[0.20] Modular Fuel System 1.3/realistic fuels, reconfigurable fuel tanks and engines

[CaptainArbitrary]

I haven't done very extensive testing, but first impressions -- I'm not sure why I would want to use anything but LF/OX combo.

Because the specific impulse of LH2/LOX is on the order of 30% higher than the specific impulse of RP1/LOX or RP1/N2O4. You get more thrust per unit of weight of propellant if you're using LH2/LOX.

The trick to exploiting this, though, is to not do what you did. You compared two rockets with the same propellant volume. Instead, compare rockets with the same propellant mass. Since LH2 is so much lighter than RP1, you need a much bigger tank to carry an appropriate propellant mass. But that's okay, because that much bigger tank of LH2 will end up being lighter than the smaller tank of RP1, because you need less propellant mass if you're using LH2, because the specific impulse is higher.

To really get an intuitive grasp of how liquid hydrogen works in the real world, look at things like the Space Shuttle external tank and the Delta IV common booster core. They're both bloody enormous, but because they're filled with mostly liquid hydrogen, they're very lightweight for their size, and make very efficient rockets.

But use having nuclear engines operate on "Oxidizer" and "Fuel" is stupid hence why I love RealFuels.

It's stupid, yeah, but not quite as stupid as you might think. One of the virtues of a thermal rocket is that you can use anything as reaction mass. The chemical properties of your reaction mass don't matter, because you're not reacting them chemically. If you want, you can use plain water as reaction mass in a thermal rocket, and get perfectly acceptable results.

What is very stupid, though, is pumping "liquid fuel" and "oxidizer" through a thermal rocket and getting an Isp in the 800s. In a thermal rocket (whether it be nuclear or any other kind) the exhaust velocity of your reaction mass goes by the square root of the exhaust's enthalpy. At NERVA temperatures and pressures, pure hydrogen has an enthalpy of nearly 40,000 kilojoules per kilogram, giving it an exhaust velocity of over 8,500 meters per second! That's where the sky-high Isps in the 800s come from. When you pump through other fluids, you get lower exhaust velocities. Pumping pure water through a NERVA-type engine, for instance, would give you an Isp in the 200s, if I remember right. Which frankly isn't that bad when you consider that you're using water.

Edited July 19, 2013 by CaptainArbitrary

[Greys]

wonder how that would work? Would it restrict the normal engines to a limited number of restarts, unless you changed it over to use hypergolic fuels? and wouldn't changing it over reduce mass on an engine that no longer needs the "plumbing" for an ignition source?

i know the third stage in the Saturn V could be restarted and i think the LEMs decent stage as welll, im not sure though, along with the space shuttles main engines but i cant remember how many times for any of them and they are the only ones i can think of at the moment unless any of SpaceXs stuff can...

If it were me I would make a system where restarts are a resource, there's a one time charge for igniting, randomized lesser charges during the burn, and above some throttle level or heat value a constant low level charge.

So say an engine has 5 units of Ignition, and every time the throttle goes from 0 to not 0, 1 unit of Ignition is removed.

Randomly during a burn, check period based on throttle level (100%, every 2 minutes, 50% every 10 minutes, etc) if the check fails 0.05 units of Ignition are removed.

If the throttle is greater than 80% remove 0.001*((throttle%-0.8)*5) per second.

Ignition could be treated as a part resource and made transferable to maintain space programs, and variance in the numbers above would serve as a point of balance between differing engines

[Designer225]

Have anyone thought of engines automatically switches from a combination to another combination (Ex: LiquidFuel+LiquidOxygen to LiquidH2+LiquidOxygen)? I think switching fuel combinations can be more effective, as LH2 is better to reach orbital speed than RP-1 (Kerosene, what OP thought as LiquidFuel), while RP-1 is more optimal on early launch than LH2 because it saves fuel.

Also, because LH2 and LOX both needs to be insulated (they are cryogenic), the insulation significantly increases the weight of a fuel tank.

I think the engines should have combination switch capability, even during flight so that it can use fuel much more efficiently (hopefully).

[CaptainArbitrary]

Have anyone thought of engines automatically switches from a combination to another combination (Ex: LiquidFuel+LiquidOxygen to LiquidH2+LiquidOxygen)?

It's really hard to imagine how you'd do that with a chemical rocket. Not only are the pumping systems completely different for different kinds of propellants, but also the combustion chamber and nozzle have to be designed completely differently. For instance, a typical RP1/LOX rocket engine might have a chamber pressure of 75 atmospheres while a typical LH2-burning engine would have a chamber pressure of 200 atmospheres or more.

[Designer225]

It's really hard to imagine how you'd do that with a chemical rocket. Not only are the pumping systems completely different for different kinds of propellants, but also the combustion chamber and nozzle have to be designed completely differently. For instance, a typical RP1/LOX rocket engine might have a chamber pressure of 75 atmospheres while a typical LH2-burning engine would have a chamber pressure of 200 atmospheres or more.

Yeah... the change is gonna be very abrupt.

[RuBisCO]

Personally I think they should have modes like Easy, Moderate and Hard in this game, or better yet a click menu when ever you start a new sandbox or career mode to choose your realism/difficulty like realistic engines (limited restarts & limited throttling), mechjeb/autopilot, communcation delay & relay stations, life support requirements, etc, etc.

[Designer225]

Speaking of which, do you have to install all of the .cfg files the mod pack comes? Or just some you need?

[RuBisCO]

Speaking of which, do you have to install all of the .cfg files the mod pack comes? Or just some you need?

Yes, I think so, you can modify them to your wishing though.

It's stupid, yeah, but not quite as stupid as you might think. One of the virtues of a thermal rocket is that you can use anything as reaction mass. The chemical properties of your reaction mass don't matter, because you're not reacting them chemically. If you want, you can use plain water as reaction mass in a thermal rocket, and get perfectly acceptable results.

IRL a nuclear thermal rocket engine is going to have problems with diffrent propellents, say LOX for example, high temperature oxygen is horribly corrosive making a lining for nuclear fuel that can survive that and 2600K+ operation and incredible neutron loads is a tall task! IRL the most likely fuels for a NTRE are Hydrogen, Ammonia and Water, even methane or CO2 would have coking problems (forming carbon soot in the reactor that would clog it up). Water would be cheap, readily minable, dense, but with a ISP only as good as LO2/LH2 engines at best. Ammonia would provied better ISP than water (550 was acheive with a russian nuclear rocket) but would be harder to find in space and a little less dense, of course your aware of all the pros and cons of hydrogen.

If it were me I would make a system where restarts are a resource, there's a one time charge for igniting, randomized lesser charges during the burn, and above some throttle level or heat value a constant low level charge.

So say an engine has 5 units of Ignition, and every time the throttle goes from 0 to not 0, 1 unit of Ignition is removed.

Randomly during a burn, check period based on throttle level (100%, every 2 minutes, 50% every 10 minutes, etc) if the check fails 0.05 units of Ignition are removed.

If the throttle is greater than 80% remove 0.001*((throttle%-0.8)*5) per second.

Ignition could be treated as a part resource and made transferable to maintain space programs, and variance in the numbers above would serve as a point of balance between differing engines

That sounds rather complex, why not just a simple counter? I guess it would make sense for engines with use a seperate fuel for the pumps, like hydrazine or H2O2. Such engines could have as many restarts as pumping fuel allows.

Edited July 19, 2013 by RuBisCO

[ialdabaoth]

Actually, LANTR solves this by ONLY pushing the H2 through the NTR; LOX is run through the coolant loop and then dumped into the propellant stream after the LH2 goes through the reactor, which effectively makes the whole thing a hypergolic LH2+LOX rocket. Your Isp drops from 920s down to about 540s, but that's still far better than you get from a standard chemical LH2+LOX rocket (and your fuel density goes way up).

Edited July 19, 2013 by ialdabaoth

[hellion13]

im glad someone knows his rocket science lol, i understand deltaV and how to calculate it, same for ISP and even enough of 2 and 3 body motion to get to places but as for how the stuff actually gets put together to make a rocket? no way

makes me wonder about the conversations about fuel types that went down during the Apollo era, those guys were all in their 20s pretty much playing Kerbal for real

Edited July 19, 2013 by hellion13

[CaptainArbitrary]

im glad someone knows his rocket science lol, i understand deltaV and how to calculate it, same for ISP and even enough of 2 and 3 body motion to get to places but as for how the stuff actually gets put together to make a rocket? no way

The broad strokes of the thing really aren't that complicated, to be honest. Conceptually, a rocket engine is somewhat less complex than the engine in a car. The actual engineering of the thing is far more complicated, of course, but as far as the overall big-idea aspect of the thing, rocket engines are really not that complex. Cold thing goes in, hot thing comes out, and the interesting stuff revolves around how you make the cold thing hot.

[RuBisCO]

The broad strokes of the thing really aren't that complicated, to be honest. Conceptually, a rocket engine is somewhat less complex than the engine in a car. The actual engineering of the thing is far more complicated, of course, but as far as the overall big-idea aspect of the thing, rocket engines are really not that complex. Cold thing goes in, hot thing comes out, and the interesting stuff revolves around how you make the cold thing hot.

Much of the complexity has been reduce these day thanks to computer controles: in the olden days an extremely complex arrangments of valves and tubes was need to keep the engines harmonics stable, expentially so as the engine's combustion chamber got large. Now we can do that with computer controlled injectors, hence the new F1 engine project droping the number the engine parts from over 5000 to less then 100!

[Chestburster]

May I suggest addition of Hypergolic Propelants?

I'm going to release a "RealWorld" edition of my configs wich will add Hydrazine (H2N4) and Dinitrogene tetroxide (N2O4) + other tweaks with the next version.

what is this compatable with?

All the bigger mods out there. KW, NP2, Firespitter, KSPX, HexCans, Spherical Fueltanks...

I haven't done very extensive testing, but first impressions -- I'm not sure why I would want to use anything but LF/OX combo.

If you want to go interplanetary you need LF/OX or you will have a nearly depleted tank after a few weeks with LOX ;-)

Speaking of which, do you have to install all of the .cfg files the mod pack comes? Or just some you need?

You can keep all of them, even if you dont have the mods. ModuleManager can't modify what doesn't exist^^

-------------------

As always: If you want support for other mods i dont have included yet or find bugs with the configs.. feel free to message me.

Next version of the configs will have few bugfixes, tweaks to HexCans and more.

[Chestburster]

MFS Configs V1.4 released! (MFS V1.3)

Currently supported mods: B9, KSPX, Spherical Fuel Tanks, HexCans, RLA, Firespitter, Kethane, KW-Rocketry, NovaPunch2, AIES, Fatman Nuclear Engines and double NERVA, TT Pitch Vector Engine

Changelog:

- reverted IntakeAir density (too much work needs to be done for a small gain)

- added single tank definitions for RCS, Oxy and Xenon

- added missing KSPX tanks

- added missing vanilla tanks

- added RLA Xenon tanks

- fixed HexCans (Liquid Barite tanks can't be 50% pre-filled with the current system)

Download V1.4

Delayed the RealWorldFuels part for now until i tested it further.

@ialdabaoth

would it be possible in some way to change the part mass via the engine switcher?

Edited July 22, 2013 by Chestburster

[ialdabaoth]

MFS Configs V1.4 released! (MFS V1.3)

Changelog:

- reverted IntakeAir density (too much work needs to be done for a small gain)

- added single tank definitions for RCS, Oxy and Xenon

- added missing KSPX tanks

- added missing vanilla tanks

- added RLA Xenon tanks

- fixed HexCans (Liquid Barite tanks can't be 50% pre-filled with the current system)

Download V1.4

Delayed the RealWorldFuels part for now until i tested it further.

@ialdabaoth

would it be possible in some way to change the part mass via the engine switcher?

I'm already considering that. Right now I'm playing with different options for when 0.21 comes out.

Oh! Also.

Note how the standard RCS tank has volume=100.

Note how the mini RCS tank has volume=40, and KSPX's mini tank has volume=50.

Note that there's NO WAY that's possible.

[Chestburster]

I'm already considering that. Right now I'm playing with different options for when 0.21 comes out.

Oh! Also.

Note how the standard RCS tank has volume=100.

Note how the mini RCS tank has volume=40, and KSPX's mini tank has volume=50.

Note that there's NO WAY that's possible.

Hu? What do you mean? I took the volumes right out of the part files.

And with no way possible you mean the 50% filling or mass changing?

[ialdabaoth]

Hu? What do you mean? I took the volumes right out of the part files.

And with no way possible you mean the 50% filling or mass changing?

I mean that the white spherical shell in the radial RCS tank has a radius of approximately 0.33 meters, while the standard RCS tank has a radius of 0.625 meters and a height of 0.625 meters.

pi * r * r * l = (3.1415926583 * 0.625 x 0.625 x 0.625) = 0.767

(4/3 * pi * r * r * r) = (3.1415926583 * 0.33 * 0.33 * 0.33) = 0.113

Thus, the radial RCS tank can't have more than about 15 fuel units in it.

The KSPX's mini-RCS tank is a straight scale-down of the standard, and is exactly the size of the Oscar, so *it* can't have more than 12.5 units in it.

I'm somewhat hesitant to make this change, because it's pretty sweeping.

If I do, I'll probably change the density of MonoPropellant to match HAN, which is 0.007. Which means the radial-RCS effectively becomes 26.25 effective-units and the Oscar-size becomes 21, but the 1m and 2m tanks almost double in effective fuel.

Edited July 19, 2013 by ialdabaoth

[Oxmaster]

Hello.

How to install this, and how can I download advanced version?

When I put folders from GameData to GameData to KSP I can change the amount of fuel, not the types.

[Chestburster]

That's interesting. I never compared the sizes of the tanks. I just took the fuel values out of their files, hoping the mod devs would have thought of that.

Btw. i noticed that im missing the mini-RCS tank with 45 Monoprop in my part folder. must have deleted it some time ago.

i will change the values of the radial tanks and the mini in the next version.

[Greys]

You really should just ignore the Oscar B tank, it's nonsense. It contains roughly a third less by volume than any other tank.

Edit: For your consideration

Just realized there's no context here

This is the stock 8 cylindrical fuel containers, the blue line is their volumes ordered from smallest to largest, Oscar B-Jumbo 64 on left, FL-T200-Jumbo 64 on right

Edited July 19, 2013 by Greys

[s20dan]

I've noticed some of the densities for real fuels seem to be way off.

The ratio between L-H2 and LOx (O2) is correct, but take a look at the density ratios for Xenon gas and Nuclear fuel, they are way off by magnitudes.

Xenon gas has a density of 0.005894 g / cm3 which is 11.876 times less dense than Liquid H2

Which would give it a density of 0.000038 in the config.

Nuclear fuel (Uranium) has a density of approx 19 g/cm3, LOx has a density of approx 1.14 g/cm3

LOx is 0.007 in the config, Nuclear fuel should be 0.11667, not 0.02.

L-H2 has a density of 0.07085 g / cm3 which is 16.1 times less dense than LOx.

LOx has 0.007 as density in the config, divide that by the mass ratio and you correctly get 0.00045 for L-H2.

Also why not just use standard density units of g/cm3? Would make things easier

Ah I just noticed this : // all density units are in kg per Kerbo (1 Kerbo = 6.25 cubic meters)

But still I think the ratios between fuel types should remain the same no matter what system you use to measure it shouldn't it?

[Greys]

s20dan, the issue is that density is a measure of mass, and nothing else. One unit of resource has a mass of 'density'. On the other hand, the volume of a unit of resource (it's volume density), has nothing to do with anything, and through out stock parts is at best inconsistent and in the case of ElectricCharge, varies from 200~Units/m^3 to 10,000~Units/m3.

The problem is with how KSP treats resources and tanks, and it's not likely to change.

[s20dan]

@Greys, I think we may be talking about different things here. My post was not meant to add any more information with regards to inconsistent volumes throughout stock/mod parts.. But rather shine light on the fact that the fuels themselves have incorrect values as far as I can see (From the Real fuels part of the mod, which is supposed to be realistic).

This is before the fuels even get applied to a tank... The inconsistent tank volumes just further exaggerate the issue.

(In fact they are their own issue entirely, as tank volume is set by guesstimation )

EG the XenonGas fuel, should have a density approx 1/10 of what it is right now. Regardless of what tanks you put it in.

But, saying that I'm rather ill at the moment and quite tired, so my math could be way off

[Greys]

What I'm saying is that you can't really begin to quantify these things when the closest you can get to a quanta is "unit". Could be that they're set correctly, you're just looking at the wrong number of units.

And then everybody Ffffffffffs.

[CaptainArbitrary]

Xenon gas has a density of 0.005894 g / cm3

At standard temperature and pressure. Xenon propellant is stored in a supercritical state at around 1450 psia, which would have a density of around 0.538 kilos per liter.

Also why not just use standard density units of g/cm3? Would make things easier

Because the game doesn't. The game requires that densities be specified in metric tons per volume unit, where the stock volume unit is arbitrary (and incredibly wrong). It's possible to do a total conversion of all volumetric units over to liters, which I've done in my game, but it's a huge pain in the butt, and you're still locked in to using metric tons per liter in the resource config file.