I've suddenly gone all wobbly

[Clear Air Turbulence]

 

So here is my first Duna mission, all tooled up and ready to go: A miner-lander and a refinery , connected with an ordinary docking port.  I flew the lander up from the Space Centre and docked off minmus, thus constituting a midsize ship that hopefully can fly to Duna and refuel there.

 

Spoiler

In high Kerbin orbit, it handled well.  Responded when I want to orient the pointy end, and while there was some pogo type vibrations under acceleration, no major stability issues.

200 days or so later, in midst of the hohmann transfer to Duna, I want to do some midcourse corrections.  And I can't turn the spacecraft at all without major wobbles and oscillations.

I can see that the docking ports could be a cause of wobbliness, but why now?

Why was the craft initially maneuvreable, but not now? 

Why did did I lose my rigidity?

And what can a man do to get his stiffness back?

 

 

 

 

Edited December 19, 2015 by Clear Air Turbulence

[Plusck]

Maybe you changed which bit you're controlling from? I have also had the impression that an over-exerted docking port can get all floppy, though I do wonder whether that is all in my head.

When I have to turn large ships with heavy bits at the ends and docking ports in the middle, I turn off SAS, turn, then warp as soon as it is pointing the right way. It isn't a cure for rigidity, but at least you don't have a hidden hand trying to twist your thing in different directions all at once.

[Tex_NL]

Sadly I'm having similar issues with a LKO station. Until a few days ago I never had any stability issues It was rock solid and did not wobble at all. Even with multiple reaction wheels running at the same time. Both the stock SAS and MechJeb were able to keep it stable and level.
Now, even with all reaction wheels turned off, MechJeb and stock SAS turned off after about a minute it starts to do the Krakendance, trying to shake itself to pieces. If it wasn't for KJR I would have lost that station multiple times already.

[Geschosskopf]

Are there, by chance, any clipped parts? This includes animated things that might hit other parts when extended.  For example, service bay doors that bump into other parts or pass through struts, or solar panels that, once extended, might rotate into other parts while tracking the sun?  I've had cases where such situations have caused Krakendances upon physics load when returning to the ship from doing other things.

[Snark]

Another possibility:

Any time you have a floppy stack (e.g. ships coupled by docking ports) where your thrust is on the back end:  the higher your acceleration, the worse the flopping gets.

There's a certain critical acceleration level, below which you're fine because oscillations tend to damp out, but above which they magnify themselves and it quickly goes all higgledy-piggledy.

It may simply be that your acceleration is higher now, because you've burned off more fuel and therefore your engine is shoving less mass.  Perhaps you just crossed that critical acceleration threshold.

If that's the problem, then it ought to get better if you reduce acceleration below the threshold again.  Does the problem persist if you throttle back a bit?

Also:  in a situation like that, one way to make things less severe is to make the two connected ships have as lopsided a mass distribution as possible (i.e. have one of them be as light as possible, preferably the one whose CoM is the farthest from the docking port junction).  If you have some fuel tanks with empty space in them, can you play games by pumping fuel around to redistribute the mass?  Might help.

[Clear Air Turbulence]

The flopping happens under zero acceleration.

In other words, I am floating in interplanetary space.  I can't change my orientation at all.  All I need to do is hit the WASD buttons and the wiggling starts. SAS makes it worse. 

Warp gets rid of it fora a while, then it comes back. 

It did cross my mind yesterday that the problem could be that I have a different centre of gravity now that the big tank is emptier.  But pumping the fuel around did not solve the problem.

[Snark]

1 hour ago, Clear Air Turbulence said:

The flopping happens under zero acceleration.

In other words, I am floating in interplanetary space.  I can't change my orientation at all.  All I need to do is hit the WASD buttons and the wiggling starts. SAS makes it worse. 

Warp gets rid of it fora a while, then it comes back. 

It did cross my mind yesterday that the problem could be that I have a different centre of gravity now that the big tank is emptier.  But pumping the fuel around did not solve the problem.

Ah.  That's a different problem.

SAS in KSP is a type of PID controller.  Without going into a bunch of math, what it boils down to is this:  It's trying to maintain your orientation by using a feedback loop between a position sensor (your control-from-here point, such as a command pod or probe core) and your reaction wheels.  The problem happens when there's floppiness between the control-from-here point and the reaction wheels.  Imagine what it would be like trying to walk, if there were a quarter-second transmission delay between your brain and your legs-- you'd trip all over yourself.

To fix this problem, you want your probe core (or command pod) to be as close as possible to the main sources of torque in your combined ship.

Even then, it's still possible to have a runaway "resonance" effect where things blow up because the natural oscillation frequency of your bendy ship is pretty close to the control frequency of SAS.

So, what do you do?

Well, to address the first problem (floppiness in between the control point and the reaction wheels):  You've got two connected ships.  One of them is the control-from-here point.  Try disabling all reaction wheels on the other ship.

If you've got a resonance problem, the way to fix that is to alter the natural wobble frequency of your ship.  The easiest way to do that, if you can, is to redistribute the mass (i.e. pump fuel from one ship to the other).

[Clear Air Turbulence]

Aha. That sounds like a possibility. I think I have a reaction wheel on my refinery. I will try disabling it. Tho I still don't why the problem cropped up in mid flight only.

[Snark]

3 minutes ago, Clear Air Turbulence said:

Aha. That sounds like a possibility. I think I have a reaction wheel on my refinery. I will try disabling it. Tho I still don't why the problem cropped up in mid flight only.

Could have something to do with your mass distribution, i.e. burning off fuel changed the dynamics of your ship so that you've stumbled into a resonance situation.

Also, if you have a ship design that involves part clipping, I've heard that that can sometimes cause some phantom forces, but I don't know much about that-- I never use part clipping myself, it feels too much like an exploit.

[Tex_NL]

My station is stable again.
Pretty sure it must indeed have been a clipping issue. Only thing I changed was opening a HG-55 High Gain Antenna that was located close next to an RCS tug. Now that I moved the tug to a different docking port the station stays stable even with the antenna closed again. Strange, but if that's what it takes I'll take it.