delta-v per stage information and rcs

[antipro]

hi, I'm trying to build a vessel to land on mun, I'm trying to build it the most light and efficient as possible, so I look at delta-v counter and I try to get the most high.
but I'm experiencing some strange behaviors in the delta-v per stage indicator:
adding or removing a little percentage of fuel results in a big change in the delta-v values, or for example,
adding or not rcs thrusters to the ship, results in the delta-v indicator go crazy and shows unrealistic dV in the stage 4 or remove it at all.

here screenshots:
https://www.dropbox.com/sh/mnc6ud6y13cr2ou/AACfx1l-mEo0AF-r-96VaCQha?dl=0

what I'm doing wrong?

Edited January 26, 2020 by antipro

[AHHans]

O.K. First the easy part: the changes in stages 2 and 3 are as expected for adding some mass to the lander. (I guess you knew that already.)

Another issue is that the stock dV calculation often has problems getting the calculation done for all stages at first: some stages first show a wrong (I only remember: none at all, or too low) dV value, but if you do a small change it will give you the value. This small change could be taking away a part and attaching it again (e.g. a nose-cone), or just left(!)-shift clicking on the vessel to move it around a bit. That usually "fixed" it for me.

I have the impression, that this happens when the program has trouble figuring out for how long each stage will burn. Usually because (as you have in your case) multiple engines run at the same time, but some burn out at some time and are staged away. It is even worse if you use fuel-cross feed to feed a stage with tanks from another. In this case it works better to use the external fuel pipe instead of the crossfeed in the decouplers and (manual) fuel feed priorities on the tanks. (But "better" does not mean "always correct".)

To check if KSP got it right I usually check if dV, start-mass, end-mass, and burn time all make sense and are consistent. But wiggling the vessel a bit with a left-shift click usually works for me.

[Superfluous J]

It looks like you are draining the side tanks first via crossfeed-enabled decouplers. The problem here is YOU know you want to decouple those as soon as the side tanks are empty, but the game assumes you will not decouple them until their entire available fuel supply is gone. Because they can still run on the fuel from the center tank, the engines will still run until that entire booster stage is out of fuel.

To get around this, I've used fuel ducts in the past. Turn off crossfeed, put fuel ducts on the side tanks going into the inner tank, and then your dV values will be correct. Once you've finished your ship, delete the fuel ducts and turn on crossfeed again on the decouplers. The values will go back to being "wrong" but you'll know the ship will get the job done because you saw the numbers that matched your staging plan.

Too bad there's not a way in the game to tell the staging system when you'll stage something in these situations, but I for one can't think of a good one.

Edited January 27, 2020 by 5thHorseman

[antipro]

I did some other tests and as you said, the problem is in the hydraulic detachment manifold fuel crossfeed and the calculation of the end mass/burn time.
if I disabled the crossfeed, the 4th stage end mass is well calculated, 16 tons.
when you see "end mass 0.1" in the following screenshots, so 0.1 is the mass of the RCS thrusters only, that's why without RCS the mass go "0.0 tons" and the calculation fails, the dV go 0.0 too and disappear.
using external fuel pipes doesn't fix it, I see same result of using the enable crossfeed option.
 

https://www.dropbox.com/sh/qtv6me90ghhcevg/AAAS1-FfJwWMrskSUEH500bCa?dl=0

edit: I've installed KER so I can use its own dV calculator that is seems accurate.

Edited January 27, 2020 by antipro