What are all of the exact mechanics of Temp Shift plates?

[Mathgeekburch]

I made this post on the subreddit and I was told to go here with my questions for more technical answers.  So this is just the copy of what my reddit post.

 

"What I understand (at least I think I understand) is that temp shift plates interact in 3x3 area, but do not exchange heat with buildings or objects. That they have thermal mass proportional to their specific heat (since they all have the same mass). And the better their conductivity, the faster they are at transferring heat.

Although I do have some questions about how they interact with different mediums and about any abnormal interactions.

Is there any noticeable differences between them interacting with gas, liquids, and solids?

--------------------------- Feel free to skip the entire mountain of text below. Just me being scatterbrained.

To my limited knowledge, the heat transfer depends on both the conductivity of the temp plate, the conductivity of the medium, and the temp difference between the two. And that the equation for its heat transfer is the same for all temp shift plates and mediums (IE there is no special multipliers for specific mediums or temp shift materials outside of the raw conductivity values). Is this correct?

 

To my limited knowledge, the only abnormal interaction with gas is that when a gas is heated and has they same type of gas above it, some of the heating is transfer to the above gas; meanwhile when a gas is cooled and has the same type of gas below it, part of the cooling is transferred to the below gas. This in turn means that when moving heat upwards through gases, that it can be advantageous to put temp shift plates particularly at where two different gases meet, as the conductivity of the plate will assist in allowing heat to pass through the boundary of lower heat transfer.

As for liquids I was under the impression that the plates treat them exactly the same as gases, just that liquids have different conductivity. The exception is that liquids have a really high effective conductivity with other liquids (I think 625 times higher than what would be the case based on their conductive values). Which basically means temp shift plates for liquid to liquid transfer isn't as useful when the liquids are in contact; it also means the liquids have a higher 'effective' thermal mass when interacting with individual temp plates as whatever heat they absorb/lose will quickly be shared with the other liquids it is in contact with.

Is there any special interaction with moving/falling liquids?

Although I have seen multiple weird interactions with temp shift plates and liquids. Particularly that ice temp shift plates instantly melt when placed in a liquid above their melting point. Base on my understanding of their mechanics, this shouldn't occur; yet it does, so clearly there is a large gap in mechanics I am missing. As such I have many sub-questions:

• Does the result depend on the type of liquid and shift-plate? IE, will above 0c water, pwater, oil, ethanol; all insta-melt an Ice temp shift plate?

• Does it purely depend on the melt temp of the shift-plate and does it occur with all shift-plate? IE, will lead plate insta-melt in 400c petrol, will iron plate insta-melt in 2000c liquid copper?

• What are the spacial requirements for it? Does it just need the liquid in one of its 3x3 area, or does it need liquid in it's 1x1 area? Does it need a full tile of liquid, or just a little bit?

• Will it occur with falling liquids?

• Does any similar effect happen with solid tiles?

• Can a similar thing happen with super heated gases (for instance if a gas is above the boiling point of the temp shift plate)?

• How much greater is the heat transfer through this effect compared to conventional heat transfer through shift-plates.

As for solids blocks/tiles I haven't seen much on temp-shift plates used with them. Does temp-shift plates treat solids the same as they do gasses and liquids, as in the heat transfer formula is the same for all three? Or do temp-shift plates not interact with said block/tiles (and doors) with its 3x3 area?

I have heard of a weird effect between temp-shift plates and insulated tiles. To my limited knowledge, insulated tiles have an additional "insulation coefficient" on top of their low conductivity, which in turn reduces their heat transfer with nearby cells. Does the temp-shift plate circumnavigate this insulation coefficient therefor greatly increase heat transfer to and from the insulated tile? Does this circumnavigation of the insulation coefficient also occur for other buildings as well such as conveyors and pipes?

Outside of that, I heard there is 25 conductance multiplier between solid blocks and gases, does this mean tempshift plates aren't as useful for conducting heat between a solid and a gas, but more useful when conducting liquid to a gas or solid to a liquid?

Also is there any special interactions with falling blocks?

Finally I presume there isn't anything particularly special about temp-shift plate made out of insulation? Presumably it only acts like a drywall with effectively zero heat capacity (your heat capacity doesn't matter if you don't absorb or emit heat). I am not to sure if there is particular uses for wanting a drywall with a heat capacity of zero. There isn't any additional "strange" interactions insulation-temp shift plates have, right?

My apologies to those who actually read the mountain of text of me being scatterbrained. Although in my defense I did warn you and said feel free to skip. Anyways since you read the mountain of text (or skipped to the bottom), do you notice any misconceptions I have? Do you happen to know any answers to my question? Do you know of any good ways to test some of the questions? How stupid are my questions? Did anything spark your interest?"

[Gracefulmuse]

I can't answer any of this. But I am totally surprised you don't have answers yet, this is normally the kind of question @mathmanican would jump on. 

[beowulf2010]

5 minutes ago, Gracefulmuse said:

I can't answer any of this. But I am totally surprised you don't have answers yet, this is normally the kind of question @mathmanican would jump on. 

It is Sunday and the forums are slow on Sundays. 

I don't have a lot of the specific answers to @Mathgeekburch's questions but in general, here's how tempshift plates work.

1) They normalize the heat energy between themselves, the tile they are built in, and the 8 surrounding tiles. 

2) They do not exchange heat energy with other tempshift plates so a checkerboard pattern of every other tile is the best layout. 

3) They do not exchange heat with other buildings, but if the gas/liquid sharing the tiles has a good thermal capacitance, it will indirectly cool/heat the buildings decently fast. 

4) They do interact with tiles (both natural and dupe built) and doors, allowing fun heat transfer through walls and between rooms. 

[Gus Smedstad]

Tempshift plates only interact with liquids, gasses, and adjacent tiles. The "adjacent tiles" bit is often fairly important.

At one point tempshift plates ignored the "insulated" flag on tiles, and would rapidly heat up insulated tiles as a result. I don't believe this is still true.

Since tempshift plates interact with 9 tiles, their own tile and the 8 adjacent ones, they tend to exchange heat at a x9 rate compared to a basic tile. There's nothing special about the center tile, they're all treated the same.

Falling liquids / debris do not interact with the environment in any way until they reach the destination tile. This is something that's significant in some oil boiler builds, that oil being dropped into the boiler goes directly to the tile below the vent without heating.

I seem to recall reading that gasses exchange heat more readily with solids than the thermal conductivity would lead you to believe. There's a hidden x25 multiplier, is what I recall, but I don't have a solid source, just a vague recollection of something someone else said in passing.

Note that I'm saying "solids," and tempshift plates count as solids. So a row of tempshift plates adjacent to a wall of tiles ends up acting as a radiator. A tile by itself exchanges heats with the gas in the adjacent square. A tile next to a row of tempshift plates interacts with 3 tempshift plates, each of which is exchanging heat with 6 gas tiles. In effect the "surface area" of each tile goes up a factor of 18.

I think that's how it works. Proper science would involve experiments in a sandbox.

In practice I'm often just interested in the high mass of tempshift plates, which at 800k is 8x that of an ordinary tile. A granite tempshift plate in a cool steam geyser chamber can be substantially cooled between eruptions, and then in turn cool a great deal of steam because the high mass requires a lot of energy to heat.

[Mathgeekburch]

Quote

Note that I'm saying "solids," and tempshift plates count as solids.

I was under the impression that tempshift plates counted as buildings like pipes, and that the gas get the 25x multiplier with solid blocks/tiles and not solid buildings. 

I think the radiator effect has nothing to do with the 25x multiplier; but it still deals with the 18x surface area, except that conduction through that 18 surface area doesn't get the 25x multiplier.  A temp shift plate with a decent conductivity typically is going to have at least 9 times the conductivity value of most solids.  I think heat calculations are calculated with geometric means now, so 9 times the conductivity would be about 3 times the heat transfer so that helps off set on having the 25x multiplier. There is also the added advantage that it is heating up gas 2 tiles away from the wall, instead of 1 tile, gases have a low conductivity so heat transfer from the wall to the gas 1 tile away, will have low transfer to the gas 2 tiles away, the tempshift plate avoids that.  Plus I think gases don't as flow as well by walls, so you might get more motions from the gas 2 tiles away which might effect heat transfer.  I don't know too much about heat gets moved through circulation so I don't know how important that is.
 

49 minutes ago, Gus Smedstad said:

Falling liquids / debris do not interact with the environment in any way until they reach the destination tile.

Ok that is good to know.  Scratches some ideas I had for heat transfer systems though.

Quote

In practice I'm often just interested in the high mass of tempshift plates, which at 800k is 8x that of an ordinary tile.

Aren't normal tiles 200kg, so only 4x?  Still will be less mass than a natural solid block which I think is about 1800kg for granite.

54 minutes ago, Gus Smedstad said:

A granite tempshift plate in a cool steam geyser chamber can be substantially cooled between eruptions, and then in turn cool a great deal of steam because the high mass requires a lot of energy to heat.

Wouldn't dirt temp shift plate be even better for a higher thermal mass?  Or do you need granite's higher thermal conductivity?

[beowulf2010]

5 minutes ago, Mathgeekburch said:

Aren't normal tiles 200kg, so only 4x?  Still will be less mass than a natural solid block which I think is about 1800kg for granite.

Wouldn't dirt temp shift plate be even better for a higher thermal mass?  Or do you need granite's higher thermal conductivity?

Remember, tempshift plates also get the mass of the squaresit occupies and thus can be significantly more than the 200kg of a normal tile. 

Yeah, dirt tempshift plates are the best for holding thermal energy for stability purposes. Granite is good for spreading heat energy around. Diamond and refined metal (especially aluminum) plates are best for transferring heat energy. 

[Lilalaunekuh]

2 minutes ago, beowulf2010 said:

Yeah, dirt tempshift plates are the best for holding thermal energy for stability purposes.

I always look at plastic when it comes to thermal stability, but most of the time I settle for dirt aswell.

=> How do you guys think about plastic tempshift plates ? [Plastic offers a higher capacity than dirt and has still a wider temperature range than ice.]

[Mathgeekburch]

2 minutes ago, beowulf2010 said:

Remember, tempshift plates also get the mass of the squaresit occupies and thus can be significantly more than the 200kg of a normal tile. 

Yeah, dirt tempshift plates are the best for holding thermal energy for stability purposes. Granite is good for spreading heat energy around. Diamond and refined metal (especially aluminum) plates are best for transferring heat energy. 

Yeah, dirt tempshift plates are the best for holding thermal energy for stability purposes.  Provided you are withing the 0c to 300c temp range.  Above 300c igneous rock is the best if you don't need the conductivity.

Below 0c,  Ice actually is better.  And below -17, brine ice is even better than ice.  I think I might save whatever brine and salt water on my map for freezing into brine ice later if I ever plan on having a cold sink.  

[beowulf2010]

4 minutes ago, Lilalaunekuh said:

=> How do you guys think about plastic tempshift plates ? [Plastic offers a higher capacity than dirt and has still a wider temperature range than ice.]

Honestly? I forgot that you can make tempshift plates out of plastic.  The 160C melting point makes it less useful in warmer applications but, yeah, they're a good alternative to dirt when temps are under 150C.

[Mathgeekburch]

7 minutes ago, Lilalaunekuh said:

I always look at plastic when it comes to thermal stability, but most of the time I settle for dirt aswell.

=> How do you guys think about plastic tempshift plates ? [Plastic offers a higher capacity than dirt and has still a wider temperature range than ice.]

It is an interesting consideration.  Although it has a crippling low conductivity.  Also it is somewhat expensive and if it does melt, you need to dump it into a sour gas boiler.

The only tempshift plate with lower conductivity is insulation, which I am pretty sure is just there to be really expensive drywall that has no "effective" thermal mass, since if you don't transfer any heat, what does you thermal mass matter anyways?

[Nebbie]

The thing to understand is that buildings (machines, pipes, etc.) in ONI only transfer heat with the material of each tile they occupy, which means if you've got a 1x3 building with the bottom inside a tile, middle inside a liquid, and top inside a gas, it transfers heat with all three, but not surrounding tiles (most notably, your robominers need a foundation but do not actually touch it or pipes behind them).
Tempshift plates are just a 1x1 building on the drywall layer that counts as 3x3 for heat exchange purposes.

Gases and liquids have boosts to thermal transfers. This is to simulate complicated interactions ONI can't do accurately, especially for liquids.

As for instantly melting ice tempshift plates in water, I believe this is related to a bug in the current version where you can drop water onto hot abyssalite and it'll instantly cool the abyssalite and turn to steam. My suspicion is that the phase change rules are screwed up so that the game is detecting it should finish a heat exchange with a phase change before the heat has actually had time to transfer between the two materials.

[Lilalaunekuh]

16 minutes ago, Mathgeekburch said:

It is an interesting consideration.  Although it has a crippling low conductivity. 

That´s my main question, is it still worth building a tempshift plate if the buffer has to much of a delay. (Low permeability)

There are just a few cases in which the temperature range wouldn´t allow me to use plastic but still allow me to use dirt.

=> [Not considering the effort to produce the material] Is plastic for pure thermal stability superior ?

Spoiler

Ice/brine tempshift plates are to "fragile" for my taste, a little power shortage/heat surge and I have to mop/rebuild my whole setup.

 

 

 

23 minutes ago, Mathgeekburch said:

The only tempshift plate with lower conductivity is insulation, which I am pretty sure is just there to be really expensive drywall that has no "effective" thermal mass, since if you don't transfer any heat, what does you thermal mass matter anyways?

Keep in mind: Insulation isn´t always a perfect insulator. [Heat transfer for non insulated buildings uses an average.]

=> "It is an interesting consideration.  Although it has a crippling low conductivity. "

[Mathgeekburch]

3 minutes ago, Lilalaunekuh said:

That´s my main question, is it still worth building a tempshift plate if the buffer has to much of a delay.

I'd normally say yes in the case of temp shift plates with conductivity down to 1, simply due to the fact that you still have heat transfer with 9 cells..

But 0.15 is pretty low; lower than hydrogen gas.  It might it still would work, although I haven't tested it before.  If homogeneity of temperature is still something you want in whatever area it is occupying, you could try mixing it with other temp-shiftplates with higher conductivity..

The low conductivity of the plastic plate wouldn't matter that much if whatever it was in contact with had a decent conductivity.  Might be a problem for gases though since gases typically have the lowest conductivity, so you might want to consider placing a single metal tile in the center of 8 plastic temp-shift plates.  Gases conduct well with solid blocks, and the metal tile should easily conduct more than enough into the plastic temp-shift plates.  

 

2 hours ago, Gus Smedstad said:

In practice I'm often just interested in the high mass of tempshift plates, which at 800k is 8x that of an ordinary tile.

Yeah you were correct on the 8x for at least metal tiles (which are 100kg).  Normal tiles are 200kg though. 

22 minutes ago, Lilalaunekuh said:

Keep in mind: Insulation isn´t always a perfect insulator. [Heat transfer for non insulated buildings uses an average.

I believe geometric averages.  So a perfect insulator would still be a perfect insulator.  Although I am certain the value isn't actually 0 for insulation temp shift plates, just really really really low.  In which case it will still conducting a bit.  Although the amount of time it will take for heat transfer is probably so low, and its thermal mass so high, that you'd probably wouldn't notice such a change except over a really long period of time.
 

The exception would probably be if it was in contact with a metal tile.  Which would produce an interesting question would an insulated temp-shift plate surrounded by 8 metal tiles actually act like a unique high temp heat sink?

[Coolthulhu]

1 hour ago, Nebbie said:

As for instantly melting ice tempshift plates in water, I believe this is related to a bug in the current version where you can drop water onto hot abyssalite and it'll instantly cool the abyssalite and turn to steam.

This is almost certainly not a bug due to how specific the mechanic is, and is also very unlikely to be related to tempshift plates melting.

The "sweating"/("bubbling"/"flaking") mechanic represents surface level phase change. It happens in exactly 5kg amounts. Without it, you couldn't have ice drip water until the entire 1000kg block of ice melted instantly, even if the ice itself was surrounded by scalding oxygen.

Ice tempshift plates can melt in water for a much simpler reason: the ice exchanges heat with water even before the construction is finished. If the ice is above 0 at the moment construction finishes, it melts instantly.

[Mathgeekburch]

4 minutes ago, Coolthulhu said:

Ice tempshift plates can melt in water for a much simpler reason: the ice exchanges heat with water even before the construction is finished.

Do all tempshift plates exchange heat before they finish construction, or is it unique to this specific interaction?

5 minutes ago, Coolthulhu said:

If the ice is above 0 at the moment construction finishes, it melts instantly.

So does this mean if I get a cold enough ice to make the plate out of, it won't instantly melt if I build it underwater?  

[Coolthulhu]

50 minutes ago, Mathgeekburch said:

Do all tempshift plates exchange heat before they finish construction, or is it unique to this specific interaction?

It's not the plate that exchanges heat, just the material. When the material is placed in a construction designation, it "exists" at that spot until construction is finished. This can be easily seen with nearly-molten metal straight from the volcano, when used to build wires: the dupes will most likely be scalded by the gases heated by the hot metal in designations.

56 minutes ago, Mathgeekburch said:

So does this mean if I get a cold enough ice to make the plate out of, it won't instantly melt if I build it underwater?  

Yes, if the ice is cold enough, it might survive the build time, especially if the water is cold too. Supercooled water (-2C water is possible) will not melt the ice plate.

[Lifegrow]

4 hours ago, Gus Smedstad said:

Falling liquids / debris do not interact with the environment in any way until they reach the destination tile. This is something that's significant in some oil boiler builds, that oil being dropped into the boiler goes directly to the tile below the vent without heating.

This isn't entirely true - there are 2 kinds of falling liquids.

To test this you can build two vents being fed into a super-hot-gas-filled room. Under one of the vents place a mesh tile. You should get very different results.

[Mathgeekburch]

44 minutes ago, Coolthulhu said:

Yes, if the ice is cold enough, it might survive the build time, especially if the water is cold too. Supercooled water (-2C water is possible) will not melt the ice plate.

If the water is too cold, it wouldn't test the whether the hypothesis is true or not.  As it is already known that the water needs enough thermal energy to melt the ice; and it would need to be at the temperature to melt the ice.

What I just need is super cooled ice, like -200C, and a dupe with 26 or higher in construction and need to get it such that they build it immediately.  The water could be fairly chilly, maybe 10 C, but I'd need a large pool of it so I know it has the thermal energy needed.

If it isn't instant heat transfer, and just heat being transferring during construction, this should have a decent chance of testing whether that is true.

[Gus Smedstad]

2 minutes ago, Lifegrow said:

To test this you can build two vents being fed into a super-hot-gas-filled room. Under one of the vents place a mesh tile. You should get very different results.

Since you don't spell it out, it sounds like you're saying that falling liquids stop when they hit a mesh tile, and then start falling again if there's a path through the mesh tile. Though visually, it's always looked like the liquid was falling straight through to me.

[Lifegrow]

12 minutes ago, Gus Smedstad said:

Since you don't spell it out, it sounds like you're saying that falling liquids stop when they hit a mesh tile, and then start falling again if there's a path through the mesh tile. Though visually, it's always looked like the liquid was falling straight through to me.

I thought giving you the means to test it for yourself would get you to edit your post, not quote me and make a new one  

Too many assumptions and not enough testing on the forums nowadays.

Less sandbox, more debug please  

[Mathgeekburch]

Looking back at an older forum post.  It seems I am very likely gravely wrong in my understanding of the heat transfer equations, at least involving buildings.
 

Since I believe temp-shift plates are considered buildings, I believe the equation that is supposed to be used is:

ΔQ = Δt * k'1 * k2 * Chot * (T1 - T2) * 0.005 * x


The part that I gravely misunderstood was Chot. Which from what I can tell is that it is proportional to the thermal mass of what is hotter.  

Now this forum was from 2017, so the heat equations are likely fairly different from what they are now.  But if the current equation still has a Chot as a multiple for heat transfer; it would mean a lot.  It would certainly explain the insta melting when placed in liquids (as liquids should have a really large chot if in a full tile).  And how well they are at being radiators for solids.

But it would also have big implications for when using temp shift plates to resist changes in temp.  When the temp shift plate is hotter than the medium, the heat transfer will be greater for a tempshift plate with greater shc and thermal mass of medium has no impact on the amount of heat transferred, but when the shiftplate is colder than the medium; the heat transferred to it will be proportional to the heat mass of the medium and the shc as no impact on the amount of heat transferred to it.

Can anyone do a quick test to see if Chot is still part of the heat equation?

[nakomaru]

Tempshifts are merely 3x3 buildings.

Tempshifts have 160kg effective thermal mass, compared to a normal tile's 200kg. This is because buildings count for 1/5th their mass.

They have normal building interactions when buried in solid (200x heat transfer) or submerged in fluid (1x heat transfer).

When utilizing them for thermal capacity, it is best to cover as much area as possible.

10 minutes ago, Mathgeekburch said:

Can anyone do a quick test to see if Chot is still part of the heat equation?

It is. You can heat up neutronium with tempshifts, but not cool them.

[Mathgeekburch]

11 minutes ago, nakomaru said:

It is. You can heat up neutronium with tempshifts, but not cool them.

How does that even work, isn't the conductivity supposed to be zero?  Like actual zero, not almost zero?

Don't building heat transfers depend on both the conductivity of the building and the conductivity of the medium?

What does heating something with zero heat capacity even mean?

I guess if it  does make some sense if it has both a really really really low heat capacity, and a really really really low conductivity.    The super low conductivity means the heat transfer is almost zero; but the super low capacity means that almost zero heat transfer does change the temp.

 

18 minutes ago, nakomaru said:

They have normal building interactions when buried in solid (200x heat transfer) or submerged in fluid (1x heat transfer)

I haven't seen someone mentioned these reactions for buildings before.  Can you refer me to a forum/source that I can learn more about this from?

[Mathgeekburch]

42 minutes ago, nakomaru said:

Tempshifts have 160kg effective thermal mass, compared to a normal tile's 200kg. This is because buildings count for 1/5th their mass.

Does that mean I'd actually create heat if I take -100c ice, turn it into a shift plate, heat it up to melting point.  As it would take 5 times the cooling to get it back down into -100c ice?

47 minutes ago, nakomaru said:

It is. You can heat up neutronium with tempshifts, but not cool them.

Also this doesn't completely answer my question.  I am more curious about Chot increasing with mass, than it increasing with SHC.  As the former has huge implications for heating buildings with gas (and also liquids and solids).

 

[nakomaru]

Yes, you can create and destroy heat by building and deconstructing buildings at different temps. (Independent of the 15C/45C temperature clamping.)

The 200x/1x factors are parsimonious interpretations of how other buried solids and buildings work. Mathman has talked about burial in solid being 200x. I'd have do an experiment to prove them with confidence.

I believe the unidirectionality of heating neutronium shows that Chot is operational (I understand it doesn't tell you about the scaling, sorry.) If that formula is correct, it must have a non zero conductivity. The fact that it can have a varying temperature probably disproves the database claims of 0 SHC, so TC is probably similarly imprecise.

If it's not correct, we could do some testing to see if tempshifts are a special case or not by comparison with the other building that is bigger than its occupied space.

15 hours ago, Mathgeekburch said:

temp-shift plate made out of insulation? Presumably it only acts like a drywall with effectively zero heat capacity (your heat capacity doesn't matter if you don't absorb or emit heat).

Tempshifts from insulation conduct heat just fine when buried in (i.e. adjacent to) solids, thanks to a large multiplier and probably non zero TC. This is one way to produce tungsten.