) . Thus ((cos t)y) = 1 and taking antiderivatives of both
sides gives (cos t)y = t + c where c R is a constant. Now multiply by
1/ cos t = sec t and we get y = (t + c) sec t for the general solution.
1 2 9 3
y= t t .
5 5
and
1
y= + (4a 2)t2 .
t
27. Let V (t) denote the volume of fluid in the tank at time t. Initially, there
are 10 gal of brine. For each minute that passes there is a net decrease of
4 3 = 1 gal of brine. Thus V (t) = 10 t gal.
gal lbs lbs
input rate: input rate = 3 1 =3 .
min gal min
gal y(t) lbs 4y(t) lbs
output rate: output rate = 4 = .
min V (t) gal 10 t min
Since y = input rateoutput rate, it follows that y(t) satisfies the initial
value problem
4
y = 3 y(t) , y(0) = 2.
10 t
Put in standard form, this equation becomes
4
y + y = 3.
10 t
4
R
The coefficient function is p(t) = 10t , P (t) = p(t) dt = 4 ln(10 t) =
ln(10 t)4 , and the integrating factor is (t) = (10 t)4 . Multiplying
the standard form equation by the integrating factor gives
29. Let V (t) denote the volume of fluid in the container at time t. Initially,
there are 10 L. For each minute that passes there is a net gain of 42 = 2
L of fluid. So V (t) = 10 + 2t. The container overflows when V (t) =
10 + 2t = 30 or t = 10 minutes.
L g g
input rate: input rate = 4 20 = 80 .
min L min
L y(t) g 2y(t) g
output rate: output rate = 2 = .
min 10 + 2t L 10 + 2t min
Since y = input rateoutput rate, it follows that y(t) satisfies the initial
value problem
2y
y = 80 , y(0) = 0.
10 + 2t
16 1 Solutions
33. Let y1 (t) and y2 (t) denote the amount of salt in Tank 1 and Tank 2,
respectively, at time t. The volume of fluid at time t in Tank 1 is V1 (t) =
10 + 2t and Tank 2 is V2 (t) = 5 + t.
L g g
input rate for Tank 1: input rate = 4 10 = 40 .
min L min
1 Solutions 17
L y1 (t) g
output rate for Tank 1: output rate = 2 =
min 10 + 2t L
2y(t) g
. The initial value problem for Tank 1 is thus
10 + 2t min
2
y1 = 40 y1 , y1 (0) = 0.
10 + 2t
Simplifying this equation and putting it in standard form gives
1
y1 + y1 = 40.
5+t
The integrating factor is (t) = 5 + t. Thus ((5 + t)y1 ) = 40(5 + t).
Integrating and simplifying gives y1 (t) = 20(5 + t) + c/(5 + t). The initial
condition y(0) = 0 implies c = 500 so y1 = 20(5 + t) 500/(5 + t).
L y1 (t) g
input rate for Tank 2: input rate = 2 = 20
min 10 + 2t L
500 g
.
(5 + t)2 min
L y2 (t) g y2 (t) g
output rate for Tank 2: output rate = 1 = .
min 5 + t L 5 + t min
The initial value problem for Tank 2 is thus
1
y2 = 20 500/(5 + t)2 y2 , y2 (0) = 0.
(5 + t)
500 ln(5 + t) c
y2 (t) = 10(5 + t) + .
5+t 5+t
The initial condition y2 (0) = 0 implies c = 500 ln 5 250 so
Section 1.5
y 2 + yt + t2
1. In standard form we get y = which is homogeneous since
t2
the degrees of the numerator and denominator are each two. Let y = tv.
Then v + tv = v 2 + v + 1 and so tv = v 2 + 1. Separating variables gives
dv dt
= . Integrating gives tan1 v = ln |t| + c. So v = tan(ln |t| + c).
v2 + 1 t
Substituting v = y/t gives y = t tan(ln |t| + c). The initial condition
implies 1 = y(1) = 1 tan c = tan c and hence c = /4. Therefore y(t) =
t tan(ln |t| + /4).
2 z
19. Let z = 2t2y +1. Then z = 22y and so y = . Substituting we
2
2 z
get = z 1 and in standard form we get z = 2 2z 1 , a separable
2
differential equation. Clearly, z = 1 is an equilibrium solution. Assume
for now that z 6= 1. Then separating variables
and simplifying using
1 z 1 1
1/(1 z ) = z1 = 1 + z1 gives 1 + z1 dz = 2 dt. Integrating we
get z + ln |z 1| = 2t + c. Now substitute z = 2t 2y + 1 and simplify to
get 2y + ln |2t 2y| = c, c R. (We absorb the constant 1 in c.) The
equilibrium solution z = 1 becomes y = t.
23. This is the same as Exercise 16 where the Bernoulli equation technique
there used the substitution z = y 2 . Here use the given substitution to
get z = 2yy + 1. Substituting we get z 1 = z and in standard form
z = 1+z. Clearly, z = 1 is an equilibrium solution. Separating variables
dz
gives = dt and integrating gives ln |1 + z| = t + c, c R. Solving
1+z
for z we get z = ket 1, where k 6= 0. Since z = y 2 + t 1 we get
y 2 + t 1 = ket 1 and solving for y gives t
y = ke t. The case
k = 0 gives the equilibrium solutions y = t.
y
25. If z = ln y then z = . Divide the given differential equation by y.
y
y
Then + ln y = t and substitution gives z + z = t. An integrating factor
y
1 Solutions 21
Section 1.6
1. This can be written in the form M (t, y) + N (t, y)y = 0 where M (t, y) =
y 2 + 2t and N (t, y) = 2ty. Since M/y = 2y = N/t, the equation is
exact, and the general solution is given implicitly by V (t, y) = c where the
function V (t, y) Ris determined by the solution method for exact equations.
Thus V (t, y) = (y 2 + 2t) dt + (y) = y 2 t + t2 + (y). The function (y)
satisfies
V 2 d d
= (y t + t2 ) + = 2ty + = N (t, y) = 2ty,
y y dy dy
7. This can be written in the form M (t, y) + N (t, y)y = 0 where M (t, y) =
2ty+2t3 and N (t, y) = t2 y. Since M/y = 2t = N/t, the equation is
exact, and the general solution is given implicitly by V (t, y) = c where the
function V (t, y) Ris determined by the solution method for exact equations.
Thus V (t, y) = (2ty + 2t3 ) dt + (y) = t2 y + t4 /2 + (y). The function
(y) satisfies
V 2 d d
= (t y + t4 /2) + = t2 + = N (t, y) = t2 y,
y y dy dy
9. This can be written in the form M (t, y) + N (t, y)y = 0 where M (t, y) =
y and N (t, y) = y 3 t. Since M/y = 1 = N/t, the equation is
exact, and the general solution is given implicitly by V (t, y) = c where the
function V (t, y)Ris determined by the solution method for exact equations.
Thus V (t, y) = (y) dt + (y) = yt + (y). The function (y) satisfies
22 1 Solutions
V d d
= (yt) + = t + = N (t, y) = y 3 t,
y y dy dy
Section 1.7
1. We first change the variable t to u and write y (u) = uy(u). Now integrate
Rt Rt
both sides from 1 to t to get 1 y (u) du = 1 uy(u) du. Now the left side
Rt Rt
is 1 y (u) du = y(t) y(1) = y(t) 1. Thus y(t) = 1 + 1 uy(u) du.
u y(u)
3. Change the variable t to u and write y (u) = . Now integrate
u + y(u)
Rt R t u y(u)
both sides from 0 to t to get 0 y (u) du = 0 du. The left side
u + y(u)
R t u y(u)
is y(t) 1 so y(t) = 1 + 0 du.
u + y(u)
Rt
5. The corresponding integral equation is y(t) = 1 + 1 uy(u) du. We then
have
y0 (t) = 1
t t
u2 t2 1 + t2
1
Z
y1 (t) = 1 + u 1 du = 1 + = 1 + =
1 2 1 2 2 2
Z t t
1 + u2 u2 u4 5 t2 t4
y2 (t) = 1 + u du = 1 + + = + +
1 2 4 8 1 8 4 8
Z t 3 5 6 t
2 4
5u u u 5u u u
y3 (t) = 1 + + + du = 1 + + +
1 8 4 8 16 16 48 1
29 5t2 t4 t6
= + + + .
48 16 16 48
Rt
7. The corresponding integral equation is y(t) = 0 (u + y 2 (u)) du. We then
have
1 Solutions 23
y0 (t) = 0
Z t
t2
y1 (t) = (u + 0) du =
0 2
Z t 2 2 ! Z t
u4 t2 t5
u
y2 (t) = u+ du = u+ du = +
0 2 0 4 2 20
Z t 2 2 ! Z t
u5 u4 u7 u10
u
y3 (t) = u+ + du = u+ + + du
0 2 20 0 4 20 400
t2 t5 t8 t11
= + + + .
2 20 160 4400
We then have
y0 (t) = 0
t t
u3 t3
Z
1 + (u 0)2 du =
y1 (t) = u+ = t+
0 3 0 3
Z t 3
2 ! Z t
u6
u
y2 (t) = 1+ u u+ du = 1+ du
0 3 0 9
t
u7 t7
= u+ =t+
63 0 7 32
Z t
u14 t15
y3 (t) = 1 + 2 4 du = t +
0 7 3 15 72 34
Z t 30
t31
u
y4 (t) = 1 + 2 4 8 du = t +
0 15 7 3 31 152 74 38
Z t 62
t63
u
y5 (t) = 1+ 2 du = t +
0 31 154 78 31 6 63 312 154 78 316
11. The right hand side is F (t, y) = y. If R is any rectangle about (1, 0)
then there are y-coordinates that are negative. Hence F is not defined on
R and Picards theorem does not apply.
ty 2t
13. The right hand side is F (t, y) = . Then Fy (t, y) = . Choose
t+y (t + y)2
a rectangle R about (0, 1) that contains no points on the line t + y = 0.
Then both F and Fy are continuous on R. Picards theorem applies and
we can conclude there is a unique solution on an interval about 0.
24 1 Solutions
Rt
15. The corresponding integral equation is y(t) = 1 + 0
ay(u) du. We thus
have
y0 (t) = 1
Z t
y1 (t) = 1 + a du = 1 + at
0
t t
a 2 t2
Z Z
y2 (t) = 1 + a(1 + au) du = 1 + (a + a2 u) du = 1 + at +
0 0 2
Z t 2 2
2 2 3 3
a u a t a t
y3 (t) = 1 + a 1 + au + du = 1 + at + +
0 2 2 3!
..
.
a 2 t2 a n tn
yn (t) = 1 + at + + + .
2 n!
Pn a k tk
We can write yn (t) = k=0 . We recognize this sum as the first n
k!
terms of the Taylor series expansion for eat . Thus the limiting function
is y(t) = limn yn (t) = eat . It is straightforward to verify that it is a
solution. If F (t, y) = ay then Fy (t, y) = a. Both F and Fy are continuous
on the whole (t, y)-plane. By Picards theorem, Theorem 5, y(t) = eat is
the only solution to the given initial value problem.
17. Let F (t, y) = cos(t + y). Then Fy (t, y) = sin(t + y). Let y1 and
y2 be arbitrary real numbers. Then by the mean value theorem there
is a number y0 in between y1 and y2 such that |F (t, y1 ) F (t, y2 )| =
|sin(t + y0 )| |y1 y2 | |y1 y2 | . It follows that F (t, y) is Lipschitz on
any strip. Theorem 10 implies there is a unique solution on all of R.
6 4 2 0 2 4 6
t
21. No. Both y1 (t) and y2 (t) would be solutions to the initial value problem
y = F (t, y), y(0) = 0. If F (t, y) and Fy (t, y) are both continuous near
(0, 0), then the initial value problem would have a unique solution by
Theorem 5.
23. For t < 0 we have y1 (t) = 0 and for t > 0 we have y1 (t) = 3t2 . For t = 0
y1 (h) y1 (0) y1 (h)
we calculate y1 (0) = limh0 = limh0 . To compute
h0 h
this limit we show the left hand and right hand limits agree. We get
y1 (h) h3
lim+ = lim+ = lim+ h2 = 0
h0 h h0 h h0
y1 (h) 0
lim = lim+ = 0
h0 h h0 h
(
0, for t < 0
It follows that y1 (t) = and so
3t2 for t 0
(
0, for t < 0
ty1 (t) =
3t3 for t 0
Section 2.1
1. Apply the Laplace transform to both sides of the equation. For the left
hand side we get sY (s) 2 4Y (s), while the right hand side is 0. Solve
2
for Y (s) to get Y (s) = . From this we see that y(t) = 2e4t .
s4
3. Apply the Laplace transform to both sides of the equation. For the left
hand side we get sY (s) 4Y (s), while the right hand side is 1/(s 4).
1
Solve for Y (s) to get Y (s) = . Therefore, y(t) = te4t .
(s 4)2
5. Apply the Laplace transform to both sides of the equation. For the left
hand side we get sY (s) 2 + 2Y (s), while the right hand side is 3/(s 1).
Solve for Y (s) to get
2 3 1 1
Y (s) = + = + .
s + 2 (s 1)(s + 2) s+2 s1
7. Apply the Laplace transform to both sides. For the left hand side we get
Hence,
3s + 3 3(s + 1) 3
Y (s) = = = ,
s2 + 3s + 2 (s + 1)(s + 2) s+2
and therefore, y(t) = 3e2t .
9. Apply the Laplace transform to both sides. For the left hand side we get
1 Solutions 27
We now get
(s2 + 25)Y (s) s + 1 = 0.
Hence,
s1 s 1 5
Y (s) = 2
= 2 2
,
s + 25 s + 25 5 s + 25
1
and therefore, y(t) = cos 5t 5 sin 5t.
11. Apply the Laplace transform to both sides. For the left hand side we get
We now get
(s + 4)2 Y (s) (s + 4) = 0.
Hence,
s+4 1
Y (s) = =
(s + 4)2 s+4
and therefore y(t) = e4t
13. Apply the Laplace transform to both sides. For the left hand side we get
1
(s + 2)2 Y (s) 1 = .
s+2
Hence,
1 1 1 1 2
Y (s) = 2
+ 3
= 2
+
(s + 2) (s + 2) (s + 2) 2 (s + 2)3
Section 2.2
1.
L {3t + 1} (s)
Z
= (3t + 1)est dt
0
Z Z
st
= 3 te dt + est dt
0 0
1 st
t st 1 st
Z
= 3 e + e dt + e
s 0 s 0 s 0
1 1 st 1
= 3 e +
s s 0 s
3 1
= + .
s2 s
3.
5
5. L 5e2t = 5L e2t =
s2
2 5 4
7. L t2 5t + 4 = L t2 5L {t} + 4L {1} = 3 2 +
s s s
1 7
9. L e3t + 7te4t = L e3t + 7L te4t =
+
s + 3 (s + 4)2
s 2 s+2
11. L {cos 2t + sin 2t} = L {cos 2t}+L {sin 2t} = + 2 = 2
s2 +2 2 s +2 2 s +4
2
13. L (te2t )2 (s) = L t2 e4t (s) =
(s + 4)3
1 Solutions 29
21. Here we use the transform derivative principle twice to get L t2 sin 2t (s) =
12s2 16
2 4s
(L {sin 2t}) = = =
s2 + 4 (s2 + 4)2 (s2 + 4)3
2
s 2s 2
23. L {tf (t)} (s) = L {f (t)} (s) = ln 2
= 2
s +1 s +1 s
1 1 ln((s/6) + 1) ln(s + 6) ln 6
25. L {Ei(6t)} (s) = L {Ei(t)} (s)|s7s/6 = =
6 6 s/6 s
1 1
27. We use the identity sin2 = (1cos 2). L sin2 bt (s) = L {1 cos 2bt} (s) =
2 2
2b2
1 1 s
= ;
2 s s2 + 4b2 s(s2 + 4b2 )
1
29. We use the identity sin at cos bt = (sin(a + b)t + sin(a b)t).
2
1
L {sin at cos bt} = (L {sin(a + b)t} + L {sin(a b)t})
2
1 ab a+b
= + .
2 s2 + (a b)2 s2 + (a + b)2
1 1 1 1 b
L ebt ebt =
31. L {sinh bt} = =
2 2 s+b sb s2 b2
33. Let f (t) = sinh bt. Then f (t) = b cosh t and f (t) = b2 sinh t. Fur-
ther, f (t)|t=0 = 0 and f (t)|t=0 = b. Thus b2 L {sinh bt} = L {f (t)} =
s2 L {f (t)} sf (0) f (0) = s2 L {f (t)} b. Solving for L {f (t)} gives
b
L {sinh bt} = 2 .
s b2
Rt R0
35. Let g(t) = 0 f (u) du and note that g (t) = f (t) and g(0) = 0 f (u) du =
0. Now apply the input derivative formula to g(t), to get
39. Suppose a and K are real and |y(t)| Keat . Then y(t)eat is bounded
by K. But
2 2 2 a2
at+ a4
et eat = et e 4
a 2 a2
= e(t 2 ) e 4
2 a2
= eu e 4 ,
2
where u = t 2a . As t approaches infinity so does u. Since limu eu =
2
it is clear that limt et eat = , for all a R, and hence y(t)eat is
not bounded. It follows that y(t) is not of exponential type.
2
Since y(t) = sin(et ) is bounded and y(0) = 0 it follows that
M
lim y(t)est 0 = 0.
M
Section 2.3
The s 1 -chain
5s + 10 3
1. (s 1)(s + 4) s1
2
s+4
The s 5 -chain
1 1/7
3. (s + 2)(s 5) (s 5)
1/7
(s + 2)
32 1 Solutions
The s 1 -chain
3s + 1 2
5. (s 1)(s2 + 1) s1
2s + 1
s2 + 1
The s + 3 -chain
2
s +s3 3
(s + 3)3 (s + 3)3
s2 5
7. (s + 3)2 (s + 3)2
1 1
s+3 s+3
0
The s + 1 -chain
s 1
(s + 2)2 (s
+ 1)2 (s + 1)2
9. s+4 3
(s + 2)2 (s + 1) s+1
3s 8
(s + 2)2
1 Solutions 33
The s 5 -chain
1 1
(s 5)5 (s 6) (s 5)5
1 1
(s 5)4 (s 6) (s 5)4
1 1
11. (s 5)3 (s 6) (s 5)3
1 1
(s 5)2 (s 6) (s 5)2
1 1
(s 5)(s 6) s5
1
s6
s2 + s 3
where A1 = =3
1
s=3
1 1
and p1 (s) = (s2 + s 3 (3)(1)) = (s2 + s 6) = s 2
s+3 s+3
Continuing gives
34 1 Solutions
s2 A2 p2 (s)
= +
(s + 3)2 (s + 3)2 s+3
s 2
where A2 = = 5
1 s=3
1 1
and p2 (s) = (s 2 (5)(1)) = (s + 3) = 1
s+3 s+3
s2 + s 3 3 5 1
Thus 3
= 3
2
+
(s + 3) (s + 3) (s + 3) s+3
Alternate Solution: Write s = (s + 3) 3 so that
s2 6s + 7 s2 6s + 7
25. 2 2
= , so use Theorem 1 to compute the
(s 4s 5) (s + 1)2 (s 5)2
(s + 1)-chain:
s2 6s + 7 A1 p1 (s)
2 2
= +
(s + 1) (s 5) (s + 1)2 (s + 1)(s 5)2
s2 6s + 7
7
where A1 = 2
=
(s 5)
s=1 18
1
and p1 (s) = (s2 6s + 7 (7/18)(s 5)2 )
s+1
1 1
= (11s2 38s 49)/18 = (11s 49)
s+1 18
Continuing gives
1 11s 49 A2 p2 (s)
= +
18 (s + 1)(s 5)2 s+1 (s 5)2
1 11s 49
where A2 = = 5/54
18 (s 5)2 s=1
1
and p2 (s) = ((11s 49)/18 (5/54)(s 5)2 ) = (5s 22)/54
s+1
1 Solutions 35
s A1 p1 (s)
= +
(s + 2)2 (s + 1)2 (s + 2)2 (s + 2)(s + 1)2
s
where A1 = = 2
(s + 1)2 s=2
1
and p1 (s) = (s (2)(s + 1)2 )
s+2
2s2 + 5s + 2 (2s + 1)(s + 1)
= = = 2s + 1
s+2 s+2
Continuing gives
2s + 1 A2 p2 (s)
= +
(s + 2)(s + 1)2 s+2 (s + 1)2
2s + 1
where A2 = = 3
(s + 1)2 s=2
1
and p2 (s) = (2s + 1 (3)(s + 1)2 ) = 3s + 2
s+2
s 2 3 3s + 2
Thus = + . Now continue using
(s + 2)2 (s + 1)2 (s + 2)2 s + 2 (s + 1)2
Theorem 1 or replace s by (s + 1) 1 in the numerator of the last fraction
s 2 3 1 3
to get 2 2
= 2
2
+
(s + 2) (s + 1) (s + 2) s + 2 (s + 1) s+1
29. Use Theorem 1 to compute the (s 3)-chain:
36 1 Solutions
8s A1 p1 (s)
= +
(s 1)(s 2)(s 3)3 (s 3)3 (s 1)(s 2)(s 3)2
8s
where A1 = = 12
(s 1)(s 2) s=3
1
and p1 (s) = (8s (12)(s 1)(s 2))
s3
12s2 + 44s 24 (12s + 8)(s 3)
= = = 12s + 8
s3 s3
For the second step in the (s 3)-chain:
12s + 8 A2 p2 (s)
= +
(s 1)(s 2)(s 3)2 (s 3)2 (s 1)(s 2)(s 3)2
12s + 8
where A2 = = 14
(s 1)(s 2) s=3
1
and p2 (s) = (12s + 8 (14)(s 1)(s 2))
s3
14s2 54s + 36 (14s 12)(s 3)
= = = 14s 12
s3 s3
Continuing gives
14s 12 A3 p3 (s)
= +
(s 1)(s 2)(s 3)2 s3 (s 1)(s 2)
14s 12
where A3 = = 15
(s 1)(s 2) s=3
1
and p3 (s) = (14s 12 (15)(s 1)(s 2)) = 15s + 14
s3
8s 12 14 15 15s + 14
Thus 3
= 3
2
+ + .
(s 1)(s 3)(s 3) (s 3) (s 3) s 3 (s 1)(s 2)
The last fraction has a denominator with distinct linear factors so we get
8s 12 14 15 16 1
= + + + +
(s 1)(s 3)(s 3)3 (s 3)3 (s 3)2 s3 s2 s1
31. Use Theorem 1 to compute the (s 2)-chain:
1 Solutions 37
s A1 p1 (s)
= +
(s 2)2 (s 3)2 (s 2)2 (s 2)(s 3)2
s
where A1 = =2
(s 3)2 s=2
1
and p1 (s) = (s (2)(s 3)2 )
s2
2s2 + 13s 18 (2s + 9)(s 2)
= = = 2s + 9
s2 s2
Continuing gives
2s + 9 A2 p2 (s)
= +
(s 2)(s 3)2 s3 (s 3)2
2s + 9
where A2 = =5
(s 3)2 s=2
1
and p2 (s) = (2s + 9 (5)(s 3)2 ) = 5s + 18
s2
s 2 5 5s + 18
Thus = + + . Now continue
(s 2)2 (s
3)2 (s 2)2 s2 (s 3)2
using Theorem 1 or replace s by (s 3) + 3 in the numerator of the last
s 2 5 3 5
fraction to get = + +
(s 2)2 (s 3)2 (s 2)2 s 2 (s 3)2 s3
33. Apply the Laplace transform to both sides. For the left hand side we get
L {y + 2y + y} = L {y } + 2L {y } + L {y}
= s2 Y (s) sy(0) y (0) + 2(sY (s) y(0)) + Y (s)
= (s2 + 2s + 1)Y (s).
9
Since L 9e2t =
we get
s2
9
Y (s) = .
(s + 1)2 (s 2)
The (s + 1) -chain
9 3
(s + 1)2 (s 2) (s + 1)2
3 1
(s + 1)(s 2) s+1
1
(s 2)
It follows that
3 1 1
Y (s) = +
(s + 1)2 s+1 s2
and
y(t) = 3tet et + e2t .
35. Apply the Laplace transform to both sides. For the left hand side we get
L {y 4y 5y} = L {y } 4L {y } 5L {y}
= s2 Y (s) sy(0) y (0) 4(sY (s) y(0)) 5Y (s)
= (s2 4s 5)Y (s) + s 5.
For the second term we start with the s-chain to get the following partial
fraction decomposition
1 Solutions 39
The s -chain
150 30
s2 (s + 1)(s 5) s2
30(s 4) 24
s(s + 1)(s 5) s
244s + 124 25
(s + 1)(s 5) s+1
1
s5
Thus
30 24 26 1
Y (s) = + +
s2 s s+1 s5
and Table 2.2 gives y(t) = 30t + 24 26et + e5t .
37. Apply the Laplace transform to both sides. For the left hand side we get
L {y 3y + 2y} = L {y } 3L {y } + 2L {y}
= s2 Y (s) sy(0) y (0) 3(sY (s) y(0)) + 2Y (s)
= (s2 3s + 2)Y (s) 2s + 3.
Section 2.4
1 B1 s + C1 p1 (s)
= + 2
(s2 + 1)2 (s2 + 2) (s2 + 1)2 (s + 1)(s2 + 2)
1 1
where B1 i + C1 = 2
= 2 =1
(s + 2) s=i i +2
B1 = 0 and C1 = 1
1
and p1 (s) = 2 (1 (1)(s2 + 2))
s +1
s2 1
= = 1.
s2 + 1
1
We now apply Theorem 1 on the remainder term .
(s2 + 1)(s2 + 2)
1 B2 s + C2 p2 (s)
= + 2
(s2 + 1)(s2 + 2) (s2 + 1) (s + 2)
1
where B2 i + C2 = = 1
(s2 + 2) s=i
B2 = 0 and C2 = 1
1
and p2 (s) = 2 (1 (1)(s2 + 2))
s +1
s2 + 1
= 2 = 1.
s +1
Thus the (s2 + 1)-chain is
The s2 + 1 -chain
1 1
(s2 + 1)2 (s2 + 2) (s2+ 1)2
1 1
(s2 + 1)(s2 + 2) (s2 + 1)
1
2
s +2
3. Note that s = 3i is a root of s2 + 3. Applying Theorem 1 gives
1 Solutions 41
8s + 8s2 B1 s + C1 p1 (s)
= + 2
(s2 + 3)3 (s2 + 1) (s2 + 3)3 (s + 3)2 (s2 + 1)
8s + 8s2 8 3i + 8( 3i)2
where B1 3i + C1 = =
(s2 + 1) s=3i ( 3i)2 + 1
= 4 3i + 12
B1 = 4 and C1 = 12
1
and p1 (s) = (8s + 8s2 (4s + 12)(s2 + 1))
s2 + 3
4s3 4s2 + 12s 12
= = 4(s 1).
s2 + 3
4s 4
Apply Theorem 1 a second time on the remainder term .
(s2 + 3)2 (s2 + 1)
4s 4 B2 s + C2 p2 (s)
= + 2
(s2 + 3)2 (s2 + 1) (s2 + 3)2 (s + 3)(s2 + 1)
4s 4
where B2 3i + C2 = 2
= 2 3i + 2
(s + 1) s= 3i
B2 = 2 and C2 = 2
1
and p2 (s) = 2
(4s 4 (2s + 2)(s2 + 1))
s +3
2s3 2s2 + 6s 6
= = 2s 2.
s2 + 3
2s 2
A third application of Theorem 1 on the remainder term
(s2 + 3)(s2 + 1)
gives
2s 2 B3 s + C3 p3 (s)
= + 2
(s2 + 3)(s2 + 1) (s2 + 3) (s + 1)
2s 2
where B3 3i + C3 = = 3i + 1
(s2 + 1) s=3i
B3 = 1 and C3 = 1
1
and p3 (s) = (2s 2 (s + 1)(s2 + 1))
s2 + 3
s3 s2 + 3s 3
= = s 1.
s2 + 3
Thus the (s2 + 3)-chain is
42 1 Solutions
The s2 + 3 -chain
8s + 8s2 12 4s
(s+ 3)3 (s2
+ 1) (s2 + 3)3
4(s 1) 2 2s
(s2 + 3)2 (s2 + 1) (s2 + 3)2
2(s 1) 1s
(s2 + 3)(s2 + 1) s2 + 3
s1
s2 + 1
(s2 + 2s + 4)
Now apply Theorem 1 to the remainder term .
(s2 + 2s + 2)(s2 + 2s + 3)2
1 Solutions 43
(s2 + 2s + 4) B2 s + C2 p2 (s)
= +
(s2 + 2s + 2)(s2 + 2s + 3)2 (s2 + 2s + 2) (s2 + 2s + 3)2
(s2 + 2s + 4)
where B2 (1 + i) + C2 = = 2
(s2 + 2s + 3) s=1+i
B2 = 0 and C2 = 2
(s2 + 2s + 4) (2)(s2 + 2s + 3)2
and p2 (s) =
s2 + 2s + 2
(2(s + 1) + 5)((s + 1)2 + 1)
2
=
s2 + 2s + 2
2
= 2s + 4s + 7.
The s2 + 2s + 2 -chain
1 1
(s2+ 2s + 2)2 (s2
+ 2s + 3)2 (s2+ 2s + 2)2
(s2 + 2s + 4) 2
(s2 + 2s + 2)(s2 + 2s + 3)2 s2 + 2s + 2
2s2 + 4s + 7
(s2 + 2s + 3)2
s A1 p1 (s)
= +
(s2 + 1)(s 3) s3 s2 + 1
s 3
where A1 = =
s2 + 1 s=3 10
1 1
and p1 (s) = (s (3/10)(s2 + 1)) = (3s2 + 10s 3)
s3 10(s 3)
3s + 1
=
10
3s + 1
Since the remainder term is already a simple partial fraction,
10(s2 + 1)
we conclude
s 1 3 1 3s
2
= + 2
(s + 1)(s 3) 10 s3 s +1
9s2 B1 s + C1 p1 (s)
= +
(s2 + 4)2 (s2 + 1) (s2 + 4)2 (s2 + 4)(s2 + 1)
9s2
where B1 (2i) + C1 = = 12
s2 + 1 s=2i
B1 = 0 and C1 = 12
1 3(s2 + 4)
and p1 (s) = 2
(9s2 12(s2 + 1)) =
s +4 s2 + 4
= 3
3 B2 s + C2 p2 (s)
= + 2
(s2 + 4)(s2 + 1) 2
s +4 s +1
3
where B2 (2i) + C1 = =1
s2 + 1 s=2i
B2 = 0 and C2 = 1
1
and p2 (s) = 2
(3 (s2 + 1))
s +4
(s2 + 1)
= = 1.
s2 + 1
1
Since the remainder term is a simple partial fraction, we conclude
s2 + 1
that the complete partial fraction decomposition is
9s2 12 1 1
= 2 + 2
(s2 + 4)2 (s2 + 1) (s + 4)2 s +4 s+1
2 A1 p1 (s)
= +
(s2 6s + 10)(s 3) s3 (s2 6s + 10)
2
where A1 = 2
=2
(s 6s + 10) s=3
1 2s2 + 12s 18
and p1 (s) = (2 (2)(s2 6s + 10)) =
s3 s3
= 2s + 6
1 Solutions 45
2s + 6
Since the remainder term is a simple partial fraction, we con-
s2 6s + 10
2
clude that the complete partial fraction decomposition is 2 =
(s 6s + 10)(s 3)
2 6 2s
+
s 3 (s 3)2 + 1
13. Note that s2 4s+8 = (s2)2 +2 so s = 22i are the roots of s2 4s+8.
We will use the root s = 2+2i to compute the (s2 4s+8)-chain. Applying
Theorem 1 gives
25 B1 s + C1
=
(s2 4s + 8)2 (s 1) (s2 4s + 8)2
p1 (s)
+ 2
(s 4s + 8)(s 1)
25
where B1 (2 + 2i) + C1 =
s 1) s=2+2i
25
= = 5 10i
2i + 1
B1 = 5 and C1 = 15
25 (5s + 15)(s 1)
and p1 (s) =
s2 4s + 8
2
(5)(s 4s + 8)
=
s2 4s + 8
= 5.
5
Now apply Theorem 1 to the remainder term .
(s2 4s + 8)(s 1)
5 B2 s + C2 p2 (s)
= +
(s2 4s + 8)(s 1) (s2 4s + 8) s 1
5)
where B2 (2 + 2i) + C2 = = 1 2i
s 1 s=2+2i
B2 = 1 and C2 = 3
5 (3 s)(s 1)
and p2 (s) =
s2 4s + 8
(1)(s2 4s + 8)
=
s2 4s + 8
= 1.
46 1 Solutions
25 5s + 15
Thus the partial fraction expansion is = 2 +
(s2 4s + 8)2 (s 1) (s 4s + 8)2
s + 3 1
+
s2 4s + 8 s 1
15. Note that s2 + 4s + 5 = (s + 2)2 + 1 so s = 2 i are the roots of
s2 + 4s + 5. We will use the root s = 2 + i to compute the (s2 + 4s + 5)-
chain. Applying Theorem 1 gives
s+1 B1 s + C1
=
(s2 + 4s + 5)2 (s2 + 4s + 6)2 (s2+ 4s + 5)2
p1 (s)
+
(s2 + 4s + 5)(s2 + 4s + 6)2
s+1
where B1 (2 + i) + C1 = 2 2
(s + 4s + 6) ) s=2+i
= 1 + i
B1 = 1 and C1 = 1
s + 1 (s + 1)(s2 + 4s + 6)2
and p1 (s) =
s2 + 4s + 5
(s + 1)((s2 + 4s + 6)2 1)
=
s2 + 4s + 5
(s + 1)(s2 + 4s + 7)(s2 + 4s + 5)
=
s2 + 4s + 5
2
= (s + 1)(s + 4s + 7).
(s + 1)(s2 + 4s + 7)
Now apply Theorem 1 to the remainder term .
(s2 + 4s + 5)(s2 + 4s + 6)2 )
(s + 1)(s2 + 4s + 7) B2 s + C2 p2 (s)
= +
(s2 + 4s + 5)(s2 + 4s + 6)2 (s2 + 4s + 5) (s2 + 4s + 6)2
(s + 1)(s2 + 4s + 7)
where B2 (2 + i) + C2 = = 2 2i
(s2 + 4s + 6)2
s=2+i
B2 = 2 and C2 = 2
(s + 1)(s2 + 4s + 7) (2s 2)(s2 + 4s + 6)2
and p2 (s) =
s2 + 4s + 5
(s + 1)(2(s + 4s + 6) + 1)(s2 + 4s + 5)
2
=
s2 + 4s + 5
2
= (s + 1)(2(s + 4s + 6) + 1).
(s + 1)(2(s2 + 4s + 6) + 1) s+1 2s + 2
= 2 + 2
(s2 + 4s + 6)2 (s + 4s + 6)2 s + 4s + 6
17. Apply the Laplace transform to both sides. For the left hand side we get
L {y + 4y + 4y} = L {y } + 4L {y } + 4L {y}
= s2 Y (s) sy(0) y (0) + 4(sY (s) y(0)) + 4Y (s)
= (s2 + 4s + 4)Y (s) 1.
4s 1
(s2 + 4)(s + 2)2 s2 + 4
1
(s + 2)2
Thus
1
Y (s) =
s2 + 4
1
and Table 2.2 gives y(t) = 2 sin 2t
19. Apply the Laplace transform to both sides. For the left hand side we get
48 1 Solutions
L {y + 4y} = L {y } + 4L {y}
= s2 Y (s) sy(0) y (0) + 4Y (s)
= (s2 + 4)Y (s) 1.
3 3/5
(s2 + 9)(s2 + 4) s2 + 9
3/5
s2 + 4
Thus
8 1 3 1 4 2 1 3
Y (s) = 2
2
=
5s +4 5s +9 5 s + 4 5 s2 + 9
2
4 1
and Table 2.2 gives y(t) = 5 sin 2t 5 sin 3t
Section 2.5
2s 5
7. First, we have s2 + 6s + 9 = (s + 3)2 . Partial fractions gives 2 =
s + 6s + 9
11 2 2s 5
2
+ . So L1 = 11te3t + 2e3t
(s + 3) s+3 (s + 3)2
6 6 1 1 6
9. 2 = = + . So L1 =
s + 2s 8 (s 2)(s + 4) s+4 s2 s2 + 2s 8
e2t e4t
2s2 5s + 1
2
1 3 2 1 2s 5s + 1
11. = + + . So L =
(s 2)4 (s 2)4 (s 2)3 (s 2)2 (s 2)4
1 3 2t 3 2 2t
t e + t e + 2te2t
6 2
4s2 4s2
1 1 1 1 1
13. = + + . So L =
(s 1)2 (s + 1)2 (s 1)2 s 1 (s + 1)2 s + 1 (s 1)2 (s + 1)2
tet + et + tet et
8s + 16 4 1 s 2 1 8s + 16
15. 2 = + . So L =
(s + 4)(s 2)2 (s 2)2 s 2 s2 + 4 s2 + 4 (s2 + 4)(s 2)2
4te2t e2t + cos 2t sin 2t
12 6 3 4 1 1 12
17. 2 = 2+ + . So L =
s (s + 1)(s 2) s s s+1 s2 s2 (s + 1)(s 2)
2t t
3 6t + e 4e
2s 2s
19. First we have s2 + 2s + 5 = (s + 1)2 + 4. So = =
s2 + 2s + 5 (s + 1)2 + 4
2(s + 1) 2 s+1 2
=2 . The First Translation prin-
(s + 1)2 + 4 (s + 1)2 + 4 (s + 1) 2+4
1 2s 1 s+1 1 2
ciple gives L = 2L L =
s2 + 2s + 5 (s + 1)2 + 4 (s + 1)2 + 4
t t
2e cos 2t e sin 2t
s1 s4 1 1 s1
21. 2 = +3 . Thus L =
s 8s + 17 (s 4)2 + 1 (s 4)2 + 1 s2 8s + 17
e4t cos t + 3e4t sin t
s1 s1 1 s1
23. 2 = . Thus L = et cos 3t
s 2s + 10 (s 1)2 + 32 s2 2s + 10
1 8s 1 s 8
25. L = 8L = (2t sin 2t) = 2t sin 2t
(s2 + 4)2 (s2 + 22 )2 2 22
29. We first complete the square s2 +8s+17= (s+4)2 +1. By the translation
1 2s 1 (s + 4) 4
principle we get L 2 2
= 2L 2 2
=
(s + 8s + 17) ((s + 4) + 1)
s 1 1 1
2e4t L1 4L 1
= 2e 4t
t sin t 4( (sin t t cos t) =
(s2 + 1)2 (s2 + 1)2 2 2
4te4t cos t + (t 4)e4t sin t
2 2 2
s 2s + 5 =
31. We first complete the square (s 1) + 2 . By the transla-
1 1
tion principle we get L1 2 3
= L 1
2 2 3
=
(s 2s +5) ((s 1) + 2 )
1 1 2
et L1 = et L1
(s2 + 22 )3 2 (s2 + 22 )3
1
= et (3 (2t)2 ) sin 2t 6t cos 2t
282 4
1
(3 4t2 )et sin 2t 6tet cos 2t
=
256
2
s 8s + 17
33. We first complete the square = (s 4)2+ 1. By the transla-
s4 s4
tion principle we get L1 2 4
= L 1
=
(s 8s + 17) ((s 4)2 + 1)4
s 1
e4t L1 = e4t (3t t3 ) sin t 3t2 cos t
(s2 + 1)4 48
1
(t3 + 3t)e4t cos t 3t2 e4t cos t
=
48
35. Apply the Laplace transform to get
4
s2 Y (s) s + 1 + Y (s) = .
s2 + 1
Solving for Y (s) we get
s1 4
Y (s) = + 2 .
s + 1 (s + 1)2
2
1 1/(a b) 1/(b a)
39. Compute the partial fraction = + .
(s a)(s b) sa sb
Then
eat ebt
1 1 1 1/(a b) 1/(b a)
L =L + = + .
(s a)(s b) sa sb ab ba
41. Apply the inverse Laplace transform to the partial fraction expansion
1 1 1 1 1 1 1
= + + .
(s a)(s b)(s c) (a b)(a c) s a (b a)(b c) s b (c a)(c b) (s c)
43. Apply the inverse Laplace transform to the partial fraction expansion
s2 a2 1 b2 1 c2 1
= + + .
(s a)(s b)(s c) (a b)(a c) s a (b a)(b c) s b (c a)(c b) (s c)
49. Apply the inverse Laplace transform to the partial fraction expansion
s2 ((s a) + a)2 1 2a a2
3
= 3
= + 2
+ .
(s a) (s a) s a (s a) (s a)3
Section 2.6
3. q(s) = s2 + 5s = s(s + 5), hence roots of q(s) are 0 and 5 each with
the
multiplicity 1. Thus Bq = 1, e5t
5. q(s) = s2 6s+ 9 = (s 2
3) , hence the root of q(s) is 3 with multiplicity
3t 3t
2. Thus Bq = e , te
9. q(s) = 6s2 11s + 4 = (3s 4)(2s 1) so the roots are 4/3 and 1/2, each
with multiplicity 1. Hence Bq = et/2 , e4t/3
4 12
11. The quadratic formula gives roots = 2 3. Hence Bq =
n o 2
e(2+ 3)t , e(2 3)t
13. q(s) = 4s2 + 12s+ 9 = (2s + 3)2 ; so the root is 3/2 with multiplicity 2
and hence Bq = e3t/2 , te3t/2
15. q(s) = 4s2 + 25 = 4(s2 + (5/2)2 ). Therefore q(s) has complex roots 25 i.
Hence Bq = {cos(5t/2), sin(5t/2)}
21. q(s) =
(s 1)3 has root 1 with multiplicity 3. Hence
Bq = e , tet , t2 et .
t
23. We complete the square to get q(s) = ((s + 2)2 + 1)2 . Thus q(s) has
root 2 i with multiplicity 2 It follows
complex that
Bq = e2t cos t, e2t sin t, te2t cos t, te2t sin t .
Section 2.7
1. Yes.
t
3. Yes; = tet .
et
2 2
5. Yes; t sin(4t ) = t( sin 4t cos 4t).
4 2 2
1 Solutions 53
7. No.
1
9. No; t 2 is not a polynomial.
1
11. No. is not in E.
sin 2t
13. s4 1 = (s2 1)(s2 + 1) = (s 1)(s + 1)(s2 + 1); The roots ate 1, 1,
and i each with multiplicity 1. Hence Bq = {et , et , cos t, sin t}.
15. The roots are 1 with multiplicity 3 and 7 with multiplicity 2. Hence
Bq = et , tet , t2 et , e7t , te7t .
17. The roots are 2 with multiplicity 3 and 2i with multiplicity 2. Hence
Bq = e2t , te2t , t2 e2t , cos 2t, sin 2t, t cos 2t, t sin 2t .
19. We must gather the roots together to get the correct multiplicity. Thus
q(s) = (s 2)2 (s + 3)3 . The roots are 2 with multiplicity 2 and 3 with
multiplicity 3. Hence Bq = e2t , te2t , e3t , te3t , t2 e3t .
21. By completing the square we may write q(s) = (s+4)2 ((s+3)2 +4)2 The
roots are 4 with multiplicity 2 and 3 2i with multiplicity 2. Hence
Bq = e4t , te4t , e3t cos 2t, e3t sin 2t, te3t cos 2t, te3t sin 2t .
23. First observe that s2 +2s+10 = (s+1)2 +32 and hence q(s) = (s3)3 ((s+
1)2 +32 )2 . Theroots are 3 with multiplicity 3 and 13i with multiplicity
2. Thus Bq = e3t , te3t , t2 e3t , et cos 3t, et sin 3t, tet cos 3t, tet sin 3t .
27. s4 +5s2 +6 = (s2 +3)(s2 +2); hence Bq = cos 3t, sin 3t, cos 2t, sin 2t
p1 (s) p2 (s)
29. r1 (s) = with deg p1 (s) < deg q1 (s) and r2 (s) = with
q1 (s) q2 (s)
p1 (s)p2 (s)
deg p2 (s) < deg q2 (s). Thus, r1 (s)r2 (s) = and
q1 (s)q2 (s)
deg(p1 (s)p2 (s) = deg p1 (s)+deg p2 (s) < deg q1 (s)+deg q2 (s) = deg(q1 (s)q2 (s)).
p(s)
31. If r(s) R then r(s) = where deg p(s) = m < n = deg q(s). Then
q(s)
f (t) = ntn1 eat cos bt btn eat sin bt + atn eat cos bt.
Section 2.8
1.
Z t
tt = x(t x) dx
0
Z t
= (tx x2 ) dx
0
2 x=t
x x3
= t
2 3 x=0
t2 t3 t3
= t =
2 3 6
3.
Z t
3 sin t = sin t 3 = (sin x)(3) dx
0
x=t
= 3 cos x|x=0
= 3(cos t cos 0)
= 3 cos t + 3
e2t e4t
e2t e4t =
2 (4)
1 2t
(e e4t ).
6
56 1 Solutions
11.
1 b
L eat sin bt (s) =
s a s2 + b 2
1 b bs + ba
= 2
s + b 2 s a s2 + b 2
1
bL eat (bL {cos bt} + aL {sin bt})
= 2 2
s +b
Thus
1
eat sin bt = (beat b cos bt a sin bt).
a2 + b 2
a2 b2
1
= 2 + 2
b a2 s2 + a2 s + b2
2 2 4
17. f (t) = t2 sin 2t so F (s) = 3 2
= 3 2 .
s s +4 s (s + 4)
6 1 6
19. f (t) = t3 e3t so F (s) = 4
= 4
s s+3 s (s + 3)
2 2 4
21. f (t) = sin 2t sin 2t so F (s) = 2
s2 2 2
+2 s +2 2 (s + 4)2
23.
1 1 1 1
L 2
= L
s 6s + 5 (s 1)(s 5)
1 1 1 1
= L L
s1 s5
= et e5t
et e5t
=
15
1
= (et + e5t )
4
25.
1 s 1 1 1 s
L = L L
(s + 1)2
2 2
s +1 2
s +1
= sin t cos t
1
= t sin t
2
27.
58 1 Solutions
1 2 1 1 1 2
L = L L
(s 3)(s2 + 4) s3 s2 + 4
= e3t sin 2t
1
= (2e3t 2 cos 2t 3 sin 2t)
13
29.
1 1 1 1 1 1
L = L L
(s a)(s b) sa sb
= eat ebt
eat ebt
= .
ab
31.
( )
1 G(s) 1 1 s
L = L {G(s)} L 2
s2 + 2 s2 + 2
= g(t) cos( 2)t
Z t
= g(x) cos 2(t x) dx
0
t
1
Z
L1 = 1 cos x dx
s2 (s2 + 1) 0
= t sin t dx
1 t
1
Z
L1 2
= 1 e3x dx
s (s + 3) 3 0
1 e3t
1
= t
3 3
1
= (3t 1 + e3t ).
9
1 t
1
Z
L1 = 3x 1 + e3x dx
s3 (s + 3) 9 0
2
e3t 1
1 t
= 3 t
9 2 3
1
= (2 6t + 9t2 2e3t )
54
1 1
37. First, L1 = (sin 3t 3t cos 3t). Thus
(s2 + 9)2 54
t
1 1
Z
L1 = (sin 3x 3x cos 3x) dx
s(s2 + 9)2 54 0
t
1 cos 3x cos 3x
= x sin 3x +
54 3 3
0
1 2 cos 3t 2
= t sin 3t +
54 3 3
1
= (2 cos 3t 3t sin 3t + 2).
162
Section 3.1
15. et and e4t are homogeneous solution so yh = c1 et +c2 e4t are homogeneous
solutions for all scalars c1 and c2 . A particular solution is yp = cos 2t.
Thus y(t) = yp (t) + yh (t) = cos 2t + c1 et + c2 e4t where c1 , c2 are arbitrary
constants.
1 = y(0) = 1 + c1 + c2
3 = y (0) = c1 + 4c2 ,
19. L(ert ) = a(ert ) + b(ert ) + cert = ar2 ert + brert + cert = (ar2 + br + c)ert
21. Let t = a be the point 1 and 2 are tangent. Then 1 (a) = 2 (a) and
1 (a) = 2 (a). By the existence and uniqueness theorem 1 = 2 .
Section 3.2
dependent.
linearly independent.
11. Let q(s) = s(s 1)(s + 1). Then Bq = {1, et , et } which is linearly
independent.
1 Solutions 61
follows since t2 et , t3 et , t4 et Bq .
15.
t t ln t
w(t, t ln t) = det = t ln t + t t ln t = t
1 ln t + 1
17.
t10 t20
w(t10 , t20 ) = det = 20t29 10t29 = 10t29
10t9 20t19
19.
21.
t t2 t3
1
0 1 2t 3t2
w(1, t, t2 , t3 ) = det
= 12
0 0 2 6t
0 0 0 6
25c1 + 10c2 = 0
25c2 = 25
a1 = a2
3 = a1
a2 = 3
62 1 Solutions
Section 3.3
1. The
2t characteristic polynomial is q(s) = s2 s 2 = (s 2)(s + 1) so Bq =
t
and the general solution takes the form y(t) = c1 e2t + c2 et ,
e ,e
c1 , c2 R
3. The characteristic
6t 4t polynomial
is q(s) = s2 + 10s + 24 = (s + 6)(s + 4)
so Bq = e , e and the general solution takes the form y(t) =
c1 e6t + c2 e4t , c1 , c2 R
5. The
4tcharacteristic polynomial is q(s) = s2 + 8s + 16 = (s + 4)2 so Bq =
e , te4t and the general solution takes the form y(t) = c1 e4t +
c2 te4t , c1 , c2 R
9. The characteristic
9t 4t polynomial
is q(s) = s2 + 13s + 36 = (s + 9)(s + 4)
so Bq = e , e and the general solution takes the form y(t) =
c1 e9t + c2 e4t , c1 , c2 R
2 2
11. The characteristic polynomial is q(s) = s + 10s + 25 = (s + 5)5tso
5t 5t
Bq = e , te and the general solution takes the form y(t) = c1 e +
c2 te5t , c1 , c2 R
e3t te3t
det 3t = (1 + 3t)e6t 3te6t = e6t . So K = 1.
3e (1 + 3t)e3t
21. Let q(s) = (s 1)2 + 22 = s2 2s + 5. Then Bq = {et cos 2t, et sin 2t}.
et cos 2t et sin 2t
t t
w(e cos 2t, e sin 2t) = det t
e (cos 2t 2 sin 2t) et (sin 2t + 2 cos 2t)
= e2t (sin 2t cos 2t + 2 cos2 2t)
e2t (cos 2t sin 2t 2 sin2 2t)
= 2e2t .
So K = 2.
Section 3.4
Since te2t is the only function in the first set but not in the second
yp (t) = a1 te2t .
5)2 (s2 + 25) so Bqv = e5t , te5t , cos 5t, sin 5t while
5. q(s)v(s)
5t= (s
Bq = e , te5t . Since cos 5t and sin 5t are the only functions in the first
set that are not in the second yp (t) = a1 cos 5t + a2 sin 5t.
7. q(s)v(s) = (s2 + 4)2 so Bqv = {cos 2t, sin 2t, t cos 2t, t sin 2t} while Bq =
{cos 2t, sin 2t}. Since t cos 2t and t sin 2t are the only functions in the first
set that are not in the second yp (t) = a1 t cos 2t + a2 t sin 2t.
9. q(s)v(s) = (s2 + 4s + 5)(s 1)2 so Bqv = et , tet , e2t cos t, e2t sin t
while Bq = {et , tet }. Since e2t cos t and e2t sin t are the only functions
in the first set that are not in the second yp (t) = a1 e2t cos t+a2 e2t sin t.
2
2t polynomial is q(s) = s 3s 10 = (s 5)(s + 2).
11. The characteristic
Since L 7e = 7/(s + 2), we set v(s) = s + 2. Then
q(s)v(s) =
(s 5)(s + 2)2 . Since Bqv = e5t , e2t , te2t and Bq = e5t , e2t we
have yp = a1 te2t , a test function. Substituting yp into the differential
64 1 Solutions
17. The
2 characteristic polynomial is q(s) = s2 1 = (s 1)(s + 1). Since
L t = 2/s , we set v(s) = s3 . Then q(s)v(s) = (s 1)(s + 1)s3 . Since
3
19. The
2tcharacteristic polynomial is q(s) = s2 4s + 4 = (s 2)2 . Since
q(s)v(s) = (s 2)3 . Since
L e = 1/(s 2), we set v(s) =s2t 2. 2tThen
2t 2t 2 2t
we have yp = a1 t2 e2t , a test
Bqv = e , te , t e and Bq = e , te
function. Substituting yp into the differential equation gives 2a1 e2t = e2t .
It follows that a1 = 1/2. The general solution is y = 12 t2 e2t +c1 e2t +c2 te2t
21. The
characteristic polynomial is q(s) = s2 + 6s + 9 = (s + 3)2 . Since
2t
= 25/(s 2)2 , we set v(s) = (s 2)2
L 25te . Then q(s)v(s) = (s +
2 2 3t 3t 2t 2t
and Bq = e3t , te3t
3) (s 2) . Since Bqv = e , te , e , te
we have yp = a1 e2t + a2 te2t , a test function. Substituting yp into the
differential equation gives (25a1 + 10a2 )e2t + 25a2 te2t = 25te2t . Linear
independence implies 25a1 + 10a2 = 0 and 25a2 = 25. We get a2 = 1 and
a1 = 2/5 The general solution is y = te2t 25 e2t + c1 e3t + c2 te3t
2 2
23. The characteristic polynomial is q(s)
3t= s + 6s + 13 = (s + 3) +2 4,
an irreducible quadratic. Since L e cos 2t = (s + 3)/((s + 3) +
1 Solutions 65
tion. Substituting yp into the differential equation gives 5a1 e2t = 10e2t
and hence a1 = 2. The general solution is y = 2e2t + c1 cos t + c2 sin t.
Since y = 4e2t c1 sin t + c2 cos t the initial conditions imply
2 + c1 = 0
4 + c2 = 0
Section 3.5
1. The characteristic
6t polynomial is q(s) = s2 4 = (s 2)(s + 2) and
L e = 1/(s + 6). Thus
1
1 p(s)
L {y} = = 32 + .
(s 2)(s + 2)(s + 6) s + 6 (s 2)(s + 2)
1 6t
A particular solution is yp = 32 e and the general solution is y =
1 6t
32 e + c1 e2t + c2 e2t
66 1 Solutions
3. The
characteristic polynomial is q(s) = s2 + 5s + 6 = (s + 2)(s + 3) and
L e2t = 1/(s + 2). Thus
1 1 p(s)
L {y} = = + .
(s + 2)2 (s + 3) (s + 2)2 (s + 2)(s + 3)
5. The
characteristic polynomial is q(s) = s2 + 2s 8 = (s 2)(s + 4) and
4t
L 6e = 6/(s + 4). Thus
6 1 p(s)
L {y} = 2
= 2
+ .
(s 2)(s + 4) (s + 4) (s 2)(s + 4)
7. The
characteristic polynomial is q(s) = s2 + 6s + 9 = (s + 3)2 and
2t
= 25/((s 2)2 ). Thus
L 25e
25 1 2 1 p(s)
L {y} = = + .
(s 2)2 (s + 3)2 (s 2)2 5 s 2 (s + 3)2
9. The
characteristic
polynomial is q(s) = s2 8s + 25 = (s 4)2 + 9 and
L 36te sin 3t = 216(s 4)/((s 4) + 9)2 . Thus
4t 2
216(s 4)
L {y} = .
((s 4)2 + 9)3
This is a partial fraction. Table 2.5 gives y = 3t2 e4t cos 3t + te4t sin 3t.
A particular solution is yp = 3t2 e4t cos 3t + te4t sin 3t and the general
solution is y = 3t2 e4t cos 3t + te4t sin 3t + c1 e4t cos 3t + c2 e4t sin 3t
s 1 1 p(s)
L {y} = = + .
(s + 1)2 (s2 + 1) 2 s + 1 (s + 1)2
2
1
A particular solution is yp = 2 sin t and the general solution is y =
1 t
2 sin t + c1 e + c2 tet
1 Solutions 67
Section 3.6
1. The force is 16 lbs. A length of 6 inches is 1/2 ft. The spring constant is
k = 16/(1/2) = 32 lbs/ft.
3. The force exerted by the mass is 40 9.8 = 392 N. Thus k = 392/.8 = 490
N/m.
represents the model for the motion of the body. The characteristic poly-
2 2
p 2 p
nomialpis q(s) = 6s + 20 = 6(s + 10/3 ). Thus y = c1 cos 10/3 t +
c2 sin 10/3 t. The initial conditions imply c1 = 1/10 and c2 = 0. Thus
1 p
y= cos 10/3 t.
10
The motion is undamped free or simple harmonic motion. Since y is
written in the form y = A cos t + we can readpoff the amplitude,
frequency, and phase shift; they are A = 1/10, = 10/3, and = 0.
11. The mass is m = 16/32 = 1/2 slugs. The spring constant k is given
by k = 16/(6/12) = 32. The damping constant is given by = 4/2 =
2. Since no external force is mentioned we may assume it is zero. The
initial conditions are y(0) = 1 and y (0) = 1. The following equation
1
2 y + 2y + 32y = 0, y(0) = 1, y (0) = 1 models the motion of the body.
The characteristic polynomial is q(s) = 21 s2 + 2s + 32 = 12 (s2 + 4s + 64) =
1 2
2
2 ((s + 2) + 60 ). Thus
y = c1 e2t cos 60t + c2 e2t sin 60t.
The initial conditions imply c1 = 1 and c2 = 3/ 60. Thus
3
y = e2t cos 60t + e2t sin 60t.
60
68 1 Solutions
13. The mass is m = 2/32 = 1/16 slug. The spring constant k is given by k =
2/(4/12) = 6 and the damping constant is = 0. The initial conditions
1
are y(0) = 0 and y (0) = 8/12 = 2/3. The equation 16 y + 6y = 0 or
equivalently y + 96y = 0 with initial conditions y(0) = 0, y (0) = 2/3
Section 4.1
11. From Exercise 9 we have y(t) = yp (t) + yh (t) = te2t + c1 e2t + c2 e2t + c3 .
Since
we have
2 = y(0) = c1 + c2 + c3
1 = y (0) = 1 + 2c1 2c2
16 = y (0) = 4 + 4c1 + 4c2 ,
Section 4.2
7. The
2tcharacteristic polynomial is q(s) = (s +2t 2)(s2 + 25). Thus Bq =
e , cos 5t, sin 5t . It follows that y(t) = c1 e + c2 cos 5t + c3 sin 5t.
we have
1 = y(0) = c1 + c2 + c3
6 = y (0) = c1 c2 + c4
3 = y (0) = c1 + c2 c3
2 = y (0) = c1 c2 c4
Section 4.3
y = ctet
y = c(1 + t)et
y = c(2 + t)et
y = c(3 + t)et
Thus Bqv \ Bq = te2t . It follows that y = cte2t is the test function and
y = cte2t
y = c(1 + 2t)e2t
y = c(4 + 4t)e2t
y = c(12 + 8t)e2t
y (4) = c(32 + 16t)e2t .
Incomplete (s 1)-chain
1 1/2
s(s + 1)(s 1)2 (s 1)2
p(s)
s(s + 1)(s 1)
n o
1 1 1 1 t
It follows that yp = 2L (s1)2 = 2 te and the general solution is
1 t t t
y= 2 te + c1 e + c2 e + c3
Incomplete s-chain
1 1/4
s3 (s2 + 4) s3
s/4
0
s2 (s2 + 4)
1/4
s(s2 + 4)
t2 t2
It follows that yp = 8 and the general solution is y = 8 + c1 + c2 cos 2t +
c3 sin 2t.
Incomplete s2 + 1-chain
4s 2s + 2
(s 1)(s2 + 1)2 (s2 + 1)2
p(s)
(s 1)(s2 + 1)
1 Solutions 73
It follows from Table 2.9 that yp = (t sin t + sin t t cos t). But since
sin t is a homogeneous solution we can write the general solution as y =
t(sin t + cos t) + c1 et + c2 cos t + c2 sin t.
Section 4.4
y1 (0) = 2 y2 (0) = 1
y1 (0) = 16 y2 (0) = 4
c1 + c2 = y(0)
2c1 + 4c2 = y (0)
For y1 we get
c1 + c2 = 2
2c1 + 4c2 = 16
which gives c1 = 4 and c2 = 6. Thus y1 (t) = 4e2t + 6e4t . For y2 we
get
c1 + c2 = 1
2c1 + 4c2 = 4
which gives c1 = 4 and c2 = 3. Thus y2 (t) = 4e2t + 3e4t .
c1 = y(0)
2c2 = y (0)
For y1 we get
c1 = 1
2c2 = 2
74 1 Solutions
y1 (0) = 0 y2 (0) = 2
y1 (0) = 20 y2 (0) = 2
y1 (0) = 60 y2 (0) = 34.
If y = c1 et + c2 et + c3 e2t then
c1 + c2 + c3 = y(0)
c1 + c2 + 2c3 = y (0)
c1 + c2 + 4c3 = y (0).
For y1 we get
c1 + c2 + c3 = 0
c1 + c2 + 2c3 = 20
c1 + c2 + 4c3 = 60.
which gives c1 = 20, c2 = 40, and c3 = 20. Thus y1 (t) = 20et +
40et 20e2t . For y2 we get
c1 + c2 + c3 = 2
c1 + c2 + 2c3 = 2
c1 + c2 + 4c3 = 34.
y1 (0) = 0 y2 (0) = 6
y1 (0) = 0 y2 (0) = 6
y1 (0) = 30 y2 (0) = 24
y1 (0) = 30 y2 (0) = 84
1 Solutions 75
c1 + c2 + c3 = y(0)
c1 c2 + 3c4 = y (0)
c1 + c2 9c3 = y (0)
c1 c2 27c4 = y (0)
For y1 we get
c1 + c2 + c3 = 0
c1 c2 + 3c4 = 0
c1 + c2 9c3 = 30
c1 c2 27c4 = 30
c1 + c2 + c3 = 6
c1 c2 + 3c4 = 6
c1 + c2 9c3 = 24
c1 c2 27c4 = 84
y1 + 3y1 = y2
y2 + 2y2 = 2y1 .
c1 + c3 = y(0)
c2 + 2c4 = y (0)
c1 4c3 = y (0)
c2 8c4 = y (0)
76 1 Solutions
For y1 we get
c1 + c3 = 3
c2 + 2c4 = 3
c1 4c3 = 9
c2 8c4 = 9
c1 + c3 = 0
c2 + 2c4 = 0
c1 4c3 = 6
c2 8c4 = 6
and therefore
Y1 (s) 1 q2 (s) 1 p1 (s)
=
Y2 (s) q1 (s)q2 (s) 1 2 2 q1 (s) p2 (s)
1 p1 (s)q2 (s) + 1 p2 (s)
= .
q1 (s)q2 (s) 1 2 p2 (s)q1 (s) + 2 p1 (s)
Section 4.5
13. The characteristic mode is et so y(t) = cet . For the unit impulse we
have y(0) = 1 and this implies c = 1. Thus y(t) = et
where C and p(t) are as in Exercise 18, which also implies that tk et cos t
f is bounded. The argument for tk et sin t f is the same.
Section 5.1
5. yes, nonhomogeneous, no
7. yes, nonhomogeneous, no
a = y(0) = 1 + c1
b = y (0) = 1 + c1 + c2 .
(3)a. y(t) = et et + 2t
(3)b. y(t) = et + (0)et + (1)t = et + t
(3)c. y(t) = et + et + 3t
(3)d. y(t) = et + (a 1)et + (b a + 2)t
27. The assumptions say that y1 (t0 ) = y2 (t0 ) and y1 (t0 ) = y2 (t0 ). Both y1
and y2 therefore satisfies the same initial conditions. By the uniqueness
part of Theorem 6 y1 = y2 .
Section 5.2
9. f1 (t) = 2 sin(2
t
ln t)
, f1 (t) = 2 sin(2 ln t)4
t2
cos(2 ln t)
, f2 (t) = 2 cos(2
t
ln t)
, and
4 sin(2 ln t)2 cos(2 ln t) 2
f2 (t) = t2 . Thus t f1 + tf1 + 4f1 = 2 sin(2 ln t)
4 cos(2 ln t) 2 sin(2 ln t) + 4 cos(2 ln t) = 0. Similarly, t2 f2 + tf2 + 4f2 =
4 sin(2 ln t) 2 cos(2 ln t) + 2 cos(2 ln t) + 4 sin(2 ln t) = 0. Now,
82 1 Solutions
cos(2 ln t) sin(2 ln t) 4
w(t) = 2 sin(2 ln t) 2 cos(2 ln t) = .
t
t t
On the other hand the coefficient function of y in theRstandard form of
t
the differential equation is a1 (t) = 1t Integrating gives 1 x1 dx = ln t and
Rt
e 0 a1 (x) dx = 1/t. At t = 1 we have w(1) = 1 so Abels formula is veri-
fied. It follows from Proposition 4 that f1 and f2 are linearly independent.
By Theorem 2 the solution set is {c1 cos(2 ln t) + c2 sin(2 ln t) : c1 , c2 R}
11. 1. Suppose at3 + b t3 = 0 on (, ). Then for t = 1 and t = 1 we
get
a+b = 0
a + b = 0.
Section 5.3
c2 t2 sin(3 ln t).
c1 = 2
1
c1 + c2 = 0.
2
Thus c1 = 2 and c2 = 1. Hence y = 2t1/2 t1/2 ln t
15. The coefficient functions for the given equation in standard form are
a1 (t) = 4/t and a2 (t) = 6/t2 both of which are not defined at the initial
condition t0 = 0. Thus the uniqueness and existence theorem does not
guarantee a solution. In fact, the condition that y (0) exist presupposes
that y is defined near t = 0. For t positive the indicial polynomial is
Q(s) = s2 5s + 6 = (s 6)(s + 1) and therefore y(t) = c1 t6 + c2 t1 .
The only way that y can be extended to t = 0 is that c2 = 0. In this case
y(t) = c1 t6 cannot satisfy the given initial conditions. Thus, no solution
is possible.
84 1 Solutions
Section 5.4
bt at
1. By LHospitals rule limt0 e e
t = b a. So Theorem 4 applies and
gives
bt Z
e eat
1 1
L (s) = d
t s b a
M b sb
= lim ln ln
M M a sa
sa
= ln
sb
cos bt cos at
3. Apply LHospitals rule twice to get lim 2 = a2 b2 . Now
t t2
use Exercise 2 to get
Z 2
+ a2
cos bt cos at
L 2 (s) = ln d
t2 s 2 + b2
Z M Z M !
2 2 2 2
= lim ln( + a ) d ln( + b ) d .
M s s
x2 +a2
2. lim x ln x2 +b2 =0
x
The first fact is shown by integration by parts and the second fact is
shown by LHospitals rule. We now get (after some simplifications)
2
s + b2
cos bt cos at 1 a
1 b
L 2 (s) = s ln 2 + 2a tan 2b tan
t2 s + a2 s s
y(t) = y0 et + C(t 1 + et )
= (y0 + C)et + C(t 1).
y0
7. Applying the Laplace transform we get Y (s) = and therefore
(s + 2)2
y0
Y (s) = + C. However, since lims Y (s) = 0 we must have C = 0.
s+2
Hence y(t) = y0 e2t .
6s
9. Applying the Laplace transform we get Y (s) + Y (s) = 0. An
s2
+1
C
integrating factor is I = (s2 + 1)3 . We then get Y (s) = 2 , and
(s + 1)3
2
y(t) = (C/8) (3 t ) sin t 3t cos t
y0
1 1
11. Apply the Laplace transform to get Y (s) = = y0 .
s(s 1) s s1
s
Then Y (s) = y0 ln + C. Take C = 0 since lims Y (s) = 0.
s1
et 1
Hence y(t) = y0 .
t
y0
13. Apply the Laplace transform, simplify, and get Y (s) = 2 =
(s 5s + 6)
1 1 s2
y0 . Then Y (s) = y0 ln + C. Take C = 0. Then
s 2 s 3 s3
e3t e2t
y(t) = y0 .
t
sy0
15. Apply the Laplace transform, simplify, and get Y (s) = 2 + 1)
s(s
2y1 1 1 s
= y0 2 2y1 2 . Integrating gives Y (s) =
s(s2 + 1 s + 1
2 s s +1
s +1
y0 tan1 (s) + y1 ln + C. Since lims Y (s) = 0 we must have
s2 2
1 s +1
C = y0 and hence Y (s) = y0 tan1 + y1 ln . Therefore
2 s s2
sin t 1 cos t
y(t) = y0 + 2y1
t t
17. We use the formula
n k
dn X n d dnk
(f (t)g(t)) = f (t) g(t).
dtn k dtk dtnk
k=0
Observe that
dk t
e = (1)k et
dtk
and
dnk n
t = n(n 1) (k + 1)tk .
dtnk
86 1 Solutions
21. Hint: Take the Laplace transform of each side. Use the previous exercise
and the binomial theorem.
L {tn } (s)
(s 1)n
=
sn+1
(s 1)n1
= (n + 1 s).
sn+2
nL {n1 } (s)
(s 1)n1
= n
sn
(s 1)n1
= (ns2 ).
sn+2
(s 1)n1
We have written each so that the common factor is . The
sn+2
coefficients are
1 Solutions 87
(s 1)n1
1
(n + 1)(s 1)2
n+1 sn+2
(s 1)n+1
=
sn+2
= L {n+1 } (s).
L et n (s)
1 (s 1)n
=
s 1 sn+1
(s 1)n1
=
sn+1
(s 1)n1
s1
= 1
sn s
n1
(s 1) (s 1)n
=
sn sn+1
= L1 {n1 (t)} L1 {n (t)} .
Section 5.5
7. Let y2 (t) = u(t) sin t2 . Then u(t) satisfies t sin t2 u +(4t2 cos t2 sin t2 )u =
u 1 cos t2
0 and hence = 4t . It follows that u = t csc2 t2 and therefore
u t sin t2
u(t) = 1 2 1 2
2 cot t . We now get y2 (t) = 2 cos t . The general solution can
2 2
be written y(t) = c1 sin t + c2 cos t .
11. The functions tan t and sec t are continuous except at points of the
form 2 + 2n, n Z. We will work in the interval (/2, /2).
Let y2 (t) = u(t) tan t. Then u tan t + u (tan2 t + 2) = 0 and hence
u
u = tan t 2 cot t. It follows that ln |u | = ln |cos t| 2 ln |sin t| and
2 1
thus u = cos t sin t. Further u(t) = sin t and we have y2 (t) = sec t.
The general solution can be written y(t) = c1 tan t + c2 sec t.
sin 2t u
13. Let y2 = u 1+cos 2t . Then u(t) satisfies u sin 2t + 4u = 0 and hence u =
4 csc 2t. We now get ln u = 2 ln |csc 2t + cot 2t|. Thus u = (csc 2t +
cot 2t)2 = csc2 2t + 2 csc 2t cot 2t + cot2 2t = 2 csc2 2t + 2 csc 2t cot 2t 1.
By integrating we get u = cot 2t csc 2t t = 1+cos 2t
sin 2t t. It now
t sin 2t
follows that y2 = 1 1+cos 2t . The general solution can be written
sin 2t t sin 2t
y(t) = c1 1+cos 2t + c2 1 + 1+cos 2t .
1 1 1+t
y2 (t) = t + (1 t2 ) ln .
2 4 1t
1 Solutions 89
Section 5.6
1. sin t and cos t form a fundamental set for the homogeneous solutions. Let
yp (t) = u1 cos t +u2 sin t.Thenthe matrix equation
u1
cos t sin t 0
= implies u1 (t) = sin2 t = 12 (cos 2t 1)
sin t cos t u2 sin t
and u2 (t) = cos t sin t = 12 (sin 2t). Integration give u1 (t) = 14 (sin(2t)
2t) = 21 (sin t cos t t) and u2 (t) = 1 1
4 cos 2t = 4 (2 cos 2t 1). This
1 1 1
implies yp (t) = 4 sin t 2 t cos t. Since 4 sin t is a homogeneous solution
we can write the general solution in the form y(t) = 1 2 t cos t + c1 cos t +
c2 sin t. We observe that a particular solution is the imaginary part of a so-
lution to y + y = eit . We use the incomplete partial fraction method and
p(s)
get Y (s) = (si)12 (s+i) . This can be written Y (s) = 2i 1 1
(si)2 + (si)(s+i) .
1 1 1
From this we get yp (t) = Im 2i L { (si) 2} = Im i 2 te
it
= 1
2 t cos t.
1
The general solution is y(t) = 2 t cos t + c1 cos t + c2 sin t.
3. The functions et cos 2t and et sin 2t form a fundamental set. Let yp (t) =
c1 et cos 2t + c2 et sin 2t.
Then the matrix equation
u1
0
t t
W (e cos 2t, e sin 2t) = implies that u1 (t) = 1 2 sin 2t and
u2 et
u2 (t) = 12 cos 2t. Hence, u1 (t) = 14 cos2t and u2 (t) = 14 sin 2t. From this
we get yp (t) = 14 et cos2 2t + 14 et sin2 2t = 14 et . On the other hand, the
method of undetermined coefficients implies that a particular solution
is of the form yp (t) = Cet . Substitution gives 4Cet = et and hence
C = 41 . It follows that yp (t) = 41 et . Furthermore, the general solution is
y(t) = 14 et + c1 et cos 2t + c2 et sin 2t.
t 2t
5. A
t fundamental set is e ,e . The matrix equation
2t
e e u1 0
= implies u1 (t) = e2t and u2 (t) = et . Hence
et 2e2t u2 e3t
u1 (t) = 1 2t t 1 2t t
2 e , u2 (t) = e , and yp (t) = 2 e e + e e
t 2t
= 12 e3t . The
1 3t t 2t
general solution is y(t) = 2 e +c1 e +c2 e . The method of undetermined
coefficients implies that a particular solution is of the form yp = Ce3t .
Substitution gives 2Ce3t = 3e3t and hence C = 12 . The general solution
is as above.
90 1 Solutions
t
A fundamental
7. is {e
set , tet}. The matrix equation
t t 0
e te u1
t t t = et implies u1 (t) = 1 and u2 (t) = 1t . Hence,
e e + te u2 t
u1 (t) = t, u2 (t) = ln t, and yp (t) = tet + t ln tet . Since tet is a
homogeneous solution we can write the general solution as y(t) = t ln tet +
c1 et + c2 tet .
15. After put in standard form the forcing function f is 4t4 . The matrix
equation
cos t2 sin t2
u1 0
= implies u1 (t) = 2t3 sin t2 and
2t sin t2 2t cos 2t u2 4t4
u2 (t) = 2t3 cos t2 . Integration by parts gives u1 (t) = t2 cos t2 sin t2 and
u2 (t) = t2 sin t2 +cos t2 . Hence yp (t) = t2 cos2 t2 cos t2 sin t2 +t2 sin2 t2 +
cos t2 sin t2 = t2 . The general solution is y(t) = t2 + c1 cos t2 + c2 sin t2 .
17. Let a and t be in the interval I. Let z1 and z2 be the definite integrals
defined as follows:
1 Solutions 91
t
y2 (x)f (x)
Z
z1 (t) = dx
a w(y1 , y2 )(x)
t
y1 (x)f (x)
Z
z2 (t) = dx.
a w(y1 , y2 )(x)
19. Let y1 (t) = eat and y2 (t) = eat . Then {y1 , y2 } is a fundamental set. We
have
y1 (x) y2 (x) ax at
y1 (t) y2 (t) = e
e eax eat = ea(tx) ea(tx) = 2 sinh(a(t x))
and
eat eat
w(y1 , y2 )(x) =
= 2a.
aeat aeat
Thus
t
2 sinh a(t x)
Z
yp (t) = f (x) dx
0 2a
1
= f (t) sinh at.
a
Applying the Laplace transform to y a2 y = f , with initial conditions
y(0) = y (0) = 0, gives s2 Y (s) a2 Y (s) = F (s). Solving for Y (s) we get
F (s) 1 a
Y (s) = = F (s).
s2 a2 a s2 a2
The convolution theorem gives a particular solution
1
yp (t) = sinh at f (t).
a
92 1 Solutions
21. Let y1 (t) = eat and y2 (t) = ebt . Then {y1 , y2 } is a fundamental set. We
have
y1 (x) y2 (x) ax bt bx at ax+bt
y1 (t) y2 (t) = e e e e = e
ebx+at
and ax
ebx
e
w(y1 , y2 )(x) = ax
= (b a)e(a+b)x .
ae bebx
Thus
t
eax+bt ebx+at
Z
yp (t) = (a+b)x
f (x) dx
0 (b a)e
Z t
1
= (eb(tx) ea(tx) )f (x) dx
ba 0
1
= f (t) (ebt eat ).
ba
Applying the Laplace transform to y (a + b)y + aby = f , with initial
conditions y(0) = y (0) = 0, gives s2 Y (s) (a + b)sY (s) + abY (s) = F (s).
Solving for Y (s) we get
F (s) 1 1 1
Y (s) = = F (s).
(s a)(s b) ab sa sb
Section 6.1
1. Graph (c)
3. Graph (e)
5. Graph (f)
7. Graph (h)
R5 R2 R3 R5 2
9. 0 f (t) dt = 0 (t2 4) dt + 2 0 dt + 3 (t + 3) dt = t3 /3 4t 0 + 0 +
5
t2 /2 + 3t = (8/3 8) + (25/2 + 15) (9/2 + 9) = 22/3.
3
R 2 R R 2 2
11. 0
|sin x| dx = 0
sin x dx +
sin x dx = cos x|0 + cos x| = 4.
1 Solutions 93
R5 R3 R4 R6
13. 2
f (t) dt = 2
(3 t) dt + 3
2(t 3) dt + 4
2 dt = 1/2 + 1 + 4 = 11/2
R6 R1 R2 R6
15. 0
f (u) du = 0
u du + 1
(2 u) du + 2
1 du = 1/2 + 1/2 + 4 = 5.
17. A is true since y(t) satisfies the differential equation on each subinterval.
B is true since the left and right limits agree at t = 2. C is not true since
y(0) = 1 6= 2.
19. A is true since y(t) satisfies the differential equation on each subinterval.
B is false since limt2 y(t) = 1 + e8 while limt2+ y(t) = 1. C is false
since B is false.
21. A is true since y(t) satisfies the differential equation on each subinterval.
B is true since limt1 y(t) = 2e + e2 = limt1+ y(t). C is false since
limt1 y (t) = 3e + 2e2 while limt1+ y (t) = 3e2 2e. D is false since
C is false.
23. A is true since y(t) satisfies the differential equation on each subinterval.
B is true since limt1 y(t) = 2e + e2 = limt1+ y(t). C is true since
limt1 y (t) = 3e + 2e2 = limt1+ y (t). D is true since y(0) = y (0) =
0.
2) = 1 so there is one sequence tn 0 with f (tn ) 0 while another
sequence tn 0 with f (tn ) 1 so f (t) cannot be continuous at 0.
3. To be piecewise continuous, f (t) would have to have a limit at t
approaches 0 from above, and this is not true as shown in part 2.
Section 6.2
0
if t < 2,
1. f (t) = 3h(t 2) h(t 5) = 3 if 2 t < 5, Thus, the graph is
2 if t 5.
y
3
2
1
0 t
0 1 2 3 4 5 6 7 8
0 t
0 1 2 3
where the dashed line is the part of the t2 graph that has been truncated.
It is only shown for emphasis and it is not part of the graph.
96 1 Solutions
7. This is the function cos 2t shifted units to the right and then truncated
at t = . The graph is
y
1
0 t
2 3
1
e3s
2 2 4
= 3 e2s 3
+ 2
2 .
s s s s
P
25. (a) n=0 P (t n)[n,n+1) (t);
(b) t Ph(t n);
n=1 P
(c) L {t n=1 h(t n)} = L {t} n=1 L {h(t n)}
1 P ens 1 1 P n
= 2 n=1 = 2 (es )
s s s s n=1
1 es
= 2 .
s s(1 es )
27. (a)
P P
n=0 (2n + 1 t)[2n,2n+2) (t); (b) (t + 1) + 2 n=0 h(t 2n);
1 1 2
(c) 2 + .
s s s(1 e2s )
3s
1 e 1 1
29. L 2
= h(t 3) L
s s2 tt3
(
0 if 0 t < 3,
= h(t 3) (t)|tt3 = (t 3)h(t 3) =
t 3 if t 3.
es
1
31. L1 2
= h(t ) L 1
2
s +1 s + 1 tt
= h(t ) (sin t)|tt = h(t ) sin(t )
( (
0 if 0 t < , 0 if 0 t < ,
= =
sin(t ) if t sin t if t .
es
1 1 1
33. L 2
= h(t ) L 2
s + 2s + 5 s + 2s + 5 tt
1 1 1 t
= h(t ) L = h(t ) ( 2 e sin 2t)
(s + 1)2 + 22 tt
tt
(
0 if 0 t < ,
= 12 e(t) sin 2(t )h(t ) = 1 (t)
2e sin 2t if t .
e2s
1
35. L1 2
= h(t 2) L 1
2
s +4 s + 4
tt2
1 1
= h(t 2) ( 2 sin 2t) tt2 = 2 h(t 2) sin 2(t 2)
(
0 if 0 t < 2,
= 1
2 sin 2(t 2) if t 2.
98 1 Solutions
se4s
s
37. L1 = h(t 4) L1
s2 + 3s + 2 s2 + 3s + 2
tt4
1 2 1
= h(t 4) L
s + 2 s + 1 tt4
= h(t 4) (2e2t et )tt4 = h(t 4) 2e2(t4) e(t4)
(
0 if 0 t < 4,
=
2e2(t4) e(t4) if t 4.
1 e5s
1 1
39. L1 = L1 h(t 5) L1
s2 s2 s2 tt5
(
t if 0 t < 5,
= t h(t 5) (t)|tt5 = t (t 5)h(t 5) =
5 if t 5.
1 s 2s + 1 1 2s + 1
41. L e 2
= h(t )L 2
s + 6s + 13 s + 6s + 13 tt
2(s + 3) 5
= h(t ) L1
(s + 3)2 + 22 tt
1 2(s + 3)
= h(t ) L 2 2
(s + 3) + 2 tt
1 5
+ h(t ) L
(s + 3)2 + 22 tt
= h(t ) (2e3t cos 2t 52 e3t sin 2t)tt
3(t)
2 cos 2(t ) 52 sin 2(t )
( )e
= h(t
0 if 0 t < ,
= 3(t) 5
e 2 cos 2t 2 sin 2t if t .
43. Let b > 0. Since f1 and f2 are piecewise continuous on [0, ) they
only have finitely many jump discontinuities on [0, b). It follows that
f1 + cf2 have only finitely many jump on [0, b). Thus f1 + cf2 is piecewise
continuous on [0, ).
Section 6.3
1. We write the forcing function as f (t) = 3h(t 1). Applying the Laplace
transform, partial fractions, and simplifying gives
3 s 3 1 1
Y (s) = e = es .
s(s + 2) 2 s s+2
e3s
Y (s) =
s(s2 + 9)
1 1 s
= e3s .
9 s s2 + 9
9. Write the forcing function as f (t) = 6[1,3) = 6h(t 1) 6h(t 3). Now
apply the Laplace transform, partial fractions, and simplify to get
6
es e3s
Y (s) =
s(s + 2)(s + 3)
1 3 2
es e3s
= +
s s+2 s+3
11. Apply the Laplace transform, partial fractions, and simplify to get
1 1
Y (s) = + e3s
(s + 1)2 s(s + 1)2
1 1 1 1 3s
= + + e
(s + 1)2 (s + 1)2 s+1 s
lb gal
13. For the first three minutes, source one adds salt at a rate of 1 gal 2 min =
lbs
2 min . and after that source two takes over and adds salt at a rate of
lb gal lbs
5 gal 2 min = 10 min . Thus the rate at which salt is being added is given
by the function
(
2 if 0 t < 3
f (t) =
10 if 3 t < .
= 2[0,3) + 10[3,)
= 2(1 h(t 3)) + 10(h(t 3))
= 2 + 8h(t 3).
y(t) y(t)
The output rate of salt is given by 4 2 = 2 lbs/min. We are thus
led to the differential equation
1
y + y(t) = 2 + 8h(t 3), y(0) = 0.
2
We take the Laplace transform of both sides and use partial fractions to
get
2 8e3s
Y (s) = +
s(s + 1/2) s(s + 1/2)
4 4 3s 16 16
= +e .
s s + 1/2 s s + 1/2
15. For the first two minutes, source one adds salt at a rate of 1 kg L
L 3 min =
kg
3 min . Thereafter source two takes over for two minutes but the input rate
of salt is 0. Thereafter source on take over again and adds salt to the tank
kg
at a rate of 3 min . Thus the rate at which salt is being added is given by
the function
3
if 0 t < 2
f (t) = 0 if 2 t < 4
3 if 4 t < .
= 3[0,2) + 3[4,)
= 3(1 h(t 2) + h(t 4)).
102 1 Solutions
y(t) 3
The output rate of salt is given by 10 3 = 10 y(t) kg/min. We are thus
led to the differential equation
3
y + y(t) = 3(1 h(t 2) + h(t 4)), y(0) = 2.
10
We take the Laplace transform of both sides, simplify, and use partial
fractions to get
2 3
Y (s) = + (1 e2s + e4s )
s + 3/10 s(s + 3/10)
10 8 10 10
= + (e4s e2s ).
s s + 3/10 s s + 3/10
Section 6.4
1. Take the Laplace transform, solve for Y (s), and simplify to get Y (s) =
es
s+2 . Laplace inversion then gives
y = e2(t1) h(t 1)
(
0 if 0 t < 1
= .
e2(t1) if 1 t <
3. Take the Laplace transform, solve for Y (s), and simplify to get Y (s) =
2 e4s
s4 + s4 . Laplace inversion then gives
1 1
y = sin 2t + sin 2(t )h(t )
2 2
1 1
= sin 2t + sin(2t) h(t )
2
( 2
sin 2t
2 if 0t<
= .
sin 2t if t <
s+3 2e2s
Y (s) = +
(s + 1)(s + 3) (s + 1)(s + 3)
1 1 1
= + e2s .
s+1 s+1 s+3
9. Take the Laplace transform, apply partial fractions, and simplify to get
s+1 3
Y (s) = + es
(s + 2)2 (s + 2)2
1 1 3
= + es .
(s + 2)2 s + 2 (s + 2)2
11. The input rate of salt is 6 + 43 while the output rate is 3 y(t)
12 . We thus
have the differential equation y + 14 y = 6 + 43 , y(0) = 0. Take the
Laplace transform, apply partial fractions, and simplify to get
104 1 Solutions
6 4
Y (s) = + e3s
s(s + 1/4) s + 1/4
24 24 4
= + e3s .
s s + 1/4 s + 1/4
13. Clearly, y(0) = 0. The input rate is 0 +2 +4 +6 while the output rate is
P3
y. We are thus led to the differential equation y +y = k=0 2k , y(0) =
0. Take the Laplace transform and solve for Y (s) to get
3
X e2ks
Y (s) = .
s+1
k=0
1r n+1
Using the formula 1 + r + r2 + + rn = 1r we get
3
X 3
X 3
X
y(6) = e(62k) = e2k = (e2 )k
k=0 k=0 k=0
2 4
1 e
= = 1.156 lb.
1 e2
.1 1
Y (s) = + e4s .
s + 2 (s + 2)2
17. Clearly
Pm = 1, = 0, and k = 1. The forcing function is 0 + + +
5
5 = k=0 k . The differential equation that describes the motion is
5
X
y + y = k .
k=0
0
2 3 4 5 6
1
106 1 Solutions
19. The value of y(t) at t = c for y given by Exercise 18 is y0 eac . Thus the
differential equation we need to solve is y + ay = 0, y(c) = y0 eac + k, on
the interval [c, ). We get the general solution y(t) = beat . The initial
condition implies beac = y0 eac + k. Solving for b gives b = y0 + keac .
Thus
y = y0 eat + kea(tc) ,
on the interval [c, ).
Section 6.5
1 1 es 1 e s
1. F (s) = s1 and G(s) = s s .
Thus F (s)G(s) = (s1)s (s1)s . Par-
1 1 1 1 s
tial fractions gives F (s)G(s) = s1 s s1 s e and Laplace
inversion gives
1 1
3. F (s) = es L{t + 1} = es and G(s) = 1s e3s e4s . Thus
s2 + s
1 1
e4s e5s
F (s)G(s) = 3
+ 2
s s
(t 4)2 (t 5)2
f g = + (t 4) h(t 4) + (t 5) h(t 5)
2 2
0
if 0 t < 4
(t4)2
= 2 + (t 4) if 4 t < 5 .
t 7/2 if 5 t <
1
5. F (s) = s 1s e2s and G(s) = 1
s 1s e2s . Thus
1
1 2e2s + e4s
F (s)G(s) = 2
s
1 Solutions 107
7.
n o
1
9. The unit impulse response function is (t) = L1 s3 = e3t . The
1 e2s
homogeneous solution is yh = 2e3t . Observe that Yp (s) = s3 s =
1 1 1 2s
3 s3 s e . It follows that the particular solution is
yp = (h(t 2))
= L1 {Yp (s)}
1 3(t2)
= e 1 h(t 2).
3
and
y = yh + yp
1 3(t2)
= 2e3t + e 1 h(t 2)
( 3
2e3t if 0 t <
= 3t 1 3(t2)
.
2e + 3 e 1 if 1 t <
n o
1
11. The unit impulse response function is (t) = L1 s+8 = e8t . The
homogeneous solution is yh = 2e8t . The particular solution is yp =
[3,5) . Observe that
108 1 Solutions
Yp (s) = L {yp }
3s
e5s
1 e
=
s+8 s s
1 1 1
e3s e5s .
=
8 s s+8
It follows that
yp = L1 {Yp (s)}
1 1
= 1 e8(t3) h(t 3) 1 e8(t5) h(t 5).
8 8
and
y = yh + yp
1 1
= 2e8t + 1 e8(t3) h(t 3) 1 e8(t5) h(t 5)
8 8
8t
2e if 0 t < 3
2e8t + 81 1 e8(t3)
= if 3 t < 5
2e8t + 18 e8(t5) e8(t3)
if 5 t <
n o
13. The unit impulse response function is (t) = L1 s21+9 = 13 sin 3t. The
homogeneous solution is yh = cos 3t. For the particular solution yp we
have
1
yp = sin 3t [0,2)
3
1 t
Z
= sin(3(t u))[0,2) du
3 0
(R t
1 0 sin(3(t u)) du if 0 t < 2
= R 2
3 sin(3(t u)) du if 2 t <
0
(
1 1 cos 3t if 0 t < 2
= .
9 0 if 2 t <
15. Let y be the homogenous solution to q(D)y = 0 with the given initial
conditions. Observe that
(
n
k
o sk Y (s) if k < n
L D y = n
.
s Y (s) 1/an if k = n
and thus
1 Solutions 109
(
n o ak sk Y (s) if k < n
L ak D k y = .
an sn Y (s) 1 if k = n
a0 + a1 s + an1 sn1
= 0.
q(s)
n o k1
X
L Dk y = sk Y (s) sk1 y0 yk1 = sk Y (s) sl yk1l ,
l=0
Therefore
n k1
X X sl
Y (s) = ak yk1l .
q(s)
k=1 l=0
n1 n
!
X X
y(t) = ak ykl1 (l)
l=0 k=l+1
n1 nl1
!
X X
= ak+l+1 yk (l) ,
l=0 k=0
1 2 1
2 3
1 2 1
2 3
y = c0 + c1
= 7tet + 2(et + tet )
= 2et 5tet .
0 1 0 1
1 0 4
0 1 0 1
1 0 4
0 1 0 1
1 0 4
y = c0 + c1 + c2
= 5(1 cos t) + 0(sin t) + 1(cos t)
= 5 4 cos t.
Section 6.6
1.
X
(t n)2 [n,n+1) (t) = (< t >1 )2 .
n=0
3.
X 1X
n2 [3n,3(n+1)) (t) = (3n)2 [3n,3(n+1)) (t)
n=0
9 n=0
1
= ([t]3 )2 .
9
5.
X X 3
(t + n)[2n,2(n+1)) (t) = (t 2n + 2n)[2n,2(n+1)) (t)
n=0 n=0
2
3
= < t >2 + [t]2 .
2
7.
1
L et et h(t 3)
L {f (< t >3 )} = 3s
1e
1 1 3s
t+3
= e L e
1 e3s s 1
e3s e3
1 1
=
1 e3s s 1 s1
1 e3(s1) 1
= .
1 e3s s 1
112 1 Solutions
9.
Z 2p
1
L {f (< t >2p )} = est f (t) dt
1 e2ps 0
Z p 2p
1
Z
= est dt est dt
1 e2ps 0 p
eps 1 e2ps eps
1
= +
1 e2ps s s
2
(1 eps ) 1
=
1 e2ps s
1 eps 1
= .
1 + eps s
11. Since < t >p = t [t]p we have [t]p = t < t >p . Hence
We now have
1 Solutions 113
e2ns e(2n+2)s
X
L {f ([t]2 )} = e2n
n=0
s
X 1 e2s
= e2n e2ns
n=0
s
1 e2s X 2n(s+1)
= e
s n=0
1 e2s X 2(s+1) n
= e
s n=0
1 e2s 1
= 2(s+1)
s 1e
1 e2s 1
=
1 e2(s+1) s
15.
X
L{f ([t]p )} = f (np)L{[np,(n+1)p) }
n=0
X enps e(n+1)ps
= f (np)
n=0
s
1 eps X
= f (np)enps .
s n=0
1e4(s2)
17. Let F (s) = (1e4s )(s2) . We first write
X 1 e4(s2) 4ns
F (s) = e
n=0
s2
X e4ns e8 e4(n+1)s
=
n=0
s2
X
L1 {F (s)} = e2(t4n) h(t 4n) e8 e2(t4(n+1)) h(t 4(n + 1))
n=0
X
= e2(t4n) (h(t 4n) h(t 4(n + 1)))
n=0
X
= e2t e2(4n) [4n,4(n+1))
n=0
19.
epns
X
1 n 1
L {F (s)} = (1) L
n=0
s+a
X
= (1)n ea(tpn) h(t pn)
n=0
X
X N
at
= e (1)n eapn [N,(N +1)p)
N =0 n=0
X 1 (eap )N +1
= eat
1 eap
N =0
X 1 (1)N +1 ea(N +1)p
= eat
1 + eap
N =0
1+ea(Nap+1)p if t [N p, (N + 1)p), (N even)
1+e
= eat a(N +1)p
1e if t [N p, (N + 1)p), (N odd)
1+eap
[t]p
1 + (1) p ea([t]p +p)
= eat .
1 + eap
Section 6.7
1. On the interval [0, 2) the input rate is 2 4 = 8 lbs salt per minute.
On the interval [2, 4) the input rate is 1 4 = 4 lbs salt per minute.
The input function f (t) is periodic with period 4. We can thus write
f (t) = 4 + 4 sw2 (t). The output rate is y(t)
10 4. The resulting differential
equation that models this problem is
1 Solutions 115
4
y + y = 4 + 4 sw2 (t), y(0) = 0.
10
Taking the Laplace transform and simplifying gives
4 4 1
Y (s) = 2 + 2 1 + e2s .
s(s + 5 ) s(s + 5 )
and hence
10 10 10 1 10 1
Y (s) = 2 + s 1 + e2s 2 1 + e2s .
s s+ 5 s+ 5
Let
10 10
Y1 (s) =
s s + 25
10 1
Y2 (s) =
s 1 + e2s
10 1
Y3 (s) = 2 1 + e2s .
s+ 5
Example 6.6.2 and Exercise 6.6.19 are useful for taking the inverse
Laplace transforms of Y2 and Y3 . We get
2t
y1 (t) = 10 10e 5
whose graph bounds the graph of y from above. Now observe that
4 4
10e 5 10e 5
lim l(t) = 20 4 13.10 and lim u(t) = 10 + 4 16.9
t 1 + e5 t 1 + e5
Thus the amount of salt fluctuates from 13.10 pounds to 16.90 pounds in
the long term.
P
3. The input function is 5 n=1 2n = 50 (< t >2 ). and therefore the dif-
ferential equation that models this system is
1
y + y = 50 (< t >2 ), y(0) = 0.
2
By Proposition 6.6.6 the Laplace transform gives
5 1
Y (s) = 1 .
s+ 2
1 e2s
This means that the salt fluctuation in the tank varies between 2.91 and
7.91 pounds for large values of t.
y1 (t) = 3000ert
N
!
X X
y2 (t) = 40 er(tn)
[N,N +1)
N =0 n=0
X 1 er(N +1)
= 40ert [N,N +1)
1 er
N =0
rt r([t]1 t+1)
e e
= 40 .
1 er
It follows now that
e60r er
y(60) = 3000e60r 40
1 er
= 7464.96 3988.16
3477
Section 6.8
1. Since c = 2 is not an odd multiple of we get
y(t) = 2 2 sw1 (t) (1)[t]1 (cos < t >1 sin < t >1 )
2 (cos t + sin t)) ,
sin 2 c 2
where = 1+cos 2
. Since = is irrational the motion is non periodic
2
5. Since c = is an odd multiple of we get y(t) = 2 (sw1 (t) [t]1 cos t cos t).
Resonance occurs.
Resonance occurs.
cos(N + 1) 1 + i sin(N + 1)
=
cos 1 + i sin
cos(N + 1) 1 + i sin(N + 1) cos 1 i sin
=
cos 1 + i sin cos 1 i sin
The product of the denominators simplifies to 2 2 cos . The product
of the numerators has a real and imaginary part. Call them R and I,
120 1 Solutions
respectively. Then
and
15. Let
N N
X 1 X
R(v) = cos nv = (1 + cos N v + sin N v) = Re einv
n=0
2 n=0
N N
X 1 X
I(v) = sin nv = (sin N v + (1 cos N v)) = Im einv ,
n=0
2 n=0
N
X N
X
cos(u + nv) = Re ei(u+nv)
n=0 n=0
N
!
X
iu inv
= Re e e
n=0
= Re ((cos u + i sin u)(R(v) + iI(v)))
= (cos u)R(v) (sin u)I(v)
1
= (cos u + cos u cos N v + cos u sin N v)
2
1
(sin u sin N v + (sin u sin u cos N v))
2
1
= (cos u + cos(u + N v) + ( sin u + sin(u + N v)) .
2
Similarly,
N
X N
X
sin(u + nv) = Im ei(u+nv)
n=0 n=0
N
!
X
= Im eiu einv
n=0
= Im ((cos u + i sin u)(R(v) + iI(v)))
= (sin u)R(v) + (cos u)I(v)
1
= (sin u + sin u cos N v + sin u sin N v)
2
1
+ (cos u sin N v + (cos u cos u cos N v))
2
1
= (sin u + sin(u + N v) + (cos u cos(u + N v))) .
2
Section 7.1
(n+1)2
1. The ratio test gives n2 1. R = 1.
2n n! 1
3. The ratio test gives 2n+1 (n+1)! = 2(n+1) 0. R = .
(n+1)!
5. The ratio test gives n! = n + 1 . R = 0.
P (1)n t2n
7. t is a factor in this series which we factor out to get t (2n)! . Since
n=0
t is a polynomial its presence will not change the radius of convergence.
122 1 Solutions
(1)n un
Let u = t2 in the new powers series to get
P
(2n)! . The ratio test
n=0
(1)n+1 (2n)! 1
gives (1) n (2n+2)! = (2n+1)(2n+2) 0. The radius of convergence in u
and hence t is .
135(2n+1)
9. The expression in the denominator can be written 1 = 1234(2n+1)
242n =
(2n+1)! (2n+1)! P (n!)2 2n tn
2n (123n) = 2n n! and the given power series is (2n+1)! . The ratio
n=0
((n+1)!)2 2n+1 (2n+1)! (n+1)2 2 1
test gives (n!)2 2n (2n+3)! = (2n+3)(2n+2) 2 . R = 2.
1 1 1 1 t n
P
11. Use the geometric series to get ta = a 1 at = a a =
n=0
P tn
an+1 .
n=0
sin t 1 P (1)n t2n+1 P (1)n t2n
13. t = t (2n+1)! = (2n+1)! .
n=0 n=0
1
15. Recall that tan1 t = 1+t
R
2 dt. Using the result of Exercise 10 we get
R 2n+1
1 n 2n
(1)n t2n+1 + C. Since tan1 0 = 0
P P
tan t = (1) t dt + C =
n=0 n=0
2n+1
it follows that C = 0. Thus tan1 t = (1)n t2n+1 .
P
n=0
d2n+1 t2n+1 and hence sin t =
P
17. Since tan t is odd we can write tan t =
n=0
t3 t5
d2n+1 t2n+1 . Writing out a few terms gives t
P
cos t 3! + 5! =
n=0
2 4
(1 t2! + t4!
)(d1 t + d3 t3 + d5 t5 ). Collecting like powers of t gives
the following recursion relations
d1 = 1
d1 1
d3 =
2! 3!
d3 d1 1
d5 + =
2! 4! 5!
d5 d3 d1 1
d7 + = .
2! 4! 6! 7!
Solving these equations gives
1 Solutions 123
d1 = 1
1
d3 =
3
2
d5 =
15
17
d7 = .
315
Thus tan t = 1 + 13 t3 + 2 5
15 t + 17 7
315 t + .
2 3 4 3 5
19. et sin t = (1 + t + t2! + t3! + t4! + )(t t3! + t5! ) = t + (1)t2 + ( 1
3! +
1 3 1 1 4 1 1 1 1 5 2 1 3 1 5
2! )t + ( 3! + 3! )t + ( 5! 2! 3! + 4! )t = t + t + 3 t 30 t .
21. The ratio test gives infinite radius of convergence. Let f (t) be the func-
tion defined by the given power series. Then
X n+1 n
f (t) = (1)n t
n=0
n!
X n 1
= (1)n + tn
n=0
n! n!
X 1 X 1
= (1)n tn + (1)n tn
n=1
(n 1)! n=0
n!
X tn X (t)n
= t (1)n+1 +
n=0
n! n=0 n!
= tet + et
23. It is easy to check that the interval of convergence is (1, 1). Let f (t) be
the function defined by the given power series. Then
tn+1
Z X
f (t) dt = (n + 1) +c
n=0
n+1
X
= t tn + c
n=0
t
= + c.
1t
Differentiation gives
1
f (t) = .
(1 t)2
124 1 Solutions
25. It is not hard to see that the interval of convergence is (1, 1). Let f (t)
be the given power series. A partial fraction decomposition gives
1 1 1 1
= .
(2n + 1)(2n 1) 2 2n 1 2n + 1
Therefore
1 X t2n+1 1 X t2n+1
f (t) = .
2 n=0 2n 1 2 n=0 2n + 1
t2n+1 t2n+1
P P
Let f1 (t) = n=0 2n1 and f2 (t) = n=0 2n+1 . Then f1 (t) = t +
t2 1+t
2 ln 1t by Exercise 24. Observe that f2 (0) = 0 and
X
f2 (t) = t2n
n=0
1 1 1 1 1
= 2
= + .
1t 21+t 21t
1 1+t
Integration and the fact that f2 (0) = 0 gives f2 (t) = 2 ln 1t . It follows
now that
1
f (t) = (f1 (t) f2 (t))
2
t2 1 + t 1 1 + t
1
= t + ln ln
2 2 1t 2 1t
t t2 1 1 + t
= + ln .
2 4 1t
Pn n
27. The binomial theorem: (a + b)n =
k nk
k=0 k a b .
cn+1
29. The ratio if n is even and 2 if n is odd. Thus lim cn+1
is 1
does
cn 2 n cn
not exist. The ratio test does not apply. The root test gives that cn is
n
1 if n is odd and n 2 if n is even. As n approaches both even and odd
terms approach 1. It follows that the radius of convergence is 1.
1
31. Suppose f (n) (t) = e t pn ( 1t ) where pn is a polynomial. Then f (n+1) (t) =
1 1 1
e t t12 pn ( 1t )+pn ( 1t ) 1 = e t pn+1 ( 1t ), where pn+1 (x) = x2 (pn (x)
t2 e
t
1
pn (x)). By mathematical induction it follows that f (n) (t) = e t pn 1t
for all n = 1, 2, . . ..
33. Since f (t) = 0 for t 0 clearly lim f (n) (t) = 0 The previous problems
t0
imply that the right hand limits are also zero. Thus f (n) (0) exist and is
0.
1 Solutions 125
Section 7.2
cn
cn tn . Then cn+2 =
P
1. Let y(t) = (n+2)(n+1) . Consider even and odd
n=0
c0 c1 t2n
P
cases to get c2n = (2n)! and c2n+1 = (2n+1)! . Thus y(t) = c0 (2n)! +
n=0
t2n+1
P
c1 (2n+1)! = c0 cosh t + c1 sinh t. (see Example 7.1.7) We observe that
n=0
the characteristic polynomial is s2 1 = (s 1)(s + 1) so {et , et } is
t t t t
a fundamental set. But cosh t = e +e 2 and sinh t = e e
2 ; the set
{cosh t, sinh t} is also a fundamental set.
cn tn . Then cn+2 (n + 2)(n + 1) + k 2 cn = 0 or cn+2 =
P
3. Let y(t) =
n=0
2
k cn
(n+2)(n+1) . We consider first the even case.
2
n=0 c2 = k(21
c0
2
k 4 c0
n=2 c4 = k43c2 = 4!
6
n=4 c6 = k 6!c0
.. ..
. .
2n
From this it follows that c2n = (1)n k(2n)!
c0
. The odd case is similar. We
2n+1
k
get c2n+1 = (1)n (2n+1)! . The power series expansion becomes
X
y(t) = c n tn
n=0
X k 2n t2n
= c0 (1)n
n=0
(2n)!
X k 2n+1 t2n+1
+ c1 (1)n
n=0
(2n + 1)!
= c0 cos kt + c1 sin kt.
cn tn . Then the recurrence relation is
P
5. Let y(t) =
n=0
or
n2
cn+2 = cn .
n+2
126 1 Solutions
Since there is a difference of two in the indices we consider the even and
odd case. We consider first the even case.
n=0 c2 = c0
n=2 c4 = 04 c2 = 0
n=4 c6 = 26 c4 = 0
.. ..
. .
n=1 c3 = 13 c1
1 1
n=3 c5 = 5 c3 = 53 c1
3 1
n=5 c7 = 7 c5 = 75 c1
n=7 c9 = 59 c5 = 97
1
c1
.. ..
. .
c1
From this we see that c2n+1 = (2n+1)(2n1) . Thus
X X t2n+1
c2n+1 t2n+1 = c1
n=0 n=0
(2n + 1)(2n 1)
t2n+1
P
and hence y1 (t) = (2n+1)(2n1) . By Exercise 7.1.25 we can write y1
n=0
as
t2 1
t 1+t
y1 (t) = ln .
2 4 1t
The general solution is
t2 1
2 t 1t
y(t) = c0 (1 t ) c1 + ln .
2 4 1+t
2 n1
cn+2 = cn+1 cn .
n+2 (n + 2)(n + 1)
1 Solutions 127
In general,
1
cn = c0 , n = 2, 3, . . . .
n!
We now get
X
y(t) = c n tn
n=0
X
= c0 + c1 t + c n tn
n=2
n
X t
= (c1 c0 )t + c0 + c0 t + c0
n=2
n!
= (c1 c0 )t + c0 et
= c0 (et t) + c1 t.
cn tn . Then the recurrence relation is
P
9. Let y(t) =
n=0
(n 2)(n 3)
cn+2 = cn .
(n + 2)(n + 1)
n=0 c2 = 62 c0 = 3c0
n=2 c4 = 0
n=4 c6 = 0
.. ..
. .
Hence
X
c2n t2n = c0 + c2 t2 = c0 (1 3t2 ).
n=0
n=1 c3 = 13 c1
n=3 c5 = 0
n=5 c7 = 0
.. ..
. .
Hence
X t3
c2n+1 t2n+1 = c1 t + c3 t3 = c1 (t ).
n=0
3
t3
y(t) = c0 (1 3t2 ) + c1 (t ).
3
Only the odd powers of i contribute to the imaginary part. It follows that
X
2
n
n+1
sin(n + 1)x = Im cosn+1(2j+1) x(i2j+1 ) sin2j+1 x
j=0
2j + 1
X
n
2
n+1
= (1)j cosn2j x(1 cos2 x)j sin x,
j=0
2j + 1
where we use the greatest integer function x to denote the greatest inte-
ger less than or equal to x. Now replace t = cos x and using the definition
P n2 n+1 n2j
(1)j 2j+1
sin xUn (cosx) = sin(n + 1)x to get Un (t) = j=0 t (1
2 j
t ) . It follows that sin nx is a product of sin x and a polynomial in cos x.
15. We have
1 Solutions 129
sin((n + 2)x) = sin((n + 1)x + x) = sin((n + 1)x) cos x + cos((n + 1)x) sin x
and hence
(sin x)Un+1 (cos x) = (sin x)Un (x) cos x + (sin x)Tn+1 (cos x).
17. By using the sum and difference formula it is easy to verify the following
trigonometric identity:
and hence
2(cos x)Un (cos x)/ sin x = Un+1 (cos x)/ sin x + Un1 (cos x)/ sin x.
19. By using the sum and difference formula it is easy to verify the following
trigonometric identity:
and hence
2Un1 (cos x) = Tn1 (cos x) Tn+1 (cos x).
Now let t = cos x, replace n by n + 1, and divide by 2.
Section 7.3
t
1. The function 1t 2 is analytic except at t = 1 and t = 1. The function
1
1+t is analytic except at t = 1. It follows that
t = 1 and t = 1 are
t t
the only singular points. Observe that (t 1) 1t2 = 1+t is analytic
130 1 Solutions
1
at 1 and (t 1)2 is analytic at t = 1. It follows that 1 is a regular
1+t
t t
singular point. Also observe that (t + 1) 1t 2 = 1t is analytic at 1
2 1
and (t + 1) 1+t = (1 + t) is analytic at t = 1. It follows that 1 is
a regular singular point. Thus 1 and 1 are regular points.
t
3. Both 3t(1 t) and 1e
t are analytic. There are no singular points and
hence no regular points.
13. Let y(t) = t2 v(t). Then y (t) = 2t3 v(t) + t2 v (t) and y (t) =
6t4 v(t) 4t3 v (t) + t2 v (t). From which we get
r=3:
n odd cn = 0
m
n = 2m c2m = 3c0 (1) (2m+2)
(2m+3)!
P (1)m (2m+2)t2m+3
y(t) = 3c0 m=0 (2m+3)! = 3c0 (sin t t cos t).
r=0: One is lead to the equation 0c3 = 0 and we can take c3 = 0. Thus
n odd cn = 0
m+1
n = 2m c2m = c0 (1) (2m)!
(2m1)
P (1)m+1 (2m1)t2m
y(t) = c0 m=0 (2m)! = c0 (t sin t + cos t).
General Solution: y = c1 (sin t t cos t) + c2 (t sin t + cos t).
n1 n2 cn ncn1 = 0.
1
This is easy to solve. We get cn = n! c0 and hence
X 1 n+1
y(t) = c0 t = c0 tet .
n=0
n!
We write out the power series for t2 et and add corresponding coefficients
to get
132 1 Solutions
1
n2 cn ncn1 + .
(n 1)!
The following list is a straightforward verification:
n=1 c1 = 1
1 1
n=2 c2 = 2 1+ 2
1 1 1
n=3 c3 = 3! 1+ 2 + 3
1 1 1 1
n=4 c4 = 4! 1+ 2 + 3 + 4 .
1
Let sn = 1 + 2 + + n1 . Then an easy argument gives that
sn
cn = .
n!
We now have a second independent solution
X s n tn
y2 (t) = tet ln t t .
n=1
n!
General Solution:
!
X s n tn
y = c1 tet + c2 tet ln t t .
n=1
n!
n = 0 c0 (r 2)(r + 1) = 0
n 1 cn (n + r 2)(n + r + 1) = cn1 (n + r 1)
r=2:
(1)n (n + 1)
cn = 6c0 , n1
(n + 3)!
P (1)n (n+1)tn t
y(t) = 6c0 n=0 (n+3)! = 6c0 (t+2)e
t + t2
t .
r=-1: The recursion relation becomes cn (n 3)(n) = cn1 (n 2) = 0.
Thus
n = 1 c1 = c20
n = 2 c2 = 0
n = 3 0c3 = 0
We can take c3 = 0 and then cn = 0 for all n 2. We now have y(t) =
c0 t1 (1 2t ) = c20 2t
t .
t
(t+2)e
General Solution: y = c1 2t
t + c2 t .
1 Solutions 133
Substitution leads to
X
t3 + 2t2 + (c1 2c0 )t + (cn (n)(n 2) + cn1 (n 3))tn = 0
n=2
n=1 c1 2c0 = 0
n=2 2 c1 = 0
n=3 1 + 3c3 = 0
n4 n(n 2)cn + (n 3)cn3 = 0.
gives
2(1)n
cn = .
n!(n 2)
A second independent solution is
!
2 t3 X 2(1)n tn
y2 (t) = t ln t + 1 + 2t + .
3 n=4 n!(n 2)
(n 2)(n 2 + 2i)
cn = cn1 .
n(n + 2i)
n = 1 c1 = c0
n 2 cn ((n + i)2 + 1) + cn1 (2n 2i + 1) + cn2 = 0
n=1 c1 = c0
1
n=2 c2 = 2! c0
1
n=3 c3 = 3! c0
1
n=4 c4 = 4! c0 .
X
y(t) = cn tn+i
n=0
n+i
X t
= c0
n=0
n!
i t
= c0 t e .
Section 7.4
1. Let
X
X
y(t) = t2k+1 c n tn = cn tn+2k+1
n=0 n=0
Then
X
ty (t) = (n + 2k)(n + 2k + 1)cn tn+2k
n=0
X
2ity (t) = 2i(n + 2k)cn1 tn+2k
n=1
X
2ky (t) = 2k(n + 2k + 1)cn tn+2k
n=0
X
2iky(t) = 2ikcn1tn+2k
n=1
By assumption the sum of the series is zero. The n = 0 terms in the first
and third sum give (2k)(2k+1)c0 2k(2k+1)c0 = 0. Thus we can start all
the series at n = 1. For n 1 we get n(n + 2k + 1)cn + 2i(n + k)cn1 = 0
which implies
2i(n + k)
cn = cn1 .
n(n + 2k + 1)
Since c0 6= 0 it follows from this recursion relation that cn 6= 0 for all
n 0. Therefore the Frobenius solution y(t) is not a polynomial.
s
5. Let g(t) = L1 . Then
(s2 1)k
d d s
L {tg(t)} = (L {g(t)}) =
ds ds (s 1)k
2
Divide by 2k, solve for the second term in the last line, and apply the
inverse Laplace transform to get
1 1 t (2k 1) 1 1
L = g(t) L
(s2 1)k+1 2k 2k (s2 1)k
t 1 s (2k 1) 1 1
= L L .
2k (s2 1)k 2k (s2 1)k
n = (1)n n .
Pk+1 n1
t
Let f (t) = n=1 n (n1)! . Then
1
L1 = f (t)et f (t)et .
(s2 1)k+1
1 53
c1 (0) = 2 c3 (0) = 5c2 (0) = 2
3 753
c2 (0) = 3c1 (0) = 2 c4 (0) = 7c3 (0) = 2
Inductively, we get
(1)k
ck (0) = (2k 1) (2k 3) (2k 5) 1
2
(1)k (2k)!
=
2 2k k!
(1)k (2k)!
= .
2k+1 k!
2. From Exercise 4 it is easy to see that dk (t) = tck1 and so dk (0) =
(1)k1 (2(k1))!
ck1 (0) = 2k (k1)!
.
Section 8.1
1 1 1 3 2 8
1. B + C = 1 7, B C = 5 1, and 2B 3C = 13 6
0 3 2 1 5 1
3 1 7
3 4
3. A(B + C) = AB + AC = , (B + C)A = 3 1 25
1 13
5 0 12
6 4 1 8
5. AB = 0 2 8 2
2 1 9 5
8 0
4 5
7. CA = 8 14
10 11
8 9 48
9. ABC = 4 0 48 .
2 3 40
0 0 1
15. 3 5 1
0 0 5
0 1 0 0
17. (a) Choose, for example, A = and B = .
0 0 1 0
2 2 2
(b) (A + B) = A + 2AB + B precisely when AB = BA.
n 1 n
19. B =
0 1
0 1 v 1 c
21. (a) A = 2 ; the two rows of A are switched. (b) A =
1 0 v1 0 1
v1 + cv2
; to the first row is added c times the second row while the
v2
second row is unchanged, (c) to the second row is added c times the first
row while the first row is unchanged. (d) the first row is multiplied by a
while the second row is unchanged, (e) the second row is multiplied by a
while the first row is unchanged.
23.
140 1 Solutions
F (1 )F (2 )
cosh 1 sinh 1 cosh 2 sinh 2
=
sinh 1 cosh 1 sinh 2 cosh 2
cosh 1 cosh 2 + sinh 1 sinh 2 cosh 1 sinh 2 + sinh 1 cosh 2
=
sinh 1 cosh 2 + cosh 1 sinh 2 sinh 1 sinh 2 + cosh 1 cosh 2
cosh(1 + 2 ) sinh(1 + 2 )
=
sinh(1 + 2 ) cosh(1 + 2 )
= F (1 + 2 ),
We used the addition formulas for sinh and cosh in the second line.
Section 8.2
1 4 3 2 1 4 3 2
x
1 1 1 4 1
, x = y, b = , and [A|b] = 1 1 4
1. A =
2
.
0 1 1 2 0 1 1
z
0 1 1 6 0 1 1 6
x1 x3 + 4x4 + 3x5 = 2
5x1 + 3x2 3x3 x4 3x5 = 1
3.
3x1 2x2 + 8x3 + 4x4 3x5 = 3
8x1 + 2x2 + 2x4 + x5 = 4
5. RREF
0 1 0 3
7. m2 (1/2)(A) = 0 0 1 3
0 0 0 0
1 0 1 0 3
9. t1,3 (3)(A) = 0 1 3 4 1
0 0 0 0 0
1 0 0 11 8
11. 0 1 0 4 2
0 0 1 9 6
1 2 0 0 3
0 0 1 0 2
13. 0 0 0 1 0
0 0 0 0 0
1 Solutions 141
1 0 2
0 1 1
15. 0 0 0
0 0 0
0 0 0
1 4 0 0 3
17. 0 0 1 0 1
0 0 0 1 3
x 1 3
19. y = 1 + 1, R
z 0 5
x 2
21. = ,R
y 1
x 14/3
23. y = 1/3
z 2/3
0 1
25. 3 + 0, R
4 0
27.
1
29. 0
1
34
40
31.
39
1
5 1 1
33. The equation 1 = a 1 + b 1 has solution a = 2 and b = 3. By
4 2 0
5
Proposition 7 1 is a solution.
4
142 1 Solutions
7/2 3/2
35. If xi is the solution set for Ax = bi then x1 = 7/2, x2 = 3/2,
3/2 1/2
7
and x3 = 6.
3
Section 8.3
4 1
1.
3 1
3. not invertible
5. not invertible
6 5 13
7. 5 4 11
1 1 3
29 39/2 22 13
7 9/2 5 3
9.
22 29/2 17 10
9 6 7 4
0 0 1 1
1 0 0 0
11.
0 1 1 1
1 1 0 1
4 1 2 5
13. x = A1 b = =
3 1 3 3
2 4 4 2 16
1 1
15. x = A1 b = 10 2 1 4 1 = 10 11
6 2 2 2 18
58 39 44 26 1 19
1 1 14
9 10 6 0 4
17. x = A b = 2 =
44 29 34 20 1 15
18 12 14 8 2 6
21. F ()1 = F ()
1 Solutions 143
Section 8.4
1. 1
3. 10
5. 21
7. 2
9. 0
1 s3 1
11. s2 6s+8 s = 2, 4
1 s3
(s 1)2
3 s1
1
13. (s1) 3
0 (s 1)2 0 s = 1
0 3(s 1) (s 1)2
2
s + s 4s + 4 0
15. s3 +s21+4s+4 s 1 s2 + s 0 s = 1, 2i
s 4 4s + 4 s2 + 4
17. no inverse
4 4 4
19. 81 1 3 1
5 1 3
2 98 9502
21. 61 0 3 297
0 0 6
55 95 44 171
1 50
85 40 150
23. 15 70 125 59 216
65 115 52 198
2 1 1 2
25. det A = 1, det A(1, b) = det = 5, and det A(2, b) = det =
3 4 3 3
3. It follows that x1 = 5/1 = 5 and x2 = 3/1 = 3
144 1 Solutions
2 0 2
27. det A = 10, det A(1, b) = det 1 2 0 = 16, det A(2, b) =
2 2 1
1 2 2 1 0 2
det 2 1 0 = 11, and det A(3, b) = det 2 2 1 = 18. It
1 2 1 1 2 2
follows that x1 = 16/10, x2 = 11/10, and x3 = 18/10.
Section 8.5
Section 9.2
7. First note that y1 (0) = 0 and y2 (0) = 1, so the initial condition is satis-
fied. Then t
e 3e3t
y1 (t)
y (t) = =
y2 (t) 2et 3e3t
while
t
5(e e3t ) 2(2et e3t )
5 2
y(t) =
4 1 4(et e3t ) (2et e3t )
t
e 3e3t
= .
2et 3e3t
5 2
Thus y (t) = y(t), as required.
4 1
9. First note that y1 (0) = 1 and y2 (0) = 3, so the initial condition is satis-
fied. Then
146 1 Solutions
y1 (t) et + et + tet
y (t) = =
y2 (t) 3et + et + tet
while
2 1 et + tet
t t
2e + 2tet 3et tet + et
e
+ t =
3 2 3et + tet e 3et + 3tet 6et 2tet + et
et + tet + et
= .
3et + tet + et
t
2 1 e
Thus y (t) = y(t) + t , as required.
3 2 e
y y
In solutions, 1115, y = 1 = .
y2 y
11. Let y1 = y and y2 = y . Then y1 =y =
y2 and y2 = y = 5y 6y +
y
e2t = 6y1 5y2 + e2t . Letting y = 1 , this can be expressed in vector
y2
form as
0 1 0 1
y = y + 2t , y(0) = .
6 5 e 2
2 1
15. Let y1 = y and y2 = y . Then
y1 = y = y2 and y2 = y = t y t2 y =
y
t12 y1 2t y2 . Letting y = 1 , this can be expressed in vector form as
y2
0 1 2
y = y, y(1) = .
t12 2t 3
3e3t 1
17. A (t) =
2t 2e2t
1
19. y (t) = 2t
t1
e2 e2 e2 + e2 2
1
23. 2 2
42e e e2 e2
4 8
25.
12 16
"1 1
#
s s2
27. 2 1
s3 s2
" 3! 2s 1
#
s4 (s2 +1)2 (s+1)2
29. 2s s3 3
s3 s2 6s+13 s
1 1
2
31. s2 1
1 1
33. 1 2t 3t2
2s 2 1 1 1 1
2 2
+ +
s 1 s 1 s + 1 s 1 s + 1 s 1
35. We have = . Laplace
2 2s 1 1 1 1
2 2
+ +
s "1 s 1 s +#1 s 1 s+1 s1
et + et et et
inversion gives t
e et et + et
Section 9.3
1.
1 0
A =
0 2
2 1 0
A =
0 4
3 1 0
A =
0 8
..
.
n 1 0
A = .
0 (2)n
A2 2 A3 3
eAt = I + At + t + t +
2! 3!
1 t2 1 tn
1 0 t 0 0 0
= + + + + +
0 1 0 2t 2! 0 (2t)2 n! 0 (2t)n
t
e 0
= .
0 e2t
3.
0 1
A =
1 0
1 0
A2 = =I
0 1
0 1
A3 = =A
1 0
5.
1 1
A =
1 1
2 2
A2 =
2 2
4 4
A3 =
4 4
..
.
2n1 2n1
An = .
2n1 2n1
A2 2 A3 3
eAt = I + At + t + t +
2! 3! 2
1 2t 2t2 1 2n1 tn 2n1 tn
1 0 t t
= + + + + +
0 1 t t 2! 2t2 2t2 n! 2n1 tn 2n1 tn
7.
150 1 Solutions
0 1 0
A = 1 0 0
0 0 2
1 0 0
A2 = 0 1 0
0 0 4
0 1 0
A3 = 1 0 0
0 0 8
1 0 0
A4 = 0 1 0
0 0 16
0 1 0
A5 = 1 0 0
0 0 32
The (1, 1) entry and the (2, 2) entry of eAt are equal and are
t2 t4
cos t = 1 + .
2! 4!
The (1, 2) entry and the (2, 1) entry of eAt have opposite signs. The (2, 1)
entry is
t3 t5
sin t = t + .
3! 5!
The (3, 3) entry is
(2t)2
e2t = 1 + 2t + + .
2!
All other entries are zero thus
cos t sin t 0
eAt = sin t cos t 0
0 0 e2t
s1 5 2
" # " #
(s1)2 +1 (s1)2 +1 (s1)2 +1 0
(sI A)1 = 1 s1 + 2 .
(s1)2 +1 (s1)2 +1 0 (s1)2 +1
Therefore
t
e cos t + 2et sin t 5et sin t
eAt =
et sin t et cos t 2et sin t
s 1 1
13. The characteristic polynomial is cA (s) = s3 and sI A = 0 s 1.
0 0 s
1 1 s+1 2
s s2 s3 1 t t + t2
Thus (sI A)1 = 0 1s 1
s2 and e
At
= 0 1
t
1 0 0 1
0 0 s
0 1 cos t sin t
15. Let M = and N = 2. Then by Example 6 eMt = .
1 0 sin t cos t
Thus
Mt
At e 0
e =
0 eN t
cos t sin t 0
= sin t cos t 0 .
0 0 e2t
Section 9.4
eAt = M et + N tet .
Differentiating we obtain
I = M
A = M + N.
M = I
N = A I.
Thus,
M1 = I
M2 = A.
Thus " #
At
1 + 2t t
e = .
4t 1 2t
I = M
A = M + N.
M = I
N = A I.
Thus,
I = M +N
A = 2M 2N.
1 1
eAt = (A + 2I)e2t (A 2I)e2t
4 4
1 7 11 2t 1 11 11 2t
= e e
4 7 11 4 7 7
1 7e2t + 11e2t 11e2t 11e2t
=
4 7e2t + 7e2t 11e2t 7e2t
2 2 2
2t= s 4s2t+ 13 = ((s 2) + 3 ).
9. The characteristic polynomial is cA (s)
The standard basis of EcA is BcA = e cos 3t, e sin 3t . It follows that
I = M
A = 2M + 3N.
M = I
1 8 13
N = (A 2I) = .
3 5 8
Thus,
1
eAt = Ie2t cos 3t + (A I)e2t sin 3t
3
2t
8e2t sin 3t 13e2t sin 3t
e cos 3t 0
= +
0 e2t cos 3t 5e2t sin 3t 8e2t sin 3t
2t
e cos 3t + 8e2t sin 3t 13e2t sin 3t
=
5e2t sin 3t e2t cos 3t 8e2t sin 3t
2
11. The characteristic
2t 2t polynomial is cA (s) = (s + 2) . The standard basis is
BcA = e , te . It follows that
I = M
A = 2M + N.
M = I
1 1
N = A + 2I = .
1 1
Thus,
eAt = M + N et + P et .
M +N +P = I
N P = A
N + P = A2
M = I A2
A2 + A
N =
2
A2 A
P = .
2
Thus,
eAt = M + N et + P et
2 0 1 0 0 0 1 0 1
= 0 0 0 + 0 1 0 et + 0 0 0 et
2 0 1 0 0 0 2 0 2
t t
2e 0 1 + e
= 0 et 0
2 2et 0 1 + 2et
Therefore
eAt = M et + N et sin t + P et cos t.
Differentiating twice and simplifying we get the system:
I = M +P
A = M +N +P
A2 = M + 2N
N = AI
M = A2 2A + 2I
P = A2 + 2A I.
1 1
2 1 2
A straightforward calculation gives A2 = 2 0 2 and
1 1
2 1 2
0 21
1 1
1 1
0 2 0 2 2 0 2
N = 1 0 1 , M = 0 0 0 , and P = 0 1 0 .
1 1 1 1 1
0 2 0 2 0 2 2 0 2
Hence,
1
0 12 0 12
1 1
2 0 2 0 2
eAt = 0 0 0 et + 1 0 1 et sin t + 0 1 0 et cos t
1 1 1 1 1
2 0 2 0 2 0 2 0 2
t
e + et cos t et sin t et et cos t
1
= 2et sin t 2et cos t 2et sin t .
2 t
e e cos t et sin t et + et cos t
t
17. In this case BcA = {et , cos 2t, sin 2t}. It follows that
M +N = I
M + 2P = A
M 4N = A2
Thus,
19. In this case BcA = {cos t, sin t, t cos t, t sin t}. It follows that
M = I
N +P = A
M + 2Q = A2
N 3P = A3
Thus,
21. The standard basis is BcA = {ert , tert } so that eAt = M ert + N tert .
Fulmers method gives
I = M
A = rM + N
M =I and N = (A rI).
Hence,
eAt = Iert + (A rI)tert = (I + (A rI)t) ert . (1)
1 Solutions 159
Section 9.5
et
0
1. It is easy to see that eAt = . Thus
0 e3t
t
et
e 0 1
y(t) = = .
0 e3t 2 2e3t
and
eAt = M1 e2t + M2 te2t .
Differentiating and evaluating at t = 0 gives
I = M1
A = 2M1 + M2 .
e2t te2t
At
and hence M1 = I and M2 = A 2I. We thus get e = and
2t 0 e2t
te2t 1
2t
e + 2te2t
e
y(t) = =
0 e2t 2 2e2t
I = M1 + M2
A = M1 + M2 .
and
hence M1 = 21 (A +I) and M2 = 1
2 (A + I). We thus get eAt =
t t
1 3e e et + et
t t and
2 3e 3e et + 3et
1 3et et et + et 1
t
e
y(t) = = .
2 3et 3et et + 3et 3 3et
I = M1
A = M1 + M2 .
et + 2tet 4tet
At
and hence M1 = I and M2 = AI. We thus get e =
t te e 2tet
t t
I = M1 + M2
A = M1 + 2M3
A2 = M1 4M2 .
and hence
0 0 0
1 2
M1 = (A + 4I) = 0 0 0
5 1 0 1
1 0 0
1 2
M2 = (A I) = 0 1 0
5 1 0 0
0 2 0
1 2
M3 = (A + 5A + 4I) = 1/2 0 0 .
10
0 2 0
cos 2t 2 sin 2t 0
It follows that
1 Solutions 161
y h (t) = eAt y 0
t 1 0 t 0 1 1
= e cos 2t + e sin 2t
0 1 1 0 0
t
e cos 2t
= .
et sin 2t
and
y p = eAt f (t)
t 1 0 t 0 1 5
= e cos 2t + e sin 2t
0 1 1 0 0
5 0
= et cos 2t 1 + et sin 2t 1
0 5
1 0
= (1 et cos 2t + 2et sin 2t) + (2 2et cos 2t et sin 2t)
0 1
1 et cos 2t + 2et sin 2t
= .
2 + 2et cos 2t + et sin 2t
y(t) = y h + y p
1 + 2et sin 2t
=
2 + 2et cos 2t
It follows that
y h (t) = eAt y 0
t 1 0 t 0 2 2
= e cos 2t + e sin 2t
0 1 1/2 0 1
t 2 t 2
= e cos 2t + e sin 2t
1 1
and
162 1 Solutions
y p = eAt f (t)
t 1 0 t 0 2 4
= e cos 2t + e sin 2t
0 1 1/2 0 1
4 2
= ((et cos 2t) 1) + ((et sin 2t) 1)
1 2
1 t t
4 1 t t
2
= 1 e cos 2t + 2e sin 2t + 2 2e cos 2t e sin 2t
5 1 5 2
2et sin 2t
= .
1 et cos 2t
y(t) = y h + y p
2et sin 2t
t 2 t 2
= e cos 2t + e sin 2t +
1 1 1 et cos 2t
2et cos 2t + 4et sin 2t
= .
1 + et sin 2t 2et cos 2t
It follows that
y h (t) = eAt y 0
1 0 4 2 0
= et + tet
0 1 8 4 1
t
2te
=
et 4tet
and
y p = eAt f (t)
t 1 0 t 4 2 1
= e + te t
0 1 8 4 2
t 1
= e t
2
t 1
= (e t 1) .
2
y(t) = y h + y p
2tet
t 1
= + (e t 1)
et 4tet 2
2tet + et t 1
=
4tet et + 2t + 2
Clearly
0
y h (t) = 0
0
while
y1 = 2 2y1
y2 = 2y1 y2
2 0
with initial conditions y1 (0) = 4 and y2 (0) = 0. Let A = ,
2 1
2 4
f (t) = , and y(0) = . We need to solve the system y = Ay + f ,
0
0
4
y(0) = . It is easy to check that
0
0 0 t 1 0 2t
eAt = e + e .
2 1 2 0
It follows that
0 0 t 1 0 2t 4 0 t 4 2t
y h (t) = e + e = e + e
2 1 2 0 0 8 8
and
At 2 t 0 0 2 2t 1 0 2
yp = e = (e 1) + (e 1)
0 2 1 0 2 0 0
0 1
= (1 et ) + (1 e2t )
4 2
1 0 t 1 2t
= + e + e .
2 4 2
y1 = 2 + 3y2 5y1
y2 = 2 + 5y1 7y2
1 Solutions 165
5 3
with initial conditions y1 (0) = 0 and y2 (0) = 0. Let A = ,
5 7
2 0
f (t) = , and y(0) = . We need to solve the system y = Ay + f ,
2 0
0
y(0) = . It is easy to check that
0
e2t 5 3 e10t
At 3 3
e = + .
8 5 3 8 5 5
Clearly y h = 0 while
Thus
y1 (t) = 1 e2t
y2 (t) = 1 e2t .
Section 9.6
0 = Du + Ev + F
= D(ax + by) + E(cx + dy) + F
= (Da + Ec)x + (Db + Ed)y + F
= D x + E y + F,
13. Let C be the graph of a power curve in the (u, v) plane and P (C) the
transform of C.Let
(x,y)be a point of P (C)
and (u,
v) the point on
u x 1 a b u ax + by
C such that P = . If P = then = .
v y c d v cx + dy
Replace u and v in the equation Au + Bv = (Cu + Dv)p by ax + by
and cx + dy, respectively. We then get (Aa + Bc)x + (Ab + Bd)y =
((Ca + Dc)x + (Cb + Dd)y)p . Thus P (C) is the graph of a power curve.
1 Solutions 167
15. The characteristic polynomial takes the form cA (s) = s2 (tr A)s+det A.
Let = tr A. Since det A = 0 we have cA (s) = s2 s = s(s ). Now
consider two cases:
6= 0: In this case A has two distinct eigenvalues, 0 and . Let v1 be an
eigenvector with eigenvalue 0 and v2 an eigenvector with eigenvalue
. Then v1 and v2 are linearly independent. If P = v1 v2 then P
is invertible and
0 0
AP = Av1 Av2 = 0 v2 = v1 v2 = P J1 .
0
Av2 = A2 v1 = 0.
Now let P = v1 v2 . Then
0 0 0 0
AP = A v1 v2 = Av1 Av2 v
= 2 0 v v
= 1 2 =P .
1 0 1 0
21. Assume c1 > 0 and thus x > 0 (the case where c1 < 0 is similar). We
ln x/c1 1 c2
have y = + + . It follows that lim y = if > 0 and
c1 x0+
lim y = if < 0.
x0+
1
23. y = and the result follows.
x
Section 9.7
t
2e2t 2 1 et e2t
e
1. Observe that (t) = = = A(t)(t).
et 8e2t 4 3 et 4e2t
Also, det (t) = 4et et = 3et 6= 0. Thus (t) is a fundamental matrix.
The general
solution can be written in the form y(t) = (t)c, where
c1
c= is a constant vector. The initial condition implies y(0) = (0)c
c2
or
1 1 1 c1 c1 + c2
= = .
2 1 4 c2 c1 + 4c2
Solving for c we get c1 = 2 and c2 = 1. It follows that
t 2t t
2e e2t
e e
y(t) = (t)c = 2 t = .
e 4e2t 2et 4e2t
3. Observe that
1 Solutions 169
Also, det (t) = sin2 (t2 /2) cos2 (t2 /2) = 1 6= 0. Thus (t) is a
fundamental matrix. The general
solution can be written in the form
c1
y(t) = (t)c, where c = is a constant vector. The initial condition
c2
implies y(0) = (0)c or
1 0 1 c1 c
= = 2 .
0 1 0 c2 c1
5. Observe that
cos t + t sin t sin t t cos t
(t) =
sin t + t cos t cos t + t sin t
1/t 1 t cos t t sin t
=
1 1/t t sin t t cos t
= A(t)(t).
7. Observe that
0 tet
(t) =
0 et
1 (t 1)et
1 1
=
1/t 1/t 1 et
= A(t)(t).
1 (t 1)et (t 1)et1 3
1
y(t) = (t)c = 3 + = .
1 e et et1 + 3
It follows that BcA = {eu , ueu }. Using Fulmers method we have eAu =
eu M1 + ueu M2 . Differentiating and evaluating at u = 0 gives
I = M1
A = M1 + M2 .
1 0 1 1
It follows that M1 = I = and M2 = A I = . Thus
0 1 1 1
Au u 1 0 u 1 1
e =e + ue .
0 1 1 1
y h (t) = (t)y(1)
t t ln t t ln t 2 2t 2t ln t
= = .
t ln t t + t ln t 0 2t ln t
It follows that
2t 2t ln t t ln t 2t t ln t
y(t) = y h (t) + y p (t) = + = .
2t ln t t ln t t ln t
3 5
11. Let A = . Then A(t) = sec(t)A. The characteristic polynomial
1 3
of A is cA (s) = s2 4 = (s 2)(s + 2). Hence BcA = e2t , e2t and
and
1 5e2u e2u 5e2u 5e2u
Au
e = .
4 e2u + e2u e2u + 5e2u
Rt
If b(t) = 0 sec u du = ln |sec t + tan t| then (t) = eAb(t) . If X = (sec t +
tan t)2 then X 1 = (sec t tan t)2 and
1 1
5X 5X 5
1 X X
(t) =
4 1 1
X + X + 5
X X
sec t + 3 sec t tan t + tan2 t
2
5 sec t tan t
= .
sec t tan t sec2 t 3 sec t tan t + tan2 t
13. Let v1 (t) and v2 (t) denote the volume of brine in Tank 1 and Tank
2, respectively. Then v1 (t) = v2 (t) = 2 t. The following differential
equations describe the system
1 Solutions 173
3 1
y1 (t) = y1 (t) + y2 (t) + 6
2t 2t
1 3
y2 (t) = y1 (t) y2 (t) + 0,
2t 2t
with initial conditions y1 (0) = 0 and y2 (0) = 20. In matrix form, y (t) =
A(t)y(t) + f (t), we have
3/(2 t) 1/(2 t) 6 0
A(t) = , f (t) = , y(0) = .
1/(2 t) 3/(2 t) 0 20
1
Let a(t) = . Then we can write A(t) = a(t)A where
2t
3 1
A= .
1 3
2
2t 4t is cA (s) = s 6sAu
The characteristic polynomial + 8 = (s 2)(s 4).
We now have BcA = e , e It follows that e = M1 e2u + M2 e4u .
Differentiating and setting u = 0 we get
I = M1 + M2
A = 2M1 + 4M2 .
and
e2u 1 1 e4u
Au 1 1
e = + .
2 1 1 2 1 1
Rt R t 1
Let b(t) = 0 a(u) du = 0 du = ln 2t2 . Then the standard funda-
2u
mental matrix is
(2 t)2 1 1 (2 t)4
Au 1 1
(t) = e |u=b(t) = + .
8 1 1 32 1 1
y h (t) = (t)y(0)
(2 t)2 1 1 (2 t)4
0 1 1 0
= +
8 1 1 20 32 1 1 20
(2 t)2 5 (2 t)4 5
= + .
2 5 8 5
(2 u)2 + 4 (2 u)2 4
1 2
(u) = ,
(2 u)4 (2 u)2 4 (2 u)2 + 4
(2 u)2 + 4 (2 u)2 4 6
1 2
f (u) =
(2 u)4 (2 u)2 4 (2 u)2 + 4 0
(2 u)2 + 4
12
=
(2 u)4 (2 u)2 4
(2 u)2 + 4(2 u)4
= 12 ,
(2 u)2 4(2 u)4
and
t
3(2 t)2 + 4
4
Z
8
1 (u)f (u) du = .
0 (2 t)3 3(2 t)2 4 4
Finally, we get
Z t
y p (t) = (t) 1 (u)f (u) du
0
Z t
(2 t)2 1 1 (2 t)4
1 1
= + 1 (u)f (u) du
8 1 1 32 1 1 0
(2 t)2 3 (2 t)4
4 1
= (2 t) .
2 2 3 8 1
We now add the homogeneous and particular solutions together and sim-
plify to get
y(t) = y h + y p
(2 t)4 3
4 2 1
= (2 t) + (2 t) + .
2 1 4 3
We now get
1 Solutions 175
3
y1 (t) = 4(2 t) + (2 t)2 (2 t)4
4
3
y2 (t) = 2(2 t) + (2 t)2 + (2 t)3 .
4
The amount of fluid in each tank after 1 minute is v1 (1) = v2 (1) = 1.
Thus the concentrations (grams/L) of salt in Tank 1 is y1 (1)/1 and in
Tank 2 is y2 (1)/1, i.e.
y1 (1) 17 y2 (1) 15
= and = .
1 4 1 4
http://www.springer.com/978-1-4614-3617-1