What would happen if you put a ship on the borderline of SOI?

[EditorRUS]

So let's imagine we have a really long ship and somehow put it at exactly 84 159 286 m obit (hyperedit, anyone?) so that one part of the vessel would lie inside Kerbin's SOI and another inside Kerbol's SOI. What would happen here? My assumption is that these two parts would start to experience different gravititional pull and basically the vessel would get phantom thrust until the entire ship is in only one SOI.

Edited December 9, 2015 by EditorRUS

[Jeanjvs]

I believe there are two possibilities:

1: The game would crash

2: The Kraken would be awakened, causing some sort of apocalyptic glitch. Then the game would crash.

[RocketPilot573]

I would assume the location of the center of mass (CoM) of the vessel would determine which SOI it's in, so you can only be in one SOI at a time. Try undocking the ship right down the middle and see what happens. Though even being near the SOI border could summon the Kraken. I should try this sometime, maybe tomorrow! 

[Choctofliatrio2.0]

What I've wondered, and I wonder if this happens in real life, since you move at different speeds depending on how high up an orbit you are in, if you had a very long ship, and had one end closer to Earth, would it start to move faster than the farther away end? How drastic is the change?

[Temstar]

In real life a very long ship in LEO will orientate on its own so that its long axis points in radial direction. This is called gravity-gradient stabilization. In higher orbit the effect will be weaker and at edge of the SOI you will notice negligible tidal effect so nothing special will happen. You orbit will be easily perturbed by other stuff and you will tend to go into a sun orbit or go further in into Earth's gravity well.

[Clear Air Turbulence]

3 hours ago, EditorRUS said:

So let's imagine we have a really long ship and somehow put it at exactly 84 159 286 m obit (hyperedit, anyone?) so that one part of the vessel would lie inside Kerbin's SOI and another inside Kerbol's SOI. What would happen here? My assumption is that these two parts would start to experience different gravititional pull and basically the vessel would get phantom thrust until the entire ship is in only one SOI.

Try it and see?

[CalculusWarrior]

As the ship's altitude is shown with respect to its centre of mass (I believe, not 100% certain, but the game definitely models it as a point when dealing with orbital mechanics), whenever this point goes above the Sphere's limit, you leave that Sphere. It doesn't matter if your ship is 20 metres or 2 km long.

[Geher]

Would be cool if it would be determined by the active part. Then you could just change the SOI by clicking Control from here .

[Hupf]

5 hours ago, RocketPilot573 said:

I would assume the location of the center of mass (CoM) of the vessel would determine which SOI it's in, so you can only be in one SOI at a time. Try undocking the ship right down the middle and see what happens.

When testing undocking and also docking, be sure that a SOI shift will be involved: e.g. make sure that when undocking, focus will be switched to the part now outside of the SOI (if previously inside) or when docking, approach from outside the SOI and dock with a much heavier part that is inside. That way you could check if the rapid vehicle assembly code will properly handle SOI switches...

[radonek]

7 hours ago, RocketPilot573 said:

 Try undocking the ship right down the middle and see what happens.

Now this is an interresting idea. You'd have one ship in SOI, other outside it, BUT both in same physics bubble. I smell Kraken.

[Red Iron Crown]

Gravity magnitude and direction is equal for all parts on a vessel. When on rails it is calculated for the root part's position. When off rails it is calculated for the CoM's position (used to be root part) and applied to all parts. So there are no tidal effects, and a vessel can only be in one SoI at a time. 

[FyunchClick]

I wonder if the entire bubble isn't either in one SOI or the other (and transfers at once too, including all objects in it). It's the only way I can imagine implementing that system given the SOI premise KSP uses, since it's also your reference wrt position and speed vectors, I imagine having to calculate that for two frames of reference for the objects in a physics bubble kinda defeats the purpose of having a SOI.

[Red Iron Crown]

The whole bubble's gravity is determined as mentioned above when off rails.

[Tw1]

This is some science I can get behind. 

Experiment In Progress! Shaping up to be a kracker. 

Edited December 9, 2015 by Tw1

[adsii1970]

I have actually thought about this since .22 - with the idea of putting a space station at the edge of Kerbin's SOI. Not sure what I would use it for, but it would be neat to do.

[J.Random]

39 minutes ago, FyunchClick said:

I wonder if the entire bubble isn't either in one SOI or the other (and transfers at once too, including all objects in it). It's the only way I can imagine implementing that system given the SOI premise KSP uses, since it's also your reference wrt position and speed vectors, I imagine having to calculate that for two frames of reference for the objects in a physics bubble kinda defeats the purpose of having a SOI.

 

29 minutes ago, Red Iron Crown said:

The whole bubble's gravity is determined as mentioned above when off rails.

Are you sure? I distinctly remember my EVA kerbal suddenly shooting away from the ship after crossing the SOI boundary.

[Red Iron Crown]

7 minutes ago, J.Random said:

Are you sure? I distinctly remember my EVA kerbal suddenly shooting away from the ship after crossing the SOI boundary.

Were you timewarping, by chance? It's my understanding that it will do so when on rails, but not off. When on rails each individual vessel has its own set of orbital parameters and uses its own root part for position and gravity determination. When off rails all gravity in the bubble is the same as at the bubble's center.

[KerBlammo]

9 hours ago, Temstar said:

In real life a very long ship in LEO will orientate on its own so that its long axis points in radial direction. This is called gravity-gradient stabilization. In higher orbit the effect will be weaker and at edge of the SOI you will notice negligible tidal effect so nothing special will happen. You orbit will be easily perturbed by other stuff and you will tend to go into a sun orbit or go further in into Earth's gravity well.

The classic Larry Niven Sci-Fi story Neutron Star is basically about this effect (and the Kraken).

https://en.wikipedia.org/wiki/Neutron_Star_(short_story)

 

[EditorRUS]

Quote

Gravity magnitude and direction is equal for all parts on a vessel. When on rails it is calculated for the root part's position. When off rails it is calculated for the CoM's position (used to be root part) and applied to all parts. So there are no tidal effects, and a vessel can only be in one SoI at a time. 

What if somehow CoM manages to get into the very exact borderline of SOI where it's neither in Kerbol's SOI nor in Kerbin's SOI?

[Red Iron Crown]

4 minutes ago, EditorRUS said:

What if somehow CoM manages to get into the very exact borderline of SOI where it's neither in Kerbol's SOI nor in Kerbin's SOI?

It can't be. Precision of location is finite and the bubble center will always fall in one SOI or the other.

[Temstar]

23 minutes ago, KerBlammo said:

The classic Larry Niven Sci-Fi story Neutron Star is basically about this effect (and the Kraken).

https://en.wikipedia.org/wiki/Neutron_Star_(short_story)

 

This tidal effect is also the exact same thing that causes spaghettification around black holes. Only around a black hole the gravity gradient is so enormous that not only does it rip apart your body and your spacecraft (overcoming the electromagnetism that holds molecules together) into ribbons, it eventually overcomes even the strong force and rips apart the nucleus of all the atoms falling in.

[Tw1]

Conclusion: You can't get there.

Hyperedit refuses to go that high. Determining the height by going to an escape trajectory, then circularising can result in interesting weirdness, but the escape seems to happen at different heights. 

Similar results seem to happen when placing into orbit using file editing. This is probably not plausible, as the SOI doesn't seem to act like an exact boundary.

Having to objects right next to each other didn't result in anything special most of the time. One flies away, one doesn't. Except one time when warping. That made one object disappear. 

May write it up in more detail if I can be bothered later.

[MarvinKitFox]

For purpose of ENTERING a SOI the radius is a teensy bit less than for purpose of EXITING the same SOI. A difference of a few tens of meters

If you are skimming part a planet's SOI, you will either hit the SOI and stay in it for several minutes, or miss it altogether. It will *never* flick you in and out in a small few seconds.

(tested with a 10-ton ship, burning a single ion drive for as short a duration as I could manually manage, when 10 minutes from SOI entry.

The displayed velocity did not change even in the 0.01 digit on KER's display.

[J.Random]

3 hours ago, Red Iron Crown said:

Were you timewarping, by chance? It's my understanding that it will do so when on rails, but not off. When on rails each individual vessel has its own set of orbital parameters and uses its own root part for position and gravity determination. When off rails all gravity in the bubble is the same as at the bubble's center.

Now this is something I don't recall even vaguely, it was long time ago and I simply don't repeat that mistake anymore. So, yeah, I guess you're right.

[Dispatcher]

Are you looking to find and exploit Lagrangian points?