Chapter 5: Solutions to Problems

Problem 5.1. Solution to Laplace's equation.

In general for the given function we can write:
2
∇2 V = ∂ f(x)g(y) + ∂2 f(x)g(y) = g ∂2 f + f ∂2 g = 0
∂x 2 ∂y 2 ∂x 2 ∂y 2
2 2 2 2
a. ∇2 (2fg) = ∂ 2fg + ∂ 2fg = 2g ∂ f2 + 2f∂ g2
∂x 2 ∂y 2 ∂x ∂y
and, since
2 2 2 2
g ∂ f2 + f ∂ g2 = 0 → 2 g ∂ f2 + 2f ∂ g2 = 0
∂x ∂y ∂x ∂y
V = 2fg is also a solution to Laplace's equation.
2 2 2 2
b. ∇2 (fg +x+yg) = ∂ (fg +x+yg) + ∂ (fg +x+yg) = g ∂ f2 + ∂ (yg) ≠0
2 2
∂x ∂y ∂x ∂y 2
V = fg + x + yg does not satisfy Laplace's equation.
2 2 2 2
c. ∇2 (2f + cg) = ∂ (2f +2 cg) + ∂ (2f +2 cg) = 2 ∂ f2 + c∂ g2 ≠ 0
∂x ∂y ∂x ∂y
Thus V = 2f + cg satisfies Laplace's equation. (f and g must be linear in x and y respectively)
2 2 2 2 2 2 2
d. ∇2 (f 2 g) = ∂ (f 2g) + ∂ (f 2g) = g∂ (f2 ) + f 2 ∂ g2 ≠ 0
∂x ∂y ∂x ∂y
2
V = f g does not satisfy Laplace's equation.

Problem 5.2. Solution to Laplace's equation.

To satisfy Laplace’s equation, the sum of second derivatives with respect to x, y, and z must equal zero:
∂2 V(x,y,z) + ∂2 V(x,y,z) + ∂2 V(x,y,z) =
∂x 2 ∂y 2 ∂z 2
∂2 5xy + y 3 z +5kz 2 2 2
3
∂ 5xy + y z +5kz 2
∂ 5xy + y z +5kz 2
3
∂2 5kz 2 ∂2 5kz 2 ∂2 5kz 2
+ + = + 6y z + +
∂x 2 ∂y 2 ∂z 2 ∂x 2 ∂y 2 ∂z 2
Note that k is assumed to be, in general, x, y, and z dependent.
Now we note that k cannot be a second order function of x or y. If it were, the second derivative of x or y would result in a
term dependent on z 2 . Therefore, the second derivatives with respect to x and y are zero and we get:
2 2 2 2 2
∂ V(x,y,z) + ∂ V(x,y,z) + ∂ V(x,y,z) = 6y z + ∂ 5kz
∂x 2 ∂y 2 ∂z 2 ∂z 2
For this to be zero we must have: k = − 0.2y z. Substitution back into the expression shows that this is correct.
The problem can be solved as follows as well. Starting from the same considerations we get:

∂2 ∂2
6yz +
∂z 2
( )
5kz 2 = 0 or : 6yz = −
∂z 2
(5kz 2 )
We can now integrate both sides of this relation with respect to z. Integrating once:

6yz 2 ∂
2
+ C1 = −
∂z
(
5kz 2 + C2 )
Integrating again:

115
6yz 3
+ C1 z + C3 = −5kz 2 + C2 z + C4
6
From this, we get:
k = −0.2yz

which, of course, is the result obtained above.

Problem 5.3. Potential distribution in a capacitor.

The general solution for potential between the plates is given by Laplace's equation.
a. Because there are no charges between the plates, the solution is:
∇2 V = − ρ = 0 → V(x) = ax + b
ε
From the boundary conditions:
V(x = 0) = b = 0, V(x = 0.002) = 0.002a = 5 → a= 5 = 2,500
0.002
Thus, the solution is: V(x) = 2500x V
b. The solution is independent of permittivity because there is no charge density inside the capacitor. If a charge density
existed, even if it were a constant charge density, the solution would depend on permittivity.
Check: V(x = 0) = 2500×0 = 0, V(x = 0.002) = 2500×0.002 = 5 V

Problem 5.4. Uniform space charge density in a capacitor.

The potential between the plates is found by superposition of the potential on the plates and that produced by the charge
between the plates. Starting with Poisson’s equation, integrating twice and setting the boundary conditions to the plate
potentials gives the correct solution due to both sources. The equation describing the potential between the plates is a one
dimensional Poisson equation:
d2 V(x) = − ρ0
dx 2 ε0
Integrating the equation once with respect to x:
dV(x) = − ρ0 x + c
1
dx ε0
where c1 is a constant of integration. Integrating again:
2
V(x) = − ρ0 x + c1 x + c2
2ε 0
To find c1 and c2 , we write the boundary conditions.
V(x=0) = 0 → V(0) = 0 + c1 0 + c2 → c2 = 0
2
V(d) = V0 → V0 = − ρ0 d + c1 d → c1 = V 0 + ρ0 d
2ε 0 d 2ε 0
The solution is therefore:
2
V(x) = − ρ0 x + V 0 + ρ0 d x V
2ε 0 d 2ε 0

Problem 5.5. Nonuniform space charge in a capacitor.

The potential between the plates is found by superposition of the potential on the plates and that produced by the charge
between the plates. Starting with Poisson’s equation, integrating twice and setting the boundary conditions to the plate
potentials gives the correct solution due to both sources.
The equation describing the potential between the plates is:
2
∇2 V(x) = d V(x) = − ρ = − 10−6 x(x − d)
dx 2 ε0 ε0

116
The equation is clearly one dimensional since the potential can only vary across the capacitor as can be seen from Figure
5.33. Now we integrate the equation once:
dV(x) = − 10−6 x 3 − x 2 d + a
dx ε0 3 2
where a is a constant of integration. Integrating again:
−6 4 3
V(x) = − 10 x − x d + ax + b V
ε 0 12 6
To find a and b, we write the boundary conditions.
−6
V(x=0) = 0 → V(0) = − 10 0 − 0 + a0 + b → b=0
ε0
−6 4 4 −6 3 3 −6 3
V(d) = V0 → V0 = − 10 d − d + ad → a = V 0 + 10 d − d = V 0 − 10 d
ε 0 12 6 d ε 0 12 6 d 12ε 0
The solution is therefore:
−6 4 3 −6 3
V(x) = − 10 x − x d + V 0 − 10 d x V
ε 0 12 6 d 12ε 0
12
With the given values V 0 = 100 V, d = 0.002 m, ε 0 = 8.854× 10 , the solution is:
3
V(x) = − 10−6 x 4 − x ×0.002 + 100 − 10−6 0.0023 x =
−1 2
8.854× 10 12 6 0.002 12×8.854× 10−1 2
− 9411.94x 4 + 37.64777x 3 + 50,000 − 7.52955× 10−5 x V
−5
or: V(x) = − 9411.94x 4 + 37.64777x 3 + 50,000 − 7.52955× 10 x V
Note: The smaller term in the brackets may be neglected for practical purposes.

Problem 5.6. Potential and field in coaxial cables.

The potential distribution in the cable only varies in the r direction. Therefore, a one dimensional Laplace equation in
cylindrical coordinates may be used.
a. The equation to solve is:
∇2 V(r) = 1 ∂ r ∂V(r) = 0
r ∂r ∂r
Since r ≠ 0, we can also write immediately:
∂ r ∂V(r) = 0
∂r ∂r
Now we integrate this once:
∂ r ∂V(r) = r ∂V(r) = c ∂V(r) = c1
1 →
r ∂r ∂r ∂r ∂r r
Integrating again we get:
V(r) = c1 = c1 ln r + c2 V
x
r

To evaluate the constants we set V = 64 V at r = 0.00025 m and V = 0 V at r = 0.004 m:

V(0.00025) = c1 ln 0.00025 + c2 = 64 → ln(0.00025)c1 + c2 = 64

V(0.004) = c1 ln 0.004 + c2 = 0 → ln(0.004)c 1 + c2 = 0

Solving for c1 and c2 : From the first equation:

c2 = 64 − ln(0.00025)c1
Substituting into the second equation:
ln(0.004)c1 + 64 − ln(0.00025)c1 = 0
and
117
 c1 = − 64 =− 64 = − 23.083
ln(0.004) − ln(0.00025) ln 0.004
0.00025
c 2 = 64 c 1ln 0.00025 = 64 + 23.083ln 0.00025 = 127.45
The solution is:
V(r) = − 23.083 ln (r) − 127.45 V
b. The electric field intensity is found by calculating the gradient of potential:
∂ − 23.083 ln(r) − 127.45
E = − ∇V(r) = − r = r 23.083 V
∂r r m

The maximum electric field intensity is at r = 0.00025 m and minimum at r = 0.004 m. These are:

E(r=0.00025m) = r 9.233× 104 E(r=0.004m) = r 5.77× 103 V

m
c. See plots in Figures A and B.
V E
80 electric field [V/m] x 10 4 10
8
60
potential [V]

6
40
4
20 2
r r
0 0
0 1 2 3 4 0 1 2 3 4
distance from center [mm] distance from center [mm]
Figure A Figure B

Problem 5.7. Potential due to large planes.

Because the planes are infinite the potential between the plates can only vary with the angle φ. Thus, we solve Laplace’s
equation in cylindrical coordinates. For the one dimensional Laplace’s equation in φ, we get:
2
1 ∂ V =0 ⇒ V ( φ ) = aφ + b V
r2 ∂φ 2
The boundary conditions are:
V(φ = 0) = V0 V( φ = π /2) = 0 V
Substituting these in the general solution we get:

V(φ = 0) = 0 + b = V0 , V( φ = π /2) = aπ + V0 = 0 ⇒ a = − 2V 0 , b = V0
2 π
Substituting these back into the general solution, the potential anywhere between the plates is:

V(φ ) = − 2V 0 φ + V0
π
Or, in Cartesian coordinates:
V (x,y) = − 2V 0 tan −1 y + V0 V
π x

Problem 5.8. Potential and field in spherical capacitor.

Because the potential in a spherical capacitor only varies with the radial position between the plates, it satisfies a one
dimensional Laplace equation in spherical coordinates.
a. The required equation is:

118
 ∇2 V(R) = 12 ∂ R 2 ∂V(R) = 0
R ∂R ∂R
Since R ≠ 0 anywhere in the solution domain, we can write:
∂ R 2 ∂V(R) = 0
∂R ∂R
To find the general solution we integrate this twice. Integrating once:

R 2 ∂V(R) = c1 → ∂V(R) = c1
2
∂R ∂R R
Integrating again,
V(R) = − c1 + c2 V
R
From the boundary potentials on the plates:
c1 + c2
V(R=0.005) = 50 = − → 50 = − 200c1 + c2
0.005
V(R = 0.00505) = 0 = − c1 + c2 → − 198.019802c1 + c2 = 0
0.00505
Solving for c1 and c2 :
− 200c1 + 198.019802c1 = 50 → c1 = − 50 = − 25.25
1.98019802
and:
c2 = − 198.019802×25.25 = − 5000
Thus, the solution is:
V(R) = 25.25 − 5000 V
R
Check: V(R = 0.005) = 25.25/0.005 − 5000 = 50, V(R = 0.00505) = 25.25/0.00505−5000 = 0 V
The potential is independent of permittivity since there is no charge density in the dielectric (right hand side of Poisson's
equation is zero, that is, the solution is obtained using Laplace's equation.
b. The electric field intensity is found from the potential as the gradient of potential:

E = − ∇V(r) = − R ∂ 25.25 − 5000 = R 25.25 V

∂R R R2 m

The maximum value occurs at R = 0.005 m and minimum at R = 0.00505 m. The fields are:

E(R = 0.005m) = R 25.252 = 1.01× 106 E(R=0.00505m) = R 25.25 = 9.9× 105 V

0.005 0.005052 m
c. The plots are shown in Figures A and B. Note that the curves for the electric potential and electric field intensity are not
straight lines but have a slight curvature. However, in the very short distance between 5.0 mm and 5.01 mm this is barely
noticeable.

V E
electric field [V/m] x 10 6

50 1.01
40
1.005
potential [V]

30
1
20
10 0.995

0 R R
5.04 5.06 0.995 5.02
5 5.02 5.04 5.06
radial position [mm] radial position [mm]
Figure A Figure B

119
Problem 5.9. Point charge above a conducting plane.
The charge density below the conductor is calculated from the normal component of the electric field intensity. Because the
charge above the plane is positive, the surface charge density is negative.
a. Directly below the charge. Using images (see Figure A)

ρ = Dn = ε 0 En = Q1 2 + Q2 2 = Q 2 [C/m2 ]
4π a 4π a 2π a
With the given values, the surface charge density is negative
Q = − 5×10−9 = − 1.99× 10−1 0
ρ=− [C/m2 ]
2π h22 2π ×22
b. When charge moved d = 1 m to the side, the configuration is as in Figure B. Again, we calculate the components of the
electric field intensity at the surface but it is clear that the tangential (horizontal) component cancels. The vertical
components add up and we have:

ρ = − ε 0 En = − 2Q cos α = − 2Qh = − Qh = − 5×10−9 ×2 = − 10−9 = − 1.424× 10−1 0 C

4π r2 4π r3 2π (h 2 + d 2 )3/2 2π (22 + 1 2 )
3/2
π 5 m2

5 nC 5 nC
(0,0, a) (0,0, a)

a a

E2 α
a E1
a
E1 E2
− 5 nC − 5 nC
(0,0, −a) (0,0, −a) d
Figure A Figure B

Problem 5.10. Point charges above a conducting plane.

By replacing the conductor with the images of the two point charges, we obtain the situation in Figure A. Now the potential
at a general point in space (x,y,z) is evaluated, then the electric field intensity is found through the gradient and finally, setting
z = 0, we obtain the field on the surface of the conductor.
a. The potential due to the four point charges (see Figure A) at a general point in space is:
q − 1 1 1 1
V= + + − V
4πε 0 x + y + (z− 2a)2
2 2
x + y + (z+2a)2
2 2 2 2
x + y + (z− a)
2
x + y + (z+a)2
2 2

Now the electric field intensity is found by calculating the gradient of V:

qx − 1 1 1 1
E = − ∇V = x 3/2
+ 3/2
+ 3/2
− 3/2
4πε 0 x 2 + y 2 + (z− 2a)2 x 2 + y 2 + (z+2a)2 x 2 + y 2 + (z− a)2 x 2 + y 2 + (z+a)2
+ y qy − 1 + 1 + 1 − 1
2 3/2 2 3/2 2 3/2 3/2
4πε 0 2 2
x + y + (z− 2a) 2 2
x + y + (z+2a) 2 2
x + y + (z− a) x + y + (z+a)2
2 2

+z q − z − 2a + z + 2a + z−a − z+a V
4πε 0 x 2 + y 2 + (z− 2a)2
3/2
x 2 + y 2 + (z+2a)2
3/2
x 2 + y 2 + (z− a)2
3/2
x 2 + y 2 + (z+a)2
3/2 m

b. Setting z = 0 gives:

q 4a 2a V
E=z −
4πε 0 x 2 + y 2 + 4a 2 3/2 x 2 + y 2 + a 2 3/2 m
and, as expected, the field is normal to the conducting
surface everywhere.
120
 z
−q P(x,y,z)

a
+q
a
y x
⊗
a
−q
a
+q
Figure A

Problem 5.11. Charged line above a conducting plane.

One method of solution is to assume a general point (x,y) anywhere in space and two position vectors r1 , r2 from the charge
line and its image to this point. Calculate the potential due to the two charged lines at the general point. Calculate the electric
field intensity at that point by taking the gradient of the electric potential at this point. Set y = 0 in the result. Now, the charge
density is ρs = ε 0 E at y = 0. The second method calculates the electric field intensity directly and adds the fields of the two
lines vectorially.
Method A. The potential due to a charged line at a distance r with reference to some arbitrary reference point r0 is given as
(see Eq. (5.29)):
a. V=−
ρl ln r V
2πε 0 r0
The potential at P(x,y) is

V (x,y) = −
ρl ln r1 + ρl ln r2 = − ρl ln r + ρl ln r V
1 2
2πε 0 r0 2πε 0 r0 2πε 0 2πε 0
where
r1 = x 2 + (y − h)2 r2 = x 2 + (y + h)2 m
Note that the reference position r0 has disappeared. The electric field intensity at a general point P(x,y) is now

E(x,y) = − ∇V(x,y) = − x ∂ − ρl ln(r1 ) + ρl ln(r2 ) − y ∂ − ρl ln(r1 ) + ρl ln(r2 ) V

∂x 2πε 0 2πε 0 ∂y 2πε 0 2πε 0 m

Performing the differentiation gives

E(x,y) = − x −
ρlx +
ρlx −y− ρl(y−h) +
ρl(y+h) V
2
2πε 0 x +(y− h)2 2
2πε 0 x +(y+h) 2 2
2πε 0 x +(y− h)2
2πε 0 x 2 +(y+h)2 m

Setting y = 0 gives the potential at the surface of the conductor:

E(x,0) = − x −
ρlx +
ρlx −y ρlh +
ρlh V
2
2πε 0 x +h
2 2
2πε 0 x +h
2 2
2πε 0 x +h
2
2πε 0 x 2 +h
2 m

The x component is zero and the y component is the only term:

E(x,0) = − y
ρlh V
πε 0 x 2 +h 2 m

Note that the y component is in the negative y direction as required.

Now the charge density is found from the interface conditions for a perfect conductor:

ρs = ε 0 En (x,0) = − ρlh C
π x 2 +h 2 m2

where E(x,0) is the normal component of the electric field intensity. Note that the charge density is negative as required.

121
Method B. In this particular case, the calculation of the electric field is actually easier than that of the electric potential.
Using the two image charges in Figure A, calculate the electric field intensity at a general point P(x,y). These are:

a. E(x,y) = r1 ρlr1 − r2 ρlr2 2 V

2πε 0 r1 2 2πε 0 r2 m

Taking the x and y components of this general expression (see Figure B) gives

E(x,y) = − x −
ρlx +
ρlx −y− ρl(y−h) +
ρl(y+h) V
2πε 0 x 2 +(y− h)2 2πε 0 x 2 +(y+h)2 2πε 0 x 2 +(y− h)2 2πε 0 x 2 +(y+h)2 m

and from here on the solution is exactly the same as previously.

Some economy in solution is achieved if instead of using a general point P(x,y) we use a point on the surface P(x,0) where x
is left general. Now, observing Figure B, it is obvious that the x components of the electric field intensity cancel each other
while the y components are in the same direction. Using Gauss' law we can write:

E(x,0) = − y ρl sinα + ρl sinα = − y ρlh sinα + ρlh V

2πε 0 R 1 2πε 0 R 2 2
2πε 0 R 1
2
2πε 0 R 2 m

and, substituting r1 and r2 from above (with y = 0) we get

E(x,0) = − y ρlh V
πε 0 x 2 +h 2 m

And the charge density is: ρs = ε 0 En (x,0) = − ρlh C

π x 2 +h 2 m2

b. To show that the total charge per unit length of the conductor's surface equals ρl, we integrated the charge density from x =
−∞ to x = +∞:
∞ ∞
ρlh dx = − ρlh ∞
− dx = − ρlh 1 tan −1 x = − ρlh 1 π + π = − ρl C
x=−∞ π x 2 +h 2 π x=−∞ x +h 2
2 π h h −∞ π h 2 2 m

P(x,y)
E1
y r1 y
E2
ρl ρl
r2 r1

.
h h
x P(x,0) x
E2 α α E1
h h r2

− ρl − ρl
Figure A Figure B

Problem 5.12. Sheet of charge above a conducting plane.

Replace the conductor by an image plate, with negative ρs
charge density at a distance − d below the upper surface ρs 2 ε0
of the conductor as shown in Figure A. Now calculate the ρs
electric field intensity due to two infinite plates using d 2 ε0
Gauss' law. Referring to Figure A, the electric field
intensity everywhere between the plate and conductor is: ρs
d
2 ε0
E = − y ρs V − ρs ρs
ε0 m
2 ε0 Figure A
Outside the plates: E = 0.

122
Problem 5.13. Charged line and multiple conducting planes.
We start by replacing the conductor’s surface with an image charge as shown in Figure A. Now we calculate the electric
field intensity at a general point in space by using Gauss’ law on each conductor and adding the fields of the two conductors
vectorially. Then, by setting the coordinates of the dotted line as (x = 2d,y) we obtain the required solution. In (b), the line of
charge is in front of two conducting surfaces and there are three image line charges. But the electric field intensity now has
only a normal component on the dotted line which coincides with the surface of the conductor.
a. Using the image charged line in Figure A, we calculate the electric field intensity at a general point P(x,y). These are
(From Gauss’ law, by taking a Gaussian surface of radius r1 and, separately, r2 ):

E(x,y) = r1 ρlr1 − r2 ρlr2 2 V (1)

2πε 0 r1 2
2πε 0 r2 m
where:
r1 = x 2 + (y− d)2 r2 = x 2 + (y+d)2 m (2)
Taking the x and y components of this general expression (see Figure A) gives

E(x,y) = x
ρlx −
ρlx +y
ρl(y−d) −
ρl(y+d) V (3)
2
2πε 0 x +(y− d)
2 2
2πε 0 x +(y+d)
2 2
2πε 0 x +(y− d)
2
2πε 0 x 2 +(y+d)2 m
Now, setting x = 2d:
E(x,y) = x
ρld −
ρld +y
ρl(y−d) −
ρl(y+d) V (4)
πε 0 4d 2 +(y− d)2 πε 0 4d 2 +(y+d)2 2πε 0 4d 2 +(y− d)2 2πε 0 4d 2 +(y+d)2 m

The potential may be found by integrating the electric field intensity. It is however easier to start with the original
configuration in Figure A and calculate the electric potential directly. This avoids the need to evaluate the reference
potential. We argue as follows: The potential due to a charged line at a distance r with reference to some arbitrary reference
point r0 is given as (see Eq. (5.29)):
V=−
ρl ln r V (5)
2πε 0 r0
The potential at P(x,y) is
V (x,y)= −
ρl ln r1 + ρl ln r2 = − ρl ln r + ρl ln r V (6)
1 2
2πε 0 r0 2πε 0 r0 2πε 0 2πε 0
Note that the reference position r0 has disappeared as expected. Setting x = 2d we get the potential on the dotted line:

V (2d,y) =
ρl ln r2 = ρl ln 4d 2 + (y+d)2 V (7)
2πε 0 r1 2πε 0 4d 2 + (y− d)2
Note in particular, that at y = 0 (at the surface of the conductor), V(2d,0) = 0 as required.
b. Now we place a second conducting surface at x = 2d. The configuration and the image charges are shown in Figure B.
Now we have four image line charges and the electric field intensity at P(2d,y) is due to all four line charges. The general
electric field intensity at any point in space due to four line charges as shown is (from (a)):

E(x,y) = ρlr1 − ρlr2 − ρlr3 + ρlr4 V

2πε 0 r1 2 2πε 0 r2 2 2πε 0 r3 2 2πε 0 r4 2 m

Although it is easy to see that only a horizontal electric field intensity can exist at that location, we will calculate the total
electric field intensity and see that the vertical components cancel. To do so we calculate first the distances r1 , r2 , r3 , r4 and
the vector r1 , r2 , r3 , r4 .

r1 = 4d 2 + (y− d)2 r2 = 4d 2 + (y− d)2 r 3 = 4d 2 + (y+d)2 r4 = 4d 2 + (y+d)2 m

r1 = x2d + y(y− d) r2 = − x2d + y(y− d) r3 = x2d + y(y+d) r4 = − x2d + y(y+d) m
Substituting these, we get the electric field intensity at (2d,y):

E(2d,y) = ρl x2d2 + y(y− d) − − x2d + y(y− d) − x2d + y(y+d) + − x2d + y(y+d) V

2πε 0 4d + (y− d)2 2
4d + (y− d)
2 2
4d + (y+d)
2 2
4d + (y+d)
2 m

Adding components, we get:

123
 E(2d,y) = x 4dρl 1 − 1 V
2πε 0 4d 2 + (y− d)2 4d 2 + (y+d)2 m
Thus, the electric field intensity is in the x direction, anywhere on the vertical line as expected.
The electric potential must be zero everywhere on the vertical line because this line is now the surface of a conductor,
midway between equal and opposite charges.

. E4 y
P(x,y) E1

y
r1
E1
r1
α2
. α1

r2
E2 E3 E2
ρl ρl − ρl
r2 r3
d d r4
x=2d x
x
d d
− ρl − ρl ρl
2d 2d
Figure A Figure B

Problem 5.14. Point charge between parallel conducting planes.

By calculating the electric field intensity everywhere q −q
between the plates and then setting the coordinates for (−b,a) (b,a)
6 1
one or the other plate, we should be able to show that the
electric field intensity is normal to the plates. The 3
(a,b)
potential is calculated by first finding the images of the −q (−a,b) q
original charge in the absence of the plates. d
a. The two plates are placed as shown in Figure A. The
images are found by reflecting the original charge about
(a,−b)
each of the plates. This produces images at points (1) and q 4 7
−q
(7). Then repeating the reflection of the two image (−a,−b)
charges obtained, we obtain the charges at points (2), and q
(6). Repeating the process again we obtain the charges at −q 2
5
points (5) and (3). Finally, reflecting the last two charges, (−b,−a) (b,−a)
both images are positive and located at point (4),
terminating the process. With the coordinates shown, the Figure A
potential at a general point in the plane is:

V(x,y) = q 1 − 1 +
4πε 0 2
(x−a) + (y− b )
2
(x−b ) + (y− a)2
2

1 − 1 + 1 −
2 2 2 2
(x+a) + (y+b )2
2
(x+b ) + (y− a) (x+a) + (y− b )
1 + 1 − 1 V
2 2 2 2
(x−a) + (y+b )2
2
(x+b ) + (y+a ) (x−b ) + (y+a )
The electric field intensity is found by calculating the gradient of this as follows:

E = −∇ V(x,y) = − x ∂V(x,y) − y ∂V(x,y) = − x q x− a − x− b +

∂x ∂y 4πε 0 (x−a)2 + (y− b)2 3/2 (x−b) + (y− a)2
2 3/2

x+b − x+a + x+a −

2 2 3/2 2 2 3/2 3/2
(x+b ) + (y− a) (x+a) + (y− b) (x+a) + (y+b )2
2

x+b + x− b − x− a
2 2 3/2 2 2 3/2 2 2 3/2
(x+b ) + (y+a ) (x−b) + (y+a ) (x−a) + (y+b )

124
 − y q y− b − y− a +
2 2 3/2 3/2
4πε 0 (x−a) + (y− b ) (x−b ) + (y− a)2
2

y− a y− b y+b
3/2
− 3/2
+ 3/2
−
(x+b ) + (y− a)2
2
(x+a)2 + (y− b )2 (x+a)2 + (y+b )2
y+a y+a y+b V
+ −
2
(x+b ) + (y+a ) 2 3/2
(x−b ) + (y+a ) 2 2 3/2
(x−a) + (y+b ) 2 2 3/2 m
This is long but straightforward. To show that the electric field intensity is perpendicular to the plates, we set y = 0 on the
lower plate and y = x on the upper plate. On the lower plate:

E=− x q x− a − x− b + x+b − x+a + x+a −

4πε 0 (x−a)2 + b 2 3/2 2
(x−b) + a 2 3/2 2
(x+b ) + a 2 3/2 2
(x+a) + b
2 3/2 2
(x+a) + b
2 3/2

x+b x− b x− a q −b −a
+ − − y − +
2 2 3/2 2 2 3/2 2 2 3/2 4 2 2 3/2 2 3/2
(x+b ) + a (x−b) + a (x−a) + b πε 0 (x−a) + b (x−b) + a 2
−a − −b + b − a + a − b V
(x+b )2 + a 2
3/2
(x+a)2 + b 2
3/2
(x+a)2 + b 2
3/2
(x+b )2 + a 2
3/2
(x−b )2 + a 2
3/2
(x−a)2 + b 2
3/2 m

The x component cancels since the negative terms equal the positive terms and we get:

E=− y q a − a + b − b V
2πε 0 (x−b )2 + a 2
3/2
(x+b )2 + a 2
3/2
(x+a)2 + b 2
3/2
(x−a)2 + b 2
3/2 m

This is in the negative y direction and therefore normal to the interface.

On the upper plate, we set y = x:

E(x,y = x) = − x q x−a − x− b +
4πε 0 (x − a )2 + (x − b )2 3/2 (x − b ) + (x − a )2
2 3/2

x+b − x+a + x+a −

2 2 3/2 2 2 3/2 3/2
(x + a ) + (x + b )2
2
(x + b ) + (x − a ) (x + a ) + (x − b )
x+b + x− b − x−a
3/2 3/2 3/2
(x + b )2 + (x + a )2 (x − b )2 + (x + a )2 (x − a )2 + (x + b )2
− y q x− b − x−a +
4πε 0 (x − a )2 + (x − b )2 3/2 (x − b ) + (x − a )2
2 3/2

x−a − x− b + x+b −
2 2 3/2 2 2 3/2 3/2
(x + a ) + (x + b )2
2
(x + b ) + (x − a ) (x + a ) + (x − b )
x+a + x+a − x+b V
(x + b )2 + (x + a )2
3/2
(x − b )2 + (x + a )2
3/2
(x − a )2 + (x + b )2
3/2 m

This gives the electric field intensity on the upper plate:

E(y = x) = − x q b−a + b+a − b+a − b−a

4πε 0 (x−a)2 + (x− b)2 3/2 (x−a)2 + (x+b )2
3/2
(x+a)2 + (x− b)2
3/2
(x+a)2 + (x+b )2
3/2

+y q b−a + b+a − b+a − b−a V

4πε 0 (x−a)2 + (x− b)2 3/2 2
(x−a) + (x+b ) 2 3/2 2
(x+a) + (x− b) 2 3/2 2
(x+a) + (x+b ) 2 3/2 m

Inspection of the terms shows that Ey = Ex. Therefore, the total electric field intensity is perpendicular to the surface.
b. To calculate the charge density on any of the plates, we write ρs = ε 0 E: On the lower plate:

ρs = − q a
3/2
− a
3/2
+ b
3/2
− b
3/2
C
2π (x − b )2 + a 2 (x + b )2 + a 2 (x + a )2 + b 2 (x − a )2 + b 2 m2

On the upper plate the charge density is positive:

ρs = q 2 b−a
3/2
+ b+a
3/2
− b+a
3/2
− b−a
3/2
4π (x−a) + (x− b)2
2
(x−a) + (x+b )2
2
(x+a) + (x− b)2
2
(x+a) + (x+b )2
2

125
The charge density on the lower plate is equal but negative. From Figure A:
b = dsin22.5° = 0.3827d a = dcos22.5° = 0.9239d m
Substituting these gives a result in d alone:

0.3827d−0.9239d 0.3827d+0.9239d
ρs = q 2 2 2 3/2
+
2 2 3/2
4π (x−0.9239d) + (x− 0.3827d) (x+0.9239d ) + (x− 0.3827d)
− 0.3827d−0.9239d − 0.3827d−0.9239d C
2 2 3/2 2 2 3/2
(x+0.9239d ) + (x− 0.3827d) (x+0.9239d ) + (x+0.3827d ) m2

Problem 5.15. Point charge between intersecting conducting planes.

The system of image charges is found by repeated reflection of the charge and its images. Each time a charge is reflected, its
image changes sign and the process terminates when no new images are generated. The process is not finite if new charges
are always generated or if the existing charges or their images are modified in subsequent steps.
a. The system of charges is shown in Figure A and is obtained by repeated reflections of the original charge and its images.
There are 2n − 1 image charges and one original charge where n = 180°/30° = 6. With the coordinates of the twelve points
shown, the potential at a general point in the x-y plane is:

V(x,y) = q 1 1 + ......+ 1 V
− (1)
4πε 0 (x−x 1 )2 + (y− y 1 )2 (x−x 2 )2 + (y− y 2 )2 (x−x 1 2)2 + (y− y 1 2)2
or, in more general terms:
q 12 (−1)i+1
V(x,y) = ∑ V (2)
4πε 0 i = 1 (x − xi)2 + (y − yi)2
The electric field intensity is found through the gradient as:

E = −∇ V(x,y) = − x ∂V(x,y) − y ∂V(x,y) =

∂x ∂y (3)
12 i+1 12 i+1
−x q ∑ (−1) (x−x i) − y q
∑ (−1) (y−y i) V
4πε 0 i=1 2
(x−x i) + (y− y i)2 3/2 4πε 0 i=1 (x−x i) + (y− y i)2
2 3/2 m

In these relations, the values of xi, yi are:

x i = a cos 15° + (i− 1)*30° , xi = a sin 15° + (i− 1)*30° m

For the configuration in Figure A. With these:

q 12
(−1)i+1(x−a cos 15° + (i− 1)*30° )
E=− x ∑ 3/2
−
4πε 0 i=1 (x−a cos 15° + (i− 1)*30° )2 + (y− a sin 15° + (i− 1)*30° )2 V
q 12
(−1)i+1(y−a sin 15° + (i− 1)*30° ) m
y ∑ 2 2 3/2
4πε 0 i=1 (x−a cos 15° + (i− 1)*30° ) + (y− a sin 15° + (i− 1)*30° )
b. The configuration now is as shown in Figure B. The only difference between this and the configuration in (a) is in the
values of the coordinates x i, x j. The solution in Eq. (3) applies here as well.

126
 −q −q
3 3 q
q q q 10
8 10
8
−q 5
−q −q
5 1
1 −q
q b
q A q a B
6 a
q 6 c

7
−q 11 −q
7 −q
11 −q
2
q 4 2 q 4
q
9 q 9
−q −q
Figure A Figure B
c. In this case there is no finite sequence of charges. Figure C shows the reason. After one or more reflections, image charges
fall on the conducting surfaces (or their extension). This modifies the requirements that these surfaces must be at constant
potential and therefore the method of images cannot be applied. In this particular case, charge No. 7 falls on the extension of
the lower plate. This is not possible physically since it would alter the constant potential condition on the plate. In some
cases, the sequence of image charges is not finite and the sequence does not terminate. While there is nothing wrong with an
infinite sequence of image charges, it is not possible to compute their effect (electric field or potential).

q −q
4 1
B
q

7
−q 5 −q

−q 3
6 q
2
q Figure C

Problem 5.16. Point charge between parallel conducting planes.

a. The image charges are found by reflecting first the original charge about the two conducting surfaces. This gives charges at
point (1) and (2). These are reflected each about the opposite conductor to obtain images (positive) at points (3) and (4) and
so on. In effect, each reflection generates a “cell” with the reflected image in it as shown in Figure A.
b. Using the images in (a), and denoting the location of the original charge at (x = d/2,y = 0) and the location of the potential
at (x,y), the potential may be written as (we also denote d1 = d − d2 for convenience):
q 1 1 1
V(x,y) = − − +
4πε 0 2
(x−d2 ) + y 2 2
(x−d−d1 ) + y 2
(x+d2 )2 + y 2
1 + 1 − 1 − 1 V
(x−2d−d2 )2 + y 2 2 2
(x+d+d1 ) + y 2
(x−3d−d1 ) + y 2
(x+2d+d2 )2 + y 2

127
Now we set: x = d/2, y = 0, d1 = d/4, d1 = 3d/4 and get:

V(d/2,0) = q 1 − 1 − 1 + 1 + 1 − 1 − 1
4πε 0 d/2−3d/4 d/2−d−d/4 d/2+3d/4 d/2−2d−3d/4 d/2+d+d/4 d/2−3d−d/4 d/2+2d+3d/4
= q 1 − 1 − 1 + 1 + 1 − 1 − 1
4πε 0 d/4 3d/4 11d/4 9d/4 7d/4 5d/4 13d/4
q
= 41− 1 − 1 + 1 + 1 − 1 − 1 = 2.2112q V
4πε 0 d 3 11 9 7 5 13 4πε 0 d
Note that the absolute values are due to the fact that the values are first squared and then the square root is calculated.
After collecting terms this gives:

V(d/2,0) = 2.2112q V
4πε 0 d

.P(x,y)
q −q q −q q −q
d2 d1 d1 d2 d2 d1 d1 d2 d2 d1 d1 d2
(4) (2) (1) (3) (5)

Figure A
x=0

Problem 5.17. Point charge between parallel conducting planes.

The image charges are found by reflecting first the original charge about the two conducting surfaces. This gives charges at
point (1) and (2). These are reflected each about the opposite conductor to obtain images (positive) at points (3) and (4) and
so on. In effect, each reflection generates a “cell” with the reflected image in it as shown.
a. With the sequence shown in Figure A in Problem 5.16, the potential at P(x,y) may be written as:
N N
q 1 1
V(x,y) = ∑ − ∑ −
4πε 0 i=1,5,9,13,... (x−(i−1)d/2−d2 )2 + y 2 i=2,6,10,14,... (x−id/2−d1 )2 + y 2
N N
∑ 1 + ∑ 1 V
i=3,7,11,15,... (x+(i−3)d/2+d2 )2 + y 2 i=4,8,12,16,... (x+(i−2)d/2+d1 )2 + y 2
where d1 = d − d2 and d2 is the x coordinate of the charge. For d2 =3d/4,
N N
q 1 1
V(x,y) =
4πε 0
∑ 2 2
− ∑ 2
i=1,5,9,13,...
(x−(i−1)d/2−3d/4) + y i=2,6,10,14,... (x−id/2−d/4) + y 2 V
N N
1 1
− ∑ 2
+ ∑
i=3,7,11,15,... (x+(i−3)d/2+3d/4) + y 2 i=4,8,12,16,... (x+(i−2)d/2+d/4)2 + y 2
b. A simple MATLAB script (listed below as script para(m,d1,d2,d,x,y,q) available from
http://extras.springer.com/2014/978-3-319-07806-9), gives the following results:
N=7 → V(0.05,2.0) = 0.041 V
N = 100 → V(0.05,2.0) = 0.021 V
N = 1000 → V(0.05,2.0) = 0.023 V
N = 10000 → V(0.05,2.0) = 0.023 V
The number of image charges necessary to obtain a relative error of 1% is 1577. This result is obtained by repeatedly
running the program above. To simplify the process, the MATLAB script para1(m,d1,d2,d,x,y,q) (available from
http://extras.springer.com/2014/978-3-319-07806-9)does this automatically. The relative error is calculated as follows:
er = abs(Vprevious − Vcurrent)/V current

128
 where V current is the potential calculated with N charges and V previous is the potential calculated with N − 1 charges.

%script para.m disp ('q=value of point charge, N=number of image charges to use')
% disp ('x,y are the coordinates at which the electric field intensity is
% script used in problem 5.17 calculated')
disp ('Solution to problem 5.17b') %
disp ('Refer to Figure 5.40 for configuration') d2=input('enter distance of point charge from upper plate ---> ')
disp ('d1=distance of point charge from upper plate') d1=input('enter distance of point charge from lower plate ---> ')
disp ('d2=distance of point charge from lower plate') q=input('enter value of point charge ---> ')
disp ('q=value of point charge, N=number of image charges to use') er1=input('enter error to use ---> ')
disp ('x,y are the coordinates at which the electric field intensity is c1=input('enter coordinates at which field is calculated [x y] ---> ')
calculated') d=d1+d2;
% x=c1(1);
d2=input('enter distance of point charge from upper plate ---> ') y=c1(2);
d1=input('enter distance of point charge from lower plate ---> ') x
q=input('enter value of point charge ---> ') y
m=input('enter number of image charges to use ---> ') vxy=0;
c1=input('enter coordinates at which field is calculated [x y] ---> ') vxy1=0;
d=d1+d2; aa=q/(4*pi*8.854e-12);
x=c1(1); n=0;
y=c1(2); m=0;
vxy=0; for l=1:100000
aa=q/(4*pi*8.854e-12); m=m+1;
n=0; n=0;
for k=1:100000 vxy=0.0;
n=n+1; for k=1:100000
if n>m, break, end n=n+1;
v1=aa/(sqrt((x-(n-1)*d/2-d2)^2+y^2)); if n>m, break, end
vxy=vxy+v1; v1=aa/(sqrt((x-(n-1)*d/2-d2)^2+y^2));
n=n+1; vxy=vxy+v1;
if n>m, break, end n=n+1;
v1=aa/(sqrt((x-n*d/2-d1)^2+y^2)); if n>m, break, end
vxy=vxy-v1; v1=aa/(sqrt((x-n*d/2-d1)^2+y^2));
n=n+1; vxy=vxy-v1;
if n>m, break, end n=n+1;
v1=aa/(sqrt((x+(n-3)*d/2+d2)^2+y^2)); if n>m, break, end
vxy=vxy+v1; v1=aa/(sqrt((x+(n-3)*d/2+d2)^2+y^2));
n=n+1; vxy=vxy+v1;
if n>m, break, end n=n+1;
v1=aa/(sqrt((x+(n-2)*d/2+d1)^2+y^2)); if n>m, break, end
vxy=vxy-v1; v1=aa/(sqrt((x+(n-2)*d/2+d1)^2+y^2));
end vxy=vxy-v1;
vxy end
er=abs(vxy-vxy1)/vxy;
if er<er1, break, end
%Script para1.m
vxy1=vxy;
%
end
disp ('Solution to problem 5.17b')
vxy
disp ('Refer to Figure 5.40 for configuration')
m
disp ('d1=distance of point charge from upper plate')
disp ('d2=distance of point charge from lower plate')

Problem 5.18. Point charge between parallel conducting planes.

a. Yes, it is possible to solve the problem using the method of images. In a way, this is the same as saying: can you see the
image in a mirror if the material in front of the mirror is not air?. To solve the problem we assume that the whole of space is
made of a material with permittivity ε 1 and proceed as with any other parallel plate reflection problem.
b. The first four image charges are shown in Figure A. A general point at which the potential is calculated is also shown
together with the various distances to the point charge and images. The solution then is:
5
1 qi q 1 1
V(x,y) =
4πε 1
∑
i=1 R
=
4πε 1 2
−
2
−
i (x−d2 ) + y 2 (x−d−d1 ) + y 2
1 + 1 + 1 V
2 2 2
(x+d2 ) + y 2 (x−2d−d2 ) + y 2
(x+d+d1 ) + y 2

129
 R5
y
.P (x,y) R4
R1 R2
q −q R3 q −q q x
d2 d1 d1 d2 d2 d1 d1 d2 d2 d1

Figure 5.18a

Problem 5.19. Arbitrary charge between

parallel conducting planes.
a. The charged line is reflected in the upper an lower plates. That gives the first two reflections and their signs as shown in.
Then the reflections are reflected again, each in the opposite surface as shown in Figure A.
b. This method takes into account the conditions imposed by the conducting surface and provides the correct potential or
electric field intensity everywhere between the plates. The actual calculation may, however be very difficult to obtain.
c. After finding the solution in (a) (which assumes zero potential on the two surfaces), the solution due to the external
potential (a linear distribution between the plates) is added to obtain the total solution.

50mm
200mm
ρl
(3)
45
50mm
50mm
200mm − ρl
45 (1)
50mm
50mm
200mm
ρl
45
50mm
50mm
200mm −ρl
45 (2)
50mm
50mm
200mm
ρl
45 (4)
50mm
Figure A

Problem 5.20. Point charge in infinite closed channel.

a. The system of image charges necessary is infinite because of the parallel plates involved. The first few layers of charges
necessary are shown in Figure A.
b. The nearest four image charges are at points (a+c,b−d), (a−c,b+d), (−(a−c),b−d) and (a−c,−b+d) and all four are negative
as shown in Figure B. The charge itself is at (a−c,b−d). The solution at the center of the box, which is at (a/2,b/2) is:

130
 q 1 1 1
V(a/2,b/2) = − −
4πε 0 2
(a−c−a/2) + (b− d−b/2)
2 2
(a−c−a/2) + (b+d−b/2)
2
(a+c−a/2) + (b− d−b/2)2
2

− 1 − 1 V
2 2
(−a+c−a/2) + (b− d−b/2)2
2
(a−c−a/2) + (− b+d−b/2)
However, to calculate the electric field intensity as the gradient of the potential we must first calculate the potential at a
general point (x,y):

V(x,y) = q 1 − 1 − 1
4πε 0 2
(a−c−x ) + (b− d−y )
2 2
(a−c−x ) + (b+d−y )
2 2
(a+c−x ) + (b− d−y )
2

− 1 − 1 V
2 2
(a−c−x ) + (− b+d−y ) (−a+c−x ) + (b− d−y )2
2

Note that the contribution due to the charge itself is positive, all others are negative.
The electric field intensity is calculated as the negative of the gradient of the potential:

E(x,y) = −∇V(x,y) = − x ∂V(x,y) − y ∂V(x,y) =

∂x ∂x
− x q a−c−x − a−c−x − a+c−x
2 2 3/2 2 2 3/2 3/2
(a+c−x ) + (b− d−y )2
4πε 0 2
(a−c−x ) + (b− d−y ) (a−c−x ) + (b+d−y )
− a−c−x − −a+c−x
2 3/2
2
(a−c−x ) + (− b+d−y ) (−a+c−x )2 + (b− d−y )2 3/2
−y q
b−d−y b+d−y b−d−y
− −
3/2 3/2 3/2
4πε 0 (a−c−x )2 + (b− d−y )2 (a−c−x )2 + (b+d−y )2 (a+c−x )2 + (b− d−y )2
−b+d−y b−d−y V
− −
2
(a−c−x ) + (− b+d−y )
2 3/2
(−a+c−x )2 + (b− d−y )2 3/2 m
Setting x = a/2, y = b/2:

E(a/2,b/2) = − x q a/2−c
3/2
− a/2−c
3/2
− a/2+c
3/2
4πε 0 (a/2−c)2 + (b/2− d)2 (a/2−c)2 + (b/2+d )2 (a/2+c)2 + (b/2− d)2
− a/2−c − −3a/2+c
2 3/2
2
(a/2−c) + (− 3b/2+d) (−3a/2+c)2 + (b/2− d)2 3/2
−y q b/2−d − b/2+d − b/2−d
4πε 0 ((a/2−c)2 + (b/2− d)2 3/2 2
((a/2−c) + (b/2+d ) 2 3/2
(a/2+c) + (b/2− d)2
2 3/2

− −3b/2+d − b/2−d V
(a/2−c)2 + (− 3b/2+d)2
3/2
(−3a/2+c)2 + (b/2− d)2 3/2 m

131
 y Figure A

q −q q −q q y
3b−d

b+d
−q

−q q −q q −q
b+d −q −q
b−d q
q −q q q x
b−d −q −(a−c) a−c a+c

x −q
−a−c −(a−c) a−c a+c 3a−c −(b−d)

q −q q −q
−q
−(b−d) Figure B
q −q q q
−b−d −q

Problem 5.21. Point charge in infinite closed channel.

The method employed here is to find all images in the The original point charge is considered to be the zeroth
system by reflecting the charge first by all four sides of the layer.
geometry. Then, the image charges are reflected about the y
sides (actually their extensions), flipping their signs for
each reflection.
q −q q
a. The final system of charges is infinite in all directions b+d
and is shown schematically in Figure A in Problem 5.20.
−q q
Once the coordinates for each image have been identified, it b−d −q
is a question of summing up a suitable number of charge
contributions. In a problem of this type it is best to consider x
layers of charges rather than individual point charges. A −(a−c) a−c a+c
layer consists of all point charges surrounding a previous
layer. Thus, the first layer consists of 8 charges surrounding q −q q
the original charge. The second layer consists of 16 charges −(b−d)
surrounding the first layer (see Figure A in Problem 5.20). Figure A

b. The system of charges is shown in Figure A in Problem 5.20. To simplify the evaluation of the expressions, we divide the
geometry into four quadrants and calculate the contributions to the potential due to charges in each quadrant separately. Placing
the original charge at x = h, y = g and calculating the potential at x,y, we get:

1. Due to charges in the first quadrant: Note: i,j are layers of charges. That is, i,j = 0 refer to the original charge (zeroth layer).
i,j = 1 refer to the first layer surrounding the original charge (the nearest 8 image charges in Figure A in Problem 5.20). i,j = 2
is the second layer (the next 16 image charges surrounding the first layer) and so on.
N N N N
q 1 1
V 1 (x,y) =
4πε 0
∑ ∑
j=0,2,4,6... i=0,2,4,6... 2 2
− ∑ ∑
j=0,2,4,6... i=1,3,5,7... 2 2
(x−ia −h) + (y− jb−g) (x−ia −c) + (y− jb−g)
N N N N
1 1
− ∑ ∑ + ∑ ∑ V
(x−ia −h)2 + (y− jb−d)2
j=1,3,5,7... i=0,2,4,6... j=1,3,5,7... i=1,3,5,7... (x−ia −c)2 + (y− jb−d)2
where h = a − c and g = b − d.
132
2. Due to charges in the second quadrant:
N N N N
q 1 1
V 2 (x,y) = − ∑ ∑ + ∑ ∑
4πε 0 j=0,2,4,6... i=1,3,5,7... (x+(i−1)a +h)2 +(y− jb−g)2 j=0,2,4,6... i=2,4,6,... (x+(i−1)a +c)2 +(y− jb−g)2
N N N N
+ ∑ ∑ 1 − ∑ ∑ 1 V
j=1,3,5,7... i=1,3,5,7... (x+(i−1)a +h)2 + (y− jb−d)2 j=1,3,5,7... i=2,4,6,... (x+(i−1)a +c)2 + (y− jb−d)2
3. Due to charges in the third quadrant:
N N N N
q 1 1
V 3 (x,y) = ∑ ∑ − ∑ ∑
4πε 0 j=1,3,5,7... i=1,3,5,7... (x+(i−1)a +h)2 +(y+(j− 1)b+g)2 j=1,3,5,7... i=2,4,6... (x+(i−1)a +c)2 +(y+(j− 1)b+g)2
N N N N
− ∑ ∑ 1 + ∑ ∑ 1 V
j=2,4,6... i=1,3,5,7... (x+(i−1)a +h) + (y+(j− 1)b+d)2
2 j=2,4,6... i=2,4,6... (x+(i−1)a +c) + (y+(j− 1)b+d)2
2

4. Due to charges in the fourth quadrant:

N N N N
q 1 1
V 4 (x,y) = ∑ ∑ − ∑ ∑
4πε 0 j=2,4,6... i=0,2,4,6... 2
(x−ia −h) + (y+(j− 1)b+d)
2 j=2,4,6... i=1,3,5,7... (x−ia −c) + (y+(j− 1)b+d)2
2
N N N N
− ∑ ∑ 1 + ∑ ∑ 1 V
2 2
j=1,3,5,7... i=0,2,4,6... (x−ia −h) + (y+(j− 1)b+g) j=1,3,5,7... i=1,3,5,7... (x−ia −c) + (y+(j− 1)b+g)2
2

The total potential is therefore the sum of these four potentials:

V (x,y) = V1 (x,y) + V2 (x,y) + V3 (x,y) + V4 (x,y) V
c. The charge and its eight nearest image charges (first layer of images) are shown in Figure A. Writing the potential at
(a/2,b/2) with the coordinates in Figure A, we get:

V(a/2,b/2) = q 1 − 1
4πε 0
(a/2−a+c)2 + (b/2− b+d)2 (a/2−a−c)2 + (b/2− b+d)2
− 1 − 1 − 1
2 2 2 2
(a/2−a+c) + (b/2+b− d) (a/2+a−c) + (b/2− b+d) (a/2−a+c) + (b/2− b−d)2
2

+ 1 + 1 + 1
2 2 2 2
(a/2−a−c) + (b/2− b−d) (a/2−a−c) + (b/2+b− d) (a/2+a−c) + (b/2+b− d)2
2

+ 1 V
(a/2+a−c) + (b/2− b−d)2
2

d. A computer program written as a script in MATLAB is listed in Figure B. This program calculates the potential for any
number of layers in the configuration. The program is executed by typing para2d.m. In the script, m is the number of point
charges to be used, including the original charge, a is the horizontal dimension of the box, b the vertical dimension of the box,
x,y, the coordinates at which the potential is calculated and q the original point charge. Note that if m is smaller than the
number of charges in an integer number of layers, the program adds the point charges to complete the incomplete layers. Thus,
for example, if m = 1, only the original charge is included in the calculation. If m = 3, the program adds 6 more charges to
complete the first layer and the number of charges will be 9. The results using this program are as follows:
N = 9 (in two layers) → V(2.5,2.5)=0.5689 V
N = 100 (121 charges in 5 layers) → V(2.5,2.5)=0.2280 V
N = 1,000 (1,089 charges in 16 layers) → V(2.5,2.5)=0.1274 V
N = 100,000 (100,489 charges in 158 layers) → V(2.5,2.5)=0.0841 V
Note: the number in brackets are the actual numbers used by the program and represent the closest number of charges to the
number given but in complete layers.
The number of image charges necessary to obtain a relative error of 1% is 2601 in 25 layers. The potential is 0.1107 V.
This result is obtained by repeatedly running the program above. To simplify the process, the MATLAB script
para2d1.m does this automatically. (Both para2d.m and para2d1.m are available from
http://extras.springer.com/2014/978-3-319-07806-9.) The way the relative error is calculated is as follows:
133
 er=abs(Vprevious−V current)/V current
where V current is the potential calculated with N charges and V previous is the potential calculated with N − 1 layers of charges.

% script para2d.m vxy=vxy+v1;

% end
% Used in problem 5.21d. Refer to problem 5.21 and Figure 5.43 for use end
of end
% variables end
% Calculated the potential at point (x,y) using m point charges. for jx=0:2:n1
% Note that if m is smaller than the number of charges in an integer for ix=1:2:n1
number of layers, if n>=m, break, end
% the program adds the necessary point charges to complete the n=n+1;
incomplete layers. v1=aa/(sqrt((x-ix*a-c)^2+(y-jx*b-g)^2));
% Thus, for example, if m = 1, only the original charge is included in the vxy=vxy-v1;
calculation. if jx~=0
% If m = 3, the program adds 6 more charges to complete the first layer n=n+1;
and the number of charges will be 9 v1=aa/(sqrt((x-ix*a-c)^2+(y+(jx-1)*b+d)^2));
disp ('Solution to problem 5.21') vxy=vxy-v1;
disp ('Refer to Figure 5.43 for configuration') end
disp ('m=number of point charges to be used, including the original n=n+1;
charge') v1=aa/(sqrt((x+(ix-1)*a+h)^2+(y-jx*b-g)^2));
disp ('a=horizontal dimension of the box in meters') vxy=vxy-v1;
disp ('b=vertical dimension of the box in meters') if jx~=0
disp ('c=distance of point charge from right plate') n=n+1;
disp ('d=distance of point charge from upper plate') v1=aa/(sqrt((x+(ix-1)*a+h)^2+(y+(jx-1)*b+d)^2));
disp ('q=value of the original point charge in Coulombs') vxy=vxy-v1;
disp ('x,y are the coordinates at which the electric potential is end
calculated') end
a=input('enter horizontal dimension of box ---> ') end
b=input('enter vertical dimension of box ---> ') for jx=1:2:n1
c=input('enter distance of point charge from right plate ---> ') for ix=0:2:n1
d=input('enter distance of point charge from upper plate ---> ') n=n+1;
q=input('enter value of point charge ---> ') v1=aa/(sqrt((x-ix*a-h)^2+(y-jx*b-d)^2));
m=input('enter number of point charges to use (including original vxy=vxy-v1;
charge) ---> ') n=n+1;
c1=input('enter coordinates at which field is calculated [x y] ---> ') v1=aa/(sqrt((x-ix*a-h)^2+(y+(jx-1)*b+g)^2));
x=c1(1); vxy=vxy-v1;
y=c1(2); if ix~=0
n=1; if n>=m, break, end
for i=0:10000 n=n+1;
n=n+i*8; v1=aa/(sqrt((x+(ix-1)*a+c)^2+(y-jx*b-d)^2));
if n>=m, break, end vxy=vxy-v1;
end end
n1=i if ix~=0
vxy=0; n=n+1;
aa=q/(4*pi*8.854e-12); v1=aa/(sqrt((x+(ix-1)*a+c)^2+(y+(jx-1)*b+g)^2));
n=0; vxy=vxy-v1;
g=b-d; end
h=a-c; end
for jx=0:2:n1 %if n>=m, break,end
for ix=0:2:n1 end
n=n+1; for jx=1:2:n1
v1=aa/(sqrt((x-ix*a-h)^2+(y-jx*b-g)^2)); for ix=1:2:n1
vxy=vxy+v1; n=n+1;
if jx~=0 v1=aa/(sqrt((x-ix*a-c)^2+(y-jx*b-d)^2));
n=n+1; vxy=vxy+v1;
v1=aa/(sqrt((x-ix*a-h)^2+(y+(jx-1)*b+d)^2)); n=n+1;
vxy=vxy+v1; v1=aa/(sqrt((x-ix*a-c)^2+(y+(jx-1)*b+g)^2));
end vxy=vxy+v1;
if ix~=0 n=n+1;
n=n+1; v1=aa/(sqrt((x+(ix-1)*a+h)^2+(y-jx*b-d)^2));
v1=aa/(sqrt((x+(ix-1)*a+c)^2+(y-jx*b-g)^2)); vxy=vxy+v1;
vxy=vxy+v1; n=n+1;
end v1=aa/(sqrt((x+(ix-1)*a+h)^2+(y+(jx-1)*b+g)^2));
if jx~=0 vxy=vxy+v1;
if ix~=0 end
n=n+1; end
v1=aa/(sqrt((x+(ix-1)*a+c)^2+(y+(jx-1)*b+d)^2)); vxy

% script para2d1.m %

134
% Used in problem 5.21d. Refer to problem 5.21 and Figure 5.43 for use for jx=0:2:n1
of for ix=1:2:n1
% variables. Calculates the number of charges needed to obtain a if n>=m, break, end
relative n=n+1;
%error of 1% for the potential calculated at point (x,y) v1=aa/(sqrt((x-ix*a-c)^2+(y-jx*b-g)^2));
% the result is m = number of required point charges. vxy=vxy-v1;
% if jx~=0
disp ('Solution to problem 5.21') n=n+1;
disp ('Refer to Figure 5.43 for configuration') v1=aa/(sqrt((x-ix*a-c)^2+(y+(jx-1)*b+d)^2));
disp ('a=horizontal dimension of the box in meters') vxy=vxy-v1;
disp ('b=vertical dimension of the box in meters') end
disp ('c=distance of point charge from right plate') n=n+1;
disp ('d=distance of point charge from upper plate') v1=aa/(sqrt((x+(ix-1)*a+h)^2+(y-jx*b-g)^2));
disp ('q=value of the original point charge in Coulombs') vxy=vxy-v1;
disp ('x,y are the coordinates at which the electric potential is if jx~=0
calculated') n=n+1;
a=input('enter horizontal dimension of box ---> ') v1=aa/(sqrt((x+(ix-1)*a+h)^2+(y+(jx-1)*b+d)^2));
b=input('enter vertical dimension of box ---> ') vxy=vxy-v1;
c=input('enter distance of point charge from right plate ---> ') end
d=input('enter distance of point charge from upper plate ---> ') end
q=input('enter value of point charge ---> ') end
c1=input('enter coordinates at which field is calculated [x y] ---> ') for jx=1:2:n1
x=c1(1); for ix=0:2:n1
y=c1(2); n=n+1;
m=1; v1=aa/(sqrt((x-ix*a-h)^2+(y-jx*b-d)^2));
vxy1=0; vxy=vxy-v1;
for l=1:1000000 n=n+1;
m=m+(l-1)*8; v1=aa/(sqrt((x-ix*a-h)^2+(y+(jx-1)*b+g)^2));
n=1; vxy=vxy-v1;
for i=0:10000 if ix~=0
n=n+i*8; if n>=m, break, end
if n>=m, break, end n=n+1;
end v1=aa/(sqrt((x+(ix-1)*a+c)^2+(y-jx*b-d)^2));
n1=i vxy=vxy-v1;
vxy=0; end
aa=q/(4*pi*8.854e-12); if ix~=0
n=0; n=n+1;
g=b-d; v1=aa/(sqrt((x+(ix-1)*a+c)^2+(y+(jx-1)*b+g)^2));
h=a-c; vxy=vxy-v1;
for jx=0:2:n1 end
for ix=0:2:n1 end
n=n+1; %if n>=m, break,end
v1=aa/(sqrt((x-ix*a-h)^2+(y-jx*b-g)^2)); end
vxy=vxy+v1; for jx=1:2:n1
if jx~=0 for ix=1:2:n1
n=n+1; n=n+1;
v1=aa/(sqrt((x-ix*a-h)^2+(y+(jx-1)*b+d)^2)); v1=aa/(sqrt((x-ix*a-c)^2+(y-jx*b-d)^2));
vxy=vxy+v1; vxy=vxy+v1;
end n=n+1;
if ix~=0 v1=aa/(sqrt((x-ix*a-c)^2+(y+(jx-1)*b+g)^2));
n=n+1; vxy=vxy+v1;
v1=aa/(sqrt((x+(ix-1)*a+c)^2+(y-jx*b-g)^2)); n=n+1;
vxy=vxy+v1; v1=aa/(sqrt((x+(ix-1)*a+h)^2+(y-jx*b-d)^2));
end vxy=vxy+v1;
if jx~=0 n=n+1;
if ix~=0 v1=aa/(sqrt((x+(ix-1)*a+h)^2+(y+(jx-1)*b+g)^2));
n=n+1; vxy=vxy+v1;
v1=aa/(sqrt((x+(ix-1)*a+c)^2+(y+(jx-1)*b+d)^2)); end
vxy=vxy+v1; end
end er=abs(vxy-vxy1)/vxy
end if er<0.01, break, end
end vxy1=vxy; n; end; vxy; m
end

2 2
Problem 5.22. Application: Cable in a tunnel. b=a d = 5 = 6.25 m
The charged cable suspended from the ceiling may be d 4
viewed as the image of a charged line with negative charge The result is the charged lines as shown in Figure B. For
as shown in Figure A. Taking the radius to be a and the simplicity in calculation we take the reference axis as the
distance of the charged line to be b [m] from the center of x axis in the downward direction and write using Gauss’
the tunnel, we can use the results in Eq. (5.35): law for each one of the charged lines:
135
 E (0,0) =
ρl − ρl = ρl 1 − 1
2πε 0 r2 2πε 0 r1 2πε 0 r2 r1
−9
= 10×10 1 − 1 = 16.178 V
2π ×8.854× 10−1 2 6.25 4 m

− ρl A
− ρl

d r1

ρl B ρl
b r2
O
a O
a

Figure A Figure B
The electric field intensity is downwards as expected from a positively charged line.

Problem 5.23. Two charged wires next to a grounded conducting cylinder.

By viewing each charged line separately, the solution is the superposition of two geometries. One includes the left charged line
and the cylinder (Figure A) and the second includes the right charged line and the cylinder (Figure B). Each one of these is
solved separately as the image problem of a single line outside a cylindrical conductor and then summed up to obtain the
solution. Starting with Figure A, the location of the image charge due to the left (negative charge is at a distance b from the
center of the cylinder). From Eq. (5.35):
2
b=r = r m
2r 2
where the notation in Figure A was used. The image charged line inside the cylinder is positive since the charge that created it
is negative. Turning now to Figure B the charged line is positive and therefore the image charge is negative and at the same
distance b from the center of the cylinder. The conducting cylinder and the two line charges are now replaced with the two
original charged lines and the two image lines as shown in Figure C. Figure C also shows the electric fields due to the four
charged lines at a general point in space. Because the charged lines are very long, we only consider the electric field in the
plane shown. The electric field intensity due to charged line No. 1 at (x,y) is:

E1 = ρl = ρl V
2πε 0 r1 2πε 0 (x − 2r)2 + y 2 m

The direction is as shown in Figure C. Separating this into the x and y components:

E1x = E1 cosθ1 =
ρl(x−2r) = ρl(x−2r) V
2πε 0 r12 2πε 0 (x − 2r)2 + y 2 m
Similarly:
E1y = E1 sinθ1 =
ρly = ρly V
2πε 0 r1 2πε 0 (x − 2r)2 + y 2
2 m

Repeating this for the other three lines and taking into account the polarity, we get:

E2x = E2 sinθ2 = −
ρl(x−r/2) = − ρl(x−r/2) V
2πε 0 r22 2
2πε 0 (x − r/2) + y 2 m
ρl y ρ l y V
E2y = E2 cosθ2 = − =−
2πε 0 r22 2πε 0 (x − r/2)2 + y 2 m
ρ (x+r/2) =
E3x = E3 sinθ3 = l
ρl(x+r/2) V
2πε 0 r32 2
2πε 0 (x + r/2) + y 2 m

136
 E3y = E3 cosθ3 =
ρly = ρly V
2πε 0 r32 2πε 0 (x + r/2)2 + y 2 m

E4x = E4 sinθ4 = − ρl(x+2r) =−

ρl(x+2r) V
2πε 0 r42 2πε 0 (x + 2r)2 + y 2 m
E4y = E4 cosθ4 = − ρly 2 = −
ρly V
2πε 0 r4 2
2πε 0 (x + 2r) + y 2 m
The total field is the sum of the components:
(x−2r) (x−r/2) (x+r/2) (x+2r)
E = E 1 + E 2 + E 3 + E 4 = x ρl − + −
2πε 0 (x − 2r)2 + y 2 (x − r/2)2 + y 2 (x + r/2)2 + y 2 (x + 2r)2 + y 2
+ y ρly 1 − 1 + 1 − 1 V
2πε 0 (x − 2r)2 + y 2 (x − r/2)2 + y 2 (x + r/2)2 + y 2 (x + 2r)2 + y 2

This is the field outside the cylinder only! The field inside the cylinder is zero.
y E1
y

ρl
y

r
E4 . θ1
θ3
E3
x
− ρl x − ρl ρl r1
x r4
r3 r2 E 2
b r b − ρl ρl θ1
2r 2r θ4 θ3 θ2 x
ρl r
− r
− − ρl
2 2
2r 2r
Figure A Figure B Figure C

Problem 5.24. Point charge in a conducting shell.

The point charge inside the sphere generates an image charge outside the sphere at a distance d from the center of the sphere
as shown in Figure A. The image charge and the distance d are calculated from Eq. (5.42).
a. The location and magnitude of the image charge outside the sphere are calculated from Eq. (5.42)”

q' = − Qa
2
b=a m, C
d d
These equations calculate the distance b from the center of the sphere where the image charge due to a point charge outside
the sphere at a distance d from the center is located. In other words, referring to Figure 5.24b in the text, here b = a/2 and q’
= q. With these:
a = a2 → d = 2a and q = − Qa → Q = − 2q
2 d a/2
Now the configuration in Figure A may be used to calculate the electric field intensity everywhere in space although, the
solution is only valid inside the sphere. There are two ways to solve this problem. One is to simply calculate the electric field
intensity at the center of the sphere due to two charges, one equal to q, the other to −2q as shown in Figure B. Since both
charges are on the x axis. This will give a quick result. Another is to calculate the electric field intensity at a general point in
space and then substitute the coordinates of the center of the sphere to obtain the values there. We perform both calculations.
Method A. From Figure B The electric field intensity at x = a, y = 0 due to the positive charge is in the negative x direction,
that due to the negative charge is in the positive x direction. Thus:
q
=− x q 2
2q
E1 = − x E2 = x =x q 2 V
4πε 0 (a/2)2 πε 0 a 4πε 0 (2a)2 8πε 0 a m

The total electric field intensity is the sum of the two fields:
q − x q = − x 7q V
E(a/2,0) = E 1 + E 2 = x
8πε 0 a2 πε 0 a2 8πε 0 a2 m
Method B: write the electric potential at a general point in space P(x,y,z) due to the two point charges in Figure B. The
potentials are:
137
 q 2q q 1 − 2
V1 = V2 = − → V= V
4πε 0 R 1 4πε 0 R 2 4πε 0 R 1 R 2
The distances R 1 and R 2 are:

R 1 = (x − 3a/2)2 + (y− 0)2 + (z− 0)2 = (x − 3a/2)2 + y 2 + z 2 R2 = (x − 3a)2 + y 2 + z 2 m

The potential at a general point is therefore

V (x,y,z) = q 1 − 2 V
4πε 0 (x − 3a/2)2 + y 2 + z 2 2
(x − 3a) + y 2 + z 2
The electric field intensity is calculated as the gradient of potential:

E(x,y,z) = − ∇V (x,y,z) = − x ∂V(x,y,z) − y ∂V(x,y,z) − z ∂V(x,y,z) =

∂x ∂y ∂z
q x − 3a/2 2(x − 3a)
− x −
4πε 0 (x − 3a/2)2 + y 2 + z 2 3/2 (x − 3a)2 + y 2 + z 2
3/2

− y qy 1 − 2
4πε 0 (x − 3a/2)2 + y 2 + z 2 3/2 (x − 3a)2 + y 2 + z 2
3/2

− z qz 1 − 2 V
4πε 0 (x − 3a/2)2 + y 2 + z 2 3/2 (x − 3a)2 + y 2 + z 2
3/2 m
Now by setting: x = a, y = 0, z = 0 for the location of the center of the sphere:
q
− 2(a − 3a)3/2 = − x q
a − 3a/2 − a/2 2( − 2a)
E(x,y,z) = − x −
2 3/2 3/2 3/2
4πε 0 (a − 3a/2) (a − 3a)2 4πε 0 ( − a/2)2 ( − 2a)2
=− x q − a/2 + 4a = − x q 4 − 1 = − x 7q V
4πε 0 (−a/2) 3
(−2a)
3
4πε 0 a 2
2a
2
8πε 0 a2 m

Which, not surprisingly gives the same result. The advantage in doing it this way is that the general result is obtained.
b. The electric field intensity outside the sphere is zero. There are surface charges induced due to the charge inside the sphere
but the net total charge is zero.
c. A plot of the field is shown in Figure C.
− −
P(R,θ)
E −
y −
R a −
R2 R1 −
a
a q −
+q q' θ z
x a q
a 2 −2q −
2
d 2a − −
a/2
Figure A Figure B Figure C

Problem 5.25. Point charge outside conducting sphere.

The point charge outside the sphere generates an image charge inside the sphere at a distance b from the center of the sphere
as shown in Figure A. The image charge and the distance b are calculated directly from Eq. (5.42).
a. The location and magnitude of the image charge outside the sphere are calculated from Eq. (5.42).
2 qa
b=a m, q' = − C (1)
d d
In this case, the charge q is given and d = a + r. Therefore:

q' = − q r
2
b= r m, C (2)
a+r a+r
b. To simplify calculation we will use here spherical coordinates. This decision is based on the following observation: In
Figure A or B, by placing the image charges on the z axis, the sphere becomes symmetric about the axis and the charge
138
density will not vary with the angle φ (circumferentially) although it will vary with θ. Using Figure B, we write the potential
at a general point in space (R,θ):
q' q
V(R,θ) = + V (3)
4πε 0 R 2 4πε 0 R 1
The distances R 1 between q and P(R,θ) and R 2 between q' and P(R,θ) are:

R 2 = b 2 + R 2 − 2bRcosθ and R1 = d2 + R 2 − 2dRcosθ m (4)

where:
q' = − q r
2
b= r d = a+r (5)
a+r a+r
The potential then becomes:
q' q
V(R,θ) = + V (6)
2 2
4πε 0 d + R 2 − 2dRcosθ
2
4πε 0 b + R − 2bRcosθ
Now we can find the general expression for the electric field intensity before substituting the values in Eq. (5). However,
since we need the field intensity at the surface of the sphere, and since we know that the electric field intensity at the surface
of a conductor is normal, then the electric field intensity at R = r must be radial. Therefore we calculate only the R
component of the field:
∂V(R,θ) q' R − bcosθ q R − dcosθ
ER(R,θ) = − = 1 + V
∂R 4πε 0 b 2 + R 2 − 2bRcosθ 3/2 2 2
d + R − 2dRcosθ
3/2 m

Now we set R = r and substitute the values in Eq. (5):

qr 2
r − r cosθ q r − (a + r)cosθ
ER(r,θ) = 1 − a+r a+r
3/2
+ 3/2
4πε 0 4 3 2
r + r 2 − 2r cosθ (a + r) + r 2 − 2(a + r)rcosθ
(a + r)
2 a+r
(a + r) (a/r + 1 − cosθ)q q r − (a+r)cosθ
= 1 − 3/2
+ 3/2
4πε 0 2 2
r2 + (a + r) − 2r(a + r)cosθ (a + r) + r 2 − 2(a + r)rcosθ
= q − a2 − 2ar V
4πε 0 r r + (a + r)2 − 2r(a + r)cosθ 3/2
2 m

c. The surface charge density on the surface of the sphere equals ε 0 E (where E is the electric filed intensity on the surface of
the sphere calculated in (b)) and we get:

ρs(r,θ) = ε 0 ER(r,θ) = − q a2 + 2ar C

4π r r + (a + r)2 − 2r(a + r)cosθ 3/2
2 m2

d. The electric field intensity outside the sphere is shown in Figure C. Note that there is no field inside the sphere.

P(R,θ)

R
r R2 R1
q' a q θ
q z
b b q'
d=a+r d=a+r

Figure A Figure B

139
 − −
−
−
− − q
−
−
− −

Figure C

Problem 5.26. Point charge inside hollow, charged conducting sphere.

We separate the problem as the superposition of a grounded spherical shell with a charge as in Figure A and a spherical
surface with a point charge at its center, Q, so that it produces a voltage equal to V 0 at R = a. The superposition is shown in
Figure A. The first part in Figure A (that of the point charge inside the grounded sphere) is solved using the method of
images and the point charge then added to produce the constant electric potential on the sphere. The potential on the sphere
does not affect the field inside the sphere while the point charge +q does not affect the electric field intensity outside the
sphere because the sphere is conducting.
a. To calculate the electric field intensity inside the sphere we ignore the potential on the surface because the electric field
intensity inside the sphere due to the potential on the surface is zero. Thus (see Problem 5.24) the point charge inside the
sphere generates an image charge outside the sphere at a distance d from the center of the sphere as shown in Figures A and
B in Problem 5.24. The image charge and the distance d are calculated from Eq. (5.42).
The location of the image charge outside the sphere and its magnitude are calculated from Eq. (5.42):

q' = − Qa
2
b=a m, C
d d
These equations calculate the distance b from the center of the sphere where the image charge due to a point charge outside
the sphere at a distance d from the center is located. In other words, Referring to Figure 5.24b in the text, here b = a/2 and q’
= q. With these:
a = a2 → d = 2a and q = − Qa → Q = − 2q
2 d a/2
Now the configuration in Figure B in Problem 5.24 may be used to calculate the electric field intensity everywhere in space
although, the solution is only valid inside the sphere. The simplest method is to calculate the electric field intensity at the
center of the sphere due to two charges, one equal to q, the other to −2q as shown in Figure B in Problem 5.24. Since both
charges are on the x axis. This will give a quick result. From Figure B in Problem 5.24, the electric field intensity due to the
positive charge is in the negative x direction, that due to the negative charge is in the positive x direction at x = a, y = 0. Thus:
q
=− x q 2
2q
E1 = − x E2 = x =x q 2 V
4πε 0 (a/2)
2
πε 0 a 4πε 0 (2a)
2
8πε 0 a m

The total electric field intensity is the sum of the two fields:
q − x q = − x 7q V
E(a/2,0) = E 1 + E 2 = x
8πε 0 a2 πε 0 a2 8πε 0 a2 m
b. The electric field intensity outside the sphere is only due to the potential on the sphere and this in turn is due to an
equivalent point charge at the center of the sphere. The calculation in (a) does not affect this calculation because the image
charge does not exist in reality and the induced charges on the surface of the sphere due to the internal charge cancels any
electric field intensity outside the sphere due to the internal charge. Using Figure A, the equivalent charge at the center of the
sphere that will produce a voltage V 0 on the sphere is:
V0 = Q → Q = 4 π ε 0 aV0
4πε 0 a
Now, the electric field intensity outside the sphere (R > a) is:

140
 E=R Q = R 4πε 0 aV02 = R aV20 V R≥a
4πε 0 R 2 4πε 0 R R m

Figure B shows the electric field intensity outside the shell, at the surface.

E
a
V=V0
a a a
+q +q q' Q
a a
2 2
d

Figure A Figure B

Problem 5.27. Point charge outside charged conducting sphere.

The point charge outside the sphere generates an image charge inside the sphere at a distance b from the center of the sphere
as shown in Figure A. The image charge and the distance b are calculated directly from Eq. (5.42). We will view this as a
separate problem from that of the surface charge density ρs. That is, we will solve without the surface charge density and then
add this surface charge density. The electric field inside the sphere is not altered by this addition but outside the sphere an
electric field due to this charge density is added to the field generated by the charge and its image.
a. We first assume that ρs = 0 and find the electric field intensity and charge density on the surface of the sphere (see
Problem 5.25). The location of the image charge outside the sphere and the magnitude of the image charge are calculated
from Eq. (5.42).
q' = − qa
2
b=a m, C (1)
d d
In this case, the charge q is given and d = a + r. Therefore:

q' = − qr
2
b= r (2)
a+r a+r
b. To simplify calculation we will use here spherical coordinates. This decision is based on the following observation: In
Figure A, by placing the image charges on the z axis, the sphere becomes symmetric about the axis and the charge density
will not vary with the angle φ (circumferentially) although it will vary with θ. Using Figure B, we write the potential at a
general point in space (R,θ):
q' q
V(R,θ) = + V (3)
4πε 0 R 2 4πε 0 R 1
The distances R 1 between q and P(R,θ) and R 2 between q' and P(R,θ) are:

R 2 = b 2 + R 2 − 2bRcosθ and R1 = d2 + R 2 − 2dRcosθ m (4)

where:
q' = − qr
2
b= r d=a+r (5)
a+r a+r
The potential then becomes:
q' q
V(R,θ) = + V (6)
2 2
4πε 0 b + R − 2bRcosθ 4πε 0 d + R 2 − 2dRcosθ
2

Now we can find the general expression for the electric field intensity before substituting the values in Eq. (5). However,
since we need the field intensity at the surface of the sphere, and since we know that the electric field intensity at the surface
of a conductor is normal, then the electric field intensity at R = r must be radial. Therefore we calculate only the R
component of the field:
q' R − bcosθ q R − dcosθ
ER(R,θ) = − ∂V(R,θ) = 1 + V
∂R 2 2
4πε 0 b + R − 2bRcosθ 3/2
d + R 2 − 2dRcosθ
2 3/2 m

Now we set R = r and substitute the values in Eq. (5):

141
 qr r − r2 cosθ
q r − (a + r)cosθ
ER(r,θ) = 1 − a+r a+r
3/2
+ 3/2
4πε 0 4 3 2
r + r 2 − 2r cosθ (a + r) + r 2 − 2(a + r)rcosθ
(a + r) 2 a + r
(a + r) (a/r + 1 − cosθ)q q r − (a + r)cosθ
= 1 − 3/2
+ 3/2
4πε 0 2 2 2
r + (a + r) − 2r(a + r)cosθ (a + r) + r 2 − 2(a + r)rcosθ
= q − a2 − 2ar V
4πε 0 r r + (a + r)2 − 2r(a + r)cosθ 3/2
2 m

Now we add the surface charge density ρs . Using Gauss' law, this is equivalent to a point charge equal to the total charge on
the sphere. The electric field intensity for any point R > r is:
2 2
Eρs = 4π r ρs2 = r ρs2 V
4πε 0 R ε0R m
At R = r this gives:
2
Eρs(R=r) = r ρs2 = ρs V
ε0R ε0 m

Thus, the electric field intensity anywhere on the sphere is the sum of the two electric fields and is everywhere radial:
ρs − q a2 + 2ar V
E(r,θ) = R
ε 0 4πε 0 r r2 + (a+r)2 − 2r(a+r)cosθ 3/2 m

c. The surface charge density on the surface of the sphere equals ε 0 E (where E is the electric filed intensity on the surface of
the sphere calculated in (b)) and we get:
a2 + 2ar
ρ(r,θ) = ε 0 Ε = ρs − q C
4π r r2 + (a+r)2 − 2r(a+r)cosθ 3/2 m2

d. The electric field intensity outside the sphere is shown in Figure C in Problem 5.25 as a superposition of the two fields.
Note that there is no field inside the sphere. Although Figure C in Problem 5.25 shows the field without the surface charge
density given in this problem, the shape of the distribution remains essentially the same unless the surface charge density is
very large. In the latter case the field will tend to be radial. The magnitudes and exact distributions will vary depending on the
relative values of q and ρs .
P(R,θ)

R
r R2 R1
q' a q θ
q z
b b q'
d=a+r d=a+r

Figure A Figure B

Problem 5.28. Two point charges outside grounded conducting sphere.

Each one of the point charges may be viewed as acting separately and generating an image charge inside the sphere. The size
of the image charge and its location are defined as in Eq. (5.42) with the configuration shown in Figure A.
2 qa
b=a m, q' = − C
d d
a. For the left charge, the image charge is positive, for the right charge, the image charge is negative. In this case, d = 2a and
therefore:
2 qa q
b=a = a q' = − = −
2a 2 2a 2

142
The image charges, their locations and magnitudes are shown in Figure B.
To calculate the electric field intensity we first calculate the electric potential at a general point in space but do so in spherical
coordinates. Choosing the axis to coincides with the z axis (the charges and their images are on the z axis) makes the potential
independent of the φ axis. Taking a point at P(R,θ) and using Figure B, we can write for the distances from the four charges
to the point P:

R 1 = R 2 + b 2 − 2Rbcosθ = R 2 + (a/2)2 − Racosθ m

R 2 = R 2 + d 2 − 2Rdcosθ = R 2 + 4a 2 − 4Racosθ m
R 3 = R 2 + b 2 − 2Rbcos(180− θ) = R 2 + (a/2)2 + Racosθ m
R 4 = R 2 + d 2 − 2Rdcos(180− θ) = R 2 + 4a 2 + 4Racosθ m
With these, the potential at P(R,θ) is:

V(R,θ) = q − 1 + 1
4πε 0 2 2
2 R + (a/2) − Racosθ
2 2
R + 4a − 4Racosθ
+ 1 − 1 V
2 2 2 2
2 R + (a/2) + Racosθ R + 4a + 4Racosθ
The electric field intensity is calculated from this through the gradient in spherical coordinates:
E = − ∇V = − R ∂V − θ 1 ∂V − φ 1 ∂V V
∂R R ∂θ R sonθ ∂φ m

The last term is zero since the potential is independent of φ. Calculating each of the remaining terms separately, we get:

E R = − R ∂V = − R q acosθ − 2R − 2R − 4acosθ
∂R 8πε 0 2 R 2 + (a/2)2 − Racosθ 3/2 R 2 + 4a 2 − 4Racosθ
3/2

− 2R + acosθ + 2R + 4acosθ V
2 R 2 + (a/2)2 + Racosθ
3/2
R 2 + 4a 2 + 4Racosθ
3/2 m

E θ = − θ 1 ∂V = − θ q Rasinθ − 4Rasinθ
R ∂θ 8πε 0 R 2 R 2 + (a/2)2 − Racosθ 3/2 R 2 + 4a 2 − 4Racosθ
3/2

+ R asinθ − Rasinθ V
2 2
2 R + (a/2) + Racosθ
3/2 2 2
R + 4a + 4Racosθ
3/2 m

b. The electric field intensity inside the sphere is zero. The image charge only represents the field outside the sphere. In
reality, there is an induced charge density on the surface of the sphere which, together with the original charge generate the
field outside. Another way to look at it is that the sphere is at a constant potential and therefore the field inside must be zero.

a
.P(x,y)
−q q' − q' q z R4 R3 R1 R2
b b R
2a 2a −q q/2 θ − q/2 q z
a a
2a 2 2 2a
Figure A Figure B

Problem 5.29. Potential in infinite channel.

The geometry described here is similar to the geometry in example 5.15. We will however solve it from scratch even though
the solution in example 5.15 may be easily modified to answer this problem as well. The general solution is:
V(x,y) = A le kx + Ble −kx A 2 sinky + B2 cosky V

where k x2 = − ky2 = k 2 was used. For the solution in the x-direction we used the exponential form. Because we anticipate
using values of x that tend to infinity. For such values, exponential forms are more convenient than the hyperbolic forms. To
satisfy the boundary conditions we write:

143
1) At y = 0, ⇒ V(0,y) = B1 C le ky + Dle −ky ⇒ B1 = 0
2) At y = b, ⇒ V(x,b) = A 2 sinkb A le kx + Ble −kx ⇒ A2 sinkb = 0
This gives:
kb = mπ ⇒ k = mπ
b
where m is any integer, including zero. We will however exclude m = 0 from the solution because it leads to k = 0 and a
linear solution of the form Ax + B. Similarly, the negative values of m need not be considered because negative m will only
change the sign of the solution. The general solution at this stage looks as:
V(x,y) = A2 sinmπ y A le mπx/b + Ble −mπx/b V
b
3) At x = ∞, ⇒ m
V(∞,y) = A2 sin π y Ale mπ∞/b → Al = 0
b
The solution at this stage is:
V(x,y) = Csinmπ y e −mπx/b V
b
where C = A 2 A 1 .
4. At x = 0, V(0,y) = V 0 : To satisfy this condition, we cannot simply substitute x = 0 in the general solution. If we did, the
solution would be sinusoidal in the x direction and no constant C can satisfy the boundary condition. However, the solution
may also be written as a superposition of solutions of the form above. We write:
∞
C msinmπ y e − b
mπx
V(x,y) = ∑
m=1 b
V
Now we substitute x = 0:
∞
V(0,y) = V0 = ∑ Cmsinmbπ y
m=1
V

The latter form is a Fourier sine series which, in effect, approximates the pulse V(0,y) = V 0 , 0 ≤ y ≤ b, by an infinite series. In
this sense, Cm are the amplitudes of the coefficients of the series. To obtain Cm, we multiply both sides by sin(pπy/a) where p
is an integer, and integrate both sides from zero to b. This is a general technique we will use again and is due to Fourier
himself:
b b ∞ ∞ b
pπ y mπ y pπ y mπ y pπ y
V 0 sin
b
dy = ∑ Cmsin
m=1 b
sin
b
dy = ∑
m=1
C msin
b
sin
b
dy
y=0 y=0 y=0

where the integration and the sum were interchanged. Each side of the relation is integrated separately. The left hand side
gives:
2bV0 for p odd
b
pπ y pπ
V 0 sin dy =
y=0 b 0 for p even
For the right hand side, we integrate each integral in the sum. For any value of m we get::
C mb for p=m
b

C msinmπ y sinpπ y dy = 2
y=0 b b 0 for p ≠m

To satisfy both conditions above, m must be odd and p = m. Any other value yields zero. Thus:

C m = 4V 0 m=1,3,5,....
mπ
If we substitute this in the general solution we obtain the general solution inside the box:
∞
V(x,y) = 4V 0 ∑ 1 sinmπ y e −mπx/b V
π m=1,3,5... m b

Problem 5.30. Potential in infinite channel.

144
a. The general solution may be written directly from Eq. (5.59) or (5.61) (see also Problem 5.31) as:
V(x,y) = A 1 sin(kxx) + A2 cos(kxx) . A 3 sinh(kyy) + A4 cosh(kyy) V
Using: k x2 = − ky2 = k 2 we can write:
V(x,y) = A 1 sin(kxx) + A2 cos(kxx) . A 3 sinh(kyy) + A4 cosh(kyy) V
To find the constants, we use the following boundary conditions:
V(0,y) = 0, V(a,y) = 0, V(x,0) = 0 V(x,b) = V0 = 10 V
At x = 0:
V(0,y) = A 1 sin(0) + A2 cos(0) . A 3 sinh(k y) + A4 cosh(ky) = 0 → A2 = 0
At x = a:
V(a,y) = A 1 sin(k a) . A 3 sinh(k y) + A4 cosh(ky) = 0 → A 1 sin(k a) = 0
This gives:
sin(k a) = 0 → k a = mπ → k= mπ
a
The solution at this point looks like:

V(x,y) = A1 sin mπ x . A 3 sinh mπ y + A4 cosh mπ y V

a a a
From the third boundary condition at y = 0:

V(x,0) = A1 sin mπ x A 3 sinh 0 + A4 cosh mπ y = 0 → A4 = 0

a a

The solution now is of the form:

V(x,y) = A sin mπ x sinh mπ y V (1)
a a
where A = A 1 A 3 is the only remaining constant.
It is not possible to find a single constant A that will satisfy the condition at y = b. Therefore, we rewrite the general form as
an infinite sum of the form of (1) as:
∞
V (x,y) = ∑ A m sin mπa x sinh mπ y V
m=1 a
This is admissible provided that we can find the constants Am that will satisfy the solution at y = b.
At y = b we have:
∞
V (x,b) = V0 = ∑ A m sin mπ x sinh mπ b V
m=1 a a

Multiply both sides by sin(pπx/a) and integrate between 0 and a:

a ∞ a
pπ x pπ x
V0 sin =∑ A m sin mπ x sin sinh mπ b
0
a m=1 0
a a a
Left hand side:
a 2aV0 if p = odd
pπ x pπ
V0 sin =
0
a 0 if p = even
Right hand side:
a
pπ x
A m sinh mπ b sin mπ x sin = Am sinh mπ b ⋅ a for p = m
0 a a a a 2
Both conditions are satisfied if p = m and m is odd. Substituting this back for m = p = 1,3,5,.. we get:
2aV0 = A sinh mπ b ⋅ a → Am = 4V 0
m
mπ a 2 mπ sinh mπ b
a
145
Thus the general solution is:
mπ y
∞ sinh
V (x,y ) = 4V 0 ∑ 1 sin mπ x a V
π m = 1,3,5,... m a
sinh m π b
a
b. A sketch of the potential is shown in Figure 6.9a in the textbook.

Problem 5.31. Potential in infinite channel.

The general solution may be written directly from Eq. (5.59) or (5.61) as:
V(x,y) = A 1 sin(kxx) + A2 cos(kxx) . A 3 sinh(kyy) + A4 cosh(kyy) V
2 2
Using: − ky = k we can write:

V(x,y) = A 1 sin(k x) + A2 cos(kx) . A 3 sinh(k y) + A4 cosh(ky) V

To find the constants, we use the following boundary conditions:

V(0,y) = 0, V(a,y) = 0, V(x,0) = 0 V(x,b) = V0 sinπ x

a
At x = 0:
V(0,y) = A 1 sin(0) + A2 cos(0) . A 3 sinh(k y) + A4 cosh(ky) = 0 → A2 = 0
At x = a:
V(a,y) = A 1 sin(k a) . A 3 sinh(k y) + A4 cosh(ky) = 0 → A 1 sin(k a) = 0
This gives:
sin(k a) = 0 → k a = mπ → k= mπ
a
The solution at this point looks like:
V(x,y) = A1 sin mπ x . A 3 sinh mπ y + A4 cosh mπ y V
a a a
From the third boundary condition at y = 0:

V(x,0) = A1 sin mπ x A 3 sinh 0 + A4 cosh mπ y =0 → A4 = 0

a a
and the solution now is of the form:
V(x,y) = A sin mπ x sinh mπ y V (1)
a a
where A = A 1 A 3 is the only remaining constant. Although in general it is not possible to find a constant A that will satisfy the
condition at y = b, in this case we can do so. Nevertheless, we rewrite the general form as an infinite sum of the form of (1)
as:
∞
V (x,y ) = ∑ A m sin mπ x sinh mπ y V
m=1 a a
This is admissible provided that we can find the constants Am that will satisfy the solution at y = b At y = b we have:
∞
V (x,b ) = V0 sin π x = ∑ A m sin mπa x sinh mπ b V
a m=1 a

Multiply both sides by sin(pπx/a) and integrate between 0 and a:

a ∞ a
pπ x pπ x
V0 sin π x sin =∑ A m sin mπ x sin sinh mπ b
0
a a 1 0
a a a
a V 0 a if p = 1
pπ x
Left hand side: V 0 sin π x sin = 2
0
a a 0 if p = otherwise
Note: in this case m = 1 since a single half sinusoid matches the solution exactly. That is to say, that a sinusoid can be
approximated by a single sinusoid - an obvious answer. The right hand side:

146
 a

A m sinh mπ b sin mπ x sin pπ x = Am sinh mπ b ⋅ a for p = m

0 a a a a 2

Both conditions are satisfied if m = p = 1. Substituting this back for m = p = 1 we get:

V 0 a = A1 sinhπ b . a → A1 = V0
2 a 2
sinhπ b
a
Thus the general solution is:
V (x,y) = V 0 sinh π y sin π x V
a a
sinhπ b
a

Problem 5.32. Potential in infinite channel.

The potential in the infinite channel may be viewed as the superposition of the solution of the two problems described in
Figures A and B. In Figure A, all sides are at zero potential except the side at x = a which is at potential V 2 . In Figure B, all
sides are at zero potential except the side at x = 0 which is at a constant potential V 1 . Solving for each of them separately we
obtain the complete solution. We start with Figure A (see also Problem 5.30). The general solution may be written directly
from Eq. (5.59) or (5.61) (see also Problem 5.31) as:
V A(x,y) = A 1 sinh(kxx) + A2 cosh(kxx) . A 3 sin(kyy) + A4 cos(kyy) V
Note that we chose the hyperbolic variation in the x direction because in the y direction we expect a simple sinusoidal
solution (which would guarantee that the potential is zero at y = 0 and y = a). Using: k x2 = − ky2 = k 2 we can write:
V A(x,y) = A 1 sinh(k x) + A2 cosh(kx) . A 3 sin(k y) + A4 cos(ky) V
To find the constants, we use the following boundary conditions (from Figure A):
V(0,y) = 0, V(a,y) = V2 , V(x,0) = 0 V(x,b) = 0
At x = 0:
V A(0,y) = A 1 sinh(0) + A2 cosh(0) . A 3 sin(k y) + A4 cos(ky) = 0 → A2 = 0
At y = 0:
V A(x,y) = A 1 sinh(k x) . A 3 sin(0) + A4 cos(0) = 0 → A4 = 0
The solution at this stage, setting A = A 1 A 3 is: V A(x,y) = Asinh(k x) sin(k y) V
At y = b: V A(x,b) = Asinh(k x) sin(k b) = 0 → sin(k b) = 0
This gives: sin(k b) = 0 → k b = mπ → k = mπ
b
The solution at this point looks like:
V A(x,y) = Asinh mπ x sin mπ y V
b b
It is not possible to find a constant A that will satisfy the condition at y = b. Therefore, we rewrite the general form as an
infinite sum of the form of (1) as:
∞
V A (x,y ) = ∑ A m sinh mπ x sin mπ y V
m=1 b b
This is admissible provided that we can find the constants Am that will satisfy the solution at x = a.
At x = a we have:
∞
V A (a,y ) = ∑ A m sinh mπ a sin mπ y V
m=1 b b
Multiply both sides by sin(pπy/b) and integrating between 0 and b:
b ∞ b
pπ y mπ y pπ y
V2 sin =∑ A m sin sin sinh mπ a
0 a 1 0 b b b

147
 b 2bV2 if p = odd
Left hand side: V2 sin pπ y = pπ
0 b 0 if p = even
Right hand side:
b
A m sinh mπ a sin nπ y sin pπ y = Am sinh mπ b ⋅ b for p = m
0 b b b a 2

Both conditions are satisfied if p = m and m is odd. Substituting this back for m = p = 1,3,5,.. we get:
2bV2 = Am sinh mπ a ⋅ b → Am = 4V 2
mπ b 2 mπ sinh mπ a
b
Thus the general solution is:
∞ sinh mπ x
V A (x,y) = 4V 2 ∑ 1 sin mπ y b V (1)
π m = 1,3,5,... m b sinh mπ a
b
We could now repeat the whole process for the configuration in Figure B. However, the solution is obviously of the same
form as in (1) if we substitute (a − x) for x and V 1 for V 2 in Eq. (1). With this we can write immediately:

∞ sinh mπ (a−x)
V B (x,y) = 4V 1 ∑ 1 sin mπ y b V
π m = 1,3,5,... m b sinh mπ a
b
The sum of V A(x,y) and V B(x,y) is the solution:
V(x,y) = VA (x,y ) + VB (x,y ) =
mπ (a−x)
∞ sinh mπ x ∞ sinh
4 V2 m π y m π y
∑ 1 sin b
m = 1,3,5,... m
b + V1 ∑ 1 sin b V
π sinh mπ a m = 1,3,5,... m b sinh mπ a
b b

y y
V=0 V=0

b V=0 V=V 2 b V=V 1 V=0

V=0 x V=0 x
a a
Figure A Figure B

Problem 5.33. Potential in a box with conducting sides.

The solution is the superposition of two separate problems shown in Figures A and B. In Figure A, all sides are at zero
potential except the side at z = a which is at V = V 0 . In Figure B, all sides are at zero potential except the side at x = a which
is at potential V 0 . Both of these are solved in a manner similar to Example 5.16. After calculating the potential distribution in
Figure A and B, we sum them up to obtain the total potential distribution.
V A(x, y ) + VB(x, y ) = V (x, y )
We start with the geometry in Figure A since it is essentially the same as in Example 5.16.
The general solution is given in Eq. 5.59:
V A(x,y,z) = A 1 sin(kxx) + A2 cos(kxx) A 3 sin(kyy) + A4 cos(kyy) A 5 sinh(kzz) + A6 cosh(kzz) V
Boundary conditions are:
at x = 0, V(0,y,z) = 0 at x = a, V(a,y,z) = 0
at y = 0, V(x,0,z) = 0 at y = b, V(x,b,z) = 0
at z = 0, V(x,y,0) = 0 at z = c, V(x,y,c) = V0
148
We start by satisfying the boundary conditions at x = 0, x = a, y = 0, y = b and z = 0.
1. At x = 0: →
V A(0,y,z) = A 2 A 3 sin(kyy) + A4 cos(kyy) A 5 sinh(kzz) + A6 cosh(kzz) = 0 → A2 = 0
2. At x = a: → V A(a,y,z) = A1 sin(kxa) . A 3 sin(kyy) + A4 cos(kyy) . A 5 sinh(kzz) + A6 cosh(kzz) = 0
This condition can be satisfied if:
sin(kxa) = 0 → kxa = mπ → kx = mπ , m= 1, 2, 3,...
a
3. At y = 0: → V A(x,0,z) = A1 sin m π x A 4 A 5 sinh(kzz) + A6 cosh(kzz) = 0 → A4 = 0
a
4. At y = b: → V A(x,b,z) = A1 sin mπ x A3 sin(kyy) A 5 sinh(kzz) + A6 cosh(kzz) = 0
a
This condition can be satisfied if:
sin(kyb) = 0 → kyb = nπ → ky = nπ , n= 1, 2, 3,...
b
5. At z = 0: → V A(x,y,0) = A1 sin mπ x A 3 sin nπ x A 6 = 0 → A6 = 0
a b
Note: both m and n can be zero but this leads to a linear solution (see text). Also, m,n, can be negative but, since negative
values merely change the sign of the solution, we exclude these values. Now, since k z2 = − k x2 + ky2 , (see Eq. (5.52)) and
combining the constants A 1 , A 3 , and A 5 into a single constant, we can write the general solution at this stage as:
2 2
V A(x,y,z) = Csin mπ x sin nπ x sinh k x2 +k y2 z = Csin mπ x sin nπ x sinhπ m2 + n 2 z V
a b a b a b
First, we note that there is only one constant to be defined. We can do that by using our sixth and final boundary condition at
z = c. Before doing so, we note that any superposition of solutions is also a solution to the original equation. The following
sum may be taken as a general solution:
∞ ∞
m 2 + n2 z
V A(x,y,z) = ∑ ∑ Cmnsin maπ x sin nbπ x sinhπ V
m=1 n= 1 a2 b2
6. The sixth boundary condition is substituted into this sum in order to evaluate the constant Cmn. For z = c, V(x,y,c) = V 0 ,
and using the short form notation for kz:
kz = π m 2 + n2
a2 b2
we get the general solution as:
∞ ∞
V A(x,y,c) = V0 = ∑ ∑ Cmnsin mπa x sin nπ y sinh(kzc) V
m=1 n= 1 b
As in Example 5.15, this is a Fourier series but now it is a series in two variables. This solution applies anywhere on the
plane z = c (0 ≤ x ≤ a, 0 ≤ y ≤ b). To find the constant C mn, we perform two operations. First we multiply both sides by the
term:
pπ x qπy
sin sin
a b
where p and q are integers (see Example 5.15). This gives:
∞ ∞
V 0 sin pπ x sin qπy = ∑ ∑ C mnsin mπ x sin nπ y sin pπ x sin qπy sinh(kzc)
a b m=1 n= 1 a b a b
In the second step both sides of the expression are integrated on the plane z = c in the limits 0 ≤ x ≤ a, 0 ≤ y ≤ b. To simplify
evaluation we perform the integration on the left hand side and right hand side separately. For the left hand side:

V 0 2a 2b = 4abV20 for p,q, odd

b a a b pπ qπ pqπ
V0 sin pπ x sin qπy = V0 sin pπ x dx sin qπy dy =
a b a ba
y=0 x=0 x=0 y=0 0 for p,q, even

For the right hand side, interchanging between the summation and integration and evaluating each of the integrals in the sum
separately yields:
149
 a b
C mn sin mπ x sin pπ x sin nπ y sin qπy sinh(kzc) dxdy =
x=0 y=0
a a b b
a b
pπ x nπ y qπy
C mnsinh(kzc) sin mπ x sin dx sin sin dy
x=0
a a y=0 b b
To evaluate the integrals, we write first for the first integral, assuming m ≠ p:
pπ x sin mπ + pπ x a
a sin mπ −
sin mπ x sin pπ x dx = a a − a a =0
a a pπ
x=0 2 mπ − 2 mπ + pπ
a a a a x=0
If m = p, we get:
a
a a sin2 mπ x
sin mπ x sin pπ x dx = sin mπ x dx = x − a =a
2

x=0
a a x=0
a 2 4 m π 2
a x=0
Similarly, for the second integral we have:
For n ≠ q:
b
b sin nπ − qπ x sin nπ + qπ x
sin nπ x sin qπx dx = a a − a a =0
a a nπ qπ nπ qπ
x=0 2 − 2 +
a a a a x=0
For n = q:
b
b b sin2 nπ x
sin nπ x sin qπx dx = sin nπ x dx = x −
2 a =b
x=0 a a x=0 a 2 4 nπ 2
a x=0
Thus, we obtain the following:
a b

C mn sin mπ x sin pπ x sin nπ y sin qπy sinh(kzc) dxdy

x=0 y=0 a a b b
C mnabsinh(kzc)
a b for m=p, n=q
sin mπ x sin pπ x dx sin nπ y sin qπy dy = 4
= Cmnsinh(kzc)
x=0 a a y=0 b b 0 m≠p, n≠q
Now, satisfying both the conditions m = p, n = q, and m,n, odd, we get:

C mn = 16V 0
mnπ 2 sinh(kzc)
Substituting this back into the general solution we get by equating the left and right hand sides:

sin mπ x sin nπ y sinh π z m 2+ n 2

∞ ∞
V A(x,y,z ) = 16V2 0 ∑ ∑ a a a a
π 2 2
m = 1, 3, 5 n = 1, 3, 5
m n sinh π a m + n
a a
V B is the same except that now the x and z coordinates interchange. Without repeating the process above, we can write:
∞ ∞ sin mπ z sin nπ y sinh π x m 2+ n 2
16V a a a a
V B(x,y,z) = 0
∑ ∑
π 2 m = 1, 3, 5 n = 1, 3, 5 m n sinh π a m 2 + n 2 V
a a
The total potential is the sum of the two:
∞ ∞ sin nπ y
16V a
V (x,y,z) = 0
∑ ∑
π 2 m = 1, 3, 5 n = 1, 3, 5 m n sinh π m 2 + n 2

sin mπ x sinh π z m 2 + n 2 + sin mπ z sinh π x m 2 + n2 V

a a2 a2 a a2 a2

150
 y y
y=a y=a
V =0 V =0

x= a x x= a x

z z=a z z=a
V=V 0 V=V 0
Figure A Figure B

Problem 5.34. Application: Electrostatic precipitator.

View the right plate as being at +V 0 or +50 kV, the left
2
plate at − V 0 or −50 kV (see Figure A). Because the a
2 − 50 kV +50 kV
potential at the center is zero, the attached solution applies
(see example 5.17). φ
r Figure A.
a.
∞
k
V (r,φ ) = 2V 0 ∑ 1 r sin kφ , r<a V
π k = 1, 3, 5,... k a
∞
k
V (r,φ ) = 2V 0 ∑ 1 a sin kφ , r>a V
π k = 1, 3, 5,... k r
b.
∞
k−1
E = −∇V = − r∂V − φ 1 ∂V = − 2V 0 ∑ 1 r rsin kφ + φcos kφ 0<r<a V
∂r r ∂φ aπ k = 1, 3, 5,... k a m
∞
k−1
E = −∇V = − r∂V − φ 1 ∂V = 2V 0 ∑ 1 r rsin kφ − φcos kφ r>a V
∂r r ∂φ aπ k = 1, 3, 5,... k a m

Problem 5.35. Application: Electrostatic precipitator.

Use superposition of two solutions. By solving for the two conditions shown in Figures A and B, the grounded quarter
spheres are at zero potential (ground) and the solution to each of the two figures is essentially the same as in Problem 5.34.
a. From Figure A, and taking V 1 = 50 kV, (see solution to Problem 5.34):

∞
k
V (r,φ ) = 4V 1 ∑∞ 1 r sin nφ r<a V
π k = 1, 3, 5,... k a k
V (r,φ ) = 4V 1 ∑ 1 a sin nφ r>a V
π k = 1, 3, 5,... k r

Add to this the fields due to Figure B: This has the same solution, but V r is different and the solution is shifted by φ " = π /2
∞
V 2 (r,φ ) = 4V 2 r ksin k φ + π
π ∑ 1
k = 1, 3, 5,... k a 2
r<a V
∞
V 2 (r,φ ) = 4V 2 a ksin k φ + π
π ∑ 1
k = 1, 3, 5,... k r 2
r>a V

Final solution is obtained by summing the two solutions above

V (r,φ ) = V1 (r,φ ) + V2 (r,φ )

151
 ∞
4 V 1 + V2 1 r k
V (r,φ ) = ∑ sin kφ + sin k φ + π r<a V
π k = 1, 3, 5,... k a 2
∞
4 V 1 + V2 1 a k
V (r,φ ) = ∑ sin kφ + sin k φ + π r>a V
π k = 1, 3, 5,... k r 2
In this case,
V 1 + V2 = 100 kV
We get:
5 ∞
k
V (r,φ ) = 4×10 ∑ 1 r sin kφ + sin k φ + π , r<a V
π k = 1, 3, 5,... k a 2
5 ∞
k
V (r,φ ) = 4×10 ∑ 1 a sin kφ + sin k φ + π , r>a V
π k = 1, 3, 5,... k r 2
b. The electric field intensity is found through the gradient as:

E = −∇V = − r∂V − φ 1 ∂V
∂r r ∂φ
∞
4 V 1 + V2 k−1
=−
aπ ∑ 1 r
a
rsin kφ + φcos k(φ + π /2) , r<a V
m
k = 1, 3, 5,... k

E = −∇V= − r∂V − φ 1 ∂V
∂r r ∂φ
∞
4 V 1 + V2 k−1
=−
aπ ∑ 1 a
r
rsin kφ − φcos k(φ + π /2) , r>a V
m
k = 1, 3, 5,... k

+50 kV

a a
− 50 kV +50 kV

φ
r Figure A. −50 kV Figure B.

152