PH 4011 Electromagnetic Fields II: Semester 2 - 2017
PH 4011 Electromagnetic Fields II: Semester 2 - 2017
PH 4011 Electromagnetic Fields II: Semester 2 - 2017
Semester 2 - 2017
45 hours 3 credits
Coulombs Law
Experimentally it has been observed that in classical electrostatics the interaction between two charges
can be given by a mechanical force.
q'
1
=
40
2
where 0 = 8.851012 So
.2
1 ( )
= 3 1.2
40 | |
Linear Superposition
Force on a charge at due to charges 1 , 2 , 3 located at 1 , 2 , 3 . is the
vector sum of
1 ( )
= 3 1.3
40 | |
=1
We can write this as
= () 1.4
Where
1 ( )
() = 3 1.5
40 | |
=1
() is the electrostatic field at the field point (known as the observation point), due to the field
producing charges at the source point .
From (1.1) we can see that for a single charge,
= ()
1 1
() = ( ) 1.6
40 | |
( ) 1
3 = ( )
| | | |
If the discrete charge are numerous enough, we can introduce the term charge density located at
.
Then we can rewrite the equation (1.5) as
1 ( ) 3
() = ( ) 3 1.7
40 | |
Note that charge of a volume element 3 is ( ) 3
( ) 3
Gausss Law
da
da
Then
.
(). = = 1.10
40 2 40 2
(). = 1.11
40
( = 2
) total element of solid angle subtended by da
da
Then
(). = = . 4 = 1.12
40 40 0
1
(). = () 3 1.13
0
Gausss Law is given in equation (1.13). But according to the divergence theorem,
1
(). = (
. ) 3 = ( ) 3 1.14
0
Since this is valid for any volume element 3 , integrands should be equal
( )
. =
1.15
0
Scalar Potential
We know that
( ) 1
(
= ) 1.17
3
| | | |
Then,
1 1
() = ( )
( ) 3 1.18
40 | |
1 ( ) 3
() = . 1.19
40 | |
Since Curl of any gradient of any well-behaved scalar vanishes, we get immediately,
= 0
1.20
As we know (1.15) is not enough to find the 3 components of , we now have equation (1.20).
In equation (1.19), we can see that a vector field is derived from a scalar.
Using (1.19), we can define a potential called a scalar potential (),which is given by,
() =
() 1.21
where,
1 ( ) 3
() = 1.22
40 | |
= ( ) 1.23
W is i.d. of path
Proof
2
(2)
1
(1)
, 2 = .
1 = .
(1) (2)
1 2 = .
+ .
= .
(1) (2)
).
1 2 = (
1 2 = 0 1 = 2
. = 2
Therefore,
2 = 1.24
0
2 = 0 1.25
1 ( ) 3
() =
40 | |
1 ( ) 1 1
2 () = 2 ( ) 3 = 2 ( ) ( ) 3
40
| | 4 0
| |
1 4 =
2 ( )={
| | 0
So
1
2 ( ) = 4( ) 1.26
| |
1
2 ( ) 3 = 4 ( ) 3 = {4 =
| | 0
Thus if
1 ( ) 3
() =
40 | |
1 1 1 1
2 () = 2 ( ) ( ) 3 = ( )( ) 3 = 4()
40 | | 40 40
()
2 () = 1.27
0
So, the equation 1.22 satisfy the Poisson equation for discrete or continuous charge distribution.
Surface distribution of charges and dipoles & discontinuities in the electric field and potential
Let us find the electric field or potential due to a given surface charge distribution
(2)
2
( )
(1)
1 ( ) 2
() = (1.29)
40 | |
For volume or surface charge distribution, the potential is everywhere continuous, even within the charge
distribution.
Another problem of interest is the potential due to dipole layer distribution on a surface S
Dipole layer can be considered as a two surfaces S and S with opposite charge densities separated by a
distance ().
Dipole layer of strength () (surface dipole moment density) is formed by letting S approach
infinitesimally close to S, while the surface charge density () becomes infinite in such a manner that,
lim ()() = ()
)0
(
() ()
-()
( )
( )
( )
1 ( ) 1 ( )
() = 1.30
40 | | 40 | + |
1
For small d we can expand | + | as
1 1 1
| + | = + (nd). ( ) + 1.31
| | | |
As 0, we can neglect the higher order terms. Substituting (1.31) to (1.30), we get
1 ( ) 1 ( ) 1 1
() = ( ). . ( )
40 | | 40 | | 40
| |
1 1 1 1
() = ( ). . ( ) = ( ). [ . ( )]
40 | | 40 | |
So,
1 1
() = ( ). . ( ) 1.32
40 | |
Integrand of equation (1.32) is the potential of a point dipole with a dipole moment = Dda. The
potential caused by a dipole at is
1 1 1
() = . . ( )= . 1.33
40 | | 40 | |3
1 [( )] | | cos
. ( ) = .
. = 3 cos =
2 =
| | | |
| |
| |
| |
1
() = ( ) 1.34
40
If the observation point views the inner side of the dipole layer, is positive.
S
| |
There is a discontinuity in potential crossing a double layer. From (1.34), it is clear that the potential has
a discontinuity in crossing the inner to the outer side.
0
2 1 = 1.35
0
20
20
Equation (1.35) agrees with the equation (1.28), which explain the discontinuity of electric field at
boundary.
Greens Theorem
Equation (1.22) gives a straightforward convenient solution for electrostatic problems, which involve
continuous or discrete charge distribution without any boundaries.
However, most of problems associated with electrostatic comes with boundary conditions.
George Green (1824), introduce a new mathematical procedure to solve these problems. This follows a
simple application of Divergence theorem.
. 3 = . 2
1.36
=
1.37
. (
) 3 = . 2 =
. 2
But
. (
) = 2 + .
and
. =
Then we have,
2
. 3 + 2 3 = 1.38
=
and equation (1.38) changed in to
2
. 3 + 2 3 = 1.39
2
. . ) 3 + (2 2 ) 3 = (
( )
2
(2 2 ) 3 = ( ) 1.40
Equation (1.40) is called the Second Greens identity or the Greens Theorem
The Poisson differential equation for potential can be converted in to integral by selecting,
1 1
= = |
and scalar potential as .
|
Also, we have
()
2 () =
0
and
1
2 () = 2 ( ) = 4( )
| |
( ) 3 1 1 2
(4( ) + ) = ( ( ) )
0 R
4( ) = 4()
Then
( ) 3 1 1 2
4() = + ( ( ) )
0 R
1 ( ) 3 1 1 1
() = + ( ( )) 2 1.41
40 0 4 R
The first term gives the potential due to the charge distribution inside the volume
If the point lies outside the surface S the LHS of the (1.41) is zero.
if V is all space so that S is at infinity, then the surface integral will vanish and we recover the volume
integral over all space. Then equation (1.41) drops back to (1.22)
Note also that if (x) = 0 for all x in V, then the potential is found simply from the surface integral of
and its normal derivative.
1 ( ) 3
() = 1.22
40 | |
For a charge-free region (i.e. for ( ) = 0 for all )equation (1.41) change into a surface integral in
terms of potential and its normal derivatives at the surface.
If the specification of potential on a closed surface defines a unique potential, that is called Dirichlet
problem or a Dirichlet B.C. can be achieved by conductors
If the electric field (normal component of the potential) everywhere on the surface is defined a unique
problem, that is called Neumann B.C. not so common in electrostatics
() specified on S Dirichlet
V
S
Suppose 2 = in V and is defined on S or is defined on S.
0
In 1st case is uniquely determined, in 2nd case is uniquely determined within a constant.
Let 2 satisfy 2 2 = in V and 2 = ( ) specified on S
0
So 1 and 2 are two solutions of 2 = satisfying the same B.C.
0
Let = 2 1 1.42
Then
2 = 2 2 2 1 = 0 in V, and = 0 on S
3 U 3
.
3 + 2 3 = U.
U 3 + U2 U 3 = U
U
U 3 = 0 since U = 0 on the surface
2
U.
U 3 = 0 |
| 3 = 0
= 0
i.e., = in V. But = 0 on S for Dirichlet B.C.
Therefore, = 0
For Neumann BC
2 1
=0 =0
Therefore = in V
The preceding proof is also valid for the case of mixed boundary conditions because the surface integral
vanishes in this case also. Finally, one cannot in general specify both () and / everywhere on S
(Cauchy boundary conditions); either one alone is sufficient to determine a unique solution and the two
solutions so determined are not necessarily the same. However, if Cauchy boundary conditions are given
on just an appropriate part of S, that can be sufficient to give a unique solution for the potential.
For Dirichlet BC, we saw that 2 = 1 and the solution is unique. In Neumann BC, still the solution is
unique apart from an unimportant constant.
The solutions of the Poisson or Laplace equation in a finite volume V with Dirichlet or Neumann BC on
the boundary surface S can be obtained by Greens theorem (1.40) and so caaled Greens function.
Suppose we have
2 () = in a volume
0
() is specified on
Take
1
(, ) = = and () =
| |
Then
2 (, ) = 4( )
(, ) ( ) 2
[( )2 (, ) (, , )2 ( )] 3 = [( )
, )
( ]
Then we have
1 (, ) ( ) 2
4() + [ (, , )( )] 3 = [( )
( ,
) ]
0
1 1 (, ) ( ) 2
() = [ (, , )( )] 3 [( )
( ,
) ] 1.44
40 4
(, ) = 0 for on
1 1 (, ) 2
() = [ (, , )( )] 3 [( ) ] 1.45
40 4
= ( )
2 () = in a volume
0
1 1 (, ) 2
() = [ (, , )( )] 3 [( ) ]
40 4
GD (, ) = 0
GD (, )
1 1 (, ) ( ) 2
() = [ (, , )( )] 3 [( )
( ,
) ]
40 4
2 (, ) = 4( )
This gives us
2 ( , ) 3 = 4 ( ) 3 = 4
2 ( , ) 3 =
.
(, ) 3 =
(, ). 2 = 4
2
(, ). 2 =
= 4
)
,
(
Therefore, we cannot use the BC = 0, unless the surface is at infinity
( )
, 4
= for on
1 1 4 1 ( ) 2
() = [ (, , )( )] 3 ( ) ( ) 2 + [ (, ) ]
40 4 4
1 1 ( ) 2
() = + [ (, , )( )] 3 + [ (, ) ] 1.46
40 4
Useful relationship
(, ) = ( , )
Proof
Let = ( , ) and = (, )
3
(2 2 ) 3 = ( )
, y) 2 GD(X
GD(X , y) GD (X
, y) 2 GD(X
, ) 3
, y)
GD (X , y)
(X
, y)
= ( (X , y)
GD (X ) 3
, y) = 0 for y on S
(X
, y) = 0 for y on S
(X
So
4 (, ) = 4 ( , )
(, ) = ( , )
Example 1
Use (1.45) to find the potential due to a point charge q at = in infinite space.
Example 2
1
2
1 1
2 = 2 ( ) = 2 (2 ) 1.47
40 |1 2 |
1 1 2
3 = 3 [ + ]
40 |1 3 | |2 3 |
2
3
3 = 1.48
40 |2 |
=1
Then the total work to be done to assemble N charges against the charge is
1
= ( ) = 1.49
=1
40 | |
If you have a continuous charge distribution, the summation can be replaced by an integration
1 ()( ) 3 3
= 1.51
80 | |
But
1 ( ) 3
() =
40 | |
1
= ()(X
) 3 1.52
2
)
(
We know that 2 =
0
Substituting to 1.52,
1
= 0 2 . () 3
2
. [(
)] = . + 2
1
. [(
= 0 [ )]
.
] 3
2
Applying Divergence theorem to the first term
1 1
)d3 X + 0
= 0 ( .
d3 X
2 2
At large distances the first term goes to zero (depend on 1/ 2 )
Then
1 2 0 ||2
= 0 d3 X = 0
. | ()| 3 = 1.53
2 2 2
At the integration of (1.53) all terms with charges disappeared. Only the electric field all over the space is
available now.
Note: It is clear that the energy is always positive regardless of the charges involved. But according to the
(1.50), we can see that the energy can be negative depending on the polarity of charges involved.
This is due to the fact that (1.53) is derived including the self-energy term.
Let us calculate the energy due to two charges 1 and 2 located at 1 and 2 respectively.
1 1 ( 1 ) 2 ( 2 )
() = [ + 3 ] 1.54
40 | |3 | |
1 2
2
2
1 12 22 2
= [ +
2 2 0 4 4 + 21 2 ( 1 ). 3 3] 1.55
32 0 | 1 | | 2 | | 1 | | 2 |
We can see that the first two terms are the self-energies and the third term is the interaction energy.
1 2 2
= 1 ).
(
16 2 0 3 3
| 1 | | 2 |
1 1 2
=
40 | |3
1 2
Capacitance
We know that
1
= ()() 3
2
Consider a system of conductors in space. There is no charges on space except on conductors. Potential
of each conductor is with total charge
1
2 3 i
Since the potentials linearly depends on , the potential of the ith conductor can be written as
= ( = 1,2, )
=1
depend on the geometry of the conductors. These n equations can be inverted to yield the charge on
the ith conductor in terms of all the potentials.
= ( = 1,2, . . )
=1
1
= 1.56
2
=1 =1
The coefficient are called capacities or capacitance while the , are called coefficients of
induction.
=
40
So, capacitance
= 40 =
We know that
0 ||2
=
2
=
40 2
3
= <
40 2 3
= <
40 3
1
= >
40 2
0
2 2
1 2 2
= [ ( ) 4 + ( ) ]
2 0 40 3 40 2
2 1 1 3 2
= [ + ]=
80 5 20 0
Variational methods are widely used in classical and quantum physics. Variational method can be used to
obtain an accurate solution.
Function ()is well behaved inside the volume V and on its surface S. () is a specified source without
singularities within V.
The first order changes in the functional when we change from + (the modification ()
is infinitesimal within V), is
= .
() 3 3 + 1.58
Proof
( + ) = + ()
( + ).
( + ) =
.
+
.
() +
.
() +
().
()
[ + ] []
1 1 1
= .
3 +
.
() 3 +
.
() 3 .
3
2 2 2
3 3 + 3 +..
= .
() 3 3 +
= , = , yields
(2 + . ) 3 = da
().
(2 + ) 3 = da ().
3 = (2 ) 3 + da
Substituting to (1.58),
= (2 ) 3 + da 3
= ( 2 ) 3 + da
Provided = 0 on (so that the surface integral vanishes) the first order change in [] vanishes if
2 () = 1.60
Neglected terms in (1.58) are semi-positive ( > 0). We see that [] is a stationary minimum if satisfy
Poisson like equation within and vanishes on .
With the and /0 , the minimization of the functional [], yields the equation of motion
of the electrostatic potential in the presence of a charge density and Dirichlet B.C. ( gives on and =
0 there)
Using the variational functional method one can derive the Poisson equation
The stationary nature of the extremum of [] permits a practical approach to an approximate solution
for ()
1
[] = . d3 X 3 1.61
2
= (2 ) 3 + ( ()) da 1.62