284 Magnetostatic Fields

TABLE 7.2 Maxwell's Equations for Static EM Fields

Differential (or Point) Form Integral Form Remarks

V • D = pv D • dS = pv dv Gauss's law

V-B = 0 B • dS = Nonexistence of magnetic

monopole

VXE= 0 E • d\ = ' Conservativeness of

electrostatic field

Vx H= J H • d\ = J • dS Ampere's law

The choice between differential and integral forms of the equations depends on a
given problem. It is evident from Table 7.2 that a vector field is defined completely by
specifying its curl and divergence. A field can only be electric or magnetic if it satisfies the
corresponding Maxwell's equations (see Problems 7.26 and 7.27). It should be noted that
Maxwell's equations as in Table 7.2 are only for static EM fields. As will be discussed in
Chapter 9, the divergence equations will remain the same for time-varying EM fields but
the curl equations will have to be modified.

7.7 MAGNETIC SCALAR AND VECTOR POTENTIALS

We recall that some electrostatic field problems were simplified by relating the electric po-
tential V to the electric field intensity E (E = — VV). Similarly, we can define a potential
associated with magnetostatic field B. In fact, the magnetic potential could be scalar Vm or
vector A. To define Vm and A involves recalling two important identities (see Example 3.9
and Practice Exercise 3.9):
V X (VV) = 0 (7.35a)

V • (V X A) = 0 (7.35b)

which must always hold for any scalar field V and vector field A.
Just as E = - VV, we define the magnetic scalar potential Vm (in amperes) as related
to H according to

H= -VVm if J = 0 (7.36)

The condition attached to this equation is important and will be explained. Combining eq.
(7.36) and eq. (7.19) gives

J = V X H = V X ( - VVm) = 0 (7.37)
 7.7 MAGNETIC SCALAR AND VECTOR POTENTIALS 285

since Vm must satisfy the condition in eq. (7.35a). Thus the magnetic scalar potential Vm is
only defined in a region where J = 0 as in eq. (7.36). We should also note that Vm satisfies
Laplace's equation just as V does for electrostatic fields; hence,
V2Vm = 0, (J = 0) (7.38)

We know that for a magnetostatic field, V • B = 0 as stated in eq. (7.34). In order to

satisfy eqs. (7.34) and (7.35b) simultaneously, we can define the vector magnetic potential
A (in Wb/m) such that

B = VXA (7.39)

Just as we defined

dQ
V = (7.40)
4ireor

we can define

for line current (7.41)

for surface current (7.42)

for volume current (7.43)

Rather than obtaining eqs. (7.41) to (7.43) from eq. (7.40), an alternative approach
would be to obtain eqs. (7.41) to (7.43) from eqs. (7.6) to (7.8). For example, we can derive
eq. (7.41) from eq. (7.6) in conjunction with eq. (7.39). To do this, we write eq. (7.6) as

IdV X R
(7.44)
R3

where R is the distance vector from the line element dV at the source point (x1, y', z') to the
field point (x, y, z) as shown in Figure 7.19 and R = |R|, that is,
R = |r - r'| = [(x - x'f + (y - y'f + (z - z'f]1 (7.45)

Hence,

v| 1\ = (x - x')ax + (y - /)a y R
Rj [(x - x'f - y'f + (z - z')T2
286 • Magnetostatic Fields

Figure 7.19 Illustration of the source point

(*', y', z') and the field point (x, y, z).

R=r r'

(x, y, z)

or

(7.46)
R R

where the differentiation is with respect to x, y, and z. Substituting this into eq. (7.44), we
obtain

B= - - M IdV X v ( - (7.47)
4TT R

We apply the vector identity

V X ( / F ) = / V X F + (V/)XF (7.48)

where/is a scalar field and F is a vector field. Taking / = \IR and F = dV, we have

Since V operates with respect to (x, y, z) while dV is a function of (x\ y', z'), V X dV = 0.
Hence,

(7.49)

With this equation, eq. (7.47) reduces to

B = Vx (7.50)
4irR

Comparing eq. (7.50) with eq. (7.39) shows that

verifying eq. (7.41).

 7.7 MAGNETIC SCALAR AND VECTOR POTENTIALS • 287

By substituting eq. (7.39) into eq. (7.32) and applying Stokes's theorem, we obtain

= B • dS = (V X A) • dS = <P A • d\
4
or

(7.51)

Thus the magnetic flux through a given area can be found using either eq. (7.32) or (7.51).
Also, the magnetic field can be determined using either Vm or A; the choice is dictated by
the nature of the given problem except that Vm can only be used in a source-free region.
The use of the magnetic vector potential provides a powerful, elegant approach to solving
EM problems, particularly those relating to antennas. As we shall notice in Chapter 13, it
is more convenient to find B by first finding A in antenna problems.

Given the magnetic vector potential A = —p2/4 az Wb/m, calculate the total magnetic flux
EXAMPLE 7.7
crossing the surface <f> = -ir/2, 1 < p < 2 m , 0 < z < 5 m .

Solution:
We can solve this problem in two different ways: using eq. (7.32) or eq. (7.51).
Method 1:

B = V x A = —-* a0 = 2J a0, dS = dp dz a 0
op

Hence,

1 15
TP = j B • dS = - p dp dz = — p (5) =
z=0 •>=
¥ = 3.75 Wb

Method 2:
We use

<p = I A • d\ = fi + v 2 + r3 + v4
'L

where L is the path bounding surface S; V 1( f 2, ¥3, and V4 are, respectively, the evalua-
tions of /A • d\ along the segments of L labeled 1 to 4 in Figure 7.20. Since A has only a
2-component,

¥>, = 0 = y 3