Math 22B, Spring 2024

Midterm 1 Practice Problems: Solutions

1. Suppose the population y(t) of a species at time t satisfies the ODE

′
 y  y
y = −ry 1 − 1−
T K
where r > 0 and 0 < T < K are positive constants.
(a) What are the equilibrium solutions of the ODE?
(b) Sketch the phase line of the ODE for y ≥ 0. Determine the stability of
the equilibria in (a).
(c) Suppose that y(t) satisfies the initial condition y(0) = y0 . For what values
of y0 > 0 does the population approach extinction and for what values of y0
does the species approach its maximum sustainable population as t → ∞.

Solution

ˆ (a) The equilibria are y = 0, y = T , y = K.

ˆ (b) By looking at the signs of the factors in

 y  y
f (y) = −ry 1 − 1− ,
T K
we see that y ′ > 0 if T < y < K and y ′ < 0 if 0 < y < T or y > K.
The phase line therefore looks like this:

ˆ The equilibria y = 0 and y = K are asymptotically stable, and the

equilibrium y = T is unstable.

ˆ (c) If 0 ≤ y0 < T , then y(t) → 0 as t → ∞, so the population becomes

extinct. If y0 > T , then y(t) → K as t → ∞, where K is the maximum
sustainable population.

1
2. Suppose that for certain continuous functions p(t), g(t) the functions
y1 (t), y2 (t) are solutions of the ODEs

y1′ + p(t)y1 = 0, y2′ + p(t)y2 = g(t).

(a) Give an expression for the general solution y(t) of the ODE

y ′ + p(t)y = 3g(t)

in terms of y1 (t) and y2 (t), and justify your answer.

(b) If y1 (0) = 2 and y2 (0) = 1, express the solution y(t) of the initial value
problem
y ′ + p(t)y = 3g(t), y(0) = 1
in terms of y1 (t) and y2 (t).

Solution

ˆ (a) Provided that y1 ̸= 0, the general solution is

y(t) = Cy1 (t) + 3y2 (t),

where C is an arbitrary constant.

ˆ To show that y(t) is a solution of the ODE, we compute

y ′ + py = (Cy1 + 3y2 )′ + p (Cy1 + 3y2 )

= Cy1′ + 3y2′ + Cpy1 + 3py2
= C (y1′ + py1 ) + 3 (y2′ + py2 )
= C ·0+3·g
= 3g.

ˆ (b) Setting t = 0 in the general solution, we get that

y(0) = Cy1 (0) + 3y2 (0).

Use of the initial conditions implies that 1 = 2C + 3, so C = −1 and

y(t) = −y1 (t) + 3y2 (t).

2
3. (a) Find the solution of the initial value problem

y ′ = y ln y, y(0) = e.

Hint. Use the substitution u = ln y.

(b) For what values of t is the solution defined?

Solution

ˆ Separating variables we get

Z Z
dy
= dt.
y ln y

The substitution u = ln y gives du = dy/y and

Z Z
dy du
= = ln u + C = ln(ln y) + C.
y ln y u
Hence,
ln(ln y) = t + C.

ˆ At t = 0, we have y = e, so

C = ln(ln e) = ln 1 = 0.

ˆ Solving for y, we get

t
y(t) = ee .

ˆ (b) The solution is defined for all −∞ < t < ∞.

Remark. This ODE is interesting because, although the right-hand side

grows faster than a linear function of y as y → ∞, the solution exists for all
t. It grows very quickly (doubly-exponentially) as t → ∞. This ODE has
close to the maximal growth in y that allows solutions to be defined for all
t. For example, a similar calculation shows that any solution of

y ′ = y(ln y)n

with initial data y(0) = y0 > 1 goes to infinity in finite time for any n > 1.

3
4. (a) Solve the initial value problem

y ′ + y = t, y(0) = 1.

(b) Find the time t at which the solution attains its minimum value.

Solution

ˆ (a) This is a linear, nonhomogeneous first-order ODE. Omitting an

unneeded constant of integration, we get the integrating factor
R
1 dt
µ(t) = e = et .

Multiplying the equation by et and rearranging the left-hand side, we

get
′
et y = tet .
Integrating this equation, and using an integration by parts, we get
Z Z
t
ey = te dt + C = te − et dt + C = tet − et + C.
t t

ˆ Imposing the initial condition, we get 1 = −1 + C, so C = 2, and

solving for y(t), we obtain that

y(t) = t − 1 + 2e−t .

ˆ (b) Differentiating the solution, we get

y ′ (t) = 1 − 2e−t , y ′′ (t) = 2e−t .

If follows that y ′ (t) = 0 if et = 2 or

t = ln 2.

This critical point corresponds to a local minimum of y(t) since y ′′ (ln 2) >
0, and it is a global minimum since y ′ (t) < 0 for t < ln 2 and y ′ (t) > 0
for t > ln 2.

4
5. (a) Solve the initial value problem

y ′ = e−y cos t, y(0) = y0 .

(b) For what initial values y0 is the solution y(t) defined for all −∞ < t < ∞?

Solution

ˆ (a) The equation is nonlinear and separable. Separating variables we

get Z Z
y
e dy = cos t dt.

Evaluation of the integrals gives

ey = sin t + C.

Imposing the initial condition (this can be done now or after solving
for y), we get
C = ey0 .
The solution for y is therefore

y(t) = ln (sin t + ey0 ) .

ˆ (b) Since ln x is differentiable if x > 0 and sin t ≥ −1, the solution is

defined for all −∞ < t < ∞ if and only if ey0 > 1, meaning that

y0 > 0.

5
6. (a) Find all equilibria of the ODE

y2 − 1
y′ = .
y2 + 1

(b) Sketch the phase line, and determine the stability of the equilibria you
found in (a).
(c) Sketch the graph versus t of the solution y(t) that satisfies the initial
condition y(0) = 0.

Solution

ˆ (a) The equilibria satisfy f (y) = 0, where

y2 − 1
f (y) = 2 ,
y +1

which implies that y 2 = 1. Therefore the equilibria are y = −1, y = 1.

ˆ (b) The denominator y 2 + 1 is always positive, so the function f (y) is

positive for y > 1 or y < −1, and negative for −1 < y < 1. The phase
line therefore looks like this:

The equilibrium y = −1 is asymptotically stable, and the equilibrium

y = 1 is unstable.

ˆ (c) The graph of y(t) has negative slope everywhere. It asymptotes to

the line y = −1 as t → +∞ and the line y = 1 as t → −∞. Also, y(t)
is an odd function of t, since z(t) = −y(−t) satisfies the same IVP as
y(t), so z(t) = y(t).