Physics 2415 Lecture 5: Using Gauss' Theorem: Spheres, Lines, Planes
Physics 2415 Lecture 5: Using Gauss' Theorem: Spheres, Lines, Planes
Physics 2415 Lecture 5: Using Gauss' Theorem: Spheres, Lines, Planes
Spherical Shell
A good example is finding the electric field from a uniformly charged spherical shell, say charge Q and
radius R. Since the sphere is uniformly charged, it has perfect spherical symmetry, it is not altered by
turning the sphere through some angle. Therefore, the electric field must also be spherically symmetric.
The only spherically symmetric electric field has the field pointing directly outwards (or inwards) from
the center at all points.
Let’s apply ∫ ( total charge inside surface ) / ε 0
E ⋅ dA =
closed surface
to a spherical surface of radius r bigger than the sphere of charge, but with the same center.
The field E points outwards everywhere on the surface, so it’s parallel to dA , and has the same
strength everywhere on the sphere, by symmetry. The total area of the sphere is 4π r 2 , so the integral
is equal to 4π r 2 E , and outside the sphere of charge:
Q
E (r ) = rˆ
4πε 0 r 2
the same as for a point charge at the center. It’s worth mentioning that since gravity is also an inverse
square force, this same result is true for the gravitational field from a spherical shell of mass. (This can
be proved using Coulomb’s Law or its gravitational equivalent, but it’s quite difficult—it’s done here.)
What about the electric field inside the sphere? We do the same trick: integrate over a spherical surface
with the same center as the sphere of charge. This time, though, there is no charge inside our smaller
spherical surface, so the electric field must be exactly zero inside the sphere.
The complete picture of the electric field for a uniformly charged shell is therefore:
2
no field in
shell
E ∝ 1/ r 2 outside
Solid Sphere
The key for any spherically symmetric charge distribution is superposition: the distribution can be
expressed as the sum of (or integral over) spherical shells. The contribution from each shell is zero
inside that shell, and equal to that from a point charge at the center outside the shell. So, for the case
of a uniformly charged (throughout the volume) sphere, outside the whole sphere the field is the same
as if all the charge were at the center, inside the solid sphere, at distance r from the center, it’s the same
field as from a point charge at the center equal to the amount of charge in a sphere of radius r : in other
words, there is no contribution from those shells the point is inside. This uniformly charged sphere is
not a likely object to find in electrostatics, but it is exactly equivalent to the gravitational field for a
sphere of uniform density, a much more realistic problem. And, in fact, the electrostatic uniformly
charged sphere was a subject of intense interest a century ago, as a possible model for the atom: before
the nucleus was discovered, but it was already known that the atom contained negatively charged
electrons, it was suggested that the positive charge was spread over a sphere, and the electrons were
inside this sphere: this was called the plum pudding model.
3
For a nonuniform spherical distribution, the same approach works: the field at any point is equivalent to
a point charge at the center equal to all the charge between the point and the center.
from which
Electric field vectors and Gaussian surface
of integration for a positively charged λ
E (r ) = .
infinite straight wire 2π rε 0
This same method applies for finding the electric field from a uniformly charged cylinder of charge. Just
imagine the wire in the picture above being replaced by a fatter wire, then by a hollow cylinder, but
staying inside the Gaussian cylindrical surface we integrate over. We get the identical result for E(r),
now we must interpret λ as the charge on one meter of the whole cylinder. If this is a hollow cylinder, a
pipe, taking a Gaussian surface inside it, the surface encloses no charge, so the electric field inside a
hollow cylinder from the charge on the cylinder is zero.
Coaxial Cable
Of course, we could add a line of charge, or even another cylinder, inside our charged cylinder, in which
case the total electric field would be the sum of the electric fields from the two cylinders, using
superposition. In fact, this is a coaxial cable, the cable used to transmit TV signals. etc. A coaxial cable
(the word means “same axis”) has a central copper wire, inside a hollow copper cylinder (see figure
below). Between the two is a nonconducting dielectric—we’ll discuss dielectrics shortly. The
transmission of electromagnetic waves, the TV signal, is of course not an electrostatic situation, but
nevertheless Gauss’ Law still holds, and at any moment there are equal amounts of charge per unit
length of cylinder on the surface of the central wire and the inner surface of the cylinder, and
consequently an electric field as shown (there are also currents in the copper producing magnetic
fields—more about that later).
4
E
From symmetry, taking the plane to have infinite extent, the field must be perpendicular to the plane as
shown above, where the plane of charge is seen in cross section, that is, the plane is perpendicular to
the paper. Of course, the charge is distributed uniformly over the plane, with area density σ coul/m2.
To use Gauss’ Law, we choose a surface shaped like a pillbox, represented in cross section by the
5
rectangle above. The top and bottom surfaces both have area A, and an area A of the charged plane is
included. The electric field is parallel to the area vectors on both the top and bottom surfaces, so the
total contribution from those surfaces to ∫ ⋅ dA =
E 2 EA . There is no contribution to the integral from
the sides of the pillbox, as the electric field is parallel to those sides, thus E ⋅ dA is zero there. It follows
immediately that
∫
surface
E ⋅ dA= 2 EA= Aσ / ε 0 , so E = σ / 2ε 0 .
For an actual physical finite plane of charge, this value of E is a good approximation at points close to the
surface relative to the size of the plane. For distances large compared to the extent of the plane, the
field becomes more like that from a point charge.
Let us consider first the case where both sheets are insulators, the charge has been sprayed on. On
bringing the two sheets close, the charges will be unable to move, and the electric fields from the two
planes add, from the Principle of Superposition, giving:
E
Electric field from two uniform parallel planes of charge: one positive, one
negative, with equal charge densities σ : the field E = σ / ε 0 between the
planes, E = 0 outside the planes.
Actually, in practice, the sheets are usually conductors—in fact, almost all capacitors have this basic
structure.
To see how charges move around as conducting charged sheets are brought close, we’ll first look at the
charge distribution on a single isolated conducting sheet of finite thickness—the charges repel each
other, and form equal layers on the two sides:
6
E
Suppose we now take two such conducting planes with equal charge densities, but of opposite signs,
and put them close and parallel. What happens?
The positive and negative charges will attract each other, and move to be as close together as possible.
That is, all the charges will move to the inside surfaces of the conductors:
E
Note the charge density σ on the lower conductor’s top surface generates a field of strength σ / ε 0 .
This can be understood by considering a pillbox Gaussian surface which encloses that top surface (see
diagram): the Gaussian surface has field E from its top, but no electric field through its bottom, which is
inside the conductor, where E = 0.
7
The absence of an electric field inside the conductor means there can be no net charge contained in any
closed surface lying entirely inside the conducting material. If there are holes in the material inside this
surface, and charge is placed in these holes, charges will move inside the conductor to the surfaces of
these holes, to exactly compensate for the charges placed in the holes. Lines of force from these added
charges will terminate at the surfaces of the holes.