Patched Conics error if gravity is changed?

[Kergarin]

I was looking for some more challenge, and started a career game with gravity set to 2x using the alt-F12 menu.
Everything was working fine, i could fly to the mun without a problem.

But then i have upgraded the Tracking Station and Mission Control to Level2, so that patched conics were visible in map.
From this moment on there was a ghost mun, which was only visible in the patched conics (no matter if i use a manouever node or not) and this ghost mun affects my flight path and has SOI.
 

The position of this ghost mun changes, when the gravity is changed.

I would really like to get the game working correct on 2x gravity, because it's so much more fun.  

Stock KSP 1.3.1.1891

Edited February 11, 2018 by Kergarin

[OHara]

I see the same behavior, even in older versions.  It looks to me like the celestial-body orbits are not handled consistently by the hack-gravity cheat.   

I am trying to think of how all the orbits should be treated, to be consistent, and I'm not sure there is any good way.

Sliding the hack-gravity value leaves the orbital shapes the same for all craft that you are not currently flying, those that are 'on-rails'. Then when you switch to such a craft while gravity is hacked, its velocity is computed to be whatever gives that orbital shape in the hacked gravity.  To make physics consistent, celestial bodies would have their velocities scaled in a similar way.   But that would also scale all their orbital periods, get their positions out of sync with the calendar, and probably confuse players who hack gravity and then use a transfer-window planner.

Alternatively, KSP could keep the celestial bodies orbits the same shape and size and period, making celestial bodies move according the original, unhacked gravity.  But then as you exit the sphere of influence of a moon, you would find yourself in an orbit about the moon's parent that is very different form the moon's orbit, because you feel the hacked gravity and the moon feels the original gravity.

Maybe better if you use Kopernicus to make a system with doubled celestial-body masses.  Then the stronger gravity is consistent over everything, and persistent between starts of KSP.

[Kergarin]

1 hour ago, OHara said:

I see the same behavior, even in older versions.  It looks to me like the celestial-body orbits are not handled consistently by the hack-gravity cheat.   

I am trying to think of how all the orbits should be treated, to be consistent, and I'm not sure there is any good way.

Sliding the hack-gravity value leaves the orbital shapes the same for all craft that you are not currently flying, those that are 'on-rails'. Then when you switch to such a craft while gravity is hacked, its velocity is computed to be whatever gives that orbital shape in the hacked gravity.  To make physics consistent, celestial bodies would have their velocities scaled in a similar way.   But that would also scale all their orbital periods, get their positions out of sync with the calendar, and probably confuse players who hack gravity and then use a transfer-window planner.

Alternatively, KSP could keep the celestial bodies orbits the same shape and size and period, making celestial bodies move according the original, unhacked gravity.  But then as you exit the sphere of influence of a moon, you would find yourself in an orbit about the moon's parent that is very different form the moon's orbit, because you feel the hacked gravity and the moon feels the original gravity.

Maybe better if you use Kopernicus to make a system with doubled celestial-body masses.  Then the stronger gravity is consistent over everything, and persistent between starts of KSP.

Exactly. Thanks for your effort.

But have you also tried a career game with all facilities at level 1? Without patched conics, everything works fine for me.

 

Thanks for the Kopernicus idea. I really fell in love with 2x gravity. It feels much more realistic and challenging at payload fractions, rocket sizes, reentry heating and so on but without the need of part mods.

Edited February 12, 2018 by Kergarin

[OHara]

9 hours ago, Kergarin said:

But have you also tried a career game with all facilities at level 1? Without patched conics, everything works fine for me.

I have not tried 2× gravity, except for the patched conics.  I am a curious how KSP handles the orbits of planets.  If the Kerbin keeps its old orbit and old orbital speed, accelerating toward the sun with 1× gravity, then a craft barely escaping Kerbin would be an a solar orbit with periapsis near that of Moho as soon as it feels the 2× solar gravity.

[Capt. Hunt]

Celestial bodies in game follow pre-programmed orbits, they are not effected by gravity

[Kergarin]

11 hours ago, OHara said:

I have not tried 2× gravity, except for the patched conics.  I am a curious how KSP handles the orbits of planets.  If the Kerbin keeps its old orbit and old orbital speed, accelerating toward the sun with 1× gravity, then a craft barely escaping Kerbin would be an a solar orbit with periapsis near that of Moho as soon as it feels the 2× solar gravity.

 

5 hours ago, Capt. Hunt said:

Celestial bodies in game follow pre-programmed orbits, they are not effected by gravity

If you change the gravity, the planets speed changes too.

But nevertheless, if you leave kerbin soi, you end up in an solar orbit like the one of Moho (pe AND ap)

There seem to be some functions, which do and some which don't recognize the gravity slider.

[OHara]

1 hour ago, Kergarin said:

if you leave kerbin soi, you end up in an solar orbit like the one of Moho (pe AND ap)

Oh!  It seems that hack gravity is, as advertised, a hack.  It is kind of surprising that the SOI-changing system worked well enough to let you go to the Mun with a level-1 tracking station, given how confused it gets when patched conics are available.

I see that the speed displayed when you hover the mouse over a planed does increase with 2× hacked gravity, but the planet doesn't actually move any faster.  I can sit in an orbit the same shape as that of the Mun, and its reported speed is 750m/s just like my speed, but I catch up to the Mun.  

[Kergarin]

Well, ok. Thanks for your efforts. Looks like this gravity multiplier is not integrated in the same way into all calculations I think?

I will accept that this is a hack and not a consistent way to change gravity, and will look for another solution like Kopernicus or so.