Cooling in vacuum needs some attention

[TunderLock]

TLDR: Heat conduction between touching buildings in vacuum doesn't work but conduction between 300C debris and buildings does.

Building on the new changes to light penetration, I wanted to update the automated solar farm for two improvements:

1. Dropping from a 3-tier to 2-tier approach and running the panels in a vacuum by cooling from below. 
2. Adding some cooled auto sweepers to pick up the refined iron.

Neither of these two worked because of the lack of heat conduction between objects when one of them is in vacuum. My implementation is below and I'll talk through these two additions. For the basics of the base automated solar farm, see the post linked above. 

 

Cooling the panels from below
The solar farm was originally a three tier approach (blastdoors-glass-glass), but with the changes to light penetration (glass removes 10% of the lux), I wanted to remove the second layer of glass. Doing so meant the panels would run in a vacuum. 

The panels are built on and touching metal tiles that are cooled from below by hydrogen gas. There is zero heat conduction between the metal plates and the solar panels. You can even delete the metal tiles and have the hydrogen gas directly in contact with the solar panels and there is still zero heat conduction. This struck me as really counter-intuitive.

For those wishing to build a solar farm, currently I would recommend building a 3-tier farm, then once you have cooled the panels, it's safe to remove the second layer of glass and run them in a vacuum after. Building a 2-tier farm immediately is possible, but risky, if one 20 tonne bit of rubble falls on the panels during construction, you'll have to build a 3-tier anyway just to cool them. 

Cooling the auto-sweepers
The image also shows an auto-sweeper implementation designed to pick up the refined iron and take it away for cooling. The iron is +300C, the sweepers can't survive that as they heat up handling it over time, so they need active cooling. I ran a chilled oil loop behind them, including radiant pipes directly underneath the sweepers and there was zero heat conduction. Building dry-wall or temp-shift plates also did nothing. 

I can not see a way to make auto-sweepers work in a solar farm at the moment. 

What does work in a vacuum?
So far, all I can that works in a vacuum is heat conduction between debris and buildings. This is shown when the glass layer heats up with meteor debris on it, and the auto-sweeper arms heating up as they pickup the debris. 

The fact that this works, but neither of the top two attempts work struck me as strange. Clearly I don't understand the math behind heat conduction (has anyone got a link to the latest confirmed formulas? Can we de-compile and look easily?). But I feel that something isn't right if debris in vacuum can transfer heat, but metal tile from below or pipes underneath can't. 

Not sure how difficult of change it would be for the devs to handle heat conduction between touching buildings more intuitively when the buildings are in vacuum. I think building in the vacuum could be even more enjoyable/challenging as we can't rely on radiation into an atmosphere to cool, but I feel heat conduction between touching buildings needs to work for this to be true.

[Capsup]

Doesn't liquids conduct temperature in a vacuum? I know that's how Sevio cools his lossless heat forge: 

He also uses a sweeper in there by only submerging 1 tile in a liquid.

 

[PhailRaptor]

1 hour ago, Capsup said:

Doesn't liquids conduct temperature in a vacuum?

But if there's liquid, it's not a vacuum.  It's a liquid.  Same thing with gas.

[Capsup]

11 minutes ago, PhailRaptor said:

But if there's liquid, it's not a vacuum.  It's a liquid.  Same thing with gas.

But why is the actual vacuum important? Add a bit of liquid at the bottom to conduct between your metal tile and the solar plant and then you have the cooling that you need for the solar plants, right? Everything above is still a vacuum and thus the water won't conduct through that. The debris is still out of reach, so the heat from there isn't a problem.

[abud]

1 hour ago, TunderLock said:

running the panels in a vacuum by cooling from below

Panels doesn't need cooling because they didn't produce heat. My solar panels working for few hundreds cycles without cooling.

[Sevio]

Which of a solar panel's tiles are checked for light level to determine the amount of power generated?

[abud]

8 minutes ago, Sevio said:

Which of a solar panel's tiles are checked for light level to determine the amount of power generated?

All those base tiles, If some not receiving light their output wattage is reduced.

 

Edit: When I think again about it @Sevio, it is possible that actually in the code they check tile just above base tiles. To prevent generating power when the light is from below. I'm not sure anymore

 

[Sevio]

That makes liquid cooling problematic, but if you're okay with taking a small performance hit you could pour a layer of water on your solar farm, then mop up so that only 1 corner of each panel is covered in water. And then pray that your panels don't go over 100 C when regolith falls onto them by accident.

[KittenIsAGeek]

4 hours ago, Capsup said:

But why is the actual vacuum important? Add a bit of liquid at the bottom to conduct between your metal tile and the solar plant and then you have the cooling that you need for the solar plants, right? Everything above is still a vacuum and thus the water won't conduct through that. The debris is still out of reach, so the heat from there isn't a problem.

Unless you build base walls and seal the room, you will not be able to put a bit of liquid at the bottom to conduct heat -- it will be sucked out into space.  The vacuum itself isn't important -- rather its the mechanics of building at the top of the asteroid.

[Sevio]

You don't need to seal the room off from space to have liquid on top of your solar panels, all you need is a row of drywalls to prevent the water from spilling out from the tile it occupies.

[heckubis]

the vacuum with liquid trick would be your best option just enough water to cool the solar panels not enough to flood them in a vacuum will cause the vacuum to cool the water and the water to cool the panels

4 minutes ago, Sevio said:

You don't need to seal the room off from space to have liquid on top of your solar panels, all you need is a row of drywalls to prevent the water from spilling out from the tile it occupies.

 

[_Q_]

4 hours ago, abud said:

Panels doesn't need cooling because they didn't produce heat. My solar panels working for few hundreds cycles without cooling.

So the panels don't require foundation, just like steam turbine. I wonder if that will change in the future.

[abud]

18 minutes ago, _Q_ said:

So the panels don't require foundation, just like steam turbine.

Panel base already solid, dupe can walk on it. Just like steam turbine.

[Capsup]

57 minutes ago, KittenIsAGeek said:

Unless you build base walls and seal the room, you will not be able to put a bit of liquid at the bottom to conduct heat -- it will be sucked out into space.  The vacuum itself isn't important -- rather its the mechanics of building at the top of the asteroid.

Pretty sure only tiles marked with "exposed to space" will delete liquid and gas. So just the bottom tiles of the solar panels will need to have drywall behind it, that should keep the water safe.

[martosss]

As I already replied in a bug report topic on the same issue, I'll copy my answer here:

2 hours ago, martosss said:

Here's a little theory about heat transfer, note that some things are outdated, but the general idea should be the same - cell-cell vs cell-building vs cell-entity!

1. is perfectly normal - tile-tile transfer, both materials non-vacuum
2. works, because the sweeper actually "touches" the materials, so there is direct contact, this is building-entity transfer(same for pipe-content)
3. Ah, I did this last, now I remember - buildings transfer heat only to their cells => vacuum = no heat transfer behind them.
4. I think this shouldn't work, because vacuum will block the heat transfer between the 2 buildings - Basically they transfer heat to the cell they occupy, but since it's vacuum, no heat transfer happens. So this is normal.

Now I'm thinking how to actually cool those panels/sweepers. I'm levitating around tempshift plates, but they might still not work, since they're basically a building and transfer heat to the cell(vacuum => no transfer).

Another idea might be using liquid. I'm not sure how space works(if liquid will flow down, or up, or disappear), but assuming it falls down, you can drip water on the panels. This should mean that the tile behind the building contains water instead of vacuum and heat transfer will start.

Sweepers might be a bit harder, though, as they're "floating" around, so I'm not sure how water dripping on them works.

Basically heat conduction works "as expected" - when there's direct contact - there's heat transfer. When there isn't - no heat transfer in vacuum.

My point is that although some formulas might have changed(e.g. log average thermal conductivity instead of taking MIN value), The general logic should be the same:

• (vacuum tile) next to (non-vacuum tile) - no transfer;
• vacuum tile behind building - no transfer(as buildings transfer heat only with cells directly behind them, not beneath them)
• building -building doesn't work in vacuum, as there isn't such a type of transfer - buildings only transfer heat with cells behind them.(at least that's what I gather from the decryption thread)
• building-entity works - in the case of pipe-content and in the case of sweeper-content, since the content is "inside" and "touches" the building - there is physical contact.
• Cell - entity works(kind of) - any debris will transfer heat to the cell they're occupying, or the cell below them with ¼ coefficient in case of vacuum.

So far everything in the decryption thread checks out with your scenario.

[TunderLock]

3 hours ago, martosss said:

As I already replied in a bug report topic on the same issue, I'll copy my answer here:

My point is that although some formulas might have changed(e.g. log average thermal conductivity instead of taking MIN value), The general logic should be the same:

• (vacuum tile) next to (non-vacuum tile) - no transfer;
• vacuum tile behind building - no transfer(as buildings transfer heat only with cells directly behind them, not beneath them)
• building -building doesn't work in vacuum, as there isn't such a type of transfer - buildings only transfer heat with cells behind them.(at least that's what I gather from the decryption thread)
• building-entity works - in the case of pipe-content and in the case of sweeper-content, since the content is "inside" and "touches" the building - there is physical contact.
• Cell - entity works(kind of) - any debris will transfer heat to the cell they're occupying, or the cell below them with ¼ coefficient in case of vacuum.

So far everything in the decryption thread checks out with your scenario.

This is great, I was hoping someone had dug into the functions. 

Facinated that there is no building-building interaction. Seems like everything goes through the cell, which is why vacuum is such an issue. 

I'd really like to see a:

If cell == vacuum then call building-building transfer. But seems like a fundamental change just add cooling in a vacuum. 

 

Off topic. Is there a tute I can follow to decompile? I have experience in ruby and python by nothing in C and its derivatives. 

[PhailRaptor]

5 hours ago, heckubis said:

the vacuum with liquid trick would be your best option just enough water to cool the solar panels not enough to flood them in a vacuum will cause the vacuum to cool the water and the water to cool the panels

That's the thing though.  Vacuum isn't cold because it's cold.  It's cold because there is no medium present to actually hold the heat.  Heat cannot exist without something to act as a bucket to hold it.  Solids, liquids, and gases can do this.  The absence of these things cannot.  Vacuum cannot cool.

The issue is that the mechanics surrounding heat transfer make cooling objects in vacuum problematic, because there is no medium through which heat can be exchanged.  If they are in vacuum, solids in contact with one another cannot exchange heat (which is wrong, but a necessary simplification for the game).

[Sevio]

4 hours ago, martosss said:

building-entity works - in the case of pipe-content and in the case of sweeper-content, since the content is "inside" and "touches" the building - there is physical contact.

I have to note here that this is actually only the case for pipes, not other buildings. Resources inside buildings such as the metal refinery count as "entities" that are occupying the building's "base" tile. (bottom middle for metal refinery and glass forge) They exchange heat with the cell behind them or with the tile below them. If both of those are vacuum or insulated abyssalite, no heat is transferred to/from the building.

[trukogre]

43 minutes ago, PhailRaptor said:

That's the thing though.  Vacuum isn't cold because it's cold.  It's cold because there is no medium present to actually hold the heat.  Heat cannot exist without something to act as a bucket to hold it.  Solids, liquids, and gases can do this.  The absence of these things cannot.  Vacuum cannot cool.

The issue is that the mechanics surrounding heat transfer make cooling objects in vacuum problematic, because there is no medium through which heat can be exchanged.  If they are in vacuum, solids in contact with one another cannot exchange heat (which is wrong, but a necessary simplification for the game).

If you actually put water into vacuum irl, the liquid water would turn into water vapor and ice, so vacuum *can* cool, but ONI doesn't model pressure realistically, so yeah.

 

"The absence of these things cannot.  Vacuum cannot cool."  

More specifically, if you put a solid object into vacuum IRL (and not in direct sunlight of course), the solid object would radiate heat, estimated heat loss given by the Stefan-Boltzman formula, absolutevalue(dE/dt) = Area * constant *( (temperature) ^ 4), where the constant is 5.67 * 10^-8 (W/m^2/K^4).  approximate radiative heat loss for a human in vacuum would be around 500 watts, which would turn you into an icicle rather quickly.  Therefore, vacuum can cool, IRL.  Just as the Earth's atmosphere keeps us from exploding, the constant radiative bath of energy at the background temperature near 300 K found here on the Earth's surface keeps us from freezing to death super quickly.  Or, put another way, we are always radiating 500 Watts of heat, and vacuum doesn't per se cool us, it just doesn't keep us warm like we're used to.  Of course, ONI doesn't model radiative heat transfer, so yeah, none of this actually applies ingame, but I was replying to part of your post that seemed to be about IRL, mainly.

[Keylan]

1 minute ago, trukogre said:

If you actually put water into vacuum irl, the liquid water would turn into water vapor and ice, so vacuum *can* cool, but ONI doesn't model pressure realistically, so yeah.

Chaning phase is very diffrent from cooling, and water is quite a special case as change of phsa acctually is a thermal reaction changing tempreature. (but turning to ice due to pressuere would accutually inrease tempreature).

 

Water and other stuff in space cools down, but thats not due to vacum but due to radiation. And the lach of radiation in ONI is the mayor reason why this game need AETN and weezewort or other heatg deletion mechanics wich defy real world physics.

 

[trukogre]

4 minutes ago, Keylan said:

Chaning phase is very diffrent from cooling, and water is quite a special case as change of phsa acctually is a thermal reaction changing tempreature. (but turning to ice due to pressuere would accutually inrease tempreature).

 

Water and other stuff in space cools down, but thats not due to vacum but due to radiation. And the lach of radiation in ONI is the mayor reason why this game need AETN and weezewort or other heatg deletion mechanics wich defy real world physics.

 

Yes, you'll see that I edited in a section about radiation while you were replying  You are indeed correct about the consequences of the lack of radiation in ONI.

[heckubis]

4 hours ago, PhailRaptor said:

That's the thing though.  Vacuum isn't cold because it's cold.  It's cold because there is no medium present to actually hold the heat.  Heat cannot exist without something to act as a bucket to hold it.  Solids, liquids, and gases can do this.  The absence of these things cannot.  Vacuum cannot cool.

The issue is that the mechanics surrounding heat transfer make cooling objects in vacuum problematic, because there is no medium through which heat can be exchanged.  If they are in vacuum, solids in contact with one another cannot exchange heat (which is wrong, but a necessary simplification for the game).

funny we dont have giant chunks of liquid water zooming through our solar system we have comets made of "ice" and rock but mostly Ice. cold is an absence of heat not an abundance of cold the vacuum is preventing the transfer of heat thus being cold in the absence of direct sunlight. thermal energy converts to a radiant form that requires no medium if not we would receive no heat from the sun

[psusi]

So the game doesn't consider that under a vacuum, even freezing cold water will boil?

 

[blash365]

On 29.7.2018 at 12:28 PM, abud said:

Panels doesn't need cooling because they didn't produce heat. My solar panels working for few hundreds cycles without cooling.

How do you keep your battery bank from overheating?

[heckubis]

On 7/30/2018 at 12:40 PM, psusi said:

So the game doesn't consider that under a vacuum, even freezing cold water will boil?

 

look up the composition of asteroids and comets. notice how boiling water isn't one of the components. although they do melt fast under direct sunlight. on a table in a vacuum the water is still receiving loads of thermal radiation modifying state change conditions. Even without direct sunlight all things emit thermal radiation known as infrared. this constant leak of thermal energy is why things in space freeze so long as they have the mass to stay together