Infinite Heat by Freezing

[ghkbrew]

10 minutes ago, Nebbie said:

conveyor rail loops holding something with a high SHC and thermal conductivity* are the best single thing you can add on top, in case you want even more heat transfer.

Yeah, though I think 10kg of super coolant in a pipe is the best.  The only thing has more thermal mass is 20kg of frozen super coolant on conveyor. Its hard to beat that 8+ SHC

[Nebbie]

6 minutes ago, ghkbrew said:

Yeah, though I think 10kg of super coolant in a pipe is the best.  The only thing has more thermal mass is 20kg of frozen super coolant on conveyor. Its hard to beat that 8+ SHC

I'd not be too sure without testing; conveyor rails to my understanding get a massive boost to thermal interactions, along with the added effect of their contents exchanging heat with solid tiles directly below because it's treated the same as debris resting on them.

[mathmanican]

5 hours ago, ghkbrew said:

this one is definitely the Mini version

This build is bringing up nostalgia for me.  My first post to the forums was the steam behemoth, showcasing how to use the minimum heat transfer issue (which you've used very nicely here with lead) to get turbines to run forever, without any inputs (back when they spit out steam, and ran at full power if they ran at all).  We now have a new turbine build, providing massive amounts of steady power, that requires zero inputs. Build and forget (you do have to have thermium to run the aquatuner though).  Nicely done! 

If we can power it off a metal refinery, instead of an aquatuner, then I think we can start promoting building it long before cycle 100 (shouldn't be hard, though uptime may not be 100%.  Super sustainable has never been simpler. I do like how you moved all the liquid vents to the middle. Central heating at the source makes for a much simpler heat dispersion method.

[ghkbrew]

4 hours ago, Nebbie said:

I'd not be too sure without testing; conveyor rails to my understanding get a massive boost to thermal interactions,

you made me doubt so I did test it

 

 

 

 

 

 

 

 

Spoiler

Starting with 1000kg/tile 0C and 100C water

After 10 cycles: Super coolant 30C/70C, Dirt conveyor 9C/91C

Super coolant wins, but not by as much as I was expecting.

39 minutes ago, mathmanican said:

If we can power it off a metal refinery,

This is running a super coolant aqua tuner with 100% uptime (though it could likely use a lot more).  That is 1,181,600 DTU/s according to the wiki.  Also according to the wiki, making steel produces 2,339,162.0 DTU/s with an unskilled duplicant. You can triple that with a skilled duplicant.  Definitely possible. Of course, you'll still need molten metal as a coolant, so maybe not the easiest method.

 

50 minutes ago, mathmanican said:

I do like how you moved all the liquid vents to the middle. Central heating at the source makes for a much simpler heat dispersion method.

Thanks, I am pretty proud of the central steam production idea too.  I'm planning to use it in a geothermal power plant in survival.

[TripleM999]

9 hours ago, ghkbrew said:

Note to self, convection > conduction.

I think, this is cause of the post patch gas behavior. Adjacent uniform gas tiles now swap places on 5% chance, i'm not even sure, if a pressure or temperature gradient is needed for this, didn't tested that case without. Though gas movement itself (and the heat within) through pressure gradients is far superior to conduction.

[mathmanican]

59 minutes ago, TripleM999 said:

I think, this is cause of the post patch gas behavior.

It's been true for years.  Here's an example where I used it last year. If you make sure that all heat is concentrated at one point, then when stuff leaves the heating chamber, it all has the same heat (hopefully).  The temp swap bug actually resulted in a tiny bit of heat loss (which is what I noticed when we started working on turbine builds), and precisely the reason I was heating stuff up down the chimney chutes while testing builds. 

[Nebbie]

5 hours ago, ghkbrew said:

...

Super coolant wins, but not by as much as I was expecting.

...

Ah. In hindsight, it was silly of me to expect that any conductivity bonus rails may provide would matter. This result quite clearly says that unless the conductivity bonus is the same, it's just mass (dirt 2x) and SHC (supercoolant ~6x) that matter.

[TripleM999]

18 hours ago, ghkbrew said:

Nevermind, this one is definitely the Mini version

Spoiler

Note to self, convection > conduction.  A solid 8.5MW from running 1 aqua tuner full time, +/- active cooling.  The steam settles out at about 207C, so the turbines are wasting just a little power.  But 10 is a sweet spot. Adding more turbines definitely gets diminishing returns.  I tried a second row of 10 above the first, it dropped the steam temperature enough that I only got about 15% more power (for 2x the turbines).

Sorry, but had to tinker again with your work.  Take it as deep admiration.

The steam chamber on the left is far from optimal, but both chambers should in theory get the same or at least similar amounts of heat.

FreezerFurnaceDouble2.sav

[ghkbrew]

17 minutes ago, TripleM999 said:

 

Oooh I like. I considered using the bottom tile, but discarded the idea because it's in contact with the heating chamber. In retrospect that's not an issue with the near vacuum quantity of molten lead.  Very nice!

[ghkbrew]

OCD successfully un-triggered:

The left side is still a couple degrees cooler than the right, but I think I'll just have to live with that

FreezerFurnaceDouble 003.sav

[TripleM999]

8 minutes ago, ghkbrew said:

The left side is still a couple degrees cooler than the right, but I think I'll just have to live with that

I think, i missed a tempshift plate under the bridge on the left side.

[ghkbrew]

46 minutes ago, TripleM999 said:

think, i missed a tempshift plate under the bridge on the left side.

Ha. I think that did it.  204.5 in both the far left lower tile and the far right lower tile. 

[TripleM999]

It can run with 3 rows too, i think. Seems to be stable at around 174°C. Cooling needed one more stage though. 

But i think, you are right, 2x10 is the sweet spot here.

[TripleM999]

19 hours ago, mathmanican said:

If we can power it off a metal refinery, instead of an aquatuner, then I think we can start promoting building it long before cycle 100 (shouldn't be hard, though uptime may not be 100%.

I'm not sure, how feasible that would be, but maybe cornerloaded kilns could be an fully automated option. But probably heat isn't enough.

[ghkbrew]

2 hours ago, TripleM999 said:

cornerloaded kilns

I've tried this in sandbox but had trouble getting high temperatures from it.  The issue is that you have to put in low temperature coal which cools the kiln room off.  You have to run the kiln in a vacuum to really get it hot, but then there's no way of getting the heat out. I guess you could periodically flood it with molten steel to extract the heat, but that seems complicated.

Does anyone have a good auto-kiln set up for high temperatures?

2 hours ago, TripleM999 said:

But probably heat isn't enough.

It isn't for this setup.  The thermium AT is putting out 1e6 DTU/s.  A kiln does 20DTU/s.

[mathmanican]

So far the builds require space materials to abuse (or aluminum, which not all worlds have).  As such, it doesn't affect regular game play, rather just adds extra fun tools to play with for those who have "completed" the game. Personally, if that is all it majorly affected, then I'd leave the bug in the game as "emergent gameplay." 

Any ideas on builds that a newbie could easily copy, in the first 50 cycles, to abuse this? To get real attention (and eliminate emergent gameplay as an option), we probably need to get a design that does precisely this. I think we've got the bug fully digested, and abusing it completely harnessed. What we need now is a simple way to abuse it early on. Ideally, something tiny and easily reproducible, using very basic materials, with huge useable gains for the builder. 

In Rime, I can quickly think of a slew of ways to abuse it, but slapping a tepidizer down to heat stuff up to 20C is probably simpler. Where could we use this on Terrra, before we hit space? Thoughts?

[wachunga]

1 hour ago, ghkbrew said:

You have to run the kiln in a vacuum to really get it hot, but then there's no way of getting the heat out.

Kilns are a two wide building so you could stick a bit of molten lead or something on the cell that doesn't contain the building contents. IMO, the bigger issue with kilns in vacuum is that most of the kiln heat is "Excess produced" which directly heats the surrounding gas/liquid. When in vacuum, this heating doesn't happen. I never bothered figuring out exactly what happens in a partial vacuum. Does a percentage of the excess produced heat the molten lead? What percentage? Could you stack two different liquids to double that percentage? And so on. Something you might find interesting to look into.

 

 

[ghkbrew]

1 hour ago, mathmanican said:

Where could we use this on Terrra, before we hit space? Thoughts?

The problem is that nothing solidifies in the ideal pre-space range of 125C - 325C.  That means for this to work without thermium we either need to use a different heat source or fundamentally change the build.

A few possibilities I'm considering:

1) Replace aluminum with lead and use metal refinery heated petroleum as the heat source

2) Go all in on the refinery heating and use molten steel for the coolant and replace aluminum with iron or magma. This would work on Terra and technically be pre-space, but probably too complicated for a newbie.

1 & 2 We could easily due right now, it's just a question of if they're a good idea. These require some R&D

3) Start using the almost identical bug with condensation by evaporting/condensing phosphorus using a steel AT (Note if anyone has a efficient way to quickly evaporate and condense liquids I'd love to hear about it for use in a molten salt reactor)

4) Counterflow debris and molten metal.  By using the molten metal to reheat the debris we can reduce the needed heat input to approximately 0, making less efficient high temperature heat sources feasible (kiln, glass forge, metal refinery).

 

[mathmanican]

1 hour ago, ghkbrew said:

anyone has a efficient way to quickly evaporate and condense liquids I'd love to hear about it for use in a molten salt reactor)

Search discord for this one. I think they have several. @Manarz could probably point you to the best resource quickly.

[wachunga]

Follow up to my kiln post. I ran a few tests to explore the excess produced mechanic.

Top kiln is in vacuum, middle has a blob of 50kg molten lead, bottom has 50kg molten steel on top of 50kg molten lead. Everything started at 1400K with enough coal to do 20 jobs of refined carbon, or 800 seconds worth of running.

 

Final temps and heat production rate by doing all the math with heat capacities and such:

Top: 1563.9K, ~4 kdtu/s

Middle:1647.3, ~8 kdtu/s

Bottom:1613K, ~12 kdtu/s

 

So it seems the excess produced heat is evenly split across the building footprint (in this case 16 kdtu/s between 4 cells). Potential heat is lost for each of those cells that is vacuum. Meaning you can triple the heat production of a kiln in vacuum by putting liquids on it in the right places. Note I originally used 1kg of liquids and got 2.7 kdtu/s per cell instead of 4. Perhaps there is some kind of clamping going on.

 

[mathmanican]

1 hour ago, wachunga said:

Perhaps there is some kind of clamping going on

There definitely is. To add to this, the "excess produced" heat from the kiln is also reduced if the mass is below 1500g (see here from another clamping). I use this when I want to run kilns without having to worry about the excess heat (just put 10g of salt water and water over them, and they effectively can be ignored forever). Knowing that vacuum results in just 4kdtu/s is interesting. I wonder if this affect is true for other buildings as well. 

[wachunga]

Quote

building/cell transfer of "excess" heat: each cell receives the excess heat, divided by the area of the building, and this transferred heat is proportionally scaled down if the cell as less than 1.5kg of content (source: C# code)

From the famous decrypting heat transfer thread. I've read it so many times and relearned all sorts of stuff I've forgotten. The problem is I've forgotten what I've forgotten. Code still looks largely the same all this time later.

The excess heat mechanic is a goofy one. Look at the polymer press, 32.5kdtu/s but only 500 as excess. The other 32k heats the building and must be shed as building<->cell conduction. When combined with how important SHC is with hot building conduction, you get gold polymer presses that overheat readily. Because gold has terrible SHC. If you want a brain workout, try to figure out how multiple cell buildings conduct heat. SHC is doubly important there, what a mess is all I can say.

[Nebbie]

2 hours ago, ghkbrew said:

The problem is that nothing solidifies in the ideal pre-space range of 125C - 325C.  That means for this to work without thermium we either need to use a different heat source or fundamentally change the build.

A few possibilities I'm considering:

1) Replace aluminum with lead and use metal refinery heated petroleum as the heat source

2) Go all in on the refinery heating and use molten steel for the coolant and replace aluminum with iron or magma. This would work on Terra and technically be pre-space, but probably too complicated for a newbie.

1 & 2 We could easily due right now, it's just a question of if they're a good idea. These require some R&D

3) Start using the almost identical bug with condensation by evaporting/condensing phosphorus using a steel AT (Note if anyone has a efficient way to quickly evaporate and condense liquids I'd love to hear about it for use in a molten salt reactor)

4) Counterflow debris and molten metal.  By using the molten metal to reheat the debris we can reduce the needed heat input to approximately 0, making less efficient high temperature heat sources feasible (kiln, glass forge, metal refinery).

 

Molten steel is definitely too much for newbie stuff, especially now that ice makers actually overheat. As well, getting magma involved puts things in "you could've just ran turbines off the magma" territory. It sounds like you want bases to be created with the following priority goals:

1. Sustainable food
2. Steel
3. POWER SOLVER 9000
4. Sustainable water
5. Sustainable filtration medium
6. Space stuff

So pretty much you want something people can put together with just the sandstone and oil biomes plus dreckos, so that things are broken on every map without any "dangerous" exploration; something you can slap down in next to the home base to solve power the entire game the moment steel, lead, and plastic are acquired in the right amounts. Phosphorus condensation is definitely the prime suspect for doing this.

By the way, sulfur's freezing point is only a little outside the ideal range; perhaps there could be a separate build to exploit that as a way to cheat on heat by raising turbine output water from 95 C to 110+ C? It would probably be especially useful in heat desalination setups to lower the power cost, and it might be a new way to prime systems that turn refinery usage into power.

[0xFADE]

Amazing.  Reducing the liquid part to pretty much 1 tile.  Amazing.

[Neutronium]

@mathmanican  Why not try the condensation process XD