Tensor Analysis and Geometry
Tensor Analysis and Geometry
Spherical Coordinates
x1 = r sin θ cos φ , x 2 = r sin θ sin φ , x 3 = r cosθ , x 4 = ct
3-dimensional line element: (
ds 2 = dr 2 + r 2 dθ 2 + sin 2 θdφ 2 )
Christoffel Symbols
1 ∂g rm ∂g rn ∂g mn
Christoffel Symbol of the first kind: Γrmn = [mn, r ] = + m − .
2 ∂x n ∂x ∂x r
Christoffel Symbol of the second kind: Γmn
z
= z = g rz Γmnr .
î mn
Now we interchange j and k (which is the other possible way of determining this second derivative):
∂Γiks
(Vi:k ): j = Vi ,kj − j Vs − Γiks Vs , j − ΓijrVr ,k + Γijr Γrks Vs − Γkjr Vi ,r + Γkjr Γirs Vs .
∂x
We now find the difference between these two. On the RHS, the first, third, fourth, sixth, and seventh terms
cancel out, thus giving the result:
∂Γijs ∂Γiks ∂Γiks ∂Γijs
Vi: jk − Vi:kj = − V + Γ Γ
r s
V + V − Γ Γ
r s
V = j − k + Γik Γrj − Γij Γrk Vs .
r s r s
∂x ∂x ∂x ∂x
k s ik rj s j s ij rk s
We define the Riemann (or Riemann-Christoffel) Curvature tensor by:
∂Γiks ∂Γij
s
R = − s
+ Γikr Γrjs − Γijr Γrks .
∂x j ∂x k
ijk
The difference between the covariant derivatives can thus be written as Vi: jk − Vi:kj = Rijk Vs . The Riemann
s
tensor used in this equation is called the Riemann curvature tensor of the second kind. The curvature tensor of
the first kind is defined as:
Rijkl = g ir R rjkl .
Symmetry Properties:
First skew symmetry Rijkl = − R jikl
Second skew symmetry Rijkl = − Rijlk
Since the LHS is now in terms of v, partial differentiation can be replaced by normal differentiation:
2
dx a dx b dx c dx a d 2 v dv
⇔ 2 + Γbca =− .
dv dv dv dv du 2 du
2
d 2v dv
By letting λ=− , we arrive at the final equation:
du 2 du
dx a b
a dx dx
c
dx a
+ Γ bc = λ .
dv 2 dv dv dv
General Relativity
1 0 0 0
η µν = 0 − 1 0 0 .
0 0 − 1 0
0 0 0 − 1
Γmn
z
= z = g rz Γmnr
î mn
we need only calculate them for values when r = z.
Γ100 = g 00 Γ100 = 12 A −1c −2 [g 00,1 + g10 , 0 − g10, 0 ] = 12 A −1 A′
Γ00
1
= g 11Γ001 = − 12 B −1 [g 10 ,0 + g 01, 0 − g 00,1 ] = 12 B −1 g 00,1 = 12 B −1 A′c 2
Γ111 = g 11Γ111 = − 12 B −1 [g11,1 + g 11,1 − g11,1 ] = − 12 B −1 g 11,1 = 12 B −1 B ′
Γ22
1
= g 11Γ221 = − 12 B −1 [g12 , 2 + g 21, 2 − g 22 ,1 ] = − 12 B −1 g 22,1 = − 12 B −1 ⋅ (2r ) = − rB −1
Γ33
1
(
= g 11Γ331 = − 12 B −1 [g13,3 + g 31,3 − g 33,1 ] = − 12 B −1 (− g 33,1 ) = − 12 B −1 2r sin 2 θ = − rB −1 sin 2 θ )
Γ212 = g 22 Γ212 = − 12 r −2 [g 21, 2 + g 22,1 − g 21, 2 ] = − 12 r −2 ⋅ (− 2r ) = r −1
Γ332 = g 22 Γ332 = − 12 r −2 [g 23,3 + g 32,3 − g 33, 2 ] = − 12 r −2 ⋅ 2r 2 cosθ sin θ = − cosθ sin θ
Γ133 = g 33 Γ133 = − 12 r −2 sin −2 θ [g 33,1 + g13,3 − g 13,3 ] = − 12 r −2 sin −2 θ ⋅ − 2r sin 2 θ = r −1( )
Γ = g Γ233 = − 12 r sin θ [g 33, 2 + g 23,3 − g 23,3 ] = − 12 r sin θ ⋅ − 2r cosθ sin θ = cot θ
3
23
33 −2 −2 −2 −2
( 2
)
We now solve the field equations:
[ (
R00 = Γ000 ,0 − Γ000 ,0 + Γ000 Γ000 − Γ000 Γ000 + Γ00
1
) (
Γ100 − Γ00
1
) (
Γ100 + Γ002 Γ200 − Γ002 Γ200 + Γ003 Γ300 − Γ003 Γ300 ) ( )]
[
+ Γ01
1 1 0 1
( 0 1 1 1 1
)(1 2 1 2 1
) 3 1
(
, 0 − Γ00 ,1 + Γ01 Γ00 − Γ00 Γ01 + Γ01 Γ10 − Γ00 Γ11 + Γ01 Γ20 − Γ00 Γ21 + Γ01 Γ30 − Γ00 Γ31
3 1
) ( )]
+ [Γ 2
02 , 0 (
− Γ002 , 2 + Γ020 Γ002 − Γ000 Γ022 + Γ02
1
) (
Γ102 − Γ00
1
)(
Γ122 + Γ022 Γ202 − Γ Γ ) + (Γ Γ
2
00
2
22
3
02
2
30 − Γ003 Γ322 )]
+ [Γ 3
03, 0 − Γ003 ,3 + (Γ 0
03 Γ003 − Γ000 Γ033 )+ (Γ
1 3
03 10Γ − Γ00
1
Γ133 )+ (Γ Γ
2
03
3
20 − Γ Γ ) + (Γ Γ
2
00
3
23
3
03
3
30 − Γ003 Γ333 )]
( )( ) (
R11 = Γ100 ,1 − Γ110 ,0 + Γ100 Γ010 − Γ110 Γ000 + Γ101 Γ110 − Γ111 Γ100 + Γ102 Γ210 − Γ112 Γ200 + Γ103 Γ310 − Γ113 Γ300
) ( )]
[
+ Γ11,1 − Γ111 ,1 + Γ110 Γ01
1 1
(
− Γ110 Γ01
1
) (
+ Γ111 Γ111 − Γ111 Γ111 + Γ112 Γ21
1
− Γ112 Γ21
1
+ Γ113 Γ31
1
) (
− Γ113 Γ31
1
) ( )]
(
+ Γ122 ,1 − Γ112 , 2 + Γ120 Γ012 − Γ110 022 2 1
Γ ) + (Γ
12 Γ11 − Γ11 Γ12 + Γ12 Γ21 − Γ11 22
1 2 2 2 2 2 3 2
)(
12 Γ31 − Γ11 32
3 2
Γ )+ (Γ Γ )]
( ) ( )(
+ Γ133 ,1 − Γ113 ,3 + Γ130 Γ013 − Γ110 Γ033 + Γ131 Γ113 − Γ111 Γ133 + Γ132 Γ213 − Γ112 Γ233 + Γ133 Γ313 − Γ113 Γ333
) ( )]
[ (
R22 = Γ200 , 2 − Γ220 ,0 + Γ200 Γ020 − Γ220 Γ000 + Γ20
1
Γ120 − Γ22
1
) ( )
Γ100 + Γ202 Γ220 − Γ222 Γ200 + Γ203 Γ320 − Γ223 Γ300 ( ) ( )]
[
+ Γ21
1 0 1 0
(
1
)( 1 1 1
)( 1 2 1
)] 2 1
) (
, 2 − Γ22 ,1 + Γ21 Γ02 − Γ22 Γ01 + Γ21 Γ12 − Γ22 Γ11 + Γ21 Γ22 − Γ22 Γ21 + Γ21 Γ32 − Γ22 Γ31
1 3 1 3 1
+ [Γ −Γ
2
Γ − Γ Γ ) + (Γ Γ − Γ Γ )+ (Γ Γ − Γ Γ ) + (Γ Γ − Γ Γ )]
2
+ (Γ 0 2 0 2 1 2 1 2 2 2 2 2 3 2 3 2
Γ − Γ Γ ) + (Γ Γ − Γ Γ )+ (Γ Γ − Γ Γ ) + (Γ Γ − Γ Γ )]
22 , 2 22 , 2 22 02 22 02 22 12 22 12 22 22 22 22 22 32 22 32
+ [Γ −Γ
3
23, 2
3
22 , 3 + (Γ 0
23
3
02
0
22
3
03
1 3
23 12
1 3
22 13
2
23
3
22
2
22
3
23
3
23
3
32
3
22
3
33
R = [Γ − Γ + (Γ Γ − Γ Γ ) + (Γ Γ − Γ Γ )+ (Γ Γ − Γ Γ ) + (Γ Γ − Γ Γ )]
0 0 0 0 0 0 1 0 1 0 2 0 2 0 3 0 3 0
+ [Γ − Γ + (Γ Γ − Γ Γ ) + (Γ Γ − Γ Γ )+ (Γ Γ − Γ Γ ) + (Γ Γ − Γ Γ )]
33 30 , 3 33, 0 30 03 33 00 30 13 33 10 30 23 33 20 30 33 33 30
1 1 0 1 0 1 1 1 1 1 2 1 2 1 3 1 3 1
31, 3 33,1 31 03 33 01 31 13 33 11 31 23 33 21 31 33 33 31
+ [Γ − Γ + (Γ Γ − Γ Γ ) + (Γ Γ − Γ Γ )+ (Γ Γ − Γ Γ ) + (Γ Γ − Γ Γ )]
2
32 , 3
2
33, 2
0
32
2
03
0
33
2
02
1
32
2
13
1 2
33 12
2
32
2
23
2
33
2
22
3
32
2
33
3
33
2
32
[ (
+ Γ333 ,3 − Γ333 ,3 + Γ330 Γ033 − Γ330 Γ033 + Γ331 Γ133 − Γ33
1
) ( ) (
Γ133 + Γ332 Γ233 − Γ332 Γ233 + Γ333 Γ333 − Γ333 Γ333 ) ( )]
We are left with
R00 = −Γ00
1
,1 + Γ01Γ00 − Γ00 Γ11 − Γ00 Γ12 − Γ00 Γ13 ,
0 1 1 1 1 2 1 3
R11 = Γ100 ,1 + Γ100 Γ010 − Γ111 Γ100 + Γ122 ,1 − Γ111 Γ122 + Γ122 Γ212 + Γ133 ,1 − Γ111 Γ133 + Γ133 Γ313 ,
R22 = −Γ22
1
Γ100 − Γ22
1
,1 − Γ22 Γ11 + Γ21Γ22 + Γ23, 2 − Γ22 Γ13 + Γ23Γ32 ,
1 1 2 1 3 1 3 3 3
dB dA dB dA 1
=− ⇔ =− ⇒ B = ⇒ AB = 1 ,
Bdr Adr B A A
upon solving the simple differential equation. We use this solution to simplify equation 3:
A′r A′r ′
0= + + A − 1 ⇔ 1 = A′r + A = A′r + Ar ′ = (Ar )
2 2A
We now integrate:
∫ dr = ∫ dr (Ar )dr ⇒ r + k = Ar ,
d
element equation at the start of this section, we have the Schwarzchild solution:
−1
2GM
c 2 dτ 2 = 1 −
2 2 2GM
c dt − 1 − 2
(
dr 2 − r 2 dθ 2 + sin 2 θdφ 2 . )
rc 2 rc
We wish to determine the motion of the particle relative to coordinate time, denoted by t. The equation would
give the path of the particle in accordance to what other observers would see if they thought they were in a
gravitational field.
With the above said, the following equation can be immediately written, transforming from proper time to
coordinate time:
d 2xµ ν
µ dx dx
σ d 2t dt dx
2 µ
+ Γ = 2 .
dt 2
νσ
dt dt dτ 2 dτ dt
First, we expand and consider its spatial components:
d 2 xi j
i dx dx
k
i dx dx
0 k 0
i dx dx
0 d 2t dt dx
2 i
+ Γ jk + 2Γ0 k + Γ00 = 2 .
dt 2 dt dt dt dt dt dt dτ 2 dτ dt
0
j
i dx dx dx 0 dx k
One of the Christoffel symbols has a coefficient of two since Γ j 0 = Γ0i k . We simplify the
dt dt dt dt
equation:
d 2 xi j
i dx dx
k
dx k d 2t dt dx
2 i
+ Γ + 2 Γ i
c + c Γ
2 i
= 2 .
dt 2
jk
dt dt
0k
dt
00
dτ 2 dτ dt
We assume that the gravitational field is quasi-static, i.e. that it doesn't change with respect to time. Therefore,
any derivatives of the metric tensor with respect to time can be left out. Now we evaluate the connection
coefficients:
1 iρ g iρ ∂g ρj ∂g ρk ∂g jk
Γ = g Γρjk =
i
+ − ρ . These derivatives are quite small, and can be neglected.1
2 ∂x k ∂x j ∂x
jk
2
g iρ ∂g ρ 0 ∂g ρk ∂g 0 k ∂hρ 0 ∂h ρk ∂h
2Γ0i k = 2 k +
2 ∂x
( )
− ρ ≈ η iρ + h iρ k + 0 − 0ρk
∂x 0
∂x ∂x ∂x ∂x
∂h0 ρ ∂h ρk ∂h ∂h ∂h ∂hµν
( )
≈ η iρ + h iρ k + 0 − 0ρk ≈ −δ is 0ks − 0sk on neglecting terms involving .
∂x ∂x ∂x ∂x ∂x ∂x 0
g iκ ∂g κ 0 ∂g 0κ ∂g 00 δ iz ∂h00
Γ00i = + − κ ≈ .
2 ∂x 0 ∂x 0 ∂x 2 ∂x z
Now we need to evaluate the RHS of the equation. We start by first looking at the following line element:
dτ dx µ dxν dx µ dxν
2
c dτ = g µν dx dx ⇔ = 2 (η µν + hµν ) = 2 (1 + hµν )
2 2 µ ν 1 1
. Neglecting
dt c dt dt c dt dt
terms involving the spatial components, which are small in comparison to the temporal components, we get:
dτ dτ
2
dt dt
d τ dh00
2
dh
Now 2
= = c 000 , so the RHS of our major equation becomes:
dt dt dx
dh
i 2
c 000
dx d t dt
2
dx i dx i
≈
dx dh00
(
≈ c 0 − 2 h00
1 1
) .
dt dτ 2 dτ 2 1 + 12 h00 dt dx dt
This is negligible since it involves temporal derivatives of the gravitational field.
d 2 xi ∂V
= −δ iz z . We want the metric tensor to be flat when there is no gravitational field present. The
dt 2
∂x
1
The metric tensor g µν ≈ η µν + h µν , where η µν is the familiar metric tensor of Special Relativity, and
h µν are small terms which include the action of any gravitational fields which may be present, and are small in
µν
comparison to the η in weak gravitational fields, as opposed to the awesome sucking power of a black hole!
equation for the potential leads us to the expression g 00 = η 00 + h00 = 1 + 2V c 2 . To finally get the actual
expression for the potential in terms of mass, we need to use the line element from the Schwarzchild solution.
The temporal component gives V = − GM r . With this expression, we can easily obtain an approximation of
the gravitational force:
F = − m∇V = − GMm r 2 .
As a test for General Relativity, at velocities which are small in comparison to the speed of light, there must be
an approximation to Poisson’s equation: ∇ V = 4πGρ . To achieve this requires the weak field
2
G µν = κT µν
R µν − 12 g µν R = κT µν
g µν R µν − 12 g µν g µν R = κT µν g µν
R − 2 R = κT µν g µν
∴ R = −κT µν g µν
We substitute this back into our original equation:
R µν = κT µν + 12 g µν R = κT µν − 12 g µν κT µν g µν
We are approximating that the material energy tensor has a negligible value in for all µ, ν except when µ = ν =
0, so we get:
R 00 = κT 00 − 12 g 00κT 00 g 00 = 12 κT 00 .
We now make the Ricci tensor covariant:
g 00 g 00 R 00 = 12 κT 00 g 00 g 00 ⇔ R00 = 12 κT00
Weak field approximation.
Rµν ≈ 1
2 (g αα , µν − gνα , µα − g µα ,να + g µν ,αα ) ≈ 12 g µν ,αα
1 2
∴ R00 ≈ 12 g 00,αα ≈ 12 ∇ 2 g 00 = ∇V
c2
1 2
∴ 2
∇ V = 12 κT00 = 12 κρv0 v0
c
The particles are traveling at the travelling at the speed of light through time, so we get:
∇ 2V = 12 κρc 4
This must equal Poisson’s equation, stated earlier
8πG
This gives a value κ= .
c4
2
Note that this particular calculation would be shorter if we took the Einstein tensor and the material energy
tensor to be covariant as opposed to contravariant, but due to the actual form of the material energy tensor, I
prefer it to be contravariant.