Why is one of my fuel tanks exploding when I detach it?

[Noname117]

EDIT: The original problem of fuel flow is fixed, now I have a new problem where the launch stage is exploding when I decouple 2 fuel tanks (in a stack). Go to the second page for details

So, I've been working on a semi-NASA styled orbiter for the past few hours, and I feel it has a lot of potential. Except the fuel is not draining properly. It should be draining from a specific tank and that specific tank only, but is instead draining from two more tanks. I have repeatedly modified the fuel lines to no avail.

What I know and what should be happening:

The shuttle is made of 5 stages:

The Orbiter: Has 6 engines with 4 fuel tanks, 2 of them centered and the other two attached to the side. 4 of the engines are attached to the center tanks, the other 2 (which are angled) are attached to the side tanks. Designed to be the final stage.

An external fuel tank made of 5 fuel tanks, designed to drain from the outer 2 tanks first, then the next 2, then the middle one. Designed to be the second to last stage.

The twin liquid fueled rocket boosters, made up of 3 fuel tanks each. Designed to be the stage decoupled before the external fuel tank stage.

The twin solid rocket booster stage, made up of 6 SRBs each with a small fuel tank on top. Detached before the liquid fuel boosters detach

An extra fuel tank beneath the 5 tanks of the external fuel tank, detaches from the rocket before any of the other stages

Fuel is supposed to drain (as you can see in the imgur album) from the extra fuel tank and the fuel tank on the SRB stages into the top of the liquid fuel booster stage. Fuel lines then connect the lower tank of the liquid fueled rocket boosters into both the upper and lower tanks of the external fuel tank. Lines then attach the middle of the external fuel tank to wing panels of the orbiter, leading into all 3 sets of fuel tanks. The route is set up so that the ends are, no matter what, the lower fuel tank designed to detach and the fuel tanks on the SRB stage.

When testing, fuel begins draining from one, not both, of the upper tanks of the liquid rocket booster stage immediately following the depletion of fuel from the fuel tanks on the SRB stage (That stage has not yet detached because the SRBs are still going). It also begins draining from the lower tank of the external fuel tank stage. The detachable lower fuel tank has not been detached at this point, and fuel is still draining from it. However, fuel should be draining from only it at this time.

Once the lower tank is detached, fuel seems to flow from the liquid booster stage which has already begun to drain faster than it does from the other one, causing the shuttle to spiral out of control until detached.

All fuel lines are bilaterally symetrical. This has been checked

I do have one plugin installed: Kerbpaint. However, I highly doubt that a plugin modifying textures is causing the fuel lines to act weirdly. Otherwise, I am mod-free.

Edits from later tests:

Adding I-beams between the boosters and decouplers made no difference. Fuel crossfeed between the two is apparently not the problem

Imgur album showing the fuel line connections:

(The in-flight screenshots are in order from the earliest taken to the latest taken. Proof is in the mission time counter)

Javascript is disabled. View full album

So, I am asking for help on both recognizing and solving the problem. I do not know if this is some weird mechanic with how the fuel lines drain or if this is a glitch. I just want to solve this problem without having to reconstruct the entire booster stages of my spacecraft. So if there are any tips on how I should re-arrange my fuel lines so everything drains properly, please tell.

Thanks!

Edited May 4, 2014 by Noname117

[Claw]

Hmm, I tried to follow your thought process but was getting a bit confused. It seems you have some crisscrossing fuel lines between your main tank and the liquid boosters, so I'm not surprised you're having some fuel flow misfeed.

It might be easier if I point you toward this thread by Kasuha. It gives the fuel flow rules in the most basic setups. Hopefully you can get an idea of the rules and how they apply to the complex setup you're aiming for.

[EdFred]

Looks like you have some sort of recursive fuel lines happening. Why do you have two fuel lines running from each bottom outside tanks to the center tank? And why are you drawing fuel away from the bottom tanks?

[Noname117]

Hmm, I tried to follow your thought process but was getting a bit confused. It seems you have some crisscrossing fuel lines between your main tank and the liquid boosters, so I'm not surprised you're having some fuel flow misfeed.

It might be easier if I point you toward this thread by Kasuha. It gives the fuel flow rules in the most basic setups. Hopefully you can get an idea of the rules and how they apply to the complex setup you're aiming for.

Looked at the thread, and it appears the radial decouplers are causing the unwanted fuel exchange. That's good because it simply means I have to add a piece without fuel crossfeed to the decoupler and attach the rocket to that, preventing a whole re-design.

At least, I hope... Don't consider this solved yet, let me test first

Looks like you have some sort of recursive fuel lines happening. Why do you have two fuel lines running from each bottom outside tanks to the center tank? And why are you drawing fuel away from the bottom tanks?

The multiple lines running from the bottom tanks to the center tank are connected to both the top and bottom of the center tank (which drains from the middle), so that the farthest back fuel tanks are the ones on the boosters (and the jettison-able fuel tank below the main, external fuel tank). Also, which bottom tanks are you talking about, the jettison-able one, or the booster ones?

EDIT: Nope, fuel still drains from the tank I wish it not to drain from.

Edited May 4, 2014 by Noname117

[sdj64]

If you want to avoid re-designing the whole thing, try putting a structural panel 2x2 or 1x1 in between your stacked tanks. They have no fuel crossfeed so they stop fuel from draining up or down in the boosters. Though, KSP really hates loops in fuel lines, so if possible try to make them go one way only.

[Noname117]

At the moment I only have 2 loops in my fuel lines, both located on the orbiter, and thus not likely causing the problem (especially since I added them after taking the screenshots)

Please, can someone get Kasuha in here to tell me what I am doing wrong? I barely understood anything from his thread

Edited May 4, 2014 by Noname117
My own ignorance

[EdFred]

I'm going to reference your tanks as follows L for left, R for right.

LB1/RB1 - the small fuel tank on top of your SRB packs

L1/R1 - Bottom most tank connected to the 4-pack rocket engine

L2/R2 - Tank above L1/R1

L3/R3 - Tank above L2/R2

C1 - bottom most center tank and as we go up the center

C2 C3 C4 C5 C6

L1 has two fuel lines going to C2, R2 has two fuel lines going to C2.

If you removed all your fuel lines, the fuel would drain from R3 --> R2--> R1 --> R-Engine and L3 --> L2 --> L1 --> L-Engine

it appears you are trying to empty R1 and L1 into C2 which feeds C4 which feeds C6 which presumably dumps into C5 which dumps into C4 which feeds C6 which dumps into C5 which dumps into C4 which feed C6, which...(loop)

In addition to that C3 is apparently dumping into C2, which goes to C4 which goes to C6....(loop)

Meanwhile, you are feeding R3 and L3 from C1. R3 and L3 dump into R2 and L2 and then R1 and L1 respectively. And of course L1 and R1 are feeding C2 which feeds into C4 which feeds to C6 (loop)

I don't know how RB1 and LB1 fit into the mix.

How do you want your tanks to empty using the naming convention I'm using?

Edited May 4, 2014 by EdFred

[Noname117]

I'm going to reference your tanks as follows L for left, R for right.

LB1/RB1 - the small fuel tank on top of your SRB packs

L1/R1 - Bottom most tank connected to the 4-pack rocket engine

L2/R2 - Tank above L1/R1

L3/R3 - Tank above L2/R2

C1 - bottom most center tank and as we go up the center

C2 C3 C4 C5 C6

L1 has two fuel lines going to C2, R2 has two fuel lines going to C2.

If you removed all your fuel lines, the fuel would drain from R3 --> R2--> R1 --> R-Engine and L3 --> L2 --> L1 --> L-Engine

it appears you are trying to empty R1 and L1 into C2 which feeds C4 which feeds C6 which presumably dumps into C5 which dumps into C4 which feeds C6 which dumps into C5 which dumps into C4 which feed C6, which...(loop)

In addition to that C3 is apparently dumping into C2, which goes to C4 which goes to C6....(loop)

Meanwhile, you are feeding R3 and L3 from C1. R3 and L3 dump into R2 and L2 and then R1 and L1 respectively. And of course L1 and R1 are feeding C2 which feeds into C4 which feeds to C6 (loop)

I don't know how RB1 and LB1 fit into the mix.

How do you want your tanks to empty using the naming convention I'm using?

Firstly, just to help me out, I will give the engines the names RE1, LE1, and SE(1-6) (though I'll just say shuttle engines). For simplicity purposes, consider the SEs connected to C4 (Yes, I wish to blow them up, MUAHAHAHAHA!)

Secondly, LB1 and RB1 dump into L3 and R3 (just so you know where they go. Also, L1 and R1 are feeding into C6, the second set of fuel lines you see is diverted upwards

Thirdly, I would like the fuel to empty from C1 and RB1/LB1 first, followed by L/R3, L/R2, and L/R1. Then empty C2 and C6, then C3 and C5, then C4

Edited May 4, 2014 by Noname117

[EdFred]

That just seems like a hot mess. Any reason why you want 2 and 6 emptying before 3 4 and 5? There's no advantage to doing so since they aren't being jettisoned.

Were it me, I would only run lines from C1 to L3/R3 and RB1/LB1 to L3/R3 and see if R/L1,2,3 empty before the center tanks. If not run a line from R1 to C6 and L1 to C2 - and that's it for fuel lines.

it *should* then empty as you want....should.

Edited May 4, 2014 by EdFred

[Noname117]

Well, they are being jettisoned later in the flight. I need them to drain like that so the shuttle keeps its balance. I just said treat the SEs like they are connected because the problem occurs before the liquid rocket boosters are jettisoned, and thus the orbiter's fuel tanks would not yet be in play.

I will try to modify the fuel lines so R1 goes to C6 and L1 goes to C2.

EDIT: The fuel started draining from R3 with that setup (Instead of L3). On the other hand, C2 is no longer draining, so that is at least fixed

Edited May 4, 2014 by Noname117

[EdFred]

Well, they are being jettisoned later in the flight. I need them to drain like that so the shuttle keeps its balance. I just said treat the SEs like they are connected because the problem occurs before the liquid rocket boosters are jettisoned, and thus the orbiter's fuel tanks would not yet be in play.

I will try to modify the fuel lines so R1 goes to C6 and L1 goes to C2

Upon further review, just send R1 to C2 as well. They are both further away from C4 than C6, so they should empty before fuel gets pulled from C6.

With no fuel lines your center tanks (oustide of C1) should empty like this C6 --> C5 --> C4 <-- C3 <-- C2 which seems to be what you want. By adding the fuel lines as I described your system should empty like this:

C6 --> C5 --> C4 <-- C3 <--C2 <--L/R1 <--L/R2 <-- L/R3 <-- C1+L/RB1

RE and LE will also pull from L1/R1 but not from C2 through 6.

C4 then feeds your shuttle engines, and it always pulls from the furthest tanks away, so that should solve the problem.

Edited May 4, 2014 by EdFred

[Noname117]

With that configuration, C6 does actually start draining at the beginning of launch.

So far, C2's drainage has been fixed, but not the drainage of one of the fuel tanks at the top of the LRB stage.

Also, just to make your head hurt more, I guess I should add in the orbiters fuel tanks too:

O1, O2, RO1, and LO1.

The ORBITER Engines will now be OE(1-4), ROE and LOE.

O1 and O2 are stacked, RO1 and LO1 radially attached (but with a buffer preventing fuel exchange) the engines are respectively attached to them.

C4 connects twice to O2 and once to each RO1 and LO1.

I am starting to think that the OEs are pulling from C6 or C2 (depending on the configuration), and somehow either L3 or R3

Maybe the problem lies in here?

EDIT: Just hooked up L1 to C6 again to fix its drainage issue

Edited May 4, 2014 by Noname117
Forgot a single word

[EdFred]

Yes, the problem is C4 also feeding to other tanks, completely changing the fuel tree.

Remove the fuel lines LEAVING C4 to any other fuel tanks to see if everything but the O tanks drain correctly. I thought C4 only fed the shuttle engines.

Edited May 4, 2014 by EdFred

[Noname117]

Kerbpaint is a plugin, and is currently not being used with the craft, so you should be fine. Besides, I believe one with stock ships can still download ships with kerbpaint, they just wouldn't be painted. Let me get working on the .craft

https://www.dropbox.com/s/wqphq0q2wsb04hk/WIP%20space%20shuttle.craft

C4 is not feeding other tanks, only the shuttle engines. Due to my own stupidity, I forgot to add "orbiter" before "engines."

The .craft I sent you has R1 connected to C2 and L1 connected to C6, so far the most successful configuration (as in it doesn't drain from the central fuel tank)

Judging by the rate R3 drains, I would assume it may just be one engine.

Edited May 4, 2014 by Noname117

[Noname117]

Further tests have shown that R3's fuel is draining into OE(1-6), not LE1 or RE1

[EdFred]

I got the tanks to drain correctly...sort of...

You shuttle engines empty out the tanks on the other side first though...and if you dont run the shuttle engines, it wont fly straight.

I think you need to start from scratch....even getting all tanks to drain correctly, and pulling equally from 2 and 6 after jettisoning LE and RE...major major imbalance. Unless you like doing loops.

Edited May 4, 2014 by EdFred

[O-Doc]

I have some craft with truly monstrous fuel line arrangements so, I'll share this tip. Keep the craft fastened to the ground and study the burn on each engine all the way through till it can't access anymore fuel. You will then know exactly what tanks each engine can access and the precise order in which do this from full. This is particularly helpful with SSTO's and other craft that use action keys instead of staging.

I'll check out your craft. There's a possibility that your fuel lines are popping due to use of symmetry and the craft wobble on physics load. So again, test your craft using unbreakable joints and see if the fuel burns correctly using that debug method.

[O-Doc]

So, I've spent a bit of time looking at this craft. The thing that stands out is the fuel lines are not connected properly to cubic struts. A couple of fuel lines even seem to be connected to the end of other fuel lines and not the cubic struts themselves. These are all asymmetrically connected so that there are more steps between each side of the craft lines. I cleaned it all up but, there's still some issues where the center shuttle rockets pull fuel out of the top left tank until the bottom one is drained. Then they start pulling out of the top left and top right tanks at the same time. It definitely looks like a fuel line symmetry split bug.

I'd say there's two solutions.

1. Redo all of the fuel lines in the VAB without symmetry

2. Remove that bottom tank as there is way too much fuel on board anyway

[Kasuha]

The problem here is identical to the one in this example. The central bottom tank fuel is split to the two side columns and then joined back together in the central upper tank. When central engines go looking for fuel, they scan one branch all the way to the central bottom tank, but the other branch is cut short one tank earlier because the central bottom tank has already been scanned.

There are several ways how to fix it. The simplest one is probably to put the two booster engines on a decoupler (to separate them from tanks above them), then draw a fuel line from the central tank just to these engines so they use fuel from it first. And then to draw another pipe from the tank above the engine to the decoupler, that will allow the engine to start drawing fuel from the column above it after it's done with the central tank. This means main engines will draw fuel from side tanks while there is still some in the central bottom tank, but they will at least do so symmetrically.

[Noname117]

Thanks for all of your help, will try a couple of ways to fix it (I think I can do it without draining from the LRB tanks until the lower center tank is used up).

[Noname117]

EDIT: By changing the one central droppable tank into 2, with a decoupler separating them, I have fixed the problem. BUT I will not be leaving this thread as answered yet... why?

Because I am changing the question:

Why does either the decoupler or fuel tank on the central tank jettisoned towards the beginning of launch explode when decoupled, sometimes taking the entire launch stage with it?

New .craft file for those who wish to test it:

https://www.dropbox.com/s/wqphq0q2wsb04hk/WIP%20space%20shuttle.craft

[Kasuha]

Why does either the decoupler or fuel tank on the central tank jettisoned towards the beginning of launch explode when decoupled, sometimes taking the entire launch stage with it?

Because there's a bug with the new decoupler. It is set up as "physicsless" and the game does not handle it well. Claw has assembled a nice list of bugs in the current release and fixes for them where fix was found. Definitely recommended read for everybody.

1) With a text editor (such as notepad) open up the part.cfg file for the TR-38-D. It should be located in KSP\GameData\NASAmission\Parts\Size3Decoupler\part.cfg

2) Find the line that says “PhysicsSignificance = 1†and change it to “PhysicsSignificance = 0â€Â

3) Save the file and restart KSP.

[Noname117]

Alright, swapping out the decoupler solved the problem. Thanks again

[Kasuha]

By the way, congratulations to the solution you found for the fuel feed. It was even simpler than what I was proposing.

[Noname117]

Now I just have to fix some balance issues, and the shuttle should work fine. Watch the spacecraft exchange for the next couple of hours

Edited May 4, 2014 by Noname117