Whirligig Girl Posted March 11, 2014 Share Posted March 11, 2014 I'll keep this short.IRL, a small asteroid and a large ship orbiting it would wobble just a little bit, probably not noticable to the naked eye. Things attract each other, things have barycenters, even in 2-body physics, becuase we're only talking about 2-body physics.In KSP, all planets are on rails, so no matter how massive your ship is, it's not going to make Gilly wobble AT ALL. Only 1 body of the two bodies is actually having gravity simulated at a time, that's the spacecraft. Gilly is not attracted to the other ship, even if you use hacked parts with masses greater than that of the Sun. Of course, the same is said for Kerbin, Mun, Minmus, Eve, Duna, etc... Huh. I thought I said this would be short. Oh well. Link to comment Share on other sites More sharing options...
Kasuha Posted March 11, 2014 Share Posted March 11, 2014 Assuming your ship is built of 4000 jumbo tanks, it has mass 144,000 tons.Gilly has mass 124,205,120,000,000 tons.Well... I think it doesn't really matter Gilly is on rails. Link to comment Share on other sites More sharing options...
Nathair Posted March 11, 2014 Share Posted March 11, 2014 In KSP, all planets are on rails, so no matter how massive your ship is, it's not going to make Gilly wobble AT ALL.Gilly has a mass of about 1.2 million million million kilograms. Exactly how massive is that ship you're flying that it would induce a detectable wobble?Edit:Ninja'd! Link to comment Share on other sites More sharing options...
MalfunctionM1Ke Posted March 11, 2014 Share Posted March 11, 2014 Harvester talked on UNITE 13 about how and why KSP works how it does.Enjoy! Link to comment Share on other sites More sharing options...
Whirligig Girl Posted March 11, 2014 Author Share Posted March 11, 2014 Videos don't work on my slow internet, can you give me the gist of it? Link to comment Share on other sites More sharing options...
problemecium Posted March 11, 2014 Share Posted March 11, 2014 Someday I will watch that whole video.For now, to which part should we skip to hear about 1/2/n-body gravity? Link to comment Share on other sites More sharing options...
MalfunctionM1Ke Posted March 11, 2014 Share Posted March 11, 2014 The whole Video is all about why there is no n-body-physic.Skip to around 18:00.@gregroxmunyou might install a download-Manager with which you can download Videos from YouTube.It really helps with a very slow Internet because Youtube doesnt buffer the whole Video anymore. Link to comment Share on other sites More sharing options...
metaphor Posted March 12, 2014 Share Posted March 12, 2014 Nice video! That should be required viewing for anyone wanting to make a Plugin mod. Link to comment Share on other sites More sharing options...
Red Iron Crown Posted March 12, 2014 Share Posted March 12, 2014 I find the physics model in KSP a satisfactory approximation of reality. The only significant thing n-body physics would give us is Lagrange points, which are interesting but not overly useful. Link to comment Share on other sites More sharing options...
PhaserArray Posted March 12, 2014 Share Posted March 12, 2014 Gilly has a mass of about 1.2 million million million kilograms. Exactly how massive is that ship you're flying that it would induce a detectable wobble?Anyone could mod a Jumbo Fuel tank to have the same mass as Gilly and then HyperEdit it into Gilly orbit. Link to comment Share on other sites More sharing options...
Seret Posted March 12, 2014 Share Posted March 12, 2014 I'm not quite sure what the OP's point is TBH. Yes, KSP physics model is a rough approximation of reality. That's not really the point though, the question is whether it's fit for purpose. I'd say, yes, it definitely is. There are some things that are definitely lacking (eg: re-entry heat) but the gravity side of it does exactly what it's required to do. Link to comment Share on other sites More sharing options...
Kasuha Posted March 12, 2014 Share Posted March 12, 2014 What we have is a 2-body gravity in a system where mass of one body (spaceship) is negligible compared to mass of the other body (planet). That is actually pretty close to what you actually get in real world when close enough to any planet or moon or asteroid. Gilly's wobbling by hundredths of millimeter around common barycenter of it and your spaceship really does not play any role. Probes to way less massive asteroids (e.g. Eros) do not move them around measurably, either. The pressure of solar radiation is affecting that asteroid's trajectory more than the presence of the probe body. Link to comment Share on other sites More sharing options...
PUNiSH3R Posted March 12, 2014 Share Posted March 12, 2014 The issue would be with larger systems like Duna and Ike, where your barycenter ends up 186 km off Duna's CoM. While I will admit to [badly] wanting to 'set-and-forget' a trojan refueling station at L4 around Kerbin, n-body would introduce far more issues than features... Link to comment Share on other sites More sharing options...
Justin Kerbice Posted March 12, 2014 Share Posted March 12, 2014 I think IMHO physics are already a serious bottleneck in KSP, no need FOR NOW to add more "weight" to the physics engine resource needs as the balance is far more over already broken. Unless you only play with 10 parts vessels. Link to comment Share on other sites More sharing options...
FenrirWolf Posted March 12, 2014 Share Posted March 12, 2014 Clearly we do have N-body physics!...For when N = 1 =P Link to comment Share on other sites More sharing options...
Red Iron Crown Posted March 12, 2014 Share Posted March 12, 2014 While I will admit to [badly] wanting to 'set-and-forget' a trojan refueling station at L4 around Kerbin, n-body would introduce far more issues than features...You can approximate L4 and L5 by just putting the station in the same orbit as the smaller body, just 60 degrees ahead of or behind it. Link to comment Share on other sites More sharing options...
Guest Posted March 12, 2014 Share Posted March 12, 2014 The whole Video is all about why there is no n-body-physic.Well, I call all this stuff BS. "N-body is impossible/unfeasible/not good for the game"? Squad, meet Principia: http://forum.kerbalspaceprogram.com/threads/68502-WIP-Principia-N-Body-Gravitation-and-Better-Integrators-for-Kerbal-Space-ProgramAn undergrad math student just proved better at physics than anyone else on the team. It's currently in early alpha, but does it's job well, despite being built upon KSP engine, not into it. Turns out that numerical integration, done right, is quite manageable and can be used for both the real and accelerated time. When it's done and released, I'm gonna suggest Squad to integrate those integrators into KSP. Link to comment Share on other sites More sharing options...
Camacha Posted March 12, 2014 Share Posted March 12, 2014 (edited) Well, I call all this stuff BS. "N-body is impossible/unfeasible/not good for the game"? Squad, meet Principia: http://forum.kerbalspaceprogram.com/threads/68502-WIP-Principia-N-Body-Gravitation-and-Better-Integrators-for-Kerbal-Space-ProgramAn undergrad math student just proved better at physics than anyone else on the team. It's currently in early alpha, but does it's job well, despite being built upon KSP engine, not into it. Turns out that numerical integration, done right, is quite manageable and can be used for both the real and accelerated time.\I tend to agree. There are numerous other desktop simulations available that show that it is quite possible to do, so using that as an argument is not really convincing. There might be other reasons, but I can not think of them. Edited March 12, 2014 by Camacha Link to comment Share on other sites More sharing options...
Guest Posted March 12, 2014 Share Posted March 12, 2014 There are no other reasons, Unity can do all that (as shown by Principia) and do it in a playable way. The real explanation is "Squad doesn't have a dedicated mathematician/physicist on payroll". They should get one, and given their recent cooperation with NASA, it should be rather easy. Scientists are, for most part, nerds, and even in my institute (I'm a physics student) there's a "computer theoretical physics" specialization, which sounds like the exactly right sort of people to implement sympletic integrators into a game. Link to comment Share on other sites More sharing options...
Red Iron Crown Posted March 12, 2014 Share Posted March 12, 2014 Do we know how n-body affects performance? Won't an orbit have to be calculated for each part of a craft? Link to comment Share on other sites More sharing options...
RC1062 Posted March 12, 2014 Share Posted March 12, 2014 It isn't a case of can't do, it's a case of don't need to.Some things can be left out because it will just become a pain for the user, so why bother? For example why not add a non-uniform gravitational field? Because it would be a pain for users and really add very little. Link to comment Share on other sites More sharing options...
*Aqua* Posted March 12, 2014 Share Posted March 12, 2014 An undergrad math student just proved better at physics than anyone else on the team.Did you take a look a these unstable orbits? How am I supposed to leave a craft in orbit, knowing it will do everything but not staying there?N-body-physics mean orbits will not last. They will resonate with celestials changing the paths up to a point where they will quickly end in a complete chaos. There is no way for a player to set up an orbit which will last an in-game year. Link to comment Share on other sites More sharing options...
Wallace Posted March 12, 2014 Share Posted March 12, 2014 Did you take a look a these unstable orbits? How am I supposed to leave a craft in orbit, knowing it will do everything but not staying there?N-body-physics mean orbits will not last. They will resonate with celestials changing the paths up to a point where they will quickly end in a complete chaos. There is no way for a player to set up an orbit which will last an in-game year.We would need to have new AI that could run in background craft to keep satelites in orbit using their rcs. This would certainly add a new level of challenge, but I suspect it would all get out of hand soon enough. Link to comment Share on other sites More sharing options...
Camacha Posted March 12, 2014 Share Posted March 12, 2014 (edited) Do we know how n-body affects performance? Won't an orbit have to be calculated for each part of a craft?I think you could get away with calculating the centre of mass (as is already done) and using that as a single point in the calculations Maybe there are some exceptions, but you could probably work around those.I would look at other simulations (such as Universe Sandbox) to gauge performance. Edited March 12, 2014 by Camacha Link to comment Share on other sites More sharing options...
Seret Posted March 12, 2014 Share Posted March 12, 2014 It isn't a case of can't do, it's a case of don't need to.This. KSP is a game. Increasing simulation accuracy past a certain point doesn't really bring more fun. Link to comment Share on other sites More sharing options...
Recommended Posts