Tempshift plate material selection

[suxkar]

3 hours ago, storm6436 said:

  It's different for the reasons I outlined in my first few paragraphs.  People, in general, do not understand the physics involved.

In terms of real life, literally everything made of bosonic matter conducts heat, including insulation, it's only a matter of how quickly. 

A lot of people think "Insulation is what keeps your house warm in the winter" and they're incorrect.  Your furnace keeps your house warm. Insulation only slows the heat leaving which means you don't have to run the furnace constantly.  Similarly the same thing applies to A/C in the summer.  

 All insulation is, is a layer of material with a lower thermal conductivity.  That's it. 

This might be helpful.  It's worth noting what most people "feel" as temperature is actuslly a rate of change.  dU/dS, to be specific.  change of kinetic energy per unit entropy changed. 

 Basically, if it takes twice as long for the far side to reach equilibrium, the shift plate is acting as insulation. Anything that slows it down is an insulator.  If it takes half as long, the shift plate is acting as a conductor.   That we have some materials that don't move heat at all is a result of the temp clamping, not their thermal conductivity.  

You'd be surprised to know how many of us people involved in these forums have a math/physics/engineering academic background, so "explanations" are not really necessary, especially those about real life: we are talking about how things work in this game.
 

3 hours ago, storm6436 said:

 Basically, if it takes twice as long for the far side to reach equilibrium, the shift plate is acting as insulation. Anything that slows it down is an insulator.  If it takes half as long, the shift plate is acting as a conductor.   That we have some materials that don't move heat at all is a result of the temp clamping, not their thermal conductivity. 

And yet I've proven that an "insulation" tempshift makes things worse in just 1 cycle. The heat absorbed by the tempshift doens't offset the extra heat transfer due to tempshift mechanic. I'm still waiting on a counter example as it would, I swear, just make me happy to have an extra way to insulate stuff.

 

[StarSquid]

Okay, I think the confusion here is arising because of one specific difference between ONI and real life. In real life, any matter will conduct heat to varying extents. In ONI, however, matter in the natural back wall does not. A tempshift plate with an insulating material is insulating compared to one made of a non-insulating material... but no tempshift plate at all is the same as a perfectly insulating tempshift plate.  It’s like how even Abyssalite is a worse insulator than vacuum, except that you don’t need tempshift plates to hold in vacuum. In real life that back wall will be conducting heat, and therefore replacing it with a better insulator would reduce heat transfer.

The other note is that tempshift plates work by “cheating” at heat transfer. Normally for heat to transfer between two tiles it would need to pass through the intervening tiles, but since a tempshift plate is bigger for heat transfer than it actually is the heat travels across the tempshift plate more-or-less instantly.

[storm6436]

5 hours ago, suxkar said:

You'd be surprised to know how many of us people involved in these forums have a math/physics/engineering academic background, so "explanations" are not really necessary, especially those about real life: we are talking about how things work in this game.
 

And yet I've proven that an "insulation" tempshift makes things worse in just 1 cycle. The heat absorbed by the tempshift doens't offset the extra heat transfer due to tempshift mechanic. I'm still waiting on a counter example as it would, I swear, just make me happy to have an extra way to insulate stuff.

 

 I wouldn't be surprised because I've read quite a bit on this forum since I started playing.  Unfortunately, even though one would hope that people with a "math/physics/engineering academic background" would understand the physics, that's seldom the case and my original statement still applies (See below.)

Small temperature gradient test to minimize amount of time (ie. I literally have to step out the door in 3 minutes)... but...Closed system, 1kg/tile O2, dirt tempshift tile.  T_hot=298.15K, T_cold=273.15K

Results? Dirt Tempshift plate retards thermal gradient flow, therefore it's an insulator.  Yes, there are subtleties in using them, I don't have time to expand... but it is an  insulator, it just doesn't have the same R-value as insulated tiles do.

Spoiler

 

 

[BLACKBERREST3]

There's something I'm missing here... "if any of the elements has the "isolated" flag in its temp (e.g. neutronium), no transfer is done"

Is this just visual?

Spoiler

Also found another shc difference by liquefying dirt. Nearly froze my hydrogen from 2000C.

Spoiler

 

We are getting into theoretical use case territory which I'm fine with, but I do admit that the practicality of it as a thermal insulator is very limited. Same with rounding errors on incredibly small amounts.

[storm6436]

7 hours ago, nakomaru said:

This is a whole lot of bombastic nonsense and irrelevant story telling. I really cannot imagine how we would ever communicate with any success.

Generally speaking, when one mistakes (purposefully or not) an explanation of the both the real world physics and possible ways the simulation implements them as "bombastic nonsense and irrelevant storytelling", we can't. 

Given the screenshots I posted of the in-game experiment worked out exactly as predicted (to 1st order), one might conclude that perhaps all that "nonsense" wasn't as irrelevant as you asserted.

[nakomaru]

No. You have simply demonstrated you still do not understand the difference between heat flow and temperature change. Nothing more.

[suxkar]

19 minutes ago, storm6436 said:

Generally speaking, when one mistakes (purposefully or not) an explanation of the both the real world physics and possible ways the simulation implements them as "bombastic nonsense and irrelevant storytelling", we can't. 

Given the screenshots I posted of the in-game experiment worked out exactly as predicted (to 1st order), one might conclude that perhaps all that "nonsense" wasn't as irrelevant as you asserted.

Well in that case any object with mass with lower temperature then environment is an insulator. And again, a tempshift might act as an "insulator" while equalizing temperature, but after that it will help spreading heat, so if you want to "insulate" using non-tile objects, you are better off using non-tempshifts. Hell, even a thermium tempshift is an insulator for a micro second, then it spreads more heat then if it wasn't there.

I repeat, in response to your test, any object with lower then environment temperature acts as an insulator, temp****s though will help spreading heat as they equalize temperature, so are basically the worst object to use to insulate.

[StarSquid]

1 hour ago, BLACKBERREST3 said:

There's something I'm missing here... "if any of the elements has the "isolated" flag in its temp (e.g. neutronium), no transfer is done"

Is this just visual?

  Reveal hidden contents

Also found another shc difference by liquefying dirt. Nearly froze my hydrogen from 2000C.

  Hide contents

 

We are getting into theoretical use case territory which I'm fine with, but I do admit that the practicality of it as a thermal insulator is very limited. Same with rounding errors on incredibly small amounts.

Wow, yeah, dirt to sand is a lot of heat deletion

[JoeW]

Let's keep posts polite and on-topic please. 

Thanks. 

[storm6436]

8 hours ago, suxkar said:

Well in that case any object with mass with lower temperature then environment is an insulator. And again, a tempshift might act as an "insulator" while equalizing temperature, but after that it will help spreading heat, so if you want to "insulate" using non-tile objects, you are better off using non-tempshifts. Hell, even a thermium tempshift is an insulator for a micro second, then it spreads more heat then if it wasn't there.

I repeat, in response to your test, any object with lower then environment temperature acts as an insulator, temp****s though will help spreading heat as they equalize temperature, so are basically the worst object to use to insulate.

 Not quite correct.  Any object colder than the environment can be a heat sink, which is what your assertion is actually describing, regardless of its thermal properties.  Insulator, conductor, doesn't matter, the question then becomes how fast you can put energy into it and get it out if you so choose, as that rate determines what one can actually use it for.. 

Still, heat sinks are not the same thing as an insulator.  You would have noted if I'd made a screen shot of it that the temp shift plates were precisely the same temperature as the air they were in at the start.  The fact that I have the building option check-marked does sort of imply that though since that sets building temps when you paint over them..  The point being is that 1 tile at either end is the source of flux while the opposite side is the measurement point.  Provided the experiment is configured correctly, what happens in the middle is largely irrelevant, only the end results matter, as all we're doing is classifying insulator vs conductor, and we're checking if the height affects flux rate or not, all in an isolated system. 

Now, I could add up the change in temperature on the relative test point convert that into watts  then divide by the number of squares used... and that would give me t_flux.  Doing the same for the control group above would give me a baseline for that particular size, then dividing the plate flux by the baseline gives us a number.  If it's less than 1 and left in a fraction (ie. 1/r), then oddly enough r=R, where R is the real life R-value you see tacked on to insulation.  It literally tells you how good at slowing down temperature movement a thing is.  Obviously if your goal is to slow down heat, having a number larger than 1 means its speeding things up, not slowing them down.

Still, I could, in fact take those three different fluxes, run this experiment a number of times to get sufficient number of points, and use the accumulation of data to generate an accepted R-value for dirt tempshift plates.

Overall, as configured, T_cold is the environment, not T_hot, but ultimately that wouldn't matter.  Because SHCs in ONI are treated as constants and we don't deal with enthalpies of vaporization or condensation, the system and the math are symmetric.  Setting the left most or the right most tile to whatever hot or cold temp you decide to use while setting the other 4 to the opposite temperature will not change the results. Measured thermal flux in the dirt shift plate  system is lower than the control group, therefore when normalized, T_flux < 1.  That meets the definition of insulator.  If you'd like I can go re-run the same thing with a refined metal or diamond plate to show that normalized T_flux with those is >1 and as such T_flux(insulator)<1<T_flux(conductor) holds true in ONI just the same as it does in real life.

 

I will point out that while this proves only insulating action on dirt tiles, it doesn't tell us everything we'd like to know..  On one hand, because the neutronium is non-conductive, this only tells us about shift tiles interacting with gasses, so nothing on "What happens if we place a plate near a wall.  Similarly it doesn't directly tell us anything about conductors, but one can draw some more than reasonable assumptions from the data nonetheless..

 As a second phase test, putting in a layer of insulated tiles with neutronium outside those would help highlight a number of potential issues for deployment.  Suffice it to say, it seems likely that spacing your plates to where they don't overlap with insulated walls is probably the smartest course of action. 

That said, it does highlight a somewhat obvious reason why you would put them next to a wall and a potential non-obvious use as well.: The obvious? You're intending on conducting through the wall.  It seems quite possible doing so would effectively augment the thermal conductance of metal tiles significantly.  The less-obvious case is, oddly enough, the same idea but in reverse: one might be able to use high conductance shift plates to push heat into insulated tiles much faster than would be otherwise possible, and use them as heat sinks.  This would make heat deletion much easier for people trying to dump heat into elements with lower thermal conductivities and then spacing them, but in that regard unless one has a massive surplus of said elements because for the most part, there's a decent correlation between low conductivity and low specific heat on a gram->gram basis.   Still, it does suggest that one could, if one were patient enough, design a system where one uses large insulated blocks as thermal batteries, and seeing as the insulated materials aren't terribly eager to lose the heat without the plates, the rate of loss on these thermal batteries would be fairly low if one could come up with a satisfactory way to toggle the feed on/off.  Other than vaccum, only construction/deconstruction comes to mind.

[BLACKBERREST3]

3 hours ago, JoeW said:

Let's keep posts polite and on-topic please. 

Thanks. 

Is it taboo to ask the devs how they programmed the tempshift plates to work?

Should we drop it or start another topic for tempshift mechanics?

I haven’t found a single reference to the mechanics behind how tempshift plates calculate temperature in a 3x3 area. Yothriel did some nice work on decrypting cell-cell transfer, but there is no mention of any topic (or could not find one) on the forums of how tempshifts lock their calculated temperature to a single number in all 9 tiles. If that is all they do while following the same rules of cell-cell temp transfer, then the math should be simple for us to figure out, but I have yet to see the empirical math for this, only testing that explains in game consequences based on the test used. There might not even be a way to test this for all I know because of all the bugs/exploits/chain of events that cause the experiment to have variability.

Will this help you in game? probably not. Could this be used to do something new? could be.

I myself would have more fun reading about it and coming up with other things than to perform operations beyond my capability so I think I will leave it here for now.

To the original post;

Unless the topic title changes, I believe the questions @StarSquid was asking is if there are any use cases currently for other materials. I had posted on shc and conductivity. The other aspect to this is that you can use them to manually move a heat/cooling source where you want. Then you have to take in to consideration the rarity and temperature range of the materials used. I use oni-db to filter what materials can be used for the goal I have in mind. Examples have been given by everyone here. I hope this helps you.

[wachunga]

From my testing there is nothing special about tempshift plates. They are simply a 3x3 building that can be built virtually anywhere without interfering with anything else. Under a tile being the exception.

For the actual calculations, "q = conductivity1 * conductivity2 * dT * (mass*SHC of the hotter thing /5 if a building) / 10 per tick" is still accurate last I checked. Buildings that span multiple cells are a bit more involved but still follow that principle.

In a building that spans n cells, the heat transfer from any single cell is hard capped at dT/n. So for a tempshift plate interacting with one cell with a difference of 90C, 10C per tick is the most the plate can change by. The change of the cell is calculated off the heat transfer necessary to achieve the plate's temperature change. If the plate is interacting with two cells each with a 90C difference, it then becomes a 20C change at most per tick.

Note that there is a soft cap in addition to this hard cap. At 60% of dT/n the heat transfer (as governed by the above q=k1*k2...) starts to become reduced. I've never bothered to figure out what this reduction is. The math nerds can gather a bunch of data, plot a graph, and figure out the relationship if they really want to.

Presumably the game sums up all the individual cell interactions to give a final temperature for the tempshift plate (or any other multi-cell building). I'm no longer inclined to put in the time gathering data, but anyone curious about this stuff can easily test it for themselves in debug. Use Alt- to advance one tick and the sample tool to read temps.