15.

2 Divergence and Curl 1231

15.2 Divergence and Curl

In this section we will look at two ways of measuring the rate of change of a vector
field F: the divergence of F at a point P and the curl of F at P. The divergence and
curl of a vector field play a very important role in describing fluid flow, heat conduc-
tion, and electromagnetism.

Divergence
Suppose that F is a vector field in 2- or 3-space and P is a point in its domain. For the
purpose of this discussion, let’s also suppose that the vector field F describes the flow
of a fluid in 2- or 3-space. Then the divergence of F at P, written div F(P), measures
the rate per unit area (or volume) at which the fluid departs or accumulates at P. Let’s
consider several examples.

EXAMPLE 1
a. Figure 1a shows the vector field F(x, y)  xi  yj described in Example 1
of Section 15.1. Let P be a point in the plane, and let N be a neighborhood
of P with center P. Referring to Figure 1b, observe that an arrow entering N
along a streamline is matched by one that emerges from N and has a greater
length (because it is located farther away from the origin). This shows that
more fluid leaves than enters a neighborhood of P. We will show in Example 2a
that the vector field F is “divergent” at P; that is, the divergence of F at P is
positive.

P
N

P
x

(a) The vector field F(x, y)  x i  yi (b) Flow through a neighborhood of P

FIGURE 1 (enlarged and not to scale)

b. Figure 2a shows the vector field F(x, y)  yi for x  0 and y  0. Observe that
the streamlines are parallel to the x-axis and that the lengths of the arrows on
each horizontal line are constant. We can think of F as describing the flow of a
river near one side of a riverbank. The velocity of flow is near zero close to the
bank (the x-axis) and increases as we move away from it. You can see from Fig-
ure 2b that the amount of fluid flowing into the neighborhood N of P is matched
by the same amount that exits N. Consequently, we expect the “divergence” at P
to be zero. We will show that this is the case in Example 2b.
1232 Chapter 15 Vector Analysis

P P

N
0 1 2 3 4 x

(a) The vector field F(x, y)  yi (b) Flow through a neighborhood of P

FIGURE 2 (enlarged and not to scale)

c. Figure 3a shows the vector field

1
F(x, y)  i
x1
for x  0 and y  0. Observe that the streamlines are parallel to the x-axis and
that the lengths of the arrows on each horizontal line get smaller as x increases.
From Figure 3b you can see that the “flow” into a neighborhood N of P is greater
than the flow that emerges from N. In this case, more fluid enters the neighbor-
hood than leaves it, and the “divergence” is negative. We will show that our intu-
ition is correct in Example 2.

N P
1
P

N
0 1 2 x

1
(a) The vector field F(x, y)  x  1 i (b) Flow through a neighborhood of P
FIGURE 3 (enlarged and not to scale)

Up to now, we have looked at the notion of “divergence” intuitively. The diver-

gence of a vector field can be defined as follows.

DEFINITION Divergence of a Vector Field

Let F(x, y, z)  Pi  Qj  Rk be a vector field in space, where P, Q, and R
have first-order partial derivatives in some region T. The divergence of F is the
scalar function defined by
P Q R
div F    (1)
x y z

(We will justify this definition of divergence in Section 15.8.) In two-dimensional space,
P Q
F(x, y)  Pi  Qj and div F  
x y
 15.2 Divergence and Curl 1233

As an aid to remembering Equation (1), let’s introduce the vector differential oper-
ator § (read “del”) defined by
  
§ i j k
x y z
If we let § operate on a scalar function f(x, y, z), we obtain
  
§f(x, y, z)  a i j kb f(x, y, z)
x y z
  
 f(x, y, z)i  f(x, y, z)j  f(x, y, z)k
x y z
f f f
 (x, y, z)i  (x, y, z)j  (x, y, z)k
x y z
which is the gradient of f. If we take the “dot product” of § with the vector field
F(x, y, z) Pi  Qj  Rk, we obtain
  
§ⴢF a i j kb ⴢ (Pi  Qj  Rk)
x y z
   P Q R
 P Q R  
x y z x y z
which is the divergence of the vector field F. Thus, we can write the divergence of F
symbolically as
div F  § ⴢ F (2)
Let’s apply the definition of divergence to the vector fields that we discussed in
Example 1.

EXAMPLE 2 Find the divergence of (a) F(x, y)  xi  yj, (b) F(x, y)  yi, and
1
(c) F(x, y)  i. Reconcile your results with the intuitive observations that
x1
were made in Example 1.

Solution
 
a. div F  (x)  (y)  1  1  2. Here, div F 0, as expected.
x y
 
b. Here, F  yi  0j, so div F  (y)  (0)  0. In this case, div F  0, as
x y
was observed in Example 1b.
c. With F  (x  1)1i  0j we find
  1
div F  (x  1)1  (0)  (x  1)2  
x y (x  1)2
and div F  0, as we concluded intuitively in Example 1c.

We turn now to an example involving a vector field whose streamlines are not so
easily visualized.
1234 Chapter 15 Vector Analysis

EXAMPLE 3 Find the divergence of F(x, y, z)  xyzi  x 2y 2zj  xy 2k at the point

(1, 1, 2).

Solution
  2 2 
div F  (xyz)  (x y z)  (xy 2)
x y z
 yz  2x 2yz
In particular, at the point (1, 1, 2) we find
div F(1, 1, 2)  (1)(2)  2(1)2 (1)(2)  6

The divergence of the vector field F(x, y)  yi of Examples 1b and 2b is zero. In

general, if div F  0, then F is called incompressible. In electromagnetic theory a vec-
tor field F that satisfies § ⴢ F  0 is called solenoidal. For example, the electric field
E in Example 4 is solenoidal. We will study the divergence of a vector field in greater
detail in Section 15.8.
kQ
EXAMPLE 4 Show that the divergence of the electric field E(x, y, z)  r, where
冟 r 冟3
r  xi  yj  zk, is zero.

Solution We first write

kQx kQy kQz
E(x, y, z)  i j k
(x  y  z )
2 2 2 3>2
(x  y  z )
2 2 2 3>2
(x  y 2  z 2) 3>2
2

Then
  
c 2 d c 2 d c 2 df
x y z
div E  kQe
x (x  y  z )
2 2 3>2 y (x  y  z )
2 2 3>2 z (x  y  z 2)3>2
2

But
 
c 2 d
x
[x(x 2  y 2  z 2)3>2]
x (x  y  z )
2 2 3>2 x

 (x 2  y 2  z 2) 3>2  x ⴢ a b (x 2  y 2  z 2)5>2 (2x)

3
2
 (x 2  y 2  z 2) 5>2[(x 2  y 2  z 2)  3x 2]
2x 2  y 2  z 2

(x 2  y 2  z 2) 5>2
Similarly, we find
 x 2  2y 2  z 2
c 2 d 2
y
y (x  y  z )
2 2 3>2
(x  y 2  z 2) 5>2
and
 x 2  y 2  2z 2
c 2 d
z

z (x  y 2  z 2) 3>2 (x 2  y 2  z 2) 5>2
Therefore,
2x 2  y 2  z 2 x 2  2y 2  z 2 x 2  y 2  2z 2
div E  kQc   d0
(x 2  y 2  z 2) 5>2 (x 2  y 2  z 2)5>2 (x 2  y 2  z 2) 5>2
 15.2 Divergence and Curl 1235

Axis
Curl
We now turn our attention to the other measure of the rate of change of a vector field
F. Let F be a vector field in 3-space, and let P be a point in its domain. Once again,
let’s think of the vector field as one that describes the flow of fluid. Suppose that a
small paddle wheel, like the one shown in Figure 4, is immersed in the fluid at P. Then
the curl of F, written curl F, is a measure of the tendency of the fluid to rotate the
device about its vertical axis at P. Later, we will show that the paddle wheel will rotate
most rapidly if its axis coincides with the direction of curl F at P and that its maxi-
mum rate of rotation at P is given by the length of curl F at P.

FIGURE 4 EXAMPLE 5
A paddle wheel
a. Consider the vector field F(x, y, z)  yi for x  0 similar to that of Example 1b.
This field is shown in Figure 5a. Notice that the positive z-axis points vertically
out of the page. Suppose that a paddle wheel is planted at a point P. Referring
to Figure 5b, you can see that the arrows in the upper half of the circle with
center at P are longer than those in the lower half. This shows that the net
clockwise flow of the fluid is greater than the net counterclockwise flow. This
will cause the paddle to rotate in a clockwise direction, as we will show in
Example 6.

5
4
P
3 P
2
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 x

(a) The vector field F(x, y, z)  yi (b) Flow through a neighborhood of P

at which a paddle wheel is located
FIGURE 5 (enlarged and not to scale)

b. Consider the vector field F(x, y, z)  yi  xj shown in Figure 6a. Observe
that it is similar to the spin vector field of Example 2 in Section 15.1. Again,
the positive z-axis points vertically out of the page. If a paddle wheel is placed
at the origin, it is easy to see that it will rotate in a counterclockwise direction.
Next, suppose that the paddle wheel is planted at a point P other than the ori-
gin. If you refer to Figure 6b, you can see that the circle with center at P is
divided into two arcs by the points of tangency of the two half-lines starting
from the origin. Notice that the arc farther from the origin is longer than the
one closer to the origin and that the flow on the larger arc is counterclockwise,
whereas the flow on the shorter arc is clockwise. Furthermore, the arrows ema-
nating from the longer arc are longer than those emanating from the shorter
arc. This shows that the amount of fluid flowing in the counterclockwise direc-
tion is greater than that flowing in the clockwise direction. Therefore, the pad-
dle wheel will rotate in a counterclockwise direction, as we will show in
Example 6.
1236 Chapter 15 Vector Analysis

y y

P P
0 x

0 x

(a) The vector field F(x, y, z)  yi  xj (b) Flow through a neighborhood of P at
which a paddle wheel is located (enlarged
FIGURE 6 and not to scale)

c. Consider the vector field F(x, y, z)  xi  yj shown in Figure 7a. Note that it is
similar to the vector field in Example 1 in Section 15.1. Suppose that a paddle
wheel is placed at a point P. Then referring to Figure 7(b) and using an argument
involving symmetry, you can convince yourself that the paddle wheel will not
rotate. Again, we will show in Example 6 that this is true.

y
y

P
N

0 x

(a) The vector field F(x, y, z)  xi  yj (b) Flow through a neighborhood of P at

which a paddle wheel is located (enlarged
FIGURE 7 and not to scale)

The following definition provides us with an exact way to measure the curl of a
vector field.

DEFINITION Curl of a Vector Field

Let F(x, y, z)  Pi  Qj  Rk be a vector field in space, where P, Q, and R
have first-order partial derivatives in some region T. The curl of F is the vector
field defined by
R Q P R Q P
curl F  § Fa  bi  a  bj  a  bk
y z z x x y

(We will justify this definition in Section 15.9.)

 15.2 Divergence and Curl 1237

As in the case of the cross product of two vectors, we can remember the expres-
sion for the curl of a vector field by writing it (formally) in determinant form:
i j k
  
curl F  § F∞ ∞
x y z
P Q R
R Q P R Q P
a  bi  a  bj  a  bk
y z z x x y
Let’s apply this definition to the vector fields that we discussed in Example 5.

EXAMPLE 6 Find the curl of (a) F(x, y, z)  yi for x  0, (b) F(x, y, z)  yi  xj,
and (c) F(x, y, z)  xi  yj. Reconcile your results with the intuitive observations that
were made in Example 5.

Solution
a. i j k
  
curl F  § F∞ ∞
x y z
y 0 0
     
c (0)  (0)di  c (0)  (y)d j  c (0)  (y)dk  k
y z x z x y

This shows that curl F is a (unit) vector that points vertically into the page. Apply-
ing the right-hand rule, we see that this result tells us that at any point in the vector
field, the paddle wheel will rotate in a clockwise direction, as was observed earlier.
b. i j k
  
curl F  § F∞ ∞
x y z
y x 0
     
c (0)  (x)di  c (0)  (y)d j  c (x)  (y)dk
y z x z x y
 2k
The result tells us that curl F points vertically out of the page, so the paddle
wheel will rotate in a counterclockwise direction when placed at any point in the
vector field F.
c. i j k
  
curl F  § F∞ ∞
x y z
x y 0
     
c (0)  (y)di  c (0)  (x)d j  c (y)  (x)dk  0
y z x z x y
This shows that a paddle wheel placed at any point in F will not rotate, as
observed earlier.
1238 Chapter 15 Vector Analysis

The vector field F in Example 6c has the property that curl F  0 at any point P.
In general, if curl F  0 at a point P, then F is said to be irrotational at P. This means
that there are no vortices or whirlpools there.

EXAMPLE 7
a. Find curl F if F(x, y, z)  xyi  xzj  xyz 2k.
b. What is curl F(1, 2, 1)?

Solution
a. By definition,
i j k
  
curl F  § F ∞ ∞
x y z
xy xz xyz 2
     
c (xyz 2)  (xz)di  c (xyz 2)  (xy)d j  c (xz)  (xy)dk
y z x z x y
 (xz 2  x)i  yz 2j  (z  x)k
 x(z 2  1)i  yz 2j  (z  x)k
b. curl F(1, 2, 1)  (1)(12  1)i  (2)(12)j  [1  (1)]k  2j  2k

The div and curl of vector fields enjoy some algebraic properties as illustrated in
the following examples. Other properties can be found in the exercises at the end of
this section.

EXAMPLE 8 Let f be a scalar function, and let F be a vector field. If f and the com-
ponents of F have first-order partial derivatives, show that
div( f F)  f div F  F ⴢ §f

Solution Let’s write F  Pi  Qj  Rk, where P, Q, and R are functions of x, y,

and z. Then
f F  f(Pi  Qj  Rk)  f Pi  fQj  fRk
so the left-hand side of the given equation reads
  
div( f F)  § ⴢ ( f F)  a i j kb ⴢ ( fPi  fQj  fRk)
x y z
  
 ( fP)  ( fQ)  ( fR)
x y z
P f Q f R f
f  Pf  Qf  R
x x y y z z
P Q R f f f
 fa   ba P Q Rb
x y z x y z
 f( § ⴢ F)  ( §f ) ⴢ F
 f div F  F ⴢ §f
which is equal to the right-hand side.
 15.2 Divergence and Curl 1239

EXAMPLE 9 Let F  Pi  Qj  Rk be a vector field in space, and suppose that P,

Q, and R have continuous second-order partial derivatives. Show that
div curl F  0

Solution Direct computation shows that

div curl F  § ⴢ ( § F)
   R Q P R Q P
a i j kb ⴢ c a  bi  a  bj  a  bkd
x y z y z z x x y
 R Q  P R  Q P
 a  b a  b a  b
x y z y z x z x y
 2R  2Q  2P  2R  2Q  2P
     
x y x z y z y x z x z y
0
Here we have used the fact that the mixed derivatives are equal because, by assump-
tion, they are continuous.

15.2 CONCEPT QUESTIONS

1. a. Define the divergence of a vector field F and give a for- 2. a. What is meant by a vector field F that is incompressible?
mula for finding it. Give a physical example of an (almost) incompressible
b. Define the curl of a vector field F, and give a formula for field.
finding it. b. Repeat part (a) for an irrotational vector field.
c. Suppose that F is the velocity vector field associated with
the airflow around an airfoil. Give an interpretation of
§ ⴢ F and § F.

15.2 EXERCISES
In Exercises 1–4, you are given the vector field F and a plot 2. F(x, y, z)  xj
of the vector field in the xy-plane. (The z-component of F is 0.)
(a) By studying the plot of F, determine whether div F is posi-
tive, negative, or zero. Justify your answer. (b) Find div F, and 5
reconcile your result with your answer in part (a). (c) By study- 4
ing the plot of F, determine whether a paddle wheel planted at 3
a point in the field will rotate clockwise, rotate counterclockwise, 2
or not rotate at all. Justify your answer. (d) Find curl F, and 1
reconcile your result with your answer in part (c). 432 1 1 2 3 4x
x y 2
1. F(x, y, z)  i, x0
冟x冟 3
2
4
1 5

3 2 1 1 2 3 x
1

2

V Videos for selected exercises are available online at www.academic.cengage.com/login.

1240 Chapter 15 Vector Analysis

x y 18. Show that the vector field F(x, y, z)  f(x)i  t(y)j  h(z)k,
3. F(x, y, z)  i j
2x  y
2 2
2x  y
2 2 where f, t and h are differentiable, is irrotational.
y
3
In Exercises 19–26, prove the property for vector fields F and G
and scalar fields f and t. Assume that the appropriate partial
2
derivatives exist and are continuous.
1
19. div(F  G)  div F  div G
3 2 1 1 2 3 x 20. curl(F  G)  curl F  curl G
1
21. curl( f F)  f curl F  ( §f ) F
2
3 22. curl( §f )  0
23. div(F G)  G ⴢ curl F  F ⴢ curl G
y x
4. F(x, y, z)   i j 24. div( §f §t)  0
2x 2  y 2 2x 2  y 2
25. § (§ F)  § ( § ⴢ F)  §2F, where
y  2
2 2
3 §2F  a 2  2  2 bF
x y z
2
26. § [ §f  ( § F)]  § (§ F)
1
Hint: Use the results of Exercises 20 and 22.

3 2 1 1 2 3 x 27. Show that there is no vector field F in space such that

1 curl F  xyi  yzj  xyk.
Hint: See Example 9.
2
28. Find the value of the constant c such that the vector field
3
G(x, y, z)  (2x  3y  z 2)i  (cy  z)j  (x  y  2z)k
In Exercises 5–12, find (a) the divergence and (b) the curl of the is the curl of some vector field F.
vector field F. 29. Show that F  (cos x)yi  (sin y)xj is not a gradient vector
5. F(x, y, z)  yzi  xzj  xyk field.
Hint: See Exercise 22.
6. F(x, y, z)  x 2yi  xy 2j  xyzk
30. Let f be a differentiable function, r  xi  yj  zk, and
7. F(x, y, z)  x 2y 3i  xz 2k
r  冟 r 冟.
8. F(x, y, z)  yz 2i  x 2zj a. Find curl[ f(r)r] by interpreting it geometrically.
9. F(x, y, z)  sin xi  x cos yj  sin zk b. Verify your answer to part (a) analytically.
10. F(x, y, z)  x cos yi  y tan xj  sec zk In Exercises 31–34, let r  xi  yj  zk and r  冟 r 冟.
11. F(x, y, z)  ex cos yi  ex sin yj  ln zk 31. Show that §r  r>r.
12. F(x, y, z)  exyz i  cos(x  y)j  ln(x  z)k 32. Show that §(1>r)  r>r 3.
In Exercises 13–15, let F be a vector field, and let f be a scalar 33. Show that §(ln r)  r>r 2.
field. Determine whether each expression is meaningful. If so, state 34. Show that §r n  nr n2r.
whether the expression represents a scalar field or a vector field.
13. a. § f b. § ⴢ f In Exercises 35–38, the differential operator §2 (called the
c. § ( §f ) d. grad F 2 2 2
Laplacian) is defined by §2  § ⴢ §  2  2  2 .
14. a. div( §f ) b. grad( §f ) x y z
c. § (grad f ) d. curl(curl F)  2
f  2
f  2f
It acts on f to produce the function §2f  2  2  2 .
15. a. § ( § F) b. § ⴢ ( §f ) x y z
Assume that f and t have second-order partial derivatives.
c. § ⴢ ( § ⴢ F) d. § [ § ( §f )]
35. Show that § ⴢ ( §f )  §2f.
16. Find div F if F  grad f, where f(x, y, z)  2xy 2z 3.
36. Show that §2( ft)  f §2t  t§2f  2§f ⴢ §t.
17. Show that the vector field F(x, y, z)  f(y, z)i  t(x, z)j 
h(x, y)k, where f, t, and h are differentiable, is incompress- 37. Show that §2r 3  12r, where r  冟 r 冟 and r  xi  yj  zk.
ible.