Jump to content

Piping Comparisons Testing w/ wierd results


Recommended Posts

Don't want to hijack the other thread (where I was asking @Risu about the new selection info panel in debug tool), but I'm not sure if there's just something wrong with the selection info or that there's somehow some leakage of heat going on in my test system...

Here's the piping:

The sections inside the testing chamber is made of the testing piping material.  Everything else is made with Gold Amalgam and Abyssalite.

T1Piping.thumb.jpg.902fd4482625fbdcea7ee6aaca6233fe.jpg

Test Setup:

T1Setup.thumb.jpg.1b9525d3354ea9b488b1962adb0391de.jpg

Result:

T1Result.thumb.jpg.105853347a42fcc289186184472271a2.jpg

I really don't know what's going on here...  The external medium for Wolframite gained less heat compare to Granite and Sandstone, and yet, the internal medium lost more heat than the other two...  Yes, I know there are some extra heat from the pump, but it should be the same for all 3.  And yes, it's true that the Abyssalite bridge changes temperature (something @Kasuha mentioned before), but that still has nothing to do with the wolframite part of the system.

I've tried to rebuild the whole thing, retest with same and different amount of internal medium, etc, but it's still giving me pretty much the same result.

Here's the save file:

Huh.sav

 

 

Link to comment
Share on other sites

When rebuilding, did you rebuild in the same order and shape? As in, wolframite on top, that bridge in there etc.

Are pipes not in medium made of tested material or abyssalite? Probably shouldn't explain the magnitude, but apparently pipe going through a tile can exchange surprising amount of heat with it, even if the tile is abyssalite.

Could be good to test it in a loop: pump the liquid once, but don't drop it through a vent, instead loop it back with a bridge. If it keeps losing heat, it's obviously a bug, but if it stabilizes it means it's just an engineering error.

Link to comment
Share on other sites

1 hour ago, Coolthulhu said:

When rebuilding, did you rebuild in the same order and shape? As in, wolframite on top, that bridge in there etc.

Are pipes not in medium made of tested material or abyssalite? Probably shouldn't explain the magnitude, but apparently pipe going through a tile can exchange surprising amount of heat with it, even if the tile is abyssalite.

Could be good to test it in a loop: pump the liquid once, but don't drop it through a vent, instead loop it back with a bridge. If it keeps losing heat, it's obviously a bug, but if it stabilizes it means it's just an engineering error.

The save file contained the first rebuild, where I moved one of the bridge to see if it'd help with it not changing temperature.  (It didn't help)

All pipes and bridges not in the test chamber are made with abyssalite, and I've actually clicked through all the pipes inside the tiles, and none of them changed temperature (mainly because both of abyssalite tiles and the abyssalite pipes inside are both at 20C).

I'm going to try and rebuild it again, this time without any bridges, and see if this is still happening...

Link to comment
Share on other sites

9 minutes ago, Reaniel said:

I'm going to try and rebuild it again, this time without any bridges, and see if this is still happening...

If everything but pipes in the water chambers is made of abyssalite, changes to layout shouldn't change anything.

The difference is negligible, about 0.2%. 

Thermal conductivity of the material doesn't play role since everything is capped by thermal conductivity of water. Heat capacity of the material plays negligible role since you're warming 1000 kg of water and 5 kg of pipe per tile. With water heat capacity 4.179 J/K/g and heat capacities of pipe materials 0.134, 0.79, and 0.8 J/K/g the difference is actually on the level of 0.1% - even less than what you measured. So I would guess you've ran into an issue with the dripping heat transfer bug somehow. Or maybe into the liquid pump/hot water interaction?

When I was making such measurements, I was holding all the material in a pipe - pre-fill an abyssalite pipe from a constant temperature reservoir, hold it with a valve. Then send it through the testing chamber, and collect it in a meandering pipe again. That way I could check temperatures of all packets, not just the bulk, and both before and after the pass.

Link to comment
Share on other sites

It's just counterintuitive if what we're observing is caused somehow by the cooling bug since this scenario is probably as far away from the causes as we can.  As for the heat drip loss, It's possible, but still doubtful, since the beginning and end packets of the internal medium had always been 88.7...  And it's just baffling how the heat loss is mostly focused in wolframite system but not so much in the other 2.

I've did another rebuild and gotten rid of the bridges altogether, and the results are pretty much the same...

T2Piping.thumb.jpg.3a8385daa3d9dad45728b85b1cfc0c6a.jpg

T2Result.thumb.jpg.5f69ff6d422a81c5a0ce38d09d993ce4.jpg

I've made the system run longer, and... The wolframite system lost even more heat...

Roughly 7% Loss in heat for wolframite pipes.

WITW.sav

Link to comment
Share on other sites

@Reaniel Did you take into account the heat capacity of the pipes in the testing chamber when calculating the net energy gains for the testing chambers? Although if you hadn't, I would expect lower net gains for sandstone and granite than for wolframite which is not what's happening here.

Link to comment
Share on other sites

14 minutes ago, Sevio said:

@Reaniel Did you take into account the heat capacity of the pipes in the testing chamber when calculating the net energy gains for the testing chambers?

Not in these new tests, no.  I figured they would be negligible from a pure comparison point of view.  If anything, the low specific heat of the Wolframite pipes should make a larger portion the heat lost/gain in the internal medium move onto the external medium, which is clearly not the case.

At this time, I'm more inclined to agree with @Kasuha that it's the cooling bug at work somehow.  Maybe the small specific heat of wolframite is somehow making each section of the pipes (with stuff in there) mimicking the thin layer of cold medium or whatever.  It's just annoying, and doesn't occur when you have low mass tiles in the external medium (aka gas).

Going to put suggestions Kasuha and Saturnus to test, but it appears that radiator systems will inadvertently cause the cooling bug to strike, making it a bad way to heat a pool of liquid and kind of cheaty to cool liquid.

 

Link to comment
Share on other sites

@Reaniel The liquid cooling bug happens as a result of smaller masses on top of larger masses. Even if you paint a tank full of water in debug mode, the liquid will settle down into larger masses at the bottom and less-than-full tiles at the very top due to how liquid pressure works.

An easy way to test if this is affecting your results is to use a horizontal tank of only 1 tile high and run your radiator pipe through that.

Link to comment
Share on other sites

11 hours ago, Sevio said:

@Reaniel The liquid cooling bug happens as a result of smaller masses on top of larger masses. Even if you paint a tank full of water in debug mode, the liquid will settle down into larger masses at the bottom and less-than-full tiles at the very top due to how liquid pressure works.

An easy way to test if this is affecting your results is to use a horizontal tank of only 1 tile high and run your radiator pipe through that.

I understand the top row would have less mass, but that layer isn't THAT low in mass, and if it is indeed the cause, it should still be uniformly present in all materials, which clearly isn't the case.

This was also happening, to a smaller extent, even in single height rows.  I did notice it in my old tests of 4000kg water runs, but I figure it was just some neglictable margin of error in calculation.  Now that I ran tests with longer runs (32000kg water) on the same setup, the heat loss is much more apparent.

So It's definitely something caused explicitly by the behavior of wolframite pipes in liquids (high mass external medium).

Here's the new tests I've done:

Retest1Piping.thumb.jpg.d8e05d4f7b848464b2bf0c59ca0261ed.jpgRetest1Heating.thumb.jpg.3fc2bb232aff04effaa4cd39f6b97171.jpgRetest1Cooling.thumb.jpg.4266bacfc9f941bcd71aab37ac2ada38.jpg

The wolframite system is losing/destroying heat in all instances, whether we're heating or cooling, and even in horizontal single depth tiles, we are seeing lost/destroyed heat.

Another inexplicable result is using granite pipes to cool liquid.  Why would there be such a huge difference between sandstone and granite in this scenario when common sense says granite should be close, if not slightly better.

At first I thought it could be that the granite pipes are holding the heat, but calculation doesn't give you THAT much either.

100*100*1000*0.79*(95-93.1) = 15MJ

And the pump-produced heat doesn't account for the rest either:

10*200*32000/10 = 6.4MJ

So where did the extra ~78MJ come from? 

Link to comment
Share on other sites

Did another rebuild to see try and see if single height rows will be a factor or not.

Piping:  The testing tiles are now the same for both single chamber and single height rows (100 pipe sections)

Retest2Piping.thumb.png.3c945f5b65e361ab262dc3e349571f93.png

Result:

Retest2Result.thumb.png.1cd33f9a2e0636a7187f68c9621b7132.png

Save File:

AdvPipe.sav

Conclusion:

It's pretty obvious that both granite pipes and wolframite pipes will trigger some kind of bug if used to cool a liquid, with wolframite destroying heat and granite adding heat.  Single height rows play little to no role in the causing the bug.  Sandstone is probably the one with the least bugginess, but the bottom line is liquid heat transfer is currently a total mess, and the interaction with different pipes make it even more so.

Interestingly, granite pipes aren't bugged if used to heat a liquid for whatever reason, but wolframite is just bugged like crazy.

Link to comment
Share on other sites

More liquid heat transfer mess...

I was just wanting to test how viable it is to use plastic tiles as  a heat conductor (since I'm still a bit iffy when it comes to spamming bridges).  And... SURPRISE!  It's freaking bugged!

Retest3Result.thumb.png.86887917b6d1ac49becdae6ff40f7dec.png

So very much like the case with Wolframite, plastic would take away more heat than the other testing materials, but then destroys a portion of that heat in the process.

I don't know, it just seems like liquid heat transfer is extremely buggy when you have a highly conductive material involved.  I guess it's best to stick to cooling/heating liquids with a aquatuner if I want to avoid unintended heat destruction.  I do wonder, though, that if wolframite aquatuner would be bug-free or not.  EDIT: It's not bugged (after testing it), nice.

Link to comment
Share on other sites

@Reaniel Interesting, I think I recall a thread in which @Risu and/or other code divers looked at the code for heat transfer and discovered that it was likely due to some strange temperature clamping being done and that it apparently can happen whenever a small mass and a large mass exchange heat. But perhaps it might be more correct to say that a small total heat capacity is actually what causes it.

With buildings having 1/5th of their construction mass  and wolframite having a relatively tiny heat capacity, this might explain why it's more prominent with Wolframite and not so much with Granite and Sandstone. The tiny heat capacity wolframite pipes are exchanging heat with 1000 kg tiles of water, which has one of the largest heat capacities in the game.

Link to comment
Share on other sites

2 minutes ago, Sevio said:

@Reaniel Interesting, I think I recall a thread in which @Risu and/or other code divers looked at the code for heat transfer and discovered that it was likely due to some strange temperature clamping being done and that it apparently can happen whenever a small mass and a large mass exchange heat. But perhaps it might be more correct to say that a small total heat capacity is actually what causes it.

With buildings having 1/5th of their construction mass  and wolframite having a relatively tiny heat capacity, this might explain why it's more prominent with Wolframite and not so much with Granite and Sandstone. The tiny heat capacity wolframite pipes are exchanging heat with 1000 kg tiles of water, which has one of the largest heat capacities in the game.

I remember that as well, but then again, it's happening to plastic also, which has a fairly high specific heat that'd more than compensate for its smaller mass compare to tiles and doors.

However, the "using granite pipe to cool liquid" bug where heat is added just completely stumped me.  I can't understand how it's happening to granite but not sandstone, when the two are so similar.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Please be aware that the content of this thread may be outdated and no longer applicable.

×
  • Create New...