Insulation Tiles gain heat quickly

[lemodrops]

I have a bunch of insulated tiles that are gaining temperature faster than normal. From the images you can see that the tile raised 1 K in roughly 90 in-game seconds. Is this a known issue, did insulation tiles change or should I report this as a bug?

 

[Gurgel]

Looks like a bug to me...

Update: Forget that, it is the temp-shift plates as others already said. I missed that you did not keep them 1 tile way from the walls.

[qqqqqqqqqqqq]

this is tempshift plates mechanics
they have an area of 3x3
remove the ones that touch the insulated tile

[OxCD]

18 hours ago, qqqqqqqqqqqq said:

this is tempshift plates mechanics
they have an area of 3x3
remove the ones that touch the insulated tile

I don't agree with you about this mechanism beeing normal. Sounds like a bug for me also.

EDIT: my bad. It appears that you was right. Wasn't expecting tempshift to break the "insulated" rule.

 

[Neotuck]

Igneous rock is not a good insulator for dealing with extreme heat like steam, switch to ceramic

Also the tempshift plates will exchange heat with nearby tiles, including insulated tiles.  Only space insulation is immune from this mechanic

[DonDegow]

As mentioned you should remove tempshift plates in a 1 tile range of any insulated tiles. However you should also be mindful of liquid bridges overlapping the steam room and insulated tiles, same for automation gates.

[TheMule]

7 hours ago, OxCD said:

I don't agree with you about this mechanism beeing normal. Sounds like a bug for me also.

It's normal in the sense it's well known. My understanding is that shiftplates behave like 3x3 buildings. in this case, they act as if entombed. Thermal transfer for entombed objects gets a bonus.

[OxCD]

I've never spotted such fast transfer between a plate and an insulated tile.

I know the area of a plate, but it doesn't deny the fact that insulated tile should be the TC taken into account to calculate the actual transfer.

[BaZERGer]

The speed of both the tile and the plate's conductivity matter, a single aluminum plate can shift even ceramics at a notable speed.

And, though not relevant to this case case, tempshift plates at the walls makes the turbine less effecient, since it's pushing heat into the tiles.

[Badpip]

You can even cool down or warm up neutronium with tempshift plate

[SamLogan]

18 hours ago, lemodrops said:

I have a bunch of insulated tiles that are gaining temperature faster than normal. From the images you can see that the tile raised 1 K in roughly 90 in-game seconds. Is this a known issue, did insulation tiles change or should I report this as a bug?

 

It's not a bug, it's only maths :

2 AT = 1170120 DTU/ sec

To increase by 1°C 400Kg of igneous rock, you need to generate 400 000 DTU. So you gain 2°C / sec, and then 180°C in 90 secondes.

But given you have also steam, tempshift and other igneous rock, you didn't reach this temperature yet, but your 370K are totally normal.

[OxCD]

39 minutes ago, SamLogan said:

It's not a bug, it's only maths :

2 AT = 1170120 DTU/ sec

To increase by 1°C 400Kg of igneous rock, you need to generate 400 000 DTU. So you gain 2°C / sec, and then 180°C in 90 secondes.

But given you have also steam, tempshift and other igneous rock, you didn't reach this temperature yet, but your 370K are totally normal.

The question is not the amount of DTU needed to warm up insulated tiles (SHC relevant only), but the speed at which the transfer is operating between those elements (both SHC & TC taken into account).

[SamLogan]

26 minutes ago, OxCD said:

The question is not the amount of DTU needed to warm up insulated tiles (SHC relevant only), but the speed at which the transfer is operating between those elements (both SHC & TC taken into account).

[lemodrops]

I have been using this setup for a while and have never had this happen before. Also the aquatuners are not at 100% uptime, probably around 30%, I haven't checked. The steam is also losing its temperature very quickly too, much faster than before. I guess I've been playing with a glitch until now?

[OxCD]

32 minutes ago, SamLogan said:

Mmmh did I say I wasn't trusting the number you brought ? ^^

Oni assistant isn't taking into account any TC excepting the one from the element (raw status) that has to be cooled.

Raw Igneous TC is 2, insulated tile with Igneous TC is 0,02. So the time needed to heat up 400kg of the first one cannot be the same as the time needed to heat up 400kg of the second one. Only DTU amount is the same because SHC is the same.

TC is implied in both gaining some heat, and giving some.

I'm sorry but I think your calculation doesn't apply in this case, without including additionnal parameters.

[KittenIsAGeek]

Because of the comparitively low TC of insulated ceramic, you'll find that your steam will always be hotter than your insulated tile.  However..

1. The TC of insulated igneous is more than 30 times greater than the TC of insulated Ceramic.  Also, since both numbers are greater than 0, energy will transfer provided there's enough of it. 
2. Igneous has a SHC of 1, meaning it takes comparatively low heat energy to raise its temperature.
3. Water and Steam have 4 times the heat capacity per gram of igneous, meaning that it holds a LOT of energy.
4. Your turbine shows fully green, meaning that the steam is around 180c to 190c, while your insulated tiles are at 97c.
5. Temp-shift plates break the rules, meaning the thermal transfer math takes the TC of the log average of the tempshift plate and the insulated tile, rather than the "insulated" rules which takes only the lowest TC.  So your actual TC is probably somewhere around 1 or 2 because of the plates.

Lets assume that the TC of the equation is actually 0.02 as shown on the insulated igneous stats.  Then the math says heat will transfer at a rate of 0.02(DTU/m*s/C) W.  Lets also assume that (m*s) is 1 (Its actually larger).  Then we can transfer 0.02 DTU/C.  We've got a thermal differential of about 100c, so we can transfer 2 DTU per second to the insulated tile.  This is enough to raise 1g of insulated igneous by 2c, or 2g by 1c.

Without the tempshift plate in the equation, then, with a 100c delta, it will take about 200 seconds to raise the temperature of an insulated igneous tile by 1c.  However, if (m*s) is 2, then it only takes 100 seconds.  You saw a 1 degree shift in 90 seconds, which seems entirely reasonable to me.

[lemodrops]

33 minutes ago, KittenIsAGeek said:

Because of the comparitively low TC of insulated ceramic, you'll find that your steam will always be hotter than your insulated tile.  However..

1. The TC of insulated igneous is more than 30 times greater than the TC of insulated Ceramic.  Also, since both numbers are greater than 0, energy will transfer provided there's enough of it. 
2. Igneous has a SHC of 1, meaning it takes comparatively low heat energy to raise its temperature.
3. Water and Steam have 4 times the heat capacity per gram of igneous, meaning that it holds a LOT of energy.
4. Your turbine shows fully green, meaning that the steam is around 180c to 190c, while your insulated tiles are at 97c.
5. Temp-shift plates break the rules, meaning the thermal transfer math takes the TC of the log average of the tempshift plate and the insulated tile, rather than the "insulated" rules which takes only the lowest TC.  So your actual TC is probably somewhere around 1 or 2 because of the plates.

Lets assume that the TC of the equation is actually 0.02 as shown on the insulated igneous stats.  Then the math says heat will transfer at a rate of 0.02(DTU/m*s/C) W.  Lets also assume that (m*s) is 1 (Its actually larger).  Then we can transfer 0.02 DTU/C.  We've got a thermal differential of about 100c, so we can transfer 2 DTU per second to the insulated tile.  This is enough to raise 1g of insulated igneous by 2c, or 2g by 1c.

Without the tempshift plate in the equation, then, with a 100c delta, it will take about 200 seconds to raise the temperature of an insulated igneous tile by 1c.  However, if (m*s) is 2, then it only takes 100 seconds.  You saw a 1 degree shift in 90 seconds, which seems entirely reasonable to me.

Thanks! That explains it very well. Weird that m > 1. Overall the temp rate of change was new with this recent update for me and I was wondering if something happened(bug).

[KittenIsAGeek]

Just now, lemodrops said:

 Weird that m > 1. 

Its (m*s) that is greater than 1.  S is the cross-section surface area of thermal contact.  m is a material type constant (building, floor, debris, etc).  As far as I know, m*s is never actually separated out in ONI.  For debris, I think the value of m*s is 1, but for most other transfers its 2.  There was a post a long while back that I'm completely failing to find now where someone did experiments to figure out what it was.

[Gurgel]

18 hours ago, qqqqqqqqqqqq said:

this is tempshift plates mechanics
they have an area of 3x3
remove the ones that touch the insulated tile

Ah, yes, you are right. I keep these 1 tile away from any insulated walls, so I never thought to look for that possibility. Temp-shift plates will even heat Neutronium. Took me a while figuring that out way back.

The thing is that the insulation of a tile is effectively on the outside. A temp-shift plate bypasses that in its 3x3 area and goes directly to the inside of the tiles in that radius. Hence instead of the thermal conductivity of the tile, you get that of the temp-shift plate.

[OxCD]

1 hour ago, KittenIsAGeek said:

1. Temp-shift plates break the rules, meaning the thermal transfer math takes the TC of the log average of the tempshift plate and the insulated tile, rather than the "insulated" rules which takes only the lowest TC.  So your actual TC is probably somewhere around 1 or 2 because of the plates.

Ok I do accept this explanation. It feels really "ONI" indeed, so makes sense to me. I didn't expect this at all, despite hundreds of thermal setups I've already done...

However, it sounds a bit awkward for me. From my point of view, I was expecting insulated tile to keep its insulated properties in this case.

[BaZERGer]

Insulated tiles are still slowing it, had it been regular tiles, it would probably have equalized in a few seconds.

[bobucles]

Insulated tiles are great when nothing touches them. Anything that comes in thermal contact with an insulated tile provides another temperature calculation, and thus will rapidly increase the transfer rate.

Do embedded conveyor items still have the 200x buff on temperature transfer? That's the biggest one.

[psusi]

On 3/13/2020 at 3:05 AM, Neotuck said:

Only space insulation is immune from this mechanic

It is?  If tempshift plates can work on neutronium I'm pretty sure they work on Insulation.

 

[Neotuck]

Just now, psusi said:

It is?  If tempshift plates can work on neutronium I'm pretty sure they work on Insulation.

The thermal conductivity is so low that it keeps rounding down to 0

I once tested it by separating LOX and magma, even with temp shift plates the insulated tile stayed at an even 20C and nether the LOX or the Magma changed temps.

Then again I only ran the test for 10 cycles so it's also possible that the thermal transfer is just very very VERY slow

[psusi]

1 minute ago, Neotuck said:

The thermal conductivity is so low that it keeps rounding down to 0

I once tested it by separating LOX and magma, even with temp shift plates the insulated tile stayed at an even 20C and nether the LOX or the Magma changed temps.

Then again I only ran the test for 10 cycles so it's also possible that the thermal transfer is just very very VERY slow

Isn't the conductivity of neutronium actually zero?