Chapter 5: Magnetostatics, Faraday’s Law,

Quasi-Static Fields
5.1 Introduction and Definitions
We begin with the law of conservation of charge:
3 Q  d 3x
v   J d x   J  d a   t
  t v B
, J

   J  t  0 conservation  da (5 2)
(5.2)
off charge
h 
arbitrary volume
Magnetostatics is applicable under the static condition. Hence,

t
 0 and (5.2) gives   J  0 [for magnetoststics] (5.3)
Assuming a magnetic force FB is experienced by charge q moving
at velocity v, we define the magnetic induction B by the relation:
FB  qv  B,
which is consistent with the definition in (5.1). 1
 5.2 Biot and Savart Law
The Biot
Biot-Savart
Savart law states that the differential magnetic field ddB
B
at point P (see figure) due to a differential current element d  2 in
0 d  2  x12 loop
p1
l
loop 2 is
i given
i b dB 
by I2 (5 4)
(5.4)
4 | x12 |3 I1
P loop 2
x12
Thus the total field at P due to I 2 in
Thus, d 1
 d  x d 2
loop 2 is: B  0 I 2  2 312 
linear superposition, 
(1)
4 | x12 |  an experimental fact  I2
Integrating the force on I1 in loop 1 due to I 2 in loop 2, we obtain
F12  I1  d 1  B (5 7)
(5.7)
0 d 1  ( d  2  x12 )
 I1I 2   d 1  x12 (d 1  d  2 )x12
4 | x12 | 3   2 
d  3
  
  | x12 | | x12 |3
 
0 d x12 1
(d 1  d  2 )x12      d ( )0
 I1I 2   x12 2 x12
4 | x12 |3 2
 5.3 Differential Equations of Magnetostatics
and Ampere’s
Ampere s Law x12  x1  x 2
B
Gauss Law of Magnetism :
x1 x d 2
0 d  2
xx12 cross section 2
Rewrite (1): B  4 I 2  of wire
|x12|3 I2
 0
Change x1 to x, x 2 to x, and let I 2 d  2  J da d  2  Jd 3 x, we obtain
0 x  x 3 0 1 3
B ( x)   J ( x )  d x      J ( x ) d x
4 | x  x |3 4 | x  x |
x x   1 B x  x
  a    a    a |x x|
|x x|3
x J ( x)
0 J (x) 1 3 x
  [     J ( x )]d x
4 | x  x  | | x  x |  
0 0
0 J (x) 3
   d x (5.16)
4 | x  x |  operates
p on x
   B  0 [Gauss law of magnetism] (5.17)3
 5.3 Differential Equations of Magnetostatics and Ampere’s Law (continued)

Ampere'ss Law : Rewrite (5.16):

Ampere
J ( x) 3
0 J (x) 3  |xx| d x
 
B ( x)    d x 0
4 | x  x |   [J ( x)   |x1x|  |x1x|   J ( x)]d 3 x
   B( x)
   J (x)  |x1x| d 3 x
0 J ( x)
      |xx| d 3 x  1   J ( x) d 3 x    J ( x) d 3 x
4  |xx|     |xx|
  
 0
  0 0
0 J ( x)
 [    |xx| d 3 x   J (x)  2 |x1x| d 3 x]
4  
  (  a)  (  a)   2a 4 ( x  x)
n
   B ( x)  0 J ( x) arbitrary (5.22)
    B  n da  0  J  n da da
     loop
 B  d  I (through the loop) d
 Ampere
Ampere'ss law , a much more elaborate 
  B  d   0 I   (5.25)
 representation of the Biot-Savart law  4
 5.4 Vector Potential
0 J ( x ) 3
Vector Potential : Rewrite (5.16): B(x)    d x
4 | x  x |
 B    A, (5 27)
(5.27)
where the vector potential A is given by
 0 J ( x ) 3 
A  d x   , (5.28)
4 | x  x |
which shows that A may be freely transformed (without changing B)
according to A  A   (gauge transformation) (5.29)
W may exploit
We l it this
thi freedom
f d by h i a  so that
b choosing th t
A  0 (Coulomb gauge) (5.31)
0
See proof on previous page.
page
  
 
  (5.28)    A  40    |Jx(xx)| d 3 x   2   2 ,
 Coulomb gauge requires  2  0 everywhere and hence   const.
5
 5.4 Vector Potential (continued)

  B  0 J
Rewrite: 
B    A
     A  0 J
 (  A)   2 A  0 J
Choose the Coulomb gauge (  A  0)
  2 A   0 J (5 31)
(5.31)
0 J ( x) 3
A  d x ((5.32))
4 | x  x |
Note:
(5.32) is valid in unbounded (infinite) space, i.e. the volume of
integration must include all currents. If there is a boundary surface,
the currents on the boundary must be accounted for by application
of boundary conditions (See example in Sec. 5.12.) 6
 5.4 Vector Potential (continued)

A Comparison of Electrostatics and Magnetostatics :

Electrostatics Magnetostatics
Definition of E:  Definition
D fi i i off B:
J
FE  qE FB  qv  B
x x x x
Coulomb's law: Biot-Savart law:
1  ( x )( x  x ) 3 0 x  x 3
E( x )   d x B( x )   J ( x )  d x
4 0 | x  x |3 4 | x  x |3

     
  E   0 E  0 B  0   B  0 J
   
 E  da  q  0  E  d   0  B  da  0  B  d   0 I
G
Gauss's ' llaw Gauss's Law Ampere's law
of electrostatics of magnetism 7
 5.6 Magnetic Field of Localized Current
Distribution, Magnetic Moment
Magnetic (Dipole) Moment : A  (B  C)  B( A  C)  C( A  B)
0 J (x) 3 1 1 x  x
A  d x    3   [Eq. [ (5),
( ) Ch.
Ch 4]
4 | x  x | | x  x | | x | |x|
 1 1
 0 [  J (x)d 3 x  3 x   xJ (x)d 3 x  ] J
4 | x |  | x |   x
x
0  1  x[ xJ ( x)]d 3 x 0
2
Proved on next page.
page
Proved on p.185 under the
0  x[xJ (x)]d 3x
 3  1. J is localized
8 |x|  within volume
0 m  x  If x is far  conditions:  of integration
  (5.55)
4 | x |3 from source. 2 J  0
2.
where m  12  x  J (x)d 3 x [magnetic (dipole) moment] (5.54)
Note
N t : In
I (5
(5.54),
54) m iis ddefined
fi d with
ith respectt to
t a point
i t off reference.
f
Here, it coincides with the origin of the coordinates (x  0). 8
 5.6 Magnetic Field of Localized Current Distribution, Magnetic Moment (continued)

Problem : Prove the relation  J ( x) d 3 x  0 under the conditions:

  J  0 and J is localized within volume of integration.
Proof : Since J  0 outside the volume of integration, we may extend
the volume of integration to  without changing the integral value.
3   
 J ( x ) d x   dx  dy  dz(J x e x  J y e y  J z e z )
Consider the x-component first:
3   
e x   J ( x) d x   dyy  dz 


J x dx
 
   J x   J x
   dy  dz  x x dx  xJ x   x dx
 x
   J J y J z
   dy  dz  x( xx  y  z )dx


   x  Jd 3 x  0 The insertion of these 2 terms will not
Similarly, the y - and change the value of the integral because
z -components vanish   ( J y )dy  J y   0 &   ( J z )dz  J z   0
components also vanish.
 y   z 
3
Thus,  J (x) d x  0 9
 5.6 Magnetic Field of Localized Current Distribution, Magnetic Moment (continued)

Anti - symmetric unit tensor ( ijk ): (used on pp.185

185 and pp.188)
188)
0 , if two or more indices are equal

 ijk  1 , if i, j , k is
i an even permutation
t ti off 1,
1 2,3
23 (2)
Levi-Civita symbol 1 , if i, j , k is an odd permutation of 1, 2,3
E
Examples l : 112  0,
0 123  1, 1 132  1,
1  312  1
( A  B)i    ijk 
j A j Bk , (  A )i    ijk
j x j
Ak
jkk jkk

  ( A  B)    ijk
ijk

xi  A j Bk 
A j
   ijk
Bk 
Bk   ijk A j
j  
 xi xi 
ijk
A j
   kij Bk
Bk 
  jik A j
ijk 
 xi xi 

 B  (  A )  A  (  B)
10
 5.6 Magnetic Field of Localized Current Distribution, Magnetic Moment (continued)

Example 1 of magnetic moment: plane loop

I
1 3 I
m  2  x  J (x)d x  2  x  d 
   d
2( area ) x
da
 m  I  (area ) (5.57)

m is normal (by right hand rule) to the plane of the loop.
loop
Example 2 of magnetic moment: a number of charged particles
i motion
in ti
angular momentum
J   qi vi (x  xi ) L  M x v
i i i i i

1 3 1 qi
  
 m  2  x  J (x )d x  2  qi xi  vi   Li (5.58)
i i 2M i

e if qi / M i  e / M for all particles.

 L (5.59)
2M
2M
L: total angular momentum
11
 5.6 Magnetic Field of Localized Current Distribution, Magnetic Moment (continued)

Dipole Field : (valid far from the source)

0 mx
Rewrite (5.55) : A  (5.55)
4 |x|3   x3  1   x  x 
3  13
|x| |x| |x|
0  x   3  x  3x  0
 B   A  m 3 
4  |x|  |x|3 |x|5

 0 (
  m is a constant.)
0
 0
 x x
 x  x
 0 [m   3  3   m   3    m   m  3]
4 |x| |x|  |x|  |x|
  x x x 
 0   mx x 3  m y y 3  mz z 3    ( A  B)  (B ) A  ( A )B
4  |x| |x| |x|   A(  B)  B(  A )

0   ex  3 x  
   mx  3  x 5    y    z  
4   |x| |x|  
0 3n(n  m)  m x  magnetic
g p 
dipole
 n  (5 56)
(5.56)
4 |x| 3 |x| 
 field 

12
 5.6 Magnetic Field of Localized Current Distribution, Magnetic Moment (continued)
As in the case of the electric dipole moment, here we characterize
a localized current distribution by a constant quantity, the magnetic
moment m, which turns an otherwise complicated field calculation
(
(see, ffor example,l Sec.
S 55.5) 5) iinto
t a simple
i l one (with
( ith limited
li it d validity.)
lidit )
Consider, for example, a circular loop carrying current I . Using
(5 57) we hhave m  I  a 2e z ((see fi
(5.57), figure).
) HHence, th
the dipo
di le
l field
fi ld is
i
0 3n(n  m)  m z
B n  e r (5.56)
4 3 e
|x| m  I  a 2e z r

0 2 3e r (e r  e z )  e z r
 I a
4 r3
a
(e z  e r cos   e sin  )
I
0 2 cos  e  sin
i  e
 I a2 r (5.41)
4 r 3

Wh r  a, the
When th dipole
di l field fi ld isi a goodd approximation
i ti off the th total
t t l
field [see Jackson (5.40).] 13
 5.7 Forces and Torque on and Energy of a Localized
Current Distribution in an External Magnetic Induction
Magnetic Force in External Field :
F   J (x)  B(x)d 3 x (5 12)
(5.12)
Expanding B : [see lecture notes, Ch. 4, Appendix A, Eq. (A.4)]
B(x)  B(0)  (x  )B(0)  

This implies
p "After differention of B(x), ) set x in results to 0," i.e.
(x )B(0)  x  x B(x)   y  y B(x)   z  z B(x) 
x 0   x 0 x 0

 F  [  J (x)d 3 x]  B(0)   J (x)   (x )B(0)  d 3 x  


 0 (p
(proved in Sec. 5.6))
  J (x)   (x )B(0)  d 3 x    (m  B)
 
U , See derivation on pp.188
pp.188-189
189 (5.69)
where U  m  B  potential energy. (5.72) 14
 5.7 Forces and Torque… (continued)

Magnetic
g Torque
q in External Field:
N   x  f (x)d 3 x 2
  [ x J (x)]
  x  [J (x)  B ( x)]d 3 x 2 x

0
 (5 13)
(5.13)
 2 2
 J ( x )    x   x     J ( x )
B ((0))  ( x  )B (0)
( )    B ((0))  2x  J (x)
  x  [J (x)  B(0)]d 3 x
 
  [B(0)  x]J (x)d 3 x  B(0)  x  J (x) d 3 x
2
=  [B(0)  x]J (x)d 3 x  12 B(0)    [ x J (x)]d 3 x

   
  

  12  B(0)  [x  J (x)]d 3 x 2

  s x J ( x)da  0
[Using the formula at the
J is localized.
bottom of p.185, replacing  J  0 on surface
x with B(0).]
 m  B(0) (5.71)15
 5.7 Forces and Torque… (continued)

A Comparison between Electric and Magnetic Potential Energy,

Force, and Torque in External Field :
Potential energy Force Torque
U   p  E (4.24) F  U N  pE
 p + E
U   m  B (5.72)
(5 72) F  
U N  mB
Both p and m tend to orient along the
positive field direction under the action of
the torque (see figures on the right). This I 
results in a state of minimum potential m B
energy. In this state, p reduces E, whereas I
m enhances B.
Questions:
(1) How does a permanent magnet attract a piece of iron?
(2) How does it attract another permanent magnet?
16
 5.7 Forces and Torque… (continued)
Force in Self -Consistent Field;; Magnetic
g Pressure and Tension :
A self-consistent field is the combined field
generated by the source J under consideration
and
d the
h externall source (if present).
) Thus,
Th using i
  B  0 J , we may express the magnetic force
density f (force / unit volume) in terms of B.
B-field
f  J  B  1 (  B)  B a solenoid
lines
0
a uniform
if
( a  b )  ( a   ) b + ( b   )a + a  (   b ) + b  (   a ) electron beam
2
  2   1 (B )B [[see pp. 320)]
B )] ((3))

 0  0 

magnetic tension force density,

magnetic pressure
as if a curved B-field line tended
force density B-field
to become a straight line. B lines
I regions
In i h J  0,
where 0 we have
h f 0
[pressure and tension force densities cancel out]. uniform current 17
 5.8 Macroscopic Equations, Boundary
Conditions on B and H
Macroscopic Equations : To be more general, we move the
point of reference for m from x  0 to x  x0 and write
p
1 1 ( x  x 0 )  ( x  x 0 )
    [See Sec. 4.1]
x  x x  x0 x  x0
3

Sub. this relation into How about Jfree? J

x
x
0 J ( x) 3 0
A  d x [(5.32)]
4 | x  x |
we obtain vanishes
i h only l if J is
i localized
l li d J x  x 0
within the volume of integration.
 x  x0

0  J (x)d 3 x 0 m (x0 )  (x  x0 )
A   , x x (4)
4 | x  x0 | 4 | x  x0 |3 x 0
where m(x0 )  12  ( x  x0 )  J ( x) d 3 x.
0 18
 5.8 Macroscopic Equations, Boundary Conditions on B and H (continued)

To proceed, we consider the orbital motion of atomic/molecular

electrons, which can collectively give rise to a permanent or induced
magnetization M (total magnetic moment / unit volume) given by
M ( x)   N i  m i  (5.76)
i
volume density of magnetic moment per type i molecule
type i molecules averaged over a small volume
As will be shown in (5.79), a current density (J M ) is associated
with M. In addition, there is also a current density due to the flow of
free charges, which we denote by J ffree (Jackson denotes it by J in
Sec. 5.8). By the principle of linear superposition, we may write
A (x)  A free (x)  A M (x)),
where A free and A M are due to J free and J M , respectively.
f ( x) 3
0 J free
Obviously, A free (x)   d x
4 | x  x | 19
 5.8 Macroscopic Equations, Boundary Conditions on B and H (continued)

For A M , we have the expression

p for M, but not yyet for J M . So
we approximate A M by the dipole term in (4).
3
0  M
J ( x) d x  0 m ( x 0 )  ( x  x 0 )
A M ( x)    ,
4 | x  x0 | 4 | x  x0 | 3

h we hhave sett  J M (x)d 3 x  0 because

where b J M isi formed
f d off currentt
loops of atomic dimensions (  volume of integration). Under this
condition m is independent of the point of reference because
condition,
m(x0 )  12  (x  x0 )  J M (x)d 3 x
 0
  (5.54)
 12  x  J M (x)d 3 x  12x0   J M (x)d 3 x  m(0).
term, we must have x  the
To represent A M by the dipole term
dimension of the dipole. So, we divide the source into infinistesimal
volumes. In each small volume V , the dipole
p moment is MV ,
which generates a small A M at x given by
20
 5.8 Macroscopic Equations, Boundary Conditions on B and H (continued)
0 M (x)( xx) V
A M ( x )  V
4 |xx|3
where we have replaced
p the notation x0 with x. This x
x
0 M (x)( xx) 3
gives A M ( x)   3 d x
4 |xx|
0 Volume of integration
  M (x)  |x1x| d 3 x includes all sources.
4
0 M ( x) 3 0 ( x) 3
  |xx| d x     M |x  x 
|
d x 
4 4  
   a    a     a ( x)
  s n  M
|xx|
da  0
3
v   Ad x   s n  Ada (M  0 on S )
0 M ( x) 3
  |xx| d x What if M ≠ 0 on S?
4
Question: Does this relation still hold as x  x ? 21
Griffith
1/4 6.2 The Field of a Magnetized Object
621B
6.2.1 Bound dC
Currents
t
Suppose we have a piece of magnetized
material (i.e. M is given). What field does
this object produce?

The vector potential of a single dipole m is

0 m  rr̂
A(r ) 
4 r 2
IIn the
th magnetized
ti d object,
bj t each h volume
l element
l t carries
i a
dipole moment Md’, so the total vector potential is

0 M (r)  rˆ
A(r ) 
4  r 2
d 

22
Griffith 2/4
Vector potential and Bound Currents
Can the equation be expressed in a more illuminating form,
as in the electrical case? Yes!
By exploiting the identity,    1
(xˆ   yˆ   zˆ  )
1 rˆ x y z  ( x  x) 2  ( y  y) 2  ( z  z) 2
  2 xˆ ( x  x)  yˆ ( y  y)  zˆ ( z  z ) rˆ
r r  
(( x  x) 2  ( y  y) 2  ( z  z ) 2 )3/ 2 r 2
0 1
The vector potential is A(r ) 
4  M (r)  ( )d 
r
U i th
Using the product l   ( fA)  f  A  f (  A)
d t rule
and integrating by part, we have
  1  M (r) 
A(r )  0  [  M (r )]d       [ ]d  
4  r r 
How? See next page.
  1   0 1
 0  [  M (r )]d    [M (r)  nˆ ]da
4  r  4 r 23
Griffith 3/4

Gauss's law  (  E)d   E  da

v S

 (  ( v  c))d  c  (  v )d
 v v
Let E  v  c, 
  ( v  c)  da  c   v  da
S S

Since c is a constant vector, so  (  v )d    v  da

24
v S
Griffith
4/4 Vector potential and Bound Currents
0 1 0 1
A(r )  
4 r
[  M (r)]d   
4 r
[M (r)  nˆ ]da

J b    M (r) K b  M (r)  nˆ 
volume current surface current

With these definitions

definitions, bound currents
0 Jb 0 Kb
A(r ) 
4 
v r

d 
4  S r
da

The electrical analogy

volume charge density  b    P
surface charge density σ b  P  nˆ
25
 5.8 Macroscopic Equations, Boundary Conditions on B and H (continued)

Th
Thus, A (x)  A free (x)  A M (x)
0 J free (x)    M (x) 3
  d x (5 78)
(5.78)
4 | x  x |
p
For comparison, in Sec. 5.3, we have  2 A   0 J , ((5.31))
0 J (x) 3
which has the solution: A( x)   d x (5.32)
4 | x  x |
In (5.31) and (5.32), J represents the current due to both free
and bound (atomic) electrons, whereas in (5.78) contributions from
free and bound electrons are separated into two terms.
p
Comparing g (5.78)
( ) and (5.32),
( ), we find that the bound electrons
contribute to A (x) through a magnetization current density (J M )
given by J M    M, (5.79)
which is the macroscopic exhibition of the atomic currents.
26
 5.8 Macroscopic Equations, Boundary Conditions on B and H (continued)

Hence, by g free and bound electrons,   B( x)  0 J (x)

y separating
p
[(5.22)] can be written   B  0 ( J free    M ) (5.80)
Defining a new quantity called the magnetic field H :
 Effects of the atomic 
H  1 B  M ,  currents are implicit
p in H. (5.81)
0 
we obtain from (5.80) the macroscopic version of (5.22):
  H  J free
f (5 82)
(5.82)
Question: Does H have a physical meaning?
Diamagnetic Paramagnetic
Diamagnetic, Paramagnetic, and Ferromagnetic Substances :
The counterpart of (5.81) in electrostatics is D   0E  P [(4.34)].
In Sec.
Sec 4.3,
4 3 it is shown that,
that for small displacement of the bound
electrons, we have the linear relations:
P   0  e E (4 36)
(4.36)

D   E, with    0 (1   e ) (4.37), (4.38) 27
 5.8 Macroscopic Equations, Boundary Conditions on B and H (continued)

However, the magnetic properties of materials are such that M is

not always proportional to B, depending on the type of the material.
We summarize, without derivation, possible relations between B and H.
1. For diamagnetic and paramagnetic substances, M is propor-
tional to B and we express the linear relation as
  0    0  M  B, paramagnetic 
M B   (5)
0    0  M  B, diamagnetic 
Substituting M into H  1 B  M, we get the linear relation:
0
B   H, (5.84)
h  is
where i called
ll d the
th magneticti permeability.
bilit
Question : The plasma is diamagnetic. Why?
B
2 For
2. F the h fferromagnetici substance
b we
have a nonlinear relation (see figure): H
B  F (H ),
) (5 85)
(5.85)
which exhibits the hysteresis phenomenon shown in the figure. 28
 5.8 Macroscopic Equations, Boundary Conditions on B and H (continued)
Boundary Conditions : n : unit normal pointing
(i)   B  0  v   Bd 3 x   s B  da  0 from region 1 into region 2
 (B 2  B1 )  n  B1  B2 B2 ((5.86))
n
(ii)   H  J free region 2 pillbox of
infinitesimal
    H  da   J free  da region 1 B thickness
hi k
1
( LHS )   H  d  (see lower figure) n
d
d
 (H 2  H1 )  (n  n)L K free rectangular
 n  [n  (H 2  H1 )]L n loop of
infinitesimal
a  (b  c )  b  (c  a ) d height
 L  

( RHS )   J ffree  nda  K free
f  nL
K free : surface
f currentt off
 n  ( H 2  H1 )  K free free charges (unit: A/m) (5.87)
Special case: K free  0  H t 2  H t1 (6)
t: tangential to surface 29
 5.8 Macroscopic Equations, Boundary Conditions on B and H (continued)

Application to a Solenoid :
l
i i n turns permeable
unit length material

 H in  Bin /   Bout

Approximate the manetic field by that of

an infinite solenoid.
solenoid So,
So H in = constant.
constant

 H  dl  I free  H inl  nil

 H in  ni  Bin   H in   ni
 Bout  Bin   ni
Question : " Bout   ni " implies that
filling the solenoid core with   0
material (while keeping i at a constant
B-field lines value) can greatly enhance Bout . Why?30
 5.9 Methods of Solving Boundary-Value
Problems
ob e s in Magnetostatics
g e os cs
We put the basic equations :   B  0 and   H  J free (5.90)
in forms suitable for 2 types of boundary - value problems.
Type 1 : Linear medium with   const (in each region).
(a) Equation for vector potential (with or without J free )
B   H    A  H  1   A
   H  1     A  1 [(  A )   2 A ]  J free
  2 A    J free  use Coulomb gauge,   A  0 (7)
(b) Equation for scalar potential (only for J free  0)
  B  0    H  0 and   H  0  H   M (5.93)
  2 M  0 (8)
Typically, we use (7) or (8) to solve for A or  M in each uniform
region and find the coefficients by applying conditions (5
(5.86)
86) and
(5.87) on the boundary. An example will be provided in Sec. 5.12. 31
 5.9 Methods of Solving Boundary-Value Problems in Magnetostatis (continued)

Discussion:
In a vacuum medium, we have
 2 A    0J free I  (5.31)
m B
In a uniform- medium, we have I
 2 A    J free . (7)
Hence, the effect of    0 medium is to increase the ability of
J free to produce B by a factor of  /  0 (see figure above).
In electrostatics, we have
2  free  p + E
   (vacuum medium) (1.13)
0
 free
and     f
2
(uniform dielectric medium) (4.39)
Hence, an    0 medium reduces the ability of  ffree to produce
E by a factor of  /  0 (see figure above).
32
 5.9 Methods of Solving Boundary-Value Problems in Magnetostatis (continued)

Type 2 : Hard ferromagnets (permanent magnet, M given, J free = 0)

(a) Vector potential
  H    ( B  M )  0 real current
0
   B  0  M  0 J M , where J M    M [see (5.79)]
B    A    B      A  (  A)   2 A  0 J M
 J (x ) 
  2 A   0 J M  A(x)  40  |xMx| d 3 x, ((5.102))
(b) Scalar potential  M is a mathematical
  H  0  H   M tool, not real charge.
  B  0  (H  M )  0   2 M    H    M (5.95)
where  M  
  M (effective magnetic charge density) (5
(5.96)
96)
 ( x)  
  M  41  |xMx| d 3 x   41  |xM (x ) 3

x| d x   a    a + a 

 41  M (x)  |x1x| d 3 x   41    |M (x ) 3
xx| d x
 (5.98)
33
 5.9 Methods of Solving Boundary-Value Problems in Magnetostatis (continued)
Effective
ff magnetic
g surface
f chargeg densityy  M :
Rewrite (5.96):   M    M (5.96)
 v   Md 3 x   s M  da   v  M d 3 x ((see ppillbox below))
0 M
  n surface area  A
 (M 2  M1 )  nA   M A M2  0
 M  nM M (5.99)
M1  M
a mathematical tool thickness  0
Surface current density K M due to magnetization M:
In Sec. 5.8, Here  by the same algebra  ,
  H  J free   M  JM
real current real current 0 M
 
 K free  n  (H 2  H1 )  KM  n  (M 2  M1 )  M  n
n H2 n M2  0
K free KM
H1 M1  M 34
 5.10 Uniformly Magnetized Sphere
Consider a ppermanent magnet
g with magnetization
g : discontinuous!
M
0
M 0e z , r  a
, ra
r P

B H

 M vanishes everywhere M a
z
except on the surface.
1 M ( x) 3 1 nM ( x)

 M  4  |xx| d x  4  s |xx | da

d 
by (5.95) 4
(3.70)

= 3 Y 10( ,  )  1 1  4 r 
  r
M 0a 2    1 M a 2 r cos 
|x  x |  3 r 2

 4  d  |cos
xx| 3 0 r2 [Y1,1 ( ,  )Y1,1 ( ,  )
 13 M 0 r cos   13 M 0 z , r  a Y10 ( ,  ) Y10 ( ,  )
 1 3 cos
  (5.104)
 3 M 0 a r 2 , ra  43 cos

Inside: H in   1 M Y11 ( ,  )Y11 ( ,  )]  

3

 Bin  0 H in  0M  32 0M ( H in  Bin )
Outside: dipole field with dipole moment m  4 a3 M.
 3 35
 5.12 Magnetic Shielding, Spherical Shell of Permeable
Material in a Uniform Field
Consider a spherical  -shell in an external B0 . B0 
im  r
 r l   m
P (cos  ) e   a
 2 M  0   M   l 1   l m     b
  r  Ql (cos  )  e  im
Eq. (8)  
 H r cos     
1 Pl (cos  ) , r  b (5.117)
l
 0
l 0 r l   H 0 r cos 
  l gives the
 M    (  l   l r l 1 ) Pl ((cos  ), a  r  b
r 1 ((5.118))
l 0 external B0 .

   l r l Pl (cos  ), r  a   (5.119)
l
l 0  M    M
  
boundary conditions b b
H   M (5.93)    M  M

  H t 2  H t1 (6)    a   a 
B  0 H (outside)  +      
B   H (inside)   B1  B2 (5.86)   0 rM    rM 
 b b
The shell is assumed to    M  M
be a linear medium.  
 r a  0 r
a  36
 5.12 Magnetic Shielding, Spherical Shell of Permeable Material in a Uniform Field (continued)
Boundary conditions result in solutions for the coefficients:
 l  l   l   l  0 if l  1
 (2  1)(  1)(b3 a3 ) 
  0 0 H  b H03

 1 (2  1)(   2)2 a3 (  1)2 0   (5.121) 
  0 0 b3 0   0  
   & 
9  9 0
1  0 H  H  (5.122) 
  a 3  0 3 0
 (2  1)(   2)2 3 (  1) 2 a
2  (1  3 )
 0 0 b 0 b 
Bin  as  , implying that   0 B  B + dipole field
0 out 0
materials tend to "absorb"
absorb B-field
field lines and
thereby provide the shielding effect. High-
materials can have  / 0 as highg as 103  106.
When   0 , B reduces to B0 everywhere,
i.e. a static megnetic field penetrates into the Bin: uniform field
shell as if there were no shell present (even if the (  1/  if   0 )
shell is made of good conductor, such as copper). 37
 5.15 Faraday’s Law of Induction
The Biot
Biot-Savart
Savart (or Ampere's)
Ampere s) law relates the magnetic field to
electrical current. Faraday then discovered experimantally that
time-varying magnetic flux through an electrical circuit could
induce an electric field around the circuit. This not only provided
the first link between electric and magnetic fields, but also led to a
new mechanismh i to t generate t the
th E-field,
fi ld i.e.
i a time-varying
ti i B-field.
fi ld
Referring to the figure, let loop C loop C
be an electrical circuit (as in Faraday
Faraday'ss B( x , t )
original experiment) or any closed path d n
p
in space ((a ggeneralization of the original
g da
observation with immense consequences).
Faraday's law states
 B  S : an arbitrary surface 
 c E  d    s t  nda,  bounded by loop C  (5.141)
 
h E iis the
where h electric fi ld at d  in
l i field i the
h frame
f hi h d 
iin which  is
i at
rest, and B is the magnetic induction in the lab frame. 38
 5.15 Faraday’s Law of Induction (continued)

loop C
R i (5.141):
Rewrite (5 141)
B(x, t )
 c E  d    s Bt  nda d n (5.141)
da
Assume loop C is at rest in the lab
frame, then E  E (electric field in the
lab frame) and (5.141) becomes

 c E  d    s Bt  nda integral form of Faraday's law  ((9))

where both E and B are lab-frame quantities.
(9) can be
b written
itt (by
(b Stokes's theorem:  c E  d   s   E  nda
St k ' th d )

s   E  nda   s Bt  nda

Thus,
  E   B  differential form of Faraday's y law))  ((5.143))
t
39
 5.16 Energy in the Magnetic Field
To find the energy associated with a magnetic field,
field we evaluate
the work needed to establish the current J (x), which produces the
magnetic field. We break up J (x) into a network of thin loops. In
the build-up process, an E field will be induced by B / t. The
rate of work done by E within each loop is
 is the cross section of the loop (same as integration over the area
Jackson's  ). J &  may vary along d . encircled by the loop
dWloop 
n a thin
hi
   J  E  d    s J  
  E  n da
dt B / t loop
 s J  n  B da
t St k ' th
Stokes's thm. S
Work done within each loop to generate  B :
 d  (d   J )
 Wloop  s J  n  B da  s J    A  nda  : cross section of loop
  A
  J  A  d  d   Jd 3 x
J  d   J 
 loop  A  Jd 3 x d 3x (10) 40
 5.16 Energy in the Magnetic Field (continued)

As shown in  Wloopp  loopp  A  Jd 3 x [(10)], the work done within

each loop is an integral over the volume of the loop. Thus, an inte-
gration over all space gives the total work done to generate  B :
Assume the rate of build-up  0  H obeys the static law   H  J.
Otherwise, the static law breaks down and there will be radiation loss.
 W    A  Jd 3 x    A  (  H )d 3 x (5 144)
(5.144)
  H  (   A ) d 3 x     (H   A ) d 3 x For this integral
   
B   s ( H A )da0 to vanish,
vanish the
volume of integ-
  H   Bd 3 x  12   (H  B)d 3 x ration must be .

Assume linear medium: B   H or B  μ  H
Total work done to bring the field up from 0 to the final value B :
W  12  (H  B)d 3 x  By B conservationti off energy, this i 
thi is
(5.148)
 the total magnetic field energy. 
 w  12 H  B [[field energy gy density] y] ((11))
Note: w  12 H  B  12 (  H j )  (  B j )  12  ( H j  B j )
41
j j j
 5.17 Energy and Self- and Mutual Inductances
Assume linear relation between J and A
for nonpermeable
 W    A  Jd x  12   ( A  J )d x
3 3
(μ0) medium (5.144)

1 0 J (x) 3 
3 
4  |xx|
 W  2  A  Jd x A ( x ) d x (5.32) (5.149)
 J (x)J (x)
 0  d 3 x  d 3 x f N current-
for t (5.153)
8 |xx| carrying circuits
0 N 3 N 3  J (xi )J (xj ) 1 N 2 N N
  d xi   d x j |x x |  2  Li I i    M ij I i I j , (5.152)
8 i
1 j 1 i j i 1 i 1 j i

where self-inductance for a thin wire

0 3 3 J ( xi )J ( xi )  0 d  i d i 
Li  
4 Ii2 C i
d xi C d xi
|xi xi |  4  C i  C i |xi xi | 
(5.154)
i

0 3 3 J (xi )J (xj )  0 d  i d j 

M ijj   d xi C d xj 
|xi xj |  4 C i C j |xi xj | 
(5.155)
4 Ii I j i
C j

mutual inductance (Mij = Mji) for thin wires 42

 5.17 Energy and Self- and Mutual Inductances (continued)

0 J (x) 3 
A ( x) 
4  |xx|
d x (5 32)
(5.32)
 Vector potential at circuit i due to current in circuit j :
0 J (xj ) 3
A ij (xi )  4  C j |x x | d xj (12)
i j
From (12) and (5.155), we obtain M ij  I 1I C Aij (xi )  J (xi )d 3 xi
i j i
Assume J flows along wire d  of negligible cross section da
 J (xi )d 3 x i  J  dad   I i d  magnetic flux from circuit j
B passing through circuit i
ij
 
 M ij  I1  C Aij  d   I1  s (  Aij )  nda  I1 Fij (5.156)
j i j i j
d d ij: induced voltage in circuit i due
  ij   Fij   M ij I j to current variation in circuit j.
dt dt
Th “–”
The “ ” sign
i implies
i li that
th t the i d d ij tends
th induced t d tot drive
d i a currentt ini
circuit i to inhibit the flux change caused by circuit j (Lenz’s law). 43
 Homework of Chap. 5

Problems: 1,, 3,, 6,, 11,, 13,,

18, 19, 20, 22, 30

44