Phys 514 - Assignment 6

Solutions
Prepared by Bryce Cyr
April 3, 2018

1. The metric for the three-sphere in coordinates (ψ, θ, φ) is

ds2 = dψ 2 + sin2 (ψ)(dθ2 + sin2 (θ)dφ2 ) (1)

a) Calculate the Christoffel symbols

b) Calculate the Riemann tensor, Ricci tensor, and Ricci scalar.

Solution
a) Recall that

1 σρ
Γσµν = g (∂µ gνρ + ∂ν gρµ − ∂ρ gµν )
2

And with our metric, the nonzero metric elements are

gψψ = 1 gθθ = sin2 ψ gφφ = sin2 ψ sin2 θ

We will approach this systematically, by check the upper index of the connection coefficients first.

1 ψψ
Γψ
µν = g (∂µ gνψ + ∂ν gψµ − ∂ψ gµν )
2

The first two terms are 0 since derivatives of gψψ vanish. This gives us two choices for the last term,
µ = ν = θ and µ = ν = φ. These yield the coefficients

Γψ
θθ = − sin ψ cos ψ

Γψ 2
φφ = − sin ψ cos ψ sin θ

That’s it for the upper ψ index. Lets move onto θ. We have

1 θθ
Γθµν = g (∂µ gνθ + ∂ν gθµ − ∂θ gµν )
2

Lets start with the last term. This is only nonzero if µ = ν = φ. This yields the symbol

Γθφφ = − sin θ cos θ

1
This takes care of the φ indices. The only other way to get a nonzero symbol is by µ = θ or ν = θ. In the
first instance, we get

1 θθ
Γθθν = g (∂θ gνθ + ∂ν gθθ − ∂θ gθν )
2

The first term and the final term are 0 by inspection. The middle term is nonzero if ν = ψ. By the symmetry
of the symbols we get

Γθθψ = Γθψθ = cot ψ

Finally, lets look at the last upper index. It is

1 φφ
Γφµν = g (∂µ gνφ + ∂ν gφµ − ∂φ gµν )
2

The final term is always 0 since no part of the metric is φ dependent. Now, either µ or ν must be φ to get
a nonzero symbol. The derivative will alternate between the symmetric indices, and so a straightforward
computation yields the final four symbols

Γφψφ = Γφφψ = cot ψ

Γφθφ = Γφφθ = cot θ

To recap, the nonzero Christoffel symbols here are

Γψ
θθ = − sin ψ cos ψ

Γψ 2
φφ = − sin ψ cos ψ sin θ

Γθφφ = − sin θ cos θ

Γθθψ = Γθψθ = cot ψ
Γφψφ = Γφφψ = cot ψ
Γφθφ = Γφφθ = cot θ

b) The Riemann tensor is defined as

ρ
Rσµν = ∂µ Γρνσ − ∂ν Γρµσ + Γρµλ Γλνσ − Γρνλ Γλµσ (2)

The techniques to find the Riemann tensor were illustrated in one of the solutions for assignment 4, and
so we will compute as many as my energy will allow for these Christoffel symbols. We recall the useful
expression

λ
Rρσµν = gλρ Rσµν

Which allows us to fully exploit the symmetries

Rρσµν = −Rρσνµ Rρσµν = −Rσρµν Rρσµν = Rµνρσ Rρ[σµν] = 0

2
With this in mind, lets start by computing the upper ψ index Riemann tensor indices. Since gψψ = 1, we
don’t have to do any conversion between the two forms. We get

ψ ψ
Rψσµν = ∂µ Γψ ψ λ λ
νσ − ∂ν Γµσ + Γµλ Γνσ − Γνλ Γµσ

Now, to keep the first term nonzero, we note that we have three options. First, we can set ν = σ = θ and
µ = ψ to find

Rψθψθ = ∂ψ (− sin ψ cos ψ) − ∂θ Γψ ψ λ ψ λ

ψθ + Γψλ Γθθ − Γθλ Γψθ

= sin2 ψ − cos2 ψ + cos2 ψ

= sin2 ψ

The middle two terms are 0 in the top line. By the symmetries, we can write a few more terms

Rψθψθ = sin2 ψ Rθψψθ = Rψθθψ = − sin2 ψ Rψψθθ = 0

As our second check, we note that choosing ν = σ = φ and µ = ψ yield a nonzero first term as well. This
expression is

Rψφψφ = ∂ψ (− sin ψ cos ψ sin2 θ) − ∂φ Γψ ψ λ ψ λ

ψφ + Γψλ Γφφ − Γφλ Γψφ

= sin2 θ(sin2 ψ − cos2 ψ) + cos2 ψ sin2 θ

= sin2 θ sin2 ψ

The symmetries then yield

Rψφψφ = sin2 θ sin2 ψ Rφψψφ = Rψφφψ = − sin2 θ sin2 ψ Rψψφφ = 0

Finally, the last option to keep the first term nonzero is ν = σ = φ with µ = θ This yields

Rψφθφ = ∂θ (− sin ψ cos ψ sin2 θ) − ∂φ Γψ ψ λ ψ λ

θφ + Γθλ Γφφ − Γφλ Γθφ
= −2 sin ψ cos ψ sin θ cos θ + sin ψ cos ψ sin θ cos θ + sin ψ cos ψ sin θ cos θ
=0

So this one vanishes, as well as all its forms related by symmetries.

Now we move to the second term, ∂ν Γψ µσ . We have multiple choices again, so first, set µ = σ = θ and
ν = ψ. This means we would be computing Rψθθψ . We already know what this is from above! Its simply
− sin2 θ sin2 ψ, thus no computation is necessary. Moving on, we can choose µ = σ = φ and ν = ρ. This is
the computation of Rψφφψ which is also done. Finally, the last thing that keeps the second term nonzero
is µ = σ = φ with ν = θ. This computes Rψφφθ , which is 0 by an antisymmetric transformation of our
previous result. Thus, we didn’t have to perform any explicit calculation for this term!

Moving now to the third term,

Γψ λ ψ ψ ψ θ ψ φ ψ θ ψ φ
µλ Γνσ = Γµψ Γνσ + Γµθ Γνσ + Γµφ Γνσ = 0 + Γµθ Γνσ + Γµφ Γνσ

3
We have options here, first, we can set µ = θ and ν = σ = φ. This will mean we compute Rψφθφ We have
computed this already, and its 0. Next option, set µ = θ, ν = θ, and σ = ψ. This computation is of Rψψθθ ,
also 0. Our next choice is µ = φ, ν = ψ, σ = φ. This is computing Rψφφψ = − sin2 θ sin2 ψ from symmetry.
Finally, we can also make the choice µ = φ, ν = θ, σ = φ. This yields Rψφφθ , which is 0 once again by
symmetry.

Lets move onto the final term, expanded it is

Γψ λ ψ ψ ψ θ ψ φ ψ θ ψ φ
νλ Γµσ = Γνψ Γµσ + Γνθ Γµσ + Γνφ Γµσ = 0 + Γνθ Γµσ + Γνφ Γµσ

We have four choices again, so lets speed through them. First, choose ν = θ, µ = σ = φ. This corresponds
to computing Rψφφθ , again 0. Next choose ν = θ, µ = θ, σ = ψ. This computes Rψψθθ = 0. Next choice
would be µ = φ, ν = ψ, σ = φ. This computes Rψφφψ , already computed above. Finally, choose ν = φ,
µ = θ, σ = φ. This computes Rψφθφ = 0.

This concludes the computation of the Riemann symbols for the upper index of ψ. As you can see, (and
I’m sure you know after doing it yourself) it is a bit of a slow process. I will quote the results of the other
nonzero Riemann tensor elements here, but rest assured they come from the exact same process as above
but for upper indices θ and φ instead. Our nonzero elements are

Rψθψθ = Rθψθψ = sin2 ψ Rθψψθ = Rψθθψ = − sin2 ψ

Rψφψφ = Rφψφψ = sin2 θ sin2 ψ Rφψψφ = Rψφφψ = − sin2 θ sin2 ψ
Rθφθφ = Rφθφθ = sin2 θ sin4 ψ Rφθθφ = Rθφφθ = − sin2 θ sin4 ψ

ρ
To get back to the proper form, we take the transformation Rσµν = g ρλ Rλσµν , where we have

1 1
g ψψ = 1 g θθ = g φφ =
sin2 ψ sin2 ψ sin2 θ

With these transforms, we get our final Riemann tensor elements.

ψ ψ ψ ψ
Rθψθ = sin2 ψ Rθθψ = − sin2 ψ Rφψφ = sin2 θ sin2 ψ Rφφψ = − sin2 θ sin2 ψ
θ
Rψθψ =1 θ
Rψψθ = −1 θ
Rφθφ = sin2 θ sin2 ψ θ
Rφφθ = − sin2 θ sin2 ψ
φ φ φ φ
Rψφψ =1 Rψψφ = −1 Rθφθ = sin2 ψ Rθθφ = − sin2 ψ

λ
The Ricci tensor is determined by a contraction of the Riemann tensor, Rµν = Rµλν Since we know this
is a symmetric tensor in three dimensions, we can calculate the six independent components individually
without much work.

ψ θ φ
Rψψ = Rψψψ + Rψθψ + Rψφψ
=0+1+1
=2

ψ θ φ
Rθθ = Rθψθ + Rθθθ + Rθφθ
= sin2 ψ + 0 + sin2 ψ
= 2 sin2 ψ

4
 ψ θ φ
Rφφ = Rφψφ + Rφθφ + Rφφφ
= sin2 ψ sin2 θ + sin2 ψ sin2 θ
= 2 sin2 ψ sin2 θ

ψ θ φ
Rψθ = Rψψθ + Rψθθ + Rψφθ
=0+0+0
=0

ψ θ φ
Rψφ = Rψψφ + Rψθφ + Rψφφ
=0+0+0
=0

ψ θ φ
Rθφ = Rθψφ + Rθθφ + Rθφφ
=0+0+0
=0

So the independent components of our Ricci tensor are

Rψψ = 1
Rθθ = 2 sin2 ψ
Rφφ = 2 sin2 ψ sin2 θ
Rψθ = Rψφ = Rθφ = 0

Our Ricci scalar is defined as R = g µν Rµν . Using the inverse metric written above, this yields

R = g ψψ Rψψ + g θθ Rθθ + g φφ Rφφ

=2+2+2
=6

5
2. Do the same calculation using the tetrad basis.

Solution
Recall our line element from the previous problem

ds2 = dψ 2 + sin2 ψdθ + sin2 ψ sin2 θdφ2

To make use of the tetrad formalism, we would like to have a line element that looks like

ds2 ea eb δab

Where a, b run from 1 to 3. This allows us to define our tetrad basis. It will be

eψ = dψ eθ = sin ψdθ eφ = sin ψ sin θdφ

Note that we are now in a noncoordinate basis. The appendix J of the book provides a good background on
the application of the tetrad formalism, so refer to it if you are having any confusion. We wish to compute
the Riemann tensor, which by equation J.29 is

Rba = dωba + ωca ∧ ωbc (3)

If we expect to find this, we had better start by first computing the spin connection, ω. Note that the
Riemann tensor in the above expression has been expressed in a basis of one forms, so Rba = Rbµν
a
dxµ dxν .
We can determine the spin connection by the expression

dea = eb ∧ ωba

Note that the spin connections are antisymmetric (see appendix J), so ωaa = 0. Clearly we need the
differential of our basis to compute the spin connections, but at least that we can do! This basis is

deψ = 0
deθ = cos ψdψ ∧ dθ
deφ = cos ψ sin θdψ ∧ dφ + sin ψ cos θdθ ∧ dφ

Lets write down our three equations now, and deduce the elements of the spin connection.

0 = sin ψdθ ∧ ωθψ + sin ψ sin θdφ ∧ ωφψ

cos ψdψ ∧ dθ = dψ ∧ ωψθ + sin ψ sin θdφ ∧ ωφθ
cos ψ sin θdψ ∧ dφ + sin ψ cos θdθ ∧ dφ = dψ ∧ ωψφ + sin ψdθ ∧ ωθφ

From the third line, comparing the left and right hand sides, we can see that ωθφ = cos θdφ and ωψφ =
cos ψ sin θdφ. Now from the second line we can see that ωψθ = cos ψdθ. Since the spin connection is a 3 × 3
antisymmetric object, there are only three independent components, which we have found. For clarity, they
are

6
 ωθφ = cos θdφ
ωψφ = cos ψ sin θdφ
ωψθ = cos ψdθ

In order to compute the Riemann tensor, we must take the differential of these objects as well. They are

dωθφ = − sin θdθ ∧ dφ

dωψφ = − sin ψ sin θdψ ∧ dφ + cos ψ cos θdθ ∧ dφ
dωψθ = − sin ψdψ ∧ dθ

From here, we can finally compute the Riemann tensor, Rba = dωba + ωca ∧ ωbc . We note this object is also
antisymmetric in a and b, so we only need to compute the three independent components once again.

Rθψ = dωθψ + ωφψ ∧ ωθφ

= sin ψdψ ∧ dθ − cos θ sin θ cos ψdφ ∧ dφ
= sin ψdψ ∧ dθ

Rφψ = dωφψ + ωθψ ∧ ωφθ

= sin ψ sin θdψ ∧ dφ − cos ψ cos θdθ ∧ dφ + cos θ cos ψdθ ∧ dφ
= sin ψ sin θdψ ∧ dφ

Rφθ = dωφθ + ωψθ ∧ ωφψ

= sin θdθ ∧ dφ − cos2 ψ sin θdθ ∧ dφ
= sin θ(1 − cos2 ψ)dθ ∧ dφ
= sin θ sin2 ψdθ ∧ dφ

Our Riemann tensor is thus

0
Rθψ0 = sin ψdψ ∧ dθ
0
Rφψ0 = sin ψ sin θdψ ∧ dφ
0
Rφθ 0 = sin θ sin2 ψdθ ∧ dφ

Where I have introduced the prime notation to specify that these are in the noncoordinate basis. We wish
to switch back to the coordinate basis now, since the difficult part of finding the tensor is taken care of.
The expression to do so is (equation J.49 in the book)

ρ
Rσµν = eρa ebσ Rbµν
a

Recall that we had expressed everything on a basis of one-forms, so we had ea = eaµ dxµ . This means that
the ebσ is the coefficient attached to the differential in our tetrad coordinates, and eρa the reciprocal (since

7
the metric is diagonal). Note also that the differentials in the wedge product are our basis one-forms, µ and
0
ν, so for example, Rθψ0 ψθ = sin ψdψ ∧ dθ. Lets compute the upper ψ Riemann metric elements now.

0
ψ
Rσµν = eψ b ψ
ψ 0 eσ Rbµν
0 0 0 0 0 0
= eψ ψ ψ ψ θ ψ ψ φ ψ
ψ 0 eσ Rψ 0 µν + eψ 0 eσ Rθ 0 µν + eψ 0 eσ Rφ0 µν

We note that since the tetrad basis is diagonal in the coordinate basis (no cross-terms), we only have three
ψ0
nonzero, diagonal e terms. Since Rψ 0 = 0, this yields the two Riemann tensor symbols from the last two

terms above as

0 0
ψ
Rθµν = eψ θ ψ
ψ 0 eθ Rθ 0 µν
= (1)(sin ψ)(sin ψdψ ∧ dθ)
ψ
Rθψθ = sin2 ψ

For one of them. The other given by the φ term

0 0
ψ
Rφµν = eψ φ ψ
ψ 0 eφ Rφ0 µν
= (1)(sin ψ sin θ)(sin ψ sin θdψ ∧ dφ)
ψ
Rφψφ = sin2 ψ sin2 θ

Moving onto upper θ terms yields

0 0
θ
Rψµν = eθθ0 eψ θ
ψ Rψ 0 µν
 
1
= (1)(− sin ψdψ ∧ dθ)
sin ψ
θ
Rψψθ = −1

and

0 0
θ
Rφµν = eθθ0 eφφ Rφθ 0 µν
 
1
= (sin ψ sin θ)(sin θ sin2 ψdθ ∧ dφ)
sin ψ
θ
Rφθφ = sin2 ψ sin2 θ

Lastly, we compute the upper φ components

0 0
φ
Rψµν = eφφ0 eψ φ
ψ Rψ 0 µν
 
1
= (1)(− sin ψ sin θdψ ∧ dφ)
sin ψ sin θ
φ
Rψψφ = −1

and finally

8
 0 0
φ
Rθµν = eφφ0 eθθ Rθφ0 µν
 
1
= (sin ψ)(− sin θ sin2 ψdθ ∧ dφ)
sin ψ sin θ
φ
Rθθφ = − sin2 ψ

Putting all these symbols together, and exploiting the antisymmetry of the final two indices yields a full
Riemann tensor of

ψ ψ ψ ψ
Rθψθ = sin2 ψ Rθθψ = − sin2 ψ Rφψφ = sin2 θ sin2 ψ Rφφψ = − sin2 θ sin2 ψ
θ
Rψθψ =1 θ
Rψψθ = −1 θ
Rφθφ = sin2 θ sin2 ψ θ
Rφφθ = − sin2 θ sin2 ψ
φ φ φ φ
Rψφψ =1 Rψψφ = −1 Rθφθ = sin2 ψ Rθθφ = − sin2 ψ

Exactly as we had in the previous problem, thus the Ricci tensor and scalar are obviously the same as
before.

9
3. Consider 3-dimensional (i.e. 2 + 1) gravity.
a) How many degrees of freedom are in the Riemann tensor?
b) How many degrees of freedom are in the Ricci tensor?
c) The Riemann tensor can be decomposed into the Ricci tensor and the Weyl tensor (see textbook, Page
130). Using this fact, how many degrees of freedom are in the Weyl tensor?
d) Are there gravity waves (gravity waves are fluctuations of space-time without associated matter pertur-
bations)?

Solution
a) The first the realize when approaching this problem, is that the number of degrees of freedom (or analo-
ρ
gously, the number of independent components) is the same in either Rσµν and Rρσµν . The latter expression
has its symmetries manifest, so lets consider how many degrees of freedom are in that object.

Please refer to pg 128 of the textbook for a great derivation on the number of independent degrees of freedom
for the Riemann tensor in n dimensional spacetime. The end result is

1 2 2
D.O.F. = n (n − 1)
12

For 3 dimensional spacetime, we get that the Riemann tensor has 6 degrees of freedom.

b) The Ricci tensor is a symmetric 2 tensor. In three dimensional spacetime, the Ricci tensor has 9 elements,
and since the tensor is symmetric, this leaves 6 independent degrees of freedom.

c) The Weyl tensor in n dimensions is defined as

2 2
Cρσµν = Rρσµν − (gρ[µ Rν]σ − gσ[µ Rν]ρ ) + gρ[µ gν]σ R
n−2 (n − 1)(n − 2)

The easiest way to see the number of degrees of freedom in the Weyl tensor is to note the decomposition.
The question states that you can decompose the Riemann tensor into the Weyl tensor and the Ricci tensor.
Since we know from above that the Riemann tensor has 6 degrees of freedom, and the Ricci tensor also has
6 degrees of freedom, this leaves 0 degrees of freedom for the Weyl tensor to possess. Thus, the Weyl tensor
has no degrees of freedom and must vanish in 2 + 1 dimensions.

Cρσµν = 0

d) Since the Weyl tensor is 0, let us rewrite the Riemann tensor in terms of the Ricci tensor and scalar. We
get (in n = 3 dimensions)

Rρσµν = 2(gρ[µ Rν]σ − gσ[µ Rν]ρ ) − gρ[µ gν]σ R

Now, in vacuum we know that Einstein’s equations reduce down to

1
Gµν = Rµν − gµν R = 0
2
1
Rµν = gµν R
2

Now we recall that the Ricci scalar is defined as R = g µν Rµν . Multiply both sides of the above equation by
g µν to find

10
 1 µν
g µν Rµν =g gµν R
2
3
R= R
2

Where we have used the fact g µν gµν = 3 in three dimensions, by definition. This is clearly only satisfied if
R = 0. This implies that Rµν = 0 as well. Now, since the Riemann tensor is defined by the Ricci tensor
and scalar, it too must vanish.

The Riemann tensor is a measure of curvature. Having it vanish implies that there is no curvature, so
gµν = ηµν where ηµν is the Minkowski metric. Recall that gravitational waves can be investigated by
perturbing a Minkowski metric, such that

gµν → ηµν + hµν

Where hµν are small perturbations ontop of a Minkowski background. Since we have just shown that the
metric goes exactly to a Minkowski one, there can be no extra perturbations, and so it is impossible to get
gravitational waves within this theory.

11
4. Find the solution to Einstein’s field equations for a massive point particle (at rest) in 2 + 1 dimensional
gravity. Hint: Write the stress-tensor for a point particle, make an ansatz for the metric making use of the
symmetries of the problem (diagonal, static, polar symmetry), and then find the resulting Riemann tensor.
Then solve for the metric.

Solution
The stress energy tensor for a point particle is T 00 = mδ 2 (~x) in 2 + 1 dimensions.

As an ansatz, we take our metric to be

ds2 = e2A(r) dt2 − dr2 − e2B(r) dφ2

Where we have eliminated the parameter attached to the r variable in the same method as used in Carroll,
page 194. A, B are arbitrary functions of r, and we will drop the r brackets in what follows. To find the
Riemann tensor, we make use of the tetrad formalism. Our tetrad basis is defined as

θt = eA dt θr = dr θφ = eB dφ

Their differentials are

dθt = A0 eA dr ∧ dt dθr = 0 dθφ = B 0 eB dr ∧ dφ

Where A0 = dA/dr. To deduce the spin connections, we must solve dθa = θb ∧ ωba . Noting the antisymmetry
of ω, we can write down our three equations

dθt = A0 eA dr ∧ dt = dr ∧ ωrt + eB dφ ∧ ωφt

dθr = 0 = eA dt ∧ ωtφ + eB dφ ∧ ωφr
dθφ = B 0 eB dr ∧ dφ = eA dt ∧ ωtφ + dr ∧ ωrφ

From these three equations, we can deduce the elements of the spin connection and its exterior derivative

ωrt = A0 eA dt dωrt = eA (A00 + A02 )dr ∧ dt

ωφt = 0 dωφt = 0
ωφr = −B 0 eB dφ dωφr = −eB (B 00 + B 02 )dr ∧ dφ

To find the elements of the Riemann tensor, we use Rba = dωba + ωca ∧ ωbc . Doing so yields

0
Rrt 0 = eA (A00 + A02 )dr ∧ dt
0
Rφt 0 = −A0 B 0 eA+B dt ∧ dφ
0
Rφr 0 = −eB (B 00 + B 02 )dr ∧ dφ

Now we can find the nonzero components of the Riemann tensor in a coordinate basis, using the expression
ρ
Rσµν = θaρ θσb Rbµν
a
, where we make note that

θaρ = diag(e−A , 1, e−B ) θσb = diag(eA , 1, eB )

12
So lets compute each element.

0 0
t
Rrµν = θtt0 θrr Rrt 0 µν
= (e−A )(1)(A00 + A02 )eA dr ∧ dt)
t
Rrtr = −A00 − A02

0 0
t
Rφµν = θtt0 θφφ Rφt 0 µν
= (e−A )(eB )(−A0 B 0 )eA+B dt ∧ dφ
t
Rφtφ = −e2B A0 B 0

0 0
r
Rφµν = θrr0 θφφ Rφr 0 µν
= (1)(eB )(−(B 00 + B 02 )eB dr ∧ dφ)
r
Rφrφ = −e2B (B 00 + B 02 )

Where the other symbols are related by symmetries and the metric. Lets compute the Ricci tensor. The
metric and inverse metric are

gµν = diag(e2A , −1, −e2B ) g µν = diag(e−2A , −1, −e−2B )

So our Ricci tensor is

t r φ
Rtt = Rttt + Rtrt + Rtφt
= 0 + gtt g rr Rrtr
t
+ gtt g φφ Rφtφ
t

= e2A (A00 + A0 B 0 + A02 )

t r φ
Rrr = Rrtr + Rrrr + Rrφr
t
= Rrtr + 0 + grr g φφ Rφrφ
r

= −(A00 + A02 + B 00 + B 02 )

t r φ
Rφφ = Rφtφ + Rφrφ + Rφφφ
t r
= Rφtφ + Rφrφ +0
= −e2B (A0 B 0 + B 00 + B 02 )

And our Ricci scalar

R = g µν Rµν = g tt Rtt + g rr Rrr + g φφ Rφφ

= 2(A00 + A02 + B 00 + B 02 + A0 B 0 )

13
Calculate Einstein tensor Gµν = Rµν − 12 gµν R

Gtt = e2A (A00 + A0 B 0 + A02 ) − e2A (A00 + A02 + B 00 + B 02 + A0 B 0 )

= −e2A (B 00 + B 02 )

Grr = −(A00 + A02 + B 00 + B 02 ) − (−1)(A00 + A02 + B 00 + B 02 + A0 B 0 )

= A0 B 0

Gφφ = −e2B (A0 B 0 + B 00 + B 02 ) − (−e2B )(A00 + A02 + B 00 + B 02 + A0 B 0 )

= e2B (A00 + A02 )

From the stress tensor, we know that the Grr and Gφφ components must be zero, for r nonzero, this implies
that A0 = 0. We can also set A to 0. This satisfies the two lower equations. Now, the first equation is

Gtt = 8πGTtt = 8πGg tt g tt T tt

With A = 0 we have gtt = g tt = 1, so we must solve

−B 00 − B 02 = 8πmGδ 2 (~r)

We note that the elemental area in polar coordinates is eB drdφ, and so we can write the normalization of
the delta function as

Z ∞ Z 2π
δ 2 (~r)eB drdφ = 1
0 0

Now, define a new variable Λ = eB so that Λ00 = (B 00 + B 02 )eB . The temporal part of the Einstein tensor
becomes

Λ00 e−B = −8πmGδ 2 (~r)

Clearly from this, δ 2 (~r)eB = −Λ00 /8πmG so from the normalization we have

∞ 2π
Λ00
Z Z
1=− drdφ
0 0 8πmG
1
=− (Λ0 (∞) − Λ0 (0))
4mG

So

Λ0 (∞) = Λ0 (0) − 4mG

Since Λ0 is constant, we can say that Λ = Dr. Integrating not to infinity but to r yields

14
 Λ0 (r) − Λ0 (0) = −4mG

So Λ0 will be discontinuous across the origin. so we can take Λ(r) = dr for r 6= 0 and Λ0 (0) = 1 to find
Λ(r) = (1 − 4mG)r. With this, we have solved the metric. It is

ds2 = dt2 − dr2 − (1 − 4mG)2 r2 dφ2

15
5. In class I sketched the derivation of the Einstein tensor for the spherically symmetric metric

ds2 = e2a(r) dt2 − [e2b(r) dr2 + r2 dΩ2 ] (4)

using the tetrad formalism. Complete the derivation.

Solution
This will follow the same structure as problem 2. The full metric is (for clarity)

ds2 = e2a(r) dt2 − [e2b(r) dr2 + r2 dθ2 + r2 sin2 θdφ2 ]

Our tetrad basis here is

σ t = ea(r) dt σ r = eb(r) dr σ θ = rdθ σ φ = r sin θdφ

The differentials of this basis are thus

dσ t = a0 (r)ea(r) dr ∧ dt
dσ r = b0 (r)eb(r) dr ∧ dr = 0
dσ θ = dr ∧ dθ
dσ φ = sin θdr ∧ dφ + r cos θdθ ∧ dφ

Where we note that a0 (r) = da(r)/dr. Now as before, we need to find the spin connections, ω. The
expression is dσ a = σ b ∧ ωba . Writing the four equations that we get yields

dσ t = a0 ea dr ∧ dt = eb dr ∧ ωrt + rdθ ∧ ωθt + r sin θdφ ∧ ωφt

dσ r = 0 = ea dt ∧ ωtr + rdθ ∧ ωθr + r sin θdφ ∧ ωφr
dσ θ = dr ∧ dθ = ea dt ∧ ωtθ + eb dr ∧ ωrθ + r sin θdφ ∧ ωφθ
dσ φ = sin θdr ∧ dφ + r cos θdθ ∧ dφ = ea dt ∧ ωtφ + eb dr ∧ ωrφ + rdθ ∧ ωθφ

Now from these equations, and the fact that ωba = −ωab , we can deduce the six independent elements of the
spin connection. From the last equation, we can easily deduce ωrφ = e−b sin θdφ and ωθφ = cos θdφ. From
the third equation we find that ωrθ = e−b dθ. Now, from the first equation we can deduce ωrt = a0 ea−b dt.
That’s it for the easy to compute elements of the spin connection. We currently have

ωrφ = e−b sin θdφ

ωθφ = cos θdφ
ωrθ = e−b dθ
ωrt = a0 ea−b dt

We are still missing ωθt and ωφt . From the first line, ωθt is either 0 or proportional to dθ. From the third
line, ωθt is either 0 or proportional to dt. To satisfy both constraints, we must have ωθt = 0. By the same
reasoning, we find that ωφt = 0, so our full spin connection is characterized by

16
 ωrφ = e−b sin θdφ
ωθφ = cos θdφ
ωrθ = e−b dθ
ωrt = a0 ea−b dt
ωθt = ωφt = 0

To find the Riemann tensor, we have to now find the differentials of this. They are

dωrφ = −b0 e−b sin θdr ∧ dφ + e−b cos θdθ ∧ dφ

dωθφ = − sin θdθ ∧ dφ
dωrθ = −b0 e−b dr ∧ dθ
dωrt = a00 ea−b dr ∧ dt + a0 (a0 − b0 )ea−b dr ∧ dt = (a00 + (a0 )2 − a0 b0 )ea−b dr ∧ dt
dωθt = dωφt = 0

The Riemann tensor is defined by

Rba = dωba + ωca ∧ ωbc

and is antisymmetric, meaning we must compute only the six independent elements once again. Let us
begin

Rrt = dωrt + ωθt ∧ ωrθ + ωφt ∧ ωrφ

= (a00 + (a0 )2 − a0 b0 )ea−b dr ∧ dt + 0 + 0
= (a00 + (a0 )2 − a0 b0 )ea−b dr ∧ dt

Rθt = dωθt + ωrt ∧ ωθr + ωφt ∧ ωθφ

= 0 + a0 ea−b dt ∧ (−e−b dθ) + 0
= a0 ea−2b dθ ∧ dt

Rφt = dωφt + ωrt ∧ ωφr + ωθt ∧ ωφθ

= 0 + a0 ea−b dt ∧ (−e−b sin θdφ) + 0
= a0 sin θea−2b dφ ∧ dt

Rθr = dωθr + ωtr ∧ ωθt + ωφr ∧ ωθφ

= b0 e−b dr ∧ dθ + 0 + (−e−b sin θdφ) ∧ (cos θdφ)
= b0 e−b dr ∧ dθ

17
 Rφr = dωφr + ωtr ∧ ωφt + ωθr ∧ ωφθ
= b0 e−b sin θdr ∧ dφ − e−b cos θdθ ∧ dφ + 0 + (−e−b dθ) ∧ (− cos θdφ)
= b0 e−b sin θdr ∧ dφ

Rφθ = dωφθ + ωtθ ∧ ωφt + ωrθ ∧ ωφr

= sin θdθ ∧ dφ + 0 + e−b dθ ∧ (−e−b sin θdφ)
= sin θ(1 − e−2b )dθ ∧ dφ

Ok, let us list our Riemann tensor elements once more for clarity (in the tetrad basis).

0
Rrt 0 = (a00 + (a0 )2 − a0 b0 )ea−b dr ∧ dt
0
Rθt 0 = a0 ea−2b dθ ∧ dt
0
Rφt 0 = a0 sin θea−2b dφ ∧ dt
0
Rθr0 = b0 e−b dr ∧ dθ
0
Rφr 0 = b0 e−b sin θdr ∧ dφ
0
Rφθ 0 = sin θ(1 − e−2b )dθ ∧ dφ

We have primed the variables to distinguish between coordinate and tetrad basis now. We now want to
switch back to the coordinate basis to compute the rest of our objects. We switch back with the usual
expression

0
t
Rσµν = ett0 ebσ Rbµν
t

0 0 0 0 0 0 0 0
= ett0 etσ Rtt0 µν + ett0 erσ Rrt 0 µν + ett0 eθσ Rθt 0 µν + ett0 eφσ Rφt 0 µν

Where we have illustrated the formula to solve for an upper t, for clarity. Recall that in our tetrad basis,
all the e’s are diagonal, explicitly they are

0 0 0 0
ett = ea(r) err = eb(r) eθθ = r eφφ = r sin θ
1 1
ett0 = e−a(r) err0 = e−b(r) eθθ0 = eφφ0 =
r r sin θ

In coordinate basis, our Riemann tensor becomes

0 0
t
Rrµν = ett0 err Rrt 0 µν
= (e−a )(eb )(a00 + (a0 )2 − a0 b0 )ea−b dr ∧ dt
t
Rrrt = a00 + (a0 )2 − a0 b0

0 0
t
Rθµν = ett0 eθθ Rθt 0 µν
= (e−a )(r)(a0 ea−2b dθ ∧ dt)
t
Rθθt = ra0 e−2b

18
 0 0
t
Rφµν = ett0 eφφ Rφt 0 µν
= (e−a )(r sin θ)(a0 sin θea−2b dφ ∧ dt)
t
Rφφt = ra0 sin2 θe−2b

0 0
r
Rtµν = err0 ett Rtr0 µν
= (e−b )(ea )(−(a00 + (a0 )2 − a0 b0 )ea−b dr ∧ dt)
r
Rtrt = −e2(a−b) (a00 + (a0 )2 − a0 b0 )

0 0
r
Rθµν = err0 eθθ Rθr0 µν
= (e−b )(r)(b0 e−b dr ∧ dθ)
r
Rθrθ = rb0 e−2b

0 0
r
Rφµν = err0 eφφ Rφr 0 µν
= (e−b )(r sin θ)(b0 e−b sin θdr ∧ dφ)
r
Rφrφ = rb0 sin2 θe−2b

0 0
θ
Rtµν = eθθ0 ett Rtθ0 µν
 
1
= (ea )(−a0 ea−2b dθ ∧ dt)
r
1
θ
Rtθt = − a0 e2(a−b)
r

0 0
θ
Rrµν = eθθ0 err Rrθ0 µν
 
1
= (eb )(−b0 e−b dr ∧ dθ)
r
1
θ
Rrrθ = − b0
r

0 0
θ
Rφµν = eθθ0 eφφ Rφθ 0 µν
 
1
= (r sin θ)(sin θ(1 − e−2b )dθ ∧ dφ)
r
θ
Rφθφ = sin2 θ(1 − e−2b )

0 0
φ
Rtµν = eφφ0 ett Rtφ0 µν
 
1
= (ea )(−a0 sin θea−2b dφ ∧ dt)
r sin θ
φ 1
Rtφt = − a0 e2(a−b)
r

19
 0 0
φ
Rrµν = eφφ0 err Rrφ0 µν
 
1
= (eb )(−b0 e−b sin θdr ∧ dφ)
r sin θ
φ 1
Rrrφ = − b0
r

0 0
φ
Rθµν = eφφ0 eθθ Rθφ0 µν
 
1
= (r)(− sin θ((1 − e−2b )dθ ∧ dφ)
r sin θ
φ
Rθθφ = −(1 − e−2b )

So our full coordinate Riemann tensor is thus

t
Rrtr = −(a00 + (a0 )2 − a0 b0 ) t
Rθtθ = −ra0 e−2b t
Rφtφ = −ra0 sin2 θe−2b
t
Rrrt = (a00 + (a0 )2 − a0 b0 ) t
Rθθt = ra0 e−2b t
Rφφt = ra0 sin2 θe−2b
r
Rtrt = e2(a−b) (a00 + (a0 )2 − a0 b0 ) r
Rθrθ = rb0 e−2b r
Rφrφ = rb0 sin2 θe−2b
r
Rttr = −e2(a−b) (a00 + (a0 )2 − a0 b0 ) r
Rθθr = −rb0 e−2b r
Rφφr = −rb0 sin2 θe−2b
1 1
θ
Rtθt = − a0 e2(a−b) θ
Rrθr = b0 θ
Rφθφ = sin2 θ(1 − e−2b )
r r
1 0 2(a−b) 1
θ
Rttθ = a e Rrrθ = − b0
θ θ
Rφφθ = − sin2 θ(1 − e−2b )
r r
φ 1 φ 1 0 φ
Rtφt = − a0 e2(a−b) Rrφr = b Rθφθ = (1 − e−2b )
r r
φ 1 0 2(a−b) φ 1 φ
Rttφ = a e Rrrφ = − b0 Rθθφ = −(1 − e−2b )
r r

The Einstein tensor is Gµν = Rµν − 12 gµν R so we need to compute the Ricci tensor and scalar. The Ricci
λ
tensor is Rµν = Rµλν . Lets do so

t r θ φ
Rtt = Rttt + Rtrt + Rtθt + Rtφt
1 1
= 0 + e2(a−b) (a00 + (a0 )2 − a0 b0 ) − a0 e2(a−b) − a0 e2(a−b)
r r
2a0
 
2(a−b) 00 0 2 0 0
=e a + (a ) − a b +
r

t r θ φ
Rrr = Rrtr + Rrrr + Rrθr + Rrφr
1 1
= −(a00 + (a0 )2 − a0 b0 ) + 0 + b0 + b0
r r
00 0 2 0 0 2b0
= −a − (a ) + a b +
r

t r θ φ
Rθθ = Rθtθ + Rθrθ + Rθθθ + Rθφθ
= −ra0 e−2b + rb0 e−2b + 0 + 1 − e−2b
= 1 + e−2b (r(b0 − a0 ) − 1)

20
 t r θ φ
Rφφ = Rφtφ + Rφrφ + Rφθφ + Rφφφ
= −ra0 sin2 θe−2b + rb0 sin2 θe−2b + sin2 θ(1 − e−2b )
= sin2 θ(1 + e−2b (r(b0 − a0 ) − 1) = sin2 θRθθ

Finally, the Ricci scalar is R = g µν Rµν

R = g tt Rtt + g rr Rrr + g θθ Rθθ + g φφ Rφφ

2a0 2b0
   
−2a 2(a−b) 00 0 2 0 0 −2b 00 0 2 0 0
=e e a + (a ) − a b + −e −a − (a ) + a b +
r r
1 1
− 2 (1 + e−2b (r(b0 − a0 ) − 1)) − 2 2 (sin2 θ(1 + e−2b (r(b0 − a0 ) − 1))
r r sin θ
2(a0 − b0 ) 1 − e2b
 
= 2e−2b a00 + (a0 )2 − a0 b0 + +
r r2

Since Rµν = 0 outside the spherically symmetric object, we can solve for the coefficients in the same way
as page 196 of Carroll. This yields

 −1
2a Rs 2b Rs
e =1− e = 1−
r r

Where Rs comes from an undetermined constant of integration, and is interpreted as the Schwarzschild
radius.

The Einstein tensor is Gµν = Rµν − 12 Rgµν . Since we know (and used) the fact that Rµν = 0, a straightfor-
ward computation in mathematica shows that R = 0 as well. The Einstein tensor is thus

Gµν = 0

As expected for the vacuum outside a spherically symmetric source.

21
6. Derive the Einstein tensor for the metric of the above problem, this time using the coordinate approach.
After solving this problem you should be convinced that it is easier to use the tetrad formalism.

Solution
For this problem, we will just compute the Christoffel symbols, showing that they match those given in
equation (5.12) of Carroll, and leave the rest as an exercise, as we have gone from symbols to Riemann
tensor elements many times in previous problems during the course. The line elements is

ds2 = e2A dt2 − e2B dr2 − r2 dθ2 − r2 sin2 θdφ2

gtt = e2A grr = −e2B gθθ = −r2 gφφ = −r2 sin2 θ

The Christoffel symbols are defined by

1 σρ
Γσµν = g (∂µ gνρ + ∂ν gρµ − ∂ρ gµν )
2

As usual, lets start with an upper t index.

1 tt
Γtµν = g (∂µ gνt + ∂ν gtµ − ∂t gµν )
2

The final term is 0 since there is no time dependence on the metric. In this cas we can take µ = r and ν = t
(or vice versa) to find

1 −2A
Γtrt = Γttr = e (2A0 e2A )
2
= A0

That’s all for this upper index. Next lets look at upper r

1 rr
Γrµν = g (∂µ gνr + ∂ν grµ − ∂r gµν )
2

There are many options here. For the first two terms, to be nonzero we can set µ = ν = r to find

1 rr 1
Γrrr = g ∂r grr = (−e−2B )(−2B 0 e2B ) = B 0
2 2

Now we get a term for each µ = ν = t, θ, φ (we already did the case where it was r). In each case, only the
−2B
last term contributes and the symbols are Γrµµ = − 21 g rr ∂r gµµ = e 2 ∂r gµµ

Γrtt = A0 e2(A−B)
Γrθθ = −re−2B
Γrφφ = −r sin2 θe−2B

22
Now lets consider an upper θ. This gives us the equation

1 θθ
Γθµν = g (∂µ gνθ + ∂ν gθµ − ∂θ gµν )
2
The first two terms are symmetric, and require either µ = r, ν = θ, or vice versa. This yields

1 θθ 1
Γθrθ = Γθθr = g ∂r gθθ = − r−2 (−2r)
2 2
1
=
r
For the last term, we require that µ = ν = φ to be nonzero, and so we find

1 −2
Γθφφ = r (−2r2 sin θ cos θ) = − sin θ cos θ
2
Now lets move onto the last terms, with upper index φ

1 φφ
Γφµν = g (∂µ gνφ + ∂ν gφµ − ∂φ gµν )
2
The last term is always 0. The first two terms are nonzero for µ = r, ν = φ AND µ = θ, ν = φ (and vice
versa). This leads to the symbols

1 φφ 1
Γφrφ = Γφφr = g ∂r gφφ = (−r−2 sin−2 θ)(−2r sin2 θ)
2 2
1
=
r
1 φ 1
Γφθφ = Γφφθ = g ∂θ gφφ = (−r−2 sin−2 θ)(−2r2 sin θ cos θ)
2 2
= cot θ

To recap, our symbols were

Γtrt = Γttr = A0
Γrrr = B 0
Γrtt = A0 e2(A−B)
Γrθθ = −re−B
Γrφφ = −r sin2 θe−2B
1
Γθrθ = Γθθr =
r
Γθφφ = − sin θ cos θ
1
Γφrφ = Γφφr =
r
Γφθφ = Γφφθ = cot θ

These 9 independent symbols match (5.12) of Carroll. From here it is straightforward (and very tedious) to
go to the Riemann tensor, and eventually to the Ricci tensor and scalar. If you have any specific questions
on those steps, feel free to email me at [email protected] and I can help you out, but I won’t be
continuing the full derivation here.

23