[1.3.x] [WIP] Stock Jet Adjustments [23 NOV 2017]

[JadeOfMaar]

@Gordon Fecyk You're onto something with those options there.

• Have the compressor parts carry a supply of Zeolite (any adsorbent stone) and have the converter consume it, IntakeAir and ElectricCharge to produce Oxidizer (at whatever efficiency and with whatever output rate you want). The Zeolite will also have the non-transfer, no-refill properties of SolidFuel. Also, (more for Urses than for you) don't worry about CRP. Some high qualifications need to be met before anyone can think of having resources added to it.
• Perhaps ease back on the fidelity to realism and real proportions with the current concept for the converter. LiquidFuel, Oxidizer and IntakeAir have the same densities, and as you mentioned, engines that shouldn't, all use LF, so go ahead and take some liberties to tune your part a little more in favor of gameplay over realism. With this in mind, consider that the actual SABRE engine should translate into 5m size or greater in KSP? The Skylon spaceplane body is simply gigantic to be able to carry enough LH2 that it can spare a lot to generate Oxidizer in-flight.
• New (crazy) idea. Use ModuleEngines and an alternator for the converter, then you can have the output controlled by the various engine curves. Using a non-engine part for it shouldn't cause any actual issue. It just won't produce any real thrust (which we don't need anyway) because non-engine parts don't have a thrust transform.

[Foozle]

As I understand it, on an industrial scale the extraction of O2 from air is usually done by fractional distillation, that is, by cooling air until the O2 and nitrogen fall out at their own liquefaction points.  So cooling, lots of of it.  With radiators to carry away the heat.

Edited November 5, 2017 by Foozle

[Gordon Fecyk]

20 hours ago, JadeOfMaar said:

Perhaps ease back on the fidelity to realism and real proportions with the current concept for the converter.

At least separating out oxidizer as a gas seems like it would take a lot less power.

OK, I invented a generic mineral resource "Zeolite" and changed the advanced compressor to use this material as a sieve to separate out gaseous oxidizer from intake air. The conversion rate is the same, so this will still need some tuning, and I have to figure out a respectable rate of zeolite depletion. I think I have it configured like Ablator, which isn't transferrable and can only be loaded onto the part in the VAB or SPH.

The part changes are on the Jet Engine Curve Extender repository.

19 hours ago, Foozle said:

As I understand it, the extraction of O2 from air is usually done by factional distillation, that is, by cooling air until the O2 and nitrogen fall out at their own

That was my original thought, until it was demonstrated that cooling the surrounding intake air would be very power-expensive. To get the rates I had, I'd need the same power output as a hydroelectric dam.

So now I'm investigating separating the O₂ out using essentially a molecular sieve. Apparently this is a real thing that is used small scale for things like home oxygen supplies, but has been scaled up to industrial levels.

I spent about 90 minutes on the idea and found some decent papers on the topic, but the math is still above me. I'll play with the part changes later on, but I just posted a theoretical part to the repository I've been working on.

[Update 05 NOV] This is a little better at least. An industrial oxygen concentrator can produce 130 cubic feet/hour of O₂, which I translated to about 1.02 L/second. O₂ has a pretty light density at sea level of 0.001429 Kg/L. That example concentrator is about 50 W, and there are apparently 45 W devices out there. Back-of-the-napkin calculation is about 170 kW/KSP mass unit of O₂, which is under a tenth of the power needed to cool the air into a liquid that we came up with last time.

That's still a buttload of power. Maybe we can make some assumptions about KSP engine alternators, or have a turbine in this part or something to generate internal electricity while venting waste air, making the power usage maybe half this. 90 kW or 90 EC/s might be reasonable, but that's still per KSP mass unit. 

Edited November 5, 2017 by Gordon Fecyk
More math

[Gordon Fecyk]

20 hours ago, JadeOfMaar said:

Use ModuleEngines and an alternator for the converter, then you can have the output controlled by the various engine curves.

I'm a bit confused as to how I'd use this module in this non-engine part. Maybe to make that turbine? You also suggested using ModuleAlternator.

What about something like this? Assuming the part produces waste gas as well as oxidizer, I could consume waste gas to produce electric charge, maybe produce extra heat as a cost.

MODULE
{
	name = ModuleEngines
	thrustVectorTransformName = thrustTransform
	exhaustDamage = False
	ignitionThreshold = 0.0
	minThrust = 0
	maxThrust = 0
	heatProduction = [something]
	// fxOffset = 0, 0, 0.4
	EngineType = LiquidFuel // Not sure what engine type to use, could use nuclear even
	//exhaustDamageDistanceOffset = 1.75
	PROPELLANT
	{
		name = WasteGas
		ratio = 108 // Assuming we use 12 of 120 to make oxidizer
		DrawGauge = True
	}
	atmosphereCurve
	{
    	// Make up a real atmosphereCurve here
	}
}
MODULE
{
	name = ModuleAlternator
	RESOURCE
	{
		name = ElectricCharge
		rate =	// some percentage of 170 EC/s depending on version of the part
           		// so the rest needs to be supplied by the craft
	}
}

 

[Urses]

And if you try the way of chemichal O2 generation as example NaClO3? If you get atmospherical O2 generate the NaClO3(a salt) you only have to add FE and the outcome is NaCl, FeO and pure O2. This way your Processor use Intake Air, Chemikals and Ore and produces Pure O2? If we use the ingame terminology. This procces is used for O2--masks in modern airplanes.

You only need Ec for heating up the mass but need tankage for the materials.

Additionaly the chemicals works as Katalysator only ore will be discarded.

Edited November 6, 2017 by Urses

[Gordon Fecyk]

24 minutes ago, Urses said:

And if you try the way of chemichal O2 generation as example NaClO3? If you get atmospherical O2 generate the NaClO3(a salt) you only have to add FE and the outcome is NaCl, FeO and pure O2.

Hadn't thought of that, just need to make sure the mass of the materials we take up there doesn't cost more than the benefit of generated oxidizer. It'd be an interesting secondary use for Ore. As for heat, at speeds above Mach 3 we can get plenty of heat from friction. The intake air might be quite warm coming in.

Reminds me of the really old ChemOx respirators we used to play with in shipboard fire drills. 

[JadeOfMaar]

1 hour ago, Gordon Fecyk said:

I'm a bit confused as to how I'd use this module in this non-engine part. Maybe to make that turbine? You also suggested using ModuleAlternator.

What about something like this? Assuming the part produces waste gas as well as oxidizer, I could consume waste gas to produce electric charge, maybe produce extra heat as a cost.

 

Give it at least 1 for maxThrust but remove thrustVectorTransformName (not needed at all). The engine will need at the very least an atmosphereCurve {} for its Isp (to help you control the "propellant" consumption, then any other curve to further influence its performance.

[Gordon Fecyk]

While I haven't introduced anything using ModuleEngines, I think I have a more balanced set of parts. The v0.1.1 release includes the zeolite resource and increases the electric charge usage to about half of what a modern industrial oxygen concentrator would use, assuming one EC/s is one kilowatt. You'll need roughly 1000 EC, or one Z-1K battery, and two to three units of zeolite to power one compressor's use on an ascent between 20 and 30 km, at least at Kerbin. Two compressors will supply a single Rapier engine with resources to spare, but if you can gather enough intake air one compressor per Rapier could possibly do it.

There's a resource flow problem with the zeolite resource. For some reason two or more compressors will share zeolite mineral and one part will drain before the others. I tried turning off resource flow functionality but FlowMode = NO_FLOW doesn't seem to have an effect.

Oxidizer concentration via adsorption and compression seems simpler to model than ChemOx, and seems a lot less power-hungry than condensation by cooling.

[MatterBeam]

Why not something like this?:

MODULE
	{
		 name = ModuleResourceConverter
		 ConverterName = Oxygen
		 StartActionName = Adsorb Oxygen
		 StopActionName = Stop
		 INPUT_RESOURCE
		 {
			ResourceName = IntakeAir
			Ratio = 1
			FlowMode = STAGE_PRIORITY_FLOW
  		 }
		 INPUT_RESOURCE
		 {
			ResourceName = Zeolite
			Ratio = 0.0001
		 }
		 OUTPUT_RESOURCE
		 {
			ResourceName = Oxidizer
			Ratio = 0.21
			DumpExcess = true
			FlowMode = STAGE_PRIORITY_FLOW
		 }
	}

Simple.

Edited November 7, 2017 by MatterBeam

[Gordon Fecyk]

8 minutes ago, MatterBeam said:

Why not something like this?

[snip otherwise good example]

I found a straight one to 0.2 conversion rate was taken literally, and 0.2 OX/s is way too low to sustain any stock rocket engine except maybe the Ant. I had to multiply the conversion rate to something that could sustain an engine.

And no electric charge? Or could this work without an electric compressor pump as long as the dynamic pressure was high enough? Adsorption apparently depends on strong pressures to encourage adhesion to whatever mineral was used to trap nitrogen; they seem to use a cycle of high pressure to capture N₂, then evacuation or collection of O₂, then low pressure to release N₂ and vent it out.

As for the mineral use rate, that's a wild guess. The industrial concentrators do require regular replenishment or replacement of zeolite, but it's in the order of every few months. I figured this part was pushing air so hard through it that it would erode away or even ablate away with use.

[MatterBeam]

6 hours ago, Gordon Fecyk said:

[snip otherwise good example]

I found a straight one to 0.2 conversion rate was taken literally, and 0.2 OX/s is way too low to sustain any stock rocket engine except maybe the Ant. I had to multiply the conversion rate to something that could sustain an engine.

And no electric charge? Or could this work without an electric compressor pump as long as the dynamic pressure was high enough? Adsorption apparently depends on strong pressures to encourage adhesion to whatever mineral was used to trap nitrogen; they seem to use a cycle of high pressure to capture N₂, then evacuation or collection of O₂, then low pressure to release N₂ and vent it out.

As for the mineral use rate, that's a wild guess. The industrial concentrators do require regular replenishment or replacement of zeolite, but it's in the order of every few months. I figured this part was pushing air so hard through it that it would erode away or even ablate away with use.

The numbers are just for illustrative purposes 

I wanted to present an alternative to whatever ModuleEngines you were discussing beforehand. ModuleResourceConverter is simple and convenient.

[Gordon Fecyk]

20 minutes ago, MatterBeam said:

I wanted to present an alternative to whatever ModuleEngines you were discussing beforehand. ModuleResourceConverter is simple and convenient.

ModuleResourceConverter is what I'm using now. It isn't impacted by air speed and altitude though, which is why Jade suggested adding ModuleEngines to have it be affected by these somehow.

However, I'm finding the stock intakes are affected enough by air speed and altitude that the conversion rate is quite dependent on them. That seems to produce the effect I'm looking for.

Are these ready to be released as a beta for others? I'm thinking of permanently combining these two things, jet engine curve extensions and this compressor, into a general jet engine enhancement package. Not sure what to call it though.

[MatterBeam]

9 hours ago, Gordon Fecyk said:

ModuleResourceConverter is what I'm using now. It isn't impacted by air speed and altitude though, which is why Jade suggested adding ModuleEngines to have it be affected by these somehow.

However, I'm finding the stock intakes are affected enough by air speed and altitude that the conversion rate is quite dependent on them. That seems to produce the effect I'm looking for.

Are these ready to be released as a beta for others? I'm thinking of permanently combining these two things, jet engine curve extensions and this compressor, into a general jet engine enhancement package. Not sure what to call it though.

If you make the resource converter super-fast, it will consume all of the available intake air. The intake air, as you said, will depend on the intake's performance, so it is sufficient to handle the varying flight conditions.

[Gordon Fecyk]

It's been a while since I looked at this. Have the interested folks such as @MatterBeam, @JadeOfMaar had a chance to evaluate the current iteration of these parts? The original concern I had, that of intake spam, is addressed by the drag being significantly greater with the more added intakes. There is a crossfeed problem still, where zeolite gets consumed on one part at a time and I'm not sure how to prevent that.

I'd like to put together a release for these compressor parts and for the jet engine curve extensions as a general atmospheric engine extension. 

[MatterBeam]

6 minutes ago, Gordon Fecyk said:

It's been a while since I looked at this. Have the interested folks such as @MatterBeam, @JadeOfMaar had a chance to evaluate the current iteration of these parts? The original concern I had, that of intake spam, is addressed by the drag being significantly greater with the more added intakes. There is a crossfeed problem still, where zeolite gets consumed on one part at a time and I'm not sure how to prevent that.

I'd like to put together a release for these compressor parts and for the jet engine curve extensions as a general atmospheric engine extension. 

Sorry, the part balancing and mod implementation is up to you!

[Gordon Fecyk]

1 minute ago, MatterBeam said:

Sorry, the part balancing and mod implementation is up to you!

OK I'll try a simpler question: Is this set of parts interesting enough that you'd add them to your 'normal' modded KSP installation?

[JadeOfMaar]

On 11/20/2017 at 9:53 PM, Gordon Fecyk said:

OK I'll try a simpler question: Is this set of parts interesting enough that you'd add them to your 'normal' modded KSP installation?

I haven't tested your iteration yet which includes the Zeolite, and yes, a part like this would have a place in my 'normal' modded install(s).  Along with the rest of Jet Engine Curve Extender as I inch ever closer to getting to do playthrough missions on Catullus and Tellumo.

[Gordon Fecyk]

I posted a new release of this part pack / part tweak. This release includes the latest adjustments to the mineral and electric charge consumption needed to produce oxidizer.

If you use this, delete the old JetEngineCurveExtender folder from your installation as well.

[0something0]

Going back to cooling the oxidizer, how about using Liquidfuel to do so instead of atrempting to cool it down using raw power?

[Gordon Fecyk]

1 hour ago, 0something0 said:

Going back to cooling the oxidizer, how about using Liquidfuel to do so instead of atrempting to cool it down using raw power?

The KSP Wiki suggests Liquid Fuel is a variant of RP-1, not liquid hydrogen, and this isn't a cryogenic fuel... at least I don't believe it is. So I wouldn't have the real world SABRE's cheat to condense oxidizer out of the air.

Perhaps variants of these parts could be made using the Real Fuels add-on, or the electric power could be reduced if liquid hydrogen was used instead of liquid fuel.

[0something0]

12 hours ago, Gordon Fecyk said:

The KSP Wiki suggests Liquid Fuel is a variant of RP-1, not liquid hydrogen, and this isn't a cryogenic fuel... at least I don't believe it is. So I wouldn't have the real world SABRE's cheat to condense oxidizer out of the air.

Perhaps variants of these parts could be made using the Real Fuels add-on, or the electric power could be reduced if liquid hydrogen was used instead of liquid fuel.

To be fair, stock KSP is supposed to be "realistic enough" and liquidfuel is already used for NERV.