2.

2 Separable Equations 45

If the differential equation is

dy F(x, y)
= , (27)
dx G(x, y)
then, by comparing numerators and denominators in Eqs. (26) and (27), we obtain the
system
dx/dt = G(x, y), dy/dt = F(x, y). (28)
At first sight it may seem unlikely that a problem will be simplified by replacing a single
equation by a pair of equations, but, in fact, the system (28) may well be more amenable
to investigation than the single equation (27). Chapter 9 is devoted to nonlinear systems
of the form (28).
Note: In Example 2 it was not difficult to solve explicitly for y as a function of x
and to determine the exact interval in which the solution exists. However, this situation
is exceptional, and often it will be better to leave the solution in implicit form, as in
Examples 1 and 3. Thus, in the problems below and in other sections where nonlinear
equations appear, the terminology “solve the following differential equation” means to
find the solution explicitly if it is convenient to do so, but otherwise to find an implicit
formula for the solution.

PROBLEMS In each of Problems 1 through 8 solve the given differential equation.

1. y  = x 2 /y 2. y  = x 2 /y(1 + x 3 )
 2
3. y + y sin x = 0 4. y  = (3x 2 − 1)/(3 + 2y)
 2 2
5. y = (cos x)(cos 2y) 6. x y  = (1 − y 2 )1/2
−x
dy x −e dy x2
7. = y 8. =
dx y+e dx 1 + y2
In each of Problems 9 through 20:
(a) Find the solution of the given initial value problem in explicit form.
(b) Plot the graph of the solution.
(c) Determine (at least approximately) the interval in which the solution is defined.
� 9. y  = (1 − 2x)y 2 , y(0) = −1/6 � 10. y  = (1 − 2x)/y, y(1) = −2
−x 2
� 11. x d x + ye dy = 0, y(0) = 1 � 12. dr/dθ = r /θ, r (1) = 2
� 13. y  = 2x/(y + x 2 y), y(0) = −2 � 14. y  = x y 3 (1 + x 2 )−1/2 , y(0) = 1
  2 3
√
� 15. y = 2x/(1 + 2y), y(2) = 0 � 16. y = x(x + 1)/4y , y(0) = −1/ 2
� 17. y  = (3x 2 − e x )/(2y − 5), y(0) = 1
 −x x
� 18. y = (e − e )/(3 + 4y), y(0) = 1
� 19. sin 2x d x + cos 3y dy = 0, y(π/2) = π/3
2 2 1/2
� 20. y (1 − x ) dy = arcsin x d x, y(0) = 0
Some of the results requested in Problems 21 through 28 can be obtained either by solving
the given equations analytically, or by plotting numerically generated approximations to the
solutions. Try to form an opinion as to the advantages and disadvantages of each approach.
� 21. Solve the initial value problem
y  = (1 + 3x 2 )/(3y 2 − 6y), y(0) = 1
and determine the interval in which the solution is valid.
Hint: To find the interval of definition, look for points where the integral curve has a vertical
tangent.
46 Chapter 2. First Order Differential Equations

� 22. Solve the initial value problem

y  = 3x 2 /(3y 2 − 4), y(1) = 0
and determine the interval in which the solution is valid.
Hint: To find the interval of definition, look for points where the integral curve has a vertical
tangent.
� 23. Solve the initial value problem
y  = 2y 2 + x y 2 , y(0) = 1
and determine where the solution attains its minimum value.
� 24. Solve the initial value problem
y  = (2 − e x )/(3 + 2y), y(0) = 0
and determine where the solution attains its maximum value.
� 25. Solve the initial value problem
y  = 2 cos 2x/(3 + 2y), y(0) = −1
and determine where the solution attains its maximum value.
� 26. Solve the initial value problem
y  = 2(1 + x)(1 + y 2 ), y(0) = 0
and determine where the solution attains its minimum value.
� 27. Consider the initial value problem
y  = t y(4 − y)/3, y(0) = y0 .
(a) Determine how the behavior of the solution as t increases depends on the initial
value y0 .
(b) Suppose that y0 = 0.5. Find the time T at which the solution first reaches the value
3.98.
� 28. Consider the initial value problem
y  = t y(4 − y)/(1 + t), y(0) = y0 > 0.
(a) Determine how the solution behaves as t → ∞.
(b) If y0 = 2, find the time T at which the solution first reaches the value 3.99.
(c) Find the range of initial values for which the solution lies in the interval 3.99 < y <
4.01 by the time t = 2.
29. Solve the equation
dy ay + b
= ,
dx cy + d
where a, b, c, and d are constants.

Homogeneous Equations. If the right side of the equation dy/d x = f (x, y) can be expressed
as a function of the ratio y/x only, then the equation is said to be homogeneous. Such equations
can always be transformed into separable equations by a change of the dependent variable.
Problem 30 illustrates how to solve first order homogeneous equations.

� 30. Consider the equation

dy y − 4x
= . (i)
dx x−y
2.3 Modeling with First Order Equations 47

(a) Show that Eq. (i) can be rewritten as

dy (y/x) − 4
= ; (ii)
dx 1 − (y/x)
thus Eq. (i) is homogeneous.
(b) Introduce a new dependent variable v so that v = y/x, or y = xv(x). Express dy/d x
in terms of x, v, and dv/d x.
(c) Replace y and dy/d x in Eq. (ii) by the expressions from part (b) that involve v and
dv/d x. Show that the resulting differential equation is
dv v−4
v+x = ,
dx 1−v
or
dv v2 − 4
x = . (iii)
dx 1−v
Observe that Eq. (iii) is separable.
(d) Solve Eq. (iii) for v in terms of x.
(e) Find the solution of Eq. (i) by replacing v by y/x in the solution in part (d).
(f) Draw a direction field and some integral curves for Eq. (i). Recall that the right side
of Eq. (1) actually depends only on the ratio y/x. This means that integral curves have the
same slope at all points on any given straight line through the origin, although the slope
changes from one line to another. Therefore the direction field and the integral curves are
symmetric with respect to the origin. Is this symmetry property evident from your plot?

The method outlined in Problem 30 can be used for any homogeneous equation. That is, the
substitution y = xv(x) transforms a homogeneous equation into a separable equation. The latter
equation can be solved by direct integration, and then replacing v by y/x gives the solution to
the original equation. In each of Problems 31 through 38:
(a) Show that the given equation is homogeneous.
(b) Solve the differential equation.
(c) Draw a direction field and some integral curves. Are they symmetric with respect to
the origin?
dy x 2 + x y + y2 dy x 2 + 3y 2
� 31. = � 32. =
dx x2 dx 2x y
dy 4y − 3x dy 4x + 3y
� 33. = � 34. =−
dx 2x − y dx 2x + y
dy x + 3y
� 35. = � 36. (x 2 + 3x y + y 2 ) d x − x 2 dy = 0
dx x−y
dy x 2 − 3y 2 dy 3y 2 − x 2
� 37. = � 38. =
dx 2x y dx 2x y

2.3 Modeling with First Order Equations

Differential equations are of interest to nonmathematicians primarily because of the
possibility of using them to investigate a wide variety of problems in the physical,
ODE biological, and social sciences. One reason for this is that mathematical models and

ODE ODE
48 Chapter 2. First Order Differential Equations

their solutions lead to equations relating the variables and parameters in the problem.
These equations often enable you to make predictions about how the natural process will
behave in various circumstances. It is often easy to vary parameters in the mathematical
model over wide ranges, whereas this may be very time-consuming or expensive
in an experimental setting. Nevertheless, mathematical modeling and experiment or
observation are both critically important and have somewhat complementary roles in
scientific investigations. Mathematical models s are validated by comparison of their
predictions with experimental results. On the other hand, mathematical analyses may
suggest the most promising directions to explore experimentally, and may indicate
fairly precisely what experimental data will be most helpful.
In Sections 1.1 and 1.2 we formulated and investigated a few simple mathemati-
cal models. We begin by recapitulating and expanding on some of the conclusions
reached in those sections. Regardless of the specific field of application, there are three
identifiable steps that are always present in the process of mathematical modeling.

Construction of the Model. This involves a translation of the physical situation into
mathematical terms, often using the steps listed at the end of Section 1.1. Perhaps
most critical at this stage is to state clearly the physical principle(s) that are believed
to govern the process. For example, it has been observed that in some circumstances
heat passes from a warmer to a cooler body at a rate proportional to the temperature
difference, that objects move about in accordance with Newton’s laws of motion, and
that isolated insect populations grow at a rate proportional to the current population.
Each of these statements involves a rate of change (derivative) and consequently, when
expressed mathematically, leads to a differential equation. The differential equation is
a mathematical model of the process.
It is important to realize that the mathematical equations are almost always only an
approximate description of the actual process. For example, bodies moving at speeds
comparable to the speed of light are not governed by Newton’s laws, insect popula-
tions do not grow indefinitely as stated because of eventual limitations on their food
supply, and heat transfer is affected by factors other than the temperature difference.
Alternatively, one can adopt the point of view that the mathematical equations exactly
describe the operation of a simplified physical model, which has been constructed (or
conceived of) so as to embody the most important features of the actual process. Some-
times, the process of mathematical modeling involves the conceptual replacement of a
discrete process by a continuous one. For instance, the number of members in an insect
population changes by discrete amounts; however, if the population is large, it seems
reasonable to consider it as a continuous variable and even to speak of its derivative.

Analysis of the Model. Once the problem has been formulated mathematically, one
is often faced with the problem of solving one or more differential equations or,
failing that, of finding out as much as possible about the properties of the solution.
It may happen that this mathematical problem is quite difficult and, if so, further
approximations may be indicated at this stage to make the problem mathematically
tractable. For example, a nonlinear equation may be approximated by a linear one,
or a slowly varying coefficient may be replaced by a constant. Naturally, any such
approximations must also be examined from the physical point of view to make
sure that the simplified mathematical problem still reflects the essential features of
the physical process under investigation. At the same time, an intimate knowledge
of the physics of the problem may suggest reasonable mathematical approximations