WiSe 2012 29.10.

2012

Prof. Dr. A.-S. Smith

Dipl.-Phys. Ellen Fischermeier
Dipl.-Phys. Matthias Saba
am Lehrstuhl für Theoretische Physik I
Department für Physik
Friedrich-Alexander-Universität
Erlangen-Nürnberg

Theoretische Physik 2: Elektrodynamik

(Prof. A.-S. Smith)

Home assignment 3

Problem 3.1 Dirac delta function

Using Dirac delta function in the appropriate coordinates, express the following charge distributions
as three-dimensional charge densities ρe (~x).
a) In spherical coordinates, a charge Q uniformly distributed over a spherical shell of radius R.
b) In cylindrical coordinates, a charge λ per unit length uniformly ditributed over a cylindrical
surface of radius b.
c) In cylindrical coordinates, a charge Q spread uniformly over a flat circular disc of negligible
thickness and radius R.
d) The same as part (c), but using spherical coordinates.

Problem 3.2 Properties of the Dirac delta function

The function h(x) has only one simple root x0 . Explain the relation
1
δ(h(x)) = δ(x − x0 ).
|h0 (x 0 )|

Prove the following properties of the δ-function:

a) xδ(x) = 0

b) ϕ(x)δ(x − a) = ϕ(a)δ(x − a)
Z
c) dy δ(x − y)δ(y − z) = δ(x − z)
Problem 3.3 Legendre polynomials and spherical harmonics
You have been told about the mathematic properties of Legendre polynomials [spherical harmonics] as
angular solutions of the Laplace equation in spherical coordinates in the presence [absence] of circular
symmetry. To obtain an intuitive understanding of these functions, go to the web pages demonstra-
tions.wolfram.com/PlotsOfLegendrePolynomials and demonstrations.wolfram.com/ComplexSphericalHarmonics
and download the demonstration in *.cdf format. The CDF Player necessary to use the demonstrations
is freely available at www.wolfram.com/cdf. If you have a Mathematica licence, you can also download
the source code of the respective demonstration in *.nb format.
a) Use the Legendre polynomial plotter to convince yourself that
1. the polynomials Pl with even [odd] l are symmetric [antisymmetric] with respect to the
origin.
2. Pl (1) = 1 and Pl (x) < 1 for |x| < 1.
3. the polynomial Pl has exactly l roots and that there is precisely one root of the polynomial
Pl+1 between two roots of the polynomial Pl
b) The spherical harmonics demonstration plots the surface parameterized by the absolute value
of the respective Ylm (Ω) and the solid angle Ω itself by
 
sin θ cos φ
S = {~r ∈ R3 : ~r = |Ylm (Ω)| ~er } ~er =  sin θ sin φ 
cos θ

The complex phase of Ylm (Ω) is coded by a rainbow color scheme. The real and imaginary part
of Ylm (Ω) are shown in the same manner.
1. Convince yourself that the absolute value of Ylm is always axially symmetric around z.
2. Convince yourself that parity relation for inversion at the origin Ylm (−~er ) = (−)l Ylm (~er )
holds.
3. How is Ylm related to Yl,−m ?

Problem 3.4 Angular momentum conservation law

The angular momentum density of the electromagnetic field is defined by the antisymmetric tensor
field
1
Lij (~x, t) = 2 (xi Sj − xj Si ) ,
c
where S ~ denotes the Poynting vector.
a) Employ the momentum balance law to construct a local balance law for the angular momentum
density of the form
∂t Lij + ∇k Mijk = −Dij .
Determine the angular moment current tensor Mijk as well as the mechanical torque tensor Dij .
Rewrite the balance law in terms of the pseudo-vector field
1
Li (~x, t) = εijk Ljk ,
2
and suitable Mik and Di .
R
b) Formulate the angular momentum conservation law in integral form, for Li = V Li dV .
 c) Demonstrate that in the radiation gauge, i.e., ϕs = 0, the angular momentum of the field can
be decomposed, L = LS + LB , in a ’spin’ part
Z
1 ~×A ~˙ dV ,
LS = A
4πc2 V

and an ’orbital’ part LB that depends explicitly on the point of reference of the coordinate
system.

Due date: Tuesday, 6.11.12

WiSe 2012 29.10.2012

Prof. Dr. A.-S. Smith

Dipl.-Phys. Ellen Fischermeier
Dipl.-Phys. Matthias Saba
am Lehrstuhl für Theoretische Physik I
Department für Physik
Friedrich-Alexander-Universität
Erlangen-Nürnberg

Theoretische Physik 2: Elektrodynamik

(Prof. A.-S. Smith)

Solutions to Home assignment 3

Solution of Problem 3.1 Dirac delta function

a) Consider the charge ∆q on a spherical shell. The charge density at (r, θ, ϕ) for the point charge
∆q positioned at (R, θ0 , ϕ0 ) is
∆q
∆ρe = δ(r − R)δ(θ − θ0 )δ(ϕ − ϕ0 )
R2 sin θ
The charge distribution over the spherical shell is uniform. Therefore,
∆q Q
=
R2 sin θ0 ∆θ0 ∆ϕ0 4πR2
so that the charge density becomes

Q δ(r − R) 2π 0 π 0 Q δ(r − R) Q δ(r − R)

Z Z
ρe (~x) = dϕ dθ sin θ0 δ(θ − θ0 )δ(ϕ − ϕ0 ) = sin θ =
4π R2 sin θ 0 0 4π R 2 sin θ 4π R2

b) Consider the charge ∆q on a cylindrical surface. The charge density at (ρ, ϕ, z) for the point
charge ∆q positioned at (b, ϕ0 , z 0 ) is
∆q
∆ρe = δ(ρ − b)δ(ϕ − ϕ0 )δ(z − z 0 )
b
The charge distribution over the cylindrical surface is
∆q λ
0 0
=
b∆ϕ ∆z 2πb
while the charge density now obtains following form

λ δ(ρ − b) 2π 0 +∞ 0 λ δ(ρ − b)
Z Z
ρe (~x) = dϕ dz δ(z − z 0 )δ(ϕ − ϕ0 ) =
2π b 0 −∞ 2π b

1
 c) Consider the charge ∆q on a thin disc in the z 0 = 0 plane. In the cylindrical coordinates the
charge density at the (ρ, ϕ, z) for the point charge ∆q positioned at (ρ0 , ϕ0 , 0) we can write as

∆q
∆ρe = δ(ρ − ρ0 )δ(ϕ − ϕ0 )δ(z)
ρ
The charge distribution in the disc is uniform
∆q Q
= 2 , ρ0 ≤ R
ρ0 ∆ρ0 ∆ϕ0 R π
so that the total charge density equals
Z 2π Z R
( Q
Q δ(z) 0 0 0 0 0 δ(z) if ρ ≤ R
ρe (~x) = 2 dϕ dρ ρ δ(ρ − ρ )δ(ϕ − ϕ ) = R2 π
R π ρ 0 0 0 if ρ > R
Q
= δ(z)u(R − ρ),
R2 π
where u(R − ρ) is the unit step function defined as

1 if ρ ≤ R
u(R − ρ) =
0 if ρ > R

d) Consider the charge ∆q on a thin disc in the θ0 = π/2 plain. In the spherical coordinates the
charge density at the (r, θ, ϕ) for the point charge ∆q positioned at (r0 , π/2, ϕ0 ) is

∆q
∆ρe = δ(r − r0 )δ(θ − π/2)δ(ϕ − ϕ0 )
r2
The charge distribution over the disc is again uniform, thus
∆q Q
= 2 , r0 ≤ R
r0 ∆r0 ∆ϕ0 R π
By substituting this equation into the expression for ∆ρe we obtain

Q δ(θ − π/2)
Z 2π Z R  Q δ(θ − π/2)
0 0 0 0 0 if r ≤ R
ρe (~x) = 2 dϕ dr r δ(r − r )δ(ϕ − ϕ ) = 2
R π r2 0 0  0R π r
if r > R
Q δ(θ − π/2)u(R − ρ)
=
R2 π r

Solution of Problem 3.2 Properties of the Dirac delta function

a) Let f (x) be an arbitrary function. h(x) is invertible near the zero of order one at x0 .
Z Z
y=h(x) 1 1
dx δ[h(x)]f (x) = dy δ(y)f (h−1 (y)) 0 −1 = f (h−1 (0)) 0 −1 =
|h (h (y))| |h (h (0))|
h(x0 )=0 1
= f (x0 ) 0
|h (x0 )|
1
⇒ δ[h(x)] = δ(x − x0 )
|h0 (x 0 )|

2
 (i) δ(x) = 0 for x 6= 0 and x = 0 for x = 0:

⇒ xδ(x) = 0 for any x

.
(ii) δ(x − a) = 0 for x 6= a and ϕ(x = a) = ϕ(a):

⇒ ϕ(x)δ(x − a) = ϕ(a)δ(x − a)
R
(iii) f (x) is arbitrary, dy f (x − y)δ(y − z) = f (x − z). Now set f (x − y) = δ(x − y):
Z
⇒ dy δ(x − y)δ(y − z) = δ(x − z)

Solution of Problem 3.4 Angular momentum conservation law

The angular momentum density of the electromagnetic field is defined as

’angular momentum density’ ≡ ~x × ’momentum density’ ⇒ ~ = ~x × 1 S.

L ~
c2
~ is related to a tensor of 2nd rank by the Hodge duality,
The pseudo-vector L
1 1 1 1 1
Li = 2
εijk xj Sk = εijk 2 (xj Sk − xk Sj ) = εijk Ljk , Lij = (xi Sj − xj Si ).
c 2 c 2 c2

a) Recall momentum balance (Problem 3.4),

1
∂t Si + ∇k Tik = −Fi .
c2

c
Poynting vector: Si = εijk Ej Bk
4π  
1 1 ~ +B
2 ~ )
2
Maxwell stress tensor: Tik = − Ei Ek + Bi Bk − δik (E
4π 2
1
Lorentz force density: Fi = ρEi + εikl jk Bl
c
Then,
1 1
∂t Lij = xi 2
∂t S j − xj 2 ∂t S i
c c
= xi (−Fj − ∇k Tjk ) − xj (−Fi − ∇k Tik )
= −(xi Fj − xj Fi ) − ∇k (xi Tjk − xj Tik ) + δik Tjk − δkj Tik
| {z } | {z }
Dij Mijk
= −Dij − ∇k Mijk + ( Tji − Tij )
| {z }
=0 by symmetry

Thus, ∂t Lij + ∇k Mijk = −Dij , where we have introduced the

torque density: Dij = xi Fj − xj Fi ,

angular momentum current: Mijk = xi Tjk − xj Tik .

3
 Note the symmetries Dij = −Dji , Mijk = −Mjik . The Hodge duality yields pseudo-vectors and
-tensors,
1 1
Di ≡ εijk Djk , Mik = εimn Mmnk ,
2 2
then
∂t Li + ∇k Mik = −Di

b) The angular momentum conservation law in integral form reads

Z Z Z Z
d
Li dV + Di dV = − ∇k Mik dV = − Mik dfk
dt V V V ∂V

Interpretation: The angular momentum of the field can be converted to mechanical angular
momentum or flow through the surface.
c) The radiation gauge, ϕs = 0, implies

~ =∇
B ~ ×A ~ = − 1 A.
~ and E ~˙
c
Poynting vector
c ~ ~ k = c (E
~ ×∇ ~ k = c El ∇k Al − c El ∇l Ak
~ × A)
Sk = (E × B)
4π 4π 4π 4π

Z Z
1
Li = Li dV = εijk xj Sk dV
V c2 V
Z
1
= εijk xj (El ∇k Al − El ∇l Ak )dV partial integration
4πc V
−1
Z
= εijk (Al ∇k xj El − Ak ∇l xj El )dV + surface terms
4πc V
−1
Z
= εijk (Al δkj El + Al xj ∇k El − Ak δlj El − Ak xj ∇l El )dV + surface terms
4πc V
−1
Z h i
= ~ i El + (A
Al (~x × ∇) ~ × E)
~ i + (A~ × ~x)i div E
~ dV
4πc V

~ = L
Which may be rewritten as L ~S + L
~ B with a part independent of the point of reference
(’spin’),
Z
~ 1 ~×A ~˙ dV,
LS = A
4πc2 V

and an orbital part,

Z h
~B = 1
i
L ~ div E
(~x × A) ~ −A
~ · (~x × ∇)
~ E~ dV.
4πc V