Bug with liquid fuel tanks in service modules

[dumuzid]

Update: here is a more minimal way to reproduce the bug:
- start a new sandbox game, enter VAB
- place a Mk1-3 command pod
- attach an SM-25 service module to the bottom
- place an FL-T400 fuel tank inside the service module on a bottom node
- attach a Swivel to the bottom of the service module
- observe that you have 829m/s of delta V
- move to launch pad, observe that you have no delta-V and the engine will not fire.

Original post:
I was playing through the "Munar 1" mission in the "Making History" DLC, when I encountered a 100% reproducible bug where the editor claims that an engine can talk to a fuel tank, but when I actually get to the launch pad, it cannot. Here's how I did it:

- build the Munar 1 spacecraft as directed, with a Mk3 pod and the service module
- connect a cheetah engine to the bottom of the service module
- put at least one fuel tank in the service module (doesn't matter what type or where in the service module)
- observe that the editor shows that the spacecraft currently has some amount of delta-V
- connect a stack decoupler to the bottom of the Cheetah
connect a fairing to the stack decoupler, with the decoupler being at an interstage node
- observe that the editor still shows that the spacecraft has some delta-V
- enter the mission
- observe that the service module does not have any delta-V
- revert back to the editor and load the auto-saved rocket
- observe that the service module still does not have any delta-V, and continues not to have any delta-V until the fuel tanks and engine are detached and re-attached in the editor.

This reddit post (https://www.reddit.com/r/KerbalSpaceProgram/comments/c1ddoy/sooo_im_doing_the_munar_1_mission_right_now_i/) describes the symptoms I'm seeing (with a screenshot), but someone there claimed to have worked around this by attaching the fuel tanks to the base of the service module, which did NOT appear to fix the problem for me.

Anyone else encountering this? Known issue?

[Wobbly Av8r]

Playing around with a Saturn V / Apollo reproduction I noticed the same thing with SM-25 Service Module - fuel tanks, despite being added at the nodes, do not feed the engine connected below.

[dumuzid]

That matches what I'm seeing. The monoprop tank functions as expected, the liquid fuel/oxidizer tanks do not. Maybe that's the actual bug and the interstage nodes thing is a red herring. I'll see if I can reproduce that way.

[GunsForBucks]

Monoprop isn't really a good indication since it doesn't get effected by crossfeed anyway.

Had some odd issues with the structural tubes as well with crossfeed of fuel and it messed up both the combo tanks and just liquid fuel ones. They don't have any fuel options and are set as allowing fuel flow. One time it worked, one time it didn't... I did have a probe coupled inside the tube and set its coupling on disallow fuel feed but that shouldn't have effected the outer structural tube it was inside of. The docking part was even on robotic hinges off the tank in front of the tube.

[dumuzid]

Okay, more targeted repro steps:
- start a new sandbox game, enter VAB
- place a Mk1-3 command pod
- attach an SM-25 service module to the bottom
- place an FL-T400 fuel tank inside the service module on a bottom node
- attach a Swivel to the bottom of the service module
- observe that you have 829m/s of delta V
- move to launch pad, observe that you have no delta-V and the engine will not fire.

I'll update the OP.

[Wobbly Av8r]

FWIW, the interim fix that works is an FTX-2 External Fuel Duct attached from the tank to the floor of the SM-25...

[AoD_lexandro]

Had similar issues myself with engines getting no fuel even though attached correctly. From my observations it seems to be an issue when the service module actually gets launched. As if the internal fuel line point is not attached when constructed on the pad.

As said above the work around is to place a fuel line to counteract it. Extremely frustrating when you have a design based on the service module and it refuses to work.