C161 Practice Midterm Problems

These are practice problems that should be mostly relevant to the problems you
will see on the C161 2023 midterm. The actual midterm will not include as
many problems (more likely around 4 to 5 such problems will appear).

Problem 1: The metric for a certain 2D spacetime is, using polar-type

coordinates
c2 dt2
ds2 = − + (1 + 2αr)2 dr2 + (1 + αr)2 r2 dϕ2 (1)
(1 − 2αr)2

where α is a constant.
1a) Consider a circle centered about r = 0 that has radius with a proper
distance dp . What is the circumference of the circle in terms of dp ?

Solution: We calculate first the proper distance from r = 0 to r = r. Since it

is a proper distance, we take dt = 0 and since it is radial we take dϕ = 0 (i.e.,
ϕ is constant along the path we are integrating).

ds2 = dr(1 + 2αr)2 =⇒ ds = (1 + 2αr)dr

Integrating this from the center r = 0 to r = r

Z Z r
dp = ds = (1 + 2αr)dr = r + αr2
0

To get the circumference we take a proper distance dt = 0 and since we are

moving around at fixed r we have dr = 0

ds2 = (1 + αr)2 r2 dϕ2 =⇒ ds = (1 + αr)rdϕ = (r + αr2 )dϕ

Integrating this around in phi, from 0 to 2π

Z Z 2π
C = ds = (r + αr2 )dϕ = 2π(r + αr2 )
0

Comparing to dp we see that C = 2πdp

1b) How much time t does it take for light emitted radially from r = 0 to get
to a radial coordinate r?

1
Solution: Light moves along a null (or lightlight) geodesic which has ds2 = 0.
Since the light moves in the radial direction, dϕ = 0, and the metric gives

c2 dt2
ds2 = − (1 + 2αr)2 dr2 = 0
(1 − 2αr)2

Taking the square root

cdt
= (1 + 2αr)dr
(1 − 2αr)

Separating the variables

dr(1 + 2αr)(1 − 2αr) = cdt

dr(1 − 4α2 r2 ) = cdt

Integrating from the emission of light at r = 0 and t = 0 to the receiving of
light at r = r and t = t
Z r Z t
2 2
dr(1 − 4α r ) = cdt
0 0

4
r + α2 r3 = ct
3
r 4 α2 r 3
t= −
c 3 c

2
Problem 2: Imagine we live in an expanding flat universe with a single
energy density component with equation of state parameter w = +1.
2a) Solve for the time evolution of the scale factor, a(t) (don’t just quote an
answer for a(t), but show how you arrived at it from the cosmological equations
we have used).

Solution:
The evolution of the energy density follows the solution of the fluid equation
dϵ ϵ
= −3 (1 + w)
da a
The solution for constant w is

ϵ = ϵ0 a−3(1+1) = ϵ0 a−6

We have seen this solution enough times that it is OK to simply quote it, but we
should keep in mind cases where it may not hold (i.e., when w is not constant
as in the next problem).
We can use the result in the Friedmann equation in the form

ȧ2
 
2 Ω0 (1 − Ω0 )
= H 0 +
a2 a6 a2
Since this is a flat universe, Ω0 = 1

ȧ2 H2 H2 da H0
2
= 60 =⇒ ȧ2 = 40 =⇒ = 2
a a a dt a
Separating the variables and integrating
Z a Z t
da a2 = H0 dt
0 0

a3
= H0 t =⇒ a(t) = (3H0 t)1/3
3
Note that if we had instead used the Friedmann equation in the form

ȧ2 8πG kc2

= ϵ(a) − 2
a2 3c2 R0 a(t)2
Then the condition that the universe if flat means that k = 0 and plugging in
the expression for ϵ(a)
ȧ2 8πG ϵ0
=
a2 3c2 a6
Further the condition that the universe if flat means ϵ0 = ϵ0,c where the critical
energy density is
3c2 H02
ϵ0,c =
8πG

3
And the same solution follows.
2b) What is the current age of this universe, t0 , in terms of the Hubble time
tH = 1/H0 ?

Solution: We use the fact that a(t0 ) = 1. Then our solution for a(t) gives
1 tH
a(t0 ) = (3H0 t0 )1/3 = 1 =⇒ 1 = 3H0 t0 =⇒ t0 = =
3H0 3

2c) An alien civilization lives in a far away galaxy at a current proper distance
dp = 32 ctH from us. If we try to communicate with them by sending radio waves,
what redshift will they observe for the light signal we send?

Solution:
As we considering light, we use the null geodesic of the FRW metric for radial
traveling light (dΩ = 0)

ds2 = −c2 dt2 + a(t)2 dr2 = 0

Which implies
c2 dt2 = a(t)2 dr2 =⇒ cdt = ±a(t)dr
We place ourselves at r = 0 and the source at r. The light is thus moving in
the positive r direction and thus we take the positive sign. Integrating from
emission time of now, t0 , to some future observed time tf
Z tf Z r
dt
c = dr
t0 a(t) 0
Z tf
c dt(t/t0 )−1/3 = r
t0
tf
3
ct0 (t/t0 )2/3 =r
2 t0
3
ct0 [(tf /t0 )2/3 − 1] = r = dp
2
Where we used the fact that the current proper distance is dp = ra(t0 ) = r.
So we have determined the relationship between dp in terms of tf . To get the
redshift we note that z is related to scale factor by
a(tobs ) a(tf )
1+z = = = a(tf )
a(temit ) a(t0 )
We needed to be a bit careful here – in many problems we consider a source
emitted in the past and observed now, but this problem specifies that the light is
emitted now and observed in the future. So by the above arguement 1+z = a(tf )
and not 1 + z = 1/a(tf ).

4
 To calculate the redshift, we need the scale factor at the observed time,
a(tf ). One way to go about it is to solve the above equation for tf

2 dp 2 dp
[(tf /t0 )2/3 − 1] = =⇒ (tf /t0 )2/3 = 1 +
3 ct0 3 ct0
Before going any further, we may realize that since our solution for the scale
factor can be written a(t) = (t/t0 )1/3 the left hand side of the above equation
is just a(tf )2 . So we can save some work by just using this

2 dp
a(tf )2 = 1 +
3 ct0
(We would get the same result if we were to solve explicitly for tf and then plug
into our expression for a(t)). We showed above that t0 = 1/3H0 = tH /3, and
the problem says that dp = 3/2ctH so

2 32 ctH
a(tf )2 = 1 + =1+3=4
3 ctH /3
And so a(tf ) = 2. Then

1 + z = a(tf ) = 2 =⇒ z = 1

5
Problem 3: Matter that accumulates on the surface of a neutron star can
sometimes ignite and undergo a thermonuclear explosion in what is called a
Type 1 x-ray burst. For this problem, assume the Schwarzchild metric applies
and that the neutron star surface is at r = 4rs .
3a) If the explosion occurs over a period of 1 second (in a frame at the surface
of the neutron star) how long would the explosion be observed to last according
to an observer at infinity?

Solution: This is an effect of gravitational time dilation. We note that the

proper time for the explosion is ∆τ = 1 second and we want to get the observed
time. We write the Schwarzchild metric in terms of a proper time

ds2  rs  dr2 r2 2
dτ 2 = − 2
= dt2
1 − − 2 rs

− dΩ
c r c 1− r
c2

For proper time, we set dr = 0 and dΩ = 0

 rs 
dτ 2 = dt2 1 −
r
which gives us the standard expression for gravitational time dilation in the
Schwarzchild metric. Since the events are happening at r = 4rs we have
 1/2  1/2 r
rs 1 3
dτ = dt 1 − = dt 1 − = dt
4rs 4 4
Integrating gives
r r r
3 4 4
∆τ = ∆τ =⇒ ∆t = ∆τ = second
4 3 3

3b) Say the explosion radially ejects some amount of mass, m, at coordinate
speed dr/dt = v0 from the surface of the neutron star. What is the value of
the constant of motion E (energy) for this ejected material in terms of mc2 and
v0 /c?

Solution: The constant of motion E is

E  rs  dt
= c2 1 −
m r dτ
We will evaluate this at the moment of ejection, since the problem tells us then
that the radius r = 4rs and the coordinate velocity is dr/dt = v0 . Evaluating
the above at r = 4rs gives
3 dt
E = mc2
4 dτ
To go further we need to determine what dt/dτ is. This is similar to the previous
problem relating dt and dτ , but now we have consider the motion of the ejected

6
material, so dr ̸= 0. We turn to the Schwarzchild metric, written in terms of
proper time, and keeping only the radial part (since dΩ = 0 for radial motion
of the mass)

ds2  rs  dr2  rs −1

dτ 2 = − = dt2
1 − − 1 −
c2 r c2 r
Plugging in r = 4rs
3 2 4 dr2
dτ 2 =dt −
4 3 c2
This metric provides the relationship between dτ , dt and dr. The problem
specifies dr/dt = v0 so to get rid of the dr we pull out a factor of dt2 from the
right hand side

4 dr2 1 4 v02
   
2 2 3 2 3
dτ = dt − = dt −
4 3 dt2 c2 4 3 c2

Taking the square root

1/2 −1/2
3 4 v02 3 4 v02
 
dt
dτ = dt − =⇒ = −
4 3 c2 dτ 4 3 c2

So the energy is
−1/2
3 3 4 v02

E = mc2 −
4 4 3 c2
There are other ways to go about this. Beginning with the metric in the form
we found above
3 4 dr2
dτ 2 = dt2 −
4 3 c2
we could have divided through by dτ 2 to get

3 dt2 4 dr2 1
1= 2
−
4 dτ 3 dτ 2 c2
And solving for dt2 /dτ 2 we find

dt2 4 16 dr2 1
= +
dτ 2 3 9 dτ 2 c2
This isn’t exactly what we want, though, since it gives dt/dτ in terms of dr/dτ ,
whereas the problem specified dr/dt. (Furthermore, dr/dτ is not a constant of
motion in this problem, since the Schwarzchild metric depends on r). But we
could proceed by using the chain rule to write
dr dr dt
=
dτ dt dτ

7
Then the above becomes
dt2 4 16 dr2 dt2 1 4 16 dt2 v0
= + = +
dτ 2 3 9 dt2 dτ 2 c2 3 9 dτ 2 c2
We can now solve this for dt/dτ

dt2
 
16 v0 4
2
1− 2
=
dτ 9 c 3
r  −1/2
dt 4 16 v0
= 1−
dτ 3 9 c2
This can be shown to be identical to the form found above, since
r   −1/2  −1/2
dt 4 4 3 4 v0 3 4 v0
= − = −
dτ 3 3 4 3 c2 4 3 c2

3c) Say that the mass m ejected in the explosion expands to a radius 8rs
before stopping and falling back onto the neutron star. What is the energy of
this ejected mass (in terms of mc2 ). To determine the energy released in the
explore, compare this to the energy the mass m had before the explosion (i.e.,
when it was sitting at rest on the surface of the neutron star).

Solution: At the turnaround point, rm , the metric is

−1
dr2
  
2 2 rs rs
dτ = dt 1− − 2 1−
rm c rm
or " −1 #
dr2 1
 
rs rs
dτ 2 = dt2 1− − 2 2 1−
rm dt c rm
At the turn-around point, the matter stops and starts to fall back in, so dr/dt =
0. This implies  
2 2 rs
dτ = dt 1 −
rm
Or  −1/2
dt rs
= 1−
dτ rm
Then plugging this into the expression for our constant or motion, we get
  −1/2  1/2
E rs rs rs
= c2 1 − 1− = c2 1 −
m rm rm rm
 1/2
rs
2
E = mc 1 −
rm

8
The problem says the turnaround happens at rm = 8rs , so
 1/2 r
2 1 7 2
E = mc 1 − = mc
8 8

For a mass sitting at rest on the surface of the star we use the same expression
but for r = 4rs . That gives
 1/2 r
2 1 3 2
Einitial = mc 1 − = mc
4 4

This is consistent with our result of part a) if we plug v0 = 0 into the expression.
The difference in energy is
r r "r r #
7 2 3 2 7 6 mc2
∆E = mc − mc = − mc2 = √
8 4 8 8 8

Problem 4: The metric for a certain spacetime is given by

 2
2 2 2 2r  r 2 2
ds = −c dt + 1 + dr2 + r2 1 + dΩ (2)
α α

where dΩ2 = dθ2 + sin2 θdϕ2 and α is a constant.

4a) Find an expression for the proper distance in the radial direction (i.e.,
along r) from coordinate r = 0 to coordinate r = r0 .

Solution: To calculate this proper distance we set dt = 0 and since it is radial

dΩ = 0. The metric becomes
 2
2r
ds2 = 1 + dr2
α
so  
2r
ds = 1+ dr
α
Integrating ds from r = 0 to r = r0 will give us the proper distance, which we’ll
call R Z r0  r
r2 0 r2

2r
R= 1+ dr = r + = r0 + 0
0 α α 0 α
and so  r0 
R = r0 1 +
α

4b) Say you are at coordinate r = r0 and you draw in the equatorial plane
(θ = π/2) a circle in the ϕ direction (such that dθ = 0 and dr = 0 along the
circle). What is the proper distance circumference of the circle?

9
Solution: For this proper distances we have dt = dr = dθ = 0 and θ = π/2.
The metric becomes
 r 2 2  r 2  r 2 2
ds2 = r2 1 + dΩ = r2 1 + (dθ2 + sin2 θ2 dϕ2 ) = r2 1 + dϕ
α α α
and here r = r0 so  r0 
ds = r0 1 + dϕ
α
Integrating we get a proper distance we call C
Z 2π 
r0   r0 
C= r0 1 + dϕ = 2πr0 1 +
0 α α

4c) Are the properties that you have just calculated for the circle consistent
with the spacetime of this metric being flat?

Solution: We want to check how the circumference of the circle relates to its
radius. We found  r0 
C = 2πr0 1 +
α
and the radius of the circle
 r0 
R = r0 1 +
α
So together we have
C = 2πR
So this is consistent with the spacetime being flat.
Indeed we saw above  
2r
dR = 1 + dr
α
So we can make a change of variables r → R and write the metric

ds2 = −c2 dt2 + dR2 + R2 dΩ2

which is just the metric of flat space in spherical coordinates. So the original
metric given was just that of flat Minkowski space, but using a different radial
coordinate to label r.

Problem 5: The Schwarzchild metric describes the spacetime outside a

spherical mass M . Far away from the mass (r ≫ rs , where rs is the Schwarzchild
radius) the full Schwarzchild metric can be approximated by one having the form

ds2 = −(1 − rs /r)c2 dt2 + (1 + rs /r) dx2 + dy 2 + dz 2

(3)
p
where r = x2 + y 2 + z 2 .

10
5a) For an object moving in the above metric, is the special relativistic linear
momentum in the x direction, px = m dx
dτ a constant of motion? Explain why or
or why not.

Solution: The terms in metric depend on x, since they depend on r which

depends on x. So this means that the px is not a constant of motion.

A source (say, a pulsar) sends a collimated beam of light through our galaxy
in the x-direction, as seen in the figure. The beam passes by the supermassive
black hole at our galactic center with an impact parameter of 1 parsec, and
eventually reaches an observer 1000 parsecs away from the black hole. The
mass of the black hole is M (for this problem you can ignore any other mass in
the galaxy).
5b) How far below the original beam direction (shown as ∆y in the figure)
would an observer be if they see this light beam? Express your answer for ∆y in
terms of M and fundamental constants. (You can assume that the light beam
bends by a very small angle, and use key results we found in class without
rederiving them.)

Solution: For a weak gravitational lensing, the deflection angle is

4GM
α=
bc2
The distance we seek is
∆y
tan α =
D
where D = 1000b in the figure. Using small angle approximation
∆y
α≈
D
4GM ∆y
=
bc2 D

11
So
4GM D 4GM 4000GM
∆y = 2
= 2
100 =
c b c c2

5c) Find an expression for the (coordinate) speed, dx/dt of light when the
light beam is just passing by the black hole (i.e., is at coordinate r = b). Since
the bending of the light is minor, you can approximate the beam here as moving
only in the x direction (i.e., approximate its path as the dashed line).

Solution:
For light we set ds2 = 0 and the metric along the x direction (setting dz =
dy = 0) is
0 = −(1 − rs /r)c2 dt2 + (1 + rs /r) dz 2
So
(1 − rs /r)c2 dt2 = (1 + rs /r) dz 2
dz 2 (1 − rs /r) 2
2
= c
dt (1 + rs /r)
s
dz (1 − rs /r)
=± c
dt (1 + rs /r)

5d) If the black hole had a mass of 2M instead of M would it take the light
pulse more or less coordinate time ∆t to travel from the source to the observer?
Briefly explain why.

Solution: Increasing M would increase the Schwarzchild radius rs . We can

see from the above equation that this reduces the coordinate speed of light.
Thus it would take longer for the pulse to travel from pulsar to observer. This
is the Shapiro delay we explored on the homework.
In addition, the deflection of the angle could lead to a longer distance to
travel, depending on the position of the observer.

Problem 6: The Robertson-Walker metric of a flat universe is

ds2 = −c2 dt2 + a(t)2 (dr2 + r2 dΩ2 ) (4)

Let’s say that the scale factor evolves as a(t) = eH0 (t−t0 ) where t0 is the time
now and H0 is the Hubble constant. (This describes a universe that contains
only a cosmological constant).

12
6a) We observe light from a galaxy a comoving distance r away from us. Find
an expression for the time it took for this light to travel to us (that is, the time
observed minus the emitted time).

Solution: For light ds2 = 0 and choosing the light to move radially dΩ = 0
the metric is
c2 dt2 = e2H0 (t−t0 ) dr
cdt = ±eH0 (t−t0 ) dr
Separating variables
dr
dt e−H0 (t−t0 ) = ±
c
We integrate both sides, taking a bit of care in picking the limits. Let’s choose
the source to be at r = r and us at r = 0. Then the starting spacetime point is
t = te , r = r and the ending is t = r0 , r = 0. The integral is
Z t0 Z 0
−H0 (t−t0 ) dr
dt e =±
te r c

Given our choice of coordinates, the light moves from r to 0 and so it moves in
the negative r direction. We thus choose negative sign in the ±
Z t0 Z 0
−H0 (t−t0 ) dr
dt e =−
te r c

(We could have also put the galaxy at r = 0 and us at r = r. In that case light
moves in the positive direction, so we take the plus sign, and the limits of the
integral on the right hand side would be flipped, so the end result is the same).
Doing the integrals Z t0
r
e H 0 t0 dt e−H0 t =
te c
t0
−1 −H0 t r
e H 0 t0 e =
H0 te c
 H0 r
−eH0 t0 e−H0 t0 − e−H0 te =

c
H0 r
−1 + eH0 (t0 −te ) =
c
H0 r
eH0 (t0 −te ) = 1 +
c
Taking the log of both sides
 
r
H0 (t0 − te ) = ln 1 +
cH0

13
  
1 H0 r
(t0 − te ) = ln 1 +
H0 c
 
1 H0 r
te = t0 − ln 1 +
H0 c
We check the units; 1/H0 is the Hubble time, so has units of time, and the
quantity inside the log is dimensionless, so this works out.
6b) The redshift of a photon is related to the scale factor by 1 + z = 1/a(te ).
Show that in this universe, the redshift z of an observed photon is linearly
proportional to the comoving coordinate r from which it was emitted.

Solution: Given the expression given for a(t) it’s value at te is

a(te ) = eH0 (te −t0 )

so
1
1+z = = e−H0 (te −t0 ) = eH0 (t0 −te )
a(te )
Above we found that
H0 r
eH0 (t0 −te ) = 1 +
c
so
rH0
1+z =1+
c
or
rH0
z=
c
which is linear in r

14
Problem 7: In class we showed that the innermost stable circular orbit
around a (non-spinning) black hole of mass M occurs at a Schwarzchild coor-
dinate r = 3rs . Consider a planet of mass m in a circular orbit around a black
hole of mass M . The orbit lies in the equatorial plane (θ = π/2) and at ra-
dial coordinate r = 3rs . We’ll use the Schwarzchild metric written in standard
Schwarzchild coordinates
 rs  dr2
ds2 = −c2 dt2 1 − + + r2 dΩ2
r (1 − rs /r)

We discussed in class two constants of motion for this metric, E/m (energy per
unit mass) and L/m (angular momentum per unit mass).
7a) An observer using the standard Schwarzchild coordinates measures the
planet to move in the ϕ direction with coordinate speed vϕ = rdϕ/dt. Use the
chain rule
dϕ dϕ dτ
= (5)
dt dτ dt
and the definitions of the constants of motion to find an expression for vϕ in
terms of L, E, rs , and c.

Solution: We can relate the terms in the expression for dϕ/dt to the constants
of motion. The angular momentum constant of motion is
L dϕ
= r2
m dτ
From which we have
dϕ L
=
dτ mr2
The energy constant of motion is
E  rs  2 dt
= 1− c
m r dτ
From which we have
dτ mc2  rs 
= 1−
dt E r
Putting these together we have

dϕ dϕ dτ L mc2  rs 
= = 2
1−
dt dτ dt mr E r
So
dϕ L c2  rs 
vϕ = r = 1−
dt E r r
Using r = 3rs
L c2 2 2 L c2
vϕ = =
E 3rs 3 9 E rs

15
 p
By the way, it turns out that L/E = 27/8rs /c
r
27 rs 2c2 c
vϕ = =√
8 c 9rs 6

7b) The period of the orbit (in Schwarzchild coordinates) is ∆torb = C/vϕ
where C is the circumference of the orbit. Find the expression for C and use it
to write the period ∆torb in terms of L, E, rs and c.

Solution: Looking at the metric, we see the angular term is r2 dΩ2 and r is
a circumferential coordinate. So the circumference here is just C = 2πr = 6πrs
so
6πrs E 9rs rs2 E
T = = 6πrs = 27π
vϕ L 2c2 c2 L

7c) The proper time describes the invariant spacetime distance of some path
through spacetime. Let ∆τorb be proper time of the path of a single orbit of
the planet. Find an expression relating ∆τorb to ∆torb , (Note that dϕ ̸= 0 for
this path in Schwarzchild coordinates.)
comment: In class we have found that the energy and angular momentum
constants of motion for a circular orbit at r = 3rs take on the specific values
√
r
8 2
E= mc L = 3mrs c (6)
9
√
And if you plugged these into your expression
√ you’d find ∆torb = 2 ∆τorb . So
for each year you spend on this planet, 2 ≈ 1.4 years elapse for someone far
away from the black hole.

Solution: There are several ways to address this, and different approaches
may lead to different looking expressions that are equivalent and acceptable.
Perhaps the most straightforward approach is to write the metric in terms
of proper time, dτ 2 = −ds2 /c2
ds2  rs  dr2 r2
dτ 2 = − 2
= dt2 1 − − 2 − 2 dϕ2
c r c (1 − rs /r) c
For a circular orbit dr = 0 and in the equatorial plane dϕ = 0, ϕ = π/2 and this
is  rs  r2 2
dτ 2 = dt2 1 − − 2 dϕ
r c
For something held stationary, dϕ = 0 and we get a gravitational time-dilation
formula we have looked at before. But here the planet moves in the ϕ direction.
As we did on the homework, however, we can pull out a factor of dt2 from the
right hand side "  2 #
2 2
 rs  r2 dϕ
dτ = dt 1− − 2
r c dt

16
which is " #
2 2
 rs  vϕ2
dτ = dt 1− − 2
r c
So " #1/2
 rs  vϕ2
dτ = dt 1− − 2
r c

Plugging in 1 − rs /r = 2/3
" #1/2
2 vϕ2
dτ = dt − 2
3 c

We can integrate this over an orbit, which is trivial since the term in brackets
is a constants
" #1/2
2 vϕ2
∆τorb = ∆torb − 2
3 c
√
with vϕ given above. If we wanted to check this answer, we can use vϕ = c/ 6
and the√term in brackets 2/3 − 1/6 = 3/6 = 1/2 so we get the result ∆τorb =
∆torb / 2.

If we didn’t think to pull out the dt2 factor, we could have used the common
technique of dividing through by dτ 2 . In which case the metric becomes
2  2
rs  r2
 
dt dϕ
1= 1− − 2
dτ r c dτ

And we can use the angular momentum constant of motion to replace the dϕ/dτ
2 
L2

dt rs 
1= 1− − 2 2 2
dτ r m c r

Using rs /r = 1/3 to simplify

2
L2

dt 2
1= −
dτ 3 9m2 c2 rs2

Rearranging
  2
L2

3 dt
1+ =
2 9m2 c2 rs2 dτ
which gives
−1/2
L2

3
dτ = dt +
2 6m2 c2 rs2

17
And integrating
−1/2
L2

3
∆τorb = ∆torb +
2 6m2 c2 rs2
The expression looks different than
√ above, but is equivalent. We can check by
plugging in the value given L = 3mcrs for which this becomes
 −1/2
3 3
∆τorb = ∆torb + = ∆torb (2)−1/2
2 6

Which is the expected result.

Note that we could have come to the same sort of argument by using the
chain rule to write dϕ = (dϕ/dτ )dτ in which case the metric becomes
 rs  r2 dϕ2 2
dτ 2 = dt2 1 − − 2 2 dτ
r c dτ
and then replacing dϕ/dτ by the constant of motion we could have done a bit
of algebra to find an expression relating dτ and dt

Another, more direct approach to this problem would be to just note that the
relation between dt and dτ is given by the constant of motion E
E dt  rs 
= 1 − (7)
mc2 dτ r
which implies
mc2  rs 
dτ = dt 1−
E r
and integrating
mc2  rs 
∆τorb = 1− ∆torb
E r
p
If we plug in r = 3rs and E/mc2 = 8/9
r
92 1
∆τorb = ∆torb = √ ∆torb
83 2

A related, but less direct, approach could use the angular momentum constant
of motion
dϕ dϕ dt dt
L = mr2 = mr2 = mrvϕ
dτ dt dτ dτ
So
mrvϕ 3mrs c vϕ
dτ = dt = dt
L L c

18
 √ √
and since L = 3mrs c and vϕ = c/ 6
√
3 1
dτ = √ dt = √ dt
6 2

Problem 8: A tortoise and a hare had a race (through flat spacetime) along
a straight line of length L. The race started at t = 0, and the tortoise moved at
a constant velocity v finishing at tf = L/v (in the earth frame). The hare stayed
at rest at the starting line until the tortoise reached the halfway mark (L/2),
and then rushed at faster constant speed to reach the finish line at exactly the
same time as the tortoise.
a) Draw a spacetime diagram of the race, showing the spacetime paths of the
tortoise and the hare.

Solution: The spacetime diagram is

b)Who will have aged more during the race, the tortoise or the hare? Give a
reason why.

Solution: The tortoise and hare both started and ended at the same point in
spacetime. However, the tortoise followed a spacetime geodesic (which we saw
in class was a constant dx/dτ ) and geodesics maximize the proper time. So the
tortoise took the “long way” through spacetime, and aged more during the race.
c) Find an expression for how much the hare aged during the race, in terms of
v and tf .

19
Solution:
The proper time interval for flat spacetime is

ds2 dx2
dτ 2 = − 2
= dt2 − 2 (8)
c c
which gives
r r r
dx2 1 dx2 v2
dτ = dt2 − 2 = dt 1 − 2 2 = dt 1 − 2 (9)
c c dt c
The total proper time for the tortoise is
Z Z tf p p
τT = dτ = dt 1 − β 2 = tf 1 − β 2 (10)
0

where we used the fact that β = v/c is constant for the tortoise.
To calculate the proper time for the hare, we split it into two segments, the
time when the hare is at rest (βH = 0) and the time when it is moving at twice
the speed of the tortoise βH = 2β
Z tf /2 Z tf
tf tf p
q
τH = dt + 2 =
1 − βH + 1 − 4β 2 (11)
0 rf /2 2 2

If we wanted to, we could show that τT > τH from these expressions.

Problem 9: Imagine you live in a (made up) spacetime described by the

metric
ds2 = −c2 e2r/a dt2 + e2r/a dr2 + r2 dΩ2 (12)
where a is a constant. You draw a circle centered at r = 0, and measure its
proper distance circumference, C, with a measuring tape. You then measure
the proper distance R from the center of the circle to its edge.
a) Find an expression for the circumference C of a circle centered at r = 0 in
terms of the radius R and the constant a.

Solution: We are looking for proper distances so set dt = 0. The circumference

is Z 2π
C= rdϕ = 2πr (13)
0
The radius is Z r r h i
R= er/a dr = aer/a = a er/a − 1 (14)
0 0

so using r = C/2π we have a relationship between C and R of

h i
R = a eC/2πa − 1 (15)

20
We can solve for C
eC/2πa = 1 + R/a (16)
C/2πa = ln(1 + R/a) (17)
C = 2πa ln(1 + R/a) (18)

b) Is this space around r = 0 positively curved, negatively curved, or flat?

Show how you determined this fact (you could use the Taylor expansion ln x ≈
x − x2 /2 + ... for x ≪ 1

Solution: We expand the circumference expression

R2
 
R
C = 2πa ln(1 + R/a) ≈ 2πa − 2 + .. (19)
a 2a

πR2
C = 2πR − + ... (20)
a
And we see that C < 2πR so this is a positively curved space...
b) Consider an observer A located at the center of the circle r = 0 and an
observer B located a proper distance R away. Whose clock runs more slowly?
By what factor (in terms of R/a)?

Solution: We want the proper time, where dr = dΩ = 0 so

−ds2
dτ 2 = = e2r/a dt2 (21)
c2
So
dτ = er/a dt (22)
We can see that dτ gets larger as r increases, so the ticks of the clock will get
larger with r and we expect the clock to run slower for B. Specifically, for
observer A at at r = 0, dτA = dt while observer B at r, dτB = er/a dt the ratio
is
dτB
= er/a (23)
dτA
So dτB > dτA and observer B’s clock runs more slowly. To get the actual factor
we would have to use the result above that
h i
R = a er/a − 1 (24)

In which case
R R
= er/a − 1 =⇒ er/a = 1 + (25)
a a
So
dτB R
=1+ (26)
dτA a

21
Problem 10: Say the metric of the universe is

ds2 = −c2 dt2 + (t/t0 )dr2 (27)

where t0 is a constant equal to the time now. We receive a light beam from a
source that is a proper distance dp away from us now. How long ago was the
light emitted?

Solution: For light ds2 = 0 so

c2 dt2 = (t/t0 )dr2 (28)

dr
dt(t/t0 )−1/2 = (29)
c
Integrating both sides gives
Z t0 Z r0
dt(t/t0 )−1/2 = dr/c (30)
te re

where te and re are the time of emission and the comoving coordinate of emis-
sion, and t0 and r0 are the time of observation and the comoving coordinate of
observation. Doing the integral
 1/2 t0
t r0 − re
2t0 = (31)
t0 te c
" 1/2  1/2 #
t0 te ∆r
2t0 − = (32)
t0 t0 c
where ∆r = r0 − re is the distance between the source and observer in comoving
coordinates. The proper distance is dp = ∆ra(t) and at the present time a(t0 ) =
1 so ∆r = dp . Then we can write
 1/2
te 1 dp
1− = (33)
t0 2 ct0
 2
te 1 dp
= 1− (34)
t0 2 ct0
 2
1 dp
te = t0 1 − (35)
2 ct0

Problem 11: (24 pts) Consider a flat universe filled with only one component
of energy, which has an equation of state parameter w = 2/3.
11a) Derive an expression that gives the scale factor, a(t), in terms of t and
the Hubble constant H0 , with initial condition a(t = 0) = 0.

22
Solution:
We use the Friedmann equation in the form
ȧ2 Ω0
= H02 3(1+w)
a2 a
Here w = 2/3 so 3(1 + w) = 3(1 + 2/3) = 3(5/3) = 5. Given that the universe
is flat, Ω0 = 1. Thus
ȧ2 H2 H2 da
2
= 50 =⇒ ȧ = 30 =⇒ = H0 a−3/2
a a a dt
Separating variables and integrating
Z a Z t
3/2
da a = H0 dt
0 0

2 5/2
a = H0 t
5
5
a5/2 = H0 t
2
 2/5
5
a(t) = H0 t
2

11b) Derive an expression for the age of this universe at the present time, t0 ,
in terms of the Hubble constant H0 .

Solution: To do this we use the constraint that a(t0 ) = 1 at the present time
 2/5
5
1= H0 t 0
2
5
1= H0 t0
2
2 1
t0 =
5 H0

11c) You observe light from a galaxy in this universe and measure a redshift
z = 3. At what time was this light emitted, te , in terms of the present time t0 ?

Solution: The relationship between redshift and scale factor is

a(t0 ) 1
1+z = =
a(te ) a(te )

Using our result above that a(t) = (t/t0 )2/5

 −2/5
te
1+z =
t0

23
  
te
= (1 + z)−5/2
t0
 
te 1 1
= (4)−5/2 = 2−5 = 5
=
t0 2 32

11d) Calculate the proper distance (at the present time) to the galaxy at
redshift z = 3, in terms of c and t0 . (You need not bother combining fractions
to simplify the numerical expressions, unless you want to).

Solution: The proper distance is given by dp = a(t)r where r is the comoving

coordinate. At the present time a(t0 ) = r so dp = r. We use the fact that light
emitted at time te at a galaxy at coordinate r = 0 arrives at us at time t0 and
coordinate r. Since light moves on on a null geodesic, ds2 = 0

0 = −c2 dt2 + a(t)2 dr2

cdt = ±a(t)dr
where the ± depends on which direction the light is traveling. We put the
galaxy at r = 0 and us at r and so the light moves in the +r direction and we
take the plus sign. Separating variables and integrating
Z t0 Z r
dt
c = dr
te a(t) 0

.Plugging in our expression for a(t) = (t/t0 )2/5

Z t0
c dt(t/t0 )−2/5 = r
te

0 t
5
c t0 (t/t0 )3/5 = r
3 te
"  3/5 #
5 te
c t0 1 − =r
3 t0
So we found the expression for the comoving coordinate r, which is itself the
proper distance dp . Using the above result that te /t0 = 1/25 for z = 3 this is
"  3/5 #
5 1
dp = ct0 1−
3 25

If we wanted to (but not necessary) we could clean things up a bit

 
5 −3
 5 1 57 35
dp = ct0 1 − 2 = ct0 1 − = ct0 = ct0
3 3 8 38 24

24
11e) The cosmic horizon is the furthest distance from which an observer may
have received light. Calculate the proper-distance of the cosmic horizon of this
universe as a function of time t.

Solution: The cosmic horizon comes from light emitted at t = 0. Returning

to our integral above for general time t
Z t  −2/5
t
dt = r/c
0 t0
 3/5
5 t
ct0 =r
3 t0
The proper distance at general time t is
 2/5  2/5  3/5
t t 5 t 5 t
dp (t) = a(t)r = r= ct0 = ct0
t0 t0 3 t0 3 t0
5
dp (t) = ct
3

Problem 12: (12 pts) In the flat, static spacetime of special relativity, we
define the energy, Esr , and the (radial) momentum, psr , of an object of rest mass
m to be
Esr dt psr dr
= c2 = (36)
m dτ m dτ
where τ is the proper time (we used the subscript sr to denote that these
are the definitions in special relativity). Consider now the general relativistic
Robertson-Walker metric that describes a dynamical universe

ds2 = −c2 dt2 + a(t)2 dr2 (37)

where we set dΩ = 0 as we will only consider radial motion.

12a) Explain why neither Esr nor psr as defined in Equation (??) are constants
of motion for an object moving along a geodesic in a dynamical RW metric.

Solution: We learned that if a metric does not depend on some spacetime

coordinate, x, then the quantity gx dx/dτ is a constant of motion, where gx is
the metric element in front of the dx2 term. The RW metric depends on the
spacetime coordinate t, so the quantity E/m will not be a constant of motion.
The RW does not depend on the spacetime coordinate r, and so there is a
constant of motion (call it p/m) related to the r coordinate which is
p dr
= a(t)2
m dτ
which means
psr dr p
= = a(t)−2
m dτ m

25
and thus psr changes with time because of the a(t) term and so is not a COM.
12b) Using the RW metric, find an equation that describes how the Esr defined
in Eq. (??) changes as the universe expands. Your equation should give Esr in
terms of a(t) and constants (both fundamental constants and true constants of
motion of the particle).

Solution: We write the metric in terms of proper time dτ 2 = −ds2 /c2

a2 2
dτ 2 = dt2 − dr
c2
Dividing through by dτ we have
 2  2
dt a2 dr
1= − 2
dτ c dτ
As usually, we use our constants of motion to reduce the equation. We found
above dr/dτ = p/ma(t)2 , where p/m is a COM, so plugging this in
 2 2
a2 p 1

dt
1= − 2
dτ c m a2
To get a term like Esr we multiply through by m2 c4
 2
dt
2 4
m c =m c 2 4
− c2 a−2 p2
dτ
which is
(mc2 )2 = Esr
2
− a−2 c2 p2
Rearranging
2 p2 c2
Esr = (mc2 )2 +
a(t)2
Comment: This could be compared to the special relativistic equation for energy

E 2 = (mc2 )2 + p2 c2

The expression in the RW metric is similar, but we notice that as the universe
expands (i.e., a(t) gets bigger), the momentum term gets smaller. This might
be thought of as a “expansion damping” – as spacetime grows underneath you,
it appears like you are not getting as far per unit time, and so your speed and
hence kinetic energy is dropping. In the limit a(t) → ∞, Esr becomes simply
the rest mass energy mc2 .

Problem 13: Some cosmologists have postulated that sheets of energy called
“cosmic domain walls” may be scattered throughout the universe1 . These do-
main walls would act as a fluid with equation of state parameter w = −2/3.
1 Domain walls are topological defects that may develop during phase transitions during

symmetry breaking as the universe expands

26
Imagine that we lived in a flat universe dominated by the energy density of
domain walls.
13a) Solve for the scale factor of the universe a(t) for this universe, with initial
condition a(t = 0) = 0.

Solution: We use the Friedmann equation for a flat universe

ȧ2
= H02 a−3(1+w) (38)
a2
which becomes for w = −2/3
ȧ2 da
2
= H02 a−1 =⇒ = H0 a1/2 (39)
a dt
Solving Z Z
−1/2
daa = H0 dt =⇒ 2a1/2 = H0 t (40)

And so we have  2
H0 t
a(t) = (41)
2

13b) In this present day universe, you emit a light signal out radially. What
is the maximum comoving distance r that your signal could reach given infinite
time? Write your answer in terms of the Hubble constant H0 and fundamental
constants.

Solution: For light on a null geodesic ds2 = 0 the Robertson-Walker metric

gives
ds2 = −c2 dt2 + a2 (t)dr2 = 0 =⇒ cdt = a(t)dr (42)
Rearrange and integrating from us (assumed to be at r = 0) to some furthest
distance r0 that light traveled from its emission time at t0 to infinity
Z r0 Z ∞
c
dr = dt (43)
0 t0 a(t)
Integrating Z ∞ ∞
4 4 1 4c
r0 = c dt = −c = (44)
t0 H02 t2 H02 t t0 H02 t0
We need to solve for the time now t0 . From a(t) = 1 we have
 2
H0 t 0
a(t0 ) = = 1 =⇒ t0 = 2/H0 (45)
2
So my maximum comoving distance is
2c
r0 = (46)
H0

27