[First Release] Realistic-Fuel Tank Pack

[nivvydaskrl]

Have you ever lamented that the stock engine is just too powerful? Have you ever wondered what it would feel like to have the limitations and realism of real-life fuels?

If so, this fuel pack is for you. Three common liquid bipropellants are modeled in 1.75m diameter and 3m diameter tanks, with realistic mass with regards to volume and realistic energy with regards to weight [EDIT: all based on the stock engine's thrust and fuel consumption values].

• 
  [li]Liquid Hydrogen/Liquid Oxygen: Used by the Space Shuttle, this is a very lightweight fuel with a large amount of energy for its weight. While bigger, heavier tanks are required, this is the best mass to energy ratio around.[/li]
  [li]RP-1/Liquid Oxygen: Used by the Saturn V and Falcon 1, this fuel replaces liquid hydrogen with high-grade kerosene. While the energy per kilogram suffers a significant hit, each tank holds a lot more fuel mass, which means a lot more total energy per tank.[/li]
  [li]Hydrazine/Hydrogen Peroxide: Rarely used, this fuel mixture is even more dense than RP-1/LOX, but its instability makes it a frightening thing to strap yourself to the top of. Still, if you need the maximum energy per fuel tank possible, regardless of explosive death, this is the propellant of choice.[/li]
  

Details: The LH2/LOX tanks are 88% fuel by mass, while the RP-1/LOX and Hydra/Hydro tanks are both 95% fuel by mass, and the fuel units of each tank represent 25 N*s of impulse per fuel unit. Total fuel volume is based on the volume of the tank cylinder, and fuel mass is based on that volume times the fuel's density as found on Wikipedia.

Note: I'd suggest NovaSilsko's Silisko Industries Doughnut Research & Spacecraft Development Pack and SundayPunch's Sunday Punchs KSP Parts packs, as they have a wide variety of adapters, shrouds, and really cool stuff for use with 1.75m and 3m tanks. NovaSilsko's pack was used in the rocket built for the below screen capture.

Second Note: If either of the above people have any concerns with me mentioning them or using their parts in the screencap, please contact me and I'll remedy the situation ASAP.

Edit 1: v.1.1 -- Corrected stacking issues, renamed 1.75 meter diameter tank resources appropriately.

Edit 2: v 1.2 -- Altered part names to be underscoreless. Made tank connections rediculously strong to correct stack falling apart issues.

[DoctorEvo]

We need to make some engines to go with this. Multiple engine types (pressure-fed, gas-generator, staged-combustion...) of varying efficiencies could be simulated by tweaking the fuel consumption parameters.

Here, check out what I made in Sketchup:

If I can figure out how to get it in-game, then configure it with the parameters you outlined in the other thread, it should be directly compatible with these fuel tanks, correct?

[Andras]

Your '2m' tanks are really 1.75m, if you redo these in the future could you rename the parts folders and components to reflect this?

I also find it impossible to stack one of these tanks on another, it flashes green and then goes red.

[nivvydaskrl]

Yeah, I was following someone else's naming convention. It's correct in the in-game name and description, but...yeah.

The flash-green-then-red thing is something I'd noticed too, but I'm not sure what the problem is. I'll try shrinking the collision mesh a tiny bit and see what happens.

This is my first addon and my first 3D modeling job ever, so I'm sure bugs will abound.

Edit: Okay, I renamed things properly and corrected the issues with stacking. New zip is attached.

We need to make some engines to go with this. Multiple engine types (pressure-fed, gas-generator, staged-combustion...) of varying efficiencies could be simulated by tweaking the fuel consumption parameters.

Here, check out what I made in Sketchup:

If I can figure out how to get it in-game, then configure it with the parameters you outlined in the other thread, it should be directly compatible with these fuel tanks, correct?

Yep! The fuel figures assume that the stock engine has 100% efficiency, and that the fuel energy density is 25 N*s per unit, as listed above. Any engine which consumes 1/25th or more of its thrust in fuel units per second is technically compatible, at varying efficiencies, IIRC.

[Andras]

Cool, I got the revised pack and I'll give them a spin.

[nivvydaskrl]

Cool, I got the revised pack and I'll give them a spin.

Sweet!

One thing to note -- right now, the tanks aren't liking it when they're stacked high, and break apart. I'm going to tweak the breakapartforce numbers upwards quite a bit, since these are supposed to be main fuselage components and quite heavy besides.

[Andras]

Yeah, I just found the 'weakest link'

I'm going to try breakingforce=10000

[Andras]

REMOVE ALL UNDERSCORES FROM ITEM NAMES

http://kerbalspaceprogram.com/forum/index.php?topic=1451.0

PS- 10000 wasn't enough to prevent break ups

[DoctorEvo]

Yep! The fuel figures assume that the stock engine has 100% efficiency, and that the fuel energy density is 25 N*s per unit, as listed above. Any engine which consumes 1/25th or more of its thrust in fuel units per second is technically compatible, at varying efficiencies, IIRC.

25 N*s/unit, huh?

... Oh, I see. You made it compatible with the stock engine. That actually means it is NOT in-spec with the RD-180 parameters you derived in the other thread, since those were in compliance with a 1 kN*s/unit spec that would satisfy the thrust=consumption condition.

So yeah, not EXACTLY what I had in mind (I was thinking just the straight 1 kN*s/unit for the sake of simplicity, and perhaps retroactively edit the stock parts to comply), but I guess it's not TOO difficult of a conversion. The only other issue was that if I had to pick one of the stock parts to be 'broken' or unrealistic, I wanted it to be the engine and not the fuel tank, since it's TWR is all out of whack anyways (actually offsetting its absurd ISP and ultimately making it 'balanced' anyways, whereas by the 25 N*s/unit spec it is seriously underpowered).

But, I dunno. I'd love to hear your thoughts as well.

Also, one more thing - I think I'd like to knock down the propellant V_e values a little more than what's on that chart I showed you, as they seem to be significantly more than real engines actually get out of said propellants. I'd like a generic engine with an exact-to-spec fuel consumption to actually be representative of real-world performance and not overpowered if possible.

I'm thinking: LH2/LOX = 4200 m/s, RP-1/LOX = 3000 m/s, Hypergol = 2800 m/s as a starting point, based on performances of typical engines using these propellants (for instance, despite the higher performance of hypergolic fuel relative to RP-1, hypergols are generally used in pressure-fed orbital maneuvering rockets rather than in main engines, and thus achieve lower ISPs in their most common role), and engines that achieve exceedingly-high efficiencies may be modeled as such on their own.

*edit*

Actually, scratch that about penalizing hypergolic fuel based on the performance of pressure-fed engines; it's better for gameplay if we have all the various engine types cross-compatable with different fuel types. Generic fuel tanks could be designated by their fuel type, and generic engines can be designated by their engine type (with pressure-fed engines being the lightest but least efficient, and staged combustion being the most efficient but somewhat heavy). Then allow players to mix and match to their liking.

Though it DOES appear that I can't find any high-ISP hypergolic engines whatsoever... not even the Viking from the Ariane or LR-87 from the Titan II seem to match RP-1-fuelled rocket engines (RD-275 is the best I can find). Maybe 2900 m/s would be appropriate... even THAT is a little high.

[nivvydaskrl]

REMOVE ALL UNDERSCORES FROM ITEM NAMES

http://kerbalspaceprogram.com/forum/index.php?topic=1451.0

PS- 10000 wasn't enough to prevent break ups

Ah, crud. I didn't see that. Okay. underscore removal and upping of connection strength by a further factor of 10 should be accomplished sometime tonight.

25 N*s/unit, huh?

... Oh, I see. You made it compatible with the stock engine. That actually means it is NOT in-spec with the RD-180 parameters you derived in the other thread, since those were in compliance with a 1 kN*s/unit spec that would satisfy the thrust=consumption condition.

So yeah, not EXACTLY what I had in mind (I was thinking just the straight 1 kN*s/unit for the sake of simplicity, and perhaps retroactively edit the stock parts to comply), but I guess it's not TOO difficult of a conversion. The only other issue was that if I had to pick one of the stock parts to be 'broken' or unrealistic, I wanted it to be the engine and not the fuel tank, since it's TWR is all out of whack anyways (actually offsetting its absurd ISP and ultimately making it 'balanced' anyways, whereas by the 25 N*s/unit spec it is seriously underpowered).

But, I dunno. I'd love to hear your thoughts as well.

My main reasoning behind this is a) I suck at modeling anything more complicated than a cigar box, and the stock engine is an engine everyone is guaranteed to have. TWR is important, but even a 100% efficiency is somewhat overblown. It just ups the difficulty a bit.

Also, one more thing - I think I'd like to knock down the propellant V_e values a little more than what's on that chart I showed you, as they seem to be significantly more than real engines actually get out of said propellants. I'd like a generic engine with an exact-to-spec fuel consumption to actually be representative of real-world performance and not overpowered if possible...

Hm. That could work, but my rationale is that energy potential is a trait of the fuel, while fuel economy is a trait of the engine. If the fuel numbers are clearly some scaled figure with respect to actual chemical potential energy available, creating an engine with a given efficiency is simple math. If I build an inefficiency factor into the fuel tank, then pinpointing and implementing a given efficiency for an engine is a more complicated computation.

[DoctorEvo]

My main reasoning behind this is a) I suck at modeling anything more complicated than a cigar box, and the stock engine is an engine everyone is guaranteed to have. TWR is important, but even a 100% efficiency is somewhat overblown. It just ups the difficulty a bit.

I kinda figured that was your reasoning. But that just means that the LFT is ridiculously overpowered. Also, it means it involves more math (i.e. another arbitrary conversion factor) for anyone who wishes to adopt this system.

Also, are you sure it's not 25 kN*s/unit? I went back and figured what it'd be if you wanted to make the stock tank behave like a proper LH2/LOX tank, and I got roughly 19 KILOnewton-seconds per unit (of course, this is presuming that the game units are tons and kilonewtons - and seconds, of course).

Hm. That could work, but my rationale is that energy potential is a trait of the fuel, while fuel economy is a trait of the engine. If the fuel numbers are clearly some scaled figure with respect to actual chemical potential energy available, creating an engine with a given efficiency is simple math. If I build an inefficiency factor into the fuel tank, then pinpointing and implementing a given efficiency for an engine is a more complicated computation.

'Simple math' to you and me is still another step for other developers who might want to use these standards. Developers who, more likely than not, are just looking to make a simple, fictional, generic-yet-functional part. And if we can make it so these generic parts are NOT inherently unbalanced and overpowered, I think there's something to be said for that.

Also, the 'chemical potential' from that chart is in no way absolute.

That was my plan. Fuel is fuel is fuel. How a given engine handles it is based on engine design and type and whatnot. :3

Alright, but do you agree that the numbers could afford to be knocked down a bit to represent more 'typical' than 'ideal' figures? Because as it is, it appears that if you use any of those values from the table outright without penalizing your engine, you've already got unrealistically efficient performance (in fact, I can't find a single I can only find one real engine that exceeds the performance listed in the table, despite the table's somewhat low combustion pressure of 1000 PSI).

[nivvydaskrl]

I kinda figured that was your reasoning. But that just means that the LFT is ridiculously overpowered. Also, it means it involves more math (i.e. another arbitrary conversion factor) for anyone who wishes to adopt this system.

Also, are you sure it's not 25 kN*s/unit? I went back and figured what it'd be if you wanted to make the stock tank behave like a proper LH2/LOX tank, and I got roughly 19 KILOnewton-seconds per unit (of course, this is presuming that the game units are tons and kilonewtons - and seconds, of course).

Yeah, it's kN per second. I dorked that up.

As for stock LFT vs. LFE -- something has to be overpowered between the two, because the pairing is certainly unrealistic.

'Simple math' to you and me is still another step for other developers who might want to use these standards. Developers who, more likely than not, are just looking to make a simple, fictional, generic-yet-functional part. And if we can make it so these generic parts are NOT inherently unbalanced and overpowered, I think there's something to be said for that.

Also, the 'chemical potential' from that chart is in no way absolute.

Well, it's either force other developers to divide by twenty five, or make these tanks completely incompatible with every other engine and engine mod released to date, and thus temporarily (or permanently, if no-one else creates engines using the same values) useless. Most engine addons that come recommended are less efficient than the stock LFE, but are still in its general realm of consumption to thrust values.

Alright, but do you agree that the numbers could afford to be knocked down a bit to represent more 'typical' than 'ideal' figures? Because as it is, it appears that if you use any of those values from the table outright without penalizing your engine, you've already got unrealistically efficient performance (in fact, I can't find a single I can only find one real engine that exceeds the performance listed in the table, despite the table's somewhat low combustion pressure of 1000 PSI).

Well, I'd rather model the available energy of the fuel as accurately as possible given my resources -- as I said, energy is a fuel trait, efficiency is an engine trait. Similarly, as I said, most respected developer's engines seem to undershoot the stock engine's efficiency by a decent percentage. But if you give me a good 'typical' efficiency percentage, I'll consider it.

I'm tweaking connection strength values now.

[DoctorEvo]

Yeah, it's kN per second. I dorked that up.

As for stock LFT vs. LFE -- something has to be overpowered between the two, because the pairing is certainly unrealistic.

Unrealistic, yes; overpowered, maybe not. Their specific impulse is ridiculous, but their mass fractions are somewhat low...

Well, it's either force other developers to divide by twenty five, or make these tanks completely incompatible with every other engine and engine mod released to date, and thus temporarily (or permanently, if no-one else creates engines using the same values) useless. Most engine addons that come recommended are less efficient than the stock LFE, but are still in its general realm of consumption to thrust values.

Perhaps this is a question best saved for the other addon developers. Changing the .cfgs would be easy, but ultimately it would be their choice whether it's worth the trouble modifying their existing parts to comply or not.

Well, I'd rather model the available energy of the fuel as accurately as possible given my resources -- as I said, energy is a fuel trait, efficiency is an engine trait. Similarly, as I said, most respected developer's engines seem to undershoot the stock engine's efficiency by a decent percentage. But if you give me a good 'typical' efficiency percentage, I'll consider it.

Sure. Gimme a minute and I'll average a bunch of basic (I'm thinking gas-generator-cycle) engines in each propellant. That'd make a good 'generic' baseline.

*edit*

LH2/LOX: 4145 m/s (four engines, gas-generator cycle)

RP-1/LOX: 2775 m/s (four engines, gas-generator cycle)

Hypergol: 2917 m/s (one engine, gas-generator cycle)

Weighted engine type efficiencies:

Pressure-fed: ~90% of gas-generator, ~80% of chart value (one engine)

Gas-generator: 100% of gas-generator, 79-94% of chart value (four engines)

Expander: ~107% of gas-generator, 95-102% of chart value (three engines)

Staged-combustion: ~109% of gas-generator, ~94-100% of chart value (lots of engines)

And overall, it appears that hydrogen achieves closest to theoretical performance most often, frequently exceeding 100% of chart value amongst staged-combustion and expander cycle engines. RP-1 and hypergol propellants never reach the chart value, and their proportionately-lower performance can be seen by the propellant-specific V_e values above.

[nivvydaskrl]

v1.2 up. Underscore name issue resolved. Strength values set to very high values to avoid, uhm, 'random stack disassembly.'

[Jouzu]

It is a shame this mod isn\'t getting more attention, I\'ve modded the standard engine and some engines from Sunday Punch\'s pack to quite realistic levels and using your fuel tanks I am trying to make a 'Triple Launch' rocket (First stage brings two more stages into space, both capable of re-launching themselves again), and It is proving to be exceedingly difficult (a GOOD thing), main problem now seems to be landing the second stage (It consisting of two 3m RP1/LOX tanks, carrying the third stage, a single 2m RP1/LOX tank with boosters).

If you had the time you could release a package with you fueltanks and engines with realistic Isp and (quite importantly) realistic thrust to weight ratios (Perhaps just a modified standard engine).

I, at least are having great fun with this mod, cheers!

[rdfox]

There\'s only two significant problems with this pack so far. First, there\'s no engines out that use consumption figures tuned correctly for them. Secondly, they only come in the same size as the standard tank, with no extended tanks or large-radius tanks. If we could get mod .cfg files to allow us to use NovaSilisko\'s and Sunday Punch\'s tanks and engines with the more realistic Isp. and T:W figures, I think there would be more attention to it. (Even if I then had to completely recalculate my VK Nova family\'s performance... all 230-odd variants...)

[nivvydaskrl]

It is a shame this mod isn\'t getting more attention, I\'ve modded the standard engine and some engines from Sunday Punch\'s pack to quite realistic levels and using your fuel tanks I am trying to make a 'Triple Launch' rocket (First stage brings two more stages into space, both capable of re-launching themselves again), and It is proving to be exceedingly difficult (a GOOD thing), main problem now seems to be landing the second stage (It consisting of two 3m RP1/LOX tanks, carrying the third stage, a single 2m RP1/LOX tank with boosters).

If you had the time you could release a package with you fueltanks and engines with realistic Isp and (quite importantly) realistic thrust to weight ratios (Perhaps just a modified standard engine).

I, at least are having great fun with this mod, cheers!

Thanks a lot! I\'ve recently become pretty busy with grad school related stuff, and the idea of modeling and texturing an engine makes me quail with fear -- in case you hadn\'t noticed, my models and textures on this mod are...unintentionally kerbal-like.

That said, if this mod starts to see more interest from others, I\'d happily get back into it. As for the triple launch thing -- good luck, man. The very idea of landing two of those 3m RP1 tanks makes me shudder. They\'re, like, 40 tons a pop, I\'m usually lucky if I can lift a stack without the force of the required engines ripping the engine mount off. :-[

There\'s only two significant problems with this pack so far. First, there\'s no engines out that use consumption figures tuned correctly for them. Secondly, they only come in the same size as the standard tank, with no extended tanks or large-radius tanks. If we could get mod .cfg files to allow us to use NovaSilisko\'s and Sunday Punch\'s tanks and engines with the more realistic Isp. and T:W figures, I think there would be more attention to it. (Even if I then had to completely recalculate my VK Nova family\'s performance... all 230-odd variants...)

The engine issue is completely valid, but I\'m a noob/crap modeler and engines have a lot of fiddly bits. I approached the Isp side from the fuel tank, as chemical energy is technically a fuel trait. I was hoping others would approach efficiency from the engine side, but no one\'s done so with my spec numbers -- most of the realism buffs like the idea of abandoning the stock parts entirely, instead of putting in a scale factor to allow the stock (and most modded engines) to be automatically balanced to some degree.

The tank size issue is news to me. When I laid out the models and .cfgs in 0.9, the tanks were in proportion to the standard tank, but 1.75m and 3m in diameter, respectively. I don\'t actually have any 1m tanks in this pack -- they\'re so tiny that they\'re really not worth it. They\'d weigh nothing and provide fuel for a handful of seconds.

[rdfox]

The tank size issue is news to me. When I laid out the models and .cfgs in 0.9, the tanks were in proportion to the standard tank, but 1.75m and 3m in diameter, respectively. I don\'t actually have any 1m tanks in this pack -- they\'re so tiny that they\'re really not worth it. They\'d weigh nothing and provide fuel for a handful of seconds.

I\'m an idiot. I hadn\'t actually tried them out yet, so I thought they were actually just mods of the standard tank. :-[ Excuse me whilst I go beat myself with a wet noodle over that one.

[nivvydaskrl]

Ehehehehe. Oh no, they\'re not. They\'re actually very low poly (~50) cylinders with textures made from -- and I\'m not kidding -- a picture of a soup can lid and a picture of house siding. They\'re golden ratio, too: the 1.75m cylinders are 2.83 meters tall, and the 3m cylinders are 4.85 meters tall.

And, as previously stated, the 3 meter tanks are a few dozen tons/kg in weight and have 3000-4000 units of fuel, as appropriate to their size and the density of the fuel modeled.

[mak10z]

is anyone else getting a broken file when this is downloaded? I was looking forward to trying this pack out

[madmat]

Mak10z,,This thread\'s last post was 7mnths ago. There\'s a very good chance it has been left to die. It probably wouldn\'t work with todays version of KSP anyway..Sorry mate

[Tiberion]

[mod]Locked due to Necro-post. If the original author wishes it reopened, contact any moderator with the thread link and we'll do so.[/mod]