Gravity Accuracy

[MythProgrammer]

I was playing with the sandbox, which is great but I was playing with the solar system and deleted the sun. It appeared all the planets were effected simultaneously and wanted to know if the sandbox would change to use the method where gravity effects the planets at the speed of light. For example, if the sun was deleted, it is expected to effect the Earth after 8 minutes. Just wanted to inquire great stuff.

[Dan Dixon]

Right... If the sun suddenly disappeared it should take 8 minutes before Earth would feel that effect.

Currently in Universe Sandbox gravity is instantaneous (contrary to reality).

It's a great question MythProgrammer... unfortunately moving to a simulation model where gravity moves at the speed of light would be a fairly complicated addition. It's a very low priority at the moment.

[mattie]

Does gravity actually propagate at the speed of light?

[Dan Dixon]

That's the theory:
http://en.wikipedia.org/wiki/Speed_of_gravity

[tony873004]

That's the theory:...

If it were significant to your software, you could prove this by both coding for it, and ignoring it, and seeing which model best fits actual observations.

In post 3 of this BAUT thread, it is Publius claims that adding the delay to a Newtonian-derived system only destabilizes things:  http://www.bautforum.com/questions-answers/76306-yet-another-gravity-question.html


Added by Dan on 2012-04-29, working link:
http://web.archive.org/web/20080709042103/http://www.bautforum.com/questions-answers/76306-yet-another-gravity-question.html

[Dan Dixon]

Their conclusions are interesting.

"GR (General Relativity) gravity does indeed propagate at the speed of light through space-time (it travels along null paths). But Newtonian gravity must be instantaneous. If you naively add a propagation delay to that, having the acceleration point to the light speed retarded positions of the sources, you'll indeed find that orbits aren't stable and quickly decay."

Thanks for the information Tony.