Heat Energy Doubling - Molten Lead for the Win

[0xFADE]

Ah, I'm on a map with random oil pockets so I have ludicrous amounts of lead.

Does this work in a gas/liquid transition or is it more specific to something with an actual change like ethanol?  You wouldn't have to worry about it rising since it would do that for free.

Phosphorus at 280c and supercoolant at 436

 

[metallichydra]

2 hours ago, mathmanican said:

You guys are too much fun.  Thanks for all the fun banter (I hope you're having fun). ONI is a blast, but the forums are even better. 

half the fun is on the forums.

[mathmanican]

12 minutes ago, metallichydra said:

half the fun is on the forums

Maybe 4/5ths or (4/5) P-(6V+12C)/5......  Hmm.... DIdn't you all love reading that math intensive post.  

[Zarquan]

Have you considered not having the lead flow at all?  With doors, airflow tiles, and temp shift plate, you can force all the heat through lead.  When it is a solid block, you heat it.  When it is liquid, you transfer heat to the steam turbine.  You could use an arbitrary large amount of lead and it would be more compact.

[mathmanican]

41 minutes ago, 0xFADE said:

Does this work in a gas/liquid transition or is it more specific to something with an actual change like ethanol?

This particular build is a solid/liquid transfer issue. The point of the build was to make the phase changes happen, and then almost instantly stop heat transfer (otherwise you have to pay the price for any extra heating cooling). With gas, it's much harder to contain perfectly the heat transfer (though I have been thinking about ways to do it with liquid mechanics). @socooo's build (posted in the original part) utilized phosphorous for something quite similar, and I think a redesign, not focused on the steam turbine itself, may do this... (or maybe it already does). 

I also don't know if the phase change mechanics are the same -3K, followed by instant up jump of 1.5K. 

7 minutes ago, Zarquan said:

Have you considered not having the lead flow at all?

I have. But I haven't found a nice way to make the phase changes hit at exactly (or near exactly) their phase change temps.  The doors melters are 100K above the phase change temp, but the liquid flows out at almost EXACTLY the phase change temp.  If I left the liquid in contact with these door, at all, then we equalize temps too fast, and loose out on the temperature difference of our cooling and heating plate.  The point is to spend as little energy as possible going beyond the transitions.  

Do you have any ideas to achieve this?  The flow you see in my build was designed precisely to make this happen.  When I saw water shoot out below my mesh tile doors at 272.5 (the transition temp), I was completely surprised. All other melting techiques I've used would add a few degrees, if I want it to happen FAST. 

That's the key problem to solve. You want FAST transitions (500kg/s or more) that keep the phase change super close to the actual phase change temps. Currently, the steam turbine part (solidifying) is loosing me about 1K. I think I can raise the temp from 585K upwards to maybe 595K, but the room needs to solidify all liquid before hiting the waiting plate. 

I was hoping to use a liquid-solid bypass ( from @Tonyroid's volcano tamer), and even had one built that did a great job, but it didn't help with the water transition (because the steam turbine was on the wrong part)..... It may actually have the needed properties for this build.  I'll play with it tomorrow (after I dream about it tonight).  If this works, then I'll need about 6 tiles to do all the heating/cooling, and get perfect transitions with no flow at all....  (probably too optimistic).  Here's to hoping.

[Zarquan]

23 minutes ago, mathmanican said:

I have. But I haven't found a nice way to make the phase changes hit at exactly (or near exactly) their phase change temps.  The doors melters are 100K above the phase change temp, but the liquid flows out at almost EXACTLY the phase change temp.  If I left the liquid in contact with these door, at all, then we loose the gain.  The point is to spend as little energy as possible going beyond the transitions.  

I see.  It would be tricky to immediately stop the heat transfer.  What about valves?  When the lead becomes solid, the valve would be immediately deactivated and heat would stop transferring (I believe).  Though the valve would overheat without space materials...  Perhaps some idea like this could be made to work.

What if the molten lead is in a mesh tile?  Then the solid lead will be forced out and should stop heat transfer.

Another question:  Have you considered using aluminum?  It has a much higher specific heat (0.91 as opposed to 0.128), which I believe is a multiplicative factor on the heat multiplication.  It isn't unreasonably hotter.  327 C and 660 C are not that far apart, as volcanoes are much hotter and any aquatuner that can achieve 327 C can also achieve 660 C.  It is a bit hot for straight use in a steam turbine, but it could heat a buffer that connects to the turbine when it gets cold enough.

[mathmanican]

2 minutes ago, Zarquan said:

Have you considered using aluminum?

Haven't yet.  I made all my doors and metal tiles in the build out of aluminum. I'll give it a try and swap the doors/tiles to thermium, and then other less expensive materials, after it starts working. 

FYI, the aquatuner is not needed at all. Remove it, and you get more power, but you have to attach another heat source (magma for example). So the build can be done without any space materials. 

[0xFADE]

Those are some tight tolerances but it wouldn't be as fun if it were easy.

[Zarquan]

2 minutes ago, mathmanican said:

FYI, the aquatuner is not needed at all. Remove it, and you get more power, but you have to attach another heat source (magma for example). So the build can be done without any space materials. 

That's why I mentioned volcanoes .

[metallichydra]

44 minutes ago, mathmanican said:

Maybe 4/5ths or (4/5) P-(6V+12C)/5......  Hmm.... DIdn't you all love reading that math intensive post.  

I like these types of posts.

 

except that I usually don't understand half of it 

 

[Tonyroid]

1 hour ago, mathmanican said:

I was hoping to use a liquid-solid bypass ( from @Tonyroid's volcano tamer), and even had one built that did a great job, but it didn't help with the water transition (because the steam turbine was on the wrong part)..... It may actually have the needed properties for this build.  I'll play with it tomorrow (after I dream about it tonight).  If this works, then I'll need about 6 tiles to do all the heating/cooling, and get perfect transitions with no flow at all....  (probably too optimistic).  Here's to hoping.

Dude. Awesome. I have an idea for a heater/cooler that uses the water/ice thing. I'm going to sandbox some stuff...

[mathmanican]

1 hour ago, Zarquan said:

Have you considered not having the lead flow at all?

30 minutes ago, Tonyroid said:

Dude. Awesome. I have an idea for a heater/cooler that uses the water/ice thing.

Well, luck panned out. The liquid-solid bypass from @Tonyroid's volcano tamer pulled through.  Just dump 10t of lead into each open spot below (make sure it starts molten), and BOOM!  You don't get the same perfect start/stop, but you completely avoid having to move the lead.  It falls (500kg/s) in 100kg/tick easily frozen chunks. The fact that the liquid touches the heating plate while fallling is not too bad, and actually makes sure you keep a tiny blob that won't freeze, which enables the rest of the stuff to heat back up, and fall back in.  It's an endless cycle, and I've got two of them running in less space, with one heater.  Mwhahaha!!!!!!  

3 hours ago, 0xFADE said:

Can you tile this so that the melting just goes right back to another cooling section and repeats for awhile.

This version is 100% tileable.  Just pop two crude coolers, one turbine, two crude coolers, 1 turbine, etc., etc., etc.  The ends should have only one crude cooler (otherwise your lead will solidify in a diggable chunk inside a mesh tile....)  You have to keep the cooling chamber cold enough to insta-solidify the falling stuff.  My thermosensors inside the turbine room are at 585K. 

Now I should sleep.   I hope to awake to a smaller, more perfect design.  Oh, we could probably fit 4 coolers between each two turbines, with a little rearranging, amplifying the power generated. I didn't spend much time on redesigning the hot plate area, rather just slapped crap down, and it all worked.  Now that we know this design will work, we can massively improve it.  The cooling region with crude could probably be just one tile of crude, provided the walls distribute the heat fast enough. You want the liquid to fall LEFT so that waterfalls form (doubling the speed - getting you 500kg/s instead of 250kg/s).

Thanks for pushing me on @Zarquan.  This was extremely fun.  It will serve as a power doubler for basically everything...  (though doubling may be much more exagerrated than with the previous build - as you can watch the liquid raise in temp, but not too much...). 

 

[Tonyroid]

11 minutes ago, mathmanican said:

Well, luck panned out.

...

When the lead freezes, it just pops back up where it came from and starts melting again? That's brilliant.

[Chthonicone]

3 hours ago, mathmanican said:

Tiling it, as in stacking two on top of each other?  It can be done, but realize that the max mass of molten lead is over 10t (so each tile you see on the left is >10t - lots of lead). Most maps won't have enough mass for the build I made. That's why a door pump would work better. You also want to throttle the input a bit (as you can't freeze 15t chunks very well), hence the beads falling, and viscosity throttling, cap it at 100kg/0.04s. 

If we put the melter ABOVE the solidifier, and use sweepers to move 1000kg chunks of solid lead up, then we can delete the waterfall/door pump entirely (this is @Zarquan's idea that I will try). We'll want to use some kind of liquid catch, and distribution system, to make sure we still throttle the flow, but that should be simple to do if the sweeper arms can keep up. 

If I can get this thing to melt 2000kg/s, then the power potential is probably around 3 turbines running almost full time.  I may be able to do that in less space (minus the extra turbine) than what you see above, but more testing is needed. I figure @Saturnus will probably grossly simplify this, compact it, and max out the juice.  I'll play with it some more till then. 

That's crazy, I think I have enough lead on my map for this. I stopped digging it because I was at around 70 tons.

[0xFADE]

This may lead to some hotfixes

[DonDegow]

You're definitely too good at this.

[Coolthulhu]

6 hours ago, 0xFADE said:

Flaking without visible damage is still a thing.

You can see the damage by mousing over the lowest pixel of the block. It's just the "properties" tab which doesn't show it.

[BLACKBERREST3]

7 hours ago, mathmanican said:

Once you use all the "premade" machines, and want to start experimenting with making your own machines, the ONI universe completely transforms into another game.

This speaks to the engineer in all of us.

btw, pwr/tile of a steam gen is 56.7w/tile not accounting for steam chamber. Petrol/Ethanol is 166.7w/tile and coal/hydrogen/nat-gas are 66.7w/tile. I haven't found a way to get infinite petrol/ethanol yet, but it seems to be the most space efficient not accounting for infrastructure or byproducts. To be the most space efficient with steam would probably be to have 1 giant steam chamber. I might try this on my next map. steam gens are basically free pumps, so why not use this to be even more power efficient.

[mathmanican]

6 hours ago, Tonyroid said:

When the lead freezes, it just pops back up where it came from and starts melting again? That's brilliant.

Unfortunately, the current new compact design is not even power positive....  The extra heat added is not enough to power the aquatuner....  I'll have to go back to the drawing board. I may be able to use a hydrosensor, with a door above the liquid, to enable,disable heating, which might make it power positive, but then the newly made metal will cool the liquid metal... Gonna have to play with it. 

[mathmanican]

1 hour ago, Saturnus said:

I'm surprised you're not blocking all but one vent on the steam turbine on the heater side and having one port open on the other side.

I haven't tried optimizing things yet, so this might be the way to go. My current thoughts, with placement, were to keep the cool stuff (liquid vent) on the cooling plate side, and then let the heat that can't be cooled move right, and get sucked into the turbine (far from the cooling plate).

1 hour ago, Saturnus said:

and split the steam turbine water output in two. The aquatuner side can heat the return water to steam by counteracting the heat that the steam turbine thinks it deleting.

I played with this idea quite a bit to see if I could understand, and I probably just missed something. However, this suggestion just moves heat from the heating plate to the cooling plate directly (a 1 - 1 transfer), and causes you to burn through your heat, at the normal rate (no doubling). Yes, you get immediately more power per second, but your heating plate drops accordingly.  If you don't follow this suggestion, then you can get twice the power out of the same heat energy. With the aquatuner for heating, the heating plate continues to drop, and the cooling plate rises, to the point where the entire "doubling" effect stops and phase changes no long occur.  

This build tries to isolate the phase change issue and extract all possible heat from it.  The power gain is currently throttled by 500kg/s (downward velocity). I might be able to trick it 1000kg/s, but not sure. The point being to freeze molten lead without any extra cooling (the steam turbine eats up the 3K drop in temp), and then melt lead without any extra heating done (the heating plate provides a 1.5K increase).  

If you connect the steam turbine output to the aquatuner heating chamber, all that does it move 550K of water heat from the aquatuner chamber to the steam turbine chambers.  Yes, you get a massive burst of power, but it eventually disappears completely, and if you let it run it's course, both rooms will fall to 125C eventually. 

Maybe I'm missing something in your suggestion. 

[0xFADE]

If you can make it a square you can call it a doubling cube. 

[Zarquan]

44 minutes ago, Saturnus said:

It doesn't work with two open ports but it does work with 3. If you have a heat source that can keep the steam under the 2 open port on the left over 300C then this will give a constant 710W-ish output and be self-cooling. Steam on the right will be 102-104C, and it will sometimes fill up with water backflow but the steam right under the turbine remains, and will never condense.

It's basically in perfect balance, as long as there's a heat source obviously.

What I think happens here is that the steam turbine doesn't count the steam being lower than 125C but still sucks it up so and the return water is plenty to cool the aquatuner the little it runs. Essentially acting as only two ports where open in regards to temperature to power generation.

 

That is nice, but it isn't using the trick we are trying optimizing.  It is exploiting a flaw in the building rather the a flaw/feature of the physics. 

I personally believe that, in most cases, flaws in buildings are closer to bugs than flaws in the physics.  The trick with lead is how the math works out for the phase changes given the physics setup in ONI.  It might not make sense to us, but it is how the ONI universe works with no indication that it is supposed to work any other way. 

The trick in the picture works because of imperfect coding of the steam turbine not checking its stated conditions properly.

[mathmanican]

35 minutes ago, Zarquan said:

I personally believe that, in most cases, flaws in buildings are closer to bugs than flaws in the physics.

I completely agree. 

The phase change problem could (possibly) easily be "fixed" as follows:

• Assume for now that the liquid and solid have the same SHC (true for lead). Let the stated transition from liquid to solid happen at temp X, and the stated transition from solid to liquid happen at temp Y. For the actually phase change to happen, we require the liquid to reach X-1.5K temp, and then the solid appears at temp X.  Then to melt the solid, we must heat it Y+1.5K, and the liquid appears at temp Y.  A drop of 1.5K below freezing is required to melt, and that in turn requires an increase of 1.5K above freezing to solidify.  Problem solved, at least from a DTU energy stance. 
• Now let the SHC's be different, say L for liquid and S for solid. Water is an example where L is almost 4, and S is almost 2  (L=4.179, S=2.05) . Again use X and Y for the stated transition temps. The 1.5K decrease/increase occurs when the material is in the form with the smaller, and the other is adjusted appropriately.
  • If L is larger (4.179 is larger than 2.05), then we require a 1.5K increase to get from solid to liquid, where the liquid appears at the transition point. so at Y+1.5K. To solidify the higher SHC liquid, we reduce the temp by just 1.5K*(S/L), so at X-1.5K*(S/L), and the solid appears at temp X. In this way, the exact amount of energy needed to melt the solid matches perfectly the energy needed to freeze the liquid (and the phase change temps can remain different, for example with steel). 
  • If S is larger than L, then we just reverse this.  A decrease of 1.5K (so at X-1.5K) gets us from liquid to solid. An increase of 1.5K* (L/S) gets us from solid to liquid. The energy required to phase change is the exact same, both directions.

I'm guessing this suggestion is not new, but I don't really want to search for it.... 

As is, we can abuse rapidly toggling phase changes to duplicate thermal energy. It's not a lot, but when you know how to move 1000kg/s of liquid (hence liquid taming is my new hobby), then it can add up. The element molten lead (and possibly aluminum) is an ideal substances (new to the game since launch) that allows this trickery to happen in mass form, in conjunction with the turbine. 

We're not abusing SHC here in this build.  We're abusing phase changes (regardless of SHC).  We just need to be able to use the steam turbine on the cooling plate side to extract the generated heat, without having to use power hungry means to move it. I'm sure we'll find a way to make countercurrent heat exchanger (probably from an ice biome, to a magma field), that extracts the extra thermal energy from the phase change of liquid to ice (at 725kg/s or 1500kg/s if you want to get fancy), netting over 14M DTU/s of bonus heat for power.  The hard part is keeping the cooling plate ready to deal with the increase in temp of 18.8M DTU/s (we only have to pay 4.7M DTU/s to melt the ice). And of course, all of this needs to be done without using aquatuners on the cooling plate, otherwise you flat out loose power (even with supercoolant). 

It's probably not possible.... But this is ONI.  

 

[mathmanican]

12 hours ago, Zarquan said:

the trick we are trying optimizing

Did you play with flash freezing any? I think we can use flash freezing to keep the cold plate steady and return to water. The possible heat we can extract from ice and water is crazy. It's gonna take a massive heat exchanger too.

[Oozinator]

Heya Captain Sir / Mam / X, have you ever thought about using dupes as cooling medium?
Let them run in a circle, born in a cold biome and died in a magma chamber?
How many dupes would it take to tame a volcano?
Perhaps clothing swapping / transfer would be ok?
Carry / drop a bottle / take heat (water) and then release the bottle of preheated steam into a steam engine chamber..
Manual labor - much micro - perhaps something for an automation pro chief master like you.