[1.1.3] RealHeat (Minimalist) v4.3 July 3

[kpstjeev]

So, I enabled all kinds of debug that was in RH sources, built it and noticed the double amount of radiation assigned to backgroundRadiationTempExposed, which explains a lot.

It was multiplied with comment "blunt bodies multiply BRT by 0.5, so multiply by 2 here." - apparently this does not apply to my *clean steam installation* for whatever reason, particularly to mk1 Cockpit.

I removed the multiplier, rebuilt and the flux looked normal again, rad flux is around 1.3 on kerbin's runaway.

So...does https://github.com/KSP-RO/RealHeat/commit/5248b74b0795a17cee848ead9c67372baae4243a still apply? How can I check if BRT is multiplied by 0.5 in KSP engine?

[NathanKell]

The 0.5 multiplier is done internally by flightintegrator when running its occlusion system and then passed on to its internal radiation method. You won't see anything related to it anywhere, but unless you are using a mod that changes that bit of Flight Integrator (and to my knowledge no such mods exist) it's still there.

[kpstjeev]

The 0.5 multiplier is done internally by flightintegrator when running its occlusion system and then passed on to its internal radiation method. You won't see anything related to it anywhere, but unless you are using a mod that changes that bit of Flight Integrator (and to my knowledge no such mods exist) it's still there.

Obviously it does not for me.

A log line from RH modified to report data only, without any assignments to 'fi':

At rho 1.12227822776333/1.12227822776333, gamma 0, DTL 0.222761160125366, BT = 242.83/242.83 AT = 311.286419315105 ET = 311.286419315105 VET = 311.286419315105

AT/ET/VET - temperature, atmospheric/external/Vessel.external

I expected second value in BT = 242.83/242.83 to be twice as large, since RH pushes the second value over 500 which is causing overheating on runaway for me.

- - - Updated - - -

To add, I only use ModularFlightIntegrator, RealHeat and ModuleManager. All installed by latest CKAN. No other mods.

[Starwaster]

Obviously it does not for me.

A log line from RH modified to report data only, without any assignments to 'fi':

At rho 1.12227822776333/1.12227822776333, gamma 0, DTL 0.222761160125366, BT = 242.83/242.83 AT = 311.286419315105 ET = 311.286419315105 VET = 311.286419315105

AT/ET/VET - temperature, atmospheric/external/Vessel.external

I expected second value in BT = 242.83/242.83 to be twice as large, since RH pushes the second value over 500 which is causing overheating on runaway for me.

- - - Updated - - -

To add, I only use ModularFlightIntegrator, RealHeat and ModuleManager. All installed by latest CKAN. No other mods.

Well something's making some significant changes to your PhysicsGlobals from what I see in my own, and they should be matching up. Maybe CKAN is pulling up something old, maybe there's a versioning issue somewhere. I've duplicated your RH/FI only environment and I'm not seeing the symptoms you're seeing and the conduction factors in your PhysicsGlobals are very low, Convection settings are different too.

I installed RH manually so I'm looking at RH on Github and I'll compare it to what's on CKAN to see if I can find something there.

Edit: Suggestion: Delete all ModuleManager.* in your GameData folder EXCEPT FOR the dll itself. (i.e. Delete ModuleManager.ConfigCache, ModuleManager.Physics, ModuleManager.TechTree and ModuleManager.ConfigSHA - obviously the very first two are the one's that are really important to nuke but nuke them all from orbit - it's the only way to be sure)

Edit #2: Somewhere along the way, things changed in your configs and the MM caches didn't get cleared. Generally, what's in ModuleManager.ConfigCache/PhysicsGlobals should match up with what's in ModuleManager.Physics, and in your case they do not. When I spliced in the contents of your cached PhysicsGlobals into my ModuleManager.Physics file and loaded it from the debug menu, my temps started climbing. Right now on the pad at night, it's past 350K which is about 170 F and that's at night. Even on the equator, that's way too hot and it's not stopping.

So, my diagnosis is that you need to clear that cache out.

Edited October 23, 2015 by Starwaster

[kpstjeev]

Ok...I just made sure that I'm not going crazy and my hardware is sane by testing KSP installation on nearby laptop. With same result.

The exact conditions/steps to reproduce the issue:

OS: Win7/Win8.1

KSP: 32bit, from Steam

1. Install KSP from Steam (remove whole directory and verify integrity to be sure)

2. Copy KSP dir to another location

3. Get RH 1.1 from Git Releases and get module manager 2.6.8

4. Install RH, FL Integrator and ModuleManager in GameData

5. Run KSP.exe, start new sandbox carrier, default settings

6. Go to Space Hangar

7. Place Mk1 Cockpit

8. Click Launch (no other parts)

9. Alt-F12 -> Physics -> Thermal > Display Thermal Data...

10. Right click on Cockpit

11. Observe "Rad Flux" value

Expected Result:

Rad Flux < 2 (less than two)

Actual Result:

Rad Flux ~ 27.29 (way over twenty)

Run simulation for a bit on 100x time-warp and observe how Skin Temp slowly rises to 350 K and beyond.

@Starwaster

What was value of "Rad Flux" before you loaded my ModuleManager.Physics file?

Could you please post your Physics.cfg? My Physics.cfg matches perfectly to ModuleManager.Physics and matches to what steam downloads.

I tried to clear ModuleManager caches and also tried to remove MM along with all its files and got exactly same result, MM is not necessary to reproduce the issue.

[Starwaster]

This is a link to Physics.cfg from 1.0.4: https://www.dropbox.com/s/0qrtkf9altzb8r4/Physics.cfg?dl=1

I don't remember what my rad was; have to run stat. BBL

[kpstjeev]

This is a link to Physics.cfg from 1.0.4: https://www.dropbox.com/s/0qrtkf9altzb8r4/Physics.cfg?dl=1

I don't remember what my rad was; have to run stat. BBL

This config matches perfectly to the config from "clear_realheat" folder, from the zip I posted above.

The one you loaded was probably from "modded" folder, which is altered by Realism Overhaul, that config just makes the issue more obvious, so please check Rad Flux value.

[Starwaster]

This config matches perfectly to the config from "clear_realheat" folder, from the zip I posted above.

The one you loaded was probably from "modded" folder, which is altered by Realism Overhaul, that config just makes the issue more obvious, so please check Rad Flux value.

No, it wasn't. I was pulling those values from your clear_realheat folder. Once I copied them over to my own physics settings I experienced what you were describing.

The rad flux I'm seeing right now, on the same part is ~5-7, holding steady at 314K, noon. That's pretty much where it's going to peak and comes out to about 105 F, which is downright cool for a metal surface with the sun beating down on it all day. That's RH+FI plus a few other mods like mechJeb and hyperedit but neither of those touches physics.

[SancoPanza]

I don't know whats going on but Real Heat is basicaly doing Earth to Mercury conversion for me.

http://imgur.com/a/BjrJt

Notice the radiating flux values and skin temp.

Near zero for stock, and ridiculous values for Real Heat with RO+RSS.

For Stock + Real Heat they are not so ridiculous but still a lot higher than expected, even at night.

KSP is latest from Steam, mods installed using CKAN, started new sandbox game for each test.

Hi,

i had exactly the same problem. After trying out this and that, i found a solution for me.

I installed this version from the modular flight integrator http://ksp.sarbian.com/jenkins/job/ModularFlightIntegrator/lastSuccessfulBuild/artifact/ModularFlightIntegrator-1.1.0.0.zip, and the overheating was gone.

Maybe this will help u too.

I'm not sure this is the right fix for it, but if it is, maybe someone can change it on ckan.

[kpstjeev]

No, it wasn't. I was pulling those values from your clear_realheat folder. Once I copied them over to my own physics settings I experienced what you were describing.

But this is impossible. Your physics config perfectly matches to what was in clear_realheat folder, byte to byte. It also matches to what PHYSICSGLOBALS section contains in ConfigCache of clear_realheat. You can verify it with diff tools.

By logic if your PHYSICSGLOBALS is different then mine from clear_realheat ConfigCache, then this mean that you are using mods that alter PHYSICSGLOBALS, not me.

It feels like your radiationFactor is 1/10 of default value, which is '1'

:confused::confused:

Using the steps from few posts above the issue was reproduced on two Windows PCs, with fresh KSP 1.04 install. If you are using Linux build then it may be different.

- - - Updated - - -

Hi,

i had exactly the same problem. After trying out this and that, i found a solution for me.

I installed this version from the modular flight integrator http://ksp.sarbian.com/jenkins/job/ModularFlightIntegrator/lastSuccessfulBuild/artifact/ModularFlightIntegrator-1.1.0.0.zip, and the overheating was gone.

Maybe this will help u too.

I'm not sure this is the right fix for it, but if it is, maybe someone can change it on ckan.

My guess is that you just disabled RH and probably FAR along with it, due old and new version of MFI not being compatible, check your logs:

AddonLoader: Instantiating addon 'RealHeat' from assembly 'RealHeat'

(Filename: C:/buildslave/unity/build/artifacts/StandalonePlayerGenerated/UnityEngineDebug.cpp Line: 56)

FileNotFoundException: Could not load file or assembly 'ModularFlightIntegrator, Version=1.0.0.0, Culture=neutral, PublicKeyToken=null' or one of its dependencies.

UnityEngine.GameObject:Internal_AddComponentWithType(Type)

UnityEngine.GameObject:AddComponent(Type)

AddonLoader:StartAddon(LoadedAssembly, Type, KSPAddon, Startup)

AddonLoader:StartAddons(Startup)

AddonLoader:OnLevelWasLoaded(Int32)

(Filename: Line: -1)

Error looking up signature for method RealHeat.UpdateThermodynamicsPre

(Filename: Line: 243)

FileNotFoundException: Could not load file or assembly 'ModularFlightIntegrator, Version=1.0.0.0, Culture=neutral, PublicKeyToken=null' or one of its dependencies.

UnityEngine.GameObject:Internal_AddComponentWithType(Type)

UnityEngine.GameObject:AddComponent(Type)

AddonLoader:StartAddon(LoadedAssembly, Type, KSPAddon, Startup)

AddonLoader:StartAddons(Startup)

AddonLoader:OnLevelWasLoaded(Int32)

(Filename: Line: -1)

[SancoPanza]

Yea, u'r right. I should really check those kind of things before posting.

Maybe i was a little bit to excited as i see that the rad flux stays normal.

Sorry i'm just a kind of noob in those things

Do you have checked the partEmissivityExponent? Its value is default on 4, whats the same is as in RO-Physics, but to lower it effects the rad flux for sure. i set it on 2, and it still looks not right, but away better the on 4.

[Starwaster]

@kpstjeev Okay, thanks for the overwhelming vote of confidence there. I know what folders I was using and what's going on in my own install. An install I only created to replicate your problem. It's a fresh install and I'm only getting a fraction of the heating you're seeing. I don't know what's happening on your side but I can't spend any more time on it right now. Good luck to you

[kpstjeev]

@SancoPanza try this dll https://www.dropbox.com/s/w52v0veboza150i/RealHeat.dll?dl=0 it the same as latest version but with backgroundRadiationTempExposed 2x multiplier removed. At least it does not look ridiculous.

@Starwaster I feel like I'm being trolled, honestly.

You did not provide any details about your installation, not even platform/OS, your overconfidence in "I know what folders I was using" terrifies me.

You loaded stock physics config and you had overheating all of sudden, while you where claiming that you already use stock physics, this is simply ridiculous, don't you think?

You also tested on god know how configured installation with "plus a few other mods", while I tested on clean&fresh installation.

I'm done proving that there is an elephant in the room, too much time wasted.

It is enough for me that at least two persons have this problem, so it not just my PC/Config/Software/Whatever. Others are probably does not care or not notice it.

[StoryMusgrave]

Well....I made a fresh install of KSP. Win 7 64bit. Running 32-bit KSP. RSS + RO (Dependencies with Deadly Reentry). I'm getting skin temps in the 500-550K range and temps <300K.

[NathanKell]

Please try not to bite the hands that help you, folks. Whether the help is exactly as you ordered it to be or not.

[StoryMusgrave]

I'm just saying that those issues described by kpstjeev and others are identical to those experience by myself when following their duplication steps. Grabbing the latest version of everything as provided by either github or when not available there kerbalstuff.

I'm not really going to worry about it at this point because 1. My ships are *NOT* exploding. 2. I know that 1.05 and/or 1.1 are going to provide vastly better thermal simulation than what 1.04 does. Not to mention you Nathan apparently leading that charge:)

It is interesting that those steps done by another *are not* producing the same effects. There has got to be something different in the versions of RealHeat or MFI used by each party because something doesn't jive.

I'm also going to agree that the physics.cfg provided by starwaster and that same file provided by kpstjeev is identical absolutely no question to it. In fact the physics.cfg of starwaster is identical to both that in the 'clear_reheat' folder AND in the 'modded' folder.

Edited October 25, 2015 by StoryMusgrave
Missed a big key word there.

[Starwaster]

I'm just saying that those issues described by kpstjeev and others are identical to those experience by myself when following their duplication steps. Grabbing the latest version of everything as provided by either github or when not available there kerbalstuff.

I'm not really going to worry about it at this point because 1. My ships are exploding. 2. I know that 1.05 and/or 1.1 are going to provide vastly better thermal simulation than what 1.04 does. Not to mention you Nathan apparently leading that charge:)

Please clarify, are you saying in general that your ships explode under 1.0.4 or are they only exploding with Real Heat?

[StoryMusgrave]

Whoops, missed a big word there. They are *NOT* exploding. So while temps are higher than what they realistically should be, it doesn't prevent me from playing.

[NathanKell]

Ah, sorry! Thought you were saying you didn't get those errors. Now I'm confused, because I figure if anyone will follow installation instructions to a T it's you...but I have yet to see those issues myself, like Starwaster.

Well, at present my hope is that it's 1.0.4 bugs which 1.0.5 will fix. Let's reevaluate shortly.

[StoryMusgrave]

I'd hope it's a 1.04 bug too. Still very strange indeed that clean installs for some are having issues, while others are not. I know some devs from time to time have had 'new' yet to be released stuff in files that have created troubleshooting nightmares because their new stuff has fixed the errors seen by others. Not saying this is the case this time, but still strange strange strange.

[StoryMusgrave]

Well...this is odd. Same test as before. Now my temps (both skin and 'regular') barely crack above 300K.

Difference. The new RealFuels that was JUST released. (v10.7).

The test was done using the Mk1 Cockpit which DOES have a "ModuleFuelTanks" of RealFuels in the RO config...

Is it possible both you NathanKell and Starwaster were using this new version of RealFuels??? While some others were using the 'old' version.

[Starwaster]

Well...this is odd. Same test as before. Now my temps (both skin and 'regular') barely crack above 300K.

Difference. The new RealFuels that was JUST released. (v10.7).

The test was done using the Mk1 Cockpit which DOES have a "ModuleFuelTanks" of RealFuels in the RO config...

Is it possible both you NathanKell and Starwaster were using this new version of RealFuels??? While some others were using the 'old' version.

Not when I duplicated the other guy (sorry forget name)'s RH only environment.

It shouldn't matter though unless you had cryogenic propellants in the part being monitored, but that probably goes without saying. (you will definitely notice the cooling effect on parts)

If it's a cryogenic tank type WITHOUT cryo resources then... the effect would be that there would be insulation between skin and internal so you'd see lower internal but somewhat higher skin?

structural and fuselage have no such insulation (this is the type where it actually modifies the part's conductivity between skin-internal)

Edited October 25, 2015 by Starwaster

[StoryMusgrave]

Well, now I'm really confused. Redid my install, just KSP, RH, MFI, and MM and I have no issues really. What I have noticed is that warping really messes with temperature. At 100x temps will climb to 350/375K. Now kinda toggle back and forth between say 50x and 1000x quickly and those 350+ temps magically drop to the 315K range. The rad flux figure is still 20+, fwiw.

I also noticed another little bit. The physics.cfg that comes with a fresh install of KSP 1.04 is different than one generated by KSP under stock conditions if the original physics.cfg is deleted.

What differences that makes, no clue.

I'm at a loss. All I can say is BRING ON 1.05.

EDIT: WHOA...so using the auto generated physics.cfg causes exactly what is happening (skyrocking temps). Using the physics.cfg that comes with a fresh install temps stay nice and cool.

Edited October 25, 2015 by StoryMusgrave

[Svm420]

EDIT: WHOA...so using the auto generated physics.cfg causes exactly what is happening (skyrocking temps).

Known issue never let KSP generate that file. Always replace from a fresh install i.e. reDL from steam if you use that.

[Starwaster]

The problem with regenerated physics.cfg is that several values are zeroed out

Also issues present when warping above 100 might not be present at all at 100 and below. The reverse is also true