Flow control doesn't preserve RCS fuel in capsule

[Zeiss Ikon]

A while back, I learned about setting flow control for resources that are shared over an entire vessel, such as RCS fuel.

Today, however, I came to wonder if I knew what I thought I knew.

My ship, Far Traveler, has a 2.5 m RCS tank in the transfer stage (250 units), four "roundified" external tanks on the lander ascent stage (60 units), and the usual 10 units stored in the command pod.  On its first real mission, I never had a problem with RCS until after decoupling the transfer stage to test the lander during aerobraking -- and then I found out both the lander's roundified tanks and the small capacity of the Mk. 1 command pod were depleted.

Now, either I ran everything dry just about the time I decoupled (unlikely, since I was turning off RCS between intentional uses to avoid having it cause orbit drift when used for stability, though I can't say I actually checked; last time I'd looked, after using RCS for a very, very small correction burn, there was about 40% left), or my flow priority (30 for the transfer stage tank, 10 for the roundified tanks, and -10 for the command pod's tank) worked backward from what I expect.

Suggestions?

[IncongruousGoat]

For an immediate solution, try turning off fuel flow from the tanks you don't want drained, then turning it back on when you need them. In the long run, though...

I ran some quick experiments to make sure RCS obeys fuel flow, and it would appear that it does, draining from higher-priority tanks first on my little mock-up ship. So it seems there's something strange going on with the fuel flow on your ship.

[Zeiss Ikon]

Hmm.  I used to use manual shutoff before I discovered flow priority -- problem is, I had several instances (in my sandbox game) where I "ran out of monopropellant" only to later discover that I had about 2/3 of my original supply remaining in tanks that had been shut off.

I suppose it's possible that I burned a lot more of the lander's monopropellant than I thought doing the turn-around and redock to the nose-to-nose configuration (during that operation, the transfer stage is unguided -- no probe core, no crew -- so it tumbles slowly).  Once redocked, the transfer stage RCS tank is available again, which prevents me from noticing the state of the lander's tanks.  I might have to reexamine my "no automation" policy for my science save -- or just put another command pod on the transfer stage so it can stabilize on a gyro heading and/or hold "point at target".  That would save time as well as RCS fuel.  Docking is fairly easy if both vessels can maintain target pointing; it's not terrible if one just holds orientation -- and it's a PITA if one is tumbling free (because every port bump that doesn't latch adds to the tumble).  That transfer stage isn't intended to be recovered; if I put a command pod on it, I'm committing to at least bringing the command pod back (or I have to redesign the lander for one or more additional seats, and then scale up the launcher for the extra mass).  OTOH, it is intended to bring the lander's ascent stage back to Kerbin from wherever the landing was done, so having a kerbal aboard wouldn't be completely off the wall.

Hmmm.  Back to the VAB, I guess...

[FullMetalMachinist]

6 hours ago, Zeiss Ikon said:

Once redocked, the transfer stage RCS tank is available again,

I'm not certain, but that might be your problem. When two vessels dock, their stages can (and usually do) get jumbled up in some way. And the default numbers for fuel flow priority are based on the staging. 

You might want to lunch your transfer/lander stage and test that out. Use the 'set orbit' feature to save time getting to orbit. 

[Zeiss Ikon]

I'll try to give this a test, but it'll have to be during an actual mission -- Val has another date with an asteroid tonight.

[Zeiss Ikon]

Following up on this, when I redocked my lander (Mk. 1 command pod with Clamp-o-Tron Jr.) to my transfer stage (Mk. 1-2 command pod with Clamp-o-Tron Jr., even though the nose has room for the 1.25 m port), I found the flow priority was no longer what I set in the VAB -- the transfer stage's 2.5 m monopropellant tank had the same flow priority after redocking as the lander's four Roundified tanks, which would reliably result in the Roundified tanks (4x60 units) running dry while there's lots of fuel in the 2.5m tank (750 units).

Oddly, at some time (possibly a game restart) after manually adjusting the flow priority, I found the transfer stage tank had regained its higher priority, plus the 10 priority units I'd added in my manual adjustment.

I also found that this particular lander does use a lot of RCS fuel when docking to the transfer stage, even when the transfer stage is stabilized; apparently, the RCS isn't well balanced or my movements are fighting with the "point to target" SAS setting.  Next time I fly this vessel I'll try to remember to set the transfer stage to "point to target" and then switch it back to "hold orientation" so I can dock to it without it turning to follow my every correction.  Maintaining pointing has worked well in the past, but I've never had vessels with the mass distribution (base-heavy) that these do.

[FullMetalMachinist]

2 hours ago, Zeiss Ikon said:

apparently, the RCS isn't well balanced

Something that can help with that is to use 'fine controls'. Press caps lock to toggle it on, you'll see the pitch yaw roll indicators in the bottom left go from orange to blue. It makes the RCS thrusters fire at less than full power, which is usually enough for the reaction wheels (if there are any) to counteract any imbalance with the placement of the RCS thrusters. 

[Zeiss Ikon]

8 hours ago, FullMetalMachinist said:

Something that can help with that is to use 'fine controls'. Press caps lock to toggle it on, you'll see the pitch yaw roll indicators in the bottom left go from orange to blue. It makes the RCS thrusters fire at less than full power, which is usually enough for the reaction wheels (if there are any) to counteract any imbalance with the placement of the RCS thrusters. 

I'm aware of the cap-lock setting (and don't use it much, preferring light taps on the keys in regular mode).  The problem I had in this case was that the combination of "point to target" on both transfer stage and lander led to the two circling each other and the rotation rate building up as I tried to translate the lander to get the stack-to-be straight enough for the docking clamps to latch.

I'll find a suggestion thread to repeat this, but it would be very, VERY helpful if there were some way to indicate pitch and yaw axes on the spacecraft in the external view.  The nav ball can say my vector is perfectly aligned, but if I'm approaching at an angle, I won't get a latch -- I'll get a bump that adds tumble to an unstabilized part, or a "tail chase" if both parts are on target point.  And when I try to straighten out, the nav ball is no help.  Yes, I know, there are mods for that, relative to docking -- but what I'm suggesting is that the ships have a clear indicator of pitch and yaw axis, which would be useful for any maneuver -- like getting into orbit the correct direction when launching from Mun, or launching to polar orbit from Kerbin.

[FullMetalMachinist]

6 hours ago, Zeiss Ikon said:

but it would be very, VERY helpful if there were some way to indicate pitch and yaw axes on the spacecraft in the external view. 

Press 'v' to cycle through the view modes until you get to 'locked'. Then rotate the camera so you're behind and slightly above your ship. Now pitch up is up for the ship, left is left, and so on. Same for translation directions. 

[Spricigo]

22 hours ago, Zeiss Ikon said:

I also found that this particular lander does use a lot of RCS fuel when docking to the transfer stage, even when the transfer stage is stabilized; apparently, the RCS isn't well balanced or my movements are fighting with the "point to target" SAS setting.

 

Disable yaw/pitch/roll control for RCS thruster, let reaction wheels handle the rotation. It makes a huge difference.