replacing water with mercury

[EluSiOn]

is it possible replacing water with mercury? It has 8.3 W·m-1·K-1 thermal conductivity which is better than water (0.6)...

so if get a big ass pump like iwaki 30 rzx that can push... would it improve cooling much better?

Thermal conductivity of other common materials:
Silver 430 W·m-1·K-1
Copper 390
Gold 320
Aluminium 236
Platinum 70
Quartz 8
Glass 1
Water 0.6
Wool 0.05

[crimedog]

what's the density of mercury though??

[Raider84]

also you know that mercury is extremely toxic to humans right?

[EluSiOn]

mercury itself is not toxic... but when it mix with stream/ air / water vapor it because extremely toxic..

and mercury's density is 13.6 where water is 1.... so just get a powerful pump.... it should be fine right?

[johan851]

The density, or viscosity, of mercury would make it very difficult for a pump to push. Also, mercury's size is pretty responsive to temperature differences - look at a thermometer. Also, since working with it would be life threatening...it's probably not a good idea.

[UberBlue]

Please not this again! The mercury horse has been beaten to an unrecognizable pulp many times in this forum. If you really want to know, search the forum. You will find a couple of huge threads with all the reasons it shoudn't be used.

http://www.ocforums.com/showthread.php?s=&threadid=188239&highlight=mercury

http://www.ocforums.com/showthread.php?s=&threadid=178661&highlight=mercury

http://www.ocforums.com/showthread.php?s=&threadid=201613&highlight=mercury

There are many, many more where that came from.

[{BC}Helix]

I think there was also some posts on procooling and i started a thread on gideontech about it. and i think someone said that it would cost somthing like $1100 to fill a system with it.

[mata2974]

mercury itself is not toxic... but when it mix with stream/ air / water vapor it because extremely toxic.. this is outright untrue, :eek: :eek: mercury is deadly, whether in elementary, salt or as organomercury and besides where are you going to buy it? Its regulated and you must have valid reasons to convince the EPA.:eek: LOL

[AngryAlpaca]

Specific heat capacity of mercury is terrible. It would be hard to pump, dangerous (When it mixes with air? When could THAT happen?) expensive, not a great cooler, might react with metals, umm... looks really freakin' cool...

[vonkaar]

Originally posted by UberBlue
Please not this again! The mercury horse has been beaten to an unrecognizable pulp many times in this forum. If you really want to know, search the forum. You will find a couple of huge threads with all the reasons it shoudn't be used.

http://www.ocforums.com/showthread.php?s=&threadid=188239&highlight=mercury

http://www.ocforums.com/showthread.php?s=&threadid=178661&highlight=mercury

http://www.ocforums.com/showthread.php?s=&threadid=201613&highlight=mercury

There are many, many more where that came from.

thx much UberBlue. Yes... tired argument... getting stupid. NOT PRACTICAL.

[kingsack]

Totally impracticle.

Mercury is toxic in all forms.
Mercury is mutagenic as well.
Mercury is extremely high in density and viscosity (beyond hard to pump)
Mercury is electrically conductive (a spill and it's all over!)
Mercury is very expensive (I concurr on > $1000 for a setup)
Flowing Mercury can create a Magnetic Field!

Etc..........

[wildfrogman]

Waterblock design is where you will be getting the improvement in temps anyways. Or, if you just HAVE to use some exotic cooling thats not water based you could buy some high end ammonia based closed loop heatpipes. They cost alot but are well over a million times safer than mercury. You have to understand...talking about filling your hoses with mercury is like talking about filling them with some botulism/cyanide/ricin/plutonium sludge that would evaporate and sink into the watertable and cause a poisonous cloud of death causing fema to close off your city and then bull doze everthing then remove the worst soil by large dump trucks then pave over the entire area and put a fence around it. Mercury, especially in that amount just shows most people you dont know what mercury really is.

[thorilan]

how would murc react inside the mag driven pumps anyways?

[Cathar]

I remember that at my University, they had the old CSIRAC computer, which was one of the older computers in the world.

For its main memory it used to use mercury delay lines. These were large tubes of mercury, with a pulsator at one end, and a sensor at the other. Stuff to be "stored" in memory was pulsed in at one end, and the pulses would take some time to travel to the other end, where they were sensed, and if the CPU needed the information at that time, it would use it, otherwise the information got cycled back. The computer had multiples of these lines, and they would slowly leak mercury, with the operators having to come in on a morning and "mop up" small puddles of mercury from the floor.

One of our University lecturers used to work in the room where the computer was during his younger years. He was the crankiest s.o.b. you could ever hope to meet.

When Coca Cola first came out, one of its main ingredients was Cocaine, obtained from the "coca" plant, hence the "Coke" and the "Coca" derivations of its name. Santa Clause was also never really perceived as a fat man dressed in red & white before a wildly successful 1930's Coca Cola advertising campaign.

I guess what I'm saying is that people's perceptions of the dangers of various substances changes over time, and with experience. Cocaine and mercury, which were once deemed to be not so harmful, have a completely different perception nowadays.

[Raider84]

good point. Is mercury magnetic?!!? and if it is.....

[Korndog]

mercury is a metal.. yes its magnetic
but that doesn't matter
the fact that it can create a magnetic field when it moves around, in say a watercooling setup

[kouch]

forget everthing else,

a one foot column of mercury would have a pressure of 5.88 psi or 40561 pascals. And thats just one foot column, without any other tubing or drag or pressure loss. It would be real hard to find ANY pump that would pump that with any kinda flow rate and if you did you would need some metal tubing of good quality to hold that knida pressure.

[johan851]

Good point Cathar. They used to use radioactive paint on watch hands too, to make them glow in the dark. The painters would lick the tips of their brushes to make them pointier. :eek:

Then there's the stories of little kids playing in asbestos piles...

[JFettig]

is aluminum magnetic?


I think that shoudl answer the magnetic question, what is the only materials that are magnetic?









IRON

[Korndog]

hehe.. forgot about that
but they can create magnetic waves, thats what counts right?
.. :cough:

[UberBlue]

Originally posted by JFettig
is aluminum magnetic?


I think that shoudl answer the magnetic question, what is the only materials that are magnetic?









IRON

Sort of.

Stainless Steel is an iron/nickel alloy, but it's not magnetic. Add a little cobalt to that mix and you've got an INCONEL magnet

A cobalt/platinum alloy is one of the most magnetic materials on earth.

[AngryAlpaca]

Well, seeing as you've only listed one third of the materials that are magnetic... :rolleyes:

[UberBlue]

Originally posted by AngryAlpaca
Well, seeing as you've only listed one third of the materials that are magnetic... :rolleyes:

Wasn't my intent to list every magnetic material, only illustrate that iron isn't the only magnetic material.

[Cathar]

http://www.dansdata.com/magnets.htm

[AngryAlpaca]

I wasn't referring to you.

[captain_sHiFTy]

I think the main reason that mercury hasnt been used to date is because of its insane inpractability. When you think about it, how many hundreds or thousands of people do you think they have researching new waterblock/heatsink/cooling designs? As computers needs for heat dissipation get greater and greater those designs are going to be needed anymore.

And with all the different kinds of cooling methods/designs that are out there, you and i all know that theres a very large industry for this stuff.

So if liquid nitrogen was such a viable solution for cooling, don't you think that someone out there would have come up with a commercial mercury cooler?

or at least made a research project out of it by now. If it was viable someone out there would have posted results right?

[captain_sHiFTy]

....on the other hand, looking at all the information, might it not be easier and better to have liquid nitrogen as a heatspreader? It has a low specific heat meaning it would not be able to gain heat very fast, so say if you had some means of removing the heat from the mercury really quickly, it would cool it down quick?

Sortof the same concept as having a copper plug in an aluminum heatsink?

...just curious...

oh, and silver has a better heat conductivity than copper right. How come nobody has tried to make a silver core in an aluminum heatsink, instead of the copper

[Cathar]

Originally posted by captain_sHiFTy
oh, and silver has a better heat conductivity than copper right. How come nobody has tried to make a silver core in an aluminum heatsink, instead of the copper

First time I've touched the site in over a year, but did so to dig up this "review" of just what you're asking for:

http://www.tomshardware.com/cpu/20010306/cooler-10.html

and

http://www.tomshardware.com/cpu/20010306/cooler-11.html

[squeakygeek]

Originally posted by EluSiOn
is it possible replacing water with mercury? It has 8.3 W·m-1·K-1 thermal conductivity which is better than water (0.6)...

so if get a big ass pump like iwaki 30 rzx that can push... would it improve cooling much better?

Thermal conductivity of other common materials:
Silver 430 W·m-1·K-1
Copper 390
Gold 320
Aluminium 236
Platinum 70
Quartz 8
Glass 1
Water 0.6
Wool 0.05

We all know that water is a pretty poor conductor of heat, but that doesn't matter because the heat does not have to transfer through the water; the water is moved physically, the heat along with it. Water has a very high heat capacity; something like 4.1 J/gC.

Originally posted by thorilan
how would murc react inside the mag driven pumps anyways?

It wouldn't budge. Most likely the impeller wouldn't even turn, and if it did by some miracle, the oil-less impeller shaft is designed to be lubricated by water. The lack of movement would also cause the pump to overheat, and live a very short life.

Originally posted by AngryAlpaca
I wasn't referring to you.

There is a "quote" button under every post. It eliminates the confusion.

Originally posted by captain_sHiFTy
....on the other hand, looking at all the information, might it not be easier and better to have liquid nitrogen as a heatspreader? It has a low specific heat meaning it would not be able to gain heat very fast, so say if you had some means of removing the heat from the mercury really quickly, it would cool it down quick?

Sortof the same concept as having a copper plug in an aluminum heatsink?

...just curious...

oh, and silver has a better heat conductivity than copper right. How come nobody has tried to make a silver core in an aluminum heatsink, instead of the copper

Actually, the lower the specific heat, the more quickly it gains temperature.

[AngryAlpaca]

Silver is MUCH more expensive than copper, and with only a TINY bit more performance. Aluminum heatsinks are built for their prices. A better option than using a silver core in an aluminum heatsink would be using a copper heatsink.

[jhl3]

If you guys want something like this just take your hs or wb to a custom jeweler and get them to coat its base several times with molten silver. just be aware that you may need to lap it afterwards as the heat may warp it a bit but not much.

[UberBlue]

Originally posted by Cathar
http://www.dansdata.com/magnets.htm

Think those are strong...

I did some more looking and found that cast Alnico V magnets have a field strenght of 1.28T. The 76.7/23.3% by weight platinum/cobalt magnets have a field strength of 2.5T, but are prohibitively expensive and potentialy dangerous.

Source: CRC Handbook of Chemistry and Physics 76th ed.

[Gentleman]

Using mercury as a coolant isn't out of the question. Putting toxicity aside. Mercury wouldn't expand that much if its delta is not that high. Changing from say 25C (room temp) to 30C CPU is not gonna increase the volume of the mercury that much. The reasone people seem to think so is because of what the see in thermometers, these have very narrow shafts to overemphasize the effect. Mercury would pump just as well given a larger pump. A simple swap of water for mercury in your WC system is not gonna do, you gotta upgrade your pump and radiator. And people talking about specific heat capacity, what are you talking about? All you are concerned about is the how fast heat can get into the mercury. ONLY major advantage i see to this is the possibility of getting your temps down to that of ambient temps without chillers and insulation or pelts and temp modulator cards.

There are several studies right now on pratical applications of using mercury as a coolant in metal smelting. So we know that it is possible. Only problem I see is the cost of getting the amount of mercury needed. Can't really buy this stuff in your local hardware store. Then again so is liquid nitrogen, but that has be done by a few already. Wonder how many thermometers are needed to get a liter of this stuff. Better start saving ;)

[clocker2]

Outside of the toxicity, pumping and availability issues, another real problem with using mercury is the weight.
Your rig would have to be built like a Mack truck just to support the the mass of the mercury and all the attendent piping, etc.

[Borisw37]

You will never be able to get that much mercury. Im sure to order it from chem. suppliers you need to have a license, be a registered institution/firm. And probably have permission from your township/city. Because when you spill 1/2 gallon of mercury, its gona be a mess for everybody to clean up. It will cost a lot. And for that much money you can get a phase changing system, that will drive CPU temps below ambient.

[clocker2]

Originally posted by Borisw37
You will never be able to get that much mercury. Im sure to order it from chem. suppliers you need to have a license, be a registered institution/firm. And probably have permission from your township/city.

Not true.
I have used significant amounts of mercury in the past in the pendulums of my clocks.
It was easily available, although I did have to sign some sort of paper ( not even sure what it was... I only had to do it the first time), but the price is prohibitive. Mercury is sold by the pound- a one pound container is approx. the size of a 35mm film canister.

The whole idea of using mercury as a coolant is completely impractical...just fabricating the transmission lines would be a nightmare. The best/safest tubing to use would Pyrex and every single joint is a potential leak ( and mercury will find the smallest porosity due to it's great weight).

Give it up.:p

[Big_Baller]

"Daddy whats that leak out of the hose...?" Daddy: "Ohh ssshh"

I would think the risks are a little too much. How would you go about leak testing a system like that? Not exactly somthing you would do in your house/garage....and with that much mercury sitting around what happens why your (insert person/relation/friend/pet) knocks over the container?

But if you found a way to make a safe bong/evap system for a mercury setup...you would be the man.

[squeakygeek]

I think he gets the idea that replacing water with mercury won't work. Let's talk about something else now... in some other thread...