Math 432 HW 2.

2 Solutions

Assigned: 1, 3, 5, 6, 7, 11, 16, 19, 23, 26, 27, 28, 37(a&b), and 38.

NOTE: For #27 (d) you may use your calculator as a substitute for a numerical integration algorithm.

Selected for Grading: 11, 16, 19, 28

Solutions:
1. dy/dx = 4y
2
3y + 1 is separable: dy/(4y
2
3y + 1) = dx

3. dy/dx = ye
x+y
/(x
2
+ 2) is separable: dy/(ye
y
) = e
x
/(x
2
+ 2)

5. s
2
+ ds/dt = (s + 1)/st
Solving for ds/dt gives

.
This is not of the form g(t)p(s) so the equation is not separable.

6. (xy
2
+ 3y
2
)dy 2x dx = 0 is separable:
(x + 3)y
2
dy = 2x dx
y
2
dy = [2x/(x
2
+ 3)] dx

7. Given: dy/dx = y(2 + sin x). Note before starting that y(x) 0 is one solution.

ln | y | = 2x cos x + C
| y | = e
2x cos x + C
= Ae
2x cos x
for some A > 0.
y = Be
2x cos x
for some B 0. (And we can "grab" the constant solution above by allowing B = 0.)
Solution: y = Be
2x cos x
for some constant, B.

11. Given: dy/dx = sec
2
y/(1 + x
2
). I'm going to divide by sec
2
y, so I'm assuming that sec
2
y 0. This is always
true, so I can just "plough ahead". When I divide by sec
2
y and "multiply" by dx, I get something to
integrate:

(1/2) y + (1/4) sin 2y = arctan(x) + C

Solution: 2y + sin 2y = 4 arctan(x) + C.

16. Given:

{Dividing by 1 + y
2
presents no problem. Same for the exponential term.}

{Since 1 + y
2
> 0 we don't need the absolute values symbols.}

for some positive C.

for some positive C.

Now, the function

has a global maximum value of 1/2 at x = 0.

is e
1/2
, which in turn implies that the global minimum
for

is e
1/2
. So . . .

Solution:

for some C e
1/2
.

19. IVP:

, y() = 0.
Note: We don't have to fuss around with dividing by zero since (at least on some interval containing )
y(x) + 1 will be strictly greater than zero. {It would take some time for y(x) + 1 to move from y() + 1 = 1
to y(x) + 1 = 0.} Anyways, we can just jump right in.

y + 1 = (sin x + C)
2

y = (sin x + C)
2
1

The initial condition gives us that
0 = (sin + C)
2
1
(C)
2
= 1
C = 1.
And we have to decide which value to use for C.
Look back to the line before we squared both sides of the equation:
There we see that we need sin x + C 0.
If we were to use C = 1, then we'd need sin x 1 0 or, equivalently, sin x 1.
But this happens only for isolated x-values, x = k /2, for k = any odd integer.
Since our (guaranteed) solution is to be defined on an entire open interval containing , then we're forced to
use C = 1.

Finally, the solution is y = (sin x + 1)
2
1 = sin
2
x + 2 sin x + 1 1 = sin
2
x + 2 sin x.

23. Given: dy/dt = 2t cos
2
y, y(0) = /4.
Since y(0) = /4 and cos
2
(/4) = 1/2 0, we can go ahead and divide to separate the variables:

Using the initial condition:
tan(/4) = 0
2
/2 + C
C = 1
So

which gives

Solution:

.

26. , y(0) = 1. Here I'll be dividing by both x + 1 and . For the initial condition,
we'll have x + 1 0 and 0. So we're OK. Here we go. . . .

C = 2

{And since x
0
= 0, and hence x
0
+ 1 > 0, we can "drop" the absolute value.}

Solution:

.

27. (a) Given:

, y(0) = 0.
I'll use the initial condition's x
0
for my lower limit of integration.

Then use the initial condition to evaluate C: 0 = 0 + C, C = 0.

Solution:

(b) Given:

, y(0) = 1.
First separate the variables. Then integrate (using the tool presented in this exercise).

Use the initial condition to evaluate C: 1/3 = 0 + C
So we have

Solution:

(c) Given:

, y(0) = 1.
Separate:

Use the initial condition: arctan(1) = 0 + C, C = /4.
So we have

Solution:

(d) I used the "numerical integration algorithm" used by my calculator it's called fnInt on my TI-83 and
here's what I got.
For the differential equation given in (b) our solution is

.
So

.
Using my calculator, I entered 3*fnInt(e^(X^2), X, 0, 0.5), waited a bit, and got y(0.5) 0.544987102.

28. Before I sketch the solution, I'll have to find it.
dy/dt = 2y 2yt, y(0) = 3.
dy/dt = 2y(1 t)

ln |y| = (1 t)
2
+ C, and we can drop the absolute values since y(0) > 0.
ln y = (1 t)
2
+ C
Using the initial condition: ln 3 = (1)
2
+ C, C = 1 + ln 3.
ln y = 1 + ln 3 (1 t)
2


Solution:

.

Here is a sketch of this solution.

You can find the maximum value using the first-derivative test from calculus.
4 3 2 1 1 2 3 4 5
1
1
2
3
4
5
6
7
8
x
y

The sole critical point is t = 1.
For t < 1, y'(t) is positive [so y(t) is increasing].
For t > 1, y'(t) is negative [so y(t) is decreasing].
So y(t) must have a global maximum at t = 1, and that maximum is 3e.

37. Given: dP/dt = (r/100)P, r = 5, P(0) = 1000.
Before answering their questions, I'll solve this IVP. The DE is separable.

ln(P) = 0.05t + C

The initial condition gives C = ln(1000). So we have
ln(P) = 0.05t + ln(1000)
ln(P/1000) = 0.05t
P/1000 = e
0.05t


Solution: P = 1000e
0.05t
.

(a) In two years there will be P(2) = 1000e
0.1
= $1105.17 (when rounded appropriately) in the account.
(b) Set P(t) equal to 4000 and solve for t:
1000e
0.05t
= 4000
e
0.05t
= 4
0.05t = ln(4)
t = ln(4) 0.05 27.73
There will be $4000 in the account in approximately 27.73 years.
(c) This part was not assigned.

38. We start with 100(dv/dt) = 980 5 and v(0) = 10 m/sec. This is separable too. Here are the details.

20 ln(980 5v) = t + C
C = 20 ln(980 50) = 20 ln(930)
20 ln(980 5v) = t 20 ln(930)
980 5v = 930e
t/20


Solution: v = 196 186e
t/20
.

The limiting velocity is 196 m/sec.