15 Multiple Integrals

Copyright © Cengage Learning. All rights reserved.

 Applications of Double
15.5 Integrals

Copyright © Cengage Learning. All rights reserved.

Density and Mass

3
Density and Mass
We were able to use single integrals to compute moments
and the center of mass of a thin plate or lamina with
constant density.

But now, equipped with the double integral, we can

consider a lamina with variable density.

Suppose the lamina occupies a region D of the xy-plane

and its density (in units of mass per unit area) at a point
(x, y) in D is given by  (x, y), where  is a continuous
function on D.

4
Density and Mass
This means that

where Δm and ΔA are the

mass and area of a small
rectangle that contains (x, y)
and the limit is taken as the
dimensions of the rectangle
approach 0. (See Figure 1.)
Figure 1

5
Density and Mass
To find the total mass m of the lamina we divide a rectangle
R containing D into subrectangles Rij of the same size
(as in Figure 2) and consider  (x, y) to be 0 outside D.

If we choose a point in
Rij, then the mass of the part of
the lamina that occupies Rij is
approximately  ΔA,
where ΔA is the area of Rij.
Figure 2

If we add all such masses, we get an approximation to the

total mass:
6
Density and Mass
If we now increase the number of subrectangles, we obtain
the total mass m of the lamina as the limiting value of the
approximations:

Physicists also consider other types of density that can be

treated in the same manner.

7
Density and Mass
For example, if an electric charge is distributed over a
region D and the charge density (in units of charge per unit
area) is given by (x, y) at a point (x, y) in D, then the total
charge Q is given by

8
Example 1
Charge is distributed over the triangular region D in Figure
3 so that the charge density at (x, y) is (x, y) = xy,
measured in coulombs per square meter (C/m2). Find the
total charge.

Figure 3
9
Example 1 – Solution
From Equation 2 and Figure 3 we have

10
Example 1 – Solution cont’d

Thus the total charge is C.

11
Moments and Centers of Mass

12
Moments and Centers of Mass
We have found the center of mass of a lamina with
constant density; here we consider a lamina with variable
density.

Suppose the lamina occupies a region D and has density

function  (x, y).

Recall that we defined the moment of a particle about an

axis as the product of its mass and its directed distance
from the axis.

We divide D into small rectangles.

13
Moments and Centers of Mass
Then the mass of Rij is approximately  A, so we
can approximate the moment of Rij with respect to the
x-axis by

If we now add these quantities and take the limit as the

number of subrectangles becomes large, we obtain the
moment of the entire lamina about the x-axis:

14
Moments and Centers of Mass
Similarly, the moment about the y-axis is

As before, we define the center of mass so that

and

The physical significance is that the lamina behaves as if its

entire mass is concentrated at its center of mass.

15
Moments and Centers of Mass
Thus the lamina balances
horizontally when supported
at its center of mass
(see Figure 4).
Figure 4

16
Moment of Inertia

17
Moment of Inertia
The moment of inertia (also called the second moment)
of a particle of mass m about an axis is defined to be mr2,
where r is the distance from the particle to the axis.

We extend this concept to a lamina with density function

 (x, y) and occupying a region D by proceeding as we did
for ordinary moments.

We divide D into small rectangles, approximate the moment

of inertia of each subrectangle about the x-axis, and take
the limit of the sum as the number of subrectangles
becomes large.
18
Moment of Inertia
The result is the moment of inertia of the lamina about
the x-axis:

Similarly, the moment of inertia about the y-axis is

19
Moment of Inertia
It is also of interest to consider the moment of inertia
about the origin, also called the polar moment of inertia:

Note that I0 = Ix + Iy.

20
Example 4
Find the moments of inertia Ix, Iy, and I0 of a homogeneous
disk D with density  (x, y) = , center the origin, and radius a.

Solution:
The boundary of D is the circle x2 + y2 = a2 and in polar
coordinates D is described by 0    2, 0  r  a.

Let’s compute I0 first:

21
Example 4 – Solution cont’d

Instead of computing Ix and Iy directly, we use the facts that

Ix + Iy = I0 and Ix = Iy (from the symmetry of the problem).

Thus
Ix = I y = =
22
Moment of Inertia
In Example 4 notice that the mass of the disk is

m = density  area =  (a2)

so the moment of inertia of the disk about the origin (like a

wheel about its axle) can be written as

Thus if we increase the mass or the radius of the disk, we

thereby increase the moment of inertia.

23
Moment of Inertia
In general, the moment of inertia plays much the same role
in rotational motion that mass plays in linear motion.

The moment of inertia of a wheel is what makes it difficult

to start or stop the rotation of the wheel, just as the mass of
a car is what makes it difficult to start or stop the motion of
the car.

The radius of gyration of a lamina about an axis is the

number R such that

mR2 = I

24
Moment of Inertia
Where m is the mass of the lamina I and is the moment of
inertia about the given axis. Equation 9 says that if the
mass of the lamina were concentrated at a distance R from
the axis, then the moment of inertia of this “point mass”
would be the same as the moment of inertia of the lamina.

25
Moment of Inertia
In particular, the radius of gyration with respect to the x-
axis and the radius of gyration with respect to the y-axis
are given by the equations

Thus is the point at which the mass of the lamina can

be concentrated without changing the moments of inertia
with respect to the coordinate axes. (Note the analogy with
the center of mass.)

26
Probability

27
Probability
We have considered the probability density function f of a
continuous random variable X.

This means that f (x)  0 for all x, and the

probability that X lies between a and b is found by
integrating f from a to b:

28
Probability
Now we consider a pair of continuous random variables
X and Y, such as the lifetimes of two components of a
machine or the height and weight of an adult female
chosen at random.

The joint density function of X and Y is a function f of two

variables such that the probability that (X, Y) lies in a region
D is

29
Probability
In particular, if the region is a rectangle, the probability that
X lies between a and b and Y lies between c and d is

(See Figure 7.)

Figure 7
The probability that X lies between a and b and Y lies between c and d is the volume that
lies above the rectangle D = [a, b]  [c, d] and below the graph of the joint density function.
30
Probability
Because probabilities aren’t negative and are measured on
a scale from 0 to 1, the joint density function has the
following properties:

f (x, y)  0

The double integral over is an improper integral defined

as the limit of double integrals over expanding circles or
squares, and we can write

31
Example 6
If the joint density function for X and Y is given by

C(x + 2y) if 0  x  10, 0  y  10

f (x, y) =
0 otherwise

find the value of the constant C. Then find P(X  7, Y  2).

32
Example 6 – Solution
We find the value of C by ensuring that the double integral
of f is equal to 1. Because f (x, y) = 0 outside the rectangle
[0, 10]  [0, 10], we have

Therefore 1500C = 1 and so C = 33

Example 6 – Solution cont’d

Now we can compute the probability that X is at most 7 and

Y is at least 2:

34
Probability
Suppose X is a random variable with probability density
function f1(x) and Y is a random variable with density
function f2(y).
Then X and Y are called independent random variables if
their joint density function is the product of their individual
density functions:
f (x, y) = f1(x)f2(y)

We have modeled waiting times by using exponential density

functions
0 if t < 0
f (t) = –1e –t / if t  0
where  is the mean waiting time. 35
Expected Values

36
Expected Values
Recall that if X is a random variable with probability density
function f, then its mean is

Now if X and Y are random variables with joint density

function f, we define the X-mean and Y-mean, also called
the expected values of X and Y, to be

37
Expected Values
Notice how closely the expressions for 1 and 2 in
resemble the moments Mx and My of a lamina with density
function  in Equations 3 and 4.

In fact, we can think of probability as being like

continuously distributed mass. We calculate probability the
way we calculate mass—by integrating a density function.

And because the total “probability mass” is 1, the

expressions for and in show that we can think of the
expected values of X and Y, 1 and 2, as the coordinates
of the “center of mass” of the probability distribution.

38
Expected Values
In the next example we deal with normal distributions.
A single random variable is normally distributed if its
probability density function is of the form

where  is the mean and  is the standard deviation.

39
Example 8
A factory produces (cylindrically shaped) roller bearings that
are sold as having diameter 4.0 cm and length 6.0 cm.
In fact, the diameters X are normally distributed with mean
4.0 cm and standard deviation 0.01 cm while the lengths Y
are normally distributed with mean 6.0 cm and standard
deviation 0.01 cm. Assuming that X and Y are independent,
write the joint density function and graph it.

Find the probability that a bearing randomly chosen from the

production line has either length or diameter that differs
from the mean by more than 0.02 cm.
40
Example 8 – Solution
We are given that X and Y are normally distributed with
1 = 4.0, 2 = 6.0 and 1 = 2 = 0.01.

So the individual density functions for X and Y are

Since X and Y are independent, the joint density function is

the product:
f (x, y) = f1(x)f2(y)

41
Example 8 – Solution cont’d

A graph of this function is shown in Figure 9.

Figure 9
Graph of the bivariate normal joint
density function in this example
42
Example 8 – Solution cont’d

Let’s first calculate the probability that both X and Y differ

from their means by less than 0.02 cm. Using a calculator
or computer to estimate the integral, we have

43
Example 8 – Solution cont’d

Then the probability that either X or Y differs from its mean

by more than 0.02 cm is approximately
1 – 0.91 = 0.09

44