Electromagnetic Theory 2018-2019 Prof. dr.

Ali Hadi Hassan Al-Batat 1-15

Chapter Six
Electrostatic Energy
In general, the energy of a system of charges, just like that of any other mechanical
system, may be divided into its potential and kinetic contributions. Under static
conditions, however, the entire energy of the charge system exists as potential
energy, and we are particularly concerned with that potential energy which arises
from electrical interaction of the charges, the so-called electrostatic energy.

In chapter four, it was shown that the electrostatic energy of a point charge is closely
related to the electrostatic potential U at the position of the point charge. In fact, if q
is the magnitude of a particular point charge, then the work done on the charge in
moving it from position A to position B is
𝐵 𝐵 𝐵
𝑊𝑜𝑟𝑘 = ∫𝐴 𝐹⃗𝒎 . 𝑑𝑙⃗ = − 𝑞 ∫𝐴 𝐸⃗⃗ . 𝑑𝑙⃗ = 𝑞 ∫𝐴 𝒈𝒓𝒂𝒅𝑈. 𝑑𝑙⃗ = 𝑞(𝑈𝐵 − 𝑈𝐴 ) (1)

Here, the mathematical force, 𝐹⃗𝒎 has been chosen so as to balance exactly the
electric force 𝑞𝐸⃗⃗ at each point along the path. Under these conditions the charged
particle does not accelerate, and equation (1) represents the change in electrostatic
energy of the charge over the path interval 𝐴 → 𝐵.
In fact, the e1ectrostatic energy of an arbitrary charge distribution may be calculated
as the work required to assemble this distribution of charge without imparting to it
other forms of energy.
1. Potential Energy of a Group of Point Charges
By the electrostatic energy of a group of m point charges, we mean the potential
energy of the system relative to the state in which all point charges are infinitely
separated from one another. This energy can be obtained by calculating the work
done to assemble the charges. The first charge 𝑞1 may be placed in position without
the expenditure of energy;

To place the second charge, 𝑞2 , requires,

𝑞1 𝑞2
∆𝑊2 = (2)
4𝜋𝜀0 𝑟12

Where 𝑟12 = ⃒𝑟⃗1 − 𝑟⃗2 ⃒

For the third charge, 𝑞3

1
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

𝑞1 𝑞2
∆𝑊3 = 𝑞3 [ + ] (3)
4𝜋𝜀0 𝑟13 4𝜋𝜀0 𝑟23

The work required to bring in the fourth charge, fifth charge, etc, written in similar
fashion.
It will be convenient to symmetrize each of these terms by repeating each entry with
indices reversed and having the result. In other words, we rewrite equation (2) as

1 𝑞1 𝑞2 +𝑞2 𝑞1
∆𝑊2 = [ ] (4)
8𝜋𝜀0 𝑟12

The succeeding terms are similarly rewritten

1 𝑞1 𝑞3 +𝑞3 𝑞1 𝑞2 𝑞3 +𝑞3 𝑞2
∆𝑊3 = [ + ] (5)
8𝜋𝜀0 𝑟13 𝑟23

1 𝑞1𝑞4 +𝑞4 𝑞1 𝑞2 𝑞4 +𝑞4 𝑞2 𝑞3 𝑞4 +𝑞4 𝑞3

∆𝑊4 = [ + + ] (6)
8𝜋𝜀0 𝑟14 𝑟24 𝑟34

Finally, to put the nth charge into place,

1 𝑞1 𝑞𝑛 +𝑞𝑛 𝑞1 𝑞2 𝑞𝑛 +𝑞𝑛 𝑞2 𝑞3 𝑞𝑛 +𝑞𝑛 𝑞3 𝑞𝑛−1 𝑞𝑛 +𝑞𝑛 𝑞𝑛−1

∆𝑊𝑛 = [ + + + ⋯…….+ ] (7)
8𝜋𝜀0 𝑟1𝑛 𝑟2𝑛 𝑟3𝑛 𝑟𝑛−1,𝑛

We add the terms to obtain the work to assemble all the charges,

𝑊 = ∆𝑊2 + ∆𝑊3 + ∆𝑊4 + ⋯ … + ∆𝑊𝑛 (8)

1 𝑞 𝑞 𝑞 𝑞 𝑞 𝑞 𝑞
=
8𝜋𝜀0
[𝑞1 (𝑟 2 + 𝑟 3 + 𝑟 4 + ⋯ … … + 𝑟 𝑛 ) + 𝑞2 (𝑟 1 + 𝑟 3 + 𝑟 4 + ⋯ … +
12 13 14 1,𝑛 21 23 24
𝑞𝑛 𝑞1 𝑞2 𝑞4 𝑞𝑛 𝑞1 𝑞2 𝑞4
𝑟2,𝑛
) +𝑞3 (𝑟 + 𝑟 + 𝑟 + ⋯ … … + 𝑟 ) + ⋯ +𝑞𝑛−1 (𝑟 +
𝑟𝑛−1,2
+
𝑟𝑛−1,4
+
31 32 34 3,𝑛 𝑛−1,1
𝑞𝑛
⋯……+ )] (9)
𝑟𝑛−1,𝑛

1 𝑛 𝑞𝑘 1 𝑛 𝑞𝑗 𝑞 𝑘
= ∑𝑛−1
𝑗=1 𝑞𝑗 ∑𝑘=1 = ∑𝑛𝑗=1 ∑𝑘=1 (10)
8𝜋𝜀0 𝑟𝑗𝑘 8𝜋𝜀0 𝑟𝑗𝑘
𝑘≠𝑗 𝑘≠𝑗

2
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

where in the last step we have increased the upper limit of the 𝑗 summation by one
which adds nothing as the (n,n) term is eliminated from the sum by the 𝑘 ≠ 𝑗
requirement. It is useful to rewrite W as

1 𝑞𝑘 𝑞𝑗 1 1 𝑞𝑘
𝑊 = ∑𝑛𝑗=1 ∑𝑛𝑘=1 𝑛
= ∑𝑛𝑗=1 𝑞𝑗 ∑𝑘=1 (11)
2 4𝜋𝜀0 𝑟𝑘𝑗 2 4𝜋𝜀0 𝑟𝑘𝑗
𝑘≠𝑗 𝑘≠𝑗

Where 𝑊 is the total electrostatic energy of the assembled n-charge system which
is the sum of the ΔW’s.

Equation (11) may be written in a somewhat different way by noting that the final
value of the potential U at the jth point charge is
1 𝑞𝑘
𝑈𝑗 = ∑𝑛𝑘=1 (12)
4𝜋𝜀0 𝑟𝑘𝑗
𝑘≠𝑗

Thus, the electrostatic of the system is

1
𝑊 = ∑𝑛𝑗=1 𝑞𝑗 𝑈𝑗 (13)
2

2. Electrostatic Energy of a Charge Distribution

Suppose we assemble the charge distribution by bringing in charge increments 𝛿𝑞
from a reference potential 𝑈𝐴 = 0. If the charge distribution is partly assembled and
the potential at a particular point in the system is 𝑈′(𝑥, 𝑦, 𝑧), then, from equation (1),
the work required to place 𝛿𝑞 at this point is

𝛿𝑊 = 𝑈′(𝑥, 𝑦, 𝑧)𝛿𝑞 (14)

The charge increment 𝛿𝑞 may be added to a volume element located at (𝑥, 𝑦, 𝑧),
such that 𝛿𝑞 = 𝛿𝜌∆𝑣 or 𝛿𝑞 may be added to a surface element at the point in
question, whereby 𝛿𝑞 = 𝛿𝜎∆𝑎. The total electrostatic energy of the assembled
charge distribution is obtained by summing contributions of the form (14).

1 1
𝑊 = ∫𝑣 𝜌𝑈 𝑑𝑣 + ∫𝑆 𝜎𝑈 𝑑𝑎 (15)
2 2

It was stipulated that conductors are present in the system. The last integral involve,
in part, integrations over the surface of these conductors; since a conductor is an
equipotential region, each of these integrations may be done:

3
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

1 1
∫ 𝜎𝑈 𝑑𝑎 = 𝑄𝑗 𝑈𝑗 (16)
2 𝑐𝑜𝑛𝑑𝑢𝑐𝑡𝑜𝑟 𝑗 2

Where 𝑄𝑗 is the charge on the jth conductor. Hence equation (15) becomes

1 1 1
𝑊 = ∫𝑣 𝜌𝑈 𝑑𝑣 + ∫𝑆 ′ 𝜎𝑈 𝑑𝑎 + ∑𝑗 𝑄𝑗 𝑈𝑗 (17)
2 2 2

where the last summation is over all conductors, and the surface integral
is restricted to nonconducting surfaces.

Since, in many problems of practical interest all of the free charge resides on the
surfaces of conductors. In these circumstances equation (17) reduces to

1
𝑊 = ∑𝑗 𝑄𝑗 𝑈𝑗 (18)
2

3. Energy density of an electrostatic field

Consider an arbitrary distribution of free charge characterized by the densities 𝜌 and
𝜎. The densities 𝜌 and 𝜎 are related to the electric displacement;
⃗⃗⃗ throughout the dielectric regions, and 𝜎 = 𝐷
𝜌 = div D ⃗⃗. 𝑛̂ on the conductor surfaces.
Hence, equation (15) becomes

1
⃗⃗ 𝑑𝑣 + 1 ∫ 𝑈 𝐷
𝑊 = ∫𝑣 𝑈 div ⃗D ⃗⃗. 𝑛̂ 𝑑𝑎 (19)
2 2 𝑆

⃗⃗⃗ is different from zero, and

The volume integral here refers to the region where div D
this is the region external to the conductors. The surface integral is over the
conductors.

The integrand in the first integral of equation (19) may be transformed by means of
the following vector identity; 𝑑𝑖𝑣 𝜑𝐹⃗ = 𝜑 𝑑𝑖𝑣 𝐹⃗ + 𝐹⃗ . ∇
⃗⃗𝜑, into

𝑈 div ⃗D
⃗⃗ = div UD
⃗⃗⃗ − 𝐷
⃗⃗. ∇
⃗⃗𝑈

Substituting in equation (19), we obtain

1
⃗⃗⃗ 𝑑𝑣 − 1 ∫ 𝐷
𝑊 = ∫𝑣 div UD ⃗⃗𝑈 𝑑𝑣 + 1 ∫ 𝑈 𝐷
⃗⃗. ∇ ⃗⃗. 𝑛̂ 𝑑𝑎
2 2 𝑣 2 𝑆

4
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

By using the divergence theorem, the first term in last equation can be transformed
to surface integral, thus the last equation becomes

1
⃗⃗⃗. 𝑛̂′ 𝑑𝑣 + 1 ∫ 𝐷
𝑊 = ∫𝑆+𝑆′ UD ⃗⃗. ⃗E⃗ 𝑑𝑣 + 1 ∫ 𝑈 𝐷
⃗⃗. 𝑛̂ 𝑑𝑎 (20)
2 2 𝑣 2 𝑆

This equation may be simplified substantially. The surface S + S' over which the
first integral of (20) is to be evaluated is the entire surface bounding the volume V.
It consists, in part, of S (the surfaces of all conductors in the system), and also of S'
(a surface which bounds our system from the outside, and, which may choose to
locate at infinity). In both cases the normal n' is directed out of the volun1e V. In the
last integral the normal n is directed out of the conductor, hence into V. Thus, the
two surface integrals over S cancel each other. It remains to show that the integral
over S' vanishes.
If the charge distribution, which is arbitrary but bounded, bears a net charge, then at
large distances from the charge system the potential falls off inversely as the
distance, i.e., as r-1. D falls off as r-2. The area of a closed surface which passes
through a point at distance r is proportional to r2. Hence the value of the integral over
S' which bounds our system at distance r is proportional to r-1 and S' is moved out
to infinity, its contribution vanishes.

Thus, for the electrostatic energy, we have

1
⃗⃗ . ⃗E⃗ 𝑑𝑣
𝑊 = ∫𝑣 𝐷 (21)
2

Hence, the energy density in an electrostatic field is

1
𝜔= 𝐷⃗⃗. ⃗E⃗ (22a)
2

Since equation (21), was derived on the basis o:f linear dielectrics, each dielectric is
characterized by a constant permittivity 𝜀. Furthermore, the discussion in preceding
chapters has been limited to isotropic dielectrics. Thus equation (22a) is equivalent
to
1
𝜔 = 𝜀𝐸 2 (23)
2

5
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

Problem.1: If the electric field in a vacuum is defined by 𝐸⃗⃗ = 2𝑟 sin 𝜃𝑎̂𝑟 − 𝑟 cos 𝜃 𝑎̂𝜃 .
Find the energy in the region 0 ≤ 𝑟 ≤ 5 and 0 ≤ 𝜃 ≤ 𝜋/3.
𝜀
Solution: by using 𝑊 = 0 ∫ 𝐸2 𝑑𝑣, we obtain 2
𝜀0 2𝜋 𝜋/3 5 9
𝑊= ∫ ∫0 ∫0 [4𝑟 2
2 0
sin 𝜃 + 𝑟 2 cos2 𝜃]𝑟 2 𝑑𝑟 sin 𝜃𝑑𝜃𝑑𝜑 = 54 𝜋𝜀0
2
5
H.W: If the electrostatic potential in a vacuum is defined by 𝑈 = 𝑥 − 𝑦 + 𝑥𝑦 + 2𝑧. Find
the electric field at the point (1,2,3) and the energy stored in a cube of length a centered at
the origin.
Problem.2: A charge distribution with spherical symmetry has density
𝜌
𝜌={ 0 0≤𝑟≤𝑅
0 𝑟>𝑅
Determine U everywhere and the energy stored in region r < R.

Solution:
⃗⃗ field has already been found in chapter two using Gauss's law.
The 𝐷
𝜌 𝑅3
a) For 𝑟 ≥ 𝑅 , 𝐸⃗⃗ = 0 2 𝑎̂𝑟
3𝜀0 𝑟
Once 𝐸⃗⃗ is known, U is determined as

𝜌 𝑅 1 3
𝑈 = − ∫ 𝐸⃗⃗ . 𝑑𝑙⃗ = − 0 ∫ 2 𝑑𝑟
3𝜀 𝑟 0

𝜌0 𝑅3
= + 𝐶1
3𝜀0 𝑟
Since, 𝑈(𝑟 = ∞) = 0, 𝐶1 = 0
𝜌0 𝑟
b) For 𝑟 ≤ 𝑅, 𝐸⃗⃗ = 𝑎̂𝑟
3𝜀0
hence,
𝜌0
𝑈 = − ∫ 𝐸⃗⃗ . 𝑑𝑙⃗ = − ∫ 𝑟𝑑𝑟
3𝜀0

𝜌0 𝑟 2
=− + 𝐶2
6𝜀0
To determine the constant 𝐶2 equating the potential at r=R
𝜌0 𝑅2
From part (a) , 𝑈(𝑟 = 𝑅 ) = , hence
3𝜀0

𝜌0 𝑅2 𝜌0 𝑅2 𝜌0 𝑅2
=− + 𝐶2 , thus, 𝐶2 =
3𝜀0 6𝜀0 2𝜀0

and

6
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

𝜌0
𝑈= (3𝑅2 − 𝑟 2 )
6𝜀0

Thus, from part (a) and (b)

𝜌0 𝑅3
𝑟≥𝑅
3𝜀0 𝑟
𝑈={ 𝜌0
(3𝑅2 − 𝑟 2 ) 𝑟≤𝑅
6𝜀0

(c) The energy stored is given by

1
⃗⃗ . ⃗E⃗ 𝑑𝑣 = 1 𝜀0 ∫ 𝐸2 𝑑𝑣
𝑊 = ∫𝑣 𝐷
2 2

𝜌 𝑟
For 𝑟 ≤ 𝑅 𝐸⃗⃗ = 0 𝑎̂𝑟
3𝜀0

H.W: Find the energy stored in free space for the region 2𝑚𝑚 < 𝑟 < 3𝑚𝑚, 0 < 𝜃 <
200 300 𝑐𝑜𝑠 𝜃
900 , 0 < 𝜑 < 900 , given the potential field (a) 𝑉, (b) 2
𝑉. Ans.: 1.391pJ, 36.7J
𝑟 𝑟

7
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

Problem.3:

8
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

4. Energy of a system of charged conductors. Coefficients of potential

It was shown that a linear relationship exists between the potentials and charges on
a set of conductors. In fact, in a system composed of N conductors, the potential of
one of them is given by

𝑈𝑖 = ∑𝑁
𝑗=1 𝑝𝑖𝑗 𝑄𝑗 (24)

The derivation of equation (24) was carried out for N conductors in vacuum;
however, it is clear that, this derivation also holds when dielectrics are present in the
system, so long as these dielectrics are both linear and devoid of free charge. The
coefficient 𝑝𝑖𝑗 is the potential of the ith conductor due to a unit charge on conductor
j. These coefficients are usually referred to as coefficients of potential. Combing
equations (18) and (24), we obtain
1
𝑊 = ∑𝑁 𝑁
𝑖=1 ∑𝑗=1 𝑝𝑖𝑗 𝑄𝑖 𝑄𝑗 (25)
2

Thus the energy is a quadratic function of the charges on the various conductors.

A number of general statements can be made about the coefficients 𝑝𝑖𝑗 , which are:
1- 𝑝𝑖𝑗 = 𝑝𝑗𝑖
2- all of the 𝑝𝑖𝑗 are positive
3- 𝑝𝑖𝑖 − 𝑝𝑖𝑗 ≥ 0

To prove the first statement we can expresses 𝑊 as 𝑊(𝑄1 , … . . , 𝑄𝑁 ); thus

𝜕𝑊 𝜕𝑊
𝑑𝑊 = ( ) 𝑑𝑄1 + ⋯ … … . + (𝜕𝑄 ) 𝑑𝑄𝑁
𝜕𝑄1 𝑁

If 𝑑𝑄1 only is changed, then

𝜕𝑊 1
𝑑𝑊 = (
𝜕𝑄1
) 𝑑𝑄1 = 2 ∑𝑁
𝑗=1(𝑝1𝑗 + 𝑝𝑗1 )𝑄𝑗 𝑑𝑄1 (26)

Bringing in 𝑑𝑄1 from a zero potential reservoir can also be obtained using the
expression 𝑊 = 𝑄𝑈, thus,

𝑑𝑊 = 𝑈1 𝑑𝑄1 = ∑𝑁
𝑗=1 𝑝1𝑗 𝑄𝑗 𝑑𝑄1 (27)

9
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

The last two equations must be equivalent for all possible values of the 𝑄𝑗 , which
implies that

1
2
(𝑝1𝑗 + 𝑝𝑗1 ) = 𝑝1𝑗 , or

𝑝𝑗1 = 𝑝1𝑗 (28)

H.W: Prove the second and third statements.

The usefulness of the coefficients 𝑝𝑖𝑗 may be illustrated by means of a simple

example. Problem: to find the potential of an uncharged spherical conductor in the
presence of a point charge 𝑞 at distance r, where r > R, and R is the radius of the
spherical conductor. The point charge and sphere are taken to be a system of two
conductors, and use is made of the equality 𝑝12 = 𝑝21 . If the sphere is charged (Q)
𝑄
and the ‘’point" uncharged, then the potential of the point" is ; thus
4𝜋𝜀0 𝑟

1
𝑝12 = 𝑝21 =
4𝜋𝜀0 𝑟

Evidently, when the ‘’point" has charge 𝑞 and the sphere is uncharged, the potential
𝑞
of the latter is .
4𝜋𝜀0 𝑟

5. Coefficients of capacitance and induction

Equation (24) is a set of N linear equations giving the potentials of the conductors
in terms of their charges. This set of equations may be solved for the 𝑄𝑖 ’𝑠, yielding

𝑄𝑖 = ∑𝑁
𝑗=1 𝑐𝑖𝑗 𝑈𝑗 (29)

where 𝑐𝑖𝑗 is called a coefficient of capacitance and 𝑐𝑖𝑗 (𝑖 ≠ 𝑗) is a coefficielt of

induction. The actual inversion of equation (24) expressing each c in terms of the
𝑝𝑖𝑗 , is most easily done by using deterlninants.

Properties of the 𝑐𝑖𝑗 follow from those of the 𝑝𝑖𝑗 , which are
1- 𝑐𝑖𝑗 = 𝑐𝑗𝑖 2- 𝑐𝑖𝑖 > 0 3- The coefficients of induction are negative or
zero.

10
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

Equation (29) may be combined with equation (18) to give an alternative expression
for the electrostatic energy of an N-conductor system:
1
𝑊 = ∑𝑁 𝑁
𝑖=1 ∑𝑗=1 𝑐𝑖𝑗 𝑈𝑖 𝑈𝑗 (30)
2

6. Capacitors
Two conductors which can store equal and opposite charges(∓𝑄), independently of
whether other conductors in the system are charged, form what is called a capacitor.
Such a situation is depicted in fig.(1)where conductors 1 and 2 form a device of this
type. In general, if two conductors, 1 and 2, form a capacitor, we may write

𝑈1 = 𝑝11 𝑄 + 𝑝12 (−𝑄) + 𝑈𝑥

(31)
𝑈2 = 𝑝12 𝑄 + 𝑝22 (−𝑄) + 𝑈𝑥

where ∓𝑄 are the charges stored and 𝑈𝑥 is the common potential contributed by
other (external) charges.

If equations (31) are subtracted, we find

∆𝑈 = 𝑈1 − 𝑈2 = (𝑝11 + 𝑝22 − 2𝑝12 )𝑄 (32)

Thus the difference in potential between the conductors of a capacitor is proportional

to the charge stored, Q. (Obviously, the total charge stored is zero but by convention,
the absolute value of the charge on one of the two conductors is called the charge on
the capacitor.)

Fig.(1): Conductors 1 and 2 form a capacitor. 𝑝31 = 𝑝32 since, by Gauss' law, when 1
and 2 are uncharged they must be at the same potential, independently of the charge on 3.
Similarly, 𝑝41 = 𝑝42 .

11
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

Equation (32) may be written

𝑄 = 𝐶∆𝑈 (33)

Where 𝐶 = (𝑝11 + 𝑝22 − 2𝑝12 )−1 is called the capacitance of the capacitor.
Evidently C is the charge stored per unit of potential difference; in the mks system
C is measured in coulombs/volt, or farads (1 farad ≡ 1 coulomb/volt).

Using the results of previous sections in this chapter, the energy of a charged
capacitor may be written as

1 1 1 𝑄2
𝑊 = 𝐶(∆𝑈)2 = 𝑄∆𝑈 = (34)
2 2 2 𝐶

If the two conductors making up the capacitor have simple geometrical shapes, the
capacitance may be obtained analytically. Thus, for example, it is easy to calculate
the capacitance of two parallel plates, two coaxial cylinders, two concentric spheres,
or that of a cylinder and a plane. The capacitance of a parallel-plate capacitor (Fig.
2) will be derived here;

Fig.2: The electric field between oppositely charged parallel plates of finite area.

Except for the fringing field at the edge of the parallel plates, the electric field
between them is uniform. An ideal parallel-plate capacitor is one in which the plate
separation d is very small compared with dimensions of the plate; thus, the fringing
field may be neglected in the ideal case. If the region between the plates is filled,
with dielectric of permittivity 𝜀, then the electric field between the plates is

12
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

1 𝑄
𝐸= 𝜎= (35)
𝜀 𝜀𝐴

where A is the area of one plate. The' potential difference ∆𝑈 = 𝐸𝑑. Therefore,

𝑄 𝜀𝐴
𝐶= = (36)
∆𝑈 𝑑

is the capacitance of this capacitor.

Possible ways of joining two capacitors are by parallel connection (Fig.3a), or by
series connection (Fig.3b). After the capacitors are joined, the described by the
equivalent capacitance of the combination.
In the parallel case, the same voltage ∆𝑈 appears across each capacitor also appears
across the combination; hence the equivalent capacitance is given by

𝑄𝑡𝑜𝑡𝑎𝑙
𝐶= = 𝐶1 + 𝐶2 (37a)
∆𝑈

If two uncharged capacitors are connected in series and subsequently charged,

conservation of charge requires that each capacitor acquire the same charge. Thus
the equivalent capacitance C of the combination is related to C 1 and C2 by the
expression

1 1 1
= + (37b)
𝐶 𝐶1 𝐶2

Fig.(3): (a) Parallel and (b) series connection of two capacitors.

7. Forces and Torques

13
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

In this chapter we have developed a number of alternative procedures for calculating

the electrostatic energy of a charge system. We shall now show how the force on
one of the objects in the charge system may be calculated from a knowledge of this
electrostatic energy. Let us suppose we are dealing with an isolated system
composed of a number of parts (conductors, point charges, dielectrics), and we allow
one of these parts to make a small displacement 𝑑𝑟⃗ under the influence of the
electrical forces acting upon it.

The mechanical work performed by the system in these circumstances is

𝑑𝑊𝑚 = 𝐹⃗ . 𝑑𝑟⃗

= 𝐹𝑥 𝑑𝑥 + 𝐹𝑦 𝑑𝑦 + 𝐹𝑧 𝑑𝑧 (38)
Because the system is isolated, this work is done at the expense of the electrostatic
energy W; in other words,

𝑑𝑊 + 𝑑𝑊𝑚 = 0 (39)

Combining equations (38) and (39)

−𝑑𝑊 = 𝐹𝑥 𝑑𝑥 + 𝐹𝑦 𝑑𝑦 + 𝐹𝑧 𝑑𝑧

𝜕𝑊
and 𝐹𝑥 = − (40)
𝜕𝑥
with similar expressions for Fy and Fz.

If the object under consideration is constrained to move in such away that it rotates
about an axis, then equation (38) may be replaced by

𝑑𝑊𝑚 = 𝜏⃗. 𝑑𝜃⃗ (41)

Where 𝜏⃗ is the electrical torque and 𝑑𝜃⃗ is the angular displacement, writing
𝜏⃗ and 𝑑𝜃⃗ in terms of their components (𝜏1 , 𝜏2 , 𝜏3 ), (𝑑𝜃1 , 𝑑𝜃2 , 𝑑𝜃3 ),
and combining equations (39) and (41), we obtain
𝜕𝑊
𝜏1 = − and similar to 𝜏2 and 𝜏3 .
𝜕𝜃1

14
Electromagnetic Theory 2018-2019 Prof. dr. Ali Hadi Hassan Al-Batat 1-15

The electrostatic energy W of a system of charged conductors has been given earlier,
in equation (13). If, now, some part of the system is displaced while at the same time
the potentials of all conductors remain fixed,

1
𝑑𝑊 = ∑𝑗 𝑈𝑗 𝑑𝑄𝑗 (42)
2

In general, the energy of the system may be calculated in terms of the electric field
vector from the following equation;

1
𝑊 = ∫𝑣 𝜀𝐸 2 𝑑𝑣 (43)
2

and, the electrostatic energy of a system of charged conductors and dielectrics has
been obtained in terms of the field vectors as follows

1
⃗⃗ . 𝐸⃗⃗ 𝑑𝑣
𝑊 = ∫𝑣 𝐷 (44)
2

Depending References:
1- Foundations of Electromagnetic Theory, Second Edition, by Reitz
2- Elements of Electromagnetics, Sadiku, 2000

15