Homework 2

Dahyeon Lee
February 14, 2018

Problem 1
Jackson 11.23
In a collision process a particle of mass m2 , at rest in the laboratory, is struck by a particle of mass m1 ,
momentum pLAB and total energy ELAB . In the collision the two initial particles are transformed into
two others of mass m3 and m4 . The configurations of the momentum vectors in the center of momentum
(cm) frame (traditionally called the center-of-mass frame) and the laboratory frame are shown in the
figure.

Figure 1: Jackson 11.23

(a) Use invariant scalar products to show that the total energy W in the cm frame has its square given
by
W 2 = m21 + m22 + 2m2 ELAB
and that the cms 3-momentum p0 is
m2 pLAB
p0 =
W

Consider the invariant scalar product pµ pµ = −p20 + p2 in the two frames, before the collision. In the lab
frame,
 2
µ ELAB
pµ p = − + m2 c + p2LAB .
c
In the cm frame,
 2
W
pµ pµ = − .
c
Therefore,
 2  2
W ELAB
= + m2 c − p2LAB
c c
E2
= LAB + m22 c2 + 2m2 ELAB − p2LAB .
c2

1
 PHYS7320 Homework 2

Now, consider a third frame in which m1 is at rest. In this frame, the invariant scalar product pµ pµ for
m1 is m21 c2 . But this should be equal to the invariant scalar product for m1 in the lab frame:
2
ELAB
m21 c2 = − p2LAB .
c2
Using this information, we get
 2
W
= m21 c2 + m2 c2 + 2m2 ELAB ,
c
or,
W 2 = m21 c4 + m22 c4 + 2m2 c2 ELAB .
If we wanted to, we could let c = 1 and get the result as stated in the problem.
Proving the second equation (for momentum) is a bit more involved. If we denote the energies of m1 and
m2 in the cm frame as W1 and W2 , respectively,
q
W1 = m21 c4 + p02 c2 ,
q
W2 = m22 c4 + p02 c2 .

Since the total energy is W = W1 + W2 ,

W 2 = W12 + W22 + 2W1 W2

= 2p02 c2 + m21 c4 + m22 c4 + 2W1 W2 .

Combining this with the first result,

2p02 c2 + 2W1 W2 = 2m2 c2 ELAB .

Rearranging and squaring both sides,

(m22 c4 + p02 c2 )(m21 c4 + p02 c2 ) = (m2 c2 ELAB − p02 c2 ) .

q
Using ELAB = m21 c4 + p2LAB c2 and after some trivial algebra,

p02 (m21 c2 + m22 c2 + 2m2 ELAB ) = m22 p2LAB c2 .

The term in the parenthesis is what we found to be (W/c)2 . Therefore,

m2 c2 pLAB
p0 = ,
W
where I used the fact that p0 and pLAB point in the same direction. Again, if we wanted to, we could let
c = 1 and obtain the equation as stated in the problem.

(b) Show that the Lorentz transformation parameters βcm and γcm describing the velocity of the cm
frame in the laboratory are
pLAB m2 + ELAB
βcm = , γcm =
m2 + ELAB W

Let’s call the speed of the cm frame with respect to the lab frame vcm . The momenta of each particle in the
two frames are related by Lorentz transformation. First, for m1 ,
 vcm 
p0 = γ pLAB − 2 ELAB
c

2
 PHYS7320 Homework 2

For m2 ,  vcm 
−p0 = γ 0 − 2 m2 c2 .
c
Combining these two relations, we get
pLAB
vcm = .
m2 + ELAB /c2

Therefore,
vcm pLAB /c
βcm = = .
c m2 + ELAB /c2
Letting c = 1, we obtain the desired result.
By looking at m2 in the cm frame, we see that p0 = γcm m2 vcm . Therefore,

p0 m2 c2 pLAB 1 m2 c2 + ELAB
γcm = = = ,
m2 vcm W m2 vcm W
where I have used the result of part a.

(c) Show that the results of parts a and b reduce in the nonrelativistic limit to the familiar expressions,
  2
m2 pLAB
W ≈ m1 + m2 +
m1 + m2 2m1
 
m2 pLAB
p0 ≈ pLAB , βcm ≈
m1 + m2 m1 + m2

We found in part a that

W 2 = m21 c4 + m22 c4 + 2m22 ELAB .
In the nonrelativistic limit,
p2LAB
ELAB ≈ m1 c2 + .
2m1
Therefore, in this limit,
m2 2 2 m2 2
W 2 ≈ m21 c4 + m22 c4 + 2m1 m2 c4 + c pLAB = (m1 c2 + m2 c2 )2 + p
m1 m1 LAB
(m2 /c2 /m1 )p2LAB
 
= (m1 c2 + m2 c2 )2 1 +
(m1 c2 + m2 c2 )2

and
1/2
(m2 c2 /m1 )p2LAB

W = (m1 c2 + m2 c2 ) 1 +
(m1 c2 + m2 c2 )2
m2 c2 p2LAB
 
≈ (m1 c2 + m2 c2 ) 1 +
(m1 c2 + m2 c2 )2 2m1
  2
2 2 m2 pLAB
= m1 c + m2 c + .
m1 + m2 2m1

Letting c = 1, we get the desired result.

Also from part a,
m2 c2 pLAB
p0 = .
W

3
 PHYS7320 Homework 2

In the nonrelativistic limit, W ≈ m1 c2 + m2 c2 . So,

m2 c2
 
m2
p0 ≈ p = pLAB
m1 c2 + m2 c2 LAB m1 + m2

and therefore  
0 m2
p ≈ pLAB .
m1 + m2
From part b,
pLAB c
βcm = .
m2 c2 + ELAB
In the nonrelativistic limit, ELAB ≈ m1 c2 . Therefore,
pLAB
βcm ≈ .
m1 c + m2 c

Problem 2
Jackson 11.13
An infinitely long straight wire of negligible cross-sectional area is at rest and has a uniform linear
charge density q0 in the inertial frame K 0 . The frame K 0 (and the wire) move with a velocity v parallel
to the direction of the wire with respect to the laboratory frame K.

(a) Write down the electric and magnetic fields in cylindrical coordinates in the rest frame of the
wire. Using the Lorentz transformation properties of the fields, find the componenets of the electric and
magnetic fields in the laboratory.

In the rest frame of the wire, there is no magnetic field because there is no moving charge. In this frame,
E can be calculated from Gauss’ law as usual. Taking a cylindrical gaussian surface of height l, we get

2πrlE = 4πq0 l ,
2q0
E0 = r̂ , B0 = 0 .
r
In the lab frame K (moving with velocity −v with respect to the frame K 0 ), the fields get transformed:

γ2β
E = γ(E0 − β × B0 ) − (β · E0 ) ,
γ+1
γ2β
B = γ(B0 + β × E0 ) − (β · B0 ) .
γ+1

Using E0 and B0 we found before and the fact that β · E0 = 0 (since β ⊥ E0 ),

2q0 /r
E = γE0 = p r̂
1 − (v/c)2

and
v/c 2q0
B = γβ × E0 = γβE 0 θ̂ = p θ̂ .
1 − (v/c) r
2

4
 PHYS7320 Homework 2

(b) What are the charge and current densities associated with the wire in its rest frame? In the
laboratory?

The charge density in the rest frame is given to be q0 . There is no current density because there is no moving
charge.
In the lab frame K, the charge density increases by a factor of γ because of length contraction. So the
charge density in this frame is γq0 . This charge density is moving at a velocity v, which produces current
density,
γq0 v
J = δ(r) x̂ ,
2πr
R
I found this by requiring I = J da and setting the x axis along the wire.

(c) From the laboratory charge and current densities, calculate directly the electric and magnetic fields
in the laboratory. Compare with the results of part a.

This is a straightforward calculation. E can be calculated by Gauss’ law:

2γq0
E= r̂ ,
r
which agrees with what we have found in part a.
The magnetic field in K can be calculated from Ampere’s law:
4π
2πr · B = γq0 v ,
c
which leads to
2γq0 v 2γβq0 v/c 2q0
B= = =p ,
rc r 1 − (v/c) r
2

This agrees with the result from part a.

5
 PHYS7320 Homework 2

Problem 3
Jackson 11.14
(a) Express the Lornetz scalars F αβ Fαβ , F αβ Fαβ , and F αβ Fαβ in terms of E and B. Are there any
other invariants quadratic in the field strengths E and B?

The explicit expressions for the field tensors are

0 Ex Ey Ez 0 Bx By Bz
   
−Ex 0 Bz −By  −Bx 0 −Ez Ey 
F µν =
−Ey
 , F µν =  .
−Bz 0 Bx  −By Ez 0 −Ex 
−Ez By −Bx 0 −Bz −Ey Ex 0

Just like for four-vectors, when lowering an index to make it covariant, the tensor obtains a minus sign if
the index is zero. With this in mind,

F αβ Fαβ = F 00 F00 + F 01 F01 + F 02 F02 + F 03 F03

+ F 10 F10 + F 11 F11 + F 12 F12 + F 13 F13 + · · ·
= −2(Ex2 + Ey2 + Ez2 ) + 2(Bx2 + By2 + Bz2 ) = −2(E2 − B2 ) .

Following the same process,

F αβ Fαβ =F 00 F00 + F 01 F01 + F 02 F02 + F 03 F03

+ F 10 F10 + F 11 F11 + F 12 F12 + F 13 F13 + · · ·
=0 − Ex Bx − Ey By − Ez Bz
− Ex Bx + 0 − Ez Bz − Ey By
− Ey By − Ez Bz + 0 − Ex Bx
− Ez Bz − Ey By − Ex Bx + 0
= − 4(Ex Bx + Ey By + Ez Bz ) = −4E · B .

Finally, again by the same process,

F αβ Fαβ = 2(E2 − B2 ) .

Another invariant quantity quadratic in the field strength strengths is F αβ Fαβ , although I’m not being
very creative here.

(b) Is it possible to have an electromagnetic field that appears as a purely electric field in one inertial
frame and as a purely magnetic field in some other inertial frame? What are the criteria imposed on E
and B such that there is an inertial frame in which ther is no electric field?

If the electromagnetic fields appear purely electric in one inertial frame, that implies F αβ Fαβ = 2(E2 −B2 ) >
0. But since F αβ Fαβ is lorentz invariant, its sign is always the same in any other inertial frame. But if
there exists an inertial frame in which this field looks purely magnetic, that would imply F αβ Fαβ < 0 in
that frame, which is a contradiction. Therefore, there is no frame in which this field looks purely magnetic.
Also note that in the frame where the field is purely electric, F αβ Fαβ = −4E · B = 0. But this quantity
is also invariant, i.e., E · B = 0 in any other inertial frame. Therefore, in order for there to exist a frame in
which the field is purely electric, the condition E · B should be satisfied.

6
 PHYS7320 Homework 2

Problem 4
Jackson 11.16
In the rest frame of a conducting medium the current density satisfies Ohm’s law, J0 = σE0 , where σ is
the conductivity and primes denote quantities in the rest frame.

(a) Taking into account the possibility of convection current as well as conduction current, show that
the covariant generalization of Ohm’s law is
1 σ
Jα − (Uβ J β )U α = F αβ Uβ
c2 c
where U α is the 4-velocity of the medium.

Let’s check that the given covariant expression reduces to the usual Ohm’s law in the rest frame of the
conductor. In this frame, U α = (c, 0, 0, 0). So, Uβ J β = c2 ρ and F αβ Uβ = cF α0 .1 First look at the spatial
components:
1
J i = 2 (c2 ρ)U i = σF i0 = σE i .
c
Since U i = 0, we get the desired result J i = σE i .
Now look at the time component:
1 2 σ
cρ − (c ρ)(c) = (−cF 00 )U0 ,
c2 c
which is just equivalent to 0 = 0.
Therefore, the given covariant generalization of Ohm’s law works in the particular case of the rest frame
of the conductor. But since the equation is manifestly covariant, it is also valid in any other inertial frame.

(b) Show that if the medium has a velocity v = cβ with respect to some inertial frame that the 3-vector
current in that frame is
J = γσ[E + β × B − β(β · E)] + ρv
where ρ is the charge density observed in that frame.

In this frame, U = (cγ, U), where U = γv = γcβ. So Uβ J β = c2 γρ − u · J. The spatial part of the
covariant Ohm’s law becomes (for α = 1)
1 2 σ σ
Jx − (c γρ − U · J)Ux = F 1β Uβ = (γcE + U × B)x .
c2 c c
Combining it with similar equations for α = 2 and α = 3,
1 1
J = γσ(E + u × B) + (γρ − 2 U · J)U . (1)
γc c

Now, look at the time component (α = 0):

1 2 σ
cρ − (c γρ − U · J)cγ = E · U ,
c2 c
which can be written as
1 1 σ
γρ − 2
U · J = ρ − 2E · U . (2)
c γ γc
1I am using Jackson’s convention of the metric, (+, −, −, −). This is required to get the expected equation as stated in the
problem.

7
 PHYS7320 Homework 2

Combining (1) and (2),

1 1 σ
J = γσ(E + U × B) + ρU − 2 U(U · E) .
γc γ γc
Using U = γv = γcβ,

J = γσ(E + β × B) + ρv − γσβ(β · E)
= γσ[E + β × B − β(β · E)] + ρv .

(c) If the medium is uncharged in its rest frame (ρ0 = 0), what is the charge density and the expression
for J in the frame of part b? This is the relativistic generalization of the equation J = σ(E + v × B)
(see p. 320).

In the rest frame of the conductor, Uβ J β = c2 ρ0 = 0. But since this quantity is invariant, it is zero in all
inertial frames. The Ohm’s law now reads
σ αβ
Jα = F Uβ .
c
Letting α = 0, we get
σ 0β σ
cρ = F Uβ = (β · E) .
c c
Letting α = i, we get
σ iβ
Ji = F Jβ = γσ(E + β × B)i ,
c
which becomes
J = γσ(E + β × B) .
We could have directly applied Lorentz transformation to the four-vector J 0α , but that gives J α in terms
of E0 and B0 , which has to be transformed in terms of E and B. This method requires a significantly more
work.

Problem 5
(a) Show that the Lagrangian density for the vector field φµ ,

1 1 µ2
L = − ∂µ φν ∂ µ φν + ∂µ φµ ∂ν φν + φµ φµ ,
2 2 2
leads to the field equation [ηµν (∂λ ∂ λ + µ2 ) − ∂µ ∂ν ]φν (x) = 0.

Lagrange’s equation for the field φλ is

∂L ∂L
− ∂σ =0.
∂φλ ∂∂σ φλ
Let’s calculate each relevant term:
∂L µ2 ∂ µ µ2 ∂
= (φ µ φ ) = (φσ ησµ φµ )
∂φλ 2 ∂φλ 2 ∂φλ
µ2
= (ησµ δλσ φµ + ησµ δλµ φσ )
2
µ2
= (φλ + φλ ) = µ2 φλ .
2

8
 PHYS7320 Homework 2

∂ ∂
(∂µ φν ∂ µ φν ) = (ηνα ∂µ φα η µβ ∂β φν )
∂∂σ φλ ∂∂σ φλ
= ηνσ η µβ δµσ δλα ∂β φν + ηνα η µβ δβσ δλν ∂µ φα
= δµσ δλα ∂ µ φα + δβσ δλν ∂ β φν
= 2∂ σ φλ .

Therefore,  
∂ 1
∂σ − ∂µ φν ∂ µ φν = −∂σ ∂ σ φλ .
∂∂σ φλ 2
The second term contributes
∂
(∂µ φµ ∂ν φν ) = δµσ δλµ ∂ν φν + δνσ δλν ∂µ φµ = 2∂ν φν .
∂∂σ φλ
and  
∂ 1
∂σ ∂µ φµ ∂ν φν = ∂σ ∂ν φν .
∂∂σ φλ 2
Combining these together, the Lagrange equation is

µ2 φλ = −∂σ ∂ σ φλ + ∂σ ∂ν φν .

Rearranging the terms and renaming dummy indices, we get

(∂λ ∂ λ + µ2 )φµ − ∂µ ∂ν φν = 0 .

The covariant index can be brought upstairs by adding a metric tensor:

[ηµν (∂λ ∂ λ + µ2 ) − ∂µ ∂ν ]φν = 0 .

(b) Show that ∂ν φν = 0 for a solution of these equations.

Rewrite the field equations:

ηµν ∂λ ∂ λ φν + ηµν µ2 φν − ∂µ ∂ν φν = 0 .
Differentiate the whole thing with ∂ν :

ηµν ∂λ ∂ λ ∂ν φν + ηµν µ2 ∂ν φν − ∂ν ∂µ ∂ν φν = 0 .

But ∂ν = ηµν ∂ µ , so the first and the third terms cancel, leaving

ηµν µ2 ∂ν φν = 0

and therefore
∂ν φν = 0 .