Draining docked fueltanks?

[Tokay Gris]

So, my new "Heavy Freighter" has made its maiden flight.

Quite huge... yes. (Don't ask about lag)

In this model it has a TWR of 0.27, weights in fully fueled and equipped 632 tons and has 592 parts.

BUT! the delta-V without any modifications is an abysmal (measured by MechJeb) 1700 m/s.

Yes, I know that this is not exactly true. I can of course transfer fuel from the "payload" to the tanks feeding the engines.

But I can't do that while burning without offsetting the balance. With that many SAS modules it might still work, but it is inelegant.

(Hm.... it IS possible, the balance should not be offset, since I would be transferring fuel down a "line" of tanks, so it should not change the center of mass... still. Inelegant.)

(EDIT: It is not possible, I think. Because while I am filling one tank, the other will run dry... I would have to go around in circles. Doable. But not elegant.)

I would very much prefer that the docked tanks of the three payload sections were to be drained from the start. I was kind of hoping it would do that, because if I build a vessel in the VAB with a docking port in between instead of decoupler, fuel is drained from top to bottom.

But not so when the top tanks are added later by docking.

Am I overlooking some feature? Can I tell the fuel which way it should flow through the docking ports?

EDIT: Oh, and there some other things that need to be changed in this design. Mainly the small docking ports on the structural cones. Not stable enough. They don't break, but kind of wobble in line. But that is just a problem of design, not concept.

And still another EDIT: Yes, those docking ports are all docked.

Edited December 12, 2013 by Tokay Gris

[Tokay Gris]

Hm.... some testing showed that it DOES work. The problem is, it doesn't work for all the tanks.

Possibly just the "first" one to dock. I can't really check that, but that would be an explanation.

I tried disabling and reenabling cross feed, but that doesn't seem to tell the ports to let the stuff through.

It is possible and doable to transfer fuel while burning without changing the center of mass too much and without letting tanks run dry. I just have to go 'round an 'round refueling the "feeding" tanks from top to bottom.

"easiest" would be to use fuel lines... but of course, those can't be added after launch. And I would like to not use Mods except MechJeb. But is there one for that purpose? KAS is not it... it doesn't transfer fuel without actually click-ALT-click.... or does it? Haven't used it much lately.

[DMagic]

Are you docking each segment together individually, or are you connecting multiple docking ports at the same time? Only the "primary" connection will transfer fuel in these multiple-port type of connections.

The only realistic solution is to use TAC fuel balancer. It is perfect for this kind of complicated situation where the only alternative is to constantly balance the fuel by hand.

[Kasuha]

Unfortunately, the fact that your docking ports are all docked doesn't really help because there are two ways how a docking port may dock, if docked one way it lets fuel flow, if docked the other way it doesn't.

The first way (which lets the fuel flow) is the standard way - if you dock a ship using just one docking port, all is ok.

The second way is when you dock two ships using two or more docking ports at once, making cycles of parts in your ship. Out of all those docked ports only one is docked the "first way" and lets fuel flow, the whole rest is docked the second way. These ports are not technically attached, the game just puts an invisible strut between them so they appear docked. Fuel cannot flow there.

So if you make sure your freighter is docked using only one docking port to the payload, you can draw fuel from the payload with no problem. With using more docking ports, it will cause chaos.

[Tokay Gris]

Are you docking each segment together individually, or are you connecting multiple docking ports at the same time? Only the "primary" connection will transfer fuel in these multiple-port type of connections.

At the same time. Well, you nailed it, it seems. Thanks for that. Will mark as "answered".

But that gives me another idea. I could make a fuel part that consists of six tanks with docking ports and nothing else. If I attached them all individually, they should then - according to you - feed the engines. And I could still use the upper modules to attach later. IF such a six-part module is stable enough. It could introduce the same wobbling I was trying to avoid. But will test that.

The only realistic solution is to use TAC fuel balancer. It is perfect for this kind of complicated situation where the only alternative is to constantly balance the fuel by hand.

I will try that one. Thanks!

[Tokay Gris]

So if you make sure your freighter is docked using only one docking port to the payload, you can draw fuel from the payload with no problem. With using more docking ports, it will cause chaos.

Not exactly chaos. I just tried and it does work to go 'round an 'round and refuel without offbalancing the vessel. It just is work and inelegant. But it works which is worth something.

But thanks. You also solved my problem. Namely: No solution without mods...

[Kasuha]

No solution without mods...

There's plenty of solutions which don't require mods.

[Tokay Gris]

There's plenty of solutions which don't require mods.

Not for that vessel. Well, not that I see how. I can't really dock each port separately. The payload modules all have six ports.

And the solution I currently use (refueling in circles) is not really elegant. It does work, though.

[Kasuha]

Well of course, you can't expect it to start behaving differently without design change.

1/ I don't understand why can't you use just one docking port.

2/ The payload seems to consist of a lot of equal modules. Maybe instead of connecting them all together radially, you could connect them as individual columns to the engine frame. There are even ways how to dock them all at once and then to dock to the frame by columns, it's all up to how you build the payload.

3/ If it needs to draw fuel from its payload, maybe you could just give it bigger base fuel reserve. Pulling all those batteries and SAS units along just to throw them away at your destination is a waste.

Edit: here's just one of possible options:

Edited December 12, 2013 by Kasuha

[Tokay Gris]

1/ I don't understand why can't you use just one docking port.

The idea was to reduce the wobble. Which works. With six docking ports (but I guess it would work just as well with three) there is no noticeable wobble in the tangential direction. Because of the design flaw with the normal ports instead of the senior ones there is longitudinal wobble. But that can be changed with different design.

2/ The payload seems to consist of a lot of equal modules. Maybe instead of connecting them all together radially, you could connect them as individual columns to the engine frame. There are even ways how to dock them all at once and then to dock to the frame by columns, it's all up to how you build the payload.

The first part is what I will try next.

The second part... how does that work? The whole payload docks as is with one port being the one steered from. How do you dock that by column afterwards?

3/ If it needs to draw fuel from its payload, maybe you could just give it bigger base fuel reserve. Pulling all those batteries and SAS units along just to throw them away at your destination is a waste.

That poses the problem of getting that monster into orbit. The engine module weights in at 220 tons. Slightly above the actual capacity of my heavy lifter and that only works because I used a docking port instead of a decoupler, so part of the fuel of the engine part goes to the big mainsails. Another version of this design idea dropped the lifting stages at apogee and then turned around to use the nuclear engines to circularize.

The SAS modules are actually needed (though not that many) because without them this beast is way to sluggish orienting itself in space. Probably not the batteries, though. Granted. But put on the outer columns I don't get the stability problems I get when I put it in the center column. Probably could use some smaller batteries and not the inline ones.

The SAS modules and batteries on the payload modules are mainly because it needs to dock. But that is an idea. Design a payload that can jettison those no longer needed parts once docked to the engine module. Will think about that and eventually try that out.

Edit: here's just one of possible options:

Hm... let me try to understand that. You mean docking the fuel tanks individually (not as a whole module) and then docking a module like mine and decouple the center column afterwards?

[Kasuha]

The second part... how does that work? The whole payload docks as is with one port being the one steered from. How do you dock that by column afterwards?

It's in that picture. If you have modules docked in a circle and you undock the "main" port, it will detach completely. But because the magnetism is off just for that one just undocked port, it will redock and make another "main" connection. If you build it the right way in the VAB, you will know which port to undock so it redocks the right way.

Edited December 12, 2013 by Kasuha

[Tokay Gris]

If you have modules docked in a circle and you undock the "main" port, it will detach completely. But because the magnetism is off just for that one just undocked port, it will redock and make another "main" connection.

Sorry. I don't get it.

If I decouple the "main" port, it will be disconnected, but the others are still connected. Right?

Or hm.... I must try this.

EDIT and later. I couldn't test it yet.

If I understand you correctly, the "main" port is undocked, magnetism is off, but because of the structure it redocks and changes the "main" port. That could actually be the solution. If it works that way. I was assuming that a once undocked port had a kind of "dormant" state until you get a certain distance to it.

But it could be that this "dormant" state is only the magnetism (not the docking itself) and that it actually redocks again.

If you build it the right way in the VAB, you will know which port to undock so it redocks the right way.

Oh, that is not a problem. I can find out which port is the "main" just fine. Just fire up the engines and see which tank is drained. If it is the one on the engine module, then it is not the "main". If it is the one in the fuel module, then it is the "main".

And - if this works the way you say - I can test if it actually does do what I how it will do.

Edited December 12, 2013 by Tokay Gris

[Tokay Gris]

It's in that picture. If you have modules docked in a circle and you undock the "main" port, it will detach completely.

Just tested this. The "main" port undocks. The rest of the module stays docked.

But because the magnetism is off just for that one just undocked port, it will redock and make another "main" connection.

Not in my test. The undocked part stays undocked and another connection is now "main", i.e. it drains fuel from the tank above it.

So - unless I understood you wrongly - I end up with five of the six ports still docked, a new "main" port that draws fuel from the tank above.

So, basically, same as before, except one port is no longer attached. In view of stability (ok, so with six ports not such a big issue) a negative result.

It would be an improvement if the other five columns were now draining the tank above, but still only one does this.

So, back to the drawing board.

Next try is attaching six single tanks seperately to the engine module and putting one combined payload module on top for stability.