Solutions to Problems in Jackson,

Classical Electrodynamics, Third Edition

Homer Reid
June 15, 2000
Chapter 3: Problems 11-18
Problem 3.11
A modied Bessel-Fourier series on the interval 0 a for an arbitrary function
f() can be based on the homogenous boundary conditions:
At = 0, J

(k)
d
d
J

(k

) = 0
At = a,
d
d
ln[J

(k)] =

a
( real)
The rst condition restricts . The second condition yields eigenvalues k = y
n
/a,
where y
n
is the nth positive root of xdJ

(x)/dx +J

(x) = 0.
(a) Show that the Bessel functions of dierent eigenvalues are orthogonal in the
usual way.
(b) Find the normalization integral and show that an arbitrary function f() can
be expanded on the interval in the modied Bessel-Fourier series
f() =

n=1
A
n
J

_
y
n
a
_
with the coecients A
n
given by
A
n
=
2
a
2
_
_
1

2
y
2
n
_
J
2

(y
n
) +
_
dJ

(y
n
)
dy
n
_
2
_
1
_
a
0
f()J

_
y
n

a
_
d.
1
Homer Reids Solutions to Jackson Problems: Chapter 3 2
(a) The function J

(k) satises the equation

1

d
d
_

d
d
J

(k)
_
+
_
k
2

2
_
J

(k) = 0. (1)
Multiplying both sides by J

(k

) and integrating from 0 to a gives

_
a
0
_
J

(k

)
d
d
_

d
d
J

(k)
_
+
_
k
2


2

_
J

(k

)J

(k)
_
d = 0. (2)
The rst term on the left can be integrated by parts:
_
a
0
J

(k

)
d
d
_

d
d
J

(k)
_
d
=

(k

)
d
d
J

(k)

a
0

_
a
0

_
d
d
J

(k

)
_ _
d
d
J

(k)
_
d. (3)
One of the conditions were given is that the thing in braces in the rst term
here vanishes at = 0. At = a we can invoke the other condition:
d
d
ln[J

(k)]

=a
=
1
J

(k)
d
d
J

(k)

=a
=

a
a
d
d
J

(ka) = J

(ka).
Plugging this into (3), we have
_
a
0
J

(k

)
d
d
_

d
d
J

(k)
_
d
= J

(k

)J

(k)
_
a
0

_
d
d
J

(k

)
_ _
d
d
J

(k)
_
. (4)
This is clearly symmetric in k and k

, so when we write down (2) with k and

k

switched and subtract from (2), the rst integral (along with the
2
/ term)
vanishes, and we are left with
(k
2
k
2
)
_
a
0
J

(k

)J

(k) d = 0
proving orthogonality.
(b) If we multiply (1) by
2
J

(k) and integrate, we nd

_
a
0
J

(k)
d
d
[J

(k)]d+k
2
_
a
0

2
J

(k)J

(k)d
2
_
a
0
J

(k)J

(k)d = 0.
(5)
Homer Reids Solutions to Jackson Problems: Chapter 3 3
The rst and third integrals are of the form
_
f(x)f

(x)dx and can be done

immediately. In the second integral we put f() =
2
J

(k), g

() = J

(k)
and integrate by parts:
_
a
0

2
J

(k)J

(k)d =

2
J
2

(k)

a
0
2
_
a
0
J
2

(k)d
_
a
0

2
J

(k)J

(k)d

_
a
0

2
J

(k)J

(k)d =
1
2
a
2
J
2

(ka)
_
a
0
J
2

(k)d.
Using this in (5),
a
2
2
J
2

(ka) +
(ak)
2
2
aJ
2

(ka) k
2
_
a
0
J
2

(k)d

2
2
J
2

(ka) = 0
so
_
a
0
J
2

(k)d =
_
a
2
2


2
2k
2
_
J
2

(ka) +
a
2
2k
2
J
2

(ka)
=
a
2
2
_
_
1

2
(ka)
2
_
J
2

(ka) +
_
d
d(ka)
J

(ka)
_
2
_
This agrees with what Jackson has if you note that k is chosen such that ka =
y
nm
.
Problem 3.12
An innite, thin, plane sheet of conducting material has a circular hole of radius a
cut in it. A thin, at, disc of the same material and slightly smaller radius lies in
the plane, lling the hole, but separated from the sheet by a very narrow insulating
ring. The disc is maintained at a xed potential V , whilc the innite sheet is kept
at zero potential.
(a) Using appropriate cylindrical coordinates, nd an integral expression involving
Bessel functions for the potential at any point above the plane.
(b) Show that the potential a perpendicular distance z above the center of the disc
is

0
(z) = V
_
1
z

a
2
+z
2
_
(c) Show that the potential a perpendicular distance z above the edge of the disc
is

a
(z) =
V
2
_
1
kz
a
K(k)
_
where k = 2a/(z
2
+ 4a
2
)
1/2
, and K(k) is the complete elliptic integral of the
rst kind.
Homer Reids Solutions to Jackson Problems: Chapter 3 4
(a) As before, we can write the potential as a sum of terms R()Q()Z(z). In
this problem there is no dependence, so Q = 1. Also, the boundary conditions
on Z are that it vanish at and be nite at 0, whence Z(z) exp(kz) for
any k. Then the potential expansion becomes
(, z) =
_

0
A(k)e
kz
J
0
(k) dk. (6)
To evaluate the coecients A(k), we multiply both sides by J
0
(k

) and inte-
grate over at z = 0:
_

0
(, 0)J
0
(k

) d =
_

0
A(k)
__

0
J
0
(k)J
0
(k

) d
_
dk
=
A(k

)
k

so
A(k) = k
_

0
(, 0)J
0
(k) d
= kV
_
a
0
J
0
(k)d.
Plugging this back into (6),
(, z) = V
_

0
_
a
0
k

e
kz
J
0
(k)J
0
(k

) d

dk. (7)
The

integral can be done right away. To do it, I appealed to the dierential

equation for J
0
:
J

0
(u) +
1
u
J

0
(u) +J
0
(u) = 0
so
_
x
0
uJ
0
(u) du =
_
x
0
uJ

0
du
_
x
0
J

0
(u) du
= |uJ

0
(u)|
x
0
+
_
x
0
J

0
(u) du
_
x
0
J

0
(u) du
= |uJ

0
(u)|
x
0
= xJ

0
(x) = xJ
1
(x).
(In going from the rst to second line, I integrated by parts.) Then (7) becomes
(, z) = aV
_

0
J
1
(ka)J
0
(k)e
kz
dk. (8)
Homer Reids Solutions to Jackson Problems: Chapter 3 5
(b) At = 0, (7) becomes
(0, z) = V J
0
(0)
_
a
0

__

0
ke
kz
J
0
(k

)dk
_
d

= V
_
a
0

z
_

0
e
kz
J
0
(k

)dk
_
d

= V
_
a
0

z
_
1
_

2
+z
2
__
d

= V
_
a
0
z

(
2
+z
2
)
3/2
d

Here we substitute u =
2
+z
2
, du = 2

d:
(0, z) =
V zJ
0
(0)
2
_
a
2
+z
2
z
2
u
3/2
du
= V z

1
u
1/2

a
2
+z
2
z
2
= V z
_
1
z

1

z
2
+z
2
_
= V
_
1
z

a
2
+z
2
_
(b) At = a, (8) becomes
(a, z) = aV
_

0
J
1
(ka)J
0
(ka)e
kz
dk
Problem 3.13
Solve for the potential in Problem 3.1, using the appropriate Green
function obtained in the text, and verify that the answer obtained
in this way agrees with the direct solution from the dierential
equation.
For Dirichlet boundary value problems, the basic equation is
(x) =
1

0
_
V
G(x; x

)(x

) dV

+
_
S
(x

)
G(x; x

)
n

dA

. (9)
Here there is no charge in the region of interest, so only the surface integral
contributes. The Greens function for the two-sphere problem is
G(x; x

) =

l=0
l

m=l
Y

lm
(

) Y
lm
(, )
2l + 1
R
l
(r; r

) (10)
Homer Reids Solutions to Jackson Problems: Chapter 3 6
with
R
l
(r; r

) =
1
_
1
_
a
b
_
2l+1
_
r
l
<

a
2l+1
r
l+1
<
__
1
r
l+1
>

r
l
>
b
2l+1
_
. (11)
Actually in this case the potential cannot have any dependence, so all terms
with m = 0 in (10) vanish, and we have
G(x; x

) =
1
4

l=0
P
l
(cos )P
l
(cos

)R
l
(r; r

).
In this case the boundary surfaces are spherical, which means the normal to a
surface element is always in the radial direction:

n
G(x; x

) =
1
4

l=0
P
l
(cos )P
l
(cos

)

n
R
l
(r; r

).
The surface integral in (9) has two parts: one integral S
1
over the surface of
the inner sphere, and a second integral S
2
over the surface of the outer sphere:
S
1
=
1
4

l=0
P
l
(cos )
R
l
n

=a
__

0
_
2
0
(a,

)P
l
(cos

)a
2
sin

dd

_
=
V
2

l=0
a
2
P
l
(cos )
R
l
n

=a
__
1
0
P
l
(x) dx
_
=
V
2

l=0
a
2

l
P
l
(cos )
R
l
n

=a
where

l
=
_
1
0
P
l
(x) dx
= (
1
2
)
(l1)/2
(l 2)!!
2[(l + 1)/2]!
, l odd
= 0, l even.
A similar calculation gives
S
2
=
V
2

l=0
b
2
P
l
(cos )
R
l
n

=b
__
0
1
P
l
(x) dx
_
=
V
2

l=0
b
2

l
P
l
(cos )
R
l
n

=b
because P
l
is odd for l odd, so its integral from -1 to 0 is just the negative of
the integral from 0 to 1. The nal potential is the sum of S
1
and S
2
:
(r, ) =
V
2

l=0

l
P
l
(cos )

r
2
R
l
n

=b
r

=a
(12)
Homer Reids Solutions to Jackson Problems: Chapter 3 7
Since the point of interest is always between the two spheres, to nd the
normal derivative at r = a we dierentiate with respect to r
<
, and at r = b
with respect to r
>
. Also, at r = a the normal is in the +r direction, while at
r = b the normal is in the negative r direction.
a
2

n
R
l
(r; r

=a
= (2l + 1)a
2
a
l1
_
1
_
a
b
_
2l+1
_
1
r
l+1

r
l
b
2l+1
_
b
2

n
R
l
(r; r

=b
= (2l + 1)b
2
b
(l+2)
_
1
_
a
b
_
2l+1
_
r
l

a
2l+1
r
l+1
_
Combining these with some algebra gives
(r, ) =
V
2

l=0
(2l + 1)
l
P
l
(cos )
_
(ab)
l+1
(b
l
+a
l
)r
(l+1)
(a
l+1
+b
l+1
)r
l
b
2l+1
a
2l+1
_
in agreement with what we found in Problem 3.1.
Problem 3.14
A line charge of length 2d with a total charge Q has a linear charge
density varying as (d
2
z
2
), where z is the distance from the mid-
point. A grounded, conducting spherical shell of inner radius b > d
is centered at the midpoint of the line charge.
(a) Find the potential everywhere inside the spherical shell as an
expansion in Legendre polynomials.
(b) Calculate the surface-charge density induced on the shell.
(c) Discuss your answers to parts a and b in the limit that d << b.
First of all, we are told that the charge density (z) = (d
2
z
2
), and that
the total charge is Q, whence
Q = 2
_
d
0
(d
2
z
2
)dz =
4
3
d
3

=
3Q
4d
3
.
In this case we have azimuthal symmetry, so the Greens function is
G(x; x

) =
1
4

l=0
P
l
(cos

)P
l
(cos )R
l
(r; r

) (13)
Homer Reids Solutions to Jackson Problems: Chapter 3 8
with
R
l
(r; r

) = r
l
<
_
1
r
l+1
>

r
l
>
b
2l+1
_
.
Since the potential vanishes on the boundary surface, the potential inside the
sphere is given by
(r, ) =
1

0
_
V
G(r, ; r

)(r

)dV.
In this case is only nonzero on the z axis, where r = z. Also, P
l
(cos )=1 for
z > 0, and (1)
l
for z < 0. This means that the contributions to the integral
from the portions of the line charge for z > 0 and z < 0 cancel out for odd l,
and add constructively for even l:
(r, ) =
1
4
0

l=0,2,4,...
P
l
(cos )
_
2
_
d
0
R
l
(r; z)(z) dz
_
We have
_
d
0
R
l
(r; z)(z) dz =
_
d
0
r
l
<
_
1
r
l+1
>

r
l
>
b
2l+1
_
(d
2
z
2
) dz
This is best split up into two separate integrals:
=
_
d
0
r
l
<
r
l+1
>
(d
2
z
2
) dz

b
2l+1
_
d
0
r
l
<
r
l
>
(d
2
z
2
) dz
The second integral is symmetric between r and r

, so we may integrate it
directly:

b
2l+1
_
d
0
r
l
<
r
l
>
(d
2
z
2
) dz =
r
l
b
2l+1
_
d
0
z
l
(d
2
z
2
) dz
=
r
l
b
2l+1
_
d
l+3
l + 1

d
l+3
l + 3
_
=
r
l
d
l+3
(l + 1)(l + 3)b
2l+1
(14)
The rst integral must be further split into two:

_
d
0
r
l
<
r
l+1
>
(d
2
z
2
) dz
Homer Reids Solutions to Jackson Problems: Chapter 3 9
=
_
1
r
l+1
_
r
0
z
l
(d
2
z
2
) dz +r
l
_
d
r
d
2
z
2
z
l+1
dz
_
=
_
1
r
l+1
_
d
2
r
l+1
l + 1

r
l+3
l + 3
_
+r
l

d
2
lz
l
+
1
(l 2)z
l2

d
r
_
=
_
d
2
l + 1

r
2
l + 3
+
_
r
d
_
l
d
2
2
l(l + 2)

d
2
l
+
r
2
l + 2
_
=
_
r
2
(l + 2)(l + 3)

d
2
l(l + 1)
+
_
r
d
_
l
d
2
2
l(l + 2)
_
Combining this with (14), we have
_
d
0
R
l
(r; z)(z) dz =
_
r
2
(l + 2)(l + 3)

d
2
l(l + 1)
+
_
r
d
_
l
d
2
2
l(l + 2)

r
l
d
l+3
(l + 1)(l + 3)b
2l+1
_
(15)
But something is wrong here, because with this result the nal potential will
contain terms like r
0
P
l
(cos ) and r
2
P
l
(cos ), which do not satisfy the Laplace
equation.
Homer Reids Solutions to Jackson Problems: Chapter 3 10
Problem 3.15
Consider the following spherical cow model of a battery con-
nected to an external circuit. A sphere of radius a and conductiv-
ity is embedded in a uniform medium of conductivity

. Inside
the sphere there is a uniform (chemical) force in the z direction
acting on the charge carriers; its strength as an eective electric
eld entering Ohms law is F. In the steady state, electric elds
exist inside and outside the sphere and surface charge resides on
its surface.
(a) Find the electric eld (in addition to F) and current density
everywhere in space. Determine the surface-charge density
and show that the electric dipole moment of the sphere is
p = 4
0
a
3
F/( + 2

).
(b) Show that the total current owing out through the upper
hemisphere of the sphere is
I =
2

+ 2

a
2
F
Calculate the total power dissipation outside the sphere. Us-
ing the lumped circuit relations, P = I
2
R
e
= IV
e
, nd the
eective external resistance R
e
and voltage V
e
.
(c) Find the power dissipated within the sphere and deduce the
eective internal resistance R
i
and voltage V
i
.
(d) Dene the total voltage through the relation V
t
= (R
e
+R
i
)I
and show that V
t
= 4aF/3, as well as V
e
+ V
i
= V
t
. Show
that IV
t
is the power supplied by the chemical force.
(a) Whats going on in this problem is that the conductivity has a discontinu-
ity going across the boundary of the sphere, but the current density must be
constant there, which means there must an electric eld discontinuity in inverse
proportion to the conductivity discontinuity. To create this electric eld discon-
tinuity, there has to be some surface charge on the sphere, and this charge gives
rise to extra elds both inside and outside the sphere.
Since there is no charge inside or outside the sphere, the potential in those
two regions satised the Laplace equation, and may be expanded in Legendre
polynomials:
Homer Reids Solutions to Jackson Problems: Chapter 3 11
for r < a, (r, ) =
in
(r, ) =

l=0
A
l
r
l
P
l
(cos )
for r > a, (r, ) =
out
(r, ) =

l=0
B
l
r
(l+1)
P
l
(cos )
Continuity at r = a requires that
A
l
a
l
= B
l
a
l+1
B
l
= a
2l+1
A
l
so
(r, ) =
_

in
(r, ) =

l=0
A
l
r
l
P
l
(cos ), r < a

out
(r, ) =

l=0
A
l
a
2l+1
r
(l+1)
P
l
(cos ), r > a.
(16)
Now, in the steady state there can be no discontinuities in the current den-
sity, because if there were than there would be more current owing into some
region of space than out of it, which means charge would pile up in that region,
which would be a growing source of electric eld, which would mean we arent
in steady state. So the current density is continuous everywhere. In particular,
the radial component of the current density is continuous across the boundary
of the sphere, i.e.
J
r
(r = a

, ) = J
r
(r = a
+
, ). (17)
Outside of the sphere, Ohms law says that
J =

E =

out
.
Inside the sphere, there is an extra term coming from the chemical force:
J = (E +F

k) = (
in
+F

k).
Applying (17) to these expressions, we have

in

r=a
+F cos
_
=

out

r=a
Using (16), this is
FP
1
(cos )

l=0
lA
l
a
l1
P
l
(cos ) =
_

l=0
(l + 1)A
l
a
l1
P
l
(cos ).
Multiplying both sides by P
l
(cos ) and integrating from to , we nd
F A
1
=
_

_
2A
1
(18)
Homer Reids Solutions to Jackson Problems: Chapter 3 12
for l=1, and
lA
l
=
_

_
(l + 1)A
l
(19)
(20)
for l = 1. Since the conductivity ratio is positive, the second relation is impos-
sible to satisfy unless A
l
= 0 for l = 1. The rst relation becomes
A
1
=

+ 2

F.
Then the potential is
(r, ) =
_

+2

Fr cos , r < a

+2

Fa
3
r
2
cos , r > a
(21)
The dipole moment p is dened by
(r, )
1
4
0
p r
r
3
as r . (22)
The external portion of (21) can be written as
(r, ) =

+ 2

Fa
3
z
r
3
and comparing this with (22) we can read o
p = 4
0

+ 2

Fa
3

k.
The electric eld is found by taking the gradient of (21):
E(r, ) =
_

+2

k, r < a

+2

F
_
a
r
_
3
(2 cos r + sin

), r > a
The surface charge
s
() on the sphere is proportional to the discontinuity
in the electric eld:

s
() =
0
[E
r
(r = a
+
) E
r
(r = a

)]
=
3
0

+ 2

F cos .
(b) The current owing out of the upper hemisphere is just
_
J dA =
_
(E
in
+F

k) dA
=
_
1

+ 2

_
F
_
/2
0
_
2
0
cos sin a
2
dd
= 2

+ 2

a
2
F (23)
Homer Reids Solutions to Jackson Problems: Chapter 3 13
The Ohmic power dissipation in a volume dV is
dP = E
2
dV (24)
To see this, suppose we have a rectangular volume element with sides dx, dy,
and dz. Consider rst the current owing in the x direction. The current density
there is E
x
and the cross-sectional area is dydz, so I = E
x
dydz. Also, the
voltage drop in the direction of current ow is V = E
x
dx. Hence the power
dissipation due to current in the x direction is IV = E
2
x
dV . Adding in the
contributions from the other two directions gives (24).
For the power dissipated outside the sphere we use the expression for the
electric eld we found earlier:
P
out
=

_

a
_

0
_
2
0
E
2
(r, , )r
2
sin dd dr
= 2

_

+ 2

_
2
F
2
a
6
_

a
_

0
1
r
4
(4 cos
2
+ sin
2
) sin d dr
=
8
3

_

+ 2

_
2
F
2
a
3
Dividing by (23), we nd the eective external voltage V
e
:
V
e
= P
out
/I =
4
3
aF

+ 2

and the eective external resistance:

R
e
= P
out
/I
2
=
2
3a

.
(c) The power dissipated inside the sphere is
P
in
=
_
(E +F

k)
2
dV =
4
2

( + 2

)
2
F
2
_
dV
=
16
2

3( + 2

)
2
a
3
F
2
Since were in steady state, the current owing out through the upper hemi-
sphere of the sphere must be replenished by an equal current owing in through
the lower half of the sphere, so to nd the internal voltage and resistance we
can just divide by (23):
V
i
= P
in
/I =
8
3

+ 2

aF
R
i
= P
in
/I
2
=
4
3a
.
Homer Reids Solutions to Jackson Problems: Chapter 3 14
(c)
(R
e
+R
i
)I =
2
3a
_
1

+
2

+ 2

a
2
F =
4
3
aF
(V
i
+V
e
) =
4aF
3( + 2

)
+ 2

=
4
3
aF
Problem 3.17
The Dirichlet Green function for the unbounded space between the planes at 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

) =
1
L

n=1

m=
e
im(

)
sin
_
nz
L
_
sin
_
nz

L
_
I
m
_
n
<
L
_
K
m
_
n
>
L
_
.
(b) Show that an alternative form of the Green function is
G(x, x

) =
1
2

m=
_

0
dk e
im(

)
J
m
(k)J
m
(k

)
sinh(kz
<
) sinh[k(L z
>
)]
sinh(kL)
.
In cylindrical coordinates, the solutions of the Laplace equation look like
linear combinations of terms of the form
T
mk
(, , z) = e
im
Z(kz)R
m
(k). (25)
There are two possibilities for the combination Z(kz)R
m
(k), both of which
solve the Laplace equation:
Z(kz)R
m
(k) = (Ae
kz
+Be
kz
)[CJ
m
(k) +DN
m
(k)] (26)
or
Z(kz)R
m
(k) = (Ae
ikz
+Be
ikz
)[CI
m
(k) +DK
m
(k)]. (27)
The Greens function G(x; x

) must be a solution of the Laplace equation,

and must thus take one of the above forms, at all points x

= x. At x

= x,
G must be continuous, but have a nite discontinuity in its rst derivative.
Homer Reids Solutions to Jackson Problems: Chapter 3 15
Furthermore, G must vanish on the boundary surfaces. These conditions may
be met by dividing space into two regions, one on either side of the source point
x, and taking G to be dierent linear combinations of terms T (as in (25)) in
the two regions. The question is, in which dimension (i.e., , z, or ) do we
dene the two sides of the source point?
(a) The rst option is to imagine a cylindrical boundary at

= , i.e. at the
radius of the source point, and take the inside and outside of the cylinder (i.e.,

< and

> ) as the two distinct regions of space. Then, within each

region, the entire range of z must be handled by one function, which means this
one function must vanish at z = 0 and z = L. This cannot happen with terms of
the form (26), so we are forced to take Z and R as in (27), with B = A and k
restricted to the discrete values k
n
= n/L. Next considering the singularities
of the functions in (27), we see that, to keep G nite everywhere, for the
inner region (

< ) we can only keep the I

m
(k) term, while for the outer
region we can only keep the K
m
(k) term. Then G(x; x

) will consist of linear

combinations of terms T as in (25) subject to the restrictions discussed above:
G(x; x

) =
_

mn
A
mn
(x)e
im

sin(k
n
z

)I
m
(k
n

),

<

mn
B
mn
(x)e
im

sin(k
n
z

)K
m
(k
n

),

> .
Clearly, to establish continuity at

= , we need to take A
mk
(x) =
mk
(z, )K
m
(k)
and B
mk
(x) =
mk
(z, )I
m
(k), where
mk
is any function of z and . Then
we can write G as
G(x; x

) =

mk

mk
(z, )e
im

sin(kz

)I
m
(k
<
)K
m
(k
>
).
The obvious choice of
mk
needed to make this a delta function in z and is

mk
= (4/L)e
im
sin(kz). Then we have
G(x; x

) =
4
L

mk
e
im(

)
sin(kz) sin(kz

)I
m
(k
<
)K
m
(k
>
).
What I dont quite understand is that this expression already has the correct
delta function behavior in , even though I never explicitly required this. To
obtain this expression I rst demanded that it satisfy the Laplace equation for
all points x

= x, that it satisfy the boundary conditions of the geometry, and

that it have the right delta function behavior in z

and

. But I never demanded

that it have the correct delta function behavior in

, and yet it does. I guess

the combination of the requirements that I did impose on this thing is already
enough to ensure that it meets the nal requirement.
(b) The second option is to imagine a plane boundary at z

= z, and take the

two distinct regions to be the regions above and below the plane. In other words,
the rst region is that for which 0 z

z, and the second region that for which

z z

L. In this case, within each region the entire range of

(from 0 to )
must be handled by one function. This requirement excludes terms of the form
Homer Reids Solutions to Jackson Problems: Chapter 3 16
(27), because K
m
is singular at the origin, while I
m
is singular at innity, and
there is no linear combination of these functions that will be nite over the whole
range of

. Hence we must use terms of the form (26). To ensure niteness at

the origin we must exlude the N
m
term, so D = 0. To ensure vanishing at z

= 0
we must take A = B, so the z

function in the region 0 z

z is proportional
to sinh(kz

). To ensure vanishing at z

= L we must take A = Be
2kL
, so
the z

function in the region z z

L is proportional to sinh[k(z

L)].
With these restrictions, the dierential equation and the boundary conditions
are satised for all terms of the form (25) with no limitation on k. Hence the
Greens function will be an integral, not a sum, over these terms:
G(x

; x) =
_

m=0
_

0
A
m
(k, , , z)e
im

sinh(kz

)J
m
(k

) dk, 0 z

m=0
_

0
B
m
(k, , , z)e
im

sinh[k(z

L)]J
m
(k

) dk, z z

L
Problem 3.18
The conguration of Problem 3.12 is modied by placing a conducting plane held
at zero potential parallel to and a distance L away from the plane with the disc
insert in it. For deniteness put the grounded plane at z = 0 and the other plane
with the center of the disc on the z axis at z = L.
(a) Show that the potential between the planes can be written in cylindrical coor-
dinates (z, , ) as
(z, ) = V
_

0
dJ
1
()J
0
(/a)
sinh(z/a)
sinh(L/a)
.
(b) Show that in the limit a with z, , L xed the solution of part a reduces
to the expected result. Viewing your result as the lowest order answer in an
expansion in powers of a
1
, consider the question of corrections to the lowest
order expression if a is large compared to and L, but not innite. Are there
diculties? Can you obtain an explicit estimate of the corrections?
(c) Consider the limit of L with (L z), a and xed and show that the
results of Problem 3.12 are recovered. What about corrections for L a, but
not L ?
(a) The general solution of the Laplace equation in cylindrical coordinates with
angular symmetry that vanishes at z = 0 is
(, z) =
_

0
A(k)J
0
(k) sinh(kz) dk. (28)
Homer Reids Solutions to Jackson Problems: Chapter 3 17
Multiplying both sides by J
0
(k

) and integrating at z = L yields

_

0
J
0
(k

)(, L) d =
_

0
A(k) sinh(kL)
__

0
J
0
(k

)J
0
(k) d
_
dk
=
_

0
A(k) sinh(kL)
_
1
k
(k k

)
_
dk
=
1
k

A(k

) sinh(k

L)
so
A(k) =
k
sinh(kL)
_

0
J
0
(k)(, L) d
=
V k
sinh(kL)
_
a
0
J
0
(k) d
=
V
k sinh(kL)
_
ka
0
uJ
0
(u) du. (29)
I worked out this integral earlier, in Problem 3.12:
_
x
0
uJ
0
(u) du = xJ
1
(x).
Then (29) becomes
A(k) =
V
k sinh(kL)
(ka)J
1
(ka)
and (28) is
(, z) = V
_

0
aJ
1
(ka)J
0
(k)
sinh(kz)
sinh(kL)
dk
= V
_

0
J
1
()J
0
(/a)
sinh(z/a)
sinh(L/a)
d. (30)
(b) For x 1,
J
0
(x) 1
1
4
x
2
+
and for x 1 and y 1,
sinh(x)
sinh(y)
=
x +
1
6
x
3
+
y +
1
6
y
3
+
=
x
y
_
1 +
1
6
(x
2
y
2
)
_
+O(x
4
)
With these approximations we may expand the terms containing a in (30):
J
0
(/a)
sinh(z/a)
sinh(L/a)

_
1
1
4
_

a
_
2
_
_
z
L
_
_
1 +
1
6
_

a
_
2
(x
2
y
2
)
_
(31)
=
z
L
_
1
_

a
_
2
_
1
6
(L
2
z
2
) +
1
4

2
_
+
_
(32)
Homer Reids Solutions to Jackson Problems: Chapter 3 18
Then the potential expansion (30) becomes
(, z) =
V z
L
__

0
J
1
() d
1
a
2
_
1
6
(L
2
z
2
) +
1
4

2
_ _

0

2
J
1
() d +
_
The rst integral evaluates to 1, so for a innite the potential becomes simply
(z) = V z/L. This is just what we expect to get for the potential between two
innite sheets, one grounded and the other at potential V.
The second integral, unfortunately, has a bit of an innity problem. Its not
hard to see where the problem comes: I derived the expansion above based on
the premise that /a is small, but the integral goes over all up to , so for
any nite a the expansions eventually become invalid in the integral.
Im still trying to work out a better procedure for estimating corrections for
nite a.
(c) In this part were interested in taking L and looking at the potential
a xed distance away from the plane with the circular insert. Calling the xed
distance z

, the z coordinate of the point were interested in is Lz

. We have
sinh k(L z

)
sinh kL
=
sinh(kL) cosh(kz

) + cosh(kL) sinh(kz

)
sinh kL
= cosh(kz

) coth(kL) sinh(kz

) (33)
Now, coth(kL) diers signicantly from 1 only for kLa 1, in which region
kz

z/L 1, so cosh(kz

) 1 and sinh(kz

) 0. By the time k gets big

enough that kz

is starting to get signicant, coth(kL) has long since started to

look like 1, so the two terms in (33) add directly. The result is that, for all k,
(33) can be approximated as exp(kz

). Then (30) becomes

(, z) = aV
_

0
J
1
(ka)J
0
(k)e
kz

dk
as we found in Problem 3.12.