Physics 131 Problem Set 2

Remulla, Katherine Isabel T.

March 12, 2017

1 Griffiths Problem 3.3 dV

s = (12)
ds
In its most general sense, Laplaces equation is where is some constant. Again, solving the differ-
ential equation yields
2 V = 0 (1) Z

V = ds (13)
In spherical coordinates, (1) may be expanded to s

1
    V = ln s + d (14)
2 V 1 V
r + 2 sin +
r2 r r r sin where d is some constant.
1 2V
=0 (2)
r2 sin2 2 2 Griffiths Problem 3.5
The condition that V is only dependent on r implies
Consider two fields satisfying the problem, both obey-
that
V V ing Gauss Law, both in the space in between the
= =0 (3) conductors,

E~1 = (15)
Therefore, (2) may be reduced to 0

1 d
  E~2 = (16)
2 dV 0
r =0 (4)
r2 dr dr and on the boundary surfaces.
I
dV Qi
r2 = (5) E~1 d~a = (17)
dr ith conducting surface 0
Where is some constant. Solving the differential I
Qi
equation yields E~2 d~a = (18)
dV ith conducting surface 0
= 2 (6) Let us define their difference:
dr r
Z
E~3 E~1 E~2 (19)
V = dr (7)
r2
By extension,
Finally, the general solution in spherical coordinates E~3 = 0 (20)
is I
E~3 d~a = 0 (21)
V = +c (8)
r
Let us now consider when V is specified on each
Where c is a constant. In spherical coordinates, how- boundary surface. Using the following product rule
ever, (1) expands to identity,
1 2V 2V
 
1
s
V
+ 2 + =0 (9) (F~ ) = () F~ + ( F~ ) (22)
s s s s 2 z 2
The divergence of the product of V3 and E~3 becomes
Similar to the method used for spherical coordinates,
the condition that V depends only on s implies that (V3 E~3 ) = (V3 ) E~3 + V3 ( E~3 ) (23)

V V ~
Because of (20) and the fact that V = E,
= =0 (10)
z
(V3 E~3 ) = (E3 )2 (24)
(9) is then reduced to If we integrate this,
 
1 d dV
Z Z
s =0 (11) (V3 E~3 )d = (E3 )2 d (25)
s ds ds V V

1
Due to the divergence theorem, For a collection of four point charges, as shown in
Z I Figure 1, the potential now becomes
(V3 E~3 )d = V3 E~3 d~a (26) 
V q 1
V (x, y, z) = p
Because of (21), 40 (x a)2 + (y b)2 + z 2
1
Z p
(E3 )2 d = 0 (27) (x + a)2 + (y b)2 + z 2
V 1
+p
For this to be valid, E~3 must be zero everywhere. (x + a)2 + (y + b)2 + z 2
Therefore,

1
E~2 = E~1 (28) p
(x a)2 + (y + b)2 + z 2
V
and uniqueness holds. On the other hand, if n is (34)
specified on the boundary surfaces, The force on the original point charge may be calcu-
lated via Coulombs Law. The force due to the charge
V1
= E~1 (29) located at (-a, b, 0) is
n
q2
V2 F~1 = (35)
= E~2 (30) 40 4a2
n
Substituting these equations into (19) yields The force due to the charge at (a, -b, 0) is
q2
V2 V1 F~2 = (36)
E~3 = =0 (31) 40 4b2
n n
The force due to the charge at (a, -b, 0) is
E~2 = E~1 (32)
q2 1 a + b

F~3 =
3 Griffiths Problem 3.11 40 4(a2 + b2 ) a2 + b2
(37)
q2 1
=
Due to the fact that the charge is flanked by two 160 (a2 + b2 ) 32
grounded conducting plates, the method of images
may be used to determine the potential. The image Adding (35), (36), and (37) yields
configuration is shown in Figure 1.
q2
 
1 a
F~ (x, y, z) = 3
160 a2 (a2 + b2 ) 2
 ! (38)
1 b
2 3
b (a2 + b2 ) 2

In general, the energy of a point charge distribution

is known to be
n
1X
W = qi Vi (r) (39)
2 i=1

From (39), the total energy of the system is equal to

q2
 
1 1 1
Wtot = +
40 2a2 2b2 4a2 + 4b2
(40)
q2
 
1 1 1

40 a+ b2 a b

Note that (40) represents the energy everywhere.

Only the energy in the first quadrant, where x > 0
Figure 1: Image charge configuration and y > 0, is needed.

For a collection of point charges, the potential can 1

W = Wtot (41)
be calculated by 4
n
q2
 
1 X qi 1 1 1
V = (33) W = 2+ (42)
40 i=1 r 320 a2 b a + b2
2

2
4 Griffiths 3.15 5 Griffiths 3.24
We shall find the potential everywhere using (1). Laplaces equation (1) expressed in cylindrical coor-
Note that the potential is independent of z. In this dinates is (9). Because we are deriving the general
case, the solution is solution when V is independent of z, (9) becomes
V (x, y) = X(x)Y (y) (43) 1

V

1 2V
s + 2 =0 (55)
X(x) = Aekx + Bekx (44) s s s s 2
Y (y) = C sin ky + D cos ky (45) Using the method of separation of variables, let
The boundary conditions of the given charge distri-
bution are as follows: V (s, ) = S(s)() (56)

0 y=0 r2


Multiplying (56) by S ,
0 y=a
V = (46) s d

dS

1 d2

0 x=0 s + =0 (57)
S ds ds d2

V0 (y) x=b

 
Due to the first boundary condition, s d dS
s = k2 (58)
S ds ds
Y (0) = 0 = 0 + D (47)
1 d2
Therefore D = 0. Due to the second boundary con- = k 2 (59)
d2
dition,
Y (a) = 0 = C sin ka (48) For (58), let us propose the following solution:
n
Therefore k = a . Due to the third boundary condi-
S(s) = sn (60)
tion,
X(0) = 0 = A + B (49) Substituting (60) into (58) yields
Therefore B = A. With these constants, the poten-  
tial becomes d d
s s sn = k 2 sn (61)
n n
n ds ds
V (x, y) = Ae a x Ae a x C sin y
a  
n n n d
= A e a x e a x C sin snsn1 = k 2 sn


y (50) s (62)
a ds
n n
= 2AC sinh x sin y d
 
a a s nsn = k 2 sn (63)
Let 2AC = cn . (50) now becomes ds

n2 sn = k 2 sn (64)
X n n
Vn (x, y) = cn sinh x sin y (51) Therefore, n = k, and (58)becomes
n=1
a a
S(s) = Ask + Bsk (65)
To complete (51), we must find an expression for cn .
To do this, we shall employ the fourth boundary con- The general solution of (59) is
dition.
X n n () = C sin k + D cos k (66)
V0 (y) = cn sinh b sin y (52)
n
a a
Which works because sine and cosine are periodic
We may now use Fouriers Trick and exploit orthog- functions. However, (66) and (65) do not work when
onality. k = 0. For (65),
Z a
m dS
V0 (y) sin ydy = (67)
a ds s
0
Z aX S = ln s + c1 (68)
n n m (53)
= cn sinh b sin y sin ydy
0 n=1 a a a where and c1 are constants. For (66),
a m
= cm sinh b d2
2 a =0 (69)
d2
Therefore, the potential everywhere is (51), where cn
is defined as d
= (70)
2
Z a
n d
cn = V0 (y) sin ydy (54)
a sinh n
a b 0 a = + c2 (71)

3
where and c2 are constants. Do note, however, that Let us now consider the first boundary condition. At
(71) is not possible because it is not periodic. Consid- s = R,
ering all of these, the general solution in cylindrical
  !
coordinates is X Cn Dn
0= sin n + E0 R cos n (80)
n=1
Rn Rn
V (s, ) = A ln s + B

X For this to hold,
sn An sin n + Bn cos n


+
n=1 (72) Cn = 0
1
X Dn = 0, (n 6= 1) (81)
s n Cn sin n + Dn cos n


+
2
n= D1 = E0 R

Applying all of these, the potential becomes

6 Griffiths 3.25  2 
R
Let the metal pipe have its axis on the z-axis, while V (s, ) = E0 cos s (82)
s
the electric field Eo points in the positive x direction.
Note that because the pipe is made from metal, it is In general, the expression for the induced surface
a conductor. This implies that the potential on it is charge is known to be
constant. Let us set V = 0 at s = R. Very far from
V
the pipe, the potential becomes = 0 (83)
n
Vf ar = E0 x + c (73)
In this case, the expression (83) becomes
where c is an arbitrary constant. Recall the following
relation between rectangular and cylindrical coordi- V 
= 0 (84)
s s=R

nates:
x = s cos (74) 
R2 

= 0 Eo cos 1 2  (85)
Substituting (74) into (73) and simplifying yields s s=R

Vf ar = E0 cos + c (75) Therefore, the induced surface charge is

Again, when s = R, V = 0. () = 20 Eo cos (86)

0 = E0 R cos + c (76)
Therefore, c = 0. The boundary conditions are now
as follows:
(
0 s=R
V = (77)
E0 s cos s

Let us now apply the second boundary condition to

(72). Note the rearrangement of terms. This is pos-
sible due to the fact that sine is an even function and
cosine is an odd function.
Eo s cos = A ln s + B

X Cn

+ An sn n sin n
s (78)
n=1

Dn

+ Bn sn + cos n
sn
For the second boundary condition to hold, the fol-
lowing values must be true. Note that Csnn and D n
sn
were disregarded for now, as they disappear when s
approaches infinity.
A=0
B=0
An = 0 (79)
Bn = 0, (n 6= 1)
B1 = E0