ECE 307: Electricity and Magnetism

Fall 2012

Instructor: J.D. Williams, Assistant Professor

Electrical and Computer Engineering
University of Alabama in Huntsville
406 Optics Building, Huntsville, Al 35899
Phone: (256) 824-2898, email: john.williams@uah.edu
Course material posted on UAH Angel course management website

Textbook:
M.N.O. Sadiku, Elements of Electromagnetics 5th ed. Oxford University Press, 2009.
Optional Reading:
H.M. Shey, Div Grad Curl and all that: an informal text on vector calculus, 4th ed. Norton Press, 2005.

All figures taken from primary textbook unless otherwise cited.

Chapter 6: Electrostatic Boundary Value Problems
• Topics Covered

– Poisson’s and Laplace’s

Equations
– Uniqueness Theorem
– General procedures for solving
Poisson’s or Laplace’s equations
– Resistance and Capacitance
– Method of Images (quick look
re-look into Coulomb’s law)

• Homework: 3, 10, 14, 16, 19, 23, 24,

30, 32, 43, 45, 46, 47,

8/17/2012 2
All figures taken from primary textbook unless otherwise cited.
 Resistance
• Recall from Chapter 5, that we defined Resistance as R = L/S
• We can also define it using Ohm’s law as
 
V
R 
 E  dl
 
I  E  dS
• The actual resistance in a conductor of non-uniform cross section can be
solved as a boundary value problem using the following steps
– Choose a coordinate system
– Assume that Vo is the potential difference between two conductor terminals
– Solve Laplace’s Eqn. to obtain V. Then Determine E = -V and solve I from
 
I   E  dS
– Finally, R = Vo/I

8/17/2012 3
 Capacitance
• Capacitance is the ratio of the magnitude of charge on two separated plates
to the potential difference between them
 
Q   E  dS
C   
 
V  E  dl
• Note that V    E  dl The negative sign is dropped in the definition above
because we are interested in the absolute value of the voltage drop
• Capacitance is obtained by one of two methods
– Assuming Q, determine V in terms of Q
– Assuming V, determine Q in terms of V
• If we use method 1, take the following steps
– Choose a suitable coordinate system
– Let the two conducting plates carry charges +Q and –Q
– Determine E using Coulomb’s or Gauss’s Law and find the magnitude of
the voltage, V, via integration
– Obtain C=Q/V 4
8/17/2012
 Parallel Plate Capacitor
• Assume two parallel plates separated by a distance, d, with
charges, +Q and –Q, on them
Q
S 
• The charge density on each
S
 plate is
D   S aˆ x
  Q
E   S aˆ x   aˆ x
 S
  d
Q Qd
V    E  dl     dx 
x 0
S S
Q S
C 
V d 1 1 Q2
WE  CV  QV 
2
C  2 2 2C
 r 
C0  0 1 Q2 Q 2 Sd
WE    2 2 dv 
2v  S 2 2 S 2
Q2  d  Q2 1
    QV
8/17/2012 2  S  2C 2 5
 Coaxial Capacitor
• Assume to cylindrical plates of inner radius a and outer
radius b with +Q and –Q on them
• The charge density on each plate is
 
Q    E  dS  E 2 L
 Q
E aˆ 
2 L
  b
Q
V    E  dl    aˆ   d aˆ 
a
2 L
b
Q Q b
 
a
2 L
d  ln  
2 L  a 
Q 2 L
C 
V b
ln  
a
8/17/2012 6
 Coaxial Line Resistance
From the resistance if a coaxial line from the electric potential and
current density

  L
2 V0
I   J dS    ddz
z 0
0  ln b / a 
2LV0
I
ln b / a 
V ln b / a 
 dV  V0 R 0 
E  V   aˆ   aˆ  2L
d  ln b / a  I
   V0 V 1 ln b / a  Resistance per
J  E  aˆ  R 0 
 ln b / a  I L 2 unit length
apply 1 2
 G 
dS   ddzaˆ  R ln b / a  Conductance

8/17/2012 7
 Spherical Capacitor
• Assume two spherical plates of inner radius a and outer radius b with +Q
and –Q on them
• The charge density on each plate is
 
Q    E  dS  Er 4r 2
 Q
E aˆ
4 r 2 r

  b
Q
V    E  dl   ra 4 r 2 aˆ  draˆ
b
Q Q 1 1
 dr    
r a
4 r 2
4  a b 
Q 4
C 
V 1 1
8/17/2012    8
a b
Series and Parallel Capacitance

Series Capacitance Parallel Capacitance

k k
1 1
 C   Cn
C n 1 Cn n 1
8/17/2012 9
Series and Parallel Capacitance
(Example)

 0 r1S 2 0 r1S 2 0 r 2 S
C1   C2   0 r1S / 2  0 r1S  0 r 2 S
d /2 d d C1   C2 
d 2d 2d
 2 0 S 
2

   0S
C1C2  d    r1 r 2  C  C1  C2   r1   r 2 
C   
C1  C2 2  0 S   r1   r 2  d
d
CC 2 S    
C  1 2  0  r1 r 2 
8/17/2012 C1  C2 d   r1   r 2  10
 Capacitance vs. Resistance
For Common Geometries

  S d
C ,R 

Parallel Plates
V E  dl d S
R    b
I  E  dS ln  
2 L
,R   
a
  C Coaxial Cylinders
b 2 L
Q   E  dS ln  
C    a
V  E  dl 1 1
  
 4
,R  
a b
RC  C Between 2 Spheres
1 1 4
   
a b
1
C  4 a , R  Isolated Sphere
4 a
8/17/2012 11
 Poisson’s and Laplace’s
Equations for Electrostatics
• Solving for the potential, V, using charge density

 
 
  D    E  v    V    v
  v
E  V  2V 

• Uniqueness theorem: Although there are many ways to solve Laplace’s
equation, there is only one solution for any given set of boundary conditions
• General Procedures for solving:
– Laplace’s Eqn.  2V  0 Poison’s Eqn.  V  0
2

• Direct integration of V is a function of only 1 variable

• If V is a function of 2 or more variables, then apply separation of
variables method to reduce dimensionality and solve
– Apply BCs to determine unique solution for V
– Having V, find E,D,J
– Find Q from D: C, from Q and V: and R from V and J

8/17/2012 12
 Uniform Charge Density in
Cylindrical Wire Biased at z=0
Apply Boundary Conditions
  v 0 
2
Vo ( z  0)   A0   Vo
Z  0,V  Vo 2
z  d ,V  0   B  Vo
 v d 2
V z  d   0   Ad  Vo
2
d V
 A v  o
2 d
General 1-D solution in z Find V,E,F
 v   v z 2   v d Vo 
 2V  V    z  Vo
 2  2 d 
 2V   v  V   v z   v d Vo  
 E  V   az      a z
z 2  z    2 d 
dV   v z
 A
dz     v z 2   v d Vo 
d

F    v Edv  S v  z  
 v z 2 2  2 d 0
V  Az  B
8/17/2012 2 S  cross _ sec tional _ area
13
1-D Capacitor with Two Dielectrics
Apply Boundary Conditions
x  d ,V1  0 V  x  d   0  A1d  B1
x  0,V2  0  B1   A1d
xa
 V  x  0  A2 (0)  B2
 V1  A1 x  B1 , x  a
 V2  A2 x  B2 , x  a
 B2  0
V1 ( x  a )  V2 ( x  a ) V ( x  a )  A1a  B1  A2 a
D1n  D2 n   s , x  a  s   1 A1   2 A2
Find E
  S aˆ x
E1   A1aˆ x  ,a  x  d
 2 d 2 
 1 1   
 1 a 1 
d 
General 1-D solution in x   S   1aˆ x
E2   A2 aˆ x  a  ,0  x  a
 2V  0
 2 d 2 
V  Ax  B  1 1   
  1 a 1 
xa    d   Is the weighting fraction of the
 1 1  2   1  dielectric constant for the distance
 V1  A1 x  B1 , x  a  1  a  
8/17/2012 d+a=1
 V2  A2 x  B2 , x  a Can solve for V, and F using algebra 14
 1-D Capacitor with Two Dielectrics
Lets solve the algebra showing that we can get E from V(x=a) and s
  S aˆ x
E1   A1aˆ x  ,a  x  d
V x  d   0  A1d  B1  B1   A1d  2 d 2 
1 1   
V x  0  A2 (0)  B2  B2  0   1 a 1 
V ( x  a)  A1a  B1  A2 a
 s   1 A1   2 A2 S
 A1 
  d  
 1 1  2  2 
V ( x  a)  A1 (a  d )  A2 a  1 a 1 
 s  1 A1   2 A2 V ( x  a )  A1 (a  d )  A2 a
ad  d 
S    S   1
V ( x  a )  A1 (a  d )  A2 a A2   a   a 
  d     d  
ad   1 1  2  2   1 1  2  2 
 s   1 A1   2   A1  1 a 1   1 a 1 
 a 
     a  d   
  A1   1 1   2    
   
   1 a
     d     a    d 
 S   1aˆ x
  A1   1 1   2     2      a 
 
    1  a    1  a    E2   A2 aˆ x 
  d  
,0  x  a
     d        1 1  2  2 
  A1   1 1   2     2     1 a 1 
 
    1  a    1   
8/17/2012 15
 1-D Semi-infinite
Intersecting Planes
Apply Boundary Conditions
V (  0)  0  A(0)  B
B0
V (  0 )  V0  A0
V0
 A
0
V0
V  A  B  
0
Find V and E
 1 V 1 V0
E  V  aˆ  aˆ
   0
General 1-D solution in  
V0  100V , 0  ,   1m
6
1 V2
V
2
 2V   0  0 600
 2  2  2 V  *Volts

V  A  B   600
8/17/2012 E aˆ *Volts / m 16

 2-D conducting cones
(1-D Problem)
Apply Boundary Conditions
V   1   0  A ln tan 1 / 2  B
 B   A ln tan 1 / 2
V    2   V0  A ln tan  2 / 2  A ln tan 1 / 2
 tan  2 / 2   tan 1 / 2 
V0  A ln    A  V0 ln  
 tan 1 / 2   tan  2 / 2 
Find V and E
 tan 1 / 2   tan 1 / 2 
General 1-D solution in  V  V0 ln   ln tan  / 2  V0 ln   ln tan 1 / 2 
 tan  2 / 2   tan  2 / 2 
 tan 1 / 2   tan  / 2 
V  V0 ln   ln  
1   V    V   tan  2 / 2   tan 1 / 2 
2V   sin  0  sin  0
r sin   
2
       tan 1 / 2 
V0 ln 
int egration _ yields  1 V A  tan  2 / 2 
V E  V   aˆ   aˆ   aˆ
sin  A r  r sin  r sin 
 1   / 10, 2   / 6,V0  50V
Further _ int egration _ yields
 tan  / 2 
d d tan  / 2  V  95.1 ln  V
V  A  A  A ln tan  / 2   B  0.1584 
sin  tan  / 2
8/17/2012 
E  V  
95.1aˆ V / m 17
http://en.wikipedia.org/wiki/List_of_integrals_of_trigonometric_functions
r sin 
 Integral of 1/sin(a)
1
From our Trig Identities we know that sin  
csc
d
Thus  sin    csc d

And again from trig identities

d 
Thus  sin   ln tan
2
C

8/17/2012 18
 Integral of 1/sin(a) (Method 2)
d sin d
 sin   sin 2  , let _ u   cos 


du 1  1 1 
     du
1 u 2
2 1 u 1 u 


1
ln 1  u  ln 1  u   C  1 ln 1  u  C
2 2 1 u
1 1  cos  1  cos  1 
 ln  C  ln  C  ln tan  C
2 1  cos  1  cos  2 2

Using the known trig. identity

8/17/2012 19
 Integral of 1/sin(a)
1
From our Trig Identities we know that sin  
csc
d
Thus  sin    csc d

And again from trig identities

d 
Thus  sin   ln tan
2
C

8/17/2012 20
 Spherical Capacitor Using
Laplace’s Equation
Apply Boundary Conditions
r  b, V  0 r  a,V  V0
A
0    B  B  A/b 1 1
V0  A  
b b a
yields
Vo
1 1 A
V  A   1 1
b r    
b a
Find E and C  
Q    E  dS
General 1-D solution in r 1 1
V0    
V 
r b 2  0 rV0
1 d  2 dV  1 1
Q   1 1
0
r 2 sin dd
 2V  r 0    0 r2  
r dr  dr
2
 b a b a
 dV 4 0 rV0
dV E  V   aˆ r Q
r2 A dr 1 1
dr  V0   
E a b
A
V  B 1 1
r2   Q 4 0 r
r C 
8/17/2012 b a V0  1 1  21
  
a b
 Resistance in a Curved
Metal Bar (1)
Solve for E, J, I, R

 dV  V0
E  V   aˆ   aˆ 
d  ln b / a 
   V0
1-D Laplace’s Eqn. in  J  E  aˆ 
 ln b / a 
dV
1 d  dV   A apply
V
2
    0 d 
 d  d  dS   ddzaˆ 
V  A ln   B
  t
 /2 V0
Boundary Conditions I   J dS    ddz
V (   a )  0  B   A ln a z 0
0  ln b / a 
V (   b)  V0  A ln b  B  A ln b  A ln a tV0
I
 A
V0 2 ln b / a 
ln b / a 
V 2 ln b / a 
V  A ln   B  A(ln   ln a )  A ln  / a  R 0 
V ln  / a 
I t
V 0
ln b / a 
22
8/17/2012
 Resistance in a Curved
Metal Bar (2)
Solve for E, J, I, R

 dV  V0
E  V   aˆ z  aˆ z
dz t
   V0
J  E  aˆ z
t
1-D Laplace’s Eqn. in z apply
d 2V

 V  2 0
2 dS   ddaˆ z
dz
V  Az  B
  b
 /2 V0
I   J dS    dd
0 t
Boundary Conditions a

V ( z  0)  0  A0   B V0 b 2  a 2 
B0 V I
V 0z 4t
V ( z  t )  V0  At t V0 4t
R 
 A 0
V 
I  b 2  a 2 
t 23
8/17/2012
 2-D Rectangular Waveguide
Separation of Variables
V  x, y   X ( x )Y ( y )
X ' 'Y  Y ' ' X  0
X '' Y ''
  
X Y

Where  is separation constant

Laplace equation for
X ' '  X  0
Cartesian coordinates in 2-D
 2V  2V Y ' '  Y  0
V  2  2 0
2

x y Separated Boundary Conditions

Boundary Conditions V 0, y   X (0)Y ( y )  0  X (0)  0
V  x  0,0  y  a   0 V b, y   X (b)Y ( y )  0  X (b)  0
V  x  b,0  y  a   0 V  x,0   X ( x )Y (0)  0  Y (0)  0
V 0  x  b, y  0   0 V  x, a   X ( x )Y ( a )  V0
8/17/2012 V 0  x  b, y  a   V0 24
Vo term is inseparable
 2-D Rectangular Waveguide
The separated set of differential Case number 2
equations has three possible   0     2
solutions:
X ' ' 2 X  0
case 1: =0
case 2: <0  D 2   2 X  0
case 3: >0 d
D
dx
Case number 1 DX  X
X ''
 0 Gives 2 solutions
X
Integrating twice yields dX
X  Ax  B  X   dx
Applying Boundary Conditions
ln X  x  ln A1
X x  0  0  0  B
X x  b  0  A  b  0  0  A  X  A1ex
Yields the trivial Solution The second solutions is
X ( x)  0
8/17/2012
 X  A2 ex 25
 2-D Rectangular Waveguide
Case number 2
where
X ( x)  A1ex  A2 ex
X ( x)  B1 cosh x  B2 sinh x
B1  A1  A2
B2  A1  A2
X  x  0   0  B1 1  B2  0
 0  B1
X  x  b   0  B1 cosh x  B2  sinh b
 0  B2  sinh b
 B2  0
X ( x)  0 Again, the trivial solution http://en.wikipedia.org/wiki/Hyperbolic_function

8/17/2012 26
 2-D Rectangular Waveguide
Case number 3
 0  2 X  x  0   0  g 0 1  g1  0
 0  g0
X ' '  2 X  0
X  x  b   0  g 0 cos x  g1  sin b

 D  X 0
2 2
  0  g1  sin b
d
D  0  g1
dx
DX   j X for _ g1  0, trivial _ but :
sin  b  0  sin n  b  n
X ( x )  C0 e jx  C1e  jx
n
X ( x )  g 0 cos  x  g1 sin  x  , n  1,2,3,... Modal periodic
b solutions allowed
g 0  C0  C1
 nx 
g1  C0  jC1 X n ( x)  g n sin  
 b 

8/17/2012 27
http://en.wikipedia.org/wiki/Sine_wave
 2-D Rectangular Waveguide
Case number 3 (cont.) Case number 3
Solution for Y

 nx 
X n ( x)  g n sin  
 b   ny 
Yn ( y )  hn sinh 
n 2 2  b 
 0    2 2

b
 nx   ny 
Y ' '  2Y  0 V  x, y   g n hn sin  sinh  
 b   b 
Y ( y )  h0 cosh  y  h1 sinh  y 
 nx   ny 
Y  y  0   0  g 0 1  g1  0 V  x, y    cn sin  sinh 
n 1  b   b 
 0  g0 
 nx   na 
Y  y  a   V0  h1 sinh a V  x, y  a   V0   cn sin  sinh 
n 1  b   b 
 n 
V0  h1 sinh  a
 b 

8/17/2012 28
 2-D Rectangular Waveguide
Case number 3 (cont.) Solution for Y

 nx   na 
V  x, y  a   V0   cn sin  sinh 
n 1  b   b 
 nx   na  2  nx 
b b

b
 m x  
 n a 
b
 mx   nx 
0 V0 sin 
 b 
dx  c n sinh 
 b 0
   b dx
sin
0 0  b  
V sin dx  cn sinh   sin
 b 0  b   b 
 sin dx
b
n 1
 nx   na  1   2nx  
b
 V0 cos   cn sinh    1  cos  dx
 b 0  b 2 0  b 
 0, m  n
1  cos n   cn sinh na  b
b

0 sinmxsinnx dx   , m  n

V0b
2 n  b 2
 4V0
 na  2V0
cn sinh   1  cos n    n , n  1,3,5,..
 b  n  0, n  2,4,6,...
 4V0
 , n  odd
 na 
cn   n sinh  
  b 
 0, n  even
 nx   ny 

sin   sinh  
V  x, y  
4V0  b   b 

 n odd  na 
8/17/2012 n sinh   29
 b 
 2-D Rectangular Waveguide
Solutions using the solved PDE
 3x    nx   ny 
V ( y  a )  V0  10 sin    cn sin  sinh 
Given an applied bias condition,  b  n 1  b   b 
one can find the potential  10
 ,n  3
 3a 
throughout the waveguide using cn   sinh 
  b 
the relations presented here  0, n  3
 3y 
sinh 
 3x   b 

 nx   n y  V  x, y   10 sin 
V  x, y    cn sin  sinh   b  sinh 3a 
 b   b   
n 1
 b 
 4V0
 , n  odd  x  1  5x    n x   n y 
 n a  V ( y  a )  V0  2 sin   sin    cn sin  sinh 
cn   n sinh   b  10  b  n 1  b   b 
  b   2
 0, n  even  ,n  1
 a 
 sinh  
 b 
 nx   n y  

 sin   sinh  cn   0, n 1,5

V  x, y  
4V0  b   b  

 n odd  n a 

1
10 sinh 5a 
,n  5
n sinh    b 
 b 
 y   5y 
sinh  sinh 
 x   b   5x   b 
8/17/2012 V  x, y   2 sin   sin  30
 b  sinh a   b  10 sinh 5a 
   
 b   b 
 3-D Cylindrical Separation
of Variables
1 V  V  1  2V  2V  V  R  2 2 1  2
2V      2 2  2  0        0
       z R       2

V  ,  , z   R(  ) ( ) Z ( z )
 V  R  2 2 1  2
 Z   R  RZ  2 2Z            2

V 
2
    2  R 2  0 R        2
       2
z
1   R  1  2 1 2Z  ' '   2  0
    2   2
R        2
Z z 2
Yielding
Z ' '2 Z  0  2 R' ' R'  22   2 R  0
Pr ovides _ V _ u sin g
Z ( z )  c1 cosh z   c2 sinh z 
 ( )  c3 cos    c3 sin  
8/17/2012 R  Bessel _ function 31
 Method of Images
• Image theory: A given charge configuration above an infinite grounded perfect
conducting plane may be replaced by a mirror image of the charge configuration and
an equipotential line in place of the conducting plane
• This theory is of significant importance b/c it allows one to significantly simplify
complex problems using symmetry

8/17/2012 32
 Method of Images
• Symmetry allows us to examine the total
force at point A (+Q) by placing conducting
ground mirrors half way between A and its
nearest neighbors
• The result is a method of images approach x
PA
that allows one to quickly determine the exact
force acting on charge +Q located at point A
given either y
– The relative orientation of a grounded
conducting surface near A
– The number and location of charges
near A allowing one to quickly establish
a conducting ground plane evaluation
 360o  Number of charges
N    1 applying force on +Q at
   A
N
F   Fi
i 1

F  kq
N
qi rˆi

 xˆ
2

yˆ

2 x xˆ  2 y  yˆ  x cos90   xˆ   y sin 90   yˆ  x cos90   xˆ  2 y sin  yˆ 
     
kq
ri 2  2 x 2 2 y 2 2 x 2  2 y 2 3 / 2  x cos 90   2   y sin 90   2 3 / 2 2 x 2  2 y 2  8 xy cos 2 2 3 / 2 
i 1
 
8/17/2012 33
 Method of Images

 360o   360o 
N    1   o  1  3
    90 
N
qi rˆi
N
F   Fi  kq 2  kq
 ˆ
2  x

 yˆ

2 x xˆ  2 y  yˆ 
i 1 r
i 1 i

 2 x 2
2 y  2
2 
x 2
 2 y  
2 3/ 2 

8/17/2012 34