Homework 1

§1.1 Basic Concepts. Modeling

1. Verify that y is a solution of the ODE and determine from y the particular
solution of the IVP.
(a) y ′ + 4y = 1.4, y = ce−4x + 0.35, y(0) = 2
(b) y ′ = y − y 2 , y = 1+ce 1
−x , y(0) = 0.25
′
(c) yy = 4x, y − 4x = c (y > 0), y(1) = 4
2 2

Sol. (a) Since y ′ = −4ce−4x = −4(y − 0.35) = −4y + 1.4, y = (x + c)ex is a

solution. From y(0) = 2, we have c = 1.65 and hence the particular solution is
y = 1.65e−4x + 0.35.
ce−x ce−x ce−x
(b) Since y ′ = (1+ce −x )2 = 1+ce−x − (1+ce−x )2 = y − y , y = 1+ce−x is a solution. From
2 1

1
y(0) = 0.25, we have c = 3 and hence the particular solution is y = 1+3e −x .
′
(c) By differentiating both sides, 2yy −8x = 0 and it infers that y −4x = c (y > 0)
2 2

is a solution of yy ′ = 4x. From y(1) = 4, we have c = 12 and the particular solution

is y 2 − 4x2 = 12 (y > 0).

2. The half-life of a radioactive substance is the time in which half of the given
amount disappears. What is the half-life of radium 88 Ra226 (in years)?
(Hint : y ′ = ky, k = −1.4 · 10−11 sec−1 for radium 88 Ra226 .)
Sol. Let y(t) be the amount of substance at any time t. By the physical law, the
time rate of change y ′ (t) is proportional to y(t), that is,
y ′ = ky.
Thus y(t) = y0 ekt . The half-life of a radioactive substance is
1 ln 2
y0 ekt = y0 =⇒ t = − .
2 k
The half-life of radium 88 R
226
is ln 2
1.4×10−11 sec ≈ 1570year.

3. If we drop a stone, we assume air resistance to be negligible. Experiments show

2
that under that assumption the acceleration y ′′ = ddt2y of this motion is constant(g =
9.80m/sec2 ). State this as an ODE for y(t), the distance fallen as a function of time
2
t. Solve the ODE to get the familiar law of free fall, y = gt2 .
Sol.
1
y ′′ = g ⇒ v = y ′ = gt + v0 ⇒ y = gt2 + v0 t + y0 .
2
′
In particular, if the initial velocity is zero, y (0) = v0 = 0 and the initial position is
zero, y(0) = y0 = 0, then we have
1
y(t) = gt2 .
2

Computational Science & Engineering (CSE) T. Jeong

 Homework 2

§1.2 Geometric Meaning of y ′ = f (x, y). Directional Fields

1. Graph a direction field by hand. In the field graph approximate a solution curve
through the given point by hand.
(a) y ′ = 1 − y 2 , (0, 0), (2, 12 ) (b) yy ′ + 4x = 0, (1, 1), (0, 2)

§1.3 Separable ODEs. Modeling

1. Find a general solution.
2 2
(a) y 3 y ′ + x3 = 0 (b) y ′ = (y + 4x)2 (c) y ′ = 4xxy+y
(d) y ′ + (x + 2)y 2 = 0 (e) y ′ = x+2y+5
x+2y+7 (f) xy ′ = 21 y 2 + y
Sol.(a) By separation and integration,
y4 x4
y dy = −x dx,
3 3
= + c∗ , y 4 = x4 + c
4 4
(b) Taking u = y + 4x, we have u′ = y ′ + 4. The equation is reduced and separated
as follows
du 1 −1 u
u′ − 4 = u2 , 2
= dx, tan = x + c∗ .
4+u 2 2
Hence the solution is

u = 2 tan(2x + c), y = 2 tan(2x + c) − 4x.

(c) The ODE is written

4x y
y′ = + .
y x
Set u = xy , i.e., y = ux. Then u′ x + u = 4
u + u and
4 u2
udu = dx, = 4 ln |x| + c∗ .
x 2
The general solution is y 2 = 8x2 ln |x| + cx2 .

Computational Science & Engineering (CSE) T. Jeong

 Homework 3

(d)
dy 1 (x + 2)2
− = (x + 2)dx, = + c∗
y2 y 2
2
The general solution is y = x2 +4x+c .

v ′ −1
(e) Set v = x + 2y. Note that y ′ = 2 . Then

v′ − 1 v + 5 v+7 1 4/3
= , dv = dx, [ + ]dv = dx.
2 v+7 3v + 17 3 3v + 17
By integration, we have
1 4
v + ln |3v + 17| = x + c∗ .
3 9
Substituting v = x + 2y, we obtain the general solution

3x − 3y − 2 ln |3x + 6y + 17| = c.

(f) xy ′ = y(y+2)
2 ⇒ y(y+2)
2
dy = dxx ⇒ ( y − y+2 )dy =
1 1 dx
x,
ln |y| − ln |y + 2| = ln |x| + c∗ ⇒ y+2
y
= cx ∴ y= 2cx
1−cx .

2. Find the particular solution.

(a) xy ′ + y = 0, y(4) = 6 (b) y ′ = 1 + 4y 2 , y(0) = 6
(c) xy ′ = y + 4x5 cos2 (y/x), y(2) = 0 (d) yy ′ = (x − 1)e−y , y(0) = 1
2

Sol. (a) By separation,

y′ 1 c
=− , ln |y| = − ln |x| + c∗ , y = .
y x x
24
From initial condition y(4) = 6, we have c = 24 and y = x.
(b) By separation,
dy 1 1
= dx, tan−1 (2y) = x + c∗ , y= tan(2x + c)
1 + 4y 2 2 2
From initial condition y(0) = 6, we have c = −2 and the particular solution y =
2 tan(2x − 2).
1

(c) Reducing the ODE to y ′ = xy + 4x4 cos2 ( xy ) and setting u = y/x, i.e., y = ux,

u′ x + u = u + 4x4 cos2 u, sec2 udu = 4x3 dx.

y
tan u = x4 + c tan = x4 + c.
x
Computational Science & Engineering (CSE) T. Jeong
 Homework 4

From the condition y(2) = 0, we have c = −16 and y = x tan−1 (x4 − 16).
(d) Reducing the ODE,

y2 1 y 2 x2
ye dy = (x − 1)dx ⇒ e = − x + c.
2 2
2
From the condition y(0) = 1, we have c = e/2 and ey = x2 − 2x + e.

3. A tank contains 400 gal of brine in which 100 lb of salt is dissolved.Fresh water
runs into the tank at a rate of 2 gal/min. The mixture, kept practically uniform by
stirring, runs out at the same rate. How much salt will there be in the tank at the
end of 1 hour?
Sol. Let y(t) be the amount of the salt in the tank after t minutes. Then
y y
y ′ = inflow/min − outflow/min = 0 − 2 × =− , y(0) = 100.
400 200
The general solution is
y = ce− 200 t .
1

By the initial condition, c = 100, and the solution is

y = 100e− 200 t .
1

The salt left in the tank after 1 hours(60 minutes) is

100e− 200 ×60 = 100e− 10 ≈ 74.08.

1 3

Computational Science & Engineering (CSE) T. Jeong

 Homework 1

§1.4 Exact ODEs. Integrating Factors

1. Test for exactness. If exact, solve. If not, find an integrating factor and solve.
(a) ex (cos ydx − sin ydy) = 0 (b) (x2 + y 2 )dx − 2xydy = 0
(c) 2x tan ydx + sec2 ydy = 0 (d) (− xy2 + 2 cos 2x)dx + ( x1 − 2 sin 2y)dy = 0
2
(e) (ey − yex )dx + (xey − ex )dy = 0 (f) (2xydx + dy)ex = 0, y(0) = 2
Sol. (a) Set P = ex cos y and Q = −ex sin y. Since Py = −ex sin y = Qx the ODE is
exact.
∫
u = ex cos ydx = ex cos y + g(y)
∂u
= −ex sin y + g ′ (y) = −ex sin y
∂y
Hence we have g(y) = c and the general solution

ex cos y + c = 0.

(b) Set P = x2 + y 2 and Q = −2xy. Since Py = 2y ̸= Qx = −2y the ODE is not

exact. Note that R = Q1 (Py − Qx ) = − x2 .
The integrating factor is
∫
F (x) = e Rdx
= e−2 ln x = x−2 .

Multiplying F (x) = x−2 , we get the exact ODE

x2 + y 2 2y
2
dx − dy = 0.
x x
∫
x2 + y 2 y2
u= dx = x − + g(y)
x2 x
∂u 2y 2y
= − + g ′ (y) = −
∂y x x
Hence we have g(y) = c and the general solution

x2 − y 2 = cx.

(c) Set P = 2x tan y and Q = sec2 y. Since Py = 2x sec2 y ̸= Qx = 0 the ODE is not
2
exact. Note that R = Q1 (Py − Qx ) = 2xsecsec2 y y = 2x.
The integrating factor is ∫ 2
F (x) = e Rdx = ex .
2
Multiplying F (x) = ex , we get the exact ODE
2 2
2xex tan ydx + ex sec2 ydy = 0.

Computational Science & Engineering (CSE) T. Jeong

 Homework 2
∫
2 2
u= 2xex tan ydx = ex tan y + g(y)
∂u
= ex sec2 y + g ′ (y) = ex sec2 y
2 2

∂y
Hence we have g(y) = c and the general solution
2
ex tan y = c.

(d) Set M = − xy2 + 2 cos 2x and N = 1

x − 2 sin 2y. Since
1
My = − = Nx ,
x2
the ODE is exact.
Hence ∃u(x, y) such that ux = M, uy = N . Then
∫
y
u = M dx = + sin 2x + k(y)
x
and
1
uy = + k ′ (y) = N, k ′ (y) = −2 sin 2y ⇒ k(y) = cos 2y + c∗ .
x
Thus u(x, y) = x + sin 2x + cos 2y + c∗ and the general solution is
y

y
+ sin 2x + cos 2y = c.
x
(e) Set M = ey − yex and N = xey − ex . My = ey − ex = Nx . The ODE is exact.
Hence ∃u(x,
∫ y) such that ux = M, uy = N .
u = M dx = xey − yex + k(y), uy = xey − ex + k ′ (y) = N , k ′ (y) =
0 ⇒ k(y) = c∗ , u(x, y) = xey − yex + c∗ , ∴ xey − yex = c.
2 2 2
(f) Set P = 2xyex , Q = ex . Py = 2xex = Qx . The ODE is exact.

∫
2 2
u= 2xyex dx = yex + g(y)
∂u
= ex + g ′ (y) = ex
2 2

∂y
Hence we have g(y) = c and the general solution
2
yex + c = 0.

Computational Science & Engineering (CSE) T. Jeong

 Homework 3

2. Solve the the ODE 4ydx − 3xdy = 0 by following methods.

(a) By finding an integrating factor of the type F (x, y) = xa y b
(b) By finding an integrating factor of the type different from (a)
Sol. (a) Note that P = 4y, Q = −3x. Since

Py = 4 ̸= Qx = −3,

the ODE is not exact.

To find an integrating factor F (x, y) = xa y b for the ODE, consider (assuming) the
exact ODE
4xa y b+1 dx − 3xa+1 y b dy = 0.
Since the ODE is exact,

4(b + 1)xa y b = −3(a + 1)xa y b

holds. Thus 4(b + 1) = −3(a + 1), we choose a = 3, b = −4. An integrating factor

for the ODE is
x3
F (x, y) = 4 .
y
x3
Multiplying y4 , we have

x4
4x3 y −3 dx − 3x4 y −4 dy = du = 0, where u = .
y3
The general solution is x4 = cy 3 .

(b) Note that P = 4y, Q = −3x and

Py − Qx 7
F (x) = x− 3 .
7
R= = ,
Q −3x
Multiplying F (x), we have

4x− 3 ydx − 3x− 3 dy = du = 0, where u = −3x− 3 y.

7 4 4

The general solution is −3x− 3 y = c.

4

Computational Science & Engineering (CSE) T. Jeong

 Homework 4

§1.5 Linear ODEs. Bernoulli Equations. Population Dynamics

1. Find the general solution. If an initial condition is given, find the particular
solution.
(a) y ′ + 2y = 4 cos 2x (b) y ′ + y sin x = ecos x y(0) = −2.5
(c) xy ′ + 4y = 8x4 y(1) = 2 (d) x3 y ′ + 3x2 y = 5 sinh 10x √
(e) y ′ + y = y 2 , y(0) = −1 (f) 2yy ′ + y 2 sin x = sin x, y(0) = 2
Sol. (a) The general solution of the linear ODE is
∫
−2x
[ ] [ ]
y=e c + e2x 4 cos 2xdx = e−2x e2x cos 2x + e2x sin 2x + c .

(b) The general solution of the linear ODE is

∫ ∫
∫
− sin xdx
[ ]
y=e c + e sin xdx ecos x dx = (x + c)ecos x .

(c) For the given linear ODE,

4
y ′ + y = 8x3
x
the general solution is
∫
−
∫ [ ∫ ]
8x3 dx = x4 + cx−4 .
4 4
y=e x dx c+ e x dx

From the initial condition y(1) = 2, we have c = −1 and the particular solution
y = x4 − x−4 .
(d) For the given linear ODE,
3 5
y ′ + y = 3 sinh 10x
x x
the general solution is
∫
−
∫ 3 [ ∫ 3 5 ] 1[ 1 ]
y=e x dx c+ e x dx sinh 10x dx = c + cosh 10x .
x3 x3 2
(e) Set u = y 1−2 = y −1 . Then the ODE reduces to the linear ODE

u′ = −y −2 y ′ = −y −2 (y 2 − y) = −1 + y −1 = u − 1, u′ − u = −1.

The solution of u is
∫
∫
− −1dx
[ ∫
−1dx
] [ ]
u=e c+ e (−1) dx = ex c + e−x = cex + 1.

Hence
1
y= .
1 + cex
Computational Science & Engineering (CSE) T. Jeong
 Homework 5

From the initial condition y(0) = −1, c = −2. The solution is

1
y= .
1 − 2ex
(f) Set u = y 2 . The ODE reduces to the linear ODE

u′ + sin x u = sin x.

The general solution is

∫
∫ [ ∫ ]
u = y 2 = e− sin xdx
c+ e sin xdx
sin x dx = cecos x + 1.
√
From the initial condition y(0) = 2, c = e−1 . The solution is y 2 = ecos x−1 + 1.

2. If in a population y(t) the death rate is proportional to the population, and the
birth rate is proportional to the chance encounters of meeting mates for reproduc-
tion, what will the model be? Solve the model.
Sol.
y ′ = birth rate − death rate = k1 (ay)(b(y) − k2 y = −Ay + By 2
for some constants k1 , k2 , a, b.
Set u = y 1−2 = y −1 .

u′ = −y −2 y ′ = −y −2 (−Ay + By 2 ) = Ay −1 − B = Au − B, u′ − Au = −B.
∫
Then u = eAt ( e−At (−B)dt + c) = ceAt + B A and the general solution is

1
y= B
.
ceAt + A

3. A model for the speed of contagious diseases is obtained by assuming that the rate
of spread is proportional to the number of contacts between infected and noninfected
person, who are assumed to move freely among each other. Set up the model. Find
the equilibrium solutions and indicate their stability or instability. Solve the ODE.
Find the limit of the proportion of infected persons as t → ∞ and explain what it
means.
Sol. Let y(t) be a rate of infected person at time t.
Modeling:
y ′ = ky(1 − y), k > 0 or y ′ = ky − ky 2
Set u = y 1−2 = y −1 . The ODE reduces to

u′ = −y −2 y ′ = −y −2 (ky − ky 2 ) = −ky −1 + k = −ku + k, u′ + ku = k.

Computational Science & Engineering (CSE) T. Jeong

 Homework 6
∫
Then u = e−kt (c + kekt dt) = ce−kt + 1 and the general solution is
1
y = −kt .
ce + 1
The equilibrium solutions are y = 0, y = 1. (⇐ y ′ = ky(1 − y) = 0)
Notice that y → 1 as t → ∞. Hence y = 0 is unstable and y = 1 stable. If the
prevention of contagious diseases is not made, everyone will be infected eventually.

§1.6 Orthogonal Trajectories

1. Sketch or graph some of the given curves. Guess what their orthogonal trajectories
may look like. Find these trajectories.
(a)y = ce−x
2
(b) x2 + (y − c)2 = c2 (c) x2 − y 2 = c2
Sol. (a) The slope of the curves y = ce−x is y ′ = −2cxe−x = −2xy. The orthogonal
2 2

trajectories are obtained from

1 1
ỹ ′ = , ỹdỹ = dx,
2xỹ 2x
and so the orthogonal trajectories are ỹ 2 = ln |x| + c∗ .

(b) The slope of the curves x2 + y 2 − 2cy = 0 is

′ ′ ′ x 2xy x2 + y 2
2x + 2yy − 2cy = 0, y = = 2 (c = ).
c−y x − y2 2y
The orthogonal trajectories are obtained from
′ x2 − ỹ 2 x ỹ
ỹ = − =− + .
2xỹ 2ỹ 2x
Set u = xỹ , i.e., ỹ = ux. Then u′ x + u = − 2u1
+ u2 , 1+u
2u
2 du = − x . By integration,
dx

we have
c∗
ln(1 + u ) = − ln |x| + c̃, 1 + u = ,
2 2
x
Computational Science & Engineering (CSE) T. Jeong
 Homework 7

and the orthogonal trajectories are

x2 + ỹ 2 = c∗ x.

(c) The slope of the curves x2 − y 2 = c2 is

x
2x − 2yy ′ = 0, y′ = .
y
The orthogonal trajectories are obtained from
ỹ dỹ dx
ỹ ′ = − , =− .
x ỹ x
By integration, we have
ln |ỹ| = − ln |x| + c̃,
c∗
and the orthogonal trajectories are ỹ = x.

Computational Science & Engineering (CSE) T. Jeong

 Homework 8

2. Find the conditions under which the orthogonal trajectories of families of ellipses
x2 /a2 + y 2 /b2 = c are again conic sections. Illustrate your result graphically. What
happens if a → 0? If b → 0?
Sol. The slope of the curves x2 /a2 + y 2 /b2 = c is

2x 2y ′ b2 x
+ 2 = 0, y =− 2 .
a2 b ay
The orthogonal trajectories are obtained from

′ a2 ỹ dỹ dx
ỹ = 2 , 2
=− 2 .
bx a ỹ bx

By integration, we have b2 ln |ỹ| = −a2 ln |x| + c̃, and the orthogonal trajectories are
2
/b2
ỹ = cxa .

If a2 = b2 then the orthogonal trajectories are ỹ = cx which are strait lines. If

a2 = 2b2 then the orthogonal trajectories are ỹ = cx2 which are parabolas. If a → 0
then the given curves are strait lines(x = 0) and the orthogonal trajectories are the
strait lines (ỹ = c). If b → 0 then the given curves are strait lines(y = 0) and the
orthogonal trajectories are the strait lines (x = c).

3. The lines of electric force of two opposite charges of the same strength at (−1, 0)
and (1, 0) are the circles through (−1, 0) and (1, 0). Show that these circles are given
by x2 + (y − c)2 = 1 + c2 . Show that the equipotential lines(orthogonal trajectories
of those circles) are the circles given by (x + c∗ )2 + ỹ 2 = c∗ 2 − 1.
Sol. The circle is expressed as (x−A)2 +(y −B)2 = C 2 . Since (−1, 0) and (1, 0) lie on
the circle, (−1−A)2 +B 2 = C 2 , (1−A)2 +B 2 = C 2 holds. Thus A = 0, B 2 +1 = C 2 .
Set B = c, then these circles are given by

x2 + (y − c)2 = 1 + c2 , x2 + y 2 − 2cy = 1.

By differentiation, we have

′ ′ x 2xy
′ x2 + y 2 − 1
2x + 2yy − 2cy = 0, y = = 2 (c = ).
c−y x − y2 − 1 2y
The orthogonal trajectories is obtained from

′ x2 − ỹ 2 − 1
ỹ = − , (x2 − ỹ 2 − 1)dx + 2xỹdỹ = 0.
2xỹ
Computational Science & Engineering (CSE) T. Jeong
 Homework 9

Let P = x2 − ỹ 2 − 1, Q = 2xỹ. Since Pỹ = −2ỹ ̸= Qx = 2ỹ the ODE is not exact.
The integrating factor F (x) is
1 2 ∫ 1
R(x) = [Pỹ − Qx ] = − , F (x) = e R(x)dx
= .
Q x x2
1
Multiplying x2 in the both side,

ỹ 2 1 2ỹ
(1 − − )dx + dỹ = 0,
x2 x2 x
and the solution is
ỹ 2 1
x+ + = −2c∗ , x2 + ỹ 2 + 1 = −2c∗ x.
x x
Therefore the equipotential lines are the circles given by

(x + c∗ )2 + ỹ 2 = c∗ 2 − 1.

Computational Science & Engineering (CSE) T. Jeong

 Homework 1

§1.7 Existence and Uniqueness of Solutions

1. (a) Does the initial value problem (x − 2)y ′ = 2y, y(2) = 1 have a solution? Does
your result contradict Existence Theorem and Uniqueness Theorem? What happens
if you replace y(2) = 1 with y(2) = k?
Sol. The ODE is written
dy dx
= .
2y x−2
By integration, ln |2y| = ln |x − 2| + c∗ , y = c(x − 2). But this does not satisfy the
initial condition y(2) = 1. Thus the IVP has no solution.
Notice that f (x, y) of the ODE
2y
y ′ = f (x, y) =
x−2
is not continuous on any neighborhood of (2, 1). Therefore this result does not con-
tradict Existence and Uniqueness Theorems.
If k ̸= 0, then the ODE has no solution. If k = 0, then the ODE has an (infinitely
many) solution.

(b) Find all initial conditions such that

(x2 − 4x)y ′ = (2x − 4)y

has no solution, precisely one solution, and more than one solution.
Sol. The ODE is written
dy 2x − 4 1 1
= 2 dx = ( + )dx.
y x − 4x x x−4
By integration, ln |y| = ln |x| + ln |x − 4| + c∗ , y = cx(x − 4).
Thus the IVP

(x2 − 4x)y ′ = (2x − 4)y, y(0) ̸= 0 or y(4) ̸= 0

has no solution, and the IVP

(x2 − 4x)y ′ = (2x − 4)y, y(0) = 0 or y(4) = 0

has more than one solution(infinitely many solutions). Otherwise the IVP

(x2 − 4x)y ′ = (2x − 4)y, y(a) = ya

has precisely one solution.

Computational Science & Engineering (CSE) T. Jeong

 Homework 2

2. Consider an initial value problem

y ′ = f (x, y), y(x0 ) = y0 .

By integrating the ODE, we obtain

∫ x
y(x) = y0 + f (t, y(t))dt.
x0
∫x
Set yn (x) = y0 + x0 f (t, yn−1 (t))dt, n = 1, 2, · · · . If yn (x) converges, then the limit
is a solution of the ODE. It is called Picard’s iteration method.
Apply the iteration to y ′ = x + y, y(0) = 0, and also solve the problem exactly.

Sol. Set y0 = 0. ∫ ∫x
x 2
y1 = y0 + x0 f (t, y0 )dt = 0 + 0 tdt = x2 .
∫x ∫x 2 2 3
y2 = y0 + x0 f (t, y1 )dt = 0 + 0 (t + t2 )dt = x2 + x6 .
∫x ∫x 2 3 2 3 4
y3 = y0 + x0 f (t, y2 )dt = 0 + 0 (t + t2 + t6 )dt = x2 + x6 + x24
······
Note: y ′ = x + y, y = −x − 1 + cex , y(0) = 0, c = 1 ∴ y = −x − 1 + ex .

Computational Science & Engineering (CSE) T. Jeong