Are you sure?
This action might not be possible to undo. Are you sure you want to continue?
Electrodynamics
Russell Bloomer
1
University of Virginia
Note: There is no guarantee that these are correct, and they should not be copied
1
email: rbloomer@virginia.edu
Contents
1 Problem Set 1 1
1.1 Problem 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Problem 2: Jackson 1.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Problem 3: Jackson 1.6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4 Problem 4: Jackson 1.10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Problem Set 2 7
2.1 Problem 1: Jackson 1.7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Problem 2: Jackson 1.9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3 Problem 3: Jackson 2.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4 Problem 4: Jackson 2.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3 Problem Set 3 13
3.1 Problem 1: Jackson 2.8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2 Problem 2: Jackson 2.10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
3.3 Problem 3: Jackson 2.13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.4 Problem 4: Jackson 2.23 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4 Problem Set 4 19
4.1 Problem 1: Jackson 3.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.2 Problem 2: Jackson 3.6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.3 Problem 3: Jackson 3.10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5 Problem Set 5 23
5.1 Problem 1: Jackson 3.17 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5.2 Problem 2: Jackson 4.7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
6 Problem Set 6 27
6.1 Problem 1: Jackson 4.9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
6.2 Problem 2: Jackson 4.10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
6.3 Problem 3: Distressed Simple Cubic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
7 Problem Set 7 31
7.1 Problem 1: Jackson 4.13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
8 Problem Set 8 33
8.1 Problem 1: Jackson 5.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
8.2 Problem 2: SemiInﬁnite Cylinder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
8.3 Problem 3: Jackson 5.6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
8.4 Problem 4: Jackson 5.13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
i
9 Problem Set 9 39
9.1 Problem 1: Jackson 5.17 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
9.2 Problem 2: Jackson 5.19 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
9.3 Problem 3: Jackson 5.22 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
10 Problem Set 10 45
10.1 Problem 1: Jackson 5.14 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
10.2 Problem 2: Jackson 5.27 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
10.3 Problem 3: Jackson 5.29 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
11 Problem Set 11 51
11.1 Problem 1: Jackson 6.8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
11.2 Problem 2: Jackson 6.20 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
11.3 Problem 3: Jackson 6.11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
12 Problem Set 12 55
12.1 Problem 1: Jackson 6.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
12.2 Problem 2: Jackson 6.14 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
A Special Functions 61
ii
Chapter 1
Problem Set 1
1.1 Problem 1
Consider a vacuum diode consisting of two parallel plates of area A which are large compared to their separation
d. Neglecting edge eﬀects, all quantities will be assumed to depend only on x a coordinate perpendicular to the plates.
Electrons are “boiled” oﬀ the cathode plate which is heated and held at potential Φ = 0. They are attracted
to the anode plate which is held at a positive potential Vo. The cloud of electrons within the gap (called the space
charge) quickly builds up to a point where it reduces ﬁeld at the cathode to zero. From then on a steady current
ﬂows between the plates.
Show that the current in the diode is given by
I = KV
3/2
o
where K =
4oA
9d
2
2e
m
(1.1)
The charge density is
ρ = −o∇
2
V →ρ = −o
∂
2
V
∂x
2
(1.2)
because the charge density only depends on the position x. Next the velocity of the electron at any position in the
potential.
1
2
mv
2
= eV →v =
2eV
m
(1.3)
Because this is a steady state, the current is independent of the position x.
dq = ρAdx →
dq
dt
= ρA
dx
dt
→I = ρAv (1.4)
Combining equations 3 and 4
1
I = ρA
2eV
m
→ρ =
I
A
m
2eV
Now adding in equation 2
ρ =
I
A
m
2eV
→
∂
2
V
∂x
2
= −
I
oA
m
2eV
All that remains is to solve the diﬀerential. Start by deﬁning W =
∂V
∂x
. The diﬀerential becomes
W
∂W
∂x
= −
I
oA
2
me
V
−1/2
∂V
∂x
⇒W∂W = −
I
oA
2
me
V
−1/2
∂V
1
If the ‘e’ on equation is lower case, it is from the homework set, but if it is upper case, it is from ‘Jackson’ 3
rd
ed.
1
Integrating once to ﬁnd
1
2
W
2
= −
2I
oA
2
me
V
1/2
Solving for W
1
2
W
2
= −
2I
oA
2
me
V
1/2
→W =
8m
e
I
oA
2
1/4
V
1/4
Now reintroduce W =
∂V
∂x
∂V
∂x
=
8m
e
I
oA
2
1/4
V
1/4
⇒V
−1/4
∂V =
8m
e
I
oA
2
1/4
∂x
Integrating
V
−1/4
∂V =
8m
e
I
oA
2
1/4
∂x ⇒
4
3
V
3/4
=
8m
e
I
oA
2
1/4
x +C
The ﬁrst boundary condition is at V (0) = 0, which determines the constant to be C = 0. The other boundary
condition is at V (d) = Vo. The equation becomes
4
3
V
3/4
o
=
8m
e
I
oA
2
1/4
d
Solving for the current
4
3
4
V
3
o
=
8m
e
I
oA
2
d
4
⇒
I
oA
2
=
e
8md
4
4
3
4
V
3
o
⇒
I
oA
=
e
8m
4
3d
2
V
3/2
o
I =
4oA
9d
2
2e
m
V
3/2
o
→I = KV
3/2
o
1.2 Problem 2: Jackson 1.5
The timeaveraged potential of a neutral hydrogen atom is given by
Φ =
q
4πo
e
−αr
r
1 +
αr
2
(1.5)
where q is the magnitude of the electric charge, and α
−1
= ao/2, ao being the Bohr radius. Find the distribution of
charge (both continuous and discrete) that will give this potential and interpret your result physically.
The charge density is related to the potential by
ρ = −o∇
2
Φ →ρ = −o∇
2
¸
q
4πo
e
−αr
r
1 +
αr
2
= −
q
4π
∇
2
¸
e
−αr
r
1 +
αr
2
The Laplace in spherical coordinates for a spherically symmetrical object is
∇
2
=
1
r
2
∂
∂r
r
2
∂
∂r
2
Then the charge density becomes
ρ = −
q
4π
1
r
2
∂
∂r
r
2
∂
∂r
¸
e
−αr
r
1 +
αr
2
= −
q
4π
1
r
2
∂
∂r
r
2
∂
∂r
¸
e
−αr
r
+
αe
−αr
2
= −
q
4π
1
r
2
∂
∂r
¸
−αr
2
2
e
−αr
+r
2
e
−αr
∂
∂r
1
r
−
αr
2
2
e
−αr
= −
q
4π
1
r
2
¸
−α
2
re
−αr
+
α
3
r
2
2
e
−αr
−αe
−αr
+α
2
r
2
e
−αr
−αr
2
e
−αr
∂
∂r
1
r
+e
−αr
∂
∂r
r
2
∂
∂r
1
r
= −
q
4π
¸
α
3
2
e
−αr
−
α
r
2
e
−αr
+
αr
2
r
2
e
−αr
r
2
+e
−αr
1
r
2
∂
∂r
r
2
∂
∂r
1
r
With ∇
2
1
r
= −4πδ(r), the factor in front of the delta function is 1 at the origin and 0 everywhere else. The above
equation reduces to
ρ =
−q
4π
¸
α
3
2
e
−αr
−4πδ(r)
= qδ(r) −
qα
3
8π
e
−αr
(1.6)
The ﬁrst term is a positive point charge at the origin, which would be the proton. The second term is a negative
spherically symmetry object that is decaying oﬀ rapidly after the Bohr radius, which is a good description of the
electric cloud.
1.3 Problem 3: Jackson 1.6
A simple capacitor is a device formed by two insulated conductors adjacent to each other. If equal and opposite charges
are placed on the conductors, there will be a certain diﬀerence potential between them. The ratio of magnitude of
the charge on one conductor to magnitude of the potential diﬀerence is called the capacitance. Using Gauss’s law,
calculate the capacitance of
(a) two large, ﬂat, conducting sheets of area A, separated by a small distance d
The electric ﬁeld outside the plate two plates is zero, because if pillbox containing both plates has no charge therefore
no electric ﬁeld. Forming a pillbox around one plate by Gauss’s law has [E[ = σ/2o, where σ = Q/A. There are two
plates adding constructively in between, so [E[ = σ/o. The potential is then
E =
V
d
=
σ
o
V =
σd
o
→V =
Qd
Ao
The capacitance is then
C =
Q
V
=
Q
Qd
Ao
→C =
Ao
d
(b) two concentric conduction spheres with radii a, b (b > a)
Gauss’s law for a sphere is
E =
Q
4πo
1
r
2
The potential is then
V = −
a
b
ˆ
E d
ˆ
l = −
Q
4πo
a
b
dr
r
2
=
Q
4πo
1
r
a
b
=
Q
4πo
1
a
−
1
b
Now the capacitance
C =
Q
V
=
Q
Q
4πo
1
a
−
1
b
= 4πo
ab
b −a
3
(c) two concentric conducting cylinders of length L, large compared to their radii a, b (b > a)
Using Gauss’s law for cylindrical surface is
E =
Q
2πo
1
r
The potential becomes
V = −
b
a
ˆ
E d
ˆ
l = −
Q
2πo
b
a
dr
r
= −
Q
2πo
ln r
b
a
=
Q
2πo
(ln b −ln a) =
Q
2πo
ln (b/a)
The capacitance is
C =
Q
V
=
Q
Q
2πo
ln (b/a)
=
2πo
ln (b/a)
(1.7)
(d) What is the inner diameter of the outer conductor in an airﬁlled coaxial cable whose center conductor is a
cylindrical wire of diameter 1 mm and whose capacitance is 3 10
−11
F/m? 3 10
−12
F/m?
Using equation 7, and rearranging it
b = ae
2πo
C
For 3 10
−11
F/m
b = 1 e
2π·8.85×10
−12
/3×10
−11
= e
1.854
= 6.38 mm
For 3 10
−12
F/m
b = 1 e
2π·8.85×10
−12
/3×10
−12
= e
18.54
= 1.12 10
8
mm
1.4 Problem 4: Jackson 1.10
Prove the mean value theorem: For chargefree space the value of the electrostatic potential at any point is equal to
the average of the potential over the surface of any sphere centered on that point.
From Equation 1.36
Φ(x) =
1
4πo
V
ρ(x
)
R
d
3
x
+
1
4π
S
¸
1
R
dΦ
dn
−Φ
d
dn
1
R
da
(1.8)
For a chargefree space ρ = 0, then
1
4πo
V
ρ(x
)
R
d
3
x
=
1
4πo
V
0
R
d
3
x
= 0 (1.9)
Using Equation 1.33 for the second part of equation 8
1
R
dΦ
dn
=
1
R
∇Φ ˆ n
From equation 1.16
1
R
∇Φ ˆ n
= −
1
R
E ˆ n
The second part of equation 8 becomes
1
4π
S
1
R
dΦ
dn
da
= −
1
4π
S
1
R
E ˆ n
da
4
Because this is over the surface it is independent of R, then
−
1
4πR
S
E ˆ n
da
By the divergence theorem
−
1
4πR
S
E ˆ n
da
=
1
4πoR
V
ρ(x)d
3
x
Again in a chargefree space
1
4πoR
V
ρ(x)d
3
x
= 0 (1.10)
For the last part of equation 8,
d
dn
1
R
= −
1
R
2
, because the diﬀerential is before the surface. At the surface the
radius is constant so the last part of equation 8 becomes
−
1
4π
S
Φ
d
dn
1
R
da
=
1
4πR
2
S
Φda
(1.11)
So combining equations 9, 10, and 11 to ﬁnd
Φ(x) =
1
4πR
2
S
Φda
5
6
Chapter 2
Problem Set 2
2.1 Problem 1: Jackson 1.7
Two long, cylindrical conductors of radii a1 and a2 are parallel and separated by a distance d, which is large compared
with either radius. Show that the capacitance per unit length is given approximately by
C · πo
ln
d
a
−1
(2.1)
where a is the geometrical mean of the two radii. Approximate what gauge wire (state diameter in millimeters) would
be necessary to make a twowire transmission line with capacitance of 1.2 10
−11
F/m in the separation of the wires
was 0.5 cm? 1.5 cm? 5.0 cm?
The electric ﬁeld for a wire is given by
E =
Q
2πo
ˆ r
r
The potential between the two wire is
Φ =
E d
l =
Q
2πo
¸
d−a
2
a
1
dr
r
+
d−a
1
a
2
dr
r
=
Q
2πo
¸
ln
d −a2
a1
+ ln
d −a1
a2
=
Q
2πo
¸
ln
(d −a2) (d −a1)
a1a2
The capacitance is given by C =
Q
Φ
. Then
C =
Q
Φ
=
Q
Q
2πo
ln
(d−a
2
)(d−a
1
)
a
1
a
2
=
2πo
ln
(d−a
2
)(d−a
1
)
a
1
a
2
Since the distance between the wires is much large than their radii, d−a1 ≈ d−a2 ≈ d. Then capacitance reduces to
C =
2πo
ln
(d−a
2
)(d−a
1
)
a
1
a
2
≈
2πo
ln
d
2
a
1
a
2
=
πo
ln
d
√
a
1
a
2
Deﬁning the average radius as a =
√
a1a2. The capacitance is then
C =
πo
ln
d
√
a
1
a
2
=
πo
ln
d
a
7
For C = 1.2 10
−11
F/m and d = 0.5cm the radius is
1.2 10
−11
=
πo
ln
.005
a
→a = .49mm
The diameter is .98 mm. Now for d = 1.5cm
1.2 10
−11
=
πo
ln
.015
a
→a = 1.5mm
The diameter is 3 mm. Lastly for d = 5.0cm
1.2 10
−11
=
πo
ln
.050
a
→a = 4.9mm
The diameter is 9.8 mm.
2.2 Problem 2: Jackson 1.9
Calculate the attractive force between conductors in the parallel plate capacitor (Problem 1.6a) and parallel cylinder
capacitor (Problem 1.7) for
(a) ﬁxed charges on each conductor
For parallel plate capacitor, the work is
W =
1
2
CV
2
(2.2)
From the last problem set, C =
Ao
d
, and V =
Qd
Ao
. Therefore
W =
1
2
Ao
d
Qd
Ao
2
=
1
2
Q
2
d
Ao
Force is given by
F = −
∂W
∂l
Q
+
∂W
∂l
V
(2.3)
Because the charge is constant
∂Q
∂l
= 0, then the force is
F = −
∂W
∂l
= −
∂
∂l
1
2
Q
2
d
Ao
= −
1
2
Q
2
Ao
For parallel cylinder capacitor, the work is given by equation 2, with C = πo
ln
d
a
−1
and V =
Q
πo
ln
d
a
. The work
is then
W =
1
2
πo
ln
d
a
−1
Q
πo
ln
d
a
2
=
1
2
Q
2
πo
ln
d
a
Because the charge is constant, the force becomes from equation 3
F = −
∂
∂l
1
2
Q
2
πo
ln
d
a
= −
1
2
Q
2
πod
(b) ﬁxed potential diﬀerence between conductors.
For parallel capacitors, the charge in terms of potential is
V =
Qd
Ao
→Q =
AoV
d
The work becomes
W =
1
2
Q
2
d
Ao
→W =
1
2
oAV
2
d
8
From equation 3, the force is
F =
∂W
∂l
V
= −
1
2
oAV
2
d
2
For parallel cylinder capacitor, the charge in terms of potential is
V =
Q
πo
ln
d
a
→Q =
πoV
ln
d
a
The work becomes
W =
1
2
Q
2
πo
ln
d
a
→W =
1
2
πoV
2
ln
d
a
From equation 3, the force is
F =
∂W
∂l
V
=
∂
∂l
1
2
πoV
2
ln
d
a
= −
1
2
πoV
2
d
ln
d
a
2
2.3 Problem 3: Jackson 2.1
A point charge q is brought to a position a distance d away from inﬁnite plane conductor held at zero potential.
Using the method of images ﬁnd:
(a) the surfacecharge density induced on the plane, and plot it.
From the method of images
Φ(x) =
q/4πo
[x −y[
+
q
/4πo
[x −
y
[
(2.4)
Because it is an inﬁnite plane conductor, by symmetry of problem y =
d and
y
= −
d when deﬁning the capacitor on
the xy plane. The charge is then also q
= −q. The potential becomes
Φ(x) =
q/4πo
[x −
d[
−
q/4πo
[x +
d[
Deﬁning the potential in cylindrical coordinates
Φ(ρ, θ, z) =
q/4πo
(z −d)
2
−ρ
2
1/2
−
q/4πo
(z +d)
2
+ρ
2
1/2
The surfacecharge density induced on the plane is when z = 0
σ = −o
∂
∂z
Φ
z=0
=
∂
∂z
¸
−q
4π
(z −d)
2
−ρ
2
1/2
−
−q
4π
(z +d)
2
+ρ
2
1/2
¸
z=0
= −
q
4π
¸
2(z −d)
−1
2
(z −d)
2
+ρ
2
3/2
−
2(z +d)
−1
2
(z +d)
2
+ρ
2
3/2
¸
z=0
= −
q
4π
¸
2d
(d
2
+ρ
2
)
3/2
¸
= −
q
2πd
2
1
1 +
ρ
d
2
3/2
See ﬁgure 1.
(b) the force between the plane and the charge by using Coulomb’s law for the force between the charge and its
image.
9
Figure 2.1: s(r) = −1/(1 +r
2
)
3/2
The force between the charge and the image is
F =
q1q2
4πo
x1 −x2
[x1 −x2[
3
The separation x1 −x2 = −2dˆ z. The charges are q1 = −q2 = q. The force becomes
F =
q(−q)
4πo
−2dˆ z
[ −2d[
3
=
q
2
4πo
ˆ z
(2d)
2
=
q
2
ˆ z
16πod
2
(c) the total force acting on the plane by integrating σ
2
/2o over the whole plane.
The force is
dF =
σ
2
2o
da →dF =
1
2o
¸
¸
q
2πo
1
1 +
ρ
d
2
3/2
¸
2
da
=
q
2
8π
2
od
4
1 +
ρ
d
2
3
da
Integrating over the entire plane
F =
q
2
8π
2
od
4
2π
0
∞
0
1
1 +
ρ
d
2
3
ρdρdθ
= 2π
q
2
8π
2
od
4
∞
0
1
1 +
ρ
d
2
3
ρdρ
=
q
2
4πod
4
−d
6
4(ρ
2
+d
2
)
2
∞
0
=
q
2
16πod
2
(d) the work necessary to remove the charge q from its position to inﬁnity.
From part b, the force is
F =
q
2
ˆ z
16πoz
2
So the work is the amount of force to move a particle a distance, then the work is
W =
∞
d
F dˆ z =
∞
d
q
2
dz
16πoz
2
=
q
2
16πo
−1
z
∞
d
=
q
2
16πod
(e) the potential energy between the charge q and its image [compare the answer to part d and discuss]
10
The work is given by
W =
1
2
1
4πo
¸
i,j,i=j
qiqj
[xi −xj[
For this system,
W =
1
2
1
4πo
¸
−q
2
[ −2d[
+
q
2
[2d[
= −
q
2
8πod
The diﬀerence between part d and e is that in d the image charge is being moved, and in e the image charge is
stationary.
(f ) ﬁnd the answer to part d in electron volts for an electron originally on angstrom from the surface.
In part d,
W =
q
2
16πod
Then for this case
e 1.60 10
−19
16π (8.85 10
−12
) 10
−10
= 3.60eV
2.4 Problem 4: Jackson 2.2
Using the method of images, discuss the problem of a point charge q inside a hollow, grounded, conducting sphere of
inner radius a. Find
(a) the potential inside the sphere;
The potential is
Φ(x) =
1
4πo
¸
q
[x −y[
+
q
[x −y
[
Now the boundary condition
Φ(x = a) = 0 =
¸
q/4πo
a[ − ˆ n +y/aˆ n
[
+
q
/4πo
y
[ − ˆ n
+a/y
ˆ n[
This is identical to the charge on the outside of the sphere, because the magnitude is the concern. Therefore q
= −q
a
y
and y
=
a
2
y
. Then the potential is
Φ(x) =
1
4πo
¸
q
[x −y[
+
q
[x −y
[
(b) the induced surfacecharge density;
The surfacecharge density is given by
σ = o
∂Φ
∂x
x=a
because it is opposite the normal to the inside. So
σ =
∂
∂x
¸
oq/4πo
(x
2
+y
2
−2xy cos γ)
1/2
+
oq
/4πo
(x
2
+y
2
−2xy
cos γ)
1/2
¸
=
1
4π
¸
q(−1/2)(2x −2y cos γ)
(x
2
+y
2
−2xy cos γ)
3/2
+
q
(−1/2)(2x −2y
cos γ)
(x
2
+y
2
−2xy
cos γ)
3/2
¸
=
q
4πa
y
2
−a
2
(y
2
+a
2
−2ay cos γ)
3/2
11
(c) the magnitude and direction of the force acting on q.
The separation between the charge and its image y −y
, so the magnitude is
[
F[ =
1
4πo
qq
(y −y
)
2
=
1
4πo
q
2 a
y
(y −
a
2
y
)
2
=
1
4πo
q
2
ay
(a
2
−y
2
)
The direction is radially towards the image charge, because it is attractive.
(d) Is there any change in the solution if the sphere is kept at a ﬁxed potential V ? If the sphere has a total charge
Q on its inner and outer surfaces?
Because this is inside of a sphere, there is no change if there is a ﬁxed potential or a charge on the sphere. This is
due to spherical symmetry. The only change will be a constant,V , add to the potential
Φ(x) =
1
4πo
¸
q
[x −y[
+
q
[x −y
[
+V
12
Chapter 3
Problem Set 3
3.1 Problem 1: Jackson 2.8
A twodimensional potential is deﬁned by two straight parallel line charges separated by a distance R with equal and
opposite linear charge densities λ and −λ,
(a) Show by direct construction that the surface of constant potential V is a circular (circle in the transverse
dimensions) and ﬁnd the coordinates of the axis of the cylinder and its radius in terms of R, λ, and V .
The potential for a line charge is
Φ(r) =
λ
2πo
ln
r
r
= V (3.1)
Solving for r
,
r
r
2
= e
4πoV/λ
. But from the geometry r
2
= (r −
R)
2
. Now
r
2
= r
2
e
4πoV/λ
→(r −
R)
2
= r
2
e
4πoV/λ
From this
rmin =
R
1 +e
2πoV/λ
rmax =
R
1 −e
2πoV/λ
This yields a radius of radius =
R
2 sinh(2πoV/λ)
. Now the oﬀset will be
rcenter = −
R
e
2πoV/λ
−1
The circle becomes with the axis of the cylinder in the ˆ z direction
r +
R
e
2πoV/λ
−1
2
=
R
2 sinh (2πoV/λ)
2
(b) Use the results of part a to show that the capacitance per unit length C of two rightcircular conductors, with
radii a and b, separated by a distance d > a +b, is
C =
2πo
cosh
−1
d
2
−a
2
−b
2
2ab
(3.2)
In this case deﬁne d = R +d1 +d2, where
d1 =
R
e
V
1
/λ
; d2 =
R
e
−V
2
/λ
13
where V1 = 4πoVa and V2 = 4πoV
b
. From the previous part the radii can be deﬁned as
a =
Re
V
1
/λ
e
V
1
/λ
−1
; b =
Re
−V
1
/λ
e
−V
1
/λ
−1
Then
d
2
−a
2
−b
2
=
R
2
e
(V
1
−V
2
)/λ
+ 1
(e
V
1
/λ
−1) (e
−V
2
/λ
−1)
Then
d
2
−a
2
−b
2
2ab
=
e
V
1
/λ
−1
e
−V
2
/λ
−1
2R
2
e
V
1
/λ
e
−V
2
/λ
R
2
e
(V
1
−V
2
)/λ
+ 1
(e
V
1
/λ
−1) (e
−V
2
/λ
−1)
=
e
(V
1
−V
2
)/2λ
+e
−(V
1
−V
2
)/2λ
2
= cosh
V1 −V2
2λ
Then the potential per length becomes
d
2
−a
2
−b
2
2ab
= cosh
V1 −V2
2λ
⇒2 cosh
−1
d
2
−a
2
−b
2
2ab
=
V1 −V2
λ
The capacitance per unit length becomes
C =
λ
V1 −V2
=
1
2 cosh
−1
d
2
−a
2
−b
2
2ab
=
2πo
2 cosh
−1
d
2
−a
2
−b
2
2ab
3.2 Problem 2: Jackson 2.10
A large parallel plate capacitor is made up of two plane conducting sheets with separation D, one of which has a small
hemispherical boss of radius a on its inner surface (D a). The conductor with the boss is kept at zero potential,
and the other is at a potential such that far from the boss the electric ﬁeld between the plates is Eo.
(a) Calculate the surfacecharge densities at an arbitrary on the plane and on the boss, sketch their behavior as a
function of distance (or angle).
This is similar to a conducting sphere in an electric ﬁeld, which from class is
Φ(r, θ) = −Eo
r −
a
3
r
2
cos θ
For 0 ≤ θ ≤ π/2, the potential can be written in Cartesian coordinates
Φ(x, y, z) = −Eoz +
Eoa
3
z
(x
2
+y
2
+z
2
)
3/2
The plate is
σ(x, y) = oEz = −o
∂Φ
∂z
z=0
= oEo
¸
1 −
a
3
(x
2
+y
2
)
3/2
¸
For the case of the boss,
σ(r, θ) = oEr = −o
∂Φ
∂r
r=a
= 3oEo cos θ
14
Figure 3.1: σ/
o
E
o
= 1 −1/x
3
(b) Show that the total charge on the boss has the magnitude 3πoEoa
2
.
For the total charge on the boss is found from the integral over the area.
Q =
2π
0
π/2
0
3oEo cos θa
2
sin θdθdφ
=
2π
0
1
2
oEoa
2
dφ = 3πoEoa
2
(c) If, instead of the other conducting sheet at a diﬀerent potential, a point charge q is placed directly above the
hemispherical boss at a distance d from its center , show that the charge induced on the boss is
q
= −q
¸
1 −
d
2
−a
2
d
√
d
2
+a
2
(3.3)
The potential is
Φ(r) =
1
4πo
¸
q
[r −
d[
−
qa/d
[r −a
2
/d
d[
−
q
[r +
d[
+
qa/d
[r +a
2
/d
d[
¸
Now in terms of r and θ
Φ(r, θ) =
1
4πo
¸
q
√
r
2
+d
2
−2rd cos θ
−
qa/d
r
2
+ (a
2
/d)
2
−2r(a
2
/d) cos θ
−
q
√
r
2
+d
2
+ 2rd cos θ
+
qa/d
r
2
+ (a
2
/d)
2
−2r(a
2
/d) cos θ
Now the surfacecharge density needs to be found
σ = −o
∂Φ
∂r
r=a
=
q
4π
d
2
−a
2
a
2
¸
1
(a
2
+d
2
−2ad cos θ)
3/2
−
1
(a
2
+d
2
+ 2ad cos θ)
3/2
15
So the induced charge on the boss is
q
=
π/2
0
q
4π
d
2
−a
2
a
2
¸
1
(a
2
+d
2
−2ad cos θ)
3/2
−
1
(a
2
+d
2
+ 2ad cos θ)
3/2
2πa
2
sin θdθ
=
−qa
2
d
2
−a
2
2a
¸
1
ad
1
d −a
−
1
√
a
2
+d
2
+
1
a +d
−
1
√
a
2
+d
2
= −q
¸
1 −
d
2
−a
2
d
√
a
2
+d
2
¸
3.3 Problem 3: Jackson 2.13
(a) Two halves of a long hollow cylinder of inner radius b are separated by a small lengthwise gap on each side, and
are kept at diﬀerent potentials V1 and V2. Show that the potential inside is given by
Φ(ρ, φ) =
V1 +V2
2
+
V1 −V2
π
tan
−1
2bρ
b
2
−ρ
2
cos φ
(3.4)
The starting point of this is Equation 2.71. In this case, the origin is located inside the cylinder, so bn = 0 to prevent
divergence at the origin. Equation 2.71 becomes
Φ(ρ, φ) = a0 +
∞
¸
n=1
anρ
n
sin (nφ +αn)
Here Φ(ρ, θ) = Φ(ρ, −θ), so αn = 0. For this case, a0
3π/2
−π/2
Φ(b, φ) = 2πa0 = πV1 +πV2 →a0 =
V1 +V2
2
Now for the remaining terms in the sum
3π/2
−π/2
cos nθ cos mθdθ = δmnπ
Then for Φ
3π/2
−π/2
V (θ) cos mθdθ =
¸
n
anb
n
3π/2
−π/2
cos nθ cos mθdθ = anb
n
δmnπ
The coeﬃcients become
an =
V1
b
n
π
π/2
−π/2
cos(nθ)dθ −
V2
b
n
π
3π/2
π/2
cos(nθ)dθ
=
2(V1 −V2)
b
n
π
sin (nπ/2) =
2(V1 −V2)
nπb
m
(−1)
n+3
2
Because of Φ’s symmetry only odd values for n are allowed. The potential becomes
Φ(ρ, θ) = −i
2(V1 −V2)
π
¸
n
ρ
b
n
i
n
n
cos(nθ)
= −i
2(V1 −V2)
π
iIm
¸
n odd
Z
n
n
where Z = i(ρ/b)e
iθ
16
¸
n odd
Z
n
n
=
1
2
ln
1+Z
1−Z
1
. From Im ln(A+iB) = tan
−1
(B/A), then
Im
ln
1 +Z
1 −Z
= tan
−1
2ρb cos θ
b
2
−ρ
2
−i
2(V1 −V2)
π
iIm
¸
n odd
Z
n
n
=
2(V1 −V2)
π
tan
−1
2ρb cos θ
b
2
−ρ
2
Combining the two parts of the potential to ﬁnd
Φ(ρ, θ) =
V1 +V2
2
+
2(V1 −V2)
π
tan
−1
2ρb cos θ
b
2
−ρ
2
(b) Calculate the surfacecharge density on each half of the cylinder
The surfacecharge density
σ = −o
∂Φ
∂ρ
ρ=b
= −2o
V1 −V2
π
b cos θ
b
2
+ρ
2
b
4
−2b
2
ρ
2
+ρ
4
+ 4ρ
2
b
2
cos
2
θ
= −o
V1 −V2
bπ cos θ
3.4 Problem 4: Jackson 2.23
A hollow cube has conducting walls deﬁned by six planes x = 0, y = 0, z = 0, and x = a, y = a, z = a. The walls
z = 0 and z = a are held at a constant potential V . The other four sides are at zero potential.
(a) Find the potential Φ(x, y, z) at any point inside the cube.
From class
Φ(x, y, z)top =
∞
¸
m=1
∞
¸
n=1
Anm sin
nπx
a
sin
mπy
a
sinh (γmnz)
where γ =
nπ
a
2
+
mπ
a
2
. Following Equation 2.58
Anm =
4
a
2
sinh (γnma)
a
0
dx
a
0
dyV sin
nπx
a
sin
mπy
a
=
16V
π
2
nmsinh(γnma)
, where m,n are both odd
By symmetry the bottom has to be the same except for z →a −z. Therefore the bottom is
Φ(x, y, z)
bottom
odd
¸
n,m=1
Anm sin
nπx
a
sin
mπy
a
sinh (γmn(a −z))
The total potential is the sum of both top and bottom potential
Φtop + Φ
bottom
=
odd
¸
n,m=1
Anm sin
nπx
a
sin
mπy
a
[sinh (γmnz) + sinh (γmn(a −z))]
(b) Evaluate the potential at the center of the cube numerically, accurate to three signiﬁcant ﬁgures. How many
terms in the series is it necessary to keep in order to attain this accuracy? Compare your numerical results
with the average value of the potential on the walls. See Problem 2.28.
1
Jackson p745
17
To have three digits of accuracy, the ﬁrst four terms need to be kept, which is
Φ
a
2
,
a
2
,
a
2
=
16V
π
2
sinh(
√
2π)
(1)(1)(2 sinh(π
√
2/2))
+
16V
3π
2
sinh(
√
10π)
(1)(−1)(2 sinh(π
√
10/2))
+
16V
3π
2
sinh(
√
10π)
(−1)(1)(2 sinh(π
√
10/2))
+
16V
9π
2
sinh(3
√
2π)
(1)(1)(2 sinh(3π
√
2/2))
≈ .3329V = .333V
From Problem 2.28 the potential should be
2
6
V = .33
¯
3V
(c) Find the surfacecharge density on the surface z = a.
The surfacecharge density in the cube is given by
σ = o
∂Φ
∂z
z=a
=
16oV
π
2
odd
¸
n,m=1
Anm sin
nπx
a
sin
mπy
a
[cosh (γmna) −1]
=
16oV
π
2
odd
¸
n,m=1
1
nm
sin
nπx
a
sin
mπy
a
¸
cosh (γmna) −1
sinh (γmna)
=
16oV
π
2
odd
¸
n,m=1
1
nm
sin
nπx
a
sin
mπy
a
tanh
γmna
2
18
Chapter 4
Problem Set 4
4.1 Problem 1: Jackson 3.3
A thin, ﬂat, conducting, circular disc of radius R is located is located in the x −y plane with its center at the origin,
and is maintained at a ﬁxed potential V . With the information that the change density on a disc at ﬁxed potential
is proportional to (R
2
−ρ
2
)
−1/2
, where ρ is the distance out from the center of the disc,
(a) show that for r > R the potential is
Φ(r, θ, φ) =
2V
π
R
r
∞
¸
l=0
(−1)
l
2l + 1
R
r
2l
p
2l
(cos θ) (4.1)
The surface charge is given by
σ(ρ) =
λ
R
2
−ρ
2
The volume charge is found by consider a shell of radius r
dq = σ(r)2πrdr =
2π
0
dφ
1
−1
d(cos θ)r
2
drf(r)δ(cos θ)
σ(r)2πrdr = 2πr
2
f(r)dr
The volume charge density is ρ(r) =
σ(r)
r
δ(cos θ). The potential at the origin is
V = Φ(0) =
1
4πo
R
0
λ2πrdr
r
√
R
2
−r
2
=
λ
2o
sin
−1
r
R
R
0
=
λπ
4o
The potential at any point becomes
Φ(x) =
1
4πo
4oV
π
∞
¸
l=0
l
¸
m=−l
Y
lm
(θ, φ)
1
r
l+1
v
r
l
Y
∗
lm
(θ
, φ
)
δ(cos θ
)r
2
dr
dΩ
r
√
R
2
−r
2
Solving the integral
Φ(x) =
V
π
2
∞
¸
l=0
P
l
(cos θ)
2π
r
l+1
R
0
1
−1
r
l+1
P
l
(cos θ
)
δ(cos θ
)dr
d(cos θ
)
√
R
2
−r
2
=
2V
π
∞
¸
l=0
P
l
(cos θ)
P(0)
r
l+1
R
0
r
l+1
dr
√
R
2
−r
2
=
2V
π
∞
¸
n=0
P2n(cos θ)
(−1)
n
(2n + 1)!!
2
n
n!
R
2n+1
n!2
n
(2n + 1)!!r
2n+1
19
Then
Φ(x) =
2V
π
∞
¸
n=0
(−1)
n
2n + 1
R
r
2n+1
P2n(cos θ)
(b) ﬁnd the potential for r < R.
To ﬁnd the potential for inside the circular radius, use the boundary condition when r = R. The potential inside and
outside has to be equal, therefore
A
l
R
l
=
2V
π
(−1)
n
2n + 1
= B
l
R
−l−1
where l = 2n. Then the equation for inside becomes
Φ =
2V
π
∞
¸
n=0
(−1)
n
2n + 1
r
R
2n
P2n(cos θ)
(c) What is the capacitance of the disc?
The capacitance is given by C = Q/V . From part (a) V =
λπ
4o
. The charge is
Q =
R
0
λ
√
R
2
−r
2
2πrdr = −2πλ
R
2
−r
2
R
0
= 2πλR
The capacitance becomes
C =
Q
V
=
2πλR
λπ
4o
→C = 8oR
4.2 Problem 2: Jackson 3.6
Two point charges q and −q are located on the z axis at z = +a and z = −a, respectively.
(a) Find the electrostatic potential as an expansion in spherical harmonics and powers of r for both r > a and r < a.
From Equation 3.38
Φ(x) =
q
4πo
¸
1
[x −a[
−
1
[x +a[
=
q
4πo
1
r>
∞
¸
l=0
¸
r<
r>
l
P
l
(cos θ) −
r<
r>
l
P
l
(cos(π −θ))
¸
From equation Equation 3.57
Φ =
q
4πo
1
r>
∞
¸
l=0
r<
r>
l
4π
2l + 1
(Y
l0
(θ, φ) −Y
l0
(π −θ, φ))
From parity of the spherical harmonics
Φ(x) =
q
4πo
1
r>
∞
¸
l=0
r<
r>
l
4π
2l + 1
1 + (−1)
l+1
Y
l0
(θ, φ)
where r< is the smaller of mod[x, a] and r> is the larger of mod[x, a].
(b) Keeping the product qa ≡ p/2 constant, take the limit of a → 0 and ﬁnd the potential for r ,= 0. This is by
deﬁnition a dipole along the z axis and its potential.
20
So for a < r
Φ =
q
4πo
1
r
∞
¸
l=0
a
r
l
4π
2l + 1
1 + (−1)
l+1
Y
l0
(θ, φ)
In the case a →0 only the ﬁrst power of a are examined. The potential is then
Φ =
q
4πo
1
r
a
r
4π
2 + 1
(1 + 1) Y10 (θ, φ)
=
q
4πo
a
r
2
4π
3
(2)
3
4π
cos θ
=
p cos θ
4πor
2
(c) Suppose now that the dipole of part b is surrounded by a grounded spherical shell of radius b concentric with
the origin. By linear superposition ﬁnd the potential everywhere inside the shell.
The potential is then Φ = Φ
d
+ Φs. Here only the A
l
terms because we are inside a sphere. Then
Φ =
p cos θ
4πor
2
+
∞
¸
l=0
A
l
r
l
P
l
(cos θ)
At the boundary condition Φ(b) = 0, then
0 =
p cos θ
4πob
2
+
∞
¸
l=0
A
l
b
l
P
l
(cos θ)
The only term needed is l = 1 because the dipole, so
A1b
1
cos θ = −
p cos θ
4πob
2
A1 = −
p
4πob
3
The potential then becomes
Φ =
p cos θ
4πo
1
r
2
−fracrb
3
4.3 Problem 3: Jackson 3.10
For the cylinder in Problem 3.9 the cylindrical surface is made of two equal halfcylinder, one at potential V and the
other at potential −V , so that
V (φ, z) =
¸
V for −π/2 < φ < π/2
−V for π/2 < φ < 3π/2
(4.2)
(a) Find the potential inside the cylinder
From class
Φ =
∞
¸
n
∞
¸
ν=0
(Anν sin νφ +Bnν cos νφ) Iν
nπρ
L
sin
nπz
L
Fourier analyst can yield the coeﬃcients. For Anν
Anν =
2
πL
1
Iν
L
0
2π
0
V (φ, z) sin
nπz
L
sin νφdφdz
21
But this has two intervals from −π/2 to π/2 and π/2 to 3π/2. In these intervals cosine is symmetric so Anν = 0.
Now for Bnν
Bnν =
2
πL
1
Iν
L
0
2π
0
V (φ, z) sin
nπz
L
sin νφdφdz
over the two ranges, only the odd terms survive for ν. For the z component
L
0
sin (nπz/L) dz =
2L
(2l+1)π
, because
only odd terms survive from n, also. So Bnν becomes
Bnν =
2L
(2l + 1)π
4(−1)
k
2k + 1
2V
πLI
2k+1
=
16V (−1)
k
(2k + 1)(2l + 1)π
2
I
2k+1
(b) Assuming L b, consider the potential at z = L/2 as a function of ρ and φ and compare it with twodimensional
Problem 2.13.
From Jackson
I
2k+1
(2l + 1)πρ
L
≈
1
Γ(2k + 2)
¸
(2l + 1)πρ
2L
k+1
and for small angles sin
(2l+1)π
L
≈ (−1)
l
. Then
Φ =
∞
¸
l,k=0
16(−1)
k+l
V
π
2
(2l + 1)(2k + 1)
ρ
b
2k+1
cos((2k + 1)φ)
Using the identity
π
4
= tan
−1
(1) =
∞
¸
l=0
(−1)
l
2l + 1
(1)
2l+1
The potential becomes
Φ =
4V
π
∞
¸
k=0
(−1)
k
2k + 1
ρ
b
2k+1
cos((2k + 1)φ)
From problem 2.13
∞
¸
k=0
(−1)
k
2k + 1
ρ
b
2k+1
cos((2k + 1)φ) =
1
2
tan
−1
¸
2bρ cos φ
b
2
−ρ
2
The potential becomes
Φ =
2V
π
tan
−1
¸
2bρ cos φ
b
2
−ρ
2
22
Chapter 5
Problem Set 5
5.1 Problem 1: Jackson 3.17
The Dirichlet Green function for the unbounded space between the planes as z = 0 and z = L allows discussion of a
point charge or a distribution of charge between parallel conducting planes held at zero potential.
(a) Using cylindrical coordinates show that one form of the Green function is
G(x, x
) =
4
L
∞
¸
n=1
∞
¸
m=−∞
e
im(φ−φ
)
sin
nπz
L
sin
nπz
L
LIm
nπ
L
ρ<
Km
nπ
L
ρ>
(5.1)
We know that
∇
2
G(x, x
) =
−4πδ(ρ −ρ
)
ρ
δ(φ −φ)δ(z −z
) (5.2)
Then expanding the δfunctions
∇
2
G(x, x
) = −
4
L
∞
¸
m=−∞
∞
¸
n=1
e
−imφ
e
imφ
sin
nπz
L
sin
nπz
L
δ(ρ −ρ
)
ρ
=
∞
¸
m=−∞
∞
¸
n=1
¸
−
4
L
e
−imφ
sin
nπz
L
δ(ρ −ρ
)
ρ
e
imφ
sin
nπz
L
(5.3)
Now expanding the Green function
∇
2
G(x, x
) = ∇
2
∞
¸
m=−∞
∞
¸
n=1
e
imφ
sin
nπz
L
Ψ
where Ψ is the eigenfunction. Expanding the Laplacian
∇
2
G(x, x
) =
∞
¸
m=−∞
∞
¸
n=1
¸
1
ρ
∂
∂ρ
ρ
∂
∂ρ
−
m
2
ρ
2
−
n
2
π
2
L
2
Ψe
imφ
sin
nπz
L
(5.4)
Then equating equation 3 and equation 4
¸
1
ρ
∂
∂ρ
ρ
∂
∂ρ
−
m
2
ρ
2
−
n
2
π
2
L
2
Ψ
1
e
−imφ
sin
nπz
L
= −
4
L
δ(ρ −ρ
)
ρ
Then deﬁne gmn =
Ψ
e
−imφ
sin
nπz
L
. From Equation 3.98 and 3.99 gmn = CIm(kρ<)Km(kρ>) where k = nπ/L. Now
to ﬁnd C
∂gmn
∂ρ
ρ+
−
∂gmn
∂ρ
ρ−
=
−4
ρL
⇒Eq3.147
−4
ρL
= −Ck
1
kρ
→C =
4
L
Now combining the above constant with gmn and equation 3 to ﬁnd
G(x, x
) =
4
L
∞
¸
m=−∞
∞
¸
n=1
e
imφ
e
−imφ
sin
nπz
L
sin
nπz
L
Im
nπ
L
ρ<
Km
nπ
L
ρ>
23
(b) Show that an alternative form of the Green function is
G(x, x
) = 2
∞
¸
m=−∞
∞
0
dk e
im(φ−φ
)
Jm(kρ)Jm(kρ
)
sinh(kz<) sinh[k(L −z>)]
sinh(kL)
(5.5)
Expanding equation 2 in φ and ρ
∇
2
G(x, x
) = −2
∞
¸
m=−∞
∞
0
e
−imφ
e
imφ
Jm(kρ
)Jm(kρ)δ(z −z
)kdk
=
∞
¸
m=−∞
∞
0
−2ke
−imφ
Jm(kρ
)δ(z −z
)
e
imφ
Jm(kρ)dk (5.6)
Now expanding the Green function
∇
2
G(x, x
) = ∇
2
∞
¸
m=−∞
∞
0
e
imφ
Jm(kρ)Ψdk
where Ψ is the eigenfunction. Expanding the Laplacian
∇
2
G(x, x
) =
∞
¸
m=−∞
∞
0
¸
∂
2
∂z
2
−
m
2
ρ
2
+
1
ρ
∂
∂ρ
ρ
∂
∂ρ
e
imφ
Jm(kρ)Ψdk
=
∞
¸
m=−∞
∞
0
¸
∂
2
∂z
2
−k
2
Ψe
imφ
Jm(kρ)dk (5.7)
Equating equation 6 and 7, to ﬁnd
¸
∂
2
∂z
2
−k
2
Ψ
1
e
−imφ
Jm(kρ
)
= −2kδ(z −z
)
Then g
m,k
= Ψ/e
−imφ
Jm(kρ
), but the expansion of the δfunction in z is g
m,k
= C sinh(kz<) sinh(k(L−z>)). Now
ﬁnding C
∂g
m,k
∂z
z+
−
∂g
m,k
∂z
z−
= −2k
−kC [sinh(kz) cosh(k(L −z)) + cosh(kz) sinh(k(L −z))] = −2k →−kC sinh(kL) = −2k →C =
2
sinh(kL)
Now combining the above equation with equation 6 and 7
G(x, x
) = 2
∞
¸
m=−∞
∞
0
e
imφ
e
−imφ
Jm(kρ)Jm(kρ
)
sinh(kz<) sinh(k(L −z>))
sinh(kL)
5.2 Problem 2: Jackson 4.7
A localized distribution of charge has a charge density
ρ(r) =
1
64π
r
2
e
−r
sin
2
θ (5.8)
(a) Make a multipole expansion of the potential due to this charge density and determine all the nonvanishing
multipole moments. Write down the potential at large distance as a ﬁnite expansion in Legendre polynomials.
For Equation 4.1 to be solved ﬁrst Equations 4.44.6 have to be solved. First notice that the charge density is
independent of φ, therefore m = 0. Now notice sin
2
θ = 1−cos
2
θ, because this is power 2, then l ≤ 2. Now beginning
with Y20
Y20 =
5
4π
3
2
cos
2
θ −
1
2
→Y20 =
5
4π
−
3
2
cos
2
θ + 1
→−
2
3
4π
5
Y20 = sin
2
θ −
2
3
24
Now the next term has no power of θ, so the next term would be from Y00. Therefore
sin
2
θ =
2
3
√
4πY00 −
2
3
4π
5
Y20
Now using Equation 4.3
1
√
4π
V
2
3
√
4π
1
64π
r
4
e
−r
sin θdrdθdφ =
1
4π
1
2
5
4π
V
−2
3
4π
5
1
64π
r
6
e
−r
sin
3
θdrdθdφ =
−30
4π
Back to Equation 4.1
Φ(x) =
1
4πo
¸
4π
1
1
4π
P0
r
−
4π
5
30
4π
P2(cos θ)
r
3
→Φ(x) =
1
4πo
¸
P0
r
−
6P2(cos θ)
r
3
(b) Determine the potential explicitly at any point in space, and show that near the origin, correct to r
2
inclusive,
Φ(r) ·
1
4πo
¸
1
4
−
r
2
120
P2(cos θ)
(5.9)
We know from Equation 4.2
Φ(x) =
1
o
¸
l,m
1
2l + 1
¸
Y
∗
lm
(θ
, φ
)r
l
ρ(x
)d
3
x
Y
lm
(θ, φ)
r
l+1
Looking ahead to the deserved solution and combined with the previous part only l = 0, 2 and m = 0 will be considers.
Then
Φ(x) =
1
o
¸
2
3
√
4πY00
∞
0
1
64π
r
3
e
−r
dr −
2
3
4π
5
Y20
r
2
5
∞
0
1
64π
re
−r
dr
¸
=
1
o
¸
2
3
√
4π
P0
√
4π
1
64π
6 −
2
3
4π
5
5
4π
r
2
P2(cos θ)
5
1
64π
1
¸
=
1
o
¸
P0
16π
−
r
2
480π
P2(cos θ)
=
1
4πo
¸
1
4
−
r
2
120
P2(cos θ)
(c) If there exists at the origin a nucleus with a quadrupole moment Q = 10
−28
m
2
, determine the magnitude of
the interaction energy, assuming that the unit of charge in ρ(r) above is the electronic charge and the unit of
length is the hydrogen Bohr radius ao = 4πo/me
2
= 0.529 10
−10
m. Express your answer as a frequency by
dividing by Planck’s constant h.
The quadrupole interaction energy is given in Equation 4.24
W = −
1
6
¸
i
¸
j
Qij
∂Ej
∂xi
(0) (5.10)
But as it has been shown earlier that m = 0 so the only terms that remain are i = j, but the trace has to be 0, which
is
W =
eQ
6
1
2
∂Ex
∂x
+
1
2
∂Ey
∂y
−
∂Ez
∂z
=
eQ
6
1
2
∇ E −
3
2
∂Ez
∂z
=
eQ
6
1
2
ρ
o
−
3
2
∂Ez
∂z
25
At the origin ρ(0) =, so the ﬁrst term drops out, so the energy is then W =
eQ
4
∂
2
Φ
∂z
2
. Now the potential has to be
transferred into Cartesian coordinates. So for the quadrupole term
Φ(x) =
1
480πo
r
2
P2 (cos θ) →Φ(x) =
1
960πo
2z
2
−x
2
−y
2
from Merzbacher Quantum Mechanics. So
∂
2
Φ
∂z
2
=
1
240πo
, then the quadrupole term becomes
W =
e
2
Q
960πoa
3
o
Then the frequency is
W
h
=
e
2
Q
960πoha
3
o
= 0.98 MHz
26
Chapter 6
Problem Set 6
6.1 Problem 1: Jackson 4.9
A point charge q is located in free space a distance d from the center of a dielectric sphere of radius a (a < d) and
dielectric constant /o.
(a) Find the potential at all points in space as an expansion in spherical harmonics.
Inside the sphere ∇ D = 0, because there is no charge inside
Φ(r, θ) =
∞
¸
l=0
A
l
r
l
P
l
(cos θ) r<a
For outside the sphere, there are two regions, a < r < d and r < d, which will dictate the expansion of 1/[x −x
, ie
what is r< and r>. Then the outside region will be
Φ(r, θ) =
∞
¸
l=0
B
l
r
−(l+1)
P
l
(cos θ) +
1
4πo
q
[x −x
[
Using the typical spherical expansion for 1/[x −x
, the expression for all space is
Φ(r, θ) =
¸
∞
l=0
A
l
r
L
P
l
(cos θ) r<a
¸
∞
l=0
B
l
r
−(l+1)
+
q
4πo
r
l
d
l+1
P
l
(cos θ) a<r<d
¸
∞
l=0
B
l
r
−(l+1)
+
q
4πo
d
l
r
l+1
P
l
(cos θ) r>d
This is the potential in terms of Lagrange Polynomials, so in Spherical Harmonics
Φ(r, θ) =
¸
∞
l=0
2l+1
4π
A
l
r
L
Y
l,0
(θ, φ) r<a
¸
∞
l=0
2l+1
4π
B
l
r
−(l+1)
+
q
4πo
r
l
d
l+1
Y
l,0
(θ, φ) a<r<d
¸
∞
l=0
2l+1
4π
B
l
r
−(l+1)
+
q
4πo
d
l
r
l+1
Y
l,0
(θ, φ) r>d
(6.1)
Now to ﬁnd the coeﬃcients. Using the normal boundary condition
o
∂Φ
∂r
r=a
−
=
∂Φ
∂r
r=a
+
(6.2)
o
lA
l
a
l−1
= −(l + 1)B
l
a
−(l+2)
+
lqa
l−1
4πod
l+1
A
l
=
o
¸
−
(l + 1)
l
B
l
a
−(2l+1)
+
q
4πod
l+1
(6.3)
27
The other boundary condition is for tangential component
∂Φ
∂θ
r=a
−
=
∂Φ
∂θ
r=a
+
A
l
a
l
= B
l
a
−(l+1)
+
q
4πo
a
l
d
l+1
(6.4)
Then the coeﬃcients can easily be found from equation 3 and 4
A
l
=
1
/o + (l + 1)/l
2l + 1
l
q
4πod
l+1
(6.5)
B
l
=
1
/o + (l + 1)/l
1 −
o
qa
2l+1
4πod
l+1
(6.6)
Combining equations 5 and 6 into equation 1 yields the potential at every point in space.
(b) Calculate the rectangular components of the electric ﬁeld near the center of the sphere.
Expanding the potential inside the sphere
Φ(r, θ) = A1rP1(cos θ) +A2r
2
P2(cos θ) +. . .
From Quantum Mechanics, rP1(cos θ) = z and r
2
P2(cos θ) = z
2
−x
2
+y
2
, then the potential becomes
Φ(x, y, z) =
q
4πo
¸
3z
(/o + 2)d
2
+
5
z
2
−x
2
−y
2
2(2/o + 3)d
3
+. . .
¸
So then Ei = −
Φdxi. Then the component in Cartesian space
Ex =
q
4πod
2
¸
5
2/o + 3
x
d
+. . .
Ey =
q
4πod
2
¸
5
2/o + 3
y
d
+. . .
Ez = −
q
4πod
2
¸
3
/o + 2
+
5
2/o + 3
z
d
+. . .
(c) Verify that, in the limit /o →∞, your result is the same as that for the conducting sphere.
Examining equation 2, the equation would blowup if A
l
,= 0, therefore A
l
= 0. This is the same for the conductor.
Then from that and equation 4,
B
l
= −
q
4πo
a
2l+1
d
l+1
which is the same as a conduction sphere.
6.2 Problem 2: Jackson 4.10
Two concentric conducting spheres of inner and outer radii a and b , respectively, carry charges ±Q. The empty
space between the spheres is halfﬁlled by a hemispherical shell of dielectric (of dielectric constant /o), as shown in
the ﬁgure.
1
(a) Find the electric ﬁeld everywhere between the spheres.
Let’s assume that E ∝
r
r
3
. Using Gauss’s Law for a sphere between the two spheres:
D da = Q
2πr
2
(oE(r) −E(r)) = Q
Q
2π(o +)
r
r
3
= E(r)
Now checking the conditions on the solution. The curl of the electric ﬁeld is obviously 0. The divergent of the ﬁeld
does yield the charge density. By the uniqueness theorem, this is the correct solution.
1
Jackson. pg 174
28
(b) Calculate the surfacecharge distribution of the inner sphere.
From Equation 4.40, in particular n ((D2 −D1) = σ
free
. The half without the dielectric is
σ
free
= r D(a) = ˆr
r
a
2
Qo
2π(o +)
=
Qo
2πa
2
(o +)
Now for part with the dielectric is directly
σ
free
=
Q
2πa
2
(o +)
(c) Calculate the polarizationcharge density induced on the surface of the dielectric at r = a
The electric polarization is given by P = ( −o)E. Because in the free space = o, the electric polarization for the
free space is P = 0. Therefore σ
pol
= 0 for the free space half. Now, the polarization density is σ
pol
= −ˆ n(P1 −P2).
For the dielectric part
σ
pol
= −ˆr P(a) = −ˆr
ˆr
a
2
( −o)Q
2π(o +)
σ
pol
= −
( −o)Q
2πa
2
(o +)
6.3 Problem 3: Distressed Simple Cubic
Repeat the analysis of section 4.5 for a cubic crystal which is subjected to a stress such that the lattice separation is
elongated along one edge of the cube (the x axis) and contracted along the other cube edges (the y and z axes) by
the amounts
ax = a(1 +δ) (6.7)
ay,z = a
1 −
1
2
δ
(6.8)
where a is the unstressed cubic lattice spacing. Find the susceptibility for two cases, considering that the molecular
polarizabilty γ
mol
is a constant scalar (Hint: Calculate Enear at a particular dipole by considering only the 6 nearest
dipoles in the cubic lattice):
(a) An electric ﬁeld applied parallel to the x axis
The polarization is given on page 161 in Jackson as P = N'p
mol
) and also 'p
mol
) = oγ
mol
(E + Ei) From the
previous page Equation 4.63, Ei =
P
3
+Enear. All that remains is to ﬁnd Enear. From Equation 4.64:
Enear =
3ˆ n(ˆ n p) −p
r
3
(6.9)
29
Considering the six nearest neighbors, which occur from the shape(from class notes):
Ei =
3ˆ n(ˆ n˙pi) −pi
a
3
i
Ex =
4px
a
3
x
−
2px
a
3
y
−
2px
a
3
z
Ey =
4py
a
3
y
−
2py
a
3
x
−
2py
a
3
z
Ez =
4pz
a
3
z
−
2pz
a
3
x
−
2pz
a
3
y
Now using the values for the expansion of the nearest neighbors and the fact that δ < 1 and the binomial expansion.
Ex =
4px
a
3
(1 +δ)
3
−
2px
a
3
1 −
1
2
δ
3
−
2px
a
3
1 −
1
2
δ
3
=
4px
a
3
(1 −3δ) −
4px
a
3
1 +
3
2
δ
= −
18pxδ
a
3
(6.10)
Ey =
4py
a
3
1 −
1
2
δ
3
−
2py
a
3
(1 +δ)
3
−
2py
a
3
1 −
1
2
δ
3
=
2py
a
3
1 +
3
2
δ
−
2py
a
3
(1 −3δ) =
9pyδ
a
3
(6.11)
Ez =
4pz
a
3
1 −
1
2
δ
3
−
2pz
a
3
(1 +δ)
3
−
2pz
a
3
1 −
1
2
δ
3
=
2pz
a
3
1 +
3
2
δ
−
2pz
a
3
(1 −3δ) =
9pzδ
a
3
(6.12)
So now Enear known. To simplify the problem, the fact that P = oχE can be used. Since Ey = Ez = 0, then
Py = Pz = 0. So only Px is needed.
Px = Nγ
mol
oEx +o
−
18
a
3
δ +
1
3o
Px
→
1 +Nγ
mol
18
a
3
δo −
Nγ
mol
3
Px = Nγ
mol
oEx
1 +Nγ
mol
18
a
3
δo −
Nγ
mol
3
χoEx = Nγ
mol
oEx → χ =
Nγ
mol
1 −
1
3
Nγ
mol
+
18
a
3
Nγ
mol
δo
(b) An electric ﬁeld applied parallel to the y axis.
For an electric ﬁeld in the y direction. As above, P = oχE, then Ex = Ez = Px = Pz = 0. So only Enear for the y
direction is needed. Then
Py = Nγ
mol
oEy +o
9
a
3
δ +
1
3o
Py
→
1 −Nγ
mol
9
a
3
δo −
Nγ
mol
3
Py = Nγ
mol
oEy
1 −Nγ
mol
9
a
3
δo −
Nγ
mol
3
χoEy = Nγ
mol
oEy → χ =
Nγ
mol
1 −
1
3
Nγ
mol
−
9
a
3
Nγ
mol
δo
30
Chapter 7
Problem Set 7
7.1 Problem 1: Jackson 4.13
Two long, cylindrical conducting surfaces of radii a and b are lowered vertically into a liquid dielectric. If the liquid
rises an average height h between the electrodes when a potential diﬀerence V is established between them, show
that the susceptibility of the liquid is
χe =
b
2
−a
2
ρghln(b/a)
oV
2
(7.1)
where ρ is the density of the liquid, g is the acceleration due to gravity, and the susceptibility of air is neglected.
The ﬁrst thing to ﬁnd is the electric ﬁeld between the cylinders. There is no charge between the cylinders, so ∇
2
Φ = 0.
From the symmetry of the problem, cylindrical coordinates are the proper choice, Further there is radial symmetry,
so there is no polar coordinate. So the Laplacian is
1
r
∂
∂r
r
∂Φ
∂r
= 0 (7.2)
The familiar solution is
Φ(r) = Aln (r/B)
Now the boundary condition that the diﬀerence of the two cylinders is V or more exact
Φ(b) −Φ(a) = Aln(b/a) = V ⇒A =
V
ln(b/a)
So the electric ﬁeld is
E = −
∂Φ
∂r
ˆr →E =
−V
ln(b/a)
ˆr
r
Now that the electric ﬁeld is known, the electric displacement is simply D = E for inside the liquid, and for above
the liquid, the electric displacement is D = oE. The work to raise the liquid is
W =
1
2
E Dd
3
x =
1
2
b
a
2πrdr
z
−d
E
2
dz +
l
z
oE
2
dz
where d is the length in the liquid and l is the length above. The work is then
W =
1
2
b
a
2πrdr
z
−d
E
2
dz +
l
z
E
2
dz
=
πV
2
ln(b/a)
( (z +d) +o (l −z))
From Equation 4.102, the force needed to left the liquid is
Fz =
∂W
∂z
V
=
( −o) πV
2
ln(b/a)
31
Now ﬁnding the volume of liquid that has been lifted is
b
2
−a
2
πh. The mass is then m =
b
2
−a
2
ρπh. At
equilibrium, the force of gravity is equal to the force of the electric ﬁeld.
b
2
−a
2
ρπh =
( −o) πV
2
ln(b/a)
b
2
−a
2
ρπh
o
=
(/o −1) πV
2
ln(b/a)
The susceptibility is deﬁned as χe = /o −1. Then
b
2
−a
2
ρπh
o
=
χeπV
2
ln(b/a)
⇒ χe =
b
2
−a
2
ρghln(b/a)
oV
2
32
Chapter 8
Problem Set 8
8.1 Problem 1: Jackson 5.2
A long, right cylindrical, ideal solenoid of arbitrary cross section is created by stacking a large number of identical
currentcarrying loops one above the other, with N coils per unit length and each loop carrying a current I.
(a) In the approximation that the solenoidal coil is an ideal current sheet and inﬁnitely long, use Problem 5.1 to
establish that any point outside the coil that H = 0, and that any point inside the coil the magnetic ﬁeld is
axial and equal to
H = NI (8.1)
This can be solved using two ampere loops. Placing a loop outside the cylinder, it yields B = 0. Placing a second
loop through a side of the surface perpendicular to the bottom surface. So
B dl = µoIenc
From the ﬁrst loop that was outside the cylinder, the induced magnetic ﬁeld outside is 0. This means that the only
ﬁeld is inside the cylinder. The integral becomes
B dl = µoIenc ⇒BL = µoNIL
Using right hand rule, Bz = µoNI. In this system M = 0, then H = B/µo, therefore Bz = µoNI ⇒H = NI.
(b) For a realistic solenoid of circular cross section of radius a (Na 1), but still inﬁnite in length, show that the
“smoothed” magnetic ﬁeld just outside the solenoid (averaged axially over several turns) is not zero, but is the
same in magnitude and direction as that of a single wire on the axis carrying a current I, even if Na → ∞.
Compare ﬁelds inside and out.
In a real solenoid, it is not simple current loops stacked on each other. There is a little bind in the loop to connect
the loops. This means there is a current in the vertical direction. So what is the current? This can be found by the
fact the total current moves up 1/N each loop. In each loop there is I/N in the vertical direction, so over a unit
length there is N loops, so the amount of vertical current is then NI/N = I. Therefore this current in the vertical
direction
B dl ⇒2πρB = µoI ⇒ B
θ
=
µoI
2πρ
This is the same as a wire along the zaxis. As the number of loops increase per unit length, the more ideal
it will become. The size of inside is many orders of magnitude large than the ﬁeld outside. To be more exact,
µoNI/(µoI/2πρ) → 2πρN. So the ﬁeld inside is 2πρN times larger. The directions of the two are perpendicular to
each other.
33
8.2 Problem 2: SemiInﬁnite Cylinder
Consider a semiinﬁnite solenoid, which we idealize as a cylindrical current sheet of radius a and azimuthal current
K per unit length, with one end at the origin and the other end at ∞ along the zaxis.
(a) Use the BiotSavart Law to show that the magnetic induction on the axis is
Bz =
µoK
2
¸
1 −
z
(a
2
+z
2
)
1/2
¸
(8.2)
Let’s begin by thinking of a single current loop at the origin. The magneticﬂux density is
Bz =
µoI
4π
dl x
[x[
3
Because this has to end the ˆ z direction, x = a cos φˆ ρ + a sin φ
ˆ
φ + zˆ z, [x[
3
=
a
2
+z
2
3/2
, and dl = −a sin φˆ ρ +
a cos φ
ˆ
φ + 0ˆ z. Then
Bz =
µoI
4π
1
(a
2
+z
2
)
3/2
2π
0
−az cos φ +az sin φ +a
2
dφ
=
µoI
2
a
2
(a
2
+z
2
)
3/2
ˆ z
Now this has to be integrated from start to ﬁnish of the cylinder remembering that I →K
Bz =
µoa
2
K
2
ˆ z
0
−∞
dz
(a
2
+z
2
)
3/2
=
µoa
2
K
2
ˆ z
z
(a
2
+z
2
)
1/2
0
−∞
=
µoK
2
ˆ z cos θ
π
θ
1
where θ is the angle that the origin makes with the edge at the end of the cylinder and cos θ =
z
(a
2
+z
2
)
3/2
, then
Bz =
µoK
2
¸
1 −
z
(a
2
+z
2
)
3/2
¸
(b) Use a Taylor series expansion of the ﬁeld near the axis together with ∇ B = 0 and ∇B = 0 to show that the
magnetic induction near the axis is
Bρ ≈
µoK
4
a
2
ρ
(a
2
+z
2
)
3/2
(8.3)
Bz ≈
µoK
2
¸
1 −
z
(a
2
+z
2
)
1/2
−
3
4
a
2
zρ
2
(a
2
+z
2
)
5/2
¸
(8.4)
The two conditions, ∇B and ∇B, lead to Laplace’s Equation ∇
2
Φmag = 0 and B = −∇Φmag. From electrostatics,
Φmag = R(ρ)Z(z)Q(θ). The radial solutions are the Bessel and Neumann functions. Then
Φmag = (AJν(kρ) +BNν(kρ)) Ce
−kz
De
iνθ
From symmetry, ν = 0 and as ρ → 0 has to be well deﬁne B = 0. Then Φmag = CJ0(kρ)e
−kz
. Then the induced
ﬁeld is
B = −∇Φmag = −
¸
ˆ ρ
∂
∂ρ
+ ˆ z
∂
∂z
Φmag
= −Ck
∂Jo(x)
∂x
e
−kx
ˆ ρ +CkJo(kρ)e
−kz
ˆ z
34
From Arfken’s book,
∂Jo(x)
∂x
= −J1(x)
1
. Expanding the two terms
2
,
Bρ = CkJ1(kρ)e
−kz
≈ Cke
−kz
kρ
2
−
k
3
ρ
3
16
Bz = CkJ0(kρ)e
−kz
≈ Cke
−kz
1 −
k
2
ρ
2
4
Now, Bz(0, z) = Cke
−kz
, this has to be the case, because it has to be continuous. It follows that
∂
n
Bz(0,z)
∂z
n
=
(−1)
n
Ck
n+1
e
−kz
. With this substitution,
Bρ ≈ Cke
−kz
kρ
2
≈
ρ
2
∂Bz(0, z)
∂z
Bz ≈ Cke
−kz
1 −
k
2
ρ
2
4
≈ Bz(0, z) −
ρ
2
4
∂
2
Bz(0, z)
∂z
Using Bz from part one ﬁnally
Bρ ≈
ρ
2
∂
∂z
µoK
2
¸
1 −
z
(a
2
+z
2
)
3/2
¸
⇒ Bρ ≈
µoK
4
a
2
ρ
(a
2
+z
2
)
3/2
Bz ≈
µoK
2
¸
1 −
z
(a
2
+z
2
)
3/2
¸
−
ρ
2
4
∂
2
∂z
¸
µoK
2
¸
1 −
z
(a
2
+z
2
)
3/2
¸¸
Bz ≈
µoK
2
¸
1 −
z
(a
2
+z
2
)
1/2
−
3
4
a
2
zρ
2
(a
2
+z
2
)
5/2
¸
8.3 Problem 3: Jackson 5.6
A cylindrical conductor of radius a has a hole of radius b bored parallel to, and centered a distance d from, the
cylinder axis (d + b < a). The current density is uniform throughout the remaining metal of the cylinder and is
parallel to the axis. Use Amp`ere’s law and principle of linear superposition to ﬁnd the magnitude and the direction
of the magneticﬂux density in the hole.
From the hint to use superposition, we will ﬁrst ﬁnd the magneticﬂux density for the cylindrical conductors. For the
one with radius a applying Amp`ere’s law in the form of Equation 5.24
C
B dl = µo
S
J nda
B2πr
2
= µoJπr
2
→B =
µo
2
Jrˆ z ˆ r
Now for the hole, it has to be the same except that it has a diﬀerent coordinate system, that is related by a linear
transformation, directly B
=
µo
2
J
r
ˆ z ˆ r
, because this is a hole the current density is in the opposite direction,
J
= −J. Next is relating rˆ r to r
ˆ r
. Deﬁning the center of the hole on the −yaxis of the cylinder, the relationship
is r
ˆ r
= rˆ r +dˆ y. Now linear superposition can be used to ﬁnd
B
final
= B+B
=
µo
2
Jˆ z (rˆ r −rˆ r −dˆ y)
=
µo
2
Jdˆ z (−ˆ y) ⇒ B
final
=
µoJd
2
ˆ x
8.4 Problem 4: Jackson 5.13
A sphere of radius a carries a uniform surfacecharge distribution σ. The sphere is rotated about a diameter with
constant angular velocity ω. Find the vector potential and magneticﬂux density both inside and outside the sphere.
1
Arfken. MMP 5
th
ed. pg.673
2
Arfken. MMP 5
th
ed. pg.670
35
The surface current is
J = σv = σ( ω r
) = σaω sin θ
(r
− a)
ˆ
φ
. The surface current is then related to l = 1 and
m = 1 only, because of the sine term. The
36
vector potential is then
A
φ
(r, θ) =
µo
4π
¸
l
P
1
l
(cos θ)
l(l + 1)
σωaδ(r
−a)
r
l
<
r
l+1
>
P
1
l
(cos θ
)r
2
sin
2
θ
dr
dθ
dφ
= −
µo
2
¸
l
P
1
l
(cos θ)
l(l + 1)
σωa
3
r
l
<
r
l+1
>
π
0
P
1
l
(cos θ
)P
1
1
(cos θ) sin θ
dθ
= −
µoσωa
3
4
r<
r
2
>
P
1
1
(cos θ)
4
3
=
µoσωa
3
r<
3r
2
>
sin θ
Now for the magnetic ﬁeld, B = ∇A. This needs to be done for the interior (a > r) and exterior (a < r). The vector
potential interior is Aint =
1
3
µoσωar sin θ
ˆ
φ . The vector potential for exterior is Aext(r, θ) =
1
3
µoσωa
4
sin θ
r
2
ˆ
φ So
for the interior
Bint(r, θ) = ˆ r
1
r sin θ
∂
∂θ
µoσωa
3
sin
2
θ
−
ˆ
θ
1
r
∂
∂r
µoσωar
2
sin θ
3
Bint(r, θ) =
2
3
µoσωa
cos θˆ r −sin θ
ˆ
θ
Now for the exterior
Bext(r, θ) = ˆ r
1
r sin θ
∂
∂θ
µoσωa
4
3r
2
sin
2
θ
−
ˆ
θ
1
r
∂
∂r
µoσωa
4
sin θ
3
1
r
Bext(r, θ) =
1
3
µoσωa
4
r
3
2 cos θˆ r + sin θ
ˆ
θ
37
38
Chapter 9
Problem Set 9
9.1 Problem 1: Jackson 5.17
A current distribution J(x) exists in a medium of unit relative permeability adjacent to a semiinﬁnite slab of material
having relative permeability µr and ﬁlling the half space, z < 0.
(a) Show that for z > 0 the magnetic induction ca be calculated by replacing the medium of permeability µr by an
image current distribution, J
∗
, with components,
µr − 1
µr + 1
Jx(x, y, −z),
µr − 1
µr + 1
Jy(x, y, −z), −
µr − 1
µr + 1
Jz(x, y, −z) (9.1)
The ﬁeld z > 0 is
B =
µo
4π
(J +J
∗
)
(x −x
)
[x −x
[
3
d
3
x
(9.2)
The ﬁeld z < 0 is
B =
µo
4π
J
(x −x
)
[x −x
[
3
d
3
x
(9.3)
At the boundary of z = 0, the normal component of the ﬁeld must be continuous, so above
B
z
+(−z) =
µo
4π
(Jx +J
∗
x
+
)(y −y
) −(Jy +J
∗
y
+
)(x −x
)
(x −x
)
2
+ (y −y
)
2
+z
2
3/2
d
3
x
below
B
z
−(z) =
µo
4π
J
∗
x
−
(y −y
) −J
∗
y
−
(x −x
)
(x −x
)
2
+ (y −y
)
2
+z
2
3/2
d
3
x
Then B
z
+(−z) = B
z
−(z). Therefore
Jx(−z) +J
∗
x
+(z) = J
∗
x
−(z) (9.4)
Jy(−z) +J
∗
y
+(z) = J
∗
y
−(z) (9.5)
Now for the tangential components of the magnetic ﬁeld must be continuous, and H = B/µ. So for above
H
x
+(−z) =
1
4π
−z
(Jy +J
∗
y
+
) −(y −y
)(Jz +J
∗
z
+
)
((x −x
)
2
+ (y −y
)
2
+z
2
)
3/2
d
3
x
H
y
+(−z) =
1
4π
(x −x
)(Jz +J
∗
z
+
) +z
(Jx +J
∗
x
+
)
((x −x
)
2
+ (y −y
)
2
+z
2
)
3/2
d
3
x
Below
H
x
−(z) =
1
4πµr
−z
(J
∗
y
−
) −(y −y
)(J
∗
z
−
)
((x −x
)
2
+ (y −y
)
2
+z
2
)
3/2
d
3
x
H
y
−(z) =
1
4πµr
(x −x
)(J
∗
z
−
) +z
(J
∗
x
−
)
((x −x
)
2
+ (y −y
)
2
+z
2
)
3/2
d
3
x
39
From H+ = H−
µrJz(−z) +µrJ
z
+(z) = J
z
−(z) (9.6)
µrJy(−z) −µrJ
y
+(z) = J
y
−(z) (9.7)
µrJx(−z) −µrJ
x
+(z) = J
x
−(z) (9.8)
From eq 4 and 7, and eq 5 and 8
J
x
+(z) =
µr −1
µr + 1
Jx(−z) J
x
−(z) =
2µr
µr + 1
Jx(z) (9.9)
J
y
+(z) =
µr −1
µr + 1
Jy(−z) J
y
−(z) =
2µr
µr + 1
Jy(z) (9.10)
Now for the z components.
∇ J
∗
−
= 0 =
∂J
∗
x
−
∂x
+
∂J
∗
y
−
∂y
+
∂J
∗
z
−
∂z
=
2µr
µr + 1
∂Jx
∂x
+
∂Jy
∂y
+µr
∂
∂z
(Jz +J
z
+) = −
2µr
µr + 1
∂Jz
∂z
+µr
∂
∂z
(Jz +J
z
+)
Then solving and integrating
−
µr −1
µr + 1
∂Jz =
∂J
z
+ ⇒ −
µr −1
µr + 1
Jz(−z) = J
z
+(z) (9.11)
From equations 9, 10, 11 for J
∗
+
matches equation 1.
(b) Show that for z < 0 the magnetic induction appears to be due to a current distribution [2µr/(µr + 1)]J in a
medium of unit relative permeability.
Combining equation 6 with equation 11 to ﬁnd
J
∗
z
−(z) =
2µr
µr + 1
Jz (z) (9.12)
Adding equation 12 with the J− parts of equations 9, 10 to ﬁnd that J
∗
−
= [2µr/(µr + 1)]J
9.2 Problem 2: Jackson 5.19
A magnetically “hard” material is in the shape of a right circular cylinder of length L and radius a. The cylinder has
a permanent magnetization Mo, uniform throughout its volume and parallel to its axis.
(a) Determine the magnetic ﬁeld H and magnetic induction B at all points on the axis of cylinder, both inside and
outside.
Because there is no free current, this is a magnetostatic potential problem. This problem then can be treated as an
electrostatic problem. Because the magnetization is uniform, ρM = −∇ M = 0. The surface ‘charge’ density is then
σM = ˆ n M = ±Mo at the caps. The potential is then
ΦM =
1
4π
σM
[x −x
[
d
3
x
This is evaluated at the caps only. Then
Φ
1
=
Mo
4π
ρ
0
2πρdρ
ρ
2
+ (z −L/2)
2
=
Mo
2
¸
ρ
2
+ (z −L/2)
2
a
0
=
Mo
2
a
2
+ (z −L/2)
2
−
(z −L/2)
2
This can easily be show for the bottom with symmetry that
Φ2 = −
Mo
2
a
2
+ (z +L/2)
2
−
(z +L/2)
2
40
The potential over all space is Φ = Φ1 + Φ2.
Φ =
Mo
2
¸
¸
a
2
+
z −
L
2
2
−
a
2
+
z +
L
2
2
−
z −
L
2
+
z +
L
2
¸
The magnetic ﬁeld has two cases, one inside
Hz =
∂Φ
∂z
Hz =
Mo
2
¸
¸
z +L/2
a
2
+ (z +L/2)
2
−
z −L/2
a
2
+ (z −L/2)
2
−2
¸
And outside
Hz =
Mo
2
¸
¸
z +L/2
a
2
+ (z +L/2)
2
−
z −L/2
a
2
+ (z −L/2)
2
¸
Now the induced ﬁeld is Bz = µo(Hz +M). In this case, M = Mo inside and M = 0 outside. Then the induced ﬁeld
both inside and outside is
Bz =
µoMo
2
¸
¸
z +L/2
a
2
+ (z +L/2)
2
−
z −L/2
a
2
+ (z −L/2)
2
¸
(b) Plot the ratios B/µoMo and H/Mo on the axis as functions of z for L/a = 5.
See last page
9.3 Problem 3: Jackson 5.22
Show that in general a long, straight bar of uniform crosssectional area A with uniform lengthwise magnetization M,
when placed with its ﬂat end against as inﬁnitely permeable ﬂat surface, adheres with a force given approximately
by
F ·
µo
2
AM
2
(9.13)
Relate your discussion to the electrostatic considerations in Section 1.11.
So consider this as two long straight bars with uniform crosssectional areas A with uniform magnetization, M. One
rod runs from z = 0 to z = −L and the other from z = 0 to z = L. These two dipoles are “head” to “tail”, so
the components of the Bﬁeld cancel everywhere except at z = 0. Now, let’s introduce a small separation zo. The
potential energy on the real rod from the image rod is
W(zo) = −
M Bid
3
x = −M
V
Bz,id
3
x
Because the image rod is not moved and the rod is much longer than it is any other direction
W(zo) = −M
V
Bz,id
3
x ≈ −MA
zo+L
zo
Bz,i(z)dz
The force from the change in the vertical separation
Fz = −
dW(zo)
dzo
zo=0
= MA(Bz,i(L) −Bz,i(0))
From the previous problem, if the shape is circular, which is not that bad of an approximation, the induced ﬁeld is
Bz,i(z) =
µoM
2
z +L
a
2
+ (z +L)
2
−
z
√
a
2
+z
2
41
Figure 9.1: H/M
o
=
1
2
z/L+1/2
√
1/25+(z/L+1/2)
2
−
z/L−1/2
√
1/25+(z/L−1/2)
2
0 z/L > 1/2 or1 z/L ≤ 1/2
The force becomes
Fz =
µoM
2
A
2
2L
√
a
2
+ 4L
2
−
L
√
a
2
+L
2
−
L
√
a
2
+L
2
+ 0
Now the condition that a <L, Fz = −
µoM
2
A
2
Now comparing the problem with 1.11. From the previous problem inside the rod, σM = M on the end above the
plain, and σM = −M below the plain. So
Bz =
µoM inside
0 outside
Because of the constant induced ﬁeld, this is similar to a constant electric ﬁeld. In fact, the magnetic potential is
similar to the electric potential of plane capacitors with σE/o →µoσM. To create a capacitor, lets separate the real
rod, which is the one above the plane, and the image rod. Now on page 43, F/A = σ
2
/2o for a capacitor. Then
F/A = σ
2
M
µ
2
o
/2µo →F/A = µoσ
2
M
/2. Finally the value of σM, F/A = µoM
2
/2 → F =
µo
2
AM
2
42
Figure 9.2: B/µ
o
M
o
=
1
2
z/L+1/2
√
1/25+(z/L+1/2)
2
−
z/L−1/2
√
1/25+(z/L−1/2)
2
43
44
Chapter 10
Problem Set 10
10.1 Problem 1: Jackson 5.14
A long, hollow, right circular cylinder of inner (outer) radius a (b), and of relative permeability µr, is placed in a
region of initially uniform magneticﬂux density Bo at right angles to the ﬁeld. Find the ﬂux density at all points
in space, and sketch the logarithm of the ratio of the magnitudes of B on the cylinder axis to Bo as a function of
log
10
µr for a
2
/b
2
= 0.5, 0.1. Neglect end eﬀects.
Because there are no end eﬀects, this problem can be reduced to a two dimensional problem. With the zaxis the
length of the cylinder and the Bo in the xdirection. With those being set, this is much like the problem done
in class. So ∇ B = ∇ (µH) = 0 and for inside and outside ∇ H = 0. The solution is H = −∇Φm where
Φ = ao + bo ln ρ +
∞
¸
n=1
anρ
n
(cos nϕ + sin nϕ) +
∞
¸
n=1
bnρ
−n
(cos nϕ + sin nϕ). All that remains is to solve for the
coeﬃcients. By symmetry the sine terms can be dropped. For ρ > b, an = 0, and for ρ < a, bn = 0. So the potential
is
Φm(ρ, ϕ) =
¸
∞
n=1
αnρ
n
cos nϕ ρ < a
¸
∞
n=1
βnρ
n
cos nϕ + γnρ
−n
cos nϕ
a < ρ < b
¸
∞
n=1
δnρ
−n
cos nϕ ρ > b
H has to be continuous at the boundaries. At ρ = b, the diﬀerential with respect to the radial is
Bo
µo
cos ϕ +
∞
¸
n=1
nb
−(n+1)
δn cos nϕ = µr
∞
¸
n=1
−n
βnb
n−1
−γnb
−(n+1)
cos nϕ
From this n = 1, so
Bo
µo
+δ1b
−2
= −µrβ1 +µrγ1b
−2
(10.1)
The tangential continuation at ρ = b
−
Bo
µo
+δ1b
−2
= β1 +γ1b
−2
(10.2)
Likewise at ρ = a
α1 = µrβ1 −µrγ1a
−2
(10.3)
α1 = β1 +γ1a
−2
(10.4)
Reducing the number of coeﬃcients in equations 1 and 2
β1 =
(µr −1)
2µr
δ1b
−2
−
(µr + 1)
2µr
Bo
µo
(10.5)
γ1 =
(µr + 1)
2µr
δ1 +
(1 −µr)
2µr
Bo
µo
b
2
(10.6)
45
Doing the same with equations 3 and 4
β1 =
(µr + 1)
2µr
α1 (10.7)
γ1 =
(µr −1)
2µr
α1a
2
(10.8)
Now combining equations 5 and 7
α1 =
(µr −1)
(µr + 1)
δ1b
−2
−
Bo
µo
(10.9)
Now for equations 6 and 8
α1 =
(µr + 1)
(µr −1)
δ1a
−2
−
b
2
a
2
Bo
µo
(10.10)
Equating 9 and 10
δ1 =
1 −
a
2
b
2
µ
2
r
−1
b
2
(µr + 1)
2
b
2
−(µr −1)
2
a
2
Bo
µo
b
2
(10.11)
With equation 9 and 11
α1 =
−4µrb
2
(µr + 1)
2
b
2
−(µr −1)
2
a
2
Bo
µo
(10.12)
Placing equation 12 into equation 7 and also equation 8
β1 =
−2 (µr + 1) b
2
(µr + 1)
2
b
2
−(µr −1)
2
a
2
Bo
µo
(10.13)
γ1 =
−2 (µr −1) b
2
(µr + 1)
2
b
2
−(µr −1)
2
a
2
Bo
µo
a
2
(10.14)
Now the ﬁeld is
H =
−α1 cos ϕˆ ρ +α1 sin ϕˆ ϕ ρ < a
−
β1 −
γ
1
ρ
2
cos ϕˆ ρ +
β1 +
γ
1
ρ
2
sin ϕˆ ϕ a < ρ < b
Bo
µo
+
δ
1
ρ
2
cos ϕˆ ρ −
Bo
µo
+
δ
1
ρ
2
sin ϕˆ ϕ ρ > b
See last page for graph
10.2 Problem 2: Jackson 5.27
A circuit consists of a long thin conducting shell of radius a and a parallel return wire of radius b on axis inside. Of
the current is assumed distributed uniformly throughout the cross section of the wire, calculate the selfinductance
per unit length. What is the selfinductance if the inner wire is a thin tube?
First let’s apply Ampere’s law
B dl = µoIenc (10.15)
So quickly, B is
B =
µoI
2π
ρ
b
2
ρ < b
µoI
2π
1
ρ
b < ρ < a
0 ρ > a
46
The energy of the magnetic ﬁeld per length of the system is given by Eq. 5.148
W =
1
2
H Bd
3
x =
1
2µo
B
2
d
3
x
=
2π
2µo
µoI
2π
2
b
0
ρ
3
dρ
b
4
+
a
b
dρ
ρ
=
π
µo
µoI
2π
2
1
4
+ ln
a
b
=
µoI
2
16π
4 ln
a
b
+ 1
From Eq. 5.152, W =
1
2
LI
2
. The selfinductance per unit length is
1
2
LI
2
=
µoI
2
16π
4 ln
a
b
+ 1
⇒ L =
µoI
2
8π
4 ln
a
b
+ 1
Now if the inner wire is a tube B = 0 for ρ < b, so there is only the middle term which yields a selfinductance per
unit length of
L =
µo
2π
ln
a
b
10.3 Problem 3: Jackson 5.29
The ﬁgure (See Jackson pg. 232) represents a transmission line consisting of two, parallel perfect conductors of
arbitrary, but constant, cross section. Current ﬂows down one conductor and returns via the other.
Show that the product of the inductance per unit length L and the capacitance per unit length C is
LC = µ (10.16)
where µ and are the permeability and the permittivity of the medium surrounding the conductors.
First using Ampere’s Law for outside the wires
B dl = µI ⇒B =
µI
2πρ
But inside the wire B = 0. The vector potential is
Az =
B
φ
dρ =
µI
2πρ
dρ ⇒Az = −
µI
2π
ln
ρ
ρo
where ρo = a, b. The work per unit length can be found from Eq. 5.149
W =
1
2
J Ad
3
x =
1
2
(JaAdaa +J
b
Ada
b
)
=
1
2
µI
2π
a
0
2πI
πa
2
ln
d
b
ρadρa +
b
0
2πI
πb
2
ln
d
a
ρ
b
dρ
b
=
1
2
µI
2π
a
2
πI
πa
2
ln
d
b
+
b
2
πI
πb
2
ln
d
a
=
µI
2
4π
ln
d
2
ab
Relating this to the selfinductance per unit length
1
2
LI
2
=
µI
2
4π
ln
d
2
ab
⇒L =
µ
2π
ln
d
2
ab
From the second problem 1.7, the capacitance is
C =
2π
ln
d
2
ab
(10.17)
47
Then
LC =
µ
2π
ln
d
2
ab
2π
ln
d
2
ab
⇒ LC = µ
Capacitance The electric ﬁeld for a wire is given by
E =
Q
2πo
ˆ r
r
The potential between the two wire is
Φ =
E d
l =
Q
2πo
¸
d−a
2
a
1
dr
r
+
d−a
1
a
2
dr
r
=
Q
2πo
¸
ln
d −a2
a1
+ ln
d −a1
a2
=
Q
2πo
¸
ln
(d −a2) (d −a1)
a1a2
The capacitance is given by C =
Q
Φ
. Then
C =
Q
Φ
=
Q
Q
2πo
ln
(d−a
2
)(d−a
1
)
a
1
a
2
=
2πo
ln
(d−a
2
)(d−a
1
)
a
1
a
2
Since the distance between the wires is much large than their radii, d−a1 ≈ d−a2 ≈ d. Then capacitance reduces to
C ≈
2πo
ln
d
2
a
1
a
2
48
Figure 10.1:
B
Bo
=
µr
(µr+1)
2
−(µr−1)
2 a
2
b
2
49
50
Chapter 11
Problem Set 11
11.1 Problem 1: Jackson 6.8
A dielectric sphere of dielectric constant and radius a is located at the origin. There is a uniform applied electric
ﬁeld Eo in the x direction. The sphere rotates with an angular velocity ω about the a axis. Show that there is a
magnetic ﬁeld H = −∇ΦM, where
ΦM =
3
5
−o
+ 2o
oEoω
a
r>
5
xy (11.1)
where r> is the larger of r and a. The motion is nonrelativistic.
First the polarization from Eq. 4.57
Px = 3
−o
+ 2o
Eoˆ x →Px = αEoˆ x (11.2)
So the volume charge density is −∇ P = 0. The surface charge density is
σ = ˆ n M = ˆ n ((r P)ωˆ z) = αoEoωa sin θ cos θ cos φ
= −αoEoωa
8π
15
1
2
[Y2,1 −Y2,−1]
The potential is
ΦM =
1
4π
σ
da
[x −x
[
= −αoEoωa
8π
15
1
2
a
2
4π
¸
4π
2l + 1
r
l
<
r
l+1
>
Y
lm
(θ, φ)
Y
∗
lm
(θ
, φ
)
Y2,1(θ
, φ
) −Y2,−1(θ
, φ
)
sin θ
dθ
dφ
= −αoEoωa
8π
15
1
2
a
2
5
r
2
<
r
3
>
[Y2,1(θ, φ) −Y2,−1(θ, φ)]
= −αoEoωa
8π
15
1
2
a
2
5
r
2
<
r
3
>
¸
−2
15
8π
sin θ cos θ cos φ
¸
=
3
5
−o
+ 2o
oEoωa
3
r
3
<
r
5
>
sin θ cos θ cos φ
ΦM =
3
5
−o
+ 2o
oEoωxz
a
r>
5
51
11.2 Problem 2: Jackson 6.20
An example of the preservation of causality and ﬁnite speed of propagation in spite of the use of the Coulomb gauge
is aﬀorded by a dipole source that is ﬂashed on and oﬀ at t = 0. The eﬀective charge and current densities are
ρ(x, t) = δ(x)δ(y)δ
(z)δ(t) (11.3)
Jz(x, t) = −δ(x)δ(y)δ(z)δ
(t) (11.4)
where the prime means diﬀerentiation with respect to the argument. This dipole is of unit strength and it points in
the negative Z direction.
(a) Show that the instantaneous Coulomb potential (6.23) is
Φ(x, t) = −
1
4πo
δ(t)
z
r
3
(11.5)
Using Eq. 6.23
Φ(x, t) =
1
4πo
ρ
x − x

d
3
x =
1
4πo
δ(x
)δ(y
)δ
(z
)δ(t)
d
3
x
x − x

= −
1
4πo
δ(t)
∂
∂z
1
z − z

= −
1
4πo
δ(t)
−
−z
− z
x − ˆ z

3
= −
1
4πo
δ(t)
z
r
3
(b) Show that the transverse current Jt with a factor of 2/3 multiplying the delta function coming from treating
the gradient of z/r
3
according to (4.20).
Jt(x, t) = −δ
(t)
¸
2
3
3δ(x) −
3
4πr
3
+
3
4πr
3
n(3 n)
(11.6)
From Eq. 6.25 and 6.29
Jt = J −J
l
= J −
1
µoc
2
∇
∂Φ
∂t
= −δ(x)δ(y)δ(z)δ
(t)ˆ z +
1
4π
∇
δ
(t)
z
r
3
= −δ(x)δ(y)δ(z)δ
(t)ˆ z +
1
4π
δ
(t)
4π
3
δ(x)δ(y)δ(z)ˆ z −
3ˆ r(ˆ z ˆ r) − ˆ z
r
3
= −δ
(t)
2
3
δ(x)ˆ z −
1
4πr
3
ˆ z +
3
4πr
3
ˆ r(ˆ z ˆ r)
(c) Show that the electric and magnetic ﬁelds are causal and that the electric ﬁeld components are (with Ey the
same as Ex except with cos φ replaced by sin φ),
Ex(x, t) =
1
4πo
c
r
¸
−δ
(r −ct) +
3
r
δ
(r −ct) −
3
r
2
δ(r −ct)
sin θ cos θ cos φ (11.7)
Ez(x, t) =
1
4πo
c
r
¸
sin
2
θδ
(r −ct) + (3 cos
2
θ −1)
δ
(r −ct)
r
−
δ(r −ct)
r
2
(11.8)
From section 6.3, A is causal, ∴ B is causal. Now for E. From equation 6.48
A =
µo
4π
1
R
[J]
ret
d
3
x
∂A
∂t
= −
µo
4π
1
R
∂
∂t
¸
δ
t −
R
c
δ(x
)ˆ z +
1
4π
∇
∂
∂z
1
r
d
3
x
= −
µo
4π
1
r
δ
t −
R
c
−
µo
4π
dk
2π
1
R
∂
2
∂t
2
e
ik(t−R/c)
1
4π
∇
∂
∂z
1
r
d
3
x
Now
∇
2
e
ik(t−R/c)
R
=
1
R
(−ik/c)
2
e
ik(t−R/c)
−e
ik(t−R/c)
4πδ(x −x
)
=
1
c
2
∂
2
∂t
2
e
ik(t−R/c)
R
−e
ikt
4πδ(x −x
)
52
With that
∂A
∂t
= −
µo
4π
1
r
δ
(t −R/c)ˆ z −
µoc
2
4π
dk
2π
∇
2
e
ik(t−R/c)
4πR
+e
ikt
δ(x −x
)
∇
∂
∂z
1
r
d
3
x
= −
µo
4π
1
r
δ
(t −R/c)ˆ z −
µo
4πµoo
dk
2π
e
ik(t−R/c)
4πR
∇
∂
∂z
∇
2
1
r
d
3
x
+δ(t)∇
∂
∂z
1
r
= −
µo
4π
1
r
δ
(t −R/c)ˆ z +
1
4πo
dk
2π
e
ik(t−R/c)
R
∇
∂
∂z
δ(x
)d
3
x
−
1
4πo
δ(t)∇
z
r
3
Now integrating the second term by parts
∂A
∂t
= −
µo
4π
1
r
δ
(t −R/c)ˆ z +
1
4πo
dk
2π
∇
∂
∂z
e
ik(t−R/c)
R
δ(x
)d
3
x
−
1
4πo
δ(t)∇
z
r
3
= −
µo
4π
1
r
δ
(t −R/c)ˆ z +
1
4πo
dk
2π
∇
∂
∂z
e
ik(t−r/c)
r
−
1
4πo
δ(t)∇
z
r
3
Again working with part of the second term
∇
∂
∂z
e
−ikr/c
r
= ∇
¸
−
ike
−ikr/c
cr
−
e
−ikr/c
r
2
∂
∂z
r
= −∇
¸
ik
cr
+
1
r
2
e
−ikr/c
cos θ
=
2
r
3
+
2ik
cr
2
−
k
2
c
2
r
cos θe
−ikr/c
ˆ r +e
−ikr/c
ik
cr
+
1
r
2
sin θ
r
ˆ
θ
Now integrating over kspace, which is just a Fourrier transform
∂A
∂t
= −
µo
4π
1
r
δ
(t −R/c)ˆ z +
1
4πo
¸
1
r
3
δ(t −R/c) −
1
cr
2
δ
(t −R/c)
2 cos θˆ r + sin θ
ˆ
θ
+
1
c
2
r
δ
(t −R/c) cos θˆ r
−
1
4πo
δ(t)∇
z
r
3
= −
µoc
3
4π
1
r
δ
(ct −R)
cos θˆ r −sin θ
ˆ
θ
+
c
4πo
¸
1
r
3
δ(ct −R) −
1
r
2
δ
(ct −R)
2 cos θˆ r + sin θ
ˆ
θ
+
1
r
δ
(ct −R) cos ˆ r
−
1
4πo
δ(t)∇
z
r
3
= −
1
4πo
c
r
¸
1
r
δ
(ct −R) −
1
r
2
δ(ct −R)
2 cos θˆ r + sin θ
ˆ
θ
−δ
(ct −R) sin θ
ˆ
θ
−
1
4πo
δ(t)∇
z
r
3
Now this can be used to ﬁnd the electric ﬁeld
E = −∇Φ −
∂A
∂t
=
1
4πo
δ(t)∇
z
r
3
−
1
4πo
c
r
¸
1
r
δ
(ct −R) −
1
r
2
δ(ct −R)
2 cos θˆ r + sin θ
ˆ
θ
−δ
(ct −R) sin θ
ˆ
θ
−
1
4πo
δ(t)∇
z
r
3
= −
1
4πo
c
r
¸
1
r
δ
(ct −R) −
1
r
2
δ(ct −R)
2 cos θˆ r + sin θ
ˆ
θ
−δ
(ct −R) sin θ
ˆ
θ
Now the electric ﬁeld in Cartesian coordinates
Ex = −
1
4πo
c
r
¸
1
r
δ
(ct −R) −
1
r
2
δ(ct −R)
(2 cos θ sin θ cos φ + sin θ cos θ cos φ) −δ
(ct −R) sin θ cos θ cos φ
=
1
4πo
c
r
¸
−δ
(r −ct) +
3
r
δ
(r −ct) −
3
r
2
δ(r −ct)
sin θ cos θ cos φ
Ey = −
1
4πo
c
r
¸
1
r
δ
(ct −R) −
1
r
2
δ(ct −R)
(2 cos θ sin θ sin φ + sin θ cos θ sin φ) −δ
(ct −R) sin θ cos θ sin φ
=
1
4πo
c
r
¸
−δ
(r −ct) +
3
r
δ
(r −ct) −
3
r
2
δ(r −ct)
sin θ cos θ sin φ
Ez = −
1
4πo
c
r
¸
1
r
δ
(ct −R) −
1
r
2
δ(ct −R)
2 cos
2
θ −sin
2
θ
+δ
(ct −R) sin
2
θ
=
1
4πo
c
r
¸
sin
2
θδ
(r −ct) + (3 cos
2
θ −1)
δ
(r −ct)
r
−
δ(r −ct)
r
2
So the electric ﬁeld is causal because the only time dependence is in the δfunctions.
53
11.3 Problem 3: Jackson 6.11
A transverse plane wave is incident normally in vacuum on a perfectly absorbing ﬂat screen.
(a) From the law of conservation of linear momentum, show that the pressure (called radiation pressure) exerted on
the screen is equal to the ﬁeld energy per unit volume in the wave.
The total electromagnetic momentum in an inﬁnitesimal volume for a small area, ∆A over a small time, ∆t,
∆P = g∆Ac∆t
The force due to the momentum, F = ∆P/∆t = g∆Ac. The pressure is then perpendicular to the force per area.
p =
g∆Ac
∆A
= gc =
S
c
= energy per unit volume or u
(b) In the neighborhood of the earth the ﬂux of electromagnetic energy from the sun is approximately 1.4 W/m
2
.
If an interplanetary “sailplane” had a sail of mass 1 g/m
2
of area and negligible other weight, what would be
its maximum acceleration in meters per second squared due to the solar radiation pressure? How does this
compare with the acceleration due to the solar “wind” (corpuscular radiation)?
The acceleration do the radiation pressure
a
rad
=
S/c
m/A
=
1.4 10
3
/(3 10
8
)
10
−6
⇒a
rad
≈ 4.67 10
−3
m/s
2
There are 10
5
protons/m
2
s with an average velocity v = 4 10
5
m/s
a
wind
=
mv/tA
m/A
=
1.67 ×10
−27
×4 ×10
5
×10
5
10
−6
= 6.68 ×10
−14
m/s
2
So a
rad
a
wind
54
Chapter 12
Problem Set 12
12.1 Problem 1: Jackson 6.4
A uniformly magnetized and conducting sphere of radius R and total magnetic moment m = 4πMR
3
/3 rotates about
its magnetization axis with angular speed ω. In the steady state no current ﬂows in the conductor. The motion is
nonrelativistic; the sphere has no excess charge on it.
(a) By considering Ohm’s law in the moving conductor, show that the motion induces an electric ﬁeld and a uniform
volume charge density in the conductor, ρ = −mω/πc
2
R
3
.
Here there are no free currents, so there is scalar potential that can describe the magnetic ﬁeld. H = B/µo −M→
∇ H = −∇ M, then ∇
2
ΦM = ∇ M. Here M = Mη (R −r) ˆ z
1
, so
∇ M =
∂
∂z
Mη(R −r) = −Mδ(r −R)
∂r
∂z
= −Mδ(r −R)
z
r
= −Mδ(r −R) cos θ
Then
∇
2
Φ
M
= −Mδ(r −R) cos θ ⇒Φ =
1
4π
Mδ(r −R) cos θ
r −r

d
3
r
This integral can be solved using a multipole expansion as done with electrostatics for inside a sphere.
Φ
M
=
1
4π
∞
¸
l=0
¸
m
4π
2l + 1
Mδ(r
− R) cos θ
r
l+1
Y
∗
lm(θ
, ϕ
)d
3
r
r
l
Y
lm
(θ, ϕ)
=
1
4π
∞
¸
l=0
l
¸
m=−l
4π
2l + 1
4π
3
M
R
2
R
l+1
Y
10
Y
∗
lm
dΩr
l
Y
lm
=
1
4π
∞
¸
l=0
l
¸
m=−l
4π
2l + 1
4π
3
Mδ
l,1
δ
m,0
r
l
Y
lm
=
1
3
Mr cos θ =
1
3
Mz
The magnetic ﬁeld inside the sphere is H = −∇ Φ =
1
3
Mˆ z. The magnetic induction is
B = µo (H+M) = µo
1
3
Mˆ z +
1
3
Mˆ z
⇒B =
2
3
µoMˆ z
From Ohm’s Law
E = −v × B = −(ω × r) × B =
2
3
µoMωˆ z × (ˆ z × r) =
2
3
µoMω (ˆ z · rˆ z − r)
E = −
2
3
µoMω (xˆ x + yˆ y) = −
1
2
µomω
R
3
(xˆ x + yˆ y)
The charge density is
∇ E =
ρ
o
= −
4
3
µoMω = −
4
3
Mω
oc
2
= −
mω
πoπc
2
R
3
∇ E =
ρ
o
= −
mω
πoπc
2
R
3
⇒ ρ = −
mω
ππc
2
R
3
1
Here η is a step function
55
(b) Because the sphere is electrically neutral, there is no monopole electric ﬁeld outside. Use symmetry arguments
to show that only a quadrupole ﬁeld exists outside and that the quadrupole moment tensor has nonvanishing
components, Q33 = −4mωR
2
/3c
2
, Q11 = Q22 = −Q33/2.
The surface of the sphere has to have positive charge some place on it, just as it has to have negative charge on
it. These charges cancel out because there is no monopole term. Now considering a ﬂip of the radial vector. This
obviously does not change the sign of the angular momentum. This will change the sign of ∇ and J, so the magnetic
ﬁeld does not change signs. From this the magnetic induction does not change sign. Because ρ is independent of the
radial vector, it does not change sign. This means that under a radial ﬂip, the problem is the same, so the odd terms
of l must vanish. Since the dipole is odd in l, it is zero, therefore the ﬁrst nonvanishing term is the quadrupole.
Because of the azimuthal symmetry about z axis, the Qxx term can not be distinguished from the Qyy term. From
this, Qxx = Qyy. Because the quadrupole term is a traceless matrix, Qxx +Qyy +Qzz = 0 ⇒Qxx = Qyy = −Qzz/2.
For the oﬀ diagonal terms, under x → −x, y, z → y, z are the Qxy and Qxz terms. From arguments before, which
are r → −r terms are 0, Qxy = Qxz = 0. Now for y → −y x, z → x, z, which are Qyx (which from before is 0) and
Qyz, are 0 from the fact that the is equivalent to r →−r.
I show that Qzz = −4mωR
2
/3c
2
in the next part.
(c) By considering the radial electric ﬁelds inside and outside the sphere, show that the necessary surfacecharge
density σ(θ) is
σ(θ) =
1
4πR
2
4mω
3c
2
¸
1 −
5
2
P2(cos θ)
(12.1)
For inside the sphere from part a
E = −
2
3
µoMω (xˆ x + yˆ y)) = −∇Φ1 ⇒Φ1 =
µoMω
3
x
2
+ y
2
+ C
In terms of spherical harmonics,
Φ1 =
µoMω
3
x
2
+y
2
+C ⇒Φ1 =
µoMω
3
r
2
sin
2
θ +C
=
µoMω
3
r
2
2
3
1 −
4π
5
Y20
+C
=
2
9
µoMωr
2
1 −
4π
5
Y20
+C
For outside the sphere, there is no charge and no dipole terms, so the ﬁrst no vanishing term is the quadrupole term.
The potential outside is
Φ2 = A2
R
r
3
Y20
The potential must be continuous at the boundaries, so
Φ
1
(R) = Φ
2
(R) ⇒
2
9
µoMωR
2
1 −
4π
5
Y
20
+C = A
2
Y
20
Then C = −
2
9
µoMωR
2
and A2 = −
2
9
4π
5
µoMωR
2
. The displacement ﬁeld is
D =
−
2
3
Mω
c
2
r sin
2
θˆ r −
2
3
Mω
c
2
r sin θ cos θ
ˆ
θ r < R
−
2
3
Mω
c
2
R
5
r
4
P
2
(cos θ) ˆ r −
2
3
Mω
c
2
R
5
r
4
sin θ cos θ
ˆ
θ r > R
56
From ∇ D = ρ with ρ evaluated at r = R, the surface charge density becomes
σ(θ, ϕ) = −
2
3
Mω
c
2
RP2(cos θ) +
2
3
Mω
c
2
Rsin
2
θ
= −
2
3
Mω
c
2
RP2(cos θ) +
2
3
Mω
c
2
R
2
3
−
2
3
P2(cos θ)
= −
4
9
Mω
c
2
R
1 −
5
2
P2(cos θ)
=
1
4πR
2
4mω
3c
2
¸
1 −
5
2
P2(cos θ)
For Qzz
Qzz = 2
ρr
2
P
2
(cos θ)d
3
r
= 2
−
mω
πc
2
R
3
η(R − r) +
1
4πR
2
4mω
3c
2
¸
1 −
5
2
P
2
(cos θ)
δ(r − R)
r
2
d
3
r
=
4mω
c
2
R
3
−η(R − r) +
R
3
¸
1 −
5
2
P
2
(cos θ)
δ(r − R)
r
4
drd(cos θ)
=
10
3
mωR
2
c
2
−1
1
(P
2
(cos θ))
2
d(cos θ)
Qzz = −
4mωR
2
3c
2
(d) The rotating sphere serves as a unipolar induction device if a stationary circuit is attached by a slip ring to the
pole and a sliding contact to the equator. Show that the line integral of the electric ﬁeld from the equator
contact to the pole contact (by any path) is c = µomω/4πR
The c´T is given by
c =
r=R,θ=0
r=R,θ=π/2
E dl =
r=R,θ=π/2
r=R,θ=0
∇Φ dl = Φ(R, π/2) −Φ(R, 0)
=
µoMωR
2
3
−
2
9
µoMωR
2
+
2
9
µoMωR
2
=
µoMωR
2
3
c =
µomω
4πR
12.2 Problem 2: Jackson 6.14
An ideal circular parallel plate capacitor of radius a and plate separation d <a is connected to a current source by
axial leads. The current in the wire is I(t) = Io cos ωt.
(a) Calculate the electric and magnetic ﬁelds between the plates to second order in powers of the frequency (or wave
number), neglecting the eﬀects of fringing ﬁelds.
As in the static case E
θ
= Eρ = 0 and Bρ = Bz = 0. From Gauss’s law in statics
Ez =
σ
o
=
Q
πa
2
o
=
Io
−iωπa
2
o
From Ampere’s law
2πρB
θ
= µoo(−iω)πρ
2
Ez ⇒B
θ
=
µoIoρ
2πa
2
From Faraday’s law, the next order correction is
∇E =
∂Ez
∂ρ
ˆ
θ = −
∂B
∂t
= iωB
θ
ˆ
θ
57
Then
Ez(ρ) = Ez(0) −iω
B
θ
dρ ⇒Ezd =
Io
−iωπa
2
o
−
iωµoIoρ
2
4πa
2
Ez(ρ) =
Io
−iωπa
2
o
1 −
1
4
ω
2
ρ
2
c
2
Now the correction to the Bﬁeld.
∇B =
1
ρ
∂(ρB
θ
)
∂ρ
= µoo
∂E
∂t
= −iµooωEz ˆ z
So
B
θ
(ρ) = −
iµooω
ρ
Ezρdρ = −
iµooω
ρ
Io
−iωπa
2
o
ρ
2
2
−
1
16
ω
2
ρ
4
c
2
B
θ
(ρ) =
µoIoρ
2πa
2
1 −
1
8
ω
2
ρ
2
c
2
Now for the time dependence, because the time dependence is of the form e
−iωt
= cos ωt −i sin ωt, the real parts are
Ez(ρ) =
Io
ωπoa
2
1 −
1
4
ω
2
ρ
2
c
2
sin ωt
B
θ
(ρ) =
µoIoρ
2πa
2
1 −
1
8
ω
2
ρ
2
c
2
cos ωt
(b) Calculate the volume integrals of we and wm that enter the deﬁnition of the reactance X, (6.140), to second
order in ω. Show that in terms of the input current Ii, deﬁned by Ii = −iωQ, where Q is the total charge on
one plate, these energies are
wed
3
x =
1
4πo
I
i

2
d
ω
2
a
2
,
wmd
3
x =
µo
4π
I
i

2
d
8
¸
1 +
ω
2
a
2
12c
2
¸
(12.2)
The energy in the ﬁelds is we =
oE
2
4
and wm =
B
2
µo4
. From part a
we =
o[E[
2
4
=
o[Ii[
2
4 (ωπoa
2
)
2
wm =
[B[
2
µo4
≈
µo[Ii[
2
ρ
2
16π
2
a
4
1 +
1
8
ω
2
ρ
2
c
2
The integral for we over all of space in between the plates
o[Ii[
2
4 (ωπoa
2
)
2
ρdρdθdz =
2
4πo
[Ii[
2
d
ω
2
a
4
a
0
ρdρ =
1
4πo
[Ii[
2
d
ω
2
a
2
For wm
µoIi
2
ρ
2
16π
2
a
4
1 +
1
8
ω
2
ρ
2
c
2
ρdρdθdz =
µo
4π
Ii
2
d
2a
4
ρ
3
+
1
8
ω
2
ρ
5
c
2
dρ
µo
4π
Ii
2
d
2a
4
a
4
4
+
1
48
ω
2
a
6
c
2
=
µo
4π
Ii
2
d
8
1 +
ω
2
a
2
12c
2
(c) Show that the equivalent series circuit has C · πoa
2
/d, L · µod/8π, and that an estimate for the resonant
frequency of the system is ωres · 2
√
2c/a. Compare with the ﬁrst root of Jo(x).
From section 6.9, the capacitance
X =
1
ωC
=
4ω
[Ii[
2
1
4πo
[Ii[
2
d
ω
2
a
2
=
d
ωπoa
2
⇒C =
πoa
2
d
For the inductance
X = ωL =
4ω
[Ii[
2
µo
4π
[Ii[
2
d
8
=
ωµod
8π
⇒L =
µod
8π
58
Resonance occurs when ωL = 1/ωC,
ωL =
1
ωC
⇒ω =
1
√
LC
=
1
µod
8π
πoa
2
d
=
8
µooa
2
=
8c
2
a
2
= 2
√
2c/a
If the next term in the Bﬁeld is kept the inductance becomes
L =
µod
8π
1 +
ω
2
a
2
12c
2
The resonance is
ω
2
=
8c
2
a
2
1 +
1
12
ω
2
a
2
c
2
−1
≈
8c
2
a
2
1 −
1
12
ω
2
a
2
c
2
=
8c
2
a
2
−
2
3
ω
2
⇒ω =
24
5
c
a
So the further the electric ﬁeld and the induced magnetic ﬁeld are expand in power of ω, the closer the resonant
frequency will be to Jo(x) = 0, where x = ω. This seems logical because the ﬁrst resonant frequency should occur
when half a wavelength is inside the cavity.
59
60
Appendix A
Special Functions
Hi
61
Index
hydrogen, 2
mean value theorem, 4
62
This action might not be possible to undo. Are you sure you want to continue?