清大物理系 “電動力學(一)” 任課老師:張存續

“Electrodynamics (I)” (PHYS 531000 )

Fall Semester, 2010

Department of Physics, National Tsing Hua University, Taiwan

Tel. 42978, E-mail: thschang@phys.nthu.edu.tw

Offi hour:
Office h 3:30-4:30
3 30 4 30 pm @Rm.
@R 417

助教
趙賢文: 0933580065, s9822817@m98.nthu.edu.tw
沈育佃 0933346965, s9822538@m98.nthu.edu.tw
沈育佃:
1
1.Textbook and Contents of the Course:
J. D. Jackson, “Classical Electrodynamics”, 3rd edition, Chapters
1-7. Other books will be referenced in the lecture notes when
needed.
2. Conduct of Class :
L t
Lecture notes
t will
ill be
b projected
j t d sequentially
ti ll on the
th screen
during the class. Physical concepts will be emphasized, while
algebraic details in the lecture notes will often be skipped.
skipped
Questions are encouraged. It is assumed that students have at
least ggone through g the algebra
g in the lecture notes before
attending classes (important!).
3. Grading Policy:
First midterm (~30%); Second midterm (~30%); Final
(~30%); Quiz (~10%); Participation (~5%, extra). The overall
score will
ill be
b normalized
li d to reflect
fl an average consistent
i with
ih
other courses. 2
4. Lecture Notes:
Starting from basic equations, the lecture notes follow Jackson
closely with algebraic details filled in.
Equations numbered in the format of (1.1), (1.2)… refer to
Jackson. Supplementary equations derived in lecture notes, which
will
ill later
l t beb referenced,
f d are numbered
b d (1),
(1) (2)...
(2) [restarting
[ t ti from
f (1)
in each chapter.] Equations in Appendices A, B…of each chapter
are numbered (A.1),
(A 1) (A.2)…and
(A 2) and (B.1),
(B 1) (B.2)…
(B 2)
Page numbers cited in the text (e.g. p. 120) refer to Jackson.
Section numbers (e.g.
(e g Sec.
Sec 1.1)
1 1) refer to Jackson.
Jackson Main topics
within each section are highlighted by boldfaced characters. Some
words are typed
yp in italicized characters for attention. Technical
terms which are introduced for the first time are underlined.

3
 Chapter 1: Introduction to Electrostatics
1 8 Green’s Theorem
1.8
Green's theorem, a powerful tool for treating electrostatic boundary-
value p
problems, is a simple p application
pp of the divergence
g theorem:
3
v   A d x   s A  n da  : theta; : phi;  : psi
Let A =  , where  and  are arbitraryy functions of pposition.
  A        2    S
  da n
 A  n     n   n v
Sub. these 2 expressions for   A and A  n into the
g
divergence theorem,, we obtain Green's first identity, y,
2 3 
v (        ) d x  
 s n da (1.34)
Interchange  and  in (1.34). (1 34)

 v ( 2    )d 3 x   s da
n
Subtract these two equations
equations, we obtain Green Green'ss second identity,
identity
2 2 3  
v (       ) d x  (
 s n   n
)da (1.35) 4
 1.8 Green’s Theorem (continued)
Green's theorem relates a volume integral g to a surface integral
g and
the volume integral contains the operator  2 . These features are useful
for the manipulateion of the Poisson equation in bounded space.
F example,
For l applyingl i Green's
G ' secondd identity:
id i
2 2 3  
v (       ) d x  (
 s n   n
)da (1.35)
we may convert the Poisson equation into an integral equation. How? H ? SeeS next
few pages.
In (1.35), letting  be 1 , ( x1x  4 (x  x)),  be the
2
|xx|

electrostatic potential (thus,  2    ), and x be the integration variable,
0
we obtain
[       1  ( x)]d 3 x  [  ( 1 )  1  ]da
v 4 ( x x )  0 x x  s n xx xx n
x inside v

  (x)  4 1

 ( x) 3  1
0 v x x
d x   1    
4  s  xx n   1  da (1.36)
n xx 
((1.36)) is an integral
g equation
q ((not a solution)) for . In infinite space,
p ,
1 . Hence, (1.36) reduces to  (x)  1  (x) 3 
we have   R 4 0 v |xx|
d x.
5
 Delta Functions (pp. 26 - 27)

Definition of delta function:

 ( x  a)

 ( x  a )  0 , if x  a
 a2
 a1  ( x  a )dx  1 , if a1  a  a2 x
a1 a a 2
Note: Since the delta function is defined in terms of an integral,
it takes an integration to bring out its full meaning
meaning.
Properties of delta function:
(i) a 2 f ( x) ( x  a )dx  f (a )
a
(2)
1
 0 
(ii) a f ( x) ( x  a ) dx  f ( x) ( x  a ) |a  a f ( x) ( x  a )dx

a2 a2 a2
1 1 1

  f (a ) (3)
6
 Delta Functions (continued)

(iii) Let x  a be the root of f ( x)  0, then f ( x)

f ( a2 )  f ( a1 )
f ( a2 )
  f ( a1 )  [ f ( x)] d f1( x) df ( x)
a2
a1 [ f ( x )]dx  a1 a a2
dx
x
  f ( a2 ) 1  ( f )df  1  1 , f (a )  0 f ( x ) f (a )  f (a )
 f ( a1 ) f  f ( a ) f ( a ) 1 2

  ff ((aa1 )) 1  ( f )df   1  1 , f (a)  0 a1 a a 2 x
 2 f f ( a ) f ( a )
Note: In both expressions above, the integration is from a samller
value to a larger value, as in the definition of the delta function.
Compare with (2)   [ f ( x)]  1  ( x  a ) [  1  ( x  a)] (4)
f ( a ) f ( x )
If f ( x) has multiple roots xi [f ( xi )  0,
0 i  1, 2 ],
1 2, ] then
 [ f ( x)]   1  ( x  x ) [   1  ( x  x )] (5)
i f ( xi ) i f ( x )
i i

Exercise: Show  (a  x)   ( x  a ) and  (cx)   ( x)/ | c | 7

 Delta Functions (continued)
Extension to 3 dimensions :
1 C
1. Cartesian
i coordinates:
di x = (x1 , x2 , x3 )
 (x  x)   ( x1  x1 ) ( x2  x2 ) ( x3  x3 ) (6)
 V  (x  x)d 3 x    ( x1  x1 )dx1   ( x2  x2 )dx2   ( x3  x3 ) dx3

 
0, if x lies outside V
1, if x lies inside V
x3 x

2. Cylindrical coordinates: x = ( ,  , z )
x2
 (x  x)  1  (    ) (   ) ( z  z ) (7)
x1
 V  (x  x)d 3 x  V  (x  x)  d  d dz z x
   (    )d    (   )d   ( z  z )dz

 
00, if x lies
li outside
t id V
1, if x lies inside V  
Q i : If x and
Question d x both
b th ha
h ve the
th di
dimension
i off cm, what
h t are the
th
dimensions of  (x) and  (x)? [See Appendix (A), Eq. (A.9).] 8
 Delta Functions (continued)
r
3. Spherical coordinates: r = (r ,  ,  )

 2 1  (r  r ) (   ) (   ), or r
 r sin

 (r  r )   (8)
1
 r 2  (r  r ) (cos   cos  ) (   ) 
By (4),  (cos   cos  )  1  (   )  1  (   ), 0    
i 
sin i 
sin

 3  (r r) 2
V  (r  r ) d x  V  ((cos   cos  ) (   ) r
drd ((cos  )d

r2
0, if r lies outside V d 3
x
 [
[see (9)
( ) below]]
1 if r lies inside V
1,
Note: Volume integration in spherical coordinates
  2  2  2 Variables are to
0 0
dr rd  0 r sin  d   0 r dr  sin  d   d 
0   0 be integrated
 1 2 from smaller to
 0 r 2 dr 1 d (cos  ) 0 d  11 d (cos  ) larger values.
values
 d 3 x  r 2 sin  drd d or r 2 drd (cos  ) d (9) 9
 Delta Functions (continued)

Approximate representations of the delta function :

The delta function,  ( x), can be represented analytically by the
following functions because they satisfy the definition of the delta
function in the limit   0 ( > 0).
1 
 ( x)  lim
 0  x 2   2
2
 x2
1
 ( x)  lim e 2
 0 2
 1 , for    x  
 ( x)  lim  2 2
 0 0, otherwise


10
 Delta Functions (continued)

Problem 1: A total charge Q is uniformly distributed around

a circular ring of radius a and infinitesimal thickness. Write the
g densityy  (x) in cylindrical
charge y coordinates.
Solution: Is there any θ-dependence?

Let  (x) = K  (r  a ) (z ) and find K as follows.

follows
3
  ( x ) d x  K   (r  a ) ( z )rdrd dz z
 2 Ka
K Q
Q
K
2 a a
Q
  ( x)   (r  a) ( z )
2 a
Note:  has the dimension of "charge/volume" as expected.

11
 Delta Functions (continued)

Problem 2: Prove  2 1r  4 (r ) ( 2 x 1x ?)

 (r )  0, if r  0 r
Solution: Definition of  (r ) :  3
  (r ) d x 1 
r
Hence, we need to prove
(i)  2 1  0,
r 0 if r  0
(ii)   2 r1 d 3 x  4   (r )d 3 x  4 
It is convenient to use the spherical coordinates. To prove (i), we
we write  2 as (see back cover of Jackson)
2 1  2  1   1  2
  2 (r ) 2 (sin  )  2 2
r r r r sin   r sin   2
21 1 d 2 d 1 1 d r 2
  r  2 (r )   2 ( 2 )  0 if r  0
r dr dr r r dr r
r 2
Note: 2 is undetermined at r  0.
0 However,
However here we are only
r
concerned with the region r  0. 12
 Delta Functions (continued)

(ii) we integrate  2 1r over a spherical volume V

To prove (ii),
21 3 1 d 3 x  e   1 da   r 2 1 d   4

v r d x  v   r  s r r  s r 2
1
divergence thm.  r 2 er r 2 d
Note: Since r > 0 on the spherical surface
surface, again we do not have
the problem of evaluating r 2 / r 2 at r = 0.
r
Change to a coordinate system in which r  x  x x
21 x
and r  x  x . We obatin from  r  4 (r )
 2 1  4 (x  x) (1 31)
(1.31)
|xx|
2  ( x) 1  (x) 3 
optional
p Problem e ve   (x)    from
oblem 3: Derive o  (x)  4  d x
0 0 |x  x|
1  (x) 2 1 d 3 x
Solution:  2 (x)  4
0
 |xx|
 41  (x)  4 (x  x)  d 3 x    ( x)
0
  0
13
 1.9 Uniqueness of Solution with Dirichlet or
Neumann Boundary Conditions
Dirichlet boundary condition:  s specified
   specified
 Neumann boundary condition: n s
Cauchy boundary condition:  and   both specified
 s n s
As another application of Green's theorem, we use it to prove the
uniqueness theorem for the solution of the Poisson equation.
Let there be two solutions, 1 and  2 , which both satisfy
2    n on S (Dirichlet b.c.), or
      0 with  

 n    n on S (Neumann b.c.) V
n

1,2 
 2 1     0 1 2   n on S , or
i.e.  2 with  
      n  1,2   
n on S S
2 0
2 U   n   n  0 on S , or
Define U  1   2 , then  U  0 with  
 
 n U   n   n  0 on S14
 1.9 Uniqueness of Solution… (continued)

  
Green s 1st identity: v  2    d 3 x   s  n da
Rewrite Green's
Let     U b.c. U  0 or U / n  0 on S
 v (U  2
   3
 U da  0  U d 3 x  0 2
 U U U ) d x 
 U
s n v
0
n
 U  0 everywhere within V V

0, if U  0 on S
 U  1   2  
const , if U / n  0 on S
S

 1 and  2 differ by at most a constant, hence are the same solution.

Note: Since the solution is uniquely determined by specifying either
 or  / n on the boundary, the Cauchy boundary condition
( and  / n both specified on the boundary) is an over-
specification, which may lead to inconsistency.
E
Exercise : Prove
P that
h there
h cannot bbe any static
i E inside
i id a closed,
l d
hollow conductor if there is no charge in the hollow region. 15
1.10 Formal Solution of Electrostatic Boundary-Value
Problem with Green Function
Green Function GD ( x, x ) : GD  0 for
In electrostatics,
electrostatics the Green function is x on S  x
x
the solution of the following problem: S
0
2GD ( x, x )  4 ( x  x) with GD ( x, x )  0 for x on S ,
where x is the variable of the differential equation and x  is treated as
constant GD ( x, x ) is the potential of a unit point source (q  4 0 )
a constant.
located at x  subject to the b. c. that GD ( x, x ) vanishes for x on S .
GD  0 for
Symmetry Property of GD ( x, x) : y on S  x
x
Consider two equations: one with a S
0
point source at x, the other with a point source at x . The variable is y.
2y GD ( y, x )  4 ( y  x ), b.c. GD ( y , x )  0 for y on S
2y GD ( y, x)  4 ( y  x), b.c. GD ( y , x )  0 for y on S 16
 1.10 Formal Solution of Electrostatic Boundary-Value Problem…(continued)

Rewrite: v   2y  2y  



3  
d y   s  n  n da

(1.35)
Let   GD (y , x) and   GD (y , x),
) where y is the variable.
4 ( y x) 4 ( y x )
 
 v [GD (y , x)  2y GD (y , x)  GD (y , x'))  2y GD (y , x)]d 3 y
  s [GD (y , x)  GD (y , x)  GD (y , x)  GD (y , x)]da
  n   n
 0 on S  0 on S
 4 [GD (x, x)  GD (x, x)]  0
 GD (x, x)  GD (x, x) [symmetry
[ property off GD (x, x)]
Questions:
1. Does  2GD (x, x)  4 (x  x) imply 2GD (x, x)  4 (x  x)?
2. Give two examples to show the physical meaning of the symmetry
property of GD (x, x).
17
 1.10 Formal Solution of Electrostatic Boundary-Value Problem…(continued)

Formal Solution of Electrostatic Boundary - Value Problem :

1  (x) 3 
The expression  (x)  4  d x is applicable only to
0 |x  x |
unbounded
b d d space. By B Green's
G ' th
theorem, we may generalize li it to
t an
expression for bounded space with prescribed boundary conditions.
Consider a general electrostatic boundary-value boundary value problem:
 2 (x)    (x) /  0 with  (x)   s (x) for x on S (10)
Green's 2nd identit
Green's identity:: s
   2
   2
  3
 
v  ( x )  ( x )   ( x )  ( x )  d x S

  s  (x) n  (x)  (x) n  (x)  da (1.35)

 
In (1 35) let  (x) be the solution
(1.35), sol tion of (10) with ith vaariable x (i e Φ(x)).
(i.e. ))
Let  (x)  G D (x, x), where G D (x, x) is the Green function satisfying
2GD (x, x)  4 (x  x) with GD (x, x)  0 for x on S (11)
Substitution of  (x) and  (x) into (1.35) gives 18
 1.10 Formal Solution of Electrostatic Boundary-Value Problem…(continued)
4 ( xx)   ( x) 
    0
2 2 3
v [  ( x)   GD ( x, x)  GD ( x, x )    ( x)]d x  s
  s [(x) n GD (x, x)  GD (x, x) n (x)]da  S
    
 0 on S
Thus,, we obtain
1 3 1 GD ( x,x)
 (x)  4 v  (x)GD (x, x)d x  
  ( x) da  (1.44)
0 4 s n
(1.44) expresses the solution  of the general electrostatic problem
in (10) in terms of the solution GD ( x, x) of the point source problem
i (11)
in ( ) andd the
h boundary
b d valuel ( s ) off  on S . To evaluate
l (1.44),
( ) we
first solve (11) for GD (x, x), then substitute GD (x, x),  (x),  s into
(1 44) It is
(1.44). i often
ft simpler
i l to l GD (x, x) from
t solve f (11) than l i 
th solving
directly from (10), because (11) has the simple b.c. of GD (x, x)  0 on
S . Applications of (1
(1.44)
44) can be found in Chs.
Chs 2 and 33. The problem
below gives an application without the need to solve (11) for G (x, x). 19
optional 1.10 Formal Solution of Electrostatic Boundary-Value Problem…(continued)
Problem: A hollow cube (see figure) has six
square sides. There is no charge inside. Five sides
are grounded. The sixth side, insulated from the
others,
th is
i held
h ld att a constant
t t potential
t ti l Φ0. Find
Fi d ththe
potential at the center of the cube.
l i : Let
Solution
S L t the
th center t off the th cube b beb att x  0 andd rewrite
it (1.44):
(1 44)
 (x)  4 1
v  ( x ) G ( x , x ) d 3
x  41  s  (x)  GD (x, x)da (1.44)
0
D n
If all 6 sides had the potential  0 , then  (x  0)   0 and by (1.44)
 0   41  s  0 n GD ((0, x)da ((12))
For the present problem, we have    0 on side 1 and   0 on
the other 5 sides. Byy ((1.44), ), the ppotential at the center is
 (x  0)   41 side 1  0 n GD (0, x)da   16 41  s  0 n GD (0, x)da

 16  0 0 by (12)
 GD (0, x) is symmetric with
respect to all six sides. 20
 1 1 Coulomb
1.1 Coulomb’ss Law
Coulomb's law, discovered experimentally, is a fundamental law
governing
i all
ll electrostatic
l i phenomena.
h It
I states that
h the
h force
f on
point charge q due to point charge q1 obeys (see figure)
1. F  q, q1 , and 12 . q
2. F is along r. r er
 q r
qq1 1
F e   (central
2 r
( t l force)
f )
4 0 r   F is attractive if q and q1 have opposite signs.
3. 
  F isi repulsive
l i if q andd q1 have
h the
th same sign.
i
Furthermore, if there are multiple charges present, the total force
on q is
i the
th vector
t sum off the
th two-body
t b d Coulomb
C l b forces
f between
b t q
and each of its surrounding charges.
Question: What is the principle of linear superposition?
21
 1.2 Electric Field
The electric field at point x due to one or more charges is defined
F
as E(x)  lim q , (1.1)
q 0
where q is a test charge and F is the total Coulomb force on q. We let
q be infinistesimal so that it will not alter the field configuration.
g
Thus, E(x) due to a single point charge q1 is er
q1 1 q1 xx1 r
E(x)  4 2 e r  4
0 r 0 |xx1|3 q1
x
x 1
re r = x  x1  ( x  x1 )e x  ( y  y1 )e y  ( z  z1 )e z

r | x  x1 | ( x  x1 )2  ( y  y1 ) 2  ( z  z1 )2 0
S
For distributed
di ib d charges,
h we have
h by
b linear
li
v
1  ( x)(xx) 3  x x
superposition: E( x)  4  3 d x (1.5)
0 v |xx|
Question: Why write "rer" as "x  x1"? 0 22
 1.3 Gauss’s Law
Consider a ppoint charge g q and a closed surface S and adopt p the
following notations:
da : infinitesimal surface area on S S n q
 n : unit vector normal to da  E  2 er
 d da 4 0r
 and pointing outward
 e r : unit vector along r q r  re r
  : The angle between n and E
q q q
E  n da
d  2 er  n da
d  2 cos
 
d = 4 d 
da
4 0r 4 0r 2 0
r d
Note: d  carries the sign of cos  .
n q outside S ,
d   0, if cos  0 q inside S , n
d   0,  d  4 q  d  0 n
 0 if cos   0 
n
S q S
q  q
 , q inside S Gauss
Gauss'ss law for 
  s E  n da  4  d    0
  (1.9)
a single charge
0 0, q outside S   23
 1.3 Gauss’s Law (continued)
By the principle of linear superposition, Gauss's law for a discrete
set of charges inside S is
 s E  n da  10  qi (1.10)
i
and Gauss's law for a distribution of charge is
  1  ( x) d 3 x
 s E n d
da  0 v (1 11)
(1.11)
Discussion: (1.11) is the integral form of Gauss's law. In the
next section, we will derive the differential form of Gauss's law.
Gauss's law is a powerful mathematical representation of Coulomb's
law (see example below). Furthermore, as will be shown in Ch. 6,
the two forms of Gauss's law are also applicable to time - dependent
cases where the original form of Coulomb's law (a static law),
1  ( x)(xx) 3 
E(x)  4  3 d x,
0 v |xx|
24
no longer applies.
 1.3 Gauss’s Law (continued)

Shell theorems - an application of Gauss’s law:

Halliday, Resnick, and Walker, “Fundamentals of Physics”:
“The two shell theorems that we found so useful in our study
y of
gravitation hold equally well in electrostatics:
Theorem 1: A uniform spherical shell of charge behaves, for
external points, as if all its charge were concentrated at its center.
Theorem 2: A uniform spherical shell of charge exerts no force on a
charge particle placed inside the shell.”
Proof :
Symmetry
S i   E  Er e r
consideration
id
 Q
  , r  a a: radius of shell
Gauss
Gauss' law   4 r Er   0
2
0, r  a Q: total charge on shell
 Q , r  a (as ( if Q were at r  0))
2
 Er   4 0r
0, r  a (Q produces no E) 25
 1.4 Differential Form of Gauss’s Law
Using the divergence theorem:
n is a unit vector normal to the 
  3
  surface element da and ppointingg 
v Ad x  s A n da ,
away from the volume v enclosed 
 by surface S . 
S
we obtain from  s E  n da  1 v  (x)d 3 x [(1.11)] da n
0
3 3 v
 s E  n da  v   Ed x  10 v  (x)d x

 v (  E   ) d 3 x  0 (1.12)
0

 E  
0
differential form of Gauss's law  (1.13)

Question: If v f ( x)d 3 x  0 for an arbitrary volume v, then

f ( x)  0 everywhere. This is the basis for obtaining
(1 13) from (1
(1.13) 12) Does  s A  da  0 for an arbitrary
(1.12).
closed surface S imply A = 0 everywhere? 26
 1.5 Another Equation of Electrostatics
and the Scalar Potential
 | x  x |n
n
 x [( x  x) 2  ( y  y) 2  ( z  z ) 2 ]2 ex  operates on x.
n
 operates on x.
 y [( x  x) 2  ( y  y) 2  ( z  z ) 2 ]2 e y  | x  x ' |n  
 | x  x |n
n
 z [( x  x) 2  ( y  y) 2  ( z  z ) 2 ]2 e
z
n
1
2 2 2 2
 2 [( x  x)  ( y  y)  ( z  z ) ] 2( x  x)e x
n
n
1
2 2 2 2
 2 [( x  x)  ( y  y)  ( z  z ) ] 2( y  y)e y
n
n
1
2 2 2 2
 2 [( x  x)  ( y  y)  ( z  z ) ] 2( z  z )e z
n
n2
 n | x  x | (x  x) (1)
Ex :  | x  x | xx ;  1   xx3 ;  1 3  3 xx5
|xx| |xx| |xx| |xx| |xx| 27
 1.5 Another Equation of Electrostatics and the Scalar Potential (continued)

1  (x)(xx) 1   ( x) d 3 x
E(x)  4  d 3 x   4  |xx|
0 |xx|3 0

  (x),  |x 1x|   x x3

|x x|
 (x)
1
where  (x)  4  d 3 x scalar potential (1.17)
0 |xx|
 E = 0 (1.14)
Question: E      E = 0. Is the reverse also true?
Below we show that q (x ) can be interpreted as the potential
energy of charge q at position x, and   E = 0 can be derived by
an alternative method using Stokes's theorem:
loop C
 c A  d   s (  A)  n da d
n
d : a line element on a closed loop C da S (open
S : arbitrary open surface bounded by loop C surface
n: unit
it vectort normall to t surface
f element
l t da
d in
i bounded
the direction given by the right -hand rule by loop C )
28
 1.7 Poisson and Laplace Equations
  E   (1.13)
Rewrite  0

E   (1 16)
(1.16)
Sub. (1.13) into (1.16), we obtain the Poisson equation
2 
    (1.28)
0
In a charge-free
g region,
g , ((1.28)) reduces to the Laplace
p euqation
q
 2  0 (1.29)

29
 Summary of Secs. 1-5 and 7:

qq1 q1
F e E 2 er
4 0 r 2 r 4 0 r  re r  x  x1
 q1 (xx1) r  x  x1
E  lim qF 
q0 4 0|xx1|3  principle of
er 1  (x)(xx) d 3 x linear superposition
 4
r  |xx|3
q1
x1 x v
S 0   | x  x |n
  41   (x) d 3 x  |   |n 2
(x  x)
x  |xx|
n x x
0 x 0
1  (x) 3 
     4 0  |xx|
d x
0
← derived in Sec. using E  qe r
  f  0  4 0r 2
  1  3
 s E n da  v
0
( x ) d x
E  0  divergence thm.
  E   0
 2     0 30
 Questions on Secs. 1-5 and 7:
1. Can one calculate E by using  · E = 0 alone?
2. Can one calculate Φ (hence E) by using 2Φ = 0 ? How?
3. Can one calculate Φ (hence E) by using  (x)  1   ( x) d 3 x ?
4 0 |xx|
How?
4. Why break one equation,  (x)  1   ( x) d 3 x, into two
4 0 |xx|
equations:  × E = 0 and  · E = 0 ?
5. Coulomb’s law gives  × E = 0 and  · E =  /0 . Can it give any
other independent relation for E?
Helmholtz s Theorem: “A
Helmholtz’s A vector is uniquely specified by giving its
divergence and its curl within a region and its normal component over
y (Arfken,
the boundary.” ( , “Math. Meth. for Physicists”,
y , 3rd Ed. pp.78))

31
 6. Is the integral form of Gauss’s law mathematically equivalent
to the differential form of Gauss
Gauss’ss law?
Answer: Yes. To prove the mathematical equivalence, we need to
show that the integral
g form of Gauss’s law is both a sufficient
ff and
necessary condition for the differential form of Gauss’s law. This
can be demonstrated as follows:
1 3
 E  nda  0
v  (x)d x (1.11)

   divergence thm.
3
v (  E    0 ) d x  0 (for arbitrary volume v)
 
  E   0 (1.13)
Downward manipulation () shows that (1.11) is a sufficient
condition for (1.13). Upward manipulation () shows that (1.11) is a
necessary condition for (1.13).
(1 13) Hence,
Hence the two forms of Gauss
Gauss’ss law
are mathematically equivalent. 32
 7. Is G
Gauss’s law mathematicallyy equivalent
q to C
Coulomb’s law?
Answer: No, because Coulomb’s law is a sufficient but not a
necessary condition for Gauss’s law. That is, we may derive Gauss’s
law from Coulomb’s law, but not the reverse.
While Coulomb’s law completely specifies the E field, we need
more information to completely specify the E field from Gauss’s law.
This is clear when we write Gauss’s law in its differential form,
 · E =  0 . By
B Helmholtz’s
H l h lt ’ Theorem,
Th we also
l needd the
th curll off E
to completely specify E. In electrostatics, this is given by  × E = 0.
In general,
general it is given by Faraday law  ×   ∂ ∂ t (Ch.
Faraday’ss law, (Ch 5).
5)
As will be shown in later chapters, while Coulomb’s law [in the
form of (1.5)
(1 5) or (1.17)]
(1 17)] deals only with the static E field,
field Gauss
Gauss’ss
law covers a much broader class of fields than Coulomb’s law, such
as the E field of an electromagnetic
g wave.
33
 8. Is Gauss’s law pphysically
y y equivalent
q to Coulomb’s law?
Answer: In the special case of electrostatics, the field
surrounding a point charge is symmetric, implying E=Erer.
Choosing a spherical surface of radius r centered at the point charge,
we may obtain Coulomb’s law from Gauss’s law,
1 3
 E  nda  v  ( x ) d x (Gauss's law)
0
  E  da  Er 4 r 2  q  0
q
 Er  2
(Coulomb's law)
4 0 r
In 1.3, we have also derived Gauss’s law from Coulomb’s law;
hence the two laws are physically equivalent in electrostatics.
hence, electrostatics
However, as discussed in question 7, the two laws are not
mathematicallyy equivalent,
q , nor are theyy pphysically
y y equivalent
q in
electrodynamics. 34
1.6 Surface Distributions of Charges and Dipoles and
Discontinuities in the Electric Field and Potential
Surface Layer of Charge :
The surface charge density is defined as charge per unit area
q
on the surface:  ( x )  lim  a
a  0
Note:  and  have different dimensions. a, q
q n2
Apply Gauss's law,  E  n da   , to an
0
infinitesimally thin pillbox, we obtain n1 E2
q
( E1  n1  E 2  n 2 )  a   E1 pillbox 
0 illb
n1   n 2 (thickness  0)
1 q 
 (E  E )  n 
2 1 2  0 a 0  (1 22)
(1.22)
The tangential component of E can be shown to be continuous
across the
h layer
l by l i  C E  d   0 to the
b applying h loop
l d
drawn in
i
35
dashed lines in the figure.
 1.6 Surface Distributions of Charges and Dipoles… (continued)

l : (see
Example ( fi
figure)) a  (uniform distribution of
 Q , ra Q on a layer of radius a )
 4 0a

Q
 4 r , r  a
 0
 
At r  a,  is continuous.
Er is discontinuous.
a r
Questions :
1. Fi ld (E andd  ) off a point
1 Fields i t charge
h diverge
di as one moves
infinistesimally close to the charge. Explain why fields of the
surface charge do not diverge as one moves infinistesimally close
to the surface.
Answer: A point charge is a finite amount of charge concentrated
at a point. However, for the surface charge, one must integrate  over
a finite surface area to obtain a finite amount of charge. Hence, there
i no finite
is fi it amountt charge
h att a single
i l point
i t on the
th layer.
l
2. Why is  continuous across the layer? 36
 1.11 Electrostatic Potential Energy and Energy
Density; Capacitance
Electric Field Energy: Let  (x) be the field due to the presence
of  . The work done to add   is
3
U i   a  a 
Using      a
 W    (x) (x)d x we obtain
   0   E    E    ( E)   E 
  0     E(x)d 3 x    (   E)  E   E
  0    (   E ) d 3 x   0  E   Ed 3 x   0  E   Ed 3 x
 
By conservation
of energy, this  
  s  E  da  0, as r    For this integral to vanish, 
must be the total 1 1 r2  the volume of integration
g 
r r2  must be infinite. 
E-field energy.
 2
 W   0  d 3 x 0 E  dE  20  E d 3 x
E
infinite volume (1.54)
2 
E  E  E  E     (E)    E    (E)  
0

0as
 r 


 W  12   (x) (x)d 3 x  20  s E  da  12   (x) (x)d 3 x (1.53)37
optional 1.5 Another Equation of Electrostatics and the Scalar Potential (continued)
Work done by bringing charge q from B
position A to position B along any path: E
B
W  A F  d  d

A
 q  A E  d 
B
   x e x   e   e ; d   dxe  dyye  dze
y y z z x y z
 q  A   d 
B
 d     d    x
dx   dy   dz
y z
 q A d 
B
 d  is an infinitesimal change of  due to an
 q ( B   A ) infinitesimal displacement d  .

Thus, W depends
Th d d onlyl on the
th values
l off  att A andd B, andd it is
i
independent of the charge's path from A to B. This justifies the
concept of potential energy
energy, which implies that the total work done
on q in a round trip along any closed path C is 0, i.e.
 C E  d   0 or,
or by Stokes theorem, S (  E)  n da  0 (1.21)
Stokes'ss theorem (1 21)
Since S is an arbitrary surface, we obtain   E  0. 38
 1.11 Electrostatic Potential Energy… (continued)

An alternative derivation of (1.53) and (1.54): Consider a state in

which a charge density  (x) has produced an electrostatic potential
Φ (x)), i.e. ρ(x) → Φ (x)).
Then, by the principle of linear superposition,

we have  (x) →  Φ (x),
where  is a constant.
To find the electric field energy, we consider the energy needed to
build up Φ (x) from   0 (no charge and no potential) to   1 (the
present state). At any stage in the build-up process, the relative
charge
h d i (hence
density (h the
h relative
l i potential)
i l) remains
i theh same;
namely, the intermediate state is characterized by the charge density
(x) and potential  Φ(x). )
In such a build-up process, when the potential is Φ(x), the work
done by adding an incremental charge (x)d is
dW  v d 3 x(x) (x)d  39
 1.11 Electrostatic Potential Energy… (continued)
Hence, the total work done from   0 to   1 is
 1 1
W   0 dW  v d 3 x  (x) (x) 0  d 
 12 v d 3 x 
 ( x)  ( x) For this integral to vanish,
vanish the (1 53)
(1.53)
volume of integration must be .
  0 2 
Green's 1st identity
y  
  ~rr2
  12  0 v  2d 3 x  12  0 [ v  d 3 x   s ( 
n
)da ]
2
 12  0 v E d 3 x iintegration
i over
~ 1 1 (1.54)
infinite volume r ~
r2
Questions: 1. If we bring q and  q toward each other, the work
 2
done is negative. Why is then W  20 v E d 3 x always positive?
2 Give one example to show that the E-field
2. field carries energy
energy.
2
Electric Field Energy Density : (1.54)  wE  12  0 E (1.55)
2
Note: wE  2  0 E  12  0 (  E j ) ( E j ) [ 12  0  (E j E j )]
1
j j j 40
 1.11 Electrostatic Potential Energy… (continued)
Force on the Surface of a Conductor: Consider conductor
a conductor with surface charge on it. At any point
on the surface, the total field (E) outside must be  E
normal to the surface and the total field inside must
be 0. Applying Gauss’s law, we find   where  is the local
surface charge density at the observation point (upper figure).
figure) But the
local  by itself will produce equal and opposite fields on
y Gauss’s law Eself Eself
both sides of  ((call it selff field self ) and by
Eself (outside)  Eself (inside)    Eself  2 , Eext Eext
0 0
which
hich is half of the total E outside.
o tside Since the total E inside is 0,
0 all the
external surface charge away from the local  must have produced an
external field with Eext=self  which cancels self inside and
thus doubles self outside. The local  can only experience a force
due to the field (extt) produced by the external surface charge.
charge Thus,
Thus
 2
force on the surface/unit area   Eext  2 (see pp. 42-43) 41
0
 1.11 Electrostatic Potential Energy… (continued)

Capacitance:
p Refer to the figure
g V1 , Q1 V2 , Q2
V   n
 n
P Q
 1 j 1 1 j j Q  C V
 1 j 1 1 j j
  V3 , Q3 V4 , Q4

n n
V2   P2 jQ j Q2   C2 jV j
 j 1  j 1
  Invert the  
 n
equations  n
Vn   PnjQ j Qn   CnjV j A system of
 j 1  j 1
n conductors
by principle of Cii : capacitance
linear superposition Cij (i  j ): coefficient of induction
Pij and Cij depend on the geometrical shape and position of the
conductors. Potential energy of the i -th conductor is [using (1.53)]
Wi  12  i (x)i (x)d 3 x  12 QiVi i (x)  Vi ;  i (x)d 3 x  Qi 

  P t ti l energy  1 n
Potential 1 n n
 2  QiVi  2   CijViV j (1.62)
of the system  i 1 i 1 j 1 42
 Homework of Chap. 1

Problems: 3,, 4,, 5,, 6,, 8,,

9, 12, 14, 16, 17

Quiz: Oct. 05, 2010

43
44
 Appendix A: Unit Systems and Dimensions
Unit Systems:
Two systems
y of electromagnetic
g units are in common use today:
y
the SI and Gaussian systems. Regardless of one’s personal
preference, it is important to be familiar with both systems
y and, in
particular, the conversion from one system to the other. Conversion
formulae can be divided into two categories: “symbol/equation
conversion such as E and E = q/(4π0r2)]” and “unit conversion
(such as coulomb)”.
Conversion formulae for symbols and equations are listed in
Table 3 on p. 782 of Jackson and conversion formulae for units in
Table 4 on p. 783 (both tables attached on next page). These two
tables are all we need to convert between SI and Gaussian systems.
Correct use of the tables requires practices.
45
 Appendix A: Unit Systems and Dimensions (continued)

Jackson, p. 782, Table 3 Jackson, p. 783, Table 4 46

 Appendix A: Unit Systems and Dimensions (continued)

Conversion off symbols

y and equations
q :
Consider, for example, the conversion of the SI equation
q
E 2 (A.1)
4 0r
into the Gaussian system.
This involves the conversion of symbols
y and equations.
q So we use
Table 3. First, we note from Table 3 (top) that mechanical symbols
(e.g. time, length, mass, force, energy, and frequency) are unchanged
in the conversion. Thus, we only need to deal with electromagnetic
symbols on both sides of (A.1).
G
From Table 3, we find E  SI E and q SI  4 0 q G (A.2)
4 0
Sub. E G / 4 0 and 4 0 q G , respectively
Sub respectively, for E and q in (A.1),
(A 1)
we obtain the corresponding equation in the Gaussian system:
4 0 qG qG
EG  E G
 2 (A.3)
4 0 4 0r 2 r 47
 Appendix A: Unit Systems and Dimensions (continued)

Conversion of units and evaluation of physical quantities:

q
Consider again the SI equation : E  2 (A.1)
4 0r
Given r  0.01 m, q  1 statcoulomb, we may evaluate E in 3 steps:
Step 1: Express r , q, and  0 in SI units. From Table 3 (bottom)
andd Table
T bl 4,
4 we find
fi d
 0  8.854  1012 Farad/m  1
9 Farad/m
 36 10
 r  0.01 m (same as given) (A.4)
 q (  1 statcoulomb)  1 9 coulomb
 310
0
Step 2: Sub. the numbers (but not the units ) from (A.4) into (A.1).
1
Thi gives
This i E
q
 3109  3  104
4 0r 2 4  1 9 (0.01)2
36 10
Step 3: Look up Table 4 for the SI unit of E. As shown in Table 4,
4
the SI unit of E is V/m. Thus, E  3 10 4 V/m (A.5) 48
 Appendix A: Unit Systems and Dimensions (continued)

As another exercise, we write (A.1) in the Gaussian system :

q
E 2 (A.3)
r
al ate E for the same r (  0.01
and eevaluate 0 01 m) and q (  1 statco
statcoulomb).
lomb)
Step 1: Express r and q in Gaussian units. From Table 4, we find

 r (  0.01
0 01 m))  1 cm
q  1 statcoulomb (same as given)
(A.6)

Step 2: Sub. the numbers (but not the units ) from (A.6)
(A 6) into (A.3).
(A 3)
q
This gives E  2  11  1
r
Step 3: Look up Table 4 for the Gaussian unit of E. We find the
unit to be statvolt/cm. Thus, E  1 statvolt/cm (A.7)
Table 4 shows 1 statvolt/cm  3 10 4 V/m. Hence, the 2 results
 E  3  104 V/m 
in (A.5)
(A 5) and (A.7):
(A 7):   are identical as expected.
expected
 E  1 statvolt/cm  49
 Appendix A: Unit Systems and Dimensions (continued)

Units and Dimensions :

In the Gaussian system, the basic units are length (), mass (m), and
time (t ). In the SI system, they are the above plus the current ( I ). [See
Table 1 (top) on p. 779 of Jackson.] All other units are derived units.
If a physical quantity is expressed in term of the basic units, we
h
have th
the dimension
di i off this
thi quantity.
tit
A mechanical quantity has the same dimension in both systems.
example the acceleration a ( d 2 x / dt 2 ) has the dimension of
For example,
t 2 . From f  ma, we obtain the dimension of force : mt 2 , which
in turn g gy m 2t 2 .
gives the dimension of work (f  ) or energy:
An electromagnetic quantity has different dimensions in different
systems. For example, the charge q has the SI dimension of It. From
the Gaussian equation f  q1q2 / r 2 and the dimensions of force and
length, we find the Gaussian dimension of q to be m1/23/2t 1. Since
q has
h the di i off energy (m 2t 2 ),
th dimension ) the t ti l  has
th potential h the th SI
dimension of m 2t 3 I 1 and the Gaussian dimension of m1/ 21/ 2t 1. 50
 Appendix A: Unit Systems and Dimensions (continued)

All physical quantities in an equation must be expressed in the

same unit system and all terms must have the same dimension. For
example, by Stokes's theorem, we have
 C E  d   S (  E)  n da (A.8)
where both terms have the dimension of   (the dimension of E ).
)
In the definition of the delta function:
a1  ( x  a)dx
a2
 1,
1 (A 9)
(A.9)
the RHS is dimensionless. Thus, if x has the dimension of ,  ( x  a )
must have the dimension of  1. However
However, "0" 0 is not to be regarded
as a dimensionless quantity. This is clear if we write (A.8) as
 C E  d   S (  E)  n da  00.
Well known equations need not be checked for dimensional
consistency However
consistency. However, for newly derived equations,
equations a dimensional
check can be a convenient way to find mistakes. 51
optional 1.6 Surface Distributions of Charges and Dipoles… (continued)

Dipole Layer :  (x)  |xx xx|

D ( x )  lim  ( x )d ( x )  ( x ) n points
d ( x ) 0 from  
x to  

0 r
x
Assume that, at any given point, the two layers have equal and
opposite surface charge densities (see figure).
1  (x) 3 1  (x)  (x)
 ( x)   d x  [ s da  s ' da]
4 0 | x  x | 4 0 | x  x | | x  ( x  n d ) |
da’=da’’
1  1 1 
   (x)    da
4 0 s
 | x  x | | x  ( x  n d ) |  52
optional 1.6 Surface Distributions of Charges and Dipoles… (continued)

Using the binomial expansion: What is the Taylor expansion?

n 1 n(n  1) n2 2
( x  y )  x  nx y 
n n
x y  ,
2!
we obtain
1 2
1 1 1 a a b 
2
  1  2  2 2 

| b  a | b 2  a 2  2a  b

12 b b b 

1 a2 a  b  1 a b
 1  2  2      3
b  2b b  b b
a/b  0

1 1 x  x
   dn  [valid for d | x  x |]
| x  (x  nd ) | | x  x | | x  x |3
b  x  x
a  nd 53
optional 1.6 Surface Distributions of Charges and Dipoles… (continued)

1 1 x  x
Sub.
b   dn 
| x  (x  nd ) | | x  x | | x  x |3
1  1 1 
into  ( x)   
 (x )    da, we obtain
4 0  | x  x | | x  ( x  n d ) | 
s

  |x1x|   |x1x|
  
1 x  x 1 1
 ( x)   

(

x )

d ( x

) n  da    D ( x ) n   da
4 0 s
| x  x |3 4 0 s | x  x |
D(x ) 
(1.24)
σand d appear as a product here, so it’s meaningful
to define the product as the dipole layer strength.
1 x  x 1 1
or  (x)   D (x) n  da    D (x)d 
4 0 s | x  x | | x  x | 2 4 0 s
(1.26)
   
0 if cos  0
 d   0,  cos 1 r2 S fi
See figure ttwo
 d   0, if cos  0  pages back. 54
 d
optional 1.6 Surface Distributions of Charges and Dipoles… (continued)

 1 
 4 s D(x )d  (1.26)
(1 26)
0
Rewrite :  (x)  
 1  D(x)n   
1
d 
da (1
(1.24)
24)
 4 0 s | x  x |
g of d  are shown below with
Note: ((1)) The direction of n and sign
respect to the polarity of the dipole layer:
direction
d ec o ofo n: sign of d  : See derivation 
n off (1.26).
(1 26) 
x

+ d  0
+ x 
 d  0
 +
(2) The RHS of (1.24) is an explicit function of x (the position
of observation). The RHS of (1.26) is an implicit function
of x, because the total solid angle depends on x.
Question: Under what condition will (1.24) and (1.26) be invalid? 55
optional 1.6 Surface Distributions of Charges and Dipoles… (continued)

Special
p case 1: A flat-disc shaped
p double layer y with D  const.
1 d

   4 s D(x )d  (1.26)
0
   2D   2D

0
      D  ( D )  D 0
2 0 2 0 0  
electric field between layers:
y E  D .
  0d
 is discontinuous across the dipole layer. p
a
S i l case 2: Point
Special P i dipole
di l d  q q
n
 
 a0 a 
p lim n D da
  a  0 d
point dipole a =q
dipole layer dipole layer point dipole
 lim a n( d )da  n d a  nqqd
a 0

  (x)  41 lim    1 da  1 p( xx) (1.25)


0 a 0 a
D ( x )n
|xx| 4 0 |xx|3
 
(1.24)  xx3 56
|xx|