Liquid metal - extreme cooling

[B NEGATIVE]

Which liquid metal did you have in mind? It would need a melting point thats below room temperature otherwise the whole 'loop' thing might not work too well. That kinda leaves Gallistan and Sodium-Potassium Alloy. Galistan costs 200$ for 50 grams and Sodium-Potassium alloy is extremely reactive with air and water so that might not be the greatest idea.

Gallium would be the next best candidate but has a melting point of around 30 Celsius. 300$ will get you 500g of that stuff but i still think its impractical and probably wont work since you might not even be able to get all the gallium in the loop to liquify in time to cool the components before overheating.

 

Dont think its gonna work... or am i missing something, im not a chemist or anything

 

 

I just pointed this out to him when he PM'd me.

[LED_Guy]

Heat is produced in the chip.  It is conducted to the exterior of the package, through the interface material between the package and the cooling block and into the cooler block.  Heat is then conducted into the boundary layer of water in the cooling block.  From there it is transported via (forced) convection to the radiator.  Inside the radiator it is conducted to the fins where it finally gets to the environment via convection (air flow through the fins) and radiation.

 

Heat transport via radiation scales as the absolute temperature raised to the 4th power.  It isn't overly effective at anything close to room temperature.

 

Conduction is heat transport through a solid (or at least non moving body)

Convection is heat transport by moving material (usually a fluid and may be forced or natural - forced is a fan, natural is due to buoyancy differences due to heating).

Radiation is heat transport via black body radiation.

 

No offense, but you need to keep your terminology straight.

 

In order for a direct die cooler to work, the flow has to be across the die surface at a high flow rate.  This will give the shear necessary to minimize the thermal and mass transfer boundary layers so that you have a very low thermal resistance.  If you don't do that then you are wasting your time.  I haven't seen the inside of the DT Sniper so I can't comment on what they did wrong if the performance was worse. 

 

On another topic, I have about a kg of gallium and 100 g of indium lying around.  Getting a little tin (+ a few other "spices") to make what I am planning on using isn't too much of a concern for me.

[Brenz]

This sounds really interesting and I know nothing about how liquid metals really work so maybe you can explain why the points below aren't problems.

 

1) You say all the metal will heat up but wouldn't this combined with a lower flow rate like you wanted result in heat going against the flow causing it to heat up more when going through a block?

 

2) You mentioned a system could be cooled by using the case as a radiator or with no radiator at all. Surely using the case would result in 1/2 the heat being put back into the case and increasing the internal case temp and using no radiator would do the same thing. The heat has to be taken somewhere, preferably outside the case or your going to have problems.

 

Like I said I don't know anything about this stuff and would be interested in seeing the results if you try it.

[LED_Guy]

When you are liquid cooling (with water) the water is above room temperature when it goes into the cooling block. The point about liquid metal is that it conducts heat much better than water. If you put a sheet of paper on top of your CPU it will only get hot directly over your CPU. Try that with a sheet of aluminum and the aluminum will warm up all the way to the edges.

It's not really that heat will travel against the flow, but it spreads out perpendicular to the flow (left-right and upwards) much more easily. This equates to better heat transfer. It's a benefit, not a problem. If it were a problem then you would use air instead of water.

As for using the case as a radiator that is a lot more speculative. I can texture the outside to improve heat transfer to the air. It is also easy enough to insulate the inside to prevent heating the inside of the case. A double layered panel with lots of fin area inside may be what I end up doing. This would let a natural chimney effect help out, but I may need a fan none-the-less. The fan would have to be ducted and distributed to give a uniform flow. Stay tuned for more on this aspect . . .

[Brenz]

Thanks for the reply. Back on to the first point again if the flow rate of the liquid metal is great enough to prevent heat transfer going against the flow then why would all the liquid metal in the block be heated like you wanted?

 

I understand if you put a pan on the cooker the heat will eventually spread to the whole pan but if you only put it there for a second the rest of the pan will remain cool. Eventually the heat would spread out but in a situation where liquid metal was flowing wouldn't it leave the block and be in the radiator before the better heat transfer capability of the liquid was able to make a noticeable difference?

[LED_Guy]

The point is that liquid metal can transfer heat better.

Water is better than air at transferring heat which is why people use it for high performance applications. Otherwise we would all just put a bunch of fans in our cases and be done with it. Liquid metal can be better than water. That's my point.

[Hatsune Miku 「 」]

oh nevermind. I thought someone made a new cpu cooler, but turns out its the same thing used to store heat in various power stations like the molten salt thermal solar power plant in spain, which heats the salt to generate power. Somewhat not feasible for a cpu cooler.

[LED_Guy]

Not even close, unless you are trolling.

How is using a metal that is already liquid in a cooling loop like melting a salt in a solar plant? This isn't heat storage, it is liquid based cooling, just with something a lot more advanced than water.

[Furball_Bunny]

Any updates on the build?