1.

5 Existence and Uniqueness of Solutions 71

y Figure 1.50
5 Euler’s method applied to
dy
= et sin y
dt
with
t = 0.05. The graph of the
0 t approximation behaves as expected
1 2 3 4 5
for t < 5, but for t slightly larger
than 5, the approximation is no
longer valid.
−5

We can easily check that the constant function y(t) = nπ is a solution for any
integer n and, hence by the Uniqueness Theorem, each solution is trapped between
y = nπ and y = (n + 1)π for some integer n. The approximations in Figure 1.50
violate this requirement. This confirms our suspicions that the numerical results in this
case are not to be believed.
This equation is unusual because of the et term on the right-hand side. When t is
large, the slopes of solutions become gigantic and hence Euler’s method overshoots the
true solution for even a very small step size.

EXERCISES FOR SECTION 1.5

In Exercises 1–4, we refer to a function f , but we do not provide its formula. However,
we do assume that f satisfies the hypotheses of the Uniqueness Theorem in the entire
t y-plane, and we do provide various solutions to the given differential equation. Finally,
we specify an initial condition. Using the Uniqueness Theorem, what can you conclude
about the solution to the equation with the given initial condition?

dy dy
1. = f (t, y) 2. = f (y)
dt dt
y1 (t) = 3 for all t is a solution, y1 (t) = 4 for all t is a solution,
initial condition y(0) = 1 y2 (t) = 2 for all t is a solution,
y3 (t) = 0 for all t is a solution,
initial condition y(0) = 1

dy dy
3. = f (t, y) 4. = f (t, y)
dt dt
y1 (t) = t + 2 for all t is a solution, y1 (t) = −1 for all t is a solution,
y2 (t) = −t 2 for all t is a solution, y2 (t) = 1 + t 2 for all t is a solution,
initial condition y(0) = 1 initial condition y(0) = 0

Copyright 2011 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
72 CHAPTER 1 First-Order Differential Equations

In Exercises 5–8, an initial condition for the differential equation

dy
= y(y − 1)(y − 3)
dt
is given. What does the Existence and Uniqueness Theorem say about the correspond-
ing solution?

5. y(0) = 4 6. y(0) = 0 7. y(0) = 2 8. y(0) = −1

9. (a) Show that y1 (t) = t 2 and y2 (t) = t 2 + 1 are solutions to

dy
= −y 2 + y + 2yt 2 + 2t − t 2 − t 4 .
dt
(b) Show that if y(t) is a solution to the differential equation in part (a) and if
0 < y(0) < 1, then t 2 < y(t) < t 2 + 1 for all t.
(c) Illustrate your answer using HPGSolver.
√
10. Consider the differential equation dy/dt = 2 |y|.
(a) Show that the function y(t) = 0 for all t is an equilibrium solution.
(b) Find all solutions. [Hint: Consider the cases y > 0 and y < 0 separately. Then
you need to define the solutions using language like “y(t) = . . . when t ≤ 0
and y(t) = . . . when t > 0.”]
(c) Why doesn’t this differential equation contradict the Uniqueness Theorem?
(d) What does HPGSolver do with this equation?

11. Consider the differential equation

dy y
= 2.
dt t
(a) Show that the constant function y1 (t) = 0 is a solution.
(b) Show that there are infinitely many other functions that satisfy the differen-
tial equation, that agree with this solution when t ≤ 0, but that are nonzero
when t > 0. [Hint: You need to define these functions using language like
“y(t) = . . . when t ≤ 0 and y(t) = . . . when t > 0.”]
(c) Why doesn’t this example contradict the Uniqueness Theorem?

12. (a) Show that

1 1
y1 (t) = and y2 (t) =
t −1 t −2
are solutions of dy/dt = −y 2 .
(b) What can you say about solutions of dy/dt = −y 2 for which the initial con-
dition y(0) satisfies the inequality −1 < y(0) < −1/2? [Hint: You could find
the general solution, but what information can you get from your answer to
part (a) alone?]

Copyright 2011 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
 1.5 Existence and Uniqueness of Solutions 73

In Exercises 13–16, an initial-value problem is given.

(a) Find a formula for the solution.
(b) State the domain of definition of the solution.
(c) Describe what happens to the solution as it approachs the limits of its domain of
definition. Why can’t the solution be extended for more time?

dy dy 1
13. = y3, y(0) = 1 14. = , y(0) = 0
dt dt (y + 1)(t − 2)
dy 1 dy t
15. = , y(0) = 1 16. = , y(−1) = 0
dt (y + 2)2 dt y−2

17. Consider a differential equation of the form dy/dt = f (y), an autonomous equa-
tion, and assume that the function f (y) is continuously differentiable.
(a) Suppose y1 (t) is a solution and y1 (t) has a local maximum at t = t0 . Let
y0 = y1 (t0 ). Show that f (y0 ) = 0.
(b) Use the information of part (a) to sketch the slope field along the line y = y0
in the t y-plane.
(c) Show that the constant function y2 (t) = y0 is a solution (in other words, y2 (t)
is an equilibrium solution).
(d) Show that y1 (t) = y0 for all t.
(e) Show that if a solution of dy/dt = f (y) has a local minimum, then it is a
constant function; that is, it also corresponds to an equilibrium solution.

18. We have emphasized that the Uniqueness Theorem does not apply to every differ-
ential equation. There are hypotheses that must be verified before we can apply
the theorem. However, there is a temptation to think that, since models of “real-
world” problems must obviously have solutions, we don’t need to worry about the
hypotheses of the Uniqueness Theorem when we are working with differential equa-
tions modeling the physical world. The following model illustrates the flaw in this
assumption.
Suppose we wish to study the formation of raindrops in the atmosphere. We
make the reasonable assumption that raindrops are approximately spherical. We also
assume that the rate of growth of the volume of a raindrop is proportional to its sur-
face area.
Let r (t) be the radius of the raindrop at time t, s(t) be its surface area at time t,
and v(t) be its volume at time t. From three-dimensional geometry, we know that
s = 4πr 2 and v = 43 πr 3 .
(a) Show that the differential equation that models the volume of the raindrop un-
der these assumptions is
dv
= kv 2/3 ,
dt
where k is a proportionality constant.

Copyright 2011 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
74 CHAPTER 1 First-Order Differential Equations

(b) Why doesn’t this equation satisfy the hypotheses of the Uniqueness Theorem?
(c) Give a physical interpretation of the fact that solutions to this equation with the
initial condition v(0) = 0 are not unique. Does this model say anything about
the way raindrops begin to form?

1.6 EQUILIBRIA AND THE PHASE LINE

Given a differential equation

dy
= f (t, y),
dt
we can get an idea of how solutions behave by drawing slope fields and sketching their
graphs or by using Euler’s method and computing approximate solutions. Sometimes
we can even derive explicit formulas for solutions and plot the results. All of these
techniques require quite a bit of work, either numerical (computation of slopes or Eu-
ler’s method) or analytic (integration).
In this section we consider differential equations where the right-hand side is in-
dependent of t. Such equations are said to be autonomous differential equations. The
word autonomous means “self-governing,” and roughly speaking, an autonomous sys-
tem is self-governing because it evolves according to differential equations that are de-
termined entirely by the values of the dependent variables. For autonomous differential
equations, there are qualitative techniques that help us sketch the graphs of the solutions
with less arithmetic than with other methods.

Autonomous Equations
Autonomous equations are differential equations of the form
dy
= f (y).
dt
In other words, the rate of change of the dependent variable can be expressed as a func-
tion of the dependent variable alone. Autonomous equations appear frequently as mod-
els for two reasons. First, many physical systems work the same way at any time. For
example, a spring compressed the same amount at 10:00 AM and at 3:00 PM provides
the same force. Second, for many systems, the time dependence “averages out” over
the time scales being considered. For example, if we are studying how wolves and field
mice interact, we might find that wolves eat many more field mice during the day than
they do at night. However, if we are interested in how the wolf and mouse populations
behave over a period of years or decades, then we can average the number of mice eaten
by each wolf per week. We ignore the daily fluctuations.
We have already noticed that autonomous equations have slope fields that have a
special form (see page 40 in Section 1.3). Because the right-hand side of the equation
does not depend on t, the slope marks are parallel along horizontal lines in the t y-plane.
That is, for an autonomous equation, two points with the same y-coordinate but differ-
ent t-coordinates have the same slope marks (see Figure 1.51).

Copyright 2011 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
760 Hints and Answers for Chapter 1

Hints and Answers for Section 1.5

1. y(t) < 3 for all t in the domain of y(t) (c) The differential equation is not defined at t = 0.
√
3. −t 2 < y(t) < t + 2 for all t 13. (a) y(t) = 1/ 1 − 2t
5. y(t) > 3 for all t in the domain of y(t), y(t) → 3 (b) t < 1/2
as t decreases, and y(t) increases without bound as (c) y(t) → ∞ as t → 1/2−
t increases. 15. (a) y(t) = (3t + 27)1/3 − 2
7. 1 < y(t) < 3 for all t, y(t) → 1 as t → ∞, and (b) t > −9 (note that the differential equation is not
y(t) → 3 as t → −∞. defined at y = −2).
9. (a) Substitute each solution into the differential (c) y(t) → ∞ as t → ∞; y(t) → −2 as t → −9+ .
equation and compute. 17. (a) Hint: Differentiate y1 (t) at t0 and remember that
(b) Use the Existence and Uniqueness Theorem. y1 (t) is a solution.
11. (a) If y1 (t) = 0, then dy1 /dt = 0 = y1 /t 2 . (b) Hint: Remember that the equation is au-
tonomous.
(b) For any real number c, let
⎧ (c) Hint: Look at the slope field, but check by sub-
⎨ 0, for t ≤ 0; stituting y2 (t) into both sides of the differential
yc (t) = equation.
⎩ ce−1/t , for t > 0.
(d) Hint: Uniqueness Theorem
The function yc (t) satisfies the equation for all (e) Hint: Do the same four steps again.
t  = 0. It is 0 for t < 0 and nonzero for t > 0.

Hints and Answers for Section 1.6

1. The point y = 0 is a sink, and the point y = 2 is a 5. The sources are w = 0, −2π , −4π . . . and π , 3π,
source. 5π . . . . The sinks are w = 1, w = −π , −3π ,
. . . and w = 2π , 4π , 6π , . . . .

y=2 source
w=π source

y=0 sink
w=1 sink
w=0 source
3. The sinks are . . . −3π/2, π/2, 5π/2, . . . , and the
sources are . . . −π/2, 3π/2, 7π/2, . . . .
7. There are no equilibrium points.

y = 3π/2 source

y = π/2 sink

y = −π/2 source

Copyright 2011 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.