T1: 4. Covariant Derivative Flashcards
In what case is the coordinate derivative of a vector, a tensor?
Under linear coordinate transformations; the non-linear term in the product rule disappears.
Define the covariant derivative ∇ (or connection)
A linear operator which takes a vector to a rank [1,1] tensors and acts like a derivative.
What does it mean for a covariant derivative to ‘act like a derivative’
Obeys the Leibniz (product) rule for derivatives, with some scalar function f:
∇(fV) = (df)V+f∇V
What define a connection to be ‘torision-free’?
The connection coefficients are symmetric in the lower indices.
What property does a ‘torsion-free’ connection have?
Its action on a scalar (function) is commutative.
[∇_µ, ∇_ν]f = 0
What two properties does the Levi-Civita connection have?
- It is torsion-free (symmetry in lower indices of connection coeffs/Christoffels)
- It is ‘metric compatible’.
What does it mean for a connection to be ‘metric compatible’?
Its action on the metric vanishes:
∇_λ g_µν = 0
For a torsion-free metric, how many equations does the cov deriv of the metric give?
The number of components of the connection coeffs/Christoffels.
Does the covariant derivative commute with raising/lowering indices
Yes!
State the covariant derivative ∇_µ V^ν (on a vector)
∇_µ V^ν = ∂µV^ν + Γ^ν(µλ) V^λ
State the covariant derivative ∇_µ ω_ν (on a covector)
∇_ν ω_µ = ∂v ω_µ - Γ^λ(vµ) ω_λ
Generally, how many connection coefficients does an n dimensional space have?
n^3
Give an expression for the Christoffel
Γ_μν^σ = 1/2 g^σλ(∂μ g_λν + ∂ν g_μλ - ∂λ g_μν)
How do we find an expression for the Christoffels?
Give cov derivative of metric and find difference.