Gravity and Cosmology

Problem 3.2 : Rindler Coordinates Consider the two-dimensional Minkowski space-time

with metric
ds2 = c2 dt2 − dx2 (1)
(a) (1 point) Change coordinates to x± with

x± = ct ± x (2)
Find the metric in these coordinates. Show that the path of a photon moving to the right is
given by x− = 0 while that of a left-moving photon by x+ = 0. This is why these coordinates
are known as light-cone coordinates.
b) (2 points) Change now again coordinates to the Rindler coordinates

x± = ρe±τ (3)

Find the metric in these new coordinates. What part of the Minkowski space-time these
coordinates cover? How can you cover the rest of the space?
(c) (4 points) Show that an observer at rest in the Rindler coordinate system, is uniformly
accelerating with respect to the original coordinate system.

a)The metric in the new coordinates is

ds′2 = g ′ ab df a df b ⇒ ds′2 = g11

′ ′
dx2+ + g12 ′
dx+ dx− + g21 ′
dx− dx+ + g22 dx2− (4)

Since x± = ct ± x we can write down the differentials

dx+ = cdt + dx, dx− = cdt − dx (5)

Substituting (5) in (4) and collecting together the common terms gives us

ds′2 = (g11
′ ′
+ g12 ′
+ g21 ′
+ g22 ′
)c2 dt2 + (g11 ′
− g12 ′
− g21 ′
+ g22 ′
)dx2 + (g11 ′
− g22 )2cdtdx (6)
′
Equating (6) with (1) gives us a set of equations for the gab
 ′ ′ ′ ′

 g11 + g12 + g21 + g22 =1 
1
′ ′ ′ ′ ′ ′ ′ ′
g11 − g12 − g21 + g22 = −1 ⇒ g11 = g22 = 0, g12 = g21 = (7)
′ ′ 2
g11 − g22 =0
 

The metric is then

ds′2 = dx+ dx− (8)
When x+ = 0 we have c = −dx/dt which means that dx must be negative and thus the
photon is moving to the left. Similarly, when x− = 0 we get c = dx/dt and as a result the
photon is moving to the right.
b) First we find the differential in Rindler coordinates

dx+ = dρeτ + ρeτ dτ, dx− = dρe−τ − ρe−τ dτ (9)

1
Substituting to (8) we get the new metric

ds2r = (dρeτ + ρeτ dτ )(dρe−τ − ρe−τ dτ ) ⇒ ds2r = dρ2 − ρ2 dτ 2 (10)

To find what part of the Minkowski space ρ and τ cover we connect them to the initial
coordinates x and t.
√
 
ct + x = ρeτ 1 ct + x
−τ ⇒ ρ = c2 t2 − x2 , τ = ln( ) (11)
ct − x = ρe 2 ct − x

Then (ct + x)(ct − x) ≥ 0 which means that −ct ≤ x ≤ ct. This is true for events inside the
light cone.
c) An observer at rest in Rindler’s system has a constant ρ = R and thus ds2r = −R2 dτ 2 .
We need to find the acceleration in the original system and show that it is constant. The
two components of the acceleration are

d 2 x0 d 2 x1
a0 = 2
, a 1
= 2
(12)
dτproper dτproper

, where the proper time can be calculated by

R2 2
c2 dτproper
2
= −R2 dτ 2 ⇒ dτproper
2
=− dτ (13)
c2
Substituting x0 = ct, x1 = x and (13) into (12) we get

c2 d2 ct 1 c2 d 2 x
a0 = − , a = − (14)
R2 dτ 2 R2 dτ 2
Now we need expressions of x, t as functions of τ . Using x± = x ± ct and x± = ρe±τ we can
show that
x+ − x− x + + x−
x= = ρsinhτ, ct = = ρcoshτ (15)
2 2
Now we can take the derivatives (14) to find

c2 c2
a0 = − coshτ, a1 = − sinhτ (16)
R R
The magnitude of the acceleration is

c2
|a| = (a0 )2 − (a1 )2 = (17)
R
which is constant.

Problem 3.3 : (a) (1 point) Find the geodesic equations on a two-dimensional plane in
polar coordinates.
(b) (6 points) Solve them and find the geodesics.

2
a)The metric of a two-dimensional space in polar coordinates is

ds2 = dρ2 + ρ2 dθ2 (18)

The distance between two points, A and B, in this space is

Z B Z Bp Z B q
SAB = ds = 2 2
dρ + ρ dθ =2 ds ρ̇2 + ρ2 θ̇2 (19)
A A A

, where the dot denotes derivative with respect to the length of the curve s. Since we want to
find the geodesics we need the minimum of SAB and thus we must solve the Euler-Lagrange
equations
d ∂L ∂L d ∂L ∂L
= , = (20)
ds ∂ ρ̇ ∂ρ ds ∂ θ̇ ∂θ
q
2 2 2
, where L = ρ̇ + ρ θ̇ (this is because ρ̇2 + ρ2 θ̇2 = 1 when we use the length of the curve
for the parametrization). Substituting L gives us the two equations we need to solve

d
(2ρ̇) = 2ρθ̇2 ⇒ ρ̈ − ρθ̇2 = 0 (21)
ds
d C
(2ρ2 θ̇) = 0 ⇒ θ̇ = 2 (22)
ds ρ
We use (8) to get rid of θ̇ in (7)
ρ̈ − C 2 ρ−3 = 0 (23)
Then we make the substitution ρ = 1/u, dρ/du = −1/u2 . The derivatives of ρ become

dρ dρ du dθ 1 du C du
= =− 2 2
= −C (24)
dt du dθ dt u dθ ρ dθ

d d du d du dθ d2 u
ρ̇ = (−C ) = (−C ) = −C 2 u2 2 (25)
dt dt dθ dθ dθ dt dθ
Substituting to (14) gives us a new, easier differential equation

d2 u d2 u
− C 2 u2 − C 2 3
u ⇒ + u = 0 ⇒ u = Acosθ + Bsinθ (26)
dθ2 dθ2
Going back to ρ we get
1
ρ(θ) = (27)
Acosθ + Bsinθ
Next we can use this result in (13) to derive an expression for θ(t)
Z
dθ C dθ
= 2
⇒ = Ct ⇒ (28)
dt (Acosθ + Bsinθ) (Acosθ + Bsinθ)2

sec2 θdθ
Z Z
1 dω 1 1
2 2
= Ct ⇒ 2
= Ct ⇒ − + C ′ = Ct ⇒ (29)
(A + Btan θ) Bω Bω
1 1
− + C ′ = Ct (30)
B A + Btanθ

3
, where in (20) we divided by cos2 θ and changed variables to ω = A + Btanθ. In (21) we
returned to the original variable. If we solve for θ we end up with
1
θ(t) = tan−1 ( ) (31)
C1 − C2 t
We can find the two constants from initial conditions and then substitute to (18) to find
ρ(t).
It is worth noting that we can check that result (18) is correct from the equation of the
geodesic in the cartesian metric y = C1 x + C2 . Using x = ρcosθ and y = ρsinθ gives us
ρ = C2 /(sinθ − C1 cosθ) which is of the same form as (18).

Problem 3.5 : (4 points) Consider the two-dimensional metric

ds2 = y 2 dx2 + x2 dy 2 (32)

Find the Gauss curvature K.

 
y2 0
The metric of this space is .
0 x2
In the special case of a diagonal 2x2 metric, as this one, the Gauss curvature is given by
1  1 h ∂g11 ∂g22 ∂g11 2 i 1 h ∂g11 ∂g22 ∂g22 2 i
K= + ( ) + + ( )
2g11 g22 2g11 ∂x1 ∂x1 ∂x2 2g22 ∂x2 ∂x2 ∂x1

1  ∂ 2 g11 ∂ 2 g22 
− + (33)
2g11 g22 (∂x2 )2 (∂x1 )2
In our case x1 = x, x2 = y g11 = y 2 and g22 = x2 and thus the derivatives are

∂g11 ∂g22 ∂g11 ∂g22 ∂ 2 g11 ∂ 2 g22

= 0, = 2x, = 2y, = 0, = 2, =2 (34)
∂x1 ∂x1 ∂x2 ∂x2 (∂x2 )2 (∂x1 )2

Substituting these results in (5) we get

1  1 2 1 2
 1
K= 2 2 2
4y + 2
4x − 2 2 (2 + 2) ⇒ K = 0 (35)
2y x 2y 2x 2y x