Problem with Duna-Lander

[Tokay Gris]

So: Here is my Duna-Vessel, complete and in the process of insertion burn:

Between the actual lander and the cruiser are two docking ports. A big one that connects the two vessels for the transfer to Duna and a small one to connect the capsule to the cruiser for the transfer back to Kerbin.

So: What happens is this: If I disconnect the big docking port (that was connected in Kerbin orbit in the first place) the lander starts to spin uncontrollably und disintegrates. Explodes, basically. Like it is in a seperate frame of reference after decoupling.

But if I disconnect the small one (that has struts stabilizing it and is basically just an adapter in between the two vessels), everything is fine. Normal decouple.

Here you see them after decoupling:

I don't understand what is happening here. If there is structural damage because of the long burn times, it would show up both with small and the large docking port. (Unlikely. TWR is 0.1)

Any ideas? Bug?

Edit: In the second picture you see the problem: If I decouple the big docking port, even AFTER I decoupled the small one, the same thing happens. Landing is still possible, I just have to leave that port as it is and take the adapter with me.

Edited June 5, 2013 by Tokay Gris

[rodion_herrera]

Question: Why are you using docking ports of differing sizes? Both sides must be the same size.

[Zelnogrard]

First thing that came to my mind was to try the same operation under slightly different conditions. Turn the SAS/ASAS on or off, fire up some of the lander engines, apply some spin to the entire ship before undocking (that's probably the silliest one). My first Duna-capable ship with a lander had a nasty habit of disintegrating randomly after switching to it via square brackets, but quicksaving before trying to switch saved me.

[Tokay Gris]

The big docking port is for stability. It docks to the landing part of the lander, which stays on the surface and acts as a kind of launch bay (power up engines, hit "Space", return vessel gets immediate power and seperator decouples).

But since the returning command pod only has a normal size port, it needs One of equal size on the cruiser.

So the cruiser needs both on the front so the idea was to decouple the big docking port to uncover the small one.

[Tokay Gris]

I tried all that. SAS, the works. Even hitting timewarp immediately after decoupling could not prevent it.

[Vanamonde]

Occasionally, SAS is applying force to the ship when you detach, and it's as if a whole bunch of torque gets transferred from the main ship to the detached part, causing violent spin. I know you said you tried turning SAS off, but try that again, and this time go into and out of warp for a few seconds. That should cancel all rotations and momentum and whatnot. Then, still with SAS left off, undock the ships. And if that doesn't work, then I got nothin'. Oh, except perhaps to undock and then instantly go into warp, stopping any rotations on the lander before they get out of hand.

[Tokay Gris]

Did that. All. Except changing orbit. It seems to be a glitch.

Once I undock, the lander turns upright with increasing speed (like incredible torque is applied, more than is even possible... the lander is usually rather sluggish, having only the command capsule as a "torqueable" component), one engine crashes into the cruiser. Ther IS no other part that can apply torque. An oversight that I would change now, but can't in orbit around Duna.

And even when avoiding that (with time warp), the part that is supposed to lift from Duna again is ripped apart from the landing stage and the six radial engines of the landing part get ripped from the structure.

I suppose it is a bug. The turning is so forceful, I can't imagine that it is meant that way.

[Tokay Gris]

Here are more pictures to show what the mission profile was:

Here we see Mission "Podkayne" during assembly in Kerbin orbit:

As you can see, the big docking rings are for the connection between the transfer cruiser and the landing stage.

Here is a picture from the Mun-test of this configuration:

The command capsule has seperated from the landing vessel and is going back to Mun orbit to dock again with the cruiser. The big docking port stays with the lander, the return vessel only has a small one.

(Here, during an aerobraking test in Kerbin orbit. Sorry, it was - again! - the night side)

It worked flawlessly on Mun. Only difference: For the Mun test, it was a "Pollux MK II", for Mission "Podkayne" it is a "Pollux MK V" with increased fuel capacity.

So what could be different? Except longer burn times, but with less TWR.

(Oh, and it is - as always it seems with my designs - way overpowered. I could use fuel from that liftoff stage to refuel the landing stage and get it intakt back to orbit to land in another place on the surface. Plenty of fuel in the Cruiser still left.)

Edited June 6, 2013 by Tokay Gris

[Zelnogrard]

... the lander turns upright with increasing speed ...

Seems like a bug to me. Perhaps, some parts of the lander clip with each other? Or the undocking itself causes the removal of some struts and (possibly, only guessing here) some collision-redetection subroutine to kick in, causing disaster.

Another thing to try: try undocking via the larger port, and after that undocking the secondary command pod. If it goes well, you could just reassemble the thing and continue from there.