What's like a heating element, but cold instead?

Factual Questions
1

A heating element, or a cartridge heater, is this solid piece with wires coming out of it, and when you put power into the wires, heat develops inside the solid piece and travels out to its environment. The thing is fairly ideal, having thin wires that don’t generate or move much heat themselves, so you could have long wires and it would be an ideal heat source out wherever you put it. I want something that does the opposite of that, and absorbs heat from its environment. A cooling element.

A typical device for cooling has big pipes or ducts running to it, which are themselves a source of cold or heat. You can get a cooling effect in the location of your choice, but also a long line of heating or cooling effect that leads to the device. If you were watching with a thermal camera, not only would you see a very cold spot on the device, but you would also see the pipes as having their own special temperature.

What would be an ideal device, as if it were exclusively generating the cold locally?

One idea is a little boiler, with fine capillary lines to send liquid to it and let gas escape from it, but they still wouldn’t be as thermally neutral as the wires on a heater, and they might blow up if their environment is too hot.

Another idea is a thermocouple junction. Forcing current through a thermocouple will chill its junction, through the Peltier effect. But this is a weak effect. It is used in a chilled thermocouple humidity sensor that causes condensation on the junction, but it’s a really poor performer.

Yet another is a cryostat. But it still has the pipe issues.

Perhaps one could bring chemicals to a little reactor in which one carries out an endothermic reaction?

I don’t think this is thermodynamically impossible. If a device converted thermal energy directly into electricity, it would fit my requirement, but it would also be a free energy machine. I don’t mind expending energy to make this thing work, I just want the business end to do nothing but chill.

2

Are you lumping thermoelectric coolers (TECs, aka Peltier coolers) in with what you are calling “thermocouple junctions”? TECs are semiconductor based instead of thermocouple based. They make great little heat pumps, though they are admittedly a bit inefficient and you need some way of drawing the heat off of the hot side or the whole thing tends to overheat and fail.

Another possibility is a vortex cooler. All you have running to it is an air line, so the line itself doesn’t heat or cool anything. Vortex coolers are annoyingly loud though, and you need to be able to vent the air somewhere.

How big is the thing you are trying to cool?

3

Anti-griddle?

4

Actually, a cartridge heater is far from ideal. A vapor compression heat pump is much more ideal than a cartridge heater and can be used equally for heating or cooling. As long as you don’t need large temperature increases or drops.

I think a vapor compression heat pump is the most ideal heat transfer mechanism.

A vapor compression heat pump does not have “large” ducts but thin refrigerant pipes, and those pipes do not necessarily act as a source of cold or heat themselves. (While in contrast, the wires of an electrical heater do actually lose some heat, unless they are superconducting, in which case they lose a lot of cold!)

With a vapor compression cooling device you have liquid refrigerant running through a thin pipe to the cooling element (evaporator). The temperature of the liquid in the pipe can be mostly equal to the outside temperature. Just before the evaporator, you have an expansion valve that drops the pressure and evaporates some of the liquid. Then the remaining liquid evaporates in the evaporator, further cooling it. The evaporated gas can mostly heat back up to outside temperature on the way back to the compressor and condenser, where it turns back into liquid.

You can reverse this cycle to get heating instead of cooling. The roles of the condenser and evaporator are then swapped, but you can still get most of the heating to happen in the condenser and most of the cooling in the evaporator, with the refrigerant pipes remaining mostly at outside temperature.

This is much more ideal than a heating cartridge because you can actually transfer more heat energy to the target than the mechanical energy you put in the compressor (as long as you don’t have a large temperature difference). E.g. you can get 3kW worth of heating or cooling while driving the compressor with only 1kW of electrical motor output.

(I heat my house with an Air to Water Heat Pump and I love it! :slight_smile: )

5

Based on that description of what issue you are trying to avoid, all you need is a heat pump or heat exchanger with insulated pipes leading to/from it. You can get vacuum-insulated pipes if necessary (basically a very long and flexible thermos bottle).

If I wanted to cool something (no heating capability necessary) and make it very cold, I might use a vacuum jacketed line to deliver liquid nitrogen to the cooling element, and another vacuum jacketed line to vent the evaporated nitrogen back out.

6

But no matter what you do, if you want to cool off your element, you’re going to have to heat up something else, and to a greater degree. You have a choice of dumping this waste heat close to the thing you’re cooling, or transporting it (via pipes or something) to a more remote location. But you can’t avoid having to dump it somewhere. In this house, we obey the laws of thermodynamics.

7

Yep. You cannot locate your air conditioner completely inside your house. And your refrigerator heats up your house a bit, because it is completely inside-- unless it’s an ice box. :wink:

8

A resistor generates heat from electricity. All you have to do is swap the +/- leads connected to the resistor and you’ll run it backwards, making it cold. :wink:

It sounds like the OP, while having a clear-eyed respect for the laws of thermodynamics, is looking for a source of “cold flux.” But as Chronos said, if you’re chilling something, you’re inevitably going to have to dump heat elsewhere.

The OP mentioned chemical cooling via an endothermic reaction. If you only need “cold flux” for a little while, you could use liquid CO[sub]2[/sub] in a pressure vessel, slowly bleeding off the pressure. Those 12-gram CO[sub]2[/sub] containers that power pellet guns and inflate bicycle tires, for example, get quite cold when you use them to inflate a tire. Mostly the cold comes from the liquid CO[sub]2[/sub] changing phase into a gas, but they get cold enough to accumulate lots of frost when operated on a humid day.

In fact, Wikipedia says that dry ice was discovered by Adrien-Jean-Pierre Thilorier, who boiled off enough liquid CO[sub]2[/sub] that it produced solid CO[sub]2[/sub] “snow”–dry ice.

If you could live with recharging a pressure vessel with liquid CO[sub]2[/sub], then you could get quite a bit of one-time phase-change cooling from that pressure vessel. You’re still dumping heat elsewhere, but you’re doing it elsewhere in time (back when the CO[sub]2[/sub] was compressed and liquefied) rather than elsewhere in space (as with a Peltier cooler, heat pipe, heat pump, etc.)

Also: while the vortex coolers engineer_comp_geek mentioned sound temptingly close to Maxwell’s demon, they’re horribly inefficient compared to other cooling methods.

While commercial heat pumps often have a coefficient of performance between 3 and 4 (for every unit of energy they consume, they move between 3 and 4 units of energy), commercial vortex chillers using air as a medium have a theoretical maximum COP of 0.42[sup]1[/sup]. But that’s theoretical. In practice, their COPs are on the order of 0.03-0.05[sup]2[/sup], or 1% that of a quality commercial heat pump.

Vortex chillers make sense if you’re bleeding off pressurized gas anyway, or if you have a particular need for a cooler with no moving parts. But the real-world efficiency of vortex coolers is abominably low.

You can get a better COP by using a gas other than air and by using the energy from the hot side of the vortex cooler to compress more air to feed the vortex cooler, but no commercially available vortex coolers do this–at least none that I’m aware of.

I guess if you really wanted to go crazy, you could cool the pressure vessel by boiling off CO[sub]2[/sub], and then feed the gas you bleed off into a vortex chiller. Then you could use harvest energy from the hot side of the vortex tube to… (This quickly becomes a rabbit hole of diminishing returns).

[sup]1[/sup] http://www.nrcresearchpress.com/doi/10.1139/cjp-2015-0089
[sup]2[/sup] M.O. Hamdan, A. Alawar, E. Elnajjar, and W. Siddique. In Proceedings of the ASME/JSAME 2011 8th Thermal Engineering Joint Conference, Honolulu, Hawaii, 13–17 March 2011. 2011. 10.1115/AJTEC2011-44393.

(N.B.: Reference 2 was shamelessly horked from reference 1).

9

OP - what temperature do you want to go down to ?

Heat pipes (Heat pipe - Wikipedia) probably gives the largest heat transfer per unit heat exchange area.

In general, you get the best heat transfer (assuming convection dominates) when there is a phase change involved. So circulating a liquid at its boiling point, where part of the boiling happens inside this black box will give you the best heat transfer.

10

Me.

I can go wherever you need me to go, and when I get there, I promise to just chill.

:smiley:

11

In this house we do indeed. And if I happen to find a device that doesn’t, screw my current career, I’m going into business with that instead.

But, returning now to planet earth, yes, yes, of course. I do want to cool off my element. And so I will have to heat up something else by a larger amount. That’s perfectly fine. Literally most of the universe is available to accept my heat if I merely transport it there to dump it.

And here we get to the problem. How to transport it? That’s my complaint. Pipes, ducts, heat pipes, all those things. They are big fat highways whose own temperature is inextricably bound to the way they are transporting my waste heat.

Picture an electric heating element. Maybe it’s a cylinder the size of an AA battery. Say we give it 24 gage wires – unusually fine but if we give the heater a high resistance and drive it with high voltage 24 gage will be just fine. See how perfect it is? A nearly self-contained cylinder, except for those skinny little wires, but they are allowed to get as hot or cold as their environment. They are thermally inert.

The plumbing on the hot side of a heat pump - yes, including the Peltier type solid state heat pumps - is anything but thermally inert. That is the problem I’m trying to fix. Chronos, when you say “You have a choice of dumping this waste heat close to the thing you’re cooling, or transporting it (via pipes or something) to a more remote location” you are framing my problem exactly. I don’t mind spending the energy at all, but i don’t want to mess up my environment with this dumping or transporting process.

Here’s my most frequent example. I design ovens, sometimes elaborate ovens, and I can sprinkle heating elements all over the place with impunity, and add as many measurement points and control loops as I want. But I can never ever use a “colding element”. It’s like trying to design an electronic circuit but never being able to access the negative side of the power supply. Several of the neato things I’d like to do are unavailable, except in a transient way if I build containers of something endothermic into it.

I think perhaps the closest I can come is a tiny boiler that operates at quite a high pressure on the liquid input side, so both the liquid supply and gas exhaust lines can be small and can run at any temperature over a wide range without exchanging much heat; in other words, have the phase change involve way more energy than any of the fluid temperature changes without phase changes.

12

OK, so you’re fine with having heat pipes, you just want them to be small and non-leaky? That’s doable. As am77494 says, the best ones contain some substance that changes phase at right around the temperature you’re working with. Precisely what substance that would be depends, of course, on your intended application.

13

That’s what I recommended above although the higher pressure may not be needed if you find the right molecule or a cocktail of molecules.

The other method is to ditch convection altogether and go with molten metals. Like molten sodium in French nuclear reactors.

14

Just like putting the batteries in your flashlight backwards makes it suck all the light out of the room.

15

No, no, no, it makes the flashlight produce antiphotons.

16

In some applications, based on LED illumination, it turns the LEDs into NEDs*.

Once.

*Noise Emitting Diodes

17

Is that anything like a Darkon?

http://lindberglce.com/tech/darkons.htm

(I have no idea why they didn’t wait a month so this could have come out in the April Fool’s issue)

Although this guy had the same idea thirteen years later, and DID put it in the April Fool’s issue:

http://wearcam.org/theory_of_darkness.html

18

While we’re at it, cold isn’t just the absence of heat, and it’s possible to remove cold from a system while decreasing the system’s total thermal energy content, or add it while increasing the energy.

Example: My mom has a chicken coop. When it’s going to be really cold overnight, she puts a big bucket of water out in the coop to keep the temperature in the coop stable. The next morning, of course, the ambient temperature warms up, and so she goes out and removes the bucket of ice. Well, even frozen, the bucket of ice has much more thermal energy in it than the air it displaces, and so she’s removing energy from the system… but she’s also removing cold.

19

It is possible to cool without contact of any sort. Laser Cooling fits the bill for that, but you have to limit the temperature you are cooling down from (substantially less than a kelvin) and the materials you can cool are limited to say the least.

If you want to limit the effect of wires penetrating the cold space, use phosphor bronze wire. If you are using low power/frequency in and out, this is what everyone uses.

Finally, if you want push a lot of current in but don’t want to have much heat leak, superconducting wire does a great job. It’s widely used as a “thermal switch” to drastically reduce (order of magnitude) the heat leak when the sample gets cold enough. Which is also a constraint on the temperature you are cooling from of course (think tens of kelvin at best). (Yeah, I know Larry Niven had a plot point that revolved on the “fact” that superconductors were perfect conductors of heat and electricity- they are neither).

In practical cryogenics, you need to know more than just how many penetrations you are going to have. Once your sample is cooled below the ambient temperature, you have to shield against radiation from that environment, and no shielding is going to be perfectly non-radiating, so you can’t get to zero radiation load. Beyond radiation, the ambient and intermediate temperatures available to you make a big difference (when cooling in a liquid, using the cold gas boiling out of the cold area to cool your wires/structure makes a big difference).

When it comes right down to it, working with thermal physics is playing at the casino from hell: you can’t win, you can’t break even, and you can’t choose another game, 'cause it’s the only game in town.

20

A million years ago I needed a cite for this from Chronos. Good times.