THE UNIVERSITY OF HONG KONG

DEPARTMENT OF MATHEMATICS

MATH1013 University Mathematics II

Tutorial 10 Solutions

1. (a) We have

y ′ + (x − 2)y 2 = (sin x)y 2

⇔ y ′ = (sin x − x + 2)y 2
1
⇔ dy = (sin x − x + 2) dx
y2
1
⇔ ∫ 2 dy = ∫ (sin x − x + 2) dx
y
1 1
⇔ − = − cos x − x2 + 2x + C
y 2
1
⇔ y= .
cos x + 2 x − 2x − C
1 2

1 2
(b) The ODE can be rewritten as a linear ODE y ′ + 2
y = − 3 . The integrating factor is
x x
µ(x) = e∫ = e− x . Therefore, we have
1 1
x2
dx

1 2
y′ + 2
y=− 3
x x
2 1
−x ′
(e y) = − 3 e− x
1
⇔
x
2 1
e y = − ∫ 3 e− x dx.
−x 1
⇔
x
1 1
Let u = − . Then du = 2 dx. This implies
x x
2 − x1
−∫ e dx = ∫ 2ueu du.
x3
Using integration by parts, this becomes
2 1
2 ∫ u d(eu ) = 2ueu − 2 ∫ eu du = 2ueu − 2eu + C = − e− x − 2e− x + C.
1

x
It follows that
2 1
y=− − 2 + Ce x .
x
4
Putting x = 1 and y = 0, we obtain 0 = −2 − 2 + Ce, and hence C = . Therefore, the
e
solution is
2
− 2 + 4e x −1 .
1
y=−
x

1
2. (a) Equivalently, we need to find α and β such that aα + bβ = −r and cα + dβ = −s. Note that

d(aα + bβ) − b(cα + dβ) = bs − dr.

bs − dr
This implies (ad − bc)α = bs − dr, and hence α = . Also, we have
ad − bc
1 1 −r(ad − bc) a(bs − dr) cr − as
β = (−r − aα) = ( − )= .
b b ad − bc ad − bc ad − bc

(b) We would like to use the result of part (a) with P (x, y) = x − y + 3 and Q(x, y) = x + y − 1.
This means (a, b, c, d, r, s) = (1, −1, 1, 1, 3, −1). Thus, we choose

(−1)(−1) − (1)(3) (1)(3) − (1)(−1)

α= = −1 and β = = 2.
(1)(1) − (−1)(1) (1)(1) − (−1)(1)

Define X = x + 1 and Y = y − 2. By the chain rule,

dY dY dy dx dy
Y′ = = ⋅ ⋅ = (1) ⋅ ⋅ (1) = y ′ .
dX dy dx dX dx

Therefore, the ODE becomes (X − Y )Y ′ = X + Y . This is a homogeneous ODE. Let

Y = vX. Then we have

(X − Y )Y ′ = X + Y
⇔ (X − vX)(v + v ′ X) = X + vX
1+v
⇔ v + v′ X =
1−v
′ 1 + v2
⇔ vX=
1−v
1−v 1
⇔ dv = dX
1 + v2 X
1−v 1
⇔ ∫ dv = ∫ dX.
1 + v2 X
For the left-hand side, we have
1−v 1 v
∫ dv = ∫ dv − ∫ dv
1+v 2 1+v 2 1 + v2
1 1
= tan−1 v − ∫ d(1 + v 2 )
2 1 + v2
1
= tan−1 v − ln (1 + v 2 ) + C1 .
2
1
For the right-hand side, we have ∫ dX = ln ∣X∣ + C2 . Therefore, the implicit solution
X
is given by
1
tan−1 v − ln (1 + v 2 ) = ln ∣X∣ + C.
2
This means
y−2 1 (y − 2)2
tan−1 − ln (1 + ) = ln ∣x + 1∣ + C.
x+1 2 (x + 1)2

2
3. By the chain rule, we have u′ = 3y 2 y ′ . Therefore, we have
x
y ′ + 2y =
y2
2 ′
⇔ 3y y + 6y = 3x
3

⇔ u′ + 6u = 3x.

The integrating factor is µ(x) = e∫ 6 dx

= e6x . Therefore, we have

u′ + 6u = 3x
⇔ (e6x u)′ = 3xe6x
⇔ e6x u = ∫ 3xe6x dx.

Using integration by parts,

1 1 6x 1 1 6x 1 6x
∫ 3xe dx = ∫ x d(e ) = xe − ∫ e dx = xe − e + C.
6x 6x 6x
2 2 2 2 12
1
1 1 1 1
+ Ce−6x and y = u 3 = ( x − + Ce−6x ) .
1 3
Therefore, u = x −
2 12 2 12

Remark. In general, if we want to solve an ODE of the form y ′ + P (x)y = Q(x)y n , we can
use the substitution u = y 1−n to transform it to a linear ODE.

4. (a) The characteristic equation is t2 − 5t + 4 = 0. This gives (t − 1)(t − 4) = 0, and hence there
are two distinct real roots t = 1, 4. By proposition 6.1,

y = Aex + Be4x (1)

for some constants A and B. This implies

y ′ = Aex + 4Be4x . (2)

Putting x = 0, y = 5 and y ′ = 14 in (1) and (2), we obtain 5 = A + B and 14 = A + 4B.

Taking the difference, we obtain 3B = 9, and hence B = 3. Then A = 5 − B = 2. The
solution is
y = 2ex + 3e4x .

(b) We have
lim y(x) = lim (2ex + 3e4x ) = 2(0) + 3(0) = 0
x→−∞ x→−∞

and
lim y(x) = lim (2ex + 3e4x ) = 2(∞) + 3(∞) = ∞.
x→∞ x→∞

5. Consider z = y + a sin x + b cos x. Then z ′ = y ′ + a cos x − b sin x and z ′′ = y ′′ − a sin x − b cos x.

The ODE becomes

(z ′′ + a sin x + b cos x) − 9(z ′ − a cos x + b sin x) + 20(z − a sin x − b cos x) = −442 sin x

3
 i.e.
z ′′ − 9z ′ + 20z + (−19a − 9b + 442) sin x + (9a − 19b) cos x = 0.
19
We choose a and b such that −19a − 9b + 442 = 9a − 19b = 0. This gives a = b, and hence
9
442
− b + 442 = 0. The solution is b = 9 and a = 19. For this pair of a and b, the ODE becomes
9
z ′′ − 9z ′ + 20z = 0.
The characteristic equation is t2 − 9t + 20 = 0, i.e. (t − 4)(t − 5) = 0. As there are two distinct
real roots t = 4, 5, the general solution is z = Ae4x + Be5x for some constants A and B by
proposition 6.1. This means
y = Ae4x + Be5x − 19 sin x − 9 cos x.

Additional problems:
6. (a) We have
y′ = 1 − y2
1
⇔ dy = 1 dx
1 − y2
1
⇔ ∫ dy = ∫ 1 dx
1 − y2
1 1 1
⇔ ∫ ( + ) dy = ∫ 1 dx
2 1+y 1−y
1
⇔ (ln ∣1 + y∣ − ln ∣1 − y∣) = x + C1
2
1+y
⇔ ln ∣ ∣ = 2x + 2C1
1−y
1+y
⇔ = ±e2x+2C1
1−y
1+y
⇔ = Ce2x
1−y
⇔ 1 + y = Ce2x (1 − y)
⇔ y + Ce2x y = Ce2x − 1
Ce2x − 1
⇔ y= .
Ce2x + 1

(b) We have
3x2 + 4x − 4
y′ =
2y − 4
⇔ (2y − 4) dy = (3x2 + 4x − 4) dx
⇔ ∫ (2y − 4) dy = ∫ (3x + 4x − 4) dx
2

⇔ y 2 − 4y = x3 + 2x2 − 4x + C
⇔ (y − 2)2 = x3 + 2x2 − 4x + 4 + C
√
⇔ y = 2 ± x3 + 2x2 − 4x + 4 + C.

4
 (c) We have

y2
y′ =
1 + x2
1 1
⇔ dy = dx
y 2 1 + x2
1 1
⇔ ∫ 2 dy = ∫ dx
y 1 + x2
1
⇔ − = tan−1 x + C
y
1
⇔ y=− .
tan−1 x + C
1
Putting x = 0 and y = 1, we obtain 1 = − , i.e. C = −1. Thus, the solution is
C
1
y=− .
tan−1 x − 1

7. (a) Let y = vx. Then we have

y−x
y′ =
x − 4y
vx − x
⇔ v + v′ x =
x − 4vx
v−1
⇔ v + v′ x =
1 − 4v
4v 2 − 1
⇔ v′ x =
1 − 4v
1 − 4v 1
⇔ dv = dx
4v 2 − 1 x
1 − 4v 1
⇔ ∫ dv = ∫ dx.
4v 2 − 1 x
For the left-hand side, we have the partial fraction decomposition
1 − 4v A B
= + .
4v 2 − 1 2v − 1 2v + 1
1 1
This means 1 − 4v = A(2v + 1) + B(2v − 1). Putting v = , we obtain A = − . Putting
2 2
1 3
v = − , we obtain B = − . Therefore,
2 2
1 − 4v 1 1 3 1
∫ dv = − ∫ ( + ) dv = − (ln ∣2v − 1∣ + 3 ln ∣2v + 1∣) + C1 .
4v − 1
2 2 2v − 1 2v + 1 4
For the right-hand side, we easily obtain
1
∫ dx = ln ∣x∣ + C2 .
x
1
Therefore, an implicit solution to the ODE is − (ln ∣2v − 1∣ + 3 ln ∣2v + 1∣) = ln ∣x∣ + C3 . We
4

5
 can further transform the equation as follows.
1
− (ln ∣2v − 1∣ + 3 ln ∣2v + 1∣) = ln ∣x∣ + C3
4
1
⇔ ln ∣(2v − 1)(2v + 1)3 ∣ = ln 4 + C4
x
1
⇔ (2v − 1)(2v + 1)3 = ± 4 eC4
x
C
⇔ (2v − 1)(2v + 1)3 = 4
x
⇔ (2y − x)(2y + x)3 = C.

(b) Let y = vx. Then we have

y(x − y)
y′ =
x2
(vx)(x − vx)
⇔ v + v′ x =
x2
′
⇔ v + v x = v(1 − v)
⇔ v ′ x = −v 2
1 1
⇔ dv = − dx
v2 x
1 1
⇔ ∫ 2 dv = − ∫ dx
v x
1
⇔ − = − ln ∣x∣ + C
v
x
⇔ − = − ln ∣x∣ + C
y
x
⇔ y= .
ln ∣x∣ − C

(c) Let y = vx. Then we have

x3 y ′ + 2xy 2 = 3y 3
⇔ x3 (v + v ′ x) + 2x(vx)2 = 3(vx)3
⇔ v + v ′ x + 2v 2 = 3v 3
⇔ v ′ x = v(v − 1)(3v + 1)
1 1
⇔ dv = dx
v(v − 1)(3v + 1) x
1 1
⇔ ∫ dv = ∫ dx.
v(v − 1)(3v + 1) x

For the left-hand side, we have the partial fraction decomposition

1 A B C
= + + .
v(v − 1)(3v + 1) v v − 1 3v + 1

This means 1 = A(v − 1)(3v + 1) + Bv(3v + 1) + Cv(v − 1). Putting v = 0, we obtain A = −1.
1 1 9
Putting v = 1, we obtain B = . Putting v = − , we obtain C = .
4 3 4

6
 Therefore,

1 −1 1 9
∫ dv = ∫ ( + + ) dv
v(v − 1)(3v + 1) v 4(v − 1) 4(3v + 1)
1 3
= − ln ∣v∣ + ln ∣v − 1∣ + ln ∣3v + 1∣ + C1 .
4 4
For the right-hand side, we easily obtain
1
∫ dx = ln ∣x∣ + C2 .
x
1 3
Therefore, an implicit solution to the ODE is − ln ∣v∣ + ln ∣v − 1∣ + ln ∣3v + 1∣ = ln ∣x∣ + C3 .
4 4
We can further transform the equation as follows.
1 3
− ln ∣v∣ + ln ∣v − 1∣ + ln ∣3v + 1∣ = ln ∣x∣ + C3
4 4
(v − 1)(3v + 1)3
⇔ ln ∣ ∣ = ln (x4 ) + C4
v4
(v − 1)(3v + 1)3
⇔ = ±x4 eC4
v4
(v − 1)(3v + 1)3
⇔ = Cx4
v4
(y − x)(3y + x)3
⇔ = Cx4 .
y4
343
Putting x = 1 and y = 2, this implies = C. Therefore, the solution is given by
16
16(y − x)(3y + x)3 = 343x4 y 4 .

1
8. (a) The ODE can be rewritten as y ′ − y = 2 sin 3x. The integrating factor is
2
− 21 dx
µ(x) = e∫ = e− 2 .
x

Therefore, we have
1
y ′ − y = 2 sin 3x
2
(e− 2 y)′ = 2e− 2 sin 3x
x x
⇔
e− 2 y = ∫ 2e− 2 sin 3x dx.
x x
⇔

By a classwork question (subsection 5.2), we know that

− 2 1
e− 2 (−3 cos 3x −
x x
∫ 2e 2 sin 3x dx = 2
sin 3x) + C.
(− 12 ) + 32 2

It follows that
4 x
y= (−6 cos 3x − sin 3x) + Ce 2 .
37

7
 24 61
Putting x = 0 and y = 1, we obtain 1 = − + C, i.e. C = . Therefore, the solution is
37 37
4 61 x
y= (−6 cos 3x − sin 3x) + e 2 .
37 37

2 cos x 1
(b) The ODE can be rewritten as y ′ + y= . The integrating factor is
sin x sin x

µ(x) = e∫ = e∫
2 cos x 2
sin x dx sin x d(sin x)
= e2 ln sin x = sin2 x.

Therefore, we have
2 cos x 1
y′ + y=
sin x sin x
⇔ ((sin2 x)y)′ = sin x
⇔ (sin2 x)y = ∫ sin x dx
⇔ (sin2 x)y = − cos x + C
⇔ y = − cos x csc2 x + C csc2 x.

(c) The integrating factor is µ(x) = e∫ = eln (x+1) = x + 1. Therefore, we have

1
x+1 dx

y
y′ + = x2 + x
x+1
⇔ ((x + 1)y)′ = x3 + 2x2 + x
⇔ (x + 1)y = ∫ (x3 + 2x2 + x) dx
1 2 1
⇔ (x + 1)y = x4 + x3 + x2 + C
4 3 2
1 1 4 2 3 1 2
⇔ y= ( x + x + x + C) .
x+1 4 3 2

9. (a) The characteristic equation is t2 − 6t + 9 = 0. It has a repeated root t = 3. By proposition

6.3,
y = Ae3x + Bxe3x (1)
for some constants A and B. This implies

y ′ = 3Ae3x + Be3x + 3Bxe3x . (2)

Putting x = 0, y = 4 and y ′ = −3 in (1) and (2), we obtain 4 = A and −3 = 3A + B.

Therefore, A = 4 and B = −15. The solution is

y = 4e3x − 15xe3x .

(b) The characteristic equation is t2 − 2t + 5 = 0. This gives (t − 1)2 = −4, and hence the roots
are t = 1 ± 2i. By proposition 6.2,

y = Aex cos 2x + Bex sin 2x (1)

8
 for some constants A and B. This implies
y ′ = Aex cos 2x − 2Aex sin 2x + Bex sin 2x + 2Bex cos 2x. (2)
Putting x = 0, y = 1 and y ′ = −3 in (1) and (2), we obtain 1 = A and −3 = A + 2B.
Therefore, A = 1 and B = −2. The solution is
y = ex cos 2x − 2ex sin 2x.

(c) The ODE is the same as y ′′ − y ′ − 6y = 0. Thus, the characteristic equation is t2 − t − 6 = 0.

This gives (t + 2)(t − 3) = 0, and hence there are two distinct real roots t = −2, 3. By
proposition 6.1,
y = Ae−2x + Be3x
for some constants A and B.

10. (a) Let y = vx. Then

(x + y)y ′ = 4x − 3y
⇔ (x + vx)(v + v ′ x) = 4x − 3vx
⇔ (1 + v)(v + v ′ x) = 4 − 3v
⇔ (1 + v)v ′ x = −v 2 − 4v + 4
v+1 1
⇔ dv = − dx
v 2 + 4v − 4 x
v+1 1
⇔ ∫ 2 dv = − ∫ dx. (1)
v + 4v − 4 x
√ √
For the left-hand side, since v 2 + 4v − 4 = (v + 2 − 2 2)(v + 2 + 2 2), we can write
v+1 A B
= √ + √
v2 + 4v − 4 v + 2 − 2 2 v + 2 + 2 2
for some constants A and B. This implies
√ √
v + 1 = A(v + 2 + 2 2) + B(v + 2 − 2 2).
√ √
√ 4− 2 √ 4+ 2
Putting v = 2 2 − 2, we obtain A = . Putting v = −2 − 2 2, we obtain B = .
8 8
Thus,
√ √
v+1 4− 2 4+ 2
∫ 2 dv = ∫ ( √ + √ ) dv
v + 4v − 4 8(v + 2 − 2 2) 8(v + 2 + 2 2)
√ √
4− 2 √ 4+ 2 √
= ln ∣v + 2 − 2 2∣ + ln ∣v + 2 + 2 2∣ + C1 .
8 8
Therefore, (1) becomes
√ √
4− 2 √ 4+ 2 √
ln ∣v + 2 − 2 2∣ + ln ∣v + 2 + 2 2∣ = − ln ∣x∣ + C.
8 8
The final solution is
√ √
4− 2 y √ 4+ 2 y √
ln ∣ + 2 − 2 2∣ + ln ∣ + 2 + 2 2∣ = − ln ∣x∣ + C.
8 x 8 x

9
(b) We have

(cos y)y ′ = 3x2

⇔ cos y dy = 3x2 dx
⇔ ∫ cos y dy = ∫ 3x dx
2

⇔ sin y = x3 + C (1)
−1
⇔ y = nπ + (−1) sin
n
(x + C)
3

where n is an integer. The last equivalence follows from proposition 1.7. Putting x = 1
and y = π in (1), we obtain 0 = 1 + C, i.e. C = −1. Thus, the general solution becomes

y = nπ + (−1)n sin−1 (x3 − 1).

Putting x = 1 and y = π in this equation, we obtain π = nπ. This implies n = 1, and hence
the unique solution is y = π − sin−1 (x3 − 1).

1 sin x
(c) Let z = y ′ . The ODE becomes z ′ + z = cos x+ . This is a linear ODE. The integrating
x x
factor is e∫ x = e
1
dx ln x
= x. Thus, we have
1 sin x
z′ + z = cos x +
x x
⇔ (xz)′ = x cos x + sin x
⇔ xz = ∫ (x cos x + sin x) dx.

Using integration by parts, we have

∫ x cos x dx = ∫ x d(sin x) = x sin x − ∫ sin x dx.

Therefore, we have xz = x sin x + C1 , or

C1
z = sin x + .
x
As z = y ′ , this implies
C1
y = ∫ (sin x + ) dx = − cos x + C1 ln ∣x∣ + C2 .
x

10