The Einstein Field Equations

Chris Ormel∗
Kapteyn Instituut
Landleven 12, 9747 AD Groningen, The Netherlands
November 20, 2001

∗ [email protected]
CONTENTS 1

Contents
1 The Einstein Field Equations 3

2 The metric gµν 3

3 The Energy-Momentum Tensor 4

4 The Riemann Curvature Tensor 5

5 The Einstein tensor 6

6 Einstein’s modification 7

7 Applications on cosmology 7
CONTENTS 2

Introduction
The Einstein equations are often refered to, even in informal (i.e. unscientific) conversations
as the (most) fundamental laws governing the structure and predicting the future of the
universe. However for most people this subject still remains schrouded in mystery, even for
some undergraduate students. My purpose is to unravel the Einstein equations, to clarify its
components and the relations they describe. In the last section I will derive symbols, used
in cosmology from the Einstein equations. It is not the intention to give a formal derivation,
but more an intuitive feeling for the ideas behind this theory. However, most students should
be able to understand the math.

prerequistites, notation
In order to be able to follow this paper, one should be equipped with a fundamental set of
’tensor analysis’. The most important properties being:
• 4-vectors e.g. xµ = (x0 , x1 , x2 , x3 ) = (t, x)1
• 4-tensor like T µν (16 components) or Rαβγδ (256 components)

• contravariant (xµ , ∂ µ ) and covariant (xµ , ∂µ ) components of vectors and (in general)
tensors or derivatives.
• Using the metric tensor to raise or lower indices: Vµ = gµν V ν
3
• contraction, which reduces the rank of a tensor with two, e.g. Rµν = Rλµλν = Rλµλν
P
λ=0
Here the λ’s are dummy indices and could be replaced by any other greek letter.
• tensor transformation from one frame to another:
∂x0µ ν ∂x0µ ∂x0ν αβ
A0µ = A A0µν = A
∂xν ∂xα ∂xβ

In this paper I will stick to the following notation

• greek letters for 4-vectors and latin indices for (spatial) 3-vectors.
∂ 2
• ∂µ for (ordinary) differentiation ∂xµ , ∇µ for the covariant derivative

1 the speed of light in x0 (ct) is to be taken 1 in this paper

2 what exactly the covariant derivative is, will be clear after reading this paper
1 THE EINSTEIN FIELD EQUATIONS 3

1 The Einstein Field Equations

The Einstein Field equations are:

Gµν = 8πGTµν (1)

In which
• Gµν = Rµν − 21 Rgµν is the Einstein tensor, which measures the curvature of spacetime
• G is Newtons constant3
• Tµν is the energy-momentum tensor.
• gµν is the metric, a generalisation of the Minkowski metric ηµν
• Rµν is the Ricci tensor, a contraction of the Riemann curvature tensor.

Rµν = Rλµλν (2)

• R is the curvature scaler, the contraction of the Ricci tensor

R = Rµµ (3)

In the following 4 chapters I will describe the varies components of the Einstein equations.
The last chapter is an application of the Einstein equations on cosmology.

2 The metric gµν

In special relativity we have the invariant

dτ 2 = c2 dt2 − dx2 − dy 2 − dz 2 = ηαβ dxα dxβ (4)

where ηαβ = diag (1, −1, −1, −1) is the Minkowski metric. The Minkowski metric only
2 α
applies in special relativity, where we have ddτξ2 = 0.
According to the principle of relativity there is always a freely falling (comoving) coor-
2 α
dinate frame ξ α in which ddτξ2 vanishes. In another coordinate frame xµ the invariant dτ 2
becomes
∂ξ α ∂ξ β
dτ 2 = ηαβ µ dxµ ν dxν = gµν dxµ dxν (5)
∂x ∂x
Where gµν is the metric tensor
∂ξ α ∂ξ β
gµν = ηαβ (6)
∂xµ ∂xν
3 In contrary to c, I will not set this constant 1
3 THE ENERGY-MOMENTUM TENSOR 4

3 The Energy-Momentum Tensor

The energy-momentum tensor Tµν is a symmetric tensor, which conserves energy and mo-
mentum. It’s described by this conservation law, rather than an exact defenition. I will give
an example of this tensor for fluids.
Consider a fluid characterised by the density field ρ(x) and four-velocity

dxµ dt dxk
 
µ
u = = , = (γ, γv) (7)
dτ dτ dτ
The boost term γ restricts the discussion to special relativity, but our goal is to show the
conservation of energy and momentum. Let’s define the momentum-energy tensor to be 4

T µν = (ρ + p)uµ uν − pg µν (8)

Where g µν will be replaced by η µν .

Now in a Newtonian case (v << c), the pressure (p) is small compared to the density of
mass energy (ρ) 5 , thus the components become

ρv j
 
µν ρ
T ≈ (9)
ρv k ρv j v k + pδ jk

Note that u0 = γ ≈ 1 in the Newtonian case. Then the conservation of energy and momentum
will appear if T µν is contracted with ∂ν

∂ν T µν = 0. (10)

This immediately gives for µ = 0

∂p
∂ν T 0ν = + ∇·v = 0 (11)
∂t
the equation of continuity. The µ = k components become:
∂ ∂ ∂
(ρv k ) + j (ρv j v k ) + p=0 (12)
∂t ∂x ∂xk
Rewriting this equation and using the equation of continuity, we get the following set of
equations:
∂ k ∂ ∂ ∂ k ∂
ρ v + v k ρ + v k j (ρv j ) + ρv j v = − kp
∂t ∂t ∂x ∂xj ∂x
∂ ∂
−v k ρ − v k j (ρxj ) =0
∂t ∂x
This finally gives Euler’s equation :
∂v 1
+ (v · ∇)v = − ∇p (13)
∂t ρ
In general relativity the ∂ν is replaced by ∇ν , so we get the following list of properties

4 the following discussion is in a viewpoint of special relativity. However the equations can be generalised

(to general relativity)

5 because I took the speed of light 1, for a fair compairison the units must be equal and ρ should be

multiplied with c2
4 THE RIEMANN CURVATURE TENSOR 5

Properties of the energy momentum tensor

• T µν is the energy-momentum (or stress-energy) tensor, in which the momentum and
energy are contained.
• T µν is symmetric: T µν = T νµ
• ∇µ T µν = 0. An expression for the conservation of energy and momentum.

4 The Riemann Curvature Tensor

The defenition of the Riemann curvature tensor (in a coordinate basis) is

Rαβγδ = ∂γ T αβδ ∂δ T αβγ + Γαµγ Γµβδ − Γαµδ Γµβγ (14)

where the Γ’s are the connection coefficients, sometimes refered to as affine connections .
The need to introduce these terms comes from the generalization of the partial derivative,
∂µ , to the covariant derivative ∇µ . This is the proper derivative in general relativity, because
it transforms as a tensor 6 . In flat spacetime, the covariant derivative reduces to the partial.
Its definition is
∇µ V ν = ∂µ V ν + Γνµλ V λ (16)
Whereas the connection coefficients itself, can be defined with the additional condition called
metric compatibility
∇ρ gµν = 0 (17)
This condition also allows for lowering and raising indices of the metric. The connection
coefficient are then defined by
1 σρ
Γσµν = g (∂µ gνρ + ∂ν gρµ − ∂ρ gµν ) (18)
2
The connection coefficients and thus the Riemann Tensor are derived from the metric. Fur-
thermore with metric compatibility contracting a covariant derivative simplifies

gµλ ∇ρ V λ = ∇ρ Vµ (19)

Thus the metric can be interchanged with the (covarient) derivative.

Now the relevance of the Riemann Tensor is its property that it is the only tensor that
includes the metric and is linear in its second derivatives. ∂µ ∂ν g αβ , which would be a
logical choice, is not right, because the result doesn’t represent a proper tensor (i.e. it
doesn’t transform as a tensor). ∇µ ∇ν g αβ is wrong as well, because this is zero by metric
compatibility. Because of (18) the Riemann tensor in its covariant form (Rαβγδ = gαµ Rµβγδ )
reduces to
1
Rαβγδ = (∂γ ∂β gαδ − ∂γ ∂α gδβ − ∂δ ∂β gαγ + ∂δ ∂α gγβ ) (20)
2
As has been said before, the Riemann Tensor is composed of 256 components, but all are
not independent. The following symmetries from the equation above reduce the independent
6 That ∂µ doens’t transform as a tensor is easily seen:
0 0 0
!
∂xµ ∂xν ν ∂xµ ∂xν ∂xµ ∂ 2 xν
 
0
∂µ0 V ν = ∂
0 µ ν
V = 0 ν
(∂µ V ν ) + 0 Vµ (15)
∂xµ ∂x µ
∂x ∂x ∂x ∂x ∂xµ
µ µ

The first part obeys tensor transformation, but the second term obviously does not.
 5 THE EINSTEIN TENSOR 6

7
components of the Riemann Tensor to 20.

symmetries of the Riemann tensor

(antisymmetry on last two indices) Rαβγδ = Rαβ[γδ] (21a)

Rα[βγδ] =0 (21b)
(antisymmetry on first two indices) Rαβγδ = R[αβ]γδ (21c)
(symmetry under pair exchange) Rαβγδ = Rγδαβ (21d)
(completely vanishing of antisymmetric part) R[αβγδ] =0 (21e)

The ’[ ]’ is a notation for:

1
Rαβ[γδ] = (Rαβγδ − Rαβδγ ) (22)
2
1
Rα[βγδ] = (Rαβγδ + Rαγδβ + Rαδβγ − Rαβδγ − Rαγδβ − Rαδγβ ) (23)
3!
Thus a ’+’ sign for an even permutation of βγδ, and a ’-’ sign for an odd permutation of
βγδ. From (22) it is easily seen where the antisymmetry comes from
The Riemann tensor is of fourth order, but can be contracted to form the Ricci tensor
Rµν or the curvature scaler R.

Rµν = Rλµλν (24)

R = Rµµ = g µν Rµν (25)

These are needed to form the Einstein tensor Gµν .

5 The Einstein tensor

The Einstein Tensor is given by
1
Gµν = Rµν − Rgµν (26)
2
properties of the einstein tensor

• Gµν vanishes when spacetime is flat

• Gµν is constructed from the Riemann tensor and the metric.
• Gµν is different from other tensors that can be constructed from the Riemann tensor
and the metric by the demands:
(1) Gµν is linear in Riemann
(2) Gµν (like Tµν ) is symmetric
7 Not all these relations are independent from each other, however. (21a) and (21b) follow directly from
the Riemann Tensor. Relation (21c) follows when the metric is included. (21d) and (21e) follow from the
first three equations
6 EINSTEIN’S MODIFICATION 7

8 9
(3) Gµν (like Tµν ) obeys the Bianchi identity
∇µ Gµν = 0 (30)

6 Einstein’s modification
Einstein modified his Field equation in order to provide a static universe. He introduced the
term Λ as cosmological constant

Gµν + Λg µν = 8πT µν (31)

A nonzero value of Λ no longer results in the vanishing of Gµν in vacuum. I will not elaborate
on the cosmological constant, but give an (simple) application of the Einstein equations in
cosmology.

7 Applications on cosmology
The Robertson-Walker metric10 is (without derivation)

−1 0 0 0
 
 0 a2 (t)
1−kr 2 0 0 
gµν =   (32)
 0 0 a2 (t)r2 0 
0 0 0 a2 (t)r2 sin2 θ

and for the inverse g µν (because g µν gµν = δνµ and the metric is diagonal).
 
−1 0 0 0
2
 0 1−kr
a2 (t) 0 0 
g µν =  (33)
 
1
 0 0 0

a2 (t)r 2 
1
0 0 0 2 2 2
a (t)r sin θ

Where a = a(t) is a time-dependent scale factor. k measures the curvature and is only
interesting for k = −1, k = 0, k = +1. The table gives an qualitive relation between the
8 That the Bianchi identity is indeed true can best be seen in an inertial frame. Because G
µν (like Tµν is
a proper tensor, the relation will then be true in every frame. In an inertial frame the connection coefficients
vanish and the derivatives are partial. The Riemann tensor is then simplified to
Rαβγδ = ∂β ∂δ gαγ − ∂β ∂γ gαδ + ∂α ∂γ gβδ − ∂α ∂δ gβγ (27)
And now contracting this (g αλ R αµλν ) and using property (19) we get for the Ricci Tensor and for the
curvature scaler
Rµν = 0 − ∂µ ∂ α gαν + gµν − ∂ α ∂ν gµα (28a)
R = g µν Rµν = −∂ ν ∂ α gαν + 0 − ∂ α ∂ µ gµα = −2∂ α ∂ β gαβ (28b)
Contraction of the Einstein tensor then gives
1
∂ µ Gµν = ∂ µ (Rµν − gµν R) = −∂ α gαν + ∂ µ gµν − ∂ µ ∂ν ∂ α gµα + ∂ α (gµν ∂ µ )∂ β gαβ
2
The first two terms cancel, as do the last two terms.
9 a more generalised version of the Bianchi Identity is

0 = ∇[λ Rρσ]µν = ∇λ Rρσµν + ∇σ Rλρµν + ∇ρ Rσλµν = 0 (29)

where I used the antisymmetry of the Riemann tensor in its first two components (i.e. Rρσµν = −Rσρµν ).
10 The RW-metric demands two assumptions about the universe: Isotropy and homogeneity of space
7 APPLICATIONS ON COSMOLOGY 8

curvature of the universe (k) and the density (Ω). The aim of this section is to derive this
relation.
k curvature Ω
-1 negative open Ω<1
0 euclidean flat Ω=1
1 positive closed Ω>1
Because gµν is given, the connection coefficients can be calculated with (18). Some are
explained in more detail.
1 0ρ 1 aȧ
Γ011 = g (∂1 g1ρ + ∂1 gρ1 − ∂ρ g11 ) = − g 00 ∂0 g11 = (34a)
2 2 1 − kr2
1 00 kr
Γ022 = − g ∂0 g22 = aȧr2 Γ111 = (34b)
2 1 − kr2
1
Γ033 = − g 00 ∂0 g33 = aȧr2 sin2 θ (34c)
2
ȧ
Γ101 = Γ10 = Γ202 = Γ220 = Γ303 = Γ330 =
1
(34d)
a
1 1ρ 1 1 − kr2 2 2
Γ122 = − g ∂1 g22 = − 2r a = −r(r − kr2 ) (34e)
2 2 a2
1
Γ133 = −r(r − kr2 ) sin2 θ Γ212 = Γ221 = Γ313 = Γ331 = (34f)
r
Γ233 = − sin θ cos θ Γ323 = Γ332 = cot θ (34g)
where ȧ is the time derivative of a(t). The nonzero components of the Ricci-Tensor are:
R00 = Rλ0λ0 = ∂λ Γλ00 − ∂0 Γλ0λ + Γλµλ Γµ00 − Γλµ0 Γµ0λ =
ȧ2 (35a)
 
∂ 3ȧ ä
=− −3 2 =3
∂t a a a
R11 = R 1λ1 = ∂0 Γ11 − ∂1 Γ11 − ∂1 Γ111 + Γ212 + Γ313
λ 0 1

+ Γ011 Γ101 + Γ202 + Γ303 + Γ111 Γ111 + Γ212 + Γ313

− Γ101 Γ011 − Γ011 Γ110 − Γ111 Γ111 − Γ221 Γ211 − Γ113 Γ313 = ... (35b)
aä + 2ȧ2 + 2k
=
1 − kr2
= r aä + 2ȧ2 + 2k
2


R22 (35c)
= r2 aä + 2ȧ2 + 2k sin2 θ


R33 (35d)
With the curvature scalar becoming
6
R = g µν Rµν = aä + ȧ2 + k


(36)
a2
Now we apply the Einstein equation (1) with Tµν for a perfect fluid (8) in a comoving frame.
Then
8πGT µν = 8πG diag(ρ, −p, −p, −p) = Gµν (37)
The two independent components of the Einstein Tensor are
1 1 0 3ȧ2 + 3k
G00 =R00 − Rg 00 = g 00 R00 − Rδ0 = (38a)
2 2 a2
1 1 1 −2aä − ȧ2 − k
G11 =R11 − Rg 11 = g 11 R11 − Rδ = (38b)
2 2 1 a2
7 APPLICATIONS ON COSMOLOGY 9

Equation (38a)11 leads to Friedmann equation

8πG 2
ȧ2 − ρa = −k (39)
3
With the following notations

ȧ 3H 2 ρ
H= ρc = Ω=
a 8πG ρc
This can be rewritten as
k
Ω−1= (40)
a2 H02

The relation between Ω and k is clear.

Of course many more interesting relations can be obtained, but this would be too much
for this paper. For those interested I would suggest the college cosmology.

11 (38b) leads to the equation of conservation of energy. I will not elaborate on this equation in this paper,

but it does play an important role in cosmology (e.g. to define the decelaration parameter q = − aä ȧ2
)
REFERENCES 10

References
[1] Misner C.W., Thorne K.S., Wheeler J.A. (1973) Gravitation
[2] Weinberg, S. (1972) Gravitation and Cosmology
[3] Peacock, J.A. (1999) Cosmological physics
[4] Sean M. Carroll (1997) Lecture Notes on General Relativity: part 1 (no-nonsense
introduction), part 3 (curvature), part 4 (gravitation) and part 8 (cosmology),
http://pancake.uchicago.edu/c̃arroll/notes/
Index
affine connections, 5

Bianchi identity, 7

connection coefficients, 5
cosmological constant, 7
curvature
universe, 8
curvature scaler, 6

dummy, 2

Einstein equations, 2
Einstein tensor, 6
energy-momentum tensor, 4
equation of continuity, 4
Euler’s equation, 4

Friedmann equations, 9

metric, 3
robertson-walker, 7
metric compatibility, 5
Minkowski metric, 3

principle of relativity, 3
properties
einstein tensor, 6
energy momentum tensor, 5

Ricci tensor, 6
Riemann
curvature tensor, 5
symmetries of .., 6
Robertson-walker metric, 7

stress energy tensor, 5

tensor transformation, 2

11