Modding Mercury viable water cooling alternative?

Specific heat: The amount of heat required to raise the temperature of 1kg of a substance through 1K (1K = 1 deg C... when talking in temperature deltas, it is always referred to in K's, even tho it's the same unit as deg C)

Specific heat of Mercury - 138
Indium - 238
Gallium - 381
Water - 4190
Hydrogen - 14300 (the only substance with a higher specific heat than water).

Water is and always will be the best coolant to use in a liquid cooling system. The answer will always be the same, no matter in what decade the question is asked (seeing as this gets asked ridiculously frequently across the net and always has).

Reference: http://www.diracdelta.co.uk/science/source/s/p/specific heat capacity/source.html

 

Does anyone really think that using a liquid metal cooling system in a PC, especially mercury, is nothing short of lunacy?

Purely from the heath and saftey aspects alone, mercury is highly toxic - take a look at this quote from Wikepedia:-

Occupational exposure

Due to the health effects of mercury exposure, industrial and commercial uses are regulated in many countries. The World Health Organization, OSHA, and NIOSH all treat mercury as an occupational hazard, and have established specific occupational exposure limits. Environmental releases and disposal of mercury are regulated in the U.S. primarily by the United States Environmental Protection Agency.

A study has shown that acute exposure (4-8 hours) to calculated elemental mercury levels of 1.1 to 44 mg/m3 resulted in chest pain, dyspnea, cough, hemoptysis, impairment of pulmonary function, and evidence of interstitial pneumonitis.

Acute exposure to mercury vapor has been shown to result in profound central nervous system effects, including psychotic reactions characterized by delirium, hallucinations, and suicidal tendency. Occupational exposure has resulted in broad-ranging functional disturbance, including erethism, irritability, excitability, excessive shyness, and insomnia. With continuing exposure, a fine tremor develops and may escalate to violent muscular spasms. Tremor initially involves the hands and later spreads to the eyelids, lips, and tongue. Long-term, low-level exposure has been associated with more subtle symptoms of erethism, including fatigue, irritability, loss of memory, vivid dreams, and depression.

Personally, I'll stick to water as my cooling medium.

 

the specific heat is how much heat it takes to raise the temp of the liquid 1C. The numbers I gave are how quickly heat can be removed the the substance. Asuming the flow rate is fast enoung, the specific heat won't really matter.

But still, I already conceded the point that mercury would be a bad choice.

 

I know a guy who did some experiments with magnetic pumps and salt water/mercury/gallium. It's not really worth the trouble.

 

your forgetting that in modding it's not about the effeciency or what not, I mean oil cooled computers aren't exactly user friendly or light, but people still build them. gallium-tin alloys would probably work a lot better, of course such a rig would be purely for the sake of doing it, not because it would be an advantageous cooling solution

 

yeah, mathematically i guess it could work, but don't try. Instead a safer/better method of cooling is vapor phase change cooling, which can be run 24/7, and for more extreme runs for shorter periods of time dice/ln2 runs. Whats mercury cost anyways? and whats the Chinese emperor who in his tomb has rivers of flowing mercury?

 

you all are forgetting that mercury expands a LOT relative to it's temperature. you'd have to make sure that you have enough resivoir space to make sure the loop doesnt burst from pressure if the pump dies. on the other hand though, you could make your res into a thermometer for the whole loop. the higher it is in the res, the hotter your comp is running.

 

That may be one of the sickest ideas I've seen for a WC loop. As if using mercury didn't have enough awesomeness in the first place.

 

i think you would get yourself killed before you kill the heat...

 

Actually it doesn't. Water's coefficient of thermal expansion is a fair bit higher above room temperature.

A mercury thermometer has a large bulb and a very fine tube to exaggerate the expansion.

 

For a metal coolant, conductance >>> specific heat. For example, a copper heat pipe still beats a WC system.

To those with concerns about pumping the metal through the loop, it's actually easier to do than water. Liquid metals (being electrically conductive, and therefore capable of electrical induction) can be pumped with a series of electromagnets much like how a MagLev train runs or a Gauss rifle fires. This means you end up with a pump that has absolutely no moving parts.

 

I researched the idea thoroughly and abandoned it as folly.

Mercury has higher conduction, but lower capacity of heat. It dissolves most metals. It has a very high surface tension which prevents proper wetting to surfaces. You can protect metals by having them coated with gas deposited diamond, but it is extremely expensive. A special type of stainless steel is used for long term mercury contact.

The best cooling you can get is with a silver water block with pure D2O; dideuterium oxide; heavy water. If you can't get it, H2O; dihydrogen oxide; water is 90% as good.

 

Mercury was the original material used in the first reactors on the planet. They had good neutron reflection properties. Problem is that it is very hazardous. You have to figure that mechanical things fail. The three mile island design is the most common type of reactor on the planet. Chernobal and three mile island both had breaches of the pressure vessel. In TMI's case it shut itself down when that happened releasing relatively little radiation. While Chernobal sprayed large areas with nuclear materials. Most recent Russian designs are more like the west. Germany first came up with the reactor ideas in the 1938-1942 time frame and their design used mercury. But with mercury you don't have a thermal neutron based process as you do with the other designs. They have to use fast breaders because you don't have the moderating affects of the water in the TMI design. You could use Graphite though as some designs do. Cooling properties I don't think are the main issue in this decision. Some reactors use gas which has the poorest heat conduction properties. Better materials like Sodium based coolants that the Russians use replaced any designs that uses Mercury.

Sodium was also used in a test of a solar power plant in California. Where mirrors were used to provide the heat source instead of Nuclear or Chemical reactions.

I believe you can find more about the subject if you search the internet for I believe it was called the "Clemantine" reactor built I think in England most likely off designs taken from fallen Nazi Germany.

The US first nuclear powered submarine featured a similar type of reactor based on the Nazi type XXI sub design from late in the war that included a nuclear power plant.

They soon replaced it with a water moderated design similar to that used at TMI.

Hope this helps it is some interesting history. Researching this also lead me to some interesting history about Nazi nuke weapons research. I still wait for the 2005 book "Hitlers Bomb" to be translated into English which discusses the subject and researched the history from German archives.

 

nice thread necro, lol! 3 years too late

 

Heheh indeed; I read all through it without checking the dates Never understood the desire to use a substance that brought about the saying "mad as a hatter"

Not only is it necro, it's off topic!

 

Correction

This portion of your statement is incorrect. I am retired from the US Navy, and while I was in I was a submarine Nuclear plant operator. The first nuclear powered submarine in the US was the USS Nautilus, and is still on display as a museum in Groton, CT. The Nautilus featured a 'pressurized water cooled reactor'.

"The newly developed S2W (Submarine, Model 2, Westinghouse) pressurized-water nuclear reactor provided her power both on the surface, where her top speed was 22 knots (41 km/hr), and underwater, where she could do 23 knots (42 km/hr)." (http://americanhistory.si.edu/subs/history/subsbeforenuc/revolution/nautilus.html)

Not trying to jump on your case, just wanted to point out that while the US Navy did experiment with other types of coolant systems, they went with water when they built the first nuclear powered submarine.


Nano

 

I had the same thoughts, the XXI was a break though in design but not a nuclear sub by any means.

Another wiki expert! lol!