Odd problem in refueller design

[Clear Air Turbulence]

Me again. Inexperienced, playing sandbox. I posted this question on an earlier thread but no one replied so I am trying again.

I am having an odd experience with my refueller.

My design:

(1) Launch craft: Solid state boosters and kickbacks, nothing too complicated.

(2) Rendezvous unit (just a small tank, poodle, inline stabliser, solar power and remote guidance unit - just enough to be able to fly around in LKO and get to space station)

(3 ) Payload

(- stack decoupler

- beeg orange tank full of fuel

- docking port

- fairing)

The idea is that I get into orbit, switch to stage 2, rendezvous, dock and then decouple and de-orbit / recover, leaving fuel payload in place.

BUT

I noticed one of two things happening in the VAB

Either the Delta-V of the whole rocket goes up insanely when I add the orange tank payload - despite the fact that the decoupler says no crossfeed. So it considers the two stages as one stage, in spite of the decoupler. So this means that (1) I am using fuel from the orange tank (2) I have no realistic sense of the d-v on the rdzv unit

OR when I insert the stack decoupler between the 2, KER tells me I have no d-v; in other words, it does not see first two stages at all, even though they are there, and part of the rocket.

Edited October 6, 2015 by Clear Air Turbulence

[Kobymaru]

KER's fuel draw mechanics seem kind of borked at the moment ...

There's a workaround, however:

Right-Click on the Orange tank, and disable fuel draw from them (the small green arrow) - this will take the tanks mass into consideration, but not as usable fuel. Plus, it prevents that your launcher draws fuel from the payload.

[Van Disaster]

Does locking the tank also work? that generally makes it look empty.

[mattssheep4]

Have you checked your staging sequence? That's what usually gets me.

[Clear Air Turbulence]

useful tips, thanks

I just put a bigger-ass stack separator in between, and that seemed to do the trick.

[Padishar]

Problems like this are usually caused by having the root part of the vessel in one of the early stages. The simulation in KER currently only simulates the part of the vessel that contains the original root part when a decoupler splits the vessel. If you change the root part of your vessel to be a part of the "payload" then it should calculate correctly. In this case it sounds like you should set the tank you are delivering as the root part and make sure that it is separated from the rdzv unit by a stack decoupler.

One other thing is that setting a decoupler as the root part can also confuse the simulation code so avoid that unless you have no choice.