M. M.

ALI

COORDINATE SYSTEM

Introduction to three dimensional coordinate systems

Electromagnetics deals with vector fields which are easy to handle when they are resolved
into components along three mutually perpendicular coordinates/directions. Cartesian or
rectangular, Cylindrical, and Spherical coordinate systems are most commonly used
coordinate systems to specify the lengths, directions, projections, angles or components of
a vector accurately. Therefore, before going to main text of electromagnetics, we will
discuss first the coordinate systems in brief.

Cartesian Coordinate System

The Cartesian or rectangular coordinate system is formed by three mutually perpendicular

straight lines. These are called the x, y, and z axes. Any point P(x, y, z) in space can be
defined uniquely by the scalar projections x, y, and z of the position vector r of the point
P on the three axes as shown in Fig. 1.1(a). Here, ax, ay, and az are unit vectors used to
indicate the directions of the components of any vector along the x, y, and z axes,
respectively. A differential volume dv=dxdydz can be obtained by six differential
planes/surfaces at x, x+dx, y, y+dy, z, z+dz as illustrated in Fig. 1.1(b). Each surface is
defined by a unit vector normal to it.
z-axis

y-axis

x-axis

(a) (b)

(c)
Fig. 1.1
 M. M. ALI

Thus the differential surfaces in the positive direction of the unit vectors are, respectively,
dsx=dydz ax, dsy=dzdx ay and dsz=dxdy az as shown in Fig. 1.1(c). The differential length
from P to Q in vector form is obtained from the difference between the position vector of
point Q and that of point P, and is given as dL=dx ax+dy ay +dz az.

Cylindrical Coordinate System

Any point in this system is specified by assigning its coordinates (ρ, , z) as illustrated in
Fig. 1.2 (a). Here ρ is the perpendicular distance of the point from z-axis,  is the angle of
the plane containing the point and the z axis and is measured from the x axis in the anti
clockwise direction and z is the perpendicular distance of the point along z axis from the
xy plane. P may also be obtained from the intersection of ρ=ρ1 plane, =1 plane, and z=z1
plane as shown in Fig. 1.2 (b).

(ρ, , z)

(a) (b)

(c) (d)
Fig. 1.2

A differential volume dv= ρdρddz is formed by the surfaces at ρ, ρ+dρ, , +d, z and
z+dz as illustrated in Fig. 1.2 (c). Thus differential surfaces in the positive direction of the
unit vectors aρ, a and az.are, respectively, dsρ = ρddz aρ, ds= dρdz a and dsz= ρdρd az.
The elemental length from P(ρ, , z) to Q(ρ+dρ, +d, z+dz) in vector form is dL= dρ aρ
+ ρd a +dz az.
 M. M. ALI

Spherical Coordinate System

A point P in space in spherical coordinate system is represented in terms of r,  and  as

shown in Fig. 1.3 (a). Here r is the radial distance from the origin to the point P,  is the
angle that r makes with the positive z axis, and  is the angle of the plane containing the
point and the z axis and is measured from the x axis in the anti clockwise direction.

(a) (b) (c)

Fig. 1.3

Six differential surfaces at r, r+dr, , +d,  and +d formed a differential volume dv=
r2sin drdd as illustrated in Fig. 1.3 (b). The differential surfaces in the positive
direction of the unit vectors ar, a and a.are, respectively, dsr = r2sindd ar, ds= r
sindrd a and ds= rdrd a as shown in Fig. 13 (c). The elemental length from P(r, ,
) to Q(r+dr, +d, +d) in vector form is dL= dr ar + rd a + r sin d a.

Transformation between Coordinate Systems

Sometimes it is easy to handle the problem of electromagnetic field in one coordinate

system but desirable to have an answer in other coordinate system. The following
discussion on transformation between the coordinate systems, therefore, plays an
important role in this subject. The transformation system involves two steps: (i) changing
of independents variables and (ii) changing of vector components.

(a) Transformation between Cartesian & Cylindrical Coordinate System

The coordinates of a point P in both Cartesian and Cylindrical coordinate systems are
shown in Fig. 2.1 (a).
 M. M. ALI

z-axis P(x,y,z)
az P(ρ, ,z)

 a
aρ

y
x ρ y-axis

(a) (b)
x-axis Fig. 2.1

The relation among the independent variables is therefore:

Cartesian to Cylindrical Cylindrical to Cartesian

ρ=(x2 + y2) x= ρ cos 
= tan-1(y/x) y= ρ sin 
z=z z=z

If a vector A=Aρ aρ + A a + Az az is given in cylindrical coordinate system, it can be

expressed in the rectangular coordinate system as A=Ax ax + Ay ay + Az az. The vector
components Ax, Ay and Az can be obtained in terms of the vector components and
independent variables of cylindrical system by dot multiplying the vector A in both
systems by ax, ay, and az, respectively.

(Ax ax + Ay ay + Az az)  ax = (Aρ aρ + A a + Az az)  ax

or Ax= (Aρ aρ + A a + Az az)  ax
From Fig. 2.1(b), we have
aρ=cos  ax+ sin  ay and a= - sin  ax+ cos  ay
Hence, aρ  ax= cos , a  ax= - sin , and az  ax= 0.
 Ax= Aρ cos  - A sin  ............................................................................(2.1)
Ay can be obtained from
(Ax ax + Ay ay + Az az)  ay = (Aρ aρ + A a + Az az)  ay
or Ay= (Aρ aρ + A a + Az az)  ay
Now,
aρ  ay= sin , a  ay= cos , and az  ay= 0.
 Ay= Aρ sin  + A cos  ............................................................................(2.2)
And Az can be obtained from
(Ax ax + Ay ay + Az az)  az = (Aρ aρ + A a + Az az)  az
or Az= (Aρ aρ + A a + Az az)  az
Hence,
aρ  az= 0, a  az= 0, az  az= 1.
 Az= Az .........................................................................................................(2.3)
 M. M. ALI

Similarly, the vector components Aρ, A and Az can be obtained in terms of the vector
components and independent variables of Cartesian system by dot multiplying the vector
A in both systems by aρ, a, and az, respectively.

(Aρ aρ + A a + Az az)  aρ =(Ax ax + Ay ay + Az az)  aρ

or Aρ =(Ax ax + Ay ay + Az az)  aρ
= Ax cos  + Ay sin  = Ax (x/(x2 + y2)1/2) + Ay (y/(x2 + y2)1/2) .............(2.4)
A can be obtained from
(Aρ aρ + A a + Az az)  a =(Ax ax + Ay ay + Az az)  a
or A =(Ax ax + Ay ay + Az az)  a
= -Ax sin  + Ay cos  = -Ax (y/(x2 + y2)1/2) + Ay (x/(x2 + y2)1/2) .............(2.5)
Az can be obtained from
(Aρ aρ + A a + Az az)  az =(Ax ax + Ay ay + Az az)  az
or Az =(Ax ax + Ay ay + Az az)  az = Az ............................................................(2.6)

The result of vector transformation from Cartesian to Cylindrical and vice-versa are
summarized in the following table.
Cartesian to Cylindrical Cylindrical to Cartesian
Aρ= Ax (x/(x2 + y2)1/2) + Ay (y/(x2 + y2)1/2) Ax= Aρ cos  - A sin 
A= -Ax (y/(x2 + y2)1/2) + Ay (x/(x2 + y2)1/2) Ay= Aρ sin  + A cos 
Az = Az Az = Az

(b) Transformation between Cartesian & Spherical Coordinate System

From the projection of r on the z axis and onto the xy-plane as shown in Fig. 2.2, we have

Fig. 2.2
x=r sin cos, y= r sin sin and z=r cos.
The relation among the independent variables is therefore:

Cartesian to Spherical Spherical to Cartesian

r=(x + y2 +z2)
2
x= r sin  cos 
= cos-1(z/(x2 + y2 +z2)) y= r sin  sin 
=tan-1(y/x) z= r cos 
 M. M. ALI

900-

Fig. 2.3
From Fig. 2.3, we have
ar =sin aρ+ cos az=sin(cos  ax+ sin  ay)+ cos az= r/r=(xax+yay+zaz)/(x2+y2+z2)1/2
a = cos aρ - sin az= cos(cos  ax+ sin  ay) - sin az and
a= - sin  ax+ cos  ay

Now if a vector A=Ar ar + A a + A a is given in spherical coordinate system, it can be

expressed in the rectangular coordinate system as A=Ax ax + Ay ay + Az az. The vector
components Ax, Ay and Az can be obtained in terms of the vector components and
independent variables of cylindrical system by dot multiplying the vector A in both
systems by ax, ay, and az, respectively.

(Ax ax + Ay ay + Az az)  ax = (Ar ar + A a + A a)  ax

or Ax= (Ar ar + A a + A a)  ax
Hence,
ar  ax= { sin(cos  ax+ sin  ay)+ cos az }  ax = sin  cos 
a  ax={ cos(cos  ax+ sin  ay) - sin az}  ax=cos  cos  a
 ax= - sin .
 Ax= Ar sin  cos +A cos  cos  - A sin
.........................................(2.7)
Ay can be obtained from
(Ax ax + Ay ay + Az az)  ay = (Ar ar + A a + A a)  ay
or Ay= (Ar ar + A a + A a)  ay
Now,
ar  ay= { sin(cos  ax+ sin  ay)+ cos az }  ay = sin  sin 
a  ay={ cos(cos  ax+ sin  ay) - sin az}  ay =cos  sin  a
 ay= cos .
 Ay= Ar sin  sin +A cos  sin  + A cos
.........................................(2.8)
And Az can be obtained from
(Ax ax + Ay ay + Az az)  az = (Ar ar + A a + A a)  az
or Az= (Ar ar + A a + A a)  az
Now,
ar  az= { sin(cos  ax+ sin  ay)+ cos az }  az = cos 
a  az={ cos(cos  ax+ sin  ay) - sin az}  az = - sin 
a  az= 0.
 M. M. ALI

 Az= Ar cos  - A sin ...........................................................................(2.9)

Similarly the vector components Ar, A and A can be obtained in terms of the vector
components and independent variables of catesian system by dot multiplying the vector A
in both systems by ar, a, and a, respectively.

(Ar ar + A a + A a)  ar =(Ax ax + Ay ay + Az az)  ar

or Ar=(Ax ax + Ay ay + Az az)  ar
= Ax sin  cos  + Ay sin  sin  + Az cos = Ax (x/r)+ Ay (y/r)+ Az(z/r)
= Ax (x/(x2 + y2 +z2)1/2) + Ay (y/( x2 + y2 +z2)1/2) + Az(z/(x2 + y2 +z2)1/2)..........(2.10)
A can be obtained from
(Ar ar + A a + A a)  a =(Ax ax + Ay ay + Az az)  a
or A=(Ax ax + Ay ay + Az az)  a
= Ax cos  cos  + Ay cos  sin  - Az sin 
= Ax (zx/((x2 + y2)1/2 (x2 + y2 +z2)1/2)) + Ay (zy/((x2 + y2)1/2 ( x2 + y2 +z2)1/2)) - Az ((x2
+ y2)1/2/(x2 + y2 +z2)1/2))............................................(2.11)
A can be obtained from
(Ar ar + A a + A a)  a =(Ax ax + Ay ay + Az az)  a
or A=(Ax ax + Ay ay + Az az)  a
= -Ax sin  + Ay cos 
= - Ax (y/(x2 + y2)1/2) + Ay (x/(x2 + y2)1/2) ....................(2.12)

The result of vector transformation from Cartesian to spherical and vice-versa are
summarized in the following table.
Cartesian to Spherical Spherical to Cartesian
Ar= Ax (x/(x2 + y2 +z2)1/2) + Ay (y/( x2 + y2 Ax= Ar sin  cos +A cos  cos  - A
+z2)1/2) + Az(z/(x2 + y2 +z2)1/2) sin 
A= Ax (zx/((x2 + y2)1/2 (x2 + y2 +z2)1/2)) + Ay Ay= Ar sin  sin +A cos  sin  + A
(zy/((x2 + y2)1/2 ( x2 + y2 +z2)1/2)) - Az ((x2 cos 
+ y2)1/2/(x2 + y2 +z2)1/2)) A = A cos  - A sin 
z r
A=- Ax (y/(x2 + y2)1/2) + Ay (x/(x2 + y2)1/2)

Divergence theorem:  A  ds   ( 
A)dv
s

Stocks theorem:  A  dl   (  A)  ds
l s
 M.M.Ali
Page No.:1/60

ELECTROSTATICS

Introduction

In the study of electrostatics, the charge, the source of electric field, is assumed to be
fixed in space and remains constant in time.
The aim of this part of study is to determine the force exerted by the charge(s) on other
charge(s), the electric field intensity, the potential distribution and the energy stored by a
capacitor.
There are numerous practical applications of electrostatic field. The static electric field is
used, for example, to accelerate a charged particle. The use of electrostatic field in the
design of an oscilloscope and an ink-jet printer is widely familiar.

1. Coulomb’s law

In an experiment, French army engineer Colonel Charles Coulomb observed that there
exists a force between two charged bodies. He showed that the electric force between two
point charges is
(a) Directly proportional to the product of the charges,
(b) Inversely proportional to the square of the distance between the charges,
(c) Directed along the line joining the charges,
(d) The magnitude of the force depends on the medium and
(e) Repulsive for charges of identical (like) sign, but attractive for charges of opposite
(unlike) sign.
On the basis of the experimental result, Coulomb formulated the force, known as
Coulomb’s electrostatic force, between two point charges in a boundless free space as
FQ  1 Qq aR Newton (N) (1)
4 0 R 2
where Q and q are two point charges in Coulomb (C), R=|R| is the distance between them
in meter (m) and ε0 is the permittivity of the free space in which the charges are situated.
If FQ is the force exerted on q (a test charge) by Q (source charge), the unit vector aR will
point in the direction from Q to q as shown in Fig. 2(a).

(x', y', z')

aR R=r-r
z Qº q
R ºq (x, y, z) r r º (x, y, z)
Qº(x', y', z') y

(a) x (b)
Fig. 2
Here R=(x- x')ax+(y- y')ay+(z- z')az and
a R  R  (x  x)ax  ( y  y)ay (z  z)az
R
(x  x)2  ( y  y)2  (z  z)2
If r and r are respectively, the position vectors of Q and q as shown in Fig. 2(b), (1) can
be expressed as
 M.M.Ali
Page No.:2/60

FQ  1 Qq R  1 QqR 1 Qq(r  (2)

4 0 R 2 a 4 0 R 3  0
r  r 3

Here r=xax + yay + za4z,)

rr= x'ax + y'ay + z'az and R=r – r=(x- x') ax + (y- y') ay + (z-
r  r
z')azand is the unit vector a inRthe direction of the vector R=r – r.
r  r
Coulomb’s law can be used over the wide range from 10-13 cm to many km. Even it
correctly describes,
1) The electric force that bind the electrons of an atom to its nucleus,
2) The forces that bind atoms together to form molecules, and
3)The forces that bind atoms or molecules together to form solids or liquids.
However, the coulomb’s law fails to explain the stability of the nucleus. The nuclear force
which is larger than the electric force described by the coulomb’s law holds the proton of
the nucleus together.

1.2 Superposition principle in Electrostatic

The force on a test charge q is a linear function of the amount of charge Q of the source.
The principle of superposition can, therefore, be applied in a system consisting of three or
more discrete charges to determine the electric force exerted on any one of the charges by
the others. If Fi be the force exerted on a test charge q by the ith source charge Qi, and aRi
is the unit vector acts in the direction from Qi to q as shown in Fig. 3, then the resultant
force on q due to several point charges Qi (where i=1,2,3….n) is given by:
n n n n
1
F   Fi  1 R Q i q a Ri   Q i qR i   Qi q(r  ri) N (3)
i1 4 2
4 1R i 3 4  r  ri 3
0 i1 i 0 i1 0 i1

Q1 R1
aR2
ºq
aR1
Q2 R2

Fig. 3
Here, in (3) Ri is the vector between the ith source charge and q, r is the position vector
of q on which the force is to be determined and ri is the position vector of the ith source
Qi.
In the above discussion, only the point charges are considered, where all the charges are
assumed to be concentrated at a point having zero volume. However, the charges may be
distributed uniformly along a line, or on a surface or in a volume.
When the charges are spread over a finite volume with Volume charge density v (Cm-3),
then it is called volume charge; where the total charge Q contained within a volume v is
given by:
Q   ρ v dv C , where the integral is performed over the volume v.
v
If the charge is spread thinly over a surface of a sheet, it is said sheet charge. If the
Surface charge density at a given point on the surface of the sheet is s (Cm-2), the total
charge Q on the whole surface S is then given by:
 M.M.Ali
Page No.:3/60

Q   ρ s ds C , where the integral is performed over the surface S.

s

When the charge is distributed along a line with a Line charge density l (Cm-1), it is
called line charge; the total charge Q in a total length L is given by:
Q   ρ l dl C , where the integral is performed along the line L.
l

Direct use of equation (1) or (2) is not possible to determine the force exerted by a line or
surface or volume charge distribution as the source charge is not in the point form.
However, the superposition principle helped us to overcome this problem easily. In case
of line, surface or volume charge distribution, we may divide the entire line, surface or
volume into several numbers of small segments. The charge of the small segment of line
or surface or volume will behave like a point charge. The forces for the individual
segments are then calculated using equation (1) or (2) and then add them as vectors to
determine the resultant. This is explained in the following in details.
For the line charge distribution shown in Fig. 4(a), if we consider a small segment dl then
ρl dl may be assumed as a point charge (ρl is the line charge density in C/m). The force dF
on q due to this small segment of line charge ρl dl C is therefore,

dF  1 qρ l dl(r  r)
4 0 N.
r  r
3

The total force on q due to the whole line charge is then given by

 
F  dF q
4  ρ l (r  r)
dl N. (4)
r  r
l 3

In Fig.
0 l
4(b), if ρs is uniform surface charge density in C/m2, the charge ρsds C on a small
surface ds, can be assumed as a point charge. The force dF on q for this point charge is
then given by
dF  1 qρs ds(r  r)
4 0 .
r  r
3

The total force on q due to the whole surface is therefore given by

 
F  dF q  ρ s (r  r)
dsN. (5)
r  r
s 3
4
In Fig.
0 s
4(c), if ρ v is uniform volume charge density in C/m3, the charge ρvdv C on an
elemental volume dv, can be assumed as a point charge. The force dF for this point
charge on q is then given by
dF  1 qρ v dv(r  r)
4 0 .
r  r
3

The total force on q due to the whole volume is therefore given by

 
F  dF q
4  ρ v (r  r)
dv N. (6)
r  r
v 3

In the0 vabove equations, r is the position vector of the point P where the test charge q is
assumed to be placed and r is the position vector of the segmented element as shown in
Fig. 4.
 M.M.Ali
Page No.:4/60

Fig. 4

1.3 Electric Field and Electric Field Intensity

If a point charge q (test charge) is brought into the vicinity of a fixed charge Q, the
charge q experiences a force. It is then said that Q has a field around it which can exert a
force on another charge q. Thus, an electric field is defined as a region where forces act.
And the electric field intensity/electric field strength/electric field E at some point in the
field is defined as the force on a unit positive charge at that point. To determine E at a
point P, we have to place a very small positive charge q at that point and measure the
force F acting on it. Then
E  lim F . As because q also creates its own electric field and perturb/alter the
initial
q0 q
field, its magnitude must be as small as possible in order to minimize the distortion.
The unit of E is N/C or more usual V/m.

For a single point charge Q

1
E 1 Q Q(r  r)
4 0 R 2 aR or E  4
V/m as obtained from (1) or (2).
0
r  r 3
The electric fields E due to several point charges, line, surface, and volume charge
distribution are, respectively, given by the following eqs. (7), (8), (9), and (10):

n n
1
E   Ei
4
 Qr(r r r )
i
3
i (7)
i1
0 i1 i

 
E  dE 1
4  ρ l (r  r)
dl (8)
r  r
l 3
0 l

 
E  dE 1
4  ρ s (r  r)
ds (9)
r  r
s 3
0 s

 
E  dE 1
4  ρ v (r  r)
dv (10)
r  r
v 3

The electric
0 v
field at a given point around a charge represents graphically by the lines.
These are known as electric field lines. It points outward from the charge if it is positive,
otherwise it is inward as shown in Fig.5. The density of field lines, however, is
proportional to the magnitude of the electric field.
 M.M.Ali
Page No.:5/60

Q+ Q- Q+ >> Q-
Fig. 5
The electric field lines have the following properties
 The tangent to the lines at any point gives the direction of the E-field at that point.
 Lines start on positive charges and finish on negative ones.
 The density of lines gives an indication of the field strength at a given point.

1.4 Electric potential or Voltage

The energy required to move a unit positive charge from one point to another, is defined
as the potential difference between the points.
If a positive test charge q is placed in an electric field E, there will be a force F=qE on
the charge q that acts in the direction of E. Assume that under this force, the charge q
moves a differential distance dl. The incremental energy expended by the electric field, or
simply the amount of work done by E is thus given by dWe= qE.dl where the subscript e
signifies that the work is done by the electric field. However, if an external force Fext =
-F= - qE acts on the charge q, it will move against the E field; the differential work done
by the external force is given as dW= - qE.dl. Assuming the test positive charge q is
moved from point b to point a against E, then by the definition of the potential difference
between the points a and b, Vab, is given by
a
a
dW
(11)
Vab  Va  Vb  b q  b E  dl
The
volts.–ve sign in the above equation signifies that the work has to be done by the external
source against the field in moving the charge q from b to a. To determine the potential
experimentally, the test charge q should be considered sufficiently small so that it does
not perturb the original field distribution.
In eq. (11), dl, in various coordinate systems, is given by,
dl=dx ax+ dy ay+ dz az in Cartesian coordinate system,
dl=dρ aρ+ ρd a+ dz az in Cylindrical coordinate system and
dl=dr ar+ rdθ aθ+ r sin θ d a in Spherical coordinate system.
In moving the test charge q in a closed path, the total work done and hence the potential
difference must be zero. In other words,
 E  dl  0 ( KVL) which on using the Stock’s theorem takes the following
form
  E  0 . It states that the static electric field is irrotational or conservative.

1.4.1 Potential due to a point charge

1
The electric field for a point charge Q is given by E  Q(r  r)
where r= x'ax +
4 0
r  r
3

y'ay + z'az and r=xax + yay + zaz are the position vectors of the location of Q and the point
where the field is to be determined.
 M.M.Ali
Page No.:6/60
Now, r- r=(x- x')ax + (y- y')ay + (z- z')az and | r- r|=[(x- x')2+ (y- y')2+ (z- z')2]1/2. Following
eq. (11) with dl=dx ax+ dy ay+ dz az the potential difference between a and b
is Vab  Va  Vb   a E 
W
dl
qab b

Now E  dl 
Q x - xdx  y - ydy  z - zdz
4 0 x - x   y - y  z - z 
2 2 2

Q a
x- xdx  y - ydy  z - zdz
3/2
(
Q 1
Vab   1 z - z  3/2
4 0 bx - x  2y - y   

0 b a
 2 2


where ra and rb are respectively, the position vectors of points a and b. If point b is
) 4
assumed to be at infinity ie. rb=, then the potential of point a is called the absolute
potential and is given by r  r
Q Q Q
Va  or, more generallyV  . . . r  r (12)
4 0 ra  r 4 0r  r 4 0R
where R is the distance of the observation point from the charge Q. In eq. (12), V is the
absolute or simply potential at a point of position vector r.
Following the above eq. and the superposition theorem, the (absolute) potential at a point,
due to a group of point charges, line charge, surface charge and volume charge are given
by eqs. (13)-(16), respectively:
n

V 1 Qi (13)
 i1 4 0 r  ri
1 ρ l dl (14)
4 0 l r  r
V

1 ρ s ds (15)
4 0 s r  r
V

1 ρ v dv (16)
V 
4 0 v r  r
In the above equations, ri is the position vector of the ith source charge, r is the position
vector of the observation point where the potential is to be determined, and r is the
position vector of the elemental source.
Equation (12) shows that the potential remains unchanged on a circle or a surface of
constant radius. A line (surface) on which the potential is same is known as an
equipotential line (surface). Thus for a point charge, the equipotential surfaces are
concentric surfaces and those for a line charge are concentric cylinders as shown in Fig.
6.

Fig. 6.
 M.M.Ali
Page No.:7/60
1.4.2 Path-independence of potential

Fig. 7
Let us consider few concentric circles around a point charge at the origin as shown in Fig.
7. The magnitude of electric field E at each and every point on each circle is constant. As
the electric force lines/field lines of a point charge is a radial one, work is only done for
movement along the line joining the two charges (one of the charges is the source of
electric field and the other is the test charge used to measure the potential). No work is
done, i.e., dW=0 for any displacement of the test charge along the circular path which is
normal to the field lines, because qE.dl=0. In the above figure, therefore, no work is done
along the arc segments AB, CD, EF and GH. Hence dW for path ABCDEFGH is
dW(BC)+ dW (DE)+ dW (FG) = dW (AH). Hence dW is independent of the path taken in
moving between two positions.

1.5 Electric Field as gradient of Potential

In case there are two points which are separated by an small distance dl, then the work
done by an external force in moving a unit positive charge from one point of potential V
to another point of potential V+dV is
dW=(V+dV)-V= - E  dl  dV= - E  dl. As V is a function of position (say x, y and z),
V V V
 dx  dy  dz  E  dl
x y z

 V
V
( x a x y  V a )
y a z  (dxa  dya  dza )  E  dl
z x y z

or, V  dl  E 
dl
  
where   x a  x a y
 ay in Cartesian
z
z coordinate system,

 
 a ρ  a φ z a in zCylindrical coordinate system and

  a in Spherical coordinate system.
 ra r a  θ rsinθ φ
rθE= - V .
Therefore (17)

1.6 Equipotential surfaces and E-Field lines

For a point charge, the lines of force/field point radially outwards and the equipotential
lines form a series of concentric circles as shown in Fig. 8. It is easy to prove that at all
points the two types of lines (force line and equipotential line) are normal to each other.
 M.M.Ali
Page No.:8/60

Fig. 8

In a direction tangential (along) an equipotential surface/line there can be no change in V.

Therefore there can be no component of E tangential to the surface (as E=-V) and hence
the only component of E must be normal to the surface.
The properties of equipotential surfaces can be summarized as follows:
(i)The electric field lines are perpendicular to the equipotentials and point from higher to
lower potentials.
(ii) By symmetry, the equipotential surfaces produced by a point charge form a family of
concentric spheres, and for constant electric field, a family of planes perpendicular to the
field lines.
(iii)The tangential component of the electric field along the equipotential surface is zero,
otherwise non-vanishing work would be done to move a charge from one point on the surface
to the other.
(iv) No work is required to move a particle along an equipotential surface.
Conceptually, a field line is a directed curve in the electric field such that the forward
drawn tangent at any point on the curve has the direction of electric field E at that point.
Therefore, if dl is a small segment on this curve, then dl=kEE where kE is a constant. By
equating the values of kE for different components of dl and E, one finds the following
differential equations in different coordinate systems:
dx dy dz
  in Cartesian coordinate system,
Ex Ey Ez

dρ ρd dz (18)
Eρ  E in
E z Cylindrical coordinate system and

dr rd rsind
Er  E Ein Spherical coordinate system.
The field lines can then be obtained by integrating the above differential equations.
Similarly, the equation of equipotential line/surface can be obtained by equating the
equation of potential to a constant, i. e, V=kv.

1.7 Electric dipole

Many physical systems are electrically neutral but still produce an electric field and are
affected when placed in an electric field. This arises because the positive and negative
charges of the molecules are physically separated when they are brought to an electric
field. Such molecules are said to be polarized. This leads to the concept of an electric
dipole which means a pair of equal charges of opposite polarities with very small
separation between them.
Two equal point charges Q and –Q when separated by a small distance s, thus results in
an electric dipole or simply dipole. In an ideal dipole s is very small compared to the
 M.M.Ali
Page No.:9/60
distances to any other charges and also to any points where we wish to find the resultant
electric potential or E-field.
az az
sin 
r-d/2 P ar
Q r 1  a

d0  r+d/2 cos 

-Q d/2=d/2 az az= cos  ar -

sin  a p=Qd=Qd az=p az

Let us find potential V and electric field E at point P due to the dipole. We can determine
E first and then V using (11) with radial distance of point b as infinity. The 2nd method is
to determine V first using (13) and then E using (17). The two methods are described in
the following.

(a) Determination of E first and then V

Using (7), E at P is given by
1 2
(r  d / 2)
E
4
 Q (r  r)
i
3
i
 Q ⎢⎡ (r  d / 
4 0 3 ⎤
⎥
0 i 1 2)r  ri ⎢⎣ r  d / 2 r  d / 2 3
Now, r  d / 2
3

⎦

 (r  d / 2) ⎥ (r  d / 2)3 / 2  r 2  r  d  d 2 /

4 ⎡
3 / 2

 r -3 1  r  d ⎤
3 /
 r -3 ⎡1  rdcos ⎤
3/ 2

⎢⎣ 2
r 2 ⎥⎦ ⎢⎣ r2
As the 2nd term in above eq. is small in comparison to 1 for an ideal dipole, using
Binomal⎥⎦
-3 ⎡ 3dcos ⎤⎥
3
3
. Similarly, we have r  d / 2
expansion we have r  d / 2  r ⎢1⎣ 2r ⎦

⎡ 3dcos ⎤
 r -3 1
⎢⎣ 2r .⎥
⎦ (r  d / 2)
Q ⎢⎡ (r  d / 3  ⎥
E 2) ⎤
r  d / 2
3


E  Qr -3 ⎡⎛ ⎥ ⎤
3dcos ⎞ ⎦ ⎛ 3dcos ⎞
4 ⎢⎜1 2r ⎟⎠(r  d / 2)  ⎜⎝1  2r ⎟⎠(r d / 2)⎥ ⎦
0 ⎣⎝
Q Q
 ⎡3dcos r  d⎤ ⎥
⎢ Q
3⎢
⎡3dcos ra r  z ⎤
da  4 r 3
3dcosa r  da z
4 0 r ⎣ r 3
40 r ⎣ r
Q Q
⎥⎦
0

 3dcosa r  d(cosa r  sina θ )
⎦
2dcosa r  dsina θ
4 0 r 3 4 0 r 3
 
Qd p
 2cosa r  sina θ   2cosar  sinθa
4 0 r 3 4 0r 3
 (19)
Now the potential at P, which is r distance away from the origin of coordinate system, is
 M.M.Ali
Page No.:10/60
P
⎞
⎛ p  ⎟  dr a  rda  r sin  da 
P

V   E  dL   
⎜

⎜
⎝ 4 r
2cos a 

sina
⎟0
3 r
θ
r θ φ

If we consider the ⎠path of integration along the radial path only

V    2pcos3 dr  pcos2  p  ar2

r

 4 0 r 40 r 4 0r

(20)
(b) Determination of V first and then E

Using (13), V at P is given by

2 ⎡ ⎤
V 1 Qi Q 1
 
1 0 r  ri 40 ⎢⎢⎣r  d / 2 r  d / 2 ⎥
i 1 4

Now, r  d / 2
1

⎦

 (r  d⎥/ 2)  (r  d / 2)1 / 2  r 2  r  d  d 2 /

4 ⎡  1 / 2

 r -1 1  r  d ⎤
1/ 2
 r -1 ⎡1  dcos ⎤
1/ 2

⎢⎣ r 2 ⎥⎦ ⎢⎣ r
As the 2nd term in above eq. is small in comparison to 1 for an ideal dipole, using
⎥⎦
Binomal
-1 ⎡ dcos ⎤⎥
1
1
. Similarly, we have r  d / 2
expansion we have r  d / 2  r ⎢1⎣ 2r ⎦

⎡ dcos ⎤
 r -1 1 2r .
⎢⎣

⎥⎦ Q ⎡1  1 ⎤ Qr 1 ⎡⎛ dcos ⎞ ⎛ dcos ⎞⎤

V 0⎢ 
4 r  d / 2 r  d / 2
⎥ 
⎥ ⎢
40 ⎣⎝
⎜1 ⎠  ⎜1
⎟ 2r ⎠⎟⎥

⎝
2rQdcos ⎢⎣pcos p  a ⎦ ⎦

40 r2  4 0 r2 40 r 2
r

Using (17) we have

⎛ 
 ⎞ pcos
E  V  -⎜ r ar  rθ aθ  rsinθ a φ⎟ 4 r 2
⎝ ⎠ 0

2pcos r psin 
p
2cosa  sina θ

4 0 r 3 a  4 0ar 3 4 0 r 3
θ r


In contrast to a single point charge the potential falls off as r-2 (not r-1). The product Qd is
defined as the electric dipole moment, symbol p. This is actually a vector p=Qd where d
is defined as pointing from -Q to +Q.
The total field is
p(4 cos2   sin 2  )1 / 2 p(1  3cos 2  )1 /
E  ER2  E2  2 
4 0 r 3 4 0r 3
In contrast to a single point charge E falls off as R-3 for an electric dipole (not r-2). It is
very important to note that the potential on the bisecting plane for which =900 is zero;
and the electric field acts perpendicularly on this surface.
 M.M.Ali
Page No.:11/60
1.7.1 E-field and equipotential lines for an electric dipole

The equation for an equipotential surface for any charge distribution is obtained by
setting the expression of V to equal a constant. Since r and  are only variable in (20), we
have
r  kV cos where kV is a constant.
By plotting r versus  for various values of kV we draw the equipotential lines (dashed
lines) in above figure. On the other hand the electric field lines represent the direction of
E in space. We, therefore get from (19)
p
E 2cosa r  sina θ   k E (2 cos a r  sin  a
4 0 r 3
)
Therefore using (18) we have
dr rd dr 2 cos d dr 2 cos d
     
2 cos  sin  r sin
r sin  ln r 
const1  ln sin   const2  r  C sin 
2 2

where CE is a constant. By plotting r versus  for various values of CE we draw the field
lines (solid lines) in above figure.

Example 1.1 A line charge density ρl(x) is given by ρl(x)=3x2 Cm-1. Calculate the total
charge contained between
L
the points x=0 and x=L.
Solution:
The total charge Q= L 1 (x) dx   3x dx  x L  L3 C
2 3 0

Example 1.2 A surface 0 charge density ρs (x,y) is given by ρs (x,y)=3x2+4y2-xy Cm-2.

Calculate the total charge
0 contained within the area bounded by x=0+a, y=0+a.
Solution:
The total charge Q   a a s (x, y) dxdy  a a (3x  4 y  xy) dxdy
2 2

0 0 0 0
a a
  (x 3  4xy 2  x 2 y / xa
dy   (a3  4ay 2  a 2 y / 2)dy  (a 3 y  4ay 3 / 3  a 2 y 2 / a
x0 0
0 0
2) 4)
 (a 4  4a 4 / 3  a 4 / 4)  25a 4 /12 C
Example 1.3 Charges of +Q, +5Q, -Q and –2Q are placed on x-y plane at the co-
 M.M.Ali
Page No.:12/60
ordinates (-a,+a), (+a,+a), (+a,-a) and (-a,-a) respectively. Calculate the E-field at the
origin.
Solution:
Q y 5Q

r1 r2

r4 r3

- 2Q -Q

Let r1 – r4 are the vectors from the position of Q to the origin of the coordinate system and
so on.
Now E  E1  E 2  E 3  E 4 where
Qr1  Qr3  2Qr4
E1  , E2  , E3 and E 4 
5Qr23
4 4 0r
3
4 0r3
3
4 0 r4
3
0 r1 2

r1  aa x  aa y , r2  aa x  aa y , r3  aa x  aa y and r4  aa x  aa y

3
3
and r1  r2  r 3 r
3 3
4
3
 a2a2  (2a
)
2 3 /2

⎡r
Q
E ⎢ 13  5r23  r33  2r4 ⎤
4 ⎥⎢ 1 r2 r3 r4 3 ⎥⎦
0⎣
r

Q 
aa  aa  5aa  5aa  aa  aa  2aa  2aa 
4 0(2a )2 3 / 2
x y x y x y x
y


Q 
 5aa  9aa V/m 
4 0(2a )
23/2 x y

Example 1.4 Two point charges -q and q/2 are placed at the origin and at (a, 0, 0)
respectively. Determine the position on x-axis where E vanishes.
Solution:

-q q/2 P(x, 0, 0)

(0, 0, 0) (a, 0, 0)
Let P be the position where the electric field E vanishes. Then the field due to –q and q/2
are, respectively, given by:
q q/2
E 1 a x and E2  a x.
4 0x 2 4 0(x  a) 2
1
E  E1 E 2 0  1   0  x 2  2(x  a) 2  x 2  4ax  2a 2 

x2 0 2(x  a) 2

or, x  a(2  2)  3.41a or 0.59a

Among the two, the right answer is x=3.41a, because E will act only in opposite
directions for considered charge configuration if x>a.
 M.M.Ali
Page No.:13/60
Example 1.5 A line charge having a constant charge density ρl Cm-1 extends along the y-
axis from - to +. Calculate the E-field at a distance a along the positive x-axis. Also
calculate the potential V.
Solution:
y
y=+
dl=dy

y r=y ay
r- r= a ax - y ay  r  r  (a 2  y 2 )1/
2

r=a ax x=a x

y= - 

Now the charge in the small segment dl=dy is ρldy C. The field for this small charge
is
dE   l dy(r  r)  E 
3   l (r  r)
 
l (aax  yay )
dy   4 0 (a2 y ) 2 3 / 2 dy
4 0 r  r  4 0 r  r
3


 l aa x 
l yay
  dy   4 (a 2 y ) 2 3 / 2 dy
4 (a2  y2 3)/ 2
 0  0
Assume y=a tan   dy=a sec2 d and (a2+y2)3/2=a3sec3. The limits of the integration
become = -/2 for y= -  and =/2 for y=.
  aa x 
yay ⎤
E  l ⎡⎢  2 dy   (a 2  y 2 )3 / 2 ⎦dy ⎥
4 0 ⎣ (a  y ) 23/2


 ⎡ / 2 a 2 sec2 a x  / 2
a 2 sec2  tana ⎤
l ⎢  a 3sec 3
d    a 3sec 3
y
d ⎥
⎥
4 0 ⎢⎣ / 2  / 2
 / 2

 cos
l 
a
⎡
4 0 ⎣⎢x/ 2 a
d  
 / 2
sin  a y d
a ⎥ 
l
sin ⎦  /2
2 a x cos
/
ay
⎦ 4 a  / 2


  / 2
⎤ /2 0

l
 ax V/m
2 0 a
To calculate V using the above equation of E, we have to express the location of the point
l
where the field has been calculated as variable which lefts E  ax V/m
2 0 x
Therefore the potential of a point, whose perpendicular distance from the line charge is x,
with respect to another point of distance x0 is
x x
 x
 dx   z x
x x y

x0 x0
2 0 x ln x0
2 0x 2 0 x0
l l

 
  l ln x  l ln 0
x 20 0
If we2
consider the potential of the point at x0 as zero, the potential about a line charge is
l
given by V   ln
2 0 x .
 M.M.Ali
Page No.:14/60
Example 1.6 A finite line charge of length L carrying uniform line charge ρl Cm-1 lies
along the z-axis. Determine E and V at a point P on the plane bisecting the line charge.
Solution: z
z=+L/2
dl=dz r-
r=
 a P
r=z az - z
z= - aL/2
z

r=a
r
z= -L/2
r
Now the charge in the small segment dl=dz is ρldz C. The
 2

dE   l dz(r  r)  E  (2l (r  r) dz 2 
3 3 field
L/
for  l (  a ρ  za z )
this small charge is
 z  L / 2 4 0 ( 2 z ) 2 3 / 2
dz
4 0 r  r 1/ 
 4 0 r  r
) L/
2
L/
 l aρ  l za z
 2
dz  
2 dz
 L / 2 4 0 (  z )  L / 2 4 0 (  z )
2 23/2 2 23/2

Assume z= tan   dz= sec2 d and ( 2+z2)3/2= 3

sec3. The limits of the
integration become = -tan-1 (L/2 ) for z= - L/2 and = tan-1 (L/2 ) for z=L/2.
 ⎡ L / 2
2 aρ 2 3/2
dz  L /
2
 2 za z
dz ⎥
⎤
E l ⎢  z )  z 2 )3 / 2
4 0 ⎣L / 2 ( L / 2 (
 ⎤
⎡ tan (L/2)
 2 sec2 a ρ d ⎦
1
tan 1 ( L / 2  )
 2 sec2  tana d ⎥
 
z
l ⎢  3sec 3  3sec 3 ⎥⎦
4 0 ⎢⎣tan 1
(L/2)
tan1 ( L / 2
)
 ⎤
⎡ tan1 ( L / 2  ) cosa ρ tan1 ( L / 2  )
sina z d ⎥
l ⎢  
d    ⎥⎦
4 0 ⎢⎣tan 1
(L/2)
tan1 ( L / 2
)
l
E l ⎡sin tan (L/2) tan1 ( L / 21 ) sin(tan 1 (L /
tan ( La/ 2⎤ 
1 z
ρ cos
tan1 ( L / 2 a) 
2 0  2 ))a
ρ
4 0 ⎢⎣ )
⎥⎦
 l L/ aρ V/m
2
2 0 (L / 2)2   2
The potential of the point P is
V   E  dl   
 l
a  (daL/da  dza )
 2 0  (L /22)2   2 ρ
ρ

φ
r z
 L/
 
2l2 0d (L / 2)2   2
Assume =(L/2) tan, therefore d= (L/2) sec2 d and (L/2)2 + 2=(L/2)2 sec2.
l
l L/2 d  l
2 0  cos ec d  2 ⎡1 ⎛ 1  cos ⎞⎤
⎢ 2ln ⎜⎝ 1  cos⎠
 2  0 (L / 2)2  2
0⎣
⎟⎥ ⎦
 M.M.Ali
Page No.:15/60
 
 ⎡ ⎡ ⎛
⎛ 1  (L / 2) / (L / 2) 2   2 ⎞⎤  l ⎢ ⎜ (L / 2) 2   2  (L / 2) ⎞⎤
V   l
⎢ln ⎜  ln⎜
4 0 ⎢⎣ ⎝ (L / 2)    (L / 2)
2 2
4 0 ⎢⎣ ⎜⎝ 1  (L / 2) / (L / 2) 2   2 ⎟⎥
⎟⎥ ⎠⎦  ⎟⎥ ⎠⎦ 
⎟⎥
⎡ ⎛
 l ⎢ ⎜ (L / 2) 2   2  (L / 2) ⎞⎟⎤⎥
 ln⎜ V
4 0 ⎢⎣ ⎝ (L / 2)    (L / 2)
2 2

⎟
Alternate Solution:
⎠⎦ ⎥
We can solve for V first and then determine E from E= -V.
For the small segment dz, potential dV is
 dz L/2
dz
dV l
4 0  z 2   2
 V  l 
4 0  L / z 2   2
2
Assume z= tan, therefore dz= sec2 d and z2 + 2
=
2
sec2.
  sec d  ln⎢  ln⎢ ⎡ 2
sin⎥⎦⎤ 21 ⎡1 z / z 2   2 ⎤⎥  z   2  z ⎤⎥
 dz
2 2
11sin
2 1 ⎣⎡
ln⎢ z2   2  z
z  2 ⎢⎣1  z /1z   2 ⎥⎦ 2
L/2
⎡  ⎡ (L / 2)2   2 ⎢ (L / 2) ⎥⎦
V   l 1 z  z⎤
2 2
l ln⎢
⎣ ⎥
⎥ 
ln⎢ 4 0 2 ⎢ z 2  2 z
L/
4 0 ⎤(L / 2)2   2  (L / 2)
⎣
2
and ⎣ ⎥⎦ ⎢ ⎥⎦
   l  ⎛ ⎡ (L / 2)2   2  (L / 2)
E  -V  -( a ρ a φ  a z)V   ln ⎟
⎟aρ
  ⎜ ⎢ ⎤(L⎞ / 2)2   2  (L / ⎥
⎥
⎝⎢ ⎦ ⎠
 z 4 0  ⎜

l ⎡ 
ln  
(L / 2) 2   2  (L / 2) 

 2)
ln (L /⎣ 2) 2   2  (L ⎤/
⎢
40 ⎣2)
l

⎥a
L/
 ⎦
ρ
 aρ
2 0  (L /22)2   2
Example 1.7 A semi-infinite line charge having a constant charge density ρl Cm-1 extends
along the y-axis from 0 to +. Calculate the E-field and V at a distance a along the
positive x-axis.
Solution:
y
y=+
dl=dy

r=y ay r- r= a ax - y ay  r  r  (a 2  y 2 )1/ 2

y=0 x
r= a ax
The problem is similar to that given in example 4. In this case, however, the lower limit
of the integration is different, which is =0 as y=0.
 ⎡ / 2 a 2 sec 2 a x  / 2
a 2 sec 2  tan a y
d    d ⎥
 E  4
l
⎢ ⎢ a 3sec 3 ⎤0 a 3 sec3  ⎥⎦
0 ⎣ 0
 M.M.Ali
Page No.:16/60
 / 2
sina y ⎤
 ⎡ / 2 cosa x d  d ⎥
  l
⎢ 
 a
4 0 ⎣ 0 a 0 ⎦


l
4 0 a
 /2
sin  0 a x cos
/2
0
ay
l
4 0a
(a x  a )y V/m


Calculation  of V:
In calculating V using V   E  dl , we have to replace a in equation of E by x to make

the position of P on the x-axis generalized.
V   E  dl   l (a  a )  (dxa  dya  dza )
P P

 4 0 x
r
 y
r a 0
a
 0
 y  
 l y
  
l dx
z
  l ln a
4 0 x    l
4 
dy0 
x   l ln4x 0 
4 0 x  4
Example
0 1.8 A semi-infinite line extending from - to 0 along the z-axis carries a
uniform charge distribution of 100 nC/m. Find the electric field intensity at point P(0, 0,
2). If a charge of 1C is placed at P, calculate the force acting on it.
Solution: Let us consider an elemental charge ldz′ at z= z′. The position vector, r, of the
observation point P and that of the charge position, r′, are given by
r=z az, r′= z′ az. Therefore r-r′=(z-z′) az.
Now using (8), E is given by
0
 dz  
E l  4 0 z
l
4 0   (z  z) z

9 109 100 109 a z  450az V/m

 2
The force on 1C charge at P is F  qE

1106  450a  450a N
z z

Example 1.9 A line charge extends from x=0 to x=L and has a density ρl(x)=ax Cm-1.
Calculate the E-field at a point b (>L) on the x-axis.
Solution:
r- r= (b–x) ax
r=x ax
x
x=0 dx x=L x=b
r=b ax
|r- r|=(b-x). Therefore the field dE at x=b due to the charge ρl(x)dx=axdx is
L
dE  axdx(r  r)  E 
ax(r  r) dx  L ax(b  x)a x3 dx
4 (b  x)
 4 0r  r
3 0
4 0r  r
3
0 0
L
xa x dx
axa x 2 dx  a L
 0 4 0 (b  x) 4 0 0 (b  x) 2
Assume y=b-x. Therefore x=b-y and dx=-dy. The limits of the integration become y= b
for x= 0 and y=b-L for x=L. Then
 M.M.Ali
Page No.:17/60
b L
a (b  y)a x a ⎡b L 1 ⎤
 
b L
E dy  a
b L
( y  b)a x dy  b
4 0 b y2 4 0 b y 2
4 0 ⎣⎢ b y dy  b y 2 dy⎦⎥ax
a ⎡
b L
a bL L
⎢ln y  b ⎥ax 
b L
 ⎢ln  a xV / m
4 b
yb ⎥ ⎤ b (b  L)
0 ⎢⎣ ⎦ 0 ⎣
⎡ ⎥
Example 1.10 A line charge has density  and extends along the x-axis from –a
to +a. Find the electric potential4 ⎦
at a point r on the x-axis (r>a). Use your result to find
the E-field at r.
Solution:
r=x ax r- r= (r-x) ax

x
x= -a x=0 dx x=a x=r
r=r ax
|r- r|=(r-x). Therefore the field V at x=r due to the charge λdx is
a
dV  dx a
V   dx  
a
 dx  
ln(r  x)
40 r  r 4 r  r 4 (r  x) 0 a
a 0
4 a 0

 ra
 ln⎛ ⎝⎜ ⎞ Volts
4 0 ra
Following
⎟ ⎠the above equation of potential on the x-axis at any position r where (x=r>a) is
given as function of x as:
 xa
V  ln⎛ ⎝⎜ ⎞ Volts
4 0 xa
⎟ ⎠
  
 E on the x - axis where x  r  a, E  V  ( xa x ay
 ya )Vz z

 ⎛ 1 1
 ⎜  x
⎝
4 0 x  a x 
⎞aa ⎠ ⎟ 1 1
 ⎛ ⎞⎟ax
Therefore E at r on the x - axis is E  0 ⎝⎜
4 r  a r 

Example 1.11 a ⎠charges, one having charge distribution l C/m
Two infinite line
and the other -l C/m, are separated by a distance a. Determine the electric field intensity
E at a point P, which is located on the line joining the charges and which is a distance 
from the positive line charge, for<a and >a.
Solution:

l -l l -l

a a
x a x a
 P  P
 M.M.Ali
Page No.:18/60
Following the result of example 5, E for an infinitely long line charge l at a perpendicular
distance r from the charge is (l/20) a, and that for -l is (-l/20 (a-
))(- a) and (-l/20 (-a)) a, respectively, for <a and >a.
l l l
Therefore E at P for r<a is E  (  )aρ  a ρ and for r>a
2
a 0  20(a   ) 2 0r(a   )
l l l a
E(  )a ρ  aρ
2 0  2 (  a) 0 2  (   a) 0
Example 1.12 Determine electric field due to a ring charge of radius b at a point
on its axis. Assume the ring is uniformly charged with density l C/m.
Solution:
z

r=z az r- r= z az - b a
dl=bd

r=b a
The field E at P due to charge of small segment dl is

2
dE   l dl(r  r)  E  2  l b(r  r)
3 l b(za z  baρ )
d  0 4 0 (z 2 b ) 2 3 / 2 d
4 0 r  r 0 4 0 r  r
3

2
 b ⎡2 zd bd ⎤
 a 
4⎢ 0 ⎣ 0 (z  b )
 l 2 23/2 z a
23/2 ρ⎥
0 (z  b )
2 ⎦
The above equation shows that the field due to a segment has two components, one acts
along the z-axis and the other acts in the -direction. If we consider a similar segment just
opposite side of the present segment, the r-components of these two segments will cancel
each other. Therefore, when the entire ring is under consideration, the field only acts in the
z-direction. This leaves
 b 2 Qz
 zd 2 bz
E  4 0 0 (z 2  b 2) 3 / 2 a z  4 0 (z  b 2 ) 3 / 2 a z  4 0 (z2  b 2)3 / 2 a z
l

Here Q is the total charge on the ring. l

Example 1.13A charged ring of radius a carries a uniform charge distribution. Determine
the potential and E at any point on the axis of the ring.
Solution: Let us consider an elemental charge ldl=lad′ at P′(a, ′, 0). The position
vector, r, of the observation point P and that of the charge position, r′, are given by
r=z az, r′= a a. Therefore r-r′=(z az -a a) and |r-r′|=(z2+a2)1/2.
 M.M.Ali
Page No.:19/60
Using (14), we obtain
2
1   ad l
  l a
V (z) 
4 0 0 (z  a )
2 2 1/ 2
2 0 (z  a )2 1/ 2
2

l a ⎤ l az
Now E  -V  - dz 2  (z 2 a ) 2 1 / 2 ⎥a z  2 (z 2  a 2 )3 / 2 a z
d ⎡⎢
⎣ 0 ⎦ 0

Example 1.14 A plane circular sheet of radius b has a uniform charge density ρs
Cm-2. Calculate the E-field and V at a point P which is a perpendicular distance a from
the centre of the sheet.

Solution:
z
P

r
-

r

=
r= a
Let us consider a small surface ds=dd a on the circular sheet placed on the z=0 xy-
plane. The electric field at P due to the charge on the surface ds is then given by

dE   s ds(r  r)  E 
3  sa(r  r)  b  2
⎡  s ( a ρ  aa z ) dd
z ds  ⎤  ⎢ 2
 a 2⎥) 3 / 2
4 0 r  r s 4 0 r  r
3
 0  0 ⎣ 4  0 ( 
However, due to symmetry the radial component ⎦ of the field is zero. Therefore,
r b ⎡  s(aa z) 0
2 ⎤-  b r b
aa z
E  ⎢   
 ⎡
 d
s aa z ⎤
d ⎥ ⎢ 2 3 / 2 d
s

⎢4 (r  a )
2 23/2
⎥ 2  (  
2
a ) ⎥  2 (  
2
a )23/2

r 0 ⎣
 ⎦   0 ⎣ 0  ⎦ 0 r 0

Assume =a tan  a

 d=a sec  d and ( +a ) =a sec3. The limits of the
2 2 2 3/2 3
0

integration become = 0 for = 0 and =tan-1 b/a for =b. Then

 tan b / a
1 r=a az
  tan b / a a 3 tan sec2 a
1 
E  s2
0

 0 a 3sec 3
z
d   s 2cos  0
az
 0

⎡

s
2 0
 

cos(tan 1 b / a)  cos 0 az   s ⎡⎢ a

⎤

2 0 a 2  b 2 1⎦⎥a z 2
s

a
⎥
⎤ V/m
a
0⎣ a  b ⎦z
2 2

⎢ 1
⎣
This result indicates that the electric field is everywhere normal to the surface of the sheet.
Also one may observe that the field E at a perpendicular distance a on the –ve z- axis,
from the surface of the sheet, is same in magnitude to the previous one but it directs in the
–ve z direction. Therefore, the field, of a sheet charge having +ve surface charge density,
directs outwards. However, it directs toward the sheet if the surface charge density is –ve.
Now using E with the position of the point on the z-axis as variable, we can find V from:
 M.M.Ali
Page No.:20/60

s ⎡ z ⎤
V   E  dl  
⎢1
⎥a z (daρ  daφ  dzaz )
 2 ⎣ 

2 2
0 z b ⎦ 2 2
a
 ⎡ dz    a

z ⎤ z zb  a2  b2 

s s

 
⎥ 
  ⎢1  0 ⎣ z  b ⎦ 2 0
2 2 s
2  0
a
Example 1.15 Show that the answer to example 6 tends to ρs/20 in the limit of
2
the sheet having an infinite radius. Also show that, the field due to two infinitely extended
sheet charge placed face to face only remains in the space between them.
Solution: In the final equation of E of question, if b is infinite, then the equation of the
field E reduces to E= (ρs/20) az.
Let us now consider two sheet charges, one having a surface charge density of + ρs C/m2
and the other having a surface charge density of - ρs C/m2, face to face as shown in the
following figure.

+ ρs

- ρs

The dashed arrows show the direction of the field due to the sheet with + ρs and the solid
arrows show the same for the sheet with - ρs. This particular system of sheet charge results
zero field outside the sheets; the field only remains in between the sheets which is qual to
(ρs/0) and will be directed from the sheet with + ρs to the sheet with - ρs.

Example 1.16 A thin annular disc of inner radius a and outer radius b carries a uniform
surface charge density s. Determine E at any point on the z-axis when z 0.

Solution: Let us consider an elemental charge sds′=s′d′d′ at (′, ′, 0). The position
vector, r, of the observation point P and that of the charge position, r′, are given by
r=z az, r′= ′ a. Therefore r-r′=(z az -′ a) and |r-r′|=(z2+′2)1/2.
Using (9), we obtain
 s b 2  d d 
(za
E
4   (z    )
2 2 3/2 z   a )
ρ
0a 0

It is easy to prove that there would be no aρ component of E because of the

symmetry of the charge distribution with respect to P.
 M.M.Ali
Page No.:21/60

 b 2
 d
 zs ⎡ 1 1 ⎤

Therefore E s   2 2
2 0 ⎣(a  z )
2 1/ 2
(b 2 z ) 2 1 / 2 ⎥⎦az
4 0 a 0 (z d  )
1 ⎤
 z⎡
For b  , E  2s ⎢ (a 2 z )2 1 / 2 ⎥a z
0 ⎣
Example 1.17 ⎦ Charge is uniformly distributed over a strip which is infinite in
length and which has a width W with a surface distribution s C/m2. Determine the electric
field at a point P which is on a line perpendicular to the strip and which is a distance D
away from the center of the strip.
Solution: z
r- r= D az - x ax x P
r=D az
W
dx y

x r= x ax sdx
Sheet is assumed as the sum of some line charges of thickness dx having a charge
distribution of sdx. The field for each line is determined first and then use the
superposition theorem to determine the required field.
Following the result of example 5, the field for an infinitely long line charge at P is give
by
dE 
l r  r s dxr  r
2 
2 r  r 2 r  r
2

0 0

 s(Da z xa )xdx  (Da z xa )xdx

W/2

 dE E  s
W / 2 2 0 {D  x )
2 2 2
 2 0 r  r
Let x=D tan 
dx=D sec2 d and D2+x2=D2 sec2.
 (Da  xa )dx s
1
W/2 tan (W 2 D )
W
E  s
 z 2 x2
  s   a  tan  a  d   (tan 1 )a z
 W / 2 2  0 {D  x )
z x
 0 2D
0 tan (W 2 D )
1

Example 1.182  Charge is uniformly distributed over a WxW square sheet with a
surface distribution s C/m2. Determine V and E at a point P on the axis perpendicular to
the sheet and through its center.
Solution:

z
r=z az xP r - r= z az - x ax - y ay
R
W r=x ax+y ay
dx y

W dy
sdxdy

x
Sheet is assumed as the sum of some elemental surface of area dxdy and having a charge
distribution of sdxdy. The field for each elemental surface is determined first and then
 M.M.Ali
Page No.:22/60
use the superposition theorem to determine the required field. The position of the
elemental surface is assumed at (x, y, 0) and the point P is taken at (0, 0, z).
The field dE at P due to the elemental surface charge is
W/2W/2
( x a x  y ay  z a z )dxdy
dE   s dxdy (r  r3 )   
E s
(x 2 y 2 z ) 2 3 / 2
4 0 r  r 4 0 W / 2 W / 2
Let us integrate the above equation first w.r.t x and then w.r.t y.
Assume x=(y2+ z2)1/2 tan 
dx=(y2+ z2)1/2 sec2 d and (x2 + y2 + z2)1/2=(y2+ z2)1/2 sec .

(xa x )dx
Now  (xa x2  ya y  za z )dx
  ( yay  zaz )dx
(x  y 2  z 2 )3 / 2  (x  y 2  z 2 )3 / 2  (x 2  y 2  z 2 )3 /
2

  2sin2d1/ 2a ( yay  za z ) cos d cos ( yay  zaz ) sin

2

(y  z )
x
 (y  z 2 )
2  a x (y 2  z 2 )
2 2 1/2
(y  z )
( yay  zaz )x
 (y 2  z 2 ) 1/2

a 
(y 2  z 2 )1 / 2 (x 2  y 2  z 2 )1 / 2 x (y 2  z 2 )(x 2  y 2  z 2 )1 /
2 ax ( yay  zaz )x
 2 a x
(x  y  z )
2 2 1/2
(y  z 2 )(x 2  y 2  z 2 )1 /
2

2
W/2
(xa x ya  za )dxz ( yay  zaz )x xW / 2

 ⎤
y
 ⎡ ax  ⎥
⎢
W / 2 (x 2  y 2  z 2 )3 / 2 ⎣(x  y  z )
2 2 2 1/2
(y 2  z 2 )(x 2  y 2  z 2 )1/ 2xW / 2
 ⎦
(y 2  z 2 )[(W / 2)2  y 2  z 2 ]1 /
2 2( yay  zaz )(W / 2)
 W/2 2( yay  zaz )(W / 2) dy
Therefore, E   s  2  z 2)[(W / 2) 2 y 2z ] 2 1 / 2
4 0 W / 2 (y
2(W / 2) W / 2 ( yay  zaz ) dy

 s 4
0
 2
 z )[(W / 2) 2 y 2z ] 2 1 / 2
2

W / 2 (y
( yay  zaz ) dy
Now
 (y 2
 z 2 )[(W / 2)2  y 2  z 2 ]1/
2  yay dy zaz dy
 
 (y 2  z 2 )[(W / 2)2  y 2  z 2 ]1 / 2  (y 2  z 2 )[(W / 2)2  y 2  z 2 ]1
To/ 2solve the first integration, let us assume t  [(W / 2)2  y 2  z 2 ]1 /
2
.Therefore
dt  y[(W / 2)2 ya y 2 dyz 2 ]1 / 2 dy and y 2  z 2  ta2  (W / 2)2
y y
 
dt
 (y 2  z2 )[(W / 2)2  y2  z 2 ]1 / 2  t 2  (W /
If we2 put t  (W/2) sin , dt  (W/2) cos d
2)
then  ya y dy  ay
  (W / 2) cos d ay  sec d a y
dt
 (y 2
 z 2 )[(W / 2) 2  y 2  z 2 ]1 / t 2
 (W / 2)  (W / 2) 2
cos2   (W / 2)
2 2
 M.M.Ali
Page No.:23/60
t
1
1  sin  y  1
 1
ln ln  (W / 2) a y
2(W / 2) 1  sin  2(W / t
1
2)
a  (W / 2)
1 ln (W / 2)  [(W / 2)  y  z 2 ]1 a
2 2
 /2
2(W / 2) (W / 2)  [(W / 2) 2  y 2  z 2 ]1 / 2 y
To solve the second integration, let us assume y  [(W / 2)2  z 2 ]1/ 2 tan.Therefore
dy  [(W / 2)2  z 2 ]1/ 2 sec2  d , y 2  z 2  [(W / 2)2 tan 2   z 2 sec2  ] and
[(W / 2)2  y 2  z 2 ]1 / 2  [(W / 2)2  z 2 ]1/ 2 sec
za z dy zaz [(W / 2)2  z2 ]1 / 2 sec 2 d
 
(y 2 z )[(W
2
/ 2) 2y  z2 ] 2 1/ 2 [(W / 2)2  z 2 ]1 / 2 sec[(W / 2)2 tan 2   z 2 sec2
 zaz d zaz ]cos d
 
sin 2   [(W / 2)2 sin 2   z
[(W / 2)2 z2
 sec ] cos
Again if we put (W/2)sin=z tan,] then (W/2)cos d=z sec2 d
2


za dy
z

2 2 21/2 zaz cos d z 2a zsec 2 d
  (y 2 z )[(W
2
/ 2)  y  z ] [(W / 2)2 sin 22  z 2]   (W / 2)z sec 2

a z d
   az  W/2
tan1 ⎡⎢ z sin ⎤a z
 (W / 2) (W 1 / 2) (W / ⎣
1 (W / 2)
 2) 1 ⎡
⎢ ⎥ ⎤
2 1 /⎦2 ⎥a z
(W / 2) tan z[(W / y
2)2
 y 
2
z ]
⎣
W/2 ⎦ ( yay  zaz ) dy
Now, 
W / 2 (y  z )[(W / 2)  y  z ]
2 2 2 2 2 1/ 2

⎡ 1 y
⎢ ln (W / 2)  [(W / 2)2  y2  z2 1 /]2 ⎥ ⎤ ay 
2 2 2 1/2

⎣2(W / 2) W(W/ 2 / 2)  [(W / 2)  y  z ]

⎦ y W / 2
⎡ y W / 2
⎢ 1 tan 1 ⎢⎡ (W / 2)2 ⎤⎤ az
(W
⎣ y / ⎣ z[(W / 2)2
 y  z ] 21/2

2) ⎥ ⎥ ⎦⎦ y W / 2
2 tan ⎡ (W / 2) 2
⎤
 1 ⎢
2 1/ 2 ⎥ z
a
⎣ z[(W / 2)  (W / 2)  z ]
2 2
(W / 2) ⎦
2(W / 2) W/2
( yay  zaz ) dy
Therefore, E s 4 0
 2
 z )[(W / 2)2  y 2  z 2 ]1/ 2
2

W / 2 (y
2
⎤
tan 1 ⎢⎡ ⎥a z
 0 ⎣ z[(W / 2)  (W / 2)  z ]
2 2 2 1/ 2

The solution⎦of E, however, can be made simple by using the knowledge of symmetry
about the z-axis. In this case we could solve only
s W / 2 W / 2 za z dxdy
E
4   (x  y 2  z 2 )3 / 2
2
0 W / 2 W / 2
 M.M.Ali
Page No.:24/60
Example 1.19 A spherical conducting shell of radius a carrying a charge Q
distributed uniformly over its surface. Determine the potential i) at an external and ii) at
an internal points.
Solution:
A
C d
a 
L r

R P
O

B D
Sphere is assumed as the sum of some rings of width ad. The potential for each ring is
determined first and then use the superposition theorem to determine the required
potential.
Potential at P due to the charged annular ring of radius CL=a sin and width AC= ad is
given by
 2CL  AC  2a sin  ad  2a 2 sind
dV  s 4 r  s 4 0 r
s
4 0 r
0

But r2=(CL)2+(PL)2=(asin)2+(R-OL)2=(asin)2+(R-a cos)2=R2+a2-2Ra cos

2rdr=2Rasin d  asin d/r=dr/R
 s 2adr Qdr
 dV  
4 0 R 8 0 a R Ra
Qdr Q Q

Ra
Therefore when P is an external point V   r  and
Ra
80 a R 8 0a R 4 0 R
Ra
a R
Qdr Q Q
when it is an internal point V  
a
 rR 
aR
80 a R 8 0a R 4 0a
a
R
Example 1.20 Show that the E-field produced by a spherical conducting shell of
radius r and carrying a charge Q is identical to that of a point charge Q placed at the centre
of the sphere for >r and is zero for <r.
Solution:
We can use E  V and the result of example 13 to determine E.

 r  a φ )V  ar which is
Therefore for >r E  V  ( ra  a θ
0 r2
rθ at the centre ofQthe sphere.
identical to that of a point charge Q placed
 rsinθ 4
  a φ)V  0 .
And for <r E  V  ( ra  a r θ
rsinθ
rθ
Example 1.21 Determine E at a point due to a volume charge density.
Assume the volume charge density is v C/m3.

Solution:
 M.M.Ali
Page No.:25/60

ar
Say OP=r
OQ=s sin P
QP=s cos s dv=r2sin d rdd
Q 

O
r0
 r
Volume charge distribution

O
If we consider the similar incremental volume dv at the similar position on the opposite
side of the line OP, it will show that the resultant field will act only in the OP direction or
in the radial direction ar. The magnitude of the field along the line OP due to the
incremental volume dv is only cos  to that along OP.Now the magnitude of the field at P
due to vdv along ar is given by
 dv  r2 sindrdd
dE 4vcos2  v
4 0s 2 cos .
0s

r2  (r  s cos ) 2  (s sin  ) 2  r 2  s 2  2rs

cos   r 2  s 2  r2
cos
2rs
Similarly, sr 22  (r 2  )  (rsin  )  r  r  2rrcos
r 2rscos
2 2 2 2
sds
 sin d 
cos  2rr r

 rdsdrd ⎛ r 2  r2 ⎟⎞
 dE  v8r 20 ⎜⎜1
s2 ⎠

2 v 2  rr⎟ ⎛ r 2  r2 ⎞⎟r dsdr d
r
⎝
0

E ⎜1 s2 
8r  0  0 r0 sr r ⎝

2 0r  ⎛ 2 2 ⎟  ⎠ v r
r r
0


3
 v r  r  ⎞ r dsdr  v(4 / 3)r
⎜1 4r 2 0
4r  0 r0
s2 ⎠ 2
4r2 dr 
4r  0 r0 sr r ⎝

Q Q
 E ar
4r 0
2
4 0 r 2
Example 1.22 The components of electric intensity in a certain region are given
Ky Kx
by E x  and E y . Prove that the lines of force are the
(x 2  y 2 )1 / 2 (x 2  y 2 ) 1/ 2
family of concentric circles.
Solution: The ratio of Ex and Ey is given by
E x dx y  xdx  -ydy  x2 y2
E y  dy  x  C1  -  C2  x2  y2 
2
2
C
 M.M.Ali
Page No.:26/60
where C, C1 and C2 are constants. The relation x2 + y2=C is an equation of circle with
radius C and having center at the origin of the coordinate system. Thus we may have a
family of concentric circles with C as variable.
Example 1.23 An electron and a proton separated by a distance of 10-11 meter are
symmetrically arranged along the z-axis with z=0 as its bisecting plane. Determine the
potential and E at P(3, 4, 12).
Solution: The position vector, r, of the observation point P(3, 4, 12) is given by
r=3 ax +4 ay +12 az and |r|=13m. The dipole moment p=1.6x10-19x10-11 az=1.6x10-30az.
Now the potential at P can be obtained from (20) as
pa
V  4 rr 2  7.865 10 V
23

0
And E is given by (19) as
p
E 2cosa r  sina θ
 4 0 r 3

Now with the help of the coordinate system we have

cos  z  12 and sin  x  y  5

2 2

r 13 r
13
E  5.04 10 24ar  5a θ V/m
25

1.8 Electric Flux Density, D

The force acting on a test charge will move it along a certain path. This path is called the
electric flux line/field line. The flux lines have no real existence but they are a useful
concept to represent, visualize, and describe the electric field. The electric flux lines have
the following properties:
a) It must be independent of the medium,
b) Its magnitude solely depends upon the charge from which it originates,
c)If a point charge is enclosed in an imaginary sphere of radius r, the electric flux must
pass perpendicularly and uniformly through the surface of the sphere, and
d)The electric flux density D, the electric flux per unit area, is then inversely proportional
to r2.
The electric field E also satisfies the above properties except that its magnitude depends
upon the permittivity of the medium. Therefore, D and E are related as D=εE. For air or
free space ε=ε0.

1.9 Gauss Law

Gauss’ law states that “The electric flux coming out from any closed surface is equal to the
total charge in the volume enclosed by that surface”.
ds

D

Assume an incremental surface dS over which flux density D is constant. If D makes an

angle θ with the normal to the surface, the number of flux lines d crossing the surface ds
 M.M.Ali
Page No.:27/60
is D ds cos θ= D·ds. Integrating over the closed surface we have the mathematical
formulation of Gauss law:
(21)
s D  ds  Q or s E  ds  Q/
To
0 . prove the Gauss law we may consider a point charge Q in free space at the origin of
the coordinate system and an imaginary sphere of radius r around it. Then,
D= εE= 1 Q 2a and ds=r2 sin θ dθ d ar
4 r
r

Therefore,
 2
Q
ψ  s D  ds  0 0 sindd 
4
Q .Q is expressed in
If
terms of the volume
charge density as Q  
v dv , using the
v

divergence theorem we have from (21),   D  v    E  v /  0

In summary, Gauss’s law provides a convenient tool for evaluating electric field.
However, its application is limited only to systems that possess certain symmetry, namely,
systems with cylindrical, planar and spherical symmetry. The use of the above equation
(21) to determine E or D is easy if we are able to choose a suitable Gaussian surface (any
closed surface; generally cylindrical surface is considered for line charge, rectangular or
cylindrical surface is considered for surface charge and spherical surface is considered for
volume charge distribution) around the charge that satisfies the following conditions:
1) D or E is everywhere either normal or tangential to the closed surface, so that
D·ds becomes either Dds or zero, respectively.
2) D has the same value at all points on the surface where D is normal.

1.10 Gauss Law in Point form

Let P(x,y,z) be a point where the flux density is D(x,y,z) and it increases in the positive
direction of the coordinate axes.
 D  ds   D  ds   D  ds   D  ds   D  ds   D  ds   D  ds
s front face back face left face right face top face bottom face

Now only the x component of D contributes to the 1st and 2nd integrations. Similarly the
y component of D contributes to the 3rd and 4th integration, and the z component of D
contributes to the 5th and 6th integrations.
On the front face abcd, the x component of flux density is
 M.M.Ali
Page No.:28/60

x x 
D x(x  2 , y, z)  D (x,
x y, z)  2 x
D x
x, y,
z
and on the back face efgh, the x component of flux density is
x x 
D x(x  2 , y, z)  D (x,
x y, z)  2 xD x
x, y ,
. z
Thus the flux leaving the volume dv=xyz through the surface abcd is

 D  ds  D (x, y, z)yz  xyz

front face
x
2

D
x x x, y , z

and the flux entering the volume dv through the surface efgh is

 D  ds  D (x, y, z)yz  xyz

back face
x
2

D
x x x, y , z

On the left face abef, the y component of flux density is

y y 
D y(x, y  2 , z)  D (x,
y y, z)  D
2 y
y
x, y ,
z
and on the right face cdgh, the x component of flux density is
y y 
D y (x, y  2 , z)  D y (x, y, z)  2 y D y x, y, z
.
Thus the flux entering the volume dv through the surface abef is

 D  ds  D (x, y, z)xz  xyz

y
2

D
y
y
left face x, y,
and
z the flux leaving the volume dv through the surface cdgh is

 D  ds  D (x, y, z)xz  xyz

right face
y
2

D
y y x, y , z

On the top face bche, the z component of flux density is

z z 
D z(x, y, z  2 )  D (x,
z y, z)  2 zD z x, y , z
and on the bottom face afgd, the z component of flux density is
z z 
D z (x, y, z  2 )  D z(x, y, z)  2 z D z x, y, z
.
Thus the flux entering the volume dv through the surface afgd is

 D  ds  D (x, y, z)xy  xyz

bottom face
z
2

D
z z x, y, z

and the flux leaving the volume dv through the surface bche is

 D  ds  D (x, y, z)xy  xyz

top face
z
2

D
z z x, y , z

Therefore, the net flux leaving the volume dv in the x direction is

 D  ds   D  ds  {D (x, y, z)yz  xyz

x
2

x D
x
front face back face x, y,
}
z
xyz  
{D x(x, y, z)yz  2 D
x x x, y, z }  xyz xD x
x, y,z
 M.M.Ali
Page No.:29/60
Similarly, the net flux leaving the volume dv in the y direction is

 D  ds   D  ds  {D
right face left face
y (x, y, z)xz  2
xyz 
yD y x, y, z
}
xyz  
{D y(x, y, z)xz  2 D y x, y, z}  xyz yD y x, y, z
y
and the net flux leaving the volume dv in the z direction is

 D  ds   D  ds  {D (x, y, z)xy  xyz

top face bottom face
z
2

z D z x, y, z

}
xyz  
{D z(x, y, z)xy  2 D
z
z }  xyz zD z x, y,z
x, y,z

Therefore net flux leaving the closed surface is

 D  ds  (x D   Dy   D z)xyz  (  D x D

x y z x 
 y
D )dv
y
z
z
s

 D  ds 
Dx

D  Dz )
or, s
 ( x  y
dv y z
In case of volume dv tends to zero, following the divergence theorem, the right hand side
of the above equation becomes

 D  ds
lim
dv0
s
dv D

Again from Gauss law

 D  ds  Q   E  ds  (22)
Q/ 0 (23)
s s v
  D  v    E 
0

where v is the volume charge density. Equation (23) is the Gauss law in point form, one
of the Maxwell’s equations. In the charge free region   E  0
From equation (22)
 D  ds  Q    v dv   (  D)dv
s v
v (24)
 D  ds   (  D)dv
s v
Equation (24) is known as divergence theorem.
Equation (23) shows that the static electric field is not solenoidal. It is noted that a
divergenceless field is called a solenoidal field whose field lines always close upon
themselves. In the source free region, the electric field however becomes solenoidal.

1.11 Use of Gauss's Law to find the E-field

(a) Spherical but non-point charge

 M.M.Ali
Page No.:30/60

A conducting sphere of radius a carries a charge Q on its surface. By symmetry the

resultant E-field must by spherically symmetric and hence can only have a radial
component. Take as a Gaussian surface (dotted) a sphere of radius r (>a) placed
concentric to the conductor. By symmetry the field E must be constant at all points on the
dotted surface and also normal to it. Flux through this surface
= field x surface area
= E.4r2
Total charge enclosed by surface is Q  E.4r2=Q/0
hence E=Q/(40r2)  V= Q/(40r) and zero at 
For r<a (within conductor) charge contained within the Gaussian surface is zero (all
charge is on surface)  E=0. Hence potential within sphere must be constant and equal to
surface potential = Q/(40a) Same result will be obtained if the sphere is hollow.

(b) Infinite plane of charge (surface charge)

An infinite (or very large) sheet carries a uniform charge density s. By symmetry the
resultant E-field must have a direction normal to the plane and must have the same size at
all points a common distance from the plane. Take as a Gaussian surface a cylinder of
cross-sectional area A and height 2h.
Flux is only non-zero through ends of the cylinder. If field at cylinder ends is E then total
flux is 2EA.
As the charge enclosed is area x charge density = As, therefore, from Gauss's law
2EA=As /0  E=s /20
(c) Infinitely long line charge

E dS3
E E

dS1 dS2

L
 M.M.Ali
Page No.:31/60
Assume an infinitely long line charge with uniformly distributed charge density of l C/m.
As seen earlier the resultant E-field must have a direction normal to the line and have the
same value at all points having a common distance from the line.
As the closed Gaussian surface, we choose a right cylinder of radius r and length L which
is coaxial with the line charge. Now from Gauss law

 D  ds   L
l   D  dS 1   D  dS 2   D  dS 3 
l L ⎛ 2 L
LHS
s  0 ⎜⎞  E  dS1 S  E  dS2 S E  dS3 S  0  E  dS3     E rddz  2  0 LE 
⎜⎟ ⎟
1 2 3
0
0 0
⎝ S1 S 2 S 3 ⎠ S 3

l l
Now, 20 LEr  l L  E  E aρ
2 0 2 0r

1.12 Conductors in Electric Fields

In conductors, electrons are free to move. If the conductor is placed into an external
electric field E, a force F=-eE acts on each free electron.
The free electrons in the conductor will move in a direction opposite to the direction of the
electric field and soon pile up on the surface on one side of the conductor. The surface on
the other side, therefore, will be depleted of electrons and have a net positive charge.
These separated positive and negative charges on opposite sides of the conductor induce
their own electric field, which opposes the external field inside the conductor. The
resultant electric field inside the conductor can be found from the superposition of the
externally applied electric field and the induced electric field. When static equilibrium is
reached, the net electric field inside the conductor is exactly zero as shown in the
following figure. The zero fields inside a conductor imply that the charge density inside
the conductor is also zero as obtained by Gauss’s law.

Even in the absence of external electric field, an electric field will be set up in the
conductor due to the free electrons that exerts force on the electrons and making them
move away from one another. The movement will continue until all the charges reach the
conductor surface and redistribute themselves in such a way that both the charges and the
electric field inside the conductor vanish (v=0, E=0). Therefore, a solid conductor
carries all its excess charge on the surface. On the other hand, the potential difference
a
between any two points in a conductor Vab  Va  Vb   E  dl  0 since E=0 inside the
b
conductor. This requires all points inside the conductor at the same potential.
As E=0 inside a conductor, the flux through Gaussian surface G is zero; and hence net
charge contained within G in Fig. (a) is also zero. A cavity in a conductor, completely
surrounded by conducting material as shown in Fig. (b), also is free of electric fields
 M.M.Ali
Page No.:32/60
unless the hollow region contains a net charge. A hollow conductor, therefore, shields its
interior from any outside electric fields; this is why we are safest inside our car during a
thunderstorm. It must also carry any excess charge on its outer surface to satisfy the
Gauss’s law for the Gaussian surface G. A hollow conductor has a charge on its inner
surface that is equal in magnitude and opposite in sign to any charge that may be enclosed
within the hollow region. This can be proved by considering a Gaussian surface that lies
within the conductor and that encircles the hollow region containing a charge as shown in
Fig. (c).

The presence of the charge on the conductor surface, however, sets an electric field
normally outward to the surface. This will be explained under the section “Boundary
Conditions” in details.

1.13 Dielectrics in Electric Fields

Dielectrics differ from conductors in that they have no free charges that can move through
the material under the influence of an electric field. Although the electrons are unable to
move far in a dielectric, a slight displacement of the positive and negative charges in
opposite directions (the positive charges move in the direction of the field and the
negative charges move in the opposite direction) happens when an electric field E is
applied. A dielectric, in which this charge displacement has taken place, is said to be
polarized. Each molecule of a polarized dielectric therefore acts as an electric dipole of
moment p which points in the same direction as the applied field. And the field produced
by these induced dipoles opposes the applied field. The total electric field within the
dielectric is thus less than the externally applied electric field.
The number of dipole moments per unit volume in a polarized dielectric is defined as the
polarization P which is proportional to the total electric field Em in the dielectric (Note
that the total electric field in the dielectric is different from the applied field.) and is given
by P=0e Em where e is a dimensionless constant known as the electric susceptibility of
the dielectric.
The effect of polarization is to produce accumulations of bound charge bv within the
dielectric and bs on the surface. Here, bv    P and bs  P  an (an is the unit
vector which is normally outward from the relevant surface) are, respectively, volume and
surface charge density. In the case of homogeneous and isotropic dielectrics where the
polarization of each molecule is to the same extent, the volume distribution of the bound
charges bv becomes zero as the positive and negative charges of dipoles neutralize each
other. The charges induced due to polarization thus left only on the dielectric surface. The
effect of polarized dielectric on the electric field may, therefore, be realized equivalently
by removing the dielectric material with the corresponding induced dipoles in free space.
Suppose a dielectric slab is placed between the conducting plates of a parallel plate
capacitor. The upper and the lower plates of the capacitor are assumed to be charged with
+fs and -fs C/m2, respectively, as shown the following figure. The dielectric will be
 M.M.Ali
Page No.:33/60
polarized and bound charges will accumulate on the surfaces of the slabs, say bs C/m2. We
could now remove the dielectric conceptually, and modify the charge density of the
capacitor plates to +(fs-bs) and -(fs-bs) with free space in between them.
In absence of dielectric, there will an electric field Evac=fs/0 an in the space between the
charged plates as explained in example 1.9; in presence of dielectric material, the field
becomes Em=(fs-bs)/0 an. Now from the definition of dielectric constant/relative
permittivity, r=(field in free space)/(field in dielectric material)= Evac/Em, we have
E m  Evac  fs  bs  fs  fs  fs
bs     0  fs 
r  0   r0  0  
r0
 0  fs  r0 bs

or Dvac   0 E m  P  Dvac  Dm   0 E m  P   0 (1   e )E m   0 r E m 
Em
because D is independent of media or material and r=(1+e).

fs

bs

bs

fs

If the field in which the dielectric is placed exceeds a certain limit, the atomic structure of
the dielectric breaks down and the electrons will come out of the atoms. These electrons
take part in avalanche breakdown mechanism of the material. The material will then
become conducting. The maximum electric field that a dielectric can withstand without
breakdown is the dielectric strength of the material.

1.14 Boundary Conditions for Electrostatic Field

Electromagnetic problem often involve media with different physical properties and
require the knowledge of the field quantities at the interface between two media (e.g.
vacuum and a dielectric, two different dielectrics etc). For instance, we may wish to
determine how E and D vectors change in crossing an interface between two media. Let
us consider an interface between two general media shown in the following figure.

t

s

h
 M.M.Ali
Page No.:34/60
For the E-field take the rectangular path as the closed loop. If its height h is made
infinitesimally small then the only contribution to the line integral is along the top and
bottom edges of length dL

 E  dl  E 1  (dL)  E 2  (dL)  E1t dL  E 2t dL  0

L

leads to E1t=E2t  D1t/1=D2t/2

(25) where Et is the tangential component of E.
Hence E1t=E2t; the tangential component of E is continuous across any interface.
In order to find a relation between the normal components of the fields at the interface, let
us construct a small cylinder with its top face in medium 1 and bottom face in medium
2. The faces have an area dS, and the height of the cylinder dt is vanishingly small.
Appling
D  dsGauss’s
 (D  law
a toDthe
 acylinder, we have

S
1 n1 )dS  (D
2 n2  D )dS  lim  (tdS )
n1 n2
t 0
v

or D n1  Dn2  lim v t   s   1 E n1   2 E n2   s
h0
(26)
where an1 and an2 are the unit vectors perpendicularly outwards to the top and bottom
surfaces of the cylinder, respectively.
In a special case where medium 2 is a perfect conductor and medium 1 is a dielectric,
electric field in the conductor E2t and hence E1t in eq.(25) become zero. From eq.(26) we
have Dn1=1En1=s as the electric field in a conductor is zero. It therefore shows that the
electric field at the interface between perfect conductor and insulator is normal to the
surface.
In the case where both media (1 and 2) are perfect dielectric, s in eq.(26) becomes zero as
no free charge can exist in the dielectric.

1.15 Calculation of Capacitance

A capacitor is formed by placing an dielectric material in between two conductors and its
capacitance is defined as the ratio of the charge on one of the conductors to the potential
difference between them i.e.
C=Q/V Farad. (27)
If once the voltage difference between the conductors is calculated and the charge on one
of the conductor is known, capacitance for any arrangement of the conductors can easily
be obtained using eq. (27).
The capacitance can also calculate from the stored electrostatic energy of the system. It is
obtained from the work done in moving a small charge dQ against a potential difference V
and is given by dWe=VdQ=QdQ/C where V=Q/C.
Q 1
Therefore W e   Q dQ  Q  CV 2
2
(28)
0
C 2C 2

Now We  2 CV 2  2QV  2  v Vdv  2(

  D)Vdv
1 1 1 1

By vector identity we have VD=DV+V(D)

 e 1
2 (  D)Vdv  2 [(  VD)  (D  V )]dv 2  (  VD)dv  2 (D  V )]dv
1 1 1
W
Using divergence theorem we have
 M.M.Ali
Page No.:35/60

We  1  VDds  1 (D
2 2   V )]dv
As D varies inversely as square of distance (1/r2), potential varies as (1/r) and the surface
area increases as r2, it follows that the first term in We is a function of (1/r). In order to
consider all the fields in calculating We, we have to consider a very big Gaussian surface
with r, the first term in We thus becomes zero.
1 1 1 1
 We    (D  V )]dv   (D  E)dv   E2 dv   D dv
2
(29)
 2 2 2 2
Therefore, C can also be obtained by using eq. (28) and (29).
Example 1.24 An infinitely long cylinder with radius a has a uniform charge
distribution s C/m2 on its surface. Determine E and V at a perpendicular distance  from
the center of the cylinder.
Solution: For >a
S3
S2

a
S1 Gaussian surface of cylindrical shape with radius .

Assume the Gaussian cylinder is of unit length. It is known that the electric field from a
cylindrical charge distribution acts in the radial direction . Therefore, using Gauss law,
we have

 D  ds  Q   2a s   D  dS 1   D  dS 2   D  dS 3   s
2⎛a
s 0 ⎜⎞  E  dS1   E  dS2  E
S  dS  Sl 1 2 S3
⎟⎜
⎝ S1 S2
3
S3
⎟
⎠ L   S0  E  dS 3  l
3

2 1 L
  0 0 0 E
 ddz  l L  2 0 E   l L

 E   sa
  sa
   E   aρ
Now for <a, the charge enclosed by the Gaussian cylinder is zero. It suggests that E is
also zero for <a.
Example 1.25 Calculate E and V as a function of radial distance for a
spherical
cloud of radius a and having a uniformly charge density v.
Solution: For r>a
Assume a Gaussian spherical surface of radius r around the charge cloud. For the spherical
charge distribution, the field E acts in the radial direction ar; now from Gauss law
 D  ds  Q  Q   v dv  a 2 v r sin2  drdd  4 a3
s v r 0  0  v 3
0
2 
 D  ds    E  dS   d d  4 0r E
2 2
and E rr sin r
 s 0 s 0  0 
4 a 3 0  a3  a3
 40 r 2E v 3
 Er 
r v
3 0 r 2  E  3v r 2 ar
0
 M.M.Ali
Page No.:36/60
For ra
 Q   v dv   r 2 v r sin drdd 
2 4 r 3
s D  ds  v r 0  0  v 3
0
Q 2 
 D  ds    E  dS   d d  4 0r E
2 2
and E rr sin r
 s 0 s 0  0 
2 4
 v r3 0 r r
 40 r E r  Er  v  E  v3a r
3 3 0 0
Calculation of V: 
For ra
r
V   E  dl  v a 3 a  (dra  rda  r sinda )
r

2 r r θ φ


V   3 0 r

va 3 r
 v a3
r
 va 3
V   3 0 r 3 0 r 
3 0
r a2
at r=a, V  3v 0 and for ra
r a
a2 r r

V   E  dl   E  dl a E  dl 3 0v   E 

a
dl r rφ  a θ2 r
r
a 2
r
r
dr
3 0 a 3 0 3 0 a 3 v0

 a2  r2  a2  r2  a2
r

V 3 0  6 a  3v 0  ( 6v  0  6v  0 )  v6(3a

 r )
v v 2 2

Example
0 1.26 Charge is
0 uniformly distributed within a spherical region of radius a. An

isolated conducting spherical shell with inner radius b and outer radius c is placed
concentrically as shown in the figure. Determine E everywhere in the region.

Solution: For 0<r<a, the charge enclosed by a spherical Gaussian surface is given by
Q=  v dv  r  2 v r sindrdd  r 
2 4
3
v 0  0 
3 v
0
Now from Gauss’s law
 2
 D  ds  Q    D  r sindd a  Q
2

r
s  0 
0
r r
4 r D  3 r 
2 r 4
 D 3   E 
3 v r ar v r v  E 
0
3 0 vr
3
For a r <b, therefore the charge enclosed by a spherical Gaussiansurface is given by
 M.M.Ali
Page No.:37/60
Q=  v dv  a  2  r sindrdd  a 
v
2 4
3
v r 0  0 
3 v
0
Now from Gauss’s law
 2
 D  ds  Q    D  r sin dd a
2

Q r
s 4  0  0
v  E a3  a
4 r Dr  a  v  Dr  a a Er
2 3 3

 3 v 0 3r 2 0
3
For b r <c, electric field must be 3r
2
zero because of the
2

3r presence of the conducting material v r

within this region. A charge –(4a3v/3) must be induced on the surface at r=b to make the
field within the conductor zero. The surface at r=c must also acquire a charge (4a3v/3) to
make the region charge neutral.
For r >c, the charge enclosed by a spherical Gaussian surface is given by
Q=  v dv  a  2 v r sindrdd  a 
2 4
3
v r 0  0 
3 v
0
Now from Gauss’s law
 2
 D  ds  Q    D  r sin dd a
2

Q r
s 4 3 0  0 E a3  a
4 r 2 Dr  a v  D  a 
3
a
3 3  E r  3r 2 0
3r 2 0
r v v
3r 2
v r by v=0(1-r /a )
2 2
Example 1.27 A spherical volume charge distribution is given
for ra and v=0 C/m3 for r>a. Find E as a function of radial distance r and also show that
the maximum value of E is at r=0.745 a.
Solution: For r>a
Assume a Gaussian spherical surface of radius r around the charge distribution. For the
spherical charge distribution, the field E acts in the radial direction ar; now from Gauss
law
 Q   v dv  a 2  0 (1  2 )r sin drdd 
r2 2
8 a 3
s D  ds  v r 0  0  a 0 15
0
Q 2  E r r sin dd  4 r E 8 a 3
Now  D  ds  0  E  dS  0  
2
2 0 15
0 0
 s s 0 r

2 8 a 3 2 a 3 2 a 3
 40 r E r 0 15 E 
 150
 E 15
r ar 2
r 0 r

2
For ra
 D  ds  Q  Q   v dv  R 2  0 (1  2
2
)r sin drdd  4 (
3
 )
r 2 r r
s v r 0  0  a 0 5 3 5a 2
0
2 
 D  ds    E  dS   d d  4 0r E
2 2
and E rr sin r
 s 0 s 0  0 
0
 E r  0( r r 0 r r3 r
3

 0 3 5a )  E   0 ( 3  5a )a
2 2
 M.M.Ali
Page No.:38/60

To determine the position of maximum E, let us first differentiate |E| w.r.t r and then
equates the result with 0; we have
0 1 3r )  0  9r 2  5a 2  r  5a  0.745a
2 0 ( 3 5a
2
Example 1.28 A spherical volume of radius a has a volume charge density v=kr
where r is the radial distance and k=constant. Find E and V in the region 0r.
Solution:
The field for the given charge distribution is radial, i.e., in the direction of ar. We have to
consider a spherical Gaussian surface.
For r>a
 Q   v dv  a    2krr
 sin  drdd  ka
2
4
s D  ds  v r 0  0 
0
Q 2 
 D  ds    E  dS   dd  4  0r 2E
2
and E Rr sin R
 s 0 s 0  0 
0 ka 4 ka 4
 40 r 2E r  ka 4  Er  E ar
4 0 r 2 4 0 r 2
For ra
 D  ds  Q  Q   v dv      krr 2sin drdd  kr 4
r  2

s v r 0  0 
0
2 
 D  ds    E  dS   d d  4 0r E
2 2
and E rr sin r
 s 0 s 0  0 
0 kr 2 kr 2
 40 r 2E r  4
kr  Er   E  ar
40 4 0
Calculation of V:
For ra
r r
V   E  dl  V ka 4
 4 r 2 a r (dra r rd a θ r sinda )φ
  0

r
 ka 4
ka  ka
4 4
V    dr 
4r r 2 4 r  4 0 r
 0 0

ka3
at r=a, V  and for ra
4 0
r a r r
ka3
V   E  dl   E  dl   E  dl   E
  a 40  a

V
dl  ka  kr ka3 kr 2
3 r 2 r

dr
4 a r  (drar  rdaθ  r sindaφ )  
0 a 0 4 0
a 4 0
R
ka4 kr 3
ka kr ka ka kr 3 3 3 3
V 3  3 (  ) 
4 0 12 0 a 4 0 12 0 12 0 3 0 12
Example
0 1.29 A point charge Q is at the center of a spherical conducting shell of
an inner radius ri and an outer radius ro. Calculate E and V as functions of the radial
distance r.
 M.M.Ali
Page No.:39/60
Solution:
ri ro

Calculation of E:
We know that the field E will act in the radial direction ar. Now, for r<ri, the Gaussian
surface shown by the dashed line encloses a total charge Q, therefore from Gauss law we
have
2 
 D  ds  Q   D  ds   E  dS   d d  4 0r E
2 2
E rr sin r
s  s 
0 s 0  0 
Q 0 Q
 40 r2 Er  Q  E r  E ar
4 0 r 2
4 0r 2
In the range ri  r  ro, the electric field E will be zero as a conducting media exists there.
To become E is zero, -ve Q charge will induce on the inner surface and +ve Q charge will
induce on the upper surface of the conducting shell. Hence no charge will be enclosed by
any Gaussian surface shown by the dashed line in this range of r.
For rro, the Gaussian surface will again enclose a charge Q, therefore the electric field in
this case is also given by
Q
E ar for rro.
4 0 r 2
Calculation of V:
For rro
r r
V   E  dl  V Q

 4 r 0
2
a r (dra r rd a θ r sinda )φ
 r
r
 Q Q Q
V    dr 
4 r 2
4 0 r 

40r
 0
Q
at r=ro , V  . As the electric field is zero in the range ri  r  r0, no work done
4 0 ro
will occur in bringing a charge from r= r0 to r= ri; the potential at ri is equal to that at ro.
Hence the potential at r= ri is given by
Q
V .
4 0 ro
And for rri
 M.M.Ali
Page No.:40/60
r ro rio r

V   E  dl    E  dl   E  dl   E 
dl
r r
V  
 Q Q r r  Q Q
       d a )    dr
o i
a (dra rd a r sin
4 0 ro r 4 0 r 4 0 ro 4 r 2
2 r r θ φ
i ri 0
r
Q Q Q Q Q Q
V  1 1 1
4 0 ro 4 0 r r 4 0 ro 4 0 r 4 0 ri 4 0 ( r0 r )
  (  ) 
i

ri 1.30 A metallic sphere of radius 10 cm has a surface charge density 10 nC/m2.

Example
Calculate the electric energy stored in the system.
Solution: The field and potential for r10 cm is obtained from Gauss’s law as
 r sin dd 
   2D r    s r sin
 2dd
2
2
s D  ds 
 0  0  0  0
Q 0.01 0.01 r
0.01 r
0.01 0.01
Dr  s  E  2 s ra  V    a  dl     sdr  s
   
2 s r 2
r 
2 r r r r
0  0  0 0
Now the potential for r10 cm is constant and is equal to that at r=10 cm: V=113 Volts.
Energy We is given by We=QV/2, where Q is given by
Q=  2 s r sindd  4r   1.257 10 C . Therefore We=71.02 nJ
2 2 9
s
 0 
0

Example 1.31 Consider two concentric spheres. The inner sphere has radius a,
and a charge Q is distributed uniformly over its outer surface. The outer sphere has radius
b, and a charge -Q is distributed uniformly over its inner surface. Determine the potential
difference between the two spheres and the capacitance of the structure.
Solution:

a b

The fields E for this system in various regions are:

E0 for r  b
Q
 ar for a  r  b
4 0r 2
E is also zero for r<a. Therefore the center and the outer surface of the inner sphere have
the same potential. Potential difference only exists in between the spheres which is
a a a
Q Q
V   E  dl    a  (dra  rd  a  r sin  d  a )    dr
4 r 2 r
4 r 2
r θ φ
b b 0 b 0
a
Q
V  Q  Q Q(b  a)
4 0 r b
4 0 a 4 0 b  4 0
ab
 M.M.Ali
Page No.:41/60

Q 4 ab
Now, C  V (b0  F
a)
Example 1.32 A capacitor is constructed in the form shown in the following
figure. The capacitor consists of two large parallel plates 3 m apart, with a dielectric slab 1
m thick located midway between the plates. The electric field in region I is E=(s/0)ax
where s is the surface charge density on one plate. The surface area of each plate is A.
Determine the capacitance.

+ I II III
+
+ 0  -
+ -
+ 0
x -
0 1 3- 2
Solution: -
The electric flux density in all regions is D=sax. The voltage difference between the
plates is
0 1 2 0
V03  - E  dl   E  dl -  E  dl -  E 
3 3 1 2
dl
 V03    s a x  dxa x-    
2 1 0
s
a x  dxa x- s
ax
 dxa x
 2
3 0 1 0

s  s  s   s ( 2  0 )
 
   0
 0 0

A 0
Now C  Q  A s 
V V
(2   0 )
03 03
Example 1.33 A system consists of two coaxial cylinders of radii a and b (b>a).
The inner cylinder is solid and has a charge + per unit length spread uniformly
throughout its volume. The outer cylinder has negligible thickness and carries a charge -
per unit length. Use Gauss’s law to find the E-field in the regions (i) <a, (ii) a<<b and
(iii) >b where  is the radial distance from the axis of the cylinders. Sketch the form of
the dependence of the E-field on .
Solution:
The length of the coaxial cylinder is assumed to be large so that the E-field is radially
outwards, i.e. E acts in the direction a and constant over a cylindrical Gaussian surface.
Change  per unit length is uniformly distributed within inner cylinder so there will be a
non-zero field for <a. Charge enclosed within a cylinder of radius  (<a) is (2)/(a2).
This gives
1 2
2 and  D  ds   0   E  ddz 
s D  ds  Q  Q  
a 2 2E  
 2

 s 0 0
 2 0 E      E 2 0a 2

a 2

 M.M.Ali
Page No.:42/60
For a<<b all of charge on inner cylinder is enclosed but no charge on outer cylinder. In
this case
1 2

 D  ds  Q  Q   and  D  ds   0   E  ddz  2 0 E  

s s 00

 2 0 E      E  
2

 cancel so E=0.
For >b charges on two cylinders
|E|

Example 1.34 A positive point charge Q is at the center of a spherical dielectric

shell of an inner radius ri and an outer radius ro. The dielectric constant of the shell is r.
Determine E, V, D and P as functions of radial position.
Solution:

ri ro

=0r

=0

Let us consider a spherical Gaussian surface of radius rro as shown by the dashed line. In
this region E can be found using Gauss’s law as
2 
          d d  4 0r E2 r
2
D ds Q now D ds E dS E rr sin
s  s 0 s  0  0 
Q Q 0
E r  E ar
4 0 r 2
4 0r 2

and V at any point in this region is given by

r r
V   E  dl  V Q
 4 r 2 a rr (dra  rd a θ r sinda )φ
 
r
0
r
 Q Q  Q
V    dr 
4 r 2
4 r  4 0 r
 0 0
Therefore, D=0E= Q/(4r2) ar and P=0 as it is free space.
 M.M.Ali
Page No.:43/60
In the region ri r ro(with in the dielectric) the electric field Em can be calculated from the
fact that D is equal in both region. Therefore, D=Dm=0r Em= Q/(4r2) ar and hence Em=
Q/(40rr2) ar. V at any point in this region is equal to V at r=ro (as obtained in the
previous section and is Q/(40ro)) plus the pd from ro to that point as
r ro r

V   E  dl    E  dl   E 
dl
r r
V   Q   Q Q Q
 4
ro
ar  (drar  rda θ r sin d a )φ 
4 r 2
4 r   dr
0 o
rr
ro 4r 2 0 o r o

Q Q Q Q Q 1 1 1
V  Q  (  ) [(1  )  ]
4 0 ro 4r or 4 0 4r 4ro 4 0  r r0  r
Now rfrom
o Dm=0Em+P, we have P=r Dm-0Em= Dm-0Dm/0r=Dm(1-1/r)= (1-1/r)
[Q/(4r2)] ar.
In the region rri, E can be found from the Gaussian surface of the sphere of radius r<ri as
2 
s D  ds   now  D  ds   E  dS   d d  4 0r E2
2
E rr sin r
Q  s 0 s  0  0 
Q Q 0
E r  E ar
4 0 r 2
4 0r 2

and V at any point in this region is equal to V at r=ri which is V= Q/(40)[(1-

1/r)1/ro+1/(rri)] as obtained from the previous section plus the pd from ri to that point as
r
obtained from V    E  dl . Therefore the potential at a point where rri,
r ri
r
V Q [(1  1 ) 1  1 ]   E  dl  Q [(1  1 ) 1  1 ]
 0
r 0 r i ri
0 r 0 r i
r 4
Q  r  r 4Q 1 1r 1
r
Q
 4 r 2
r a  (dra  rd a  θr sin da ) φ
4 0 [(1  )  ]  dr
r r 0  r ri 4 0 r 2
r r
ri 0 r i

r 
Q Q Q Q
V 1 1 1  1 1 1 1 1
4 0 [(1  )r r0 ]r ri 40 rr 4 0 [(1  )r r0 ]r ri 4 0( r )
i

Q ri
V 1 1 1 1
 4 0 [(1  )(r r0 )r
i
]r
In this region, D=0E= Q/(4r2) ar and P=0 as it is free space.
Example 1.35 The following figure gives a cross-sectional view of a co-axial
cable. The inner cylindrical conductor of radius a is charged of l C/m and the outer
cylinder of radius b is charged of -l C/m. Determine the capacitance between the
cylinders. Assume the dielectric between the cylinders has a permittivity .
Solution:

a b
 M.M.Ali
Page No.:44/60
The field E outside the outer conductor is zero as the total charge enclosed by the
Gaussian surface considered around the outer conductor is zero. In the region arb, the
field for unit length of the inner conductor is given by Gauss law as
1  2

 D  ds  Q  Q  l and  D  ds    E  dS     E  ddz  2 E 

s s s z 0  0

l l
E  E aρ
2  2 
The pd between the conductors is
r b a
V   E  dl   E  dl   E 
dl a
 l a
b
a
l l
V  0   a ρ (da ρ d a φ dza )z 0  b 2  d   2 ln  2
l
ln
b
b 2  b
a

 2
C  Vl  F/m
b
ln a
Example 1.36 Solve the above problem considering that the space between two
cylinders is filled up with two different dielectrics (one is of permittivity 1 and attached
to the inner cylinder with a thickness t1 and the other is of permittivity 2).
Solution:
The field E outside the outer conductor is zero as the total charge enclosed by the
Gaussian surface considered around the outer conductor is zero. In the region a(a+t1)
and (a+t1)b, the fields for unit length of the inner conductor
are l l
E a ρ and E  aρ, respectively.
21  2 2
Now the pd between the conductors is
at a  dl⎤
⎤ ⎡a l  
a at ρ
⎡a
2 2 
l
⎣at ⎥⎦
⎣at 21 
b b b ρ
⎦
a
 at
⎡  ⎤ l at
⎤
⎡a
l
d   
 2ln 
 ⎢ 
⎣at 1 
2  2 d ⎥ ⎦ ⎢ ln 
l

b 2
l
⎢⎣ at 2 b
⎥

2 ⎦
⎡ a  1 at⎤  at 
 ⎢ l at
l
ln b ⎥ 2l ln a  2l ln a 
⎣ 2 1 2 ⎦ 1 2
ln b
2 t
l
 ⎡ at b ⎤
212 ⎢  ln a   ln
1
⎣ at
2 
Therefore the capacitance per⎥⎦unit length of the cable C  l  21 2
Vab
⎡ at b ⎤
⎢⎣ ln a   ln
1
at
Example 1.37 Two wires having charges of l C/m and
2 -l C/m are spaced D
⎥⎦
meters apart. The radius of each wire is  meter (<<D). Calculate the capacitance
between the wires.
Solution: Consider the wires are along y-axis on the x-y plane and assume the wire with
l charge is at x=0 and the other at x=D.
 M.M.Ali
Page No.:45/60
y

(0,0) x x=D x

Assuming the wires are very long, the field at any generalized point P in between the
conductors, x meter away from the origin of the coordinate system using Gauss’s law can
l l
be obtained as E  [  ]ax.
2 0 x 2 0(D  x)
Now determine pd between the conductors from
 
V  E  dl    [  l l ]a  (dxa  dya  dza )
D  D 

2 0 x 2 0 (D 
x
x

 x) y
r
  z
 Dx


 0l x 2l  0 (D
  D [2 ]dx l [ln 2x
 ln(D
0  x)] 2l
ln 0 x D 
Dr
. .
x) D
 l
ln 0 
l  0
C  F/m
V D
ln

Example 1.38 A single core lead sheathed cable has a conductor diameter of 1 cm
and the two layers of different insulating materials each 1.1 cm thick. The relative
permittivities are 3.3 (inner) and 2.5 (outer). Calculate the potential gradient at the
conductor surface (maximum field in the inner insulator) when the potential difference
between the conductor and the lead-sheath is 65 kV.
Solution: Assume that the line charge on the inner conductor is l C/m. In the dielectrics,
maximum fields occur at the respective inner radii. Therefore, the maxm fields Emax1 and
Emax2 in inner and outer insulators are given by
l l
E max1  aρ and E max  aρ .
2  3.3  o  0.005 2 2  2.5  o 

However, if V1 and V2 are PDs across the insulators,  then they are given as
0.016

0.005 0.005
l a ρ l
V1    E1  dl     da ρ  ln(0.016 /
 0.016  0.016 2  3.3   o 2  3.3  o 0.005)
 0.00582E max1  
0.016 0.016  daρ   ln(0.027 /
l a ρ 0.016)

 E 2  dl   0.027 2  2.5   o
V2    0.027 2  2.5  
l

 0.0084E max2  o

Now at the interface of two insulators

 M.M.Ali
Page No.:46/60
l l  3.3 
E max 2  aρ aρ
2  2.5  o  0.016 2  2.5 0.005
 o  0.016  3.3 
.
0.005
Emax1  3.3 

0.005 2.5 
Example0.0161.39 Determine the maximum working voltage of a single core lead-
sheathed
V cable with conductor diameter 1 cm and sheath 5 cm inner diameter. Two
1+V2=(0.00582+0
dielectrics are used:
.0084x0.4125)E max1=
Inner layer:
65000  maximum working stress 72 kV/cm, relative permittivity r=4.
Outer layer: maximum
Emax1=7000538 V/m working stress 60 kV/cm, relative permittivity r=2.5.
70kV/cmAssume that the line charge on the inner conductor is l C/m and the thickness
Solution:
of the inner dielectric is t m. In the dielectrics, maximum fields occur at the respective
inner radii. Therefore, the magnitudes of maxm fields Emax1 and Emax2 in inner and outer
dielectrics are given by
l l
E max1   7200 kV/m and Emax 2 
2  4  o  0.005 2  2.5   o  .

 6000 kV/m
0.005t

Putting the value of l=8 C/m as obtained from the equation of Emax1 into equation of
Emax2 we obtain t=0.0046 m.
If V1 and V2 are PDs across the dielectrics, then they are given as
0.005 0.005
l a ρ l
V1    E1  dl     da ρ  ln(0.0096 /
 0.0096  0.0096
2   4  
o 2   4  o 0.005)

 23.46kV.  
0.0096 0.0096
l aρ a ρ 
 d0.0096)  ln(0.025 /
V2    E 2  dl

 0.025 2  2.5   o  2  2.5  
l


o
 0.025
Therefore
 55kV. the maximum working voltage of the cable is (23.46+55)=78.46 kV.
Example 1.40 The radius of the inner conductor of a co-axial cable is 0.40 cm.
Concentric layers of rubber (rr=3.2) with inner radius 0.40 cm and polystyrene (rp=2.6)
with inner radius p are used as insulating materials. Design the cable (i.e., determine p
and the inner radius of the outer cylindrical shell (outer radius of the polystyrene) o) that
is to work at a voltage rating of 20 kV. To avoid breakdown due to voltage surges, the
maximum electric fields in the insulating materials are not to exceed 25% of their
dielectric strengths. Assume that the dielectric strengths of rubber and polystyrene are 25
MV/m and 20 MV/m, respectively.
Solution: Assume that the line charge on the inner conductor is l C/m. In the dielectrics,
maximum fields occur at the respective inner radii. Therefore, the maxm fields Emax1 and
Emax2 in inner and outer insulators are given by
l l
E max1  aρ and Emax 2  aρ .
2  3.2  o  0.004
2  2.6  o   p
 E =0.25x25MV/m and E =0.25x20MV/m
Now using  in the above equations, we
max1 max2
have p=0.616 cm.
However, if V1 and V2 are PDs across the insulators, then they are given as
 M.M.Ali
Page No.:47/60
i i
 l aρ l
V1    E1 dl    2  3.2  o
 da ρ 
2  3.2 
ln( p/  )i
 p  p o

l  
V1 V 2 ⎡ 1 1
ln( /  )⎤  20000V
ln( p/  )i  2.6
o p

2 o⎢⎣32. ⎥⎦
Now finding the value of (l /20)=8x104 form the equation Emax1 we obtain o=0.832 cm
from the above equation.
Example 1.41 A cylindrical capacitor consists of an inner conductor of radius a
and an outer conductor whose inner radius is b. The space between the conductors is
filled with a dielectric of permittivity , and the length of the capacitor is L. Determine
the capacitance of the capacitor.
Solution: In the region ab, E can be found using Gauss’s law as E=Q/(2L) a. V
is then calculated from
a
a a
Qa ρ Q Q b
 d a ρ   ln   ln
V    bE  dl   b 2L 2L b 2L a

Q 2L
C
V
b
ln a
Example 1.42 A spherical capacitor consists of an inner conducting sphere of
radius Ri and an outer conductor with a spherical inner wall of radius Ro. The space in
between is filled with a dielectric of permittivity . Determine the capacitance. Assume a
charge Q is distributed on the surface of the inner sphere.
Solution: In the region Ri r Ro, E is found using Gauss’s law as E=Q/(2r2) ar. V is
then calculated from
r i R
Qa  dra  Q
r Q 1 1i
r i
(  )
V   r  RE  dl   r 4 2
o r
4r Ro 4 Ri R0
R
Q
o
4 r
C 
V 1
( R  1R)
i 0

1.16 Poisson’s and Laplace’s Equations

Using E= - V and D=E in D=v, we have

V =2V= - v/. (30)
Equation (30) is known as Poisson’s which becomes Laplace’s equation in absence of
free charge v. Now
 2V  2V 
2V  2   in Cartesian coordinate system
 xV 2 y 2 z
2
1  V 1  2V 
 2V    (  )
 in Cylindrical coordinate system and
V2  2 z
2

2
1  V 1  V 1 
 2V  r2V (sin ) 2
2
(r ) in Spherical coordinate
2
r  r r sin 
2
 r sin   2
2

system.
 M.M.Ali
Page No.:48/60
In most practical problems, the exact charge distribution is not known everywhere, so the
previous equations obtained from Coulomb’s law cannot be applied for finding the
potential and the field intensities. Sometimes the potentials of the conducting bodies may
be known and we wish to find the potential and field intensity in the surrounding space. In
such cases, the above differential equations must be solved subject to the appropriate
boundary conditions. These problems, therefore, are named by boundary value
problems.

Example 1.43 Two large parallel conducting plates are separated by a distance d
and maintained at potentials 0 and V0 as shown in the following figure. Assuming
negligible fringing effect at the edges, determine the potential at any point between the
plates and the surface charge densities on the plates when (a) the region between the
plates is empty and (b) the region between the plates is filled with continuous distribution
of charge having a volume charge density v= -0 y/d.
y

d
V0

0
Solution:
(a) We have to solve Laplace equation 2V=0.
2 ⎛  2 V   2 V  2 V
Now in rectangular coordinate system,  V  ⎜⎜ x 2 y 2 2 ⎟
⎝
⎞ z ⎟
⎠  2V
However, for this particular problem where V is only function of y, we 0
have y 2

V
 y  C 1  V  C1 y  2
C
Now C1 and C2 are to be determined from the boundary conditions: V=0 ay y=0 and
V=V0 at y=d. Substituiting these conditions in the above equation of V, we have C2=0
and C1=V0/d. Therefore,
V0
V y.
d
Now as the electric field and potential are related by E= -V, threfore we have
 V V
V  a )    ( V0 y)a y   0 a y .
E  -V  ( x a x y a y z z y d d
V
However, for parallel plate capacitor, |E| in between the plates is given by s/0.
 V V
Therefore  s  d0  s  d0  0
0

As E is directed in the –y direction, the upper plate is charged with positive charge i.e., s
and the lower plate is charged with negative charge i.e., -s.
 M.M.Ali
Page No.:49/60
(b) We have to solve Poissons equation 2V=-v/0.
2 ⎛  2 V   2 V  2 V⎟
In rectangular coordinate system,  V  ⎜⎜ x 2 y 2 2 ⎟
⎝
⎞ z
⎠
. However, for this particular problem where V is only function of y, we  2V
2  
v
have y 0

y 3 
  2V 0 y V 0y 2
y 2  0 d  y  20 d C  V  60  C y  C
1 0 1 2

C1 andd C2 are to be determined from the boundary conditions: V=0 ay y=0 and V=V0 at
y=d. Substituting these conditions in the above equation of V, we have C2=0 and
V 
C1 ( d0  0 d) . Therefore,
0

6 
V 0 V0  0
6 0 d y  ( d 6d)  y
3

Now 0as the electric field and potential are related by E= -V, threfore we have
V V  0
E  -V  ( x a x  y a y  z az )   y 6(  d y 3  ( d0  0 d) y)a
V V  y
0 0
 6
 V 0
 ( 2 0 d y 2  d0  d)ay .
0 0
At the6lower plate
0
d)ay . Using Gauss' s law at the lower plate 0 E  Say  S
V
E  ( d0  sl
0
 6V
0

  0 ( d  0 d)   sl
0
6 plate
At the upper
 V
E  ( 0 d  0 )ay.. Using Gauss' s law at the upper plate 0 E  S (ay )   su
 3 0 d
S
 V
  0( 30d  0 )   su
0
Heredsl and su are the surface charge densities on the lower and upper plates.
Example 1.44 A high voltage coaxial able consists of a single conductor of radius
Ri and a cylindrical metal sheath of radius 0 (0>i) with a homogeneous insulating
material between the two. Show that the potential distribution in the dielectric is
Vs ln(0 /  ) and hence determine the capacitance per unit length of the cable.
V
ln(0 / i )
Assume Vs is the voltage applied on the inner conductor and the the outer metal sheath is
grounded.
Solution:
i 0

V= Vs

V=0
 M.M.Ali
Page No.:50/60
We have to use Laplace equation in cylindrical coordinate system; therefore,
1   1 2
 2V  0  (  V )  V   V0
    2 2  2 z
2

1 
As V is only function of , we have  ( V ) 0


   V  C1  V 
 C1 Integrating this equation w.r.t  we have

V  C1 ln   C2
Now from the boundary conditions V  0 at   0 and V  Vs at   i we have

C1  Vs and C2  Vs ln 0
ln( 0/  )i  ln( 0/  )i
Vs Vs
V   ln   ln 0  Vs ln(0 / )
ln( 0/  )i ln( 0/  )i ln( 0/  )i
 
Now E  -V  -( a ρ a φ   a z )⎜⎛ Vs ln(0 /  ) ⎞
  z ⎝ ln( 0/  )i ⎟
⎠
V
 ⎛ Vs ln(0 /  ) ⎞ ⎟a ρ  s
aρ
E- ⎜⎝ 0
 ln( ⎠/  )
i  ln(0/  )i

1 1 ⎛s V 2 0 2  L 1 dddz  L⎛⎜s V ⎟ ⎞ln( 2

 / 0 )
 E dv 
2
We   ⎜ ⎟  ⎞  2 i
v
 2 2 ⎝ ln(0 / i ) ⎠ i 0 0  ⎝ ln(0 / i ) ⎠
Here We is the energy stored in the cable of L meter length. If C is the capacitance of
the 1
cable per unit length, then We is again given by W e  CLVs2 .
2

1 2

s  L ⎜
 CLV 2 ⎛ Vs ⎟⎞ ln( 0/  )i
2 ⎝ ln( 0/  )i ⎠
2 F/m
C
ln(  0 / i )
Example 1.45 A coaxial cable have a core diameter of 2 cm and an insulator
thickness of 0.5 cm for which r=3.5. Calculate the capacitance, voltage distribution in the
dielectric and the electric field at the core surface that is at a voltage of 1 kV with respect
to the outer shield.
Solution: Cross-sectional view of the cable is shown in the following:

0.5 cm

2 cm
r=3.5 V0=1 kV
 M.M.Ali
Page No.:51/60
In the insulator
1   1 2
 2 V  0  (  V )  V   V0
    2 2  2 z
2
Due to symmetry, V is only function of r,
therefore
1
  
  ( V )  0   V  C1  V  C1 ln   C 2
 the boundary conditions, V  0 at   1.5 cm and V  1 kV at   1 cm we have
For
C 2  1 000 and C1  2466.3; thereforeV  2466.3 ln   1000. Here  in cm and V is in
volts.
 1   2466.3
 E  V  -(a ρ   a φ a z )V z a ρ

Therefore at the surface of the core E  2466.3 Volts/cm

l
Now for line charge E a ρ  l 
2466.3
 2    2  2466.3  2
 2466.3  2 0  3.5  0.48
C/m
 0.48 106 9
C  Vl  1000  0.48 10
F/m
1.17 Theory of Images

In many practical problems, charge is placed near a conducting surface that induces
charges on it. The calculation of total electric field in such a region needs to consider their
effects also. For example, the effect of the earth on the fields from an open wire
transmission line cannot be ignored. Similarly, the field patterns of transmitting and
receiving antennas are greatly modified by the presence of the conducting bodies on
which they are mounted. To consider the influence of a nearby conductor on a field we
must know the charge distribution on the surface of the conductor that depends upon the
fields just above its surface. However, in the case of static fields, we know that (a) a
conductor forms an equipotential surface, (b) there are no fields inside an isolated
conductor, and (c) the fields are normal to the surface of a conductor. These observations
will help to quantify the charge distribution on the surface of a conductor and its influence
on the fields in the region.
In the calculation of potential and electric field of an electric dipole, we found that the
potential on the bisecting plane is zero, and the electric field is normal to the plane.
Therefore, the bisecting plane satisfies the requirements of a conducting plane. In other
words, if a conducting plane is inserted to coincide with the bisecting plane, the field
pattern of the dipole remains unchanged. If the negative charge below the conducting
plane is removed, the field distribution in the region above the plane still remains the
same, and the total charge induced on the surface of the conductor is –Q as shown in the
following figure. Conversely, if we are given a point charge Q at a distance d above a
conducting plane of infinite extent, we can determine the potential and the electric field at
any point above the plane by ignoring the plane and imagining a charge –Q at the same
distance away from the other side of the plane. However, these statements are only true
for a conducting plane of infinite extent and depth. For a curved surface, the imaginary
charge is neither equal in magnitude nor as far away on the other side of the conducting
 M.M.Ali
Page No.:52/60
surface. The imaginary charge –Q is said to be the image of the real charge Q. Thus, in
the method of images, a conducting plane is temporarily ignored, and an imaginary charge
is placed behind the plane.

(a) Point charge near an infinite grounded conducting plane

P’ O(x, y)

d Q -Q d O d Q
+

P
Let PP’ be an earthed conducting infinite plane. Consider Q at a distance d on the right
side of the plan. Whatever may be the charge induced on the plane, it is always at zero
potential because it is earthed. It is obvious that if we remove the grounded conductor and
replace it by a point –Q at a distance d to the left of the plane, then every point on the
plane will be equidistance from charges and will be at zero potential. Now these two
charges will give the proper solution of the problem. The charge –Q is known as the
image of the charge Q. Now the potential at a point O(x,y) is given by
Q Q
V 
4 (d  x)2  y 2 4 (d  x) 2  y 2
⎡ ⎛ ⎤
⎢ ⎟ ⎜ 1 ⎞ ⎥
⎟
⎛ 1 ⎞ 2
⎥
Q ⎢ ⎜  ⎟a 2
E  V   ⎟⎟ay ⎥
4 ⎢1 (d  x)  y y ⎜⎜1
2 2
⎢x ⎜⎝ (d  x)  y
2 2 ⎟⎠ x  ⎥
⎢⎣ ⎜ (d  x) 2  y 2 ⎟ ⎥
⎝ ⎠ ⎦

⎡ ⎜⎛ ⎤
y
2 3/2 
⎞⎟ ⎥
Q ⎢⎛ dx dx ⎞ {(d  x)  y }
2

 ⎢⎜⎜  ⎜
2 3 / 2 ⎟ ⎟a x  ⎜
⎟a ⎥
2 23/2
4 ⎢⎢⎝ {(d  x)  y } {(d  x)  y }
2
⎠ y ⎟ y⎥
⎜⎜ 2 2 3/2 ⎟ ⎥
⎣ ⎝{(d  x) ⎟ y } ⎠ ⎦

Now for x=0 i. e., on the grounded conducting plane D=E=s

dQ
  s 
2 (d  y 2 )3 / 2
2
 M.M.Ali
Page No.:53/60

(b) Point charge near a grounded conducting sphere P

r4
r3 r P3
a r2 a
r1 ’

Q
Q d d P1
P2
b
’
Let us try to find the position and magnitude of the image charge Q that will make the
potential on the spherical surface zero. The zero potential at P1 and P2 give us
Q  Q   0 and Q  Q 
d a0 ab da a
Solving
b these two equations gives Q’= - aQ/d and b=a2/d.
Let us now calculate potential of P3 and hence first prove that the image charge Q’ and its
position as determined above are valid solution of this problem.
Q ⎡ 1 a/
VP 3  Q  Q   ⎤ 2 ⎥
4r1 4 d  a 2 d 2da cos  b 2  a 2  2ba cos
⎢
4r2 ⎢⎣ ⎥⎦
Q 1 1

4 ⎡⎢ d 2  a 2  2da cos  a 2  d 2  2ad cos ⎥⎤  0
Since
⎢⎣ the original
⎥⎦ point charge and the image jointly satisfy the boundary condition, they
must give the correct field at every point in the space outside the sphere. The potential at P
is
⎡ 1 a/
VP  Q Q Q ⎤

4 r3 4  ⎢
4 d 2  r 2 d 2dr cos 
b 2  r 2  2br cos
r4 ⎥ ⎢⎣ ⎥⎦ ⎤
Q 1 a/
 ⎢ 
4 ⎡ d 2  r 2 d 2dr cos a 2  (rd / a) 2  2rd cos
⎥
Special
⎢⎣ Note:
⎥⎦ If the sphere is at a potential other than zero, we may still determine the
voltage or electric field by the method of images. We first replace the conducting sphere
with an image charge Q, as we did for a grounded sphere. This makes the surface
occupied by the sphere an equipotential. We next add a second image charge at the center
to raise the spherical surface to the required potential. If we are given a sphere of radius a
with a charge Q on it, we can replace the sphere by an image charge Q= - aQ/d at a
distance b=a2/d from the center, plus a charge (Q- Q) at the center.
(c) Line charge near a parallel conducting cylinder

P3
r1
a r2 a
-l 
l d l P2 P1
d
b
To make the surface equipotential, the image must be a parallel image line charge inside
the cylinder; and because of symmetry with respect to the line joining the charge l and
 M.M.Ali
Page No.:54/60
the center of the cylinder, the image line charge must lie somewhere on this line. If we
assume the image as -l, its position at a distance b from the center can be determined by
equating the potential of P1 with that of P1. We, therefore, have

l r0 r0 l r0
 l 
r0 0 ln d  a  2 0 ln a  b  2 0 ln d  a  2l ln
2 0 a
b a2

ab ab b
da da d
Now the potential at P3 is
l l
VP3  r0   l r0 r0
20 ln r1 2ln0 r2 20 ln r1
r1  (d  a cos ) 2  (a sin ) 2  d 2  a 2  2ad cos
and
a
r2  (b  a cos ) 2  (a sin ) 2  b 2  a 2  2ab cos  d d 2  a 2  2ad cos
l
 VP3  a
20 ln d
(31)
Example 1.46 Determine the capacitance per unit length between two long,
parallel circular conducting wires of radius a. The axes of the wires are separated by a
distance D.
Solution:
The above problem can be realized by the following system:
2

1
-l d
b
l
D if b=a2/d. Now if V1 and V2 are the
The equipotential surfaces of the wires can be realized
potentials at the surfaces of the wires 1 and 2, then using (31) we have
l l l d
V1  a

d 2  a
 
l 0 d 20 a
2 ln ln and V ln ln
2 0 d 20 a
V1  V 2  l ln d  C  V  lV


 0 F/m
 0 a 1 2 ln
d

1
Again d  D - a 2 / d d  (D  D 2  4a a
2

2  0  0
C is also given by C  F/m
 ln[(D / 2a)  (D / 2a)2  cosh (D / 2a)
1

Example 1.47 1]A transmission line of radius a and having charge density l C/m is
situated at a height h above and parallel to the surface of the infinite ground plane.
Determine the capacitance per unit length of the line between the line and the ground
plane.
Solution:
 M.M.Ali
Page No.:55/60

l l

h h
Cg

Ground
h Cg
-l

(a)

(b)

The problem is depicted in fig. (a). The conducting ground plane can be replaced by
considering an image line at a depth h below the ground plane. The problem then
becomes similar  to that in Example1.46 except D=2h. The capacitance between the
C 0
 F/m
original line
ln[(h and
/ a) 0 the
(hits
/ a)image
2
 iscosh
therefore
1
(h /given by
Now 1]the capacitance between the a)line and the ground is, therefore,
2 0
C g  2C  F/m
cosh1 (h / a)
Example 1.48 A point charge q is located at (0, 0, d) above the surface of a conducting
plane of infinite extent and depth. Determine the potential and electric field at P.Also
show that the total charge induced on the conducting surface is –q.
Solution: The problem is depicted in the following figure. Following the theory of
images, to determine the fields at any position above the plane we have to place an image
charge –q at (0, 0, -d) and ignore the existence of the plane.
Now the potential at P (x, y, z) is

V q ⎡ 1  1 ⎤, where R1=[x 2 +y2+(z-d) 2] 1/2 2and

2 2
R =[x +y + 2 1/2
4 0 ⎣⎢ R1 R2 ⎥(z+d) ]
Now the ⎦electric field E is
E  V  - q ⎢⎡x ⎛ 1 1 x ⎛ 1 1 ⎞ ⎛zd zd
4 ⎜ R 3  R 3 ⎟a  y⎜ R3  R 3 ⎟a y ⎜ 3
 3 ⎟⎥a
⎞ 0 ⎢⎣ ⎝ 2 1 ⎠ ⎞z⎝⎤ 2 1 ⎠ ⎝ R2 R 1 ⎠⎥
2qd
On the surface of the conducting plane E= - a z, where R=[x2+y⎦2+d2]1/2.
4 0 R 3
 M.M.Ali
Page No.:56/60

Form the boundary condition at the interface of free space and conducting body we have
the surface charge density induced on the conducting surface must be equal to the normal
component of D. Therefore,

 s - 2qd
4R.3 Thus the total charge induced on the conducting surface of infinite extent
2
2qd  dd 
d
is Q  s s ds  -  
 0  (  d )
2 23/2
 qd 
 (  d )
2 23/2
 q
4 0 0

1.18 Equation of Continuity

Consider an arbitrary volume V bounded by a closed surface S contains a charge Q. If a

current I flow across the surface out of this region, the charge in the volume decrease at a
rate that equals the current. The current leaving the region is the total outward flux of the
current density vector through the surface. We, therefore, have

Again I  J  ds  J  ds  
dQ d d
I  ρv dv ρv dv
v s s 
v
dt dt dt
Now with the help of divergence theorem we have

 J  ds  (  J)dv   d ρ dv    J   dρ
v v
dtContinuity equation.
(32)
s v
dt  v

As J  σE , From eq. (32) we have

dρv
dρ σ dρ σ v
  σE   dtv    v  vdt  dt    0
The solution of the above equation is
   e ( /  )t C/m3. The initial charge v0 will decay to 1/e or 36.8% of its value in a
v v0

time equal to =/. The time constant  is called the relaxation time. We are now in
position to explain why the charges introduced into the conducting material will move to
its surface and redistribute themselves in such a way as to make =0 and E=0 inside
under equilibrium conditions.
Now for dc current   J  0 that leads to  J  ds 0 . If the closed surface shrinks to
a
s

point, we have the Kirchhoff’s current law  I  0 . Putting J=E in the continuity
equation for dc current, we obtain
  J  0    E  0    E  E    0 . But for homogeneous medium =0,
therefore we obtain
  E  0    (V )  0   2V  0
The equation 2V=0 states that the potential distribution within the conducting body
satisfies Laplace’s equation as long as the medium is homogeneous and the current
distribution is time invariant.
 M.M.Ali
Page No.:57/60

1.19 Boundary Conditions for Current Density

Boundary condition for Boundary condition for

tangential component normal component

We study here how the current density J changes when passed through the interface
between two media of different conductivities. The boundary condition for tangential
components of static electric fields, Et1= Et2, relates Jt1 with Jt2, and the continuity equation
 J  ds 0 determines the relation between the normal components of current densities
s

in two different media.

Using J=E in Et1= Et2, we have

J t1 J t2 J t1 1 (33)
1  2   
Jt 2
where
2 1 and 2 are respectively, the conductivities of media 1 and media 2.
Assuming the height of the pillbox shown in the above figure is small so that the
contribution from the current leaving or entering through the curved surface is negligible.
Now from the integration of the continuity equation over the closed surface s of the
pillbox with h0, we obtain
J n1an  san  J n 2 a n  san  0  J n1  J n 2 . (34)
From (33) and (34) we obtain
 2 J t1 J n1 J t1 J n  J t1 / Jn 1  1
tan1  1 (35)
   1
2    
2 tan 2  2
1 J t 2 Jn 2 J t 2 J n1 J t 2 / Jn
Let
2
us consider a special case
2 where mediums 1 and 2 are respectively, poorly and highly
conducting. As  is directly proportional to the conductivity, 1 will be very small because
2 1. J in medium 1 is therefore almost normal to the interface. This leaves the
tangential component of J and hence that of E negligibly small in medium 1. Smaller
value of the tangential electric field in medium 1 leaves the tangential component of E in
medium 2 small as Et1= Et2. Now from (34), the normal component of E in medium 2
given by

E n2  1 E n1is also very small. This means that the electric field in a highly conducting
2

medium is practically zero. As the medium 2 is a conducting body, there must exist a free
surface charge density at the interface we can be calculated from (26) as
 M.M.Ali
Page No.:58/60

1 ⎡  1 ⎤ ⎡ 1  2 ⎤
 D D  E  E  E  E 
s n1 n2 1 n1 2 n2 1 n1
2 2
E
⎢ 1 
n1 ⎣ 2  2 ⎦⎥ J n1⎣⎢1  2 ⎥
Similar expression can be obtained in terms of
n1 ⎦
Jn2 as
⎡  ⎡  ⎤
 s  D n1  Dn 2  1 En1  2 En 2  1 12 E n 2  2En 2  En 2 ⎣⎢ 1  1  2 ⎥  J n2 ⎣ 11  2 2⎥.
⎤
 ⎢2 
⎦ ⎦
1.20 Leakage Resistance between Two Conductors

Whatever be the shapes of the conductors, the capacitance between the conductors is
given by

C
Q  D  ds  E  (36)
 ds
s
 s

- l E  dl - l E 
V
Now if thedldielectric between the conductors is loosy, its resistance is given by

V - E - E
R  dll  dll
(37)

I s J  ds s  E 
ds
  (38)
RC   R
Two infinitely conducting parallel plates, each of cross-

area A,Care separated by a distance d. The potential difference between the plates is Vab. If
sectional
the 1.49between the plates is homogeneous and has a finite conductivity ,
medium
Example
determine the resistance of the region between the plates and the capacitance.
Solution: Let us consider the following problem configuration.

d 2V
The potential distribution may be expressed as 
dz 2 0 .
Integrating the above equation twice, we obtain V=az+b where a and b are constants and
can be obtained from the boundary conditions that V=0 at z=0 and V=Vab at z=d. Now
a=Vab/d and b=0. Therefore, the potential in the medium between the plates is given by
V=(Vab/d)z.
d Vab Vab
 E  V   dz( dz)a  
z ad z

V
 J  E   dab a z
Now the current through a surface of area A normal to J is

V A
 I   J  ds   dab
 M.M.Ali
Page No.:59/60

V d
Therefore, R= Iab 
A
 A
Now following (38), we obtain C  R  where
d  is the permittivity of the medium.
Example 1.50 The conductivity of the medium between two concentric metal
spheres is
 and the permittivity is . If the radius of the inner sphere is a and the inner radius of the
outer sphere is b, determine the resistance of the medium between the spheres.
Solution:
4abFollowing the result of Example 1.42, Capacitance of the system is given by
C .
ba
 (b  a)
Now using (38), we have R  C . 4 ab
Example 1.51 A two-wire transmission line consists of wires of radius a. The
wires are separated by a distance D in a lossy media of permittivity and conductivity,
respectively,  and . Determine the leakage resistance between the conductors for unit
length of the line.
Solution: The solution has two steps. In a first step, the capacitance per unit length of the
line between the wires is to be determined which is explained in Example 1.46. C is
given by
 
C  ln[(D / 2a)   (D / 2a)2  F/m. cosh 1 (D /
1] the leakage resistance per2a)
Therefore unit length of the line
is
 1 1
R  C  cosh 1 (D / 2a)  ln[(D / 2a)  (D / 2a)2  1]  
 resistance of the line for a length of l meter is R/l.
m. total
The
Example 1.52 A coaxial cable consists of inner conductor of radius a and outer
conductor of inner radius b. The space between the conductors is a lossy media of
permittivity and conductivity, respectively,  and . Determine the leakage resistance
between the conductors for unit length of the cable.
Solution: The solution has two steps. In a first step, the capacitance per unit length of the
cable is to be determined which is explained in Example 1.41. C is given by
2
C
F/m.
b
ln a
Therefore the leakage resistance per unit length of the cable
is 1
 ln(b / a)  
R  C 
m.
2 resistance of the cable for a length of l meter is R/l.
The total

1.21 Summary

 The electric fields for any kind of charge distribution are found to obey the
following relations which are known as the fundamental postulates of
electrostatics:
 M.M.Ali
Page No.:60/60
  E  v /    E  0.
 In presence of source v, the electric field is not solenoidal but irrotational.
However, in a source free region the electric field is solenoidal as well as
irrotational.
 No free charge stays within a conducting body; the electric field is thus zero
everywhere in it.
 Electric field is always normal to the conducting surface which leaves the
conducting surface as equipotential.
 STATIC MAGNETIC FIELDS

Introduction

We will study here the magnetic field associated with a dc current carrying conductor.
The aim of this chapter is to discuss the Biot-Savart law and use it as a basic tool to
calculate the magnetic field set up by any given distribution of currents.

2.1 Magnetic field of a moving point charge

It was found experimentally that the magnetic field (magnetic flux density) B at a
qu sin 
distance R away from a charge q moving with velocity u is related by B where
the proportionality constant in SI unit is /4. Here  is the permeability of the media
where the field is to be measured. In free space = 0 and has a value of 4x10-7 Hm-1
(Henrys/meter).

or B 
 q u  aR (36)
4 R2
where B and u are the magnitude of B and u,  is the angle between u and R (here R is a
vector pointing from the charge to the point where the field is being found), and aR is the
unit vector acting along R. It is noted that the magnetic field intensity H=B/.

2.2 The Biot-Savart law

In fact, we are often interested not in the field of moving charge, but in that of a current
carrying conducting wire. Biot and Savart proposed a mathematical formula for the field
due to a single current element (a small conducting wire of length dl having current I
through it) as follows.

P

 R

dl I

The geometry is shown in the figure. Suppose cross-sectional area of the wire is A and v
is the density of charge in the wire. Assume these charges move with velocity u. Then the
charge crossing any cross-section in one second=uAv=I. Idl= uAdlv=uq.
Therefore the magnetic field dB at point P in the above figure due to dl length of wire
carrying current I is given by the following relation after substituting qu=Idl or qu=Jdv
(because Idl=IdIA/A=Jdv where dv is the volume of the elemental wire) in equation (36):
I dl  a
 (37a)
dB  4 R
R2
 JaR
or dB  4 dv
R2
(37b)
The resultant field is found experimentally to be normal to the plane containing the
vectors dl and R This expression (37) is known as the ‘Biot-Savart law’.
The unit of B is tesla (T)
By the use of a suitable integration, Biot-Savart law can be used to calculate the B-field
resulting from a current carrying conductor of any length L which can be decomposed
into an infinite number of back to back connected current elements:

µI dl  a R (38a)
B
4 
2
R

µ JaR (38b)
or B  dv
4  R 2

The direction of magnetic field can also be found by using Right Hand Rule which states
in the following way: Grasp the wire with right hand; if the thumb points in the direction
of current, the fingers will curl around the wire in the direction of B or vice versa.
If the position vectors of the field point P and the source point Q are, respectively, r and
r, we have R in (38) as R=r - r and aR=(r - r)/| r - r|.

Therefore the alternate form of (38) is obtained as

µI dl  R µI dl  r  r µ JR µ J  r  r
B  or B  dv 
dv
4  R3 4  r  r
3
4  R3 4  r  r
3

Example 2.1 Determine B-field due to an infinitely long, straight current carrying wire.

Solution: We would like to determine B at P a perpendicular distance r from the wire as

shown in the following figure.

I az aR
 P a
 R
dl=dz az z

L
Consider dB produced by an element of length dl which lies along the z axis and a
distance R rom P as shown in the figure. Assume the point P is L unit above the x-y plane.
dB=Idz az  aR/(4R2)
but az aR=az(cos  az + sin  a)= sin  a Now
sin =/{2+(L-z)2}1/2 and R2={2+(L-z)2}
Hence dB=(I/4)dz a /{2+(L-z)2}3/2

LU
B  µI  2 dz a φ   µI ⎢⎡ 2 L  2 1/ 2 ⎤ a φ
z4  {   (L  z)23/2
} 4 ⎣{  (L  z) }
⎥⎦ L
where z=LU and z=LL are the upper and lower position of the wire along z-axis. For an
L

infinitely long wire the limits of the integral are LU= and LL= - . Hence
µI
B aφ
2 
The resultant B-field, therefore, is of the form of concentric circles around the wire.

Example 2.2 A circular loop of radius a carries a direct current I in the anti-clockwise
direction as shown in the following figure. Calculate H at a point on the z-axis.

Solution:

R=z az –a a  R3=(z2+a2)3/2

dl x R= ad a x (z az –a a)= azd a + a2 d az

2 2
 I dl 3 R I azd a2
B
I dl  a R
2  a
2 d 2 3 / 2 ρ
  2 2 3 / 2 az ]
4 R 4 R 4 (z  a ) 0 (z  a )
[ 0

The first integration over a closed circular path is zero, as a varies with . It can be
verified mathematically if we replace a by (cos  ax + sin  ay) in the first integration and
then integrate it w.r.t  from 0 to 2. The field is therefore given by
B  I a2 a T . At the centre of the loop B  aT
2I (z  a 2 )3 / 2
2 z

2a z
For z>>a B  I a a T .
2

2 z3
When the point of observation is far from the loop, the size of the loop is very small in
z
comparison with the distance z. In this case, we refer to the current carrying loop as a
magnetic dipole. If the magnetic dipole moment is defined as m=Ia2az, B is then given
by
m
B  2 zT3

Example 2.3 A sheet of conducting metal carrying a dc current of IS amperes per unit
width in the z-direction as shown in the following figure. Calculate H at a point P(x0, y0,
z0).

Solution: z

P(x0,y0, z0)
R
dz
z This sheet is assumed be placed at y=0.

y
x
dx
x

R=(x0 –x) ax + y0 ay + (z0 –z) az

R3={(x0 –x)2 + y02 + (z0 –z)2}3/2
 I dl  a R   I dl  R
d (dB)  4
R2 R3
4 az
I=ISdx, dl=dz
dlxR= dz az x {(x0 –x) ax + y0 ay + (z0 –z) az}= dz {(x0 –x) ay - y0 ax}

I   {(x 0 x)a y  y 0 a x }dzdx

B  4S   
 2
{(x  x)2  y 0  0  z) 2 }3 / 2
z  x  0
B x   I  (z 
SI H  
y dzdx   S I
 
0

4
S 2 x
z  x {(x 0  x)  y 0 
2
0  z) 2 }3 / 2
2 2
B y  I   (z (x 0 x)dzdx
S
4   {(x  x) 2
 y
2
 2 3 / 2
 0 Hy  0
z  x 0 0 0  z) }
(z
I
 H   2S a x
It can be shown similarly H=IS ax/2 for P(x0, -y0, z0).
It is very interesting to note that electrostatic field due to sheet charge was also found to
be constant like magnetostatic field, irrespective of the position of the observation point
in front of the sheet.

Example 2.4 Assume that the wire in the following figure has a current I through it in the
direction shown. What is the magnitude of the magnetic field produced at point C due to
the straight segments, (b) the semicircular arc and (c) the entire wire?

x
Solution:
a.Looking at the figure, we see that the straight sections have infinitesimal lengths that lie
parallel or antiparallel to R, the vector that goes from dl to point C. Therefore, dl×aR = 0
in both cases for every dl along the straight lengths.
b.For the semicircular arc, each infinitesimal segment, dl, is an equal distance R
from point C. For each such segment, dl×aR is directed into the page (here aR is
antiparallel to a of cylindrical coordinate system) since dl and R are perpendicular at
each point around the arc. Therefore, by the Biot-Savart Law, we have

B  µI  dl  2a  µI Rd  aφ (aρ )  µI a
R 

4 R2 4 0 R z
4R
c. To get the net contribution, just add the results for parts a. and b. of the question

µI
B a
z
4R

Example 2.5 Consider the circuit shown in following figure. A current I runs through it.
Given the parameters shown in the figure, find the magnetic field at point P.

y
x

Solution:
We can take our result from the previous problem and apply it here directly. The straight
sections yield no contribution to the field at P since they lie along the radial line from
each current element to point P. We also know that a semicircular arc gives a magnetic
field of magnitude B=I/4R. An arc which is less than a semicircle gives a field of
B=I(arc-length)/4R
B
I arc  length

4R R
For our problem, we use the right-hand rule to note that the field due to the arc of radius
b points out of the page while the arc of radius a gives a field that points into the page.
Then, the net field at P

B ⎜
⎛ I  b  I  a ⎞ ⎟a z  I (a  b) a z
⎝ 4b b 4a 4a
⎠ a 2.6 Find the magnetic bfield at point P for following figure which shows a
Example

current I traveling a long, straight wire length into a semicircular wire of radius R and out
along another long, straight section.

Solution:
Let's consider the current distribution as consisting of 3 parts as labeled in the following
figure.
y

These pieces are the two long, straight sections and the semicircular arc. For pieces 1 and
3, Note that their contributions to the net B field at point P are both out of the page (along
+ve z-axis). In setting up the Biot-Savart integral for infinitesimal lengths of current
elements dx, Note that the setup of the integral will be identical for both, namely, we can
consider the position of dx in both cases w.r.t. the places where the straight sections
combine with the semi-circular arcs. Therefore, the sum of these two contributions is the
same as that of a single infinite length line carrying a current I. Hence
dl  a R
µI
BB 2
1 3

Now dl=dx a and4 R=Rcos

R2
(-a )+ Rsin a = -x a +r a , |R|=R=(x2+r2)1/2
x x y x y

 dl  a R

µI dxax  (xa x  ra y ) / R  µI rdxa 3
z


 
B1  B 3  2  µI
4 R2 2 0
2
2 0 R
R
 B1  B 3  µI  2 rdxa

z
 µI
az
2 0 (x 2 r ) 23/
2 r
To get the contribution from piece 2, the semi-circular arc, look again at figure. In this
case, each current element is a distance r away from point P and the angle between dl, the
infinitesimal arc-length of current, and R, the vector from dl to point P, is 90 because dl
is tangential to the arc and R is along the radial direction. So,
dl  a  µI rdaφ  (aρ )
R  µI 
B 2  µI
4  r22 4 0 r 4r  da z  4r a z
 µI
0

Summing all the contributions together yields

B net  B1 B 2 B 
µI ⎛ 1 1 ⎞ ⎟a z
⎝⎜ 
3
2r  2
⎠
Example 2.7 Determine the field B at a point P on the axis of a solenoid having uniform
winding with n number of turns per unit length. Assume the current through the
winding is I amps and the radius of the solenoid is r.

Solution: dx
 
2 r
R d
r
1


P P
Rd

dx
Assume the point P is R distance away from the small segment dx of the solenoid.
From the right most figure, we have
Rd Rd
 sin   dx  and also r/R=sin.
dx sin 
B  I r a where R is the distance nRd
2
of the observation point on theInRdaxis of the ring
Total 2number
R 3 of turns on the length dx=
x and the total current in dx= .
sin 
from its circumference. Here the direction of the field is assumed in the x-direction,
following the Fleming’s Right-hand rule,sin as the direction of current is assumed to be into
the page at the upper conductors.
Now the field on the axis of a ring of radius r is given as (see example 2.2)
Therefore, for the segment dx, the field at P is
InRd r 32 a x  In ax
dB  R
sin 
2 sind  2
Total field at P therefore

2
In sind In
ax  (cos1  cos 2 )ax
B   2 2
1

If the solenoid is very long and P is somewhere near the middle, then 1=0 and 2= and
B  Inax
In
If P at the right end, then 1=/2 and  = then B  ax. Same result obtains when P
2 2
at the left end where 1=0 and 2=/2.

2.3 Ampere’s Circuital Law

Ampere’s circuital law states that the line integral of the magnetic field intensity H around
a closed path equals the currents enclosed. This statement is given mathematically by,

(39)
 H  dl  I   B  dl 
µI
2.4 The differential (Point) form of Ampere’s circuital law
By applying Stoke’s theorem the lest side of (39) can be written as
 H  dl   (  H)  ds
S

or,  B  dl   (  B)  ds
S
ds
dl

Again the total current flowing through ds can be written in terms of a surface integral of
the current density J as
I   J  ds
s
Hence Ampere’s circuital law can be written in the form
  H  J or   B  J (40)
This result must be true for any choice of surface s or at any point in space and is known
as the differential form of Ampere’s circuital law.
This equation shows that the magnetic field is not irrotational.
Ampere’s circuital law is very powerful in determining the magnetic field caused by a
current I when there is a closed path around the current such that the magnitude of the
field is constant over the path.
No free charges can stay with in a good conductor; the currents due to free charges
thus only flow along the surface of a good conductor. The right hand side of eq. (39)
for any closed path inside the conductor becomes zero. It concludes that a static
magnetic field can not exist in a good conductor.

2.5 Derivation of differential (Point) form of Ampere’s circuital

law

Let us consider an infinitesimal closed path in the x-y plane as shown in the following
figure. The fields along dx and dy remain constant.
 z

H=Hx0 ax+ Hy0 ay+ Hz0 az

4 3
dx 
1 2
dy

If the magnetic field circulates in the anticlockwise direction as shown, the current will
flow in the +ve z direction. Therefore eq. (39) in this case takes the following form
(41)
 H  dl  Iz
12341
Now the field that contributes to eq. (41) on the path 1-2 only has its y-component and is

H y  H y0  xy ( 2 )   H  dl  {Hy0  xy ( 2 )}dy

H dx H dx (42)
12
The field that contributes to eq. (41) on the path 2-3 only has its x-component and is
H x H  H
x dy
 H  dl  {Hx0  H dy (43)
y ( 2)  23 y ( 2 )}(-dx)
x0 x

The field that contributes to eq. (41) on the path 3-4 only has its y-component and is

H y  H y0  xy ( 2 )   H  dl  {Hy0  xy ( 2 )}(-dy)

H dx H dx (44)
34

and the field that contributes to eq. (41) on the path 4-1 only has its x-component and is
Hx H  H
x dy
 H  dl  {Hx0  H dy (45)
y ( 2)  41 y ( 2 )}dx
x0 x

The left hand side of eq. (41) therefore is given by

H
H
 H  dl  ( x y - y x )dxdy
H H H H (46)
 ( x y - yx )dxdy  I z  J zdxdy  ( x y - yx )a z J a z z

If a closed loop is considered in the z-x plane then we have

H H (47)
 ( z x - xz )a y J a y y
Again if a closed loop is considered in the y-z plane then we have
H H (48)
 ( y z - zy )a x J a x x

On adding eqs. (46)-(48), we have

H H H H H H
 ( y z - zy )a x ( z -
x
x)a  (
z y y
x-
x
)a
y
z  J xa x J a y J ya z z
H (49)
J
2.6 Divergence of magnetic field – Gauss’s law

As the magnetic pole cannot be isolated, unlike the electric flux the magnetic flux lines
always close upon themselves. Therefore, the total outward magnetic flux through any
closed surface is zero i.e.,
 B  dS  0
S
From divergence theorem we have
 B  dS  (  B)dv .
S v
As the volume has finite value, it reveals that   B =0 (50)
To prove eq. (50) mathematically, let us consider the magnetic field due to a long current
carrying conductor, the B field of which is given by
µI dl  a R
B
4  R2
µI dl  a
Then   B   R
. Now R can be assumed as R=|r-r|=(x-x)ax+(y-y)ay+(z-
4   R 2
z)az, where (x,y,z) and (x, y, z) are respectively the coordinates of the observation and
source points, and r and r are respectively the position vectors of the observation and
source points.
2 1/ 2
1
 ( R )  {(x - x)  (y - y)  (z -
2 2

z) }
 x - }3 / 2 a x  y - }3 / 2 a y  z - }3 / 2 a z  
R a
 R 2
x{ y{ z{
3
R R
µI dl  a R  1dl 
µI
B  
4   R 2
4  ( ) R
With the help of the vector identity AB=BA - AB the above equation can be
rearrange as
µI 1
B  [( )    dl  dl    However from vector identity,
1
( )].
1
  ( )=0. 4 The first term also
R equals to zero because  is function of (x, y, z) and dl is
R R
function of (x, y, z). This
proves that   B =0.
The divergenceless property of the magnetic field indicates that the magnetic flux lines
close upon themselves. It is thus an example of solenoidal field.

Example 2.8 Six parallel aluminum wires of small, but finite radius lie in the same plane.
The wires are separated by equal distances d, and they carry equal currents I in the same
direction. Find the magnetic field at the center of the first wire. Assume that the current in
each wire is uniformly distributed over its cross section.
Solution:
A schematic layout of the problem is shown below. The magnetic field generated by a
single wire as obtained from Ampere’s law is equal to
B  µI a φ
2
where  is the distance from the center of the wire. The equation is correct for all points
outside the wire, and can therefore be used to determine the magnetic field generated by
wire 2, 3, 4, 5, and 6. The field at the center of wire 1, due to the current flowing in wire 1,
can be determined using Ampere's law, and is equal to zero. The total magnetic field at the
center of wire 1 can be found by vector addition of the contributions of each of the six
wires. Since the direction of each of these contributions is the same, the total magnetic
field at the center of wire 1 is equal to

B  B 1 B 2 B  3B  B4  B5 6 
µI ⎛ 1 1 1 1 1 ⎞
    ⎟a φ
⎝⎜
2 d 2d 3d 4d ⎠
5d

Figure: Six parallel wires.

Example 2.9 A coaxial cable consists of a long cylindrical copper wire of radius r1
surrounded by a cylindrical shell of inner radius r2 and outer radius r3 (see Figure below).
The wire and the shell carry equal and opposite currents I uniformly distributed over their
volumes. Find the formulas for the magnetic field in each of the regions r < r1, r1 < r < r2,
r2 < r < r3, and r > r3.

Solution:
The magnetic field lines are circles, centered on the symmetry axis of the coaxial cable.
First consider an integration path for Ampere’s Circuital Law (ACL) with r < r1. The path
(line) integral of H along this path is equal to
 r 2I
 H  dl
 r12
Here dl=rda and H=Ha

2
r 2I r 2I
aφ
 H  dl  2  Hrd  2
rI H 
 2 r12
r1 r 0
1

For r1<r<r2, the ACL becomes

 H  dl  I Here dl=rda and H=Ha
 2
I
 H  dl  I  0Hrd  I  H  2 r a φ

For r2<r<r3, the ACL becomes

 H  dl  I  I⎛⎜⎝ rr   r r ⎟⎟⎠⎞ I ⎜⎝⎛rr r r ⎠⎞

2 2 2 2
2 22 32 2
Here dl=rda and H=Ha
3 2 3 2
⎜ 
2⎟
⎟
 H  dl  I⎛⎜⎜rr r r ⎞ ⎛r2 r2 ⎞ I ⎜⎛ 3r2 2  r 22 ⎞ φ
2 2
2 2 ⎟ Hrd  I⎜⎜r 2 r 2⎟ H 
2 r ⎝r 3 r 2 ⎠⎟a
3 ⎝ 3 2
⎟3⎠ 0 ⎝ 3 2 ⎠
For r>r3 the ACL becomes
 H  dl  0 Here dl=rda and H=Ha

 H  dl  0  H  0aφ
Example 2.10 A very long hollow conductor of inner radius a and outer radius b is
located along the z axis and carries a current I in the z direction, as shown in the following
figure. If the current distribution is uniform, determine H at any point in space.

Solution: In the region b   a, J=I/{(b2-a2)}az.

For  a, the current enclosed by any closed path is zero. Therefore, H is also zero as
obtained by ACL.
For b   a, the current enclosed by a closed circular path of radius  is

I enc   J  ds
I  2  2  a 2 I
 b2  a 2  a 
a  d d a 
b  a 
z z 2 2
 s
0
From ACL, we have
  a I 2 2 2

 daφ
 H  d l  I  H  b  a 
2 2
l enc

 H    a

2
I  H    a
2
0
Ia
2 2

 b  a 
2  b  a 
2
2 2 2 2

For b, the current enclosed by a closed circular path of any radius  is I
Therefore, from ACL, we have φ
 2
  H  daφ I
l H  d l  
0
I I I
H H aφ
2 2
Example 2.11 A closely spaced winding (toroidal winding) with N turns is wound in the
form of a ring as shown in the following figure (a). The inner and the outer radii of the
ring are a and b, respectively. The height of the ring is h. If the winding carries a current I
amps, find H within the ring and total magnetic flux enclosed by the ring.
Solution:

The current enclosed by any closed circular path of radius  where a   b is NI.
Therefore, from ACL we obtain
2

 Hdl   H  daφ NI

NI  0
NI NI

l
H  2  H  2a φ

 B  NI a
b h
NIh
lna /
    B  d s    B  d
dza 
2 2
φ
s
0
a z
b
Exampleφ 2.12 Along cylindrical conductor of radius a carries a current of density J.
Show that the H field, inside a cylindrical cavity whose axis is displaced from that of the
conductor by a distance d, is constant and depends only on the location of the cavity and
not on its radius b.
Solution:
y The cross-sectional view is shown in
 ’ the adjacent figure. Assume that the
current is flowing in the z-direction.
We can use superposition theorem
and consider the H field inside the
b a cavity as that due to two long
x cylindrical conductors with radii b
J and a and current densities J and –J,
respectively. Let now find the H field
at a point which is  distance away
from the centre of the conductor and
d ’ distance away from the centre of
the cavity.
From ACL  H  dl  I   J  ds , H at the specified position due to J is given
by
2H  l 2 J    J J φ  H  J  ρ
2 H s
2a 2
H
Similarly, H at the specified position due to –J is given by
 J  J J  ρ
2 H    2 J     2  H   2 aφ  H   2
H
J  ρ J  ρ J  (ρ  ρ) J  d
Therefore total H field is H  2 
2 2 where
2 d is a vector
directed from the centre of the conductor to that of the cavity.

2.7 Magnetic vector potential

Since   B =0 and from the vector identity we know, the divergence of the curl of a
vector is zero, we can define B=   A where A be a vector, called the vector magnetic
potential. With the help of the previous section, we have

µI dl  a R µ Jdv  a R µ µ
B  Jdv  (
1  )
  vector 4 fF=fF 1
2
From4the
R
2 identity
R 4 we have
+ fF  ( )  Jdv R
4
 R
µ 1 1
B  4 [  ( RJdv) - ( R)  Jdv]
The second term is zero, because  is a function of (x,y,z) and J is a function of (x, y, z).
µ 1 µ 1
B
4
  ( Jdv)   [
R
 ( 4Jdv)]
R
µ 1 µ 1 (51)
hence A(r)   J(r )dv(r )   I(r)dl(r ) 4
R 4
R

Let us now consider the divergence of A. We have

µ 1
  A(r)
from  4   J(r
eq. (51) R )dv(r )
Using the vector identity   ( fF)  f  F  F  f the above equation takes the form

µ 1 1 (52)
  A(r)  4 [J(r )  ( R
)  R  J(r)]dv(r )
The second term of eq.(52) is zero; therefore eq.(52) becomes

µ 1 µ 1
  A(r)  4 [J(r )  ( )]dv(r ) 
R   A(r)   4
 [J(r )  ( )]dv(r )
R
1R 1 R
because    .
µ 1 J(r) From continuity equation the 1st term
  A(r)  4 [   J(r)   
R )]dv(r
R )
is zero.
 µ J(r) µ J(r) As the surface s(r) must
   A(r)     dv(r )     ds(r ) 4
R 4
include all currents, there will be
R no current through the surface s(r). Therefore
J  ds(r)  0    A(r)  0
(53)
From Ampere’s law
  H  J    B  J      A  J
Using the vector identity and eq. (53) we have
(54)
 2 A  J
Now    B  ds   (  A)  ds   A  dl (55)
Example 2.13 A direct current I flows in a straight wire of length 2L. Find the magnetic
flux density B at a point located at a distance  from the wire on the bisecting plane by
determining the vector magnetic potential A first.
Solution: At a height z along the wire above the bisecting plane assume a current element
Idl= Idz az. A is then given by

L
Idz ⎛ 2
L   2 L ⎞ z
A 
4 z  z  
2 I 2 a z  4 ln⎜⎟a2 . 2
⎜ L    L ⎟⎠
L
⎝
B is obtained from
a ρ aφ a z
1    ⎛ 2
Az aφ    ⎢I ⎜ L    L ⎞⎤ φ
⎡ 2
BA 
    ln ⎜ 2
  ⎢⎣ ⎝ L   2  L ⎟⎠⎥
A  4
⎟ ⎥a
z ⎦
IL φ A
 a
2  L 2   2 Az
Example 2.14 A very long straight conductor located along z axis carries a current I in the
z direction. Obtain an expression of A at a point on the bisecting plane of the conductor.
Also determine B at that point.
Solution:
The conductor is assumed to be extending from z=-L to z=L.
The vector R from the current element to the field point is
given by R= a - z az. Now A at P is given by
µ
1
A Idl where dl  dza zand R   2  z2
4  R
 A  µI zdz a
L

4 L
 2  z2
Assume z=tan  dz=sec2 d and (2 +z2)1/2=sec
 dz  2  z2  z )
   sec d  ln(sec  tan )  ln(
 2  z2 

Therefore A  I ln   L  L a z
2 2

4   L2  L
2

 2  L2  L  L  L[1  2 1] / 2  L  L[1   2 ]  2L    2L
2 2
But
L2 2L
2L 2
and  2  L 2  L  L  L[1   2 1/] 2  L  L[1   2 ] 
 L2 2L
2 2L
I 2L I 2L
 A  4ln a2 z ln2 a z

 2L
Now as B=A
aρ aφ a ρ aφ az
az 1   
1   
I a φ
BA       z  Aa
zφ

A    0 2
Example 2.15 TheA inner

z
conductor of a 100m0 longAcoaxial
z cable has a radius of 1cm and
carries a current of 80 A in theAz z direction as shown in the following figure. The outer
cylinder is very thin and has radius of 10 cm. Calculate the total flux enclosed within the
conductors.

Solution: We can use (55) and A or (55) and B to determine flux in the following way:
   A  dl or    B  ds
If we use A, then
   A  dl   A  dl   A  dl   A  dl   A  dl   A  dl   A  dl as
there will be
c1 c2 c3 c4 c1 c3

no contribution from the integration along the paths c2 and c4. Considering the cable length is
long in comparison to the separation between the conductors, we may assume A along c1 and
c3 I 2L example I as
A following 2.13 2L
2ln aa and A  ln2 a b
z z

L
I 2L L
I 2L
L
I 2L L
I 2L
   L ln a z  dzaz   Lln a z  (dz)(az )   ln dz   ln  dz
2 a 2 b 2 L a 2 bL


IL b ln


a I b L
I
ddz IL b
  B  ds  a z   ln because here ds  ddzaφ and B 
aφ
Again from 2 L   2
Example 2.16 If B=Baz, compute the magnetic flux passing through a hemisphere of
a
radius R centered at the origin and bounded by the plane z-0.
Solution:
We know    B  ds where ds=r2 sindd. In this problem
B=Baz and r=R.
   B  ds    Ba z / 2 2
 R sin  dda
2
r
 0  0

Now azar=cos

 / 2 2 /2
/2
cosd  R B  sin 2d  R B
r  sin
 0  0 2  0 2  0 2
2
Example 2.17 Two very long identical and parallel conductors carrying 1 kA in opposite
directions are strung on poles 100m apart. If the radius of each conductor is 2cm and the
separation between their axes is 1m, determine the flux passing through the region
bounded by the conductors and the two consecutive poles.
Solution:

Here a=2cm, b=1m, L=100m, I=1kA. Magnetic

flux density B at a distance y from the conductor at
y=0 is

B
I ⎡ 1  ⎤
⎢ (a x)
1 ⎣
2 y b  y
⎥ ⎦
   B  ds where ds  -dydzax
Therefore
 B
I ⎡ 1  ⎤
⎢ (a x)
⎣
2 y b  y
1
⎥
⎦
L ba
I ⎡ 1 1 ⎤ ⎡b  a ⎤
   z0 y ⎢⎣  IL⎥dydz   a ⎥⎦  155.67
a
2 y b  y
⎦ mWb
ln
8.Magnetic
⎢⎣ force acting on a differential current element:
Ampere’s force law
A charged particle in motion in a magnetic field of flux density B is found experimentally
to experience a force which is proportional to the charge q, it velocity u, the flux density
B, and to the sine of the angle between the vector u and B. The force acts in the direction
of uxB. The force may therefore be expressed as F=quxB. If an electric field E is also
present then the total force is the sum of the electrostatic and magnetic forces
F=q(E+uB).
Assume an element of current, as a length dl of wire carrying a current I. The cross-
sectional area of the wire and the density of charge in the wire are considered S and v
respectively. Assume this charge moves with velocity u. Then the charge crossing any
cross-section in one second=uSv=I. Idl= uSdlv=uq. The force on the differential
current element due to magnetic field only can be expressed as
dF=quxB=IdlxB. (56)
For a real circuit L the total magnetic force is given by the appropriate integral of the
forces acting on the individual current elements. Equation (56) is known as Ampere’s
force law.

9. Magnetic forces between current carrying elements

The B-field produced by one current element exerts a magnetic force on the other current
carrying element (and vice-versa) placed in B.
Let us consider two current elements I1dl1 and I2dl2 are separated by a distance R as shown
in the following figure:

z
R
aR R I2dl2
I1dl1 r r
y

x
The magnetic field at the position of the current element I1dl1 due to the current element
I2dl2 is found from eq. (37) as
 I 2 dl 2  a R where a R  R  r  r where r and r are the position vectors of
dB 
4 R2 R r  r
the current elements I1dl1 and I2dl2, respectively.
Following eq. (56), the force d(dF1) on the differential current element I1dl1, due to the
differential current element I2dl2 is therefore given by
d(dF1)= I1 dl1 x dB
 µI1dl 1 (I dl2  2a )
 d(dF1)  R

R
Total force on circuit4L
2
1, due to the circuit L2 is then given by
µI I (dl 2  a
F1  1 2  [dl 1   Newton (57a)
4)  L L
R2
R
] 1 2

Using the vector identity A(BC)= B(AC)- C(AB), we obtain from (57a)
⎤
F1  µI1 I2  [ dl1  (dl 2  R) ]  µI1I2 ⎢⎡ dl1 3 R dl2   dl1  3dl 2 R ⎥
4 L1 L2 R3
4 ⎣ 1 2 R L1 L 2
R ⎥
⎢L L
⎦
As because R/R3= -(1/R), the first integral on the right side of the above equation can
be write as follows with the help of Stock’s theorem.
⎡ ⎤1 ⎡ ( 1 )  ds ⎤
 ( )  d1l ⎥ dl 2     l ⎥d 2

L1 L 2 ⎢⎣ R ⎦ L 2s1 
1⎦
R
The above integral is equal to zero as the curl of a gradient of a scalar function is zero.
⎢⎣ the magnetic force on circuit 1 due to current in circuit 2 is also given by
Thus
d l 1  dl 2
F1  µI1 I2
4 
L L
R R (57b)
1 2
3

Example 2.18 A bent wire as shown in the following figure lies in the xy plane and
carries a current I. If the magnetic flux density in the region B=B az, determine the
magnetic force acting on the wire.

Solution: Let F1, F2 and F3 are the magnetic forces act on the sections of the wire from x=-
(a+L) to x=-a, from x=a to x=(a+L), and on the semicircular section of radius a,
respectively. Now
a a
F1   Idxax  Baz    BIdxay  BILay
(a L) ( a L)
(a L ) (a L)
F2   Idxax  Baz    BIdxay  BILay
a a
0 0 0 0
and F3   Idl  Baz   Iadaφ  Baz   IBad(aρ )  IBa(cosax  sin ay )d
   
 2IBaay
 F  F1  F2  F3  2BI (a  L)ay
Example 2.19 The following figure shows a current-carrying conductor of finite length L
placed at a distance b from another current carrying conductor of infinite extent.
Determine the magnetic force per unit length acting on the finite conductor.

Solution: From example 2.1 we have B at a distance b from a current carrying

conductor of infinite extent is
I
B  0 aφ. Now from (56), the force on element dz is given by
2b
2
F  Idl  B  Idz a z
a dza ρ
2b φ 2b
dza  0I La
2
F
 I
L/2 0 2

2b
ρ

L/22 b ρ

I2
Now the force per unit length  20b aρN / m
Example 2.20 A rectangular loop carrying current I2 is placed close to a straight conductor
carrying current I1, as shown in the following figure. Obtain an expression for the
magnetic force experienced by the loop.

Solution: The total magnetic force acting on the loop can be expressed as the sum of the
forces on sections AB, BC, CD, and DA. The incremental segments for sections AB and
CD are along z-axis, therefore, their dot product with the segment dz1 exist. However, the
segment dz2 for sections BC and DA is perpendicular to dz1, thus their dot product with
the segment dz1 is zero do not contribute to the force on the loop.
Let us first determine the magnetic force on section AB. The vector R is given by
R=bay - (z1 – z2)az. From (57b), we obtain the force on section AB of the loop is
FAB   µI I L a
[bay  (z1  z2 )az ]
  2
1 2
dz1
4 z  L
1 z  a [b  (z
2
1  z2 ) ]2 3 / 2 dz2

FAB  
µI1I2
4
I1 I 2
L a
bdz 2
where I1  
z  L
dz1  2
z  a [b  (z 1  z 2 ) ]2 3 / 2 y
a
1 2

L a
(z 1 z 2)dz 2
and I 2   dz1  [b  (z  z ) 2 ]3 / 2
2
az
z1  L z2  a 1

To solve I22, let us consider z1- z2= b tan , therefore dz2= -b sec2 d.
L a
a L
 I 2    dz1 sin dz1
z1 L

z 2 a
db
az  
z1 L [b  (z 1  z 2 ) ]2 1 / 2
2 az
z2  a

a L
 ln[((z  z )  b 2  (z  z ) 2 ] a z 0
1 2 1
2
z2  a z  L
1

To solve I1, let us consider now z1- z2= b tan , therefore dz2= -b sec2 d.
a
cos d
L a L
( z1  z 2)dz 1
I 1   dz1  b
ay    b[b2  (z  z ) 2 ]1 / 2
ay
z1  L z2  a z1  L 1 z2  a

Put again z1- z2= b tan , therefore dz1=b sec  d.

2 2

a L
L a

 I1    tan b  (z  z )
2 2

 1 2 ay
secday z1 L
z2  a b z 2 a z  L
1

b 2  (L  a) 2  b 2  (L  a) 2
2 ay
b
µI I µI I ⎡ b 2  (L  a) 2  b 2  (L  ⎤
 FAB   1 2 I 1  I 2   1 2 ⎢ a) 2 a y⎥
4 2 ⎢ b

⎣ ⎥⎦
The force FAB is attractive in nature as the term within the bracket is positive.
Let us now determine the magnetic force on section CD. The vector R is given by
R=cay - (z1 – z2)az. From (57b), we obtain the force on section CD of the loop is
µI1I2
L a
[cay  (z1  z2 )az ]

4 z1 L
dz1 
z 2 a [c  (z 1  z 2 ) ] 2 3 / 2
2 dz 2

FAB  
µI1I2
4
I 1 I 2
L a
cdz 2
where I1  
z  L
dz1  2
z  a [c  (z 1  z2 ) ]2 3 / 2 y
a
1 2

L a
(z 1 z )dz
and I 2   
2 2
dz1 az
[c  (z  z ) 2 ]3 / 2
2
z1  L z2  a 1

To solve I22, let us consider z1- z2= c tan , therefore dz2= -c sec2 d.
 L a
a L
 I 2    dz1 sin dz1
z1 L

z 2 a
dc
az  
z1 L [c  (z 1  z 2 ) ]2 1 / 2
2 az
z2  a

a L
 ln[((z  z )  c 2  (z  z ) 2 ] a z 0
1 2 1
2
z2  a z  L
1

To solve I1, let us consider now z1- z2= c tan , therefore dz2= -c sec2 d.
a
cos d
L a L
( z1  z 2)dz 1
I 1   dz1  c
ay    c[c 2  (z  z ) 2 ]1 / 2
ay
z1  L z2  a z1  L 1 z2  a

Put again z1- z2= c tan , therefore dz1=c sec  d.

2 2

a L
L a

 I1    tan c 2  (z  z ) 2
 1 2 ay
secday z1 L
z2 a c z 2 a z  L
1

c 2  (L  a) 2  c 2  (L 
 2 a) 2 ay
c
µI 1I 2 µI 1I 2 ⎡ c 2  (L  a) 2  c 2  (L  ⎤
 FCD  I1 I 2  ⎢ a y⎥
4 2 ⎢ a)
2
c
⎣ ⎥⎦
The force FCD is repulsive as the term in the bracket is positive. The total force on the
loop is therefore given by
F  FAB  FCD
⎡ 2
µI I b  (L  a) 2  b 2  (L  a) 2  c 2  (L  a) 2  c 2  (L  a) 2 ⎤⎥a
  2⎢
12
b c ⎥⎦ y
⎢
As because⎣ c>b, the force on the loop is attractive.

2.10 Magnetic torque

A current carrying conductor when placed in a magnetic field experiences a force that
tends to move the conductor in a direction perpendicular to both the magnetic field and
the conductor. However, if a current carrying coil is placed in the magnetic field, the
magnetic force acting on the coil may impart a rotation to the coil.
The operation of electric motor and moving type of electric meters in facts depends on
this principle.
Consider a current carrying rectangular coil in a magnetic field as shown in Fig. (a).
According to Ampere’s force law, no force will act on the coil sides bc and da. However,
sides ab and cd experience forces Fab=-ILB ay and Fcd=ILB ay, respectively. Figure (b)
shows the forces acting on the sides ab and cd of the coil. These forces exert a torque that
tends to rotate the coil about its axis. The torque experienced by the side ab is given by
Tab= (axW/2)xFab=(-WILB/2)az and that experienced by the side cd is given by Tcd= (-
axW/2)xFcd=(-WILB/2)az. Therefore, the net torque acting on the coil is T= -WILBaz. In
terms of magnetic dipole moment m=IWL ay, defined as the vector whose magnitude is
the product of the current through a conducting loop and the area of the loop and whose
direction is the direction of the thumb as the fingers of the right hand follow the direction
of the current, the torque is given by T=mxB.
Under the action of this torque, the coil begins to rotate. However, the torque becomes
zero when the plane of the coil is normal to the direction of the magnetic field.

Example 2.21 A circular coil of 200 turns has a mean area of 10 cm2, and the plane of the
coil makes an angle of 300 with the uniform magnetic flux density of 1.2 T, as shown in
the following figure. Determine the torque experienced by the coil if it carries a current of
59 A.
Solution:

Coin in B field Side view of

the coil
The side view of the coil indicating the direction of the dipole moment is shown in Figure
(b). It lies in the xy-plane and has magnitude of m=NIA=200501010-4=10 Atm2. The
torque experienced by the coil is
T  m  B  10 1.2 sin 60o a z  10.39az N  m.

2.11 Boundary condition for static magnetic field

1 B1 dS K H1
dt  dh
2 dS
dL
For the normal field components,Hlet
2 us assume a small cylinder with its top face in
medium 1 and bottom face in medium 2. The faces have area dS, and the height of the
cylinder dt is vanishingly small. Appling divergenceless nature of B field to the cylinder,
we have
  B  ds  (B  a 1 n1  B2  an 2 )dS  (Bn1  Bn2 )dS  0
S (58)
or B n 1  B   H   H
where an1 and an2 are the unit vectors perpendicularly outwards to the top and bottom
n2 1 n1 2 n 2

surfaces of the cylinder, respectively.

The boundary condition for the tangential components of magnetostatic field is obtained
from
 H  dl  I
L

Let us take the rectangular path as the closed loop. If its height h is made infinitesimally
small then the only contribution to the line integral is along the top and bottom edges of
length dL. A sheet of current having a linear current density of K A/m width is also
assumed at the interface of the two media.

Then  H  dl  H 1  (dL)  H 2  (dL)  H1t dL  H 2t dL  KdL

L

leads to H2t-H1t=K (59)

The tangential component of H is continuous (K=0) across the boundary of almost all
physical media; it is discontinuous (K0) only when an interface with an ideal perfect
conductor or a superconductor is assumed.

Example 2.22 The magnetic flux density in a finitely conducting cylinder of radius 10 cm
and with a relative permeability of 5 is found to vary as 0.2/ a T. If the region
surrounding the cylinder is characterized by free space, determine the magnetic flux
density just outside the cylinder.
Solution: The interface is at =10 cm, the B in the cylinder just beneath the boundary is
given by 2
B c  0.2 a φ  2aφ T.  H c aφ  318.31aφ kA/m
0.1 5  4 107

As the magnetic flux density is tangential to the boundary and K=0 for the finite
conductivity of the cylinder, from (59) we obtain
H a  318.31aφ kA/m

2.12 Energy stored in static magnetic field and inductance

The energy stored in a steady magnetic field is given by

2  dS  I(
2  A)  dS
1 1 1 1
Wm I22 L  I2  IB

Using stock's theorem we have m  1

2  IA  dl  A  Jdv
1
W
Now A  J  A  (  H)  H  (  A)    (A  H)  H  B    (A
 H) 2
Wm  1 
2 A  Jdv  2 [H  B    (A  H)]dv 2 vol
1
H  Bdv 2    (A 
1 1
vol
H)dv
1 1
  H  Bdv 2 s (A  H) 
2 vol 2
As
dS H varies inversely as square of distance (1/R ), A varies as (1/R) and the surface area
increases as R2, it follows that the second term in Wm is a function of (1/R). In order to
consider all the fields in calculating Wm, we have to consider a very big closed surface
with R, the second term in Wm thus becomes zero.
1 1
W m   H  Bdv   H2dv   B dv
2
(60)
12 2 vol 2 vol
vol

Example 2.23 Calculate the energy stored in the magnetic field of the toroidal winding
discussed in Example 2.11.
Solution: H inside the toroid is
NI
H  2 a φ for a    b as obtained in Example 2.11. Now from (60) we obtain
b 2 2 h
1 1 1
2
2
H dv   dddz ⎛ NI ⎞ b   dddz
2
Wm 
h
2
⎛⎜ NI ⎞⎟ 1
 a  0 z⎜02 ⎟ ⎠ ⎜
 vol
2  2 ⎝ 2  a  0 z
⎝ ⎟ 0

 N I h lnb / a
1 2 2 ⎠
4

2.13 Inductance
The self inductance of a current carrying loop is defined as the magnetic flux linkage per
unit current in the loop itself, i.e., L=/I. The self inductance can also be calculated from
Wm=I2L/2.
The mutual inductance between two circuits is the magnetic flux linkage with one circuit
per unit current in the other. If I1 current in circuit 1 makes a flux linkage 12 in circuit 2,
the mutual inductance L12 between circuit 1 and 2 is given by L12= 12/I1. If B1 be the flux
density due to current I1, the flux linkage of circuit 2 is given by
12  N 2  B1  ds2 . Therefore, L12  2  1
N B  ds2

s2
s2
I1  A )  ds
N 2  (
s2 1 2

I1
Using Stock's theorem, we have 12 
l 2

L I1
N 2  A1  dl 2
Now from (51) we have A 1  µN1 I1
4 l dl1
R
1

Where N1 and N2 are the no. of turns in circuit 1 and 2, and R is the distance between the
incremental segments dl1 and dl2.
N 2  A1  dl 2
N 1N 2 dl 1  dl 2

l2
Therefore L12   (61)
I1 l1 l 2
R
4
Equation (61) is known as Neumann formula.
Comparing eqs. (57b) and (61), we may prove that the magnitude of the force between
two conductors can be written as
dL12
F  I 1I2
dR
(62)
Example 2.24 Self inductance of some typical geometry without considering internal
inductance.
Solution: 1) Self inductance of a long solenoid
In case of long solenoid, the magnitude of B field in the solenoid is B  IN . If the
cross-sectional area of the solenoid is A, then the flux linkage of the coil is
=N=NBA=IN2A. Therefore the self inductance of the solenoid L=/I=N2A Henry.
2) Self inductance of a Toroid of circular cross-section

For a circular path of radius =R, the ampere’s circuital law gives us

 H  dl  NI NI
 B  2  Ra 
If the cross-sectional area of the toroid is A, the flux linkage of the coil is =N=NBA=
IN2A/2R Therefore the self inductance of the toroid L=/I=N2A/2R Henry.
2) Self inductance of a Coaxial cable
dr
Assume the current through the inner
conductor is I. Then the magnetic field dz
at a position  meter away from the
centre of inner conductor is

 H  dl  I I
 B  2 a 

The magnetic flux that thus produced

I
crosses the area drdz, is d  2ddz
The flux linkage per unit length of the cable is
1 D
I
  z 0  ddz.
d
2  
I 
 2ln(D / d ) and the inductance per unit length is L  ln(D / d ) Henry/m
2 unit length inductance of two-wire transmission line
2) Per
Assume the transmission line is along the z-direction and one at (x=0, y=0) carries a
current of I amp and the other at (x=0, y=R) carries –I amp. The radius of both conductors
is assumed to be d. In this case, the magnetic field B and hence flux crosses the yz-plane
in between the conductors will be additive and is directed along –x axis.
 R

dy

dz

I I
B  [ 2y  2 (R]a  x

y)linkage per unit length of the line is

The flux
1 Rd
I I I
]dydz  ln{(R  d ) 2 / d 2 }  ln{(R  d ) /
   [  I
z 0 y d 2y
d}.
2 (R  y) 2 
Therefore the inductance for unit length of the two wire transmission line is L=(/)
ln{(R-d)/d} (/) ln(R/d) H/m.
Example 2.25 Find the magnetic field at radius r within a solid copper conductor of
radius ro>r carrying current I uniformly distributed over the cross-section. Also determine
the internal inductance of the conductor.
Solution:
The current that crosses an area of r2=I r2/ r02. Therefore on applying the ampere’s
circuital law along the closed circular path of radius r is
r 2  r 2
 H  dl  ( 0 r B
) I 2 
0
( ) Ia  
Now the storage magnetic field is obtainedIaasr
1 2r r0 2 I 2 z J  L   H /
Wm   B  1   1 ⎛⎜ 2rr2 I ⎞⎟ rdrddz 
2 r 2 z
0

2v  2 0 0 0  ⎝ 2r0 16 m
8
Example⎠ 2.26 Derive an expression for the magnetic field at a point P(x, y) on the x-y
plane due to two long straight parallel wires directed along z-axis and placed at y=d/2 and
y= -d/2. Assume the currents in the wires are opposite to each other.
Solution:
Let us first consider the field due to a single infinitely long conductor which is placed at
y=-d/2 as shown in the following figure:
z

I
dz r1={x2+ (y+d/2)2+ z2}1/2={ 21+ z2}1/2
-I
z r1
-y y
1
P(x
,y)

x
 I
Using the result for infinitely long wire we have B  2a . It can be expressed in
rectangular coordinate system as B  B x a x  B y a y  B z a z where

B x  I a   x I I a  I I a  0
2 y a   2sin, B  a  cos andB z  2a
y  
2 z
2
For this particular problem, the field due to the left conducting wire is
B x   I y  d / 2 , y I x
21 1 z
 2  andB  0
1
1
AndBthe field due to the right conducting wire is
B x   I y  d / 2 , I x
y 21  2  andB z 0
1 1
Example
B  1
2.27 Calculate the internal inductance of a thick conducting cylinder of inner
radius a and outer radius b, as shown in the following figure.
Solution:
Assume that a current I be flowing through the
cylindrical tube and having a uniform current
density through out the cross-section. The
current density J is then
I
J a.
 (b  a 2 ) z
2

From ACL, we obtain  H  dl  I   J 

ds
l s
 2 I I ( 2  a
H will be along  direction;
) 2 a z  ddaz  (b2of radius
if a circular path  (a    b), we obtain
H 2  a 0
2
a 2)
 (b  a )
2

I ( 2  a 2 ) I ( 2  a
H  2 ) 2 H a
2(b  a 2 ) 2(b2  a 2 ) φ

Wm  1  ⎢⎡ I (2  a2 ) ⎤dv  1 ⎢⎡   ⎡(

2 2 b 2 z 2
2 2
⎤ 2
 a 2 )⎤
   2
 2 ⎥
 dddz
 2
I v ⎥⎣ 2 (b a )  2 ⎣ 2 (b a )
⎢ ⎦⎥
1
⎦
I
⎦  a00⎣

⎤ b 2 z ⎡
2
  a 4⎤
⎡2 ⎣ 2 (b  a ) ⎥⎦    ⎢  2 a   ⎦
2 2 3 2

a00⎣
⎢

⎡ I 
⎥⎤d⎡db
2
dz4  a  ab 2  a2  4a bln
⎢
z ⎣⎢4 2 (b  a )⎦
2 2 ⎥
4 2
 ⎤⎦
⎥
⎣  ⎡ b  a ba
 a 2 b 2  a 2  a 4 ⎤ H / m
4
L

42 (b2  a )2 2 ⎢⎣ 4
ln
a
⎥
⎦
Example 2.28 Show that the mutual inductance between a straight long conductor and a
coplanar equilateral triangular loop shown in the following figure is
0 ⎡
a  bln a b b  a⎤⎥
3
⎢⎣ ⎦
Solution:

600

x
r

600
b a

I
B at a distance r from the straight long conductor is B  aφ
2 r

I
ab

Therefore,   s B  ds  b 0 2xdr
r
However, (x/2)=(a+b-r)tan
2 ab I 300  x=(2/3){a-(r-b)}
 1 0 I
ab
⎧a  b a
 I ⎫
  
0
a  (r  b)dr  1⎬dr  0 ⎧
3 b 2 r  ⎫
 ⎩⎨ r ⎭ ⎩ b b  a⎬
⎭
3  b

0 ⎨(a  b) ln 3
 M   ⎬
⎧ (a  b ln a ⎭
b)
⎨
b I
 a⎫
3 ⎩
 M. M. ALI
Page
No.:1/11
TIME VARYING FIELDS

Introduction

In this chapter we will find that a time varying electric field can be produced by a time
varying magnetic field and vice versa. In fact, if there exists a time varying electric
(magnetic) field in a region, there also exists a time varying magnetic (electric) field in
that region. We will refer to an electric field due to time varying magnetic field as an
induced electric field. The line integral of such an induced electric field is called the
induced emf.

3.1 Motional electromotive force

Let us consider a conductor moving with a uniform velocity
u in the magnetic field B as shown in the adjacent figure.
The magnetic force acting on each of the free electrons in
the conductor is F  eu  B  euBay , where –e is the
charge of an electron. Under the influence of this force the
electrons within the conductor will move from right to left.
Such a migration of electrons will result in a net negative
charge at the left end of the conductor and a net positive
charge at the right end.
The workdone by an external source to move the electron a distance dl is dWm= -F•dl=
e(uB)•dl =eE•dl (The electric field is defined as the force on a positive unit charge and
hence E=uB). Therefore the voltage difference between two ends of the conducting bar,
named as induced voltage, is given by

em  e m  E  dl  (u  B)  dl .
dW (61a)
As this induced voltage em is due to the motion of conducting bar in the magnetic field, it
is called the motional electromotive force.
Let us now consider the case where a conductor is sliding freely over a pair of stationary
conductors as shown below. The electric circuit is closed by connecting a resistance
between the far ends of the two stationary conductors.

a
For the considered problem em   (u  B)  dl   uBdy  uB(a  b)  (61b)
b
uBL
which indicates that the terminal b is at higher potential than a. As in calculating voltage
we have to move a unit positive charge against E, the integration is carried out from b to
a.
 M. M. ALI
Page
No.:2/11

The voltage em induced in the bar results in a current I in the circuit which is known as
induced current. This current produces a magnetic flux that will oppose the original
magnetic flux B and satisfies Len’s law.

3.2 Maxwell’s equation from Faraday’s law of electromagnetic

induction
As the sliding conductor in the above figure moves a distance dx in the x-direction in time
dt, the increase in cross-sectional area made by the circuit is given by ds=Ldx. Now the
change in magnetic flux passing
d through
dx the plane ds  Ldxa z is
d  B  ds  BLdx  dt  BL dt BLu
Therefore from (61b) we obtain equation of the induced voltage as
d
ei   dt .
(62)
Equation (62) is a mathematical definition of Faraday’s Law of electromagnetic
induction. It states that the induced emf around a closed path is equal to the negative rate
of change of the magnetic flux with respect to time passing through the area enclosed by
the path.
In (62) ei=  E  dl and    B  dS
S

dB (63a)
  E  dl  S dt 
dS
Using Stokes’s theorem the above equation can be written as
dB (63b)
  E   dt
It becomes E=0 for time invariant or dc field. We identify eqs. (63a) and (63b) as
Maxwell’s equations in integral and differential or point forms, respectively.
The induced emf in a stationary circuit caused by a time-varying magnetic field can also
be calculated from (62) as
d d
e i  et    B  ds
s
dt dt
(64) et is termed as transformer emf.
where
In case of moving circuit in a time varying magnetic field, the induced emf is therefore
given by

B  ds  (v  B)  dl
d (65)
ei  
. dt 
s L

d
  E  dl  dt s  B  ds L  (u  B) 
Now the application of Stoke’s theorem yields
dl

s (  E)  ds  dt  B  ds s   (u  B)

d
s

ds
d
  E   dt B    (u 
B)
 M. M. ALI
Page
No.:3/11

This equation is the most general form of Maxwell’s equation (Faraday’s law) in the point
form.

3.3 Displacement current and Maxwell’s equation from Ampere’s

Circuital law
The ampere's circuital law for static field is no longer useful in time varying fields as is
shown below:
Taking the divergence of the ampere's circuital law give by eq (49) and using the
continuity equation (32) we have
d d
  (  H)    J   dt v  dtv  0
That is, equation (49) leads to steady state conditions in which charge density is not time
varying function. Therefore, for time dependent fields eq (49) needs some modifications.
Maxwell suggested that the definition of total current density of ampere's circuital law is
incomplete and advised to add something to J. If it is assumed to be J, eq (49) becomes
  H  J  J
d
   (  H)    (J  J)  0    (J  J)    J    J  dt v

d (  D) dD
   J  dt  J  dt
Since J' arises due to the variation of electric displacement (electric flux density) D with
time, it is termed as displacement current density. It is as effective as J. The modified
Ampere's circuital law (Maxwell’s equation), therefore, for time varying field takes the
following form
dD
HJ
dt (66a)
Applying Stoke’s (66a) can be given in integral form as

dD dD (66b)
L H  dl  s (J  dt )  ds  I  sdt 
The
ds important conclusion that can be drawn from eq (65) is that, since displacement
current is related to the electric field, it is not possible in case of time varying fields to
deal separately with electric and magnetic fields but, instead, the two fields are interlinked
giving rise to electromagnetic fields. It is to be noted that, in a good conductor J' is
negligible compared to J at frequency lower than light frequencies (1015 Hz).

Example 3.1 A copper strip of length L pivoted at one end is rotating freely with an
angular velocity  in a uniform magnetic field, as shown in the following figure. What is
the induced emf between the two ends of the strip?

Solution : The velocity at any radius ρ is u=ρ a.

Therefore, the induced electric field E=uB= ρB a.
Using (65) we obtain
0
d
eiab    s B  ds  L (u  B)  dl L  Baρ 
dt
daρ1
  2 BL2
 M. M. ALI
Page
No.:4/11

The result shows that the b end is at higher potential than the pivoted end.
Example 3.2 A copper strip of length 2L pivoted at the midpoint is rotating with an
angular velocity  in a uniform magnetic field as shown in the following figure.
Determine the induced emf between the two ends.
Solution:

The problem is similar to that in the previous

example except that we have to consider two arms of
equal length L pivoted at the centre. Therefore two
are voltage of 12BL2 volts with respect to the
tipsequal
at
midpoint. Therefore the potential difference i.e., the
induced emf between the ends is zero.

Example 3.3 A copper conductor is placed in such a way that the magnetic field is
perpendicularly outward to the surface contains the conductor as shown in the following
figure. Determine induced emf e12.
Solution:
B= B0 sin t az

1 y
B b
e12
2 x

a z
Using (65) we obtain
d dab
e12    sB  ds   (u  B)  dl
L
   B0 sin 0ta
0 z
 dxdyaz  B0 ab
dt dt
The
cospolarity
t of the terminals will be such that, the current due to this voltage will set up a
magnetic field which will oppose the original field.
Example 3.4 A circular conducting loop of radius 40 cm lies in the x-y plane and has a
resistance of 20 . If the magnetic flux density in the region is given as B=0.2 cos 500t
ax +0.75 sin 400t ay+1.2 cos 314t az Tesla. Determine the effective value of the induced
current in the loop.
Solution: The differential surface area of the loop is ds=dd az. Therefore from (65)

B  ds  (u  B)  dl   B 
d d
ei  
ds dt 
s L
dt s
d
Now dtB  100 sin 500ta x 300 cos 400ta y 376.8sin 314ta z

0.42
d
 ei   s B  ds  376.8   sin 314t0d0d  189.4 sin
dt
and i  9.47 sin 314t
314t
Therefore the required result is 6.7 Amps.
Example 3.5 The magnetic field in free space is given by H=H0sin ay A/m, where =
t-z, and  is a constant. Determine the displacement current and E.
 M. M. ALI
Page
No.:5/11
Solution: As the conduction current J in free space is zero, we obtain from Maxwell’s
equation that
ax ay
az
dD  dD  d d
0 a  x H sina
  H  dt d d   dzH sin dx
0 z

d
dt dx dy dz
 H 0 cosax A/m 2
0 H 0 sin
1
Again D   H 0cosdta x  1 H00 sina x  E  
D H 0 sinax V/m
 0 0

3.4 Maxwell’s equations in time domain

By studying the physical properties of electric (E-field) and magnetic (B-field) fields we
have been able to describe these properties by four, relatively simple, equations known as
Maxwell’s equations.
These four fundamental equations of electromagnetism can be expressed in both an
integral and differential form as tabulated below:

Sl No. Point/Differential form Integral form Equation No.

1) D=v  D  ds    v dv (67)
s v

2) B=0  B  ds  0 (68)
s

3) dB dB (69)
E  E  dl    ds
dt L s dt
4) dD dD (70)
HJ  H  dl   (J  )  ds
dt L s dt

Equation (67) results from Coulomb’s and Gauss’s laws and states that free charges act as
sources or sinks of D. It suggests that the total electric flux density or total electric
displacement through the surface enclosing a volume v is equal to the total charge within
the volume.
Equation (68) arises from the application of Gauss’s law to magnetic fields and the non-
existence of magnetic monopoles. There are no sources or sinks of B. This equation
suggests that the net magnetic flux emerging through any closed surface is zero.
Equation (69) describes Faraday’s law of electromagnetic induction and states that an
electromotance is produced in a circuit when the magnetic flux through the circuit
changes. It suggests that the electromagnetic force around a closed path is equal to the
time derivative of the magnetic flux density through any surface bounded by the path.
Equation (70) describes Ampere’s circuital law (which is derived from the Biot-Savart
law) and states that the magnetomotive force around a closed path is equal to the
conduction current J=E plus the time derivative of the electric flux density through any
surface bounded by the path.
In vacuum/free space v=0, J=0, =0 and =0. Therefore, in vacuum the Maxwell's
equations take the following forms:
 M. M. ALI
Page
No.:6/11

Sl No. Point/Differential form Integral form Equation No.

1) D=0  D  ds  0
s

2) B=0  B  ds  0
s
dH
3) E
dB
 
dH
 E  dl    0
 ds
dt dt
L s

0
dt
dE
4) H
dD

dE
 H  dl    0
 ds
dt 0 dt
L s
dt
3.5 Maxwell’s equations in frequency domain

Assuming the fields are varying harmonically with time as ejt, the maxwell’s equations
are given by

Sl No. Point/Differential form Integral form Equation No.

1) D=v  D  ds   v dv (71)
s v

2) B=0  B  ds  0 (72)
s

3)   E   jB  E  dl   j  B  ds (73)
L s

4)   H  (  j )E  H  dl  (  j ) E  ds (74)

L s

3.6 Maxwell’s equations in phasor forms

In case the field quantities are sinusoidally time varying, the electric field E can be
expressed as
E(x, y, z, t)=Ex(x, y, z, t) ax+ Ey(x, y, z, t) ay+ Ez(x, y, z, t) az
where Ex=Exm cos(t+x), Ey=Eym cos(t+y), Ez=Ezm cos(t+z). Here the magnitudes Exm,
Eym, Ezm and the phase angles x, y, z are independent of time but may depend on spatial
coordinates, e.g., Exm(x, y, z), x(x, y, z). Now Ex can be expressed as
Ex= Exm cos(t+x)=Re[Exm ej(t+x)]= Re[Exm ejt ejx] and similarly Ey and Ez. Therefore Ex,
Ey and Ez can be expressed in phasor form as follows: Êx= Exm ejx= Exm x, Êy= Eym ejy=
Eym y, Êz= Ezm ejz= Ezm z. Therefore
E(x, y, z, t)= Exm cos(t+x) ax+ Eym cos(t+y) ay+ Ezm cos(t+z) az
= Re[Exm ej(t+x)] ax+ Re[Eym ej(t+y)] ay+ Re[Ezm ej(t+z)] az
= Re[Êx ejt] ax+ Re[Êy ejt] ay+ Re[Êz ejt] az
= Re[(Êx ax+ Êy ay+ Êz az) ejt]= Re[Ê ejt]
If we remember that the fields are real part of their exponential form, we may consider
the fields as follows:
E(x, y, z, t)= Ê(x, y, z) ejt
Likewise
 M. M. ALI
Page
No.:7/11
H(x, y, z, t)= Ĥ(x, y, z) ejt
D(x, y, z, t)  Dˆ (x, y, z)e
jt

d(x,
B(x, y, z,ˆy,  Bjˆt jt
t) z)e
dtB (x, y, z)e  jB (x, y, z)e that is, the first derivative of a sinusoidal
In this cases, ˆ jt

varying field is j times the field. Therefore, the Maxwell's equations in phasor form can
be expressed as:
Sl No. Point/Differential form Integral form Equation No.
1)
  Dˆ  ˆv  D  ds   ˆv dv
ˆ (75)
s v

2)
  B ˆ 0  Bˆ ds  0 (76)
s

3)   Eˆ  jBˆ  Eˆ  dl   j  Bˆ ds (77)

L s

4)   Hˆ  (  j )Eˆ  Hˆ  dl  (  j ) Eˆ  ds (78)

L s

3.7 Boundary condition for electromagnetic field

For the time varying tangential component of E-field let us apply the Faraday’s law (61)
along the rectangular path as the closed loop in figure (a). If its height h is made
infinitesimally small then the only contribution to the line integral of the left side of (61)
is along the top and bottom edges of length dL and the right hand side becomes zero as the
area dLxh is vanishingly small.

  E  dl  E1  (dL)  E 2  (dL)  E1t dL  E 2t dL  0

L

leads to
 D1t/1=D2t/2
E1t=E2t
(79)
where Et is the tangential component of E.
Hence E1t=E2t; the tangential component of E is continuous across any interface.
For the time varying tangential component of H-field let us apply the Ampere's circuital
law (66) along the rectangular path as the closed loop in figure (b). If its height dh is made
infinitesimally small then the only contribution to the line integral of the left side of
(62) is along the top and bottom edges of length dL and second term of the right hand
side becomes zero as the area dLxh is vanishingly small.

Then  H  dl  H 1  (dL)  H 2  (dL)  H1t dL  H 2t dL  KdL leads to H2t-H1t=K

L
(80)
where a sheet of current having a linear current density of K A/m width is also assumed at
the interface of the two media.
Since the Gauss's law for time varying electromagnetic field for E and B remain same as
those for static field, the boundary conditions for normal components of E and B remain
uncharged.
 M. M. ALI
Page
No.:8/11

t

s

h

(a)

B1 H1
1 dS K
dt  dh
2 dS
dL
(b) H2

3.8 Potential due to time varying field (Retarded Potential)

dB
From Faraday’s law for time varying field (62),   E   dt , and B= A we have
dB d (  A) dA
E     (E  )0
dt dt dt
Now from the vector identity  f=0 or  -f=0, we may define
dA dA
E  V  E  V  (81)
dt dt
In static case dA/dt=0, (81) reduces to E= - V and hence E can be determined from V
alone. For time varying case E is a function of both V and A. In time varying case A can
not be determined from (51) or (55). Similarly, V is not also the solution of 2V= - v/.
From (70) we have
dA
  J  d (V  dt )
dD dE
BJ AJ
 dt
dt dV d 2A
(  A)  2 A   J  (  )
 dt dt dt 2
However, it is easy to prove from (16), (32) and (51) that A= -dV/dt
d 2A
 2 A     J (82)
 dt 2

Again from D= v we have

dA d 2V 
E= v/    (V  ) V
2
  v
(83)
dt  dt 2

The solution of eq (82) andv (83) are of similar type and it is easy to solve them in
frequency domain. Therefore the fields are assumed to be a function of exp(jt). Now
from (82) we have
 M. M. ALI
Page
No.:9/11
  2 A    2 A    J   2 A   2 A    J (84)
where  =  =  /c . Here c is the light velocity in the media whose permittivity and
2 2 2 2

permeability are  and , respectively. To find the solution let us consider a point source
J=Jz az which is function only of radial distance. Then (84) takes the following form
d dA
 2 A   2 A   J 1 (r 2 z )   2 A   J
(85)  r z2 dr
z z z z
At points far from the source
dr (85) takes the following form
1 d (r 2 dA z )   2A z 0 (86)
r 2 dr
dr if we assume Az=f/r, the dAz/dr=r-1df/dr – r-2f and (86) becomes
Now
d 2f
2f 0
dr (87)
2

Since f=C exp(-jr) is the solution of (87), therefore Az= C exp(-jr)/r. In order to
determine C, let us integrate both sides of (85) over a small spherical volume of radius r0.
 Az   Az   J z    Az dv    Az dv    J z dv
2 2 2 2

2
V V V
z J z dv
2

V V V S
V V

(
8
S Az  dSa r  lim
r 0   [  (1  jr0 )C exp( jr0 ) sindd 
8
4C
0
00
)
Again since dv=r2sindd, and Az varies as 1/r, consequently, if r0 is vanishingly small
Now Azar=dAz/dr= - (1+j r)Cexp(-jr)/r2. Therefore in the limit r0 tends to
the volume integral of Az in (88) vanishes. Therefore,
zero
  2  J z exp( j r)dv
C Jz dv  A z 
4 V 4 V
r

 J exp( jr r)dv

 A(r)  4 V
 J exp( jt  jr)dv
 A(r, t)  4 V r
 J exp j(t  r / c)dv  J(t  r / c)dv
 A(r, t)    
4 V r 4 V

r (89)
Similarly, the solution of (83) can be obtained as
1 (t  r / c)dv (90)

V (r, t)  4 V r

3.9 Power flow and the Poynting Vector

The unit of E is volt/m and that of H is A/m, therefore the product of their magnitudes
have the unit of power density. The flow of power due to electromagnetic field in a
particular direction is of prime importance, the vector product of E and H is used to
determine the power of an electromagnetic wave.
If we define the power density vector as
P=EH Watt/m2.
 M. M. ALI
Page
No.:10/11
Now P=( EH)= H(E) - E(H)
Using the Maxwell’s equations we have
P=H(-dB/dt) - E(J+dD/dt)  -P=H(dH/dt) + EJ+EdE/dt Watt/m2 (91)
Eq (91) is known as the point form of poynting theorem. Integrating both sides of (91)
over some volume v and applying the divergence theorem, we obtain the integral form of
poynting theorem as follows


   Pdv 
 Jdv   ⎛⎜ H  dH  E dE ⎞
dt ⎟dv
V E V  V⎝  dt ⎠
⎛ dH
or   P  ds   E Jdv   ⎜ H   E  dE ⎞
 s V  V ⎝  dt
⎠ ⎟dv (92)
The term on the side of (92) is the net inward flux of
watt dt P into the volume v. The first term
on the right side of (92) is a power dissipation term in that it represents the rate of
expenditure of energy by the electric field. The second term on the right side of (92) is
given by
1 2
V ⎝⎛⎜H  dH  E  dE ⎞⎟dv  V dtd (12 H  2 E )dv
2

 dt dt
⎠ represents the time rate of increase of energy stored in the magnetic and electric fields
and
respectively in the volume v.
Therefore (92) states that the net inward flux of the poynting vector through some closed
surface is the sum of the power dissipated in the volume enclosed by the surface and the
rate of change of energy stored in the volume enclosed by the surface.

3.10 Poynting Vector in phasor form

The poynting vector can be expressed in phasor form as follows:

Pˆ Eˆ Hˆ 
The density of average flow is then
1 1
Pav  2 Re[Pˆ ]  2 Re[Eˆ  Hˆ

] . can be proved as follows:
This
1
E  Re(Eˆ e jt )  2 (Eˆe jt  Eˆ  e  jt )

1
H  Re(Hˆ e jt )  2(Hˆe jt  Hˆ  e  jt )

1 1 (93)
 P  E  H  4 (Eˆ  Hˆ   Eˆ   H4ˆ )  (Eˆ  Hˆ e j 2t  Eˆ  
H ˆ  e  j 2ˆt ) ˆ  
Since (E  H )  Eˆ  and Eˆ  Hˆ   (Eˆ Hˆ ) (93) can be written
H ˆ 1 1 as (94)
P 4  Mˆ  ) 4 (Nˆ e j 2t  Nˆ  e  j

(Mˆ Mˆ 2
where t
) Eˆ and Nˆ  E
ˆ Hˆ and we obtain from
(94)
 M. M. ALI
Page
No.:11/11
1 1
P  Re(Mˆ )  Re(Nˆ e j 2t )
2 2
1 1
 2 Re(Eˆ  Hˆ  ) 2 Re(Eˆ  Hˆ e j
2 t
) T 1
 Pav  1  Pdt  Re(E ˆ  ˆH ) (95)
T 0 2
because the time average of the second term within one period is zero.
 ELECTROMAGNETIC WAVE

Introduction

One of the major achievements of Maxwell was the correct prediction of electromagnetic
waves from considerations of the forms of the four equations which bear his name.

Wave equation
We will consider a linear, isotropic, homogeneous medium. Moreover, the net free charge
in the source free region is zero (v=0) and that any currents in the region are conduction
currents (J=E). This type of regions are quite general ones and include the practical
cases of free space (=0) as well as most conductors and dielectrics. Maxwell’s equations
for this region:

Sl No. In time domain In phasor form

1) E=0   Eˆ  0 (93)
2) H=0   Hˆ  0 (94)
3) dH (95)
  E     Eˆ  jHˆ
dt
4) dE (96)
HE   Hˆ  (  j )Eˆ
dt

Taking curl of (95) and substituting (93) and (96) we have

In time domain In phasor form
d (  H)
    E       Eˆ   j(  Hˆ )
dt
(  E)   2 E  
dE
 
d2 (  Eˆ )   2 Eˆ   j(  j )Eˆ
E
dt  2 Eˆ  j(  j )Eˆ (97b)

dt 2
2
dE d E
 2 E  by taking
Similarly,   curl of (96) and substituting (94) and (95) we have
(97a)
In time domain In phasor form
dt
dt (  E)   d (  E)
H
2
    Hˆ  (  j )(  Eˆ )
dt
dH (  Hˆ )   2 Hˆ   j(  j )Hˆ
 d 2H   2   

 2 Hˆ  j(  j )Hˆ (97d )
2
dH d H
 2 H    

(97c)equations in (97) are called electromagnetic wave equations or Helmholtz equations.

The
dt and (97c) are given in time domain, and equations (97b) and (97d) are in
Equations (97a)
terms of the dt
phasor
2 form. The wave equations in E and in H have exactly the same form.
 In absence of source, the wave equation for E in time domain as obtained from (97a) is
given by
d 2E 1 d 2E
2 E   
 dt 2 (98)
2
u 2 dt
One dimensional general solution of the scalar form of the above differential equation is
given by E=f1(z-ct)+ f2(z+ct). The first term represents a wave travelling in the +ve z-
direction while the second term represents a wave travelling in the -ve z-direction.

In the following we will consider the phasor form of wave equation for its simplicity.

Uniform plane wave

An electromagnetic wave originates from a point in space, spreads out uniformly in all
directions and it forms a spherical wave. An observer, however, at a great distance from
the source is able to observe only the small part of the wave in his immediate vicinity and
it appears to him as plane wave. Strictly speaking, a uniform plane wave does not exist in
practice because a source infinite in extent would be required to creat it, and practical
wave sources are always finite in extent.
A uniform plane wave satisfy the following conditions:
(i) At every point in space E and H are perpendicular to each other and to the
direction of propagation. No fields, therefore, in the direction of wave
propagation.
(ii) Everywhere in space, the fields vary harmonically with time and at the same
frequency;
(iii) Each field has the same direction, magnitude and phase at every point in any
plane perpendicular to the direction of wave propagation. The fields,
therefore, are only function of the coordinate that represents the direction of
wave propagation.

Assume the fields lie in the xy-plane. Again, with the assumption that the wave
propagates in the z-direction, the fields those only present are given
by
 Eˆx
Eˆx (z)a and Hˆ  Hˆ y (z)a y .
The following wave equations in phasor forms:
 j (  j )E
ˆ 2E2 ˆEˆ (99a)

d 2EˆHxˆ  2 ˆHˆ
 (99b)
2 2

  Ex
dz  =j(+j), for this particular case take the following form
2
where 2
(100a)
d2
ˆ 2  H
y 2 ˆy
Hdz
(100b) . of (100) are given by
The solutions
 z ˆ e ez jz
 Eˆm e z  Em ez  jez  E
Eˆx  Eˆme m (101a)
ˆ ˆy  H me z  ˆ m e z  ˆ m e z e jz  ˆ m ezejz (101b
H
ˆ H H H )
 where   j(  j )    jm ,and  ˆ ˆ ˆ
H, E E , Hˆ  are m m mundetermined complex
constants. ,  and  are, respectively, called the propagation constant, attenuation
constant and phase constant. The unit of  is neper per meter (Np/m). =1Np/m means
the wave magnitude decreases to 36.8% of its magnitude if it travels a distance of 1m.
The unit of the phase constant  is radians per meter (rad/m). The phase constant
expresses the amount of phase shift that occurs as the wave travels one meter.

The constants Eˆ  m, Eˆ  , Hˆ  , Hˆ  are related to each other by (95) and (96). From
m

dEˆ y (
m
dEˆ dEˆ dEˆ dEˆ
( dy zdE
ˆy x
dz )a m( dz  dx  dy )a   jHˆ
x z
dx
  Eˆ )ajHˆ we have
x z y

dE ˆ  dEˆ
 dzx a y  dzx   jHˆ y
jˆHˆ z
 ˆEm ez  E me   j(H ˆ  ez  H
m
ˆ ez )
m
Matching appropriate terms from both sides in this equation we have

Eˆm j Emˆ j
   ˆ and 

Hˆm Hˆm

 
ˆ
j j j
ˆ    (  j )  

j(  j )

ˆ, the intrinsic impedance of the medium, has a unit of ohm since it is a ratio of electric
field intensity (volt/m) to magnetic field intensity (Amp/m).
Assume Eˆm and Eˆm as Eˆ m E  m  and Eˆ mE  m  , as they are in general, the fields
in
E xtime domain
 Re(E ˆ x e are
) given by
jt

 E m ez cos(t  z   )  Em ze cos(t  z   ) (102a)

H y  Re(Hˆ y e )
jt

E  E
 me z
cos(t  z      )  m ez cos(t  z      )
As  (102b)
the fields E and H are perpendicular to each other and lie in a plane perpendicular to
the direction of wave propagation, they are called TEM waves (Transverse Electro-
Magnetic waves). The first term of both fields, E and H, represent the waves traveling in
the positive z-direction, while the second term of both of them represents the wave
traveling in the negative z-direction.

Uniform plane wave in lossless media

In lossless media, =0. Therefore the propagation constant  and the intrinsic impedance
ˆfor such a media are given by
 
 j(  j )    j  j     0 and      where
u u is the

j j 
velocityof light in the considered media; andˆ    j    0 .
0

The fields in the lossless media is, therefore, given

by
E x E  cos(
m t  z    )  E  cos(
m t  z    ) (103a)

E  E 
(103b)
 t  z   )  cos(t  z   )
H  cos(
y m m

 of (103a) has been plotted in the following for two different values of time, t 1
First term
and t2>t1. At successive times the curve effectively travels in the positive z-direction. Thus
it is a traveling wave. 

1.5

1
t= t2
magnitude of Ex

0.5

0
0 2 4 6 8 10
-0.5 t=t1

-1

-1.5
beta*z

Note that corresponding points on the waveforms occur at positions and times such that
the argument of the cosine has the same value; that is
t1  z  1   t  2z  2   (t  2t ) 1  (z  z2 ) (z  z ) 
1
 (t2 1 t )
2 1

dz  

dt    u 
Here u is referred to as the phase velocity of the wave. From the equation of phase
velocity, we have =2f/u=2/.  is therefore also known as wave number. Similar result
for the phase velocity with opposite polarity can be obtained for the wave traveling in the
negative z-direction.
For typical material >0 and >0, thus, the phase velocity of the wave in material media
is slower than in free space and the wavelength is shorter.
The phase velocity, intrinsic impedance, wave number and wavelength in lossless material
media can be expressed in terms to those for free space as follows:
1 1 c r
u 0 r r
  0 r 0 r r r ,    0 r 0    r  377 r ,   
ˆ r r
0 0 0

  and
 
r0r where   c ,   2 / and c is the light velocity in freespace.
0 0 0
Uniform plane wave in lossy media

In lossy media, 0. Thus, two important differences between wave natures in lossy and
lossless media have been observed. The first difference is that the propagation constant 
has a nonzero real part
  j(  j )    j
The fields in the lossy media is, therefore, given by
E x  E m e-z cos(t  z   )  Em ze t  z   ) (104a)

cos( 
H y  E m -ez cos(t  z     ) E m ez cos(t  z     ) (104b)
 
 and second terms of the field equations are still forward and backward traveling
The first
waves; however, their magnitude decreases as the waves move away from their origin as
shown in the following figures:

1.2
1.2
1
1
0.8
0.8
magnitude of Ex

0.6
magnitude of Ex
0.6
0.4 0.4
0.2 0.2
0 0
-0.2 2 4 6 8 10 -10 -8 -6 -4 -2 -0.2
0 0
-0.4
-0.4
-0.6
-0.6
-0.8
-0.8 beta*z
beta*z

Forward traveling wave Backward traveling

wave

The second difference is in the intrinsic impedance of the medium:


 j
j j(  j ) (j

ˆ j )   
 is a complex quantity in this case and   0. This non-zero value of  results in the
ˆ
electric and magnetic fields of both forward and backward traveling waves being out of
time phase by the phase angle (0  450). In case of a good conducting media where
()>>1   j (  j )  j 2 2  (1   j ) 
j
1 j   . The intrinsic
  2  2 (1  j)    j where     2
impedance of a good conducting media is therefore given by
j j j
ˆ   j   2 (1  j).

The presence  the field equations (104) in case of lossy and good
of the term e-z(ez) in
conducting medium signifies that the wave attenuates as it propagates in the +ve(-ve) z-
direction as is shown in the above figures. The attenuation factor  depends upon the
conductivity of the medium. The higher the conductivity of the medium, the more
pronounced the attenuation. Then the question arises: How far can the wave propagate in a
conducting medium before its amplitude becomes insignificant? The answer is skin depth
that is defined as the distance traveled by the wave in a conducting medium at
which its amplitude falls to 1/e of its value on the surface of that conducting medium. If
we denote the skin depth by , the amplitude of the wave falls to 1/e when =1.
Therefore =1/. The most remarkable feature is that the phase velocity of the wave in a
conducting medium is frequency and conductivity dependent term. Thus the shape of a
wave comprising many different frequencies will keep on changing as it progresses; i.e.,
the signal is distorted by time it reaches its destination. A medium in which the signal
becomes distorted is said to be dispersive medium. The velocity of propagation of the
signal comprises a group of frequencies is defined as the group velocity.
Consider a signal that consists of two forward traveling waves having equal amplitude
and slightly different angular frequencies + and -. The phase constants, being
functions of frequency, will also be slightly different. Say the phase constants for two
different frequencies be + and -. We have from (104) with =0 for simplicity
E x  E 0 e-z (  )t  (   )z E0 e-z cos(   )t  ( 
cos
 0 t
2E e-z cos(   cos(t 
)zz)
The term cos(tz) z) represents an envelope the velocity of propagation of which is
known as group velocity ug and can be determined by setting its argument equal to a
constant: t  z =Const from which we obtain

u g  dz    d .
dt 
d velocity ug can be related with the phase velocity as follows:
The group
1 d d  1 du p 1 ⎛  du p ⎟⎞
  ( ). 

ug d d u p u p u p 2 d  u ⎜⎜1 u d ⎠⎟
 p⎝ p
up
u g .
⎛  du p⎞ ⎟
⎜⎜1 
⎝ u p d ⎟ ⎠

Power flow

The average power density is given by

1 ˆ  ]  1 Re[Eˆ
ˆ H ˆ 1 ˆ
Pav  Re[E x a x  Hy a
y ]  x Hˆy
2 2 2 Re[E z


]a
E x and Hˆy  can be obtained from (104)
 E m-ez (t  z  )  E me 
ˆ x as z
(t  z   )
 
E  E m -ez (t  z     )  E m ez (t  z     )
ˆ
y

 (E ) 2 (E E  ) ⎤
⎡(E ) 2-2z
P av  ⎢ m2
e cos m
2e cos 
2 z  mm
sin sin(2z     )⎥a
  z
H ⎣
 is denoted as the average power density in the forward traveling wave and
ˆ first⎦term
The
the second term is denoted as the average power density in the backward traveling wave.
The third term is a cross-coupling term that disappears for lossless media where =0.
Polarization of UPW

The polarization of a UPW describes the time varying behavior of the electric field vector
at a given point in space. It is the locus of the tip of the electric field E at a given point as
a function of time. A wave is said to be linearly polarized when at some point in the
medium of propagation E oscillates along a straight line as a function of time. If the tip of
E vector traces a circle, the wave is said to be circularly polarized. The wave is
elliptically polarized when the electric field follows an elliptical path. The
circularly/elliptically polarized wave may be classified into right handed and left handed
polarized waves depending on the rotating nature of the tip of E field as a function of
time. If it rotates in the clockwise (anticlockwise) direction, it is said be left handed (right
handed) polarized wave.
The electric field of a UPW in a propagating medium described by
E(z, t)  E e  z cos(t  z)a is an example of linearly polarized wave as the field varies
0 x

along x-axis only with the change of time whatever be the value of z.
Assume the E field is given by E(z, t)  E1 (z, t)a x  E 2 (z, t)a y
where
E1 (z, t)  E10 cos(t  z) and E2 ( z, t)  E20 cos(t  z   / 2)  E20 sin(t  z).In
this case [E1 (z, t) / E10 ]  cos (t  z) and [E 2 (z, t) / E20 ]  sin (t  z). Therefore,
[E 1 (z, t) / E10 ]  [E2 2 (z, t) / E 2 0 ]  1. This is an2 equation of ellipse and the wave
to
2 is2 said
2 2

be right handed elliptically polarized because with the variation of time it rotates in the
anticlockwise direction. However, if E(z, t)  E1 (z, t)a x  E 2 (z, t)a y the will be left
handed elliptically polarized wave. In the above equation if E10= E20, the wave is said to
be circularly polarized wave.
The polarization of a wave depends upon the transmitting source such as an antenna. In
the standard AM broadcast frequency band, the vertical antenna is designed to transmit a
ground wave, which is vertically polarized because the E field from the antenna to ground
is vertical. In other applications including TV signals, the antennas are placed in a
horizontal plane to transmit a horizontally polarized wave. Both vertically and
horizontally polarized waves are examples of linearly polarized waves. This is why the
wires of rooftop TV receiving antennas are horizontal. The waves radiated by FM
broadcast stations are generally circularly polarized; hence the orientation of an FM
receiving antenna is not critical as long as it lies in a plane normal to the direction of the
signal.

Incidence of UPW on a plane boundary

(a) Normal incidence

In the previous sections, we have focused our attention on the propagation of UPW in an
unbounded medium. Here we consider a UPW that travels through one medium and then
enters into another medium of infinite extent. We assume that the interface between the
two media is normal to the direction of propagation of the incoming wave. We further
presume that (a) the incoming wave, called the incident wave, is propagating in the z-
direction; (b) the interface is an infinite plane at z=0; and (c) the region to the left of the
interface (z  0) is medium 1 and the region to the right of the interface (z  0) is medium
2. At the interface, we expect a part of the wave to penetrate the boundary and continue
its propagation in medium 2. This wave is referred to as the transmitted wave. The
remainder of the incident wave is reflected at the interface and then propagates in medium
1 in the negative z-direction. This wave is called the reflected wave. Thus, both the
incident and the transmitted waves propagate in the positive z-direction, whereas the
reflected propagates in the negative z-direction. The incident and the reflected waves are
in medium 1, and the transmitted wave is in medium 2. If we treat the incident and
transmitted waves as the forward-traveling waves, the reflected wave is then the backward
traveling wave.
For the system shown in the following figure, the E field of the incident wave in phasor
 z
form is assumed to be given by E i (z)  E i 0 e a x ; the associated magnetic field H is
1

) E i  1z
then H (z)
i  0 e ay. The corresponding fields for reflected and transmitted waves

)
1
are given by:
E r 0 1z ) E t 0  2z
)   2z
t(z) 
)r e a , (z) E e a , H e ay .
E (z)  E e a , H (z)  )
r 0  z x 1 r y t t 0 x
)
)  1 E 

2
Here the subscripts i, r and t stand for the incident, reflected and transmitted waves
respectively. And the subscripts 1 and 2 are used to refer medium 1 and 2, respectively.

At the interface (z=0), the tangential components (the x and y components) of electric and
magnetic fields in medium 1 are equal to their corresponding terms in medium 2. We
therefore have
E i (z  0)  E r (z  0)  E t (z  0)  Ei 0  Er 0  Et 0
(105) ) Ei Er 0 E t
)i )r (106)
H (z  0)  H (z  0)  H (z  0) 
t 0 -
) 0
) )
1  1 2

If we replace Et0 from (106) by substituting its value from (105) we obtain
E i 0 E r 0 E i0  E r 0 E E E E
) - ) )  i0) - i 0 ) r 0 ) r 0 )  E () -) )  E ()  ) )
       i0 2
1
1 r0 2
1 1 2 1 2 1
2

Er 0  E i0 (()2)-) 1) (107)
) 2  1 )

Again by replacing Er0 from (106) by substituting its value from (105) we have
Ei 0 Et 0  Ei 0 Et 0

)- ) 
)
 2 E 
2  1 E t 0  E t  22 E i 0 (108)
1 1 
)1 2 0
)
1
i0

)2
 1
2 ) transmission coefficient,
The ratios Er0/ Ei0 and Et0/ Ei0 are called reflection coefficientand
)
respectively. They are given in termsof the intrinsic impedances of the media by

E (2 -)1 ) ) (109)

  Er 0  (  )
i0 2 1

E
T  )  )
t0 2 (110)
2E)  2 1
i0
From (109) and (110) we may prove that 1+=T.
In terms of the reflection coefficient, the E field in medium 1 is
 z
 z (111)
2 z
 z
(z) case
InE1this E i (z)  Eand
1=0  E1 
r (z)hence ( z)  Ei 0 e a x  E i0 e a x  (1  e )E i 0 e a x
1=0, and 1 =j1. The remarkable feature of this system is
1 1 1 1

that ,  both
(i) When and Tmedia
are allare non-conducting
real.
From (111) we have
j 2 z  j z
E1 (z)  (1  e )E i0 e a x
1 1

The magnitude of E1 (z) is given by

(1  e j 2 z )E 1
E (1   cos 2 z) 2  (sin 2 z) 2  E (1   2  2 cos 2
z)
i0 i0 1 1 i0
1

 E maximum
ItEbecomes (1  2 for
2)  Ewhen
>0 (1 )
21.z=-2n where n=0, 1, 2, ……and is given by
i0 i0
1 max
) magnitude of E (z) becomes minimum for >0 when 21z=-(2n+1) where n=0, 1,
The 1

2, ……and is given by E 1 min  i00 (1    2)  Ei (1  ) .

) 2

E maximum and minimum values of E (z) for <0 are given by

The E1  Ei 0 (1 
1
)
and E1 min
 Ei 0 (1  ) . max

In generalized form, the maximum and minimum values of E1 (z) are therefore given by
E1  Ei 0 (1   ) and E1  Ei0 (1   ) .
max min
The ratio of maximum value of electric field to its minimum value is defined as the
standing wave ratio (SWR) S and is then given by
E) 1  (1   )
S max
. (112)
E1 (1   )
min

S -1
From this equation we obtain   S  1.
(ii) When media 1 is non-conducting and media 2 is good conductor
The equations of reflection and transmission coefficients and SWR remain same, but in
this case =-1 and T=0 as 2  0 .

(b) Oblique incidence

When an electromagnetic wave strikes a plane boundary any arbitrary angle, we refer to it
as oblique incident. We again consider that a xy-plane at z=0 separates two media; media
1 is on the left side on the plane and media 2 is on the right side. In our following
discussion we will limit to two special cases: i) First we will assume a perpendicularly
polarized wave in which E field of the incident wave is normal to the plane on incident
and ii) In the second case a parallel polarized wave where E field lies on the incident
plane will be considered.

Perpendicular Polarization
Now the plane that includes the unit vector an normal to the boundary and the propagation
constant of the incidence wave is called the plane of incidence as shown in the following
figure.

(a) Oblique incidence on a plane boundary (b) Perpendicularly polarized wave

Assume the incident wave propagates in the z’ direction and makes an angle I with an
unit vector. If Ei0 is the amplitude of the incident E field, we can express it at any point in
’
medium 1 as E i  Eio e 1z a x. From the following figure, the exponent  z can1 be
expressed as

 1 z   1a z  zaz
 1 cos ia  1  ai 
z  sin y za z ya  y

  1 z cosi  y sini 

 E  E e 1 ( z cos i  y sin i ) a

i io x

Therefore from   E i   j1H i and

1  ( z cos   y sin
H i   ) Eioe  ) i
1 i

 cosi ay  sin  iaz 
)  1

(c)
Let us assume that the reflected wave propagates in the z” direction and makes an angle
r with an as shown in figure (b). If Er0 is the amplitude of the reflected E field, we can
express it at any point in medium 1 as E  E e  za . From the following figure, the 1

r ro x

exponent 1z” can be expressed as

 1 z   1a z  zaz

  1 cosr a z  sin 1  ar 
y  
 z(a ) z ya  y

  1  z cos r  y sin r

 1( z cos r y sin r )
 Er  E roe ax
Therefore from   E r   j1H r and

H r   1)E ero
1 ( z cosr  y sin r )
cos a  sin a 
r y rz

)  1

The transmitted electric and magnetic fields are of similar expressions to those of
incidence wave; the exception is in the angle. The fields for transmitted wave are,
therefore,
 ( z cos  y sin  )
E t  E to e 2 2
ax 2

1
H t   ) Eto e

 2 ( z cos2  y sin  2 )  cos a  sin
 a

At

) z=0, we obtain from the boundary
2
conditions that 2 y
  1 y sin i  1 y sin  r   2 y sin 22 z
E e E e E e and
io ro to
1  y sin  i
1 E ioe 1 cos  i
1
E ro e  y sin  cos  r  Eto e  2y sin  2 cos  2
) ) )
1 r

1  1  2
For any values of y to satisfy the above equations the following conditions must fulfill:
 1 y sini   1 y sinr   2 y sin2
 i   r and  1 sin i   2 sin 2 (113)
Equation (113) is known as Snell’s law of reflection.
The boundary conditions at z=0 are given with the help of Snell’s law by
Eio  Ero  Eto and (114a)

1 1 1 (114b)
)1 E iocos  E
ro   iE cos to
cos 2

)
 1 2
Replacing
i Eto from (114b) by (114a) we have
Eio (2 cosi 1 )cos2 )  Ero (2 cosi 1 cos2 )
) ) (115)
   roE 2 ( cosi  1cos )2
)
Eio (2 cos i )1cos )2
Replacing Ero from (114b) by (114a) we have
 (2 cosi  1 cos 2 )
2E io cosi  E to
)
2 (116)
)
22 cosi
   E to  )
T E io (2 cos i )1cos )2
Equations (115) and (116) are known as Fresnel equations. Again we find that 1+=T as
was the case for normal incidence.

Parallel Polarization

The case where the Ei field is parallel to

the plane of incidence (yz-plane) is
shown in the adjacent figure. As the
direction Ei field will be perpendicular to
az’, the unit vector along it can be
resolved into (-cosi ay+ sini az). Again
from the previous article we have
 1 z   1 z cosi  y sini
 Ei  E eio

)   1 ( z cos  i  y sin i )

(cos  i a y  sini a z )
The corresponding H i can be obtained
from   E i   jHi and is
E io   1 ( z cos  i  y sin  i )
Hi  )1 e a x
(a) 
)
The direction of propagation of the
reflected wave is az” and the Er field is
assumed to be as shown in figure (c). The
unit vector along Er can be given by
(cosr ay+ sinr az). Again from the
previous article
 1 z   1  z cos r  y sin  r 
)
 Er  Ero e  ( z cos   y sin 
1 r r )

(b) (cos r a y  sin  r a z )

The corresponding H r is obtained from
  E r   jH r and is
r E ro
H  e  1 (  z cos  r  y sin  r ) a x
1
)  )
The equation of transmitted electric and
magnetic fields are similar to those of
Er incident fields with an exception in
angle.
(c)

These are given

by
  ( z cos  y sin  )
Et  E to e 2 2 2
( cos 2 a y  sin 
2 az )

Eto   2 ( z cos  2  y sin  2 )

Ht  )2 e a x
) 
Applying the Snell’s law we obtain from the boundary conditions at z=0
 y sin   y sin   y sin 
 Eio e 1
cosi  E ro ei
cosi  E to e cos21 i 1 i

 Eio cosi  Ero cosi  Eto cos2 (117)

and
Eio E ro Eto
) e  1y sin  i
 e)  1 y sin  r  e )  2 y sin  2
  
1 1 2


Eio Ero Eto (118)
)  1)
1 2
From (117) and (118) we obtained
)  cos i cos
E 2 
||  ro  1 2
(119)
E io 1)cosi ) 2cos 2
and

T||  E to  ) 22) cos

)
i
(120)
E io 1cos i  cos
2  2
From (119) and (120) it can prove that

1  ||  T|| ⎜⎛ cos2 ⎞⎟ (121)

⎝ cos i ⎠
Critical Angle

Let us consider the Snell’s law  1 sin i   2 sin2 for two non-conducting media; in this
case we have
2  
j  
sin    sin 2  2 sin 2  2 sin 2   2 2 sin  2   2 sin 2
i
1 1 1 11 1
j 
We have assumed here the typical non-ferromagnetic dielectric media for which r 1. If a
wave in media 1 is incident on a less dense media 2(1> 2), in this case 2 becomes
greater than i. The value of i that makes 2=900 is called the critical angle in which the
transmitted wave will propagate entirely parallel to the interface. Simply stated, there will
be no power propagating in the z direction in media 2. This is called total reflection. This
is why the critical angle is also called the angle of total reflection. The critical angle is
given by

 c sin -1  2 .
1

Brewster Angle

For both perpendicular and parallel polarization, there is a value of the incidence angle for
which the reflection coefficient becomes zero. This angle of incidence is called the
Brewster angle. No reflection occurs, i.e., all the waves transmitted into the second media.
For perpendicular polarization =0 if
2 cosi  1 cos 2

) 1) ⎞ )
 cos  i  ) 1 ⎛ ⎛
2
2 2 1 2
)
 cos   ⎜  ⎟ cos2 2  1  sin i  ⎜) ⎟ (1  sin  2 )
cos  2 2 i
⎝2 2 ⎝ 2
⎞
⎠ ⎠
) 2 ) 2
⎛  ⎞ ⎛  ⎞ (122)
 sin 2 i  ⎜ ) 1 ⎟ sin 2  2 1  ⎜ ) 1 ⎟
⎝ 2⎠ ⎝ 2
. ⎠
Using the Snell’s law i   r and  1 sin i   2 sin 2 we have for non-conducting media

  /u
sin 2   1 sin i  sin 2  1 sin i  sin 2   / 1 sin  i
2 2 2
u
 sin 2  11
  sin 
i
2 2

Using this value of sin2 in (122) we obtain

) 2
⎛ 1 ⎞  1 1
sin   ⎜ ) ⎟   sin   1  ⎜ 
2 i
2
i
)
⎛)1 ⎟ 
 sin2 i 1   1 / 2  1
2
 12
 21
⎝2 2 ⎠ 2 2 ⎝ 2 ⎠
⎞
1  (1 2/  2 1)
 sin 2 i
1  1 / 2
2

The incidence angle satisfying this equation is referred to as the Brewster angle and is
denoted by B. However, for non-ferromagnetic matter 1=2=0, therefore the Brewster
angle for perpendicular polarization does not exist.

For parallel polarization ||=0 if

1 cosi  2 cos 2
) ) 2 ) 2
 cos   2 2 ⎛)2 ⎞ 2 2 ⎛)2 ⎞ 2

)1 2  cos   ⎜  ⎟ cos  2

i  1  sin   ⎜  ⎟ (1  sin  2 )
i
cos
i
⎝ 1 ⎝ 1
⎠ ⎠
)2 2 )2 2
⎛  ⎞ 2 ⎛  ⎞
⎟ (123)
 sin 2 i  ⎜ ) ⎟ sin  2 1  ⎜ )
⎝ 1⎠ ⎝ 1
. ⎠
Using the Snell’s law i   r and  1 sin i   2 sin 2 we have for non-conducting media

  /u
sin 2   1 sin i  sin 2  1 sin i  sin 2   / 1 sin  i
2 2 2
u
 sin 2  11
  sin 
i
2 2

Using this value of sin2 in (123) we obtain for non-ferromagnetic materials

 ) 2
⎛ 2 ⎞ 
) 2
⎛)2 ⎞⎟  sin 2  ⎡1  2  1/  2 2 ⎤ 
sin  sin   1  ⎜  i⎢
1 2
2
 ⎜) ⎟  1  21
⎣ 1 ⎥ 12
i i
⎝ 1⎠ 2 ⎝ 1   ⎦
⎠ 1
 sin 2 i 1  (2 1/  12) ⎤ 1 1(1 / 12/) 2
⎡ 2 1  1 / 2
1   / 
2
⎥
⎢
1  1 2 ⎦
2
⎣
 1 /  2
 cos 2 i 1  sin 2 i
1   /  
 /  i  1 B|| 2 tan 
tani   2 / 2 1
1

1 incidence angle satisfying this equation is referred to as the Brewster angle and is
The
denoted by B||.
Thus when an incident wave having both parallel and perpendicular components
impinges upon a plane boundary between two dielectric media at the Brewster angle B||,
the transmitted wave will have only parallel component where as the reflected wave
contains only the perpendicular component. That is an elliptically or circularly polarized
wave will become a linearly polarized wave upon reflection. For this reason, the Brewster
angle is also called the polarizing angle.

See the Examples 8-13 & 8-14 of Cheng