[WIP] Nert's Dev Thread - Current: various updates

[Psycho_zs]

I agree, it is a great discussion.

Nertea, knowing your work, whatever you end up with will be top class.

I just hope that radiators would remain good conductive passive emitters (0.95/0.95), at least when attached to non-reactor parts.

And let's hope that stock heat mechanics will mature with next KSP versions.

[void_error]

I agree, it is a great discussion.

Nertea, knowing your work, whatever you end up with will be top class.

Agreed. My view is to keep the reactor heat management as simple as possible without compromising functionality. These two NF packs are going to be even more awesome then the 0.90 ones. Keep up the good work!

[johnsonwax]

As well, I was a bit concerned about the direction of the discussion. Unless you're going to go the whole FAR route and model heat dissipation within the VAB, then the parts need to be predictable in how they'll behave just based on the usual description. That could be as simple as requiring 4 R10 radiators or 2 R20 radiators or 1 R40 radiator on an N40 reactor, but if we're trying to stand on a knife-edge on reactor temp between efficiency and self-destruction, and our only mechanism to work that is to repeatedly launch ships and test in space, that's going to become unfun fairly quick.

They can be realistic, but tuned in such a way that they can snap together pretty straightforwardly heat-wise. That is, I think you want a modular, discrete set of reactor/radiator combinations rather than an infinite continuum that requires a lot of trial/error to get working properly.

[Starbuckminsterfullerton]

Ok, see there's this problem now. I can't really just keep iterating and testing and redoing another system and testing again. This isn't very interesting or fun for me. Also, this system is starting to become needlessly complex, which is partly my fault for allowing all this feature creep. I'm going to write up a post about the features I particularly want in the system (which is LESS than there are currently) and I'm going to allow comments on it. I'm not implementing anything until we've looked a bit at it, but to all those that want a really realistic thermal and reactor sim, you're going to be disappointed.

You do you Nertea; no one will be disappointed because whatever you do will be infinitely better than the (lack of a) stock system.

[Nertea]

I agree, this was useful.

The meat and potatoes of the system will be the same as the current build. Reactor & radiators will have low conductivity, radiators will "steal" heat from whatever they're attached to. This allows two things.

1. Easy stock compatibility
   
2. Easy arithmetic reactor radiator counts
   

Heat stealing will be limited to temperature differential between the two parts - that is, if the radiators are hotter than the attached part, the conduction will throttle. Reactor will keep the nominal/critical temperature system, it's not hard to understand and encourages not using the whole ship as a radiator. Reactors will also go from 0-100% instead of 30-100%, it's artificial and kinda annoying. Spinup speeds will increase, but reactor thermal mass will go down so that it's easier to heat up the object. Reactor fuel transfer will be as implemented in current version.

Sound good? This covers pretty much all the bases.

[GabrielG.A.B.Fonseca]

As long as I can also use radiators for other parts, which seems like the case, and the reactor maintenance is in, it's all fine by me, despite being a realism freak myself.

I trust your judgment.

[Fraz86]

Sound good?

Yes, that sounds great!

Minor cosmetic issues:

When the XR-2000 heats up, the most distal segment of the radiator becomes more red than the rest. Also, it looks like the XR-500 does not become as red as other radiators.

Edited May 11, 2015 by Fraz86

[RedParadize]

Sounds great, looks good. As always!

[Psycho_zs]

So, basically you're swapping stock conduction with your own conduction for radiators? And no core in reactors?

[Streetwind]

I agree, this was useful.

The meat and potatoes of the system will be the same as the current build. Reactor & radiators will have low conductivity, radiators will "steal" heat from whatever they're attached to. This allows two things.

1. Easy stock compatibility
   
2. Easy arithmetic reactor radiator counts
   

Heat stealing will be limited to temperature differential between the two parts - that is, if the radiators are hotter than the attached part, the conduction will throttle. Reactor will keep the nominal/critical temperature system, it's not hard to understand and encourages not using the whole ship as a radiator. Reactors will also go from 0-100% instead of 30-100%, it's artificial and kinda annoying. Spinup speeds will increase, but reactor thermal mass will go down so that it's easier to heat up the object. Reactor fuel transfer will be as implemented in current version.

Sound good? This covers pretty much all the bases.

Sounds good. With low conductivity the active mode radiators will work much better than they do now - with the heat transfer amount and temperature differential based throttle keeping them from scaling out of bounds. And AFAIK, active cooling is still not available from other mods (though I may have missed something, always possible with the huge number of mods this game has). Should only add to the popularity of Near Future

Question: are you going to include the thermal management parts (insulators, conductors, heatpipes) or are those going the way of the Dodo? I suppose you could argue whether or not they fit in NF Electrical thematically. On one hand, you're supposed to thermally manage reactors. On the other hand, the pack is about electricity, not heat, or spacecraft building options. You could conceivably also put them in NF Construction for instance.

Mostly I'm asking though whether or not I should still test them. The heatpipes are working kind of oddly, I haven't been able to put my finger on it just yet...

[4nxs]

... I suppose you could argue whether or not they fit in NF Electrical thematically. ...

You could change the name to NF Energy Management.

Joking aside, the new outlay seem interesting and not too expansive to comprehend, so looking forward to using it in game.

[kiwiak]

When I try to replace current test version of near future propulsion, will it break my save completly in future (if i try upgrading), or will it just remove ships containing parts from this mod?

[RedParadize]

This is the dev tread. Assume that it will!

[Starbuckminsterfullerton]

radiators will "steal" heat from whatever they're attached to.

Sounds just fine. Does this mean that radiators will have to be attached directly to the engine/reactor that they're cooling?

[Nertea]

Yes, that sounds great!

Minor cosmetic issues:

http://i.imgur.com/D72kxdd.jpg

When the XR-2000 heats up, the most distal segment of the radiator becomes more red than the rest. Also, it looks like the XR-500 does not become as red as other radiators.

Yes, I'm in the process of redoing all the animation curves for the emissives, you caught me part of the way through it.

So, basically you're swapping stock conduction with your own conduction for radiators? And no core in reactors?

No core - no point. It would just be adding another heat transfer coefficient. Better to not simulate it.

Sounds good. With low conductivity the active mode radiators will work much better than they do now - with the heat transfer amount and temperature differential based throttle keeping them from scaling out of bounds. And AFAIK, active cooling is still not available from other mods (though I may have missed something, always possible with the huge number of mods this game has). Should only add to the popularity of Near Future

Question: are you going to include the thermal management parts (insulators, conductors, heatpipes) or are those going the way of the Dodo? I suppose you could argue whether or not they fit in NF Electrical thematically. On one hand, you're supposed to thermally manage reactors. On the other hand, the pack is about electricity, not heat, or spacecraft building options. You could conceivably also put them in NF Construction for instance.

Mostly I'm asking though whether or not I should still test them. The heatpipes are working kind of oddly, I haven't been able to put my finger on it just yet...

Splitting those to HeatControl, which will be separately installable. So yeah, test as you wish.

When I try to replace current test version of near future propulsion, will it break my save completly in future (if i try upgrading), or will it just remove ships containing parts from this mod?

Nothing should break, but you might have minor issues with overheating and power supply.

Sounds just fine. Does this mean that radiators will have to be attached directly to the engine/reactor that they're cooling?

Not really. It'll be most effective to do this, but with some small penalty, you could run a heat pipe to the other end of the ship and install radiators there.

[RedParadize]

Not really. It'll be most effective to do this, but with some small penalty, you could run a heat pipe to the other end of the ship and install radiators there.

That sound soo cool. Make me want to try to have to pipe heat for a non ablative heat sheild to the top of a capsule and try a reentry! I know, will not work, but I am still gonna try it!

I wonder... With a "small" pluggin, something that generate heat from on energy consumtion, like electricity, combustion and even Kerbal metabolism, we could have a very nice heat management chalange here. Its not a request for you, just something I always wanted in KSP.

[Tau137]

Guys, this was a long and interesting discussion, but here is a small idea:

DRE already has core vs skin heat mechanics, may be that is something worth lookinging into and piggyback on (for better cross-mod compatibility as well)? Simple conductive drain for radiators for core temp of the part they are attached to (ignoring skin T) with high thermal insulation for reactors - this should do the trick while being simple and transparent enough.

Or is there anyone here who would rather NOT use DRE in their games!?

Edited May 12, 2015 by Tau137

[Britich]

I searched the thread, however the AFTER Ion Thruster appears to be upside down

[Nertea]

I searched the thread, however the AFTER Ion Thruster appears to be upside down

No to me! It was like.... 10 versions ago.

Guys, this was a long and interesting discussion, but here is a small idea:

DRE already has core vs skin heat mechanics, may be that is something worth lookinging into and piggyback on (for better cross-mod compatibility as well)? Simple conductive drain for radiators for core temp of the part they are attached to (ignoring skin T) with high thermal insulation for reactors - this should do the trick while being simple and transparent enough.

Or is there anyone here who would rather NOT use DRE in their games!?

Me.

NFP X.5.3

Assuming this is done, barring two things: NTR patches and LH2 tank switchability (to LH2/OX)

Changes

• Mipmaps for all dds textures
  
• Upgraded CRP version
  
• Upgraded MM version
  
• Fixed hydrogen tank misnaming
  
• Fixed stock ion patch mistargeting
  
• Fixed some Isp/heat production values for PITs and VASIMRs
  

NFE X.4.10

Please test:

• Functioning/balance of heat pipe
  
• Functioning/balance of heat insulators
  
• Appropriate conformal radiator count for reactors
  
• Lack of overheat under standard radiator count conditions
  

I know the Tau radiator doesn't radiate at all at the moment and am investigating why.

Changes

• Upgraded CRP version
  
• Model for Heat Pipe
  
• Model for Heat Insulator (textures WIP)
  
• Added 1.25 and 3.75m Heat Insulators (textures WIP)
  
• Mipmaps for all dds textures
  
• Adjusted radiator heat animations with new curve
  
• Tweaked radiator transfer capacities and method
  
• Reactor generator spinups greatly increased
  

[Streetwind]

Haha, I was about to post a bug report about the super weird hydrogen tank behavior Gonna test again with the new versions now.

I searched the thread, however the AFTER Ion Thruster appears to be upside down

That's the thing though, it only appears to be.

It's just that the model looks a bit different than you'd expect. When you actually mount it you'll find that it produces thrust just fine... and the exhaust comes from the flat bottom area. Because that one is a grid, and this engine is a gridden ion thruster.

[Streetwind]

Okay, I adore the heatpipe texture and model

Getting some weeeird effects though. Took a MX-4 with 6 conformal radiators to space. One there, and the reactor offline, they would happily cool the reactor down all the way to 4 K, which is equal to ambient. The fact that I sandwiched it between double insulators made the radiator's job that much easier.

https://dl.dropboxusercontent.com/u/44754370/screenshot51.png

But then - you'd think that at 4 K, the heat transfer would stop, right? Because the reactor is clearly not getting any heat input aside from the conductive backflow from the radiators, as you can see in the image. But no, the radiators keep heating up at full speed. They're at about 600 K now, even though they were barely over 400 K when the reactor reached 4 K! It seems like there's an infinite reservoir of heat to draw from a part that sits at minimum temperature. Inside the atmosphere, when the radiators keep the reactor at ambient, this does not happen.

I watched for a while and the radiators went all the way to 1048 K, at which point the rad flux + cond flux outgoing was equal the internal flux incoming.

With the reactor running at 100%, the radiators went up to 1065 K, a mere 17 K more. The reactor itself rose from 4 K to 390.7 K - nowhere near a critical temperature. Note, again, that this was double-insulated from the rest of the ship, which worked exceedingly well. The rest of the ship did not contribute anything whatsoever towards cooling the reactor.

Don't have more time than this to test today, I'm afraid. More tomorrow.

Edited May 12, 2015 by Streetwind

[MatterBeam]

Wow Nertea, you're like the superhuman blend of hard work and talent.

Sorry if it might sound indiscreet, but how are things holding up on the financial end? I see you have a Donate button, but there's been little discussion of it.

[Britich]

It's just that the model looks a bit different than you'd expect. When you actually mount it you'll find that it produces thrust just fine... and the exhaust comes from the flat bottom area. Because that one is a grid, and this engine is a gridden ion thruster.

hmm well ill be damned.

[Fraz86]

Nertea,

It may be helpful to include a ModuleManager file that adds nuclear reactors and decaying RTGs to the list of modules recognized as power sources for contracts. I believe this should work:

@Contracts
{
	@MODULE_DEFINITIONS
	{
		Power = ModuleRadioisotopeGenerator
		Power = FissionReactor
	}
}

[Starbuckminsterfullerton]

Me.

HAH. I agree, I think requiring heat shielding is very awkward in a sandbox game.

Assuming this is done, barring two things: NTR patches and LH2 tank switchability (to LH2/OX)

...

• Reactor generator spinups greatly increased
  

I think tank switchability is better left to Cryo Engines. Without them there is little use for it, and you already bundle IFS with that mod.

So is the generator separate from the core now in terms of warmup then?

You probably know, but the description for the small universal is incomplete.

Heat pipe looks exactly how I thought it would!

The Tau is a nice addition, is that what you would recommend for use in atmosphere?

I'll try and test more, still getting used to the new system. I find myself wishing for a 3m conformal.

Edit: is the heatpipe meant to shunt to a part with radiators attached, or the rads themselves? Also it's kinda hard to use vertically along a stack. With fuel lines you could always stick a cubestrut or something in there to let it reach further, but with heat the part either insulates the engine/reactor from the pipe or explodes due to overheat.

Edited May 13, 2015 by Starbuckminsterfullerton