A. Polnarev. (SPA7019 ). 2014. Solutions to course work 1.

————————————————————————————————————————————————-

A. Solutions to course work 1

Q1
a) Formulate the equivalence principle and explain what is the difference in interpretation of this principle in Newtonian theory
and in General relativity.[5 marks]
Solution:
The equivalence principle states that a uniform gravitational field is equivalent, i.e. is not distinguishable from a
uniform acceleration.
It is explained within frame of newtonian theory just by the following ”coincidence”: inertial mass min is equal to
gravitational mass mgr .
The GR gives a very simple and natural explanation of the Principle of Equivalence: in curved space-time all bodies
move along geodesics, that is why their world lines are the same in given gravitational field.
b) Explain the similarity between an ”actual” gravitational field and a non-inertial reference system. Give the definition of a
locally Galilean coordinate system.[5 marks]
Solution:
The similarity is that for a given gravitational and at a given event one can always choose such a non-inertial frame
that the motion of particles or propagation of light with respect to this non-inertial frame in the vicinity of that event
(i.e. locally) is the same as in the given ”actual” gravitational field.
The local galilean frame of reference (the freely falling frame of reference) is a such frame in which

gik → ηik ≡ diag(1, −1, −1, −1). (A.1)

c) Explain why an ”actual” gravitational field cannot be eliminated by any transformation of coordinates over all space-time.[5
marks]
Solution:
In general ”actual” gravitational field is non-uniform and one should produce different transformations to the local
galilean frame of reference at different events (points) of space-time.
0 0 0
d) Show that in a uniformly rotating system of coordinates x , y , z , such that
0 0 0 0 0
x = x cos Ωt − y sin Ωt, y = x sin Ωt + y cos Ωt, z = z ,

the interval ds has the following form:

0 0 0 0 0 0 0 0 0 0
ds2 = gik dxi dxk = gik dx i dx k
= [c2 − Ω2 (x 2 + y 2 )]dt2 − dx 2 − dy 2
− dz 2
+ 2Ωy dx0 dt0 − 2Ωx dy 0 dt0 .

[10 marks]
Solution:

φ = Ωt = Ωt0 , x = x0 cosφ − y 0 sin φ, y = x0 sin φ + y 0 cosφ; (A.2)

dφ = Ωdt0 , dx = dx0 cosφ − dy 0 sin φ − (x0 sin φ − y 0 cos φ)dφ, dy = dx0 sin φ + dy 0 cosφ + (x0 cos φ − y 0 sin φ)dφ.(A.3)

Hence, ds2 = c2 dt2 − dx2 − dy 2 − dz 2 = c2 dt02 − [dx0 cosφ − dy 0 sin φ − (x0 sin φ − y 0 cos φ)dφ]2 −

−[dx0 sin φ + dy 0 cosφ + (x0 cos φ − y 0 sin φ)dφ]2 − dz 02 = cd t02 − dx02 − dy 02 − dφ2 (x02 + y 02 )+

+2dφ[(dx0 cosφ − dy 0 sin φ)(x0 sin φ − y 0 cos φ) − (dx0 sin φ + dy 0 cosφ)(x0 cos φ − y 0 sin φ)] − dz 02 =

0 0 0 0 0 0 0
= [c2 − Ω2 (x 2 + y 2 )]dt2 − dx 2 − dy 2 − dz 2 + 2Ωy dx0 dt0 − 2Ωx dy 0 dt0 . (A.4)

2
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 1.
————————————————————————————————————————————————-
Q2
a) Formulate the covariance principle and explain the relationship between this principle and the principle of equivalence.[5
marks]
Solution:
The Principle of Covariance says: the shape of all physical equations should be the same in an arbitrary frame of
reference.
The Principle of Covariance is more general and more mathematical version of the Principle of Equivalence.
b) Give the definition of a contravariant vector in terms of the transformation of curvilinear coordinates.[5 marks]
Solution:
The Contravariant four-vector is the combination of four quantities (components) Ai , which are transformed like
differentials of coordinates:
∂xi 0k
Ai = A = Ski A0k . (A.5)
∂x0k

c) Give the definition of a covariant vector in terms of the transformation of curvilinear coordinates.[5 marks]
Solution:
The Covariant four-vector is the combination of four quantities (components) Ai , which are transformed like com-
ponents of the gradient of a scalar field:

∂x0k 0
Ai = A = S̃ik A0k . (A.6)
∂xi k

d) What is the mixed tensor of the second rank in terms of the transformation of curvilinear coordinates (you can assume that
a mixed tensor of the second rank is transformed as a product of covariant and contrvariant vectors).[5 marks]
Solution:
Mixed tensor of the 2 rank has 42 = 16 components and 2 indices, 1 contravariant and 1 covariant. Corresponding
transformation law is

Aik = Sni S̃km A0n

m, (A.7)

we see 2 transformation matrices in the transformation law. It really looks like transformation of the product of two
vectors

 
B i Ck = Sni B 0n S̃km Cm
0
= Sni S̃km B 0n Cm
0
. (A.8)

e) Explain why the principle of covariance implies that all physical equations should contain only tensors.[5 marks]
Solution:
The Principle of Covariance predetermines the mathematical structure of General Relativity: all equations should
contain tensors only, because tensors are objects which are transformed properly in the course of coordinate transfor-
mations from one frame of reference to another keeping the shape of any physical equation being unchanged.

3
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 1.
————————————————————————————————————————————————-
Q3
a) Prove that the metric tensor is symmetric. Give a rigorous proof that the interval is a scalar. [8 marks]
Solution:

1 1 1
ds2 = gik dxi dxk = (gik dxi dxk + gik dxi dxk ) = (gki dxk dxi + gik dxi dxk ) = (gki + gik )dxi dxk =
2 2 2

= g̃ik dxi dxk , (A.9)

where
1
(gki + gik ),
g̃ik = (A.10)
2
which is obviously a symmetric one. Then we just drop ” ˜ ”.
0
ds2 = gik dxi dxk = (S̃in S̃km gnm )(Spi dx0p )(Sw
k
dx0w ) = (S̃in Spi )(S̃km Sw
k 0
)(gnm dx0p dx0w ) =

m 0
= δpn δw (gnm dx0p dx0w ) = gpw
0
dx0p dx0w = gik
0
dx0i dx0k = ds02 , (A.11)
0
hence ds = ds which means that ds is a scalar.
b) Give the definition of the reciprocal tensors of the second rank. What is the contravariant metric tensor g ik .[5 marks]
Solution:
Two tensors Aik and B ik are called reciprocal to each other if
Aik B kl = δil . (A.12)
ik
A contravariant metric tensor g is reciprocal to the covariant metric tensor gik :
gik g kl = δil . (A.13)
c) Show that in an arbitrary non-inertial frame
g ik = S(0)0
i k
S(0)0 i
− S(0)1 k
S(0)1 i
− S(0)2 k
S(0)2 i
− S(0)3 k
S(0)3 ,
i
where S(0)k is the transformation matrix from locally inertial frame of reference (galilean frame) to this non-inertial frame.[7
marks]
Solution:
i
Let S(0)k be the transformation matrix from locally inertial frame of reference (galilean frame) to an arbitrary
non-inertial frame, let us denote it as . In the galilean frame of reference

1 0 0 0
 
0 −1 0 0 
g ik = η ik = ≡ η ik ≡ diag(1, −1, −1, −1), (A.14)
0 0 −1 0 
0 0 0 −1
hence

g ik = S(0)n
i k
S(0)m η lm = S(0)0
i k
S(0)0 i
− S(0)1 k
S(0)1 i
− S(0)2 k
S(0)2 i
− S(0)3 k
S(0)3 . (A.15)

d) Demonstrate how using the reciprocal contravariant metric tensor g ik and the covariant metric tensor gik you can form
contravariant tensor from covariant tensors and vice versa.[5 marks]
Solution:
With the help of the metric tensor and its reciprocal we can form contravariant tensors from covariant tensors and
vice versa, for example:
Ai = g ik Ak , Ai = gik Ak , (A.16)
in other words we can rise and descend indices as we like, like a kind of juggling with indices. We can say that
contravariant, covariant and mixed tensors can be considered as different representations of the same geometrical
object.

4
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 1.
————————————————————————————————————————————————-
Q4
a) In the local Galilean frame xi[G] of reference a mixed tensor of the second rank, Cki has the only one non-vanishing
0
component, C0[G] = 1, and all other components are equal to zero. Write down all components of this mixed tensor in arbitrary
frame of reference. Express your result in terms of transformation matrix.[10 marks]
Solution:

Cki = S(0)n
i m
S̃(0)k n
Cm[G] i
= S(0)n m
S̃(0)k δ0n δm
0 i
= S(0)0 0
S̃(0)k . (A.17)

b) In the non-rotating system of Cartesian coordinates (x, y, z) the only non-vanishing component of some tensor Aik is
A11 = 1 and all other components vanish. Using coordinate transformation from Cartesian to the uniformly rotating cylindrical
coordinates (r, θ, φ)
x = r cos(θ + Ωt), y = r sin(θ + Ωt), z = Z,
show that in the latter coordinates
rΩ
A01
0 = − sin 2(θ + Ωt).
2c
[15 marks]
Solution:

Aik = Sni S̃km A0n

m. (A.18)

Multiplying both sides of (A.18) by S0k S̃i1 one obtains

S0k S̃i1 Aik = Sni S̃i1 S0k S̃km A0n

m, (A.19)

hence

S0k S̃i1 δ1i δk1 = δn1 δ0m A0n 01

m = A0 . (A.20)

Thus
∂x1 ∂x01 ∂(r cos(θ + Ωt)) ∂r
A01 1 1
0 = S0 S̃1 = 00 1
= . (A.21)
∂x ∂x c∂t ∂x
Taking into account that
∂r
r2 = x2 + y 2 and 2r = 2x, (A.22)
∂x
one finally obtains
rΩ rΩ
A01
0 = [− sin(θ + Ωt)][cos(θ + Ωt)] = − sin 2(θ + Ωt). (A.23)
c 2c

5
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 2.
————————————————————————————————————————————————-

B. Solutions to course work 2

Q1.
a) What is meant by proper time and what is the relationship between the proper time and coordinate time?[10 marks]

Solution:
Proper time, τ , is defined as
ds
dτ = . (B.1)
c
For the proper time between any two events occurring at the same point in space (dxα = 0) we have
Z t
√ √
ds2 = g00 (dx0 )2 , hence, dτ = g00 dt, and τ = τ0 + g00 dt0 . (B.2)
t0

A light signal emitted at the moment corresponding to time coordinate x0 + ∆x0(1) propagates from some point B with spatial
coordinates xα + ∆xα to a point A with spatial coordinates xα and then, after reflection at the moment corresponding to time
coordinate x0 , the signal propagates back over the same path and is detected at point B at the moment corresponding to time
coordinate x0 + ∆x0(2) . Given that g0α = 0, express the physical distance between A and B, lAB , in terms of the metric tensor,
gik , and ∆xα . You may assume that gik is the same at points A and B.[15 marks]

Solution:
Separating the space and time coordinates in ds we have

ds2 = gαβ dxα dxβ + 2g0α dx0 dxα + g00 (dx0 )2 = gαβ dxα dxβ + g00 (dx0 )2 . (B.3)

The interval between the events which belong to the same world line of light in Special and General Relativity is
always equal to zero:

ds=0.

Solving this equation with respect to dx0 we find two roots:

1
q
dx0(1) = − −gαβ g00 dxα dxβ (B.4)
g00
and
1
q
dx0(2) = −gαβ g00 dxα dxβ , (B.5)
g00
hence
2
q
dx0(2) − dx0(1) = −gαβ g00 dxα dxβ . (B.6)
g00
Then
√
c c g00
dl = dτ = (dx0(2) − dx0(1) ), hence dl2 = −gαβ dxα dxβ (B.7)
2 2 c
and finally
Z A q
lAB = dl = −gαβ ∆xα ∆xβ . (B.8)
B

6
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 2.
————————————————————————————————————————————————-
Q2.
a) Motivate the necessity to introduce parallel translation for proper differentiation of tensors and explain the geometrical and
physical meaning of the Christoffel symbols.[15 marks]
Solution:
Differentiating, say, vectors we take difference between components of this vector at infinitesimally close but never-
theless different points. At different points in general case matrices of transformation are different and the differential
of vector (tensor) is not a vector (tensor). In order to make the differential of a tensor be also a tensor, we should
take difference of two objects at the same point. To do this we need produce a parallel translation.
The Christoffel symbols from geometrical point of view appear as coefficients in corrections to components of tensors
due to the parallel translations in curved space-time (in flat space-time such corrections obviously are equal to zero).
From physical point of view the Christoffel symbols represent non-inertial ”forces” in non-inertial frames of reference
or gravitational ”forces”.
b) List all physical and geometrical arguments, you know, to demonstrate that the Christoffel symbols do not form a tensor.[10
marks]
Solution:
1) Corrections for the parallel translations are added to non-tensors to obtain tensors, hence the are non-tensors
(geometrical argument);
2) Gravitational field can be eliminated locally, which means that in locally inertial frame of reference the Christoffel
symbols are equal to zero, while according to the definition of tensors, the tensors vanishing in one frame of reference
should vanish in any other frame of reference, hence the Christoffel symbols do not form a tensor (physical argument).
Q3
a) Write down the covariant derivative of the mixed tensor of the second rank in terms of Christoffel symbols.[10 marks]
Solution:
Aik;n = Aik,n + Γinm Am m i
k − Γkn Am . (B.9)
b) Explain why for the derivation of physical equations in the presence of a gravitational field one can simply replace partial
derivatives by covariant derivatives. Take any physical equation by your own choice and write down it in the presence of
gravitational field.[15 marks]
Solution:
According to the principle of covariance all physical equation should have the same shape in arbitrary frames of
reference, but in locally inertial frames where the Christoffel symbols are equal to zero partial derivatives are identical
to covariant derivatives.
For example, the energy-momentum conservation without gravity (in flat space time) and in inertial frame of reference
i i
Tk,i should be replaced by Tk;i .
Q4
Show by straightforward calculations that √
1 ∂g ∂ ln −g
Γiki = = .
2g ∂xk ∂xk
You can use here without proof that the differential of g can be expressed as
dg = gg ik dgik = −ggik dg ik .
[25 marks]
Solution:
1 in
Γiki =
g (gkn,i + gin,k − gki;n ). (B.10)
2
Replacing indices of summation in the first term in the brackets, i → n and n → i we obtain
1 ni 1 g − ,k 1
(g gki,n + g in gin,k − g in gki;n ) = g in gin,k = = (ln |g|),k . (B.11)
2 2 2g 2
Taking into account that g < 0, i.e. |g| = −g, we finally obtain
√
1 ∂g ∂ ln −g
Γiki = = . (B.12)
2g ∂xk ∂xk

7
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 3.
————————————————————————————————————————————————-

C. Solutions to course work 3

Q1
a) Give the definition of a test particle. Consider a test particle of mass m moving in arbitrary gravitational field. Derive the
Hamilton-Jacobi equation. Explain how using this equation one can obtain the world line of the test particle.[15 marks]
Solution:
Test particle by definition generates negligible gravitational field.
From

ds2 = gik dxi dxk , (C.1)

introducing the 4-velocity

dxi
ui = . (C.2)
ds
one obtains

gik ui uk = 1. (C.3)

Introducing then the 4-momentum

pi = mcui , (C.4)

one obtains

gik pi pk = m2 c2 gik ui uk = m2 c2 . (C.5)

Using covariant presentation of 4-momentum:

pi = g in pn , (C.6)

one obtains

gik g in pn g kv pv = δkn g kv pn pv = g nv pn pv = m2 c2 . (C.7)

Taking into account that there always exists such a scalar S that

pi = −S,i , (C.8)

one finally obtains the Hamilton-Jacobi equation for S:

g ik S,i S,k − m2 c2 = 0. (C.9)

As soon as the solution for S is obtained we have

pi 1 in 1 in
1) pi = −S,i , 2) pi = g in pn = −g in S,n , 3) ui = =− g S,n 4) dxi = − g S,n ds. (C.10)
mc mc mc
Finally
Z
1
xi (s) = − g in S,n ds. (C.11)
mc

8
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 3.
————————————————————————————————————————————————-
b) In some gravitational field the transformation from the local galilean frame of reference xi(0) to some non-inertial frame xi
i
is given by the following transformation matrix: S(0)k = F δki , where F is a scalar field. Consider propagation of light using
mentioned above non-inertial frame of reference to show that the solution of the Eikonal equation does not depend on F and
can be written as Ψ = (δ αβ Bα Bβ )1/2 x0 + Bα xα , where Bα is a constant 3-vector.[10 marks]
Solution:

g ik = S(0)n
i k
S(0)m ηnm = F 2 δni δm
k
ηnm = F 2 ηik , (C.12)
hence the Eikonal equation is
F 2 ηik Ψ,i Ψ,k = 0, hence, ηik Ψ,i Ψ,k = 0. (C.13)
One can try a solution in the following form:
Ψ = Ax0 + Bα xα . (C.14)
Substituting this into the Eikonal equation, one obtains
A2 − δ αβ Bα Bβ = 0, (C.15)
hence,
q
A=± δ αβ Bα Bβ (C.16)
p
and A = δ αβ Bα Bβ is one of these two solutions.
Q2
Using the locally-inertial coordinate system prove that the Riemann tensor has the following symmetry properties:
Riklm = −Rkilm = −Rikml ,
Riklm = Rlmik ,
Riklm + Rimkl + Rilmk = 0.
[25 marks]
Solution:
In the local galilean frame of reference
n

1 1
Riklm = ηin Rklm = ηin Γnkm,l − Γnkl,m = ηin [g np (gkp,m + gmp,k − gkm,p ],l − ηin [g np (gkp,l + glp,k − gkl,p ],m =
2 2
1
= ηin η np (gkp,m,l + gmp,k,l − gkm,p,l − gkp,l,m − glp,k,m + gkl,p,m ) =
2
1 p 1
= δ (gmp,k,l − gkm,p,l − glp,k,m + gkl,p,m ) = (gim,k,l + gkl,i,m − gil,k,m − gkm,i,l ) . (C.17)
2 i 2
1 1
Rkilm = (gkm,i,l + gil,k,m − gkl,i,m − gim,k,l ) = − (gim,k,l + gkl,i,m − gil,k,m − gkm,i,l ) = −Riklm . (C.18)
2 2
1 1
Rikml = (gil,k,m + gkm,i,l − gim,k,l − gkl,i,m ) = − (gim,k,l + gkl,i,m − gil,k,m − gkm,i,l ) = −Riklm . (C.19)
2 2
1 1
Rlmik = (glk,m,i + gmi,l,k − gli,m,k − gmk,l,i ) = (gim,k,l + gkl,i,m − gil,k,m − gkm,i,l ) = Riklm . (C.20)
2 2

Riklm + Rimkl + Rilmk = ηin Γnkm,l − Γnkl,m + Γnml,k − Γnmk,l + Γnlk,m − Γnlm,k =

= ηin Γnkm,l − Γnmk,l − Γnkl,m + Γnlk,m + Γnml,k − Γnlm,k =

= ηin Γnkm,l − Γnkm,l − Γnkl,m + Γnkl,m + Γnml,k − Γnml,k = 0.

(C.21)

9
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 3.
————————————————————————————————————————————————-
Q3
Show that
Rik = Γlik,l − Γlil,k + Γlik Γm m l
lm − Γil Γkm .
[25 marks]
Solution:
Starting from
i
Rklm = Γikm,l − Γikl,m + Γinl Γnkm − Γinm Γnkl , (C.22)
l
(which is obtained by straightforward calculation of Ak;l;m −Ak;m;l ), and using definition of the Ricci tensor Rik = Rilk ,
one obtains
n
Rik = Rink = Γnik,n − Γnin,k + Γnpn Γpik − Γnpk Γpin =

= Γlik,l − Γlil,k + Γlik Γm m l

lm − Γil Γkm . (C.23)

Q4
a) Using a locally-inertial coordinate system prove the Bianchi identity:
n n n
Rikl;m + Rimk;l + Rilm;k = 0.
[15 marks]
Solution:
The Bianchi identity,
n n n
Rikl;m + Rimk;l + Rilm;k = 0, (C.24)
in the local galilean frame of reference, where all Christoffel symbols are equal to zero, can be re-written as
n n n
Rikl,m + Rimk,l + Rilm,k =0 (C.25)
and the Riemann tensor in this frame can be written as
i
Rklm = Γikm,l − Γikl,m + Γinl Γnkm − Γinm Γnkl = Γikm,l − Γikl,m , (C.26)

n n n
= Γnil,k − Γnik,l + Γnim,l − Γnil,m + Γnik,m − Γnim,k




Rikl,m + Rimk,l + Rilm,k ,m ,k ,l
=

= Γnil,k,m − Γnik,l,m + Γnim,l,k − Γnil,m,k + Γnik,m,l − Γnim,k,l =

= [Γnil,k,m − Γnil,m,k ] + [Γnik,m,l − Γnik,l,m ] + [Γnim,l,k − Γnim,k,l ] = [0] + [0] + [0] = 0. (C.27)

b) Using the Bianchi identity, prove that the Ricci tensor and the scalar curvature R = g ik Rik satisfy the following identity:
l 1
Rm;l = Rm .
2
[10 marks]
Solution:
After contracting the Bianchi identity we obtain
g kl Rklm;i
i
+ g kl Rkil;m
i
+ g kl Rkmi;l
i
= g in g kl (Rnklm;i + Rnkil;m + Rnkmi;l ) =

= g kl g in (−Rknlm;i + Rnkil;m − Rnkim;l ) = −g in Rnm;i + g kl Rkl;m − g kl Rkm;l = −Rm;i

i l
+ R,m − Rm;l =

i
= −2Rm;i + R,m = 0.

10
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 4.
————————————————————————————————————————————————-

D. Solutions to course work 4

Q1
Starting from the Einstein equations in the form
1 8πG
Rik − gik R = 4 Tik , (D.1)
2 c
where G is the gravitational constant, prove that

c4
 
1
Tki = Rki − δki R . (D.2)
8πG 2

[20 marks]
Solution:
Contracting with g ik , we have the Einstein equations in mixed form
8πG i 1 i
Rki = (Tk − δk T ). (D.3)
c4 2

     
ik 8πG 1 8πG 1 8πG 1 8πG
R = g Rik = 4 g Tik − g ik gik T
ik
= 4 Tii − δii T = 4 T − 4 = − 4 T. (D.4)
c 2 c 2 c 2 c

Thus
c4
T =− R. (D.5)
8πG
Hence
c4
 
1
Tik = Rik − gik R , (D.6)
8πG 2

then in mixed form we have

c4
 
1 i
Tki = i
Rk − δk R . (D.7)
8πG 2

11
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 4.
————————————————————————————————————————————————-
Q2
The stress-energy tensor of ideal fluid has the following form:

Tki = ( + P )ui uk − P δki , (D.8)

where  is energy density, P is pressure (if P > 0) or tension (if P < 0) and ui is 4-velocity of the fluid. Using the Einstein
equations express the scalar curvature in terms of  and P .[20 marks]
Solution:
Contracting the EFEs written in mixed form (see the previous question)we have

m 1 m 8πG m
Rm − δm R = 4 Tm , (D.9)
2 c
hence
4 8πG 8πG
R − R = 4 T = 4 (ε − 3P ), (D.10)
2 c c
and
8πG
R − 2R = (ε − 3P ), (D.11)
c4
finally
8πG
R=− (ε − 3P ). (D.12)
c4

Q3
Prove that the covariant divergence of an arbitrary contravariant vector can be written as
1 √
Ai;i = √ ( −gAi ),i . (D.13)
−g

[20 marks]
Solution:

1
Ai;i = Ai,i + Γiin An = An,n + g im (gim,n + gmn,i − gin,m )An =
2

1
= Ai,i + g im gim,n An . (D.14)
2
Taking into account that
dg
g ik dgik = (D.15)
g

(see CW2 question 4), we have

1 √ 1 √ √
Ai;i = An,n + g,n An = An,n + (ln −g),n An = √ [An,n −g + ( −g),n An ] =
2g −g

1 √ 1 √
= √ ( −gAn ),n = √ ( −gAi ),i . (D.16)
−g −g

12
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 4.
————————————————————————————————————————————————-
Q4
Let φ is an arbitrary scalar field. Prove that
1 √ 
g ik φ;k;i = √ −gg ik φ,k .
−g ,i

[20 marks]
Solution:
Let us introduce the following covariant vector:
Ai = φ,i = φ;i . (D.17)
After that
1
g ik φ;k;i = g ik Ak;i = g ik Ak = Ai;i = Ai,i + Γiin An = Ai,i + g im (gim,n + gnm,i − gin,m ) An =


;i 2

1 1
= Ai,i + g im gim,n An + g im (gnm,i − gin,m ) An . (D.18)
2 2
Taking into account that
√
1 im √
( −g),n
g gim,n = ln −g ,n = √ . (D.19)
2 −g
And taking into account that
g im (gnm,i − gin,m ) = g im gnm,i − g im gin,m = g mi gni,m − g im gin,m = g im gin,m − g im gin,m = 0, (D.20)
we obtain
√
ik ( −g),n n 1 √ 1 √ 1 √
Ai,i −gAn ,n = √ −gg ni Ai ,n = √ −gg ni φ,i ,n =




g φ;k;i = + √ A == √
−g −g −g −g

1 √
−gg ik φ,k ,i .


=√ (D.21)
−g

Q5
Let Aik is a symmetric tensor. Prove that
1 √  1
Aki;k = √ −gAki − gkl,i Akl .
−g ,k 2
[20 marks]
Solution:

1 kn 1 mn
Aki;k = Aki,k + Γkkm Am m k k
i − Γik Am = Ai,k + g (gkn,m + gmn,k − gkm,n ) Am
i − g (gin,k + gkn,i − gik,n ) Akm =
2 2

g,m m 1 kn 1 mn k 1
= Am
i,m + Ai + g (gmn,k − gkm,n ) Am
i − g Am gkn,i − g mn Akm (gin,k − gik,n ) =
2g 2 2 2
√
( −g),m m 1 m kn
1 1
= Am
i,m + √ Ai + Ai g gmn,k − g kn gkm,n − Akn gkn,i − Akn (gin,k − gik,n ) =
−g 2 2 2

1 √ 1
−gAki ,k − gkl,i Akl .


=√ (D.22)
−g 2

13
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 5.
————————————————————————————————————————————————-

E. Solutions to course work 5

Q1
A photon emitted outward from r = r0 at time t = 0 propagates radially in the Schwarzschild space-time. Show that the
world-line of the photon is given by the following equation
r − rg
ct = r − r0 + rg ln .
r0 − rg

[25 marks]
Solution:
From ds = 0 for θ, φ = const, we have
rg 2 rg
c2 (1 − )dt − (1 − )−1 dr2 = 0, (E.1)
r r
hence
Z
rg −1
cdt = (1 − ) dr = r(r − rg )−1 dr = r(r − rg )−1 dr =
r

Z
= (r − rg + rg )(r − rg )−1 dr = (r − rg ) + rg ln(r − rg ) + C. (E.2)

If at t = 0 r = r0 , then

C = −[(r0 − rg ) + rg ln(r0 − rg )], (E.3)

and finally
r − rg
ct = r − r0 + rg ln . (E.4)
r0 − rg

14
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 5.
————————————————————————————————————————————————-
Q2
A particle starts to fall freely from infinity and moves along a radial geodesic in the Schwarzschild metric. Using the
expression for ds and an appropriate component of geodesic equation, show that
 2/3
3/2 3
r(τ ) = r (τ0 ) − crg1/2 (τ − τ0 ) ,
2

where τ is the proper time (ds = cdτ ).[25 marks]

Solution:
A particle moves along radial geodesic in the Schwarzschild metric, then

cd2 t dt dt dr dr
2
+ Γ000 c2 ( )2 + 2Γ001 c + Γ011 ( )2 = 0. (E.5)
ds ds ds ds ds

1 00
Γ000 = g (g00,0 + g00,0 − g00,0 ) = 0, (E.6)
2

rg
1 00 1 dg00 1 rg d(1 − r ) rg rg
Γ001 = g (g00,1 + g10,0 − g01,0 ) = g 00 = (1 − )−1 = 2
(1 − )−1 , (E.7)
2 2 dr 2 r dr 2r r

1 00
Γ011 = g (g10,1 + g10,1 − g11,0 ) = 0, (E.8)
2
so we have
d2 t rg rg dt dr
+ + 2 (1 − )−1 = 0, (E.9)
ds2 r r ds ds
or
d dt rg dt d rg rg d dt rg
( ) + (1 − )−1 (1 − ) = (1 − )−1 [ (1 − )] = 0, (E.10)
ds ds r ds ds r r ds ds r
hence
dt rg
(1 − ) = C. (E.11)
ds r
At infinity dt
ds = c−1 , hence C = c−1 . Substituting this into eq. for ds, we have

rg 2 rg rg dr
1 = (1 − )c (1 − )−2 c−2 − (1 − )−1 ( )2 , (E.12)
r r r ds

r
rg dr dr rg
1− = 1 − ( )2 ⇒ ( ) = −c , (E.13)
r ds dτ r

we take ”−” for falling objects, then

2 3/2
r (τ ) − r3/2 (τ0 ) = −crg1/2 (τ − τ0 ), (E.14)
3
and finally
3
r(τ ) = [r3/2 (τ0 ) − crg1/2 (τ − τ0 )]2/3 . (E.15)
2

15
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 5.
————————————————————————————————————————————————-
Q3
Using the coordinate transformations
1/2
rg r1/2 dr r3/2 dr
Z Z
cτ = ct + , R = ct +
r − rg rg
1/2
(r − rg )
show that the Schwarzschild metric takes the form
rg
ds2 = c2 dτ 2 − dR2 − r2 (dθ2 + sin2 θdφ2 ).
r
Expressing r in terms of R − cτ , demonstrate that the latter metric is non-stationary. What can be said about the true character
of the Schwarzschild space-time metric at r = rg ? [25 marks]
Solution:
By differentiating
1/2
rg r1/2 dr r3/2 dr
cdτ = cdt + , dR = cdt + 1/2
. (E.16)
r − rg rg (r − rg )
Subtracting the first from the second we have
!  1/2
dr r3/2 1/2 1/2 r1/2 dr r  r 1/2
g
dR − cdτ = − rg r = (r − rg ) = dr, hence dr = (dR − cdτ ).(E.17)
r − rg rg1/2 1/2
(r − rg )rg rg r

Subtracting the first multiplied by r/rg from the second we have

r r crdτ − rg dR
cdτ − dR = cdt( − 1), hence cdt = . (E.18)
rg rg r − rg

Then substituting the expressions for dr and cdt into ds2 in the Schwarzschild form we obtain
 2  
r − rg rcdτ − rg dR rg 2 1 1
ds2 = − (dR − cdτ ) −r2 (dθ2 +sin2 θdφ2 ) = (rcdτ − rg dR)2 − rg (dR − cdτ )2 −
r r − rg r − rg r − rg r

" #
2 rg r r2
2 g
2 2 2 2 2
−r (dθ + sin θdφ ) = c dτ (r − rg ) − 2cdRdτ ( − rg ) − dR ( − rg ) − r2 (dθ2 + sin2 θdφ2 ) =
r r

rg 2
c2 dτ 2 − dR − r2 (dθ2 + sin2 θdφ2 ). (E.19)
r
From
2 3/2
r1/2 dr = rg1/2 d(R − cτ ) we have r = C + rg1/2 (R − cτ ), (E.20)
3
then choosing the constant of integration C = 0 so that r = 0 → R − cτ = 0, we have
 2/3
3 1/2
r= r (R − cτ ) . (E.21)
2 g
Finally, putting this into the metric in new coordinates we have
 2/3  4/3
2rg 3 1/2
2
ds = c dτ −2 2
− rg (R − cτ ) (dθ2 + sin2 θdφ2 ), (E.22)
3(R − cτ ) 2
we can see that the metric depends on τ , which means that the gravitational field is non- stationary. We can see also
that there is no physical singularity at r = rg .

16
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 5.
————————————————————————————————————————————————-
Q4
A rotating black hole is described by the Kerr metric.
a) Find the location of the event horizon and the limit of stationarity. Sketch the rotating black hole as projected on i) the
equatorial plane, θ = π/2, and ii) the perpendicular plane, φ = 0 (indicate the event horizon, the limit of stationarity and the
ergosphere).[25 marks]
Solution:
Location of the event horizon corresponds to

g 11 = 0. (E.23)

Taking into account that all out of diagonal components g1i = 0 (if i 6= i), one can see that g 11 = 1/g11 and the
location of event horizon can be determined from

g11 = ∞, (E.24)

hence

∆ = r2 − rg r + a2 = 0. (E.25)

There are two solutions

q
rg ± rg2 + 4a2
r± = . (E.26)
2
The outer event horizon, rhor , corresponds to the sign “+”, hence
rg  p 
rhor = 1 + 1 − α2 , (E.27)
2
where
2a ac2 Jc
α= = = . (E.28)
rg GM GM 2
The location of limit of stationarity corresponds to

g00 = 0. (E.29)

In the case of Kerr metric this corresponds to

rg r
1− = 0, (E.30)
ρ2
where

ρ2 = r2 + a2 cos2 θ, (E.31)

hence from

r2 − rg r + a2 cos2 θ = 0.

There are two solutions

q
rg ± rg2 + 4a2 cos2 θ
r± = . (E.32)
2
The outer limit of stationarity, rst , corresponds to the sign “+”, hence
q
rg + rg2 + 4a2 rg p
rst = = (1 + 1 − α2 cos2 θ). (E.33)
2 2

17
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 6.
————————————————————————————————————————————————-

F. Solutions to course work 6

Q1
A weak gravitational wave is a small perturbation of the Minkowski metric, gik = ηik + hik . Show that, to terms of first order
in hik , the contravariant metric tensor is g ik = η ik − η in η km hnk . Consider a linear transformation xi = x0i + ξ i , where ξ i are
small functions of xi . Show that hik = h0ik − ξi,k − ξk,i . Prove that it is always possible to find such ξ i that the Ricci tensor
takes the following simple form:
1
Rik = − η lm hik,l,m .
2
[30 marks]
Solution:

If gik = ηik + hik , where hik are small, contravariant metric tensor can be written as g ik = η ik + aik , where aik are
also small. Taking into account that gik g kn = δin we have

(ηik + hik )(η kn + akn ) = δin , δin + ηik akn + hik η kn = δin , (F.1)

ηik akn = −hik η kn , η im ηik akn = −η im hik η kn , δkm akn = −η im η kn hik , (F.2)

amn = −η mi η nk hik , finally g ik = η ik − η in η km hnk . (F.3)

Then

0 ∂x0i
gik = S̃in S̃km gnm , where S̃ki = = δki − ξ,k
i
, (F.4)
∂xk
hence
ηik + hik = (δin − ξ,in )(δkm − ξ,k
m
)(ηnm + h0nm ); (F.5)
to terms of first order in the h0ik
hik = −ηik + δin [δkm (ηnm + h0nm ) − ξ,k
m
ηnm ] − ξ,in δkm ηnm = −ηik + ηik + h0ik − ξ,k
m
ηim − ξ,in ηnk =

= h0ik − ξi,k − ξk,i . (F.6)

Writing the Ricci tensors to terms of first order (in linear approximation) we have
1 lm
Rik = Γlik,l − Γlil,k = η (him,k,l + hkm,i,l − hik,m,l − him,l,k − hlm,i,k + hil,m,k ) =
2

1 1
= − η lm hik,m,l + η lm (hkm,i,l − hlm,i,k + hil,m,k ) =
2 2

1 1 l
= − η lm hik,m,l + hk,i,l − h,i,k + hm l


i,m,k , where h = hl . (F.7)
2 2
We have four arbitrary functions ξ, thus we can impose on hik four supplementary conditions:
hki,k − 1/2h,i = 0, (F.8)
hence
 
1 1 1 1 1
Rik = − η lm hik,l,m + h,k,i − h,i,k + h,i,k = − η lm hik,l,m . (F.9)
2 2 2 2 2

18
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 6.
————————————————————————————————————————————————-
Q2
Two bodies of equal mass, m1 = m2 = m, attracting each other according to Newton’s law, move in circular orbits around their
common centre of mass with orbital period P . Using the quadrupole formula for the generation of gravitational waves, show
that to an order of magnitude, h ∼ (rg /R)(rg /cP )2/3 , where R is the distance to the system and rg = 2Gm
c2
is the gravitational
radius.[10 marks]
Solution:
See solution to the next question and take into account that
2π
ω= . (F.10)
P
Q3
The quadrupole formula for the metric perturbation associated with gravitational waves is given by
2G d2 Dαβ
hαβ = − (t − R/c),
3c4 R dt2
where R is the distance to the source of the gravitational waves and
Z
Dαβ = (3xα xβ − r2 δαβ )dM

is the quadrupole tensor of the source. Consider a mass m moving along circular orbit around the black hole of mass M ,
assuming that m  M .
a) Show that all the amplitudes hαβ of gravitational wave, emitted by such system, are periodic functions of time with ω = 2ω0 ,
where ω0 = 2π/T , and T is the orbital period.[30 marks]

Solution:

1 2
From x1 = r cos ω0 t, x2 = r sin ω0 t, D11 = mrc2 (3 cos2 ω0 t − 1) = mr (1 + 3 cos 2ω0 t),
2
1 2 3
D22 = mrc2 (3 sin2 ω0 t − 1) = mr (1 − 3 cos 2ω0 t), D12 = mrc2 sin 2ω0 t, (F.11)
2 2
then
2Gmr2 3 4ω02 Gmr2 4ω02 Gmr2
h11 = −h22 = − 4
(2ω0 )2 cos 2ω0 t = 4
cos 2ω0 , h12 = sin 2ω0 , (F.12)
3c R 2 c R c4 R
it is clear, that ω = 2ω0 .
b) Show that, to an order of magnitude (omitting the indices α and β)
 2/3
rg R g ω
h≈ ,
R c
where rg is the gravitational radius of the mass m and Rg is the gravitational radius of the black hole.[30 marks]

Solution:
From
GM
rω02 = , (F.13)
r2
we have
1 ω02
= , (F.14)
r3 GM
and finally
rc−1 = (4GM )−1/3 ω 2/3 . (F.15)
Thus
2/3
4ω02 Gmr2

rg Rg rg Rg ω
h≈ 4
= ≈ . (F.16)
c R rR R c

19
 A. Polnarev. (SPA7019 ). 2014. Solutions to course work 6.
————————————————————————————————————————————————-
Q4
The future LISA mission will be able to detect gravitational waves with h > 10−23 , if 10−4 Hz < ω < 3 · 10−3 Hz. From what
distance will it be possible to detect gravitational radiation from the binary system, containing the black hole of mass m = 3M ,
moving along a circular orbit with radius r = 104 Rg around the massive black hole of mass M = 103 M ?[30 marks]
Solution:

GM c2 2GM Rg c2
ω02 = 3
= 2 3
= , (F.17)
r 2 c r 2r3
hence,
r s
Rg Rg 10−6 c 10−4 Hz
ω0 = c = c = √ = √ , (F.18)
2r3 2 · 1012 Rg3 2Rg 2

thus
√
ω = 2ω0 = 210−4 Hz ≈ 10−4 Hz, (F.19)

which means that the radiation is within LISA frequency range.

Then
2/3
3 · 105 3 · 105 · 10−4

m R −1 M 2/3 ω
h= ( )( ) ( ) ( −4 )2/3
3 · 1018 3 · 1010 M 1pc M 10 Hz

m R −1 M 2/3 ω
≈ 10−19 ( )( ) ( ) ( −4 )2/3 . (F.20)
M 1 pc M 10 Hz

Hence
3 · 105 cm 3 · 105 · 103 · 1.4 · 10−4 s−1 cm 2/3
h= ( ) > 10−23 , (F.21)
R 3 · 1010
if

R < 3 · 1023 · 105 cm · 10−4 ≈ 1M pc. (F.22)

20