Today in Physics 217: the method of images

Solving the Laplace and

Poisson equations by
a y a
sleight of hand -q q
Introduction to the
method of images b b
Caveats
Example: a point charge x
and a grounded
b b
conducting sphere
Multiple images
q a a -q

9 October 2002 Physics 217, Fall 2002 1

 Solving the Laplace and Poisson equations by
suuuu of hand
The guaranteed uniqueness of solutions has spawned several
creative ways to solve the Laplace and Poisson equations for
the electric potential. We will treat three of them in this class:
Method of images (today).
Very powerful technique for solving electrostatics
problems involving charges and conductors.
Separation of variables
Perhaps the most useful technique for solving partial
differential equations. You’ll be using it frequently in
quantum mechanics too.
Multipole expansion
Fermi used to say, “When in doubt, expand in a power
series.” This provides another fruitful way to approach
problems not immediately accessible by other means.
9 October 2002 Physics 217, Fall 2002 2
 Introduction to the method of images
A point charge lies a distance d above
a infinite, conducting, grounded
plane. Calculate the potential V
everywhere above the plane. z
This looks like a Laplace-equation
problem, and we know some
q
boundary conditions at the plane:
V = 0, Ex = Ey = 0.
d
But there’s charge induced on the
grounded plane.The electrostatic
potential can not be calculated
directly without knowing the
induced charge distribution.

9 October 2002 Physics 217, Fall 2002 3

 Introduction to the method of images (continued)
Consider alternatively the situation
of two point charges q and –q,
separated by 2d. z
The potential can be calculated
directly and is equal to
q
q
V ( s,φ , z ) = d
2
s2 + ( z − d )
−q
+ d
2
s + (z + d)
2
-q
Note that ∇ 2V = −4πρ is
automatically satisfied. This also
gives V = 0 on the plane z = 0, just
as it would need to be for the
grounded plane.
9 October 2002 Physics 217, Fall 2002 4
 Introduction to the method of images (continued)
The potential yields the electric field, as usual:
∂V 1 ∂V ∂V
E ( s , φ , z ) = −— V = sˆ + φˆ+ zˆ No φ dependence
∂s s ∂φ ∂z
qs qs
= 32
− 32
 s 2 + ( z − d )2   s 2 + ( z + d )2 
   
q ( z − d ) zˆ q ( z + d ) zˆ
+ 3 2
− 32
s 2 + ( z − d )2   s 2 + ( z + d )2 
   
2 qdzˆ E perpendicular to z = 0 plane,
Note E ( s , φ , 0 ) = − ,
( 2
s +d )2 32 at z = 0, just as it would need to
be with the grounded plane.
E⊥ , above ( s , φ , 0 ) qd
whence σ ( s , φ ) = =− .
( )
4π 3 2
2π s 2 + d 2
9 October 2002 Physics 217, Fall 2002 5
 Introduction to the method of images (continued)
z

Thus, for z ≥ 0 , the two-charge

q
potential satisfies the Poisson
equation and the boundary d
conditions for the single charge
– grounded plane problem: it is
a solution to this problem. d

But there is no “a” solution, -q

=
z
only “the” solution, because
solutions of electrostatics
q
problems are unique.
d

9 October 2002 Physics 217, Fall 2002 6

 Introduction to the method of images (continued)

How did we know this would

work? Is there a method by
z
which we could guess these
solutions in general?
q
Yes. The auxiliary charge is
the image of the original
Mirror
charge in the “mirror” that
comprises the grounded
conducting plane. Other
configurations of charges and -q
grounded conductors can be
treated similarly: as if they
were objects, images and
mirrors in geometrical optics.

9 October 2002 Physics 217, Fall 2002 7

 Caveats for the method of images

The solution for the images is only the same as that for the
conductor, in the region outside the conductor! In
particular, the field is still zero, and the potential constant,
inside the conductor. Remember that the image charge
doesn’t really exist.
In particular, the potential energy of the
charge+conductor arrangement is quite different from the
charge-image charge combination, because the field is
finite for the latter in locations where the field of the
former is zero. (Got it?)
Remind yourself of these facts by noting, in every image
solution, that you can calculate the induced charge on the
surface of the grounded conductor.

9 October 2002 Physics 217, Fall 2002 8

Potential for a point charge and a grounded sphere
(Example 3.2 + Problem 3.7 in Griffiths)
A point charge q is situated a
distance Z from the center of
a grounded conducting
sphere of radius R.
Find the potential
everywhere.
Find the induced surface R q
V=0
charge on the sphere, as
Z
function of θ. Integrate
this to get the total
induced charge.
Calculate the potential
energy of the system.

9 October 2002 Physics 217, Fall 2002 9

Potential for a point charge and a grounded sphere
(continued)
We need to find a position to put
charge q’ such that the boundary
conditions on the sphere are satisfied.
R q
Start by determining where this
V=0
charge should lie, using the points
Z
along the z axis:
q q′
VP = 0 = +
Z − R R − z′
q
⇒ q′ = − ( R − z′ ) Q
Z−R d
q q′ R d' q
VQ = 0 = +
Z + R R + z′ V=0 z' q'
q Z
⇒ q′ = − ( R + z′ ) P
Z+R

9 October 2002 Physics 217, Fall 2002 10

 Potential for a point charge and a grounded sphere
(continued)
q q Q
So ( R − z ′ ) = ( R + z′ ) P’
Z−R Z+R d
( Z + R ) ( R − z′ ) = ( Z − R ) ( R + z′ ) R d' q
θ
Rearrange: V=0 z' q'
Z
z′ ( Z − R ) + z′ ( Z + R ) = 2Zz′
P
= R ( Z + R ) − R ( Z − R ) = 2 R2
R2 q q  R2  R
⇒ z′ = , q′ = − ( R + z′ ) = −  R +  = −q
Z Z+R Z+R Z  Z

Now consider an arbitrary point on the sphere, P’:

d = R 2 + Z 2 − 2 RZ cosθ

2 R2 R4
′2
d ′ = R + z − 2 Rz′ cosθ = R + 2 − 2 R 2
cosθ
Z Z
9 October 2002 Physics 217, Fall 2002 11
Potential for a point charge and a grounded sphere
(continued)
The potential should come out to be zero there, and sure enough,
R
q q′ q − q
VP′ = + = + Z
d d′ R 2 + Z 2 − 2 RZ cos θ 4
R2
2 R
R + 2 − 2R cosθ
Z Z
R
q − q
= + Z
R 2 + Z 2 − 2 RZ cosθ 2 R2 R4
R + 2 − 2R cosθ
Z Z
q q
= − =0
2 2 2 2
R + Z − 2 RZ cosθ R + Z − 2 RZ cosθ
Thus the potential outside the grounded sphere is given by the
superposition of the potential of the charge q and the image charge q’.

9 October 2002 Physics 217, Fall 2002 12

Potential for a point charge and a grounded sphere
(continued)
So the potential at some point ( r ,θ , φ ) outside the sphere is given by
R
q −q
V= + Z
2 2
r + Z − 2rZ cosθ 2 R4
R2
r + 2 − 2r cosθ
Z Z
q q
= −
r 2 + Z 2 − 2rZ cosθ  rZ  2
2
  + R − 2 rZ cosθ
 R
Now for the induced charge density:
1 1 ∂V
σ= Er = −
4π 4π ∂r

Differentiate the formula above for the potential, and evaluate it at r = R:

9 October 2002 Physics 217, Fall 2002 13
Potential for a point charge and a grounded sphere
(continued)
 
 2 
rZ
 − 2 + Z cos θ 
∂V −r + Z cosθ R
= q − 
( )
∂r  2 3 /2 3 /2 
r =R r + Z 2
− 2 rZ cosθ   rZ 2 
 2
   + R − 2 rZ cos θ  
  R   
   r =R
Z2
− + Z cosθ
− R + Z cosθ R
=q −q
( ) ( )
3 /2 3 /2
R 2 + Z 2 − 2 RZ cosθ Z 2 + R 2 − 2 RZ cosθ

q Z2 − R2
=
(R )
R 2 2 3 /2
+ Z − 2 RZ cosθ

1 ∂V q Z2 − R2
σ =− =−
(R )
4π ∂r 4π R 2 2 3 /2
r =R + Z − 2 RZ cosθ
9 October 2002 Physics 217, Fall 2002 14
Potential for a point charge and a grounded sphere
(continued)
From this we get the total charge induced on the grounded sphere:
q
( ) ∫0 sin θ dθ
π
∫
Q = σ R sin θ dθ dφ = − R Z 2 − R 2
2

(R )
2 2 2 3 /2
+ Z − 2 RZ cosθ
π
 1 
 − 
q
(
= − R Z2 − R2 
2
) ZR
 R 2 + Z 2 − 2 RZ cos θ


  0
q Z2 − R2  1 1  R Goes to zero as Z
=  −  = − q
2 Z  R + Z R−Z  Z goes to infinity.

9 October 2002 Physics 217, Fall 2002 15

Potential for a point charge and a grounded sphere
(continued)
Potential energy: first, the force between the charge and image,

 R 
q− q
qq ′  Z  ZRq 2
Fqq′ = zˆ = zˆ = − zˆ
( )
2 2 2
( Z − z′ )  R2  Z2 − R2
 Z − 
 Z 

and then the work to bring q in from infinity:

Z
Z Z zRq 2 − Rq 2 1 Rq 2
W = ∫ − Fqq′ ⋅ dl = ∫ dz = =−
∞ ∞
( z2 − R )
2 2 (
2 z2 − R 2 )∞ 2 Z2 − R2

9 October 2002 Physics 217, Fall 2002 16

 Multiple images
Here’s a hint for Problem 3.10, on
this week’s homework:
y a
Two semi-infinite grounded q
conducting planes meet at right
angles. In the region between V=0
them, there is a point charge q, b
situated as shown at right. Set up
the image configuration, and
calculate the potential in this x
region. What charges do you need,
and where should they be located?
What is the force on q? How much V=0
work did it take to bring q in from
infinity?

9 October 2002 Physics 217, Fall 2002 17

 Multiple images (continued)

In order to satisfy the

boundary conditions, three
image charges must be a y a
-q q
added to the system.
The net force on q can be
calculated by determining
b b
the vector sum of the
forces on q due to the
three image charges.
x
The electrostatic energy b b
of the real system is equal
to 1/4 of the electrostatic q -q
a a
energy of the image-
charge system.
9 October 2002 Physics 217, Fall 2002 18