Gravity simulation software?

[ceramicweasel]

Hello,

I'm not sure if this is the right place to ask, but can anyone here suggest some good gravity simulation software?

Ideally, I'd like to be able to model everything from a simple solar system to an entire galaxy, but I understand the number of calculations involved with galaxy simulations are... well, astronomical. Nevertheless, I want to do this for known systems (with a high degree of accuracy) as well as hypothetical ones.

Open source would be great, as that would allow me to tinker, but I'll take what I can get. :) It doesn't have to be pretty, or fast, so long as it's functional and accurate.

Any help is greatly appreciated.

[sheeny]

I had a free one on my old computer, which had a number of systems already modelled in it - (our) moon formation, solar system formation, 3 body instability demo, etc. You can also model what you like. I'll try to find out what it was called.

Al.

[sheeny]

You'll never guess what it's called...

Gravity Simulator! Who'd a thunk it?:P

I can't find the installer on my computer:(. I think I got a disk of free astro software with it on it. Try Googling... I'll have a look for a disk.

:)

Al.

[sheeny]

Not sure if this is the one...

http://www.orbitsimulator.com/gravity/articles/download.html

It looks very similar though.

Al.

[Auster1b]

Follow this post http://www.iceinspace.com.au/forum/showthread.php?t=38108 (hope I did this right)
Has a link to an astro software list, go to educational, even has Gravity Simulator.

Neil.

[g__day]

I had software (freeware that models two galaxies interacting. Google is your friend - there are a handful of interesting (but limited) programs out there.

The number of calculations depend on your algorithim but are order n cubed for a very unsophisticated approach (calculate acceleration and speed in x,y,z (6 operations) for every body interacting with every other body. Half that if you're clever as the forces between every two bodies are equal so 1/2 * 6 * (3 position + 3 acceleration) = 0rder 3 n^3

http://www.synthetic-reality.com/galaxy.htm


An n-tree alogrithim of greater sophistication might manage order n squared * log n without losing accuracy - which is quite an accomplishment. Better still if such a program is well written to use a modern computers CPU and GPU - you can probably manage simultaion between 50,00 to 100,000 stars in real time!

http://www.cs.princeton.edu/courses/archive/fall03/cs126/assignments/barnes-hut.html