tidal forces and unstable otbits

[noobsrtoast]

if they ever add them how do you think that they would work. in my opinion the places it would be most noticeable would be laythe,jool, kerbin,eve and gilly, and duna and ike on kerbin there would probably be high and low tide and the mun would gradually be moving away from kerbin im not sure if the presence of minmus would affect what kind of tidal forces kerbin endured but I bet it would be something similar to a spacecraft receiving a gravitational assist from kerbin the planet would be affected minutely while minmus got affected greatly same news for gilly this could create a fun new game mechanic in my opinion attempting to save the muns from escaping there orbits.

different news with ike and duna due to ikes size im pondering whether or not something that size could orbit duna without destabilizing its orbit in the first place probably not...if someone could explain how something like that might work that would be great.

then tidal forces could contribute to the reason why life wouldn't be at laythe tidal forces hurt the planet,cause volcanism,mess things up for whatever would have been living there then you got radiation

as for jool it would exert massive tidal forces on things I would assume I haven't a clue whether it would have the capability to rip a ship apart or not..

anyway looking for your ideas on this thanks

[BudgetHedgehog]

So basically, N-Body physics?

Yeah, they won't add it. Too much of a strain on an already overworked system and I don't think it'll be terribly fun for players either ("god dammit, gotta tweak my stations height AGAIN")

[PDCWolf]

Before they close this i want to clear a few things up:

•N-body won't make orbits unstable unless you purposely place craft in places where they could get any "perturbance" not even on the duna-ike system would they get flung out unless you want it to happen or carelessly place them there (we are talking about high orbits in this case)

•N-body won't cause orbital decay (No alterations to the atmosphere algorithm) so there's no need for the station-keeping everybody is afraid of, the only case would be launching on a retrograde orbit and the decay would take thousands of years.

•N-body could potentially cause apsidal recession, yes, but it would be noticeable only on high energy/eccentricity orbits. The almost-perfect circular orbits KSPers normally use would have a totally negligible effect

•N-body won't cause the "chaotic orbits you can see on this mod's thread", those orbits are seen with a changed plane of reference (one of them is a kerbin orbit seen from the mun's rotating plane of reference, so yeah, it obviously looks crazy), freaking read the thread before coming to conclusions. The Jool system would be fun tho and those things could potentially happen if you either want them to happen or if the entire system aligns for it to happen. In fact, you can already get pretty crazy stuff in the Jool system with patched conics.

•N-body doesn't tax the CPU that much, that's why integrators and symplectic integrations exist.

•N-body won't make neither Jool or Duna system a nightmare. You ever tried playing on jupiter/saturn with Orbiter? yeah, nothing "crazy" or "nightmarish" happens either, unless you want it to happen or you want an unrealistically high orbit over a moon on a multiple-moon system.

So yeah, most concerns around N-body are either fallacies or plain ignorance (or maybe fear of change).

[vexx32]

One point that seemingly never gets mentioned is that N-body physics quickly (very quickly) becomes quite CPU heavy. Yes, there are ways to do it quite easily with today's computing power... in real-time (mostly)... however... the more bodies you must account for, the more complex the situation gets -- exponentially, if I recall correctly. In addition, it will quickly become far too much for pretty much any machine in existence to simulate real-time if you have to do it for every craft in the game... which means unfocused craft will probably still have to use patched conics (the 'rails' system)... which means that you'd only be able to utilise the useful bits of n-body simulations (i.e., Lagrange points) while you are focused on that craft. Soon as you switch craft, it'll be in a 'standard' orbit which will very likely not stay in the L-point. In addition, you couldn't run n-body physics at the speeds required to be able to use time-warp beyond about 100x... and even then, it'd be unstable. Time warp would also have to switch back to the regular patched conics 'rails' system...

All of that basically means the n-body simulation would only work while you are not time-warping, and have the craft in focus... and since the primary use of L-points is to have craft sitting there for long periods of time for various purposes (with some small adjustments every now and again, as most L-points are not truly stable), the limitations of the system would make it, well, useless...

That said, the implication of this is an n-body simulation... which is on the [thread=30064]What Not to Suggest List[/thread]. As such, we're gonna have to close this for now -- it's been brought up a million times (well, it feels like it, anyway), and every single time, the above limitations render the discussion either useless or destructive as arguments ensue, without many real ideas of how to circumvent the limitations such a simulation would impose. Mind you, the limitations are largely due to the engine KSP is presently using -- namely Unity, and PhysX. Some other engines could potentially work (good luck finding them useable enough), but... changing engines at this stage in development would mean basically recoding almost everything from scratch. Not the best of plans, really.

All that said, I think that's probably enough on this for now... let's close this, shall we?