|PHY 514||Electromagnetic Field Equations and Radiation||Spring 2012|
|This page: Home ⇒ Week 02|
The problem of radiation from a point charge is very clearly described in Feynman Lectures I: 28-1,2.
The goal this week is to derive and understand the Heaviside-Feynman formulas for the electromagnetic fields of a point charge and the Liénard-Wiechert potentials: Jackson chapter 6 sections 3,4,5, chapter 12 section 11, and chapter 14 section 1. Landau-Lifshitz (Fields) chapter 8, sections 62,63. Misner-Thorne-Wheeler chapter 3 section 4, chapter 4 section 4. Panofsky-Phillips chapter 19, section 1, chapter 20, section 1. Schwartz chapter 6 sections 1,2,3.
The article by J.J. Monaghan, The Heaviside-Feynman expression for the fields of an accelerated dipole, J. Phys. A: 1, 112, 1968 extends the Heaviside-Feynman expressions to the case of a point object with electric and magnetic dipole moments.
Coulomb's Law for the electric field holds exactly only for static charges. It implies instantaneous action at a distance and does not take into account the finite speed of light.
The inverse square law of gravity was discovered by Newton, but he was never comfortable with it.
That one body may act upon another at a distance through a vacuum without the mediation of anything else, by and through which their action and force may be conveyed from one another, is to me so great an absurdity that, I believe, no man who has in philosophic matters a competent faculty of thinking could ever fall into it. — Isaac Newton, third letter to Richard Bentley, 1692.
The fundamental problem of classical electrodynamics is to derive the exact electromagnetic field due to a point charge moving with arbitrary acceleration. The electromagnetic field due to a system of charges with given accelerations can then be derived by linear superposition.
The concept of a potential field was first introduced by Lagrange (in a 1773 article on the secular equation of the Moon) to explain gravitational action at a distance. Coulomb's Law is obtained by solving Poisson's equation for the electrostatic potential.
The fully relativistic analog of Poisson's equation in electrostatics is
Subscript 4-vector indices denote covariant components. I write 4-vectors with the time component last. The contravariant components of the space-time point 4-vector are written
The covariant time component has the opposite sign
Einstein's summation convention is used for repeated indices, one upper and the other lower:
The covariant gradient is
Note the superscript in the denominator!
The d'Alembertian wave operator is
The contravariant and covariant photon field components are
with SI units volts.
The contravariant and covariant current density components are
with SI units .
The decoupled form of the electromagnetic wave equation assumes the Lorenz gauge fixing condition
Note that the every equation has exactly the same form as the nonrelativistic equations in Jackson Chapter 6.
Note also that I use different units and conventions compared with Jackson Chapters 11-16. The time component is written following the space components, and the metric tensor in Minkowski space
which is used to raise and lower indices
has the opposite sign from Jackson. SI units are used consistently in my notes. These are just conventions. Choose the convention you are most comfortable with and use it consistently.
A Green's function is a particular solution of a linear partial differential equation with a point source and a choice of boundary conditions.
In relativistic physics information can only be transmitted forward in time and it cannot travel faster than the speed of light. This requirement is called causality.
The causal Green's function for the d'Alembertian operator is a particular solution of
The right hand side represents a unit source at the origin of spacetime. Causal boundary conditions require that the solution is zero at every spacetime point outside of the forward light cone with vertex at the origin of spacetime.
The Green's function equation is easily solved using Fourier integral transforms:
The solution in space is
This solution is singular for Fourier modes that propagate with the speed of light!
Singularities in mathematics need to be defined using a limiting process. Different limiting procedures will in general give different solutions. The physical solution that we require is selected by specifying causal boundary conditions.
Singularities in differential equations can be defined by analytic continuation in one or more integration variables. Consider the integral
The denominator of the integrand has two simple zeros on the axis at
The integral can be evaluted using the Methods of contour integration
The integral will vanish exponentially as if is given a small positive imaginary part. This is the behavior we need for forward propagation in time. In fact, this procedure gives causal propagation.
For any the integrand vanishes exponentially on the infinite semicircle with . The integration contour can be closed in the upper half plane, which excludes the two poles.
For the integration contour is closed in the lower half plane, which includes the two poles, and can be evaluated using the method of residues.
The causal Green's function is
The Green's function for an instantaneous source at any other spacetime point is
The causal Green's function solves the electromagnetic wave equation
with physically correct causal boundary conditions.
The observer measures the photon field at spacetime point . The effect is calculated by integrating the current density over all spacetime points , in the past and future and at every location in space.
The mass and electric charge of a particle are constants that do not depend on the reference frame of the observer. Mass and charge are relativistically invariant.
The spacetime position of the particle is a 4-vector. A spacetime displacement is also a 4-vector
The time interval is not and invariant: it depends on the frame of the observer. This means that the nonrelativistic velocity
is not the space part of a 4-vector, so it does not transform in a simple way.
To obtain the relativistic 3-vector velocity we must divide the 3-vector displacement by an invariant time interval called the proper time interval
Proper time is just the time measured in the rest frame of the particle.
The correct 4-vector velocity of the particle is
The inner product of with itself is
and is the same in every reference frame.
Multiplying 4-velocity by invariant mass gives the correct relativistic 4-momentum
The electric charge of a particle is a constant and has the same value in any reference frame.
Charge density at a point in space is defined by finding the net charge inside a small spatial volume containing the point and taking the limit
This is not Lorentz invariant because is the same in every reference frame but is not.
The spatial current density is
where is the velocity of in the limit .
We saw in defining four velocity that is not the spatial part of the relativistic velocity. Rather amazingly, the product of with (which is not an invariant, is the spatial part of a relativistic 4-vector!
To show this, consider a small time interval during which the charge in moves . The quantity
is a relativistic 4-vector, and
is relativistically invariant: under a Lorentz transformation contracts in the boost direction, and dilates by the same factor!
For a system of point particles with charges and positions
The relativistic 4-vector current density of charge is
This can be put in manifestly relativistic form using the relativistic 4-velocity and proper time of the charge .
The time in the observer's frame and the proper time in the particle's rest frame both increase continuously but at different rates. Let be the function the particle would use to compute the time in the observer's frame.
To prove the last equality use the inverse function
which also implies
This gives the manifestly relativistic formula
The causal Green's function solution for charge is
The subscript means that the proper time is evaluated using the causality constraint
To solve this constraint find the instant on the world line of the charge such that the forward light cone at this instant intersects the spacetime position at which the observer measures the field.
The electric field formula in Feynman's Lectures on Physics Chapter 28:
The first term is the Coulomb field evaluated with the distance to the charge determined by causality.
The second term is a correction to the Coulomb field which falls off more rapidly than the square of the distance.
The third term represents radiation, and falls off inversely as the distance.
J.J. Thomson gave a good physical derivation of the radiation field in his 1907 book "Mathematical Theory of Electricity and Magnetism". Figure 4.6 below from Misner-Thorne-Wheeler illustrates his explanation.
The basic concept is animated in the Mathematica demo Radiation Pulse from an Accelerated Point Charge
Please format your submission as a single PDF file according to the instructions in the syllabus and submit it before 11:59 pm on Sunday, February 5.
Problem 1: Starting from the Liénard-Wiechert potential
derive Feynman's expression
Problem 2: Understand the definitions and units used in Jackson's expression
and show that it is equivalent to Feynman's expression in Problem 1.
Problem 3: Starting from the Liénard-Wiechert potential in Problem 1, derive the Heaviside expression for the magnetic field given in Jackson
Did you need to change units or make any non-relativistic approximations?
Problem 4: Understand and summarize the derivation of the dipole potentials in section 3 of the article by J.J. Monaghan, The Heaviside-Feynman expression for the fields of an accelerated dipole, J. Phys. A: 1, 112, 1968. Write Monaghan's expressions for the potential using SI units and relativistic 4-vector notation as in the Liénard-Wiechert potential expression in Problem 1.
Problem 5: Download and experiment with the Mathematica Demonstration Radiation Pulse from an Accelerated Point Charge and explain how it relates to the Liénard-Wiechert potential. Read the discussion of the dipole radiator in Feynman Lectures I: 28-2 and use Mathematica to plot the angular dependence of the radiation in the , and planes, as is done in the Dipole Antenna Radiation Pattern Mathematica Demonstration.