Olen
07-05-2007, 04:53 PM
I was thinking again of all things cool and I was wandering why noone has ever, to my knowledge, tried this idea.
First a little background: phase change cooling, this basically means that a liquid turns into a gas. This takes heat which does not contribute to raising the temperature, for example it takes about six times more energy to turn a pan of water at boiling point into steam at the same temperature as it did to heat the pan from room temperature to boiling point. Pritty effective eh?
Now when you compress a gas its boiling point increses (this is how a pressure cooker can boil things at 120*C). So if you get the right gas you compress it and it turns to liquid giving out lots of heat which is dumped in a radiator, if you then decompress the liquid it boils taking away lots of heat. This is how a fridge works. And also how vapochill systems work.
But I have an idea which is simpler and probably more effective. What if, after the gas has been compressed it were decompressed far more voilently, say by squirting out of a jet into lower pressure. Lots of cooling. What if that jet were directed at a heatsink... It would be like cutting out the middle man and putting the component inside the vapochill's tubes.
Here's a quick sketch up of the idea: note I have only bothered showing one narrow pipe as I'm lazy, there would be two or three - the one shown is for the GPU, the CPU would have one as might other 'hotspots'.
http://i69.photobucket.com/albums/i58/Seidr/server/basicidea.jpg
The whole system is sealed with only the coolent (I'm thinking about R31 (if I can get it) just now) inside, this is drawn though the large pipe at the top of the computer case into the little box next to it which houses a compressor (taken from a large freezer) this turns the gas to liquid then cools that to ambient temperatures. The high pressure liquid is then taken back in the thin pipes and fired though thin jets at hot components - graphics card, cpu, north bridge - and random spray and general coldness should do the rest. Obvoisly drives etc would be housed outside and cables would be sealed in. The expansion slots in the case would also be sealed as would everything to stop coolent escaping.
The only possible issue I can forsee is finding a large enough compressor willing to work against that sort of pressure gradient (I'm looking at two or three bar (c. 40psi) difference from one side to the other). Also finding a suitable coolent as pressures over 5bar (75psi) above ambient worry me too much as if anything burst it would do you real damage.
Can anyone think of any improvements? Is this just to dangerous to be sensible trying? Its an unusual ideas but it seems never to have been done and might work. Does anyone know what components make of high pressure?
First a little background: phase change cooling, this basically means that a liquid turns into a gas. This takes heat which does not contribute to raising the temperature, for example it takes about six times more energy to turn a pan of water at boiling point into steam at the same temperature as it did to heat the pan from room temperature to boiling point. Pritty effective eh?
Now when you compress a gas its boiling point increses (this is how a pressure cooker can boil things at 120*C). So if you get the right gas you compress it and it turns to liquid giving out lots of heat which is dumped in a radiator, if you then decompress the liquid it boils taking away lots of heat. This is how a fridge works. And also how vapochill systems work.
But I have an idea which is simpler and probably more effective. What if, after the gas has been compressed it were decompressed far more voilently, say by squirting out of a jet into lower pressure. Lots of cooling. What if that jet were directed at a heatsink... It would be like cutting out the middle man and putting the component inside the vapochill's tubes.
Here's a quick sketch up of the idea: note I have only bothered showing one narrow pipe as I'm lazy, there would be two or three - the one shown is for the GPU, the CPU would have one as might other 'hotspots'.
http://i69.photobucket.com/albums/i58/Seidr/server/basicidea.jpg
The whole system is sealed with only the coolent (I'm thinking about R31 (if I can get it) just now) inside, this is drawn though the large pipe at the top of the computer case into the little box next to it which houses a compressor (taken from a large freezer) this turns the gas to liquid then cools that to ambient temperatures. The high pressure liquid is then taken back in the thin pipes and fired though thin jets at hot components - graphics card, cpu, north bridge - and random spray and general coldness should do the rest. Obvoisly drives etc would be housed outside and cables would be sealed in. The expansion slots in the case would also be sealed as would everything to stop coolent escaping.
The only possible issue I can forsee is finding a large enough compressor willing to work against that sort of pressure gradient (I'm looking at two or three bar (c. 40psi) difference from one side to the other). Also finding a suitable coolent as pressures over 5bar (75psi) above ambient worry me too much as if anything burst it would do you real damage.
Can anyone think of any improvements? Is this just to dangerous to be sensible trying? Its an unusual ideas but it seems never to have been done and might work. Does anyone know what components make of high pressure?