(anti)-symmetries of the Riemann curvature tensor

Click For Summary

Discussion Overview

The discussion revolves around the symmetries of the Riemann curvature tensor, specifically focusing on proving certain identities related to these symmetries. Participants explore both theoretical aspects and implications of these symmetries in the context of differential geometry.

Discussion Character

  • Technical explanation
  • Debate/contested
  • Mathematical reasoning

Main Points Raised

  • Some participants outline the known symmetries of the Riemann curvature tensor, including identities (a) through (d).
  • One participant expresses uncertainty about proving the identity Rijkl = -Rjilk, suggesting it should be obvious but finding it challenging.
  • Another participant points out that contravariant and covariant indices must match for the identity to make sense, indicating that the original identity may not be valid.
  • There is a query about deducing the identity Rijmm = 0 from the established symmetries, with a suggestion to use symmetry (b) and the properties of the metric.
  • A later reply confirms the approach using the metric's symmetry to derive the identity Rijmm = 0.

Areas of Agreement / Disagreement

Participants express differing levels of confidence regarding the proof of certain identities. While some agree on the established symmetries, the discussion on proving Rijkl = -Rjilk remains unresolved, with no consensus on its validity.

Contextual Notes

Some assumptions about the identities and their transformations are not fully explored, and the dependence on the properties of the metric is acknowledged but not resolved.

quasar987
Science Advisor
Homework Helper
Gold Member
Messages
4,796
Reaction score
32
The Riemannian curvature tensor has the following symmetries:
(a) Rijkl=-Rjikl
(b) Rijkl=-Rijlk
(c) Rijkl=Rklij
(d) Rijkl+Rjkil+Rkijl=0

This is surely trivial, but I do not see how to prove that

Rijkl=-Rjilk.

:(

Thanks.
 
Physics news on Phys.org
This is supposed to be obvious, according to a commentary in a book.
I can get

Rijkl=-Rjilk

(from memory) but not the required identity...
 
Contravariant and covariant indices must match on both sides of an equality (else the left and right hand sides do not transform the same way), so the original identity makes no sense. The identity in post #2 is as close as you can get.
 
True!

So do you know how we can deduce from identities (a)-(d) that

Rijmm=0

(summation over m implied)?
 
quasar987 said:
True!

So do you know how we can deduce from identities (a)-(d) that

Rijmm=0

(summation over m implied)?

Yes, use (b). Specifically use the fact that the metric is symmetric to show that

{R_{ijm}}^m = g^{mn} R_{ijmn} = - {R_{ijm}}^m .
 
mh, I see! Thanks fzero.
 

Similar threads

  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 9 ·
Replies
9
Views
4K
  • · Replies 6 ·
Replies
6
Views
4K
  • · Replies 3 ·
Replies
3
Views
6K
Replies
4
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 1 ·
Replies
1
Views
6K
  • · Replies 34 ·
2
Replies
34
Views
5K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 3 ·
Replies
3
Views
924