Nertea

[WIP] Nert's Dev Thread - Current: guess I'm working on Fast Fourier Transforms again

Recommended Posts

Cool!

Insulator becomes barely effective if conductivity is decreased 10 times, to 0.001

This is the test with MX-1 which has 2000 heat output (by the way, why so low?) and tweaked insulators around it.

RFd2Ba2.png

When I decreased conductivity further to 0.0001, this contraption reached equilibrium with values 542/370 which is rather good. So I guess something like 0.0003 should do.

I've also decreased thermal mass of the insulator to 1, because bumping it up is a hack, it just postpones the inevitable and makes harder to cool things afterwards. Thermal insulation does not have high mass IRL anyway, on contrary it relies on reduced density. And if I remember correctly, near-perfect heat insulation was available in Apollo era.

Share this post


Link to post
Share on other sites

Also, why nerf reactor's own emissive? It is like insulating it from space which is illogical (or am I missing something?). If 0.7 is a default value for most parts, then why not just leave it there?

I've bumped MX-1 heat production to 4000

With insulator thermal mass 1, conductivity 0.0003, reactor heat 4000, emissive 0.7, equilibrium of my test vessel is 640(reactor)/418(battery next to insulator).

What temperature do you consider nominal for reactor operating at full power?

Share this post


Link to post
Share on other sites

Another test, MX-2 (did not tweak it), 8 XR-500 radiators, insulators tweaked as above. There are also two folded XR-500 radiators on the insulator.

g2hllQa.png

Share this post


Link to post
Share on other sites

I nerfed the reactor's own emission for gameplay reasons, because it's physically large it's radiating an unreasonable amount of heat, which is in term making the radiators less required. Which is no good

Heat outputs are still supposed the same as in NFE 0.3.x, so the MX-1 should actually be 5000kW. This could be subject to change, but basically the old thing was

MX-1 -> 5000 kW

MX-2 -> 4000 kW

MX-L -> 6000 kW

MX-4 -> 1200 kW

M-EXP -> 200 kW

Radiators being:

GR-1 -> 1250 kW

GR-4 -> 300 kW

GR-EXP -> 50 kW

XR-2000 -> 2000kW

XR-500 -> 500kW

So I want to keep the "needed radiator count for full power" similar. What do you think would achieve that in terms of values?

I'll incorporate your notes about the insulator, glad the concept works though.

Share this post


Link to post
Share on other sites

Well, while previous mechanics were essentially arithmetic, now we have to find proper temperature equilibrium. And I do not know how the game calculates effective radiation flux.

I suggest tuning it step by step, then throwing insulators in the mix (stock conductivity essentially turns ships with heat sources into ovens).

First, what should be considered as nominal temperature of the reactor?

Then, tune conformal radiators so that a test vessel consisting of just reactor, radiators and probe core equalized on that value.

Then nerf radiators by some 10% accounting for radiation via other connected parts on actual vessels.

And last but not least there should be some visible stat value for radiators. I.e. how much would they radiate when heated to the nominal reactor temperature (folded, extended)

Share this post


Link to post
Share on other sites

OK, I did tests with current stats (only change from your latest version - I've bumped MX-1 heat to 5000)

Here are approximate results:

M-EXP, 4 GR-EXP

Reactor: 742K

GR-EXP radiates 45

MX-4, 6 GR-4:

Reactor: 847K

GR-4 radiates 182

MX-1, 4 GR-1:

Reactor: 675K

GR-1 radiates 1213

MX-2, 8 XR-500:

Reactor: 838K

XR-500 radiates 430

MX-L, 2 XR-2000, 4 XR-500:

Reactor: 810K

XR-2000 radiates 2000

XR-500 radiates 419

Minor issues:

XR-2000 animation is too fast.

GR-4 edge clips through MX-4 top base when placed at intended spots

GR-1 inner corners clip trhough neighbors when placed on MX-1 intended spots, maybe bevel the corners?

- - - Updated - - -

Good explanation of values here: http://forum.kerbalspaceprogram.com/threads/117930-Guide-New-temperature-rules-for-parts-in-1-0-%281-0-2-updates%29

Share this post


Link to post
Share on other sites

Is the heatmanagment plugin necessary? I don't like the glowing side-effect or the little bars. If something is going to explode, I want it to be a surprise. Can I turn that bit off?

Actually, do the normal radiators work on normal heat? From engines and such?

Edited by Mekan1k

Share this post


Link to post
Share on other sites

I've also did a quick test nerfing all radiators emissive to 0.95 ("almost perfect blackbody"). Reactors equalized at ~1100 with the exception of MX-L which stopped at 930 due to own high surface area.

Share this post


Link to post
Share on other sites
Is the heatmanagment plugin necessary? I don't like the glowing side-effect or the little bars. If something is going to explode, I want it to be a surprise. Can I turn that bit off?

Actually, do the normal radiators work on normal heat? From engines and such?

Sounds like that's a new feature from 1.01. It can be toggled with F10.

Share this post


Link to post
Share on other sites

@psycho, those numbers are eerily close to the desired ones!

Another new version.

vX.4.2

  • Testing release! Use at own risk
  • Fixed a battery attach node
  • Reduced XR-2000 extend speed
  • Added a first pass at radiator emissivity estimation in the VAB (Area*Emissivity*SBConst*MaxTemp^4). This seems like it should be correct, but produces really high values.
  • Reactor thermal emissivity increased to 0.4
  • Added a heat pipe! Transfers heat from the higher temp part to the lower temp part (looks like a fuel pipe, test model)

Edited by Nertea

Share this post


Link to post
Share on other sites

Some observations:

1. Reactors don't need currently any radiators since they cool themselves too efficiently (testing 2 x MX-L) reactors working at 100%, temp stabilises at around 1600 degrees

b3mcbwQ.png

2. Radiators emmissivity estimation if true is too generous IMHO

Biggest MX-L reactor produces 6k heat; biggest radiator XR-2000 has ability to radiate 106k heat when extended so assuming no other parts radiate their heat it is able singlehandedly to disperse heat from 17+ biggest reactors.

3. Did some calculations using plumber's guide to starships spreadsheet.

Using carbon-carbon radiator with aluminium or berylum coolant (that require least radiator area):

To radiate same amount of heat (106 MW) you would need at least 46-48m2 area.

ZJoR2w4.png

In KSP such radiator would have area similar to structure on the left (as opposed to the right):

jbsuMlZ.png

And this assuming best carbon-carbon radiators and average temperature of radiator of 2150 degrees with supply temperature around 2600K and output temp around 1800K.

So radiator would have to be connected to 2600K part either directly or via heat pipe to be most efficient.

With lower temperatures or other coolants used radiator area rises dramatically (which is I guess cooling is the main issue with nuclear engines).

spreadsheet source - it's very useful for tweaking different radiator characteristics and checking weight/temp and area. this spreadsheet is simply golden:)

more stuff:

heat pipes

radiators

Share this post


Link to post
Share on other sites

Thanks dude. I really don't have time to iteratively test these hugely so every bit helps!

X.4.6

Some major changes here.

  • Minor plugin fixes
  • Reactor heat production increased x2 - I think this was needed. they heat up so slowly you could do your engine burn and turn them off before problems started occurring. This might even go up more...
  • Improvements to VAB readout of reactors + radiators (now shows radiation at 75% of Tmax, accounts for radiator area)
  • Reduced temperature tolerance of reactors (1600) and reactor radiators (1800) - increases sensitivity to temp a bit
  • Universal radiators have MaxTemp of 1600 - reactor radiators are more effective on reactors
  • Reduced reactor emissivity to 0.05, increased conductivity to 0.15, reduced thermalMassModifier to 0.15 - artificial, but ensures that radiators are more required
  • Decreased emissives of radiators to 0.75 (closed) and 0.95 (deployed), reduced thermalMassModifier to 1.0 - these numbers seem to be working well

Share this post


Link to post
Share on other sites

Dev build of NFP is also ready.

NFP vX.4.0

  • Test Release, use at own risk!
  • Test release does not include stock ion nerf or LV-N configs atm
  • Tech tree locations + CTT patch not functioning
  • KSP 1.02
  • All textures converted to DDS
  • New version of KSPAPIExtensions included
  • CrossfeedEnabler is no longer packaged
  • If installed, CFE will be supported (LH2 radial tanks)
  • Radial tanks are now surface attachable for fuel lines
  • Community Resource Pack is now bundled and required
  • Fixed stack nodes and thrust location on HI-SNAP 2.5m engine
  • Fixed offset thrust on VF-200 (not VF-10K)
  • New Liquid Hydrogen tank parts (all previous tanks obsolete)
  • Huge 10m tank
  • Three 3.75m tanks in identical sizes to Kerbodyne tanks
  • Three 2.5m tanks in identical sizes to Jumbo-64, X200-32, X200-16 tanks
  • Two 1.25m tanks in identical sizes to FL-T800 and FL-T400 tanks
  • Three radial tanks, one for each of 1.25m, 2.5m, 3.75m
  • New Argon tank parts (all previous tanks obsolete)
  • Three 2.5m tanks in identical sizes to Jumbo-64, X200-32, X200-16 tanks
  • Three 1.25m tanks in identical sizes to FL-T800, FL-T400, FL-T200 tanks
  • Three radial tanks, one for each of 0.625m, 1.25m, 2.5m
  • New inline Xenon tank parts (old ones obsolete)
  • Three 2.5m tanks in identical sizes to Jumbo-64, X200-32, X200-16 tanks
  • Three 1.25m tanks in identical sizes to FL-T800, FL-T400, FL-T200 tanks
  • Rebalanced fuel capacity of non-replaced tanks (XE-100R)
  • Improved MPDT FX
  • ElectricEngineThrustLimiter is no longer required, removed from all relevant engines

Share this post


Link to post
Share on other sites

For the reactors, would it be an idea to base the electricity production of the reactor on the heat flow out of the reactor? That's how it works IRL and it provides a much deeper link between temperature management and electricity production.

Share this post


Link to post
Share on other sites

thanks Nertea;

quick dirty test of biggest reactor + 2 biggest radiators:

********** 1st version - reactor is surrounded by 2 thermal shields and is only using itself and 2 radiators to release heat. 100% power results in critical heat and boom, stability is achieved at about 95% reactor power (with temp almost 1500F on reactor and heat sinks). Which IMHO is quite nice balance ******************

TTgZwOs.png

***** 2nd version - no 2nd thermal shield so heat is leaking directly to LH2 tank next to reactor. And tank acts as a huge radiator - 100% reactor power no problem, temperature stabilises at around 1100-1200F. *******************

Uj3NwZX.png

No things get really weird (and interesting):

****** 3rd version, no thermal shield, no heat sinks, reactor itself is surrounded by tanks from both sides. 100% reactor power - temp stabilises at 1200F on whole ship. **************

LPxadWP.png

It seems that cryogenic tanks are much better high-temperature heat sinks (operating at 1200F hehe) which is unrealistic imho. Cryogenic insulation should prevent stock and those tanks (especially cryogenic) from 1st achieving high temperature, 2nd radiating anything in or out.

Similarly I tried 2 biggest reactors and 3 tanks (without radiators), no problem again, 2 x 100% reactors and radiation from tanks stabilises everything at 1350F.

Another extreme test, 2 argon tanks cooling 2 biggest reactors (temp stabilise at 1480F on 100% and 90% power reactor settings):

0FKtjw2.png

edit: Additional info: changing percentage of fuel in tanks even to empty doesn't change much.

The main problem I see is with all tanks (which are usually cryogenic) that 1st they could withstand such large internal temperatures (higher than reactor and heat sinks), 2nd they could radiate soo much heat being insulated from outside (and inside). Right now probably any fuel tank (stock included) is simply superior in temperature radiation and temp operating ranges compared to heat radiators hehe:)

I guess there are couple solutions for that:

1st option - reduce heat dissipation dramatically; soft rebalance - ship will need radiators to dissipate heat; tanks still can withstand quite hard atmo reentry and closeness to kerbol

2nd option - reduce max temperature dramatically; you'll need thermal barrier part otherwise they'll go boom; cannot perform reentry with those tanks, tanks dissipation is also reduced because max temp is smaller, tanks cannot be used close to kerbol (or need some kind of shield). Now you also need to check tanks against engines temps (even if they operate at lower temperatures)

3rd option - reduce both heat dissipation and max temperature - so you need thermal barrier and cannot perform reentry and close kerbol orbits. Plus you might possibly need some small radiator for multiple tanks if they acquire heat (which they cannot dissipate themselves) during long journey? You'll need thermal barrier between engine and tanks as well (or just cluster engine to reactor and then use thermal barrier just like in 'real' sci-fi designs:)

- - - Updated - - -

For the reactors, would it be an idea to base the electricity production of the reactor on the heat flow out of the reactor? That's how it works IRL and it provides a much deeper link between temperature management and electricity production.

that would require introducing electric generators (similar to KSPI) which would require new parts and balancing. now as I understand things are simplified by the fact that reactors have EC generators embedded in them.

Edited by riocrokite

Share this post


Link to post
Share on other sites

Yeah, Squad set fuel tank emissivity at 0.8! So very high :P. I set those LH2 tanks at lower, perhaps 0.4 or 0.5, I forget which exactly. So even at those levels, the tanks are really good radiators. Hm.

I don't really know how to get around this without patching everything's emissivity, and I don't really want to do that... I guess I could crank reactor heat again, go to x3 or so. That would probably push them to higher thresholds.

edit- Oh, and temperature is in Kelvins afaik ;)

Edited by Nertea

Share this post


Link to post
Share on other sites

...I wanted to write that if you drop emmisivity of radiators to 'passive' 0.95 and scale them up 2 times (4x area), it would do without venturing in the 'active cooler' emmisivity area... but I had to sleep.

So, we consider 1500K to be reactor's nominal temp?

Maybe instead of cranking up reactor heat, gradually cut electricity production when reactor heats above nominal?

- - - Updated - - -

I think XR-500 needs to be scaled up to 1/4 area of XR-2000

Share this post


Link to post
Share on other sites
Yeah, Squad set fuel tank emissivity at 0.8! So very high :P. I set those LH2 tanks at lower, perhaps 0.4 or 0.5, I forget which exactly. So even at those levels, the tanks are really good radiators. Hm.

I don't really know how to get around this without patching everything's emissivity, and I don't really want to do that... I guess I could crank reactor heat again, go to x3 or so. That would probably push them to higher thresholds.

edit- Oh, and temperature is in Kelvins afaik ;)

yah stock tanks emmissivity is really dissappointing; however since NFT props have superior dV and can't use stock tanks for them so maybe tweaking NFT tank emmisivity to much lower level wouldn't be a problem?

Share this post


Link to post
Share on other sites

that would require introducing electric generators (similar to KSPI) which would require new parts and balancing. now as I understand things are simplified by the fact that reactors have EC generators embedded in them.

Is it possible to just mix reactor's negative conductive flow and radiative flow values into energy production formula without adding new parts? When reactor is in equilibrium, it generates X heat and X heat is flowing away via conduction and radiation. So generate power as heat_outflow*reactor_efficiency (12000*0.25 for M-XL when in equilibrium)

- - - Updated - - -

You could also mix nominal temperature in. So there is a motivation to really cool the reactor with efficient cooling systems, not just leave it barely hanging below explosion point between two tanks.

I.e. generate power as heat_outflow*reactor_efficiency*(nominal_temperature/current_temperature). Does that make any sense?

- - - Updated - - -

There also was core integrity in previous versions. Maybe reintroduce that? Part explosion is an extreme condition, there could be gradually less bad bad things.

500K - nominal temperature, 500-1000K gradually reduced power output, 1000-1600K - further reduced power output, permanent core damage.

Share this post


Link to post
Share on other sites

Alright, I did this and it's pretty cool. New field "NominalTemperature" for reactors, if exceeded introduces inefficiency proportional to the amount it is exceeded by (100% losses at maxTemp). That Temp bar on the staging area now shows this (ie, keep it empty). In addition, reactors now have spinup times (about 200s to full power right now).

Working on reintroducing core damage now (if nominal is exceeded by more than 50 or 75%, take core damage, repairable with level 5 engie).

I'll have this version up later today or tomorrow.

Share this post


Link to post
Share on other sites
Alright, I did this and it's pretty cool. New field "NominalTemperature" for reactors, if exceeded introduces inefficiency proportional to the amount it is exceeded by (100% losses at maxTemp). That Temp bar on the staging area now shows this (ie, keep it empty). In addition, reactors now have spinup times (about 200s to full power right now).

Working on reintroducing core damage now (if nominal is exceeded by more than 50 or 75%, take core damage, repairable with level 5 engie).

I'll have this version up later today or tomorrow.

random suggestions - take if you like them:

- higher-level engineers onboard should mean you can run a reactor closer to redline with less risk of damage (or, alternatively, you have more precise knowledge of where the redline is for this reactor at this moment..)

- bad core damage is only partially reversible "in flight" - you might have to limp home on 25% or 50% power, but that's better than being stranded out beyond Eeloo...

Share this post


Link to post
Share on other sites

As a former U.S. Navy nuclear operator I have to say that the visuals on this:

core damage, repairable with level 5 engie

are horrifying. Bill floating next to the reactor with a leaking fuel element in one hand and a roll of duct tape in the other, with a high-range detector screeching in the background. Very Kerbal, I love it. :D

Share this post


Link to post
Share on other sites
As a former U.S. Navy nuclear operator I have to say that the visuals on this:

are horrifying. Bill floating next to the reactor with a leaking fuel element in one hand and a roll of duct tape in the other, with a high-range detector screeching in the background. Very Kerbal, I love it. :D

Oh yes, exactly ;)

random suggestions - take if you like them:

- higher-level engineers onboard should mean you can run a reactor closer to redline with less risk of damage (or, alternatively, you have more precise knowledge of where the redline is for this reactor at this moment..)

- bad core damage is only partially reversible "in flight" - you might have to limp home on 25% or 50% power, but that's better than being stranded out beyond Eeloo...

Mmm, the first, maybe a later version. The second one is implemented - you can only repair a reactor to 75% of max core integrity in the field.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.