Orbit Changes with No Force Applied

[Starhawk]

KSP Version: 0.1.1.0 21573
OS: Windows 10  21H2
CPU: AMD Ryzen 5 3600
GPU: Nvidia GeForce GTX1660 Ti
Expected Behaviour:  Orbit around airless celestial body should remain unchanged when no force is applied.
Observed Behaviour:  When craft is below a specific altitude and time warp is set to 1x, the periapsis will steadily decrease while the apoapsis steadily increases.  Changing time warp to any value other than 1x arrests the behaviour.  Behaviour reappears when time warp is returned to 1x.  At the Mun the altitude threshold is about 23 km.  At Minmus the altitude threshold is about 27 km.
Steps to Replicate:  Put a craft in Mun orbit with the periapsis below 23 km.  When (sea level) altitude drops below 23 km the orbit begins to change.  This behaviour stops when the altitude of the craft rises above 23 km.
Fixes/Workarounds:  Keep Mun orbit above 23 km (Minmus above 27 km) or use time warp values other than 1x.
Mods: None.

Tested at the Mun and Minmus.

This bug was present in 0.1.0.0 as well.

 

[Kurtvw]

I've also seen this behavior.

 

[nattkaill]

+1 Seeing the same.

[Xjph]

Experiencing this as well. Ap and Pe over minmus were drifting while no thrust applied.

[mciann]

I'm gonna call mine a duplicate of this. These are the exact same symptoms I was seeing, but it never occurred to me to try to establish a relationship to time acceleration or altitude.

 

 

[Tazooka]

+1 same

Captured it too

 

[Tedward]

I'd like to report the same behavior that I have experienced from day 1 of the game coming out.  I hopped the patch would include a fix.

I've always placed my Mun and Minmus orbits between 12 and 20k and this has become quite annoying.  Personally, it is my primary bug in the game.

I can deal with parts sometimes failing... I look at it as just an engineering flaw of the Kerbals and understand that It'll improve as they get better. 

 

On a positive note... the frustration of going to the mun or minmus has forced me to learn how to build/fly/land aircraft... something I was never liked doing in KSP1.

 

 

Edited March 19, 2023 by Tedward

[Yellowburn10]

So, I've been having this constant problem whenever I visit the Mun where my orbit is constantly changing on it's own, at least when I don't have time warp on. I've made a few tests with different kinds of vessels and have had the same issue. It's strange because I don't have the problem when I orbit kerbin, and while I haven't tested this anywhere else yet I'll be trying soon. Anyone else experiencing similar?

[The Aziz]

Yes, it's known. Current workaround is to orbit above 25km, it seems to be stable there.

[FirroSeranel]

I also have this problem, but it does happen on Kerbin, just much more subtly. ~0.5 to 1 m/s/s around Kerbin vs. 10-20 m/s/s around the Mun.

The below were taken about 15 seconds apart, with no control inputs and no thrust of any kind active.

[The Aziz]

What happens if you lock the SAS, instead of setting it to prograde?

[FirroSeranel]

Same. Also same with disabling SAS altogether.

Yet in Munar orbit, the oribits do degrade faster if I rotate the craft in various directions with reaction wheels. My guess is something isn't respecting the craft's overall center of mass and there's a phantom force, with the phantom force responsible for the always-on orbit wandering, and the CoM issue (similar to the old wobble kraken drive in KSP1) responsible for the faster decay while rotating.

It also seems to be related to the bug in which coming out of rails-warp gives a big impulse kick to the ship. The faster the orbital wander is happening, the harder the kick. Around Kerbin, coming out of time warp maybe nudges the nose aside a bit and SAS brings it right back, barely even a wobble. Around the Mun at low altitude (10-15km orbit), coming out of warp kicks the Pe to -10,000m or more, and sets the ship spinning at ~100 RPM.

Then again, if it were a phantom force, why would it go away above a certain altitude? And why would it _not_ be happening near the surface? When I'm going in to land, it feels perfectly normal, no 20 m/s/s accelerations from nowhere. Very odd...

My gut feeling is that it's to do with whatever implementation they did of the floating reference point system that lets KSP have both fine spatial resolution for crafts, but also billions of kilometers of "open world" game area... which is also the source of classic KSP tropes like orbit line jitter, and many of the "Kraken" bugs.

And maybe near the surface, it engages a higher-precision mode? And at high-orbit a semi-rails mode?

For now, I guess I'll use higher parking orbits, and call it an unintentional feature instead of a bug!

Like... an accidental implementation of relativistic frame dragging, amplified in gleeful mischievousness by the Kraken.

[FirroSeranel]

It seems to act like an outward radial force. It's doing it in a 30km munar orbit, but much more slowly, about like it did in LKO at 80km. Then the lower I got with my lander, the more intense it got, until at some point it stopped.

[pandaman]

I get this orbiting Mun and Minmus (not checked elsewhere yet). 

In my case Ap goes up and Pe goes down.

 

[FirroSeranel]

Yeah, always Ap up, Pe down, which is radial-out thrust. It seems like Ap goes up faster than Pe goes down usually, so some prograde component, too I think.

[Flush Foot]

KSP 2: 0.1.1.x [patch 1]
Windows 10 Home, 22H2 (19045.2728)
AMD Ryzen 9 5900X
ASUS TUF 3080 10G

Expected behaviour: vessel in space around atmosphereless body should not 'randomly' have orbits changing without thrust being applied.

Observed behaviour: Orbit is changing on its own, as if a small thruster (or "Kraken drive") is applying net-forces to the craft.

Steps to replicate: Place multi-craft vessel in low orbit around The Mun, detach probe/lander portion of main vehicle:

1. Decent chance this vessel winds up in a 'free' sub-orbital trajectory
2. Even if not, either immediately or upon landing and regaining altitude, your orbital parameters [or those of the 'mothership'] are likely to be fluctuating uncommanded.

Fixes/workarounds: none that I know of (except 'stay in time-warp as much as possible')

Mods: none

Other notes: as hinted above, though not in my video, as soon as I undocked the lander in LMO, it was on a suborbital trajectory. I believe I also noticed the mothership having changed / actively-changing Ap/Pe, but I was more focused on not crashing the lander.

Unlisted video on YouTube showing phantom forces at work in low Munar orbit

[Starhawk]

Overlapping bug reports merged.

[Yellowburn10]

13 hours ago, FirroSeranel said:

Same. Also same with disabling SAS altogether.

Yet in Munar orbit, the oribits do degrade faster if I rotate the craft in various directions with reaction wheels. My guess is something isn't respecting the craft's overall center of mass and there's a phantom force, with the phantom force responsible for the always-on orbit wandering, and the CoM issue (similar to the old wobble kraken drive in KSP1) responsible for the faster decay while rotating.

It also seems to be related to the bug in which coming out of rails-warp gives a big impulse kick to the ship. The faster the orbital wander is happening, the harder the kick. Around Kerbin, coming out of time warp maybe nudges the nose aside a bit and SAS brings it right back, barely even a wobble. Around the Mun at low altitude (10-15km orbit), coming out of warp kicks the Pe to -10,000m or more, and sets the ship spinning at ~100 RPM.

Then again, if it were a phantom force, why would it go away above a certain altitude? And why would it _not_ be happening near the surface? When I'm going in to land, it feels perfectly normal, no 20 m/s/s accelerations from nowhere. Very odd...

My gut feeling is that it's to do with whatever implementation they did of the floating reference point system that lets KSP have both fine spatial resolution for crafts, but also billions of kilometers of "open world" game area... which is also the source of classic KSP tropes like orbit line jitter, and many of the "Kraken" bugs.

And maybe near the surface, it engages a higher-precision mode? And at high-orbit a semi-rails mode?

For now, I guess I'll use higher parking orbits, and call it an unintentional feature instead of a bug!

Like... an accidental implementation of relativistic frame dragging, amplified in gleeful mischievousness by the Kraken.

I had something like that happen to me when I loaded a save of a craft in low mun orbit, the thing was suddenly spinning wildly when it loaded in. I had a passing thought that maybe the two bugs were connected, but it's only just a thought.

I'm going to be trying a few other tests soon just to try to recreate the bug in different scenarios, such as if inclination or highly eccentric orbits make any differences.

[Stoup]

Just wanna pitch in that I've experienced this bug too from day 1 and it's one of the biggest bugs keeping me disinterested from digging into any projects in the game

[paulysch]

I experienced the same on some occasions, especially if you are pretty low.

[Yellowburn10]

3 hours ago, Stoup said:

Just wanna pitch in that I've experienced this bug too from day 1 and it's one of the biggest bugs keeping me disinterested from digging into any projects in the game

I feel ya. I was attempting to do an apollo mission for one of my first missions, and my lander's delta v was only enough to do a return to low mun orbit, which complicate things if any low mun orbit is constantly shifting. That and having kerbals go EVA can sometimes disintegrate your craft.

[Rubenio]

KSP 2: 0.1.1.x [patch 1]
Windows 11 Pro, 22H2 (22621.1413)
Intel i7-10700F CPU @ 2.90GHz  - 48Gb RAM
Nvidia GeForce RTX3060 

Expected Behavior: Transfer ship maintains same orbit at ~140km waiting for the lander to reunite.

Observed Behavior:   I've seen the same issue with ships randomly losing changing orbits as described above.  Quite frustrating to send a lander to Mun and when taking back off to rendezvous with the transfer ship  (with Jeb & Bob waiting...) we see that their ship is on a nose dive to the Mun.  I've tried this several times and I either get this error or the following below

Second Observed Behavior:  When lander is taking back off Mun, orbit of 50km established, there are NO orbit trajectory lines.  It's as if the ship where just a point in space according to map view.  On the Nav Ball, it's a clear Orbit with AP & PE both in the ~75k range. NOTE the blue ship is the lander with NO orbit lines

[Flush Foot]

On 3/22/2023 at 12:57 PM, FirroSeranel said:

Same. Also same with disabling SAS altogether.

Yet in Munar orbit, the oribits do degrade faster if I rotate the craft in various directions with reaction wheels. My guess is something isn't respecting the craft's overall center of mass and there's a phantom force, with the phantom force responsible for the always-on orbit wandering, and the CoM issue (similar to the old wobble kraken drive in KSP1) responsible for the faster decay while rotating.

It also seems to be related to the bug in which coming out of rails-warp gives a big impulse kick to the ship. The faster the orbital wander is happening, the harder the kick. Around Kerbin, coming out of time warp maybe nudges the nose aside a bit and SAS brings it right back, barely even a wobble. Around the Mun at low altitude (10-15km orbit), coming out of warp kicks the Pe to -10,000m or more, and sets the ship spinning at ~100 RPM.

Then again, if it were a phantom force, why would it go away above a certain altitude? And why would it _not_ be happening near the surface? When I'm going in to land, it feels perfectly normal, no 20 m/s/s accelerations from nowhere. Very odd...

My gut feeling is that it's to do with whatever implementation they did of the floating reference point system that lets KSP have both fine spatial resolution for crafts, but also billions of kilometers of "open world" game area... which is also the source of classic KSP tropes like orbit line jitter, and many of the "Kraken" bugs.

And maybe near the surface, it engages a higher-precision mode? And at high-orbit a semi-rails mode?

For now, I guess I'll use higher parking orbits, and call it an unintentional feature instead of a bug!

Like... an accidental implementation of relativistic frame dragging, amplified in gleeful mischievousness by the Kraken.

Friday night I was doing my Ike-ascent as part of this week's #weeklychallenge [Jeb-level, not Watney], which is the only real reason I had my GPU recording my session (I'm not really a content creator, don't have that gift / experience to make doing that easy, so 'why make more work for myself?')... In any case, I noticed something weird with the phantom force in the earlier part of the ascent (immune to warp 'locking it down' until I cleared about 10 km ASL), then also found / believe I was able to direct the phantom force via SAS prograde/radial-out/etc. markers to give me a 'free' encounter with my interplanetary SWERV-pusher.

Unlisted video, details are/should be there, but it is still a bit long...

[Still]

KSP 2: 0.1.1.x [patch 1]
Windows 10 Home, 22H2 (19045.2728)
AMD Ryzen 9 5900X
ASUS TUF 3080 10G