Essential Mathematics

INTRODUCTION

Physical quantities are broadly classified into two categories; scalars and vectors. A scalar is
a quantity that is specified by its magnitude only for example temperature, mass and permittivity.
On the other hand, a vector is defined by both its magnitude and direction. Examples of vectors
include force, acceleration and torque.

Electromagnetics is concerned with the study of electric and magnetic fields, which are
essentially force fields, and therefore vector fields. A detailed understanding of vector analysis
and calculus is therefore essential in the analysis of these fields. The fields can only be evaluated
at specified locations in space. For this purpose, a comprehensive understanding of co-ordinate
systems, especially orthogonal, is paramount. In this section, an overview of these topics is
carried out. Some prior understanding of these concepts is however assumed.

VECTOR ALGEBRA

A vector A can be written as

A  aAA (1.1)

where A is the magnitude having the unit and dimension of A and a A is a dimensionless vector
having the direction of A and whose magnitude is unity. Such a vector is known as a unit vector.
Thus

A A (1.2)

A A
aA   (1.3)
A A

A graphical representation of vector A is given in Figure 1.1. the length of the line represents the
magnitude of A whereas the arrow points in the direction of A.

©j. k. makiche 1
 A

Figure 1.1: Vector A

Vector Addition and Subtraction

Two vectors A and B, neither in the same direction nor in opposite direction, can be added by
parallelogram rule to obtain another vector C so that A  B  C as indicated in Figure 1.2(a).

A
B
C B

(a) (b)

Figure 1.2: Vector addition and subtraction (a) Addition (b) Subtraction

Vector subtraction is defined in terms of vector addition as

A  B  A  (B) (1.6)

As indicated in Figure 1.2(b). Vector addition obeys commutative and associative laws i.e.

Commutative law: AB BA (1.4)

Associative law: A  (B  C)  ( A  B )  C (1.5)

©j. k. makiche 2
Vector Products

There are two vector products; the dot product (or scalar product) and the cross product (or
vector product).

The dot product of two vectors A and B, denoted A  B is defined as

A  B  AB cos  AB (1.7)

where θ AB is the smaller angle between A and B as indicated in Figure 1.3 (a). It is less than 
radians.

A  A  A2 (1.9)

or

A AA (1.10)

B
B
θ AB A×B
θ AB
A
A

Figure 1.3: Vector products (a) Dot product (b) Cross product

The dot product is always a scalar.

It should also be noted that the dot product obeys commutative and distributive laws but not
associative law.

Commutative law: A B B A (1.11)

Distributive law: A  (B  C)  A  B  A  C (1.12)

©j. k. makiche 3
The cross product of two vectors A and B, denoted A  B is a vector perpendicular to the plane
containing A and B as indicated in Figure 1.3 (b) i.e.

A  B  a n AB sin θ AB (1.13)

The cross product does not obey commutative law since

A  B  B  A (1.14)

However it obeys distributive law i.e.

A  (B  C)  A  B  A  C (1.15)

The cross product is not associative.

The product of three vectors may be scalar triple product or vector triple product. The scalar
triple product has the following property:

A  ( B  C )  B  (C  A )  C ( A  B ) (1.16)

The vector triple product (also known as back-cab rule) can be expanded as the difference of two
vectors as follows:

A  (B  C)  B  ( A  C)  C( A  B ) (1.17)

Division by a vector is not defined.

Orthogonal Co-ordinate Systems

Any vector A can be decomposed into three orthogonal components. The vector A may be
written as

A  au1 Au1  au 2 Au 2  au 3 Au 3 (1.18)

where au1 , au 2 and a u 3 are unit vectors in the three orthogonal directions defined such that,

a u1  a u 2  a u 2  a u 3  a u 3  a u1  0 (1.19)

©j. k. makiche 4
and

a u1  a u1  a u 2  a u 2  a u 3  a u 3  1 (1.20)

The magnitude of A is

A  A  Au21  Au22  Au23 (1.21)

Vector calculus frequently requires evaluation of line, surface, and volume integrals. An
appropriate co-ordinate system must be selected so that the required differential length, area and
surface can be obtained. The most common orthogonal co-ordinate systems are Cartesian,
Cylindrical and Spherical co-ordinate systems. Figure 1.1 shows a general point P defined in
terms of the various co-ordinate systems.

x   cos   r sin  cos    x 2  y 2  r sin  r  x2  y 2  z 2   2  z 2

y   sin   r sin  sin    tan 1 ( y / x)   tan 1 ( x 2  y 2 / z )  tan 1 (  / z )
r  z cos  zz   tan 1 ( y / x)
z z z

P ( x, y , z ) P(  ,  , z ) P(r , ,  )
 r
z z
y y y

x  
y
x x x
  x   0   0r 
  y   0    2 0  
  z     z   0    2

Figure 1.1: General point P described in (a) Cartesian co-ordinates P ( x, y, z ) (b) Cylindrical co-ordinates P (  ,  , z )

(c) Spherical co-ordinates P (r ,  ,  )

Cartesian Co-ordinates

(u1 , u2 , u3 )  ( x, y, z )

The unit vectors a x , a y and a z satisfy the relations

©j. k. makiche 5
 ax  a y  az (1.22a)

a y  az  ax (1.22b)

az  ax  a y (1.22c)

The vector A becomes

A  a x Ax  a y Ay  a z Az (1.23)

The dot product of two vectors A and B is from eq. 1.7,

A  B  (a x Ax  a y Ay  a z Az )  (a x Bx  a y By  a z Bz )
(1.24)
 Ax Bx  Ay By  Az Bz

The cross product of A and B is from eq. 1.13,

A  B  a y ( Ay Bz  Az B y )  a y ( Az Bx  Ax Bz )  a z ( Ax B y  Ay Bx )

ax ay az
 Ax Ay Az (1.25)
Bx By Bz

Since x, y and z are lengths, the expressions for the differential length, differential area and
differential volume are respectively,

d   a x dx  a y dy  a z dz (1.26)

dsx  dydz (1.27a)

ds y  dxdz (1.27b)

dsz  dxdy (1.27c)

and

©j. k. makiche 6
 dv  dxdydz (1.28)

Cylindrical Co-ordinates

(u1 , u2 , u3 )  (  ,  , z )

The following relations between unit vectors in different directions apply.

a   a  a z (1.29a)

a  a z  a  (1.29b)

a z  a   a (1.29c)

A vector is written as

A  A a   A a  Az a z (1.30)

The expressions for the differential lengths, area and volume are

d   a  d   a  d  a z dz (1.31)

ds   d dz (1.32a)

ds  d  dz (1.32b)

dsz   d  d (1.32c)

and

dv   d  d  dz (1.33)

The relations between the components of a vector in Cartesian and cylindrical co-ordinates is
given in matrix form as

©j. k. makiche 7
  Ax   cos   sin  0   A 
 A    sin  cos  0   A  (1.34)
 y 
 Az   0 0 1   Az 

Hence the following conversion formulas from cylindrical to Cartesian co-ordinates are used,

x   cos  (1.35a)

y   sin  (1.35b)

zz (1.35c)

The inverse relations (from Cartesian to cylindrical) are

ρ  x2  y2 (1.36a)

y
  tan 1 (1.36b)
x

zz (1.36c)

Spherical Co-ordinates

(u1 , u2 , u3 )  (r ,  ,  )

The directional unit vectors are related as follows;

a r  a  a (1.37a)

a  a  a r (1.37b)

a  a r  a (1.37c)

A vector in spherical co-ordinates is written as

A  Ar a r  A a  A a (1.38)

©j. k. makiche 8
The differential length, area and volume are given by,

d   a r dr  a rd  a r sin  d (1.39)

dsr  r 2 sin  d d (1.40a)

ds  r sin  drd (1.40b)

ds  rdrd (1.40c)

and

dv  r 2 sin  drd d (1.41)

A vector given in spherical co-ordinates can be transformed into Cartesian co-ordinates using the
following expressions,

x  r sin  cos  (1.42a)

y  r sin  sin  (1.42b)

z  r cos θ (1.42c)

Conversely,

r  x2  y 2  z 2 (1.43a)

x2  y 2
θ  tan 1 (1.43b)
z

y
  tan 1 (1.43c)
x

THE DEL ( ) OPERATOR

This is a vector differential operator defined in Cartesian co-ordinates as

©j. k. makiche 9
   
  ax  ay  az (1.44)
x y z

The Gradient of a Scalar

The gradient of a scalar is defined as the vector that represents both the magnitude and the
direction of the maximum space rate of increase of that scalar. The gradient of a scalar V,
denoted V is given in the various orthogonal co-ordinate systems as;

    
V   a x  a y  a z  V (Cartesian) (1.45)
 x y z 

  1   
V   a   a  a z  V (Cylindrical) (1.46)
    z 

  1  1  
V   a r  a  a  V (Spherical) (1.47)
 r r  r sin   

The Divergence of a Vector Field

In the study of vector fields, it is convenient to represent field variations graphically by directed
field lines called flux lines or streamlines. The flux lines indicate the direction of the field at
each point. The divergence of a vector field A at a point is defined as the net outward flux of A
per unit volume as the volume about the point tends to zero. The divergence of a vector A,
(denoted   A or div A) in Cartesian co-ordinates (eq. 1.23) is found using the expression

Ax Ay Az

A    (1.49)
x y z

In cylindrical co-ordinates (eq. 1.30),

1  1 A Az
A  (  A )   (1.50)
    z

In spherical co-ordinates (eq. 1.38),

©j. k. makiche 10
 1  2 1  1 A
A  ( r A )  ( A sin  )  (1.51)
r 2 r r sin   r sin  
r

Divergence Theorem

It states that the volume integral of the divergence of a vector field equals the total outward
flux of the vector through the surface that bounds the volume.

   Adv   A  dS
V S
(1.52)

The direction of dS is always that of the outward normal, perpendicular to the surface dS and
away from the volume. This theorem converts a volume integral of the divergence of a vector to
a closed surface integral. A vector field whose divergence is zero is called a solenoidal field.

Curl of a Vector Field

The curl of a vector field A, denoted  A or curl A or rot A is a vector whose magnitude is the
maximum net circulation of A per unit area as the area tends to zero and whose direction is
the normal direction of the area when the area is oriented to make the net circulation
maximum.

The curl is given in the various orthogonal co-ordinate systems as follows:

Cartesian:

ax ay az
    A Ay   Ax Az   Ay Ax 
 A   ax  z    ay     az    (1.53)
x y z  y z   z x   x y 
Ax Ay Az

Cylindrical:

©j. k. makiche 11
 a a az
  
 A 
  z
(1.54)
A  A Az
 1 Az A   A Az  1   (  A ) A 
 a     a     az   
   z   z       

Spherical:

ar a r a r sin 
1   
 A  2
r sin  r  
Ar rA r sin  A
1   ( A sin  ) A  1  1 Ar  (rA  1   (rA ) Ar 
 ar     a     a  
r sin      r  sin   r  r  r  
(1.55)

The determinantal forms are similar to the cross product exhibited in eq. 1.25. A curl-free vector
field is known as an irrotational or a conservative field.

Stoke’s Theorem

It states that the surface integral of the curl of a vector field over an open surface is equal to
the closed line integral of the vector along a contour bounding the surface. For a vector field
A,

 (  A)  dS   A  d 
S C
(1.56)

Stoke’s theorem converts a surface integral of the vector to a line integral of the vector and vice
versa.

Null Identities

Identity I: The curl of the gradient of any scalar field is zero. For the scalar V,

©j. k. makiche 12
   ( V )  0 (1.57)

This identity may conversely be stated as follows: If a vector field is curl-free, then it can be
expressed as a gradient of a scalar field. Hence, an irrotational (or conservative) vector field can
always be expressed as the gradient of a scalar field.

Identity II: The divergence of the curl of any vector field is identically zero. For a vector field A

  (  A )  0 (1.58)

This identity may conversely be stated as: if a vector field is divergenceless (or solenoidal), then
it can be expressed as a curl of another vector field.

Helmholtz’s Theorem

Vector fields may be classified in accordance with their being solenoidal and/or irrotational. A
vector field F is:

1. Solenoidal and irrotational if   F  0 and  F  0 .

2. Solenoidal but not irrotational if   F  0 and   F  0 .

3. Irrotational but not solenoidal if  F  0 and   F  0 .

4. Neither solenoidal nor irrotational if   F  0 and   F  0

A general vector field has both a non-zero divergence and a non-zero curl and may be considered
as a sum of a solenoidal field and an irrotational field.

Helmholtz’s Theorem: A vector field (or a vector point function) is determined to within an
additive constant if both its curl and divergence is specified everywhere. This implies that a
vector function is determined if both its curl and divergence are specified.

©j. k. makiche 13
 Electrostatics

COULOMB’S LAW AND THE ELECTRIC FIELD INTENSITY

Basic Facts

The source of electric and magnetic fields is the electric charge. Here are some basics facts
concerning the electric charge:

 There are two types of electric charge; positive due to protons and negative due to
electrons.
 Charge is measured in coulomb (C). One coulomb is the total charge contained by
approximately 6  1018 electrons. Therefore the electron charge is e  1.6019  1019 C .
 A proton has charge of the same magnitude as an electron but of opposite polarity.
 Like charges repel and unlike charges attract.

Coulomb’s Law

Coulomb’s law is an experimental law used to determine the amount of force that a point charge
exerts on another point charge. A point charge is charge located in a body whose dimensions are
much smaller than other relevant dimensions.

Coulomb’s law states that the force between two point charges Q1 and Q2 is directly proportional

to the product Q1Q2 of the charges and inversely proportional to the square of distance R
between them. The force acts along a straight line joining the two charges or

Q1Q2
F k (2.1)
R2

where k is a proportionality constant. When SI units are used

1
k (2.2)
4 πε0

©j. k. makiche 14
where ε0 is the permittivity of free space (in farads per metre) and has the value

ε0  8.854 1012 F/m . Eq. (2.1) becomes

Q1Q2
F (2.3)
4 πε0 R 2

If the point charges Q1 and Q2 are located at points having position vectors r1 and r2 as shown in

Fig. 2.1, then the force F12 on Q1 due to Q2 is given by

Q1Q2
F12  a R12 (2.4)
4 πε0 R 2

where

R12  r2  r1 (2.5a)

R  R12 (2.5b)

R12
a R12  (2.5c)
R

Fig. 2.1: Coulomb vector force on point charges Q1 and Q2

Substituting eq. (2.5) in (2.4),

©j. k. makiche 15
 Q1Q2
F12  R12 (2.6a)
4 πε0 R 3

or

Q1Q2 (r2  r1 )
F12  (2.6b)
4 πε0 r2  r1
3

We note that:

1. Since a R12  a R 21 , F12  F21 where F21 is the force on Q1 due to Q2 .

2. The distance R between Q1 and Q2 must be large compared with the linear dimensions of

the bodies, i.e. Q1 and Q2 must be point charges.

3. Q1 and Q2 must be static i.e. at rest.

4. The signs of Q1 and Q2 must be taken into account in eq. (2.4).

If there are more than two point charges, the principle of superposition is used to determine the
force on a particular charge. The principle states that if there are N charges Q1 , Q2 , QN located

respectively at points with position vectors r1 , r2 , rN , the resultant force F on a charge Q

located at point r is the vector sum of the forces exerted on Q by each of the charges
Q1 , Q2 , QN . Hence,

QQ1 (r  r1 ) QQ2 (r  r2 ) QQN (r  rN )

F   
4 πε0 r  r1 4 πε0 r  r2 4 πε0 r  rN
3 3 3

Q N
Qk (r  rk )
F
4 πε0

k 1 r  rk
3
(2.7)

©j. k. makiche 16
ELECTRIC FIELD INTENSITY

Consider a charge in a fixed position, say Q . If a second charge, say Q1 is moved slowly around

Q , it is noted that there exists everywhere a force on Q1 . This implies the existence of a force

field. The force on Q due to Q1 is given as

QQ1
F  aR (2.8)
4 πε0 R 2

Writing this force per unit charge gives

F Q
 aR (2.9)
Q1 4 πε0 R 2

The quantity on the right side of eq. (2.11) describes a vector field and is called the electric field
intensity or the electric field strength (denoted E ). The electric field intensity is the force per
unit charge when placed in the electric field. Thus,

F
E  lim (2.10)
Q 0 Q

or simply

F
E (2.11)
Q

The direction of E is that of F and it is measured in newtons/coulomb or volts/metre. For N

charges Q1 , Q2 , QN located at r1 , r2 , rN the electric field intensity at point r is obtained using
eq. (2.7) and (2.11) as

Q1 (r  r1 ) Q2 (r  r2 ) QN (r  rN )
E   
4 πε0 r  r1 4 πε0 r  r2 4 πε0 r  rN
3 3 3

or

©j. k. makiche 17
 1 N
Qk (r  rk )
F
4 πε0

k 1 r  rk
3
(2.12)

Example 1: Two charges Q1  150 C , Q2  100 C and are located at ( 1,2,3) and (0,3,1)
respectively. Determine the force on Q2 due to Q1 .

Given: Q1  150μC , r1  a x  2a y  3a z ,

Q2  100μC , r2  3a y  a z

1 Q1Q2
F12  a 12
4 0 r 2
r12  (0  1)a x  (3  2)a y  (1  3)a z  a x  a y  2a z
r12  12  12  22  6
1 Q1Q2
F12  r12
4 πε0 r12 3
1 150  106  100  106
F12  (a x  a y  2a z )  9.17(a x  a y  2a z )
4 π  8.85  1012 63/ 2

Example 2: Point charges 1 mC and 2 mC are located at ( 3, 2, 1) and ( 1, 1, 4) respectively.
Calculate the electric force on a 10 nC charge located at (0, 3,1) and the electric field intensity at
that point.

Given: Q  10 nC , r  3a y  a z ,

Q1  1 mC , r1  3a x  2a y  a z ,

Q2  2 mC , r2  a x  a y  4a z

r  r1  (3a y  a z )  (3a x  2a y  a z )  3a x  a y  2a z

r  r1  32  12  22  14

r  r2  (3a y  a z )  (a x  a y  4a z )  a x  4a y  3a z

©j. k. makiche 18
 r  r2  12  42  32  26

Q 2
Qk (r  rk ) Q  Q (r  r ) Q (r  r ) 
F
4 πε0
 r  rk

4 πε0
 1 1
 2 2

 r  r1 r  r2 
3 3 3
k 1

10  10 9 103 (3a x  a y  2a z ) 2  10 3 (a x  4a y  3a z ) 
   
4 π (8.854  10 12 ) 
3 3
[14] 2 [26] 2 

which gives

F  6.507a x  3.817a y  7.506a z mN

At that point,

3
F (6.507a x  3.817a y  7.506a z )  10
E 
Q 10 109
 650.7a x  381.7a y  750.6a z kV/m

Example 3: A point charge Q1  1 nC is located at the origin in free space. What charge must
be located at (2,0,0) to cause E x to be zero at (3,1,1)?

Q1

Q2
R1

R2

From the figure,

R1  (3  0)a x  (1  0)a y  (1  0)a z  3a x  a y  a z

R 2  (3  2)a x  (1  0)a y  (1  0)a z  a x  a y  a z

©j. k. makiche 19
 R1  11 , R2  3

R1 3a x  a y  a z
a R1  
R1 11

R2 ax  a y  az
aR2  
R2 3

1  Q1 Q2 
E  2 a R1  2 a R 2 
4 πε0  r1 r2 

1  109 3a x  a y  a z Q2 a x  a y  a z 
   
4 π (8.854 1012 )  11 11 3 3 

If Ex  0 at (3, 1, 1), the x-component of E must be set to zero. Hence,

109 (3) Q2
Ex  3  3 0
11 2
32

 Q2  0.427 nC

ELECTRIC FIELD DUE TO CONTINUOUS CHARGE DISTRIBUTIONS

Charge Configurations

So far we have only considered forces and electric fields due to point charges. If charge Q is
distributed over a finite volume v , we define volume charge density ρv (in C/m3) as

Q dQ
ρv  lim  (2.13)
v 0 v dv

Hence, dQ  ρv dv and the total charge in a given volume is obtained in terms of ρv as

Q   ρv ( x, y, z )dxdydz (2.14)

©j. k. makiche 20
All other charge distributions are special cases of the volume charge distribution. For example
when the charge Q is distributed over an infinitesimal volume, it is regarded as a point charge.
In this case, volume charge density becomes infinite and is therefore inappropriate.

Charge may be also be distributed over a filament (whose length may be mathematically infinite)
with infinitesimal thickness. In this case, we define the line charge density ρL (in C/m) such that

dQ  ρL dL  Q   ρL dL (2.13)
L

It is also possible to have continuous charge distribution along a line, on a surface, or in a

volume. We denote the line charge density, surface charge density, and volume charge density
by, , and, respectively.

The surface charge density, ρS (in C/m2) is defined for charge distributed over very thin surfaces
(whose area may be mathematically infinite). In this case, the surface charge density is defined
such that

dQ  ρS dS  Q   ρS dS (2.14)
S

The various charge configurations discussed are depicted in Fig. 2.3.

ρS ρv
ρL
Q

Fig 2.3: Continuous charge distributions

where

  
S

©j. k. makiche 21
   
v

The expression for the electric field becomes

1 ρL dL
E
4 πε0 
L
R2
aR (2.16)

1 ρS dS
E
4 πε0 
s
R2
aR (2.17)

1  ρv dv 
E 
4 πε0 v  R 2 
aR (2.18)

Line Charge.

Consider a finite line charge with uniform charge density ρL extending from A to B along the z-
axis as shown in Figure 2.3. The charge element dQ associated with element dl = dz of the line is

dQ  ρL dL  ρL dz

and the total charge Q is

zB
Q   ρL dz
zA

It is customary to denote the field point by ( x, y , z ) and the source point as ( x, y , z ) . Hence,
from Fig. 2.3,

dl  dz 

R  xa x  ya y  za z  z a z  xa x  ya y  ( z  z )a z

or

R  ρa ρ  ( z  z )a z

©j. k. makiche 22
 R 2  x 2  y 2  ( z  z ) 2  ρ 2  ( z  z ) 2

R ρa ρ  ( z  z)a z
aR  
R ρ 2  ( z  z)2

z
dE z dE

ρ α
(0,0, z ) T dE ρ
α1
α2 α
P ( x , y, z )

B R

dl (0,0, z)

y
0

Fig. 2.4: Evaluation of E field due to a line charge

Substituting into Eq. (2.16),

1 ρL dz ρa ρ  ( z  z)a z ρ ρa ρ  ( z  z)a z

E 
4 πε0 L ρ  ( z  z)
2 2
 L  2
ρ 2  ( z  z)2 4 πε0 L [ ρ  ( z  z) ]
2 3/ 2
dz (2.19)

From Fig. 2.3,

R  ρ 2  ( z  z ) 2  ρ sec α

z   OT  ρ tan α ,

dz    ρ sec 2 αd α

Eq. (2.19) becomes

©j. k. makiche 23
 α2
 ρL ρ sec 2 α[cos αa ρ  sin αa z ]d α
E
4 πε0 
α1
ρ 2 sec 2 α
α
 ρL 2
4 πε0 ρ α1
 [cos αa ρ  sin αa z ]d α

Thus, for finite line charge,

ρL
E [ (sin α2  sin α1 )a ρ  (cos α2  cos α1 )a z ] (2.20)
4 πε0 ρ

For the case of an infinite line charge, point B is at (0, 0,  ) and A at (0, 0,  ) so that α1  π / 2

and α2   π / 2 . Eq. (2.20) becomes

ρL
E aρ (2.21)
2 πε0 ρ

Eq. (2.21) indicates that E at a point P due to an infinite line charge only depends on the
perpendicular distance from the line charge to the point of interest.

Surface Charge.

Consider an infinite sheet of charge in the xy-plane with uniform charge density ρS . We wish to
determine the electric field at any point P off the plane of the sheet.

The charge associated with elemental area dS is

dQ  ρS dS

Hence, the total charge is

Q   ρS dS
S

From Eq. (2.14), the contribution to the E field at point P (0, 0, h) by the elemental surface 1
shown in Fig. 2.4 is

©j. k. makiche 24
 z

P (0,0, h )

ρS
h R

y
φ ρ
1

Fig. 2.4: Evaluation of E field due to a surface charge

dQ
dE  aR
4 πε0 R 2

R  ρ(a ρ )  ha z

R  R  ρ2  h2

R
aR 
R

dE   S dSd d 

 S  d d [  (a  )  ha z ]
dE 
4 o [  2  h 2 ]3/ 2

Due to symmetry of the charge distribution, for every element 1, there is a corresponding
element 2 whose contribution along a ρ cancels that of element 1, as illustrated in Fig. 2.4. Thus,

the contributions to Eρ add up to zero and the only remaining component of E is the z

component. Therefore,

©j. k. makiche 25
 2 
S h  d d 
E   dE    az
4 o  0  0 [  2  h2 ]3/ 2


ρh 1
 S 2 π  [ ρ2  h2 ]3/ 2 (d ρ)2a z
4 πε0 ρ 0
2

ρS h
 

 [ ρ 2  h 2 ]1/ 2 az
2 ε0 0

ρS
E az
2 ε0

In general, for an infinite sheet of charge,

ρS
E an (2.22)
2ε0

where an is a unit vector normal to the sheet. Eq. (2.22) indicates that E is normal to the sheet
and is independent of the distance between the sheet and the point of observation P.

In a parallel plate capacitor, the electric field existing between two plates having equal and
opposite charges is given by

ρS ρ ρ
E a n  S ( a n )  S a n
2 ε0 2 ε0 ε0

Volume charge

Consider spherical volume charge distribution with uniform charge density ρv be as shown in

Figure 2.5. Show that the electric field at P ( r ,  ,  ) is given by

Q
E ar
4 πε0 r 2

The charge dQ associated with the elemental volume dv is

©j. k. makiche 26
 dQ  ρv dv

z
dE dE z

P (0,0, z )

α R
dv at (r ,θ ,φ )
ρv

r
θ

y
φ

Fig. 2.5: Evaluation of E field due to a volume charge

Hence the total charge in a sphere of radius a is

4π a3
Q   ρv dv  ρv  dv  ρv
v v
3

The electric field dE at P(0, 0, z) due to the elementary volume charge is

ρv dv
E aR
4 πε0 R 2

where

a R  cos αa z  sin αa ρ

Due to the symmetry of the charge distribution, the contributions to Ex or Ey add up to zero. We

are left with only Ez , given by

©j. k. makiche 27
 ρv dv cos α
E z  E  a z   dE cos α  
4 πε0 R2

dv  r 2 sin  dr d d 

R 2  z 2  r 2  2 zr  cos θ 

r 2  z 2  R 2  2 zR cos α

z 2  R 2  r 2
cos α  (2.23a)
2 zR

z 2  r 2  R 2
cos θ  (2.23b)
2 zr 

Differentiating eq. (2.23b) with respect to θ  keeping z and r' fixed, we obtain

Rdr 
sin θd θ 
zr 

Carrying out substitutions,

2π z r
ρv Rdr  z 2  R 2  r 2 1
a
Ez 
4 πε0 
φ
d φ  
r 0 R  z  r 
r 2
zr 
dr 
2 zR R2

z r
ρ 2π 2  z 2  r 2 
a
 v 2
8 πε0 z   
r 0 R  z  r 
r  1 
R2 
 dRdr 

z r
ρv π  z 2  r 2 
a

4 πε0 z 2 r0 r  R  R  dr 

z r

ρv π 1 1 4 3 
a

4 πε0 z 2 
r  0
4r 2 dr    π a ρv 
4 πε0 z 2  3 

or

©j. k. makiche 28
 Q
E az (2.24)
4 πε0 z 2

This result is obtained for E at P (0, 0, z ) . Due to the symmetry of the charge distribution, the
electric field at P ( r ,  ,  ) is obtained from eq. (2.24) as

Q
E ar (2.25)
4 πε0 r 2

which is identical to the electric field at the same point due to a point charge Q located at the
origin or the center of the spherical charge distribution.

Example 4: A circular ring of radius a shown in Fig. 2.6 carries a uniform charge ρL C/m and is
placed on the xy-plane with axis the same as the z-axis.

a) Show that

ρL ah
E(0, 0, h)  az
2ε0 [ h 2  a 2 ]3/ 2

b) What values of h give the maximum value of E?

c) If the total charge on the ring is Q, find E as a  0 .

Solution:

a) We use Eq. (2.16). But first, each term in the equation must be derived. From the figure

dl  ad

R  a(a ρ )  ha z

R  R  a 2  h2

©j. k. makiche 29
 z

dE dE z

dE ρ

h R

x dl

Fig. 2.6: Charged ring; for Example 4.4.

R
aR 
R

or

a R R  aa ρ  ha z
 
R 2 R 3 [a 2  h 2 ]3/ 2

Using Eq. (2.16) with appropriate substitutions,

L
2
aa   ha z
E 
4 o  0 [a 2  h 2 ]3/ 2
ad

By symmetry, the contributions along a ρ add up to zero since for every element dl there is a

corresponding element diametrically opposite it that gives an equal but opposite dE ρ so that the

two contributions cancel each other. Thus we are left with the z-component. Hence,

©j. k. makiche 30
 2
L aha z
4 o [a  h2 ]3/ 2 0
E 2
d

 L ah
 az
2 0 [h2  a 2 ]3/ 2

 2 3 
 [a  h 2 ]3/ 2 (1)  (h)2h[a 2  h 2 ]1/ 2 
dE ρL a 2
a)   
dh 2 ε0  [a 2  h 2 ]3 
 

d E
For maximum E,  0 . This implies that
dh

3
[ a 2  h 2 ]3/ 2 (1)  ( h)2h[ a 2  h 2 ]1/ 2  0
2
[ a  h ] [ a  h 2  3h 2 ]  0
2 2 1/ 2 2

a 2  2h 2  0

or

a2
h
2

b) Since the charge is uniformly distributed, the line charge density is

Q
ρL 
2π a

so that

Qh
E az
4 πε0 [a 2  h 2 ]3/ 2

As a  0 ,

Q
E az
4 πε0 h 2

©j. k. makiche 31
 or in general,

Q
E ar
4 πε0 r 2

which is the same as that of a point charge may be expected.

©j. k. makiche 32
 Gauss Law and the Electric Potential

Electric Flux Density

It is established that E is dependent on the medium in which the charge is placed (free space in
this case). A new vector field D independent of the medium may be defined as

D  ε0 E (2.26)

The electric flux ψ is defined in terms of D as

ψ   D  dS (2.27)

In SI units, one line of electric flux emanates from +1 C and terminates on - 1 C. Therefore, the
electric flux is measured in coulombs. The vector field D is called the electric flux density and is
measured in coulombs per square meter. For historical reasons, the electric flux density is also
called electric displacement.

The formulae for E obtained using Coulomb’s law can be used to determine D by multiplying by
ε0 e.g. for a uniform infinite line charge,

ρL
D aρ
2 πρ

and for an infinite surface charge,

ρS
E an
2

GAUSS LAW

Gauss's law constitutes one of the fundamental laws of electromagnetism and it states that the
total electric flux ψ through any closed surface is equal to the total charge enclosed by that
surface i.e.

©j. k. makiche 33
 ψ  Qenc (2.28)

Using Eq. (2.27) and integrating over a closed surface

ψ   D  dS (2.29)

But,

 D  dS     Ddv
v
(2.30)

where we have used divergence theorem. Also, the general expression for Q is given in Eq.
(2.15) as

Q   ρv dv (2.31)
v

Using Eqs. (2.30) and (2.31) in (2.28),

   Ddv   ρ dv
v v
v (2.32)

The integrands in Eq. (2.32) must be equal since the integrations are carried out over the same
volume. Hence,

  D  ρv (2.33)

Eq. (2.33) is Gauss’ law in point or differential form. It states that the volume charge density is
the same as the divergence of the electric flux density and is one of the fundamental postulates in
electrostatics.

Application of Gauss Law

Gauss's law provides an easy means of finding E or D for symmetrical charge distributions such
as a point charge, an infinite line charge, an infinite cylindrical surface charge, and a spherical
distribution of charge. A continuous charge distribution has rectangular symmetry if it depends

©j. k. makiche 34
only on x (or y or z), cylindrical symmetry if it depends only on ρ , or spherical symmetry if it
depends only on r (independent of θ and φ ).

The procedure for application of Gauss’ law first involves determination of whether symmetry of
charge distribution exists. Once this is positively established, a mathematical closed surface
(known as a Gaussian surface) is constructed. A suitable Gaussian surface must be such that;

1. It passes through the particular point at which the particular component of the field is to
be evaluated. The surface must be normal to that component at the point in question.
2. It is completely enclosed and at every point on the surface, the normal component of the
electric field must either have the same value as that at the point in question or be zero.

A. Point charge

Suppose a point charge Q is located at the origin. We wish to determine E at any point P using
Gauss’ law.

To determine D at a point P, it is easily seen that choosing a spherical surface containing P will
satisfy symmetry conditions. Thus, a spherical surface centered at the origin is the Gaussian
surface in this case and is shown in Figure 2.7. Since D is everywhere normal to the Gaussian
surface,

 D  dS   DdS
D  Dr a r

dS  r 2 sin  d d a r

2 

 dS   r sin  d d  4 r 2
2

S  0  0

©j. k. makiche 35
 z
D

Q y

Fig. 2.7: Gaussian surface about a point charge.

Using Gauss’ law

 DdS  D 4 π r Q
2
r

Q
Dr 
4π r 2

or

Q
D ar
4π r 2

D Q
E  ar (2.34)
ε0 4 πε0 r 2

which is as expected from Eq. (2.9).

B. Uniform Line Charge

An infinite line charge has a uniform linear charge density  L . Using Gauss Law, derive an
expression for the electric field at a radial distance ρ from the axis of the line charge.

©j. k. makiche 36
Suppose the infinite line of uniform charge  L C/m lies along the z-axis. To determine D at a
point P, we choose a cylindrical surface containing P to satisfy symmetry condition as shown in
Figure 4.14. D is constant on and normal to the cylindrical Gaussian surface; that is, D  Dρa ρ . If

we apply Gauss's law to an arbitrary length L of the line,

Q  ρL L   D  dS  Dρ  dS Dρ 2 πρ L

where

 dS  2 πρL
z
ρL (C/m)

P D
L ρ
Q y

Fig. 2.8: Gaussian surface about an infinite line charge

Note that  D  dS evaluated on the top and bottom surfaces of the cylinder is zero. Thus,
Q
Dρ  aρ
2 πρ

and

D ρL
E  aρ (2.35)
ε0 2 πε0 ρ

©j. k. makiche 37
which is as expected from Eq. (2.21).

C. Uniform Surface Charge

Consider the infinite sheet of uniform charge ρS C/m2 lying on the z  0 plane. To determine E
at point P. We choose a rectangular box that is cut symmetrically by the sheet of charge and has
two of its faces parallel to the sheet as shown in Fig. 2.9. Since D is normal to the sheet,
D  Dz a z . Applying Gauss’ law,

 
Q  ρS  dS   D  dS  Dz   dS   dS 
 top bottom 

Note that D  d S evaluated on the sides of the box is zero since there are no a x and a y components.

If the top and bottom of the box each has area A,

Q  ρS A  Dz  A  A

Thus,

P
Infinite sheet of charge ρS C/m 2

Fig. 2.9: Gaussian surface about an infinite sheet of charge.

©j. k. makiche 38
 ρS
D az
2

and

D ρS
E  az (2.36)
ε0 2 ε0

which is the same as Eq. (2.22).

D. Volume charge

Consider a sphere of radius a that is uniformly charged with charge density ρv C/m3 . We wish to

determine E in the regions where r  a and r  a . Since the charge has spherical symmetry, a
spherical surface is an appropriate Gaussian surface. For r  a , the total charge enclosed by the
spherical surface of radius r, as shown in Fig. 2.10 (a), is

2  r
4
Qenc   v dv  v  dv  v   r
2
sin  drd d   v  r 3 C
v v  0  0 r 0
3

2 
   D  dS  Dr  dS  Dr  r
2
sin  d d  Dr 4 r 2 C
S S  0  0

From Gauss’ law, ψ  Qenc , which gives

4 3
Dr 4 π r 2  ρv πr
3

ρv
D ra r
3

and

D ρv
E  ra r 0ra
ε0 3ε0

©j. k. makiche 39
 r

r a a

Fig. 2.10: Gaussian surface for a uniformly charged sphere when: (a) r  a and

(b) r  a .

For r  a , the Gaussian surface is shown in Figure 2.10(b). The charge enclosed by the
surface is the entire charge in this case, that is,

2  a
4
Qenc   v dv  v  dv  v   r
2
sin  drd d   v  a 3 C
v v  0  0 r 0
3

whereas

2 
   D  dS  Dr  dS  Dr  r
2
sin  d d  Dr 4 r 2 C
S S  0  0

Using Gauss’ law,

4 3
Dr 4 π r 2  ρv πa
3

Hence,

ρv a 3
D ar
3r 2

and

©j. k. makiche 40
 D ρv a3
E  ar ra
ε0 3r 2 ε0

In general E is everywhere given by

 ρv
 3ε ra r 0r a
 0
E
 ρv a a
3
ra
 3r 2 ε0 r

A sketch of D against r is shown in Fig.

D
a
ρv
3

ρv a 3
ρv 3r 2ε0
r
3

0 r
a

Fig. 2.11: Sketch of D against r for a uniformly charged sphere.

Example 5: Given D  z  cos 2  a z C/m 2 calculate the charge density at (1, π / 4, 3) and the total

charge enclosed by the cylinder of radius 1 m with 2  z  2 m .

From Eq. 2.33,

Dz 
v    D     z cos2     cos2 
z z

At (1, π / 4, 3)

©j. k. makiche 41
 v   cos 2   1 cos 2 ( / 4)  0.5 C/m3

The total charge enclosed by the cylinder may be found in two ways.

Method 1: Based on the definition of the total volume charge.

2
4
2 1
Qenc   v dv      cos 2  d  d dz  C
v z 2   0   0
3

Method 2: Based on Gauss’ law.

 
Q  ψ   D  dS    +  +   D  dS  ψside  ψtop  ψbottom
S  side top bottom 

where ψside , ψtop and ψbottom are the flux through the sides, the top surface, and the bottom

surface of the cylinder, respectively. Since D does not have component along a ρ , ψside  0 .

For ψtop , dS   d  d a z so,

2
2
1
 top   
 0  0
 z cos 2  d d  
3
C
z 2

For ψbottom , dS    d  d a z , so

2
2
1
 top    
 0  0
 z cos 2  d d  
3
C
z 2

Thus

2π 2π 4π
Q  ψ  0   C
3 3 3

©j. k. makiche 42
ELECTRIC POTENTIAL

Suppose we wish to move a point charge Q from point A to point B in an electric field E along
the path indicated in Figure 2.12. From Coulomb's law, the force on Q is F = QE so that the work
done in displacing the charge by dL is

dW   F  dL  QE  dL (2.38)

A
dl
rA B
r
rB

Fig. 2.12: Displacement of point charge Q in an electrostatic field E.

The negative sign indicates that the work is being done by an external agent. Thus, the total work
done, or the potential energy required in moving the charge Q from A to B is

B
W  Q  E  dL (2.37)
A

The potential difference VAB between points A and B in an electrostatic field E is the work done
by an external agent in moving a unit positive charge between the given points. This is
determined by dividing Eq. (2.37) by Q

B
W
VBA     E  dl (2.38)
Q A

Notes:

©j. k. makiche 43
 1. In determining VBA , A is the initial point while B is the final point.

2. If VBA is negative, there is loss in potential energy in moving Q from A to B. This implies that

work is done by the field. On the other hand, if VBA is positive, there is a gain in potential energy

in the movement; an external agent performs the work.

3. VBA is measured in Joules per Coulomb or Volts.

If the E field in figure 2.12 is due to a point charge Q located at the origin, determine VBA .

Q
E ar
4 πε0 r 2

1 1
B rB
Q Q
VBA    E  dL    a  dra r     V (2.39)
A rA
4 πε0 r 2 r
4 πε0  rB rA 

or

VBA  VB  VA (2.40)

where VB and VA are the potentials (or absolute potentials) at B and A, respectively. Thus the
potential difference VBA may be regarded as the potential at B with reference to A.

The potential at any point is the potential difference between that point and a chosen point in
which the potential is zero. In problems involving point charges, it is customary to choose
infinity as reference; that is, rA   or V A  0 . Eq. (2.39) becomes

Q
VBA  V
4 πε0 rB

Therefore the potential V at any point due to a point charge Q located at the origin is

Q
V V (2.41)
4 πε0 r

©j. k. makiche 44
In other words, by assuming zero potential at infinity, the potential at a distance r from the point
charge is the work done per unit charge by an external agent in transferring a test charge from
infinity to that point. Thus

r
V    E  dl (2.42)


If the point charge Q in Eq. (2.41) is not located at the origin but at a point whose position vector
is r , the potential V (r ) at r becomes

Q
V (r )  (2.43)
4 πε0 r  r

For n point charges Q1 , Q2  Qn located at position vectors r1 , r2  rn , the potential at r is

Q1 Q2 Qn
V (r )   
4 πε0 r  r1 4 πε0 r  r2 4 πε0 r  rn

or

n
1 Qk
V (r ) 
4 πε0
 r r
k 1
(2.44)
k

For continuous charge distributions, Qk in Eq. (2.44) is replaced with ρL dl , ρS dS , ρv dv and the
summation becomes an integral so that the potential at r becomes

1 ρL (r)dl 
V (r ) 
4 πε0 L r  rk (line charge) (2.45)

1 ρS (r)dS 
V (r ) 
4 πε0 
S
r  rk
(surface charge) (2.46)

1 ρv (r)dv
V (r ) 
4 πε0 v r  rk (volume charge) (2.47)

©j. k. makiche 45
where the primed coordinates are used customarily to denote source point location and the
unprimed coordinates refer to field point (the point at which V is to be determined).

We note the following:

1. In obtaining Eqs. (2.43) to (2.47) the zero potential (reference) point has been chosen
arbitrarily to be at infinity. If any other point is chosen as reference, eq. (4.65), for
example, becomes

Q
V C (V) (2.48)
4 πε0 r

where C is a constant that is determined at the chosen point of reference. The same idea
may be applied to Eqs. (2.43) to (2.47).

2. The potential at a point can be determined in two ways depending on whether the charge
distribution or E is known. If the charge distribution is known, we use Eqs. (2.43) to
(2.47). If E is known,
V    E  dl  C (2.49)

The potential difference VBA is obtained as

B
W
VBA  VB  VA    E  dl  (2.50)
A
Q

Example 16: Two point charges 4 μC and 5 μC are located at (2,  1, 3) and (0, 4,  2)
respectively. Find the potential (1, 0,1) assuming zero potential at infinity.

Let Q1  4 μC and Q2  5 μC .

Q1 Q2
V (r )    C0
4 πε0 r  r1 4 πε0 r  r2

If V ( )  0 , C0  0 .

r  r1  (1, 0,1)  (2,  1,3)  (1,1,  2)  6

©j. k. makiche 46
 r  r2  (1, 0,1)  (0, 4,  2)  (1,  4,3)  26

Hence,

106  4 5 
V (1, 0,1)  12 
   5.872 kV
4 π (8.854 10 )  6 26 

Example 17: A point charge 5 nC is located at ( 3, 4, 0) and while line y  1 , z  1 carries a
uniform charge 2 nC/m.

a) If V  0 at O (0, 0, 0) find V at A(5, 0,1) .

b) If V  100 V at B (1, 2,1) find V at C ( 2, 5, 3) .

c) If V  5 V at O , find VBC .

Solution:

Let the potential at any point be

V  VQ  VL

where VQ and VL are the contributions to V due to the point charge and the line charge

respectively. For the point charge,

Q Q
VQ    E  dl    a  dra r   C1
4 πε0 r 2 r
4 πε0 r

For infinite line charge,

ρL ρ
VL    E  dl    a ρ  d ρa ρ  L ln ρ  C2
2 πε0 ρ 2 πε0

ρL Q
V ln ρ  C
2 πε0 4 πε0 r

©j. k. makiche 47
 where C  C1  C2  constant , ρ is the perpendicular distance from the line y  1 , z  1 to
the field point, and r is the distance from the point charge to the field point.

a) ρ is the distance between ( x, y , z ) and ( x,1,1) . Hence,

ρ  ( x, y, z )  ( x,1,1)  ( y  1) 2  ( z  1) 2

r  ( x2  x1 ) 2  ( y2  y1 ) 2  ( z2  z1 ) 2

For points O and A, we have

ρO  (0, 0, 0)  (0,1,1)  2

rO  (0, 0, 0)  ( 3, 4, 0)  5

ρ A  (5, 0,1)  (5,1,1)  1

rA  (5, 0,1)  ( 3, 4, 0)  9

ρL ρ  Q 1 1
VO  VA   ln  O     
2 πε0  ρ A  4 πε0  rO rA 
2  109  2 5  10 9 1 1 
 ln    12 
 
2 π  8.854  10 12  1  4 π  8.854 10  5 9 

1 1 
0  VA  36 ln 2  45   
5 9

V A  36 ln 2  4  8.477 V

b) If V  100 at B (1, 2,1) and V at C ( 2, 5, 3) is to be determined

ρB  (1, 2,1)  (1,1,1)  1

rO  (1, 2,1)  (3, 4, 0)  21

©j. k. makiche 48
 ρ A  (2,5,3)  (2,1,1)  20

rA  (2,5,3)  (3, 4, 0)  11

ρL ρ  Q 1 1
VC  VB   ln  O     
2 πε0  ρB  4 πε0  rC rB 

 20   1 1 
VC  100  36 ln    45     50.175 V
 1   11 21 

or

VC  49.825 V

c) VBC  VC  VB  49.825  100  50.125 V

Relationship between E and V

In Eq. (2.39) it is observed that VBA does not depend on θ or  but only on r . In other words,

VBA does not depend on the exact nature of the path between A and B, but only on the radial

distance between the points. In general, the potential difference VBA is not dependent on the path
chosen for the line integral, regardless of the source of E field. Hence,

VBA  VAB

That is

VBA  VAB   E  dl  0

or

 E  dl  0 (2.51)

©j. k. makiche 49
Physically, this implies that no net work is done in moving a charge around a closed path in an
electrostatic field. Applying Stoke’s theorem to Eq. (2.51),

 E  dl   (  E)  dS  0
S

or

E  0 (2.52)

Any vector field that satisfies eq. (2.51) or (2.52) is said to be conservative, or irrotational. Thus
an electrostatic field is a conservative field. Later, it will be shown that Eq. (2.52) is Faraday’s
law for electrostatic fields.

From the definition of potential i.e. V    E  dl , it follows that

dV  E  dl   Ex dx  E y dy  Ez dz

But for V  V ( x, y, z ) its total differential is given as

V V V
dV  dx  dy  dz
x y z

Therefore

V V V
Ex   Ey   Ez  
x y z

Thus

V V V
E  a x E x  a y E y  a z E z  a x  ay  az
x y z

or

E   V (2.53)

©j. k. makiche 50
that is, the electric field intensity is the gradient of V. The negative sign shows that the direction
of E is opposite to the direction in which V increases; E is directed from higher to lower levels of
V. Equation (2.53) shows another way to obtain the E field apart from using Coulomb's or
Gauss's law. That is, if the potential field V is known, the E can be found using eq. (2.53).

10
Example 18: Given the potential V  sin θ sin φ
r2

a) Find the electric flux density D at (2, π / 2, 0) .

b) Calculate the work done in moving a 10 μC charge from a point A(1,30 ,120 ) to

B (4,90 , 60 )

Solution:

a) D  ε0 E

But

 V 1 V 1 V 
E  V    a r  a  a 
 r r  r sin   
20 10 10
 3 sin  sin  a r  3 sin  sin  a  3 sin  a
r r r

b) At (2, π / 2, 0)

 20 
D   0 E(r  2,   / 2,   0)   0  a r  0a  0a   2.5 0a r C/m 2
 8 
B
W  Q  E  dl  QVAB  Q(VB  VA )
A

 10 10 
 10  sin 90 cos 60  sin 30 cos120  106  28.125 μJ
 16 1 

ELECTRIC DIPOLE

An electric dipole is formed when two point charges of equal magnitude but opposite sign are
separated by a small distance, d. Fig. (2.14) shows an electric dipole centered at the origin. The

©j. k. makiche 51
dipole arrangement is fundamental to the understanding of the behavior of electric field in
dielectrics.

R1
R
Q
R2
θ
p
y

Q
x d cos θ

Fig. 2.13: An electric dipole

The potential at point P may be determined from as follows:

Q Q
V1  and V2 
4 πε0 R1 4 πε0 R2

Therefore

Q  1 1  Q  R2  R1 
V  V1  V2       (2.53)
4 πε0  R1 R2  4 πε0  R2 R1 

For a distant point P, the geometry of the problem will be as shown below in Fig. 2.14. We then
have;

R1 R2  r 2
R2  R1  d cos θ

So that

©j. k. makiche 52
 Q p cos θ
V d cos θ 
4 πε0 R 2
4 πε0 R 2

or

p  ar
V (r )  V (2.54)
4 πε0 R 2

R1

R
Q
θ
R2
p
y

Q
x d cosθ

Fig. 2.14: Electric dipole for distant point P.

where p  Qd is a vector directed from –Q to +Q and is known as the electric dipole moment.
When the dipole is not centred at the origin but at a point whose position vector is r , V (r )
becomes

p  ar
V (r )  V (2.55)
4 πε0 r  r
2

To determine E, we use

E   V

©j. k. makiche 53
For spherical co-ordinates,

V 1 V 1 V
E  V   ar  a  a
r r  r sin  

We obtain

Qd cos  Qd sin 
E ar  a
4 0 r 3
4 0 r 3

Therefore,

Qd
E (2 cos θa r  sin θaθ ) (2.56)
4 πε0 r 3

or

p
E (2 cos  a r  sin  a ) V/m (2.57)
4 0 r 3

where p  p  Qd

ENERGY DENSITY IN ELECTROSTATIC FIELDS

Suppose we wish to position three point charges Q1 , Q2 and Q3 in that order in an initially empty

space as shown in Fig. 2.15. No work is done in transferring Q1 from infinity to a point P1 in the

space since the space is initially charge free and there is no electric field. Hence W1  0 . The

work done in transferring Q2 to a point P2 is W2  Q2V21 where V21 is the potential at P2 due to Q1 .

Similarly, the work done in positioning Q3 at P3 is W3  Q3 (V32  V31 ) where V32 and V31 are the

potentials at P3 due to Q2 and Q1 respectively. The total work done in positioning the three
charges therefore is

WE  W1  W2  W3  0  Q2V21  Q3 (V32  V31 ) (2.60)

©j. k. makiche 54
 Q1 P1

Q2 P2

Q3 P3

Fig. 2.13: Assembly of charges

If the charges were positioned in the reverse order,

WE  W3  W2  W1  0  Q2V23  Q1 (V12  V13 ) (2.61)

where V23 is the potential at P2 due to Q3 , V12 and V13 are respectively the potentials at P1 due to Q2

and Q3 respectively. Adding eqs. (2.60) and (2.61),

2WE  Q1 (V12  V13 )  Q2 (V21  V23 )  Q3 (V32  V31 )

 QV
1 1  Q2V2  Q3V3

or

1
WE  (QV
1 1  Q2V2  Q3V3 ) (2.62)
2

where V1 , V2 and V3 are the total potentials at P1 , P2 and P3 respectively. In general, if there are n
point charges, eq. (2.56) becomes

1 n
WE   QkVk
2 k 1
(2.63)

If, instead of point charges, the region has a continuous charge distribution, the summation in eq.
(2.63) becomes integration; that is,

1
2 l
WE  ρLVdl (2.64)

©j. k. makiche 55
 1
2 S
WE  ρSVdS (2.65)

1
2 v
WE  ρvVdv (2.66)

Using ρv    D in eq. (2.66),

1
2 v
WE    DVdv (2.67)

But for any vector A and scalar V,

  AV  A V  V (  A)

or

V (  A)    AV  A V (2.68)

Using eq. (2.68) in (2.67)

1 1
WE 
2    DVdv   D Vdv
2

Applying divergence theorem,

1 1
WE 
2S
 VD  dS   D Vdv
2v
(2.69)

For a point charge V and D vary as 1/ r and 1/ r 2 respectively. For an electric dipole, V and D
vary as 1/ r 2 and 1/ r 3 respectively and so on. Consequently, the first integral in eq. (2.69) must
tend to zero as the surface S becomes large. Hence eq. (2.69) reduces to

1 1
WE  
2v D Vdv   D  Edv
2v
(2.70)

©j. k. makiche 56
since E   V . Using D  ε0 E ,

1 1 1 D2
2 v 2 v 2 v ε0
WE  D  E dv  ε0 E 2
dv  dv (2.71)

From eq. (2.65), the electrostatic energy density wE in J/m3 may be defined as

dWE 1 1 D2
wE   D  E  ε0 E 2  (2.72)
dv 2 2 2ε0

Eq. (2.65) may therefore be written as

1
2 v
WE  wE dv (2.73)

Example 19: Three point charges - 1 nC, 4 nC, and 3 nC are located at (0,0,0), (0,0,1), and
(1,0,0), respectively. Find the energy in the system.

WE  W1  W2  W3
 0  Q2V21  Q3 (V32  V31 )
Q1 Q  Q1 Q2 
 Q2  3   
4 πε0 (1, 0, 0)  (0, 0, 0) 4 πε0  (1, 0, 0)  (0, 0, 0) (1, 0, 0)  (0, 0,1) 

1  Q3Q2 
 Q2Q1  Q3Q1  
4 πε0  2 

1  12  18 
 12 
4  3   10   13.37 nJ
4 π  8.854  10  2 

1 3
Alternatively, we may solve the problem using WE   QkVk which gives the same answer.
2 k 1

Example 20: A charge distribution with spherical symmetry has density

©j. k. makiche 57
 ρ 0r  R
ρv   0
0 R  r,

Determine V everywhere and the energy stored in the region R  r

a) Using Gauss’ law for r  R ,

4
π R 3 ρ0  Dr 4 π r 2
3

which gives

R 3 ρ0
Dr 
3r 2

or

R3 ρ0 R3 ρ0
D a  E  ar
3r 2 ε0
r
3r 2

R3 ρ0 1 R3 ρ0
V    E  dl  
3ε0  r 2
dr   C1 , rR
3ε0 r

Since V (r  )  0, C1  0 . Hence

R 3 ρ0
V , rR
3ε0 r

b) For R  r , Gauss’ law gives

4 3
π r ρ0  Dr 4π r 2
3

rρ0 rρ
D ar or E  0 a r
3 3ε0

©j. k. makiche 58
 ρ0 r 2 ρ0
V    E  dl  
3ε0 
rdr    C2 , rR
6ε0

R 2 ρ0
From part (a), V (r  R)  , Hence,
3ε0

R 2 ρ0 R 2 ρ0 R 2 ρ0
  C2  C2 
3ε0 6ε0 2ε0

and

r 2 ρ0 R 2 ρ0 ρ0
V    (3R 2  r 2 ), rR
6ε0 2ε0 6ε0

From parts (a) and (b), we have

 ρ0
 6ε (3R  r ) 0  r  R
2 2

 0
V 
 R 3 ρ0
R  r,
 3ε0 r

The energy stored is given by

1 1
WE 
2v D  Edv  ε0  E 2 dv
2 v

For r  R ,

rρ0
E ar
3ε0

 2 R
1  02  02 2 02 R 5
R
r5
WE           
2 2
r r sin d d dr 4 J
2 9 0 r 0  0  0
18 0 5 0
45 0

©j. k. makiche 59
Tutorial Problems

1. Three point charges Q1  50nC , Q2  70nC and Q3  50nC are located at (0,0,0), (0,4,2)
and (-2,1,4) respectively in the Cartesian plane. Determine the electric field at (1,1,1) due
to the charges. ( 86.02a x  37.76a y  86.89a z V/m )

2. Point charges 5 nC and 2 nC are located at (2, 0, 4) and ( 3, 0, 5) respectively. Determine
the force on a 1 nC charge located at (1, 3, 7) and the electric field intensity at that point.

(1.004a x  1.284a y  1.4a z nN , 1.004a x  1.284a y  1.4a z V/m )

3. A continuous volume charge defined as ρv  [ x 2  y 2  z 2 ]5/ 2 C/m3 is distributed in the

region 0  x  1 , 0  y  1 , 0  z  1 and is zero elsewhere. Find E x at the origin using

Coulomb’s law. (5.25  109 V/m)

4. A circular disk of radius a is uniformly charged with ρS C/m2 . If the disk lies in the z  0
plane, with its axis along the z-axis,

a) Show that

ρS  h 
E(0, 0, h)  1  2 2 1/ 2  z
a
2ε0  [h  a ] 

b) From this, derive the E field due to an infinite sheet of charge on the z = 0 plane.

c) If a  h show that E is similar to that of a point charge.

ρS
(proof; ; proof)
2ε0

5. The finite sheet 0  x  1, 0  y  1 on the z  0 plane has a charge density

ρS  xy[ x 2  y 2  25]3/ 2 C/m 2 . Find,

a) the total charge on the sheet.

©j. k. makiche 60
 b) the electric field at (0,0,5)

c) the force experienced by a 1 mC charge located at (0,0,5).

( 33.15 nC ; 1.5a x  1.5a y  11.25a z V/m ; 1.5a x  1.5a y  11.25a z mN )

6. A square plate described by 2  x  2 , 2  y  2 , z  0 carries a charge 12 y mC/m 2 .

Find the total charge on the plate and the electric field intensity at (0, 0, 10). ( 192 mC ,
16.46a z MV/m )

7. Planes x = 2 and y = -3, respectively, carry charges 10 nC/m2 and 15 nC/m2. If the line x =
0, z = 2 carries charge 10 π nC/m , calculate E at (1, 1, -1) due to the three charge

distributions. ( 162 π a x  270 π a y  54 π a z V/m )

8. Determine D at (4, 0, 3) if there is a point charge —5 π mC at (4, 0, 0) and a line charge 3

π mC/m along the y-axis.

9. The region in which 4  r  5 , 0  θ  25 and 0.9 π  φ  1.1π contains a volume charge

density ρv  10(r  4)sin θ cos 12 φμC/m3 . Outside the region, ρv  0 . Determine the total
electric flux due to the charge distribution.

10. Given that D  (2 y 2  z )a x  4 xya y  xya z C/m 2 , find

a) The volume charge density at ( 1, 0, 3)

b) The flux through the cube defined by 0  x  1 , 0  y  1 , 0  z  1 .

c) The total charge enclosed by the cube.
11. A charge distribution with spherical symmetry is given by

 ρ0 r
 0r  R
ρv   R
 0 R  r,

Determine E everywhere.

©j. k. makiche 61
12. A charge distribution has in free space ρv  2r nC/m3 for 0  r  10 m and zero otherwise.

Determine E at r  2 m and r  2 m
13. Given that

12 ρ nC/m 3 1 ρ  2
ρv  
 0 elsewhere,

Determine D everywhere.

14. Given that E  (3 x 2  y )a x  xa y kV/m find the work done in moving a 2 μC charge from

(0,5,0) to (2,-1,0) by taking the path

a) (0, 5, 0)  (2, 5, 0)  (2, 1, 0)

b) y  5  3x

15. Determine the charge density due to each of the following electric flux densities

a) D  8 xya x  4 x 2a y C/m 2

b) D   sin  a   2  cos  a  2 z 2a z C/m 2

2cos θ sin θ
c) D  3
ar  3 aθ C/m2
r r

16. A spherical charge distribution is given by

 r
ρ , r  a
ρv   0 a
 0 ra

Find V everywhere.

17. Verify whether E  yza x  xza y  xya y V/m is an electrostatic field,

a) Using   E  0 .

©j. k. makiche 62
 b) By showing that  E  dl  0
18. For a spherical charge distribution

 ρ ( r 2  a 2 ), r  a
ρv   0
 0 ra

a) Find E and V for r  a .

b) Find E and V for r  a .

c) Find the total charge.

d) Show that E is maximum when r  0.145 a .

19. If V  x  y  xy  2 z V find E at (1,2,3) and the electrostatic energy stored in a cube of

side 2m centered at the origin.

20. Determine the work necessary to transfer charges Q1 = 1 mC and Q2 = —2 mC from

infinity to points (— 2, 6, 1) and (3, —4, 0), respectively.

21. If V   2 z sin  V calculate the energy within the region defined by 1  ρ  4 , 2  z  2 ,

0   /3.

22. Find the energy stored in the hemispherical region r  2 m , where the field

E  2r sin  cos  a r  r cos  cos  a  r sin  a V/m

exists.

23. The electric flux lines (or the electric lines of force) are the lines to which the electric field
density D is tangential at every point. The direction of these lines at a given point is the
direction of the field at that point. Draw the electric flux lines for

a) a negative point charge.

©j. k. makiche 63
 b) a positive point charge.

c) an electric dipole.

©j. k. makiche 64
 Electric Fields in Material Space

ELECTRIC CURRENT

Materials are broadly classified in terms of their electrical properties as conductors and
nonconductors. Nonconducting materials are usually referred to as insulators or dielectrics. Just
as electric fields can exist in free space, they can also exist in material media. Before examining
how electric field behaves in a conductor or dielectric, it is appropriate to consider electric
current.

Any motion of electric charge constitutes an electric current. The current (in amperes) through a
given area is the electric charge passing through the area per unit time. That is

dQ
I (3.1)
dt

Thus a current of one ampere is said to be flowing if charge is being transferred at a rate of one
coulomb per second.

If current I flows through a surface S perpendicular to the surface, the current density in
ampere per square metre (A/m2) is

I
Jn 
S

or

I  J n S

If the current is not normal to the surface, S

I  J  S

In general, the total current flowing through a surface S is

©j. k. makiche 65
 I   J  dS (3.2)
S

The current density at a given point is the current through a unit normal area at that point.

Depending on how I is produced, there are different kinds of current densities: convection
current density, conduction current density, and displacement current density. We will consider
convection and conduction current densities here.

Convection current, unlike conduction current, does not involve conductors and consequently
does not satisfy Ohm's law. It occurs when current flows through an insulating medium such as
liquid, rarefied gas, or a vacuum. A beam of electrons in a vacuum tube, for example, is
convection current.

Consider a filament of Figure 3.1. If there is a flow of charge, of density ρv at velocity v  a y a y

, from eq. (3.1), the current through the filament is

Q 
I  ρv S  ρv Sv y
t t

S ρv



Fig. 3.1: Current in a filament

The y-directed current density J y is given by

I
Jy   ρv v y
S

For the general case, v  vx a x  v y a y  vz a z , so that

©j. k. makiche 66
 J  ρv v (3.3)

The current I is the convection current and J is the convection current density in amperes/square
meter (A/m2).

Conduction current requires a conductor. A conductor is characterized by a large amount of free

electrons that provide conduction current due an impressed electric field. When an electric field
E is applied, the force on an electron with charge -e is

F   eE

Since the electron is not in free space, it will not be accelerated under the influence of the electric
field. Rather, it suffers constant collision with the atomic lattice and drifts from one atom to
another. If the electron with mass m is moving in an electric field E with an average drift
velocity v , according to Newton's law, the average change in momentum of the free electron
must match the applied force. Thus,

mv
  eE
τ

or

eτ
v E
m

where τ is the average time interval between collisions. This indicates that the drift velocity of
the electron is directly proportional to the applied field. If there are n electrons per unit volume,
the electronic charge density is given by

ρv  ne

Thus the conduction current density is

ne 2 τ
J  ρv v  E
m

©j. k. makiche 67
or

J  σE (3.4)

where σ  ne2τ / m is the conductivity of the conductor. The relationship in eq. (3.4) is known as
the point form of Ohm's law.

CONDUCTORS

A conductor has abundance of charge that is free to move. Consider an isolated conductor, such
as shown in Figure 3.2(a). When an external electric field Ee is applied, the positive free charges
are pushed along the same direction as the applied field, while the negative free charges move in
the opposite direction. This charge migration takes place very quickly. The free charges do two
things. First, they accumulate on the surface of the conductor and form an induced surface
charge. Second, the induced charges set up an internal induced field Ei, which cancels the
externally applied field Ee. The result is illustrated in Figure 3.2(b). This leads to an important
property of a conductor, that a perfect conductor cannot contain an electrostatic field within it.

Because E   V  0 a conductor is called an equipotential body, implying that the potential is

the same everywhere in the conductor. According Gauss’ law, if E  0 , then ρv  0 . Hence,
under static conditions,

E  0 , ρv  0 , and Vab  0 inside a conductor. (3.5)

   
  Ee   Ee
Ei   
  Ee  ρv  0  Ee
Ei   
  Ee  E0  Ee
   
E i   
  Ee   Ee

Fig. 3.2 (a) An isolated conductor under the influence of an applied field; (b) a conductor has
zero electric field under static conditions.

©j. k. makiche 68
Suppose the conductor ends are maintained at a potential difference V as shown in Fig. 3.3.
There is no static equilibrium since the conductor is not isolated but wired to a source of
electromotive force, which compels the free charges to move and prevents the eventual
establishment of electrostatic equilibrium. Thus in this case an electric field must exist inside the
conductor to sustain the flow of current. As the electrons move, they encounter damping forces
called resistance.

I   
  

Fig. 3.3: A conductor of uniform cross section under an applied E field

Suppose the conductor has a uniform cross section S and is of length  . The direction of the
electric field E produced is the same as the direction of the flow of positive charges or current I.
This direction is opposite to the direction of the flow of electrons. The electric field applied is
uniform and its magnitude is given by

V
E (3.6)


Since the conductor has a uniform cross section,

I
J (3.7)
S

Substituting eq. (3.4) and (3.6) in (3.7)

I σV
 σE 
S 

©j. k. makiche 69
Hence,

V 
R 
I σS

or

ρc 
R (3.8)
S

where ρc  1/ σ is the resistivity of the material. Equation 3.8 is useful in determining the
resistance of any conductor of uniform cross section. If the cross section of the conductor is not
uniform, the basic definition of resistance R as the ratio of the potential difference V between the
two ends of the conductor to the current I through the conductor still applies so that we have

V  E  dl
R  (3.9)
I σ  E  dS
S

Note that the negative sign before V    E  dl is dropped in eq. (3.9) since  E  dl  0 if I  0 .
Power P (in watts) is defined as the rate of change of energy W (in joules) or force times
velocity. Hence,

 ρ dvE  v   E  ρ vdv
v v

or

P   E  Jdv (3.10)

Eq. (3.10) is known as Joule’s law. The power density wP (in watts/m3) is given by the integrand
in eq. (3.10) i.e.

dP
wp   EJ  σ E
2
(3.11)
dv

©j. k. makiche 70
For a conductor with uniform cross section, dv = dS dl. Eq. (3.10) becomes

P   Edl  JdS  VI  I 2 R (3.12)

L S

which is the more common form of Joule's law in electric circuit theory.

1
Example 3.1: If J  3
(2 cos θa r  sin θa θ ) A/m 2 calculate the current passing through
r

a) A hemispherical shell of radius 20 cm

b) A spherical shell of radius 10 cm

I   J  dS
S

a) In this case, dS  r 2 sin  d d a r

2 2 2
1 2
I  0 0 r 3 2 cos r sin  d d 
r  0
2 sin  d (sin  )
2

r  0.2 r  0.2

π /2
4 π sin 2 θ
  10 π  31.4 A
0.2 2 0

b) The only difference here is that we have 0  θ  π instead of 0  θ  π / 2 and r  0.1 .

Hence,

π
4 π sin 2 θ
I 0
0.1 2 0

DIELECTRICS AND POLARIZATION

Unlike conductors, dielectrics do not have free electrons in the atomic outermost shells to
conduct current. The charges in a dielectric are tightly bound to their parent atoms and are
referred to as bound charges. When an external field is applied to a dielectric, a displacement of
these bound charges occurs.

©j. k. makiche 71
Consider an atom or molecule of the dielectric as consisting of a negative charge Q (electron
cloud) and a positive charge Q (nucleus). Since we have equal amounts of positive and
negative charge, the whole atom or molecule is electrically neutral. When an electric field E is
applied, the positive charge is displaced from its equilibrium position in the direction of E by the
force F  QE while the negative charge is displaced in the opposite direction by the force

F  QE . A dipole results from the displacement of the charges and the dielectric is said to be
polarized. In the polarized state, the electron cloud is distorted by the applied electric field E.
This distorted charge distribution is equivalent, by the principle of superposition, to the original
distribution plus a dipole whose moment is

p  Qd (3.13)

where d is the distance vector from Q to Q of the dipole. If there are N dipoles in a volume
v of the dielectric, the total dipole moment due to the electric field is

N
Q1d1  Q2d 2    QN d N   Qk d k
k 1

We define polarization P (in coulombs/meter square) as the dipole moment per unit volume of
the dielectric; that is,

N
lim  Qk d k
v  0
P k 1
(3.14)
v

Thus we conclude that the major effect of the electric field E on a dielectric is the creation of
dipole moments that align themselves in the direction of E. This type of dielectric is said to be
non-polar. Examples of such dielectrics are hydrogen, oxygen, nitrogen, and the rare gases. Non-
polar dielectric molecules do not possess dipoles until the application of the electric field.

Other types of molecules such as water, sulfur dioxide, and hydrochloric acid have built-in
permanent dipoles that are randomly oriented as shown in Figure 3.4(a) and are said to be polar.
When an electric field E is applied to a polar molecule, the permanent dipole experiences a
torque tending to align its dipole moment parallel with E as in Figure 3.4(b).

©j. k. makiche 72
Consider a dielectric material consisting of dipoles with dipole moment P per unit volume as
shown in fig. 3.5. The potential dV at an exterior point O due to the dipole moment Pdv' is


E

Fig. 3.4: Polarization of a polar molecule: (a) permanent dipole (E = 0), (b) alignment of
permanent dipole ( E  0 ).

P  a R dv
dV  (3.15)
4 πε0 R 2

where R 2  ( x  x) 2  ( y  y) 2  ( z  z ) 2 z') and R is the distance between the volume element
dv' at ( x, y, z) and the field point O (x, y, z). But

1 aR
 
R R2

O( x , y, z )
R

aR
dv
( x , y , z )

Fig. 3.5: A block of dielectric material with dipole moment P per unit volume.

Thus

©j. k. makiche 73
 P  aR 1
 P   
R
2
R

Applying the vector identity   fA  f   A  A f ,

P  aR P   P
2
   
R R R

Substituting this into eq. (3.14) and integrating over the entire volume v' of the dielectric, we
obtain

1   P   P  
V    R  R  dv
v
4 πε0

Applying divergence theorem to the first term leads finally to

P  an   P
V  dS    dv (3.16)
S
4 πε0 R v
4 πε0 R

where an is the outward normal to the surface dS  of the dielectric. Comparing the two terms on
the right side of eq. (3.16) with eqs. (2.46) and (2.47) shows that the two terms denote the
potential due to surface and volume charge distributions with densities (upon dropping the
primes),

ρ ps  P  an

ρ pv    P

In other words, eq. (3.16) reveals that where polarization occurs, an equivalent volume charge
density ρ pv is formed throughout the dielectric while an equivalent surface charge density ρ ps is

formed over the surface of the dielectric. We refer to ρ ps and ρ pv as bound (or polarization)

surface and volume charge densities, respectively, as distinct from free surface and volume
charge densities ρs and ρv .Bound charges are those that are not free to move within the
dielectric material; they are caused by the displacement that occurs on a molecular scale during

©j. k. makiche 74
polarization. Free charges are those that are capable of moving over macroscopic distance as
electrons in a conductor; they are the stuff we control. The total positive bound charge on surface
S bounding the dielectric is

Qb   P  dS   ρ ps dS (3.17)

while the charge that remains inside S is

Qb   ρ pv dv      Pdv (3.18)

v v

The total charge Qt in the dielectric remains zero since

Qt   ρ ps dS   ρ pv dv  Qb  Qb  0
v

This is expected since the dielectric remains electrically neutral as was the case before
polarization.

Consider the case in which the dielectric region contains free charge. If ρv is the free charge

volume density, the total volume charge density ρt , is given by

ρt  ρv  ρ pv    ε0E

Hence,

ρv  ρt  ρ pv    ε0 E    P    (ε0E  P)    D

where

D  ε0 E  P (3.19)

We conclude that the net effect of the dielectric on the electric field E is to increase D inside it
by amount P. In other words, due to the application of E to the dielectric material, the flux
density is greater than it would be in free space.

©j. k. makiche 75
The polarization P would vary directly as the applied electric field E. For some dielectrics, P is
related to E as follows:

P  χ e ε0 E (3.20)

where χ e is known as the electric susceptibility of the material (dimensionless) and is a measure
of how susceptible (or sensitive) a given dielectric is to electric fields.

Substituting eq. (3.20) into eq. (3.19), we obtain

D  ε0 (1  χ e )E  ε0 εr E

or

D  εE (3.21)

where

ε  ε0 εr (3.22)

is the absolute permittivity (or simply permittivity) of the dielectric.

ε
εr  1  χ e  (3.23)
ε0

is called the dielectric constant or relative permittivity while ε0 is the permittivity of free space.

So we have assumed ideal dielectrics. Practically speaking, no dielectric is ideal. When the
electric field in a dielectric is sufficiently large, it begins to pull electrons completely out of the
molecules, and the dielectric becomes conducting. Dielectric breakdown is said to have occurred
when a dielectric becomes conducting. Dielectric breakdown occurs in all kinds of dielectric
materials (gases, liquids, or solids) and depends on the nature of the material, temperature,
humidity, and the amount of time that the field is applied. The minimum value of the electric
field at which dielectric breakdown occurs is called the dielectric strength of the dielectric
material.

©j. k. makiche 76
LINEAR, ISOTROPIC AND HOMOGENEOUS DIELECTRICS

A material is said to be linear if D varies linearly with E and nonlinear otherwise. Materials for
which ε (or σ ) does not vary in the region being considered and is therefore the same at all
points (i.e., independent of x, y, z) are said to be homogeneous. They are said to be
inhomogeneous (or non-homogeneous) when ε is dependent of the space coordinates. The
atmosphere is a typical example of an inhomogeneous medium; its permittivity varies with
altitude. Materials for which D and E are in the same direction are said to be isotropic. That is,
isotropic dielectrics are those which have the same properties in all directions. For anisotropic
(or non-isotropic) materials, D, E, and P are not parallel. For example, instead of eq. (3.21), we
have

 Dx   ε xx ε xy ε xz   Ex 
 D   ε ε yy

ε yz   E y  (3.24)
 y   yx
 Dz   ε zx ε zy ε zz   Ez 

The nine components of ε in eq. (3.24) are collectively referred to as a tensor. Crystalline
materials and magnetized plasma are anisotropic.

We will be concerned only with linear, isotropic, and homogeneous media. For such media, all
formulas derived in for free space can be applied by merely replacing ε0 with ε0 εr .

Example 3.2: A dielectric cube of side L and center at the origin has a radial polarization given
by P  ar , where a is a constant and r = xax + yay + zaz. Find all bound charge densities and
show explicitly that the total bound charge vanishes.

Solution:

For each of the six faces of the cube, there is a surface charge ρ ps .For the face located at

x = L/2, for example,

aL
ρ ps  P  a x xL / 2
 ax x  L / 2 
2

©j. k. makiche 77
 The total bound surface charge is

L/2 L/2
aL 2
Qs   ρ ps dS  6   ρ ps dydz  6 L  3aL3
L/ 2 L/ 2
2

The bound volume charge density is given by

ρ pv    P  (a  a  a)  3a

The total bound volume charge is

Qv   ρ pv dv  3a  dv  3aL3
v

Hence the total charge is

Qt  Qs  Qv  3aL3  3aL3  0

Example 3.3: The electric field intensity in polystyrene ( εr = 2.55) filling the space between the
plates of a parallel-plate capacitor is 10 kV/m. The distance between the plates is 1.5 mm.
Calculate:

a) D
b) P
c) The surface charge density of free charge on the plates
d) The surface density of polarization charge
e) The potential difference between the plates

Solution

a) D  ε0 εr E  8.854 1012  2.25  104  225.4 nC/m 2

b) P  χ e ε0 E  1.55  8.854 1012  104  137 nC/m 2

c) ρS  D  a n  Dn  225.4 nC/m 2

d) ρ ps  P  a n  Pn  137 nC/m 2

©j. k. makiche 78
 e) V  Ed  104 (1.5  103 )  15 V

CONTINUITY EQUATION

Due to the principle of charge conservation, the time rate of decrease of charge within a given
volume must be equal to the net outward current flow through the closed surface of the volume.
Thus current I out coming out of the closed surface is

 dQin
I out   J  dS  (3.26)
dt

where Qin is the total charge enclosed by the closed surface. Applying divergence theorem,

 J  dS     Jdv
v
(3.27)

But

dQin d ρ
dt
 
dt v
ρv dv    v dv
v
t
(3.28)

Substituting eqs. (3.27) and (3.28) into eq. (3.26) gives

ρv
   Jdv   
v v
t
dv

or

ρv
J   (3.29)
t

Eq. (3.29) is called the continuity of current equation. The continuity equation is derived from
the principle of conservation of charge and essentially states that there can be no accumulation of

©j. k. makiche 79
charge at any point. For steady currents, ρv / t  0 and hence   J  0 showing that the total
charge leaving a volume is the same as the total charge entering it. Kirchhoff's current law
follows from this.

BOUNDARY CONDITIONS

If the field exists in a region consisting of two different media, the conditions that the field must
satisfy at the interface separating the media are called boundary conditions. These conditions are
helpful in determining the field on one side of the boundary if the field on the other side is
known.

Dielectric-Dielectric Boundary Conditions

Consider the E field existing in a region consisting of two different dielectrics characterized by
ε1  ε0 εr1 l and ε2  ε0 εr 2 as shown in Figure 3.6.

ε1
S

En 1 E1
a w b h
Et 2 ρS
h
En 2 d c
w
E2

Et 2 ε2

Fig. 3.6: Dielectric-dielectric boundary.

E1 and E2 in media 1 and 2, respectively, can be decomposed as

E1  Et1  E n1 (3.30)

E 2  Et 2  E n 2 (3.31)

©j. k. makiche 80
Applying  E  dl  0 closed path abcda of Fig. 3.6 and assuming that the path is very small with
respect to the variation of E, we obtain

1 1 1 1
Et1w  En1h  En 2 h  Et 2 w  En 2 h  En1h  0 (3.32)
2 2 2 2

where Et  E t and En  E n . As h  0 , eq. 3.32 becomes

Et1  Et 2 (3.33)

Thus the tangential components of E are the same on the two sides of the boundary. In other
words, E undergoes no change on the boundary and it is said to be continuous across the
boundary. Since D  εE  Dt  Dn , eq. (3.33) may be written as

Dt1 Dt 2
 (3.34)
ε1 ε2

that is, Dt, undergoes some change across the interface. Hence Dt, is said to be discontinuous
across the interface.

To obtain the boundary conditions for the normal component, we apply Gauss’ law  D  dS  Q
to the pillbox (Gaussian surface) of Fig. 3.6. Allowing h  0 , we have

Q  ρS S  Dn 2 S  Dn1S

which gives

Dn 2  Dn1  ρS (3.35)

where ρS is the surface charge density at the interface. In this case, the field is directed from
medium 2 to medium 1. If the field is directed from medium 1 to medium 2, eq. (3.35) becomes

Dn1  Dn 2  ρS (3.36)

©j. k. makiche 81
If no free charges exist at the interface (which is usually the case) ρS  0 and eq. (3.36) becomes

Dn1  Dn 2 (3.37)

Thus the normal component of D is continuous across the interface. Since D  εE , eq. (3.37)
becomes

ε1 En1  ε2 En 2 (3.38)

Thus, the normal component of E is discontinuous at the boundary. Eqs. (3.33) and (3.36), or
(3.37) are collectively referred to as boundary conditions. They must be satisfied by an electric
field at the boundary separating two different dielectrics.

E2 θ ε2
E2n 2

E2t

E1n θ E
1 1

ε1
E1t

Fig. 3.7: Refraction of E at a dielectric-dielectric boundary.

Consider E1, and E2 making angles θ1 and θ2 with the normal to the interface as illustrated in
Fig. 5.11. Using eq. (3.33), we have,

E1 sin θ1  Et1  Et 2  E2 sin θ2

or

E1 sin θ1  E2 sin θ2 (3.38)

Similarly, using eq. (3.37) or (3.38),

ε1 E1 cos θ1  ε1 En1  ε2 En 2  ε2 E2 cos θ2

©j. k. makiche 82
or

ε1 E1 cos θ1  ε2 E2 cos θ2 (3.39)

Dividing eq. (3.38) by eq. (3.39) gives

tan θ1 tan θ2
 (3.40)
ε1 ε2

Since ε1  ε0 εr1 and ε2  ε0 εr 2 , eq. (3.40) may be written as

tan θ1 εr1
 (3.41)
tan θ2 εr 2

This is the law of refraction of the electric field at a boundary free of charge (since ρS  0 is
assumed at the interface). Thus, in general, an interface between two dielectrics produces
bending of the flux lines as a result of unequal polarization charges that accumulate on the sides
of the interface.

Conductor-Dielectric Boundary Conditions

The conductor is assumed to be perfect (i.e., σ   or ρc  0 ). To determine the boundary

conditions for a conductor-dielectric interface, the procedure used for dielectric-dielectric
interface is followed except that we incorporate the fact that E  0 inside the conductor. If
medium 2 in fig. 3.6 is a conductor then eqs. (3.33) and (3.36), and dropping the subscript 1,

Et  0 (3.42)

Dn  ρS (3.43)

Thus under static conditions, the following conclusions can be made about a perfect conductor:

1. No electric field may exist within a conductor; that is,

E  0, ρv  0

©j. k. makiche 83
 2. Since E   V  0 , there can be no potential difference between any two points in the
conductor; that is, a conductor is an equipotential body.
3. The electric field E can be external to the conductor and normal to its surface; that is

Dt  ε0 εr Et  0, Dn  ε0 εr En  ρS

Example 3.4: Two extensive homogeneous isotropic dielectrics meet on plane z = 0. For z  0 ,
εr1  4 and for z  0 , εr 2  4 . A uniform electric field E  5a x  2a y  3a z kV/m exists for z  0 .

Find

a) E2 for z  0 .

b) The angles E1 and E2 make with the interface

c) The energy densities in J/m3 in both dielectrics
d) The energy within a cube of side 2 m centered at (3, 4, -5)

Solution:

a) Since a z is the normal to the boundary plane, the normal components are

En1  E1  a n  E1  a z  (5a x  2a y  3a z )  a z  3

E1n  3a z

E 2 n  (E 2  a z )a z

Also

E  E n  Et

Hence,

Et1  E  E1  5a x  2a y  3a z  3a z  5a x  2a y

Using eq. (3.33)

©j. k. makiche 84
 Et 2  Et1  5a x  2a y

Dn 2  Dn1  Dn 2  εr 2 En 2  εr1En1

εr 1 4
E2 n  E1n  (3a z )  4a z
εr 2 3

E2  E2 n  E2t  5a x  2a y  4a z kV/m

b) Let α1 and α2 be the angles E1 and E2 make with the interface while θ1 and θ2 are the
angles they make with the normal to the interface; that is,

α1  90  θ1

α2  90  θ2

En1  3 and Et1  52  2 2  29

E11 29
tan θ1    1.795  θ1  60.9
En1 3

Hence,

α1  90  60.9  29.1

Similarly,

En 2  4 and Et 2  Et1  29

Et 2 29
tan θ2    1.346  θ2  53.4
En 2 4

Hence,

α2  36.6

©j. k. makiche 85
 tan θ1 εr1
Check: Confirm that  is satisfied.
tan θ2 εr 2

c) The energy densities are given by

1 1 10 9
wE1  ε1 E1   4   (25  4  9)  106  672 μJ/m 3
2

2 2 36 π

1 1 10 9
wE 2  ε2 E 2   3   (25  4  16)  106  597 μJ/m 3
2

2 2 36 π

d) At the center (3, 4, - 5) of the cube of side 2 m, z  5  0 ; that is, the cube is in region 2
with 2 < x < 4, 3 < y < 5, - 6 < z < - 4 . Hence

4 5 4
WE   wE 2 dv     wE 2 dxdydz  wE 2  (2)(2)(2)  597  8  4.776 mJ
x  2 y 3 z 6

Example 3.5: Region y  0 consists of a perfect conductor while region y  0 is a dielectric

medium ( εr 1  2) . If there is a surface charge of 2 nC/m2 on the conductor, determine E and D at

a) A(3, 2, 2)
b) B(4,1,5)

Solution:

a) Point A(3, 2, 2) is in the conductor since y  2  0 . Hence,

E0D

b) Point B(4,1,5) is in the dielectric medium since y  1  0 .

Dn  ρS  2 nC/m 2

Hence,

D  2a y nC/m 2

©j. k. makiche 86
 D 2a y  36 π
E  9
109  36 πa y V/m
ε0 εr 10  2

TUTORIAL PROBLEMS

1. A thin rod of cross section A extends along the x-axis from x = 0 to x = L. The polarization of
the rod is along its length and is given by Px = ax2 + b. Calculate ρ pv and ρ ps at each end.

Show explicitly that the total bound charge vanishes in this case.
2. A parallel-plate capacitor with plate separation of 2 mm has a 1-kV voltage applied to its
plates. If the space between its plates is filled with polystyrene ( εr  2.55 ), find E, P, and

ρ ps .

3. A dielectric sphere ( εr  5.7 ) of radius 10 cm has a point charge 2 pC placed at its center.
Calculate:
a) The surface density of polarization charge on the surface of the sphere
b) The force exerted by the charge on a 4 pC point charge placed on the sphere
4. Show that the force with which the plates of a parallel-plate capacitor attract each other is
given by

Q2
F
2ε S

where Q is the magnitude of charge on each plate, S is the area of the plate and ε is the
permittivity of the dielectric between the plates. Also determine the pressure on the surface
of the plate due to the field.

5. In a certain region, J  3r 2 cos θa r  r sin θaθ A/m find the current crossing the surface

defined by θ  30 , 0  φ  2 π , 0  r  2 m .

©j. k. makiche 87
6. The current density in a cylindrical conductor of radius a is J  10e  (1 ρ / a )a z A/m 2 . Find the
current through the cross section of the conductor.
100
7. If J  a ρ A/m 2 , find
ρ2
a) the rate of increase in the volume charge density,
b) the total current passing through a surface defined by   2, 0  z  1, 0    2 .
8. A sphere of radius a and dielectric constant εr has a uniform charge density of ρ0 .
a) At the center of the sphere, show that

ρ0 a
V (2εr  1)
6 ε0 ε r

b) Find the potential at the surface of the sphere.

9. A homogeneous dielectric ( εr  2.5 ) fills region 1 ( x  0 ) while region 2 ( x  0 ) is free
space.
a) If D1  12a x  10a y  4a z nC/m 2 find D 2 and θ2 .

b) If E2  12 V/m and θ2  60 , find E1 and θ1 .

[ (a) 12a x  4a y  1.6a z nC/m 2 , 19.75 , (b) 10.67 V/m, 77  ]

10. It is found that E  60a x  20a y  30a z mV/m at a particular point on the interface between

air and a conducting surface. Find D and ρS at that point.

[ 0.531a x  0.177a y  0.265a z pC/m 2 , 0.619, pC/m 2 ]

11. Two homogeneous dielectric regions 1 ( ρ  4 cm ) and 2 ( ρ  4 cm) have dielectric

constants 3.5 and 1.5, respectively. If D1  12a   6a  9a z nC/m2 , calculate:

a) E1 and D1

b) P2 and ρ pv 2 ,

c) the energy density for each region.

©j. k. makiche 88
12. Region 1 ( z  0 ) contains a dielectric for which εr  2.5 , while region 2 ( z  0 ) is

characterized by εr  4 . Let E  30a x  50a y  70a z V/m , find: (a) D2, (b) P2, (c) the

angle between E1 and the normal to the surface.

13. A silver-coated sphere of radius 5 cm carries a total charge of 12 nC uniformly distributed on
its surface in free space. Calculate
a) D on the surface of the sphere,

b) D external to the sphere, and

c) the total energy stored in the field.

©j. k. makiche 89
 Electrostatic Boundary Value Problems

The electric field E has been determined using Coulombs law, Gauss law or E  V when the
potential is known throughout the region of interest. In many practical situations, neither the ρv
nor V is known. We here consider practical electrostatic problems where only electrostatic
conditions (charge and potential) at some boundaries are known and it is desired to find E and V
throughout the region. Such problems are usually referred to as boundary value problems.

POISSON'S AND LAPLACE'S EQUATIONS

For a linear material medium, Gauss’ law may be stated as

  D  ε  E  ρv (3.1)

Also,

E  V (3.2)

Substituting eq. (3.2) in (3.1) gives

ε  E  ε  (V )  ρv

or

ρv
  V  
ε

Let V  V ( x, y, z )

V V V
V  ax  ay  az
x y z

  V    V    V  V V V
2 2 2
 V             2V
x  x  y  y  z  z  x
2
y 2
y 2

Hence,

©j. k. makiche 90
 ρv
 2V   (3.3)
ε

This is known as Poisson's equation. When ρv  0 , eq. (3.3) becomes

 2V  0 (3.4)

which is known as Laplace equation. The operator  2 is known as the Laplacian operator.
Laplace's equation in Cartesian, cylindrical, or spherical coordinates respectively is given by

 2V  2V  2V
  0 (3.5)
x 2 y 2 y 2

1   V  1  2V  2V
ρ   0 (3.6)
ρ ρ  ρ  ρ 2 φ2 z 2

1   2 V  1   V  1   2V 
r  2  sin   2 2  2 0 (3.7)
r r  r  r sin   
2
  r sin    

Poisson's equation in these coordinate systems may be obtained by simply replacing zero on the
right-hand side of eqs. (3.5), (3.6), and (3.7) with ρv / ε .

Laplace's equation is of primary importance in solving electrostatic problems involving a set of

conductors maintained at different potentials. Examples of such problems include capacitors and
vacuum tube diodes.

UNIQUENESS THEOREM

Since there are several methods (analytical, graphical, numerical, experimental, etc.) of solving a
given problem, it is important to know whether solving Laplace's equation in different ways
gives different solutions. Suppose a given problem has two solutions V1 and V2 both of which
satisfy the prescribed boundary conditions. Thus

 2V1  0 ,  2V2  0 (3.8)

©j. k. makiche 91
 V1  V2 at the boundary. (3.9)

Consider the difference,

Vd  V2  V1 (3.10)

which obeys

 2V2   2V1   2Vd  0 (3.11)

From divergence theorem,

   Adv   A  dS
v s
(3.12)

Let A  Vd Vd .

  A    Vd Vd  Vd ( 2Vd )  Vd Vd (3.13)

From eq. (3.11),  2Vd  0 . Hence, eq. 3.12 becomes

  A  Vd Vd (3.14)

Substituting eq. (3.14) in (3.12),

 V
v
d Vd dv   Vd Vd  dS
s
(3.15)

From eqs. (3.9) and (3.11), it is evident that the right side of eq. (3.15 vanishes. Hence,

 V dv  0
2
d (3.16)
v

Since the integrand Vd

2
is non-negative everywhere, eq. (3.16) can only be satisfied if  Vd is

identically zero. Hence,

©j. k. makiche 92
 Vd  V2  V1  a constant

To be consistent with eq. (3.9), Vd  0 or V2  V1 . Hence, V1 and V2 cannot be different solutions

of the same problem. This is the uniqueness theorem: If a solution to Laplace's equation can be
found that satisfies the boundary conditions, then the solution is unique.

A boundary-value problem is uniquely described by three things:

1. The appropriate differential equation (Laplace's or Poisson's equation in this case)

2. The solution region
3. The prescribed boundary conditions

A problem does not have a unique solution and cannot be solved completely if any of the three
items is missing. Poisson’s equation also satisfies uniqueness theorem.

The following general procedure may be taken in solving a given boundary-value problem
involving Poisson's or Laplace's equation:

1. Solve Laplace’s or Poisson’s equation for V by direct integration if V is a function of one

variable and by separation of variables if V is a function of more than one variable.
2. Apply the boundary conditions to determine the unique solution for V .
3. Determine E using E  V and D using D  εE .
4. If desired to determine the charge Q induced on a conductor, use

Q   ρS dS
S

where ρS  Dn and Dn is the component of D normal to the conductor surface. To

determine the capacitance C between the two conductors, use

Q
C
V

where V is the potential difference between the plates.

©j. k. makiche 93
Example 3.1: Two plates of a parallel plate capacitor are separated by a distance d as shown in
Fig. 3.1. The potential difference between the plates is V0 . Assuming negligible fringing effects at
the edges, determine

a) The potential at any point between the plates.

b) The surface charge densities at the plates.

d
V0 E
0

Fig. 3.1: A parallel plate capacitor

Solution:

a) Since ρv  0 between the plates, we use Laplace’s equation. Ignoring fringing effects is
tantamount to assuming that the field distribution between the plates is the same as
though the plates are infinitely large, so that there is no variation in the x and z
directions. Eq. (3.5) simplifies to

d 2V
0 (3.17)
dy 2

Integrating with respect to y gives

dV
 C1
dy

where C1 is a constant to be determined. Integrating again,

V  C1 y  C2 (3.18)

Two boundary conditions are required to determine C1 and C2 .

©j. k. makiche 94
 At y  0 , V  0 (3.19)

At y  d , V  V0 (3.20)

Substituting eq. (3.19) in (3.18),

0  C1 (0)  C2  C2  0

Hence,

V  C1 y (3.21)

Substituting eq. (3.20) in (3.21),

V0
V0  C1d  C1 
d

Thus,

V0
V y (3.22)
d

The potential increases linearly from y  0 to y  d .

b) To determine ρS , we must determine E .

dV V
E  V  a y  a y 0
dy d

εV0
D  ε E  a y
d

Dn  a n  D  ρS

At the lower plate,

©j. k. makiche 95
  εV  εV
an  a y , Dn  a y   a y 0    0
 d  d

εV0
 ρSl  
d

At the upper plate,

 εV  εV
a n  a y , Dn  a y   a y 0   0
 d  d

εV0
 ρSu  
d

Example 3.2: Semi-infinite conducting planes   0 and    / 6 are separated by an

infinitesimal insulating gap as in Fig. 3.2. If V (  0)  0 V and V (  0)  100 V , calculate V
and E in the region between the plates.

Solution:

V depends only on  . Laplace’s equation (3.6) becomes

1  2V
0
 2  2

Integrating twice gives

V  C1  C2

At   0 , V  0 V . Hence,

0  C1 (0)  C2  C2  0 .

At    / 6 , V  100 V . Thus

©j. k. makiche 96
 z

φ0

V0

Fig. 3.2: Semi-infinite conducting planes

π 600
100  C1  C1 
6 π

Hence,

600
V 


1 dV 600
E  V   a   a
 d 

Example 3.3: Two conducting cones ( θ  π /10 and θ  π / 6 ) of infinite extent are separated by
an infinitesimal gap at r  0 as shown in Fig. 3.3. If V (θ  π /10)  0 and V (θ  π / 6)  50 V
find V and E between the cones.

Solution:

The gap serves as an insulator between the two conducting cones. V depends only on θ .
Eq. (3.7) becomes

1   V 
 sin θ 0
r sin θ θ 
2
θ 

©j. k. makiche 97
 d  dV 
 sin θ 0
dθ  dθ 

Integrating once gives

dV
sin θ  C1
dθ

dV C
 1
dθ sin θ

V0

θ1
θ2

Fig. 3.3: Conducting cones

Integrating again yields

C1 dθ
V  d θ  C1 
sin θ 2sin(θ / 2) cos(θ / 2)
1/ 2sec 2 (θ / 2) d θ d [tan(θ / 2)]
 C1   C1   C1 ln tan(θ / 2)  C2
tan(θ / 2) tan(θ / 2)

V (θ  π /10)  0  0  C1 ln tan( π / 20)  C2

C2  C1 ln tan( π / 20)

©j. k. makiche 98
  tan(θ / 2) 
V  C1 ln  
 tan( π / 20) 

 tan( π /12) 
V (θ  π / 6)  50 V  50  C1 ln  
 tan( π / 20) 

50
C1 
 tan( π /12) 
ln  
 tan( π / 20) 

Hence,

 tan(θ / 2) 
50 ln 
 tan(θ / 2)   tan( π / 20)   tan(θ / 2) 
V  C1 ln     95.1ln  V
 tan( π / 20)   tan( π /12)   0.1584 
ln  
 tan( π / 20) 

and

1 dV 95.1
E  V   aθ   a θ V/m
r dθ r sin θ

PROBLEMS

1. Show that the following potentials satisfy Laplace's equation.

a) V  e 5 x cos13 y sinh12 z
1 dV 600
b) E  V   a   a
 d 
30 cos θ
c) V 
r2
2. A certain material occupies the space between two conducting slabs located at y  2 cm.

When heated, the material emits electrons such that ρv  50(1  y 2 ) μC/m 3 . If the slabs are

both held at 30 kV, find the potential distribution within the slabs. Take ε  3ε0 .

©j. k. makiche 99
3. Concentric cylinders ρ  2 cm and ρ  6 cm are maintained at V = 60 V and V  20 V ,
respectively. Calculate V, E, and D at ρ  4 cm.
4. Concentric shells r  20 cm and r  30 cm are held at V  0 V and V  50 V respectively.

If the space between them is filled with a dielectric material (ε  3ε0 , σ  1012 S/m) , find
(a) V, E, and D (b) the charge density at the shells (c) the leakage resistance.

©j. k. makiche 100

 The Steady Magnetic Field
A non-time varying magnetic field is called a magnetostatic field or a steady magnetic field. The
source of a magnetostatic field may be a permanent magnet, an electric field changing linearly
with time, or a direct current. The relationships presented here will mainly concern the magnetic
field produced by direct currents.

BIOT-SAVART LAW

Biot-Savart law is used to determine the magnetic field produced by a differential current
element in free space. This differential current element may be thought of a vanishingly small
section of a current-carrying filamentary conductor as shown in Fig. 4.1. A filamentary
conductor is the limiting case of a cylindrical conductor of circular cross section as the radius
approaches zero.

dL 1
a R12
R 12

I1
I 1 dL 1  a R12
dH 2 
4 πR122

Fig. 4.1: Magnetic field intensity dH 2 produced by a differential current element I1dL1 . The

direction of dH 2 is into the page

Biot-Savart law states that the magnetic field intensity dH at a point P in space due to a
differential current element IdL located a distance R from point P is given by

IdL  a R IdL  R
dH   (4.1)
4π R2 4π R3

where a R  R / R is a unit vector from the differential current element to point P. The unit of the
magnetic field intensity H is Amperes per metre (A/m). If the current element is located at point
1 and the point P at which the field is to be determined is denoted point 2, then

©j. k. makiche 101

 I1dL1  a R12
dH 2  (4.2)
4 π R122

It is impossible to experimentally check the validity of Biot-Savart law in the form given in eq.
(4.1) since the differential current element cannot be isolated. The continuity equation is

ρv
J  
v

Our attention is focused on direct current, so the charge density is not a function of time. Thus,

J  0

Applying divergence theorem,

 J  dS  0
S

The total current crossing any closed surface is zero. This condition can only be satisfied if the
current flows around a closed path. The integral form of Biot-Savart law, which can be verified
experimentally, becomes

IdL  a R
H   (4.3)
4π R2

Biot-Savart law may be expressed in terms of distributed sources such as the current density J
and the surface current density K . Surface current flows in a sheet of vanishingly small
thickness and therefore the current density J ( A/m 2 ) is infinite. Surface current density K is
measured in Amperes per metre width ( A/m ). The differential current element IdL may be
expressed in terms of K or J as

Id L  K dS  J dv

Alternate forms of Biot-Savart law therefore becomes,

K  a R dS
H (4.4)
S
4π R2

J  a R dv
H (4.5)
v
4π R2

©j. k. makiche 102

Example 4-1: Consider an infinitely long straight filament carrying a uniform current I and
lying along the z axis as shown in Fig. 4.2. Determine H at any point P off the axis using Biot-
Savart law.
z

dL
aR

za z

R
ρa ρ y

Fig. 4-2: An infinitely long straight filament carrying current I

Solution:

The field possesses cylindrical symmetry. Also, no variation with z or  can exist. We therefore
choose point P at which the field is to be determined to be in the z  0 plane. The field point r is
r  ρa ρ . The source point r  is given by r   z a z . Hence,

R  r  r  ρa ρ  z a z

ρa ρ  za z
a R12 
ρ 2  z 2

dL  dz a z

I1dL1  a R12 Idz a z  ρa ρ  z a z

dH 2  
4π R 2
12 4 π ( ρ 2  z 2 )3/ 2

Idz a z  (  a   z a z ) Ia
 
dz 
H2     (

4 (  2  z 2 )3/ 2 4 
2
 z 2 )3/ 2

The unit vector a is removed from under the integral sign since it changes with  co-ordinate

but not with ρ or z yet the integration is with respect to z only.

©j. k. makiche 103

 
Ia z I
H2   a
4  2  2  z 2 2


Thus, the magnitude of the field is not a function of  or z and it varies inversely with the
distance from the filament. The magnetic field intensity vector is circumferential.

In general, the magnetic field H at a point P off the axis of an infinite uniform line current I is
given by

I
H a (4.6)
2

For a finite length current element shown in Fig. 4-3, it can be shown that H at point 2 is given
in terms of the angles α1 and α2 as

I
H (sin  2  sin 1 )a (4.7)
4

y
α2 α
1

Fig. 4-3: Finite current filament

AMPERE’S CIRCUITAL LAW

Ampere’s law states that the line integral of H about any closed path is exactly equal to the direct
current enclosed by the path.

 H  dL  I (4.8)

©j. k. makiche 104

Just like Gauss’ law was used to solve electrostatic problems with a high degree of symmetry in
charge distribution, Ampere’s circuital law is used to solve magnetic field problems wherever a
high degree of symmetry in current distribution is present.

Example 4-2: Determine H due to the long straight wire carrying a current I shown in Fig. 4.2
using Ampere’s law.

Solution:

We choose to integrate H along a circle of radius ρ . Ampere’s circuital law becomes

2 2

 H  dL   H
0
  d  H   d  2 H   I
0
2 2

 H  dL   H
0
  d  H   d  2 H   I
0

I
H 
2

or

I
H a
2

Ampere’s law in Point Form

By definition,

I   J  dS
S

So that Ampere’s law becomes

 H  dL   J  dS
S
(4.9)

Applying Stokes theorem in eq. (4.8),

 H  dL     H  dS   J  dS
S S

So that,

©j. k. makiche 105

 H  J (4.10)

Eq. (4.10) is the differential or point form of Ampere’s circuital law.

MAGNETIC FLUX AND MAGNETIC FLUX DENSITY

In free space, we define the magnetic flux density B as

B  μ0 H (free space only) (4.11)

where B is measured in Webers per square metre (Wb/m2) or Tesla (T).

μ0  4 π 107 H/m

is the permeability of free space.

The magnetic flux Φ is defined as the flux passing through any designated area and is given in
terms of B as

Φ   B  dS Wb. (4.12)
S

Biot-Savart law may be written in terms of J and B as

μ0 J  a R dv
B
4π  R2

Taking the divergence of both sides,

μ0 J  a R dv
B  
4π v

R2

Using the product rule,

  ( A  B)  B  (  A)  A  (  B)

J  a R dv a R  a 
  2  (  J )  J     R2  (4.13)
 R 
2
R R

The first term on the right hand side of eq. (4.13) is zero since   J  0 . This is because 
operates on field points at ( x, y, z ) but is a function of co-ordinates ( x, y, z) . The second term is

also zero since from vector algebra,   R na R  0 so that

©j. k. makiche 106

 aR
 0.
R2

Hence,

 B  0 (4.14)

If we apply divergence theorem in eq. (4.14),

 B.dS  0
S
(4.15)

Eqs. (4.14) and (4.15) are the equivalents of Gauss’ law in point and integral forms respectively
for magnetic fields and they indicate that magnetic flux lines are closed and do not terminate on
a ‘magnetic charge’. This tells us that there are no magnetic ‘monopoles’ or ‘magnetic charges’
in nature.

Eq. (4.14) is the last of the four fundamental equations, known as Maxwell’s equations, as they
apply to non-time varying fields. The complete list is as follows;

  D  ρv
E  0
(4.16)
H  J
B  0

The corresponding set of four integral equations that apply to non-time varying fields are

 D  dS  Q   ρ dv
S v
v

 E  dL  0
(4.17)
 H  dL  I   J  dS
S

 B.dS  0
S

THE SCALAR AND VECTOR MAGNETIC POTENTIALS

The electric scalar potential V was defined in terms of E as

E   V

In a similar manner, the scalar magnetic potential Vm may defined such that

©j. k. makiche 107

 H  Vm

But

  H  J    (Vm )

The curl of any gradient is identically zero. Hence, if H is to be defined as a gradient of a scalar
potential, the current density must be zero throughout the region in which the scalar magnetic
potential is to be defined. Hence,

H  Vm (J  0) (4.18)

The scalar magnetic potential also satisfies Laplace’s equation. In free space,

  B  μ0  H  μ0  (Vm )

or

 2Vm  0 (J  0) (4.19)

The electric potential V is a single valued function of position. Once a zero reference is assigned,
there is only one value of V associated with each point in space. Such is not the case with Vm .
This is because unlike the electrostatic field, the steady magnetic field is not a conservative field.

 H  dL  I
even if J  0 along the path of integration. Hence,

b
Vm ,ba    H  dL (Specified path) (4.20)
a

To define the magnetic vector potential, we note that

 B  0

Any vector whose divergence is zero must be a curl of another vector. We let

B  A (4.22)

A is the magnetic vector potential and it is defined in region where J is zero or not zero. The H
is

©j. k. makiche 108

 1
H  A
μ0

and

1
H  J   A
μ0

The curl of the cur of the vector field A is given by the vector identity

    A   (  A )   2 A

so that

(  A)   2 A  μ0 J (4.23)

We let   A  0 . This choice of the divergence of A is not arbitrary and is known as Coulomb’s
gauge. We obtain

 2 A   μ0 J (4.24)

 2 A is the Laplacian of a vector defined in Cartesian co-ordinates as

 2 A   2 Ax a x   2 Ay a y   2 Az a z

Eq. (4.24) may be solved to determine A . For a uniform current with current density J it can be
shown that the solution is given by

μ0 Jdv
A (4.25)
v
4π R

For a surface current with density K , we have

μ0 KdS
A (4.27)
v
4π R

and for filamentary currents

μ0 IdL
A   (4.28)
4π R

©j. k. makiche 109

Eqs. (4.25), (4.26) and (4.27) express the magnetic vector potential as integration over all of its
sources. The zero reference for A is infinity.

Problems

1. A filamentary conductor is formed in formed into an equilateral triangle with sides of

length  carrying a current I . Find the magnetic current density at the centre of the
triangle.
2. A filamentary conductor is formed in formed into a square with sides of length  carrying
a current I . Find the magnetic current density at the centre of the triangle.
3. An infinite plane lying on the z  0 plane carries a uniform current with surface current
density K  K y a y . Determine H at a point P located in the z  0 region.

4. Consider an infinitely long coaxial cable whose cross-sectional area is shown in Fig. 4-4.
The inner conductor with radius a carries a uniform current with density J  K z a z . The

outer conductor has an inner radius b , outer radius c and carries a current with density
J   K z a z . Determine H in everywhere using Ampere’s law.

Fig. 4-4: Cross-sectional view of coaxial cable.

5. Given H  (3r 2 / sin  )a  54r cos  a A/m in free space find the total current in the a θ

direction through the conical surface θ  20 , 0    2 , 0  r  5 by whatever side of

Stokes theorem you like best. Verify your result using the other side of Stoke’s theorem.
6. A long straight non-magnetic conductor of 0.2 mm radius carries a uniformly distributed
current of 2 A dc.
a) Find J within the conductor.
b) Use Ampere’s circuital law to find H and B within the conductor.
c) Find H and B outside the conductor.
d) Show   H  J outside the conductor.

©j. k. makiche 110

 7. Let A  (3 y  z )a x  2 xza y Wb/m in a certain region of free space.

a) Show that   A  0
b) Find A , B , H and J at P(2, 1,3) .

©j. k. makiche 111

 Magnetic Forces, Materials and Inductance
FORCE ON A MOVING CHARGE

In an electric field, the force FE on a charged particle is

FE  QE (5.1)

The force F is in the same direction as E and is proportional to both Q and E . A

charged particle in motion in a magnetic field of flux density B is found experimentally
to experience a force FM whose magnitude is proportional to the product of the
magnitudes of the charge Q , its velocity v ,the flux density B and to the sine of the
angle between v and B . Its direction is perpendicular to both v and B and is given by
the unit vector in the direction v  B . Hence,

FM  Qv  B (5.2)

Since FM is always perpendicular to v , the acceleration vector is always normal to the

velocity vector and therefore the kinetic energy of the particle remains unchanged.
Hence, a steady magnetic field is incapable of transferring energy to a moving charge.
The electric field, on the other hand exerts a force on the particle that is independent of
the direction in which the particle is progressing.

The force on a moving particle arising from combined electric and magnetic field is
obtained by superposition as

F  FE  FM  Q(E  v  B) (5.3)

Eq. (5.3) is known as the Lorentz force equation.

Example 5.1: A point charge Q  40nC has a velocity 3  10 6 m/s in the direction
a v  0.2a x  0.75a y  0.3a z . Calculate

i) the magnitude of the force on the charge due to the field B  3a x  4a y  6a z

mT.

©j. k. makiche 112

 ii) the magnitude of the force on the charge due to the field E  2a x  4a y  5a z

kV/m

iii) the Lorentz force due to the two fields in (i) and (ii) above

Solution:

i) FM  Qv  B

 
 40  10 -9 [3a x  4a y  6a z ]  10 3  [3  10 6 (0.2a x  0.75a y  0.3a z )]


 40  10 -9  9.9  10 3 a x  0.9  10 3 a y  9.15  10 3 a z 
 0.396 a x  0.036 a y  0.366 a z mN

FM  0.3962  0.0362  0.3662  0.54mN

ii) FE  QE  40  109  (3a x  4a y  6a z )  103  0.12a x  0.16a y  0.24a z mN

FE  0.12 2  0.16 2  0.24 2  0.312mN

iii) F  FM  FE  ( 0.396a x  0.036a y  0.366a z )  (0.12a x  0.16a y  0.24a z )

 0.276 a x  0.124 a y  0.126 a z

F  0.2762  0.1242  0.1262  0.33mN

FORCE ON A DIFFERENTIAL CURRENT ELEMENT

When a charge particle is in motion in a steady magnetic field, the differential force
exerted on a differential element of charge may be written as

dF  dQv  B (5.4)

A differential element of charge physically consists of numerous small discrete charges

occupying a small volume such that the average separation between the charges is
insignificant. Hence, differential force in eq. (5.4) is the sum of the forces on the
individual charges.

©j. k. makiche 113

If the charges are electrons in motion in a conductor, the magnetic field exerts forces on
the electrons causing them to shift position slightly resulting in a small displacement
between the positive and negative charges. This displacement is however resisted by
Coulomb forces between the electrons and positive ions and the force is therefore
transferred to the crystalline lattice or to the conductor itself. Since the Coulomb forces
are much greater than the magnetic forces in good conductors, the actual displacement
of the electrons is almost immeasurable. The charge separation that does result is
however disclosed by the presence of a slight potential difference across the conductor
sample in a direction perpendicular to both the magnetic field and the velocity of the
charges. The voltage is known as Hall voltage and the effect is called Hall effect.

The conduction current density is defined as

J  ρv v

The differential element of charge in eq. (5.4) may be written as

dQ  ρv dv

Thus,

dF  ρ vdv v  B

or

dF  J  Bdv (5.5)

Since

Jdv  KdS  IdL

Eq. (5.5) may be written in terms of surface current as

dF  K  BdS (5.6)

or in terms of a differential current filament as

dF  IdL  B (5.7)

The total force F is determined by integrating eqs. (5.5), (5.6) and (5.7) to obtain

©j. k. makiche 114

 F   J  Bdv (5.8)
v

F   K  BdS (5.9)
S

F   IdL  B   I  B  dL (5.10)

Applying eq. (10) to a straight conductor in a uniform magnetic field,

F  IL  B (5.11)

Example 5.2: Consider a square loop of wire in the z  0 plane carrying 2 mA in the
field of an infinite filament on the y axis as shown in Fig. 5-1. Determine the total force
on the loop.

Solution:

The field produced in the plane of the loop by the straight filament is

I 15
H az  a z A/m
2 πx 2 πx

3  106
B  μ0 H  4 π  107 H  az T
x
z

15 A
y

(1,0,0) (1,2,0)

(3,0,0) 2 mA
x

Fig. 5-1: A square loop of wire in the xy plane carrying 2 mA subjected to a non-uniform
magnetic field.

©j. k. makiche 115

 3 a 2
a
F   I  B  dL  2  103  3  106   z  dxa x   z  dya y 
 x 1 x y 0
3
1
az
0
a 
x 3 x  dx a x  y2 3z  dya y 

9
 3 1
2
1 0 
 6  10  ln x 1 a y  y ( a x )  ln x 3 a y  y 2 (a x )
 3 0 
 2 1 
 6  109  ln 3a y  a x  ln a y  2a x   8a x nN
 3 3 

FORCE AND TORQUE ON A CLOSED CIRCUIT

The force on a filamentary closed circuit is given in eq. (5.10) as

F   I  B  dL

Assuming a uniform magnetic flux density, B may be removed from the integral:

F   IB   dL

But  dL  0 . Hence, the force on a closed filamentary circuit in a uniform magnetic

field is zero. If the field is not uniform, the total force need not be zero. This result for
also applies to distributed currents as well.

T T

O O

R F R2 R1 F1

P P1
R 21

F2  F1

©j. k. makiche 116

 Fig. 5-2: (a) Lever arm R extending from the origin O to point P where the force F is
applied. (b) If F2  F1 , then T  R 21  F1 is independent of the origin of R 1 and R 2 .

The torque T (or moment of a force) about a point O with a rigid lever arm R extending
from O to P as shown in Fig. 5-2 (a) is a cross product defined as

T  R F (5.11)

Suppose two forces F1 at P1 and F2 at P2 having lever arms R 1 and R 2 extending from a
common origin O as shown in Fig. 5-2 (b) are applied to a object of fixed shape and that
the object does not undergo any translation. The torque about the origin is

T  R 1  F1  R 2  F2

Since F2  F1 or F1  F2  0 , we have

T  (R 1  R 2 )  F1  R 21  F1

where R 21  R 1  R 2 joins the points of application of F2 to that of F1 and is independent

of the choice of the origin of R 1 and R 2 . Hence, T is independent of the choice of the
origin provided that the total force is zero.

Consider the torque on a differential current loop in a magnetic field B . The loop lies in
the xy plane with its sides parallel to the x and y axes as shown in Fig. 5.3. Let the value
of the magnetic field at the centre of the loop be B 0 . Since the loop is of differential size,

the value of B at all points may be taken as B 0 . The total force in the loop is therefore
zero and we are free to choose the origin for the torque at the centre of the loop.

©j. k. makiche 117

 y

dx
B
I 3

dy
O x
4 2

R
1

Fig. 5-3: A differential current loop in a magnetic field B.

The vector force on side 1 is

dF1  Idxa x  B0
 Idxa x  ( B0 x a x  B0 y a y  B0 z a z )
 Idx( B0 y a z  B0 z a y )

For this side of the loop, the lever

1
R 1   dya y
2

and the contribution to the torque is

1 1
dT1  R 1  dF   dya y  Idx ( B0 y a z  B0 z a y )   dydxIB0 y a x
2 2

The torque contribution on side 3 is found to be the same

1 1
dT3  R 3  dF3  dya y   Idx ( B0 y a z  B0 z a y )   dydxIB0 y a x
2 2

and

dT1  dT3  dydxIB0 y a x

Evaluating the torque on sides 2 and 4, we obtain

©j. k. makiche 118

 dT2  dT4  dydxIB0 x a y

The total torque is

dT  dT1  dT2  dT3  dT4  Idydx ( B0 x a y  B0 y a x )

The quantity in parentheses may be written as a cross product so that

dT  Idydx (a z  B0 )

or

dT  IdS  B (5.12)

where dS  dydxa z is the vector area of the differential current loop and the subscript

on B 0 has been dropped.

The product of the loop current and the vector area of the loop is defined as the
differential magnetic dipole moment dm with units A  m 2 . Thus,

dm  IdS (5.13)

and

dT  dm  B (5.14)

In general, the torque in a planar loop of any size or shape in a uniform magnetic field is
given by

T  IS  B  m  B (5.15)

The torque on the current loop always tends to turn the loop so as to align the magnetic
field produced by the loop with the applied magnetic field that is causing the torque.

NATURE OF MAGNETIC MATERIALS

To describe the nature of magnetic materials, we use a simple atomic model of an atom
that assumes a central positive nucleus surrounded by electrons in various orbits. An
electron in an orbit is analogous to a small current loop (in which the current is directed
opposite to the direction of electron travel) and as such experiences a torque in an
external magnetic field, the torque tending to align the magnetic field produced by the
orbiting electron with the external magnetic field. If there were no of other magnetic

©j. k. makiche 119

moments to consider, the orbiting electrons in the material would shift in such a way as
to add their magnetic fields to the applied field and the resultant magnetic field at any
point in the material would be greater than the external field.

A second moment is due to electron spin. It can be shown using relativistic quantum
theory that an electron may have a spin magnetic moment of about 9  1024 A/m 2 . The
plus and minus signs indicate alignment aiding or opposing the external field.

A third contribution to the moment of an atom is due to nuclear spin. Although this
factor provides negligible effect on the overall magnetic properties of materials, it is the
basis of magnetic resonance imaging (MRI) procedure provided in large hospitals.

Each atom has many different components of these moments and their combination
determines the magnetic characteristics of the material and provides its general
magnetic classification. Magnetic materials may be classified as follows:

1. Diamagnetic: The magnetic moment due to electron travel in its orbit and that
due to electron spin combine to produce a net field of zero. In these materials, the
permanent magnetic moment m 0 is zero. Consequently, no realignment of the
dipole field would occur under the application of an external field and the
internal magnetic field is the same as the applied field.
2. Paramagnetic: In these materials, the electron spin and orbital motion
moments do not quite cancel. Each atom has a small magnetic moment but the
random orientation of these moments produces an average magnetic moment of
zero. Hence, the material exhibits no magnetic properties when no external field
is applied. When an external field is applied, there is an increase in B in the
material. Examples of such materials include potassium, oxygen, tungsten and
some rare earth elements.
3. Ferromagnetic. Each atom has a relatively large dipole moment caused mainly
by uncompensated electron spin moments. Interatomic forces causes these
moments to line up in parallel fashion over regions containing large number of
atoms. These regions are called domains and may have a variety of shapes and
sizes.

©j. k. makiche 120

 4. Antiferromagnetic. In these materials, the forces between adjacent atoms
cause the atomic moments to line up in antiparallel fashion. The net magnetic
moment is zero hence these materials are affected only slightly by the application
of an external field. Example includes manganese dioxide, nickel oxide and iron
sulphide.
5. Ferrimagnetic substances show an antiparallel alignment of adjacent atomic
moments but the moments are not equal. A large response to an external field
occurs although not as large as ferromagnetic materials. Examples include iron
oxide magnetite, nickel-zinc ferrite etc.
6. Superparamagnetic. These materials are composed of an assembly of
ferromagnetic particles in a nonferromagnetic matrix. Although domains exist
within the individual particles, the domain walls cannot penetrate the intervening
matrix material to the adjacent particle. Example includes magnetic tape used in
audio tape and video tape recorders.

MAGNETIZATION AND PERMEABILITY

To quantitatively describe magnetic materials, we consider the current due to movement

of bound charges (orbital electrons, electron spin and nuclear spin). The current
produced by these bound charges is called bound current or Amperian current.

We define the magnetic dipole moment m as the dipole moment established when a
bound current I b circulates about a path enclosing a differential area dS . Thus

m  I b dS (5.16)

If there are n magnetic dipoles per unit volume and considering a volume v , the total
magnetic dipole moment is obtained by the vector sum
nv
m total   m i (5.17)
i 1

We define magnetization M as the magnetic dipole moment per unit volume,

1 nv
M  lim
v 0 v

i 1
mi (5.18)

©j. k. makiche 121

The units of M must be the same as for H . A medium for which M is not zero
everywhere is said to be magnetized. For a differential volume dv , the magnetic
moment is dm  M dv . It can be shown that the magnetic vector potential due to dm is

μ0 M  a R μ MR
dA  2
dv  0 dv
4 πR 4 πR 3

But

R 1
3
   
R R

Hence,

μ0 1
A
4π  M     dv
R
(5.19)

But

1 1 M
M         M     (5.20)
R R R

Substituting eq. (5.20) in (5.19),

μ0   M μ M
A 
4 π v R
dv  0    dv
4 π v R

Applying the identity

   Fdv   F  dS
v S

to the second integral, we obtain

μ0   M μ M
A 
4 π v R
dv  0 
4π S R
 dS

μ0   M μ M  an
 
4 π v R
dv  0 
4π S R
dS  (5.21)

μ0 Jb μ K
 
4 π v R
dv  0  b dS 
4π S R

where (upon dropping the primes)

©j. k. makiche 122

 Jb    M (5.22)

and

K b  M  an (5.23)

where Jb is the bound volume current density and K b is the bound surface current
density. In free space, M  0 and we have

B
J  H   
 μ0 

where J is the free current volume density. In a material where M  0 ,

B
     J  J b  Jt    H    M    (H  M )
 μ0 

Therefore

B  μ0 (H  M) (5.24)

Eq. (5.24) applies to all types of magnetic materials whether linear or non-linear. For
linear materials, M is linearly related to H so that

M  χM M

where χ M is a dimensionless quantity known as magnetic susceptibility. Eq. (5.24)

becomes

B  μ0 (H  χ M H)  μ0 (1  χ M )H  μH (5.25)

or

B  μ0 μr H (5.26)

where

μr  μ0 (1  χ M ) (5.27)

©j. k. makiche 123

is the relative permeability of the material (dimensionless) and μ  μ0 μr is the
permeability of the material in henrys per metre (H/m). Eq. (5.26) only applies to linear
isotropic materials.

MAGNETIC FIELD BOUNDARY CONDITIONS

Fig. 5-4 shows a boundary between two isotropic homogeneous linear materials with
permeabilities μ1 and μ2 . The boundary condition for the normal components is
determined by allowing the surface to cut a small cylindrical Gaussian surface and
applying Gauss’ law for magnetic fields,

 B  dS  0
S

We find that

BN 1 S  BN 2 S  0

0r

BN 1  BN 2 (5.28)

BN 1
μ1
S
Ht 1
a w b h

d c
w
a N 12
Ht 2

μ2
BN 2

Fig. 5-4: Boundary between two magnetic materials

Thus

μ1
HN2  HN1 (5.30)
μ2

The normal component of B is continuous but the normal component of H is

discontinuous by the ratio μ1 / μ2 .

©j. k. makiche 124

To determine the boundary conditions for the tangential components, we apply
Ampere’s circuital law

 H  dL  I
about a small closed path a-b-c-d-a in a plane normal to the boundary surface. Taking a
clockwise trip around the path,

H t 1 L  H t 2 L  K L

where we have assumed that the boundary may carry a surface current K whose
component normal to the plane of the closed path is K . Thus,

Ht 1  Ht 2  K (5.31)

Eq. (5.31) may be written using cross product to identify the tangential components;

(H1  H2 )  a N 12  K

where a N 12 is a unit vector directed from region 1 to region 2. An equivalent formulation

in terms of vector tangential components is obtained as

Ht 1  Ht 2  a N 12  K

For tangential components of B , we have

Bt 1 Bt 2
 K (5.32)
μ1 μ2

Example 5.4: Assume that μ  μ1  4 μH/m in region 1 where z  0 , while

μ2  7 μH/m in region 2 wherever z  0 . Moreover, let K  80a x A/m on the surface

z  0 . Given that a field B1  2a x  3a y  a z mT in region 1, determine B 2 .

Solution:

B N 1  (B1  a N 12 )a N 12  [(2a x  3a y  a z )  ( a z )]( a z )  a z mT

Thus

B N 2  B N 1  a z mT

©j. k. makiche 125

 Bt 1  B1  B N 1  2a x  3a y mT

3
Bt 1 (2a x  3a y )  10
Ht 1    500a x  750a y A/m
μ1 4  106

Ht 2  Ht 1  a N 12  K  500a x  750a y  ( a z )  80a x

 500a x  750a y  80a y  500a x  670a y A/m

Bt 2  μ2 Ht 2  7  106 (500a x  670a y )  3.5a x  4.69a y mT

B 2  Bt 2  B N 2  3.5a x  4.69a y  a z mT

THE MAGNETIC CIRCUIT

In this section, we identify the field equations upon which resistive circuit analysis is
based and point out or derive the analogous equations for the magnetic circuit.

The electrostatic potential is related to the electric field intensity as

E  V (5.33a)

The analogous relation to magnetic field intensity is

H  Vm (5.33b)

The electric potential difference between points A and B may be written as

B
V AB   E  dL (5.34a)
A

The corresponding relation between the mmf and the magnetic field intensity
B
VmAB   H  dL (5.34b)
A

Ohm’s law for the electric circuit in point form is given by

J  σE (5.35a)

The magnetic flux density is the analog of the current density,

©j. k. makiche 126

 B  μH (5.36a)

To find the total current, we integrate:

I   J  dS (5.37a)
S

The corresponding operation to determine the total magnetic flux flowing through the
cross section of a magnetic circuit is

Φ   B  dS (5.38a)
S

Resistance is defined as the ratio of potential difference to current or

V  IR (5.39a)

We define reluctance as the ratio of magnetomotive force to the total flux; thus

Vm  ΦR (5.39b)

where reluctance is measured in ampere-turns per Weber (A  t/Wb) . For linear,

isotropic and homogeneous material of conductivity σ , a uniform cross section of area
S and length d , the resistance is

d
R (5.40a)
σS

Similarly for linear, isotropic and homogeneous magnetic material of uniform cross
section of area S and length d , the total reluctance is

d
R (5.41a)
μS

The closed line integral of E is zero i.e.

 E  dL  0
This is Kirchhoff’s voltage law. The corresponding expression for magnetic phenomena
is

 H  dL  I total

©j. k. makiche 127

If the current I is flowing through an N turn coil, we have

 H  dL  NI (5.42)

In an electric circuit, the voltage source is part of the closed path. In a magnetic circuit,
the current carrying coil will surround or link the magnetic circuit.

POTENTIAL ENERGY AND FORCES ON MAGNETIC MATERIALS

The general expression for energy stored in an electrostatic field is

1
2 v
WE  D  Edv

where a linear relationship between D and E is assumed. The total energy stored in a
steady magnetic field in which B is linearly related to H is
2
1 1 1 B
WH 
2v B  Hdv   μH 2dv  
2v 2v μ
dv (5.43)

PROBLEMS

1. A rectangular loop of wire in free space joins points A(1,0,1) to B(3,0,1) to

C (3,0,4) to D(1,0,4) to A . The wire carries a current of 6 mA flowing in the a z
direction. From B to C . A filamentary current of 15 A flows along the entire z
axis in the a z direction. Find the total force on the loop.
2. The magnetic flux density in a region of free space is given as
B  3 xa x  5 ya y  2 za z T . Find the total force on a rectangular loop lying in the

plane z  0 and bounded by x  1 , x  3 , y  2 and y  5 axes given that a 30 A

current flows around the loop in the anticlockwise direction. All dimensions in
cm.

©j. k. makiche 128