Revisiting an old, simple one: The 10 fuel tank challenge.

[mathmavin]

I took another go at the challenge and tweaked my design slightly to give it just a little bit better performance - basically, I split up the 2 tank final stage from my earlier ship by sticking one of the tanks to the front using a decoupler and adding a couple of fuel lines. I calculated that it would give me roughly 270 m/s better delta V, with a more or less negligible effect on the rest of craft's performance in the earlier stages. The final stage has a low enough thrust that you can kick the fuel tank off sideways and get back on course within a second or so while always staying at full throttle, so you don't lose more than a few m/s there.

I also added a MechJeb unit which I used for information and for the final stage to keep the craft pointed prograde for the ~13 minutes of burn that the final stages take. It's easy enough, but rather tedious, to keep the craft pointed manually. Everything prior was fully manual, though.

Here is screenshot of the result and the craft: http://imgur.com/a/HYLgN

Final statistics are:

Altitude: 1562 km

Burnout speed: 5468.4 m/s

Orbital Energy: 13.3 MJ/kg

[mathmavin]

Last one tonight, I promise.

I did one more redesign. Basically, I took one of the fuel tanks normally attached to the aerospikes and stuck it on the top of the central stack, and made it so the fuel systems for the aerospikes and the low-powered engine in the middle were separate. I had to do some weird stuff with the positioning since I was left with 7 tanks total for the aerospikes, so I ended up with tanks in a 3-way symmetry below the aerospikes in such a way that none of them were in line with the engines. At burnout for the aerospikes, there's about 2 and 2/3 tanks left in the central stackup. At this point, the craft can barely make orbit - it actually goes past apogee and starts heading back down before it can get enough speed to raise the periapsis above the atmosphere.

I ended up with this result: http://imgur.com/a/12jwg

Altitude: 1149 km

Burnout speed: 5782 m/s

Orbital Energy: 14.70 MJ/kg

The high energy and speed are a result of having 2 and 2/3 tanks running at a vacuum ISP of 400, as well as basically doing a powered slingshot maneuver around Kerbin, compared to most other hyperbolic escape trajectories from launch.

I'm not sure how to attach files here, otherwise I'd post my craft files. I'd love for someone to duplicate my results to help verify my sanity. I'm hoping between my description here and the images that someone'll be able to do it.

[Nao]

My submission :3

Altitude: 1933 km edit:bugged results

Burnout speed: 6117 m/s edit:bugged results

Orbital Energy: 17.316 MJ/kg edit:bugged results

with this contraption:

(Edited out earlier submission since i'm bad at math and posting)

It had fun top Tank hat thou! =3

Edited August 14, 2012 by Nao

[khyron42]

I'll toss a "Wow!" and a "how does that even fly?" at both mathmavin's and Nao's designs. Nice outside-the-box thinking.

I love the fact that both are also best in all three categories at the time they were posted.

[mathmavin]

I think I've discovered a fuel bug with Nao's design (and probably mine, but I haven't retested mine yet). The first three tanks have 1200 units of fuel, and the aerospikes consume about 13.1 L/s each while the small engine consumes around 1.2 L/s in atmosphere. The three tanks should last for only about 45 seconds, but they last for well over a minute. If you remove the small engine, they only last for about 45 seconds (which is how long they should last). I think having that extra engine somehow reduces the fuel consumption.

I don't have an entirely dissimilar setup on my rocket, so it's quite possible I've inadvertently exploited this as well. I'll run some tests/do some number crunching tomorrow when it's not late to see if my fuel usage lines up with the consumption rates for the engines. I also don't know if this is a known fuel bug, but if it is not, someone should look into it.

Edit to add:

I did a test run to see what kind of performance you could get if you wanted to really exploit the bug - I ended up with this: http://imgur.com/a/oC6fk

Altitude: 2954 km

Burnout speed: 6803 m/s

Orbital Energy: 22.1 MJ/kg

Keep in these are BUGGED RESULTS.

Upon reviewing my design from my previous run, the small engine and the aerospikes never draw from the same fuel source, and the aerospike powered tanks burn out at 46 seconds, 1:17 seconds, and 1:47 seconds, for 3, 2, and 2 tanks, respectively. At a burn rate of 13.1 L/s, it should take 45.8 seconds, 30.5 seconds, and 30.5 seconds for each of those stages, which match up quite nicely with the timings on my ship.

Edited August 14, 2012 by mathmavin

[Nao]

There is indeed a fuel bug in my design! Thanks for finding that one out!

From some more testing it appears that if there is a fuel tank stack of at least 2 with engine at bottom, the fuel directed by fuel lines to other engines from every tank except the bottom one will have fuel drain halved, while still providing 100% thrust and indication of fuel usage by the engines. Engine at the bottom of the fuel tank stack will have its fuel drain counted normally for every tank in a stack.

Now I reeeeeaaaaaallllyyyyy can't wait for 0.17 T_T

Actually does anybody know if its a known bug or should we post it in bug thread? I vaguely remember hearing about it some time ago but didn't understand it until now. :/

[Clouds]

Actually does anybody know if its a known bug or should we post it in bug thread? I vaguely remember hearing about it some time ago but didn't understand it until now. :/

Better safe than sorry! Post it, I'd say.

[Ziff]

Actually does anybody know if its a known bug or should we post it in bug thread? I vaguely remember hearing about it some time ago but didn't understand it until now. :/

I am pretty sure this has been bug reported already, also, HarvesteR has already revamped the fuel flow code to work out the bugs. It's mentioned already in this post here -> http://kerbalspaceprogram.com/forum/showthread.php/15757-KSP-0-17-Update

[Nao]

Thanks Ziff, I'll pass on reporting this, I'm sure it will be gone with update.

Also moAr success This time with mechJeb because it would be a too much pain to fly for 20mins with constant corrections required by the design.

This time i hope no more bugs are involved.

Altitude: 1278 km

Burnout speed: 5855.6 m/s

Orbital Energy: 15.26 MJ/kg

There is still room for improvement but I'm satisfied with its aggressive shape =3

SS taken a just after reaching 0,0L fuel (i'm slow) - but engine is still warm and acceleration is still not 0 to make a proof.

Alzo ...letz go explore Beta Centauri

Detaching the top tanks was fun since the strength of decoupler was greater than engine so i would yaw the craft 30deg, shoot it, then return to prograde position , with empty tank merrily passing my craft 1m from the side. =3

[Caroliano]

Edited with a correct submission:

Altitude: 386 km

Burnout speed 5847 m/s

Orbital Energy: I'm getting a negative number. Will look at it latter

If I stage it I lose control, so this design ended up very inefficient. But I did reach kerbol scape velocity, as planned...

And, after all, the challenge is 10 FL-T400 fuel tanks, as this thread posts appear to be, or 4000 fuel units, as is said in the first post?

Edited August 14, 2012 by Caroliano
I didn't properly read the first post.

[Nao]

Unfortunately you need to reread OP since the challenge is more complicated than just 10 tanks.

You need to burn at max throttle whole time and take screenshot at the moment of using all fuel!

[Caroliano]

Right, I did burn at max throttle, but I overlooked the "at the moment of using all fuel". I edited my post with a correct submission of the same rocket. And what about using other types of fuel tanks? What is the current understanding?

EDIT:

Ok, I simplified the design, but it got nowhere near the same efficiency... I think I may have hit a fuel cross feed bug in the earlier design, or something like that... The 0.17 version can't come fast enough!

Edited August 14, 2012 by Caroliano

[trekkie_]

it's also important to note that slower fuel burning engines like the aerospike, won't necessarily get distance. in some configurations the higher thrust engines will get you further because of the extra speed they allow for early on. by my calculations you have to be going at least 1500-2000m/s once you reach 100K meters in order for the 20 thrust lv909 engines as a last stage, to not fall back to the planet at some point, and to go the distance. otherwise you will lose too much speed from the lack of thrust.

[khyron42]

Ok, with the new fuel bugs found, I'm not even sure how to update the results. I also wonder if this bug was a small factor in some of the earlier designs, but I doubt it since most of them were feeding inwards from two tanks towards 1 engine, instead of the other way around.

Caroliano, the challenge accepts both 10 LF-T400 results, and "4000 fuel" results. 10 tank is preferred, but if there's a 4000 fuel result that's better than the current best 10 tank result for any category, I'll list it separately in parentheses.

[Clouds]

Not related to this challenge directly...I re-piped a ship of mine and now it gets about 1/2 the fuel consumption for the same number of engines and near-identical everything else. I was trying to make a minimalist grand tour capable 3-Kerbal-capsule rocket and this pretty much means it's a pointless endeavor since this bug, coincidentally, kicks in for properly efficient designs (all engines running, sequential fuel piping).

REALLY hoping .17 comes out, soon...

[trekkie_]

all stock parts, no jet engines or aerospike.

Edited August 15, 2012 by trekkie_

[Mindofscience3]

http://imgur.com/c1N6l

velocity at burnout: 3998.1 m/s

altitude at burnout: 275km

Energy: 3.859 MJ/kg

4000 units of fuel in as few fuel tanks as possible.

[antbin]

Well, I ripped off Nao's design, except I used 18 half tanks and 1 regular tank. Also Mechjeb (as I am lazy).

That would make it 15.33 MJ/kg?

Oh well, this will all change with 0.17... any day now!

Edited September 17, 2012 by antbin

[sporkafife]

What if we have the fuel bug fix installed? Would we still have to burn at full throttle the whole time?

Without having to burn at full throttle and with the bug fix, it's probably very possiblr to take one of them ships to Minmus. Scott Manley on YouTube got to Minmus with 3 fuel tanks, and did a grand tour with 4!

[walia6]

This is gonna be fun! Doing right now

[Peppe]

Anyone getting ~4k m/s or more in .17?

[O_Prime]

UPD: mistake in formula(

Edited September 29, 2012 by O_Prime

[combatEngineer]

[ATTACH=CONFIG]33567[/ATTACH][ATTACH=CONFIG]33563[/ATTACH][ATTACH=CONFIG]33564[/ATTACH][ATTACH=CONFIG]33565[/ATTACH]

This is the best i could do with 10 tanks On V 0.17 I got in Orbit around the moon..lol. Pretty much basic parts that are in all the other Vers of KSP. May have got further without the asas on it. but I figure thats pretty good with just 10 tanks. gonna try to take the asas off and try to orbit Minmus with the same config. I Started the gravity turn at 10km, I turned to approx 50 degrees and when the AP hit 80km, I coasted to around 75km and then fired at 90 degrees till in orbit. I'm not on the forums much, but I'd love to hear what you guys think of it.

Edited September 25, 2012 by combatEngineer
forgot pic of ship in orbit..lol

[Nikozc]

Final speed = 5252.4 m/s

Final altitude = 131 884 m

Efficiency = 0.5*(5252.4)^2-3530.461/(600+131.884) = 13.793 MJ/kg