Physics 505 Electricity and Magnetism Fall 2003

Prof. G. Raithel

Problem Set 4

Problem 2.13 5 Points

a): First, note that due to the the superposition principle the problem is equivalent to the sum of a constant
potential V1 +V
2
2
and a problem of two half-cylinders on opposite potentials V1 V
2
2
and V2 V
2
1
. Thus, we only
V1 V2 V2 +V1
need to consider the case of two half-cylinders on opposite potentials V = 2 and V = 2 . Assuming
that = 0 corresponds to the middle of the half-cylinder on potential V , dropping diverging terms, and
considering the symmetry of the problem, for our case Eq. 2.71 reads

X
(, ) = an n cos(n) (1)
n=1

To obtain the coefficients an , we write the potential on the surface = b, multiply with cos(n0 ) and integrate
over :

Z 2
X Z 2
X 0
V () cos(n0 )d = an bn cos(n) cos(n0 )d = an bn n,n0 = an0 bn (2)
0 n=1 0 n=1

Thus,

Z Z (3)/2 !
/2
V 4V n 4V 0 , n even
an = cos(n)d cos(n)d = n sin( ) = n (n+3)/2
bn n /2 /2 b n 2 b n (1) , n odd

0 , n even
an = (3)
i bn4Vn in , n odd

and:

!
4V X 1 n n

X
4V 1 n
(, ) = i i cos(n) = i i Im Z with Z = ir exp(i) and r =
n a n a
n odd n odd
(4)
P
1 n 1 1+Z 1+Z
Following the elaborations on p.74f of the textbook, it is n odd n Z = 2 ln 1Z , and Im ln 1Z equals
2
the phase of 1+Z
1Z .
Since we find 1Z 1+Z
= 1r 2
+i2r cos()
with 0 r 1, the phase has a range [/2, /2]
1+r +2r sin()

2r cos() 2b cos()
and is equal to tan1 1r 2 = tan1 b2 2 . Thus,

4V i 2b cos() 2V 2b cos()
(, ) = i tan1 = tan1 . (5)
2 b 2 2 b 2 2

V1 V2
Using V = 2 and the superposition explained at the beginning, it is

V1 + V2 V1 V2 2b cos()
(, ) = + tan1 , q.e.d. (6)
2 b2 2

2b cos()
It is noted that the result has the correct limit for b: Writing = b with > 0, it is tan1 b2 2 =

tan1 b cos()
= 2 Sign(cos ), and

V1 +V2 V1 V2

2 + 2 = V1 for 2 < < 2
(, ) = V1 +V2 V1 V2 3 and b . (7)
2 2 = V2 for 2 << 2

b): The charge density is = 0 E = 0

n = +0 |=b . Thus,

V1 V2 1 (b2 2 )2b cos + 42 b cos

() = 0 2 |=b
2b cos (b2 2 )2
1+ b2 2

V1 V2 2b cos (b2 + 2 )
= 0 |=b
(b2 2 )2 + 42 b2 cos2
V1 V2 1
= 0 (8)
b cos
Problem 2.23 5 Points

a): To simplify the calculation, we choose the origin such that the upper and lower walls are at z = a/2,
and the other four are at x = 0 or a and y = 0 or a. We transform into the frame of the problem statement
after the calculation.

For the given boundary conditions, the potential is of the form

X nx my p 2
(x, y, z) = Anm sin( ) sin( ) cosh(nm z) where nm = n + m2 (9)
n,m=1
a a a

Thus, on the upper and lower surface it is

Z a Z a
n0 x m0 y
V ) sin(
sin( ) dxdy =
0 0 a a
X Z aZ a
a nx my n0 x m0 y
= Anm cosh(nm ) sin( ) sin( ) sin( ) sin( )dxdy (10)
n,m=1
2 0 0 a a a a

and

a a
a2 n0 x m0 y a a2
V 2 0 0 cos cos = An0 m0 cosh(n0 m0 )
nm a 0 a 0 2 4
( 16V
2 nm cosh( n2 +m2 )
, n and m both odd
Anm = 2 (11)
0 , otherwise

and, after transformation into the frame specified in the problem,

X 16V nx my p a
(x, y, z) = sin( ) sin( ) cosh n2 + m2 (z ) (12)
2 nm cosh 2 n2 + m2 a a a 2
n,m odd

b): At the center of the cube

X
a a a 16V n m
( , , ) = sin( ) sin( )
2 2 2 2 nm cosh n2 + m2 2 2
n,m odd 2

X 16V (1)i+j
= p (13)
2
i,j=0 (2i + 1)(2j + 1) cosh 2 (2i + 1)2 + (2j + 1)2
To achieve an accuracy of three significant digits, it is:
i+j
(1)

(i,j)
Multiplicity
2 (2i+1)(2j+1) cosh 2 (2i+1)2 +(2j+1)2
(0,0) +0.21438 1 required
(0,1) -0.00464 2 required
(1,1) 0.00028 1 required
(2,0) 0.00013 2 required
(2,1) -0.000007 2 not required
(3,0) -0.000004 2 not required
... not required

and ( a2 , a2 , a2 ) = 16V
2 0.20564 = 0.333371V . According to Problem 2.28, the exact result must be
a a a 1
( 2 , 2 , 2 ) = 3 V . We see that only four terms of the expansion approach the exact result with a rela-
tive accuracy of order 104 .

c): On the upper face, it is = 0 E = 0

n = +0 z |z=a/2 , i.e.

a 160 V

X 1 nx my p p
(x, y, z = ) = sin( ) sin( ) tanh n2 + m 2 n2 + m 2 (14)
2 a nm a a 2
n,m odd
Problem 2.24 5 Points
n o
The functions Um () = Am sin m with integer m and arbitrary non-zero constants Am form a com-
plete orthogonal set on the interval 0 with Dirichlet BC. Note the analogy of these functions with
the complete set of eigenfunctions of a quantum particle in an infinitely deep square well.
P
Consider the expansion of a function f () satisfying Dirichlet BC, f () = m=1 am Um () with coefficients
am , multiply both sides with Um0 () and integrate over :

Z
X Z
m0 m m0
f ()Am0 sin d = am Am Am0 f () sin sin d = am0 A2m0
0 m=1 0 2
Z
2 0 m0
am = f ( ) sin d0 (15)
Am 0

where, for later convenience, we have switched the primes in the second line. Inserting this expression into
the expansion of f (),

( Z )
X 2
m 0
m
0 0
f () = f ( ) sin d Am sin
m=1 0 Am
(
Z X )
2 m0 m
= sin sin f (0 ) d0
0 m=1

Thus,

X
2 m0 m
sin sin = ( 0 ) , q.e.d. (16)
m=1

Problem 3.2 5 Points

In Problems 3.2 and 3.3, use of Eq. 3.70 of the textbook is recommended.
Q
It is (x0 ) = 4R2 (r
0
R). Using Eq. 3.70 of the textbook, with r< = r and r> = r0 for the interior region,
it is

Z
1 1
(x) = (x0 ) d3 x0
40 |x x0 |
Z
1 X 4 Q rl
(x) = 2

(r0 R) 0l+1 Ylm (, )Ylm (0 , 0 )r02 dr0 d cos 0 d0 (17)
40 2l + 1 4R r
l,m

q
2l+1
Sine the charge density does not depend on , only terms with m = 0 occur. Using Yl0 (, ) = 4 Pl (cos )
R 1
and Eq. 3.28, which integrates to Pl (x)dx = 2l+1 (Pl+1 (x) Pl1 (x)),

Z cos
1 X Q rl
(x) = Pl (cos ) Pl (cos 0 )d cos 0
40 2 Rl+1 cos =1
l=0

Q X rl 1 cos
(x) = l+1
Pl (cos ) [Pl+1 (x) Pl1 (x)]x=1 (18)
80 R 2l + 1
l=0

Since Pl+1 (1) = (1)l+1 = (1)l1 = Pl1 (1),

Q X rl 1
interior (x) = l+1
Pl (cos ) [Pl+1 (cos ) Pl1 (cos )] , q.e.d. (19)
80 R 2l + 1
l=0

Explicit integration of the dubious case l = 0 shows that our result is correct if we define P1 (cos ) = 1.
In the exterior region, choose r< = r0 and r> = r to find

Q X Rl 1
exterior (x) = l+1
Pl (cos ) [Pl+1 (cos ) Pl1 (cos )] . (20)
80 r 2l + 1
l=0

b): The interior radial field Er = r is

Q X rl1 l
Er = l+1
Pl (cos ) [Pl+1 (cos ) Pl1 (cos )] . (21)
80 R 2l + 1
l=1

By symmetry, the field near the origin must point in z-direction. Therefore, the field at the origin is

E(0) = z lim Er (r, = 0)

r0

Q X rl1 l
= z [Pl+1 (cos ) Pl1 (cos )] |r0
80 Rl+1 2l + 1
l=1
 Q 1 1
= z [P2 (cos ) P0 (cos )]
80 R2 3

Q 1 1 1
= z (3 cos2 1) 1 (22)
80 R2 3 2
Q
= z sin2 (23)
160 R2

c): Limit of a small cap. For 0 it is cos = 1 12 2 and, in the case l 1,

1 1 1 1
Pl+1 (1 2 ) Pl1 (1 2 ) Pl+1 (1) 2 Pl+1
0
(1) Pl1 (1) + 2 Pl1
0
(1)
2 2 2 2
1 2 0 0 1
= (Pl+1 (1) Pl1 (1)) = 2 (2l + 1)Pl (1)
2 2
2l + 1 2
= , (24)
2

0 0
where in the last step we have used Eq. 3.28 of the textbook, (2l + 1)Pl = Pl+1 Pl1 .

In the case l = 0 it is Pl+1 (1 12 2 ) + 1 = 2 12 2 = 2 2l+1 2

2 . Thus, using r> = max(R, r) and
r> = min(R, r), the potential in the interior and the exterior region is

Q X r< l
1 2l + 1 2 Q 1
(x) = l+1
Pl (cos ) + 2
80 r> 2l + 1 2 80 r>
l
Q Q 2 X r< l
= P (cos )
l+1 l
40 r> 40 4 r>
l
Q Q 2 1
= . (25)
40 r> 40 4 |x zR|

This is the potential of a homogeneously charged sphere (first term) plus the potential of a charge q located
2
at the north pole of the sphere (second term). The charge q equals Q times the ratio 4 between the
solid angle of the cap and the solid angle of a full sphere. This behavior in the limit of a small cap is to be
expected.
2
Limit of a large cap. In this case, = with 0. Then, cos 2 1, and for l 1 it is

1 1 1 1
Pl+1 ( 2 1) Pl1 ( 2 1) = (1)l+1 Pl+1 (1 2 ) Pl1 (1 2 )
2 2 2 2
2

(1)l+1 0
Pl+1 0
(1) Pl1 (1)
2
2l + 1 2l + 1 2
= (1)l Pl (1)2 = (1)l , (26)
2 2

where Eq. 3.28 of the textbook has been used. In the special case l = 0, it also is Pl+1 ( 12 2 1)Pl1 ( 12 2 1) =
P1 ( 21 2 1) + 1 = 12 2 = (1)l 2l+1 2
2 . Thus, the potential in the interior and the exterior region is

Q X 1 l
r< 2l + 1 2
(x) = l+1
Pl (cos ) (1)l
80 2l + 1 r> 2
l=0

Q 2 X r< l
= Pl ( cos )
40 4 rl+1
l=0 >
Q 2 1
= . (27)
40 4 |x + zR|

This is the potential of a charge q located at the south pole of the sphere. The charge q equals Q times the
2
ratio 4 between the solid angle of the charged area and the solid angle of a full sphere. This behavior in
the limit of a large cap is to be expected.

2
Q
In both cases of small and large cap, the electric field at the origin is E(0) = z 16 0R
2 and E(0) =
2
Q
z 16 0R
2 , respectively. This result is consistent with the interpretations of the potentials.
Problem 3.3 5 Points
RR 2n+1
Hint: A closed expression exists for r dr (you can find it with Mathematica, for instance).
0 R2 r 2

a): Exterior potential (r > R). The surface charge density is (r) = R2r2 for r < R and zero otherwise.
The volume charge density must be of the form (x) = f (r)(cos ). The function f (r) is determined by
considering the charge in a shell of radius r and thickness dr:

Z 2 Z 1
dq = (r)2rdr = d d cos r2 drf (r)(cos ) = 2r2 f (r)dr (28)
=0 cos =1

(r)
and thus (r) = r (cos ). To find , we calculate the potential at the origin and equate the result to V :

Z
1 R
h 1 r iR
V = (0) = 2rdr = sin = (29)
40 0 r R2 r 2 20 R 0 40

40 V 40 V
Thus, = , and (r) =
r R2 r 2
(cos ) . Using Eq. 3.70 of the textbook for the case r > R we have
0
r> = r and r< = r , and we find

Z Z Z
1 40 V X 4 1 1
(x) = Ylm (, ) l+1 r0l Ylm

(0 , 0 ) (cos 0 )r02 d cos 0 d0 dr0
40 2l + 1 r r 0 R 2 r 02
l,m
Z Z
V X 2 1
= 2
P l (cos ) l+1
r0l+1 Pl (cos 0 ) (cos 0 )d cos 0 dr0
r R 2 r 02
l
X Z R
2V 1 r0l+1
= l+1
P l (cos )P l (0) dr0
r 0 R 2 r 02
l

X Z
2V 1 (1)n (2n 1)!! R r0l+1
= P 2n (cos ) dr0 .
n=0 r2n+1 2n n! 0 R2 r02

RR l+1 n
Using integral tables or software, it is found that r n!2
dr = R2n+1 (2n+1)!! , and thus
0 R2 r 2

2n+1
2V X (1)n R
r>R (x) = P2n (cos ) , q.e.d. (30)
n=0 2n + 1 r

b): Interior potential (r < R). On the surface r = R the expansions for r > R and r < R must agree, i.e. the
respective coefficient functions Bl rl1 and Al rl of the Pl (cos ) must be equal for r = R and for all l. Thus,
the interior coefficients Al = Bl R2l1 . Here, l = 2n and A2n = B2n R4n1 = R2n+1 R4n1 = R2n .
Thus,

2V X (1)n r 2n

r<R (x) = P2n (cos ) . (31)
n=0 2n + 1 R

With r< = min(r, R) and r> = max r, R the potential in all space can be written as

2V R X (1)n r< 2n
everywhere (x) = 2n+1 P2n (cos ) . (32)
n=0 2n + 1 r>

Q
c): The capacitance is C = V . From part a), we know that V = 40 . The total charge Q on the disk is
obtained as

Z R h p iR

Q= 2rdr = 2 R2 r2 = 2R . (33)
0 R2 r 2 0

Q
Thus, C = V = 2R 4
, and C = 8R0 .
0