Relativity Notes
Local flatness theorem
Given a spacetime (with a manifold and a metric), at a point p in M, there exists a coordinate
transform such that the metric at P reduces to nu(ab), and the derivative vanishes.
Mathematically, this implies that a coordinate system exists such that the Christoffel vanishes – this
is called an inertial frame. Physically, this allows us to consider surfaces as locally flat and as such,
the equivalence principle can be applied.
Geodesics
Previously there was no method to compare a vector at p to a vector at q. This is solved using
parallel transport, which consists of moving the vector to different points, keeping it constant.
However, this is clearly not path invariant.
The geodesic is the generalisation of the shortest path, and it is the path between two points in
which the tangent vector to the curve is parallel transported. To find the equation of this curve, we
can derive the geodesic equation. Given a curve x a(λ), we require dT/dλ=0, where T is the tangent
vector. In flat space we could define this as
dT ∂ T ∂ x a
= =0
dλ ∂ x a ∂ λ
In curved space, analogously,
dTa ⅆ xa ⅆ xb
dλ
=T a ∇a T b=
ⅆλ
∇a
ⅆλ ( )
=0
We therefore have
ⅆ xa ∂ ⅆ xb b c
a dx d x
+ Γ bc =0
ⅆλ ∂ x a ⅆλ dλ dλ
d2 xa b c
a dx d x
+ Γ bc =0
d λ2 dλ dλ
Where the last step uses the chain rule, and relabels indices. This is the geodesic equation. The
geodesic hypothesis states that free particles follow the aforementioned equation of motion, which
represents motion under the influence of gravity. Here, the Christoffel symbol loosely represents the
‘acceleration due to gravity’. Evidently this is non-tensorial as in local frames we cn render the effect
null. Finally, the character of the path remains invariant, i.e. the term g abuaub stays invariant along
the path, where u represents the 4-velocity of the particle.
d a b d xc a b
( gab u u )= ∇ c gab u u =0
dλ dλ
The last step is trivial to check.
A more intuitive and equally useful method is the idea of extremising the proper time τ.
Let us vary the lagrangian
S=∫ L dλ
Using
1
L= gab ua ub
2
Varying, we have
b
1 d xa d xb c d xa d (δ x )
δS =∫ δL dλ= ∫ −∂c gab
2 (dλ dλ
δ x −2 g ab
dλ dλ )
dλ
b
d xa d xb c d xa d ( δ x )
¿∫ ( −1
2
∂c gab
dλ dλ
δ x −gab
dλ dλ )
¿ ∫❑