Sign in with
Sign up | Sign in
Your question

Helium Gas Cooling

Last response: in Overclocking
Share
July 16, 2008 4:52:42 PM

Hey all,
I have a fairly out-there idea. I wanted to get peoples thoughts on using gaseous helium in a computer as a cooling medium. vs air. From what I found the thermal conductivity of air is .0257 W/(mK) where helium is .1513 W/(mK). From a quick table lookup Cottonseed oil has a thermal conductivy of .173 W/(mK) and water has .588 W/(mK). If helium is used, system fans can still be used and cooling should be much more efficient minus the liquid problems of oil or liquid cooling. However helium would have to be reused in the system due to cost if nothing else, so a container would have to be built in a manner similar to the vegetable oil container or possibly it some type of giant mylar balloon thing. Assuming the container can hold helium reasonably well, and that it is not too insulating, you may be able to get better cooling than using air.

So I was wondering if anyone had any clue at all what one could use for a helium container (I know helium is really tricky to store) and also any comments on the idea would be appreciated.

P.S. Helium has a much lower heat capacity than oil so it will heat up quickly compared to an oil type system, however as long as steady state temps are acceptable this wont be a problem..

More about : helium gas cooling

July 16, 2008 5:02:11 PM

You cant have a close system. The helium will heat up to where it actually heats up the components, thats why there is intake/exhaust fans. Unless you have some sort of radiator/compressor it won't work
July 16, 2008 5:04:38 PM

are we talkling liquied helium or gas helium? Not to pop your balloon but If it worked they might be doing it allready.
Related resources
July 16, 2008 5:07:18 PM

That was my initial thought, but the oil cooled computers were able to reach an equilibrium at a temperature below systems temps. Tom's hardware got to 40C using their oil cooled PC just letting it sit. My though was if the container was big enough you would have enough heat transfer to keep the case at a similar temperature.
July 16, 2008 5:09:43 PM

liquid helium would be really tight, like water cooling or liquid nitrogen, but the energy bill to keep it liquid would be outrageous, however cool, no pun intended.
July 16, 2008 5:09:47 PM

I was thinking gaseous helium. My thought was it may work, but I am sure it would be a pain to maintain, a pain to set up and may not have the performance of liquid cooling and thats why it probably hasn't been done. But it would be cool to be different.
July 16, 2008 5:10:25 PM

Liquid helium would be absurd, 4 Kelvin or so. Probably could super-conduct at that point.
July 16, 2008 5:14:51 PM

I'm saying you should probably give this up and get back to studying for that Thermo exam!
July 16, 2008 5:18:01 PM

I'm a mechanical engineering grad student, I have absolutely nothing better to be doing:) 
July 16, 2008 5:36:43 PM

I would suggest getting a really huge case. Preferably two and put them together. The next step would be to assure that the heat can be transfered from the gas to the outside of the cases. A bigger surface area should help with that. There are passive hdd cooler that are basically glued to hdd. If you cover the in- and outside of the case with them and make sure that there is a constant airflow you might actually succeed - unless your case is too small. Then the Helium will heat up too much.
July 16, 2008 5:38:37 PM

Lots o problems to work around . U either need a closed system (and ANOTHER) heat exchanger, or a limitless supply of cool (and free) helium.
Using LONNNNNNG heatpipes to move the heatsink fins to a box or tube with a drain and a water mist/fall/drip to use the .588 whatever of water
plus a fan for evaporation benefit etc. is mo better idea. Hurry up and build one so I can sue you for using my ideas college boy :) 

PS> dont ever listen to anyone who says that it would have been done already if it was any good. Thats pure CRAP! Somewhere @ 15 or so years ago I posted somwhere about using heatpipes for heatsinks and got the same lame reply. Who knows where I'd be today if I had just ordered me some parts and started selling my own design online bak then. :) 
July 16, 2008 5:48:55 PM

Tomfool,
Random side note, one of my professors actually was one of the guys that first pushed heatpipes, and he even has a book out on the subject. I can ask him if he first got the idea for heatpipes off of a message board :D 
July 16, 2008 5:54:36 PM

See what I mean, dammit. :) 

Hence the handle.........................TomFool
July 16, 2008 6:02:43 PM

How bout you have an air type (or Helium tight) system, bucket with the bottom cut out and make a hole at the top of the system with a 120mm fan. Then with a Push Pull system build a duct to the bottom front of your system.
July 16, 2008 6:25:06 PM

^^ I think that would be a prety good idea,
1. get your case
2. take off the bottom and top
3. create a duct with cooling fins followed by fans then another layer of cooling fins in it connecting the top and bottom
4. add fans to the top and bottom to speed the airflow
5. add passive heatsinks to everything

*Edit* I'm off to the patent office as I speak :p 
July 16, 2008 6:50:19 PM

here is my take on it...
treat the gas helium as a liquid. using a fan instead of a pump.

in a sealed box. have a hole on top that leads to a tube that goes to a rad. from there have the helium go back into the PC case at the bottom.

using fans to push the air through the rad.

hard part would be keeping the system perfectly sealed. helium atoms are very small. so they will leak out of any microscopic holes.
even party balloons start to depressurize because the helium escapes out through the this membrane of the balloon.

my 2¢

edit: something else to worry about... since it is helium filled... it might float away :lol: 
July 16, 2008 7:36:06 PM

Sealing will always be a problem with a gas system. So how bout this, don't go with a full system cooling.

You can have a tube that goes to the front of a heat sink and out the back all sealed, this way you only have to seal like 50 cm worth of "gap" per thing your cooling. Of course a full system would be cooler :) 

But remember you would have atleast 2 exhausts pm a computer, the PSU and the system, plus you might have a graphics exhaust.

Probably building a system like in the old THG video of the oil cooled pc would be a good idea.

Edit: or if you have a lot of money and space, you could always build a pressure chamber thats at 1 atmosphere.

This way you can just put you computer inside, but we still need a way to lose the heat, and maybe we can have a liquid cooled radiator in the chamber and just have tubes coming out. For this to work the helium in the chamber must be warmer then the air outside (Obviously).
July 16, 2008 8:29:34 PM

But thats liquid helium, we are thinking of a helium gas replacing air.
July 16, 2008 8:35:23 PM

it would have to be under allot of pressure so it would not collect on the top point of the system.
July 16, 2008 8:43:03 PM

MadHacker said:
here is my take on it...
treat the gas helium as a liquid. using a fan instead of a pump.

in a sealed box. have a hole on top that leads to a tube that goes to a rad. from there have the helium go back into the PC case at the bottom.

using fans to push the air through the rad.

hard part would be keeping the system perfectly sealed. helium atoms are very small. so they will leak out of any microscopic holes.
even party balloons start to depressurize because the helium escapes out through the this membrane of the balloon.

my 2¢

edit: something else to worry about... since it is helium filled... it might float away :lol: 


This has been the best (and funniest) idea so far.
July 16, 2008 9:08:45 PM

You could design a small refrigeration unit and have the evaporator coils contacting the CPU and GPU as a cooling device.

or buy a small freezer and stick your PC inside. Keeps the Vodka just right too!
July 16, 2008 9:29:16 PM

lol, This reminds me of an idea I had the other day. We use Mobo standoff all the time, but why not make better use of them. Why not make them into removeable heatsinks? If you could design them into a way that would fit with most mobos you could sell them piecemeal for pretty cheap and make a nice profit with a little bonus cooling.

Related to that I got to thinking, if you could design a system with fans on both sides of the case, you could have one side be intake and the other outake and just have a massive wind tunnel/cooling system monster for a case :D  lol. I was/am planning on doing this (eventually!).

btw just out of curiosity, does anyone know how much helium costs?
July 16, 2008 9:31:40 PM

OrderChaos said:
Related to that I got to thinking, if you could design a system with fans on both sides of the case, you could have one side be intake and the other outake and just have a massive wind tunnel/cooling system monster for a case :D  lol. I was/am planning on doing this (eventually!).



I think i have that already... one side is the front.. the other side is the back...
make a wind tunnel... nope can't do that though j/k :pt1cable: 
July 16, 2008 9:36:26 PM

^I meant the bigger sides lol. That way you could just have either several smaller fans (probly about 12 120mm fans) or just a REALLY BIG fan on each side. Basically both sides of the case would be only an outline of steel or aluminum with Fans filling in the middle.

Lol imagine how that would look with all LED fans.
July 16, 2008 9:37:18 PM

Just put your assembled system in a tube, sitting on a shelf in the middle and run your wires out the top. Then get 2 fans to match the ID (inner diameter ) of the tube and put them in a push - pull config... cake !
July 16, 2008 9:38:31 PM

OrderChaos said:
^I meant the bigger sides lol. That way you could just have either several smaller fans (probly about 12 120mm fans) or just a REALLY BIG fan on each side. Basically both sides of the case would be only an outline of steel or aluminum with Fans filling in the middle.

Lol imagine how that would look with all LED fans.



to bad the motherboard gets in the way of the direct air flow.
I guess we could just drill a bunch of holes in it to aid in air flow....
July 16, 2008 9:40:41 PM

Yeah, thats why I haven't tried it yet. That and I'm too lazy to cut out the sides of my case and i'm too cheap to buy that many fans lol.
July 16, 2008 9:53:34 PM

jjl0402 said:
Hey all,
I have a fairly out-there idea. I wanted to get peoples thoughts on using gaseous helium in a computer as a cooling medium. vs air. From what I found the thermal conductivity of air is .0257 W/(mK) where helium is .1513 W/(mK). From a quick table lookup Cottonseed oil has a thermal conductivy of .173 W/(mK) and water has .588 W/(mK). If helium is used, system fans can still be used and cooling should be much more efficient minus the liquid problems of oil or liquid cooling. However helium would have to be reused in the system due to cost if nothing else, so a container would have to be built in a manner similar to the vegetable oil container or possibly it some type of giant mylar balloon thing. Assuming the container can hold helium reasonably well, and that it is not too insulating, you may be able to get better cooling than using air.

So I was wondering if anyone had any clue at all what one could use for a helium container (I know helium is really tricky to store) and also any comments on the idea would be appreciated.

P.S. Helium has a much lower heat capacity than oil so it will heat up quickly compared to an oil type system, however as long as steady state temps are acceptable this wont be a problem..


here's you problem:
when the "hot" helium gets to one end, it's gonna diffuse and raise avg temp. If you have some kinda magical closed loop where the heated gas stayed at one end, your improved heat transfer is still going to be limited by the delta t of whatever is cooling your helium. you could just use something with a much greater surface area I suppose. you still reinvented the fridge if you do it that way. You could just use my frigid ex wife to freeze the system to absolute zero.
a b K Overclocking
July 17, 2008 12:54:37 PM

I have no idea how helium or anything along this works but I can hopefully help you a bit. Instead of closing off the whole computer why don't you isolate the heat souces and close them off and than cool those parts with your desired method helium, how ever you are planning to do that. It will help you out by allowing you to control the flow of helium while keeping the volume you have to cool down to the minimum which should help out greatly and give you better results. So isolating the graphics cards and the CPU should be the priority is you do it this way. I just hope you have some idea to create a complete isolated circuit of helium that cools itself somehow while doing this because im at a lost on how to do this.

Some links that I have no idea would be helpful but may be worth the read to give you some idea's. (not on your exact topic, but never the less you may still get useful information)

http://overclockers.com/tips1187/index.asp

http://overclockers.com/tips1193/index.asp
July 24, 2008 7:59:36 AM

maximiza said:
it would have to be under allot of pressure so it would not collect on the top point of the system.

I actually don't think this would be an issue, since ideally, the cooling circuit would not have both helium and air - it would only have helium. This means that there's no surrounding gas for helium to be lighter-than, and hence it won't tend to rise (correct me if I'm wrong). The cooling circuit filled with helium would be lighter than if it were filled with air, though.

@OP: Thanks for this fascinating idea. I think it might be doable, at least in the short term, but I think leakage of the helium from the cooling loop would become a huge problem. I think that it would conceptually not be very different from a water cooling system, though moving the helium around and having it cool on radiators may be trickier. Supposing you could set up a helium gas cooling loop across a standard air cooler, it would undoubtedly leak helium rather rapidly, necessitating constant re-filling with fresh helium. This is a rather expensive proposal, as helium is not getting cheaper. It's slowly disappearing from Earth, and is also constantly in higher demand for keeping systems (like the LHC at CERN) very cold (yes, that's a many-kilometer long ring of liquid-helium-cooled vacuum tube!).

I think the best way to test a system like this would be to build it and try it simply with air. If it didn't get way too hot, you could try putting in helium, and seeing if you noticed any changes. Chances are you wouldn't :)  But if you do try, please post lots of pics!
July 25, 2008 8:17:48 PM

You'd need to contain the helium in metal-lined tubing so it doesn't diffuse. But by the time you do that, you've basically recreated a water-cooled system. Might as well just use water since it's better, cheaper, and been tried.

I always thought helium was an insulator. Isn't that what they put between double-panned glass windows...or is that argon?
Anonymous
a b K Overclocking
February 25, 2009 5:42:18 PM

Look up "Nobel Gases" they are famous for being inert (buts thats not quite true), thats why argon is used in double glazing, its an inert gas thats an insulator, nitrogen is also often used because its inert. I guess the conductive properties must differ a lot between the nobel gases if Helium is a good conductor...
!