SPECIAL TECHNIQUES-I

Lecture 17: Electromagnetic Theory

Professor D. K. Ghosh, Physics Department, I.I.T., Bombay

Method of Images

The uniqueness theorem for Poission’s or Laplace’s equations, which we studied

in the last couple of lectures, has some interesting consequences. Frequently, it
is not easy to obtain an analytic solution to either of these equations. Even when
it is possible to do so, it may require rigorous mathematical tools. Occasionally,
however, one can guess a solution to a problem, by some intuitive method. When
this becomes feasible, the uniqueness theorem tells us that the solution must be
the one we are looking for.
One such intuitive method is the “method of images” a terminology borrowed
from optics. In this lecture, we illustrate this method by some examples.

1
Consider an infinite, grounded conducting plane occupying which occupies the
x-y plane. A charge q is located at a distance d from this plane, the location of the
charge is taken along the z axis. We are required to obtain an expression for the
potential everywhere in the region z > 0 , excepting of course, at the location of
the charge itself. Let us look at the potential at the point P which is at a distance
!" from the charge q (indicated by a red circle in the figure).

Instead of solving the problem from the first principles, let us suppose there
existed an “image charge” #’ at a distance %’ “below” the plane. This is very
similar to the image of an object located in front of a mirror, the image of the
object is “behind” the mirror. The image is virtual because we cannot capture
the image on a screen behind the mirror. In the same way, the image charge is
located at a virtual distance and the charge itself is fictitious. Let the point P be
at a distance !2 from the image charge. We take the origin on the plane, as
shown. Let the position of P be ((, *, +).

The potential ." / at P due to the real charge # and .0 / due to the image
charge #’ is given by superposition,
# # 1
." = =
4345 !" 4345 ( 0 + * 0 + + − % 0
#′ #′ 1
.0 = =
4345 !0 4345 ( 0 + * 0 + + + %′ 0

The potential P is given by . / = ." + .0 . Note that this expression for the
potential satisfies the Laplace’s equation at a general point P above the
" "
conducting plane, since ; 0 = −43> ? (!" ) and ; 0 = −43> ? (!0 ) and at a
<= <@
point P, neither of the distances is zero so that the delta function vanishes. Let us
impose the boundary condition of the potential being zero on the boundary, We
must have, on the plane, ." = −.0 . Since the distances are positive, the charges
# and #’ must have opposite sign. V Further, we must have, substituting + = 0 for
the position of the conducting plane,
#0 ( 0 + * 0 + % B0 = #B0 (( 0 + * 0 + % 0 )
Rewriting this as,
#0 − #B0 ( 0 + * 0 + #0 % B0 − #B0 % 0 = 0
This equation must be satisfied at all points on the conducting plane, I.e for
arbitrary values of x,y. Thus the two terms of the above equation must be
separately zero. This gives us the solution,
#B = −#
%B = %
Thus the image charge is equal and opposite to the object charge and, like in the
case of optical images, the image distance is equal to the object distance.

2
It is important to realize that the field exists only in the region above the
conducting plane. The expression for potential which satisfies Laplace’s
equation everywhere above the plane and also satisfies the boundary condition
on the plane, is given by

# 1 1
. / = −
4345 C0 + + − % 0 C0 + + + % 0

where we have, because of obvious reasons, switched to a cylindrical coordinates

C0 = ( 0 + * 0 as the square of the distance of the foot of the perpendicular from
the point P onto the plane from the origin.

The electric field is obtained as the negative gradient of the potential, and
is given by
E E
D = −;. = −C −F . C, +
EC E+
# C C
= C G − G
4345 C0 + + − % 0 @ C0 + + + % 0 @
+−% ++%
+F G − G
C0 + + − % 0 @C0 + + + % 0 @
We emphasize that this expression is valid only in the region + > 0.
From this expression for the electric field, we can obtain the charge
density induced on the surface. Remembering that charge density is given by the
normal component of the electric field, we only need to evaluate the z
component of the electric field at z=0,
1 −2#% −# %
IJKLMNOL = 45 DP + = 0 = G = −
43 C0 + % 0 @ 23 C0 + % 0 G@
The total induced charge is obtained by integrating this expression in the
entire xy plane,
QJKLMNOL = ∫ IJKLMNOL %( %*
#% 1
=− 23 C%C
23 5 C0 + % 0 G@

#%
= = = −#
C0 + % 0 @ 5
Thus the total induced charge has the same magnitude as the real charge,
though of opposite sign. In the following figure we have plotted the charge
density as a function of x-y coordinates on the plane for some representative
distances of the object charge. One can see that the magnitude of the charge
density is maximum at a point on the plane directly opposite to the real charge
and decreases as the distance from this point increases.

3
 The following figure shows the lines of force and equipotentials. As
expected the lines of force strike the conducting plane normally. The
equipotentials are spheres (shown as circles)

Thus in the above problem we have replaced the original problem of a

charge and a conducting plane by the charge and an image charge and eliminated
the conducting plane altogether. The solution obtained satisfies the boundary
condition that at z=0, the potential is zero. The solution does not make sense in
the region + < 0 but that is immaterial because we did not seek the solution in
that region in any case! The electric field on the plane + = 0 is purely along the

4
vertical plane because, the object and the image charges being equidistant from
a point on the conducting plane, the horizontal components cancel by symmetry.
The uniqueness theorem guarantees that this is the only possible solution.
Let us calculate the field that is generated at the position of the real
charge due to the induced charges. For this we take the general expression for
the electric field and consider only the contribution due to the image charge,
given by the second term within each curly bracket. As the real charge is located
at C = 0, + = %, only the normal component gives non-zero contribution,
# ++% # 1
D 0, % = F − G = −F
4345 C0 + + + % 0 @ 4345 4% 0
PTL,UT5
Thus the force on the charge due to the charges induced on the surface of
W@ "
the conductor is V = −F which is the same as the force exerted by the
XYZ[ XL @
image charge on the real charge (located at a distance 2d from each other.

It is interesting to calculate field at the surface due to the charge q. In this

case, we do a bit of hand waving and consider only half of the field that we have
calculated. This is because the image charge being fictitious does not create a
field on the surface of the conductor. The force on the conductor is then given by
# +−% ++% # −2%
D′P = G − G =
8345 C0 + + − % 0 @ C0 + + + % 0 @ 8345 C0 + % 0 G@
PT5

The force exerted on the surface is then,

B
#0 %0 1 #0
VP = ∫ IDP %] = + 23C%C = +
8345 5 C0 + % 0 ? 16345 % 0
which, as expected from Newton’s third law, is equal and opposite to the
force exerted on the charge by the surface.

We can now proceed to calculate the electrostatic energy of the system.

This is just the work done in bringing the charge from infinity to its position on
the z axis. We already have an expression for the force exerted on the charge
when it is at a distance d from the surface. It is a simple matter to get the
expression for the force when the charge is at a distance z along the z axis. Since
the electrostatic force is conservative, we bring the charge along the z axis. The
work done is then
L
#0 #0
_= %+ = −
16345 + 0 16345 %

Example 1:

Consider two semi infinite grounded, conducting planes intersecting

along the z axis, which is taken out of the plane of the paper (which is the xy

5
plane). A charge q is located at a point (`, a). Calculate the force on the charge q
due to the charges induced on the surfaces of the planes.

The problem is similar to the problem of image formed by two plane

mirrors intersecting at right angles. The image charge due to the y=0 plane is
formed at (`, −a) and that formed by the x=0 plane at (−`, a). Both these images
have charge – #. However, the image of these images which are formed at the
point (−`, −a) has a charge +q.
The potential can be written down at arbitrary point due to these four
charges and it is left as an exercise.
Let us calculate the force on the charge q due to the image charges. Note
that the force due to the charge – # at (`, −a) is along the negative y axis while
that due to the charge – # at (−`, −) is along the negative x axis. The force due to
the charge + q at (−`, −a) is repulsive and is along the diagonal. The distance
between this image and the given charge is 2 `0 + a 0 . Thus, the components of
the forces on the given charge can be written as

#0 1 1
Vc = − 0+ cos g
4345 4` 4 ` + a0
0

6
 #0 1 `
= − 0+ G
4345 4` 4 `0 + a0 @

#0 1 a
Vh = − 0+ G
4345 4` 4 `0 + a0 @

Example 2:

The problem is same as in Example 1, except that instead of a point

charge at (`, a), there is an infinite line charge of charge density iat that point
parallel to the z axis.

The image charges are also line charges of charge density – i each at
(`, −a) and (−`, a). In addition, there is a line charge of density+i at (−`, −a).
The potential due to a line charge parallel to z axis is independent of z coordinate
and only depends on the distance C of the point from the line charge, being given
by,
i l0
. C = ln 0
4345 C
where C is a constant and C = ( − ` + * − a 0 , where ((, *, +) is the
0 0

coordinate of the point P where the potential is sought. Similar expressions for
the potential due to the image charges can be written down and the total

7
potential obtained by superposition. Thus, the net potential at ((, *, +) is given
by
i l0 l0
. (, * = ln + ln
4345 (−` 0+ *−a 0 (+` 0+ *+a 0
0
l l0
− ln − ln
(−` 0+ *+a 0 (+` 0+ *−a 0

It can be seen that the solution satisfies the boundary condition . 0, * =

mn
. (, 0 = 0. Note that the tangential components on the * = 0 plane is –
mc hT5
mn
and that on the ( = 0 plane – are zero, as expected. The normal
mh cT5
components, from which we can infer the charge densities are along the positive
y direction for the former and along the positive x direction for the latter.
E. 4i`a( I hT5
DK hT5 = − =− 0 0 0 0
=
E* hT5 345 ( − ` + a (+` +a 45
E. 4i`a( I cT5
DK cT5 = − =− =
E( cT5 345 * − a 0 + `0 * + a 0 + `0 45

In the following figure we have plotted the charge densities for different
locations of the given charged line indicated on the graph. Consider the situation
where the line charge is at the point ` = 2, a = 1 . The axes are the coordinates
of points on the plane.

Note that the charge density is maximum on the plane at the point (2,1) and
spreads out when one goes away from this point. If you reduce the x distance to,
say, [1,1], the charge density has a smaller peak and spreads out more.

8
The total linear charge density on the y=0 plane can be calculated by integrating
the charge density
i o %( o
%(
i′ hT5 = − −
3 5 ( − ` 0 + a0 5 ( + ` 0 + a0
2i `
= − tanr"
3 a

The linear charge density on the x=0 plane can be written down by symmetry to
be

2i a
i′ cT5 =− tanr"
3 `
" Y
Adding these and using the identity tanr" ( + tanr" = , we get the total linear
c 0
charge density on the two planes to be – i.

The method of images is not restricted to only planar geometry. We will take up
different application in the next lecture.

9
 SPECIAL TECHNIQUES-II
Lecture 18: Electromagnetic Theory

Professor D. K. Ghosh, Physics Department, I.I.T., Bombay

Method of Images for a spherical conductor

The method of images can be applied to the case of a charge in front of a grounded spherical conductor.
The method is not as straightforward as the case of plane conductors but works equally well.
Consider a charge q kept at a distance a from the centre of the grounded sphere. We wish to obtain
an expression for the potential at a point P which is at the position , the potential obviously will
not depend on the azimuthal angle and hence that coordinate has been suppressed. Let the point P
be at a distance from the location of the charge q.

The image charge is located at a distance b from the centre along the line joining the centre to the charge
q. The line joining the charges and the centre is taken as the reference line with respect to which
the angle is measured. Let P be at a distance b from the image charge. Let q’ be the image charge.

The potential at P is given by . Using the property of triangle, we can express the
potential at P as,

Since the potential vanishes at for all values of θ, the signs of and must be opposite, and we
must have,

1
In order that this relation may be true for all values of θ, the coefficient of from both sides of this
equation must cancel,

Since and have opposite sign, this gives,

Substituting this in the θ independent terms above, we get,

which gives . Thus .

It follows that if the object charge is outside the sphere, the image charge is inside the sphere. Using
these, the potential at P is given by

The electric field can be obtained by computing gradient of the potential. It can be easily verified that the
tangential component of the electric field . The normal component is given by,

The charge density on the surface of the sphere is and is given by

2
The charge density is opposite to the sign of q since .

It is interesting to note that unlike in the case of a conducting plane, the magnitude of the image charge
is not equal to that of the object charge but has a reduced value,

π +1
q R( R 2 − a 2 ) dµ
Qind = 2π ∫ σ ( R, θ ) R sin θ dθ = ∫ 2
2

0
2 −1 ( R + a − 2aR µ ) (−2aR)
2 3/2

qR(a 2 − R 2 ) 2 1/2 1 R
=
4aR
(R + a 2 − 2aR µ ) −1
= −q
a

Thus the potential can be written as

The following figure shows the variation of charge density on the surface as a function of the angle θ. As
expected, when the charge comes closer to the sphere, the charge density peaks around θ=0.

Since the distance between the charge and its image is the force exerted on the charge by the
sphere is

For the force is proportional to the inverse cube of the distance of the charge from the centre.
For let α be the distance of q from the surface of the sphere,

3
Consider the limit of (one has to be careful here as the origin would now have shifted to extreme
left; instead, measure distance from the surface of the sphere. The charge q is located at a distance
from the surface,

The distance of the image charge, so that the distance of the image from the surface

as . Thus we recover the case of a charge in front of an

infinite plane.

A Few Special Cases

1. If the sphere, instead of being grounded, is kept at a constant potential , we can solve the problem
by putting an additional charge distributed uniformly over its surface.
2. If the sphere was insulated, conducting and has a charge , we put a charge over the surface
after disconnecting from the ground. The potential at any point then is the sum of the potential due
to the image problem and that due to a charge at the centre.

4
Method of Inversion

The problem of finding the image charge for a sphere indicates that there is a symmetry associated with
finding the potential. The potential expression has a symmetry about the centre of the sphere,
which can be termed as the “centre of inversion” and the potential expression is symmetric if we let
. It may be observed that if is the potential due to a collection of charges
located at ,then the potential due to the charges located at

the inversion points are related by the relation

Consider a charge q located at ( The potential at due to this charge is given by

The potential at due to charges located at is given by

Thus,

5
Example :Conducting sphere in a uniform electric field

To produce the field we put a charge – at and a charge at . The field at an

arbitrary point P is given by

We have, if the charges are far away from the sphere,

so that

so that the situation represents a constant electric field in the z direction with .

The potential due to the pair of charges is – .

6
 r2 r1

Q Q

The potential due to the two image charges (shown in the figure, inside the sphere)

Thus the net potential is given by the sum of the potentials due to the charges and the potential
due to the image charges,

as was obtained by earlier treatment.

Example :A dipole near aconducting sphere

7
The dipole is placed at the position . It has a dipole moment . The corresponding
image charges are located (the direction of the dipole moment being from the negative
charge to the positive charge, the image dipole moment vector is directed as shown in the figure
sine the image charge has opposite sign with respect to real charge.) The image charges are as
follows :

Corresponding to the charge at the image charge is

And, corresponding to the charge at the image charge is

The image approximates a dipole with dipole moment

where,

We have,

where .

The dipole moment of the image is thus calculated as follows :

8
We can rearrange these to write,

The image charges, being of unequal magnitude, there is some excess charge within the sphere. The
excess charge is given by

so that,

Example : A charge q in front of an infinite grounded conducting plane with a hemispherical boss of
radius R directly in front of it.

9
The image charges along with their magnitudes and positions are shown in the figure. Here,
.The pairs and make the potential on the plane zero, while the pairs
and – make the potential on the hemisphere vanish. By uniqueness theorem, these four
are the appropriate charges to satisfy the required boundary conditions.

If we take an arbitrary point , the potential at that point due to the four charge system is given
by

The electric field is obtained by taking the gradient of the above. As we are only interested in
computing the surface charge density, we will only compute the radial component of above and
evaluate it at r=R,

Substituting r=R and combining the first term with the third and the second with the fourth, we get,

10
The charge density can be integrated over the hemisphere to give the total charge on the
hemispherical boss,

Example : Line charge near a conducting cylinder

The potential at an arbitrary point is given by

11
Equating the tangential component of electric field on the surface to zero, we have,

Taking one of the terms to the other side and cross multiplying,

As this is valid for all values of , we have, on equating coefficient of and that of to zero,

The second relation gives, and substituting this in the first relation, we get,

SPECIAL TECHNIQUES-II
Lecture 18: Electromagnetic Theory

Professor D. K. Ghosh, Physics Department, I.I.T., Bombay

Tutorial Assignment

1. A charge q is placed in front of a conducting sphere of radius R which is maintained at a

constant potential . Obtain an expression for the potential at points outside the sphere.
2. A spherical cavity of radius R is scooped out of a block of conductor which is maintained at zero
potential. A point charge is placed inside the cavity at a distance from its centre. Obtain the
image charges, an expression for the potential inside the cavity as also the induced charge on the
cavity wall.
3. A conducting sphere of radius R, containing a charge Q, is kept at a height h above a grounded,
infinite plane. Calculate the amount and location of the image charges.
4. Two identical spheres of radius R each, contain charge Q and – . The spheres are separated by
a distance d. Locate the image charges.

12
Solution to Tutorial Assignment

1. We have seen that when the sphere is maintained at zero potential, we can introduce an image
charge at a distance from the centre of the sphere. If we want the sphere to be
maintained at a constant potential, we imagine an additional image charge at the centre of the
sphere which keeps the potential constant at . Equating this to we take this charge at
the centre to be equal to . The potential at an arbitrary point located at a distance r
from the centre is thus given by

2. Let us put an image charge at a distance along the line joining the centre of the sphere to
the position of the charge q which is taken along the direction . Take a point within the cavity
along a direction . The potential at such an arbitrary point is given by

q’

P
q
O

If the point is chosen on the surface of the cavity, this must give zero value for the potential.
Following identical method as shown in the lecture, the image charge is found to be

and the location of the image charge is . The surface charge density is found to be given by

3. Imagine the charge Q to be at the centre of the sphere. The sphere is at a constant potential. In
order to keep the plane at zero potential an image charge is there at a distance h below

13
 the plane, i.e. at a distance 2h from the centre of the sphere. This would disturb the potential on
the sphere which has to compensated by an image charge at a distance from
the centre of the sphere. As this image charge is located at a distance from the plane, the
image charge due to this is at a distance below the plane, i.e. at a distance
from the centre of the sphere. The image in the sphere is a charge at a

distance from the centre of the sphere. Thus we see that there would be infinite
number of images below the plane and inside the sphere. There seems
to be a pattern for the magnitude and the location of these images. Inside the sphere,

at , at , etc. If we define , it seems to suggest

at . This assertion can be easily proved as follows. If true, the

nth image is at a distance from the plane. The image of this in the plane is at a distance
from the centre of the sphere and has a charge – . The image in the sphere is
at a distance from the centre of the sphere. These are, respectively, the charge
and the distance .

Thus we have

Rearranging,

This is a difference equation which can be solved by assuming a solution of the form, ,

which gives , which has the non-zero solution . It can be

seen that in order that the series converges for all distances , the positive square root is to
be taken. The total charge in the sphere is

14
4. Consider the image of charge +Q on the other sphere. It is a charge – at a distance

from the centre of right hand sphere O’. The distance of this from O is

from O.

Similarly, image of in the sphere to the left is at . The image of in the

sphere to the left is at .

O O’

It can be seen that there are infinite number of images with the images on the left being
at .

SPECIAL TECHNIQUES-II
Lecture 18: Electromagnetic Theory

Professor D. K. Ghosh, Physics Department, I.I.T., Bombay

Self Assessment Quiz

1. Consider a point charge in front of an insulated, charged conducting sphere of radius R. If the
sphere is to contain a charge Q, what is the potential outside the sphere?
2. A conducting sphere of radius R has a charge Q. A charge q is located at a distance 3R from the
centre of the sphere. Calculate the potential at a distance R/2 from the centre of the sphere along
the line joining the centre with the charge q.
3. Two point charges q each are at a distance d from each other. What should be the minimum
radius R of a grounded sphere that can be put with its centre at the mid point of the line joining
the two charges so that the mutual repulsion between the point charges is compensated?
Assume d >> R.
4. Two spheres, each of radius R contain identical charge Q. The spheres are separated by a
negligible distance. Locate all image charges.

15
Solutions to Self Assessment Quiz

1. Initially, assume that the sphere is grounded. If the charge q is at the position with respect
to the centre of the sphere, an image charge located at satisfies the required
boundary conditions. In order that the sphere has a total charge Q, we now disconnect it from
the ground and add amount of charge to the sphere, which will be uniformly
distributed over the surface. The potential at is then given by,

2. The image of q in the sphere is a charge – at from the centre. Since the
sphere is a conductor all parts of it are equipotential. As the overall charge is Q, it can be
looked upon as an image charge at R/3 which cancels the potential due to the charge q
outside and a charge at the centre equal to . The potential of the sphere due to this is
.

3. Each of the point charge gives an image charge within the sphere with charge –

at a distance from the centre towards the charge.

q’ q’
q q
b/2 b/2

16
These two image charges, having opposite sign to that of the original charges, attract each of them.
The distance of these image charges from either of the charges is . The net

attraction due to these charges must cancel the repulsive force . Thus, we have,

Simplifying, we get,

Since b>> R, as a first approximation we can neglect all terms other than the first term on either
sides of this equation, and get, which gives

4. The problem is very similar to Problem 4 of the tutorial assignment with the difference that both
the charges being +Q, the image charges alternate in sign. Taking d=2R, we have, for image on
the left sphere,

etc. The total induced charge is – .

17