Real ISRU Development Thread

[Guest]

Instead an electrolyser, what about a reversible fuel cell?

Not sure I follow you. Links?

Thanks for the fixed images Red.. but the scale seems too big for our current technology, maybe that was ok for 30 years ago... not sure what year to choose for the mod equipment, that depends on you Regex.

Scale seems to big? Elektron is in use on the ISS right now. As far as I understand it scale is really determined the the desired output rate, with an expected lower minimum on size for a given reaction (no idea what that might be, but there probably is a practical lower-end), with "tech level" only having a fairly minor role to play in miniaturization (you can only convert so fast to begin with) but definitely having a decent effect on efficiency.

As far as the year/"tech level", I would say 1980s~2020s. Pretty reasonable while allowing for near-future and proposal-based projects to be made. A manned mission to Ceres might realistically be prototyped with near-Earth asteroids within the next fifteen years, for instance.

[AngelLestat]

Not sure I follow you. Links?

http://en.wikipedia.org/wiki/Regenerative_fuel_cell

Any fuel cell can work in reverse mode, but those which are not specific designed for this task will have lower efficiency.

Scale seems to big? Elektron is in use on the ISS right now. As far as I understand it scale is really determined the the desired output rate, with an expected lower minimum on size for a given reaction (no idea what that might be, but there probably is a practical lower-end), with "tech level" only having a fairly minor role to play in miniaturization (you can only convert so fast to begin with) but definitely having a decent effect on efficiency.

In the same link you give me it said that the elektron device to split water is the same used in the mir, from 1986, it seems they dint want to test nothing new.

I said 30 years just looking the picture and it has 30 years

We are talking about an alkaline electrolysis device, which are more heavy than PEM and less efficient.

This device was a pain in the ass for the ISS, it had several fails over its lifetime and multiple repairs, this include even fires.

http://en.wikipedia.org/wiki/ISS_ECLSS#Elektron

But they can not remplace it because all the ISS is designed to work with this kind of alkaline electrolysis, you need to respect the pressures, temperatures, potassium hydroxide systems needed, etc.

Also is oversized, you dont need such amount of oxygen production to maintain the max iss crew.

I will see if I can find more specifications on current PEM devices with their weight... The weight is not easy to found in the specifications, but I will try..

As far as the year/"tech level", I would say 1980s~2020s. Pretty reasonable while allowing for near-future and proposal-based projects to be made. A manned mission to Ceres might realistically be prototyped with near-Earth asteroids within the next fifteen years, for instance.

Yeah I am agree.

Edited June 3, 2015 by AngelLestat

[Guest]

http://en.wikipedia.org/wiki/Regenerative_fuel_cell

Any fuel cell can work in reverse mode, but those which are not specific designed for this task will have lower efficiency.

Awesome, I'll look into that.

In the same link you give me it said that the elektron device to split water is the same used in the mir, from 1986, it seems they dint want to test nothing new.

I said 30 years just looking the picture and it has 30 years

We are talking about an alkaline electrolysis device, which are more heavy than PEM and less efficient.

Elektron is an important piece of history for people who make replica craft, why should we not want to provide it just because something else works better? If we went by that philosophy there wouldn't be a place for a large majority of FASA, and other mods, in Realism Overhaul, which spans a very large chunk of Earth's space-faring history. Elektron is also a relatively simple model to make and can stand in for other equipment, such as a Haber-Bosch reactor.

[Dr. Jet]

Elektron is also a relatively simple model to make and can stand in for other equipment, such as a Haber-Bosch reactor.

I wouldn't call it "simple".

On the picture - Elektron system assembled for usage.

Well... if we could somehow exclude all that pipes... then it could be called more or less simple.

BTW, detailed high-poly replicas are nice... ONLY if you have working 64-bit client. We don't have it now and there is no ETA for when it will be available. Maybe some kerbalized look-a-like simplification?

Also, let's discuss the form-factor. 1.25m, 2.5m, 3.75m, surface attached or something modular like Universal Storage? Modular looks quite promising IMO. Central part will be either a tank, passable tube or huge battery, side parts (4 to 8) will be chemical reactor assemblies, radiators or tanks.

Edited June 3, 2015 by Dr. Jet

[Guest]

Well... if we could somehow exclude all that pipes... then it could be called more or less simple.

Yeah, just the general gist of the device, really. "Is it identifiable as Elektron? Awesome." Or just use it as inspiration. Remember that many of these devices are still in a prototype stage and have yet to be packaged for space-based use.

BTW, detailed high-poly replicas are nice...

Not really what I was envisioning.

Also, let's discuss the form-factor. 1.25m, 2.5m, surface attached or something modular like Universal Storage? Modular looks quite promising IMO. Central part will be either a tank, passable tube or huge battery, side parts (4 to 6) will be chemical reactor assemblies or tanks.

Modular would be fine if you want to go that route, US looks fantastic and I pretty much always install it alongside RO. In fact, following the US form factor would be pretty cool as well, but we might want to consider a double-length attachment for it.

After a little thinking I'm pretty much fine with you making the parts any size you think will work, they can always be rescaled and the attach nodes moved through Module Manager, etc... Or rather, if you want to make parts to fit into the stock spec go ahead, I'll just alter the config file to match RO and move all the stock specs to optional Module Manager configs. That is something I'm more than willing to package for the stock people since it doesn't require endless haggling over how much atmospheric oxygen, if any, Laythe has, etc...

If I end up making any models they will follow the RO form factor, 0.5m, 1m, 2m, 3m, 4m, etc...

Edited June 3, 2015 by regex

[Dr. Jet]

US is nice but

1) wedges are too small to be realistic

2) noone open wedges when playing with US except for DMagic ones.

Speaking about modular, I meant something like this:

Internal hexagon with capability to host 1 capsule tank or 2 spherical tanks or passable tube and 6 segment cages to be strapped onto it. Each segment willl hold some type of ISRU unit or additional tank. Whole assembly is 2m tall and 2.4m wide (to fit into fairings). Size and shape are discussable, it's just a draft.

Foreseen trouble: right-clicking on core tank will be really tricky if segment colliders will be made solid.

Edited June 3, 2015 by Dr. Jet

[Guest]

Speaking about modular, I meant something like this:

http://i.imgur.com/5U7bFlM.png

Internal hexagon with capability to host 1 capsule tank or 2 spherical tanks or passable tube and 6 segment cages. Each segment willl hold some type of ISRU unit or additional tank. Whole assembly is 2m tall and 2.4m wide (to fit into fairings).

That is awesome, and was kind of what I was talking about RE: a double-length US core. Those dimensions are also eminently reasonable for rescale and general use. Very, very cool.

[AngelLestat]

Good looking parts and realism is nice.. But do you really need to enter in so much detail for each instrument?

There is something that any mod needs to balance, how good the parts looks, how many parts it adds to the selection menu and how much memory it takes.

Almost all KSP players are tied to the 32b version, so our memory is limited, more if we play with realism overhaul.

There are mods that I will love to use, but the increase game crash rate added makes the game unplayable.

So not sure how do you plan to add so many good looking devices, more taking into account different tech versions added in the future for each device.

It seems too much work

There is not a chance to use other mods models for this matter? Like universal storage mod? (I guess this particular mod is not done for realism overhaul scales, but is just an example)

[FennexFox]

Also, let's discuss the form-factor. 1.25m, 2.5m, 3.75m, surface attached or something modular like Universal Storage?

There's no such form factors in RO, in real rockets. Is DCSS in what form factor? What about CST-100, Orion, Dragon? All the three capsules have different diameter. The rocket makers in real life don't consider the volume of payload in those way.

I think there's no point on discussing the actual form factor right now, especially when we're confusing on kerbal scale and real scale.

Anyway, your works are really cool!

Edited June 4, 2015 by FennexFox

[Guest]

Good looking parts and realism is nice.. But do you really need to enter in so much detail for each instrument?

I'm thinking we should have an individual part for each process, if possible, plus perhaps a few small tanks which can be duplicated with a decal or something. Not sure how Dr. Jet wants to approach that. The frame idea he proposed is outstanding and can work with stock or RO pretty easily.

The parts we'll need, from where I'm sitting right now, are:

• Water Electrolyser
  
• Water Gas Shift Unit
  
• Sabatier Reactor
  
• Methane Pyrolyser
  
• Haber-Bosch Unit
  
• Anthraquinone unit
  
• HTP-Hydrazine Unit
  
• Solid Oxide Electrolyser (this can double for any old fuel cell, actually)
  
• Bosch Reactor
  
• HTP Decomposer
  
• Ammonia Oxidizer
  
• Fisher-Tropsch Unit
  
• Atmospheric Intake (or two)
  
• Ice Drill (or two, already borrowed)
  
• Regolith Sifter (or two, can use a drill I suppose since we already have them)
  
• Cryogenic Compressor (or two)

Now, these parts don't have to match historical models in many cases since there really aren't a lot of historical models to choose from. All those pictures I posted were pretty much just for inspiration or nernie reference since many processes are either simply proposals or in the prototyping phase, lacking compacted space-ready housings. ISRU is still very much in its infancy. Sure, we may want an Elektron-esque model for replicas but they're generally inside a hab unit because they're crew-serviceable and simple vent to the cabin. The same goes for current Sabatier reactors. Obviously we'll want to have the ability to add those units, but having them look exactly like the historical models isn't necessary.

We can also double-up some models if we want to use different textures on them, as I understand things. For instance, the Ammonia Oxidizer, Fisher-Tropsch Unit, and Haber-Bosch Unit may all look similar in some regards so could use the same model with different textures in certain parts. As noted the Solid Oxide Electrolyser is pretty much just a fuel cell so we could use the stock model.

That's another thing, I'm going to be updating RO using the spreadsheet that Raptor831 added, so it will have some Sabatier equipment where appropriate. I've also updated the spreadsheet today to have a "Rate" column so we can adjust input/output rates for different ISRU units, some habs may want historically accurate units, I'll commit that here pretty soon.

Edited June 4, 2015 by regex

[Daishi]

This project sounds awesome, and i'm so glad to hear my work is inspirational to you and Jet, Regex

Subscribed!

[Dr. Jet]

That is awesome, and was kind of what I was talking about RE: a double-length US core.

It's quadruple length or even bigger. US wedges are VEEEERY tiny.

Those dimensions are also eminently reasonable for rescale and general use.

If we want central part to host kerbal-passable tube it should not be smaller then it is now. 1m is the smallest dimension for kerbal to pass through.

[Guest]

If we want central part to host kerbal-passable tube it should not be smaller then it is now. 1m is the smallest dimension for kerbal to pass through.

Oh nice, hadn't even thought of that.

[Dr. Jet]

perhaps a few small tanks which can be duplicated with a decal or something.

Explain please.

[*] Water Gas Shift Unit

Questionably useful... for now. Electrolyser, Sabatier reactor and Pyroliser cover all needed conversions between H2O, CO2, O2 and CH4. CO usefullness is questionable until RF will treat it as legitimate fuel and RISRU reach the point of operating asteroid resources. Then there will be a choice between synthezing CH4 for fuel or saving that precious hydrogen for drinkable water and burn ineffective CO instead.

[*] Methane Pyrolyser

Most pyrolisers are the same. Gas fraction pyroliser is even simpler in construction. It looks a good idea to have a general pyroliser unit for all reactions of that type. Why not to pyrolise Hydrazine when you lack Nitrogen for example?

[*] HTP Decomposer

It has something to do with ISRU? Hydrogen peroxide is decomposing by itself with a small amount of heat or light or you can just dip something silver into it for reaction start. HTP is pretty crappy monopropellant (ISP = 161s) and not very stable oxidizer.

[*] Atmospheric Intake (or two)

Useless without molecular sieves or PSA/VPSA unit or cryogenic separator. BTW we already have some in stock.

I already wrote about molecular sieves and you don't like the idea of cryogenic separator that ouputs liquid products, so.. we have a choice between PSA/VPSA unit with it's own air intake and PSA/VPSA unit without that, relying on MM-modified stock air intakes.

[*] Regolith Sifter (or two, can use a drill I suppose since we already have them)

More realistic is something like this

NASA regolith excavator concept is some subtype of general rotor excavator.

[*] Cryogenic Compressor (or two)

Why two? It's just a powerful fridge, no difference for variety of gases.

Now, these parts don't have to match historical models in many cases since there really aren't a lot of historical models to choose from. All those pictures I posted were pretty much just for inspiration or nernie reference since many processes are either simply proposals or in the prototyping phase, lacking compacted space-ready housings. ISRU is still very much in its infancy. Sure, we may want an Elektron-esque model for replicas but they're generally inside a hab unit because they're crew-serviceable and simple vent to the cabin. The same goes for current Sabatier reactors. Obviously we'll want to have the ability to add those units, but having them look exactly like the historical models isn't necessary.

But I surely want to know which parts of each assembly should be modelled and which may be omitted, how they look and what are their relative dimensions.

We can also double-up some models if we want to use different textures on them, as I understand things. For instance, the Ammonia Oxidizer, Fisher-Tropsch Unit, and Haber-Bosch Unit may all look similar in some regards so could use the same model with different textures in certain parts.

Reasonable.

Edited June 4, 2015 by Dr. Jet

[riocrokite]

hey regex,

are you going to modify stock resources system / regolith to cater for your realistic resources system needs?

[Guest]

Explain please.

Well, I personally would just go with Procedural Tanks, and will make a config for that, but stock players may want some dedicated ISRU tanks. I was thinking maybe a single model (like the large radial monoprop tank I'm using for a placeholder) that would be able to accept a different texture at a certain point, to denote a different tank type.

Questionably useful... for now.

I'll still want to include it, but the model can be doubled up on something.

Most pyrolisers are the same. Gas fraction pyroliser is even simpler in construction. It looks a good idea to have a general pyroliser unit for all reactions of that type.

Sure, that sounds good.

It has something to do with ISRU?

HTP is used in the making of Hydrazine and is also available as a fuel in Real Fuels. Someone may want to create a very, very early moon mission, for instance. vOv Might as well make it available. If you don't want to model it I can figure something out.

Useless without molecular sieves or PSA/VPSA unit or cryogenic separator. BTW we already have some in stock.

Yeah, I'm already using stock models and we can stick with those.

More realistic is something like this

http://www.element14.com/community/servlet/JiveServlet/downloadImage/38-10331-123232/226-170/robotics227x100.jpg

NASA regolith excavator concept

Sure. I'll look into how stock handles resource depletion because it's a very real concern. Realistically ISRU is only going to be able to access very shallow surface resources.

Why two? It's just a powerful fridge, no difference for variety of gases.

vOv Your call.

But I surely want to know which parts of each assembly should be modelled and which may be omitted, how they look and what are their relative dimensions.

That's something we'll have to hash out. Fortunately KSP allows us to resize models in the part configs so I would say make the models at a size you think is reasonable for stock (seeing as how you seem to want this more for stock).

IMO you should concentrate on contributing what you'd like to see. I can fill in the rest with scavenged parts.

- - - Updated - - -

are you going to modify stock resources system / regolith to cater for your realistic resources system needs?

Yeah, why reinvent the wheel? I've already moved over a global atmospheric resource scanner from an old test project, will compile and test it tonight. Tomorrow I'll be working on some latitude-based resource availability code and possibly looking into a "resource bloom" system to handle stuff like Mars' methane blooms.

[Dr. Jet]

Dedicated tanks... other then those are already planned for hex modular system?

Maybe we just use an extra dependency of InterstellarFuelSwitch or FirespitterFuelSwitch to avoid part cluttering? Right click and set what should be stored where...

[Guest]

Dedicated tanks... other then those are already planned for hex modular system?

If those can be made surface attachable (as well as node attach, which I'm assuming you're planning for the hex system?), and I see no reason why they can't, then we probably don't need any more. Shoot, we could even keep the placeholder tanks if we want, but those should more properly be used with Real Fuels tank types or FirespitterFuelSwitch.

Maybe we just use an extra dependency of InterstellarFuelSwitch or FirespitterFuelSwitch to avoid part cluttering? Right click and set what should be stored where...

I would consider that an optional, just like Procedural Tanks and Real Fuels. For stock players InterstellarFuelSwitch or FirespitterFuelSwitch would work best, for RO players we'll stick with Real Fuels, since the mods are not entirely compatible, as I understand it. It's all just Module Manager scripts anyway. I'll start hooking those up this weekend.

E: Or rather, you get to choose one of Real Fuels or FirespitterFuelSwitch/InterstellarFuelSwitch for your install, but either are hard dependencies. That should work fine.

[Guest]

Realized I was getting ahead of myself in the last post. Since I have project time at work today, the plan is:

• Square away asteroid and ice mining, adding resource definitions as needed.
  
• Integrate Real Fuels/Modular Fuel Tanks and FSFuelSwitch.
  
• Add a refrigeration unit (stock part, probably the fuel cell array).
  
• Move the Solid Oxide converter over to the Fuel Cell part.
  
• Finish up and assign parts to the processes that are left over (Anthraquinone Process, HTP-Hydrazine Production, HTP Decomposition)

No idea if I'll get to all that today, but it should provide the basic skeleton of the mod from where we can add custom-made parts and move forward.

As far as the fuel tanks are concerned, I'm thinking we should prefer Real Fuels/Modular Fuel Tanks (you can only have one installed), then shift to FSFuelSwitch if neither is installed, then prefer copying tanks (last resort). This satisfies both stock and RO needs since Firespitter is a pretty popular add-on because of it's usefulness, but not everyone has RF/MFT installed. RF/MFT are incompatible with FSFuelSwitch, as I understand it, so we will prefer those mods over FSFuelSwitch. If no fuel mods are installed you get additional part spam (with very little, if any, memory overhead). All of this can be done through Module Manager scripts.

[Dr. Jet]

Better variant of hex core. Will allow surface mount of legs, drills, etc. along with 6 ISRU modules.

Weight is only 0.49 tons, assuming it's duraluminium.

[FreeThinker]

As far as the fuel tanks are concerned, I'm thinking we should prefer Real Fuels/Modular Fuel Tanks (you can only have one installed), then shift to FSFuelSwitch if neither is installed, then prefer copying tanks (last resort). This satisfies both stock and RO needs since Firespitter is a pretty popular add-on because of it's usefulness, but not everyone has RF/MFT installed. RF/MFT are incompatible with FSFuelSwitch, as I understand it, so we will prefer those mods over FSFuelSwitch. If no fuel mods are installed you get additional part spam (with very little, if any, memory overhead). All of this can be done through Module Manager scripts.

Have you considered using IFS (Interstellar Fuel Switch), it was developed specifically to support IRSU functions.

[Guest]

Have you considered using IFS (Interstellar Fuel Switch), it was developed specifically to support IRSU functions.

I'll add support for that, yeah. So Real ISRU will prefer Real Fuels > MFT > Interstellar Fuel Switch > Firespitter Fuel Switch > extra tank spam. Shouldn't be too hard to do.

[FreeThinker]

I'll add support for that, yeah. So Real ISRU will prefer Real Fuels > MFT > Interstellar Fuel Switch > Firespitter Fuel Switch > extra tank spam. Shouldn't be too hard to do.

Ok, but what about switching in the field, to my understanding Real Fuels and MFT do not support this, they only allow VAB configuration and consider switching resources for ISRU outside their domain.

[Guest]

Ok, but what about switching in the field, to my understanding Real Fuels and MFT do not support this, they only allow VAB configuration and consider switching resources for ISRU outside their domain.

Not our problem. MFT and Real Fuels are not compatible with Firespitter Fuel Switch and, by that logic, Interstellar Fuel Switch. Since the whole point of Real ISRU is to work with Realism Overhaul, which has Real Fuels as a hard dependency, there is nothing more to say on the matter. I'll add support, but Real Fuels will be the primary and preferred mod to handle tanks.

[Dr. Jet]

Draft model of "Elektron". Lowest tank is 100 liter water storage.

Connecting all this crap with tubes will be pain in the ass... and texturing too...