MSci EXAMINATION

PHY-966(4261) Electromagnetic Theory

Time Allowed: 2 hours 30 minutes

Date: May 2008

Time: hh:mm

Course Organiser: Prof WJ Spence

Deputy CO: Dr O Soloviev

Instructions: Answer THREE QUESTIONS only. Each ques-

tion carries 20 marks. An indicative marking-
scheme is shown in square brackets [ ] after each
part of a question. A formula sheet is provided at
the end of the examination paper.

YOU ARE NOT PERMITTED TO START READING THIS QUES-

TION PAPER UNTIL INSTRUCTED TO DO SO BY AN INVIG-
ILATOR

c Queen Mary, University of London 2008

!
This page should be blank
 1. Maxwell’s equations in linear media are
∂D ∂B
∇ · D = ρ, ∇×H=J+ , ∇ · B = 0, ∇×E=− .
∂t ∂t
(i) Consider a region of space V bounded by a closed surface S, and
also let C be a closed contour in space with an open surface S !
spanning the contour. Explaining the notation used, derive from
the above equations the integral forms
∂D
! ! " !
D · dS = ρ dV, H · dl = (J + ) · dS!
S V C S! ∂t
∂B
! " !
B · dS = 0, E · dl = − · dS! . [6 marks]
S C S ! ∂t

(ii) Consider two regions, labelled by i = 1, 2, containing different

linear media, which meet at an infinite two-dimensional boundary,
with unit normal n to the boundary. Let Ei , Di , Bi , Hi for i = 1, 2
label the electromagnetic fields in the two regions.
Using a suitable small, shallow cylinder, straddling the boundary
between the two regions, with surface charge density σ, derive the
boundary conditions
(D2 − D1 ) · n = σ, (B2 − B1 ) · n = 0,
from two of the integral equations above.
Now considering a suitable small rectangle straddling the bound-
ary, with current density K on the surface of the rectangle, derive
the further boundary conditions
n × (H2 − H1 ) = K, n × (E2 − E1 ) = 0,
[8 marks]
(iii) Consider incident, refracted and reflected waves at this matter
interface, with
! !! ·x)
Einc = E0 e−i(ωt−k·x) , Eref r = E!0 e−i(ωt−k ·x) , Eref l = E!!0 e−i(ωt−k .
Assume that the matter interface is at z = 0, and that the incident
wave has electric field parallel to the z − x plane. Let the angles of
incidence, refraction and reflection be θ, θ! , θ!! respectively. Show
that the boundary conditions on the fields E at the interface imply
that
!! ! !
−E0 cos θeikx sin θ + E0!! cos θ!! eikx sin θ = −E0! cos θ! eik x sin θ
must be true for all x. Show that this implies that θ = θ!! (the
law of reflection), and k sin θ = k ! sin θ! (Snell’s law). [6 marks]

please turn to the next page

c Queen Mary, University of London 2008

! Page 1 of 6
 2. For an oscillating electric dipole with strength p, oscillating in time as
e−iωt , the vector potential is given by
1 eikr ik
Ae.d. (r, t) = − p e−iωt ,
4π& r c
#
where r = (x, y, z), r = x2 + y 2 + z 2 and k = ω/c.

(i) Show that in the far zone, where kr >> 1, this results in the
magnetic field
1 k2 eikr −iωt
Be.d. (p) = n×p e
4π& c r
where n = 1r r. [4 marks]
(ii) For the electric field Ee.d. (p), use the source-free Maxwell equation
Ė = c2 ∇ × B and the fact that the time dependence of the fields
is e−iωt to deduce that
ic
Ee.d. (p) = ∇ × Be.d. (p) e−iωt
k
and hence that in the far zone
Ee.d. (p) = c Be.d. (p) × n .
[5 marks]
(iii) The vector potential for an oscillating magnetic dipole is given by
µ0 eikr
Am.d. (r, t) = ik n × m e−iωt .
4π r
Show that this is proportional to the magnetic field for the electric
dipole, with p replaced by m:
i
Am.d. = Be.d. (p → m).
kc
Thus prove that the electric and magnetic fields for a magnetic
dipole are given by
1 e.d.
Bm.d. (m) = E (p → m),
c2
Em.d. (m) = − Be.d. (p → m).
[8 marks]
(iv) How are the polarisation vectors, the directions of the magnetic
fields, and the directions of the radiation n oriented with respect
to each other in the two cases of electric and magnetic dipole
radiation in the far zone ?
[3 marks]

please turn to the next page

c Queen Mary, University of London 2008

! Page 2 of 6
 3. (i) Show in the Lorentz gauge (∂ µ Aµ = 0), with Aµ = ( 1c Φ, A) and
j µ = (cρ, J), that the Maxwell equation ∂ µ Fµν = µ0 jν reduces to

1
∂ µ ∂µ A = µ0 J, ∂ µ ∂µ Φ = ρ.
&0
[3 marks]
$∞
(ii) Integrate the equation for A above with −∞
e−iωt to obtain the
Fourier transformed equation

(∇2 + k 2 )A(x, ω) = −µ0 J(x, ω), (1)

with k 2 = ω 2 /c2 . [4 marks]

(iii) Suppose that there exists a Green function Gk (x, x! ), satisfying

(∇2 + k 2 )Gk (x, x! ) = −4πδ 3 (x − x! ). (2)

Show that
µ0
!
A(x, ω) = Gk (x, x! )J(x! , ω)d3 x!
4π

solves equation (1) above. [3 marks]

(iv) Give an argument why Gk (x, x! ) must be purely a function of

r = |r| = |x − x! |. Show that in this case equation (2) becomes

1 d2
(rGk (r)) + k 2 Gk (r) = −4πδ 3 (r)
r dr2
and hence that when r &= 0, Gk (r) is given by
1
Gk (r) = (Aeikr + Be−ikr ), (3)
r
for some constants A, B. [5 marks]
ρ(r !
(v) A solution of Poisson’s equation ∇2 φ = − $10 ρ is φ = 4π$
1
d3 r! .
$
|r−r! |
Use this fact to show that when r → 0, (3) above remains a solu-
tion of equation (2) if
A + B = 1.
[5 marks]

please turn to the next page

c Queen Mary, University of London 2008

! Page 3 of 6
 4. Consider the Maxwell equations in a vacuum with sources -
∂B
∇ · B = 0, ∇ × E = − ,
∂t
1 ∂E
∇·E= ρ, ∇ × B = µ0 &0 + µ0 J.
&0 ∂t
(i) Show that the first two of these equations may be solved by intro-
ducing the potentials A and Φ, and writing
∂A
B = ∇ × A, E = −∇Φ − .
∂t
Show that the other two Maxwell equations then become

∂ 1
∇2 Φ +
(∇ · A) = − ρ,
∂t &0
1 ∂ 2A 1 ∂Φ
∇2 A − 2 2 − ∇(∇ · A + 2 ) = −µ0 J.
c ∂t c ∂t
[6 marks]
(ii) Show that the definitions of the potentials are unchanged if we
make the gauge transformations
∂Λ
A → A + ∇Λ, Φ → Φ −
∂t
for any function Λ. [2 marks]
(iii) In Lorentz covariant notation, Maxwell’s equations above may be
written

∂µ Fνρ + ∂ν Fρµ + ∂ρ Fµν = 0, ∂ µ Fµν = −µ0 jν .

Show that the first of these equations is solved by writing

Fµν = ∂µ Aν − ∂ν Aµ .

Write down the gauge transformations on Aµ and show that they

leave Fµν invariant. [4 marks]
(iv) Consider Maxwell’s equations in the Coulomb gauge ∇ · A = 0.
Show that the equation for A can be written
1 ∂2A
∇2 A − = −µJt ,
c2 ∂t2
where Jt is transverse (∇ · Jt = 0). You may use the result that
∇2 |x!1−x| = −4πδ 3 (x! − x) and the identities ∇2 J = ∇∇ · J − ∇ ×
∇ × J, and J(x) = δ 3 (x! − x)J(x! ) d3 x! for any vector field J.
$

[8 marks]

please turn to the next page

c Queen Mary, University of London 2008

! Page 4 of 6
 5. The electric and magnetic fields generated by a charged particle moving
with velocity cβ and acceleration cβ̇ are given by the Lienard-Wiechert
expressions
1
B = [n × E]ret ,
c
q % n−β & q 1 % n × [(n − β) × β̇] &
E= + ,
4π&0 γu2 R2 (1 − β · n)3 ret 4π&0 c (1 − β · n)3 R ret
where n is the unit vector which points from the point on the particle
trajectory to the field point x, with x = r(τ0 ) = nR, and the retarded
time is tret = t − R/c.

(i) Show that for the case when the acceleration is parallel to the
velocity, the electric field far from the charge is given by
q 1 % n × [n × β̇] &
E= .
4π&0 c (1 − β · n)3 R ret
[2 marks]
1
(ii) Show that in this case, the Poynting vector S = µ0
E × B, far from
the charge is
q2 1 % β̇ sin θ &2
S= n,
4π&0 4πc (1 − β · n)3 R ret
where θ is the angle between n and the common direction of the
velocity and acceleration of the particle. [4 marks]
(iii) Show that the power radiated per unit solid angle is given by
dP (t! ) dt
= R2 n · S ! .
dΩ dt
and hence equals
dP (t! ) q2 1 β̇ 2 sin2 θ
= .
dΩ 4π&0 4πc (1 − β cos θ)5
[6 marks]
(iv) For non-relativistic motion, deduce from this the Larmor formula
dP q2 1 2 2
= u̇ sin θ.
dΩ 4π&0 4πc3
[2 marks]
(v) Without making the non-relativistic approximation, show that the
maximum intensity of radiation is observed at the angle
% 1 #
θmax = cos−1 ( 1 + 15β 2 − 1).
3β
[6 marks]

please turn to the next page

c Queen Mary, University of London 2008

! Page 5 of 6
 Formula Sheet

a × (b × c) =(a · c)b − (a · b)c,

∇ · (ψa) = a · ∇ψ + ψ∇ · a,
∇ × (ψa) = (∇ψ) × a + ψ(∇ × a),
∇'× (∇(× a) = ∇(∇ · a) − ∇2 a,
∇ ψ(r) = nψ ! (r).

Maxwell’s equations:
∇ · B = 0, ∇ × E = − ∂B
∂t
;
∇ · D = ρ, ∇ × H = J + ∂D
∂t
.

F = q(E + v × B).
∇ · J + ρ̇ = 0.
For linear isotropic media:
1 1
D = &E = &0 E + P, H= B = B − M.
µ µ0
c2 dτ 2 = c2 dt2 − dx2 − dy 2 − dz 2 = dxα ηαβ dxβ .
)+1 if α = β = 0
ηαβ = −1 if α = β = 1, 2, 3
0 if α &= β

∂ *1 ∂ +
µ
*1 ∂ +
∂µ = = ,∇ , ∂ = , −∇ .
∂xµ c ∂t c ∂t
∂α F αβ = ∂α ∂ α Aβ − ∂ β ∂α Aα = µ0 j β ; F αβ = ∂ α Aβ − ∂ β Aα .
∂α Fβγ + ∂β Fγα + ∂γ Fαβ = 0.
−E 1 /c −E 2 /c −E 3 /c
 
0
E 1 /c 0 −B 3 B2 
'F αβ ' = 
E 2 /c
.
B3 0 −B 1 
E 3 /c −B 2 B1 0

In spherical coordinates (r, θ, φ), for a scalar field G(r, θ, φ),

1 ∂2 ∂2G
2 3
2 1 ∂ ∂G 1
∇ G= (rG) + 2 sin θ + 2 2 .
r ∂r2 r sin θ ∂θ ∂θ r sin θ ∂φ2
$ ρ(r! 3 !
A solution of Poisson’s equation ∇2 φ = − $10 ρ is φ = 4π$1
|r−r! |
d r.

End of Examination Paper Prof WJ Spence

c Queen Mary, University of London 2008

! Page 6 of 6