[1.8+] Real Fuels

[NathanKell]

Nope, the point of Real Fuels, like most of my mods, is to replace KSP ~magic~ with reality. Making things use magic fuel flow is...kinda counter to the point.

In particular, it's a very *good* thing that RCS fuel doesn't flow through interstages or solid walls in RF. With CrossFeedEnabler, there really isn't a need for magic fuel flow rules, and there's a lot of need for *not* having them.

[Starwaster]

Nope, the point of Real Fuels, like most of my mods, is to replace KSP ~magic~ with reality. Making things use magic fuel flow is...kinda counter to the point.

In particular, it's a very *good* thing that RCS fuel doesn't flow through interstages or solid walls in RF. With CrossFeedEnabler, there really isn't a need for magic fuel flow rules, and there's a lot of need for *not* having them.

I was thinking more along the lines of radially mounted tanks

[NathanKell]

Which have no problems when CrossFeedEnabler is installed, right? I mean, I radially mount all my hydrazine tanks...

[Starwaster]

Which have no problems when CrossFeedEnabler is installed, right? I mean, I radially mount all my hydrazine tanks...

I'll take your word for it; I don't use CrossfeedEnabler.

Seems to much like magic to me

Edited November 15, 2014 by Starwaster
missing smiley...thing... face

[NathanKell]

Meh, I find it less magical if you have the tank itself connected to another tank, it seems reasonable there could be non-magical connection between them, whereas an invisible fuel line that goes through empty space in an interstage...

[Starwaster]

Did a pull request for JRA's Soviet fix. Also inserted a missing smiley face in my post above

[NathanKell]

Pull accepted, with gratuitous smilie.

[Starman4308]

Hey Nathan, I've been writing up a RF config for NecroBones's Modular Rocket Systems*, and I noticed a few oddities with the ion engines/tanks along the way. The small radial xenon tank has a capacity of 200L, and the cylindrical stack tank has a capacity of 70L: empty masses are unchanged from stock. NecroBones also has some xenon tanks, so it would be helpful to know how to scale xenon volumes.

It'll also help with the ion engine config: I'm pretty sure what happens with the PB-ION as-is in Raptor's config is that it appropriately adjusts the xenon consumption, but doesn't touch the ratio of electricity to xenon, and as a result consumes a hideous amount of electricity. I figure that, for now, I'll try to wrangle it to return to stock performance, and let Raptor decide if he wants to nerf it down to something less outrageously unrealistic. Aside from all that, the only goofiness I see is that procedural tanks let you store outrageous amounts of electric charge when using stock electricity units, and there's a simple fix to that: don't use procedural tanks to store electricity.

*Is the standard procedure just to send pull requests on Github when finished?

EDIT: For clarifications, Raptor's config does not presently touch the ion engines: this was me saying "knowing the xenon scaling ratio will help me figure out how to whack the electricity/xenon ratio back in line with whatever Raptor wants".

Edited November 16, 2014 by Starman4308

[Starwaster]

Hey Nathan, I've been writing up a RF config for NecroBones's Modular Rocket Systems*, and I noticed a few oddities with the ion engines/tanks along the way. The small radial xenon tank has a capacity of 200L, and the cylindrical stack tank has a capacity of 70L: empty masses are unchanged from stock. NecroBones also has some xenon tanks, so it would be helpful to know how to scale xenon volumes.

It'll also help with the ion engine config: I'm pretty sure what happens with the PB-ION as-is in Raptor's config is that it appropriately adjusts the xenon consumption, but doesn't touch the ratio of electricity to xenon, and as a result consumes a hideous amount of electricity. I figure that, for now, I'll try to wrangle it to return to stock performance, and let Raptor decide if he wants to nerf it down to something less outrageously unrealistic. Aside from all that, the only goofiness I see is that procedural tanks let you store outrageous amounts of electric charge when using stock electricity units, and there's a simple fix to that: don't use procedural tanks to store electricity.

*Is the standard procedure just to send pull requests on Github when finished?

Raptor has a config that modifies the PBION? That's news to me... and the stock ion engine has always consumed massive amounts of electricity. It's been that way because it's more powerful in terms of thrust than a real ion engine because nobody wants to spend the time to do a realistic ion engine burn which could take weeks or even months and the stock game doesn't allow engines to operate during non-physical warp.

As to electricity storage it's that way for Realism Overhaul which scales electricity generation down. (kilowatts and joules)

[Starman4308]

Raptor has a config that modifies the PBION?

It doesn't: the effects are a result of RF scaling the xenon, and Raptor's config not scaling the electricity/xenon ratio, because he doesn't touch it. I'm just figuring I'll try to tone it back down: it presently consumes the equivalent of 8.2 Gigantor XLs of electricity, due to a calculation which has to do with scaling issues and no deliberate thought on anybody's part.

EDIT: For all I know, 8.2 Gigantors might be 100x less than what real-world ion engines would need for 2 kN of thrust. I'm cognizant, however, of that ion engines really do need to be out-of-line with real-world performance to not make them boring to use (though 4200s Isp is a little crazy).

Edited November 16, 2014 by Starman4308

[NathanKell]

Well, Orbit Manipulator allows thrust on rails (it's now a recommended mod for Realism Overhaul). And yes, Gigantors are 2.74kW panels, so you'd need about 12,000 (not a typo, 2200W = 120 milliNewtons) of them for a 2kN xenon thruster (at 1800s Isp, btw).

Pull requests are great!

(And it's generally better, if you're willing, to actually figure out the volume of the tank, and then apply a volume utilization [we use ~86%] to account for it actually being composed of various pressure vessels, though the 5x rule works decently.)

[Raptor831]

Oh, uh, didn't realize the Xenon density would affect the amount of propellant/fuel used. Though, I probably should have realized that.

Stockalike was intended to just give the stock parts the RF treatment. Thrust scaling with ISP, (more) realistic TWR/mass ratios, real fuels/propellants, etc. But, still playable on stock-scales (realistic mass = false) or the 6.4x scale. Using RF with FAR on stock-sized Kerbin lets you build ridiculously tiny rockets that make orbit!

So, really, for Stockalike the ion engines (or the resistojets and arcjets from other mods) should more or less be scaled to resemble the stock balance with a lean towards realism. Whatever that is...

@Starman4308 If you make up an engine config for the ions, you can submit a pull request at the Stockalike repo. Always could use the help!

[Starman4308]

(And it's generally better, if you're willing, to actually figure out the volume of the tank, and then apply a volume utilization [we use ~86%] to account for it actually being composed of various pressure vessels, though the 5x rule works decently.)

Well then that makes NecroBones's xenon tanks really simple, because they are spheres. The stock xenon tanks are, upon a closer look, mostly correct, although the radial tank seems to be made mostly out of hyperspace (contains 200L; a procedural tank of slightly larger size is 46L). If you're interested, here's the project so far, although I've only got basic functionality going yet.

His fuel cells are probably something which should be incorporated into the engine configs, because they're dependent on what electricity unit is being used (the stock Who-Knows-What-This-Is unit or the RO-style kJ). For the stock unit, I'll probably assume for now that LF/O is RP-1/LOX, and scale the chemical energy of that to LH2/LOX. I'm not sure if other fuel types are used for space vehicles, but LH2/LOX has an advantage in that, with a bit of coding work, it should be possible to store the resulting water, and then see if I can borrow TAC:LS's water splitter to regenerate the fuel.

EDIT: Change to repo location

Edited November 17, 2014 by Starman4308

[NathanKell]

That's weird about the xenon tank, I thought I measured it. Isn't it the same size as the radial monoprop tank?

As for electric charge, while RF doesn't go ahead and nerf solar panels (although their mass:energy ratios are fine or even a bit poor, their size is about 1/2 to 1/4 what it should be), RF *does* enforce a sane total of kJ/liter and kJ/tonne, taking EC as kJ. So it's not just an RO metric.

For the fuel cell, I'd go based on volume rather than energy-density maybe? After all it's a part of a fixed size, not fixed performance.

[Starman4308]

That's weird about the xenon tank, I thought I measured it. Isn't it the same size as the radial monoprop tank?

I suppose maybe it's bigger than it looks, but here's an image for comparison. The procedural tank is, again, 46L. I'm pretty sure the appropriate scaling would be <=40L*; the cylindrical stack tank is 70L, and that very neatly coincides with the stock volumes of 400/700 xenon. In any event, should I kick any further MRS-RF development questions to a separate thread? I wouldn't want to clog up RF or Raptor's thread with things not related to RF or RF-stockalike engines as they currently are.

*On playing with the perspective a bit, the 46L tank is much too wide for valid comparison. I can't quite get the sizes to line up (the xenon tank a bit wider than 0.25m, and a bit longer than 0.5m), but it might be closer to ~25L. Unfortunately, I don't know what tools could be used to measure the volume accurately.

Edited November 16, 2014 by Starman4308

[Raptor831]

If you know the radius/diameter of the cylinder, and the length, you can get an approximate volume. The pill shape is basically two half spheres (i.e. 1 full sphere) with radius r on the ends and a cylinder with radius r and height (total height - 2r).

With your numbers, you'd get:

r = 0.125m

H = .5m

h = H - 2r = .25m

Vs = 4/3*pi*0.125^2 = .00818 m^3 = 8.18 L

Vc = 0.25*pi*0.125^2 = .012 m^3 = 12L

Vt = 20.18 L

Check the math on that too, just in case. Since you were underestimating the dimensions as well, you could fudge that up a bit, or just not take the 86% utilization penalty. The tank is already a pill shape, which I believe is what is assumed is inside all of the tank parts (pill inside a cylinder), so I don't think there'd be much wasted space in there.

If you can figure out the dimensions in-game, you can figure out the volume. Really, the best way is to look at the 3d models themselves, but that isn't always possible if you're not the part author. But eyeballing it in-game with known-dimension parts works pretty well, if it is a bit tedious.

[Starman4308]

Pull request sent! The only decision I made not mentioned in the pull request was the kitchen-sink decoupler, which combines a 1000L RCS tank with a 2.5m decoupler and 2.5-3.75m adapter. That, I gave a base mass of 0.55t (0.4t for the decoupler, 0.15t for structural element).

EDIT: To remove any possible confusion, this did not include the ion engine tweak, which should not be in RF proper, and is not complete yet. It's just the tanks.

Edited November 16, 2014 by Starman4308

[NathanKell]

Thanks, that's most awesome!

Regarding the xenon tank, I thought at one point it was the size of the radial monoprop tanks, but now it is clearly smaller, yes. As for measuring, um, you know you can click on the bar of the size tweakables, yeah, like for fuel? You can get any width and height you like, down to millimeter accuracy.

[Starman4308]

Thanks, that's most awesome!

Regarding the xenon tank, I thought at one point it was the size of the radial monoprop tanks, but now it is clearly smaller, yes. As for measuring, um, you know you can click on the bar of the size tweakables, yeah, like for fuel? You can get any width and height you like, down to millimeter accuracy.

Ah, thanks: I'd forgotten about that, because I never needed such precision before. In any event, my best guess is ~28-30L, by playing with a procedural tank until it looked right.

Doing a bit of research into fuel cells, using the space shuttle fuel cells as my reference. Either I'm missing something, or the Space Shuttle fuel cells were fuel-inefficient, achieving only ~14.5% efficiency, which does not square with modern PEM fuel cells, which get 50-60% efficiency (more for industrial fuel cells which utilize the waste heat). I'm not 100% sure how I want to implement the fuel; I think the best way to go about it would be to use the TAC: Life Support resources and implement a Hydrogen resource; that avoids problems like "how do I create a tank which won't boil off LH2/LOX". I imagine there'd be some way to have in-house H2/O2/water resources over-ridden if TAC:LS is installed?

Fuel cellshttp://spaceflight.nasa.gov/shuttle/reference/shutref/orbiter/eps/pwrplants.html

Space shuttle fuel cells: 14"x15"x40", 255 lbs (0.3556 x 0.381 x 1.016 m, total of 0.1377 m^3 = 137.7 L, mass of 115.67 kg)

Provides 7 kW continuous, 12 kW peak (60.5 W/kg continuous)

Takes "only a few minutes" to flood the fuel cell

At liftoff, fuel cells produce 220 A using 4 lbs/hr LOX and 0.6 lbs/hr LH2, probably ~30 V.

Produced 11 kg/h of water at 7 kW (http://www.braeunig.us/space/specs/orbiter.htm and http://science.howstuffworks.com/space-shuttle.htm/printable)

http://spaceflight.nasa.gov/shuttle/reference/shutref/orbiter/eps/storage.html

The oxygen tanks were double-walled (33.43" interior, 36.7" exterior diameter), with a volume of 11.2 ft^3 and 781 lbs of O2. Dry mass 201 lbs, storage temperature -285F

The H2 tanks were 41.51" interior, 45.5" exterior, storing 21.39 ft^3, for 92 lbs LH2 (216 lbs dry). Initial temperature -420F

Up to five tank pairs/mission

CRC Handbook of Chemistry and Physics lists 141.8 MJ/kg energy capacity of hydrogen, and 286 kJ/mol (roughly square with each other)

Calculations

Produced 25200 kJ/hr

11 kg/h water would be 9.78 kg/hr LOX and 1.22 kg/hr LH2.

Theoretical generation: 173.31 MJ/hr. Approximate efficiency is 14.54%.

Either space shuttle fuel cells were terrible, or I'm missing something. Hybrid cars using PEM cells get ~50-60% efficiency

[Northstar1989]

@NathanKell

So any thought on redesigning the Thermal Fins so they'll play nicer with FAR?

Currently, they're almost entirely square, instead of having a nice triangular shape that would make them much more aerodynamic. Oh, and they're way too short (longer fins that are proportionally more powerful and a bit more cost-effective would be preferable).

EDIT: I'm sick of discussing this in abstract (mainly with other people besides yourself), without actual images. So here are screenshots of the current thermal fins alongside what I would consider a *REASONABLE* sized aerodynamic control fin for a large rocket (the Heavy Control Winglet from NovaPunch2) that doesn't rely on reaction wheel spam, and a small stock canard for reference:

As you can see, the RealFuels Thermal Fin is shorter than both, and MUCH less aerodynamic in shape (triangles are aerodynamic- squares... aren't).

What I would consider a REASONABLE thermal fin for a large rocket would be at least 5-6 times as long (so a good bit longer than the Heavy Control Winglet) and triangular-shaped with a gradual slant, much like the design of the (much larger) radiators for the real-life Jupiter Icy Moon Orbiter proposals:

JIMO was trying to dissipate the heat of a small nuclear reactor, of course- but I would consider the challenge of trying to dissipate the heat of a 5-meter diameter fuel depot (at least 10-12 meters long) holding LH2 long-term in space to be quite substantial as well... At least meriting 3-4 square meters or so of surface area per radiator (JIMO would have had 422 m^2 total radiator area), and that's BEFORE you account for the fact that RealFuels currently does absolutely nothing to scale-down boil-off of larger fuel tanks in accordance with the Square-Cube Law..

TANGENT/EXPLANATION: (The Square-Cube Law dictates that surface area increases with the 2nd power of tank dimensions, whereas volume increases with the 3rd power. Thermal leakage into the tank is proportional to surface area, and thermal leakage directly determines boil-off rate. That's BEFORE you consider the fuel tanks are also pressure-vessels, and thus larger fuel tanks have proportionally thicker walls- which allow much less heat to leak through to their contents.)

The current Thermal Fins each are exactly ONE square-meter rectangles, which is absolutely pathetic (and thus not surprising it currently takes 16-18 of them to stop boil-off for really large fuel depots holding LH2... it would still take 6-8 of them if the Square-Cube law were actually applied...) and ridiculously un-aerodynamic compared to a 5-6 meter long triangular radiator (a 6-meter long right-triangle that extends 1 meter from the rocket body has surface area of 3 m^2)

Regards,

Northstar

P.S. You may find it interesting to note that the Square-Cube law was first described by Galileo. Thus, from a certain perspective, *nowhere* is it more natural than being applied to spacecraft-engineering...

P.P.S. The whole bit about the Square-Cube Law is just a tangent. I'd like to make clear my main focus is on seeing longer, triangular Thermal Fins added so that they'll actually place nice with FAR... I only bothered explaining what I meant about the Square-Cube Law so that other players reading this post would have any idea what I was talking about. Please don't mod-hammer me for bringing that up again...

Edited November 19, 2014 by Northstar1989

[Northstar1989]

Thanks, that's most awesome!

Regarding the xenon tank, I thought at one point it was the size of the radial monoprop tanks, but now it is clearly smaller, yes. As for measuring, um, you know you can click on the bar of the size tweakables, yeah, like for fuel? You can get any width and height you like, down to millimeter accuracy.

@NathanKell

Just a thought on Xenon tanks. You know that in real life, space-grade Xenon tanks are typically highly-pressurized, right? I hope you're considering that in assigning them capacities- they should hold a *LOT* more Xenon than a comparable-sized unpressurized fuel tank.

Of course, since Xenon tanks are so pressurized, the tank mass is also quite high for the size- due to the laws of pressure vessels (tank mass is directly proportional to volume that can be held at 1 atm- double the pressure or double the volume, and the wall thickness has to increase proportionally, creating a strictly linear relationship between capacity and dry mass...)

All that being said, I'd rather have a highly-pressurized Xenon tank than one at low pressure- the mass fraction is the same, and it means I can pack more Delta-V into a tiny probe that fits inside a tiny fairing. If the tanks were less pressurized, the probe would be bulkier, and I'd need a bigger (and therefore more massive and draggier) fairing.

The current Xenon tank capacities don't seem in line with real-world Xenon tanks- the capacities and tank masses should both be higher (the pressures normally used are much higher than stock seems to assume...)

Regards,

Northstar

P.S. On Deep Space One, the Xenon was stored in a supercritical state (this is *highly* pressurized- Xenon has a critical point of 15.9 degrees Celcius, 841 psi; which equates to over 57 atmospheres pressure). Which in fact created several engineering challenges in bringing the pressure down for the ion thrusters:

http://trs-new.jpl.nasa.gov/dspace/bitstream/2014/14210/1/00-0619.pdf

Edited November 22, 2014 by Northstar1989

[lurkoholic]

Is anyone else having problems with the rescaled SRBs (FOWDS & RT-5) or launch clamps? For me, both of the SRBs still look like the full scale 1.25m Squad parts, but with the attach nodes moved to the correct locations. Maybe this is due to the scaling changes (bug fix?) in 0.25?

Even if I turn the launch clamp pumps on in the VAB, when I launch the rocket, they are always set to "Pump Enabled: False" without a toggle to turn them on.

I've even tried a fresh copy of KSP with just Real Fuels v8.1 and the latest Stockalike RF engine configs (KSP 0.25 32-Bit on Win 7 64bit), but both problems persisted.

[NathanKell]

Hah, whoops. Thanks for catching. I use either RO or RftSEngines so I never saw that.

[Deeprafter]

Im using the RSS and Overhaul mod alongr with the suggested ones. I'm having issues with the fuels. some of them just wont show up (like kerosine fe.)<- Well i fixed that by doing something right with the 4th install

This is my 3rd clean install (and i assume to have done it like described for those mods, overhaul installed last).

iI really need help here.

The Problem that remains is, that no engine is throttleble (every engine has min. max. thrust the same value)

Edited November 22, 2014 by Deeprafter

[Deeprafter]

/edit (pls delete)

Edited November 22, 2014 by Deeprafter