Lecture 2 Electrostatics and Radiation

1. Poisson equation in electrostatics

The solution can be obtained using Fourier transform:

in case that we know the charge distribution in the whole 3-D space. Unfortunately, in most electrostatic problems, we only know in a finite region and some boundary conditions of . For general shapes of boundary, solution can be extremely hard. Analytically solvable cases are mostly 3 types: box, sphere, and cylinder. 1.1 Box 1. (Problem 4 in Sec 2, 2011) Consider a box with side lengths a, b, and c along the x, y, and z axes. Suppose there is no electric charge in the box and that on the surface of the box except at z = 0 where everywhere in the box. , and at z = c where . Find

Since theres no charge inside the box, the inhomogeneous Poisson equation becomes the homogeneous Laplace equation

And the boundary conditions are

The boundary conditions on the - and - faces are homogeneous why on the -face they are inhomogeneous. We may use the separation of variables

Then, from the Laplace equation and the boundary conditions we have

and

which has the general solution

Therefore, the general solution of

is

Using

we get

and

by using the orthogonality of different eigenfunctions and are all zero. We may now find the coefficients specific solution:

. And all other coefficients to determine the

2. (Problem 5 in Sec 2, 2012) Consider an infinite pipe with a square cross section as drawn, with three sides grounded and one side at potential . Calculate the potential everywhere inside

the pipe.

In this problem, is independent of essentially two-dimensional. Using the separation of variables general solution

. The problem is therefore , we get the

To satisfy

, we require

for all

. This leads to

To determine the coefficients

, we write

Using the orthogonality of

, we get

so that

Since

is only nonzero for odd

, the final result can be simplified to

At , the two functions cancel and you recognize the familiar trigonometric expansion of a square wave, so you know everything works. Later in the question, you are asked to find when the boundary value is also nonzero. This seems to have complicated the problem a lot, since we no longer have a pair of homogeneous boundary conditions on either - or -direction. But we can actually solve each nonzero edge separately and add the two solutions at the end using the linearity of the system.

1.2 Sphere 1. (Problem 2 in Sec 2, 2012) A dielectric sphere of radius R is hollowed-out in the region and a thin, grounded, conducting shell inserted at r = s. The sphere is placed in a uniform, external Efield along the z-axis. The dielectric constant is . Calculate the potential in the region .

The potential have

has cylindrical symmetry. Inside the grounded conducting shell, we . Outside the conducting shell, we have the Laplace equation

for both

and

. But at

, due to the continuity condition of the normal , we have

component of the electric displacement

This will cause discontinuity of

at

, while the fact that

is still continuous at

implies that the continuity of the tangential component of the electric field is satisfied. Now we have to solve the Laplace equation using the separation of variables in spherical coordinates. We have

Plugging in Legendre polynomials of

, we find that and

are the

This is an Euler differential equation. By doing a change of variable coefficients become constant.

, the varying

The two linearly independent solutions are given by

The general solution of

is therefore

At

, we require that

. Therefore

At

, we require that

is continuous

and

Finally, by looking at the asymptotic behavior of

we have

. Only the . We find that

component is nonzero. We now have four equations for

The

outside the dielectric sphere is then given by

2. (Problem 2 in Sec 2, 2013) A linear uniform charge distribution of extends along the -axis from to . Find a series expansion for of Legendre polynomials valid to all orders, and for .

coulomb/meter in terms

The solution is a straightforward integral using the Coulomb potential

Now we want to do an expansion of in the region of do this, we need the generating function of the Legendre polynomials:

to all orders of

. To

In fact, other than the first few orders of Legendre polynomials, and how they appear in our Poisson equation, theres a lot more you need to review to prepare for the quals the recursive relations, the series expansions, Rodriguez formula, orthogonality relations, etc. Fortunately, you can take a cheat sheet that has everything on it. Once you remember the generating function, you get the multi-pole expansion formula below.

Youll find that this 1.3 Cylinder

also has the form of the general solution we derived previously.

1. (Problem 3 in Sec 2, 2010) Consider an infinitely long, grounded conducting cylinder, of radius a, which is introduced into a uniform electric field . The axis of the cylinder is perpendicular to . (a) Find an expression for the external potential after insertion of the cylinder; (b) Find an expression for the surface charge induced on the cylinder.

(a) The problem involves solving the Laplace equation

in the region or

. Using the separation of variables and

, we get

In case that

, we have

The general solution is therefore

Defining

as the direction of

, we have the asymptotic behavior

Therefore , we have

. The only non-zero terms will therefore be . This gives us , so that

and

. And at

(b) The surface charge induced can be found by Gausss law:

Positive charge gets propelled to the side while negative charge gets pulled to the side. This makes sense. Not every cylinder involves Bessel functions. Dont get scared. This problem is special because has no -dependence. Bessel functions will typically not appear on quals.

2. Radiation using multi-pole expansion

After introducing the electromagnetic potentials

and

, the fields are expressed as

Using the Lorentz gauge condition

the potentials

and

satisfy dAlembert wave equations as given by

Radiation theory will focus on

. The retarded solution of dAlembert equation is

In case that

(the radiation zone), we may do two approximations:

and

where and is a unit vector along the direction of . Under these two approximations, the vector potential in the radiation zone is given by

and the E and B fields and the energy current are

We can then calculate the fields and the power distribution from these.

2.2 Electric dipole radiation If the typical length-scale of the source is much smaller than the wave length of the radiation, we can do a Taylor series expansion on and have

This simplifies the retardation to of the current distribution

and all we need is find the

-th order moment

. In most radiation problems in the quals, we only need to term as given by

look at the leading order term in the above expansion, which is the

From charge conservation we have

where we have used the formulas in vector calculus

and that the total divergence term integrated over all space is zero because we have

following from our locality assumption of the source. Define the electric dipole moment of the source as

so that the leading order term of the vector potential

Since only differs from by a constant, we have potential under the electric dipole approximation is finally given by

. Therefore, the vector

From this, the E and B fields are found to be

and

Finally, the radiation power distribution is given by

Next order of expansion gives us the magnetic dipole and electric quadruple terms, and when it comes to the power radiated, there can be cross terms between them as well. These, however, will be covered in Prof Blaer s E&M class, but will not appear in the quals. 2.3 Sample problems 1. (Problem 1 in Sec 2, 2012) Two point particles (each having the same electric charge +e) travel in the x-y plane around the circumference of a circle (having radius R). Both charges travel at the same angular velocity but maintain a fixed angular separation throughout the motion. Assume that the motion of the particles is non-relativistic ( ).

The particles radiate energy at distance r far from the circular orbit. Find the time-averaged power radiated per solid angle in the direction in the radiation zone (r>>R). The condition enables us to do a multi-pole expansion in the radiation zone . So long as the angular separation of the two particles isnt 180 degrees, theres going to be change in the electric dipole moment of the system as the particles move. Therefore, an electric dipole approximation will be enough to solve the problem. We have

and this dipole moment

rotates about the origin at angular frequency

. Therefore,

and

also rotates about the origin at angular frequency . The component of that is parallel to can be found to be

where is the angle between and the projection of generates no radiation power in the direction. Therefore

in the x-y plane. This component

where we have used

. And this gives us the time-averaged power

2. (Problem 1 in Sec 2, 2013) Consider two very tiny oscillating electric dipoles which are located on the z-axis as shown in the diagram.

The first dipole is located at

and has a time-varying electric dipole moment given by

The second dipole is located at

and has a time-varying electric dipole moment given by

The two dipoles therefore oscillate in the same -direction with the same amplitude and the same angular frequency but are out of phase by phase angle . Consider a field at point in the radiation zone ( ) and keep only the leading terms. (a) Find the electric field at in the radiation zone. (b) Find the time-averaged power detected per unit solid angle ( direction in the radiation zone. (a) The vector potential in the radiation zone is given by ) in the

where retarded times need to consider

and and

are the positions of the two dipoles. There are two different because the two dipoles are located at different positions. And we or not. But we do know

because we are not told if

since we are in the radiation zone. Therefore,

Plugging these in, we get

The E and B fields are therefore

where we have used

and

(b) The energy flux is given by the Poynting vector

Therefore, the time-averaged angular distribution of the power radiated is

where

when averaged over a long time.