Liquid Metal Cooling

[E1000BASE-TX]

Has this ever been done with computers? Back when liquid metal like mercury was not well known for its toxicity but form its cooling cabablitiyes. People use to use liquid metals like mercury to cool down there equipment. Well why the hell not?, HG(mercury) has 140J/kgK Spec heat compared to H2O(Watter) which has 4184J/kgK. Here is a good link for a liquid metal cooling Intro L1nkx0r5. Another good property about liquid metal like mercury is that melting point is at -39ºC. You can have a freezer unit cool the mercury down to around -38ºC with out freezing. Also the boiling point is at 356.73 ºC so you dont have to wory of it expoliding on you. Besides the fact THIS STUFF CAN KILL YOU!!! it makes a very good cooling agent. So any ideas about making a liquid metal cooled computer?

[E1000BASE-TX]

Originally Posted by obobskivich

um...aside that fact that someone (you) has done their homework on this...it seems stupid for about 4 reasons (you may be able to dispell these but IDK)

#1: It CAN KILL YOU!!!!

#2: Cost?

#3: The pump to move it would need to be very strong, since mercury is heavier than water...since it's metal.

#4: Keeping it at -38C and still flowing....that is kinda werid and I don't fully understand it (continued below)

You are saying that at -39C it becomes liquid? I am getting hung up on something, it seems like a good idea (if it wasn't about as lethal as raw Weapons Grade Uranium...in liquid form) but I am not 100% sure of what you are saying
you are saying at -39C it becomes liquid and at 356.73 it becomes boiling?
I don't believe your -39C = liquid thingy, since a Mercury Thermomoter is liquid at roughly 30C...which is 69C above what your saying (and my understanding of what your saying) is it's "turn liquid" point

could you explain my confusion, and the cost/lethal-ness is an issue

Since it's HIGHLY TOXIC, IDK if you would be allowed to buy it (inside United States) due to the "constant" threat of terrorism (one attack 3 years ago and all of a sudden the threat is "constant" wtf?, it's always been there) but IDK if the Gov't would let you buy it/if your out of US if you would be able to buy it for a decent price

I Am guessing you can measure Liquid Mercury in Gallons (so I will )
A water cooling system uses like 1-2 gallons of water, which is like $10 for Distilled (That should b accurate pricing) I am also guessing that 1-2 gallons of Mercury would cost about $1000...since it's got to be shipped "Ground HAZ-MAT" which means you have to help cover the costs of the satilete tracking and everything (the joys of haz-mat). So, if we ignore the fact that you asking to use a haz-mat labled substance to cool your computer...wouldn't it be cheaper to just buy something fron nVentiv or Prometia? The pumps for this thing would be monsters (it's thick) and keeping it cold+XL tubes for good flow


IMO if this was a "perfect" cooling solution people would've done it by now...and what equipment are you saying was cooled by mercury? I have seen computers designed to run Liquid Freon for cooling (Cray 1A) and systems to run Controled Enviroment HVAC for their cooling (NOAA's JET, is it JET or JETT?)

A little more elaboration please...

-It freezzes at at -39C.
-Read the link. To pump it uses a magnet to move the liquid metal.
-You can buy mercury if you have a dregree in scince and know what your doing.
-yes it is expensive.

[Impaqt]

1: Mercury by itself isnt nearly as Toxic as people make it out to be. When Mixed with something like mercuric chloride it can kill you in a very short period of time though...
2: Major Obstical here..... Mercury, If you can get it, runs in the neighborhood of $350 a Flask...... Which is about 2.5Liters.......... And weighs 76 Pounds! (35 Kgs). Reinforced Desk???

3:Significantly heavier... Desity of 13.6 vs under 2 for various temps of water.... Probobly would need an industrial pump of some sort...

4: Water, Solid FOrm at and below 0c , Liquif from .1-99.9, Turns to gas at 100c
Merury, Solid at and Below -39c, Liquid from -39 to 357c, Boils and turns to gas at temps higher than 357c

Overall... Bad Idea........... ANd Mercury is NOT magnetic... Dunno how you could pump it with any sory of magnet.....

[-navairum-]

And i dunno about throwing a magnet in my comptuer anyways...

[E1000BASE-TX]

Originally Posted by -navairum-

And i dunno about throwing a magnet in my comptuer anyways...

No one said it must be inside the computer.

[E1000BASE-TX]

Originally Posted by Impaqt

1: Mercury by itself isnt nearly as Toxic as people make it out to be. When Mixed with something like mercuric chloride it can kill you in a very short period of time though...
2: Major Obstical here..... Mercury, If you can get it, runs in the neighborhood of $350 a Flask...... Which is about 2.5Liters.......... And weighs 76 Pounds! (35 Kgs). Reinforced Desk???

3:Significantly heavier... Desity of 13.6 vs under 2 for various temps of water.... Probobly would need an industrial pump of some sort...

4: Water, Solid FOrm at and below 0c , Liquif from .1-99.9, Turns to gas at 100c
Merury, Solid at and Below -39c, Liquid from -39 to 357c, Boils and turns to gas at temps higher than 357c

Overall... Bad Idea........... ANd Mercury is NOT magnetic... Dunno how you could pump it with any sory of magnet.....

Somthing i need to look up.
BTW: I'm just thinking outlound here. This is not a project or anthing that will probly be done.

[tennvols_69]

i have acess to bout 2 pints of murcury. i wonder how this would work in a heatpipe situation. as one could assume that as it heats it does like most sooling materials and flows. this might have to be my next experiment. btw its toxicity is as harmful as its made out to be. it will absorb thru the skin and into your blood stream. whatcha guys think? you think it could be a better heat pipe substance then whats currently used?

[E1000BASE-TX]

Originally Posted by tennvols_69

i have acess to bout 2 pints of murcury. i wonder how this would work in a heatpipe situation. as one could assume that as it heats it does like most sooling materials and flows. this might have to be my next experiment. btw its toxicity is as harmful as its made out to be. it will absorb thru the skin and into your blood stream. whatcha guys think? you think it could be a better heat pipe substance then whats currently used?

If you tuch it enoth times yes. Contact with the skin once wont kill you. Be shure to handle with gloves and gogles if you plan to cool somthing with it.

[-navairum-]

umm, dont f*** with the murcury, end of story.

[tennvols_69]

ive messed with mercury many times, i know the safety precautions navairum. now e1000 ive been looking somemore into this. and gathered i would need to place it in a vacuum inside the heatpipes for the reality for it to work. or i beleave. what is your opinion on this? i feel this is needed to prevent oxidation and because i feel any exacess air will cause a decine in its thermal properties. so the only way i can see doing this is to place every heatpipe in a vacuumed space and seal them. however it may not be needed, as i might be able to do it in a flooded environment with an inert gas. thus preventing oxidation but this might add impurities to the murcury. inhitbiting its cooling properties, or worse case causing a reaction between the Hg and the inert gas. i feel if i used heatpipes and the result was negative that the base heat sink will provide enought cooling still to prevent any kind of hardware failures.here is the MSDS for Hg. http://msds.pdc.cornell.edu/msds/MSD...81/M140462.htm now notice that it says in reactivity to prevent contact with ACETYLENE PRODUCTS however would this cover mapp gass as that is what i would be using for the brazing of the heat pipes.

[E1000BASE-TX]

I dont know... sory.
I'm looking this up as I speek.

[tennvols_69]

if i do this my computer thats cooled by it will be called Quicksilver

[Soundstorm]

I don't think it will work in heatpipes!
The normal (those supplied in the new commercial heatsinks) heatpipes are filled with water, yes normal water, under a low presure so the boiling temperatures get lower. So if you fill with mercury, you should need a near vacuum tube with some mercury in it. This is allmost impossible to do without the right equipment (= expensive).

Also in his experiment (or whatever it is) the pump is pretty easy. There is no magnet moving, it's just staying where it is. It works like a normal motor. Only here the shaft is the mercury (or some magnetic fluid).

[Asgorath]

Reply to #1

And it conducts electricity, if you get a leak you kill you computer and put you in danger...
Well Fluid Xp is cheaper than mercury: http://www.fluidxp.com/

[freecableguy]

Because water is a better heat conductor than Hg and will have a smaller temperature rise for an equal input heat flux. You obviously don't understand why mercury was used in the first place to cool equipment....think maintaining the medium subcooled without the need for high pressure and you will be halfway there....

[freecableguy]

And i dunno about throwing a magnet in my comptuer anyways...

Let's see...there are magnets in your floppy drive, all your hard drives (these magnets are STRONG), if you have a magneticly driven waterpump (most watercoolers do) then that has a big-*** magnet...and, OH YEAH, EVERY fan you have also produces magnetic fields....wait, one more....speakers, magnets drive the coils....uhhh, ummm...your monitor....BIG magnet in the back....have I made my point yet? Inductors on your MB....magnets wrapped in wire....flowing current, especially high-frequency SWITCHING current, just like the PWM power supply circuits on your MB and don't forget your PSU....big magnet in the HUGE stepdown transformer in there....I could go on for quite a while...

[AngryAlpaca]

Mercury, overall, is a worse coolant than water or alcohol. Also, it's too dangerous for cooling nuclear reactors. If it's too dangerous for them, too dangerous for anything else. It's too expensive. it's too heavy. It offers very little. It expands a lot.

[Topher]

so to wrap it up in a few words... forget about ever using Hg for cooling... either stick to water or get a phase change

[E1000BASE-TX]

Originally Posted by AngryAlpaca

Mercury, overall, is a worse coolant than water or alcohol. Also, it's too dangerous for cooling nuclear reactors. If it's too dangerous for them, too dangerous for anything else. It's too expensive. it's too heavy. It offers very little. It expands a lot.

This topic is not about using Mercury. Mercury was an example I came up with for a liquid metal.

[Topher]

http://www.webelements.com/webelemen...xt/Hg/key.html

sorry, but turns out Hg isn't so great at cooling anyway...

It is a rather poor conductor of heat as compared with other metals but is a fair conductor of electricity.