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

[Gordon Fecyk]

I'm playing with a custom part that acts as a sort of air compressor for rocket engines. It uses ModuleResourceConverter to turn intake air into oxidizer, turning the stock rockets into air-augmented rockets. This is supposed to be what the Sabre does, and what the Rapier's air breathing mode is supposed to emulate with its higher service ceiling and top speed.

A simple part that converts 60 IA into 12 OX, assuming 21% O², can feed a Rapier's closed cycle mode or an Aerospike on its own with a suitable stock air intake and no onboard oxidizer. On the GPP world Tellumo I could hit Mach 10 at 20 km up, which on Kerbin I think is close to 35 km up with regards to air density. I had part heating turned off at the time, otherwise the test craft would surely have melted. I'm looking at the active cooling mod, but that's a different problem.

Pulling Mach 10 on Kerbin at 35 km up is surely overpowered. How should I adjust this part to make it balanced, gameplay-wise? Right now I'm using the Engine Pre-cooler model, and I'm not currently consuming electric charge. It doesn't have an intake module of its own, so it requires separate intakes, restricting it to oxygen atmospheres. There'd also be three versions of the part, with only the advanced version doing 12 OX/s.

As of 02 NOV 2017, here's a demonstration of the advanced part in action. I also added the parts to a related GitHub repository.

Spoiler

 

 

Edited November 23, 2017 by Gordon Fecyk
Renamed OP title, changed repository link

[Nightside]

Interesting idea @Gordon Fecyk. Do you plan to consider the part to be a compressor and changing the density of intake air (variable by altitude) from a gas and to a liquid (liquid oxygen/oxidizer)?

[JadeOfMaar]

I've considered the gameplay balance problem as well with some super engine or engine-related parts myself. I believe the answer lies in part upgrade configs. Part upgrades appear as individual parts in the tech tree but once you buy them, their feature is added to the associated existing part. Part upgrades are disabled by default in sandbox saves so the air compressor's full power can't be used unless the user knows how to enable them. (I have experience with this now so send me your part config and I'll setup for you then you can study it)

I've also considered enforcing implicit requirements on other mods where appropriate. I've already done this as well. For example, my Fertilizer producing parts for USI life support. If USI LS alone is installed, they have one module to produce from Ore. But if MKS is installed too, they have two modules instead to produce from Gypsum or Minerals but not Ore.

[Gordon Fecyk]

I've added these air compressor parts to the same repository as my jet engine curve extender mods. So @JadeOfMaar you can fork this and try your part upgrade adjustments. I haven't seen upgrades applied to stock parts before, though I know the option was added in KSP 1.2, presumably for modders to allow upgrading parts at a cost of science points. I figured both the compressors and the atmCurve adjustments for the jet engines at higher densities were related somewhat.

Right now there are three parts: A Mk0 part, a Mk1 part and a Mk1 advanced part. These all use the Engine Pre-Cooler model so they look distinct, but I'll want to make a custom texture later so they look at least a little different. They are subject to balance tuning, for instance these will require some electric charge, but right now I want to get a feel for what these parts should do.

@Nightside Yes, that's right. These parts are an attempt to turn the stock rocket engines into air-augmented rockets. While all jet engines have a service ceiling (atmCurve), the stock intakes can still gather intake air at higher altitudes and speeds. Jade and I have been thinking about the problem of getting off the ground at Tellumo, a 'super-kerbin' world in Galileo's planet pack, which has an oxygen atmosphere but is super, super dense and heavy. The jet curve extensions give higher thrust in denser air at the expense of consuming a lot more fuel for low altitudes, and the compressors make oxidizer for higher altitudes so ships don't have to carry as much oxidizer up, which could save a lot of weight on Tellumo's almost 2g gravity.

While this is mostly for Tellumo, these parts and part adjustments could be abused at Kerbin or Gael. The advanced compressor plus a shock cone intake could extend the Rapier's air-breathing operating ceiling by another 10 km at Kerbin, but I'm concerned I've just re-introduced the intake spam problem that was solved by introducing velCurve and atmCurve for jets in the first place. But maybe electric charge requirements or mass adjustments can balance that out.

Edited November 23, 2017 by Gordon Fecyk
New repository link

[Nightside]

Cool. In terms of balance  I think it will come down to deciding if you want to have realistic behavior, i.e. conservation of mass, and then making the parts heavy and/or draggy or use a lot of EC so that there is a balancing downside.

[Gordon Fecyk]

Draggy? That makes some sense since they would process and expel intake air. It doesn't have to be perfect extraction of O² either. Maybe it could produce a by-product of inert gas (which would be N² in reality) that it then has to vent out but in the meantime adds mass. But if I make the conversion too complicated that will take up CPU time.

The part model I chose also isn't very draggy. I can make the stock drag worse, but someone using Ferram Aerospace will override that.

[Gordon Fecyk]

@kraden brought up another interesting Sabre tidbit that applies here, in that it uses an engine pre-cooler that still does compression, though it doesn't liquefy the air like my compressor part theoretically does. It's also coincidental that I'm using the Engine Pre-cooler model for this when this is kind-of what I'm doing here.

How about modding, or cloning, the stock pre-cooler from being an intake into being a converter, but still keeping its thermal properties and its Liquid Fuel reservoir? Then add an electric charge requirement of say 9 EC/s like Jade suggested back in the GPP thread? The resulting advanced part would do this:

• Convert 120 Intake Air and 9 Electric Charge/second into 12 Oxidizer/second (up from 60 IA)
• Have an Oxidizer reservoir of 12 OX but its default amount be zero
• Have a Liquid Fuel reservoir of 40 LF, default amount 40 LF, same as Pre-cooler, or maybe 28 LF to make room for 12 OX
• Have its intake module removed, but still have a reservoir of 5 IA
• Have mass and thermal properties identical to the current Engine Pre-cooler

The basic version of this part would have half of these values but no LF reservoir, and the small (Mk0) version a quarter of these and no LF reservoir.

Does anyone actually use the pre-cooler as an intake? When I do see them on YouTube videos, they usually have another intake attached to the front of them. Often it's shock cone, pre-cooler, and rapier.

[JadeOfMaar]

I will be avoiding trying to influence tankage in the part upgrade configs. I'll only influence the conversion module. Your formula for the re-purposing the part sounds great. In my clones of the Precooler and Engine Nacelle, I treat its intake module with low priority as it's not meant to work at the speeds that shock intakes can. The Precooler is popular in use due to: its high intake rate at lower velocities; its thermal properties; it surface attaches and is the best for a body to which the shock cone and Rapier are attached; it ultimately it holds a little extra IntakeAir, even if it can't collect enough by itself at Mach 5, and that's good for delaying the suffocation or overheat of air-breathing engines.

7 hours ago, Gordon Fecyk said:

The part model I chose also isn't very draggy. I can make the stock drag worse, but someone using Ferram Aerospace will override that.

Don't invest in that. If anything, nerf the Oxidizer output so players always have to stay a bit below full throttle to match consumption with it, or that they get creative with their Oxidizer tanks.

[kraden]

Is there a way to disallow the Rapier its air-breathing mode in lieu of your compressor?

[JadeOfMaar]

52 minutes ago, kraden said:

Is there a way to disallow the Rapier its air-breathing mode in lieu of your compressor?

That's easy to do, but why?

[Gordon Fecyk]

1 hour ago, kraden said:

Is there a way to disallow the Rapier its air-breathing mode in lieu of your compressor?

Try using an Aerospike instead of a Rapier for this, if a pure air-augmented rocket is what you're aiming for. It has better specific impulse at sea level than the Rapier's closed cycle mode, and the same specific impulse in a vacuum. It's also somewhat lighter, and has an alternator that can power the compressor.

While the compressor can free an engine of its atmCurve ceiling, it doesn't free the intakes from their altitude ceiling. I just experimented with an 'air-augmented' Aerospike, and aside from asymmetric thrust problems I found I had to reduce thrust to avoid running out of oxidizer. I was able to creep up to 1800 m/s / Mach 5.5 at 33 km up, though, at Kerbin. Perhaps a bunch of radial ramp intakes would help, which would then re-introduce intake spam as I've feared. The intake spam didn't help any, so I guess I don't have anything to fear from that.

To avoid asymmetric thrust, don't join any tanks, intakes or compressors across the centre of the craft. 

I'm going to stop for today. I'll update the advanced compressor part on the repository before turning in.

Edited November 2, 2017 by Gordon Fecyk
Test results

[kraden]

40 minutes ago, JadeOfMaar said:

That's easy to do, but why?

Because the Rapier is based off of the real-life Sabre engine which uses near-liquid compressed air as oxidizer.  The Sabre is dual-mode only in that it would switch to on-board oxidizer once it could no longer get enough from the air, there is no change in the engine portion itself.  If the compressor part Gordon wants to make is producing Ox, it's more like the real thing and the Rapier could then be more like the real thing by simply drawing Ox from on-board tanks once the compressor can't keep up with the demand. I am also aware that I could simply choose not to use the "air breathing" mode, I was just curious.

[FleshJeb]

@MatterBeam may be able to help with math and logic checking on this. He did a great piece on air scoops.

[Urses]

@Gordon Fecyk and @JadeOfMaar you have a outstanding idea and i stalk you both here and over on GPP.

For the downgrade ability of this op engines and not to add to many new things to CRP. Compressor will produce Ox + Gas(Xenon?) You use than spezialised radiators to vent the liquidised Gas and sink the heat. This way you have a realy strong engine but have to provide tanks and radiators to get rid of additional mass and have not to play with disabled heat gauge? If radiators are tweckable you can provide a cooler reserves for higher speeds.

A new ship generation which will stay a little longer in thicker layers and harvest ressources for speed runs there.

As a thoughthrough option.

Otherwise i can't wait to get ma claws in this masterpieces.

Wish you funny Kabooms 

Urses 

Edit: and as precooler idea make the tanks for 12 OX and 28 Gas that have to be Vented down or translated to other tanks or the Compressor don't work as insisted? You have to work for advantages

Edited November 2, 2017 by Urses

[MatterBeam]

3 hours ago, FleshJeb said:

@MatterBeam may be able to help with math and logic checking on this. He did a great piece on air scoops.

Ready to help.

[Gordon Fecyk]

8 hours ago, MatterBeam said:

Ready to help

Hi. You come strongly recommended.

At the original post explains, I want to make a custom part that converts intake air to oxidizer, so a craft doesn't need to carry as much up for its rocket engines. I'm trying to balance this part's power somewhat as it can turn rockets into air-breathers. I can hit 33 km altitude at Mach 5.5 on a pair of Aerospikes in a fairly light craft, but it's meant for use on super-Kerbin worlds like Tellumo in GPP. Over there, the same craft can hit Mach 10 at 20 km up, where Tellumo has almost 2g and only a 45 km high atmosphere. Tellumo's low orbit velocity is in excess of 4400 m/s, even faster than that of Eve.

So that's pretty powerful, in that it can save a fair bit of onboard oxidizer mass and increase a Rapier's air-breathing ceiling by about 30%, assuming the craft doesn't melt on the way up.

I want to know what compressing Kerbin-like air into oxidizer should cost, in terms of oxidizer per intake air, what other resource would be needed such as electric charge, and how heavy such a part should be. I'm using the Engine Pre-cooler model right now and have exchanged its air intake module for a resource converter module, exchanged some of its LF space for OX space, but otherwise haven't changed its mass or thermal properties. You see some of the specifications already here, and I have a prototype part available in a GitHub repository.

I'm going to make a 'basic' version that uses the Engine Nacelle model, and while I thought about making a Mk0 version, it doesn't seem appropriate anymore.

[JadeOfMaar]

@Gordon Fecyk I can also suggest that you apply the appropriate NEEDS tags so that the converter only works when the need is met (when a certain mod is installed).

Edited November 3, 2017 by JadeOfMaar

[Gordon Fecyk]

I just posted a video demonstrating the part in action on the otherwise-stock Aeris 4A space plane. The original post contains the video in a spoiler and a link to the GitHub repository. Let me know if the behaviour looks right, or needs tuning.

Edited November 3, 2017 by Gordon Fecyk

[MatterBeam]

9 hours ago, Gordon Fecyk said:

Hi. You come strongly recommended.

At the original post explains, I want to make a custom part that converts intake air to oxidizer, so a craft doesn't need to carry as much up for its rocket engines. I'm trying to balance this part's power somewhat as it can turn rockets into air-breathers. I can hit 33 km altitude at Mach 5.5 on a pair of Aerospikes in a fairly light craft, but it's meant for use on super-Kerbin worlds like Tellumo in GPP. Over there, the same craft can hit Mach 10 at 20 km up, where Tellumo has almost 2g and only a 45 km high atmosphere. Tellumo's low orbit velocity is in excess of 4400 m/s, even faster than that of Eve.

So that's pretty powerful, in that it can save a fair bit of onboard oxidizer mass and increase a Rapier's air-breathing ceiling by about 30%, assuming the craft doesn't melt on the way up.

I want to know what compressing Kerbin-like air into oxidizer should cost, in terms of oxidizer per intake air, what other resource would be needed such as electric charge, and how heavy such a part should be. I'm using the Engine Pre-cooler model right now and have exchanged its air intake module for a resource converter module, exchanged some of its LF space for OX space, but otherwise haven't changed its mass or thermal properties. You see some of the specifications already here, and I have a prototype part available in a GitHub repository.

I'm going to make a 'basic' version that uses the Engine Nacelle model, and while I thought about making a Mk0 version, it doesn't seem appropriate anymore.

Compression isn't the important part, it's the cooling. You need to calculate how much power is needed to cool down the air until it becomes liquid and the oxygen separates out. This happens at 90K.

External temperature is 220K on average at the altitudes you'll need this the most (20km) and the pressure is 0.07bars.

SABRE 'cheats' by using liquid hydrogen for a heat sink. The hydrogen heats up while the air cools down - the average temperature is low enough that liquid oxygen condensed inside the engine.
I ran the maths on cooling down the air without liquid hydrogen. You need a heat pump. A heat pump moves heat from a cold source to a hot source, against the temperature gradient. This requires power.

The power is the following: Amount of heat to be removed * Heat pump Coefficient of Performance * Heat pump efficiency + Other processes.
We'll now calculate how much power is needed to fully process the air going into a 1m^2 intake at various speeds, starting at subsonic speeds (275m/s) and moving up to high hypersonic (3400m/s)

The amount of heat to be removed is simply the energy contained in the air you're collecting. The heat capacity of air is roughly 1kJ/kg/K. At 275m/s, a 1m^2 air inlet collects 275m^3 of air per second. This is a quantity of air equivalent to 0.089*275: 24.475kg. It contains 5384kJ of heat energy. At 90K, it'll contain only 2202kJ. The difference is 3181kJ. Therefore, you need to handle 3181kW of heat. 

The coefficient of performance depends on the temperature gradient. It is equal to Cold Temperature/ (Cold- Hot temperature).
The CoP here is 90/(220-90): 0.69
You therefore need to feed the heat pumps at least 1.44W of power for each 1W of heat they move. 

The heat pump's own efficiency depends on the design. 30% is the usual figure, but it might be improved for aerospace applications. So, you need to give the heat pumps roughly 4.81W for each 1W of heat they move.

Putting all these together, we get a figure of 15.3MW for the heat pumps before any other losses (Other Processes) when travelling at 20km altitude at 275m/s. You'll need 2979kW to get 1kg of liquid oxygen per second

At higher velocities, you suck in more air, but the air temperature will be higher. The power requirements rise rapidly. For example, at Mach 3, you'll take in 90.78kg of air per second per m^2 of intake area, and the air temperature will be 700K. This means the CoP drops to 0.1475 and the heat difference rises to 55.38MW, so  you'll need 1251MW of power and you'll get 1kg of liquid oxygen for 65MW.

It gets much worse at Mach 10. You now handle 302.8kg/s of air. The external temperature is at least 1500K. You'll need 22GW and you'll get 1kg of liquid oxygen for 350.6MW

[Gordon Fecyk]

3 hours ago, MatterBeam said:

You'll need 22GW and you'll get 1kg of liquid oxygen for 350.6MW. [At Mach 10]

And there's the reality check. Free energy, it isn't. This is what I wanted to know, thanks so much!

One unit of oxidizer in KSP is 5 kg if the Wiki is to be believed, so to get 30 kg or six units per second I'm looking at 660 GW 10.5 GW for Mach 10. That's a lot of power! By comparison, the Kettle Generating Station in Manitoba produces 1200 MW. This comparison is great, because KSP liquid fuel is not liquid hydrogen (again, if the Wiki is to be believed) and I wouldn't have the Sabre's cheat.

I'll take a look at the ascent video I did yesterday to see what the power requirements would have been for each of those points on the ascent. At the very top of that ceiling, at 1800 m/s or close to Mach 6, I'm not converting a whole lot. Still, there isn't much hope for this being a realistic part unless the electric charge requirements can be made realistic somehow.

Is there a guide as to what one unit of electric charge is supposed to be? Real rechargeable batteries have ratings in Watt-hours (Wh), which gets reduced to Joules, a unit of energy but not of power. KSP EC is actually a unit of energy and not of electric charge.

Edited November 3, 2017 by Gordon Fecyk
More math; I'm an idiot today

[MatterBeam]

25 minutes ago, Gordon Fecyk said:

And there's the reality check. Free energy, it isn't. This is what I wanted to know, thanks so much!

One unit of oxidizer in KSP is 5 kg if the Wiki is to be believed, so to get 30 kg or six units per second I'm looking at 660 GW for Mach 10. That's a lot of power! By comparison, the Kettle Generating Station in Manitoba produces 1200 MW. This comparison is great, because KSP liquid fuel is not liquid hydrogen (again, if the Wiki is to be believed) and I wouldn't have the Sabre's cheat.

I'll take a look at the ascent video I did yesterday to see what the power requirements would have been for each of those points on the ascent. At the very top of that ceiling, at 1800 m/s or close to Mach 6, I'm not converting a whole lot. Still, there isn't much hope for this being a realistic part unless the electric charge requirements can be made realistic somehow.

Is there a guide as to what one unit of electric charge is supposed to be? Real rechargeable batteries have ratings in Watt-hours (Wh), which gets reduced to Joules, a unit of energy but not of power. KSP EC is actually a unit of energy and not of electric charge.

350.6MW/kg --> 10516MW/30kg

The calculations I did were for an intake of 1m^2. If you use a smaller intake, you'll get less air, need less power and only get as much liquid oxygen as you need. 

Everyone just assumes that 1EC/s is 1kW.

[Gordon Fecyk]

Here's some raw data I collected on that ascent. 

Altitude (m)	Speed (m/s)	Mach	Load (% of 6 OX/s)
20611	1508.0	5.045	26.89
20649	1550.0	5.183	27.42
20579	1600.3	5.351	28.68
20890	1650.6	5.521	29.47
21114	1700.2	5.677	27.43
21805	1750.3	5.384	24.66
22693	1800.7	5.986	21.29
24510	1850.0	6.122	15.38
29113	1900.0	6.104	6.46
31686	1909.0	6.012	3.99

The shock cone has 0.75 m² area. The ramp intake has 0.74, and the standard mk1 intake has 0.60. I'm supposed to be processing 10% of intake air into OX, and IA flow rates never exceed 36 IA/sec, or 180 kg/s, so I think I'm producing 3.6 OX/s or 18 kg/s at most. If I can guess based on what you posted, that's a best case of 53 622 kW and a worst case of 6 310 800 kW depending on speed.

So, I'd be discharging no less than 54 000 EC/s to condense the air, plus radiating heat and needing to emit waste gas. That's a nice tall stack of Z-1K batteries. I'll need to scale the conversion rate way back. Even the Aerospike only produces 5 EC/s at full thrust. To use 9 kW only, it seems I'd only be producing 0.00084 OX/s.

Looks like without some alternate means to generate megawatts of power, this isn't going to work realistically. At one point I thought about venting the waste gas through a turbine of some sort, which would make the part even more complex. But again I'm talking about 50 times the capacity of the Kettle Station plant. In a 1.25 m part. 

Don't get me wrong: I really appreciate the reality check. It means I don't have to waste any more time on this part and can do something dumber. 

[MatterBeam]

@Gordon Fecyk

Yeah, the power required is not in the range handled by stock KSP or even modern technology. But, if you have access to oodle of power from a nuclear reactor, you have more options.

What you might be interested in is simple air augmentation. 

BTW, get Atmospheric Autopilot. It makes flying so much more fun and all the cool streamers are using it.

[Gordon Fecyk]

On ‎11‎/‎3‎/‎2017 at 5:55 PM, MatterBeam said:

What you might be interested in is simple air augmentation.

That was my first thought, hence calling the part a compressor. But then how do I separate oxygen or oxidizer out of intake air when it's still a gas? The stock rockets don't burn compressed intake air.

KSP oxidizer might be mystical stuff, maybe a compressed gas and not a cryogenic liquid. Maybe it's raw compressed atmosphere. The game already stretches things by using the equivalent of RP-1 in jets, rockets and a nuclear engine, so who knows?

Maybe I'll come back to this later. Or Jade or someone else has a better idea. Or take a look at something like this.

Edited November 5, 2017 by Gordon Fecyk
Possible lower power gas separator

[Gordon Fecyk]

Apparently, industrial gas separators are a thing, and use some solid material plus higher pressures to adsorb (a new word I hadn't heard of before) nitrogen out from the air, leaving mostly oxygen.

Making higher pressures is simple enough with a compressor that uses electric charge, and then the cost could be in some sort of consumable solid material that does the adsorbing. These devices would need a supply of this material, then it'd be a matter of how much material adsorbs (be careful with the spelling, as this isn't "absorption") how much nitrogen at what pressure and what rate. These things can apparently supply oxidizer to oxy-acetylene torches, so perhaps they could supply it to rockets.