Sie sind auf Seite 1von 3


Homework #6
F 22 March
Due: F 29 March by 9 PM

1. Read the following article. (To open it press on the hyperlink in the word “article”
in the previous sentence. In a few sentences, summarize what have you leaned from
it, and how it complements what you have learned from Sayiang’s presentation during
one of the Cosmo in the News session.

2. In this exercise you will calculate the curvature for the simplest curved space, a two-
sphere, a.k.a S 2 . The line element for a two sphere of radius a is, in spherical coordi-
nates (what else are we going to use?): ds2 = a2 (dθ2 + sin2 θdφ2 )

(a) Calculate the non-zero Christoffel symbols.

(b) How many different independent, i.e. not related to each other via the symmetries
of the Riemann tensor, non zero components will the Riemman curvature tensor
have? Do not calculate any of them yet, just do the counting, similar to the one
we have done in class for the 4D case. Now your manifold is S 2 , a 2D manifold.
(c) Now calculate the independent components of the Riemann tensor, with all lower
(d) Calculate the non-zero components of the Ricci tensor.
(e) Without calculating, guess, up to numerical constants, what would the Ricci scalar
be, and justify your guess based on dimensional analysis. Basically there is only
one parameter describing our metric, the radius of the sphere a. So R ∼ ap . What
is p, using dimensional analysis.
(f) Bonus Points can you also find a way to guess, and justify the numerical pre-factor
relating R to ap .
(g) Calculate the Ricci scalar and see if it matches your expectations based on Part 2e.

3. Do not forget to attach a printout of your Mathematica notebook for this problem! As
you have seen from Problem 2, even for the simplest curved space, calculating the
Riemann tensor takes quite a bit of algebra. In this exercise you will learn how to use
Mathematica to avoid doing this sort of manipulations in the future. From now on we
will start focusing on the physics, and leave the symbolic algebra for the computer.
First download the file under the name Riemann and Ricci tensors with Mathematica
in the Additional Resources tab of our Moodle site, at the very bottom. Edit that
file appropriately, to calculate, for the same two-spehere of radius a, considered in
Problem 2:

(a) The non-zero Christoffel symbols


(b) The Riemann tensor components. Note that this code will calculate the Rρµν
components, not the Rσρµν . You should find two non-zero components, if you
modified the code appropriately.
(c) Use the metric to and your result from Part. 3b to calculate the non-zero com-
ponents of the Riemann tensor, when all indices are lower. Compare with your
result from Part. 2c. They should be the same.
(d) Finally, find the Ricci scalar.
4. In this problem you will demonstrate an important identity of the variation of the
Crhistoffel symbols, called the Palatini identity. Under a variation of the metric:
gµν (x) → gµν (x) + δgµν (x)
(a) Show that variation of the Christoffel symbols δΓσµν transforms as a tensor. We
know that the Crhistoffel symbols do not transform as tensors, themselves, how-
ever it turns out that their variation does. Again, a proof that sometimes two
wrongs can make a right, when subtracted.
(b) Show that the following identity holds:
δRµσλ = ∇σ δΓρµλ − ∇λ δΓρµσ
This is called the Palatini Identity. To show it, it is most efficient if you work it
out in a locally Lorentz frame (LLF), and then use the fact that if an equality
between tensors holds in any coordinate system. Namely if A = B is an equality
between two tensors you have found in a reference frame, the same equality should
hold in any reference frame. The LLF is very useful, because it greatly simplifies
the algebra.
5. In this problem you will be guided through the variation of the Hilbert-Einstein action
and find Einstein’s field equations in the absence of matter, sometimes also referred to
as Einstein’s field equations in vacuum. In class, just by using the power of tensors,
we have justified that:
√ √
δSHE = δ d x −gR = d4 x −g(ARµν + Bg µν R)δgµν , (1)

where A and B are, yet undetermined numerical constants. Using the variational
procedure we will now find A, and B. We have three terms to worry about in the
variation, since:

√ √ √ √ √
δ( −gR) = δ( −gRµν g µν ) = δ( −g)Rµν g µν + −gδRµν g µν + −gRµν δg µν

(a) Prelude 1. Show that δg = gg µν δgµν . You might find it useful to use following,
very useful, matrix identity: ln(det(M )) = T r(ln(M )), where M is any square

(b) Prelude 2. Show that δg σρ = −g σµ δgµν g νρ .

(c) Use the result from Part 5a to express δ −g in terms of δgµν .

(d) Focus now on the last term −gRµν δg µν . Use the result from Part 5b to put it
in its simplest form.
(e) Now for the real work. We have avoided long enough to tackle the second term,
dancing around it, but there is no way out now. We have to deal with it. Use the
Palatini identity from Problem 4, show that:

d4 x −gδRµν g µν = 0, (2)

(f) Put everything you found so far together and identify the values of A and B from
Eq. 1
(g) Based on the result in the previous part, write down the tensorial equation that
describes the dynamics of the metric field gµν (x) in an empty space, i.e. if S =
SHE . Congratulations, you have just derived Einstein’s field equations in vacuum!