problems after decoupling a lander

[king of nowhere]

As part of my plant to make a Mun base, I started building a "sky crane": it is a lander made of 4 rockets on the sides at a few meters from each other, and in the center there is the space to attach whatever part i want to land on Mun. I attach the sky crane and the lander in orbit, then I land the sky crane, I lower the module on the ground, then the sky crane can go back to orbit and bring down more pieces. The whole thing is authomatically controlled by a probodobodyne. On top of it there is a detacheable manned module with a bit of fuel to bring back astronauts to kerbin.

I brought down the first piece, a rover, and the whole thing worked like a dream. Perfect landing with excellent control and stability, I landed on Mun without a glitch. I decoupled the rover, did a bit of science.

the problem was coming back. the sky crane capsized on launch almost immediately. It became unstable already at a speed of 10 m/s. And I can't figure out how.

It's not a stability problem, as the thing worked just fine on landing.

Furthermore, I see two more glitches that may or may not be related. First, my navsphere shows the ground. as if, when i start, the ship was pointed downwards. which is clearly ridiculous. and again, it didn't do that on landing. Thinking that perhaps i inverted the control module, I tried to put the atronaut inside the manned module, but again, no change. the navsphere is inverted over what is should be. Second, when I deattach the manned reentry module (with the astronut inside, of course), it won't start. I can control it, I can select it normally, but the engine won't work. I checked, fuel is ok.

I made a few experiments and found out that the only real problem was with the navsphere inversion. if i deactivate the assisted stability, I can take off. It's quite difficult to achieve orbit without assisted stability and with the commands inverted, but doable. in fact, the problem is likely that when the navsphere gets inverted, the authomatic stability sysem gets all commands inverted too, and all its attempts to compensate only make things wrong. I activated it a couple of times after launch, but i have to point it to retrograde to get it to prograde and viceversa.

furthermore, i was able to determine that this inversion happens when i detach the rover, and it happens because the rover is recognized as the command center of the vehicle. which makes sense, because i devised the sky crane as something completely automated, while I wanted a pilot in the rover. so, i had a pilot in the rover, when i uncoupled the rover the sky crane inverted its navsphere. And when I designed the ship I started from the rover, because I then spaced the rockets to ensure that the rover would fit between them. if I put my pilot inside the manned reentry module, instead, i can detach the rover and the configuration stays fine, and i can takeoff afterwards.

but just because i found out how to tackle the problem, it does not mean there is no problem. in fact, future flights may not have a manned reentry stage at all (i plan to deattach it and leave it in orbit if I don't need to bring back astronauts), so i will not be able to take this solution in the future. What should I do about this annoying inversion of configuration?

 

I also have a suspect for the second problem, reentry module engine not working: I have put a SC-9001 Science Jr. between the rocket and the fuel tank. I didn't think this was a problem because I know fuel can get transferred through other spaces, and at least I can transfer crew through it. but perhaps it's not the case here? perhaps the SC-9001 Science Jr. does not allow fuel transfer? if that's the case, I will need to reload the save before I launched the whole thing, to fix the reentry module.

 

Since I am making questions, I take the chance for another, related one. I plan the sky crane to go back and forth from surface to orbit carrying modules, but of course it will need refueling. I plan to send a resupply rocket to it in the future (in the far future I plan to produce fuel directly on Mun and resupply it on landing), but how does fuel transfer work? once I attach the two ships through a clamp-o-tron cna i just right-click somewhere to manually transfer fuel from one tank to the other? Or do I need some special parts included?

[Grogs]

Sounds like the probe core in the rover is remaining the point of control after reattaching. This makes sense since you say it was the root part in the VAB. Have you tried clicking on the probe core on the skycrane and selecting 'control from here'? For a more permanent solution, you can go into the VAB and assign each crew module or probe core a priority number, with lower being more important. When you dock two pieces together, the one with the lower priority becomes the new control point. I've had to do this with complicated designs where KSP doesn't pick the correct control point.

EDIT: It's control vehicle naming, which ranges from 1 to 20, with 20 being the highest. I'm not sure it actually controls which part is in control though, more just which one is shown in the map view when they're linked.

 

No idea why the engines won't fire assuming they were staged first. I use a Science Jr. module on many landers with balloon tanks and engines attached to the module separately. I've never had a problem with fuel flowing from tank to SC-9001 to engines.

Edited May 29, 2020 by Grogs

[bewing]

What Grogs said is correct. The Navball orientation is determined by the orientation of the probe core or docking port that is in control of the vessel. If you build your craft with an upside-down probe core, then your navball will point behind your craft. In this case, down toward the ground.

There are several things you can do to fix this in flight, but they all involve choosing a particular docking port or probe core, and then clicking the "Control from here" button on the menu.

 

[king of nowhere]

yes, "control from here" was the answer. the probe core was aligned right, but the clamp-o-tron that released the rover was oriented downward (not surprising, since the rover was attached underneat), and when i deattached the rover, control was transferred authomatically to the clamp-o-tron. setting control to the probe core fixed the issue.

thanks.

as for the reentry module, its issue got fixed when i selected "interact with vehicle: yes". funny thing is, even after i revert it back to no, it still keeps working. i 'm not sure what that option even means.