How should temperature management work?

[KittenIsAGeek]

5 hours ago, Nightinggale said:

Rapid water cooling: Right now if you dump a lot of ice into a lot of hot water and wait a bunch of cycles you still have ice in hot water. Ice would be much more useful if the heat transfer was faster. 

One method that you can use is to build a rail system through your pool of hot water with a loader at one end and a bridge to loop the rails.  The loader feeds in 50kg chunks of ice, which due to the reduced mass will more readily take heat from the water and melt.

[beowulf2010]

6 minutes ago, KittenIsAGeek said:

One method that you can use is to build a rail system through your pool of hot water with a loader at one end and a bridge to loop the rails.  The loader feeds in 50kg chunks of ice, which due to the reduced mass will more readily take heat from the water and melt.

I love using conveyors to heat and cools things so this is the method I usually use.

As an aside, we need to not forget about the absolute fastest way to melt ice. Though it is the most dupe and player intensive way. Building tempshift plates out of ice in water is so good as it near instantly melts 800kg at a time. But man is it a pain continually queuing build orders... 

[KittenIsAGeek]

2 minutes ago, beowulf2010 said:

Building tempshift plates out of ice in water is so good as it near instantly melts 800kg at a time. But man is it a pain continually queuing build orders... 

Heck, even ice sculptures are a quick way to do it.  But, yeah.. you gotta keep queuing up the work.  

[Nightinggale]

12 minutes ago, KittenIsAGeek said:

One method that you can use is to build a rail system through your pool of hot water with a loader at one end and a bridge to loop the rails.  The loader feeds in 50kg chunks of ice, which due to the reduced mass will more readily take heat from the water and melt.

I was thinking of a pre-conveyor rail approach. Part of the issue with the temperature control talk is that people propose solutions, which works well in sandbox/debug mode, but when people play for real, they start out without techs and without tons of refined metal. The game should be playable and enjoyable before having done all research.

It doesn't change the fact that it actually works though.

1 minute ago, beowulf2010 said:

As an aside, we need to not forget about the absolute fastest way to melt ice. Though it is the most dupe and player intensive way. Building tempshift plates out of ice in water is so good as it near instantly melts 800kg at a time. But man is it a pain continually queuing build orders... 

Lesson 5: don't confuse "interesting" with "fun"

Sounds like you admit while your solution makes great sense from a game mechanic point of view, it's not actually fun to play with, at least not as a long term solution, which scales up to a bigger base.

 

While I agree those solutions should be in the game, I was thinking about the simple approach of just dumping ice into water. It's not really intuitive to put ice in boiling water, come back 3 days later and notice that the water is still near boiling and the ice haven't melted.

[TLW]

5 hours ago, Nightinggale said:

This means ice cooling can be made a valid option for early game to rocket "era" and matching every step in between.

You're just pushing heat around. You're not removing it. At least unless you're using exactly the same sort of fixed-temperature or mass-loss shenanigans that just got patched...

I personally like the idea of "stuff exposed to space loses heat dependent on temperature/material". It's not quite "correct", but it's well within suspension of disbelief, and encourages complex contraptions.

Or even "liquids / gasses disappearing into space cool adjacent cells", as a crude replacement for adiabatic cooling.

[beowulf2010]

3 minutes ago, Nightinggale said:

Sounds like you admit while your solution makes great sense from a game mechanic point of view, it's not actually fun to play with, at least not as a long term solution, which scales up to a bigger base.

Yeah. Ice sculptures and tempshift plates are only short term, temporary, and/or emergency solutions. I didn't mean to imply otherwise. 

I'm vaguely remembering unbuilt tempshift plates melting fairly fast (compared to 800kg of ice sitting in a pile) so maybe there's a way to automate it but since it'd need the conveyor technology it probably won't be even as good as the conveyor rail approach. 

Too bad I'm going out of town for the next few days and will completely forget to play around with it. 

[KittenIsAGeek]

3 minutes ago, Nightinggale said:

While I agree those solutions should be in the game, I was thinking about the simple approach of just dumping ice into water. It's not really intuitive to put ice in boiling water, come back 3 days later and notice that the water is still near boiling and the ice haven't melted.

Yeah, the actual mechanics aren't exactly intuitive.  Basically, the ENTIRE MASS within the cell has to surpass its phase transition point before anything happens.  In the case of ice in hot water, you have, say, 1 ton of ice sitting on the floor of the pool.  The code says, "trade heat with the water in your tile."  Well, you have 1000 times the mass of ice compared to the mass of water within that cell.  Additionally, the code limits the amount of heat energy that gets exchanged to avoid problems such as race conditions.  Thus that little square of water doesn't instantly freeze, but it does change temperature slightly.  That slight change gets dispersed through the entire pool and.. the pool itself doesn't actually change temperature because the amount of heat getting dispersed is below the threshold.  So nothing appears to change with the water itself.  Meanwhile, the ice warms up ever so slightly.  Its _just_ enough that rounding errors don't prevent it from happening entirely.  Fast forward 50 to 100 cycles later, the ice has warmed up enough to "melt" and ... bam.  Your pool suddenly overflows as 1 ton of ice becomes water and the temperature of the pool FINALLY drops.

[Darkin Coaled]

An industrial scale ice maker is a very cold metal tile that you drip water onto. And it's fully customizable and automateable and limited only by game mechanics and player ingenuity.

[Nightinggale]

3 minutes ago, KittenIsAGeek said:

Basically, the ENTIRE MASS within the cell has to surpass its phase transition point before anything happens. 

[snip]

Fast forward 50 to 100 cycles later, the ice has warmed up enough to "melt" and ... bam.  Your pool suddenly overflows as 1 ton of ice becomes water and the temperature of the pool FINALLY drops.

You are on to something here. Maybe ice should "de-gas" in water, like instead of increasing in temperature, it should take all the energy it takes from the water and melt part of the ice.

Say the ice is -20 C. It takes enough energy from the water to increase 1 kg 1 C. The same energy can increase 50 g of water by 20 C. Instead of increasing the entire ton of ice by something microscopic, remove 50 g without changing the temperature and add 50 g of water at 0 C. This will avoid the "water explosion" and generally cool the water faster because it both loses heat energy and adds 0 C water.

It's still an open question which numbers should be used for a reasonable exchange of temperature.

13 minutes ago, KittenIsAGeek said:

Additionally, the code limits the amount of heat energy that gets exchanged to avoid problems such as race conditions.  Thus that little square of water doesn't instantly freeze, but it does change temperature slightly. 

Maybe instead of making the heat exchange slow, cap the temperature change to not be rapid around the phase change temperature to avoid quick freeze/melt issues.

[BaloneyOs]

1 hour ago, Nightinggale said:

There is a late game alternative, which is called liquid pump. Early on you need a pitcher pump and a bottle emptier to move water. Later you can use a liquid pump and a liquid valve to accomplish the same task. This means the pitcher pump is reduced to deliver to buildings requiring bottled water. Hydroponic farms and sinks means the bottle requirements will drop. While bottles will not vanish entirely, most water tasks will transform into dupe-less tasks.

Heh that's true. If this is solely about what mods can add then the sky's the limit.

Quote

Unless there is some Klei statement I don't know about, then "what they want" is pure speculation and shouldn't restrict creative thinking on the forum. I will not rule out that you are right, but at the same time the whole deal with the reaction to the water sieve change is unlikely to have gone unnoticed. What Klei wanted a month ago might not be precisely the same as what they want tomorrow.

After the whole water sieve "discussion" it became clear that there are a lot of "I don't want that", but it contained very little of what people actually want. This thread is to brainstorm for ideas and figuring out what people actually want other than just being against whatever happens. Quite frankly, whatever Klei wants is irrelevant to this thread. This is about what we want and we need freedom to voice whatever idea we can come up with (good or bad).

Sure it's possible Klei will read this and go "we will never do that", but that's for Klei to decide. Since we don't know what they want, let's just assume they want to read whatever we can come up with.

What people don't want is usually made very clear but I'd argue that what people do want can be infinitely ambiguous. For this game where the theme is problem solving and design, what people don't want will much more likely cause tunnel vision and therefore create a bigger focus. I agree though that what Klei wants is irrelevant to this discussion but it's still the game's base design that pushes people to want and not want things. If Klei from the first day of Alpha told players that they wanted their space magic water sieves to always output the same regardless of input the no one would even bat an eye now but their general ambiguity and silence on the topic definitely made people question it.

I'd speculate that if we were to ask some Tetris players whether they'll want an "O" shaped block most will probably say no, and not just because they're purists. I agree that what Klei wants is irrelevant to this thread though.

 

6 hours ago, Nightinggale said:

Ice would be much more useful if the heat transfer was faster. It could be something as simple as a massive thermal conductivity multiplier.

This one had been confusing ever since early alpha lol. It's almost like heat transfer for compacted materials doesn't scale intuitively due to lacking surface area. At least there's still the ghetto solution of distributing the ice over more tiles to transfer heat more quickly.

[mathmanican]

1 hour ago, Nightinggale said:

remove 50 g without changing the temperature and add 50 g of water at 0 C.

Interesting idea. There are some possible problems:

• Does this only work with ice in water?  What if the ice is in petroleum?  Will I now get 50 g chunks of water flowing through my petroleum? This would be quite annoying (not fun). The "1 element per tile" rule in ONI (anything ONIfied respects this - for now) needs to be considered against any new mechanic. The fact that dupes drop stuff all over, makes this sound nightmarish. 
• Will this only apply to water and ice?  What about any other liquid?  Can I repeat the exact same thing with solid petro and liquid petro?  Without consistency, irritation will set in. What about liquid phosphorous and solid phosporous. I think any solution that works with heat transfer and ice needs to be consistently applied to all other liquid/solid phase changes. I am still somewhat hoping that the "ice maker" will take other liquids (polluted water, petro, crude, LOX, etc.) and return solid forms that are over 20C under their freezing point.  If dupes know how to freeze water, why not let them freeze other stuff (it's not as hard as rocket science - which they've mastered - actually, it might be)?

A building that could rapidly increase heat exchange could be a nice addition. A tempshift plate does this, yes.  But what if the building were something akin to a rock crusher. It takes in a solid and crumbles it up to increase surface area (turning large balls of ice into crushed ice), enabling greater heat exchange. There could be a dupe controlled version, as well as an industrialized version. You feed it some solid, wait for some time as an animation or something else takes place, and then the heat exchange greatly increases (some multiplier, greater than 200, to make this better than solids on conveyors inside metal doors). The building will melt things in small chunks, (1 ton at a time? - gotta make it more useful than a sweeper putting ice into 20 storage bins in 10kg chunks... been there.... done that....), and only requires power (or dupe intervention) after each melting.  If the object never melts, then the machine stays dormant (the surroundings are too cold). One version of the "Melter" could just let the stuff melt onto the ground.  An advanced version has a liquid output. Maybe let the building work underwater. 

I just took my frozen crockpot meal out of the deep freezer and needed to thaw the contents enough to get the bag off. Was thinking how this could apply to ONI.

 

[Nightinggale]

20 minutes ago, mathmanican said:

Will this only apply to water and ice?  What about any other liquid?  Can I repeat the exact same thing with solid petro and liquid petro?

Will I now get 50 g chunks of water flowing through my petroleum?

If solidElement.highTempTransitionTarget == liquidElement

To be put in normal English, if the solid will melt into the liquid element in the cell, then melt a bit. If not, then use current temperature exchange code. That should make it work for all frozen liquids while avoiding the "chunks of 50 g water in petroleum" issue.

You have some good points. Ideally the code shouldn't even mention water/ice. It should rely on the element setup as this will make the game behave consistently even with mod created elements.

41 minutes ago, mathmanican said:

A building that could rapidly increase heat exchange could be a nice addition. A tempshift plate does this, yes.  But what if the building were something akin to a rock crusher. It takes in a solid and crumbles it up to increase surface area (turning large balls of ice into crushed ice), enabling greater heat exchange. There could be a dupe controlled version, as well as an industrialized version.

That would be awesome. Those buildings should be like the steam turbine in the sense that they sit on a floor and then interact with whatever gas/liquid is below the floor. Add automation inputs to allow them to be turned on/off by sensors (added automatically if using power).