Creating a "pool" of fuel?

[CrazyCrazySam]

Before I start, I have to say that I am completely aware of the fact that what I am doing is not the most efficient way of getting into space.

I'm fairly new to the game, and I seem to have misinterpreted the way fuel lines (FTX-2) work. I needed to create a lot of thrust, and needed a lot of fuel, for a traditionally-staged (non-asparagus) heavy lifter. So I packed a lot of large orange tanks into a shape I was happy with and towered them 3 high, giving some towers 2.5 tanks to fine tune the wet mass of the stage. Each tower terminates with a Mainsail with a sprinkling of other thrusters added in balanced formations to fine tune the thrust. The whole thing is balanced on the launchpad. However, each tower obviously drains at a slightly different rate, so it was in my interest to "pool" the fuel - to treat the many tanks as one big tank so that the whole thing burns evenly and the stage runs out of fuel predictably. What I tried to do was use fuel ducts heading in both directions between neighboring tanks, such that every tank was in some way connected to every other tank. As you probably know, having more experience, this failed and the thing shredded as certain areas drained faster. Is there an easy way to pool all this fuel? Am I doing it right but missing a connection somewhere? Or will engines always just try to pull from the farthest tank?

Edited January 1, 2014 by CrazyCrazySam
Marked as answered.

[Soda Popinski]

Pictures may help.

Do the additional rockets assisting the Mainsails have their own fuel tanks? Or did you drag fuel lines directly to them. Fuel moves in the direction of the drag / arrows on the side. So if the additional rockets have their own fuel tank, the Jumbo tank will drain first, and assuming that feeds the Mainsail, it'll stop, while the additional rockets continue until their smaller fuel tanks run out.

If you want all rockets in the stage to stop at the same time, try dragging fuel lines directly to the rocket engines, so they all pull from the Jumbo and whatever is connected to it.

You may not need Mainsails for those upper stages, as you'll probably be well into your gravity turn or doing your orbital insertion. May want to save weight with a smaller rocket engine.

Note, if you're building higher vertically stacked (realistic) rockets like that, I'd suggest using Ferram's Kerbal Joint Reinforcement Mod, to give "properly rigid rockets." Works well with Ferram Aerospace Research.

[Kasuha]

An important thing to note about fuel balancing is that fuel pipes do not pump any fuel. A fuel pipe from a full tank to an empty tank will not put any fuel to that empty tank. Fuel pipes are just clues for engines. Because engines are looking for fuel around the ship, select a set of tanks according to a set of rules (involving among others also fuel pipes) and draw fuel from these tanks. As soon as a tank becomes empty, engines re-evaluate their rules and select a new set of tanks to draw fuel from.

From that perspective, putting fuel lines in both directions between tanks makes little sense in most cases.

You have a set of engines and a set of fuel stacks. And if I understand it correctly, you want to achieve the state when all engines run as long as there is any fuel in at least one of these stacks. You probably also want these stacks to be drawn in a balanced way, i.e. to prevent one side of the ship to be heavier than the other. It does not mean they cannot run empty at different times, though.

The simplest way to achieve that is to run a fuel pipe from each stack to each engine. Yes, you can end the fuel pipe on an engine.

Second simplest way - assuming you have a central tank - is to put a fuel pipe from each stack to the center stack, then a fuel pipe from center stack to each engine.

There are more ways how to balance the fuel drawing but for more details I need to know more details about design of your ship.

Edited December 30, 2013 by Kasuha

[CrazyCrazySam]

The stage was built to perform similarly to the Saturn V first stage, and as you might expect has a *few* engines. With the current configuration, running fuel pipes in straight lines (isn't that the only way?) to each engine is going to be frustrating. The stage is shaped like this:

0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0 0 0

0 0 0

If you don't get it I can get more screenshots.

[michaelhester07]

The fuel lines let you decide which tank drains first. It ultimately is supposed to let you transfer fuel between containers which are attached with components that don't allow a fuel transfer (like the metal struts). The pipes will make fuel flow in the direction of the arrows on them. The fuel will flow out of the tank in the retrograde direction of the pipe first. That is if you have 2 tanks with a pipe between them they'll drain from the first to the second. The second one doesn't drain yet.

The point of asparagus staging (from what I can tell) is to maintain a flat TWR as the rocket launches. You'll drain 6 rockets with a TWR of 2 but by the time you drain the tanks on the first 2 you'll be up to TWR of 2.6 or so. You break those rockets off and continue on the 4, back down to a TWR of 2. This burns until you're down to one rocket with a TWR of 2 for the remaining weight of the rocket. You'll get the same acceleration with less fuel and hopefully cost than if you just stuck 7 tanks together each with a rocket and let em burn. That by the way is how I've been doing it and I'll describe it below. I personally don't do asparagus staging as it is hard for the automatic stack control to understand and I don't like taking 10 minutes to figure out the stack only to find out I have to add a part and it completely resets my stack after I do that.

My 1.5 rule.

If you consider the largest tank you have available that matches your rocket's engine size as 1, the second largest as .5 you end up with the 1.5 I refer to. Always put the bigger tank on the bottom to increase the rocket's stability. Struts may be used to increase the stack stability for long rockets.

Here we go:

LV-T30/45: FL 800 and FL 400. : Provides about 2500m/s deltav at about 2.1 TWR. The T30 has a higher TWR with a lower delta v. Once I have the 45s I don't use the 30s though.

LV 909: FL 400, FL 200. If I followed the rule. I personally hate making tall lander rockets and wouldn't use the 909 for anything else. 1 FL 400 gives 2200 delta V for 1 ton of materal on the craft, enough for a return trip after a Mun landing.

Rocomax poodle: The rocomax 1600 matches this like the FL 400 matches the 909. Same purpose, bigger rocket. Also doesn't follow the rule. If it did it would take the rocomax 800 size as well (that thin one with the same fuel content as the FL 800)

Rocomax skipper: Rocomax 3200, Rocomax 1600: I primarily use this as an orbit insertion rocket. 2400 delta v for a 15 ton rocket. TWR is fairly low though, only 1.8.

Rocomax Mainsail: Rocomax 6400 (orange tank), rocomax 3200: I call it the Flying cigarette. Used for bottom stacks in a pack of 5 (center with 4 surrounding). Struts must be used to attach the upper stages, otherwise the rocomax decoupler gets crushed under the thrust and the rocket falls apart. Launches up to 50 tons of craft above it (including the weight of the skipper stage). If you need to get more weight than this consider a castle shaped launch vehicle.

With what you're doing adding extra radial rockets watch your G force doesn't go too high and keep the large tanks to the bottom. If you put your orange tanks on top of the silver ones your rocket will be unstable. Make sure you have enough struts to keep it from flying apart. If something falls off, add struts to it.

[CrazyCrazySam]

I've mocked it up, hopefully this is helpful. Like I said, running fuel lines will be tricky if they have to connect further than neighbor-neighbor.

Is there an easy way to do this? I can give you more info about the situation if you need it.

I guess I'll need a mod to be able to have it stay together reasonably? I can't really add too many more struts before my computer self destructs it seems like, running 46 engines is hard enough.

[michaelhester07]

So asparagus staging is actually a decent way to add engines to a rocket and it did what I thought it would. I got 7k delta v off 5 LV 30s, 2 FL800s and 4 FL 400s. Orbit with 3k m/s to spare.

[Smidge204]

Oscar-B tanks mounted on cubic octagonal struts are a great way to manage odd configurations like this. They are very low mass but can be used as "nodes" to drain multiple tanks into before redistributing, or to make links in cases where a fuel line can't stretch long enough.

But draining everything into a central tank may not work as expected, since the engines will still drain from their own tanks unless something is feeding into it.

For example, if you have two stacks of tanks, each with an engine, and tank "A" has a fuel line leading to tank "B", then both engines will take fuel from tank "A" until it's empty. At that point, Engine "A" will burn out and Engine "B" will start to drain the still full tank "B." All else being equal, the smaller tanks will still drain first if they are all equally preferred by the fuel search algorithm.

=Smidge=

[Sauron]

There is a plugin (TAC Fuel Balancer) that will do this for you. Otherwise, Smidge's approach will (sometimes) solve the problem depending on how your rocket is set up.

[CrazyCrazySam]

Thanks for the help.

The Solutions I've Gathered:

-Kasuha and Smidge's direct-to-engine lines with Oscar-B tanks to help with routing and logistics. This could potentially work, however for the scale of the ship the construction is going to be time consuming and the final result is going to be sloppy. Thanks for clearing up fuel transfer however, as I've always wondered why sometimes well connected tanks just wouldn't work right.

-The plugin suggested by Sauron may solve the problem, I'll have to test that out.

-The plugin suggested by Soda Popinski actually allows the concept to even work. I've been trying to use the same formula as the Saturn V and it simply couldn't do things like grav turns without exploding.

I actually was able to 'solve' this today, although I'm not satisfied. What I did was instead of scaling 1:1 from the NASA Stage, I scaled 1:10. Considering the scale of the Kerbal universe, this is logical and still functional. With the help of Ferram's mod I created this 1:10 model and its currently sitting in orbit with quite a bit of delta-V left over. Hopefully this fuel mod will help the bigger model work, but it's hard to imagine what could need that much lifting power, considering I used just about a quarter of my Stage 2 delta-v to reach orbital velocity when I planned on ditching it just outside the atmosphere.

[pvtnum11]

I second Smidge204's fuel node thing. I don't have to do that very often, but it works well when I have to.

[Sirine]

Press Alt+Click on parts to copy them at VAB. Very useful at building something big.

[Kasuha]

Okay, I prepared the following example. Each fuel stack represents one type of column - orange tank represents the central "red circle" column, the grey tank represents the "black circle" column and the thin FL-T800 represents the "blue circle" column.

The trick is that I added small tanks at their bottom and I tweaked fuel out of them so they are empty and act just as aesthetically-pleasing fuel transfer mediators. Any other part could do the trick (even not crossfeed capable). Notice all engines except the center are connected to the "mediators", not to fuel tanks.

Real fuel tanks are connected from outside to inside. Then, these empty fuel tanks are connected from the central tank towards outside. You can see that even though all engines have different fuel consumption rate, they run all the time while fuel is drawn from the thin tank first, then from the grey tank, then from Jumbo.

Your real fuel connections can then go as in the diagram below, towards the center between fuel tanks, and out from center in the "fuel distribution floor".

Javascript is disabled. View full album

Edited December 31, 2013 by Kasuha

[michaelhester07]

The mod I'd recommend for you is kerbal engineer. It will show you what you're getting out of that rocket for numbers.

Aside from that with the first picture I imagine that your fail point is the Rocomax couplers between the first and second stage. Also if there are any SAS modules in the stack they'll also be fail points and they crush even easier than the couplers do. ( I know cause I built rockets where these were the fail point)

[CrazyCrazySam]

Thanks!

Kasuha's layout works perfectly.