Physics model problem: Why is my rocket moving??

[Loren Pechtel]

I just landed on Minmus.  I'm down, engines off, SAS off.  I've even done an EVA.

However, MechJeb says my vertical speed is about -10mm/s and my horizontal speed is around 20 mm/s.  Note that these numbers keep changing--the surface speed sometimes goes as high as 40 mm/s.  My landing site isn't flat, could that have anything to do with it?  (I have a 4-legged rocket.)  Pitch, heading and roll are also varying by .2 degrees but much more slowly.

MechJeb also thinks the suicide timer is 4.6s, not jittering like the speed.  I'm on the ground!  Is it perhaps counting from the location of the root part?

[bewing]

Are you using actual landing legs? This is the problem that everyone is complaining about with the landing legs. They vibrate against the ground, and make you slide around.

[Loren Pechtel]

11 minutes ago, bewing said:

Are you using actual landing legs? This is the problem that everyone is complaining about with the landing legs. They vibrate against the ground, and make you slide around.

I'm not visibly going anywhere.  It didn't sound like the landing leg bug.

What's the workaround?

[klesh]

Plant a flag and see if the ship moves away from it after enough time.  

I think the workaround at present is to not use landing legs.  

[Loren Pechtel]

2 hours ago, klesh said:

Plant a flag and see if the ship moves away from it after enough time.  

I think the workaround at present is to not use landing legs.  

What do you support your rocket with then?

[Ultimate Steve]

11 minutes ago, Loren Pechtel said:

What do you support your rocket with then?

If it's Minmus, and you're good at landing slowly, you don't need anything. Just land on the engine bell. If you tip, it's Minmus. You can probably right the ship with the pod's reaction wheels.

[bewing]

2 hours ago, Loren Pechtel said:

What do you support your rocket with then?

Wing parts (swept wings, wing component type D, etc.) work very well as substitutes for landing legs. Or retract some/all of your landing legs after you land. As soon as something else besides a landing leg touches the ground, most of your "sliding" stops. If possible, try retracting your uphill landing legs and see what happens.

[Snark]

Also... how sure are you that this is a KSP problem and not just a MechJeb problem?  Is your ship actually moving?

(Not a MechJeb user myself, so I have no idea about this myself-- just putting it out there as a possibility.)

[Loren Pechtel]

1 hour ago, Snark said:

Also... how sure are you that this is a KSP problem and not just a MechJeb problem?  Is your ship actually moving?

(Not a MechJeb user myself, so I have no idea about this myself-- just putting it out there as a possibility.)

Since the engines aren't on I don't see what MechJeb could be doing to cause it.  MechJeb works by issuing orders to your rocket.  I'm only aware of two cases it does things you can't do:  A MechJeb-controlled rocket maintains it's aim at the burn direction through warp and MechJeb's ascent module can turn your rocket more gradually than you can.  (And the bug that the initial ascent burn can be done with a rocket with piloting.  The circularization burn is done with a normal maneuver node and fails.)

[Snark]

1 hour ago, Loren Pechtel said:

Since the engines aren't on I don't see what MechJeb could be doing to cause it.  MechJeb works by issuing orders to your rocket.  I'm only aware of two cases it does things you can't do:  A MechJeb-controlled rocket maintains it's aim at the burn direction through warp and MechJeb's ascent module can turn your rocket more gradually than you can.  (And the bug that the initial ascent burn can be done with a rocket with piloting.  The circularization burn is done with a normal maneuver node and fails.)

Actually, that's not what I meant. My question came from your comment "MechJeb says..." i.e. is it possible that your ship is fine and not moving, but MechJeb is reporting bogus numbers?

Again, I don't know anything about how MechJeb's internals work, so I could be totally off base here. Just was curious whether it's a possibility.

[Jimbo Jangles]

I don't use MechJeb, but with similar readouts using KER I see the same type of jittery readings while landed and at rest. I'm of the opinion that it is related to the rotation of the planetary body.

 

I've also seen the pitch, heading, and roll rate jumping around while in orbit. Even when I've just come out of timewarp and my ship is perfectly still. 

 

I get the feeling that it's a general issue where the math KSP is doing behind the scenes isn't accurate enough when run through an analytical program like MechJeb or KER without introducing small errors. I've heard people on the forums say something to that effect before. 

 

Hope this helps! 

Jimbo

 

[Loren Pechtel]

10 hours ago, Jimbo Jangles said:

I don't use MechJeb, but with similar readouts using KER I see the same type of jittery readings while landed and at rest. I'm of the opinion that it is related to the rotation of the planetary body.

 

I've also seen the pitch, heading, and roll rate jumping around while in orbit. Even when I've just come out of timewarp and my ship is perfectly still. 

 

I get the feeling that it's a general issue where the math KSP is doing behind the scenes isn't accurate enough when run through an analytical program like MechJeb or KER without introducing small errors. I've heard people on the forums say something to that effect before. 

 

Hope this helps! 

Jimbo

 

I think you've got it.  We know the math isn't accurate enough in orbit as observed by the jittering of the markers with nearly circular orbits.  The planetary rotation makes good sense as to what's going on as the thing that was really baffling me about this is the lack of any source of movement.

[bewing]

Did you plant a flag, as suggested by klesh, to verify whether movement is or is not, in fact, occurring?

 

[Loren Pechtel]

2 hours ago, bewing said:

Did you plant a flag, as suggested by klesh, to verify whether movement is or is not, in fact, occurring?

 

Observations:

1)  I'm actually sliding downhill.

2)  Retracting my landing legs (I'm in Minmus, I just dropped onto the engine bells) cut the jitter by more than 90% but did not stop it.  Now the peaks are about 1 mm/s.  Pitch and Heading appear constant, Roll is slowly ticking up.

3)  Slight game bug:  I tried to place a flag too close to one of the landing legs.  That overstressed the leg and it exploded.

4)  I just noticed that I have a jitter in my current orbital speed.