Skip to main content
14 events
when toggle format what by license comment
S Apr 9, 2023 at 20:13 history bounty ended Chris
S Apr 9, 2023 at 20:13 history notice removed Chris
Apr 2, 2023 at 21:52 vote accept Chris
Apr 2, 2023 at 21:51 answer added Ryan Budney timeline score: 1
Apr 2, 2023 at 21:46 comment added Chris @RyanBudney I see that makes sense. If you would like to expand on your comment with an answer I will award it the bounty
Apr 2, 2023 at 21:39 comment added Ryan Budney Right, your first paragraph, but adapted to the decomposition $TM \simeq E^t \oplus E^s$. This decomposition you get by arguing that the full-rank timelike (or spacelike) subspaces of a vector space with non-degenerate symmetric bilinear form, they are a convex space. So you piece together the decomposition of the tangent bundle fibrewise, with a partition of unity argument.
Apr 2, 2023 at 21:34 comment added Chris @RyanBudney Ok, so that's pretty much what I said in the my last paragraph?
Apr 2, 2023 at 21:33 comment added Ryan Budney Hi Chris, but I think it's the exact same argument. You have a Stiefel-Whitney class for each sub-bundle, the first sub-bundle is orientable if the $w_1$ class for that sub-bundle is trivial. The 2nd is orientable if its $w_1$ class is trivial. So if I understand which notion of orientability you are interested in, it would be the one for the timelike subspace, i.e. you want a timelike sense of direction.
Apr 2, 2023 at 21:01 comment added Chris @RyanBudney I don't think this is covered in Whiston's paper. He does not really mention time orientability in depth out side of the $(1,3)$ case
Apr 2, 2023 at 20:09 comment added Chris @RyanBudney was unaware of this paper, but I’ll check it out
Apr 2, 2023 at 20:05 comment added Ryan Budney Isn't this covered in George Whiston's paper "Lorentzian characteristic classes" GRG (1975)?
S Apr 2, 2023 at 18:36 history bounty started Chris
S Apr 2, 2023 at 18:36 history notice added Chris Draw attention
Mar 31, 2023 at 18:07 history asked Chris CC BY-SA 4.0