Crazy Coolant Idea

[FireDemon666]

I've had this idea for a few years now, and while I know it wouldn't be practical for everyday use due to toxic evaporation, I wondered if anyone has ever tried using mercury as a coolant.

It's thermal conductivity is vastly superior to water, but I imagine the milestones one would run into would be the pump struggling against It's density, not to mention the possibility of it reacting negatively against any materials in the system and having to run the system in an open area.

Can anyone speculate on this?

[Kurz]

Conductivity isn't the only factor.
There is also Thermal Capacity.

The ability of a substance to hold heat without increasing in temperature.
Example because of the H bonds of water it has a higher capacity of holding heat.
The H bonds stores that potiential energy.

http://en.wikipedia.org/wiki/Heat_capacity

[Praxis]

This comes up every few months anytime someone reads about mercury. You will eventually die or kill someone trying to use it or some other terrible thing so don't do it.

[Church]

thermal conductivity is only one of important properties of coolant. Another one is thermal capacity. I doubt it beating water in that, when just few metals in solid form only nears it. +Do you know how poisonous mercury is?

[Petra]

OP, cooling with liquid gallium would be much more practical... http://www.xtremesystems.org/forums/...ad.php?t=99240

[FireDemon666]

Guys I'm going to make this very clear right now since there are already 2 responses assuming I am an idiot who is not aware mercury is poisonous which shows they just skimmed over my original post. Please scroll up and read very carefully where I typed:

and while I know it wouldn't be practical for everyday use due to toxic evaporation

Good point on the thermal capacity BTW.

[Waterlogged]

There's also the little problem that your pretty much stuck with air as a medium for the last thermal transfer in most situations. Water would work a lot better if air wasn't such a crappy means of transferring thermal energy.

[NKrader]

where would you buy that much murcury? lol the people that you ask for it would get so confusted.. "yes can i get 1/2 gallon of mercury please?"

[NaeKuh]

OP, cooling with liquid gallium would be much more practical... http://www.xtremesystems.org/forums/...ad.php?t=99240

why is it that only u and i remember this post?

And i remember it from the time i was lurking too.

Even vapor didnt know about this post until i told him @ GOOC.

[FireDemon666]

where would you buy that much murcury? lol the people that you ask for it would get so confusted.. "yes can i get 1/2 gallon of mercury please?"

LOL! I thought of the same thing.

"Yes can I please get a 1/2 gallon of mercury...It will not be used for terrorist purposes"

[Manicdan]

wtf is liquid gallium, nm ill google it instead of having you guys convince me to go get some

and thermal capacity shouldnt be much of an issue, since liquid flow rates should be fast enough to negate that issue right? the difference between water temp going in and out is still really low, like 5C ish?

[Petra]

why is it that only u and i remember this post?

Considering it was posted in 2006, it does pre-date a rather large percentage of XS members... though, I'm a bit surprised that Eric hadn't seen it.

With that said, I do have a knack for remembering random stuff like this.

[Church]

and thermal capacity shouldnt be much of an issue, since liquid flow rates should be fast enough to negate that issue right?

I doubt it not being issue given much higher viscosity of liquid metals - so most probably low flow.

[NaeKuh]

ur cheating the flow component by changing the medium.

Gallium has such a high heat capacity, that you really wouldnt need super flow.

But the pumps i believe are far more reliable, as its magnetic, meaning it pulls the liquid via magnetic field, so there is no real moving parts inside that pump.

[FireDemon666]

I was actually considering doing this, until I saw the price of liquid gallium.

$150 for 100 grams...

[NaeKuh]

I was actually considering doing this, until I saw the price of liquid gallium.

$150 for 100 grams...

ROFLROFLROFL..

I think realtim had a simular idea. But it makes your coolant conductive.

And all the times ive leaked, spilled, did something STUPID, i dont think i can ever use a conductive coolant.

[FireDemon666]

ROFLROFLROFL..

I think realtim had a simular idea. But it makes your coolant conductive.

And all the times ive leaked, spilled, did something STUPID, i dont think i can ever use a conductive coolant.

QTF.

Like that time I accidentally yanked a hose out of a fitting with the PC powered on and pump running...

Or the time that POS coolermaster waterblock I had developed a crack in the acrylic top...

[mambo500]

under heat, doesn't mercury also expand more than other liquids?

[Akula]

Off topic, as far as computer cooling goes but on the topic of crazy coolants some fast breeder nuclear reactors use liquid sodium as the primary heat transfer fluid. About as nasty of a combination as I can imagine.

http://en.wikipedia.org/wiki/Liquid_metal_fast_reactor

http://www.youtube.com/watch?v=SiSqW6pFuR8

[JumpingJack]

Conductivity isn't the only factor.
There is also Thermal Capacity.

The ability of a substance to hold heat without increasing in temperature.
Example because of the H bonds of water it has a higher capacity of holding heat.
The H bonds stores that potiential energy.

http://en.wikipedia.org/wiki/Heat_capacity

There is also a problem with density, for the same volume Hg will be about 14x heavier than water meaning much higher pressures are needed to get equitable flow through the same elevations.

[meanmoe]

I doubt it not being issue given much higher viscosity of liquid metals - so most probably low flow.

lol, more like low flow for about 15 minutes until your pump has an aneurysm.

I worked in a lab a while ago where we used mercury manometers for high pressure differentials without knocking holes in the ceiling (for an equivalent water manometer).

We had liter bottles of it. Someone put one on a top shelf once, I didn't realize what it was and slid it off. I of course was not expecting the weight. It fell down and briefly pulled my shoulder out of socket.... fun times.

I'll take water though.

Edit: it tastes funny too...

[Vetalar]

may be we return to this idea (liquid metal cooling) after CPU's TDP reaches 1kW ... but now it's not worth it. IMHO better stick with slightly sub-ambient phase change... much less toxic and danger