Jump to content

Recommended Posts

9 hours ago, Kasuha said:

Yes, they do.

I would be damned... That is actually really useful information!

Knowing that I was inspired to make this thing.

20170620203939_1.jpg

It's not exactly boiling with thermo regulators, but it boils.

The idea here is that the top room cools down the water through the bridges, and then the water cools down the steam.

I used water because you basically can't change the temp once it's at 1000g, so it just transfers the cold through the bridge instantly. This results in steam condensing as soon as it hits the bridge, but the bridge itself can't heat up and cools right down. It should be said that it is at its most effective if you can drain the room to a vacuum before you start making steam.

When I built it I thought the top room had to run every now and then to cool it all down, but even after making all that water you see there from steam, neither of the top rooms had changed more than 0.1C*. So all that is actually running here is the 2 tepidizers on the sides whenever the hydro switch at 10kg is triggered and the input pump. Which is quite energy efficient if you ask me.

Played around with different temps on the input water as well, but even having it at 2C* it only slightly slowed down, so pre heating water is not needed.

 

20170620203914_1.jpg

Link to comment
Share on other sites

Interesting build, using the wire bridges to transfer temps between rooms! Although to accelerate it further you could also not make the walls between those out of insulated tiles. :) So if the thermoregulators are cooling the steam, are you recycling their heat anywhere or does it just go into the environment?

Link to comment
Share on other sites

A little bit more engineering today, I think I've solved The First Law's flaws with the hydrogen loop.
 

Spoiler

 

The First Law-radiatorupgrade.jpg


 

I have squared out the water basin and filled it with some warm water to provide thermal mass and area for the hydrogen radiator, moved the switch upward so the pump leaves the basin mostly filled. added some more granite tiles to prevent stress. The hydrogen radiator now runs granite gas pipes across every tile of granite and water in the basin.

I have also closed the hydrogen loop so it's running entirely self-contained, with every pipe filled with 1000g of hydrogen. This saves 480W on running the pumps from before, so power costs are down to roughly 1 kW now. (but you should still have separate transformers for the thermoregulators, the tepidizer and the pumps to prevent overload damage)

As you can see from the picture, the hydrogen gets pretty cold at -188.5 but it's still well above the dew point. It will probably get a little bit colder still as the machine keeps running. Filling the basin with warm water from a geyser when building should be a great way to keep the hydrogen in the loop from freezing while you start it up.

Link to comment
Share on other sites

1 hour ago, Sevio said:

Interesting build, using the wire bridges to transfer temps between rooms! Although to accelerate it further you could also not make the walls between those out of insulated tiles. :) So if the thermoregulators are cooling the steam, are you recycling their heat anywhere or does it just go into the environment?

The thermoregulators was only used to cool the water in the middle. Turns out you don't even need that part. All you really need is a pool of cold water that is insulated and has bridges sticking out from the bottom. The steam will condense as soon as it touches it without affecting the water at all.

Link to comment
Share on other sites

3 minutes ago, Rattja said:

The thermoregulators was only used to cool the water in the middle. Turns out you don't even need that part. All you really need is a pool of cold water that is insulated and has bridges sticking out from the bottom. The steam will condense as soon as it touches it without affecting the water at all.

When cooling that water, you maybe unknowingly exploited a bug in heat transfer between water tiles. Otherwise cooling that pool wouldn't be that easy.

And once you have that cold pool, it will warm up gradually. The game definitely tries to have reliable power transfers (heat removed equals heat added somewhere else) where appropriate so if you use that water for cooling for a long time, it will eventually warm up. There's just a lot of it, 1000 kg per tile, and with quite high heat capacity.

Link to comment
Share on other sites

Just now, Kasuha said:

When cooling that water, you maybe unknowingly exploited a bug in heat transfer between water tiles. Otherwise cooling that pool wouldn't be that easy.

And once you have that cold pool, it will warm up gradually. The game definitely tries to have reliable power transfers (heat removed equals heat added somewhere else) where appropriate so if you use that water for cooling for a long time, it will eventually warm up. There's just a lot of it, 1000 kg per tile, and with quite high heat capacity.

Not unknowingly no.

To cool the water, I used the same trick as I usually do, overfill a vent so I have small layer of water on top and exploit the weird thermal transfer down that water/liquid has by cooling the top layer. So once the water touched the bridges at the top, it cooled down very quickly. That was actually what I was testing before I got the idea of cooling steam.

It is true that over time, it would warm up if all you had was a pool of cold water, although it would last a very very long time before it would be of any concern depending on the size of the pool and the thing you are cooling down. Even with the water at 60C*+ it was still able to liquefy the steam. 

However, what you could do if you want something a bit more permanent is to just do the same thing and put another pool on top of the first one with bridges sticking down. It doesn't really need to be a hydrogen cooler.

20170620234436_1.jpg

Basically the top pool only needs to cool the top layer of the middle one, (which should be anything from a few grams to a few kg less is better) but the bottom on the other hand needs to heat up the full 1000kg. So the difference in energy needed to heat up vs cooling down is so massive that it will never transfer heat up.

This actually opens up for a lot of very low powered cooling systems, as it works the same for any liquid. All you would have to do is get it cold, after that it would stay there. 

 

Link to comment
Share on other sites

3 minutes ago, Rattja said:

Not unknowingly no.

Ah, okay, then everything is fine. I would suggest a maybe more efficient method then, using less water and taking up less space.

DwgoRm6.jpg

Upper pipe is abyssalite and the cold water just sits in it. Upper part of the lower pipe is abyssalite too, the radiator is wolframite. You can deconstruct the part of the pipe that goes to the pump once it fills up. The bridges constantly cool the water in the closed loop, it will never heat up and it cools the steam down very efficiently. You only need enough cold water to fill the pipe segments on bridge inputs.

Link to comment
Share on other sites

21 hours ago, Kasuha said:

Tepidizer has 320 W of "normal operation" heating and 20 kW of "exhaust". When you let that exhaust to be released to small amount of liquid, it gains a lot of temperature.

Neat!  I didn't know that either!  So, if you can somehow move the stuff away that you'd dripping onto the tepidizer fast enough, then the machine wouldn't heat up so much.  

Link to comment
Share on other sites

I wonder if that many bridges are even needed to use the temperature exploit, but if you started this system with a supply of 1C water it could destroy so much heat... Didn't I read somewhere they would fix this particular bug with the next update, though?

Link to comment
Share on other sites

2 minutes ago, Sevio said:

I wonder if that many bridges are even needed to use the temperature exploit

Not really. Each bridge forces the packet at output to average temperature. With 5 C water at inputs, one bridge would bring 95 C water to 50 C and five of them to 7.8 C. But in this closed loop you're dealing with way smaller temperature difference. My experience is that in this radiator the water gains about 2 C during the cycle at most. So even one bridge could be enough, keeping the contents of the loop about 2 C above input temperature.

Link to comment
Share on other sites

8 hours ago, Kasuha said:

Ah, okay, then everything is fine. I would suggest a maybe more efficient method then, using less water and taking up less space.

DwgoRm6.jpg

Upper pipe is abyssalite and the cold water just sits in it. Upper part of the lower pipe is abyssalite too, the radiator is wolframite. You can deconstruct the part of the pipe that goes to the pump once it fills up. The bridges constantly cool the water in the closed loop, it will never heat up and it cools the steam down very efficiently. You only need enough cold water to fill the pipe segments on bridge inputs.

Hm, that very interesting. So as long as you fill it till it stops it just copy that on the other side when it passes through? That sure would be useful to keep in mind.

It is indeed more efficient in regards to water usage, but came up with this to save some space. Same idea, as before but only 1 water tile needed for locking the temperature.

20170621094247_1.jpg

Don't have my main save atm but in a steam room it works very well.

Link to comment
Share on other sites

8 minutes ago, Sevio said:

Didn't I read somewhere they would fix this particular bug with the next update, though?

In the worst case it will work till next monday. But so far I have no indication that it will be fixed next monday, it's not fixed in current closed testing hotfix.

We're talking exploits here. They're fun while they last but they can be fixed any moment. I actually welcome each such fix. But personally I see the exploit you use now as a much more fix-worthy thing than this since it seriously screws up what's happening e.g. in steam geyser chamber when steam condenses on top of released water.

Link to comment
Share on other sites

11 minutes ago, Kasuha said:

In the worst case it will work till next monday. But so far I have no indication that it will be fixed next monday, it's not fixed in current closed testing hotfix.

We're talking exploits here. They're fun while they last but they can be fixed any moment. I actually welcome each such fix. But personally I see the exploit you use now as a much more fix-worthy thing than this since it seriously screws up what's happening e.g. in steam geyser chamber when steam condenses on top of released water.

I'm not sure I follow, what exploit am I using in my tepidizer steam condensing chamber? Are you referring to the tepidizer being able to steam water at all?

Edit: You mentioned a bug in heat transfer between water tiles a few posts up, I'm guessing you're referring to that but I don't understand what the bug is from your description.

Link to comment
Share on other sites

 

11 minutes ago, Sevio said:

I'm not sure I follow, what exploit am I using in my tepidizer steam condensing chamber? Are you referring to the tepidizer being able to steam water at all?

Edit: You mentioned a bug in heat transfer between water tiles a few posts up, I'm guessing you're referring to that but I don't understand what the bug is from your description.

First of all, sorry for not keeping good track of the discussion. I was referring to Rattja's cooler, not to your design.

The bug is that small amount of cold water dropped on top of full tile of hot water instantly cools that full tile down. And not just that tile but also all tiles below it. 

Link to comment
Share on other sites

@Kasuha

Ah that makes more sense to me, I meanwhile went searching on the bugtracker and found this, which corresponds very well with your explanation:

That looks like a pretty severe bug and should definitely be up there on the list for fixing, I agree.

I don't think I'm doing that in my design fortunately (neither the bridge bug nor the water dropping bug), the steam cools on touching the water surface, which I'm cooling "properly" with the hydrogen radiator. Looks like a decent chance my contraption will continue to work in the future, either with the tepidizer or a proper boiler that is yet to be introduced. :)

Link to comment
Share on other sites

So I've been working on trying to improve the throughput of the Thermo Regulator boiler by adding a heat exchanger to the design. I've also added a second thermoregulator room and connected them in parallel, 4 sets of two. This doubles the power usage to roughly 2 kW, but the boiler can be throttled to 50% by deactivating half of the regulators in each room, if only 1 kW is available.

The First Law Mk2.jpg

Overlays in the spoiler:

Spoiler

 

The First Law Mk2-liquid.jpg

The First Law Mk2-gas.jpg

Oxygen Purifier-power.jpg

At 50% throttle it can process 511 g/s of polluted water, which is a huge improvement over the old design of 271 g/s, thanks to the heat exchanger doing part of the work of the regulators. At full throttle it can process about 800 g/s. It's not quite twice as much as the half throttle setting because it's limited by the heat exchanger at the moment.

For the heat exchanger I decided to use granite tiles because they have a better heat conductivity than just having more water. Additionally, wolframite wire bridges and granite gas pipe bridges connect each water tile into a granite tile to help equalize heat across all tiles. Granite pipes for the polluted input water run through the entire heat exchanger. (I didn't want to use wolframite for those because I thought it might be very difficult to gather that much wolframite in survival)

Exchanger stats:

  • 50% throttle - 26.9 C pwater in, 59 C out. Water output 65C.
  • Full throttle - 26.9 C pwater in, 53.8 C out. Water output 67.8C.
  • Downside - Very slow to heat up.

With double the thermoregulators and running them in parallel, this design is also really pushing the limits of the hydrogen radiator that cools the output water, since each set of two has only a quarter of the entire radiator area available to it. (Stabilizes around -227 C output from the regulators) The pool is larger as a result and I used granite tiles here as well in an attempt to help heat conductivity and reduce the amount of clean water that's "wasted" in the storage tank.

So the main problem that I still need to solve with this is the heat exchanger while running at full throttle. Are there still ways to improve the exchanger's overall conductivity other than upgrading to wolframite pipes?

Link to comment
Share on other sites

2 hours ago, Sevio said:

Are there still ways to improve the exchanger's overall conductivity other than upgrading to wolframite pipes?

As far as I got it it is more effective to heat tiles and transfer from them to your fluid, compared to transfering between pipes and fluid directly. @Kasuha posted something about that recently.

Other option is to vent the heat exchanger on one side of a granite tile wall and bridge that into your target fluid with wolframite wire bridges, downside is that you'll need gas pumps to get it back into a pipe.

Link to comment
Share on other sites

I would have preferred to avoid pumping the polluted water twice, but maybe it's possible to have the polluted water intake tank be a snaking tunnel that serves as the heat exchanger at the same time... It would be incredibly slow to heat up though.

Link to comment
Share on other sites

I bit the bullet and decided to upgrade the entire polluted water radiator to wolframite pipes. To my surprise, it made no difference whatsoever to the output temperature. I guess going bigger is the only way to transfer more heat.

So I started thinking again of a heat exchanger that exchanges heat directly between water and polluted water, without the use of pipes. (or in addition to pipes)

I decided to conduct a little thermal experiment to figure out what materials could transfer heat between water and polluted water most effectively.

Two tiles wide of conductive material at exactly 50 C, 2000 kg of 90 C water below and 2000 kg of 10C water above. This will even out over time to two pools of 50 C water and does not change the conductor's temperature, eliminating its thermal mass as a variable in the experiment. Starting conditions below:

door-conductor-1.jpg

From left to right (conductivity in parentheses): Granite (3.39), Copper Mechanized Airlock (4.5), Copper Manual Airlock (4.5), Iron Manual Airlock (4), Wolframite Airlock (15).

When clicking twice on an airlock one can also see that the underlying material is called Steel Door <DO NOT TRANSLATE> with a conductivity of 54, so I half expected to see all doors show the same rate of heat transfer. To my surprise however, all doors and the granite tiles are showing the exact same rate of heat transfer after several cycles!

door-conductor-2.jpg

Here the cool water in all 5 experiments has heated up to exactly 20 C and the hot water has cooled down to exactly 80 C. All doors and tiles are still at 50C. The abyssalite insulated tiles have not changed temperature either. (still exactly 20C)

It appears that in the equilibrium state, the conductivity of the intermediate material does not matter at all! This seems like a bug to me.

So I then decided to test the Wire Bridge, which are known to transfer heat across tiles pretty well due to being in contact with multiple at once. Again, the same experiment setup. Heat all wire bridges to 50 C, then a pool of cold water at 10 C above, and a pool of hot water at 90 C below.

wirebridge-conductor-1.jpg

From left to right (conductivity in parentheses): Gold Amalgam (2), Iron Ore (4), Copper Ore (4.5), Wolframite (15)

This time the thermal transfer worked a bit faster than before, and there were also differences between the different wire bridges. But again to my surprise, they're not in order of thermal conductivity!

wirebridge-conductor-2.jpg

End temperature of the cold/hot water from left to right:

  • Gold Amalgam - 14.9 C / 85.2 C
  • Iron Ore - 30 C / 69.8 C
  • Copper Ore - 29.1 C / 70.8 C
  • Wolframite - 24.4 C / 75.8 C

I thought maybe somehow it's the specific heat capacity that matters, however the specific heats from left to right are: Gold Amalgam (0.15), Iron Ore (0.449), Copper Ore (0.386), Wolframite (0.134), the results are not in order of specific heat either.

The wire bridges themselves also did not stay at 50 C. From left to right: Gold Amalgam (75.9 C), Iron Ore (60.5 C), Copper Ore (60.1 C), Wolframite (50.1 C). Again, the wire bridge temperatures do not correspond to the results of heat transfer rate either.

Can anyone make sense of these results?

In any case, for a heat exchanger between two fluids, it looks like iron wire bridges are the way to go, rather than the poorly performing wolframite. And it doesn't seem to matter whether you use doors or granite tiles.

Edit: It gets weirder still. When comparing the temps of the cold and hot water for each experiment, a small discrepancy has built up. (Hot + cold temperature / 2) should add up to 50 C, but it doesn't. Averages below:

  • Gold Amalgam - 50.05 C (within rounding error margins)
  • Iron Ore - 49.9 C (0.1 C lost)
  • Copper Ore - 49.95 C (within rounding error margins)
  • Wolframite - 50.1 (0.1 C gained)

For Gold Amalgam and Copper Ore it looks close enough to just be a rounding error but the discrepancy for Iron Ore and Wolframite seems a bit too large.

Link to comment
Share on other sites

1 hour ago, Sevio said:

Can anyone make sense of these results?

I decompiled the relevant functions to find out and found something interesting:

public static float CalculateEnergyFlow(float source_temp, float source_thermal_conductivity, float dest_temp, float dest_thermal_conductivity, float surface_area = 1f, float thickness = 1f)
{
    return (float) ((double) (source_temp - dest_temp) * (double) Math.Min(source_thermal_conductivity, dest_thermal_conductivity) * ((double) surface_area / (double) thickness));
}

Note the Math.Min - if this function is used for tile and object heat transfer (and further sniffing around decompiled code indicated it is), this means that the lowest thermal conductivity is used, NOT the product of both conductivities.

This would explain wolframite pipes not helping with transfer over granite (granite already has higher conductivity than water).

Still doesn't say much about wire bridge transfer. I think it is caused by those lines:

float energyFlow = SimUtil.CalculateEnergyFlow(temperature1, data1.thermalConductivity, temperature2, data1.conduitThermalConductivity, this.contentsSurfaceArea * ConduitTemperatureManager.ContentsScaleFactor, 1f);
float delta_kilojoules = SimUtil.ClampEnergyTransfer(dt, temperature1, data1.heatCapacity, temperature2, data1.conduitHeatCapacity, energyFlow);

The first one is just an overload of the function I listed above, the second one doesn't decompile cleanly so I won't paste it here, but the important part is that it takes heat capacity of both source and destination into account and multiplies heat transfer by product of both.

So it seems that heat transfer depends on heat capacity, which would explain your results, as iron has the highest specific heat, then copper, then wolframite, then gold.

This would mean that sandstone and iron ore are the best materials for radiators and heaters, not granite and wolframite. Needs proper experimentation, though.

Link to comment
Share on other sites

I was trying to reproduce your results. I also tried to use permeable tiles containing vacuum for insulation as they should not be in contact with anything. To my surprise, permeable tiles above both water masses somehow exchange temperature with them. Not ones below, but in the image you can clearly see the breaking hot ones (made of copper ore) but the ones above cold water masses got colder too. All others did not change their temperature at all. So I guess the experiment has failed but it has revealed yet another peculiarity of the physics engine. I'll retry with abyssalite but just wanted to share the find.

Q3EFbew.jpg

 

 

Link to comment
Share on other sites

I am not sure about this. But for some reason wolframite transfer heat slower if the heat differens is high than for granite.

So let say the differens is 100C between materials then I did se that Granite transfers faster heat. But when it got lower Wolframite started to transfer faster than Granite.

I havent done that much testing dont have time at the moment.

I did test with floor lamp and 100kg hydrogen and 100kg CO2

Link to comment
Share on other sites

Most efficient heat exchange is directly between water tiles. Put layer of clean water directly on layer of polluted water, you'll get them at near equilibrium temperature in no time. For some reason they stop exchanging heat when they're 1 C apart (e.g. they stop on 39.3 and 40.3 C) but they can get to such state from 80 degrees difference in matter of seconds. Polluted water needs to be at the bottom, or they'll instantly swap places, but it works great for both hot at the bottom and hot at the top.

Link to comment
Share on other sites

Let's say you have polluted water at the bottom and clean water at the top, both flowing in some direction or another. Can you separate the two flows again? I guess separating the clean water is easy, just let it flow onto a tile and then into a drop. But the only way I could see to separate the bottom flow is to let it accumulate in a basin. and pump it from there, or perhaps let it flow from the bottom layer of a basin under a tile.

Flowing clean water on top of polluted water, this sounds like a terrific idea for a water purification plant. :roll:

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Please be aware that the content of this thread may be outdated and no longer applicable.

×
  • Create New...