A clarification of "On Rails"

[GarrisonChisholm]

Good evening. Knowing that the planets of Kerbol are "on rails", I understand that this means that they are immovable- 100,000 Mammoth engines with infinte fuel will not budge the tiniest moon from its track. Seeing that spent stages with sub-orbital trajectories persist until you "ride" them through enough atmospheric drag to cause them to terminate, I also concluded this applies to probes/ships that you were not "actively" piloting.

However, after 4 1/2 years of my (1.02) game, I noticed that my two earliest satellites which had been on highly elliptical, beginners-error trajectories, had been ejected just recently from the Kerbin SOI by Mun. So, my question is, what physics *are* applied to "stagnant" probes/vessels? Did these two satellites eject because I had just happened to be near them with something else, and so the game "noticed" them, or does gravity always apply to all vessels? Which, if so, begs the question of why atmospheric drag would be excluded from this application.

Edited October 23, 2015 by GarrisonChisholm

[JTpopcorn]

Generally, gravity always aplies to all veesels, even those that are on rails. The game does not do this with atmospheric drag because gravity is much easier to calculate than atmospheric drag, meaning that it is a performance saver.

-JT

[Red Iron Crown]

"On rails" means it stops using the Unity PhysX engine for trajectories and instead uses custom code for keplerian orbits. The *only* force considered is gravity, so a gravity assisted trajectory change like your satellites' ejections will still work. However, if an unfocused object falls below a certain altitude (based on pressure) on atmospheric bodies it will be deleted to simulate burn up. Are you 100% sure your suborbital spent stages persist after falling deep into the atmosphere?

[Claw]

Craft on rails are still subject to gravity physics. So orbits are still dependent on the SOI, and orbits that encounter another body's SOIs (such as a moon) will stilll be effected. This can change orbits, including flinging things out of the original SOI. Getting ejected from the solar system seems fairly unlikely though. So you either got really lucky, ot there was some other glitch (which happens from time to time) that sent them racing off.

Craft on rails are just packed and treated in such a way that they don't have physics applied to the rigid bodies directly. This means that thrust, aero forces, drag, etc are not calculated. That also means that things like Electrical Charge and resource harvesting aren't calculated either.

Cheers,

-Claw

[GarrisonChisholm]

I understand now- thank you one and all!

Are you 100% sure your suborbital spent stages persist after falling deep into the atmosphere?

The definition may be with "deep". I have not seen any persistent sub-orbital stages that got down to the 40's, but I did see one with with Pe of 58km, and numerous in the mid 60's.

[Red Iron Crown]

It's lower than that, a bit over 23km on Kerbin before it gets deleted: http://wiki.kerbalspaceprogram.com/wiki/Atmosphere#On-rails_physics

(Figures might be inaccurate, not sure if this changed with 1.0 but I don't think it did.)

[GarrisonChisholm]

Thank you for pointing out the figure RIC, I'll now have a benchmark to shoot for with my spent stages.

[Claw]

(Figures might be inaccurate, not sure if this changed with 1.0 but I don't think it did.)

The atmosphere profiles changed some, but I believe the altitudes are pretty much the same.

The big difference is the physics bubble extension. When dropping boosters on ascent, they won't necessarily insta-disappear at 23km if they are within ~23km of your active ship. (It used to be outside of 2.5km, but that range is bigger now.)

Cheers,

~Claw

[FancyMouse]

Then I'm confused... because the bubble radius seems 2.25km after the craft is in space (LS for rescue missions start calculating from that distance, and when approaching my complex bases, it's also the point where I start to see lag).

So KSP has a different bubble radius between launch and after getting into orbit?

[Claw]

So KSP has a different bubble radius between launch and after getting into orbit?

Different bubble for in atmosphere, and in space.

[magnemoe]

Different bubble for in atmosphere, and in space.

Yes, this is probably to avoid problem if you have multiple huge ship in low orbit, with 22 km its an chance multiple will load bringing total part count up to multiple thousands.