12 Chapter 1.

Introduction

a particular value of c, and is the graph of the solution corresponding to that value of c.
Satisfying an initial condition amounts to identifying the integral curve that passes
through the given initial point.
To relate the solution (18) to Eq. (2), which models the field mouse population, we
need only replace a by the growth rate r and b by the predation rate k. Then the solution
(18) becomes
p = (k/r ) + [ p0 − (k/r )]er t , (19)
where p0 is the initial population of field mice. The solution (19) confirms the con-
clusions reached on the basis of the direction field and Example 1. If p0 = k/r , then
from Eq. (19) it follows that p = k/r for all t; this is the constant, or equilibrium,
solution. If p0 = k/r , then the behavior of the solution depends on the sign of the
coefficient p0 − (k/r ) of the exponential term in Eq. (19). If p0 > k/r , then p grows
exponentially with time t; if p0 < k/r , then p decreases and eventually becomes
zero, corresponding to extinction of the field mouse population. Negative values of p,
while possible for the expression (19), make no sense in the context of this particular
problem.
To put the falling object equation (1) in the form (3), we must identify a with −γ /m
and b with −g. Making these substitutions in the solution (18), we obtain
v = (mg/γ ) + [v0 − (mg/γ )]e−γ t/m , (20)
where v0 is the initial velocity. Again, this solution confirms the conclusions reached
in Section 1.1 on the basis of a direction field. There is an equilibrium, or constant,
solution v = mg/γ , and all other solutions tend to approach this equilibrium solution.
The speed of convergence to the equilibrium solution is determined by the exponent
−γ /m. Thus, for a given mass m the velocity approaches the equilibrium value faster
as the drag coefficient γ increases.

Suppose that, as in Example 2 of Section 1.1, we consider a falling object of mass

EXAMPLE m = 10 kg and drag coefficient γ = 2 kg/sec. Then the equation of motion (1) becomes
2 dv v
= 9.8 − . (21)
A Falling dt 5
Object Suppose this object is dropped from a height of 300 m. Find its velocity at any time t.
(continued) How long will it take to fall to the ground, and how fast will it be moving at the time
of impact?
The first step is to state an appropriate initial condition for Eq. (21). The word
“dropped” in the statement of the problem suggests that the initial velocity is zero, so
we will use the initial condition
v(0) = 0. (22)
The solution of Eq. (21) can be found by substituting the values of the coefficients
into the solution (20), but we will proceed instead to solve Eq. (21) directly. First,
rewrite the equation as
dv/dt 1
=− . (23)
v − 49 5
1.2 Solutions of Some Differential Equations 13

By integrating both sides we obtain

t
ln |v − 49| = − + C, (24)
5
and then the general solution of Eq. (21) is
v = 49 + ce−t/5 , (25)
where c is arbitrary. To determine c, we substitute t = 0 and v = 0 from the initial
condition (22) into Eq. (25), with the result that c = −49. Then the solution of the
initial value problem (21), (22) is
v = 49(1 − e−t/5 ). (26)
Equation (26) gives the velocity of the falling object at any positive time (before it hits
the ground, of course).
Graphs of the solution (25) for several values of c are shown in Figure 1.2.2, with the
solution (26) shown by the heavy curve. It is evident that all solutions tend to approach
the equilibrium solution v = 49. This confirms the conclusions we reached in Section
1.1 on the basis of the direction fields in Figures 1.1.2 and 1.1.3.

v
100

80

60

40
(10.51, 43.01)

20 v = 49 (1 – e–t/5)

2 4 6 8 10 12 t

FIGURE 1.2.2 Graphs of the solution (25) for several values of c.

To find the velocity of the object when it hits the ground, we need to know the time
at which impact occurs. In other words, we need to determine how long it takes the
object to fall 300 m. To do this, we note that the distance x the object has fallen is
related to its velocity v by the equation v = dx/dt, or
dx
= 49(1 − e−t/5 ). (27)
dt
Consequently,
x = 49t + 245e−t/5 + c, (28)
14 Chapter 1. Introduction

where c is an arbitrary constant of integration. The object starts to fall when t = 0,

so we know that x = 0 when t = 0. From Eq. (28) it follows that c = −245, so the
distance the object has fallen at time t is given by
x = 49t + 245e−t/5 − 245. (29)
Let T be the time at which the object hits the ground; then x = 300 when t = T . By
substituting these values in Eq. (29) we obtain the equation
49T + 245e−T /5 − 545 = 0. (30)
The value of T satisfying Eq. (30) can be readily approximated by a numerical process
using a scientific calculator or computer, with the result that T ∼
= 10.51 sec. At this
time, the corresponding velocity vT is found from Eq. (26) to be vT ∼
= 43.01 m/sec.

Further Remarks on Mathematical Modeling. Up to this point we have related our

discussion of differential equations to mathematical models of a falling object and of
a hypothetical relation between field mice and owls. The derivation of these models
may have been plausible, and possibly even convincing, but you should remember
that the ultimate test of any mathematical model is whether its predictions agree with
observations or experimental results. We have no actual observations or experimental
results to use for comparison purposes here, but there are several sources of possible
discrepancies.
In the case of the falling object the underlying physical principle (Newton’s law of
motion) is well-established and widely applicable. However, the assumption that the
drag force is proportional to the velocity is less certain. Even if this assumption is
correct, the determination of the drag coefficient γ by direct measurement presents
difficulties. Indeed, sometimes one finds the drag coefficient indirectly, for example,
by measuring the time of fall from a given height, and then calculating the value of γ
that predicts this time.
The model of the field mouse population is subject to various uncertainties. The
determination of the growth rate r and the predation rate k depends on observations
of actual populations, which may be subject to considerable variation. The assumption
that r and k are constants may also be questionable. For example, a constant predation
rate becomes harder to sustain as the population becomes smaller. Further, the model
predicts that a population above the equilibrium value will grow exponentially larger
and larger. This seems at variance with the behavior of actual populations; see the
further discussion of population dynamics in Section 2.5.
Even if a mathematical model is incomplete or somewhat inaccurate, it may never-
theless be useful in explaining qualitative features of the problem under investigation.
It may also be valid under some circumstances but not others. Thus you should always
use good judgment and common sense in constructing mathematical models and in
using their predictions.

PROBLEMS
䉴 1. Solve each of the following initial value problems and plot the solutions for several values
of y0 . Then describe in a few words how the solutions resemble, and differ from, each
other.
(a) dy/dt = −y + 5, y(0) = y0 (b) dy/dt = −2y + 5, y(0) = y0
(c) dy/dt = −2y + 10, y(0) = y0