A question on n-body physics

[Corbald]

Well, I'm not getting the joke, sorry! ;( None the less, I have to stress that using the quantum mechanical principles to make a structure that still acts in exactly the same way as a classical transistor is miles away from what the D-Wave 2 does. Sorry if I'm not following where you're trying to lead me, but don't fret, it's not uncommon for me. :/

[dagothar]

I get why we won't probably see n-body physics in KSP and I'm not going to press it it any way, but I don't agree that involved math is really that complex. Feynman shows in his books (although in really simple and probably not accurate enough way) how to do orbit calculations on a piece of paper. Hell, I did that at times when I was really bored. And I guess it could be optimised to a point of feasibility (for Lagrange points it would be probably enough to just solve 3 bodies problem - no biggie). Anyway, if it's closed, it's closed.

I just wonder if there is a possibility of modding this in?

[artao]

Along with L-points, N-Body would allow for the fun of using the Interplanetary Transfer Network (aka Gravity Tunnels)

http://en.wikipedia.org/wiki/Interplanetary_Transport_Network

I too was going to bring up Orbiter being able to use N-Body and time-accel, but I see someone else did. .. I don't really understand the response to that tho. "Crappy orbits", which the poster clarified to "realistic orbits". Hmmm. .. While I do kinda like not having to do do regular orbital maintenance burns, on a game-play level at least (that WOULD get tedious with a large interplanetary fleet of spacecraft) I wouldn't exactly call that a crappy orbit. .. It's my understanding that Orbiter has the most accurate orbital mechanics available to the general public.

KSP of course isn't meant to. Orbiter wasn't developed as a "fun" game, but rather a research simulation. KSP certainly IS developed as a fun game, with a degree of realistic orbital mechanics. I think it works, but it sure would be nice to be able to use L-Points .. and Gravity Tunnels =D

[rkman]

We'll never get n-body physics for KSP because:

- It would be way to CPU intensive

- The math required is ridiculously complex.

Why would n-body physics be to CPU intensive and to complex for KSP but not for Orbiter?

Contrary to what some people seem to think n-body physics does not require a super computer.

It is only to CPU intensive to do it at high timewarp with sufficient accuracy. Which is good enough reason why it is not in KSP.

...Orbiter being able to use N-Body and time-accel...

Yes, but it does have issues with accuracy at high timewarp.

Also in Orbiter people do not typically have many ongoing missions at the same time as we have in KSP, the orbits of all of which would have to be calculated by n-body physics and would have to be highly accurate even at high timewarp.

[asmi]

The very basic math for n-body physics is incredibly simple - it's basically second Newton's law plus equation of motion. However the problem is with timewarp when dt becomes so large as orbits become very inaccurate, so better integration techniques have been devised, and subsampling is now an option since CPU performance is high enough to allow that. Orbiter comes with technical documentation that outlines this question very well, so I'd suggest reading it to everyone interested. I was actually considering developing mod for KSP that would introduce n-body physics, but never got enough time for that