Chapter 3 GAUSS’ LAW

• Introduction
• Flux of vector field
• Solid angle
• Gauss’s Law
• Symmetry
• Spherical symmetry
• Cylindrical symmetry
• Plane symmetry Figure 1 E field at point p due to a charge distribution
• Superposition of symmetric geometries (0  0  0 )

• Motion of point charges in electric fields

• Summary
FLUX OF VECTOR FIELD
Vector fields, such as an electric field, magnetic field,
fluid flow, etc, have both a magnitude and direction
INTRODUCTION that depend on the position. Vector fields can be char-
acterized using two important concepts, namely, flux
Coulomb’s Law and the Law of Superposition were and circulation. For example, the rate of loss of wa-
introduced in addition to the concept of electric field ter out of a draining bath tub can be related to the
defined as: litres/sec flowing down the drain and the circulation
→
− or angular momentum carried away by the flowing wa-
−
→ F ter. The concepts of flux and circulation make it pos-
E= (3.1)
 sible to express Coulomb’s Law in a form that is more
The electric field is a vector field just like the gravita- general.
tional field. It can be expressed in terms of cartesian Flux, Φ is the normal component of the vector
→
−
coordinates as: field, F through any surface. It is given by the in-
− b
→ tegral over the surface.
 + b
E =  + b  (3.2) Z
Last lecture focussed on the calculation of the electric − −
→ →
Φ= F · S (3.5)
field using Coulomb’s law plus Superposition.  
Coulomb’s Law + Superposition where the scalar product, also called dot product, was
Discrete charges: → −
− →
defined in P113, namely F ·  S =  cos  where
→
−
 is the angle between the vector F and the normal
 −
→
−−→ X 1  to the surface element S Note that the flux is the
E = c
2 r0 (3.3) normal component to the surface of the vector field, it
=1
40 0
excludes the component along the surface; the surface
Continuous charges: component figures in the definition of circulation to be
In cartesian coordinates the E field at the point discussed next lecture. The most interesting case is
() is written as an integral over the charge distri- when the surface is closed, this is called a Gaussian
bution at 0 (0  0  0 ) : surface. In this case the net flux outwards is written
Z as:
−−−−→ 1 (0  0  0 )0  0  0
E() = rd
0  (3.4)
40  20  I
− −
→ →
Φ= F · S (3.6)
This lecture will focus on exploiting Gauss’s law  
to calculate electric fields. Gauss’s Law relates the I
net flux of the electric field over a closed surface to where the symbol means a closed surface. By de-
the net charge enclosed by the closed surface. This
−
→
is of sufficient import to justify a recap of what was finition, for a closed surface the surface vector S is
discussed last lecture. assumed to point outward, that is Φ is taken outwards.
17
 Figure 3 Solid angle for a closed surface, a) enclosing
Figure 2 Solid angle subtended by an area element dS the point, and b) not enclosing the point.
and analog to the angle subtended by a line element dl.

Point enclosed by a closed surface The inte-

SOLID ANGLE gral for a closed surface surrounding a point is

I −
→
The concept of solid angle should be familiar to you. b
r · S
Ω= = 4 (3.9)
It is a measure of the apparent size of a surface area as 2
seen when viewing from some location. It is analogous
to the two-dimensional problem of the angle subtended since the surface area of a sphere is 42 then a com-
by a line when viewed from a given location. It is plete sphere must subtend a total solid angle of 4
easiest to understand solid angle by comparison with −→
steradians. The vector S is always taken to point out-
the concept of angle. wards for a closed surface. Note that any shaped closed
In two dimensions, the infinitesimal angle ∆ sub- surface subtends 4 steradians relative to a point in-
tended by a short line element at a distance r is given side the closed surface, that is, at all angles the closed
by the projection of the line element  on a concen- surface completely encloses the point independent of
tric circle divided by the radius r of the circle. That whether the enclosing surface is a sphere, a cube, or
is, ∆ =  sin


The unit of angle is the radian. some arbitrary shape.
The analogous relation for three dimensions gives
the definition of solid angle. Consider a surface area
−
→
element S at a radius r. Then the solid angle ∆Ω is Point external to closed surface For points
defined as: lying outside the closed surface, the net solid angle is
zero. This can be seen by dividing the closed surface
−
→ into two halves having the same perimeter as seen from
S · b
r  cos 
∆Ω = = (3.7) the point  illustrated in the figure. Both halves sub-
2 2 tend the same magnitude solid angle. However, since
−→
where  is the angle between the normal to the sur- the vector S is always taken to point outwards, then
face element and unit vector br. The solid angle is the the cosine has the opposite sign for the two halves and
apparent area of the surface element projected onto a thus the integral of the solid angle over the external
concentric sphere, divided by the square of the radius closed surface is zero since the contributions from the
of the sphere. The unit of solid angle is the steradian two halves cancel.
which is dimensionless. Thus the net solid angle sub-
tended by some surface is given by
Z Z −
→
Ω =
b
r · S
(3.8) FLUX OF ELECTRIC FIELD FOR AN
2
ARBITRARY GAUSSIAN SURFACE
The sun and the moon happen to subtend almost ENCLOSING A POINT CHARGE
identical solid angles at the earth even though there
is a factor of 200 difference in size which is cancelled Since the electric field for a point charge is given by
by a similar ratio of the distances from the earth. The
solid angle of a closed surface is an important special −
→ 1 
case that will be used frequently. E= b
r (3.10)
40 2
18
For an arbitrary closed surface enclosing the point
charge  the net flux is
I
→ −
− →
Φ = E · S
I Ã − →!
 b
r · S
= (3.11)
40 2

But the above discussion of the solid angle gave that

Z Z −
→
br · S
Ω = (3.12)
2
This is just the term in brackets, that is
I
 Figure 4 A single point charge surrounded by a con-
Φ= Ω (3.13)
40 centric closed spherical surface and an irregular closed
Also from above we have that for a closed surface en- surface.
closing the point charge the total solid angle, in stera-
dians, is given by can be extended by invoking the Principle of Superpo-
I I −
→ sition. Superposition can be used to integrate over the
b
r · S charge distribution throughout the volume enclosed by
Ω = = 4 steradians (3.14)
2 the closed Gaussian surface resulting in the final form
This is true for any shaped closed surface surrounding of Gauss’s Law. The Principle of Superposition allows
the point. Thus this gives that the net flux of the extension to arbitrary charge distributions. Consider 
electric field due to an enclosed point charge is given charges within the enclosed surface. Each charge pro-
by duces an  field which add to produce the net electric
I I
→ −
− →   field. That is:
Φ= E · S = Ω = (3.15)
40 0 −−→ − → − → − →
where the 4 factors cancel. E = E1 + E2 + E3 +  (3.16)
The above argument shows that the net flux out of Now since the flux Φ is just a number, that is, it is a
any shaped Gaussian surface enclosing a point charge scalar:
q is independent of the shape of the Gaussian surface
or the exact location of the point charge within the
enclosed volume. The assumptions made in obtaining I
−−→ − →
this result are: Φ = E · S
 
a) Point charge I I
b) Coulomb’s law. −
→ − → −
→ − →
= E1 · S + E2 · S + 
This is a much more powerful statement of Gauss’    
Law in that the net flux is independant of the shape of = Φ1 + Φ2 + Φ3 + 
the closed surface or the location of the enclosed point 1 2 3
= + + +  (3.17)
charge. 0 0 0
The above proof is consistent with Faraday’s de-
Thus one has that for an arbitrary distribution of
scription of the electric field in terms of electric field
 charges, the net flux is:
lines; that is, the number of field lines intersecting a
concentric sphere is constant independant of the ra-
dius of the sphere. Since the number of field lines is I Ã  !
conserved then it can be seen from figure 4 that the −−→ − → X 
Φ = E · S = (3.18)
net flux of electric field lines is the same for any closed 
  =1 0 
arbitrary closed surface enclosing a given point charge
which is what has just been proved mathematically. Finally, one can write the sum over an arbitrary
continuous charge density distribution  by taking
 =  for the infinitessimal volume   to get the
final form of Gauss’s Law.
GAUSS’ LAW Gauss’s Law:
The above proof that for a point charge the net flux I Z
out of a closed surface enclosing a point charge is in- −−→ − → 1
Φ = E · S =  (3.19)
dependent of the shape of the Gaussian surface, now   0 

19
Figure 5 Closed Gaussian surface enclosing the positive
charge (A), the negative charge (B) and both charges
(C). The net flux out of these surfaces is positive for A, Figure 6 Spherical shell of charge density 
negative for B, and zero for C.

SYMMETRY
This relates the net flux out of a closed Gaussian sur-
face to the total charge lying within the enclosed vol- Symmetry is a powerful concept in physics that can
ume. simplify solution of complicated problems. One can
Note that the assumptions used to derive this are: discuss symmetry of systems under rotation, reflection
and time reversal. For example, crystal lattices have
• Coulomb’s Law certain spatial symmetries. Coulomb’s law implies a
spherically-symmetric spatial symmetry for the elec-
• Principle of Superposition tric field around a point charge. Such very general
symmetry principles are especially powerful when com-
Note that the two crucial aspects of Coulomb’s law bined with Gauss’ law for calculating electric fields.
that lead to Gauss’ law are that the electric field for a For certain charge distributions in physics, one can
point charge is: use symmetry to identify a particular Gaussian surface
upon which the electric field is uniform and perpendic-
1 ular to the surface. When such symmetries occur, it is
• exactly proportional to 2 ,
easy to evaluate the flux integral for the special surface
• the field is radial. where the electric field is uniform and perpendicular.
This allows use of Gauss’ Law in order to determine
the electric field for this surface. Three symmetries
Gauss’s law is a restatement of Coulomb’s law in a will be considered; spherical, cylindrical, and plane.
less transparent but more useful form.
Gauss’ Law actually is one of Maxwell’s four laws
of electromagnetism. Gauss’ Law is completely equiv- SPHERICAL SYMMETRY
alent to Coulomb’s law for electrostatics or for slowly
moving charges. However, Gauss’ law is more gen-
A) Electric field for a uniformly charged spher-
eral and applies to electric fields arising from rapidly
ical shell
moving and accelerating charges. Really one should
derived Coulomb’s Law from Gauss’s Law. Consider a uniform spherical surface of radius R and
It is interesting to apply Gauss’ Law to the case of surface charge density  The net charge Q = 42 
the electric dipole. As seen in figure 5, if the Gaussian From spherical symmetry one can see that the electric
surface encloses only the positive charge then the net field must be radial and its magnitude can depend only
flux out is positive, if the Gaussian surface encloses on r, not angle. Therefore select a spherical Gaussian
only the negative charge then the net flux out is neg- surface concentric with the charged shell and apply
ative, that is the flux is flowing into the surface. If Gauss’ law.
the Gaussian surface encloses both charges then the Outside the spherical shell:  ≥ 
net flux out is zero. However, this does not mean that Gauss’ law gives
there is not electric field. Actually one has flux flowing I
inward in some locations and outwards at others, such − −
→ → 
Φ= E · S = (3.20)
that the net total is zero as seen in Figure 5.  0
20
Figure 7 Concentric Gaussian surface outside a solid
uniform spherical charge distribution.
Figure 8 Concentric spherical Gaussian surface inside a
uniform solid sphere of charge.
Since  is constant and normal to the spherical gaussian
surface of radius r, the surface integral equals 42 .
Thus one obtains


42 = (3.21)
0
That is:

−
→ 
E= b
r ≥ (3.22)
40 2
Note that the E field for a spherical shell of charge of
radius R is the same as for a point charge Q. In fact,
one gets Coulomb’s law by letting R go to zero.
Inside the spherical shell  ≤ 
Since there is no charge within a concentric spheri-
cal Gaussian surface inside the charged shell, then the
net flux equals zero from Gauss’ law. From symmetry
therefore we have that the electric field is zero inside
the spherical charged shell. Figure 9

B) Electric field outside a uniformly charged

solid sphere,  ≥  of charge in the region between  and  does not con-
tribute to . Thus one has only to consider the charge
Let the charged sphere, of radius , have a volume within the sphere of radius , that is, 43 3 . Thus
charge density  The net charge  = 43 3 . Again,
from symmetry of figure 7 one can deduce that the 4 3
electric field is radial and only depend on radius . −
→ 3  
E= b
r ≤ (3.24)
For exactly the same argument given for the spher- 40 2
ical shell, the external field is given by:
Since  = 43 3  then this can be rewritten as:

−
→  −
→ 1 −
E= b
r ≥ (3.23) E= →
r ≤ (3.25)
40 2 40 3
Thus the  field is zero at the origin, is propor-
tional to  for 0    , and then falls off as 12
like for a point charge outside of the charged sphere as
C) Electric field inside a uniformly charged solid
shown in figure 9.
sphere,  ≤ 
In the 1900’s there were two possible models for the
From the arguments for the spherical shell it is obvi- atom. Thomson’s model of the atom was of electrons
ous that only the charge within the sphere of radius distributed in a uniform positive charge distribution
r contributes to the electric field, the spherical shell of the size of the atom, that is 10−10  The Rutherford
21
 Figure 11 Gaussian surface for infinite line charge 
C/m.

Figure 10 The electric field in the Thomson and

Rutherford models of the atom.

atom envisioned electrons orbiting around a small pos-

itive nucleus. These models predict a different E field
at radii  10−10 m as shown in figure 10. In 1911, two
of Rutherford’s students, Geiger and Marsden scat-
tered alpha particles off gold atoms and proved that
the electric field due to the nucleus continued to fall
off as 12 down to the size of the nucleus of 10−15  Figure 12 Concentric Gaussian cylinder inside a cylin-
confirming the Rutherford model of the atom. drical shell of charge
The gravitational field of the earth has the same
radial form as the case of a uniform spherical charge.
The only difference between the gravitational and elec- I
trostatic cases is that 10 is replaced by −4 and the − −
→ → 
Φ= E · S = 2r  = (3.26)
mass density replaces the charge density  This leads  0
to the conclusion that inside the earth, the radial de-
pendence of the gravitational field is proportional to , This gives that the E field is:
while outside the earth it has a 12 dependence. Inside
the earth Hooke’s Law applies since the restoring force
is proportional to the displacement . If you dropped −
→ 
E= b
r (3.27)
a body into a hole through the center of the earth then 2 
the body would execute simple harmonic motion with This is the same answer as obtained previously by
a period of 1.4 hours, that is, the length of this lecture. integration of Coulomb’s law.

B) Thin cylindrical shell of charge of radius 

For    then the field is identical to the answer
CYLINDRICAL SYMMETRY above.
For    then there is no charge inside the con-
A) Electric Field for an infinite line charge centric cylindrical Gaussian surface shown in figure 12,
thus the E field is zero.
Coulomb’s law was used to derive the electric field due
to a line of charge density  Coulombs/m. This prob- C) Uniform cylinder of charge of radius 
lem has cylindrical symmetry, that is the electric field
must radiate normal to the line charge. For    the answer is as above for the infinite line
2
Note that a concentric cylinder of length L and charge. However, for    then only the fraction  2

cylindrical radius r perpendicular to the line charge, of the linear charge density is inside the cylindrical
shown in figure 11, contains a charge L. The E field gaussian surface. This causes the the E field to depend
must be uniform and perpendicular to the cylinder and linearly of  just like the case of the uniform sphere of
parallel to the ends of the cylinder. Thus the flux charge. i.e. for   
through the ends of the cylinder is zero. Using Gauss’s −
→ 
law we have the net flux through the cylinder is E= r̄ (3.28)
2 2
22
 Figure 13 Concentric charged cylindrical shells.
Figure 14 Infinite plane sheet of charge density 

D) Two concentric equal and oppositely charged

cylindrical shells; I
→ −
− → 
Consider that the inner cylindrical shell, of radius a, E · S = 2∆ = ∆ (3.31)
  
has a linear charge density of - C/m while the outer
cylindrical shell, of radius b, has a linear charge density That is:
of + C/m as shown in figure 13.
a)    
Obviously E = 0 inside the inner cylindrical shell ⊥ = (3.32)
2
since there is no charge enclosed by a concentric Gaussian
surface lying inside with    Note that this is independent of the distance from
b)      the infinite sheet of charge.
Between the concentric cylinders, Gauss’s law gives
for a concentric cylindrical Gaussian surface of radius
r perpendicular to the symmetry axis and length l; SUPERPOSITION OF SYMMETRIC
− GEOMETRIES
2 = (3.29)

Gauss’s law is very powerful for deriving the electric
Thus the electric field between the shells is:
field for those few case where a symmetry exists such
−
→  that a Gaussian surface can be used for which you
E =− b
r (3.30) know the relative magnitude and direction of the E
2 
field. Unfortunately, for most cases there is no sym-
where b is perpendicular to the symmetry axis. Note metry and thus one must use the brute-force method
that the electric field points inwards since the inner and evaluate the integral:
charge density is negative.
c)    Z
The radial electric field outside of both cylindrical −
→ 1 
E= b
r (3.33)
shells is zero since the net charge enclosed by a cylin- 40 2
drical Gaussian surface with    is zero if the two However, there are cases where superposition of
cylinders have equal and opposite line charge densities. symmetric systems can be used to calculate the fields
due to non-symmetric systems. For example, consider
the system shown in figure 15 where a cylindrical hole
PLANE SYMMETRY of radius 0 is drilled parallel to the axis of a uniform
cylindrical charge distribution of radius  and a dis-
tance  between the axes of the cylinders. This can be
Consider a infinite plane sheet of areal charge density
solved by considering the superposition of the electric
 C/m2 . By symmetry the electric field must be per-
fields due to the uniform positive cylindrical charge
pendicular to the sheet charge and equal on both sides.
distribution of radius  plus that from the negative
Consider a small cylindrical Gaussian pill box enclos-
cylindrical distribution of radius 0 displaced by  that
ing an area ∆ of the charged surface shown in Figure
cancels the charge distribution inside cylindrical hole.
14.
The net flux flowing out of each of the two flat
surface of the pillbox is ∆ and zero flux from the
cylindrical sides. Thus Gauss’s law gives:
23
Figure 15 Uniform cylindrical charge distribution of
radius  having a parallel cylindrical hole, of radius 0  Figure 16 An electron moving transverse to an electric
drilled a distance  from the axis. field

MOTION OF POINT CHARGES IN

ELECTRIC FIELDS −
→ b = 0bi +  bj
v = bi + bj +  k (3.36)

Having used Coulomb’s law, or Gauss’s law, to com-
The charged particle travels a time  in the region of
pute the electric field distribution, then it is possible to
the  field given by:
use Newton’s Laws of motion to calculate the motion
of charged objects in this electric field. The acceler- 1
ation, , of a charged object of mass,  and charge,  = (3.37)
0
 in an electric field, , is given by Newton’s Law
→
− →
− The particle’s deflection when it exits the E field at
F =  E = − →
a  That is:
time  is:
−
→ −→
a = E (3.34) −
→ 1  2b
 r =  bi +  j (3.38)
2 
For uniform electric field  one has the simple case of
The particle now continues in a straight line in the
motion with constant acceleration which you discussed
field-free region at an angle  given by:
extensively in P113. It is useful to consider a simple
case to illustrate the application of Newton’s law of
motion to a problem in electrostatics.  ()(1 0 ) 1
tan  = = = (3.39)
 0 02
Example: Deflection of moving charged parti-
cles. Thus the net deflection in the bj direction is:

Consider the case encountered in a cathode ray tube

used in a television or computer CRT display. An elec-  21
 = 1 +2 = 1 +2 tan  = ( +1 2 ) (3.40)
tron, of mass  and charge , initially at velocity v0 02 2
in the + direction, travels perpendicular to a uniform
Note the important fact that the deflection is propor-
non-zero  field between 0 ≤  ≤ 1 as shown in fig-
tional to , that is, the deflection of the electron beam
ure 16. The deflected electron then travels a further
in your television screen reproduces the magnitude of
distance 1 ≤  ≤ 1 + 2 in a field-free region before
the electric field derived from the radio signal.
hitting the screen. Calculate the deflection when the
Inserting typical values, where 1 = 3, 2 =
electron hits the screen.
12 0 = 3106 , and = 103 C, then:
Assuming that the E field points in the − direc-
tion, then, since the electron charge is − the accel-
eration vector −→
a equals:
(16 10−19 )(103 ) 1
 = ( (003)2 + (003)(012))
→
− b =  bj
a = bi + bj +  k (3.35) (911 10−31 )(3 106 )2 2

= 80 × 10−2  (3.41)
Assume that the electron enters the E field at  = 0
and  = 0. Because the initial velocity is only in the Of this 80m deflection, 1 = 09 and 2 =
 direction and the  acceleration is zero, then the 71. That is, most of the deflection comes from the
velocity vector at time t is: angular deflection of the beam.
24
 SUMMARY
Two methods have been presented for calculating the
electric field.
Coulomb’s Law:
Discrete charges:


−−→ X 1 
E = c
2 r0 (3.42)
=1
40 0

Continuous charges:
In cartesian coordinates the E field at the point
() is written as an integral over the charge distri-
bution at 0 (0  0  0 ) :
Z
−−−−→ 1 (0  0  0 )0  0  0
E() = rd
0  (3.43)
40  20 

Gauss’s Law

I Z
−−→ − → 1
Φ = E · S =  (3.44)
   0 

Gauss’s law combined with symmetry was shown to be

a simple method for calculating electric fields for those
systems having a symmetry. In particular, spherical,
cylindrical and planar symmetry were considered.
Once the electric field has been computed, then it
is possible to predict the motion of charged particles in
this field using Newton’s laws of motion. The motion
of an electron in an electric field was discussed.
Reading assignment: Giancoli, Chapter 22

25