Classical Electrodynamics

Proposed Exercises 2024/2025, Técnico Lisboa

Frederico Fiuza, Zakary Buras-Stubbs & Pablo San Miguel

Week 4 - September 30 to October 4

Radiation and special relativity

! → important
* → challenging
Problem 1. Cherenkov Consider a dispersive dielectric medium, with a relative permati-
vity given by,
a
ϵr (ω) = 1 + 2
ω0 − ω 2
in a frequency domain for which |ω02 − ω 2 | ≫ 2βω, where ω0 is the characteristic frequency
fe e 2
of the medium, β is the coefficient of friction, and a = namϵ 0
≪ ω02 is a constant.
Determine the range of frequencies of cherenkov radiation emitted by an electron travelling
through the medium with velocity v = 2c
2
A:. Cherenkov Radiation is emitted when ϵr > vc 2 is satisfied, hence in this case one has,
a
1+ 2 > 4,
ω0 − ω 2
and therefore, r
a
ω > ω0 1 − 2
3ω0
Using that a ≪ ω0 ,
a
ω > ω0 (1 − )
6ω02

Separately, as |ω02 − ω 2 | ≫ 2βω, we find

s the following conditions,
2βω βω
ω ≪ ω0 1 − 2 ≈ ω0 (1 − 2 )
ω0 ω0
or s
2βω βω
ω ≫ ω0 1 + 2 ≈ ω0 (1 + 2 )
ω0 ω0
To ensure naturalness (positive frequencies) we require that ω < ω0 . Using these and the
range we found previously, we may determine that
a βω
ω0 (1 − 2 ) < ω ≪ ω0 (1 − 2 )
6ω0 ω0
Generation of Electromagnetic Waves

1
Problem 2. Field energy of a moving particle Calculate the Poynting vector and energy
density of the electromagnetic field of a particle of charge q moving with constant velocity.
Show that:
2.a) The field energy is carried along with the particle.
Hint: Use cylindrical coordinates. For the field energy to be carried along with the
particle, it should have the velocity v of the particle, so that space x and time t
variables should always appear in the combination z − vt.
A:. Particle velocity vẑ, particle position ω = vtẑ, observation point r = sŝ + z ẑ.
This leads to
qc (c2 − v 2 )[(s)ŝ + (z − vt)ẑ]
E= , (1)
4πϵ0 [s2 (c2 − v 2 ) + c2 (z − vt)2 ]3/2
q (c2 − v 2 )vcs
B= ϕ̂, (2)
4πϵ0 c2 [s2 (c2 − v 2 ) + c2 (z − vt)2 ]3/2
2
(c2 − v 2 )2

q vs
S= [(s)ẑ − (z − vt)ŝ], (3)
4πϵ0 µ0 [s2 (c2 − v 2 ) + c2 (z − vt)2 ]3
2 2 2
s (c + v 2 ) + c2 (z − vt)2 2

ϵ0 q
u= (c − v 2 )2 . (4)
2 4πϵ0 [s2 (c2 − v 2 ) + c2 (z − vt)2 ]3
The variables z and t always appear as z − vt.
2.b) The electric and magnetic fields measured at a distance d away from an infinite
straight wire carrying a uniform line charge density λ moving at a constant speed v
down the wire reduces to the same fields as electrostatics and magnetostatics.
A:. The distribution can be considered an integral of charges dq = λdz along the
wire ˆ ∞
λ 1 − v 2 /c2 R̂
E= 2 dz. (5)
4πϵ0 −∞ (1 − v sin (θ)/c ) R2
2 2 3/2

By symmetry, horizontal components present in this system. The vertical componet

of R is simply given by sin θ, and hence the height "above"the wire is α = R sin θ,
and hence
1 sin2 θ z α
= − = cot θ =⇒ dz = dθ
R 2 d 2 α sin2 θ
1 α sin2 θ dθ
2
dz = 2 2
dθ =
R sin θ α α
ˆ π
λ 1
E= (1 − v 2 /c2 )r̂ 2 2 2 3/2
sin θdθ. (6)
4πϵ0 α 0 (1 − v sin (θ)/c )
2 2
ˆ 1 κ = cos θ and sin θ = 1 − κ , 3we
Substituting ˆ have
1
dκ c dκ 2 c2
C 2 2 2 2 2 3/2
= C a = −1 (7)
−1 (1 − v /c + v /c κ ) v 3 −1 (a2 + κ2 )3/2 v2
Substituting (thats right again!) κ = a tan ˆζ, dκ = a sec2 ζdζ we ˆsee that
3 ˆ
c a sec2 ζdζ c3 dζ c3
C 3 = C = C cos ζdζ (8)
v a3 (1 + tan2 ζ)3/2 v3 a2 (sec2 )3/2 v3
and so noting that sin θ = √κ2κ+a2 we find
1
c3

κ
E=C 3 √ (9)
v κ2 + a2 −1

2
 from which we then find
2λ
E= r̂, (10)
4πϵ0 α
v×E µ0 λv µ0 I
B= 2
= ϕ̂ = ϕ̂, (11)
c 2πα 2πα
the electrostatics and magnetostatics results.

Problem 3. Waves from an oscillating electric dipole

The leading terms in the radiation multipole expansion for the electric and magnetic fields
are
µ0 µ0
E= r̂ × [r̂ × p̈(t0 )], B = − r̂ × p̈, (12)
4πr 4πcr
where r̂ = r/r is the position unit vector and t0 = t − r/c the retarded time at the origin.
3.a) Consider the case of two point charges, +q and ˘q located on the z-axis, symmetrically
about the origin. The separation of the charges is given by acos(ωt). Find the
resulting E and B fields in spherical coordinates.

A:. Remember that the dipole moment is just the product of the charge times the
distance between the charges. The result thus comes from the equations of the fields
above.
3.b) Check whether the fields that you found satisfy Maxwell’s equations in vacuum, and
find the time-averaged Poynting vector S. Make a plot of the magnitude of the time-
averaged Poynting vector as a function of azimuthal angle ϕ for θ = π/2, as well as
a polar plot of the magnitude as a function of θ for ϕ = 0.

Problem 4. Dipole radiation from Larmor formula

The average power radiated by an oscillating dipole, with dipole moment p(t) = p0 cos(ωt)
is
µ0 p20 ω 4
⟨P ⟩ = . (13)
12πc
Derive this result from the Larmor formula
µ 0 q 2 a2
P = , (14)
6πc
by treating the dipole as an oscillating pair of charges ±q0 that oscillate 180 degrees out of
phase with an amplitude of oscillation d/2 (note that p0 = q0 d). Hint: The waves radiated
by +q0 and −q0 interfere, so in the Larmor formula you must add the qa’s and then square,
rather than adding the squares of the qa’s.

Problem 5. Radiation from an incoming charge [!]

A positive charge q is fired head-on at a distant positive charge Q (which is held stationary),
with an initial velocity v0 . It comes in, decelerates to v = 0, and returns out to infinity.
What fraction of its initial energy (mv 2 /2) is radiated away? Assume v0 ≪ c and that you
can safely ignore the effect of radiative losses on the motion of the particle.
A:.
dWr µ0 q 2 a2 qQ
= a= (15)
dt 6πc 4πϵ0 r2

3
 ˆ ∞ 2
µ0 q 2

qQ 1 2k
Wr = 2 dr r0 = (16)
6πc r0 4πϵ0 r4 v02
´ ´
Notice that the factor of 2 comes from the fact that we are integrating dt = dx/v first
from ∞ to r0 with a negative v and then from r0 to ∞ with a positive v.
8qmv02 Wr 16q  v0 3
Wr = = (17)
45c3 Q W0 45Q c
Radiation Reaction

Problem 6. Simple motion of a charged particle [!]

For the following cases, calculate the radiation reaction force and what is the power radiated.
6.a) Motion in a circle of radius R at a constant speed v. Given that, a centripetal
force mv 2 /R must be provided, what additional force Fe must you exert in order to
counteract the radiation reaction? What power does this extra force deliver?

A:.
r(t) = R(cos(ωt)x̂ + sin(ωt)ŷ)
µ0 q 2 2 µ0 q 2 2 2 µ 0 q 2 a2
Fe = ω v Pe = ω v = Prad =
6πc 6πc 6πc
6.b) Motion in a simple harmonic oscillator with amplitude A and angular frequency ω.
Explain any discrepancy that arises.

A:.
r(t) = A(cos(ωt)ẑ
2
µ0 q 2 µ0 q 2 2 2 µ 0 q 2 a2
Fe = ω v Pe = ω v ̸= Prad =
6πc 6πc 6πc
However
⟨Pe ⟩ = ⟨Prad ⟩
6.c) Free fall at constant acceleration g. Comment.

A:.
1
r(t) = gt2
2
µ0 q 2 g 2
Fe = 0 Pe = 0 ̸= Prad =
6πc
This is a paradox as our charge radiates infinitely but no power is required to sustain
it. To resolve this paradox, we first note that the reference frame in free fall is
not an inertial frame. Therefore, we cannot use Galilean/Lorentz invariance nor
the principles of special relativity. It is necessary to use coordinate transformations
that include acceleration and, by doing so, we find that the charge radiates in both a
stationary reference frame and in a moving frame with the charge. For more details on
this issue (which had extensive discussion throughout the 20th century), see Rohrlich,
Fritz (1964). Classical Charged Particle. Reading, Mass.: Addison-Wesley.

4
Problem 7. Classical tunneling through a barrier [*]
A charged particle, travelling from along the x axis, encounters a rectangular potential energy
barrier of the form (
U0 , if 0 < x < L,
U (x) = (18)
0, otherwise.
Show that, due to the radiation reaction, it is possible for a particle with kinetic energy less
than U0 to pass through, there is, the particle can tunnel through the barrier. Show that
the final velocity vf is related to the time T spent traversing the barrier by the equation

U0
L = vf T − (τ e−T /τ + T − τ ). (19)
mvf

and the initial velocity (at x = −∞) is

 
U0  1
vi = vf − 1− U0
. (20)
mvf 1+ mvf2
(e−T /τ − 1)

Hint: to solve the equation a = τ ȧ + F/m where F (x) = U0 [−δ(x) + δ(x − L)], consider
the three regions (before, during, after the barrier) separately, finding the general solution
for a, v, x and imposing the appropriate boundary conditions. Extra hint: you can simplify
the results for a specific case, such that the final kinetic energy is half the barrier height and
L = vf τ /4 leading to vi = 4vf /3 so that the initial kinetic energy is 8U0 /9 and the particle
makes it through, even if it didn’t have enough energy to get over the barrier.

Problem 8. Imagine the case of a time-dependent electric field in a case where ϕ = 0 and
A ̸= 0. In this case E = −∂A/∂t. Is the electric field conservative in this case?
A:. By Faraday’s law, the line integral of the electric field is the time derivative of the
magnetic field flux through that surface. As the case of a time varying magnetic vector
potential A leads to a time varying magnetic field B, then its flux will also depend on time.
Therefore the line integral of E won’t be zero and the field will not be conservative.

Problem 9. Enumerate all the different letters and numbers in the following electric and
magnetic fields and explain their physical significance.
9.a) Fields produced by a moving charge q:

q R  2 2
 R̂ × E
E= (c − v )u + R × (u × a) , B = . (21)
4ϵ0 (R · u)3 c

9.b) Fields produced by a nonrelativistic oscillating charge q near the origin:

qr̂ µ0 pω 2 sin(θ) cos(ω(t − r/c))θ̂

E= − , (22)
4πϵ0 r2 4πr
−µ0 pω sin(θ) sin(ω(t − r/c))p̂hi µ0 pω 2 sin(θ) cos(ω(t − r/c))ϕ̂
B= − . (23)
4πr2 4πrc