Economically Refuel Your Spacestation (How?)

[Riph]

So now that currency is in with 0.24, I'm revisiting an old topic of frequent consideration:

How best to refuel your space station in kerbin orbit?

In my case, I have a small station that's little more than an orbiting fuel tank that I keep around so that my other ships can top off their tanks before setting of for other planets. It's at roughly 100km circular orbit.

I've managed to put a 90% full big orange tank in stable orbit with a ship that costs 140k credits.

I've also managed to put a Rockomax 32 (half a big orange tank) in orbit with a ship that costs 80k, which actually makes my dollar per fuel gallon worse than the prior model.

(In both cases, getting the tank from stable orbit to station intercept is just a few more fuel units because I have a nuke tug orbiting nearby that can drag them into place very efficiently.)

So, my question to you is, how would you refuel this station, with an eye for maximizing your dollar? If you have a ship design that works really well, please post screenshots and .craft files!

[pellinor]

Recover your fuel transport. What matters is costs after recovery. What works best for me is an expendable SRB-stage and a second stage that makes it to orbit in one piece. The second stage is then recovered near KSC. So you basically just pay for fuel. The empty SRBs are negligible, just don't use too many decouplers as they cost quite some money.

Late in the tech tree, SSTO Spaceplanes come into play, cutting down the fuel costs considerably.

[Laie]

Look closely at the cost: only 20% of a fuel tank's cost is the actual fuel. The tank itself is by far the more expensive bit. If you manage to recover (most of) your tanks, engines &c, the shipment will be quite cheap.

A lander leg can be worth a dozen parachutes. If the engine is too tall for legs, try to land the assembly on the side. Or keep a few drops of fuel in the tank and kill the final few m/s with a last burst from the engine.

[RuBisCO]

If your going to be moving just fuel your going to want an SSTO, like a spaceplane, doing circuits up and down to fill up your orbital fuel depo and minimizing fuel cost by using air-breathing engines.

[Riph]

Care to suggest a specific craft, Rubisco? My spaceplanes get themselves into orbit neatly and cheaply, but they can just barely get a 1 ton satellite into orbit.

If it can't get a rockomax 32 (at minimum) into orbit in one trip, then really you're only saving kerbal currency at the expense of player time.

[Laie]

Hmmm. Depends. A single orange tank should be doable within 200 stock parts, though I wouldn't place too high a bet on it.

But with procedural wings -- no problem. I built a huge tanker (carries 160t, that's about 5 orange tanks worth) at ~270 parts.

http://forum.kerbalspaceprogram.com/threads/11214-The-K-Prize-100-reusable-spaceplane-to-orbit-and-back?p=1290218&viewfull=1#post1290218

[Mesklin]

Single orange tank and slight more (maximum 44t dumb payload) you can deliver to orbit with only 71 parts (in SSTO lifter). See my old post with "Hexagon" lifter (44t to LKO) or "SuperHexagon" (145t to LKO)

[Reactordrone]

You could also offload excess fuel from capsules on recovery missions before de-orbiting them to slowly keep topping it up.

[Wanderfound]

Care to suggest a specific craft, Rubisco? My spaceplanes get themselves into orbit neatly and cheaply, but they can just barely get a 1 ton satellite into orbit.

If it can't get a rockomax 32 (at minimum) into orbit in one trip, then really you're only saving kerbal currency at the expense of player time.

You might like this one: https://www.dropbox.com/s/jhgl6sva73ntcum/Kerbodyne%20D7%20Heavy.craft

All stock, and specifically designed to lug Rockomax 64s. Manages it easily under old FAR; checking how it goes post-nerf as I type.

[Wanderfound]

All stock, and specifically designed to lug Rockomax 64s. Manages it easily under old FAR; checking how it goes post-nerf as I type.

Hmmn. We have new-FAR issues. Might be worth keeping it to a half-full tank until the engineers get it sorted.

[Nurph]

Scott Manley had a nice reusable lifter in his Reusable Space Program series, you might check that one out for some ideas.

[Riph]

Mesklin's Hexagons look freaking sweet. I love the design. But do my wandering eyes detect some noclipping abuse in the ram air intakes?

[Wanderfound]

As mentioned by others: spaceplanes. With a bit of work, you can do a stock-parts spaceplane that will lift a Rockomax 64 into orbit. If you land on the runway, the only expense is fuel. This should be less than √10,000 if you're doing it right.

[Riph]

I'm checking out the D7 Heavy now, Wanderfound. Thanks for being the only one to actually post a .craft file ;D

I have a question already though. Do you not have action group hotkeys to control your ram air intakes in pairs (rather than all or nothing?) You can control your drag and squeeze quite a bit more ascent power by making sure you always have just enough air intakes open. Any additional, unneeded air intakes that are open at any point are just drag.

[PPR]

I've found some guiding principles to keeping things economical:

1) Recover as much as you can. You get back 98% of your funds from parts you recover. (This is what makes the space-plane concept a good idea)

2) Use Solid rockets as much as possible (while less fuel efficient, they are substantially cheaper for the amount of delta-v they generate).

3) Keep it small. The larger your ship, the less economical it is because the larger engines are much less fuel efficient. By insisting on delivering the large orange Rockomax tanks you are limiting your options.

4) When designing your rocket, ask yourself: "Do I REALLY need that part?" People tend to overbuild their rockets. For example, you might use a Gigantor XL solar array (3000 funds and .35 tons) when a couple of Photovoltaic panels (100 funds and .005 tons) will provide all the power you really need. Or you might use radial decouplers when the radially-mounted rockets have the same burn time as the central tank. Just attach the radial tanks directly to the central tank and save yourself the cost of 4 decouplers.

[Wanderfound]

I'm checking out the D7 Heavy now, Wanderfound. Thanks for being the only one to actually post a .craft file ;D

I have a question already though. Do you not have action group hotkeys to control your ram air intakes in pairs (rather than all or nothing?) You can control your drag and squeeze quite a bit more ascent power by making sure you always have just enough air intakes open. Any additional, unneeded air intakes that are open at any point are just drag.

The D7 is being redesigned; the nerfing of air-breathing engines in the latest FAR update revealed some aerodynamic issues that I'd previously been able to power through. See the mod support thread ("power or aero?") if you're interested in how it's going. My little sports runabout is still flying well, though; the only thing I needed to change on that was to set the canards as flaps to give a bit more supersonic pitch authority. See https://www.dropbox.com/s/o7z87n1i1zqlv3n/Kerbodyne%20Overshoot%20VTOL.craft if you feel like taking it for a spin. The dinky VTOL jets are for Mun landings; don't try using them on Kerbin.

I usually go for all-or-nothing on the intakes, but gradually shut engines down as I ascend to concentrate the thin air. The intakes are more than enough to feed the six engines at low altitude, but once you get to 30,000m you've only got enough for one pair, even with all intakes open. Shutting some intakes at low altitude reduces the drag, but I didn't really care about low altitude drag.

Pre-nerf, the D7 could climb near vertically, so low altitude flight usually only lasted a minute or two. The surplus drag is trivial when you have that much thrust.

My general approach was to climb as fast as possible to 18,000m, then level off into a slow climb to 30,000m, shutting off the RAPIERs in pairs as I go. Once the turbojets die (or get so weak that I start losing speed, typically around 30,000m/Mach 4.5), I pull up the nose and turn the RAPIERs back on. The ram-air effect would revive the turbojets and allow me to keep half of the RAPIERs on air-breathing mode for a bit longer. I'd typically have the apoapsis over 70km before the turbojets died completely and it was time to shut down the intakes.

Still, splitting the intakes would be useful for fine control of air braking during reentry. But with the D7, I was running into an action group shortage: 1 was intakes, 2 for turbojets, 3 and 4 were the RAPIERs, 5 and 6 switched the RAPIER modes, 7 was the emergency brakes, 0 is always science on my planes and 9 was generally used for the cargo decoupling/undocking. 8 was the only one spare, and I often wanted that to activate something in the payload.

[Wanderfound]

https://www.dropbox.com/s/et4mcneguukur4o/Kerbodyne%20D7%20revision%20WIP.craft is the current D7 prototype, BTW. Very much a work in progress, though; at the moment, the control surfaces are tearing off any time the angle of attack gets over fifteen degrees...

[Riph]

I did have a look at that thread, Wanderfound. One thing that came to mind was that there was a part re-weighting in 0.24 that may also be messing with you. For example, the D7 Heavy uses several Mk2 Fuselage pieces which got heavier in 0.24. If she's flying differently now, you might double check how much mass you magically gained at patchtime.

(Also, what the heck is up with D7's tail section? Are you also abusing part clipping? Is everyone in this game abusing part clipping? D: )

[Wanderfound]

I did have a look at that thread, Wanderfound. One thing that came to mind was that there was a part re-weighting in 0.24 that may also be messing with you. For example, the D7 Heavy uses several Mk2 Fuselage pieces which got heavier in 0.24. If she's flying differently now, you might double check how much mass you magically gained at patchtime.

(Also, what the heck is up with D7's tail section? Are you also abusing part clipping? Is everyone in this game abusing part clipping? D: )

The D7 was flying fine under .24; I didn't run into trouble until the Ferram update. And minor changes in mass aren't really an issue when you're dealing with a plane that can fly with a 70 ton payload.

The only part clipping in the D7 is what you can do without the debug menu. It's mainly just that there's a probe core hidden in the fuselage so it can be flown as a drone. Mount girder section on the back of the fuselage, mount a probe core on top of that, then rotate the girder section inside the fuselage; you can stash most things with this trick. It's also how I got the shielded docking port mounted on the underside of the runabout (a cubic octagonal would be tidier, but I hadn't unlocked them yet on the save I built that in).

The only wing overlapping was what was necessary to achieve the shape I wanted; you need to get creative when you're cooking with a limited range of ingredients. IIRC, the tailplane was made from two deltas, two rectangular structurals and two triangular structurals, mounted on another pair of rectangular structurals. It was the only way I could make the large delta form I was after.

[Riph]

Single orange tank and slight more (maximum 44t dumb payload) you can deliver to orbit with only 71 parts (in SSTO lifter). See my old post with "Hexagon" lifter (44t to LKO) or "SuperHexagon" (145t to LKO)

Any chance we can get a .craft file for the Hexagon lifter, Mesklin? I'm trying to build something similar and I can't squeeze half the performance out of it that you seem to have.

[iStickyDuck]

You could try using something like Kethane if you're okay with mods.

Otherwise, SSTOs' are the way forward

[Pecan]

It is easier to build a SSTO than spaceplane for larger loads and a lot easier to fly them. Preferably capable of powered landing, assisted by parachutes works well for precision.

[Riph]

After about a thousand iterations of testing, I've got an SSTO that mostly works. Screenshot here: http://cloud-4.steampowered.com/ugc/44226402587467898/CA53EF46A188509084445D246C40258CAD93FCE5/

Astute observers will note that it is totally ripped off inspired by Mesklin's hex lifter that I like so much. Unfortunately, while Mesk's lifter supposedly can get a big orange 64 into orbit comfortably, mine struggles and just barely gets a 32 in orbit (at about 80% full).

After flying this thing so much, I began to realize that I know almost nothing about the 'correct' way to get a mostly air breathing craft into orbit. To me it seems there are two possible avenues:

1) Fly upward until 12k altitude, then level off and fly mostly horizontal. This maximizes your horizontal speed and keeps your air breathing engines happy for as long as possible. The downside is that you're basically riding your own apoapsis and once the air runs out at around 35k, you find that you're no where near your desired altitude of 70k, and have no momentum carrying you upward.

2) Fly upward until 12k altitude, then level off way less and continue rocketing upward at maybe 30 degree angle from the horizon. You very quickly run out of air, but at least in this strategy your apoapsis is way out in front of you, and your rocket engines can readily push it to 70k and beyond. The downside is that you have very little horizontal momentum, and it takes ages and tons of fuel to circularize this very eliptical orbit.

I'm suspecting there's an option 3) here that's way better and I'd love to hear it. Some posts talk of craft getting their apoapsis to 70k before the air breathing engines finally starve, which I have actually managed to do, but only with super lightweight fightercraft type SSTOs. Doing it with a rockomax 32 (or 64!) strapped to my back seems impossible.

[Pecan]

Not impossible but an awful lot of work. That's why I use rockets, such as:

Haven't tested this yet in 0.24.2 but in 0.23.5 it goes up in less than 5 minutes on a normal rocket trajectory, it comes back to the pad on a powered+drogue landing and gives 97-100% recovery depending on how close you can be bothered to get it. Who needs the effort of spaceplanes? If you like flying them then they're great, otherwise just do it the easy way.

[einsteiner]

I've been working on this for the last week to refuel my transfer stages (which I launch empty). I'm not a great spaceplane pilot, but I can get about 1.4 orange tanks worth of LFO into orbit. It has Procedural Wings and a Mechjeb module. Please forgive the unimaginative name.

Craft File

Edit: Version without Mechjeb module or Hullcam

Edited July 28, 2014 by einsteiner