How Can Numerical Solutions to General Relativity Enhance Computational Physics?

[lokofer]

Hello could someone give some info about the "Numerical solution" to GR...is this a field of "Computational Physics"?..

- What i know is that you take the Hyper-surface, and you " split " it into triangles..and use the ¿angles? of every triangle as finite-coordinates..then you get a problem with finite degrees of freedom...but What happens with the metric, Riemann Tensor Energy-momentum tensor in this discrete space-time?..could you use discrete espace but continuous time so the usual Einstein Lagrangian becomes a finite one in the form:

$$L(q_i ,\dot q_i ,t)$$ so it's easier to "Quantize" than the continuous one?..

- Main questions: how do you define $$g_{ab}$$ $$R_{ab}$$ and other quantities into a finite "triangularized" surface..thanks

 

[robphy]

Look up "Regge Calculus".

 

[lokofer]

Look up "Regge Calculus".

I was afraid of this answer... i have looked it up in "Wikipedia" and "Arxiv.org" but i don't see or can't understand the explanation...or how you recover the Riemann Tensor in the end...